Merge remote-tracking branch 'drm-intel/drm-intel-next-queued' into drm-tip

# Conflicts: # drivers/gpu/drm/i915/i915_drv.h # drivers/gpu/drm/i915/i915_gem_gtt.c # drivers/gpu/drm/i915/i915_gem_shrinker.c # drivers/gpu/drm/i915/intel_dp.c # drivers/gpu/drm/i915/intel_engine_cs.c # drivers/gpu/drm/i915/intel_uc.h
author: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> 2017-06-12 16:26:43 +0200
committer: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> 2017-06-12 16:26:43 +0200
commit: 6cb59a79097546ee0c1880a1d1cf4d6b852005f7 (patch)
tree: 1c702e567e44d34b05442587f9654b82333d2287
parent: 6e88007e224ce51969ccf9afeec645146c638816 (diff)
parent: 7d025e0804f2cbeafeba2f5b8effa6361d7db8e4 (diff)
395 files changed, 15505 insertions, 5942 deletions
diff --git a/Documentation/devicetree/bindings/display/brcm,bcm-vc4.txt b/Documentation/devicetree/bindings/display/brcm,bcm-vc4.txt
index ca02d3e4db91..284e2b14cfbe 100644
--- a/Documentation/devicetree/bindings/display/brcm,bcm-vc4.txt
+++ b/Documentation/devicetree/bindings/display/brcm,bcm-vc4.txt
@@ -5,7 +5,7 @@ with HDMI output and the HVS (Hardware Video Scaler) for compositing
 display planes.
 
 Required properties for VC4:
-- compatible:	Should be "brcm,bcm2835-vc4"
+- compatible:	Should be "brcm,bcm2835-vc4" or "brcm,cygnus-vc4"
 
 Required properties for Pixel Valve:
 - compatible:	Should be one of "brcm,bcm2835-pixelvalve0",
@@ -54,11 +54,14 @@ Required properties for VEC:
 		  See bindings/interrupt-controller/brcm,bcm2835-armctrl-ic.txt
 
 Required properties for V3D:
-- compatible:	Should be "brcm,bcm2835-v3d"
+- compatible:	Should be "brcm,bcm2835-v3d" or "brcm,cygnus-v3d"
 - reg:		Physical base address and length of the V3D's registers
 - interrupts:	The interrupt number
 		  See bindings/interrupt-controller/brcm,bcm2835-armctrl-ic.txt
 
+Optional properties for V3D:
+- clocks:	The clock the unit runs on
+
 Required properties for DSI:
 - compatible:	Should be "brcm,bcm2835-dsi0" or "brcm,bcm2835-dsi1"
 - reg:		Physical base address and length of the DSI block's registers
diff --git a/Documentation/devicetree/bindings/display/st,stm32-ltdc.txt b/Documentation/devicetree/bindings/display/st,stm32-ltdc.txt
new file mode 100644
index 000000000000..8e1476941c0f
--- /dev/null
+++ b/Documentation/devicetree/bindings/display/st,stm32-ltdc.txt
@@ -0,0 +1,36 @@
+* STMicroelectronics STM32 lcd-tft display controller
+
+- ltdc: lcd-tft display controller host
+  must be a sub-node of st-display-subsystem
+  Required properties:
+  - compatible: "st,stm32-ltdc"
+  - reg: Physical base address of the IP registers and length of memory mapped region.
+  - clocks: A list of phandle + clock-specifier pairs, one for each
+    entry in 'clock-names'.
+  - clock-names: A list of clock names. For ltdc it should contain:
+      - "lcd" for the clock feeding the output pixel clock & IP clock.
+  - resets: reset to be used by the device (defined by use of RCC macro).
+  Required nodes:
+    - Video port for RGB output.
+
+Example:
+
+/ {
+	...
+	soc {
+	...
+		ltdc: display-controller@40016800 {
+			compatible = "st,stm32-ltdc";
+			reg = <0x40016800 0x200>;
+			interrupts = <88>, <89>;
+			resets = <&rcc STM32F4_APB2_RESET(LTDC)>;
+			clocks = <&rcc 1 CLK_LCD>;
+			clock-names = "lcd";
+
+			port {
+				ltdc_out_rgb: endpoint {
+				};
+			};
+		};
+	};
+};
diff --git a/Documentation/devicetree/bindings/display/zte,vou.txt b/Documentation/devicetree/bindings/display/zte,vou.txt
index 9c356284232b..38476475fd60 100644
--- a/Documentation/devicetree/bindings/display/zte,vou.txt
+++ b/Documentation/devicetree/bindings/display/zte,vou.txt
@@ -58,6 +58,18 @@ Required properties:
    integer cells.  The first cell is the offset of SYSCTRL register used
    to control TV Encoder DAC power, and the second cell is the bit mask.
 
+* VGA output device
+
+Required properties:
+ - compatible: should be "zte,zx296718-vga"
+ - reg: Physical base address and length of the VGA device IO region
+ - interrupts : VGA interrupt number to CPU
+ - clocks: Phandle with clock-specifier pointing to VGA I2C clock.
+ - clock-names: Must be "i2c_wclk".
+ - zte,vga-power-control: the phandle to SYSCTRL block followed by two
+   integer cells.  The first cell is the offset of SYSCTRL register used
+   to control VGA DAC power, and the second cell is the bit mask.
+
 Example:
 
 vou: vou@1440000 {
@@ -81,6 +93,15 @@ vou: vou@1440000 {
 			      "main_wclk", "aux_wclk";
 	};
 
+	vga: vga@8000 {
+		compatible = "zte,zx296718-vga";
+		reg = <0x8000 0x1000>;
+		interrupts = <GIC_SPI 86 IRQ_TYPE_LEVEL_HIGH>;
+		clocks = <&topcrm VGA_I2C_WCLK>;
+		clock-names = "i2c_wclk";
+		zte,vga-power-control = <&sysctrl 0x170 0xe0>;
+	};
+
 	hdmi: hdmi@c000 {
 		compatible = "zte,zx296718-hdmi";
 		reg = <0xc000 0x4000>;
diff --git a/Documentation/gpu/index.rst b/Documentation/gpu/index.rst
index c572f092739e..037a39ac1807 100644
--- a/Documentation/gpu/index.rst
+++ b/Documentation/gpu/index.rst
@@ -12,6 +12,7 @@ Linux GPU Driver Developer's Guide
    drm-uapi
    i915
    meson
+   pl111
    tinydrm
    vc4
    vga-switcheroo
diff --git a/Documentation/gpu/pl111.rst b/Documentation/gpu/pl111.rst
new file mode 100644
index 000000000000..9b03736d33dd
--- /dev/null
+++ b/Documentation/gpu/pl111.rst
@@ -0,0 +1,6 @@
+==========================================
+ drm/pl111 ARM PrimeCell PL111 CLCD Driver
+==========================================
+
+.. kernel-doc:: drivers/gpu/drm/pl111/pl111_drv.c
+   :doc: ARM PrimeCell PL111 CLCD Driver
diff --git a/Documentation/sync_file.txt b/Documentation/sync_file.txt
index c3d033a06e8d..496fb2c3b3e6 100644
--- a/Documentation/sync_file.txt
+++ b/Documentation/sync_file.txt
@@ -1,8 +1,8 @@
-			      Sync File API Guide
-			      ~~~~~~~~~~~~~~~~~~~
+===================
+Sync File API Guide
+===================
 
-				Gustavo Padovan
-			  <gustavo at padovan dot org>
+:Author: Gustavo Padovan <gustavo at padovan dot org>
 
 This document serves as a guide for device drivers writers on what the
 sync_file API is, and how drivers can support it. Sync file is the carrier of
@@ -46,16 +46,17 @@ Creating Sync Files
 
 When a driver needs to send an out-fence userspace it creates a sync_file.
 
-Interface:
+Interface::
+
 	struct sync_file *sync_file_create(struct dma_fence *fence);
 
 The caller pass the out-fence and gets back the sync_file. That is just the
 first step, next it needs to install an fd on sync_file->file. So it gets an
-fd:
+fd::
 
 	fd = get_unused_fd_flags(O_CLOEXEC);
 
-and installs it on sync_file->file:
+and installs it on sync_file->file::
 
 	fd_install(fd, sync_file->file);
 
@@ -71,7 +72,8 @@ When userspace needs to send an in-fence to the driver it passes file descriptor
 of the Sync File to the kernel. The kernel can then retrieve the fences
 from it.
 
-Interface:
+Interface::
+
 	struct dma_fence *sync_file_get_fence(int fd);
 
 
@@ -79,5 +81,6 @@ The returned reference is owned by the caller and must be disposed of
 afterwards using dma_fence_put(). In case of error, a NULL is returned instead.
 
 References:
-[1] struct sync_file in include/linux/sync_file.h
-[2] All interfaces mentioned above defined in include/linux/sync_file.h
+
+1. struct sync_file in include/linux/sync_file.h
+2. All interfaces mentioned above defined in include/linux/sync_file.h
diff --git a/MAINTAINERS b/MAINTAINERS
index 7a28acd7f525..f6fc18fd6e88 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -4235,6 +4235,12 @@ F:	include/drm/drm*
 F:	include/uapi/drm/drm*
 F:	include/linux/vga*
 
+DRM DRIVER FOR ARM PL111 CLCD
+M:	Eric Anholt <eric@anholt.net>
+T:	git git://anongit.freedesktop.org/drm/drm-misc
+S:	Supported
+F:	drivers/gpu/drm/pl111/
+
 DRM DRIVER FOR AST SERVER GRAPHICS CHIPS
 M:	Dave Airlie <airlied@redhat.com>
 S:	Odd Fixes
@@ -4242,6 +4248,8 @@ F:	drivers/gpu/drm/ast/
 
 DRM DRIVERS FOR BRIDGE CHIPS
 M:	Archit Taneja <architt@codeaurora.org>
+M:	Andrzej Hajda <a.hajda@samsung.com>
+R:	Laurent Pinchart <Laurent.pinchart@ideasonboard.com>
 S:	Maintained
 T:	git git://anongit.freedesktop.org/drm/drm-misc
 F:	drivers/gpu/drm/bridge/
@@ -4498,6 +4506,17 @@ S:	Maintained
 F:	drivers/gpu/drm/sti
 F:	Documentation/devicetree/bindings/display/st,stih4xx.txt
 
+DRM DRIVERS FOR STM
+M:	Yannick Fertre <yannick.fertre@st.com>
+M:	Philippe Cornu <philippe.cornu@st.com>
+M:	Benjamin Gaignard <benjamin.gaignard@linaro.org>
+M:	Vincent Abriou <vincent.abriou@st.com>
+L:	dri-devel@lists.freedesktop.org
+T:	git git://anongit.freedesktop.org/drm/drm-misc
+S:	Maintained
+F:	drivers/gpu/drm/stm
+F:	Documentation/devicetree/bindings/display/st,stm32-ltdc.txt
+
 DRM DRIVER FOR TDFX VIDEO CARDS
 S:	Orphan / Obsolete
 F:	drivers/gpu/drm/tdfx/
diff --git a/drivers/dma-buf/dma-buf.c b/drivers/dma-buf/dma-buf.c
index 512bdbc23bbb..4a038dcf5361 100644
--- a/drivers/dma-buf/dma-buf.c
+++ b/drivers/dma-buf/dma-buf.c
@@ -558,8 +558,8 @@ struct dma_buf_attachment *dma_buf_attach(struct dma_buf *dmabuf,
 	if (WARN_ON(!dmabuf || !dev))
 		return ERR_PTR(-EINVAL);
 
-	attach = kzalloc(sizeof(struct dma_buf_attachment), GFP_KERNEL);
-	if (attach == NULL)
+	attach = kzalloc(sizeof(*attach), GFP_KERNEL);
+	if (!attach)
 		return ERR_PTR(-ENOMEM);
 
 	attach->dev = dev;
@@ -1122,9 +1122,7 @@ static int dma_buf_debug_show(struct seq_file *s, void *unused)
 		attach_count = 0;
 
 		list_for_each_entry(attach_obj, &buf_obj->attachments, node) {
-			seq_puts(s, "\t");
-
-			seq_printf(s, "%s\n", dev_name(attach_obj->dev));
+			seq_printf(s, "\t%s\n", dev_name(attach_obj->dev));
 			attach_count++;
 		}
 
diff --git a/drivers/dma-buf/dma-fence.c b/drivers/dma-buf/dma-fence.c
index 0918d3f003d6..57da14c15987 100644
--- a/drivers/dma-buf/dma-fence.c
+++ b/drivers/dma-buf/dma-fence.c
@@ -402,6 +402,11 @@ dma_fence_default_wait(struct dma_fence *fence, bool intr, signed long timeout)
 		}
 	}
 
+	if (!timeout) {
+		ret = 0;
+		goto out;
+	}
+
 	cb.base.func = dma_fence_default_wait_cb;
 	cb.task = current;
 	list_add(&cb.base.node, &fence->cb_list);
diff --git a/drivers/dma-buf/sync_debug.c b/drivers/dma-buf/sync_debug.c
index c769dc653b34..82a6e7f6d37f 100644
--- a/drivers/dma-buf/sync_debug.c
+++ b/drivers/dma-buf/sync_debug.c
@@ -110,7 +110,7 @@ static void sync_print_fence(struct seq_file *s,
 		}
 	}
 
-	seq_puts(s, "\n");
+	seq_putc(s, '\n');
 }
 
 static void sync_print_obj(struct seq_file *s, struct sync_timeline *obj)
@@ -132,9 +132,11 @@ static void sync_print_obj(struct seq_file *s, struct sync_timeline *obj)
 static void sync_print_sync_file(struct seq_file *s,
 				  struct sync_file *sync_file)
 {
+	char buf[128];
 	int i;
 
-	seq_printf(s, "[%p] %s: %s\n", sync_file, sync_file->name,
+	seq_printf(s, "[%p] %s: %s\n", sync_file,
+		   sync_file_get_name(sync_file, buf, sizeof(buf)),
 		   sync_status_str(dma_fence_get_status(sync_file->fence)));
 
 	if (dma_fence_is_array(sync_file->fence)) {
@@ -161,7 +163,7 @@ static int sync_debugfs_show(struct seq_file *s, void *unused)
 				     sync_timeline_list);
 
 		sync_print_obj(s, obj);
-		seq_puts(s, "\n");
+		seq_putc(s, '\n');
 	}
 	spin_unlock_irqrestore(&sync_timeline_list_lock, flags);
 
@@ -173,7 +175,7 @@ static int sync_debugfs_show(struct seq_file *s, void *unused)
 			container_of(pos, struct sync_file, sync_file_list);
 
 		sync_print_sync_file(s, sync_file);
-		seq_puts(s, "\n");
+		seq_putc(s, '\n');
 	}
 	spin_unlock_irqrestore(&sync_file_list_lock, flags);
 	return 0;
diff --git a/drivers/dma-buf/sync_file.c b/drivers/dma-buf/sync_file.c
index 2321035f6204..545e2c5c4815 100644
--- a/drivers/dma-buf/sync_file.c
+++ b/drivers/dma-buf/sync_file.c
@@ -41,8 +41,6 @@ static struct sync_file *sync_file_alloc(void)
 	if (IS_ERR(sync_file->file))
 		goto err;
 
-	kref_init(&sync_file->kref);
-
 	init_waitqueue_head(&sync_file->wq);
 
 	INIT_LIST_HEAD(&sync_file->cb.node);
@@ -82,11 +80,6 @@ struct sync_file *sync_file_create(struct dma_fence *fence)
 
 	sync_file->fence = dma_fence_get(fence);
 
-	snprintf(sync_file->name, sizeof(sync_file->name), "%s-%s%llu-%d",
-		 fence->ops->get_driver_name(fence),
-		 fence->ops->get_timeline_name(fence), fence->context,
-		 fence->seqno);
-
 	return sync_file;
 }
 EXPORT_SYMBOL(sync_file_create);
@@ -131,6 +124,36 @@ struct dma_fence *sync_file_get_fence(int fd)
 }
 EXPORT_SYMBOL(sync_file_get_fence);
 
+/**
+ * sync_file_get_name - get the name of the sync_file
+ * @sync_file:		sync_file to get the fence from
+ * @buf:		destination buffer to copy sync_file name into
+ * @len:		available size of destination buffer.
+ *
+ * Each sync_file may have a name assigned either by the user (when merging
+ * sync_files together) or created from the fence it contains. In the latter
+ * case construction of the name is deferred until use, and so requires
+ * sync_file_get_name().
+ *
+ * Returns: a string representing the name.
+ */
+char *sync_file_get_name(struct sync_file *sync_file, char *buf, int len)
+{
+	if (sync_file->user_name[0]) {
+		strlcpy(buf, sync_file->user_name, len);
+	} else {
+		struct dma_fence *fence = sync_file->fence;
+
+		snprintf(buf, len, "%s-%s%llu-%d",
+			 fence->ops->get_driver_name(fence),
+			 fence->ops->get_timeline_name(fence),
+			 fence->context,
+			 fence->seqno);
+	}
+
+	return buf;
+}
+
 static int sync_file_set_fence(struct sync_file *sync_file,
 			       struct dma_fence **fences, int num_fences)
 {
@@ -268,7 +291,7 @@ static struct sync_file *sync_file_merge(const char *name, struct sync_file *a,
 		goto err;
 	}
 
-	strlcpy(sync_file->name, name, sizeof(sync_file->name));
+	strlcpy(sync_file->user_name, name, sizeof(sync_file->user_name));
 	return sync_file;
 
 err:
@@ -277,22 +300,15 @@ err:
 
 }
 
-static void sync_file_free(struct kref *kref)
+static int sync_file_release(struct inode *inode, struct file *file)
 {
-	struct sync_file *sync_file = container_of(kref, struct sync_file,
-						     kref);
+	struct sync_file *sync_file = file->private_data;
 
 	if (test_bit(POLL_ENABLED, &sync_file->fence->flags))
 		dma_fence_remove_callback(sync_file->fence, &sync_file->cb);
 	dma_fence_put(sync_file->fence);
 	kfree(sync_file);
-}
-
-static int sync_file_release(struct inode *inode, struct file *file)
-{
-	struct sync_file *sync_file = file->private_data;
 
-	kref_put(&sync_file->kref, sync_file_free);
 	return 0;
 }
 
@@ -422,7 +438,7 @@ static long sync_file_ioctl_fence_info(struct sync_file *sync_file,
 	}
 
 no_fences:
-	strlcpy(info.name, sync_file->name, sizeof(info.name));
+	sync_file_get_name(sync_file, info.name, sizeof(info.name));
 	info.status = dma_fence_is_signaled(sync_file->fence);
 	info.num_fences = num_fences;
 
diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
index 78d7fc0ebb57..83cb2a88c204 100644
--- a/drivers/gpu/drm/Kconfig
+++ b/drivers/gpu/drm/Kconfig
@@ -246,6 +246,8 @@ source "drivers/gpu/drm/fsl-dcu/Kconfig"
 
 source "drivers/gpu/drm/tegra/Kconfig"
 
+source "drivers/gpu/drm/stm/Kconfig"
+
 source "drivers/gpu/drm/panel/Kconfig"
 
 source "drivers/gpu/drm/bridge/Kconfig"
@@ -274,6 +276,8 @@ source "drivers/gpu/drm/meson/Kconfig"
 
 source "drivers/gpu/drm/tinydrm/Kconfig"
 
+source "drivers/gpu/drm/pl111/Kconfig"
+
 # Keep legacy drivers last
 
 menuconfig DRM_LEGACY
diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile
index 59f0f9b696eb..c156fecfb362 100644
--- a/drivers/gpu/drm/Makefile
+++ b/drivers/gpu/drm/Makefile
@@ -82,6 +82,7 @@ obj-$(CONFIG_DRM_BOCHS) += bochs/
 obj-$(CONFIG_DRM_VIRTIO_GPU) += virtio/
 obj-$(CONFIG_DRM_MSM) += msm/
 obj-$(CONFIG_DRM_TEGRA) += tegra/
+obj-$(CONFIG_DRM_STM) += stm/
 obj-$(CONFIG_DRM_STI) += sti/
 obj-$(CONFIG_DRM_IMX) += imx/
 obj-$(CONFIG_DRM_MEDIATEK) += mediatek/
@@ -96,3 +97,4 @@ obj-y			+= hisilicon/
 obj-$(CONFIG_DRM_ZTE)	+= zte/
 obj-$(CONFIG_DRM_MXSFB)	+= mxsfb/
 obj-$(CONFIG_DRM_TINYDRM) += tinydrm/
+obj-$(CONFIG_DRM_PL111) += pl111/
diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile
index 660786aba7d2..20bde726419e 100644
--- a/drivers/gpu/drm/amd/amdgpu/Makefile
+++ b/drivers/gpu/drm/amd/amdgpu/Makefile
@@ -4,7 +4,7 @@
 
 FULL_AMD_PATH=$(src)/..
 
-ccflags-y := -Iinclude/drm -I$(FULL_AMD_PATH)/include/asic_reg \
+ccflags-y := -I$(FULL_AMD_PATH)/include/asic_reg \
 	-I$(FULL_AMD_PATH)/include \
 	-I$(FULL_AMD_PATH)/amdgpu \
 	-I$(FULL_AMD_PATH)/scheduler \
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index 833c3c16501a..77ff68f9932b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -36,11 +36,11 @@
 #include <linux/hashtable.h>
 #include <linux/dma-fence.h>
 
-#include <ttm/ttm_bo_api.h>
-#include <ttm/ttm_bo_driver.h>
-#include <ttm/ttm_placement.h>
-#include <ttm/ttm_module.h>
-#include <ttm/ttm_execbuf_util.h>
+#include <drm/ttm/ttm_bo_api.h>
+#include <drm/ttm/ttm_bo_driver.h>
+#include <drm/ttm/ttm_placement.h>
+#include <drm/ttm/ttm_module.h>
+#include <drm/ttm/ttm_execbuf_util.h>
 
 #include <drm/drmP.h>
 #include <drm/drm_gem.h>
@@ -1912,10 +1912,6 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon);
 u32 amdgpu_get_vblank_counter_kms(struct drm_device *dev, unsigned int pipe);
 int amdgpu_enable_vblank_kms(struct drm_device *dev, unsigned int pipe);
 void amdgpu_disable_vblank_kms(struct drm_device *dev, unsigned int pipe);
-int amdgpu_get_vblank_timestamp_kms(struct drm_device *dev, unsigned int pipe,
-				    int *max_error,
-				    struct timeval *vblank_time,
-				    unsigned flags);
 long amdgpu_kms_compat_ioctl(struct file *filp, unsigned int cmd,
 			     unsigned long arg);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c
index a6649874e6ce..9f0247cdda5e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c
@@ -96,7 +96,7 @@ static int amdgpu_bo_list_set(struct amdgpu_device *adev,
 	int r;
 	unsigned long total_size = 0;
 
-	array = drm_malloc_ab(num_entries, sizeof(struct amdgpu_bo_list_entry));
+	array = kvmalloc_array(num_entries, sizeof(struct amdgpu_bo_list_entry), GFP_KERNEL);
 	if (!array)
 		return -ENOMEM;
 	memset(array, 0, num_entries * sizeof(struct amdgpu_bo_list_entry));
@@ -148,7 +148,7 @@ static int amdgpu_bo_list_set(struct amdgpu_device *adev,
 	for (i = 0; i < list->num_entries; ++i)
 		amdgpu_bo_unref(&list->array[i].robj);
 
-	drm_free_large(list->array);
+	kvfree(list->array);
 
 	list->gds_obj = gds_obj;
 	list->gws_obj = gws_obj;
@@ -163,7 +163,7 @@ static int amdgpu_bo_list_set(struct amdgpu_device *adev,
 error_free:
 	while (i--)
 		amdgpu_bo_unref(&array[i].robj);
-	drm_free_large(array);
+	kvfree(array);
 	return r;
 }
 
@@ -224,7 +224,7 @@ void amdgpu_bo_list_free(struct amdgpu_bo_list *list)
 		amdgpu_bo_unref(&list->array[i].robj);
 
 	mutex_destroy(&list->lock);
-	drm_free_large(list->array);
+	kvfree(list->array);
 	kfree(list);
 }
 
@@ -244,8 +244,8 @@ int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data,
 
 	int r;
 
-	info = drm_malloc_ab(args->in.bo_number,
-			     sizeof(struct drm_amdgpu_bo_list_entry));
+	info = kvmalloc_array(args->in.bo_number,
+			     sizeof(struct drm_amdgpu_bo_list_entry), GFP_KERNEL);
 	if (!info)
 		return -ENOMEM;
 
@@ -311,11 +311,11 @@ int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data,
 
 	memset(args, 0, sizeof(*args));
 	args->out.list_handle = handle;
-	drm_free_large(info);
+	kvfree(info);
 
 	return 0;
 
 error_free:
-	drm_free_large(info);
+	kvfree(info);
 	return r;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index 4e6b9501ab0a..5b3e0f63a115 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -194,7 +194,7 @@ int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data)
 		size = p->chunks[i].length_dw;
 		cdata = (void __user *)(uintptr_t)user_chunk.chunk_data;
 
-		p->chunks[i].kdata = drm_malloc_ab(size, sizeof(uint32_t));
+		p->chunks[i].kdata = kvmalloc_array(size, sizeof(uint32_t), GFP_KERNEL);
 		if (p->chunks[i].kdata == NULL) {
 			ret = -ENOMEM;
 			i--;
@@ -247,7 +247,7 @@ free_all_kdata:
 	i = p->nchunks - 1;
 free_partial_kdata:
 	for (; i >= 0; i--)
-		drm_free_large(p->chunks[i].kdata);
+		kvfree(p->chunks[i].kdata);
 	kfree(p->chunks);
 	p->chunks = NULL;
 	p->nchunks = 0;
@@ -505,7 +505,7 @@ static int amdgpu_cs_list_validate(struct amdgpu_cs_parser *p,
 			return r;
 
 		if (binding_userptr) {
-			drm_free_large(lobj->user_pages);
+			kvfree(lobj->user_pages);
 			lobj->user_pages = NULL;
 		}
 	}
@@ -571,7 +571,7 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 				release_pages(e->user_pages,
 					      e->robj->tbo.ttm->num_pages,
 					      false);
-				drm_free_large(e->user_pages);
+				kvfree(e->user_pages);
 				e->user_pages = NULL;
 			}
 
@@ -601,8 +601,9 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 		list_for_each_entry(e, &need_pages, tv.head) {
 			struct ttm_tt *ttm = e->robj->tbo.ttm;
 
-			e->user_pages = drm_calloc_large(ttm->num_pages,
-							 sizeof(struct page*));
+			e->user_pages = kvmalloc_array(ttm->num_pages,
+							 sizeof(struct page*),
+							 GFP_KERNEL | __GFP_ZERO);
 			if (!e->user_pages) {
 				r = -ENOMEM;
 				DRM_ERROR("calloc failure in %s\n", __func__);
@@ -612,7 +613,7 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 			r = amdgpu_ttm_tt_get_user_pages(ttm, e->user_pages);
 			if (r) {
 				DRM_ERROR("amdgpu_ttm_tt_get_user_pages failed.\n");
-				drm_free_large(e->user_pages);
+				kvfree(e->user_pages);
 				e->user_pages = NULL;
 				goto error_free_pages;
 			}
@@ -708,7 +709,7 @@ error_free_pages:
 			release_pages(e->user_pages,
 				      e->robj->tbo.ttm->num_pages,
 				      false);
-			drm_free_large(e->user_pages);
+			kvfree(e->user_pages);
 		}
 	}
 
@@ -761,7 +762,7 @@ static void amdgpu_cs_parser_fini(struct amdgpu_cs_parser *parser, int error, bo
 		amdgpu_bo_list_put(parser->bo_list);
 
 	for (i = 0; i < parser->nchunks; i++)
-		drm_free_large(parser->chunks[i].kdata);
+		kvfree(parser->chunks[i].kdata);
 	kfree(parser->chunks);
 	if (parser->job)
 		amdgpu_job_free(parser->job);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c
index 38e9b0d3659a..1cb52fd19060 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c
@@ -22,7 +22,7 @@
  * Authors: Alex Deucher
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_atombios.h"
 #include "amdgpu_i2c.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index f2d705e6a75a..31eddd85eb40 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -39,7 +39,7 @@
 #include <linux/module.h>
 #include <linux/pm_runtime.h>
 #include <linux/vga_switcheroo.h>
-#include "drm_crtc_helper.h"
+#include <drm/drm_crtc_helper.h>
 
 #include "amdgpu.h"
 #include "amdgpu_irq.h"
@@ -715,6 +715,16 @@ static const struct file_operations amdgpu_driver_kms_fops = {
 #endif
 };
 
+static bool
+amdgpu_get_crtc_scanout_position(struct drm_device *dev, unsigned int pipe,
+				 bool in_vblank_irq, int *vpos, int *hpos,
+				 ktime_t *stime, ktime_t *etime,
+				 const struct drm_display_mode *mode)
+{
+	return amdgpu_get_crtc_scanoutpos(dev, pipe, 0, vpos, hpos,
+					  stime, etime, mode);
+}
+
 static struct drm_driver kms_driver = {
 	.driver_features =
 	    DRIVER_USE_AGP |
@@ -729,8 +739,8 @@ static struct drm_driver kms_driver = {
 	.get_vblank_counter = amdgpu_get_vblank_counter_kms,
 	.enable_vblank = amdgpu_enable_vblank_kms,
 	.disable_vblank = amdgpu_disable_vblank_kms,
-	.get_vblank_timestamp = amdgpu_get_vblank_timestamp_kms,
-	.get_scanout_position = amdgpu_get_crtc_scanoutpos,
+	.get_vblank_timestamp = drm_calc_vbltimestamp_from_scanoutpos,
+	.get_scanout_position = amdgpu_get_crtc_scanout_position,
 #if defined(CONFIG_DEBUG_FS)
 	.debugfs_init = amdgpu_debugfs_init,
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
index 96c341670782..dca4be970d13 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
@@ -945,47 +945,6 @@ void amdgpu_disable_vblank_kms(struct drm_device *dev, unsigned int pipe)
 	amdgpu_irq_put(adev, &adev->crtc_irq, idx);
 }
 
-/**
- * amdgpu_get_vblank_timestamp_kms - get vblank timestamp
- *
- * @dev: drm dev pointer
- * @crtc: crtc to get the timestamp for
- * @max_error: max error
- * @vblank_time: time value
- * @flags: flags passed to the driver
- *
- * Gets the timestamp on the requested crtc based on the
- * scanout position.  (all asics).
- * Returns postive status flags on success, negative error on failure.
- */
-int amdgpu_get_vblank_timestamp_kms(struct drm_device *dev, unsigned int pipe,
-				    int *max_error,
-				    struct timeval *vblank_time,
-				    unsigned flags)
-{
-	struct drm_crtc *crtc;
-	struct amdgpu_device *adev = dev->dev_private;
-
-	if (pipe >= dev->num_crtcs) {
-		DRM_ERROR("Invalid crtc %u\n", pipe);
-		return -EINVAL;
-	}
-
-	/* Get associated drm_crtc: */
-	crtc = &adev->mode_info.crtcs[pipe]->base;
-	if (!crtc) {
-		/* This can occur on driver load if some component fails to
-		 * initialize completely and driver is unloaded */
-		DRM_ERROR("Uninitialized crtc %d\n", pipe);
-		return -EINVAL;
-	}
-
-	/* Helper routine in DRM core does all the work: */
-	return drm_calc_vbltimestamp_from_scanoutpos(dev, pipe, max_error,
-						     vblank_time, flags,
-						     &crtc->hwmode);
-}
-
 const struct drm_ioctl_desc amdgpu_ioctls_kms[] = {
 	DRM_IOCTL_DEF_DRV(AMDGPU_GEM_CREATE, amdgpu_gem_create_ioctl, DRM_AUTH|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(AMDGPU_CTX, amdgpu_ctx_ioctl, DRM_AUTH|DRM_RENDER_ALLOW),
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h
index dbd10618ec20..43a9d3aec6c4 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h
@@ -534,6 +534,9 @@ struct amdgpu_framebuffer {
 				((em) == ATOM_ENCODER_MODE_DP_MST))
 
 /* Driver internal use only flags of amdgpu_get_crtc_scanoutpos() */
+#define DRM_SCANOUTPOS_VALID        (1 << 0)
+#define DRM_SCANOUTPOS_IN_VBLANK    (1 << 1)
+#define DRM_SCANOUTPOS_ACCURATE     (1 << 2)
 #define USE_REAL_VBLANKSTART		(1 << 30)
 #define GET_DISTANCE_TO_VBLANKSTART	(1 << 31)
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
index ac5e92e5d59d..596e3957bdd9 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
@@ -24,7 +24,7 @@
  */
 
 #include <linux/firmware.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_psp.h"
 #include "amdgpu_ucode.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index 5db0230e45c6..b5fa003c1341 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -29,11 +29,11 @@
  *    Thomas Hellstrom <thomas-at-tungstengraphics-dot-com>
  *    Dave Airlie
  */
-#include <ttm/ttm_bo_api.h>
-#include <ttm/ttm_bo_driver.h>
-#include <ttm/ttm_placement.h>
-#include <ttm/ttm_module.h>
-#include <ttm/ttm_page_alloc.h>
+#include <drm/ttm/ttm_bo_api.h>
+#include <drm/ttm/ttm_bo_driver.h>
+#include <drm/ttm/ttm_placement.h>
+#include <drm/ttm/ttm_module.h>
+#include <drm/ttm/ttm_page_alloc.h>
 #include <drm/drmP.h>
 #include <drm/amdgpu_drm.h>
 #include <linux/seq_file.h>
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 8ecf82c5fe74..83c172a6e938 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -279,8 +279,9 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev,
 	if (!parent->entries) {
 		unsigned num_entries = amdgpu_vm_num_entries(adev, level);
 
-		parent->entries = drm_calloc_large(num_entries,
-						   sizeof(struct amdgpu_vm_pt));
+		parent->entries = kvmalloc_array(num_entries,
+						   sizeof(struct amdgpu_vm_pt),
+						   GFP_KERNEL | __GFP_ZERO);
 		if (!parent->entries)
 			return -ENOMEM;
 		memset(parent->entries, 0 , sizeof(struct amdgpu_vm_pt));
@@ -2219,7 +2220,7 @@ static void amdgpu_vm_free_levels(struct amdgpu_vm_pt *level)
 		for (i = 0; i <= level->last_entry_used; i++)
 			amdgpu_vm_free_levels(&level->entries[i]);
 
-	drm_free_large(level->entries);
+	kvfree(level->entries);
 }
 
 /**
diff --git a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
index ec93714e4524..cb508a211b2f 100644
--- a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
@@ -22,7 +22,7 @@
  */
 
 #include <linux/firmware.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_pm.h"
 #include "amdgpu_ucode.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/ci_smc.c b/drivers/gpu/drm/amd/amdgpu/ci_smc.c
index 7eb9069db8e3..b8ba51e045b5 100644
--- a/drivers/gpu/drm/amd/amdgpu/ci_smc.c
+++ b/drivers/gpu/drm/amd/amdgpu/ci_smc.c
@@ -23,7 +23,7 @@
  */
 
 #include <linux/firmware.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "cikd.h"
 #include "ppsmc.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/cik.c b/drivers/gpu/drm/amd/amdgpu/cik.c
index 9d33e5641419..6b2034533f68 100644
--- a/drivers/gpu/drm/amd/amdgpu/cik.c
+++ b/drivers/gpu/drm/amd/amdgpu/cik.c
@@ -24,7 +24,7 @@
 #include <linux/firmware.h>
 #include <linux/slab.h>
 #include <linux/module.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_atombios.h"
 #include "amdgpu_ih.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/cik_ih.c b/drivers/gpu/drm/amd/amdgpu/cik_ih.c
index c57c3f18af01..b8918432c572 100644
--- a/drivers/gpu/drm/amd/amdgpu/cik_ih.c
+++ b/drivers/gpu/drm/amd/amdgpu/cik_ih.c
@@ -20,7 +20,7 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  */
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_ih.h"
 #include "cikd.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/cz_ih.c b/drivers/gpu/drm/amd/amdgpu/cz_ih.c
index a5f294ebff5c..0c1209cdd1cb 100644
--- a/drivers/gpu/drm/amd/amdgpu/cz_ih.c
+++ b/drivers/gpu/drm/amd/amdgpu/cz_ih.c
@@ -20,7 +20,7 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  */
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_ih.h"
 #include "vid.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
index 0cdeb6a2e4a0..3c62c45f43a1 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
@@ -20,7 +20,7 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  */
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_pm.h"
 #include "amdgpu_i2c.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
index 773654a19749..c8ed0facddcd 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
@@ -20,7 +20,7 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  */
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_pm.h"
 #include "amdgpu_i2c.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
index 1f3552967ba3..3f3a25493327 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
@@ -20,7 +20,7 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  */
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_pm.h"
 #include "amdgpu_i2c.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
index 3c558c170e5e..3e90c19b9c7f 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
@@ -20,7 +20,7 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  */
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_pm.h"
 #include "amdgpu_i2c.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
index f1b479b6ac98..90bb08309a53 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
@@ -20,7 +20,7 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  */
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_pm.h"
 #include "amdgpu_i2c.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
index ee2f2139e2eb..f7414cabd4ff 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
@@ -21,7 +21,7 @@
  *
  */
 #include <linux/firmware.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_ih.h"
 #include "amdgpu_gfx.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
index 758d636a6f52..404d12785853 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@@ -21,7 +21,7 @@
  *
  */
 #include <linux/firmware.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_gfx.h"
 #include "vi.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 0c16b7563b73..125b11950071 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -21,7 +21,7 @@
  *
  */
 #include <linux/firmware.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_gfx.h"
 #include "soc15.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
index d860939152df..9776ad3d2d71 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
@@ -21,7 +21,7 @@
  *
  */
 #include <linux/firmware.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "gmc_v6_0.h"
 #include "amdgpu_ucode.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
index 2750e5c23813..fca8e77182c9 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
@@ -21,7 +21,7 @@
  *
  */
 #include <linux/firmware.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "cikd.h"
 #include "cik.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
index f56b4089ee9f..e9c127037b39 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
@@ -21,7 +21,7 @@
  *
  */
 #include <linux/firmware.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "gmc_v8_0.h"
 #include "amdgpu_ucode.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/iceland_ih.c b/drivers/gpu/drm/amd/amdgpu/iceland_ih.c
index cb622add99a7..7a0ea27ac429 100644
--- a/drivers/gpu/drm/amd/amdgpu/iceland_ih.c
+++ b/drivers/gpu/drm/amd/amdgpu/iceland_ih.c
@@ -20,7 +20,7 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  */
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_ih.h"
 #include "vid.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
index 79a52ad2c80d..3bbf2ccfca89 100644
--- a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
@@ -21,7 +21,7 @@
  *
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_pm.h"
 #include "cikd.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/kv_smc.c b/drivers/gpu/drm/amd/amdgpu/kv_smc.c
index e6b7b42acfe1..b82e33c01571 100644
--- a/drivers/gpu/drm/amd/amdgpu/kv_smc.c
+++ b/drivers/gpu/drm/amd/amdgpu/kv_smc.c
@@ -22,7 +22,7 @@
  * Authors: Alex Deucher
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "cikd.h"
 #include "kv_dpm.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c
index 60a6407ba267..eef89abc0cee 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c
+++ b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c
@@ -24,7 +24,7 @@
  */
 
 #include <linux/firmware.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_psp.h"
 #include "amdgpu_ucode.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/si.c b/drivers/gpu/drm/amd/amdgpu/si.c
index c0b1aabf282f..2431639baf47 100644
--- a/drivers/gpu/drm/amd/amdgpu/si.c
+++ b/drivers/gpu/drm/amd/amdgpu/si.c
@@ -24,7 +24,7 @@
 #include <linux/firmware.h>
 #include <linux/slab.h>
 #include <linux/module.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_atombios.h"
 #include "amdgpu_ih.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/si_dpm.c b/drivers/gpu/drm/amd/amdgpu/si_dpm.c
index 7c1c5d127281..a7ad8390981c 100644
--- a/drivers/gpu/drm/amd/amdgpu/si_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/si_dpm.c
@@ -21,7 +21,7 @@
  *
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_pm.h"
 #include "amdgpu_dpm.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/si_ih.c b/drivers/gpu/drm/amd/amdgpu/si_ih.c
index e66084211c74..ce25e03a077d 100644
--- a/drivers/gpu/drm/amd/amdgpu/si_ih.c
+++ b/drivers/gpu/drm/amd/amdgpu/si_ih.c
@@ -20,7 +20,7 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  */
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_ih.h"
 #include "sid.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/si_smc.c b/drivers/gpu/drm/amd/amdgpu/si_smc.c
index 0726bc3b6f90..4a2fd8b61940 100644
--- a/drivers/gpu/drm/amd/amdgpu/si_smc.c
+++ b/drivers/gpu/drm/amd/amdgpu/si_smc.c
@@ -23,7 +23,7 @@
  */
 
 #include <linux/firmware.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "sid.h"
 #include "ppsmc.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index 6b55d451ae7f..e945f8b07487 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -23,7 +23,7 @@
 #include <linux/firmware.h>
 #include <linux/slab.h>
 #include <linux/module.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_atomfirmware.h"
 #include "amdgpu_ih.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/tonga_ih.c b/drivers/gpu/drm/amd/amdgpu/tonga_ih.c
index 3a5097ac2bb4..923df2c0e535 100644
--- a/drivers/gpu/drm/amd/amdgpu/tonga_ih.c
+++ b/drivers/gpu/drm/amd/amdgpu/tonga_ih.c
@@ -20,7 +20,7 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  */
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_ih.h"
 #include "vid.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/vega10_ih.c b/drivers/gpu/drm/amd/amdgpu/vega10_ih.c
index 071f56e439bb..3b9740fb2c41 100644
--- a/drivers/gpu/drm/amd/amdgpu/vega10_ih.c
+++ b/drivers/gpu/drm/amd/amdgpu/vega10_ih.c
@@ -20,7 +20,7 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  */
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_ih.h"
 #include "soc15.h"
diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c
index b1132f5e84fc..3a187619286f 100644
--- a/drivers/gpu/drm/amd/amdgpu/vi.c
+++ b/drivers/gpu/drm/amd/amdgpu/vi.c
@@ -21,7 +21,7 @@
  *
  */
 #include <linux/slab.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "amdgpu.h"
 #include "amdgpu_atombios.h"
 #include "amdgpu_ih.h"
diff --git a/drivers/gpu/drm/amd/amdkfd/Makefile b/drivers/gpu/drm/amd/amdkfd/Makefile
index 7fc9b0f444cb..b400d5664252 100644
--- a/drivers/gpu/drm/amd/amdkfd/Makefile
+++ b/drivers/gpu/drm/amd/amdkfd/Makefile
@@ -2,7 +2,7 @@
 # Makefile for Heterogenous System Architecture support for AMD GPU devices
 #
 
-ccflags-y := -Iinclude/drm -Idrivers/gpu/drm/amd/include/  \
+ccflags-y := -Idrivers/gpu/drm/amd/include/  \
 		-Idrivers/gpu/drm/amd/include/asic_reg
 
 amdkfd-y	:= kfd_module.o kfd_device.o kfd_chardev.o kfd_topology.o \
diff --git a/drivers/gpu/drm/amd/powerplay/Makefile b/drivers/gpu/drm/amd/powerplay/Makefile
index 043e6ebab575..4e132b936e3d 100644
--- a/drivers/gpu/drm/amd/powerplay/Makefile
+++ b/drivers/gpu/drm/amd/powerplay/Makefile
@@ -1,5 +1,5 @@
 
-subdir-ccflags-y += -Iinclude/drm  \
+subdir-ccflags-y += \
 		-I$(FULL_AMD_PATH)/powerplay/inc/  \
 		-I$(FULL_AMD_PATH)/include/asic_reg  \
 		-I$(FULL_AMD_PATH)/include  \
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
index ff4ae3de6bb6..963a9e017a28 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
@@ -22,10 +22,10 @@
  */
 
 #include "pp_debug.h"
-#include "linux/delay.h"
-#include <linux/types.h>
+#include <linux/delay.h>
 #include <linux/kernel.h>
 #include <linux/slab.h>
+#include <linux/types.h>
 #include <drm/amdgpu_drm.h>
 #include "cgs_common.h"
 #include "power_state.h"
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/pp_acpi.c b/drivers/gpu/drm/amd/powerplay/hwmgr/pp_acpi.c
index f5e8fda964f7..f6b4dd96c0ec 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/pp_acpi.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/pp_acpi.c
@@ -21,8 +21,8 @@
  *
  */
 
+#include <linux/delay.h>
 #include <linux/errno.h>
-#include "linux/delay.h"
 #include "hwmgr.h"
 #include "amd_acpi.h"
 #include "pp_acpi.h"
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 102eb6d029fa..1f01020ce3a9 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -21,11 +21,11 @@
  *
  */
 #include "pp_debug.h"
+#include <linux/delay.h>
+#include <linux/fb.h>
 #include <linux/module.h>
 #include <linux/slab.h>
-#include <linux/fb.h>
 #include <asm/div64.h>
-#include "linux/delay.h"
 #include "pp_acpi.h"
 #include "ppatomctrl.h"
 #include "atombios.h"
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
index 2614af2f553f..ab17350e853d 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
@@ -20,10 +20,11 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  */
+
+#include <linux/delay.h>
+#include <linux/fb.h>
 #include <linux/module.h>
 #include <linux/slab.h>
-#include <linux/fb.h>
-#include "linux/delay.h"
 
 #include "hwmgr.h"
 #include "amd_powerplay.h"
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
index 1f6744a443d4..39c7091866e8 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
@@ -20,11 +20,13 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  */
-#include <linux/types.h>
+
+#include <linux/delay.h>
+#include <linux/gfp.h>
 #include <linux/kernel.h>
 #include <linux/slab.h>
-#include <linux/gfp.h>
-#include "linux/delay.h"
+#include <linux/types.h>
+
 #include "cgs_common.h"
 #include "smu/smu_8_0_d.h"
 #include "smu/smu_8_0_sh_mask.h"
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c
index c0d75766bbc8..2e954a44bac1 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c
@@ -20,15 +20,16 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  *
  */
-#include <linux/types.h>
+
+#include <linux/delay.h>
 #include <linux/kernel.h>
 #include <linux/module.h>
 #include <linux/slab.h>
+#include <linux/types.h>
 #include <drm/amdgpu_drm.h>
 #include "pp_instance.h"
 #include "smumgr.h"
 #include "cgs_common.h"
-#include "linux/delay.h"
 
 MODULE_FIRMWARE("amdgpu/topaz_smc.bin");
 MODULE_FIRMWARE("amdgpu/topaz_k_smc.bin");
diff --git a/drivers/gpu/drm/arm/malidp_drv.h b/drivers/gpu/drm/arm/malidp_drv.h
index 040311ffcaec..2e2033140efc 100644
--- a/drivers/gpu/drm/arm/malidp_drv.h
+++ b/drivers/gpu/drm/arm/malidp_drv.h
@@ -65,6 +65,6 @@ void malidp_de_planes_destroy(struct drm_device *drm);
 int malidp_crtc_init(struct drm_device *drm);
 
 /* often used combination of rotational bits */
-#define MALIDP_ROTATED_MASK	(DRM_ROTATE_90 | DRM_ROTATE_270)
+#define MALIDP_ROTATED_MASK	(DRM_MODE_ROTATE_90 | DRM_MODE_ROTATE_270)
 
 #endif  /* __MALIDP_DRV_H__ */
diff --git a/drivers/gpu/drm/arm/malidp_planes.c b/drivers/gpu/drm/arm/malidp_planes.c
index 814fda23cead..063a8d2b0be3 100644
--- a/drivers/gpu/drm/arm/malidp_planes.c
+++ b/drivers/gpu/drm/arm/malidp_planes.c
@@ -80,7 +80,7 @@ static void malidp_plane_reset(struct drm_plane *plane)
 	state = kzalloc(sizeof(*state), GFP_KERNEL);
 	if (state) {
 		state->base.plane = plane;
-		state->base.rotation = DRM_ROTATE_0;
+		state->base.rotation = DRM_MODE_ROTATE_0;
 		plane->state = &state->base;
 	}
 }
@@ -221,7 +221,7 @@ static int malidp_de_plane_check(struct drm_plane *plane,
 		return ret;
 
 	/* packed RGB888 / BGR888 can't be rotated or flipped */
-	if (state->rotation != DRM_ROTATE_0 &&
+	if (state->rotation != DRM_MODE_ROTATE_0 &&
 	    (fb->format->format == DRM_FORMAT_RGB888 ||
 	     fb->format->format == DRM_FORMAT_BGR888))
 		return -EINVAL;
@@ -315,12 +315,12 @@ static void malidp_de_plane_update(struct drm_plane *plane,
 	val &= ~LAYER_ROT_MASK;
 
 	/* setup the rotation and axis flip bits */
-	if (plane->state->rotation & DRM_ROTATE_MASK)
-		val |= ilog2(plane->state->rotation & DRM_ROTATE_MASK) <<
+	if (plane->state->rotation & DRM_MODE_ROTATE_MASK)
+		val |= ilog2(plane->state->rotation & DRM_MODE_ROTATE_MASK) <<
 		       LAYER_ROT_OFFSET;
-	if (plane->state->rotation & DRM_REFLECT_X)
+	if (plane->state->rotation & DRM_MODE_REFLECT_X)
 		val |= LAYER_H_FLIP;
-	if (plane->state->rotation & DRM_REFLECT_Y)
+	if (plane->state->rotation & DRM_MODE_REFLECT_Y)
 		val |= LAYER_V_FLIP;
 
 	/*
@@ -370,8 +370,8 @@ int malidp_de_planes_init(struct drm_device *drm)
 	struct malidp_plane *plane = NULL;
 	enum drm_plane_type plane_type;
 	unsigned long crtcs = 1 << drm->mode_config.num_crtc;
-	unsigned long flags = DRM_ROTATE_0 | DRM_ROTATE_90 | DRM_ROTATE_180 |
-			      DRM_ROTATE_270 | DRM_REFLECT_X | DRM_REFLECT_Y;
+	unsigned long flags = DRM_MODE_ROTATE_0 | DRM_MODE_ROTATE_90 | DRM_MODE_ROTATE_180 |
+			      DRM_MODE_ROTATE_270 | DRM_MODE_REFLECT_X | DRM_MODE_REFLECT_Y;
 	u32 *formats;
 	int ret, i, j, n;
 
@@ -420,7 +420,7 @@ int malidp_de_planes_init(struct drm_device *drm)
 			continue;
 		}
 
-		drm_plane_create_rotation_property(&plane->base, DRM_ROTATE_0, flags);
+		drm_plane_create_rotation_property(&plane->base, DRM_MODE_ROTATE_0, flags);
 		malidp_hw_write(malidp->dev, MALIDP_ALPHA_LUT,
 				plane->layer->base + MALIDP_LAYER_COMPOSE);
 	}
diff --git a/drivers/gpu/drm/armada/armada_overlay.c b/drivers/gpu/drm/armada/armada_overlay.c
index 424e465ff407..e9a29df4b443 100644
--- a/drivers/gpu/drm/armada/armada_overlay.c
+++ b/drivers/gpu/drm/armada/armada_overlay.c
@@ -125,7 +125,7 @@ armada_ovl_plane_update(struct drm_plane *plane, struct drm_crtc *crtc,
 				 src_x, src_y, src_w, src_h);
 
 	ret = drm_plane_helper_check_update(plane, crtc, fb, &src, &dest, &clip,
-					    DRM_ROTATE_0,
+					    DRM_MODE_ROTATE_0,
 					    0, INT_MAX, true, false, &visible);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/ast/Makefile b/drivers/gpu/drm/ast/Makefile
index 171aa0622b66..617fdd39519c 100644
--- a/drivers/gpu/drm/ast/Makefile
+++ b/drivers/gpu/drm/ast/Makefile
@@ -2,8 +2,6 @@
 # Makefile for the drm device driver.  This driver provides support for the
 # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.
 
-ccflags-y := -Iinclude/drm
-
 ast-y := ast_drv.o ast_main.o ast_mode.o ast_fb.o ast_ttm.o ast_post.o ast_dp501.o
 
 obj-$(CONFIG_DRM_AST) := ast.o
diff --git a/drivers/gpu/drm/ast/ast_ttm.c b/drivers/gpu/drm/ast/ast_ttm.c
index e879496b8a42..58084985e6cf 100644
--- a/drivers/gpu/drm/ast/ast_ttm.c
+++ b/drivers/gpu/drm/ast/ast_ttm.c
@@ -26,8 +26,9 @@
  * Authors: Dave Airlie <airlied@redhat.com>
  */
 #include <drm/drmP.h>
+#include <drm/ttm/ttm_page_alloc.h>
+
 #include "ast_drv.h"
-#include <ttm/ttm_page_alloc.h>
 
 static inline struct ast_private *
 ast_bdev(struct ttm_bo_device *bd)
diff --git a/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c b/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c
index 29cc10d053eb..1124200bb280 100644
--- a/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c
+++ b/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c
@@ -678,8 +678,8 @@ static int atmel_hlcdc_plane_atomic_check(struct drm_plane *p,
 		if (!state->bpp[i])
 			return -EINVAL;
 
-		switch (state->base.rotation & DRM_ROTATE_MASK) {
-		case DRM_ROTATE_90:
+		switch (state->base.rotation & DRM_MODE_ROTATE_MASK) {
+		case DRM_MODE_ROTATE_90:
 			offset = ((y_offset + state->src_y + patched_src_w - 1) /
 				  ydiv) * fb->pitches[i];
 			offset += ((x_offset + state->src_x) / xdiv) *
@@ -688,7 +688,7 @@ static int atmel_hlcdc_plane_atomic_check(struct drm_plane *p,
 					  fb->pitches[i];
 			state->pstride[i] = -fb->pitches[i] - state->bpp[i];
 			break;
-		case DRM_ROTATE_180:
+		case DRM_MODE_ROTATE_180:
 			offset = ((y_offset + state->src_y + patched_src_h - 1) /
 				  ydiv) * fb->pitches[i];
 			offset += ((x_offset + state->src_x + patched_src_w - 1) /
@@ -697,7 +697,7 @@ static int atmel_hlcdc_plane_atomic_check(struct drm_plane *p,
 					   state->bpp[i]) - fb->pitches[i];
 			state->pstride[i] = -2 * state->bpp[i];
 			break;
-		case DRM_ROTATE_270:
+		case DRM_MODE_ROTATE_270:
 			offset = ((y_offset + state->src_y) / ydiv) *
 				 fb->pitches[i];
 			offset += ((x_offset + state->src_x + patched_src_h - 1) /
@@ -707,7 +707,7 @@ static int atmel_hlcdc_plane_atomic_check(struct drm_plane *p,
 					  (2 * state->bpp[i]);
 			state->pstride[i] = fb->pitches[i] - state->bpp[i];
 			break;
-		case DRM_ROTATE_0:
+		case DRM_MODE_ROTATE_0:
 		default:
 			offset = ((y_offset + state->src_y) / ydiv) *
 				 fb->pitches[i];
@@ -864,11 +864,11 @@ static int atmel_hlcdc_plane_init_properties(struct atmel_hlcdc_plane *plane,
 		int ret;
 
 		ret = drm_plane_create_rotation_property(&plane->base,
-							 DRM_ROTATE_0,
-							 DRM_ROTATE_0 |
-							 DRM_ROTATE_90 |
-							 DRM_ROTATE_180 |
-							 DRM_ROTATE_270);
+							 DRM_MODE_ROTATE_0,
+							 DRM_MODE_ROTATE_0 |
+							 DRM_MODE_ROTATE_90 |
+							 DRM_MODE_ROTATE_180 |
+							 DRM_MODE_ROTATE_270);
 		if (ret)
 			return ret;
 	}
diff --git a/drivers/gpu/drm/bochs/Makefile b/drivers/gpu/drm/bochs/Makefile
index 844a55614920..98ef60a19e8f 100644
--- a/drivers/gpu/drm/bochs/Makefile
+++ b/drivers/gpu/drm/bochs/Makefile
@@ -1,4 +1,3 @@
-ccflags-y := -Iinclude/drm
 bochs-drm-y := bochs_drv.o bochs_mm.o bochs_kms.o bochs_fbdev.o bochs_hw.o
 
 obj-$(CONFIG_DRM_BOCHS)	+= bochs-drm.o
diff --git a/drivers/gpu/drm/bochs/bochs.h b/drivers/gpu/drm/bochs/bochs.h
index f626bab7f5e3..76c490c3cdbc 100644
--- a/drivers/gpu/drm/bochs/bochs.h
+++ b/drivers/gpu/drm/bochs/bochs.h
@@ -9,8 +9,8 @@
 
 #include <drm/drm_gem.h>
 
-#include <ttm/ttm_bo_driver.h>
-#include <ttm/ttm_page_alloc.h>
+#include <drm/ttm/ttm_bo_driver.h>
+#include <drm/ttm/ttm_page_alloc.h>
 
 /* ---------------------------------------------------------------------- */
 
diff --git a/drivers/gpu/drm/bridge/Makefile b/drivers/gpu/drm/bridge/Makefile
index 3fe2226ee2f2..defcf1e7ca1c 100644
--- a/drivers/gpu/drm/bridge/Makefile
+++ b/drivers/gpu/drm/bridge/Makefile
@@ -1,5 +1,3 @@
-ccflags-y := -Iinclude/drm
-
 obj-$(CONFIG_DRM_ANALOGIX_ANX78XX) += analogix-anx78xx.o
 obj-$(CONFIG_DRM_DUMB_VGA_DAC) += dumb-vga-dac.o
 obj-$(CONFIG_DRM_LVDS_ENCODER) += lvds-encoder.o
diff --git a/drivers/gpu/drm/bridge/nxp-ptn3460.c b/drivers/gpu/drm/bridge/nxp-ptn3460.c
index 351704390d02..4f64e717e01b 100644
--- a/drivers/gpu/drm/bridge/nxp-ptn3460.c
+++ b/drivers/gpu/drm/bridge/nxp-ptn3460.c
@@ -20,15 +20,13 @@
 #include <linux/module.h>
 #include <linux/of.h>
 #include <linux/of_gpio.h>
-
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_edid.h>
 #include <drm/drm_of.h>
 #include <drm/drm_panel.h>
-
-#include "drm_crtc.h"
-#include "drm_crtc_helper.h"
-#include "drm_atomic_helper.h"
-#include "drm_edid.h"
-#include "drmP.h"
+#include <drm/drmP.h>
 
 #define PTN3460_EDID_ADDR			0x0
 #define PTN3460_EDID_EMULATION_ADDR		0x84
diff --git a/drivers/gpu/drm/bridge/parade-ps8622.c b/drivers/gpu/drm/bridge/parade-ps8622.c
index 1dcec3b97e67..6f22f9fec9bf 100644
--- a/drivers/gpu/drm/bridge/parade-ps8622.c
+++ b/drivers/gpu/drm/bridge/parade-ps8622.c
@@ -24,14 +24,12 @@
 #include <linux/of_device.h>
 #include <linux/pm.h>
 #include <linux/regulator/consumer.h>
-
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
 #include <drm/drm_of.h>
 #include <drm/drm_panel.h>
-
-#include "drmP.h"
-#include "drm_crtc.h"
-#include "drm_crtc_helper.h"
-#include "drm_atomic_helper.h"
+#include <drm/drmP.h>
 
 /* Brightness scale on the Parade chip */
 #define PS8622_MAX_BRIGHTNESS 0xff
diff --git a/drivers/gpu/drm/bridge/sii902x.c b/drivers/gpu/drm/bridge/sii902x.c
index 9126d0306ab5..9b87067c022c 100644
--- a/drivers/gpu/drm/bridge/sii902x.c
+++ b/drivers/gpu/drm/bridge/sii902x.c
@@ -160,7 +160,7 @@ static int sii902x_get_modes(struct drm_connector *connector)
 		 time_before(jiffies, timeout));
 
 	if (!(status & SII902X_SYS_CTRL_DDC_BUS_GRTD)) {
-		dev_err(&sii902x->i2c->dev, "failed to acquire the i2c bus");
+		dev_err(&sii902x->i2c->dev, "failed to acquire the i2c bus\n");
 		return -ETIMEDOUT;
 	}
 
@@ -202,7 +202,7 @@ static int sii902x_get_modes(struct drm_connector *connector)
 
 	if (status & (SII902X_SYS_CTRL_DDC_BUS_REQ |
 		      SII902X_SYS_CTRL_DDC_BUS_GRTD)) {
-		dev_err(&sii902x->i2c->dev, "failed to release the i2c bus");
+		dev_err(&sii902x->i2c->dev, "failed to release the i2c bus\n");
 		return -ETIMEDOUT;
 	}
 
@@ -298,7 +298,7 @@ static int sii902x_bridge_attach(struct drm_bridge *bridge)
 
 	if (!drm_core_check_feature(drm, DRIVER_ATOMIC)) {
 		dev_err(&sii902x->i2c->dev,
-			"sii902x driver is only compatible with DRM devices supporting atomic updates");
+			"sii902x driver is only compatible with DRM devices supporting atomic updates\n");
 		return -ENOTSUPP;
 	}
 
diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
index 4e1f54a675d8..8737de8c1c52 100644
--- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
+++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
@@ -173,6 +173,8 @@ struct dw_hdmi {
 
 	unsigned int reg_shift;
 	struct regmap *regm;
+	void (*enable_audio)(struct dw_hdmi *hdmi);
+	void (*disable_audio)(struct dw_hdmi *hdmi);
 };
 
 #define HDMI_IH_PHY_STAT0_RX_SENSE \
@@ -542,13 +544,41 @@ void dw_hdmi_set_sample_rate(struct dw_hdmi *hdmi, unsigned int rate)
 }
 EXPORT_SYMBOL_GPL(dw_hdmi_set_sample_rate);
 
+static void hdmi_enable_audio_clk(struct dw_hdmi *hdmi, bool enable)
+{
+	hdmi_modb(hdmi, enable ? 0 : HDMI_MC_CLKDIS_AUDCLK_DISABLE,
+		  HDMI_MC_CLKDIS_AUDCLK_DISABLE, HDMI_MC_CLKDIS);
+}
+
+static void dw_hdmi_ahb_audio_enable(struct dw_hdmi *hdmi)
+{
+	hdmi_set_cts_n(hdmi, hdmi->audio_cts, hdmi->audio_n);
+}
+
+static void dw_hdmi_ahb_audio_disable(struct dw_hdmi *hdmi)
+{
+	hdmi_set_cts_n(hdmi, hdmi->audio_cts, 0);
+}
+
+static void dw_hdmi_i2s_audio_enable(struct dw_hdmi *hdmi)
+{
+	hdmi_set_cts_n(hdmi, hdmi->audio_cts, hdmi->audio_n);
+	hdmi_enable_audio_clk(hdmi, true);
+}
+
+static void dw_hdmi_i2s_audio_disable(struct dw_hdmi *hdmi)
+{
+	hdmi_enable_audio_clk(hdmi, false);
+}
+
 void dw_hdmi_audio_enable(struct dw_hdmi *hdmi)
 {
 	unsigned long flags;
 
 	spin_lock_irqsave(&hdmi->audio_lock, flags);
 	hdmi->audio_enable = true;
-	hdmi_set_cts_n(hdmi, hdmi->audio_cts, hdmi->audio_n);
+	if (hdmi->enable_audio)
+		hdmi->enable_audio(hdmi);
 	spin_unlock_irqrestore(&hdmi->audio_lock, flags);
 }
 EXPORT_SYMBOL_GPL(dw_hdmi_audio_enable);
@@ -559,7 +589,8 @@ void dw_hdmi_audio_disable(struct dw_hdmi *hdmi)
 
 	spin_lock_irqsave(&hdmi->audio_lock, flags);
 	hdmi->audio_enable = false;
-	hdmi_set_cts_n(hdmi, hdmi->audio_cts, 0);
+	if (hdmi->disable_audio)
+		hdmi->disable_audio(hdmi);
 	spin_unlock_irqrestore(&hdmi->audio_lock, flags);
 }
 EXPORT_SYMBOL_GPL(dw_hdmi_audio_disable);
@@ -1573,11 +1604,6 @@ static void dw_hdmi_enable_video_path(struct dw_hdmi *hdmi)
 			    HDMI_MC_FLOWCTRL);
 }
 
-static void hdmi_enable_audio_clk(struct dw_hdmi *hdmi)
-{
-	hdmi_modb(hdmi, 0, HDMI_MC_CLKDIS_AUDCLK_DISABLE, HDMI_MC_CLKDIS);
-}
-
 /* Workaround to clear the overflow condition */
 static void dw_hdmi_clear_overflow(struct dw_hdmi *hdmi)
 {
@@ -1691,7 +1717,7 @@ static int dw_hdmi_setup(struct dw_hdmi *hdmi, struct drm_display_mode *mode)
 
 		/* HDMI Initialization Step E - Configure audio */
 		hdmi_clk_regenerator_update_pixel_clock(hdmi);
-		hdmi_enable_audio_clk(hdmi);
+		hdmi_enable_audio_clk(hdmi, true);
 	}
 
 	/* not for DVI mode */
@@ -2403,6 +2429,8 @@ __dw_hdmi_probe(struct platform_device *pdev,
 		audio.irq = irq;
 		audio.hdmi = hdmi;
 		audio.eld = hdmi->connector.eld;
+		hdmi->enable_audio = dw_hdmi_ahb_audio_enable;
+		hdmi->disable_audio = dw_hdmi_ahb_audio_disable;
 
 		pdevinfo.name = "dw-hdmi-ahb-audio";
 		pdevinfo.data = &audio;
@@ -2415,6 +2443,8 @@ __dw_hdmi_probe(struct platform_device *pdev,
 		audio.hdmi	= hdmi;
 		audio.write	= hdmi_writeb;
 		audio.read	= hdmi_readb;
+		hdmi->enable_audio = dw_hdmi_i2s_audio_enable;
+		hdmi->disable_audio = dw_hdmi_i2s_audio_disable;
 
 		pdevinfo.name = "dw-hdmi-i2s-audio";
 		pdevinfo.data = &audio;
diff --git a/drivers/gpu/drm/cirrus/Makefile b/drivers/gpu/drm/cirrus/Makefile
index 69ffe7006d55..919c0a336c97 100644
--- a/drivers/gpu/drm/cirrus/Makefile
+++ b/drivers/gpu/drm/cirrus/Makefile
@@ -1,4 +1,3 @@
-ccflags-y := -Iinclude/drm
 cirrus-y  := cirrus_main.o cirrus_mode.o \
 	cirrus_drv.o cirrus_fbdev.o cirrus_ttm.o
 
diff --git a/drivers/gpu/drm/cirrus/cirrus_ttm.c b/drivers/gpu/drm/cirrus/cirrus_ttm.c
index 93dbcd38355d..1ff1838c0d44 100644
--- a/drivers/gpu/drm/cirrus/cirrus_ttm.c
+++ b/drivers/gpu/drm/cirrus/cirrus_ttm.c
@@ -26,8 +26,9 @@
  * Authors: Dave Airlie <airlied@redhat.com>
  */
 #include <drm/drmP.h>
+#include <drm/ttm/ttm_page_alloc.h>
+
 #include "cirrus_drv.h"
-#include <ttm/ttm_page_alloc.h>
 
 static inline struct cirrus_device *
 cirrus_bdev(struct ttm_bo_device *bd)
diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index f32506a7c1d6..e1637011e18a 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -57,6 +57,7 @@ void drm_atomic_state_default_release(struct drm_atomic_state *state)
 	kfree(state->connectors);
 	kfree(state->crtcs);
 	kfree(state->planes);
+	kfree(state->private_objs);
 }
 EXPORT_SYMBOL(drm_atomic_state_default_release);
 
@@ -184,6 +185,17 @@ void drm_atomic_state_default_clear(struct drm_atomic_state *state)
 		state->planes[i].ptr = NULL;
 		state->planes[i].state = NULL;
 	}
+
+	for (i = 0; i < state->num_private_objs; i++) {
+		void *obj_state = state->private_objs[i].obj_state;
+
+		state->private_objs[i].funcs->destroy_state(obj_state);
+		state->private_objs[i].obj = NULL;
+		state->private_objs[i].obj_state = NULL;
+		state->private_objs[i].funcs = NULL;
+	}
+	state->num_private_objs = 0;
+
 }
 EXPORT_SYMBOL(drm_atomic_state_default_clear);
 
@@ -425,7 +437,7 @@ drm_atomic_replace_property_blob(struct drm_property_blob **blob,
 }
 
 static int
-drm_atomic_replace_property_blob_from_id(struct drm_crtc *crtc,
+drm_atomic_replace_property_blob_from_id(struct drm_device *dev,
 					 struct drm_property_blob **blob,
 					 uint64_t blob_id,
 					 ssize_t expected_size,
@@ -434,7 +446,7 @@ drm_atomic_replace_property_blob_from_id(struct drm_crtc *crtc,
 	struct drm_property_blob *new_blob = NULL;
 
 	if (blob_id != 0) {
-		new_blob = drm_property_lookup_blob(crtc->dev, blob_id);
+		new_blob = drm_property_lookup_blob(dev, blob_id);
 		if (new_blob == NULL)
 			return -EINVAL;
 
@@ -483,7 +495,7 @@ int drm_atomic_crtc_set_property(struct drm_crtc *crtc,
 		drm_property_blob_put(mode);
 		return ret;
 	} else if (property == config->degamma_lut_property) {
-		ret = drm_atomic_replace_property_blob_from_id(crtc,
+		ret = drm_atomic_replace_property_blob_from_id(dev,
 					&state->degamma_lut,
 					val,
 					-1,
@@ -491,7 +503,7 @@ int drm_atomic_crtc_set_property(struct drm_crtc *crtc,
 		state->color_mgmt_changed |= replaced;
 		return ret;
 	} else if (property == config->ctm_property) {
-		ret = drm_atomic_replace_property_blob_from_id(crtc,
+		ret = drm_atomic_replace_property_blob_from_id(dev,
 					&state->ctm,
 					val,
 					sizeof(struct drm_color_ctm),
@@ -499,7 +511,7 @@ int drm_atomic_crtc_set_property(struct drm_crtc *crtc,
 		state->color_mgmt_changed |= replaced;
 		return ret;
 	} else if (property == config->gamma_lut_property) {
-		ret = drm_atomic_replace_property_blob_from_id(crtc,
+		ret = drm_atomic_replace_property_blob_from_id(dev,
 					&state->gamma_lut,
 					val,
 					-1,
@@ -769,7 +781,7 @@ int drm_atomic_plane_set_property(struct drm_plane *plane,
 	} else if (property == config->prop_src_h) {
 		state->src_h = val;
 	} else if (property == plane->rotation_property) {
-		if (!is_power_of_2(val & DRM_ROTATE_MASK))
+		if (!is_power_of_2(val & DRM_MODE_ROTATE_MASK))
 			return -EINVAL;
 		state->rotation = val;
 	} else if (property == plane->zpos_property) {
@@ -978,6 +990,59 @@ static void drm_atomic_plane_print_state(struct drm_printer *p,
 }
 
 /**
+ * drm_atomic_get_private_obj_state - get private object state
+ * @state: global atomic state
+ * @obj: private object to get the state for
+ * @funcs: pointer to the struct of function pointers that identify the object
+ * type
+ *
+ * This function returns the private object state for the given private object,
+ * allocating the state if needed. It does not grab any locks as the caller is
+ * expected to care of any required locking.
+ *
+ * RETURNS:
+ *
+ * Either the allocated state or the error code encoded into a pointer.
+ */
+void *
+drm_atomic_get_private_obj_state(struct drm_atomic_state *state, void *obj,
+			      const struct drm_private_state_funcs *funcs)
+{
+	int index, num_objs, i;
+	size_t size;
+	struct __drm_private_objs_state *arr;
+
+	for (i = 0; i < state->num_private_objs; i++)
+		if (obj == state->private_objs[i].obj &&
+		    state->private_objs[i].obj_state)
+			return state->private_objs[i].obj_state;
+
+	num_objs = state->num_private_objs + 1;
+	size = sizeof(*state->private_objs) * num_objs;
+	arr = krealloc(state->private_objs, size, GFP_KERNEL);
+	if (!arr)
+		return ERR_PTR(-ENOMEM);
+
+	state->private_objs = arr;
+	index = state->num_private_objs;
+	memset(&state->private_objs[index], 0, sizeof(*state->private_objs));
+
+	state->private_objs[index].obj_state = funcs->duplicate_state(state, obj);
+	if (!state->private_objs[index].obj_state)
+		return ERR_PTR(-ENOMEM);
+
+	state->private_objs[index].obj = obj;
+	state->private_objs[index].funcs = funcs;
+	state->num_private_objs = num_objs;
+
+	DRM_DEBUG_ATOMIC("Added new private object state %p to %p\n",
+			 state->private_objs[index].obj_state, state);
+
+	return state->private_objs[index].obj_state;
+}
+EXPORT_SYMBOL(drm_atomic_get_private_obj_state);
+
+/**
  * drm_atomic_get_connector_state - get connector state
  * @state: global atomic state object
  * @connector: connector to get state object for
@@ -1123,6 +1188,10 @@ int drm_atomic_connector_set_property(struct drm_connector *connector,
 		 */
 		if (state->link_status != DRM_LINK_STATUS_GOOD)
 			state->link_status = val;
+	} else if (property == config->aspect_ratio_property) {
+		state->picture_aspect_ratio = val;
+	} else if (property == connector->scaling_mode_property) {
+		state->scaling_mode = val;
 	} else if (connector->funcs->atomic_set_property) {
 		return connector->funcs->atomic_set_property(connector,
 				state, property, val);
@@ -1199,6 +1268,10 @@ drm_atomic_connector_get_property(struct drm_connector *connector,
 		*val = state->tv.hue;
 	} else if (property == config->link_status_property) {
 		*val = state->link_status;
+	} else if (property == config->aspect_ratio_property) {
+		*val = state->picture_aspect_ratio;
+	} else if (property == connector->scaling_mode_property) {
+		*val = state->scaling_mode;
 	} else if (connector->funcs->atomic_get_property) {
 		return connector->funcs->atomic_get_property(connector,
 				state, property, val);
@@ -1618,7 +1691,7 @@ int drm_atomic_commit(struct drm_atomic_state *state)
 	if (ret)
 		return ret;
 
-	DRM_DEBUG_ATOMIC("commiting %p\n", state);
+	DRM_DEBUG_ATOMIC("committing %p\n", state);
 
 	return config->funcs->atomic_commit(state->dev, state, false);
 }
@@ -1647,7 +1720,7 @@ int drm_atomic_nonblocking_commit(struct drm_atomic_state *state)
 	if (ret)
 		return ret;
 
-	DRM_DEBUG_ATOMIC("commiting %p nonblocking\n", state);
+	DRM_DEBUG_ATOMIC("committing %p nonblocking\n", state);
 
 	return config->funcs->atomic_commit(state->dev, state, true);
 }
diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c
index aa885a614e27..bb6dc4e71fe1 100644
--- a/drivers/gpu/drm/drm_atomic_helper.c
+++ b/drivers/gpu/drm/drm_atomic_helper.c
@@ -1076,8 +1076,8 @@ EXPORT_SYMBOL(drm_atomic_helper_commit_modeset_enables);
  *
  * Note that @pre_swap is needed since the point where we block for fences moves
  * around depending upon whether an atomic commit is blocking or
- * non-blocking. For async commit all waiting needs to happen after
- * drm_atomic_helper_swap_state() is called, but for synchronous commits we want
+ * non-blocking. For non-blocking commit all waiting needs to happen after
+ * drm_atomic_helper_swap_state() is called, but for blocking commits we want
  * to wait **before** we do anything that can't be easily rolled back. That is
  * before we call drm_atomic_helper_swap_state().
  *
@@ -2038,6 +2038,8 @@ void drm_atomic_helper_swap_state(struct drm_atomic_state *state,
 	struct drm_plane *plane;
 	struct drm_plane_state *old_plane_state, *new_plane_state;
 	struct drm_crtc_commit *commit;
+	void *obj, *obj_state;
+	const struct drm_private_state_funcs *funcs;
 
 	if (stall) {
 		for_each_new_crtc_in_state(state, crtc, new_crtc_state, i) {
@@ -2098,6 +2100,9 @@ void drm_atomic_helper_swap_state(struct drm_atomic_state *state,
 		state->planes[i].state = old_plane_state;
 		plane->state = new_plane_state;
 	}
+
+	__for_each_private_obj(state, obj, obj_state, i, funcs)
+		funcs->swap_state(obj, &state->private_objs[i].obj_state);
 }
 EXPORT_SYMBOL(drm_atomic_helper_swap_state);
 
@@ -3231,7 +3236,7 @@ void drm_atomic_helper_plane_reset(struct drm_plane *plane)
 
 	if (plane->state) {
 		plane->state->plane = plane;
-		plane->state->rotation = DRM_ROTATE_0;
+		plane->state->rotation = DRM_MODE_ROTATE_0;
 	}
 }
 EXPORT_SYMBOL(drm_atomic_helper_plane_reset);
@@ -3528,7 +3533,8 @@ EXPORT_SYMBOL(drm_atomic_helper_connector_destroy_state);
  *
  * Implements support for legacy gamma correction table for drivers
  * that support color management through the DEGAMMA_LUT/GAMMA_LUT
- * properties.
+ * properties. See drm_crtc_enable_color_mgmt() and the containing chapter for
+ * how the atomic color management and gamma tables work.
  */
 int drm_atomic_helper_legacy_gamma_set(struct drm_crtc *crtc,
 				       u16 *red, u16 *green, u16 *blue,
diff --git a/drivers/gpu/drm/drm_blend.c b/drivers/gpu/drm/drm_blend.c
index a0d0d6843288..db6aeec50b82 100644
--- a/drivers/gpu/drm/drm_blend.c
+++ b/drivers/gpu/drm/drm_blend.c
@@ -119,17 +119,17 @@
  * drm_property_create_bitmask()) called "rotation" and has the following
  * bitmask enumaration values:
  *
- * DRM_ROTATE_0:
+ * DRM_MODE_ROTATE_0:
  * 	"rotate-0"
- * DRM_ROTATE_90:
+ * DRM_MODE_ROTATE_90:
  * 	"rotate-90"
- * DRM_ROTATE_180:
+ * DRM_MODE_ROTATE_180:
  * 	"rotate-180"
- * DRM_ROTATE_270:
+ * DRM_MODE_ROTATE_270:
  * 	"rotate-270"
- * DRM_REFLECT_X:
+ * DRM_MODE_REFLECT_X:
  * 	"reflect-x"
- * DRM_REFELCT_Y:
+ * DRM_MODE_REFLECT_Y:
  * 	"reflect-y"
  *
  * Rotation is the specified amount in degrees in counter clockwise direction,
@@ -142,17 +142,17 @@ int drm_plane_create_rotation_property(struct drm_plane *plane,
 				       unsigned int supported_rotations)
 {
 	static const struct drm_prop_enum_list props[] = {
-		{ __builtin_ffs(DRM_ROTATE_0) - 1,   "rotate-0" },
-		{ __builtin_ffs(DRM_ROTATE_90) - 1,  "rotate-90" },
-		{ __builtin_ffs(DRM_ROTATE_180) - 1, "rotate-180" },
-		{ __builtin_ffs(DRM_ROTATE_270) - 1, "rotate-270" },
-		{ __builtin_ffs(DRM_REFLECT_X) - 1,  "reflect-x" },
-		{ __builtin_ffs(DRM_REFLECT_Y) - 1,  "reflect-y" },
+		{ __builtin_ffs(DRM_MODE_ROTATE_0) - 1,   "rotate-0" },
+		{ __builtin_ffs(DRM_MODE_ROTATE_90) - 1,  "rotate-90" },
+		{ __builtin_ffs(DRM_MODE_ROTATE_180) - 1, "rotate-180" },
+		{ __builtin_ffs(DRM_MODE_ROTATE_270) - 1, "rotate-270" },
+		{ __builtin_ffs(DRM_MODE_REFLECT_X) - 1,  "reflect-x" },
+		{ __builtin_ffs(DRM_MODE_REFLECT_Y) - 1,  "reflect-y" },
 	};
 	struct drm_property *prop;
 
-	WARN_ON((supported_rotations & DRM_ROTATE_MASK) == 0);
-	WARN_ON(!is_power_of_2(rotation & DRM_ROTATE_MASK));
+	WARN_ON((supported_rotations & DRM_MODE_ROTATE_MASK) == 0);
+	WARN_ON(!is_power_of_2(rotation & DRM_MODE_ROTATE_MASK));
 	WARN_ON(rotation & ~supported_rotations);
 
 	prop = drm_property_create_bitmask(plane->dev, 0, "rotation",
@@ -178,14 +178,14 @@ EXPORT_SYMBOL(drm_plane_create_rotation_property);
  * @supported_rotations: Supported rotations
  *
  * Attempt to simplify the rotation to a form that is supported.
- * Eg. if the hardware supports everything except DRM_REFLECT_X
+ * Eg. if the hardware supports everything except DRM_MODE_REFLECT_X
  * one could call this function like this:
  *
- * drm_rotation_simplify(rotation, DRM_ROTATE_0 |
- *                       DRM_ROTATE_90 | DRM_ROTATE_180 |
- *                       DRM_ROTATE_270 | DRM_REFLECT_Y);
+ * drm_rotation_simplify(rotation, DRM_MODE_ROTATE_0 |
+ *                       DRM_MODE_ROTATE_90 | DRM_MODE_ROTATE_180 |
+ *                       DRM_MODE_ROTATE_270 | DRM_MODE_REFLECT_Y);
  *
- * to eliminate the DRM_ROTATE_X flag. Depending on what kind of
+ * to eliminate the DRM_MODE_ROTATE_X flag. Depending on what kind of
  * transforms the hardware supports, this function may not
  * be able to produce a supported transform, so the caller should
  * check the result afterwards.
@@ -194,9 +194,10 @@ unsigned int drm_rotation_simplify(unsigned int rotation,
 				   unsigned int supported_rotations)
 {
 	if (rotation & ~supported_rotations) {
-		rotation ^= DRM_REFLECT_X | DRM_REFLECT_Y;
-		rotation = (rotation & DRM_REFLECT_MASK) |
-		           BIT((ffs(rotation & DRM_ROTATE_MASK) + 1) % 4);
+		rotation ^= DRM_MODE_REFLECT_X | DRM_MODE_REFLECT_Y;
+		rotation = (rotation & DRM_MODE_REFLECT_MASK) |
+		           BIT((ffs(rotation & DRM_MODE_ROTATE_MASK) + 1)
+		           % 4);
 	}
 
 	return rotation;
diff --git a/drivers/gpu/drm/drm_color_mgmt.c b/drivers/gpu/drm/drm_color_mgmt.c
index 533f3a3e6877..3eda500fc005 100644
--- a/drivers/gpu/drm/drm_color_mgmt.c
+++ b/drivers/gpu/drm/drm_color_mgmt.c
@@ -43,7 +43,8 @@
  *
  *	Setting this to NULL (blob property value set to 0) means a
  *	linear/pass-thru gamma table should be used. This is generally the
- *	driver boot-up state too.
+ *	driver boot-up state too. Drivers can access this blob through
+ *	&drm_crtc_state.degamma_lut.
  *
  * “DEGAMMA_LUT_SIZE”:
  *	Unsinged range property to give the size of the lookup table to be set
@@ -60,7 +61,8 @@
  *
  *	Setting this to NULL (blob property value set to 0) means a
  *	unit/pass-thru matrix should be used. This is generally the driver
- *	boot-up state too.
+ *	boot-up state too. Drivers can access the blob for the color conversion
+ *	matrix through &drm_crtc_state.ctm.
  *
  * “GAMMA_LUT”:
  *	Blob property to set the gamma lookup table (LUT) mapping pixel data
@@ -72,7 +74,8 @@
  *
  *	Setting this to NULL (blob property value set to 0) means a
  *	linear/pass-thru gamma table should be used. This is generally the
- *	driver boot-up state too.
+ *	driver boot-up state too. Drivers can access this blob through
+ *	&drm_crtc_state.gamma_lut.
  *
  * “GAMMA_LUT_SIZE”:
  *	Unsigned range property to give the size of the lookup table to be set
diff --git a/drivers/gpu/drm/drm_connector.c b/drivers/gpu/drm/drm_connector.c
index 9f847615ac74..5cd61aff7857 100644
--- a/drivers/gpu/drm/drm_connector.c
+++ b/drivers/gpu/drm/drm_connector.c
@@ -941,6 +941,10 @@ EXPORT_SYMBOL(drm_mode_create_tv_properties);
  *
  * Called by a driver the first time it's needed, must be attached to desired
  * connectors.
+ *
+ * Atomic drivers should use drm_connector_attach_scaling_mode_property()
+ * instead to correctly assign &drm_connector_state.picture_aspect_ratio
+ * in the atomic state.
  */
 int drm_mode_create_scaling_mode_property(struct drm_device *dev)
 {
@@ -961,6 +965,66 @@ int drm_mode_create_scaling_mode_property(struct drm_device *dev)
 EXPORT_SYMBOL(drm_mode_create_scaling_mode_property);
 
 /**
+ * drm_connector_attach_scaling_mode_property - attach atomic scaling mode property
+ * @connector: connector to attach scaling mode property on.
+ * @scaling_mode_mask: or'ed mask of BIT(%DRM_MODE_SCALE_\*).
+ *
+ * This is used to add support for scaling mode to atomic drivers.
+ * The scaling mode will be set to &drm_connector_state.picture_aspect_ratio
+ * and can be used from &drm_connector_helper_funcs->atomic_check for validation.
+ *
+ * This is the atomic version of drm_mode_create_scaling_mode_property().
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
+int drm_connector_attach_scaling_mode_property(struct drm_connector *connector,
+					       u32 scaling_mode_mask)
+{
+	struct drm_device *dev = connector->dev;
+	struct drm_property *scaling_mode_property;
+	int i, j = 0;
+	const unsigned valid_scaling_mode_mask =
+		(1U << ARRAY_SIZE(drm_scaling_mode_enum_list)) - 1;
+
+	if (WARN_ON(hweight32(scaling_mode_mask) < 2 ||
+		    scaling_mode_mask & ~valid_scaling_mode_mask))
+		return -EINVAL;
+
+	scaling_mode_property =
+		drm_property_create(dev, DRM_MODE_PROP_ENUM, "scaling mode",
+				    hweight32(scaling_mode_mask));
+
+	if (!scaling_mode_property)
+		return -ENOMEM;
+
+	for (i = 0; i < ARRAY_SIZE(drm_scaling_mode_enum_list); i++) {
+		int ret;
+
+		if (!(BIT(i) & scaling_mode_mask))
+			continue;
+
+		ret = drm_property_add_enum(scaling_mode_property, j++,
+					    drm_scaling_mode_enum_list[i].type,
+					    drm_scaling_mode_enum_list[i].name);
+
+		if (ret) {
+			drm_property_destroy(dev, scaling_mode_property);
+
+			return ret;
+		}
+	}
+
+	drm_object_attach_property(&connector->base,
+				   scaling_mode_property, 0);
+
+	connector->scaling_mode_property = scaling_mode_property;
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_connector_attach_scaling_mode_property);
+
+/**
  * drm_mode_create_aspect_ratio_property - create aspect ratio property
  * @dev: DRM device
  *
diff --git a/drivers/gpu/drm/drm_dp_mst_topology.c b/drivers/gpu/drm/drm_dp_mst_topology.c
index d3fc7e4e85b7..222eb1a8549b 100644
--- a/drivers/gpu/drm/drm_dp_mst_topology.c
+++ b/drivers/gpu/drm/drm_dp_mst_topology.c
@@ -737,16 +737,16 @@ static void drm_dp_mst_put_payload_id(struct drm_dp_mst_topology_mgr *mgr,
 static bool check_txmsg_state(struct drm_dp_mst_topology_mgr *mgr,
 			      struct drm_dp_sideband_msg_tx *txmsg)
 {
-	bool ret;
+	unsigned int state;
 
 	/*
 	 * All updates to txmsg->state are protected by mgr->qlock, and the two
 	 * cases we check here are terminal states. For those the barriers
 	 * provided by the wake_up/wait_event pair are enough.
 	 */
-	ret = (txmsg->state == DRM_DP_SIDEBAND_TX_RX ||
-	       txmsg->state == DRM_DP_SIDEBAND_TX_TIMEOUT);
-	return ret;
+	state = READ_ONCE(txmsg->state);
+	return (state == DRM_DP_SIDEBAND_TX_RX ||
+		state == DRM_DP_SIDEBAND_TX_TIMEOUT);
 }
 
 static int drm_dp_mst_wait_tx_reply(struct drm_dp_mst_branch *mstb,
@@ -855,7 +855,7 @@ static void drm_dp_destroy_mst_branch_device(struct kref *kref)
 	mutex_unlock(&mstb->mgr->qlock);
 
 	if (wake_tx)
-		wake_up(&mstb->mgr->tx_waitq);
+		wake_up_all(&mstb->mgr->tx_waitq);
 
 	kref_put(kref, drm_dp_free_mst_branch_device);
 }
@@ -1510,7 +1510,7 @@ static void process_single_down_tx_qlock(struct drm_dp_mst_topology_mgr *mgr)
 		if (txmsg->seqno != -1)
 			txmsg->dst->tx_slots[txmsg->seqno] = NULL;
 		txmsg->state = DRM_DP_SIDEBAND_TX_TIMEOUT;
-		wake_up(&mgr->tx_waitq);
+		wake_up_all(&mgr->tx_waitq);
 	}
 }
 
@@ -2258,7 +2258,7 @@ static int drm_dp_mst_handle_down_rep(struct drm_dp_mst_topology_mgr *mgr)
 		mstb->tx_slots[slot] = NULL;
 		mutex_unlock(&mgr->qlock);
 
-		wake_up(&mgr->tx_waitq);
+		wake_up_all(&mgr->tx_waitq);
 	}
 	return ret;
 }
@@ -2498,6 +2498,81 @@ static int drm_dp_init_vcpi(struct drm_dp_mst_topology_mgr *mgr,
 }
 
 /**
+ * drm_dp_atomic_find_vcpi_slots() - Find and add vcpi slots to the state
+ * @state: global atomic state
+ * @mgr: MST topology manager for the port
+ * @port: port to find vcpi slots for
+ * @pbn: bandwidth required for the mode in PBN
+ *
+ * RETURNS:
+ * Total slots in the atomic state assigned for this port or error
+ */
+int drm_dp_atomic_find_vcpi_slots(struct drm_atomic_state *state,
+				  struct drm_dp_mst_topology_mgr *mgr,
+				  struct drm_dp_mst_port *port, int pbn)
+{
+	struct drm_dp_mst_topology_state *topology_state;
+	int req_slots;
+
+	topology_state = drm_atomic_get_mst_topology_state(state, mgr);
+	if (topology_state == NULL)
+		return -ENOMEM;
+
+	port = drm_dp_get_validated_port_ref(mgr, port);
+	if (port == NULL)
+		return -EINVAL;
+	req_slots = DIV_ROUND_UP(pbn, mgr->pbn_div);
+	DRM_DEBUG_KMS("vcpi slots req=%d, avail=%d\n",
+			req_slots, topology_state->avail_slots);
+
+	if (req_slots > topology_state->avail_slots) {
+		drm_dp_put_port(port);
+		return -ENOSPC;
+	}
+
+	topology_state->avail_slots -= req_slots;
+	DRM_DEBUG_KMS("vcpi slots avail=%d", topology_state->avail_slots);
+
+	drm_dp_put_port(port);
+	return req_slots;
+}
+EXPORT_SYMBOL(drm_dp_atomic_find_vcpi_slots);
+
+/**
+ * drm_dp_atomic_release_vcpi_slots() - Release allocated vcpi slots
+ * @state: global atomic state
+ * @mgr: MST topology manager for the port
+ * @slots: number of vcpi slots to release
+ *
+ * RETURNS:
+ * 0 if @slots were added back to &drm_dp_mst_topology_state->avail_slots or
+ * negative error code
+ */
+int drm_dp_atomic_release_vcpi_slots(struct drm_atomic_state *state,
+				     struct drm_dp_mst_topology_mgr *mgr,
+				     int slots)
+{
+	struct drm_dp_mst_topology_state *topology_state;
+
+	topology_state = drm_atomic_get_mst_topology_state(state, mgr);
+	if (topology_state == NULL)
+		return -ENOMEM;
+
+	/* We cannot rely on port->vcpi.num_slots to update
+	 * topology_state->avail_slots as the port may not exist if the parent
+	 * branch device was unplugged. This should be fixed by tracking
+	 * per-port slot allocation in drm_dp_mst_topology_state instead of
+	 * depending on the caller to tell us how many slots to release.
+	 */
+	topology_state->avail_slots += slots;
+	DRM_DEBUG_KMS("vcpi slots released=%d, avail=%d\n",
+			slots, topology_state->avail_slots);
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_dp_atomic_release_vcpi_slots);
+
+/**
  * drm_dp_mst_allocate_vcpi() - Allocate a virtual channel
  * @mgr: manager for this port
  * @port: port to allocate a virtual channel for.
@@ -2936,6 +3011,69 @@ static void drm_dp_destroy_connector_work(struct work_struct *work)
 		(*mgr->cbs->hotplug)(mgr);
 }
 
+void *drm_dp_mst_duplicate_state(struct drm_atomic_state *state, void *obj)
+{
+	struct drm_dp_mst_topology_mgr *mgr = obj;
+	struct drm_dp_mst_topology_state *new_mst_state;
+
+	if (WARN_ON(!mgr->state))
+		return NULL;
+
+	new_mst_state = kmemdup(mgr->state, sizeof(*new_mst_state), GFP_KERNEL);
+	if (new_mst_state)
+		new_mst_state->state = state;
+	return new_mst_state;
+}
+
+void drm_dp_mst_swap_state(void *obj, void **obj_state_ptr)
+{
+	struct drm_dp_mst_topology_mgr *mgr = obj;
+	struct drm_dp_mst_topology_state **topology_state_ptr;
+
+	topology_state_ptr = (struct drm_dp_mst_topology_state **)obj_state_ptr;
+
+	mgr->state->state = (*topology_state_ptr)->state;
+	swap(*topology_state_ptr, mgr->state);
+	mgr->state->state = NULL;
+}
+
+void drm_dp_mst_destroy_state(void *obj_state)
+{
+	kfree(obj_state);
+}
+
+static const struct drm_private_state_funcs mst_state_funcs = {
+	.duplicate_state = drm_dp_mst_duplicate_state,
+	.swap_state = drm_dp_mst_swap_state,
+	.destroy_state = drm_dp_mst_destroy_state,
+};
+
+/**
+ * drm_atomic_get_mst_topology_state: get MST topology state
+ *
+ * @state: global atomic state
+ * @mgr: MST topology manager, also the private object in this case
+ *
+ * This function wraps drm_atomic_get_priv_obj_state() passing in the MST atomic
+ * state vtable so that the private object state returned is that of a MST
+ * topology object. Also, drm_atomic_get_private_obj_state() expects the caller
+ * to care of the locking, so warn if don't hold the connection_mutex.
+ *
+ * RETURNS:
+ *
+ * The MST topology state or error pointer.
+ */
+struct drm_dp_mst_topology_state *drm_atomic_get_mst_topology_state(struct drm_atomic_state *state,
+								    struct drm_dp_mst_topology_mgr *mgr)
+{
+	struct drm_device *dev = mgr->dev;
+
+	WARN_ON(!drm_modeset_is_locked(&dev->mode_config.connection_mutex));
+	return drm_atomic_get_private_obj_state(state, mgr,
+						&mst_state_funcs);
+}
+EXPORT_SYMBOL(drm_atomic_get_mst_topology_state);
+
 /**
  * drm_dp_mst_topology_mgr_init - initialise a topology manager
  * @mgr: manager struct to initialise
@@ -2980,6 +3118,15 @@ int drm_dp_mst_topology_mgr_init(struct drm_dp_mst_topology_mgr *mgr,
 	if (test_calc_pbn_mode() < 0)
 		DRM_ERROR("MST PBN self-test failed\n");
 
+	mgr->state = kzalloc(sizeof(*mgr->state), GFP_KERNEL);
+	if (mgr->state == NULL)
+		return -ENOMEM;
+	mgr->state->mgr = mgr;
+
+	/* max. time slots - one slot for MTP header */
+	mgr->state->avail_slots = 63;
+	mgr->funcs = &mst_state_funcs;
+
 	return 0;
 }
 EXPORT_SYMBOL(drm_dp_mst_topology_mgr_init);
@@ -3000,6 +3147,9 @@ void drm_dp_mst_topology_mgr_destroy(struct drm_dp_mst_topology_mgr *mgr)
 	mutex_unlock(&mgr->payload_lock);
 	mgr->dev = NULL;
 	mgr->aux = NULL;
+	kfree(mgr->state);
+	mgr->state = NULL;
+	mgr->funcs = NULL;
 }
 EXPORT_SYMBOL(drm_dp_mst_topology_mgr_destroy);
 
diff --git a/drivers/gpu/drm/drm_fb_cma_helper.c b/drivers/gpu/drm/drm_fb_cma_helper.c
index 50abd1faf38f..53f9bdf470d7 100644
--- a/drivers/gpu/drm/drm_fb_cma_helper.c
+++ b/drivers/gpu/drm/drm_fb_cma_helper.c
@@ -189,7 +189,7 @@ struct drm_framebuffer *drm_fb_cma_create_with_funcs(struct drm_device *dev,
 		obj = drm_gem_object_lookup(file_priv, mode_cmd->handles[i]);
 		if (!obj) {
 			dev_err(dev->dev, "Failed to lookup GEM object\n");
-			ret = -ENXIO;
+			ret = -ENOENT;
 			goto err_gem_object_put;
 		}
 
@@ -260,6 +260,33 @@ struct drm_gem_cma_object *drm_fb_cma_get_gem_obj(struct drm_framebuffer *fb,
 EXPORT_SYMBOL_GPL(drm_fb_cma_get_gem_obj);
 
 /**
+ * drm_fb_cma_get_gem_addr() - Get physical address for framebuffer
+ * @fb: The framebuffer
+ * @state: Which state of drm plane
+ * @plane: Which plane
+ * Return the CMA GEM address for given framebuffer.
+ *
+ * This function will usually be called from the PLANE callback functions.
+ */
+dma_addr_t drm_fb_cma_get_gem_addr(struct drm_framebuffer *fb,
+				   struct drm_plane_state *state,
+				   unsigned int plane)
+{
+	struct drm_fb_cma *fb_cma = to_fb_cma(fb);
+	dma_addr_t paddr;
+
+	if (plane >= 4)
+		return 0;
+
+	paddr = fb_cma->obj[plane]->paddr + fb->offsets[plane];
+	paddr += fb->format->cpp[plane] * (state->src_x >> 16);
+	paddr += fb->pitches[plane] * (state->src_y >> 16);
+
+	return paddr;
+}
+EXPORT_SYMBOL_GPL(drm_fb_cma_get_gem_addr);
+
+/**
  * drm_fb_cma_prepare_fb() - Prepare CMA framebuffer
  * @plane: Which plane
  * @state: Plane state attach fence to
diff --git a/drivers/gpu/drm/drm_fb_helper.c b/drivers/gpu/drm/drm_fb_helper.c
index 1f178b878e42..574af01d3ce9 100644
--- a/drivers/gpu/drm/drm_fb_helper.c
+++ b/drivers/gpu/drm/drm_fb_helper.c
@@ -378,7 +378,7 @@ retry:
 			goto fail;
 		}
 
-		plane_state->rotation = DRM_ROTATE_0;
+		plane_state->rotation = DRM_MODE_ROTATE_0;
 
 		plane->old_fb = plane->fb;
 		plane_mask |= 1 << drm_plane_index(plane);
@@ -431,7 +431,7 @@ static int restore_fbdev_mode_legacy(struct drm_fb_helper *fb_helper)
 		if (plane->rotation_property)
 			drm_mode_plane_set_obj_prop(plane,
 						    plane->rotation_property,
-						    DRM_ROTATE_0);
+						    DRM_MODE_ROTATE_0);
 	}
 
 	for (i = 0; i < fb_helper->crtc_count; i++) {
diff --git a/drivers/gpu/drm/drm_file.c b/drivers/gpu/drm/drm_file.c
index 3783b659cd38..caad93dab54b 100644
--- a/drivers/gpu/drm/drm_file.c
+++ b/drivers/gpu/drm/drm_file.c
@@ -351,9 +351,8 @@ void drm_lastclose(struct drm_device * dev)
  *
  * This function must be used by drivers as their &file_operations.release
  * method. It frees any resources associated with the open file, and calls the
- * &drm_driver.preclose and &drm_driver.lastclose driver callbacks. If this is
- * the last open file for the DRM device also proceeds to call the
- * &drm_driver.lastclose driver callback.
+ * &drm_driver.postclose driver callback. If this is the last open file for the
+ * DRM device also proceeds to call the &drm_driver.lastclose driver callback.
  *
  * RETURNS:
  *
@@ -373,7 +372,8 @@ int drm_release(struct inode *inode, struct file *filp)
 	list_del(&file_priv->lhead);
 	mutex_unlock(&dev->filelist_mutex);
 
-	if (dev->driver->preclose)
+	if (drm_core_check_feature(dev, DRIVER_LEGACY) &&
+	    dev->driver->preclose)
 		dev->driver->preclose(dev, file_priv);
 
 	/* ========================================================
diff --git a/drivers/gpu/drm/drm_gem.c b/drivers/gpu/drm/drm_gem.c
index b1e28c944637..8dc11064253d 100644
--- a/drivers/gpu/drm/drm_gem.c
+++ b/drivers/gpu/drm/drm_gem.c
@@ -521,7 +521,7 @@ struct page **drm_gem_get_pages(struct drm_gem_object *obj)
 
 	npages = obj->size >> PAGE_SHIFT;
 
-	pages = drm_malloc_ab(npages, sizeof(struct page *));
+	pages = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL);
 	if (pages == NULL)
 		return ERR_PTR(-ENOMEM);
 
@@ -546,7 +546,7 @@ fail:
 	while (i--)
 		put_page(pages[i]);
 
-	drm_free_large(pages);
+	kvfree(pages);
 	return ERR_CAST(p);
 }
 EXPORT_SYMBOL(drm_gem_get_pages);
@@ -582,7 +582,7 @@ void drm_gem_put_pages(struct drm_gem_object *obj, struct page **pages,
 		put_page(pages[i]);
 	}
 
-	drm_free_large(pages);
+	kvfree(pages);
 }
 EXPORT_SYMBOL(drm_gem_put_pages);
 
diff --git a/drivers/gpu/drm/drm_irq.c b/drivers/gpu/drm/drm_irq.c
index 8c866cac62dd..c7debaad67f8 100644
--- a/drivers/gpu/drm/drm_irq.c
+++ b/drivers/gpu/drm/drm_irq.c
@@ -54,7 +54,7 @@
 
 static bool
 drm_get_last_vbltimestamp(struct drm_device *dev, unsigned int pipe,
-			  struct timeval *tvblank, unsigned flags);
+			  struct timeval *tvblank, bool in_vblank_irq);
 
 static unsigned int drm_timestamp_precision = 20;  /* Default to 20 usecs. */
 
@@ -138,7 +138,7 @@ static void drm_reset_vblank_timestamp(struct drm_device *dev, unsigned int pipe
 	 */
 	do {
 		cur_vblank = __get_vblank_counter(dev, pipe);
-		rc = drm_get_last_vbltimestamp(dev, pipe, &t_vblank, 0);
+		rc = drm_get_last_vbltimestamp(dev, pipe, &t_vblank, false);
 	} while (cur_vblank != __get_vblank_counter(dev, pipe) && --count > 0);
 
 	/*
@@ -171,7 +171,7 @@ static void drm_reset_vblank_timestamp(struct drm_device *dev, unsigned int pipe
  * device vblank fields.
  */
 static void drm_update_vblank_count(struct drm_device *dev, unsigned int pipe,
-				    unsigned long flags)
+				    bool in_vblank_irq)
 {
 	struct drm_vblank_crtc *vblank = &dev->vblank[pipe];
 	u32 cur_vblank, diff;
@@ -194,7 +194,7 @@ static void drm_update_vblank_count(struct drm_device *dev, unsigned int pipe,
 	 */
 	do {
 		cur_vblank = __get_vblank_counter(dev, pipe);
-		rc = drm_get_last_vbltimestamp(dev, pipe, &t_vblank, flags);
+		rc = drm_get_last_vbltimestamp(dev, pipe, &t_vblank, in_vblank_irq);
 	} while (cur_vblank != __get_vblank_counter(dev, pipe) && --count > 0);
 
 	if (dev->max_vblank_count != 0) {
@@ -214,13 +214,13 @@ static void drm_update_vblank_count(struct drm_device *dev, unsigned int pipe,
 		 */
 		diff = DIV_ROUND_CLOSEST_ULL(diff_ns, framedur_ns);
 
-		if (diff == 0 && flags & DRM_CALLED_FROM_VBLIRQ)
+		if (diff == 0 && in_vblank_irq)
 			DRM_DEBUG_VBL("crtc %u: Redundant vblirq ignored."
 				      " diff_ns = %lld, framedur_ns = %d)\n",
 				      pipe, (long long) diff_ns, framedur_ns);
 	} else {
 		/* some kind of default for drivers w/o accurate vbl timestamping */
-		diff = (flags & DRM_CALLED_FROM_VBLIRQ) != 0;
+		diff = in_vblank_irq ? 1 : 0;
 	}
 
 	/*
@@ -253,7 +253,7 @@ static void drm_update_vblank_count(struct drm_device *dev, unsigned int pipe,
 	 * Otherwise reinitialize delayed at next vblank interrupt and assign 0
 	 * for now, to mark the vblanktimestamp as invalid.
 	 */
-	if (!rc && (flags & DRM_CALLED_FROM_VBLIRQ) == 0)
+	if (!rc && in_vblank_irq)
 		t_vblank = (struct timeval) {0, 0};
 
 	store_vblank(dev, pipe, diff, &t_vblank, cur_vblank);
@@ -291,7 +291,7 @@ u32 drm_accurate_vblank_count(struct drm_crtc *crtc)
 
 	spin_lock_irqsave(&dev->vblank_time_lock, flags);
 
-	drm_update_vblank_count(dev, pipe, 0);
+	drm_update_vblank_count(dev, pipe, false);
 	vblank = drm_vblank_count(dev, pipe);
 
 	spin_unlock_irqrestore(&dev->vblank_time_lock, flags);
@@ -349,7 +349,7 @@ static void vblank_disable_and_save(struct drm_device *dev, unsigned int pipe)
 	 * this time. This makes the count account for the entire time
 	 * between drm_crtc_vblank_on() and drm_crtc_vblank_off().
 	 */
-	drm_update_vblank_count(dev, pipe, 0);
+	drm_update_vblank_count(dev, pipe, false);
 
 	spin_unlock_irqrestore(&dev->vblank_time_lock, irqflags);
 }
@@ -684,6 +684,7 @@ void drm_calc_timestamping_constants(struct drm_crtc *crtc,
 
 	vblank->linedur_ns  = linedur_ns;
 	vblank->framedur_ns = framedur_ns;
+	vblank->hwmode = *mode;
 
 	DRM_DEBUG("crtc %u: hwmode: htotal %d, vtotal %d, vdisplay %d\n",
 		  crtc->base.id, mode->crtc_htotal,
@@ -700,10 +701,10 @@ EXPORT_SYMBOL(drm_calc_timestamping_constants);
  * @max_error: Desired maximum allowable error in timestamps (nanosecs)
  *             On return contains true maximum error of timestamp
  * @vblank_time: Pointer to struct timeval which should receive the timestamp
- * @flags: Flags to pass to driver:
- *         0 = Default,
- *         DRM_CALLED_FROM_VBLIRQ = If function is called from vbl IRQ handler
- * @mode: mode which defines the scanout timings
+ * @in_vblank_irq:
+ *     True when called from drm_crtc_handle_vblank().  Some drivers
+ *     need to apply some workarounds for gpu-specific vblank irq quirks
+ *     if flag is set.
  *
  * Implements calculation of exact vblank timestamps from given drm_display_mode
  * timings and current video scanout position of a CRTC. This can be called from
@@ -723,52 +724,62 @@ EXPORT_SYMBOL(drm_calc_timestamping_constants);
  * returns as no operation if a doublescan or interlaced video mode is
  * active. Higher level code is expected to handle this.
  *
- * Returns:
- * Negative value on error, failure or if not supported in current
- * video mode:
- *
- * -EINVAL    Invalid CRTC.
- * -EAGAIN    Temporary unavailable, e.g., called before initial modeset.
- * -ENOTSUPP  Function not supported in current display mode.
- * -EIO       Failed, e.g., due to failed scanout position query.
+ * This function can be used to implement the &drm_driver.get_vblank_timestamp
+ * directly, if the driver implements the &drm_driver.get_scanout_position hook.
  *
- * Returns or'ed positive status flags on success:
+ * Note that atomic drivers must call drm_calc_timestamping_constants() before
+ * enabling a CRTC. The atomic helpers already take care of that in
+ * drm_atomic_helper_update_legacy_modeset_state().
  *
- * DRM_VBLANKTIME_SCANOUTPOS_METHOD - Signal this method used for timestamping.
- * DRM_VBLANKTIME_INVBL - Timestamp taken while scanout was in vblank interval.
+ * Returns:
  *
+ * Returns true on success, and false on failure, i.e. when no accurate
+ * timestamp could be acquired.
  */
-int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev,
-					  unsigned int pipe,
-					  int *max_error,
-					  struct timeval *vblank_time,
-					  unsigned flags,
-					  const struct drm_display_mode *mode)
+bool drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev,
+					   unsigned int pipe,
+					   int *max_error,
+					   struct timeval *vblank_time,
+					   bool in_vblank_irq)
 {
 	struct timeval tv_etime;
 	ktime_t stime, etime;
-	unsigned int vbl_status;
-	int ret = DRM_VBLANKTIME_SCANOUTPOS_METHOD;
+	bool vbl_status;
+	struct drm_crtc *crtc;
+	const struct drm_display_mode *mode;
+	struct drm_vblank_crtc *vblank = &dev->vblank[pipe];
 	int vpos, hpos, i;
 	int delta_ns, duration_ns;
 
-	if (pipe >= dev->num_crtcs) {
+	if (!drm_core_check_feature(dev, DRIVER_MODESET))
+		return false;
+
+	crtc = drm_crtc_from_index(dev, pipe);
+
+	if (pipe >= dev->num_crtcs || !crtc) {
 		DRM_ERROR("Invalid crtc %u\n", pipe);
-		return -EINVAL;
+		return false;
 	}
 
 	/* Scanout position query not supported? Should not happen. */
 	if (!dev->driver->get_scanout_position) {
 		DRM_ERROR("Called from driver w/o get_scanout_position()!?\n");
-		return -EIO;
+		return false;
 	}
 
+	if (drm_drv_uses_atomic_modeset(dev))
+		mode = &vblank->hwmode;
+	else
+		mode = &crtc->hwmode;
+
 	/* If mode timing undefined, just return as no-op:
 	 * Happens during initial modesetting of a crtc.
 	 */
 	if (mode->crtc_clock == 0) {
 		DRM_DEBUG("crtc %u: Noop due to uninitialized mode.\n", pipe);
-		return -EAGAIN;
+		WARN_ON_ONCE(drm_drv_uses_atomic_modeset(dev));
+
+		return false;
 	}
 
 	/* Get current scanout position with system timestamp.
@@ -783,16 +794,17 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev,
 		 * Get vertical and horizontal scanout position vpos, hpos,
 		 * and bounding timestamps stime, etime, pre/post query.
 		 */
-		vbl_status = dev->driver->get_scanout_position(dev, pipe, flags,
+		vbl_status = dev->driver->get_scanout_position(dev, pipe,
+							       in_vblank_irq,
 							       &vpos, &hpos,
 							       &stime, &etime,
 							       mode);
 
 		/* Return as no-op if scanout query unsupported or failed. */
-		if (!(vbl_status & DRM_SCANOUTPOS_VALID)) {
-			DRM_DEBUG("crtc %u : scanoutpos query failed [0x%x].\n",
-				  pipe, vbl_status);
-			return -EIO;
+		if (!vbl_status) {
+			DRM_DEBUG("crtc %u : scanoutpos query failed.\n",
+				  pipe);
+			return false;
 		}
 
 		/* Compute uncertainty in timestamp of scanout position query. */
@@ -830,13 +842,13 @@ int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev,
 	etime = ktime_sub_ns(etime, delta_ns);
 	*vblank_time = ktime_to_timeval(etime);
 
-	DRM_DEBUG_VBL("crtc %u : v 0x%x p(%d,%d)@ %ld.%ld -> %ld.%ld [e %d us, %d rep]\n",
-		      pipe, vbl_status, hpos, vpos,
+	DRM_DEBUG_VBL("crtc %u : v p(%d,%d)@ %ld.%ld -> %ld.%ld [e %d us, %d rep]\n",
+		      pipe, hpos, vpos,
 		      (long)tv_etime.tv_sec, (long)tv_etime.tv_usec,
 		      (long)vblank_time->tv_sec, (long)vblank_time->tv_usec,
 		      duration_ns/1000, i);
 
-	return ret;
+	return true;
 }
 EXPORT_SYMBOL(drm_calc_vbltimestamp_from_scanoutpos);
 
@@ -854,9 +866,10 @@ static struct timeval get_drm_timestamp(void)
  * @dev: DRM device
  * @pipe: index of CRTC whose vblank timestamp to retrieve
  * @tvblank: Pointer to target struct timeval which should receive the timestamp
- * @flags: Flags to pass to driver:
- *         0 = Default,
- *         DRM_CALLED_FROM_VBLIRQ = If function is called from vbl IRQ handler
+ * @in_vblank_irq:
+ *     True when called from drm_crtc_handle_vblank().  Some drivers
+ *     need to apply some workarounds for gpu-specific vblank irq quirks
+ *     if flag is set.
  *
  * Fetches the system timestamp corresponding to the time of the most recent
  * vblank interval on specified CRTC. May call into kms-driver to
@@ -870,27 +883,25 @@ static struct timeval get_drm_timestamp(void)
  */
 static bool
 drm_get_last_vbltimestamp(struct drm_device *dev, unsigned int pipe,
-			  struct timeval *tvblank, unsigned flags)
+			  struct timeval *tvblank, bool in_vblank_irq)
 {
-	int ret;
+	bool ret = false;
 
 	/* Define requested maximum error on timestamps (nanoseconds). */
 	int max_error = (int) drm_timestamp_precision * 1000;
 
 	/* Query driver if possible and precision timestamping enabled. */
-	if (dev->driver->get_vblank_timestamp && (max_error > 0)) {
+	if (dev->driver->get_vblank_timestamp && (max_error > 0))
 		ret = dev->driver->get_vblank_timestamp(dev, pipe, &max_error,
-							tvblank, flags);
-		if (ret > 0)
-			return true;
-	}
+							tvblank, in_vblank_irq);
 
 	/* GPU high precision timestamp query unsupported or failed.
 	 * Return current monotonic/gettimeofday timestamp as best estimate.
 	 */
-	*tvblank = get_drm_timestamp();
+	if (!ret)
+		*tvblank = get_drm_timestamp();
 
-	return false;
+	return ret;
 }
 
 /**
@@ -1329,6 +1340,10 @@ void drm_crtc_vblank_off(struct drm_crtc *crtc)
 		send_vblank_event(dev, e, seq, &now);
 	}
 	spin_unlock_irqrestore(&dev->event_lock, irqflags);
+
+	/* Will be reset by the modeset helpers when re-enabling the crtc by
+	 * calling drm_calc_timestamping_constants(). */
+	vblank->hwmode.crtc_clock = 0;
 }
 EXPORT_SYMBOL(drm_crtc_vblank_off);
 
@@ -1760,7 +1775,7 @@ bool drm_handle_vblank(struct drm_device *dev, unsigned int pipe)
 		return false;
 	}
 
-	drm_update_vblank_count(dev, pipe, DRM_CALLED_FROM_VBLIRQ);
+	drm_update_vblank_count(dev, pipe, true);
 
 	spin_unlock(&dev->vblank_time_lock);
 
diff --git a/drivers/gpu/drm/drm_plane_helper.c b/drivers/gpu/drm/drm_plane_helper.c
index b84a295230fc..06aee1741e96 100644
--- a/drivers/gpu/drm/drm_plane_helper.c
+++ b/drivers/gpu/drm/drm_plane_helper.c
@@ -336,7 +336,7 @@ int drm_primary_helper_update(struct drm_plane *plane, struct drm_crtc *crtc,
 
 	ret = drm_plane_helper_check_update(plane, crtc, fb,
 					    &src, &dest, &clip,
-					    DRM_ROTATE_0,
+					    DRM_MODE_ROTATE_0,
 					    DRM_PLANE_HELPER_NO_SCALING,
 					    DRM_PLANE_HELPER_NO_SCALING,
 					    false, false, &visible);
@@ -381,6 +381,7 @@ EXPORT_SYMBOL(drm_primary_helper_update);
 /**
  * drm_primary_helper_disable() - Helper for primary plane disable
  * @plane: plane to disable
+ * @ctx: lock acquire context, not used here
  *
  * Provides a default plane disable handler for primary planes.  This is handler
  * is called in response to a userspace SetPlane operation on the plane with a
@@ -510,12 +511,10 @@ int drm_plane_helper_commit(struct drm_plane *plane,
 	if (plane_funcs->cleanup_fb)
 		plane_funcs->cleanup_fb(plane, plane_state);
 out:
-	if (plane_state) {
-		if (plane->funcs->atomic_destroy_state)
-			plane->funcs->atomic_destroy_state(plane, plane_state);
-		else
-			drm_atomic_helper_plane_destroy_state(plane, plane_state);
-	}
+	if (plane->funcs->atomic_destroy_state)
+		plane->funcs->atomic_destroy_state(plane, plane_state);
+	else
+		drm_atomic_helper_plane_destroy_state(plane, plane_state);
 
 	return ret;
 }
diff --git a/drivers/gpu/drm/drm_prime.c b/drivers/gpu/drm/drm_prime.c
index 954eb848b5e2..22408badc617 100644
--- a/drivers/gpu/drm/drm_prime.c
+++ b/drivers/gpu/drm/drm_prime.c
@@ -595,15 +595,18 @@ out_unlock:
 EXPORT_SYMBOL(drm_gem_prime_handle_to_fd);
 
 /**
- * drm_gem_prime_import - helper library implementation of the import callback
+ * drm_gem_prime_import_dev - core implementation of the import callback
  * @dev: drm_device to import into
  * @dma_buf: dma-buf object to import
+ * @attach_dev: struct device to dma_buf attach
  *
- * This is the implementation of the gem_prime_import functions for GEM drivers
- * using the PRIME helpers.
+ * This is the core of drm_gem_prime_import. It's designed to be called by
+ * drivers who want to use a different device structure than dev->dev for
+ * attaching via dma_buf.
  */
-struct drm_gem_object *drm_gem_prime_import(struct drm_device *dev,
-					    struct dma_buf *dma_buf)
+struct drm_gem_object *drm_gem_prime_import_dev(struct drm_device *dev,
+					    struct dma_buf *dma_buf,
+					    struct device *attach_dev)
 {
 	struct dma_buf_attachment *attach;
 	struct sg_table *sgt;
@@ -625,7 +628,7 @@ struct drm_gem_object *drm_gem_prime_import(struct drm_device *dev,
 	if (!dev->driver->gem_prime_import_sg_table)
 		return ERR_PTR(-EINVAL);
 
-	attach = dma_buf_attach(dma_buf, dev->dev);
+	attach = dma_buf_attach(dma_buf, attach_dev);
 	if (IS_ERR(attach))
 		return ERR_CAST(attach);
 
@@ -655,6 +658,21 @@ fail_detach:
 
 	return ERR_PTR(ret);
 }
+EXPORT_SYMBOL(drm_gem_prime_import_dev);
+
+/**
+ * drm_gem_prime_import - helper library implementation of the import callback
+ * @dev: drm_device to import into
+ * @dma_buf: dma-buf object to import
+ *
+ * This is the implementation of the gem_prime_import functions for GEM drivers
+ * using the PRIME helpers.
+ */
+struct drm_gem_object *drm_gem_prime_import(struct drm_device *dev,
+					    struct dma_buf *dma_buf)
+{
+	return drm_gem_prime_import_dev(dev, dma_buf, dev->dev);
+}
 EXPORT_SYMBOL(drm_gem_prime_import);
 
 /**
diff --git a/drivers/gpu/drm/drm_rect.c b/drivers/gpu/drm/drm_rect.c
index bc5575960ebc..9817c1445ba9 100644
--- a/drivers/gpu/drm/drm_rect.c
+++ b/drivers/gpu/drm/drm_rect.c
@@ -310,38 +310,38 @@ void drm_rect_rotate(struct drm_rect *r,
 {
 	struct drm_rect tmp;
 
-	if (rotation & (DRM_REFLECT_X | DRM_REFLECT_Y)) {
+	if (rotation & (DRM_MODE_REFLECT_X | DRM_MODE_REFLECT_Y)) {
 		tmp = *r;
 
-		if (rotation & DRM_REFLECT_X) {
+		if (rotation & DRM_MODE_REFLECT_X) {
 			r->x1 = width - tmp.x2;
 			r->x2 = width - tmp.x1;
 		}
 
-		if (rotation & DRM_REFLECT_Y) {
+		if (rotation & DRM_MODE_REFLECT_Y) {
 			r->y1 = height - tmp.y2;
 			r->y2 = height - tmp.y1;
 		}
 	}
 
-	switch (rotation & DRM_ROTATE_MASK) {
-	case DRM_ROTATE_0:
+	switch (rotation & DRM_MODE_ROTATE_MASK) {
+	case DRM_MODE_ROTATE_0:
 		break;
-	case DRM_ROTATE_90:
+	case DRM_MODE_ROTATE_90:
 		tmp = *r;
 		r->x1 = tmp.y1;
 		r->x2 = tmp.y2;
 		r->y1 = width - tmp.x2;
 		r->y2 = width - tmp.x1;
 		break;
-	case DRM_ROTATE_180:
+	case DRM_MODE_ROTATE_180:
 		tmp = *r;
 		r->x1 = width - tmp.x2;
 		r->x2 = width - tmp.x1;
 		r->y1 = height - tmp.y2;
 		r->y2 = height - tmp.y1;
 		break;
-	case DRM_ROTATE_270:
+	case DRM_MODE_ROTATE_270:
 		tmp = *r;
 		r->x1 = height - tmp.y2;
 		r->x2 = height - tmp.y1;
@@ -373,8 +373,8 @@ EXPORT_SYMBOL(drm_rect_rotate);
  * them when doing a rotatation and its inverse.
  * That is, if you do ::
  *
- *     drm_rotate(&r, width, height, rotation);
- *     drm_rotate_inv(&r, width, height, rotation);
+ *     DRM_MODE_PROP_ROTATE(&r, width, height, rotation);
+ *     DRM_MODE_ROTATE_inv(&r, width, height, rotation);
  *
  * you will always get back the original rectangle.
  */
@@ -384,24 +384,24 @@ void drm_rect_rotate_inv(struct drm_rect *r,
 {
 	struct drm_rect tmp;
 
-	switch (rotation & DRM_ROTATE_MASK) {
-	case DRM_ROTATE_0:
+	switch (rotation & DRM_MODE_ROTATE_MASK) {
+	case DRM_MODE_ROTATE_0:
 		break;
-	case DRM_ROTATE_90:
+	case DRM_MODE_ROTATE_90:
 		tmp = *r;
 		r->x1 = width - tmp.y2;
 		r->x2 = width - tmp.y1;
 		r->y1 = tmp.x1;
 		r->y2 = tmp.x2;
 		break;
-	case DRM_ROTATE_180:
+	case DRM_MODE_ROTATE_180:
 		tmp = *r;
 		r->x1 = width - tmp.x2;
 		r->x2 = width - tmp.x1;
 		r->y1 = height - tmp.y2;
 		r->y2 = height - tmp.y1;
 		break;
-	case DRM_ROTATE_270:
+	case DRM_MODE_ROTATE_270:
 		tmp = *r;
 		r->x1 = tmp.y1;
 		r->x2 = tmp.y2;
@@ -412,15 +412,15 @@ void drm_rect_rotate_inv(struct drm_rect *r,
 		break;
 	}
 
-	if (rotation & (DRM_REFLECT_X | DRM_REFLECT_Y)) {
+	if (rotation & (DRM_MODE_REFLECT_X | DRM_MODE_REFLECT_Y)) {
 		tmp = *r;
 
-		if (rotation & DRM_REFLECT_X) {
+		if (rotation & DRM_MODE_REFLECT_X) {
 			r->x1 = width - tmp.x2;
 			r->x2 = width - tmp.x1;
 		}
 
-		if (rotation & DRM_REFLECT_Y) {
+		if (rotation & DRM_MODE_REFLECT_Y) {
 			r->y1 = height - tmp.y2;
 			r->y2 = height - tmp.y1;
 		}
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem.c b/drivers/gpu/drm/etnaviv/etnaviv_gem.c
index fd56f92f3469..d6fb724fc3cc 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_gem.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_gem.c
@@ -748,7 +748,7 @@ static struct page **etnaviv_gem_userptr_do_get_pages(
 	uintptr_t ptr;
 	unsigned int flags = 0;
 
-	pvec = drm_malloc_ab(npages, sizeof(struct page *));
+	pvec = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL);
 	if (!pvec)
 		return ERR_PTR(-ENOMEM);
 
@@ -772,7 +772,7 @@ static struct page **etnaviv_gem_userptr_do_get_pages(
 
 	if (ret < 0) {
 		release_pages(pvec, pinned, 0);
-		drm_free_large(pvec);
+		kvfree(pvec);
 		return ERR_PTR(ret);
 	}
 
@@ -823,7 +823,7 @@ static int etnaviv_gem_userptr_get_pages(struct etnaviv_gem_object *etnaviv_obj)
 	mm = get_task_mm(etnaviv_obj->userptr.task);
 	pinned = 0;
 	if (mm == current->mm) {
-		pvec = drm_malloc_ab(npages, sizeof(struct page *));
+		pvec = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL);
 		if (!pvec) {
 			mmput(mm);
 			return -ENOMEM;
@@ -832,7 +832,7 @@ static int etnaviv_gem_userptr_get_pages(struct etnaviv_gem_object *etnaviv_obj)
 		pinned = __get_user_pages_fast(etnaviv_obj->userptr.ptr, npages,
 					       !etnaviv_obj->userptr.ro, pvec);
 		if (pinned < 0) {
-			drm_free_large(pvec);
+			kvfree(pvec);
 			mmput(mm);
 			return pinned;
 		}
@@ -845,7 +845,7 @@ static int etnaviv_gem_userptr_get_pages(struct etnaviv_gem_object *etnaviv_obj)
 	}
 
 	release_pages(pvec, pinned, 0);
-	drm_free_large(pvec);
+	kvfree(pvec);
 
 	work = kmalloc(sizeof(*work), GFP_KERNEL);
 	if (!work) {
@@ -879,7 +879,7 @@ static void etnaviv_gem_userptr_release(struct etnaviv_gem_object *etnaviv_obj)
 		int npages = etnaviv_obj->base.size >> PAGE_SHIFT;
 
 		release_pages(etnaviv_obj->pages, npages, 0);
-		drm_free_large(etnaviv_obj->pages);
+		kvfree(etnaviv_obj->pages);
 	}
 	put_task_struct(etnaviv_obj->userptr.task);
 }
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c b/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c
index 62b47972a52e..367bf952f61a 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c
@@ -87,7 +87,7 @@ static void etnaviv_gem_prime_release(struct etnaviv_gem_object *etnaviv_obj)
 	 * ours, just free the array we allocated:
 	 */
 	if (etnaviv_obj->pages)
-		drm_free_large(etnaviv_obj->pages);
+		kvfree(etnaviv_obj->pages);
 
 	drm_prime_gem_destroy(&etnaviv_obj->base, etnaviv_obj->sgt);
 }
@@ -128,7 +128,7 @@ struct drm_gem_object *etnaviv_gem_prime_import_sg_table(struct drm_device *dev,
 	npages = size / PAGE_SIZE;
 
 	etnaviv_obj->sgt = sgt;
-	etnaviv_obj->pages = drm_malloc_ab(npages, sizeof(struct page *));
+	etnaviv_obj->pages = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL);
 	if (!etnaviv_obj->pages) {
 		ret = -ENOMEM;
 		goto fail;
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c b/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c
index de80ee1b71df..ee7069e93eda 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c
@@ -345,9 +345,9 @@ int etnaviv_ioctl_gem_submit(struct drm_device *dev, void *data,
 	 * Copy the command submission and bo array to kernel space in
 	 * one go, and do this outside of any locks.
 	 */
-	bos = drm_malloc_ab(args->nr_bos, sizeof(*bos));
-	relocs = drm_malloc_ab(args->nr_relocs, sizeof(*relocs));
-	stream = drm_malloc_ab(1, args->stream_size);
+	bos = kvmalloc_array(args->nr_bos, sizeof(*bos), GFP_KERNEL);
+	relocs = kvmalloc_array(args->nr_relocs, sizeof(*relocs), GFP_KERNEL);
+	stream = kvmalloc_array(1, args->stream_size, GFP_KERNEL);
 	cmdbuf = etnaviv_cmdbuf_new(gpu->cmdbuf_suballoc,
 				    ALIGN(args->stream_size, 8) + 8,
 				    args->nr_bos);
@@ -489,11 +489,11 @@ err_submit_cmds:
 	if (cmdbuf)
 		etnaviv_cmdbuf_free(cmdbuf);
 	if (stream)
-		drm_free_large(stream);
+		kvfree(stream);
 	if (bos)
-		drm_free_large(bos);
+		kvfree(bos);
 	if (relocs)
-		drm_free_large(relocs);
+		kvfree(relocs);
 
 	return ret;
 }
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gem.c b/drivers/gpu/drm/exynos/exynos_drm_gem.c
index 55a1579d11b3..c23479be4850 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gem.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_gem.c
@@ -59,7 +59,8 @@ static int exynos_drm_alloc_buf(struct exynos_drm_gem *exynos_gem)
 
 	nr_pages = exynos_gem->size >> PAGE_SHIFT;
 
-	exynos_gem->pages = drm_calloc_large(nr_pages, sizeof(struct page *));
+	exynos_gem->pages = kvmalloc_array(nr_pages, sizeof(struct page *),
+			GFP_KERNEL | __GFP_ZERO);
 	if (!exynos_gem->pages) {
 		DRM_ERROR("failed to allocate pages.\n");
 		return -ENOMEM;
@@ -101,7 +102,7 @@ err_dma_free:
 	dma_free_attrs(to_dma_dev(dev), exynos_gem->size, exynos_gem->cookie,
 		       exynos_gem->dma_addr, exynos_gem->dma_attrs);
 err_free:
-	drm_free_large(exynos_gem->pages);
+	kvfree(exynos_gem->pages);
 
 	return ret;
 }
@@ -122,7 +123,7 @@ static void exynos_drm_free_buf(struct exynos_drm_gem *exynos_gem)
 			(dma_addr_t)exynos_gem->dma_addr,
 			exynos_gem->dma_attrs);
 
-	drm_free_large(exynos_gem->pages);
+	kvfree(exynos_gem->pages);
 }
 
 static int exynos_drm_gem_handle_create(struct drm_gem_object *obj,
@@ -559,7 +560,7 @@ exynos_drm_gem_prime_import_sg_table(struct drm_device *dev,
 	exynos_gem->dma_addr = sg_dma_address(sgt->sgl);
 
 	npages = exynos_gem->size >> PAGE_SHIFT;
-	exynos_gem->pages = drm_malloc_ab(npages, sizeof(struct page *));
+	exynos_gem->pages = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL);
 	if (!exynos_gem->pages) {
 		ret = -ENOMEM;
 		goto err;
@@ -588,7 +589,7 @@ exynos_drm_gem_prime_import_sg_table(struct drm_device *dev,
 	return &exynos_gem->base;
 
 err_free_large:
-	drm_free_large(exynos_gem->pages);
+	kvfree(exynos_gem->pages);
 err:
 	drm_gem_object_release(&exynos_gem->base);
 	kfree(exynos_gem);
diff --git a/drivers/gpu/drm/gma500/Makefile b/drivers/gpu/drm/gma500/Makefile
index 190e55f2f891..c1c8dc18aa53 100644
--- a/drivers/gpu/drm/gma500/Makefile
+++ b/drivers/gpu/drm/gma500/Makefile
@@ -1,7 +1,6 @@
 #
 #	KMS driver for the GMA500
 #
-ccflags-y += -I$(srctree)/include/drm
 
 gma500_gfx-y += \
 	  accel_2d.o \
diff --git a/drivers/gpu/drm/gma500/mdfld_tpo_vid.c b/drivers/gpu/drm/gma500/mdfld_tpo_vid.c
index d8d4170725b2..a9420bf9a419 100644
--- a/drivers/gpu/drm/gma500/mdfld_tpo_vid.c
+++ b/drivers/gpu/drm/gma500/mdfld_tpo_vid.c
@@ -30,55 +30,20 @@
 static struct drm_display_mode *tpo_vid_get_config_mode(struct drm_device *dev)
 {
 	struct drm_display_mode *mode;
-	struct drm_psb_private *dev_priv = dev->dev_private;
-	struct oaktrail_timing_info *ti = &dev_priv->gct_data.DTD;
-	bool use_gct = false;
 
 	mode = kzalloc(sizeof(*mode), GFP_KERNEL);
 	if (!mode)
 		return NULL;
 
-	if (use_gct) {
-		mode->hdisplay = (ti->hactive_hi << 8) | ti->hactive_lo;
-		mode->vdisplay = (ti->vactive_hi << 8) | ti->vactive_lo;
-		mode->hsync_start = mode->hdisplay +
-				((ti->hsync_offset_hi << 8) |
-				ti->hsync_offset_lo);
-		mode->hsync_end = mode->hsync_start +
-				((ti->hsync_pulse_width_hi << 8) |
-				ti->hsync_pulse_width_lo);
-		mode->htotal = mode->hdisplay + ((ti->hblank_hi << 8) |
-								ti->hblank_lo);
-		mode->vsync_start =
-			mode->vdisplay + ((ti->vsync_offset_hi << 8) |
-						ti->vsync_offset_lo);
-		mode->vsync_end =
-			mode->vsync_start + ((ti->vsync_pulse_width_hi << 8) |
-						ti->vsync_pulse_width_lo);
-		mode->vtotal = mode->vdisplay +
-				((ti->vblank_hi << 8) | ti->vblank_lo);
-		mode->clock = ti->pixel_clock * 10;
-
-		dev_dbg(dev->dev, "hdisplay is %d\n", mode->hdisplay);
-		dev_dbg(dev->dev, "vdisplay is %d\n", mode->vdisplay);
-		dev_dbg(dev->dev, "HSS is %d\n", mode->hsync_start);
-		dev_dbg(dev->dev, "HSE is %d\n", mode->hsync_end);
-		dev_dbg(dev->dev, "htotal is %d\n", mode->htotal);
-		dev_dbg(dev->dev, "VSS is %d\n", mode->vsync_start);
-		dev_dbg(dev->dev, "VSE is %d\n", mode->vsync_end);
-		dev_dbg(dev->dev, "vtotal is %d\n", mode->vtotal);
-		dev_dbg(dev->dev, "clock is %d\n", mode->clock);
-	} else {
-		mode->hdisplay = 864;
-		mode->vdisplay = 480;
-		mode->hsync_start = 873;
-		mode->hsync_end = 876;
-		mode->htotal = 887;
-		mode->vsync_start = 487;
-		mode->vsync_end = 490;
-		mode->vtotal = 499;
-		mode->clock = 33264;
-	}
+	mode->hdisplay = 864;
+	mode->vdisplay = 480;
+	mode->hsync_start = 873;
+	mode->hsync_end = 876;
+	mode->htotal = 887;
+	mode->vsync_start = 487;
+	mode->vsync_end = 490;
+	mode->vtotal = 499;
+	mode->clock = 33264;
 
 	drm_mode_set_name(mode);
 	drm_mode_set_crtcinfo(mode, 0);
diff --git a/drivers/gpu/drm/hisilicon/hibmc/Makefile b/drivers/gpu/drm/hisilicon/hibmc/Makefile
index f2e04c035673..3df726696372 100644
--- a/drivers/gpu/drm/hisilicon/hibmc/Makefile
+++ b/drivers/gpu/drm/hisilicon/hibmc/Makefile
@@ -1,4 +1,3 @@
-ccflags-y := -Iinclude/drm
 hibmc-drm-y := hibmc_drm_drv.o hibmc_drm_de.o hibmc_drm_vdac.o hibmc_drm_fbdev.o hibmc_ttm.o
 
 obj-$(CONFIG_DRM_HISI_HIBMC) += hibmc-drm.o
diff --git a/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c b/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c
index 20732b62d4c9..ac457c779caa 100644
--- a/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c
+++ b/drivers/gpu/drm/hisilicon/hibmc/hibmc_ttm.c
@@ -17,7 +17,7 @@
  */
 
 #include <drm/drm_atomic_helper.h>
-#include <ttm/ttm_page_alloc.h>
+#include <drm/ttm/ttm_page_alloc.h>
 
 #include "hibmc_drm_drv.h"
 
diff --git a/drivers/gpu/drm/i2c/Makefile b/drivers/gpu/drm/i2c/Makefile
index 43aa33baebed..a77acfc1852e 100644
--- a/drivers/gpu/drm/i2c/Makefile
+++ b/drivers/gpu/drm/i2c/Makefile
@@ -1,5 +1,3 @@
-ccflags-y := -Iinclude/drm
-
 ch7006-y := ch7006_drv.o ch7006_mode.o
 obj-$(CONFIG_DRM_I2C_CH7006) += ch7006.o
 
diff --git a/drivers/gpu/drm/i810/Makefile b/drivers/gpu/drm/i810/Makefile
index 43844ecafcc5..639f8596c978 100644
--- a/drivers/gpu/drm/i810/Makefile
+++ b/drivers/gpu/drm/i810/Makefile
@@ -2,7 +2,6 @@
 # Makefile for the drm device driver.  This driver provides support for the
 # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.
 
-ccflags-y := -Iinclude/drm
 i810-y := i810_drv.o i810_dma.o
 
 obj-$(CONFIG_DRM_I810)	+= i810.o
diff --git a/drivers/gpu/drm/i915/Kconfig.debug b/drivers/gpu/drm/i915/Kconfig.debug
index b00edd3b8800..78c5c049a347 100644
--- a/drivers/gpu/drm/i915/Kconfig.debug
+++ b/drivers/gpu/drm/i915/Kconfig.debug
@@ -61,6 +61,18 @@ config DRM_I915_SW_FENCE_DEBUG_OBJECTS
 
           If in doubt, say "N".
 
+config DRM_I915_SW_FENCE_CHECK_DAG
+        bool "Enable additional driver debugging for detecting dependency cycles"
+        depends on DRM_I915
+        default n
+        help
+          Choose this option to turn on extra driver debugging that may affect
+          performance but will catch some internal issues.
+
+          Recommended for driver developers only.
+
+          If in doubt, say "N".
+
 config DRM_I915_SELFTEST
 	bool "Enable selftests upon driver load"
 	depends on DRM_I915
diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile
index 2cf04504e494..16dccf550412 100644
--- a/drivers/gpu/drm/i915/Makefile
+++ b/drivers/gpu/drm/i915/Makefile
@@ -16,6 +16,7 @@ i915-y := i915_drv.o \
 	  i915_params.o \
 	  i915_pci.o \
           i915_suspend.o \
+	  i915_syncmap.o \
 	  i915_sw_fence.o \
 	  i915_sysfs.o \
 	  intel_csr.o \
@@ -57,6 +58,7 @@ i915-y += i915_cmd_parser.o \
 
 # general-purpose microcontroller (GuC) support
 i915-y += intel_uc.o \
+	  intel_guc_ct.o \
 	  intel_guc_log.o \
 	  intel_guc_loader.o \
 	  intel_huc.o \
diff --git a/drivers/gpu/drm/i915/dvo_ch7017.c b/drivers/gpu/drm/i915/dvo_ch7017.c
index b3c7c199200c..80b3e16cf48c 100644
--- a/drivers/gpu/drm/i915/dvo_ch7017.c
+++ b/drivers/gpu/drm/i915/dvo_ch7017.c
@@ -280,10 +280,10 @@ static void ch7017_mode_set(struct intel_dvo_device *dvo,
 			(0 << CH7017_PHASE_DETECTOR_SHIFT);
 	} else {
 		outputs_enable = CH7017_LVDS_CHANNEL_A | CH7017_CHARGE_PUMP_HIGH;
-		lvds_pll_feedback_div = CH7017_LVDS_PLL_FEEDBACK_DEFAULT_RESERVED |
+		lvds_pll_feedback_div =
+			CH7017_LVDS_PLL_FEEDBACK_DEFAULT_RESERVED |
 			(2 << CH7017_LVDS_PLL_FEED_BACK_DIVIDER_SHIFT) |
 			(3 << CH7017_LVDS_PLL_FEED_FORWARD_DIVIDER_SHIFT);
-		lvds_pll_feedback_div = 35;
 		lvds_control_2 = (3 << CH7017_LOOP_FILTER_SHIFT) |
 			(0 << CH7017_PHASE_DETECTOR_SHIFT);
 		if (1) { /* XXX: dual channel panel detection.  Assume yes for now. */
diff --git a/drivers/gpu/drm/i915/dvo_ch7xxx.c b/drivers/gpu/drm/i915/dvo_ch7xxx.c
index 44b3159f2fe8..7aeeffd2428b 100644
--- a/drivers/gpu/drm/i915/dvo_ch7xxx.c
+++ b/drivers/gpu/drm/i915/dvo_ch7xxx.c
@@ -217,9 +217,8 @@ static bool ch7xxx_init(struct intel_dvo_device *dvo,
 
 	name = ch7xxx_get_id(vendor);
 	if (!name) {
-		DRM_DEBUG_KMS("ch7xxx not detected; got 0x%02x from %s "
-				"slave %d.\n",
-			  vendor, adapter->name, dvo->slave_addr);
+		DRM_DEBUG_KMS("ch7xxx not detected; got VID 0x%02x from %s slave %d.\n",
+			      vendor, adapter->name, dvo->slave_addr);
 		goto out;
 	}
 
@@ -229,9 +228,8 @@ static bool ch7xxx_init(struct intel_dvo_device *dvo,
 
 	devid = ch7xxx_get_did(device);
 	if (!devid) {
-		DRM_DEBUG_KMS("ch7xxx not detected; got 0x%02x from %s "
-				"slave %d.\n",
-			  vendor, adapter->name, dvo->slave_addr);
+		DRM_DEBUG_KMS("ch7xxx not detected; got DID 0x%02x from %s slave %d.\n",
+			      device, adapter->name, dvo->slave_addr);
 		goto out;
 	}
 
diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c
index bada32b33237..6ae286cb5804 100644
--- a/drivers/gpu/drm/i915/gvt/scheduler.c
+++ b/drivers/gpu/drm/i915/gvt/scheduler.c
@@ -69,8 +69,7 @@ static int populate_shadow_context(struct intel_vgpu_workload *workload)
 	gvt_dbg_sched("ring id %d workload lrca %x", ring_id,
 			workload->ctx_desc.lrca);
 
-	context_page_num = intel_lr_context_size(
-			gvt->dev_priv->engine[ring_id]);
+	context_page_num = gvt->dev_priv->engine[ring_id]->context_size;
 
 	context_page_num = context_page_num >> PAGE_SHIFT;
 
@@ -181,6 +180,7 @@ static int dispatch_workload(struct intel_vgpu_workload *workload)
 	struct intel_engine_cs *engine = dev_priv->engine[ring_id];
 	struct drm_i915_gem_request *rq;
 	struct intel_vgpu *vgpu = workload->vgpu;
+	struct intel_ring *ring;
 	int ret;
 
 	gvt_dbg_sched("ring id %d prepare to dispatch workload %p\n",
@@ -199,8 +199,9 @@ static int dispatch_workload(struct intel_vgpu_workload *workload)
 	 * shadow_ctx pages invalid. So gvt need to pin itself. After update
 	 * the guest context, gvt can unpin the shadow_ctx safely.
 	 */
-	ret = engine->context_pin(engine, shadow_ctx);
-	if (ret) {
+	ring = engine->context_pin(engine, shadow_ctx);
+	if (IS_ERR(ring)) {
+		ret = PTR_ERR(ring);
 		gvt_vgpu_err("fail to pin shadow context\n");
 		workload->status = ret;
 		mutex_unlock(&dev_priv->drm.struct_mutex);
@@ -330,8 +331,7 @@ static void update_guest_context(struct intel_vgpu_workload *workload)
 	gvt_dbg_sched("ring id %d workload lrca %x\n", ring_id,
 			workload->ctx_desc.lrca);
 
-	context_page_num = intel_lr_context_size(
-			gvt->dev_priv->engine[ring_id]);
+	context_page_num = gvt->dev_priv->engine[ring_id]->context_size;
 
 	context_page_num = context_page_num >> PAGE_SHIFT;
 
diff --git a/drivers/gpu/drm/i915/i915_cmd_parser.c b/drivers/gpu/drm/i915/i915_cmd_parser.c
index 7af100f84410..f0cb22cc0dd6 100644
--- a/drivers/gpu/drm/i915/i915_cmd_parser.c
+++ b/drivers/gpu/drm/i915/i915_cmd_parser.c
@@ -1166,8 +1166,8 @@ static bool check_cmd(const struct intel_engine_cs *engine,
 				find_reg(engine, is_master, reg_addr);
 
 			if (!reg) {
-				DRM_DEBUG_DRIVER("CMD: Rejected register 0x%08X in command: 0x%08X (exec_id=%d)\n",
-						 reg_addr, *cmd, engine->exec_id);
+				DRM_DEBUG_DRIVER("CMD: Rejected register 0x%08X in command: 0x%08X (%s)\n",
+						 reg_addr, *cmd, engine->name);
 				return false;
 			}
 
@@ -1222,11 +1222,11 @@ static bool check_cmd(const struct intel_engine_cs *engine,
 				desc->bits[i].mask;
 
 			if (dword != desc->bits[i].expected) {
-				DRM_DEBUG_DRIVER("CMD: Rejected command 0x%08X for bitmask 0x%08X (exp=0x%08X act=0x%08X) (exec_id=%d)\n",
+				DRM_DEBUG_DRIVER("CMD: Rejected command 0x%08X for bitmask 0x%08X (exp=0x%08X act=0x%08X) (%s)\n",
 						 *cmd,
 						 desc->bits[i].mask,
 						 desc->bits[i].expected,
-						 dword, engine->exec_id);
+						 dword, engine->name);
 				return false;
 			}
 		}
@@ -1284,7 +1284,7 @@ int intel_engine_cmd_parser(struct intel_engine_cs *engine,
 
 		if (*cmd == MI_BATCH_BUFFER_END) {
 			if (needs_clflush_after) {
-				void *ptr = ptr_mask_bits(shadow_batch_obj->mm.mapping);
+				void *ptr = page_mask_bits(shadow_batch_obj->mm.mapping);
 				drm_clflush_virt_range(ptr,
 						       (void *)(cmd + 1) - ptr);
 			}
diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index d689e511744e..a6ba2100bb88 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -229,7 +229,7 @@ static int i915_gem_stolen_list_info(struct seq_file *m, void *data)
 	int ret;
 
 	total = READ_ONCE(dev_priv->mm.object_count);
-	objects = drm_malloc_ab(total, sizeof(*objects));
+	objects = kvmalloc_array(total, sizeof(*objects), GFP_KERNEL);
 	if (!objects)
 		return -ENOMEM;
 
@@ -274,7 +274,7 @@ static int i915_gem_stolen_list_info(struct seq_file *m, void *data)
 
 	mutex_unlock(&dev->struct_mutex);
 out:
-	drm_free_large(objects);
+	kvfree(objects);
 	return ret;
 }
 
@@ -1670,12 +1670,22 @@ static int i915_fbc_status(struct seq_file *m, void *unused)
 		seq_printf(m, "FBC disabled: %s\n",
 			   dev_priv->fbc.no_fbc_reason);
 
-	if (intel_fbc_is_active(dev_priv) && INTEL_GEN(dev_priv) >= 7) {
-		uint32_t mask = INTEL_GEN(dev_priv) >= 8 ?
-				BDW_FBC_COMPRESSION_MASK :
-				IVB_FBC_COMPRESSION_MASK;
-		seq_printf(m, "Compressing: %s\n",
-			   yesno(I915_READ(FBC_STATUS2) & mask));
+	if (intel_fbc_is_active(dev_priv)) {
+		u32 mask;
+
+		if (INTEL_GEN(dev_priv) >= 8)
+			mask = I915_READ(IVB_FBC_STATUS2) & BDW_FBC_COMP_SEG_MASK;
+		else if (INTEL_GEN(dev_priv) >= 7)
+			mask = I915_READ(IVB_FBC_STATUS2) & IVB_FBC_COMP_SEG_MASK;
+		else if (INTEL_GEN(dev_priv) >= 5)
+			mask = I915_READ(ILK_DPFC_STATUS) & ILK_DPFC_COMP_SEG_MASK;
+		else if (IS_G4X(dev_priv))
+			mask = I915_READ(DPFC_STATUS) & DPFC_COMP_SEG_MASK;
+		else
+			mask = I915_READ(FBC_STATUS) & (FBC_STAT_COMPRESSING |
+							FBC_STAT_COMPRESSED);
+
+		seq_printf(m, "Compressing: %s\n", yesno(mask));
 	}
 
 	mutex_unlock(&dev_priv->fbc.lock);
@@ -1684,7 +1694,7 @@ static int i915_fbc_status(struct seq_file *m, void *unused)
 	return 0;
 }
 
-static int i915_fbc_fc_get(void *data, u64 *val)
+static int i915_fbc_false_color_get(void *data, u64 *val)
 {
 	struct drm_i915_private *dev_priv = data;
 
@@ -1696,7 +1706,7 @@ static int i915_fbc_fc_get(void *data, u64 *val)
 	return 0;
 }
 
-static int i915_fbc_fc_set(void *data, u64 val)
+static int i915_fbc_false_color_set(void *data, u64 val)
 {
 	struct drm_i915_private *dev_priv = data;
 	u32 reg;
@@ -1717,8 +1727,8 @@ static int i915_fbc_fc_set(void *data, u64 val)
 	return 0;
 }
 
-DEFINE_SIMPLE_ATTRIBUTE(i915_fbc_fc_fops,
-			i915_fbc_fc_get, i915_fbc_fc_set,
+DEFINE_SIMPLE_ATTRIBUTE(i915_fbc_false_color_fops,
+			i915_fbc_false_color_get, i915_fbc_false_color_set,
 			"%llu\n");
 
 static int i915_ips_status(struct seq_file *m, void *unused)
@@ -2482,8 +2492,6 @@ static void i915_guc_client_info(struct seq_file *m,
 		client->wq_size, client->wq_offset, client->wq_tail);
 
 	seq_printf(m, "\tWork queue full: %u\n", client->no_wq_space);
-	seq_printf(m, "\tFailed doorbell: %u\n", client->b_fail);
-	seq_printf(m, "\tLast submission result: %d\n", client->retcode);
 
 	for_each_engine(engine, dev_priv, id) {
 		u64 submissions = client->submissions[id];
@@ -2494,42 +2502,34 @@ static void i915_guc_client_info(struct seq_file *m,
 	seq_printf(m, "\tTotal: %llu\n", tot);
 }
 
-static int i915_guc_info(struct seq_file *m, void *data)
+static bool check_guc_submission(struct seq_file *m)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
 	const struct intel_guc *guc = &dev_priv->guc;
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
-	u64 total;
 
 	if (!guc->execbuf_client) {
 		seq_printf(m, "GuC submission %s\n",
 			   HAS_GUC_SCHED(dev_priv) ?
 			   "disabled" :
 			   "not supported");
-		return 0;
+		return false;
 	}
 
+	return true;
+}
+
+static int i915_guc_info(struct seq_file *m, void *data)
+{
+	struct drm_i915_private *dev_priv = node_to_i915(m->private);
+	const struct intel_guc *guc = &dev_priv->guc;
+
+	if (!check_guc_submission(m))
+		return 0;
+
 	seq_printf(m, "Doorbell map:\n");
 	seq_printf(m, "\t%*pb\n", GUC_NUM_DOORBELLS, guc->doorbell_bitmap);
 	seq_printf(m, "Doorbell next cacheline: 0x%x\n\n", guc->db_cacheline);
 
-	seq_printf(m, "GuC total action count: %llu\n", guc->action_count);
-	seq_printf(m, "GuC action failure count: %u\n", guc->action_fail);
-	seq_printf(m, "GuC last action command: 0x%x\n", guc->action_cmd);
-	seq_printf(m, "GuC last action status: 0x%x\n", guc->action_status);
-	seq_printf(m, "GuC last action error code: %d\n", guc->action_err);
-
-	total = 0;
-	seq_printf(m, "\nGuC submissions:\n");
-	for_each_engine(engine, dev_priv, id) {
-		u64 submissions = guc->submissions[id];
-		total += submissions;
-		seq_printf(m, "\t%-24s: %10llu, last seqno 0x%08x\n",
-			engine->name, submissions, guc->last_seqno[id]);
-	}
-	seq_printf(m, "\t%s: %llu\n", "Total", total);
-
 	seq_printf(m, "\nGuC execbuf client @ %p:\n", guc->execbuf_client);
 	i915_guc_client_info(m, dev_priv, guc->execbuf_client);
 
@@ -2540,36 +2540,99 @@ static int i915_guc_info(struct seq_file *m, void *data)
 	return 0;
 }
 
-static int i915_guc_log_dump(struct seq_file *m, void *data)
+static int i915_guc_stage_pool(struct seq_file *m, void *data)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	struct drm_i915_gem_object *obj;
-	int i = 0, pg;
+	const struct intel_guc *guc = &dev_priv->guc;
+	struct guc_stage_desc *desc = guc->stage_desc_pool_vaddr;
+	struct i915_guc_client *client = guc->execbuf_client;
+	unsigned int tmp;
+	int index;
 
-	if (!dev_priv->guc.log.vma)
+	if (!check_guc_submission(m))
 		return 0;
 
-	obj = dev_priv->guc.log.vma->obj;
-	for (pg = 0; pg < obj->base.size / PAGE_SIZE; pg++) {
-		u32 *log = kmap_atomic(i915_gem_object_get_page(obj, pg));
+	for (index = 0; index < GUC_MAX_STAGE_DESCRIPTORS; index++, desc++) {
+		struct intel_engine_cs *engine;
 
-		for (i = 0; i < PAGE_SIZE / sizeof(u32); i += 4)
-			seq_printf(m, "0x%08x 0x%08x 0x%08x 0x%08x\n",
-				   *(log + i), *(log + i + 1),
-				   *(log + i + 2), *(log + i + 3));
+		if (!(desc->attribute & GUC_STAGE_DESC_ATTR_ACTIVE))
+			continue;
+
+		seq_printf(m, "GuC stage descriptor %u:\n", index);
+		seq_printf(m, "\tIndex: %u\n", desc->stage_id);
+		seq_printf(m, "\tAttribute: 0x%x\n", desc->attribute);
+		seq_printf(m, "\tPriority: %d\n", desc->priority);
+		seq_printf(m, "\tDoorbell id: %d\n", desc->db_id);
+		seq_printf(m, "\tEngines used: 0x%x\n",
+			   desc->engines_used);
+		seq_printf(m, "\tDoorbell trigger phy: 0x%llx, cpu: 0x%llx, uK: 0x%x\n",
+			   desc->db_trigger_phy,
+			   desc->db_trigger_cpu,
+			   desc->db_trigger_uk);
+		seq_printf(m, "\tProcess descriptor: 0x%x\n",
+			   desc->process_desc);
+		seq_printf(m, "\tWorkqueue address: 0x%x, size: 0x%x\n",
+			   desc->wq_addr, desc->wq_size);
+		seq_putc(m, '\n');
 
-		kunmap_atomic(log);
+		for_each_engine_masked(engine, dev_priv, client->engines, tmp) {
+			u32 guc_engine_id = engine->guc_id;
+			struct guc_execlist_context *lrc =
+						&desc->lrc[guc_engine_id];
+
+			seq_printf(m, "\t%s LRC:\n", engine->name);
+			seq_printf(m, "\t\tContext desc: 0x%x\n",
+				   lrc->context_desc);
+			seq_printf(m, "\t\tContext id: 0x%x\n", lrc->context_id);
+			seq_printf(m, "\t\tLRCA: 0x%x\n", lrc->ring_lrca);
+			seq_printf(m, "\t\tRing begin: 0x%x\n", lrc->ring_begin);
+			seq_printf(m, "\t\tRing end: 0x%x\n", lrc->ring_end);
+			seq_putc(m, '\n');
+		}
 	}
 
+	return 0;
+}
+
+static int i915_guc_log_dump(struct seq_file *m, void *data)
+{
+	struct drm_info_node *node = m->private;
+	struct drm_i915_private *dev_priv = node_to_i915(node);
+	bool dump_load_err = !!node->info_ent->data;
+	struct drm_i915_gem_object *obj = NULL;
+	u32 *log;
+	int i = 0;
+
+	if (dump_load_err)
+		obj = dev_priv->guc.load_err_log;
+	else if (dev_priv->guc.log.vma)
+		obj = dev_priv->guc.log.vma->obj;
+
+	if (!obj)
+		return 0;
+
+	log = i915_gem_object_pin_map(obj, I915_MAP_WC);
+	if (IS_ERR(log)) {
+		DRM_DEBUG("Failed to pin object\n");
+		seq_puts(m, "(log data unaccessible)\n");
+		return PTR_ERR(log);
+	}
+
+	for (i = 0; i < obj->base.size / sizeof(u32); i += 4)
+		seq_printf(m, "0x%08x 0x%08x 0x%08x 0x%08x\n",
+			   *(log + i), *(log + i + 1),
+			   *(log + i + 2), *(log + i + 3));
+
 	seq_putc(m, '\n');
 
+	i915_gem_object_unpin_map(obj);
+
 	return 0;
 }
 
 static int i915_guc_log_control_get(void *data, u64 *val)
 {
-	struct drm_device *dev = data;
-	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct drm_i915_private *dev_priv = data;
 
 	if (!dev_priv->guc.log.vma)
 		return -EINVAL;
@@ -2581,14 +2644,13 @@ static int i915_guc_log_control_get(void *data, u64 *val)
 
 static int i915_guc_log_control_set(void *data, u64 val)
 {
-	struct drm_device *dev = data;
-	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct drm_i915_private *dev_priv = data;
 	int ret;
 
 	if (!dev_priv->guc.log.vma)
 		return -EINVAL;
 
-	ret = mutex_lock_interruptible(&dev->struct_mutex);
+	ret = mutex_lock_interruptible(&dev_priv->drm.struct_mutex);
 	if (ret)
 		return ret;
 
@@ -2596,7 +2658,7 @@ static int i915_guc_log_control_set(void *data, u64 val)
 	ret = i915_guc_log_control(dev_priv, val);
 	intel_runtime_pm_put(dev_priv);
 
-	mutex_unlock(&dev->struct_mutex);
+	mutex_unlock(&dev_priv->drm.struct_mutex);
 	return ret;
 }
 
@@ -2855,7 +2917,8 @@ static int i915_dmc_info(struct seq_file *m, void *unused)
 	seq_printf(m, "version: %d.%d\n", CSR_VERSION_MAJOR(csr->version),
 		   CSR_VERSION_MINOR(csr->version));
 
-	if (IS_SKYLAKE(dev_priv) && csr->version >= CSR_VERSION(1, 6)) {
+	if (IS_KABYLAKE(dev_priv) ||
+	    (IS_SKYLAKE(dev_priv) && csr->version >= CSR_VERSION(1, 6))) {
 		seq_printf(m, "DC3 -> DC5 count: %d\n",
 			   I915_READ(SKL_CSR_DC3_DC5_COUNT));
 		seq_printf(m, "DC5 -> DC6 count: %d\n",
@@ -3043,36 +3106,6 @@ static void intel_connector_info(struct seq_file *m,
 		intel_seq_print_mode(m, 2, mode);
 }
 
-static bool cursor_active(struct drm_i915_private *dev_priv, int pipe)
-{
-	u32 state;
-
-	if (IS_I845G(dev_priv) || IS_I865G(dev_priv))
-		state = I915_READ(CURCNTR(PIPE_A)) & CURSOR_ENABLE;
-	else
-		state = I915_READ(CURCNTR(pipe)) & CURSOR_MODE;
-
-	return state;
-}
-
-static bool cursor_position(struct drm_i915_private *dev_priv,
-			    int pipe, int *x, int *y)
-{
-	u32 pos;
-
-	pos = I915_READ(CURPOS(pipe));
-
-	*x = (pos >> CURSOR_X_SHIFT) & CURSOR_POS_MASK;
-	if (pos & (CURSOR_POS_SIGN << CURSOR_X_SHIFT))
-		*x = -*x;
-
-	*y = (pos >> CURSOR_Y_SHIFT) & CURSOR_POS_MASK;
-	if (pos & (CURSOR_POS_SIGN << CURSOR_Y_SHIFT))
-		*y = -*y;
-
-	return cursor_active(dev_priv, pipe);
-}
-
 static const char *plane_type(enum drm_plane_type type)
 {
 	switch (type) {
@@ -3095,17 +3128,17 @@ static const char *plane_rotation(unsigned int rotation)
 {
 	static char buf[48];
 	/*
-	 * According to doc only one DRM_ROTATE_ is allowed but this
+	 * According to doc only one DRM_MODE_ROTATE_ is allowed but this
 	 * will print them all to visualize if the values are misused
 	 */
 	snprintf(buf, sizeof(buf),
 		 "%s%s%s%s%s%s(0x%08x)",
-		 (rotation & DRM_ROTATE_0) ? "0 " : "",
-		 (rotation & DRM_ROTATE_90) ? "90 " : "",
-		 (rotation & DRM_ROTATE_180) ? "180 " : "",
-		 (rotation & DRM_ROTATE_270) ? "270 " : "",
-		 (rotation & DRM_REFLECT_X) ? "FLIPX " : "",
-		 (rotation & DRM_REFLECT_Y) ? "FLIPY " : "",
+		 (rotation & DRM_MODE_ROTATE_0) ? "0 " : "",
+		 (rotation & DRM_MODE_ROTATE_90) ? "90 " : "",
+		 (rotation & DRM_MODE_ROTATE_180) ? "180 " : "",
+		 (rotation & DRM_MODE_ROTATE_270) ? "270 " : "",
+		 (rotation & DRM_MODE_REFLECT_X) ? "FLIPX " : "",
+		 (rotation & DRM_MODE_REFLECT_Y) ? "FLIPY " : "",
 		 rotation);
 
 	return buf;
@@ -3194,9 +3227,7 @@ static int i915_display_info(struct seq_file *m, void *unused)
 	seq_printf(m, "CRTC info\n");
 	seq_printf(m, "---------\n");
 	for_each_intel_crtc(dev, crtc) {
-		bool active;
 		struct intel_crtc_state *pipe_config;
-		int x, y;
 
 		drm_modeset_lock(&crtc->base.mutex, NULL);
 		pipe_config = to_intel_crtc_state(crtc->base.state);
@@ -3208,14 +3239,18 @@ static int i915_display_info(struct seq_file *m, void *unused)
 			   yesno(pipe_config->dither), pipe_config->pipe_bpp);
 
 		if (pipe_config->base.active) {
+			struct intel_plane *cursor =
+				to_intel_plane(crtc->base.cursor);
+
 			intel_crtc_info(m, crtc);
 
-			active = cursor_position(dev_priv, crtc->pipe, &x, &y);
-			seq_printf(m, "\tcursor visible? %s, position (%d, %d), size %dx%d, addr 0x%08x, active? %s\n",
-				   yesno(crtc->cursor_base),
-				   x, y, crtc->base.cursor->state->crtc_w,
-				   crtc->base.cursor->state->crtc_h,
-				   crtc->cursor_addr, yesno(active));
+			seq_printf(m, "\tcursor visible? %s, position (%d, %d), size %dx%d, addr 0x%08x\n",
+				   yesno(cursor->base.state->visible),
+				   cursor->base.state->crtc_x,
+				   cursor->base.state->crtc_y,
+				   cursor->base.state->crtc_w,
+				   cursor->base.state->crtc_h,
+				   cursor->cursor.base);
 			intel_scaler_info(m, crtc);
 			intel_plane_info(m, crtc);
 		}
@@ -3316,7 +3351,7 @@ static int i915_engine_info(struct seq_file *m, void *unused)
 
 		if (i915.enable_execlists) {
 			u32 ptr, read, write;
-			struct rb_node *rb;
+			unsigned int idx;
 
 			seq_printf(m, "\tExeclist status: 0x%08x %08x\n",
 				   I915_READ(RING_EXECLIST_STATUS_LO(engine)),
@@ -3334,8 +3369,7 @@ static int i915_engine_info(struct seq_file *m, void *unused)
 			if (read > write)
 				write += GEN8_CSB_ENTRIES;
 			while (read < write) {
-				unsigned int idx = ++read % GEN8_CSB_ENTRIES;
-
+				idx = ++read % GEN8_CSB_ENTRIES;
 				seq_printf(m, "\tExeclist CSB[%d]: 0x%08x, context: %d\n",
 					   idx,
 					   I915_READ(RING_CONTEXT_STATUS_BUF_LO(engine, idx)),
@@ -3343,28 +3377,30 @@ static int i915_engine_info(struct seq_file *m, void *unused)
 			}
 
 			rcu_read_lock();
-			rq = READ_ONCE(engine->execlist_port[0].request);
-			if (rq) {
-				seq_printf(m, "\t\tELSP[0] count=%d, ",
-					   engine->execlist_port[0].count);
-				print_request(m, rq, "rq: ");
-			} else {
-				seq_printf(m, "\t\tELSP[0] idle\n");
-			}
-			rq = READ_ONCE(engine->execlist_port[1].request);
-			if (rq) {
-				seq_printf(m, "\t\tELSP[1] count=%d, ",
-					   engine->execlist_port[1].count);
-				print_request(m, rq, "rq: ");
-			} else {
-				seq_printf(m, "\t\tELSP[1] idle\n");
+			for (idx = 0; idx < ARRAY_SIZE(engine->execlist_port); idx++) {
+				unsigned int count;
+
+				rq = port_unpack(&engine->execlist_port[idx],
+						 &count);
+				if (rq) {
+					seq_printf(m, "\t\tELSP[%d] count=%d, ",
+						   idx, count);
+					print_request(m, rq, "rq: ");
+				} else {
+					seq_printf(m, "\t\tELSP[%d] idle\n",
+						   idx);
+				}
 			}
 			rcu_read_unlock();
 
 			spin_lock_irq(&engine->timeline->lock);
-			for (rb = engine->execlist_first; rb; rb = rb_next(rb)) {
-				rq = rb_entry(rb, typeof(*rq), priotree.node);
-				print_request(m, rq, "\t\tQ ");
+			for (rb = engine->execlist_first; rb; rb = rb_next(rb)){
+				struct i915_priolist *p =
+					rb_entry(rb, typeof(*p), node);
+
+				list_for_each_entry(rq, &p->requests,
+						    priotree.link)
+					print_request(m, rq, "\t\tQ ");
 			}
 			spin_unlock_irq(&engine->timeline->lock);
 		} else if (INTEL_GEN(dev_priv) > 6) {
@@ -3704,16 +3740,10 @@ static ssize_t i915_displayport_test_active_write(struct file *file,
 	if (len == 0)
 		return 0;
 
-	input_buffer = kmalloc(len + 1, GFP_KERNEL);
-	if (!input_buffer)
-		return -ENOMEM;
-
-	if (copy_from_user(input_buffer, ubuf, len)) {
-		status = -EFAULT;
-		goto out;
-	}
+	input_buffer = memdup_user_nul(ubuf, len);
+	if (IS_ERR(input_buffer))
+		return PTR_ERR(input_buffer);
 
-	input_buffer[len] = '\0';
 	DRM_DEBUG_DRIVER("Copied %d bytes from user\n", (unsigned int)len);
 
 	drm_connector_list_iter_begin(dev, &conn_iter);
@@ -3739,7 +3769,6 @@ static ssize_t i915_displayport_test_active_write(struct file *file,
 		}
 	}
 	drm_connector_list_iter_end(&conn_iter);
-out:
 	kfree(input_buffer);
 	if (status < 0)
 		return status;
@@ -3900,6 +3929,8 @@ static void wm_latency_show(struct seq_file *m, const uint16_t wm[8])
 		num_levels = 3;
 	else if (IS_VALLEYVIEW(dev_priv))
 		num_levels = 1;
+	else if (IS_G4X(dev_priv))
+		num_levels = 3;
 	else
 		num_levels = ilk_wm_max_level(dev_priv) + 1;
 
@@ -3912,8 +3943,10 @@ static void wm_latency_show(struct seq_file *m, const uint16_t wm[8])
 		 * - WM1+ latency values in 0.5us units
 		 * - latencies are in us on gen9/vlv/chv
 		 */
-		if (INTEL_GEN(dev_priv) >= 9 || IS_VALLEYVIEW(dev_priv) ||
-		    IS_CHERRYVIEW(dev_priv))
+		if (INTEL_GEN(dev_priv) >= 9 ||
+		    IS_VALLEYVIEW(dev_priv) ||
+		    IS_CHERRYVIEW(dev_priv) ||
+		    IS_G4X(dev_priv))
 			latency *= 10;
 		else if (level > 0)
 			latency *= 5;
@@ -3974,7 +4007,7 @@ static int pri_wm_latency_open(struct inode *inode, struct file *file)
 {
 	struct drm_i915_private *dev_priv = inode->i_private;
 
-	if (INTEL_GEN(dev_priv) < 5)
+	if (INTEL_GEN(dev_priv) < 5 && !IS_G4X(dev_priv))
 		return -ENODEV;
 
 	return single_open(file, pri_wm_latency_show, dev_priv);
@@ -4016,6 +4049,8 @@ static ssize_t wm_latency_write(struct file *file, const char __user *ubuf,
 		num_levels = 3;
 	else if (IS_VALLEYVIEW(dev_priv))
 		num_levels = 1;
+	else if (IS_G4X(dev_priv))
+		num_levels = 3;
 	else
 		num_levels = ilk_wm_max_level(dev_priv) + 1;
 
@@ -4264,26 +4299,27 @@ i915_drop_caches_set(void *data, u64 val)
 {
 	struct drm_i915_private *dev_priv = data;
 	struct drm_device *dev = &dev_priv->drm;
-	int ret;
+	int ret = 0;
 
 	DRM_DEBUG("Dropping caches: 0x%08llx\n", val);
 
 	/* No need to check and wait for gpu resets, only libdrm auto-restarts
 	 * on ioctls on -EAGAIN. */
-	ret = mutex_lock_interruptible(&dev->struct_mutex);
-	if (ret)
-		return ret;
-
-	if (val & DROP_ACTIVE) {
-		ret = i915_gem_wait_for_idle(dev_priv,
-					     I915_WAIT_INTERRUPTIBLE |
-					     I915_WAIT_LOCKED);
+	if (val & (DROP_ACTIVE | DROP_RETIRE)) {
+		ret = mutex_lock_interruptible(&dev->struct_mutex);
 		if (ret)
-			goto unlock;
-	}
+			return ret;
+
+		if (val & DROP_ACTIVE)
+			ret = i915_gem_wait_for_idle(dev_priv,
+						     I915_WAIT_INTERRUPTIBLE |
+						     I915_WAIT_LOCKED);
 
-	if (val & DROP_RETIRE)
-		i915_gem_retire_requests(dev_priv);
+		if (val & DROP_RETIRE)
+			i915_gem_retire_requests(dev_priv);
+
+		mutex_unlock(&dev->struct_mutex);
+	}
 
 	lockdep_set_current_reclaim_state(GFP_KERNEL);
 	if (val & DROP_BOUND)
@@ -4296,9 +4332,6 @@ i915_drop_caches_set(void *data, u64 val)
 		i915_gem_shrink_all(dev_priv);
 	lockdep_clear_current_reclaim_state();
 
-unlock:
-	mutex_unlock(&dev->struct_mutex);
-
 	if (val & DROP_FREED) {
 		synchronize_rcu();
 		i915_gem_drain_freed_objects(dev_priv);
@@ -4776,6 +4809,8 @@ static const struct drm_info_list i915_debugfs_list[] = {
 	{"i915_guc_info", i915_guc_info, 0},
 	{"i915_guc_load_status", i915_guc_load_status_info, 0},
 	{"i915_guc_log_dump", i915_guc_log_dump, 0},
+	{"i915_guc_load_err_log_dump", i915_guc_log_dump, 0, (void *)1},
+	{"i915_guc_stage_pool", i915_guc_stage_pool, 0},
 	{"i915_huc_load_status", i915_huc_load_status_info, 0},
 	{"i915_frequency_info", i915_frequency_info, 0},
 	{"i915_hangcheck_info", i915_hangcheck_info, 0},
@@ -4834,7 +4869,7 @@ static const struct i915_debugfs_files {
 	{"i915_pri_wm_latency", &i915_pri_wm_latency_fops},
 	{"i915_spr_wm_latency", &i915_spr_wm_latency_fops},
 	{"i915_cur_wm_latency", &i915_cur_wm_latency_fops},
-	{"i915_fbc_false_color", &i915_fbc_fc_fops},
+	{"i915_fbc_false_color", &i915_fbc_false_color_fops},
 	{"i915_dp_test_data", &i915_displayport_test_data_fops},
 	{"i915_dp_test_type", &i915_displayport_test_type_fops},
 	{"i915_dp_test_active", &i915_displayport_test_active_fops},
diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index 48428672fc6e..1f802de7b94b 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -139,6 +139,9 @@ static enum intel_pch intel_virt_detect_pch(struct drm_i915_private *dev_priv)
 	} else if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) {
 		ret = PCH_SPT;
 		DRM_DEBUG_KMS("Assuming SunrisePoint PCH\n");
+	} else if (IS_COFFEELAKE(dev_priv) || IS_CANNONLAKE(dev_priv)) {
+		ret = PCH_CNP;
+		DRM_DEBUG_KMS("Assuming CannonPoint PCH\n");
 	}
 
 	return ret;
@@ -170,6 +173,9 @@ static void intel_detect_pch(struct drm_i915_private *dev_priv)
 	while ((pch = pci_get_class(PCI_CLASS_BRIDGE_ISA << 8, pch))) {
 		if (pch->vendor == PCI_VENDOR_ID_INTEL) {
 			unsigned short id = pch->device & INTEL_PCH_DEVICE_ID_MASK;
+			unsigned short id_ext = pch->device &
+				INTEL_PCH_DEVICE_ID_MASK_EXT;
+
 			dev_priv->pch_id = id;
 
 			if (id == INTEL_PCH_IBX_DEVICE_ID_TYPE) {
@@ -206,7 +212,7 @@ static void intel_detect_pch(struct drm_i915_private *dev_priv)
 				DRM_DEBUG_KMS("Found SunrisePoint PCH\n");
 				WARN_ON(!IS_SKYLAKE(dev_priv) &&
 					!IS_KABYLAKE(dev_priv));
-			} else if (id == INTEL_PCH_SPT_LP_DEVICE_ID_TYPE) {
+			} else if (id_ext == INTEL_PCH_SPT_LP_DEVICE_ID_TYPE) {
 				dev_priv->pch_type = PCH_SPT;
 				DRM_DEBUG_KMS("Found SunrisePoint LP PCH\n");
 				WARN_ON(!IS_SKYLAKE(dev_priv) &&
@@ -216,6 +222,16 @@ static void intel_detect_pch(struct drm_i915_private *dev_priv)
 				DRM_DEBUG_KMS("Found KabyPoint PCH\n");
 				WARN_ON(!IS_SKYLAKE(dev_priv) &&
 					!IS_KABYLAKE(dev_priv));
+			} else if (id == INTEL_PCH_CNP_DEVICE_ID_TYPE) {
+				dev_priv->pch_type = PCH_CNP;
+				DRM_DEBUG_KMS("Found CannonPoint PCH\n");
+				WARN_ON(!IS_CANNONLAKE(dev_priv) &&
+					!IS_COFFEELAKE(dev_priv));
+			} else if (id_ext == INTEL_PCH_CNP_LP_DEVICE_ID_TYPE) {
+				dev_priv->pch_type = PCH_CNP;
+				DRM_DEBUG_KMS("Found CannonPoint LP PCH\n");
+				WARN_ON(!IS_CANNONLAKE(dev_priv) &&
+					!IS_COFFEELAKE(dev_priv));
 			} else if ((id == INTEL_PCH_P2X_DEVICE_ID_TYPE) ||
 				   (id == INTEL_PCH_P3X_DEVICE_ID_TYPE) ||
 				   ((id == INTEL_PCH_QEMU_DEVICE_ID_TYPE) &&
@@ -350,6 +366,7 @@ static int i915_getparam(struct drm_device *dev, void *data,
 	case I915_PARAM_HAS_EXEC_SOFTPIN:
 	case I915_PARAM_HAS_EXEC_ASYNC:
 	case I915_PARAM_HAS_EXEC_FENCE:
+	case I915_PARAM_HAS_EXEC_CAPTURE:
 		/* For the time being all of these are always true;
 		 * if some supported hardware does not have one of these
 		 * features this value needs to be provided from
@@ -834,10 +851,6 @@ static int i915_driver_init_early(struct drm_i915_private *dev_priv,
 	intel_uc_init_early(dev_priv);
 	i915_memcpy_init_early(dev_priv);
 
-	ret = intel_engines_init_early(dev_priv);
-	if (ret)
-		return ret;
-
 	ret = i915_workqueues_init(dev_priv);
 	if (ret < 0)
 		goto err_engines;
@@ -855,7 +868,7 @@ static int i915_driver_init_early(struct drm_i915_private *dev_priv,
 	intel_init_audio_hooks(dev_priv);
 	ret = i915_gem_load_init(dev_priv);
 	if (ret < 0)
-		goto err_workqueues;
+		goto err_irq;
 
 	intel_display_crc_init(dev_priv);
 
@@ -867,7 +880,8 @@ static int i915_driver_init_early(struct drm_i915_private *dev_priv,
 
 	return 0;
 
-err_workqueues:
+err_irq:
+	intel_irq_fini(dev_priv);
 	i915_workqueues_cleanup(dev_priv);
 err_engines:
 	i915_engines_cleanup(dev_priv);
@@ -882,6 +896,7 @@ static void i915_driver_cleanup_early(struct drm_i915_private *dev_priv)
 {
 	i915_perf_fini(dev_priv);
 	i915_gem_load_cleanup(dev_priv);
+	intel_irq_fini(dev_priv);
 	i915_workqueues_cleanup(dev_priv);
 	i915_engines_cleanup(dev_priv);
 }
@@ -947,14 +962,21 @@ static int i915_driver_init_mmio(struct drm_i915_private *dev_priv)
 
 	ret = i915_mmio_setup(dev_priv);
 	if (ret < 0)
-		goto put_bridge;
+		goto err_bridge;
 
 	intel_uncore_init(dev_priv);
+
+	ret = intel_engines_init_mmio(dev_priv);
+	if (ret)
+		goto err_uncore;
+
 	i915_gem_init_mmio(dev_priv);
 
 	return 0;
 
-put_bridge:
+err_uncore:
+	intel_uncore_fini(dev_priv);
+err_bridge:
 	pci_dev_put(dev_priv->bridge_dev);
 
 	return ret;
@@ -991,6 +1013,8 @@ static void intel_sanitize_options(struct drm_i915_private *dev_priv)
 	DRM_DEBUG_DRIVER("use GPU semaphores? %s\n", yesno(i915.semaphores));
 
 	intel_uc_sanitize_options(dev_priv);
+
+	intel_gvt_sanitize_options(dev_priv);
 }
 
 /**
@@ -1213,9 +1237,8 @@ int i915_driver_load(struct pci_dev *pdev, const struct pci_device_id *ent)
 	struct drm_i915_private *dev_priv;
 	int ret;
 
-	/* Enable nuclear pageflip on ILK+, except vlv/chv */
-	if (!i915.nuclear_pageflip &&
-	    (match_info->gen < 5 || match_info->has_gmch_display))
+	/* Enable nuclear pageflip on ILK+ */
+	if (!i915.nuclear_pageflip && match_info->gen < 5)
 		driver.driver_features &= ~DRIVER_ATOMIC;
 
 	ret = -ENOMEM;
@@ -2454,9 +2477,6 @@ static int intel_runtime_resume(struct device *kdev)
 
 	intel_guc_resume(dev_priv);
 
-	if (IS_GEN6(dev_priv))
-		intel_init_pch_refclk(dev_priv);
-
 	if (IS_GEN9_LP(dev_priv)) {
 		bxt_disable_dc9(dev_priv);
 		bxt_display_core_init(dev_priv, true);
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 2c453a4e97d5..602fb3324484 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -55,6 +55,7 @@
 #include "i915_reg.h"
 #include "i915_utils.h"
 
+#include "intel_uncore.h"
 #include "intel_bios.h"
 #include "intel_dpll_mgr.h"
 #include "intel_uc.h"
@@ -79,8 +80,8 @@
 
 #define DRIVER_NAME		"i915"
 #define DRIVER_DESC		"Intel Graphics"
-#define DRIVER_DATE		"20170403"
-#define DRIVER_TIMESTAMP	1491198738
+#define DRIVER_DATE		"20170529"
+#define DRIVER_TIMESTAMP	1496041258
 
 /* Use I915_STATE_WARN(x) and I915_STATE_WARN_ON() (rather than WARN() and
  * WARN_ON()) for hw state sanity checks to check for unexpected conditions
@@ -114,6 +115,13 @@ typedef struct {
 	fp; \
 })
 
+static inline bool is_fixed16_zero(uint_fixed_16_16_t val)
+{
+	if (val.val == 0)
+		return true;
+	return false;
+}
+
 static inline uint_fixed_16_16_t u32_to_fixed_16_16(uint32_t val)
 {
 	uint_fixed_16_16_t fp;
@@ -152,8 +160,39 @@ static inline uint_fixed_16_16_t max_fixed_16_16(uint_fixed_16_16_t max1,
 	return max;
 }
 
-static inline uint_fixed_16_16_t fixed_16_16_div_round_up(uint32_t val,
-							  uint32_t d)
+static inline uint32_t div_round_up_fixed16(uint_fixed_16_16_t val,
+					    uint_fixed_16_16_t d)
+{
+	return DIV_ROUND_UP(val.val, d.val);
+}
+
+static inline uint32_t mul_round_up_u32_fixed16(uint32_t val,
+						uint_fixed_16_16_t mul)
+{
+	uint64_t intermediate_val;
+	uint32_t result;
+
+	intermediate_val = (uint64_t) val * mul.val;
+	intermediate_val = DIV_ROUND_UP_ULL(intermediate_val, 1 << 16);
+	WARN_ON(intermediate_val >> 32);
+	result = clamp_t(uint32_t, intermediate_val, 0, ~0);
+	return result;
+}
+
+static inline uint_fixed_16_16_t mul_fixed16(uint_fixed_16_16_t val,
+					     uint_fixed_16_16_t mul)
+{
+	uint64_t intermediate_val;
+	uint_fixed_16_16_t fp;
+
+	intermediate_val = (uint64_t) val.val * mul.val;
+	intermediate_val = intermediate_val >> 16;
+	WARN_ON(intermediate_val >> 32);
+	fp.val = clamp_t(uint32_t, intermediate_val, 0, ~0);
+	return fp;
+}
+
+static inline uint_fixed_16_16_t fixed_16_16_div(uint32_t val, uint32_t d)
 {
 	uint_fixed_16_16_t fp, res;
 
@@ -162,8 +201,7 @@ static inline uint_fixed_16_16_t fixed_16_16_div_round_up(uint32_t val,
 	return res;
 }
 
-static inline uint_fixed_16_16_t fixed_16_16_div_round_up_u64(uint32_t val,
-							      uint32_t d)
+static inline uint_fixed_16_16_t fixed_16_16_div_u64(uint32_t val, uint32_t d)
 {
 	uint_fixed_16_16_t res;
 	uint64_t interm_val;
@@ -176,6 +214,17 @@ static inline uint_fixed_16_16_t fixed_16_16_div_round_up_u64(uint32_t val,
 	return res;
 }
 
+static inline uint32_t div_round_up_u32_fixed16(uint32_t val,
+						uint_fixed_16_16_t d)
+{
+	uint64_t interm_val;
+
+	interm_val = (uint64_t)val << 16;
+	interm_val = DIV_ROUND_UP_ULL(interm_val, d.val);
+	WARN_ON(interm_val >> 32);
+	return clamp_t(uint32_t, interm_val, 0, ~0);
+}
+
 static inline uint_fixed_16_16_t mul_u32_fixed_16_16(uint32_t val,
 						     uint_fixed_16_16_t mul)
 {
@@ -677,116 +726,6 @@ struct drm_i915_display_funcs {
 	void (*load_luts)(struct drm_crtc_state *crtc_state);
 };
 
-enum forcewake_domain_id {
-	FW_DOMAIN_ID_RENDER = 0,
-	FW_DOMAIN_ID_BLITTER,
-	FW_DOMAIN_ID_MEDIA,
-
-	FW_DOMAIN_ID_COUNT
-};
-
-enum forcewake_domains {
-	FORCEWAKE_RENDER = BIT(FW_DOMAIN_ID_RENDER),
-	FORCEWAKE_BLITTER = BIT(FW_DOMAIN_ID_BLITTER),
-	FORCEWAKE_MEDIA	= BIT(FW_DOMAIN_ID_MEDIA),
-	FORCEWAKE_ALL = (FORCEWAKE_RENDER |
-			 FORCEWAKE_BLITTER |
-			 FORCEWAKE_MEDIA)
-};
-
-#define FW_REG_READ  (1)
-#define FW_REG_WRITE (2)
-
-enum decoupled_power_domain {
-	GEN9_DECOUPLED_PD_BLITTER = 0,
-	GEN9_DECOUPLED_PD_RENDER,
-	GEN9_DECOUPLED_PD_MEDIA,
-	GEN9_DECOUPLED_PD_ALL
-};
-
-enum decoupled_ops {
-	GEN9_DECOUPLED_OP_WRITE = 0,
-	GEN9_DECOUPLED_OP_READ
-};
-
-enum forcewake_domains
-intel_uncore_forcewake_for_reg(struct drm_i915_private *dev_priv,
-			       i915_reg_t reg, unsigned int op);
-
-struct intel_uncore_funcs {
-	void (*force_wake_get)(struct drm_i915_private *dev_priv,
-			       enum forcewake_domains domains);
-	void (*force_wake_put)(struct drm_i915_private *dev_priv,
-			       enum forcewake_domains domains);
-
-	uint8_t  (*mmio_readb)(struct drm_i915_private *dev_priv,
-			       i915_reg_t r, bool trace);
-	uint16_t (*mmio_readw)(struct drm_i915_private *dev_priv,
-			       i915_reg_t r, bool trace);
-	uint32_t (*mmio_readl)(struct drm_i915_private *dev_priv,
-			       i915_reg_t r, bool trace);
-	uint64_t (*mmio_readq)(struct drm_i915_private *dev_priv,
-			       i915_reg_t r, bool trace);
-
-	void (*mmio_writeb)(struct drm_i915_private *dev_priv,
-			    i915_reg_t r, uint8_t val, bool trace);
-	void (*mmio_writew)(struct drm_i915_private *dev_priv,
-			    i915_reg_t r, uint16_t val, bool trace);
-	void (*mmio_writel)(struct drm_i915_private *dev_priv,
-			    i915_reg_t r, uint32_t val, bool trace);
-};
-
-struct intel_forcewake_range {
-	u32 start;
-	u32 end;
-
-	enum forcewake_domains domains;
-};
-
-struct intel_uncore {
-	spinlock_t lock; /** lock is also taken in irq contexts. */
-
-	const struct intel_forcewake_range *fw_domains_table;
-	unsigned int fw_domains_table_entries;
-
-	struct notifier_block pmic_bus_access_nb;
-	struct intel_uncore_funcs funcs;
-
-	unsigned fifo_count;
-
-	enum forcewake_domains fw_domains;
-	enum forcewake_domains fw_domains_active;
-
-	u32 fw_set;
-	u32 fw_clear;
-	u32 fw_reset;
-
-	struct intel_uncore_forcewake_domain {
-		enum forcewake_domain_id id;
-		enum forcewake_domains mask;
-		unsigned wake_count;
-		struct hrtimer timer;
-		i915_reg_t reg_set;
-		i915_reg_t reg_ack;
-	} fw_domain[FW_DOMAIN_ID_COUNT];
-
-	int unclaimed_mmio_check;
-};
-
-#define __mask_next_bit(mask) ({					\
-	int __idx = ffs(mask) - 1;					\
-	mask &= ~BIT(__idx);						\
-	__idx;								\
-})
-
-/* Iterate over initialised fw domains */
-#define for_each_fw_domain_masked(domain__, mask__, dev_priv__, tmp__) \
-	for (tmp__ = (mask__); \
-	     tmp__ ? (domain__ = &(dev_priv__)->uncore.fw_domain[__mask_next_bit(tmp__)]), 1 : 0;)
-
-#define for_each_fw_domain(domain__, dev_priv__, tmp__) \
-	for_each_fw_domain_masked(domain__, (dev_priv__)->uncore.fw_domains, dev_priv__, tmp__)
-
 #define CSR_VERSION(major, minor)	((major) << 16 | (minor))
 #define CSR_VERSION_MAJOR(version)	((version) >> 16)
 #define CSR_VERSION_MINOR(version)	((version) & 0xffff)
@@ -813,7 +752,6 @@ struct intel_csr {
 	func(has_aliasing_ppgtt); \
 	func(has_csr); \
 	func(has_ddi); \
-	func(has_decoupled_mmio); \
 	func(has_dp_mst); \
 	func(has_fbc); \
 	func(has_fpga_dbg); \
@@ -822,8 +760,8 @@ struct intel_csr {
 	func(has_gmbus_irq); \
 	func(has_gmch_display); \
 	func(has_guc); \
+	func(has_guc_ct); \
 	func(has_hotplug); \
-	func(has_hw_contexts); \
 	func(has_l3_dpf); \
 	func(has_llc); \
 	func(has_logical_ring_contexts); \
@@ -888,6 +826,8 @@ enum intel_platform {
 	INTEL_BROXTON,
 	INTEL_KABYLAKE,
 	INTEL_GEMINILAKE,
+	INTEL_COFFEELAKE,
+	INTEL_CANNONLAKE,
 	INTEL_MAX_PLATFORMS
 };
 
@@ -1026,6 +966,9 @@ struct i915_gpu_state {
 			u32 *pages[0];
 		} *ringbuffer, *batchbuffer, *wa_batchbuffer, *ctx, *hws_page;
 
+		struct drm_i915_error_object **user_bo;
+		long user_bo_count;
+
 		struct drm_i915_error_object *wa_ctx;
 
 		struct drm_i915_error_request {
@@ -1210,6 +1153,7 @@ enum intel_pch {
 	PCH_LPT,	/* Lynxpoint PCH */
 	PCH_SPT,        /* Sunrisepoint PCH */
 	PCH_KBP,        /* Kabypoint PCH */
+	PCH_CNP,        /* Cannonpoint PCH */
 	PCH_NOP,
 };
 
@@ -1512,11 +1456,7 @@ struct i915_gem_mm {
 	/** LRU list of objects with fence regs on them. */
 	struct list_head fence_list;
 
-	/**
-	 * Are we in a non-interruptible section of code like
-	 * modesetting?
-	 */
-	bool interruptible;
+	u64 unordered_timeline;
 
 	/* the indicator for dispatch video commands on two BSD rings */
 	atomic_t bsd_engine_dispatch_index;
@@ -1567,7 +1507,7 @@ struct i915_gpu_error {
 	 *
 	 * This is a counter which gets incremented when reset is triggered,
 	 *
-	 * Before the reset commences, the I915_RESET_IN_PROGRESS bit is set
+	 * Before the reset commences, the I915_RESET_BACKOFF bit is set
 	 * meaning that any waiters holding onto the struct_mutex should
 	 * relinquish the lock immediately in order for the reset to start.
 	 *
@@ -1764,13 +1704,15 @@ struct ilk_wm_values {
 	enum intel_ddb_partitioning partitioning;
 };
 
-struct vlv_pipe_wm {
+struct g4x_pipe_wm {
 	uint16_t plane[I915_MAX_PLANES];
+	uint16_t fbc;
 };
 
-struct vlv_sr_wm {
+struct g4x_sr_wm {
 	uint16_t plane;
 	uint16_t cursor;
+	uint16_t fbc;
 };
 
 struct vlv_wm_ddl_values {
@@ -1778,13 +1720,22 @@ struct vlv_wm_ddl_values {
 };
 
 struct vlv_wm_values {
-	struct vlv_pipe_wm pipe[3];
-	struct vlv_sr_wm sr;
+	struct g4x_pipe_wm pipe[3];
+	struct g4x_sr_wm sr;
 	struct vlv_wm_ddl_values ddl[3];
 	uint8_t level;
 	bool cxsr;
 };
 
+struct g4x_wm_values {
+	struct g4x_pipe_wm pipe[2];
+	struct g4x_sr_wm sr;
+	struct g4x_sr_wm hpll;
+	bool cxsr;
+	bool hpll_en;
+	bool fbc_en;
+};
+
 struct skl_ddb_entry {
 	uint16_t start, end;	/* in number of blocks, 'end' is exclusive */
 };
@@ -2101,7 +2052,7 @@ struct i915_oa_ops {
 		    size_t *offset);
 
 	/**
-	 * @oa_buffer_is_empty: Check if OA buffer empty (false positives OK)
+	 * @oa_buffer_check: Check for OA buffer data + update tail
 	 *
 	 * This is either called via fops or the poll check hrtimer (atomic
 	 * ctx) without any locks taken.
@@ -2114,7 +2065,7 @@ struct i915_oa_ops {
 	 * here, which will be handled gracefully - likely resulting in an
 	 * %EAGAIN error for userspace.
 	 */
-	bool (*oa_buffer_is_empty)(struct drm_i915_private *dev_priv);
+	bool (*oa_buffer_check)(struct drm_i915_private *dev_priv);
 };
 
 struct intel_cdclk_state {
@@ -2128,6 +2079,7 @@ struct drm_i915_private {
 	struct kmem_cache *vmas;
 	struct kmem_cache *requests;
 	struct kmem_cache *dependencies;
+	struct kmem_cache *priorities;
 
 	const struct intel_device_info info;
 
@@ -2363,7 +2315,6 @@ struct drm_i915_private {
 	 */
 	struct mutex av_mutex;
 
-	uint32_t hw_context_size;
 	struct list_head context_list;
 
 	u32 fdi_rx_config;
@@ -2414,6 +2365,7 @@ struct drm_i915_private {
 			struct ilk_wm_values hw;
 			struct skl_wm_values skl_hw;
 			struct vlv_wm_values vlv;
+			struct g4x_wm_values g4x;
 		};
 
 		uint8_t max_level;
@@ -2455,11 +2407,14 @@ struct drm_i915_private {
 			wait_queue_head_t poll_wq;
 			bool pollin;
 
+			/**
+			 * For rate limiting any notifications of spurious
+			 * invalid OA reports
+			 */
+			struct ratelimit_state spurious_report_rs;
+
 			bool periodic;
 			int period_exponent;
-			int timestamp_frequency;
-
-			int tail_margin;
 
 			int metrics_set;
 
@@ -2473,6 +2428,70 @@ struct drm_i915_private {
 				u8 *vaddr;
 				int format;
 				int format_size;
+
+				/**
+				 * Locks reads and writes to all head/tail state
+				 *
+				 * Consider: the head and tail pointer state
+				 * needs to be read consistently from a hrtimer
+				 * callback (atomic context) and read() fop
+				 * (user context) with tail pointer updates
+				 * happening in atomic context and head updates
+				 * in user context and the (unlikely)
+				 * possibility of read() errors needing to
+				 * reset all head/tail state.
+				 *
+				 * Note: Contention or performance aren't
+				 * currently a significant concern here
+				 * considering the relatively low frequency of
+				 * hrtimer callbacks (5ms period) and that
+				 * reads typically only happen in response to a
+				 * hrtimer event and likely complete before the
+				 * next callback.
+				 *
+				 * Note: This lock is not held *while* reading
+				 * and copying data to userspace so the value
+				 * of head observed in htrimer callbacks won't
+				 * represent any partial consumption of data.
+				 */
+				spinlock_t ptr_lock;
+
+				/**
+				 * One 'aging' tail pointer and one 'aged'
+				 * tail pointer ready to used for reading.
+				 *
+				 * Initial values of 0xffffffff are invalid
+				 * and imply that an update is required
+				 * (and should be ignored by an attempted
+				 * read)
+				 */
+				struct {
+					u32 offset;
+				} tails[2];
+
+				/**
+				 * Index for the aged tail ready to read()
+				 * data up to.
+				 */
+				unsigned int aged_tail_idx;
+
+				/**
+				 * A monotonic timestamp for when the current
+				 * aging tail pointer was read; used to
+				 * determine when it is old enough to trust.
+				 */
+				u64 aging_timestamp;
+
+				/**
+				 * Although we can always read back the head
+				 * pointer register, we prefer to avoid
+				 * trusting the HW state, just to avoid any
+				 * risk that some hardware condition could
+				 * somehow bump the head pointer unpredictably
+				 * and cause us to forward the wrong OA buffer
+				 * data to userspace.
+				 */
+				u32 head;
 			} oa_buffer;
 
 			u32 gen7_latched_oastatus1;
@@ -2751,6 +2770,8 @@ intel_info(const struct drm_i915_private *dev_priv)
 #define IS_BROXTON(dev_priv)	((dev_priv)->info.platform == INTEL_BROXTON)
 #define IS_KABYLAKE(dev_priv)	((dev_priv)->info.platform == INTEL_KABYLAKE)
 #define IS_GEMINILAKE(dev_priv)	((dev_priv)->info.platform == INTEL_GEMINILAKE)
+#define IS_COFFEELAKE(dev_priv)	((dev_priv)->info.platform == INTEL_COFFEELAKE)
+#define IS_CANNONLAKE(dev_priv)	((dev_priv)->info.platform == INTEL_CANNONLAKE)
 #define IS_MOBILE(dev_priv)	((dev_priv)->info.is_mobile)
 #define IS_HSW_EARLY_SDV(dev_priv) (IS_HASWELL(dev_priv) && \
 				    (INTEL_DEVID(dev_priv) & 0xFF00) == 0x0C00)
@@ -2828,6 +2849,12 @@ intel_info(const struct drm_i915_private *dev_priv)
 #define IS_GLK_REVID(dev_priv, since, until) \
 	(IS_GEMINILAKE(dev_priv) && IS_REVID(dev_priv, since, until))
 
+#define CNL_REVID_A0		0x0
+#define CNL_REVID_B0		0x1
+
+#define IS_CNL_REVID(p, since, until) \
+	(IS_CANNONLAKE(p) && IS_REVID(p, since, until))
+
 /*
  * The genX designation typically refers to the render engine, so render
  * capability related checks should use IS_GEN, while display and other checks
@@ -2842,6 +2869,7 @@ intel_info(const struct drm_i915_private *dev_priv)
 #define IS_GEN7(dev_priv)	(!!((dev_priv)->info.gen_mask & BIT(6)))
 #define IS_GEN8(dev_priv)	(!!((dev_priv)->info.gen_mask & BIT(7)))
 #define IS_GEN9(dev_priv)	(!!((dev_priv)->info.gen_mask & BIT(8)))
+#define IS_GEN10(dev_priv)	(!!((dev_priv)->info.gen_mask & BIT(9)))
 
 #define IS_LP(dev_priv)	(INTEL_INFO(dev_priv)->is_lp)
 #define IS_GEN9_LP(dev_priv)	(IS_GEN9(dev_priv) && IS_LP(dev_priv))
@@ -2871,7 +2899,6 @@ intel_info(const struct drm_i915_private *dev_priv)
 
 #define HWS_NEEDS_PHYSICAL(dev_priv)	((dev_priv)->info.hws_needs_physical)
 
-#define HAS_HW_CONTEXTS(dev_priv)	    ((dev_priv)->info.has_hw_contexts)
 #define HAS_LOGICAL_RING_CONTEXTS(dev_priv) \
 		((dev_priv)->info.has_logical_ring_contexts)
 #define USES_PPGTT(dev_priv)		(i915.enable_ppgtt)
@@ -2910,6 +2937,7 @@ intel_info(const struct drm_i915_private *dev_priv)
 #define HAS_FW_BLC(dev_priv) 	(INTEL_GEN(dev_priv) > 2)
 #define HAS_PIPE_CXSR(dev_priv) ((dev_priv)->info.has_pipe_cxsr)
 #define HAS_FBC(dev_priv)	((dev_priv)->info.has_fbc)
+#define HAS_CUR_FBC(dev_priv)	(!HAS_GMCH_DISPLAY(dev_priv) && INTEL_INFO(dev_priv)->gen >= 7)
 
 #define HAS_IPS(dev_priv)	(IS_HSW_ULT(dev_priv) || IS_BROADWELL(dev_priv))
 
@@ -2932,6 +2960,7 @@ intel_info(const struct drm_i915_private *dev_priv)
  * properties, so we have separate macros to test them.
  */
 #define HAS_GUC(dev_priv)	((dev_priv)->info.has_guc)
+#define HAS_GUC_CT(dev_priv)	((dev_priv)->info.has_guc_ct)
 #define HAS_GUC_UCODE(dev_priv)	(HAS_GUC(dev_priv))
 #define HAS_GUC_SCHED(dev_priv)	(HAS_GUC(dev_priv))
 #define HAS_HUC_UCODE(dev_priv)	(HAS_GUC(dev_priv))
@@ -2941,6 +2970,7 @@ intel_info(const struct drm_i915_private *dev_priv)
 #define HAS_POOLED_EU(dev_priv)	((dev_priv)->info.has_pooled_eu)
 
 #define INTEL_PCH_DEVICE_ID_MASK		0xff00
+#define INTEL_PCH_DEVICE_ID_MASK_EXT		0xff80
 #define INTEL_PCH_IBX_DEVICE_ID_TYPE		0x3b00
 #define INTEL_PCH_CPT_DEVICE_ID_TYPE		0x1c00
 #define INTEL_PCH_PPT_DEVICE_ID_TYPE		0x1e00
@@ -2949,11 +2979,16 @@ intel_info(const struct drm_i915_private *dev_priv)
 #define INTEL_PCH_SPT_DEVICE_ID_TYPE		0xA100
 #define INTEL_PCH_SPT_LP_DEVICE_ID_TYPE		0x9D00
 #define INTEL_PCH_KBP_DEVICE_ID_TYPE		0xA200
+#define INTEL_PCH_CNP_DEVICE_ID_TYPE		0xA300
+#define INTEL_PCH_CNP_LP_DEVICE_ID_TYPE		0x9D80
 #define INTEL_PCH_P2X_DEVICE_ID_TYPE		0x7100
 #define INTEL_PCH_P3X_DEVICE_ID_TYPE		0x7000
 #define INTEL_PCH_QEMU_DEVICE_ID_TYPE		0x2900 /* qemu q35 has 2918 */
 
 #define INTEL_PCH_TYPE(dev_priv) ((dev_priv)->pch_type)
+#define HAS_PCH_CNP(dev_priv) (INTEL_PCH_TYPE(dev_priv) == PCH_CNP)
+#define HAS_PCH_CNP_LP(dev_priv) \
+	((dev_priv)->pch_id == INTEL_PCH_CNP_LP_DEVICE_ID_TYPE)
 #define HAS_PCH_KBP(dev_priv) (INTEL_PCH_TYPE(dev_priv) == PCH_KBP)
 #define HAS_PCH_SPT(dev_priv) (INTEL_PCH_TYPE(dev_priv) == PCH_SPT)
 #define HAS_PCH_LPT(dev_priv) (INTEL_PCH_TYPE(dev_priv) == PCH_LPT)
@@ -2978,27 +3013,26 @@ intel_info(const struct drm_i915_private *dev_priv)
 #define GT_FREQUENCY_MULTIPLIER 50
 #define GEN9_FREQ_SCALER 3
 
-#define HAS_DECOUPLED_MMIO(dev_priv) (INTEL_INFO(dev_priv)->has_decoupled_mmio)
-
 #include "i915_trace.h"
 
-static inline bool intel_scanout_needs_vtd_wa(struct drm_i915_private *dev_priv)
+static inline bool intel_vtd_active(void)
 {
 #ifdef CONFIG_INTEL_IOMMU
-	if (INTEL_GEN(dev_priv) >= 6 && intel_iommu_gfx_mapped)
+	if (intel_iommu_gfx_mapped)
 		return true;
 #endif
 	return false;
 }
 
+static inline bool intel_scanout_needs_vtd_wa(struct drm_i915_private *dev_priv)
+{
+	return INTEL_GEN(dev_priv) >= 6 && intel_vtd_active();
+}
+
 static inline bool
 intel_ggtt_update_needs_vtd_wa(struct drm_i915_private *dev_priv)
 {
-#ifdef CONFIG_INTEL_IOMMU
-	if (IS_BROXTON(dev_priv) && intel_iommu_gfx_mapped)
-		return true;
-#endif
-	return false;
+	return IS_BROXTON(dev_priv) && intel_vtd_active();
 }
 
 int intel_sanitize_enable_ppgtt(struct drm_i915_private *dev_priv,
@@ -3037,7 +3071,7 @@ extern unsigned long i915_gfx_val(struct drm_i915_private *dev_priv);
 extern void i915_update_gfx_val(struct drm_i915_private *dev_priv);
 int vlv_force_gfx_clock(struct drm_i915_private *dev_priv, bool on);
 
-int intel_engines_init_early(struct drm_i915_private *dev_priv);
+int intel_engines_init_mmio(struct drm_i915_private *dev_priv);
 int intel_engines_init(struct drm_i915_private *dev_priv);
 
 /* intel_hotplug.c */
@@ -3074,43 +3108,10 @@ void i915_handle_error(struct drm_i915_private *dev_priv,
 		       const char *fmt, ...);
 
 extern void intel_irq_init(struct drm_i915_private *dev_priv);
+extern void intel_irq_fini(struct drm_i915_private *dev_priv);
 int intel_irq_install(struct drm_i915_private *dev_priv);
 void intel_irq_uninstall(struct drm_i915_private *dev_priv);
 
-extern void intel_uncore_sanitize(struct drm_i915_private *dev_priv);
-extern void intel_uncore_init(struct drm_i915_private *dev_priv);
-extern bool intel_uncore_unclaimed_mmio(struct drm_i915_private *dev_priv);
-extern bool intel_uncore_arm_unclaimed_mmio_detection(struct drm_i915_private *dev_priv);
-extern void intel_uncore_fini(struct drm_i915_private *dev_priv);
-extern void intel_uncore_suspend(struct drm_i915_private *dev_priv);
-extern void intel_uncore_resume_early(struct drm_i915_private *dev_priv);
-const char *intel_uncore_forcewake_domain_to_str(const enum forcewake_domain_id id);
-void intel_uncore_forcewake_get(struct drm_i915_private *dev_priv,
-				enum forcewake_domains domains);
-void intel_uncore_forcewake_put(struct drm_i915_private *dev_priv,
-				enum forcewake_domains domains);
-/* Like above but the caller must manage the uncore.lock itself.
- * Must be used with I915_READ_FW and friends.
- */
-void intel_uncore_forcewake_get__locked(struct drm_i915_private *dev_priv,
-					enum forcewake_domains domains);
-void intel_uncore_forcewake_put__locked(struct drm_i915_private *dev_priv,
-					enum forcewake_domains domains);
-u64 intel_uncore_edram_size(struct drm_i915_private *dev_priv);
-
-void assert_forcewakes_inactive(struct drm_i915_private *dev_priv);
-
-int intel_wait_for_register(struct drm_i915_private *dev_priv,
-			    i915_reg_t reg,
-			    const u32 mask,
-			    const u32 value,
-			    const unsigned long timeout_ms);
-int intel_wait_for_register_fw(struct drm_i915_private *dev_priv,
-			       i915_reg_t reg,
-			       const u32 mask,
-			       const u32 value,
-			       const unsigned long timeout_ms);
-
 static inline bool intel_gvt_active(struct drm_i915_private *dev_priv)
 {
 	return dev_priv->gvt;
@@ -3458,8 +3459,9 @@ int i915_gem_object_wait_priority(struct drm_i915_gem_object *obj,
 #define I915_PRIORITY_DISPLAY I915_PRIORITY_MAX
 
 int __must_check
-i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj,
-				  bool write);
+i915_gem_object_set_to_wc_domain(struct drm_i915_gem_object *obj, bool write);
+int __must_check
+i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, bool write);
 int __must_check
 i915_gem_object_set_to_cpu_domain(struct drm_i915_gem_object *obj, bool write);
 struct i915_vma * __must_check
@@ -3722,8 +3724,8 @@ int  intel_lpe_audio_init(struct drm_i915_private *dev_priv);
 void intel_lpe_audio_teardown(struct drm_i915_private *dev_priv);
 void intel_lpe_audio_irq_handler(struct drm_i915_private *dev_priv);
 void intel_lpe_audio_notify(struct drm_i915_private *dev_priv,
-			    void *eld, int port, int pipe, int tmds_clk_speed,
-			    bool dp_output, int link_rate);
+			    enum pipe pipe, enum port port,
+			    const void *eld, int ls_clock, bool dp_output);
 
 /* intel_i2c.c */
 extern int intel_setup_gmbus(struct drm_i915_private *dev_priv);
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 462031cbd77f..aff449807399 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -46,8 +46,6 @@
 #include <linux/dma-buf.h>
 
 static void i915_gem_flush_free_objects(struct drm_i915_private *i915);
-static void i915_gem_object_flush_gtt_write_domain(struct drm_i915_gem_object *obj);
-static void i915_gem_object_flush_cpu_write_domain(struct drm_i915_gem_object *obj);
 
 static bool cpu_write_needs_clflush(struct drm_i915_gem_object *obj)
 {
@@ -145,9 +143,9 @@ i915_gem_get_aperture_ioctl(struct drm_device *dev, void *data,
 	struct i915_ggtt *ggtt = &dev_priv->ggtt;
 	struct drm_i915_gem_get_aperture *args = data;
 	struct i915_vma *vma;
-	size_t pinned;
+	u64 pinned;
 
-	pinned = 0;
+	pinned = ggtt->base.reserved;
 	mutex_lock(&dev->struct_mutex);
 	list_for_each_entry(vma, &ggtt->base.active_list, vm_link)
 		if (i915_vma_is_pinned(vma))
@@ -705,6 +703,61 @@ i915_gem_create_ioctl(struct drm_device *dev, void *data,
 			       args->size, &args->handle);
 }
 
+static inline enum fb_op_origin
+fb_write_origin(struct drm_i915_gem_object *obj, unsigned int domain)
+{
+	return (domain == I915_GEM_DOMAIN_GTT ?
+		obj->frontbuffer_ggtt_origin : ORIGIN_CPU);
+}
+
+static void
+flush_write_domain(struct drm_i915_gem_object *obj, unsigned int flush_domains)
+{
+	struct drm_i915_private *dev_priv = to_i915(obj->base.dev);
+
+	if (!(obj->base.write_domain & flush_domains))
+		return;
+
+	/* No actual flushing is required for the GTT write domain.  Writes
+	 * to it "immediately" go to main memory as far as we know, so there's
+	 * no chipset flush.  It also doesn't land in render cache.
+	 *
+	 * However, we do have to enforce the order so that all writes through
+	 * the GTT land before any writes to the device, such as updates to
+	 * the GATT itself.
+	 *
+	 * We also have to wait a bit for the writes to land from the GTT.
+	 * An uncached read (i.e. mmio) seems to be ideal for the round-trip
+	 * timing. This issue has only been observed when switching quickly
+	 * between GTT writes and CPU reads from inside the kernel on recent hw,
+	 * and it appears to only affect discrete GTT blocks (i.e. on LLC
+	 * system agents we cannot reproduce this behaviour).
+	 */
+	wmb();
+
+	switch (obj->base.write_domain) {
+	case I915_GEM_DOMAIN_GTT:
+		if (INTEL_GEN(dev_priv) >= 6 && !HAS_LLC(dev_priv)) {
+			if (intel_runtime_pm_get_if_in_use(dev_priv)) {
+				spin_lock_irq(&dev_priv->uncore.lock);
+				POSTING_READ_FW(RING_ACTHD(dev_priv->engine[RCS]->mmio_base));
+				spin_unlock_irq(&dev_priv->uncore.lock);
+				intel_runtime_pm_put(dev_priv);
+			}
+		}
+
+		intel_fb_obj_flush(obj,
+				   fb_write_origin(obj, I915_GEM_DOMAIN_GTT));
+		break;
+
+	case I915_GEM_DOMAIN_CPU:
+		i915_gem_clflush_object(obj, I915_CLFLUSH_SYNC);
+		break;
+	}
+
+	obj->base.write_domain = 0;
+}
+
 static inline int
 __copy_to_user_swizzled(char __user *cpu_vaddr,
 			const char *gpu_vaddr, int gpu_offset,
@@ -794,7 +847,7 @@ int i915_gem_obj_prepare_shmem_read(struct drm_i915_gem_object *obj,
 			goto out;
 	}
 
-	i915_gem_object_flush_gtt_write_domain(obj);
+	flush_write_domain(obj, ~I915_GEM_DOMAIN_CPU);
 
 	/* If we're not in the cpu read domain, set ourself into the gtt
 	 * read domain and manually flush cachelines (if required). This
@@ -846,7 +899,7 @@ int i915_gem_obj_prepare_shmem_write(struct drm_i915_gem_object *obj,
 			goto out;
 	}
 
-	i915_gem_object_flush_gtt_write_domain(obj);
+	flush_write_domain(obj, ~I915_GEM_DOMAIN_CPU);
 
 	/* If we're not in the cpu write domain, set ourself into the
 	 * gtt write domain and manually flush cachelines (as required).
@@ -1501,13 +1554,6 @@ err:
 	return ret;
 }
 
-static inline enum fb_op_origin
-write_origin(struct drm_i915_gem_object *obj, unsigned domain)
-{
-	return (domain == I915_GEM_DOMAIN_GTT ?
-		obj->frontbuffer_ggtt_origin : ORIGIN_CPU);
-}
-
 static void i915_gem_object_bump_inactive_ggtt(struct drm_i915_gem_object *obj)
 {
 	struct drm_i915_private *i915;
@@ -1591,10 +1637,12 @@ i915_gem_set_domain_ioctl(struct drm_device *dev, void *data,
 	if (err)
 		goto out_unpin;
 
-	if (read_domains & I915_GEM_DOMAIN_GTT)
-		err = i915_gem_object_set_to_gtt_domain(obj, write_domain != 0);
+	if (read_domains & I915_GEM_DOMAIN_WC)
+		err = i915_gem_object_set_to_wc_domain(obj, write_domain);
+	else if (read_domains & I915_GEM_DOMAIN_GTT)
+		err = i915_gem_object_set_to_gtt_domain(obj, write_domain);
 	else
-		err = i915_gem_object_set_to_cpu_domain(obj, write_domain != 0);
+		err = i915_gem_object_set_to_cpu_domain(obj, write_domain);
 
 	/* And bump the LRU for this access */
 	i915_gem_object_bump_inactive_ggtt(obj);
@@ -1602,7 +1650,8 @@ i915_gem_set_domain_ioctl(struct drm_device *dev, void *data,
 	mutex_unlock(&dev->struct_mutex);
 
 	if (write_domain != 0)
-		intel_fb_obj_invalidate(obj, write_origin(obj, write_domain));
+		intel_fb_obj_invalidate(obj,
+					fb_write_origin(obj, write_domain));
 
 out_unpin:
 	i915_gem_object_unpin_pages(obj);
@@ -1737,6 +1786,9 @@ static unsigned int tile_row_pages(struct drm_i915_gem_object *obj)
  *     into userspace. (This view is aligned and sized appropriately for
  *     fenced access.)
  *
+ * 2 - Recognise WC as a separate cache domain so that we can flush the
+ *     delayed writes via GTT before performing direct access via WC.
+ *
  * Restrictions:
  *
  *  * snoopable objects cannot be accessed via the GTT. It can cause machine
@@ -1764,7 +1816,7 @@ static unsigned int tile_row_pages(struct drm_i915_gem_object *obj)
  */
 int i915_gem_mmap_gtt_version(void)
 {
-	return 1;
+	return 2;
 }
 
 static inline struct i915_ggtt_view
@@ -2228,7 +2280,7 @@ void __i915_gem_object_put_pages(struct drm_i915_gem_object *obj,
 	if (obj->mm.mapping) {
 		void *ptr;
 
-		ptr = ptr_mask_bits(obj->mm.mapping);
+		ptr = page_mask_bits(obj->mm.mapping);
 		if (is_vmalloc_addr(ptr))
 			vunmap(ptr);
 		else
@@ -2504,7 +2556,7 @@ static void *i915_gem_object_map(const struct drm_i915_gem_object *obj,
 
 	if (n_pages > ARRAY_SIZE(stack_pages)) {
 		/* Too big for stack -- allocate temporary array instead */
-		pages = drm_malloc_gfp(n_pages, sizeof(*pages), GFP_TEMPORARY);
+		pages = kvmalloc_array(n_pages, sizeof(*pages), GFP_TEMPORARY);
 		if (!pages)
 			return NULL;
 	}
@@ -2526,7 +2578,7 @@ static void *i915_gem_object_map(const struct drm_i915_gem_object *obj,
 	addr = vmap(pages, n_pages, 0, pgprot);
 
 	if (pages != stack_pages)
-		drm_free_large(pages);
+		kvfree(pages);
 
 	return addr;
 }
@@ -2560,7 +2612,7 @@ void *i915_gem_object_pin_map(struct drm_i915_gem_object *obj,
 	}
 	GEM_BUG_ON(!obj->mm.pages);
 
-	ptr = ptr_unpack_bits(obj->mm.mapping, has_type);
+	ptr = page_unpack_bits(obj->mm.mapping, &has_type);
 	if (ptr && has_type != type) {
 		if (pinned) {
 			ret = -EBUSY;
@@ -2582,7 +2634,7 @@ void *i915_gem_object_pin_map(struct drm_i915_gem_object *obj,
 			goto err_unpin;
 		}
 
-		obj->mm.mapping = ptr_pack_bits(ptr, type);
+		obj->mm.mapping = page_pack_bits(ptr, type);
 	}
 
 out_unlock:
@@ -2967,12 +3019,14 @@ static void engine_set_wedged(struct intel_engine_cs *engine)
 	 */
 
 	if (i915.enable_execlists) {
+		struct execlist_port *port = engine->execlist_port;
 		unsigned long flags;
+		unsigned int n;
 
 		spin_lock_irqsave(&engine->timeline->lock, flags);
 
-		i915_gem_request_put(engine->execlist_port[0].request);
-		i915_gem_request_put(engine->execlist_port[1].request);
+		for (n = 0; n < ARRAY_SIZE(engine->execlist_port); n++)
+			i915_gem_request_put(port_request(&port[n]));
 		memset(engine->execlist_port, 0, sizeof(engine->execlist_port));
 		engine->execlist_queue = RB_ROOT;
 		engine->execlist_first = NULL;
@@ -3101,8 +3155,6 @@ i915_gem_idle_work_handler(struct work_struct *work)
 	struct drm_i915_private *dev_priv =
 		container_of(work, typeof(*dev_priv), gt.idle_work.work);
 	struct drm_device *dev = &dev_priv->drm;
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
 	bool rearm_hangcheck;
 
 	if (!READ_ONCE(dev_priv->gt.awake))
@@ -3140,10 +3192,8 @@ i915_gem_idle_work_handler(struct work_struct *work)
 	if (wait_for(intel_engines_are_idle(dev_priv), 10))
 		DRM_ERROR("Timeout waiting for engines to idle\n");
 
-	for_each_engine(engine, dev_priv, id) {
-		intel_engine_disarm_breadcrumbs(engine);
-		i915_gem_batch_pool_fini(&engine->batch_pool);
-	}
+	intel_engines_mark_idle(dev_priv);
+	i915_gem_timelines_mark_idle(dev_priv);
 
 	GEM_BUG_ON(!dev_priv->gt.awake);
 	dev_priv->gt.awake = false;
@@ -3324,56 +3374,6 @@ int i915_gem_wait_for_idle(struct drm_i915_private *i915, unsigned int flags)
 	return ret;
 }
 
-/** Flushes the GTT write domain for the object if it's dirty. */
-static void
-i915_gem_object_flush_gtt_write_domain(struct drm_i915_gem_object *obj)
-{
-	struct drm_i915_private *dev_priv = to_i915(obj->base.dev);
-
-	if (obj->base.write_domain != I915_GEM_DOMAIN_GTT)
-		return;
-
-	/* No actual flushing is required for the GTT write domain.  Writes
-	 * to it "immediately" go to main memory as far as we know, so there's
-	 * no chipset flush.  It also doesn't land in render cache.
-	 *
-	 * However, we do have to enforce the order so that all writes through
-	 * the GTT land before any writes to the device, such as updates to
-	 * the GATT itself.
-	 *
-	 * We also have to wait a bit for the writes to land from the GTT.
-	 * An uncached read (i.e. mmio) seems to be ideal for the round-trip
-	 * timing. This issue has only been observed when switching quickly
-	 * between GTT writes and CPU reads from inside the kernel on recent hw,
-	 * and it appears to only affect discrete GTT blocks (i.e. on LLC
-	 * system agents we cannot reproduce this behaviour).
-	 */
-	wmb();
-	if (INTEL_GEN(dev_priv) >= 6 && !HAS_LLC(dev_priv)) {
-		if (intel_runtime_pm_get_if_in_use(dev_priv)) {
-			spin_lock_irq(&dev_priv->uncore.lock);
-			POSTING_READ_FW(RING_ACTHD(dev_priv->engine[RCS]->mmio_base));
-			spin_unlock_irq(&dev_priv->uncore.lock);
-			intel_runtime_pm_put(dev_priv);
-		}
-	}
-
-	intel_fb_obj_flush(obj, write_origin(obj, I915_GEM_DOMAIN_GTT));
-
-	obj->base.write_domain = 0;
-}
-
-/** Flushes the CPU write domain for the object if it's dirty. */
-static void
-i915_gem_object_flush_cpu_write_domain(struct drm_i915_gem_object *obj)
-{
-	if (obj->base.write_domain != I915_GEM_DOMAIN_CPU)
-		return;
-
-	i915_gem_clflush_object(obj, I915_CLFLUSH_SYNC);
-	obj->base.write_domain = 0;
-}
-
 static void __i915_gem_object_flush_for_display(struct drm_i915_gem_object *obj)
 {
 	if (obj->base.write_domain != I915_GEM_DOMAIN_CPU && !obj->cache_dirty)
@@ -3394,6 +3394,69 @@ void i915_gem_object_flush_if_display(struct drm_i915_gem_object *obj)
 }
 
 /**
+ * Moves a single object to the WC read, and possibly write domain.
+ * @obj: object to act on
+ * @write: ask for write access or read only
+ *
+ * This function returns when the move is complete, including waiting on
+ * flushes to occur.
+ */
+int
+i915_gem_object_set_to_wc_domain(struct drm_i915_gem_object *obj, bool write)
+{
+	int ret;
+
+	lockdep_assert_held(&obj->base.dev->struct_mutex);
+
+	ret = i915_gem_object_wait(obj,
+				   I915_WAIT_INTERRUPTIBLE |
+				   I915_WAIT_LOCKED |
+				   (write ? I915_WAIT_ALL : 0),
+				   MAX_SCHEDULE_TIMEOUT,
+				   NULL);
+	if (ret)
+		return ret;
+
+	if (obj->base.write_domain == I915_GEM_DOMAIN_WC)
+		return 0;
+
+	/* Flush and acquire obj->pages so that we are coherent through
+	 * direct access in memory with previous cached writes through
+	 * shmemfs and that our cache domain tracking remains valid.
+	 * For example, if the obj->filp was moved to swap without us
+	 * being notified and releasing the pages, we would mistakenly
+	 * continue to assume that the obj remained out of the CPU cached
+	 * domain.
+	 */
+	ret = i915_gem_object_pin_pages(obj);
+	if (ret)
+		return ret;
+
+	flush_write_domain(obj, ~I915_GEM_DOMAIN_WC);
+
+	/* Serialise direct access to this object with the barriers for
+	 * coherent writes from the GPU, by effectively invalidating the
+	 * WC domain upon first access.
+	 */
+	if ((obj->base.read_domains & I915_GEM_DOMAIN_WC) == 0)
+		mb();
+
+	/* It should now be out of any other write domains, and we can update
+	 * the domain values for our changes.
+	 */
+	GEM_BUG_ON((obj->base.write_domain & ~I915_GEM_DOMAIN_WC) != 0);
+	obj->base.read_domains |= I915_GEM_DOMAIN_WC;
+	if (write) {
+		obj->base.read_domains = I915_GEM_DOMAIN_WC;
+		obj->base.write_domain = I915_GEM_DOMAIN_WC;
+		obj->mm.dirty = true;
+	}
+
+	i915_gem_object_unpin_pages(obj);
+	return 0;
+}
+
+/**
  * Moves a single object to the GTT read, and possibly write domain.
  * @obj: object to act on
  * @write: ask for write access or read only
@@ -3432,7 +3495,7 @@ i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, bool write)
 	if (ret)
 		return ret;
 
-	i915_gem_object_flush_cpu_write_domain(obj);
+	flush_write_domain(obj, ~I915_GEM_DOMAIN_GTT);
 
 	/* Serialise direct access to this object with the barriers for
 	 * coherent writes from the GPU, by effectively invalidating the
@@ -3806,7 +3869,7 @@ i915_gem_object_set_to_cpu_domain(struct drm_i915_gem_object *obj, bool write)
 	if (obj->base.write_domain == I915_GEM_DOMAIN_CPU)
 		return 0;
 
-	i915_gem_object_flush_gtt_write_domain(obj);
+	flush_write_domain(obj, ~I915_GEM_DOMAIN_CPU);
 
 	/* Flush the CPU cache if it's still invalid. */
 	if ((obj->base.read_domains & I915_GEM_DOMAIN_CPU) == 0) {
@@ -4000,7 +4063,7 @@ __busy_set_if_active(const struct dma_fence *fence,
 	if (i915_gem_request_completed(rq))
 		return 0;
 
-	return flag(rq->engine->exec_id);
+	return flag(rq->engine->uabi_id);
 }
 
 static __always_inline unsigned int
@@ -4199,7 +4262,7 @@ i915_gem_object_create(struct drm_i915_private *dev_priv, u64 size)
 	 * catch if we ever need to fix it. In the meantime, if you do spot
 	 * such a local variable, please consider fixing!
 	 */
-	if (WARN_ON(size >> PAGE_SHIFT > INT_MAX))
+	if (size >> PAGE_SHIFT > INT_MAX)
 		return ERR_PTR(-E2BIG);
 
 	if (overflows_type(size, obj->base.size))
@@ -4306,6 +4369,8 @@ static void __i915_gem_free_objects(struct drm_i915_private *i915,
 	intel_runtime_pm_put(i915);
 	mutex_unlock(&i915->drm.struct_mutex);
 
+	cond_resched();
+
 	llist_for_each_entry_safe(obj, on, freed, freed) {
 		GEM_BUG_ON(obj->bind_count);
 		GEM_BUG_ON(atomic_read(&obj->frontbuffer_bits));
@@ -4353,8 +4418,11 @@ static void __i915_gem_free_work(struct work_struct *work)
 	 * unbound now.
 	 */
 
-	while ((freed = llist_del_all(&i915->mm.free_list)))
+	while ((freed = llist_del_all(&i915->mm.free_list))) {
 		__i915_gem_free_objects(i915, freed);
+		if (need_resched())
+			break;
+	}
 }
 
 static void __i915_gem_free_object_rcu(struct rcu_head *head)
@@ -4419,10 +4487,9 @@ void i915_gem_sanitize(struct drm_i915_private *i915)
 	 * try to take over. The only way to remove the earlier state
 	 * is by resetting. However, resetting on earlier gen is tricky as
 	 * it may impact the display and we are uncertain about the stability
-	 * of the reset, so we only reset recent machines with logical
-	 * context support (that must be reset to remove any stray contexts).
+	 * of the reset, so this could be applied to even earlier gen.
 	 */
-	if (HAS_HW_CONTEXTS(i915)) {
+	if (INTEL_GEN(i915) >= 5) {
 		int reset = intel_gpu_reset(i915, ALL_ENGINES);
 		WARN_ON(reset && reset != -ENODEV);
 	}
@@ -4665,11 +4732,9 @@ bool intel_sanitize_semaphores(struct drm_i915_private *dev_priv, int value)
 	if (value >= 0)
 		return value;
 
-#ifdef CONFIG_INTEL_IOMMU
 	/* Enable semaphores on SNB when IO remapping is off */
-	if (INTEL_INFO(dev_priv)->gen == 6 && intel_iommu_gfx_mapped)
+	if (IS_GEN6(dev_priv) && intel_vtd_active())
 		return false;
-#endif
 
 	return true;
 }
@@ -4680,7 +4745,7 @@ int i915_gem_init(struct drm_i915_private *dev_priv)
 
 	mutex_lock(&dev_priv->drm.struct_mutex);
 
-	i915_gem_clflush_init(dev_priv);
+	dev_priv->mm.unordered_timeline = dma_fence_context_alloc(1);
 
 	if (!i915.enable_execlists) {
 		dev_priv->gt.resume = intel_legacy_submission_resume;
@@ -4803,12 +4868,16 @@ i915_gem_load_init(struct drm_i915_private *dev_priv)
 	if (!dev_priv->dependencies)
 		goto err_requests;
 
+	dev_priv->priorities = KMEM_CACHE(i915_priolist, SLAB_HWCACHE_ALIGN);
+	if (!dev_priv->priorities)
+		goto err_dependencies;
+
 	mutex_lock(&dev_priv->drm.struct_mutex);
 	INIT_LIST_HEAD(&dev_priv->gt.timelines);
 	err = i915_gem_timeline_init__global(dev_priv);
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 	if (err)
-		goto err_dependencies;
+		goto err_priorities;
 
 	INIT_LIST_HEAD(&dev_priv->context_list);
 	INIT_WORK(&dev_priv->mm.free_work, __i915_gem_free_work);
@@ -4826,14 +4895,14 @@ i915_gem_load_init(struct drm_i915_private *dev_priv)
 
 	init_waitqueue_head(&dev_priv->pending_flip_queue);
 
-	dev_priv->mm.interruptible = true;
-
 	atomic_set(&dev_priv->mm.bsd_engine_dispatch_index, 0);
 
 	spin_lock_init(&dev_priv->fb_tracking.lock);
 
 	return 0;
 
+err_priorities:
+	kmem_cache_destroy(dev_priv->priorities);
 err_dependencies:
 	kmem_cache_destroy(dev_priv->dependencies);
 err_requests:
@@ -4857,6 +4926,7 @@ void i915_gem_load_cleanup(struct drm_i915_private *dev_priv)
 	WARN_ON(!list_empty(&dev_priv->gt.timelines));
 	mutex_unlock(&dev_priv->drm.struct_mutex);
 
+	kmem_cache_destroy(dev_priv->priorities);
 	kmem_cache_destroy(dev_priv->dependencies);
 	kmem_cache_destroy(dev_priv->requests);
 	kmem_cache_destroy(dev_priv->vmas);
@@ -4868,9 +4938,10 @@ void i915_gem_load_cleanup(struct drm_i915_private *dev_priv)
 
 int i915_gem_freeze(struct drm_i915_private *dev_priv)
 {
-	mutex_lock(&dev_priv->drm.struct_mutex);
+	/* Discard all purgeable objects, let userspace recover those as
+	 * required after resuming.
+	 */
 	i915_gem_shrink_all(dev_priv);
-	mutex_unlock(&dev_priv->drm.struct_mutex);
 
 	return 0;
 }
@@ -4895,12 +4966,13 @@ int i915_gem_freeze_late(struct drm_i915_private *dev_priv)
 	 * we update that state just before writing out the image.
 	 *
 	 * To try and reduce the hibernation image, we manually shrink
-	 * the objects as well.
+	 * the objects as well, see i915_gem_freeze()
 	 */
 
-	mutex_lock(&dev_priv->drm.struct_mutex);
 	i915_gem_shrink(dev_priv, -1UL, I915_SHRINK_UNBOUND);
+	i915_gem_drain_freed_objects(dev_priv);
 
+	mutex_lock(&dev_priv->drm.struct_mutex);
 	for (p = phases; *p; p++) {
 		list_for_each_entry(obj, *p, global_link) {
 			obj->base.read_domains = I915_GEM_DOMAIN_CPU;
diff --git a/drivers/gpu/drm/i915/i915_gem.h b/drivers/gpu/drm/i915/i915_gem.h
index 5a49487368ca..ee54597465b6 100644
--- a/drivers/gpu/drm/i915/i915_gem.h
+++ b/drivers/gpu/drm/i915/i915_gem.h
@@ -25,6 +25,8 @@
 #ifndef __I915_GEM_H__
 #define __I915_GEM_H__
 
+#include <linux/bug.h>
+
 #ifdef CONFIG_DRM_I915_DEBUG_GEM
 #define GEM_BUG_ON(expr) BUG_ON(expr)
 #define GEM_WARN_ON(expr) WARN_ON(expr)
diff --git a/drivers/gpu/drm/i915/i915_gem_clflush.c b/drivers/gpu/drm/i915/i915_gem_clflush.c
index ffd01e02fe94..ffac7a1f0caf 100644
--- a/drivers/gpu/drm/i915/i915_gem_clflush.c
+++ b/drivers/gpu/drm/i915/i915_gem_clflush.c
@@ -27,7 +27,6 @@
 #include "i915_gem_clflush.h"
 
 static DEFINE_SPINLOCK(clflush_lock);
-static u64 clflush_context;
 
 struct clflush {
 	struct dma_fence dma; /* Must be first for dma_fence_free() */
@@ -157,7 +156,7 @@ void i915_gem_clflush_object(struct drm_i915_gem_object *obj,
 		dma_fence_init(&clflush->dma,
 			       &i915_clflush_ops,
 			       &clflush_lock,
-			       clflush_context,
+			       to_i915(obj->base.dev)->mm.unordered_timeline,
 			       0);
 		i915_sw_fence_init(&clflush->wait, i915_clflush_notify);
 
@@ -182,8 +181,3 @@ void i915_gem_clflush_object(struct drm_i915_gem_object *obj,
 		GEM_BUG_ON(obj->base.write_domain != I915_GEM_DOMAIN_CPU);
 	}
 }
-
-void i915_gem_clflush_init(struct drm_i915_private *i915)
-{
-	clflush_context = dma_fence_context_alloc(1);
-}
diff --git a/drivers/gpu/drm/i915/i915_gem_clflush.h b/drivers/gpu/drm/i915/i915_gem_clflush.h
index b62d61a2d15f..2455a7820937 100644
--- a/drivers/gpu/drm/i915/i915_gem_clflush.h
+++ b/drivers/gpu/drm/i915/i915_gem_clflush.h
@@ -28,7 +28,6 @@
 struct drm_i915_private;
 struct drm_i915_gem_object;
 
-void i915_gem_clflush_init(struct drm_i915_private *i915);
 void i915_gem_clflush_object(struct drm_i915_gem_object *obj,
 			     unsigned int flags);
 #define I915_CLFLUSH_FORCE BIT(0)
diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index 8bd0c4966913..c5d1666d7071 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -92,33 +92,6 @@
 
 #define ALL_L3_SLICES(dev) (1 << NUM_L3_SLICES(dev)) - 1
 
-static int get_context_size(struct drm_i915_private *dev_priv)
-{
-	int ret;
-	u32 reg;
-
-	switch (INTEL_GEN(dev_priv)) {
-	case 6:
-		reg = I915_READ(CXT_SIZE);
-		ret = GEN6_CXT_TOTAL_SIZE(reg) * 64;
-		break;
-	case 7:
-		reg = I915_READ(GEN7_CXT_SIZE);
-		if (IS_HASWELL(dev_priv))
-			ret = HSW_CXT_TOTAL_SIZE;
-		else
-			ret = GEN7_CXT_TOTAL_SIZE(reg) * 64;
-		break;
-	case 8:
-		ret = GEN8_CXT_TOTAL_SIZE;
-		break;
-	default:
-		BUG();
-	}
-
-	return ret;
-}
-
 void i915_gem_context_free(struct kref *ctx_ref)
 {
 	struct i915_gem_context *ctx = container_of(ctx_ref, typeof(*ctx), ref);
@@ -151,45 +124,6 @@ void i915_gem_context_free(struct kref *ctx_ref)
 	kfree(ctx);
 }
 
-static struct drm_i915_gem_object *
-alloc_context_obj(struct drm_i915_private *dev_priv, u64 size)
-{
-	struct drm_i915_gem_object *obj;
-	int ret;
-
-	lockdep_assert_held(&dev_priv->drm.struct_mutex);
-
-	obj = i915_gem_object_create(dev_priv, size);
-	if (IS_ERR(obj))
-		return obj;
-
-	/*
-	 * Try to make the context utilize L3 as well as LLC.
-	 *
-	 * On VLV we don't have L3 controls in the PTEs so we
-	 * shouldn't touch the cache level, especially as that
-	 * would make the object snooped which might have a
-	 * negative performance impact.
-	 *
-	 * Snooping is required on non-llc platforms in execlist
-	 * mode, but since all GGTT accesses use PAT entry 0 we
-	 * get snooping anyway regardless of cache_level.
-	 *
-	 * This is only applicable for Ivy Bridge devices since
-	 * later platforms don't have L3 control bits in the PTE.
-	 */
-	if (IS_IVYBRIDGE(dev_priv)) {
-		ret = i915_gem_object_set_cache_level(obj, I915_CACHE_L3_LLC);
-		/* Failure shouldn't ever happen this early */
-		if (WARN_ON(ret)) {
-			i915_gem_object_put(obj);
-			return ERR_PTR(ret);
-		}
-	}
-
-	return obj;
-}
-
 static void context_close(struct i915_gem_context *ctx)
 {
 	i915_gem_context_set_closed(ctx);
@@ -265,26 +199,7 @@ __create_hw_context(struct drm_i915_private *dev_priv,
 	kref_init(&ctx->ref);
 	list_add_tail(&ctx->link, &dev_priv->context_list);
 	ctx->i915 = dev_priv;
-
-	if (dev_priv->hw_context_size) {
-		struct drm_i915_gem_object *obj;
-		struct i915_vma *vma;
-
-		obj = alloc_context_obj(dev_priv, dev_priv->hw_context_size);
-		if (IS_ERR(obj)) {
-			ret = PTR_ERR(obj);
-			goto err_out;
-		}
-
-		vma = i915_vma_instance(obj, &dev_priv->ggtt.base, NULL);
-		if (IS_ERR(vma)) {
-			i915_gem_object_put(obj);
-			ret = PTR_ERR(vma);
-			goto err_out;
-		}
-
-		ctx->engine[RCS].state = vma;
-	}
+	ctx->priority = I915_PRIORITY_NORMAL;
 
 	/* Default context will never have a file_priv */
 	ret = DEFAULT_CONTEXT_HANDLE;
@@ -443,21 +358,6 @@ int i915_gem_context_init(struct drm_i915_private *dev_priv)
 	BUILD_BUG_ON(MAX_CONTEXT_HW_ID > INT_MAX);
 	ida_init(&dev_priv->context_hw_ida);
 
-	if (i915.enable_execlists) {
-		/* NB: intentionally left blank. We will allocate our own
-		 * backing objects as we need them, thank you very much */
-		dev_priv->hw_context_size = 0;
-	} else if (HAS_HW_CONTEXTS(dev_priv)) {
-		dev_priv->hw_context_size =
-			round_up(get_context_size(dev_priv),
-				 I915_GTT_PAGE_SIZE);
-		if (dev_priv->hw_context_size > (1<<20)) {
-			DRM_DEBUG_DRIVER("Disabling HW Contexts; invalid size %d\n",
-					 dev_priv->hw_context_size);
-			dev_priv->hw_context_size = 0;
-		}
-	}
-
 	ctx = i915_gem_create_context(dev_priv, NULL);
 	if (IS_ERR(ctx)) {
 		DRM_ERROR("Failed to create default global context (error %ld)\n",
@@ -477,8 +377,8 @@ int i915_gem_context_init(struct drm_i915_private *dev_priv)
 	GEM_BUG_ON(!i915_gem_context_is_kernel(ctx));
 
 	DRM_DEBUG_DRIVER("%s context support initialized\n",
-			i915.enable_execlists ? "LR" :
-			dev_priv->hw_context_size ? "HW" : "fake");
+			 dev_priv->engine[RCS]->context_size ? "logical" :
+			 "fake");
 	return 0;
 }
 
@@ -941,11 +841,6 @@ int i915_gem_switch_to_kernel_context(struct drm_i915_private *dev_priv)
 	return 0;
 }
 
-static bool contexts_enabled(struct drm_device *dev)
-{
-	return i915.enable_execlists || to_i915(dev)->hw_context_size;
-}
-
 static bool client_is_banned(struct drm_i915_file_private *file_priv)
 {
 	return file_priv->context_bans > I915_MAX_CLIENT_CONTEXT_BANS;
@@ -954,12 +849,13 @@ static bool client_is_banned(struct drm_i915_file_private *file_priv)
 int i915_gem_context_create_ioctl(struct drm_device *dev, void *data,
 				  struct drm_file *file)
 {
+	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct drm_i915_gem_context_create *args = data;
 	struct drm_i915_file_private *file_priv = file->driver_priv;
 	struct i915_gem_context *ctx;
 	int ret;
 
-	if (!contexts_enabled(dev))
+	if (!dev_priv->engine[RCS]->context_size)
 		return -ENODEV;
 
 	if (args->pad != 0)
@@ -977,7 +873,7 @@ int i915_gem_context_create_ioctl(struct drm_device *dev, void *data,
 	if (ret)
 		return ret;
 
-	ctx = i915_gem_create_context(to_i915(dev), file_priv);
+	ctx = i915_gem_create_context(dev_priv, file_priv);
 	mutex_unlock(&dev->struct_mutex);
 	if (IS_ERR(ctx))
 		return PTR_ERR(ctx);
diff --git a/drivers/gpu/drm/i915/i915_gem_dmabuf.c b/drivers/gpu/drm/i915/i915_gem_dmabuf.c
index f225bf680b6d..6176e589cf09 100644
--- a/drivers/gpu/drm/i915/i915_gem_dmabuf.c
+++ b/drivers/gpu/drm/i915/i915_gem_dmabuf.c
@@ -122,12 +122,36 @@ static void i915_gem_dmabuf_kunmap_atomic(struct dma_buf *dma_buf, unsigned long
 }
 static void *i915_gem_dmabuf_kmap(struct dma_buf *dma_buf, unsigned long page_num)
 {
+	struct drm_i915_gem_object *obj = dma_buf_to_obj(dma_buf);
+	struct page *page;
+
+	if (page_num >= obj->base.size >> PAGE_SHIFT)
+		return NULL;
+
+	if (!i915_gem_object_has_struct_page(obj))
+		return NULL;
+
+	if (i915_gem_object_pin_pages(obj))
+		return NULL;
+
+	/* Synchronisation is left to the caller (via .begin_cpu_access()) */
+	page = i915_gem_object_get_page(obj, page_num);
+	if (IS_ERR(page))
+		goto err_unpin;
+
+	return kmap(page);
+
+err_unpin:
+	i915_gem_object_unpin_pages(obj);
 	return NULL;
 }
 
 static void i915_gem_dmabuf_kunmap(struct dma_buf *dma_buf, unsigned long page_num, void *addr)
 {
+	struct drm_i915_gem_object *obj = dma_buf_to_obj(dma_buf);
 
+	kunmap(virt_to_page(addr));
+	i915_gem_object_unpin_pages(obj);
 }
 
 static int i915_gem_dmabuf_mmap(struct dma_buf *dma_buf, struct vm_area_struct *vma)
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index a3e59c8ef27b..04211c970b9f 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -1019,11 +1019,11 @@ i915_gem_execbuffer_relocate_slow(struct drm_device *dev,
 	for (i = 0; i < count; i++)
 		total += exec[i].relocation_count;
 
-	reloc_offset = drm_malloc_ab(count, sizeof(*reloc_offset));
-	reloc = drm_malloc_ab(total, sizeof(*reloc));
+	reloc_offset = kvmalloc_array(count, sizeof(*reloc_offset), GFP_KERNEL);
+	reloc = kvmalloc_array(total, sizeof(*reloc), GFP_KERNEL);
 	if (reloc == NULL || reloc_offset == NULL) {
-		drm_free_large(reloc);
-		drm_free_large(reloc_offset);
+		kvfree(reloc);
+		kvfree(reloc_offset);
 		mutex_lock(&dev->struct_mutex);
 		return -ENOMEM;
 	}
@@ -1099,8 +1099,8 @@ i915_gem_execbuffer_relocate_slow(struct drm_device *dev,
 	 */
 
 err:
-	drm_free_large(reloc);
-	drm_free_large(reloc_offset);
+	kvfree(reloc);
+	kvfree(reloc_offset);
 	return ret;
 }
 
@@ -1114,6 +1114,18 @@ i915_gem_execbuffer_move_to_gpu(struct drm_i915_gem_request *req,
 	list_for_each_entry(vma, vmas, exec_list) {
 		struct drm_i915_gem_object *obj = vma->obj;
 
+		if (vma->exec_entry->flags & EXEC_OBJECT_CAPTURE) {
+			struct i915_gem_capture_list *capture;
+
+			capture = kmalloc(sizeof(*capture), GFP_KERNEL);
+			if (unlikely(!capture))
+				return -ENOMEM;
+
+			capture->next = req->capture_list;
+			capture->vma = vma;
+			req->capture_list = capture;
+		}
+
 		if (vma->exec_entry->flags & EXEC_OBJECT_ASYNC)
 			continue;
 
@@ -1859,13 +1871,13 @@ i915_gem_execbuffer(struct drm_device *dev, void *data,
 	}
 
 	/* Copy in the exec list from userland */
-	exec_list = drm_malloc_ab(sizeof(*exec_list), args->buffer_count);
-	exec2_list = drm_malloc_ab(sizeof(*exec2_list), args->buffer_count);
+	exec_list = kvmalloc_array(sizeof(*exec_list), args->buffer_count, GFP_KERNEL);
+	exec2_list = kvmalloc_array(sizeof(*exec2_list), args->buffer_count, GFP_KERNEL);
 	if (exec_list == NULL || exec2_list == NULL) {
 		DRM_DEBUG("Failed to allocate exec list for %d buffers\n",
 			  args->buffer_count);
-		drm_free_large(exec_list);
-		drm_free_large(exec2_list);
+		kvfree(exec_list);
+		kvfree(exec2_list);
 		return -ENOMEM;
 	}
 	ret = copy_from_user(exec_list,
@@ -1874,8 +1886,8 @@ i915_gem_execbuffer(struct drm_device *dev, void *data,
 	if (ret != 0) {
 		DRM_DEBUG("copy %d exec entries failed %d\n",
 			  args->buffer_count, ret);
-		drm_free_large(exec_list);
-		drm_free_large(exec2_list);
+		kvfree(exec_list);
+		kvfree(exec2_list);
 		return -EFAULT;
 	}
 
@@ -1924,8 +1936,8 @@ i915_gem_execbuffer(struct drm_device *dev, void *data,
 		}
 	}
 
-	drm_free_large(exec_list);
-	drm_free_large(exec2_list);
+	kvfree(exec_list);
+	kvfree(exec2_list);
 	return ret;
 }
 
@@ -1943,7 +1955,7 @@ i915_gem_execbuffer2(struct drm_device *dev, void *data,
 		return -EINVAL;
 	}
 
-	exec2_list = drm_malloc_gfp(args->buffer_count,
+	exec2_list = kvmalloc_array(args->buffer_count,
 				    sizeof(*exec2_list),
 				    GFP_TEMPORARY);
 	if (exec2_list == NULL) {
@@ -1957,7 +1969,7 @@ i915_gem_execbuffer2(struct drm_device *dev, void *data,
 	if (ret != 0) {
 		DRM_DEBUG("copy %d exec entries failed %d\n",
 			  args->buffer_count, ret);
-		drm_free_large(exec2_list);
+		kvfree(exec2_list);
 		return -EFAULT;
 	}
 
@@ -1984,6 +1996,6 @@ i915_gem_execbuffer2(struct drm_device *dev, void *data,
 		}
 	}
 
-	drm_free_large(exec2_list);
+	kvfree(exec2_list);
 	return ret;
 }
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index f1989b8792dd..4ff854e6413c 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -168,13 +168,11 @@ int intel_sanitize_enable_ppgtt(struct drm_i915_private *dev_priv,
 	if (enable_ppgtt == 3 && has_full_48bit_ppgtt)
 		return 3;
 
-#ifdef CONFIG_INTEL_IOMMU
 	/* Disable ppgtt on SNB if VT-d is on. */
-	if (IS_GEN6(dev_priv) && intel_iommu_gfx_mapped) {
+	if (IS_GEN6(dev_priv) && intel_vtd_active()) {
 		DRM_INFO("Disabling PPGTT because VT-d is on\n");
 		return 0;
 	}
-#endif
 
 	/* Early VLV doesn't have this */
 	if (IS_VALLEYVIEW(dev_priv) && dev_priv->drm.pdev->revision < 0xb) {
@@ -1992,14 +1990,10 @@ void i915_ppgtt_release(struct kref *kref)
  */
 static bool needs_idle_maps(struct drm_i915_private *dev_priv)
 {
-#ifdef CONFIG_INTEL_IOMMU
 	/* Query intel_iommu to see if we need the workaround. Presumably that
 	 * was loaded first.
 	 */
-	if (IS_GEN5(dev_priv) && IS_MOBILE(dev_priv) && intel_iommu_gfx_mapped)
-		return true;
-#endif
-	return false;
+	return IS_GEN5(dev_priv) && IS_MOBILE(dev_priv) && intel_vtd_active();
 }
 
 void i915_check_and_clear_faults(struct drm_i915_private *dev_priv)
@@ -2678,14 +2672,14 @@ static size_t gen6_get_stolen_size(u16 snb_gmch_ctl)
 {
 	snb_gmch_ctl >>= SNB_GMCH_GMS_SHIFT;
 	snb_gmch_ctl &= SNB_GMCH_GMS_MASK;
-	return snb_gmch_ctl << 25; /* 32 MB units */
+	return (size_t)snb_gmch_ctl << 25; /* 32 MB units */
 }
 
 static size_t gen8_get_stolen_size(u16 bdw_gmch_ctl)
 {
 	bdw_gmch_ctl >>= BDW_GMCH_GMS_SHIFT;
 	bdw_gmch_ctl &= BDW_GMCH_GMS_MASK;
-	return bdw_gmch_ctl << 25; /* 32 MB units */
+	return (size_t)bdw_gmch_ctl << 25; /* 32 MB units */
 }
 
 static size_t chv_get_stolen_size(u16 gmch_ctrl)
@@ -2699,11 +2693,11 @@ static size_t chv_get_stolen_size(u16 gmch_ctrl)
 	 * 0x17 to 0x1d: 4MB increments start at 36MB
 	 */
 	if (gmch_ctrl < 0x11)
-		return gmch_ctrl << 25;
+		return (size_t)gmch_ctrl << 25;
 	else if (gmch_ctrl < 0x17)
-		return (gmch_ctrl - 0x11 + 2) << 22;
+		return (size_t)(gmch_ctrl - 0x11 + 2) << 22;
 	else
-		return (gmch_ctrl - 0x17 + 9) << 22;
+		return (size_t)(gmch_ctrl - 0x17 + 9) << 22;
 }
 
 static size_t gen9_get_stolen_size(u16 gen9_gmch_ctl)
@@ -2712,10 +2706,10 @@ static size_t gen9_get_stolen_size(u16 gen9_gmch_ctl)
 	gen9_gmch_ctl &= BDW_GMCH_GMS_MASK;
 
 	if (gen9_gmch_ctl < 0xf0)
-		return gen9_gmch_ctl << 25; /* 32 MB units */
+		return (size_t)gen9_gmch_ctl << 25; /* 32 MB units */
 	else
 		/* 4MB increments starting at 0xf0 for 4MB */
-		return (gen9_gmch_ctl - 0xf0 + 1) << 22;
+		return (size_t)(gen9_gmch_ctl - 0xf0 + 1) << 22;
 }
 
 static int ggtt_probe_common(struct i915_ggtt *ggtt, u64 size)
@@ -2842,13 +2836,17 @@ static int gen8_gmch_probe(struct i915_ggtt *ggtt)
 	struct pci_dev *pdev = dev_priv->drm.pdev;
 	unsigned int size;
 	u16 snb_gmch_ctl;
+	int err;
 
 	/* TODO: We're not aware of mappable constraints on gen8 yet */
 	ggtt->mappable_base = pci_resource_start(pdev, 2);
 	ggtt->mappable_end = pci_resource_len(pdev, 2);
 
-	if (!pci_set_dma_mask(pdev, DMA_BIT_MASK(39)))
-		pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(39));
+	err = pci_set_dma_mask(pdev, DMA_BIT_MASK(39));
+	if (!err)
+		err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(39));
+	if (err)
+		DRM_ERROR("Can't set DMA mask/consistent mask (%d)\n", err);
 
 	pci_read_config_word(pdev, SNB_GMCH_CTRL, &snb_gmch_ctl);
 
@@ -2899,6 +2897,7 @@ static int gen6_gmch_probe(struct i915_ggtt *ggtt)
 	struct pci_dev *pdev = dev_priv->drm.pdev;
 	unsigned int size;
 	u16 snb_gmch_ctl;
+	int err;
 
 	ggtt->mappable_base = pci_resource_start(pdev, 2);
 	ggtt->mappable_end = pci_resource_len(pdev, 2);
@@ -2911,8 +2910,11 @@ static int gen6_gmch_probe(struct i915_ggtt *ggtt)
 		return -ENXIO;
 	}
 
-	if (!pci_set_dma_mask(pdev, DMA_BIT_MASK(40)))
-		pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(40));
+	err = pci_set_dma_mask(pdev, DMA_BIT_MASK(40));
+	if (!err)
+		err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(40));
+	if (err)
+		DRM_ERROR("Can't set DMA mask/consistent mask (%d)\n", err);
 	pci_read_config_word(pdev, SNB_GMCH_CTRL, &snb_gmch_ctl);
 
 	ggtt->stolen_size = gen6_get_stolen_size(snb_gmch_ctl);
@@ -3031,10 +3033,8 @@ int i915_ggtt_probe_hw(struct drm_i915_private *dev_priv)
 		 ggtt->base.total >> 20);
 	DRM_DEBUG_DRIVER("GMADR size = %lldM\n", ggtt->mappable_end >> 20);
 	DRM_DEBUG_DRIVER("GTT stolen size = %uM\n", ggtt->stolen_size >> 20);
-#ifdef CONFIG_INTEL_IOMMU
-	if (intel_iommu_gfx_mapped)
+	if (intel_vtd_active())
 		DRM_INFO("VT-d active for gfx access\n");
-#endif
 
 	return 0;
 }
@@ -3095,13 +3095,17 @@ int i915_ggtt_enable_hw(struct drm_i915_private *dev_priv)
 
 void i915_ggtt_enable_guc(struct drm_i915_private *i915)
 {
+	GEM_BUG_ON(i915->ggtt.invalidate != gen6_ggtt_invalidate);
+
 	i915->ggtt.invalidate = guc_ggtt_invalidate;
 }
 
 void i915_ggtt_disable_guc(struct drm_i915_private *i915)
 {
-	if (i915->ggtt.invalidate == guc_ggtt_invalidate)
-		i915->ggtt.invalidate = gen6_ggtt_invalidate;
+	/* We should only be called after i915_ggtt_enable_guc() */
+	GEM_BUG_ON(i915->ggtt.invalidate != guc_ggtt_invalidate);
+
+	i915->ggtt.invalidate = gen6_ggtt_invalidate;
 }
 
 void i915_gem_restore_gtt_mappings(struct drm_i915_private *dev_priv)
@@ -3210,7 +3214,7 @@ intel_rotate_pages(struct intel_rotation_info *rot_info,
 	int ret = -ENOMEM;
 
 	/* Allocate a temporary list of source pages for random access. */
-	page_addr_list = drm_malloc_gfp(n_pages,
+	page_addr_list = kvmalloc_array(n_pages,
 					sizeof(dma_addr_t),
 					GFP_TEMPORARY);
 	if (!page_addr_list)
@@ -3243,14 +3247,14 @@ intel_rotate_pages(struct intel_rotation_info *rot_info,
 	DRM_DEBUG_KMS("Created rotated page mapping for object size %zu (%ux%u tiles, %u pages)\n",
 		      obj->base.size, rot_info->plane[0].width, rot_info->plane[0].height, size);
 
-	drm_free_large(page_addr_list);
+	kvfree(page_addr_list);
 
 	return st;
 
 err_sg_alloc:
 	kfree(st);
 err_st_alloc:
-	drm_free_large(page_addr_list);
+	kvfree(page_addr_list);
 
 	DRM_DEBUG_KMS("Failed to create rotated mapping for object size %zu! (%ux%u tiles, %u pages)\n",
 		      obj->base.size, rot_info->plane[0].width, rot_info->plane[0].height, size);
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h
index fb15684c1d83..da9aa9f706e7 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.h
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.h
@@ -255,6 +255,7 @@ struct i915_address_space {
 	struct drm_i915_file_private *file;
 	struct list_head global_link;
 	u64 total;		/* size addr space maps (ex. 2GB for ggtt) */
+	u64 reserved;		/* size addr space reserved */
 
 	bool closed;
 
diff --git a/drivers/gpu/drm/i915/i915_gem_object.h b/drivers/gpu/drm/i915/i915_gem_object.h
index 174cf923c236..915057824284 100644
--- a/drivers/gpu/drm/i915/i915_gem_object.h
+++ b/drivers/gpu/drm/i915/i915_gem_object.h
@@ -37,8 +37,8 @@
 
 struct drm_i915_gem_object_ops {
 	unsigned int flags;
-#define I915_GEM_OBJECT_HAS_STRUCT_PAGE 0x1
-#define I915_GEM_OBJECT_IS_SHRINKABLE   0x2
+#define I915_GEM_OBJECT_HAS_STRUCT_PAGE BIT(0)
+#define I915_GEM_OBJECT_IS_SHRINKABLE   BIT(1)
 
 	/* Interface between the GEM object and its backing storage.
 	 * get_pages() is called once prior to the use of the associated set
@@ -68,8 +68,23 @@ struct drm_i915_gem_object {
 
 	const struct drm_i915_gem_object_ops *ops;
 
-	/** List of VMAs backed by this object */
+	/**
+	 * @vma_list: List of VMAs backed by this object
+	 *
+	 * The VMA on this list are ordered by type, all GGTT vma are placed
+	 * at the head and all ppGTT vma are placed at the tail. The different
+	 * types of GGTT vma are unordered between themselves, use the
+	 * @vma_tree (which has a defined order between all VMA) to find an
+	 * exact match.
+	 */
 	struct list_head vma_list;
+	/**
+	 * @vma_tree: Ordered tree of VMAs backed by this object
+	 *
+	 * All VMA created for this object are placed in the @vma_tree for
+	 * fast retrieval via a binary search in i915_vma_instance().
+	 * They are also added to @vma_list for easy iteration.
+	 */
 	struct rb_root vma_tree;
 
 	/** Stolen memory for this object, instead of being backed by shmem. */
diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c
index 5ddbc9499775..8c59c79cbd8b 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.c
+++ b/drivers/gpu/drm/i915/i915_gem_request.c
@@ -61,8 +61,8 @@ static bool i915_fence_enable_signaling(struct dma_fence *fence)
 	if (i915_fence_signaled(fence))
 		return false;
 
-	intel_engine_enable_signaling(to_request(fence));
-	return true;
+	intel_engine_enable_signaling(to_request(fence), true);
+	return !i915_fence_signaled(fence);
 }
 
 static signed long i915_fence_wait(struct dma_fence *fence,
@@ -159,7 +159,7 @@ i915_priotree_fini(struct drm_i915_private *i915, struct i915_priotree *pt)
 {
 	struct i915_dependency *dep, *next;
 
-	GEM_BUG_ON(!RB_EMPTY_NODE(&pt->node));
+	GEM_BUG_ON(!list_empty(&pt->link));
 
 	/* Everyone we depended upon (the fences we wait to be signaled)
 	 * should retire before us and remove themselves from our list.
@@ -185,7 +185,7 @@ i915_priotree_init(struct i915_priotree *pt)
 {
 	INIT_LIST_HEAD(&pt->signalers_list);
 	INIT_LIST_HEAD(&pt->waiters_list);
-	RB_CLEAR_NODE(&pt->node);
+	INIT_LIST_HEAD(&pt->link);
 	pt->priority = INT_MIN;
 }
 
@@ -214,12 +214,12 @@ static int reset_all_global_seqno(struct drm_i915_private *i915, u32 seqno)
 		}
 
 		/* Finally reset hw state */
-		tl->seqno = seqno;
 		intel_engine_init_global_seqno(engine, seqno);
+		tl->seqno = seqno;
 
 		list_for_each_entry(timeline, &i915->gt.timelines, link)
-			memset(timeline->engine[id].sync_seqno, 0,
-			       sizeof(timeline->engine[id].sync_seqno));
+			memset(timeline->engine[id].global_sync, 0,
+			       sizeof(timeline->engine[id].global_sync));
 	}
 
 	return 0;
@@ -271,6 +271,48 @@ void i915_gem_retire_noop(struct i915_gem_active *active,
 	/* Space left intentionally blank */
 }
 
+static void advance_ring(struct drm_i915_gem_request *request)
+{
+	unsigned int tail;
+
+	/* We know the GPU must have read the request to have
+	 * sent us the seqno + interrupt, so use the position
+	 * of tail of the request to update the last known position
+	 * of the GPU head.
+	 *
+	 * Note this requires that we are always called in request
+	 * completion order.
+	 */
+	if (list_is_last(&request->ring_link, &request->ring->request_list)) {
+		/* We may race here with execlists resubmitting this request
+		 * as we retire it. The resubmission will move the ring->tail
+		 * forwards (to request->wa_tail). We either read the
+		 * current value that was written to hw, or the value that
+		 * is just about to be. Either works, if we miss the last two
+		 * noops - they are safe to be replayed on a reset.
+		 */
+		tail = READ_ONCE(request->ring->tail);
+	} else {
+		tail = request->postfix;
+	}
+	list_del(&request->ring_link);
+
+	request->ring->head = tail;
+}
+
+static void free_capture_list(struct drm_i915_gem_request *request)
+{
+	struct i915_gem_capture_list *capture;
+
+	capture = request->capture_list;
+	while (capture) {
+		struct i915_gem_capture_list *next = capture->next;
+
+		kfree(capture);
+		capture = next;
+	}
+}
+
 static void i915_gem_request_retire(struct drm_i915_gem_request *request)
 {
 	struct intel_engine_cs *engine = request->engine;
@@ -287,16 +329,6 @@ static void i915_gem_request_retire(struct drm_i915_gem_request *request)
 	list_del_init(&request->link);
 	spin_unlock_irq(&engine->timeline->lock);
 
-	/* We know the GPU must have read the request to have
-	 * sent us the seqno + interrupt, so use the position
-	 * of tail of the request to update the last known position
-	 * of the GPU head.
-	 *
-	 * Note this requires that we are always called in request
-	 * completion order.
-	 */
-	list_del(&request->ring_link);
-	request->ring->head = request->postfix;
 	if (!--request->i915->gt.active_requests) {
 		GEM_BUG_ON(!request->i915->gt.awake);
 		mod_delayed_work(request->i915->wq,
@@ -304,6 +336,9 @@ static void i915_gem_request_retire(struct drm_i915_gem_request *request)
 				 msecs_to_jiffies(100));
 	}
 	unreserve_seqno(request->engine);
+	advance_ring(request);
+
+	free_capture_list(request);
 
 	/* Walk through the active list, calling retire on each. This allows
 	 * objects to track their GPU activity and mark themselves as idle
@@ -402,7 +437,7 @@ void __i915_gem_request_submit(struct drm_i915_gem_request *request)
 	spin_lock_nested(&request->lock, SINGLE_DEPTH_NESTING);
 	request->global_seqno = seqno;
 	if (test_bit(DMA_FENCE_FLAG_ENABLE_SIGNAL_BIT, &request->fence.flags))
-		intel_engine_enable_signaling(request);
+		intel_engine_enable_signaling(request, false);
 	spin_unlock(&request->lock);
 
 	engine->emit_breadcrumb(request,
@@ -503,9 +538,6 @@ submit_notify(struct i915_sw_fence *fence, enum i915_sw_fence_notify state)
  *
  * @engine: engine that we wish to issue the request on.
  * @ctx: context that the request will be associated with.
- *       This can be NULL if the request is not directly related to
- *       any specific user context, in which case this function will
- *       choose an appropriate context to use.
  *
  * Returns a pointer to the allocated request if successful,
  * or an error code if not.
@@ -516,6 +548,7 @@ i915_gem_request_alloc(struct intel_engine_cs *engine,
 {
 	struct drm_i915_private *dev_priv = engine->i915;
 	struct drm_i915_gem_request *req;
+	struct intel_ring *ring;
 	int ret;
 
 	lockdep_assert_held(&dev_priv->drm.struct_mutex);
@@ -530,9 +563,10 @@ i915_gem_request_alloc(struct intel_engine_cs *engine,
 	 * GGTT space, so do this first before we reserve a seqno for
 	 * ourselves.
 	 */
-	ret = engine->context_pin(engine, ctx);
-	if (ret)
-		return ERR_PTR(ret);
+	ring = engine->context_pin(engine, ctx);
+	if (IS_ERR(ring))
+		return ERR_CAST(ring);
+	GEM_BUG_ON(!ring);
 
 	ret = reserve_seqno(engine);
 	if (ret)
@@ -598,11 +632,13 @@ i915_gem_request_alloc(struct intel_engine_cs *engine,
 	req->i915 = dev_priv;
 	req->engine = engine;
 	req->ctx = ctx;
+	req->ring = ring;
 
 	/* No zalloc, must clear what we need by hand */
 	req->global_seqno = 0;
 	req->file_priv = NULL;
 	req->batch = NULL;
+	req->capture_list = NULL;
 
 	/*
 	 * Reserve space in the ring buffer for all the commands required to
@@ -623,7 +659,7 @@ i915_gem_request_alloc(struct intel_engine_cs *engine,
 	 * GPU processing the request, we never over-estimate the
 	 * position of the head.
 	 */
-	req->head = req->ring->tail;
+	req->head = req->ring->emit;
 
 	/* Check that we didn't interrupt ourselves with a new request */
 	GEM_BUG_ON(req->timeline->seqno != req->fence.seqno);
@@ -647,10 +683,10 @@ static int
 i915_gem_request_await_request(struct drm_i915_gem_request *to,
 			       struct drm_i915_gem_request *from)
 {
-	u32 seqno;
 	int ret;
 
 	GEM_BUG_ON(to == from);
+	GEM_BUG_ON(to->timeline == from->timeline);
 
 	if (i915_gem_request_completed(from))
 		return 0;
@@ -663,9 +699,6 @@ i915_gem_request_await_request(struct drm_i915_gem_request *to,
 			return ret;
 	}
 
-	if (to->timeline == from->timeline)
-		return 0;
-
 	if (to->engine == from->engine) {
 		ret = i915_sw_fence_await_sw_fence_gfp(&to->submit,
 						       &from->submit,
@@ -673,56 +706,41 @@ i915_gem_request_await_request(struct drm_i915_gem_request *to,
 		return ret < 0 ? ret : 0;
 	}
 
-	seqno = i915_gem_request_global_seqno(from);
-	if (!seqno) {
-		ret = i915_sw_fence_await_dma_fence(&to->submit,
-						    &from->fence, 0,
-						    GFP_KERNEL);
-		return ret < 0 ? ret : 0;
-	}
+	if (to->engine->semaphore.sync_to) {
+		u32 seqno;
 
-	if (seqno <= to->timeline->sync_seqno[from->engine->id])
-		return 0;
+		GEM_BUG_ON(!from->engine->semaphore.signal);
 
-	trace_i915_gem_ring_sync_to(to, from);
-	if (!i915.semaphores) {
-		if (!i915_spin_request(from, TASK_INTERRUPTIBLE, 2)) {
-			ret = i915_sw_fence_await_dma_fence(&to->submit,
-							    &from->fence, 0,
-							    GFP_KERNEL);
-			if (ret < 0)
-				return ret;
-		}
-	} else {
+		seqno = i915_gem_request_global_seqno(from);
+		if (!seqno)
+			goto await_dma_fence;
+
+		if (seqno <= to->timeline->global_sync[from->engine->id])
+			return 0;
+
+		trace_i915_gem_ring_sync_to(to, from);
 		ret = to->engine->semaphore.sync_to(to, from);
 		if (ret)
 			return ret;
+
+		to->timeline->global_sync[from->engine->id] = seqno;
+		return 0;
 	}
 
-	to->timeline->sync_seqno[from->engine->id] = seqno;
-	return 0;
+await_dma_fence:
+	ret = i915_sw_fence_await_dma_fence(&to->submit,
+					    &from->fence, 0,
+					    GFP_KERNEL);
+	return ret < 0 ? ret : 0;
 }
 
 int
 i915_gem_request_await_dma_fence(struct drm_i915_gem_request *req,
 				 struct dma_fence *fence)
 {
-	struct dma_fence_array *array;
+	struct dma_fence **child = &fence;
+	unsigned int nchild = 1;
 	int ret;
-	int i;
-
-	if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags))
-		return 0;
-
-	if (dma_fence_is_i915(fence))
-		return i915_gem_request_await_request(req, to_request(fence));
-
-	if (!dma_fence_is_array(fence)) {
-		ret = i915_sw_fence_await_dma_fence(&req->submit,
-						    fence, I915_FENCE_TIMEOUT,
-						    GFP_KERNEL);
-		return ret < 0 ? ret : 0;
-	}
 
 	/* Note that if the fence-array was created in signal-on-any mode,
 	 * we should *not* decompose it into its individual fences. However,
@@ -731,21 +749,46 @@ i915_gem_request_await_dma_fence(struct drm_i915_gem_request *req,
 	 * amdgpu and we should not see any incoming fence-array from
 	 * sync-file being in signal-on-any mode.
 	 */
+	if (dma_fence_is_array(fence)) {
+		struct dma_fence_array *array = to_dma_fence_array(fence);
 
-	array = to_dma_fence_array(fence);
-	for (i = 0; i < array->num_fences; i++) {
-		struct dma_fence *child = array->fences[i];
+		child = array->fences;
+		nchild = array->num_fences;
+		GEM_BUG_ON(!nchild);
+	}
 
-		if (dma_fence_is_i915(child))
+	do {
+		fence = *child++;
+		if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags))
+			continue;
+
+		/*
+		 * Requests on the same timeline are explicitly ordered, along
+		 * with their dependencies, by i915_add_request() which ensures
+		 * that requests are submitted in-order through each ring.
+		 */
+		if (fence->context == req->fence.context)
+			continue;
+
+		/* Squash repeated waits to the same timelines */
+		if (fence->context != req->i915->mm.unordered_timeline &&
+		    intel_timeline_sync_is_later(req->timeline, fence))
+			continue;
+
+		if (dma_fence_is_i915(fence))
 			ret = i915_gem_request_await_request(req,
-							     to_request(child));
+							     to_request(fence));
 		else
-			ret = i915_sw_fence_await_dma_fence(&req->submit,
-							    child, I915_FENCE_TIMEOUT,
+			ret = i915_sw_fence_await_dma_fence(&req->submit, fence,
+							    I915_FENCE_TIMEOUT,
 							    GFP_KERNEL);
 		if (ret < 0)
 			return ret;
-	}
+
+		/* Record the latest fence used against each timeline */
+		if (fence->context != req->i915->mm.unordered_timeline)
+			intel_timeline_sync_set(req->timeline, fence);
+	} while (--nchild);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/i915/i915_gem_request.h b/drivers/gpu/drm/i915/i915_gem_request.h
index 129c58bb4805..7b7c84369d78 100644
--- a/drivers/gpu/drm/i915/i915_gem_request.h
+++ b/drivers/gpu/drm/i915/i915_gem_request.h
@@ -67,12 +67,18 @@ struct i915_dependency {
 struct i915_priotree {
 	struct list_head signalers_list; /* those before us, we depend upon */
 	struct list_head waiters_list; /* those after us, they depend upon us */
-	struct rb_node node;
+	struct list_head link;
 	int priority;
 #define I915_PRIORITY_MAX 1024
+#define I915_PRIORITY_NORMAL 0
 #define I915_PRIORITY_MIN (-I915_PRIORITY_MAX)
 };
 
+struct i915_gem_capture_list {
+	struct i915_gem_capture_list *next;
+	struct i915_vma *vma;
+};
+
 /**
  * Request queue structure.
  *
@@ -167,6 +173,12 @@ struct drm_i915_gem_request {
 	 * error state dump only).
 	 */
 	struct i915_vma *batch;
+	/** Additional buffers requested by userspace to be captured upon
+	 * a GPU hang. The vma/obj on this list are protected by their
+	 * active reference - all objects on this list must also be
+	 * on the active_list (of their final request).
+	 */
+	struct i915_gem_capture_list *capture_list;
 	struct list_head active_list;
 
 	/** Time at which this request was emitted, in jiffies. */
diff --git a/drivers/gpu/drm/i915/i915_gem_shrinker.c b/drivers/gpu/drm/i915/i915_gem_shrinker.c
index 57d9f7f4ef15..58f27369183c 100644
--- a/drivers/gpu/drm/i915/i915_gem_shrinker.c
+++ b/drivers/gpu/drm/i915/i915_gem_shrinker.c
@@ -35,9 +35,9 @@
 #include "i915_drv.h"
 #include "i915_trace.h"
 
-static bool i915_gem_shrinker_lock(struct drm_device *dev, bool *unlock)
+static bool shrinker_lock(struct drm_i915_private *dev_priv, bool *unlock)
 {
-	switch (mutex_trylock_recursive(&dev->struct_mutex)) {
+	switch (mutex_trylock_recursive(&dev_priv->drm.struct_mutex)) {
 	case MUTEX_TRYLOCK_FAILED:
 		return false;
 
@@ -53,21 +53,29 @@ static bool i915_gem_shrinker_lock(struct drm_device *dev, bool *unlock)
 	BUG();
 }
 
-static void i915_gem_shrinker_unlock(struct drm_device *dev, bool unlock)
+static void shrinker_unlock(struct drm_i915_private *dev_priv, bool unlock)
 {
 	if (!unlock)
 		return;
 
-	mutex_unlock(&dev->struct_mutex);
+	mutex_unlock(&dev_priv->drm.struct_mutex);
 }
 
 static bool any_vma_pinned(struct drm_i915_gem_object *obj)
 {
 	struct i915_vma *vma;
 
-	list_for_each_entry(vma, &obj->vma_list, obj_link)
+	list_for_each_entry(vma, &obj->vma_list, obj_link) {
+		/* Only GGTT vma may be permanently pinned, and are always
+		 * at the start of the list. We can stop hunting as soon
+		 * as we see a ppGTT vma.
+		 */
+		if (!i915_vma_is_ggtt(vma))
+			break;
+
 		if (i915_vma_is_pinned(vma))
 			return true;
+	}
 
 	return false;
 }
@@ -153,7 +161,7 @@ i915_gem_shrink(struct drm_i915_private *dev_priv,
 	unsigned long count = 0;
 	bool unlock;
 
-	if (!i915_gem_shrinker_lock(&dev_priv->drm, &unlock))
+	if (!shrinker_lock(dev_priv, &unlock))
 		return 0;
 
 	trace_i915_gem_shrink(dev_priv, target, flags);
@@ -241,7 +249,7 @@ i915_gem_shrink(struct drm_i915_private *dev_priv,
 
 	i915_gem_retire_requests(dev_priv);
 
-	i915_gem_shrinker_unlock(&dev_priv->drm, unlock);
+	shrinker_unlock(dev_priv, unlock);
 
 	return count;
 }
@@ -279,12 +287,11 @@ i915_gem_shrinker_count(struct shrinker *shrinker, struct shrink_control *sc)
 {
 	struct drm_i915_private *dev_priv =
 		container_of(shrinker, struct drm_i915_private, mm.shrinker);
-	struct drm_device *dev = &dev_priv->drm;
 	struct drm_i915_gem_object *obj;
 	unsigned long count;
 	bool unlock;
 
-	if (!i915_gem_shrinker_lock(dev, &unlock))
+	if (!shrinker_lock(dev_priv, &unlock))
 		return 0;
 
 	i915_gem_retire_requests(dev_priv);
@@ -299,7 +306,7 @@ i915_gem_shrinker_count(struct shrinker *shrinker, struct shrink_control *sc)
 			count += obj->base.size >> PAGE_SHIFT;
 	}
 
-	i915_gem_shrinker_unlock(dev, unlock);
+	shrinker_unlock(dev_priv, unlock);
 
 	return count;
 }
@@ -309,11 +316,10 @@ i915_gem_shrinker_scan(struct shrinker *shrinker, struct shrink_control *sc)
 {
 	struct drm_i915_private *dev_priv =
 		container_of(shrinker, struct drm_i915_private, mm.shrinker);
-	struct drm_device *dev = &dev_priv->drm;
 	unsigned long freed;
 	bool unlock;
 
-	if (!i915_gem_shrinker_lock(dev, &unlock))
+	if (!shrinker_lock(dev_priv, &unlock))
 		return SHRINK_STOP;
 
 	freed = i915_gem_shrink(dev_priv,
@@ -326,27 +332,30 @@ i915_gem_shrinker_scan(struct shrinker *shrinker, struct shrink_control *sc)
 					 sc->nr_to_scan - freed,
 					 I915_SHRINK_BOUND |
 					 I915_SHRINK_UNBOUND);
+	if (freed < sc->nr_to_scan && current_is_kswapd()) {
+		intel_runtime_pm_get(dev_priv);
+		freed += i915_gem_shrink(dev_priv,
+					 sc->nr_to_scan - freed,
+					 I915_SHRINK_ACTIVE |
+					 I915_SHRINK_BOUND |
+					 I915_SHRINK_UNBOUND);
+		intel_runtime_pm_put(dev_priv);
+	}
 
-	i915_gem_shrinker_unlock(dev, unlock);
+	shrinker_unlock(dev_priv, unlock);
 
 	return freed;
 }
 
-struct shrinker_lock_uninterruptible {
-	bool was_interruptible;
-	bool unlock;
-};
-
 static bool
-i915_gem_shrinker_lock_uninterruptible(struct drm_i915_private *dev_priv,
-				       struct shrinker_lock_uninterruptible *slu,
-				       int timeout_ms)
+shrinker_lock_uninterruptible(struct drm_i915_private *dev_priv, bool *unlock,
+			      int timeout_ms)
 {
 	unsigned long timeout = jiffies + msecs_to_jiffies_timeout(timeout_ms);
 
 	do {
 		if (i915_gem_wait_for_idle(dev_priv, 0) == 0 &&
-		    i915_gem_shrinker_lock(&dev_priv->drm, &slu->unlock))
+		    shrinker_lock(dev_priv, unlock))
 			break;
 
 		schedule_timeout_killable(1);
@@ -359,29 +368,19 @@ i915_gem_shrinker_lock_uninterruptible(struct drm_i915_private *dev_priv,
 		}
 	} while (1);
 
-	slu->was_interruptible = dev_priv->mm.interruptible;
-	dev_priv->mm.interruptible = false;
 	return true;
 }
 
-static void
-i915_gem_shrinker_unlock_uninterruptible(struct drm_i915_private *dev_priv,
-					 struct shrinker_lock_uninterruptible *slu)
-{
-	dev_priv->mm.interruptible = slu->was_interruptible;
-	i915_gem_shrinker_unlock(&dev_priv->drm, slu->unlock);
-}
-
 static int
 i915_gem_shrinker_oom(struct notifier_block *nb, unsigned long event, void *ptr)
 {
 	struct drm_i915_private *dev_priv =
 		container_of(nb, struct drm_i915_private, mm.oom_notifier);
-	struct shrinker_lock_uninterruptible slu;
 	struct drm_i915_gem_object *obj;
 	unsigned long unevictable, bound, unbound, freed_pages;
+	bool unlock;
 
-	if (!i915_gem_shrinker_lock_uninterruptible(dev_priv, &slu, 5000))
+	if (!shrinker_lock_uninterruptible(dev_priv, &unlock, 5000))
 		return NOTIFY_DONE;
 
 	freed_pages = i915_gem_shrink_all(dev_priv);
@@ -410,7 +409,7 @@ i915_gem_shrinker_oom(struct notifier_block *nb, unsigned long event, void *ptr)
 			bound += obj->base.size >> PAGE_SHIFT;
 	}
 
-	i915_gem_shrinker_unlock_uninterruptible(dev_priv, &slu);
+	shrinker_unlock(dev_priv, unlock);
 
 	if (freed_pages || unbound || bound)
 		pr_info("Purging GPU memory, %lu pages freed, "
@@ -430,12 +429,12 @@ i915_gem_shrinker_vmap(struct notifier_block *nb, unsigned long event, void *ptr
 {
 	struct drm_i915_private *dev_priv =
 		container_of(nb, struct drm_i915_private, mm.vmap_notifier);
-	struct shrinker_lock_uninterruptible slu;
 	struct i915_vma *vma, *next;
 	unsigned long freed_pages = 0;
+	bool unlock;
 	int ret;
 
-	if (!i915_gem_shrinker_lock_uninterruptible(dev_priv, &slu, 5000))
+	if (!shrinker_lock_uninterruptible(dev_priv, &unlock, 5000))
 		return NOTIFY_DONE;
 
 	/* Force everything onto the inactive lists */
@@ -460,7 +459,7 @@ i915_gem_shrinker_vmap(struct notifier_block *nb, unsigned long event, void *ptr
 	}
 
 out:
-	i915_gem_shrinker_unlock_uninterruptible(dev_priv, &slu);
+	shrinker_unlock(dev_priv, unlock);
 
 	*(unsigned long *)ptr += freed_pages;
 	return NOTIFY_DONE;
diff --git a/drivers/gpu/drm/i915/i915_gem_stolen.c b/drivers/gpu/drm/i915/i915_gem_stolen.c
index f3abdc27c5dd..681db6083f4d 100644
--- a/drivers/gpu/drm/i915/i915_gem_stolen.c
+++ b/drivers/gpu/drm/i915/i915_gem_stolen.c
@@ -414,12 +414,10 @@ int i915_gem_init_stolen(struct drm_i915_private *dev_priv)
 		return 0;
 	}
 
-#ifdef CONFIG_INTEL_IOMMU
-	if (intel_iommu_gfx_mapped && INTEL_GEN(dev_priv) < 8) {
+	if (intel_vtd_active() && INTEL_GEN(dev_priv) < 8) {
 		DRM_INFO("DMAR active, disabling use of stolen memory\n");
 		return 0;
 	}
-#endif
 
 	if (ggtt->stolen_size == 0)
 		return 0;
diff --git a/drivers/gpu/drm/i915/i915_gem_timeline.c b/drivers/gpu/drm/i915/i915_gem_timeline.c
index b596ca7ee058..c597ce277a04 100644
--- a/drivers/gpu/drm/i915/i915_gem_timeline.c
+++ b/drivers/gpu/drm/i915/i915_gem_timeline.c
@@ -23,6 +23,32 @@
  */
 
 #include "i915_drv.h"
+#include "i915_syncmap.h"
+
+static void __intel_timeline_init(struct intel_timeline *tl,
+				  struct i915_gem_timeline *parent,
+				  u64 context,
+				  struct lock_class_key *lockclass,
+				  const char *lockname)
+{
+	tl->fence_context = context;
+	tl->common = parent;
+#ifdef CONFIG_DEBUG_SPINLOCK
+	__raw_spin_lock_init(&tl->lock.rlock, lockname, lockclass);
+#else
+	spin_lock_init(&tl->lock);
+#endif
+	init_request_active(&tl->last_request, NULL);
+	INIT_LIST_HEAD(&tl->requests);
+	i915_syncmap_init(&tl->sync);
+}
+
+static void __intel_timeline_fini(struct intel_timeline *tl)
+{
+	GEM_BUG_ON(!list_empty(&tl->requests));
+
+	i915_syncmap_free(&tl->sync);
+}
 
 static int __i915_gem_timeline_init(struct drm_i915_private *i915,
 				    struct i915_gem_timeline *timeline,
@@ -35,6 +61,14 @@ static int __i915_gem_timeline_init(struct drm_i915_private *i915,
 
 	lockdep_assert_held(&i915->drm.struct_mutex);
 
+	/*
+	 * Ideally we want a set of engines on a single leaf as we expect
+	 * to mostly be tracking synchronisation between engines. It is not
+	 * a huge issue if this is not the case, but we may want to mitigate
+	 * any page crossing penalties if they become an issue.
+	 */
+	BUILD_BUG_ON(KSYNCMAP < I915_NUM_ENGINES);
+
 	timeline->i915 = i915;
 	timeline->name = kstrdup(name ?: "[kernel]", GFP_KERNEL);
 	if (!timeline->name)
@@ -44,19 +78,10 @@ static int __i915_gem_timeline_init(struct drm_i915_private *i915,
 
 	/* Called during early_init before we know how many engines there are */
 	fences = dma_fence_context_alloc(ARRAY_SIZE(timeline->engine));
-	for (i = 0; i < ARRAY_SIZE(timeline->engine); i++) {
-		struct intel_timeline *tl = &timeline->engine[i];
-
-		tl->fence_context = fences++;
-		tl->common = timeline;
-#ifdef CONFIG_DEBUG_SPINLOCK
-		__raw_spin_lock_init(&tl->lock.rlock, lockname, lockclass);
-#else
-		spin_lock_init(&tl->lock);
-#endif
-		init_request_active(&tl->last_request, NULL);
-		INIT_LIST_HEAD(&tl->requests);
-	}
+	for (i = 0; i < ARRAY_SIZE(timeline->engine); i++)
+		__intel_timeline_init(&timeline->engine[i],
+				      timeline, fences++,
+				      lockclass, lockname);
 
 	return 0;
 }
@@ -81,18 +106,52 @@ int i915_gem_timeline_init__global(struct drm_i915_private *i915)
 					&class, "&global_timeline->lock");
 }
 
+/**
+ * i915_gem_timelines_mark_idle -- called when the driver idles
+ * @i915 - the drm_i915_private device
+ *
+ * When the driver is completely idle, we know that all of our sync points
+ * have been signaled and our tracking is then entirely redundant. Any request
+ * to wait upon an older sync point will be completed instantly as we know
+ * the fence is signaled and therefore we will not even look them up in the
+ * sync point map.
+ */
+void i915_gem_timelines_mark_idle(struct drm_i915_private *i915)
+{
+	struct i915_gem_timeline *timeline;
+	int i;
+
+	lockdep_assert_held(&i915->drm.struct_mutex);
+
+	list_for_each_entry(timeline, &i915->gt.timelines, link) {
+		for (i = 0; i < ARRAY_SIZE(timeline->engine); i++) {
+			struct intel_timeline *tl = &timeline->engine[i];
+
+			/*
+			 * All known fences are completed so we can scrap
+			 * the current sync point tracking and start afresh,
+			 * any attempt to wait upon a previous sync point
+			 * will be skipped as the fence was signaled.
+			 */
+			i915_syncmap_free(&tl->sync);
+		}
+	}
+}
+
 void i915_gem_timeline_fini(struct i915_gem_timeline *timeline)
 {
 	int i;
 
 	lockdep_assert_held(&timeline->i915->drm.struct_mutex);
 
-	for (i = 0; i < ARRAY_SIZE(timeline->engine); i++) {
-		struct intel_timeline *tl = &timeline->engine[i];
-
-		GEM_BUG_ON(!list_empty(&tl->requests));
-	}
+	for (i = 0; i < ARRAY_SIZE(timeline->engine); i++)
+		__intel_timeline_fini(&timeline->engine[i]);
 
 	list_del(&timeline->link);
 	kfree(timeline->name);
 }
+
+#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
+#include "selftests/mock_timeline.c"
+#include "selftests/i915_gem_timeline.c"
+#endif
diff --git a/drivers/gpu/drm/i915/i915_gem_timeline.h b/drivers/gpu/drm/i915/i915_gem_timeline.h
index 6c53e14cab2a..bfb5eb94c64d 100644
--- a/drivers/gpu/drm/i915/i915_gem_timeline.h
+++ b/drivers/gpu/drm/i915/i915_gem_timeline.h
@@ -27,7 +27,9 @@
 
 #include <linux/list.h>
 
+#include "i915_utils.h"
 #include "i915_gem_request.h"
+#include "i915_syncmap.h"
 
 struct i915_gem_timeline;
 
@@ -55,7 +57,25 @@ struct intel_timeline {
 	 * struct_mutex.
 	 */
 	struct i915_gem_active last_request;
-	u32 sync_seqno[I915_NUM_ENGINES];
+
+	/**
+	 * We track the most recent seqno that we wait on in every context so
+	 * that we only have to emit a new await and dependency on a more
+	 * recent sync point. As the contexts may be executed out-of-order, we
+	 * have to track each individually and can not rely on an absolute
+	 * global_seqno. When we know that all tracked fences are completed
+	 * (i.e. when the driver is idle), we know that the syncmap is
+	 * redundant and we can discard it without loss of generality.
+	 */
+	struct i915_syncmap *sync;
+	/**
+	 * Separately to the inter-context seqno map above, we track the last
+	 * barrier (e.g. semaphore wait) to the global engine timelines. Note
+	 * that this tracks global_seqno rather than the context.seqno, and
+	 * so it is subject to the limitations of hw wraparound and that we
+	 * may need to revoke global_seqno (on pre-emption).
+	 */
+	u32 global_sync[I915_NUM_ENGINES];
 
 	struct i915_gem_timeline *common;
 };
@@ -73,6 +93,31 @@ int i915_gem_timeline_init(struct drm_i915_private *i915,
 			   struct i915_gem_timeline *tl,
 			   const char *name);
 int i915_gem_timeline_init__global(struct drm_i915_private *i915);
+void i915_gem_timelines_mark_idle(struct drm_i915_private *i915);
 void i915_gem_timeline_fini(struct i915_gem_timeline *tl);
 
+static inline int __intel_timeline_sync_set(struct intel_timeline *tl,
+					    u64 context, u32 seqno)
+{
+	return i915_syncmap_set(&tl->sync, context, seqno);
+}
+
+static inline int intel_timeline_sync_set(struct intel_timeline *tl,
+					  const struct dma_fence *fence)
+{
+	return __intel_timeline_sync_set(tl, fence->context, fence->seqno);
+}
+
+static inline bool __intel_timeline_sync_is_later(struct intel_timeline *tl,
+						  u64 context, u32 seqno)
+{
+	return i915_syncmap_is_later(&tl->sync, context, seqno);
+}
+
+static inline bool intel_timeline_sync_is_later(struct intel_timeline *tl,
+						const struct dma_fence *fence)
+{
+	return __intel_timeline_sync_is_later(tl, fence->context, fence->seqno);
+}
+
 #endif
diff --git a/drivers/gpu/drm/i915/i915_gem_userptr.c b/drivers/gpu/drm/i915/i915_gem_userptr.c
index 58ccf8b8ca1c..1a0ce1dc68f5 100644
--- a/drivers/gpu/drm/i915/i915_gem_userptr.c
+++ b/drivers/gpu/drm/i915/i915_gem_userptr.c
@@ -507,7 +507,7 @@ __i915_gem_userptr_get_pages_worker(struct work_struct *_work)
 	ret = -ENOMEM;
 	pinned = 0;
 
-	pvec = drm_malloc_gfp(npages, sizeof(struct page *), GFP_TEMPORARY);
+	pvec = kvmalloc_array(npages, sizeof(struct page *), GFP_TEMPORARY);
 	if (pvec != NULL) {
 		struct mm_struct *mm = obj->userptr.mm->mm;
 		unsigned int flags = 0;
@@ -555,7 +555,7 @@ __i915_gem_userptr_get_pages_worker(struct work_struct *_work)
 	mutex_unlock(&obj->mm.lock);
 
 	release_pages(pvec, pinned, 0);
-	drm_free_large(pvec);
+	kvfree(pvec);
 
 	i915_gem_object_put(obj);
 	put_task_struct(work->task);
@@ -642,7 +642,7 @@ i915_gem_userptr_get_pages(struct drm_i915_gem_object *obj)
 	pinned = 0;
 
 	if (mm == current->mm) {
-		pvec = drm_malloc_gfp(num_pages, sizeof(struct page *),
+		pvec = kvmalloc_array(num_pages, sizeof(struct page *),
 				      GFP_TEMPORARY |
 				      __GFP_NORETRY |
 				      __GFP_NOWARN);
@@ -669,7 +669,7 @@ i915_gem_userptr_get_pages(struct drm_i915_gem_object *obj)
 
 	if (IS_ERR(pages))
 		release_pages(pvec, pinned, 0);
-	drm_free_large(pvec);
+	kvfree(pvec);
 
 	return pages;
 }
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index 8effc59f5cb5..e18f350bc364 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -712,6 +712,10 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m,
 			print_error_obj(m, dev_priv->engine[i], NULL, obj);
 		}
 
+		for (j = 0; j < ee->user_bo_count; j++)
+			print_error_obj(m, dev_priv->engine[i],
+					"user", ee->user_bo[j]);
+
 		if (ee->num_requests) {
 			err_printf(m, "%s --- %d requests\n",
 				   dev_priv->engine[i]->name,
@@ -825,11 +829,15 @@ void __i915_gpu_state_free(struct kref *error_ref)
 {
 	struct i915_gpu_state *error =
 		container_of(error_ref, typeof(*error), ref);
-	int i;
+	long i, j;
 
 	for (i = 0; i < ARRAY_SIZE(error->engine); i++) {
 		struct drm_i915_error_engine *ee = &error->engine[i];
 
+		for (j = 0; j < ee->user_bo_count; j++)
+			i915_error_object_free(ee->user_bo[j]);
+		kfree(ee->user_bo);
+
 		i915_error_object_free(ee->batchbuffer);
 		i915_error_object_free(ee->wa_batchbuffer);
 		i915_error_object_free(ee->ringbuffer);
@@ -1316,12 +1324,17 @@ static void engine_record_requests(struct intel_engine_cs *engine,
 static void error_record_engine_execlists(struct intel_engine_cs *engine,
 					  struct drm_i915_error_engine *ee)
 {
+	const struct execlist_port *port = engine->execlist_port;
 	unsigned int n;
 
-	for (n = 0; n < ARRAY_SIZE(engine->execlist_port); n++)
-		if (engine->execlist_port[n].request)
-			record_request(engine->execlist_port[n].request,
-				       &ee->execlist[n]);
+	for (n = 0; n < ARRAY_SIZE(engine->execlist_port); n++) {
+		struct drm_i915_gem_request *rq = port_request(&port[n]);
+
+		if (!rq)
+			break;
+
+		record_request(rq, &ee->execlist[n]);
+	}
 }
 
 static void record_context(struct drm_i915_error_context *e,
@@ -1346,6 +1359,35 @@ static void record_context(struct drm_i915_error_context *e,
 	e->active = ctx->active_count;
 }
 
+static void request_record_user_bo(struct drm_i915_gem_request *request,
+				   struct drm_i915_error_engine *ee)
+{
+	struct i915_gem_capture_list *c;
+	struct drm_i915_error_object **bo;
+	long count;
+
+	count = 0;
+	for (c = request->capture_list; c; c = c->next)
+		count++;
+
+	bo = NULL;
+	if (count)
+		bo = kcalloc(count, sizeof(*bo), GFP_ATOMIC);
+	if (!bo)
+		return;
+
+	count = 0;
+	for (c = request->capture_list; c; c = c->next) {
+		bo[count] = i915_error_object_create(request->i915, c->vma);
+		if (!bo[count])
+			break;
+		count++;
+	}
+
+	ee->user_bo = bo;
+	ee->user_bo_count = count;
+}
+
 static void i915_gem_record_rings(struct drm_i915_private *dev_priv,
 				  struct i915_gpu_state *error)
 {
@@ -1392,6 +1434,7 @@ static void i915_gem_record_rings(struct drm_i915_private *dev_priv,
 				ee->wa_batchbuffer =
 					i915_error_object_create(dev_priv,
 								 engine->scratch);
+			request_record_user_bo(request, ee);
 
 			ee->ctx =
 				i915_error_object_create(dev_priv,
@@ -1560,6 +1603,9 @@ static void i915_capture_reg_state(struct drm_i915_private *dev_priv,
 		error->done_reg = I915_READ(DONE_REG);
 	}
 
+	if (INTEL_GEN(dev_priv) >= 5)
+		error->ccid = I915_READ(CCID);
+
 	/* 3: Feature specific registers */
 	if (IS_GEN6(dev_priv) || IS_GEN7(dev_priv)) {
 		error->gam_ecochk = I915_READ(GAM_ECOCHK);
@@ -1567,9 +1613,6 @@ static void i915_capture_reg_state(struct drm_i915_private *dev_priv,
 	}
 
 	/* 4: Everything else */
-	if (HAS_HW_CONTEXTS(dev_priv))
-		error->ccid = I915_READ(CCID);
-
 	if (INTEL_GEN(dev_priv) >= 8) {
 		error->ier = I915_READ(GEN8_DE_MISC_IER);
 		for (i = 0; i < 4; i++)
diff --git a/drivers/gpu/drm/i915/i915_guc_submission.c b/drivers/gpu/drm/i915/i915_guc_submission.c
index 1642fff9cf13..48a1e9349a2c 100644
--- a/drivers/gpu/drm/i915/i915_guc_submission.c
+++ b/drivers/gpu/drm/i915/i915_guc_submission.c
@@ -105,7 +105,7 @@ static int __reserve_doorbell(struct i915_guc_client *client)
 		end += offset;
 	}
 
-	id = find_next_zero_bit(client->guc->doorbell_bitmap, offset, end);
+	id = find_next_zero_bit(client->guc->doorbell_bitmap, end, offset);
 	if (id == end)
 		return -ENOSPC;
 
@@ -480,9 +480,7 @@ static void guc_wq_item_append(struct i915_guc_client *client,
 	GEM_BUG_ON(freespace < wqi_size);
 
 	/* The GuC firmware wants the tail index in QWords, not bytes */
-	tail = rq->tail;
-	assert_ring_tail_valid(rq->ring, rq->tail);
-	tail >>= 3;
+	tail = intel_ring_set_tail(rq->ring, rq->tail) >> 3;
 	GEM_BUG_ON(tail > WQ_RING_TAIL_MAX);
 
 	/* For now workqueue item is 4 DWs; workqueue buffer is 2 pages. So we
@@ -616,12 +614,6 @@ static void __i915_guc_submit(struct drm_i915_gem_request *rq)
 	b_ret = guc_ring_doorbell(client);
 
 	client->submissions[engine_id] += 1;
-	client->retcode = b_ret;
-	if (b_ret)
-		client->b_fail += 1;
-
-	guc->submissions[engine_id] += 1;
-	guc->last_seqno[engine_id] = rq->global_seqno;
 
 	spin_unlock_irqrestore(&client->wq_lock, flags);
 }
@@ -651,47 +643,68 @@ static void nested_enable_signaling(struct drm_i915_gem_request *rq)
 	trace_dma_fence_enable_signal(&rq->fence);
 
 	spin_lock_nested(&rq->lock, SINGLE_DEPTH_NESTING);
-	intel_engine_enable_signaling(rq);
+	intel_engine_enable_signaling(rq, true);
 	spin_unlock(&rq->lock);
 }
 
+static void port_assign(struct execlist_port *port,
+			struct drm_i915_gem_request *rq)
+{
+	GEM_BUG_ON(rq == port_request(port));
+
+	if (port_isset(port))
+		i915_gem_request_put(port_request(port));
+
+	port_set(port, i915_gem_request_get(rq));
+	nested_enable_signaling(rq);
+}
+
 static bool i915_guc_dequeue(struct intel_engine_cs *engine)
 {
 	struct execlist_port *port = engine->execlist_port;
-	struct drm_i915_gem_request *last = port[0].request;
+	struct drm_i915_gem_request *last = port_request(port);
 	struct rb_node *rb;
 	bool submit = false;
 
 	spin_lock_irq(&engine->timeline->lock);
 	rb = engine->execlist_first;
+	GEM_BUG_ON(rb_first(&engine->execlist_queue) != rb);
 	while (rb) {
-		struct drm_i915_gem_request *rq =
-			rb_entry(rb, typeof(*rq), priotree.node);
-
-		if (last && rq->ctx != last->ctx) {
-			if (port != engine->execlist_port)
-				break;
-
-			i915_gem_request_assign(&port->request, last);
-			nested_enable_signaling(last);
-			port++;
+		struct i915_priolist *p = rb_entry(rb, typeof(*p), node);
+		struct drm_i915_gem_request *rq, *rn;
+
+		list_for_each_entry_safe(rq, rn, &p->requests, priotree.link) {
+			if (last && rq->ctx != last->ctx) {
+				if (port != engine->execlist_port) {
+					__list_del_many(&p->requests,
+							&rq->priotree.link);
+					goto done;
+				}
+
+				if (submit)
+					port_assign(port, last);
+				port++;
+			}
+
+			INIT_LIST_HEAD(&rq->priotree.link);
+			rq->priotree.priority = INT_MAX;
+
+			i915_guc_submit(rq);
+			trace_i915_gem_request_in(rq, port_index(port, engine));
+			last = rq;
+			submit = true;
 		}
 
 		rb = rb_next(rb);
-		rb_erase(&rq->priotree.node, &engine->execlist_queue);
-		RB_CLEAR_NODE(&rq->priotree.node);
-		rq->priotree.priority = INT_MAX;
-
-		i915_guc_submit(rq);
-		trace_i915_gem_request_in(rq, port - engine->execlist_port);
-		last = rq;
-		submit = true;
-	}
-	if (submit) {
-		i915_gem_request_assign(&port->request, last);
-		nested_enable_signaling(last);
-		engine->execlist_first = rb;
+		rb_erase(&p->node, &engine->execlist_queue);
+		INIT_LIST_HEAD(&p->requests);
+		if (p->priority != I915_PRIORITY_NORMAL)
+			kmem_cache_free(engine->i915->priorities, p);
 	}
+done:
+	engine->execlist_first = rb;
+	if (submit)
+		port_assign(port, last);
 	spin_unlock_irq(&engine->timeline->lock);
 
 	return submit;
@@ -705,17 +718,19 @@ static void i915_guc_irq_handler(unsigned long data)
 	bool submit;
 
 	do {
-		rq = port[0].request;
+		rq = port_request(&port[0]);
 		while (rq && i915_gem_request_completed(rq)) {
 			trace_i915_gem_request_out(rq);
 			i915_gem_request_put(rq);
-			port[0].request = port[1].request;
-			port[1].request = NULL;
-			rq = port[0].request;
+
+			port[0] = port[1];
+			memset(&port[1], 0, sizeof(port[1]));
+
+			rq = port_request(&port[0]);
 		}
 
 		submit = false;
-		if (!port[1].request)
+		if (!port_count(&port[1]))
 			submit = i915_guc_dequeue(engine);
 	} while (submit);
 }
@@ -1053,8 +1068,7 @@ static int guc_ads_create(struct intel_guc *guc)
 		dev_priv->engine[RCS]->status_page.ggtt_offset;
 
 	for_each_engine(engine, dev_priv, id)
-		blob->ads.eng_state_size[engine->guc_id] =
-			intel_lr_context_size(engine);
+		blob->ads.eng_state_size[engine->guc_id] = engine->context_size;
 
 	base = guc_ggtt_offset(vma);
 	blob->ads.scheduler_policies = base + ptr_offset(blob, policies);
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index 190f6aa5d15e..4cd9ee1ba332 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -720,9 +720,7 @@ static u32 i915_get_vblank_counter(struct drm_device *dev, unsigned int pipe)
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	i915_reg_t high_frame, low_frame;
 	u32 high1, high2, low, pixel, vbl_start, hsync_start, htotal;
-	struct intel_crtc *intel_crtc = intel_get_crtc_for_pipe(dev_priv,
-								pipe);
-	const struct drm_display_mode *mode = &intel_crtc->base.hwmode;
+	const struct drm_display_mode *mode = &dev->vblank[pipe].hwmode;
 	unsigned long irqflags;
 
 	htotal = mode->crtc_htotal;
@@ -779,13 +777,17 @@ static int __intel_get_crtc_scanline(struct intel_crtc *crtc)
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	const struct drm_display_mode *mode = &crtc->base.hwmode;
+	const struct drm_display_mode *mode;
+	struct drm_vblank_crtc *vblank;
 	enum pipe pipe = crtc->pipe;
 	int position, vtotal;
 
 	if (!crtc->active)
 		return -1;
 
+	vblank = &crtc->base.dev->vblank[drm_crtc_index(&crtc->base)];
+	mode = &vblank->hwmode;
+
 	vtotal = mode->crtc_vtotal;
 	if (mode->flags & DRM_MODE_FLAG_INTERLACE)
 		vtotal /= 2;
@@ -827,10 +829,10 @@ static int __intel_get_crtc_scanline(struct intel_crtc *crtc)
 	return (position + crtc->scanline_offset) % vtotal;
 }
 
-static int i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe,
-				    unsigned int flags, int *vpos, int *hpos,
-				    ktime_t *stime, ktime_t *etime,
-				    const struct drm_display_mode *mode)
+static bool i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe,
+				     bool in_vblank_irq, int *vpos, int *hpos,
+				     ktime_t *stime, ktime_t *etime,
+				     const struct drm_display_mode *mode)
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_crtc *intel_crtc = intel_get_crtc_for_pipe(dev_priv,
@@ -838,13 +840,12 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe,
 	int position;
 	int vbl_start, vbl_end, hsync_start, htotal, vtotal;
 	bool in_vbl = true;
-	int ret = 0;
 	unsigned long irqflags;
 
 	if (WARN_ON(!mode->crtc_clock)) {
 		DRM_DEBUG_DRIVER("trying to get scanoutpos for disabled "
 				 "pipe %c\n", pipe_name(pipe));
-		return 0;
+		return false;
 	}
 
 	htotal = mode->crtc_htotal;
@@ -859,8 +860,6 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe,
 		vtotal /= 2;
 	}
 
-	ret |= DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_ACCURATE;
-
 	/*
 	 * Lock uncore.lock, as we will do multiple timing critical raw
 	 * register reads, potentially with preemption disabled, so the
@@ -944,11 +943,7 @@ static int i915_get_crtc_scanoutpos(struct drm_device *dev, unsigned int pipe,
 		*hpos = position - (*vpos * htotal);
 	}
 
-	/* In vblank? */
-	if (in_vbl)
-		ret |= DRM_SCANOUTPOS_IN_VBLANK;
-
-	return ret;
+	return true;
 }
 
 int intel_get_crtc_scanline(struct intel_crtc *crtc)
@@ -964,37 +959,6 @@ int intel_get_crtc_scanline(struct intel_crtc *crtc)
 	return position;
 }
 
-static int i915_get_vblank_timestamp(struct drm_device *dev, unsigned int pipe,
-			      int *max_error,
-			      struct timeval *vblank_time,
-			      unsigned flags)
-{
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc *crtc;
-
-	if (pipe >= INTEL_INFO(dev_priv)->num_pipes) {
-		DRM_ERROR("Invalid crtc %u\n", pipe);
-		return -EINVAL;
-	}
-
-	/* Get drm_crtc to timestamp: */
-	crtc = intel_get_crtc_for_pipe(dev_priv, pipe);
-	if (crtc == NULL) {
-		DRM_ERROR("Invalid crtc %u\n", pipe);
-		return -EINVAL;
-	}
-
-	if (!crtc->base.hwmode.crtc_clock) {
-		DRM_DEBUG_KMS("crtc %u is disabled\n", pipe);
-		return -EBUSY;
-	}
-
-	/* Helper routine in DRM core does all the work: */
-	return drm_calc_vbltimestamp_from_scanoutpos(dev, pipe, max_error,
-						     vblank_time, flags,
-						     &crtc->base.hwmode);
-}
-
 static void ironlake_rps_change_irq_handler(struct drm_i915_private *dev_priv)
 {
 	u32 busy_up, busy_down, max_avg, min_avg;
@@ -1236,7 +1200,7 @@ out:
 static void ivybridge_parity_work(struct work_struct *work)
 {
 	struct drm_i915_private *dev_priv =
-		container_of(work, struct drm_i915_private, l3_parity.error_work);
+		container_of(work, typeof(*dev_priv), l3_parity.error_work);
 	u32 error_status, row, bank, subbank;
 	char *parity_event[6];
 	uint32_t misccpctl;
@@ -1353,14 +1317,16 @@ static void snb_gt_irq_handler(struct drm_i915_private *dev_priv,
 		ivybridge_parity_error_irq_handler(dev_priv, gt_iir);
 }
 
-static __always_inline void
+static void
 gen8_cs_irq_handler(struct intel_engine_cs *engine, u32 iir, int test_shift)
 {
 	bool tasklet = false;
 
 	if (iir & (GT_CONTEXT_SWITCH_INTERRUPT << test_shift)) {
-		set_bit(ENGINE_IRQ_EXECLIST, &engine->irq_posted);
-		tasklet = true;
+		if (port_count(&engine->execlist_port[0])) {
+			__set_bit(ENGINE_IRQ_EXECLIST, &engine->irq_posted);
+			tasklet = true;
+		}
 	}
 
 	if (iir & (GT_RENDER_USER_INTERRUPT << test_shift)) {
@@ -2582,7 +2548,8 @@ gen8_de_irq_handler(struct drm_i915_private *dev_priv, u32 master_ctl)
 			I915_WRITE(SDEIIR, iir);
 			ret = IRQ_HANDLED;
 
-			if (HAS_PCH_SPT(dev_priv) || HAS_PCH_KBP(dev_priv))
+			if (HAS_PCH_SPT(dev_priv) || HAS_PCH_KBP(dev_priv) ||
+			    HAS_PCH_CNP(dev_priv))
 				spt_irq_handler(dev_priv, iir);
 			else
 				cpt_irq_handler(dev_priv, iir);
@@ -4230,11 +4197,15 @@ static void i965_irq_uninstall(struct drm_device * dev)
 void intel_irq_init(struct drm_i915_private *dev_priv)
 {
 	struct drm_device *dev = &dev_priv->drm;
+	int i;
 
 	intel_hpd_init_work(dev_priv);
 
 	INIT_WORK(&dev_priv->rps.work, gen6_pm_rps_work);
+
 	INIT_WORK(&dev_priv->l3_parity.error_work, ivybridge_parity_work);
+	for (i = 0; i < MAX_L3_SLICES; ++i)
+		dev_priv->l3_parity.remap_info[i] = NULL;
 
 	if (HAS_GUC_SCHED(dev_priv))
 		dev_priv->pm_guc_events = GEN9_GUC_TO_HOST_INT_EVENT;
@@ -4291,7 +4262,7 @@ void intel_irq_init(struct drm_i915_private *dev_priv)
 
 	dev_priv->hotplug.hpd_storm_threshold = HPD_STORM_DEFAULT_THRESHOLD;
 
-	dev->driver->get_vblank_timestamp = i915_get_vblank_timestamp;
+	dev->driver->get_vblank_timestamp = drm_calc_vbltimestamp_from_scanoutpos;
 	dev->driver->get_scanout_position = i915_get_crtc_scanoutpos;
 
 	if (IS_CHERRYVIEW(dev_priv)) {
@@ -4319,7 +4290,8 @@ void intel_irq_init(struct drm_i915_private *dev_priv)
 		dev->driver->disable_vblank = gen8_disable_vblank;
 		if (IS_GEN9_LP(dev_priv))
 			dev_priv->display.hpd_irq_setup = bxt_hpd_irq_setup;
-		else if (HAS_PCH_SPT(dev_priv) || HAS_PCH_KBP(dev_priv))
+		else if (HAS_PCH_SPT(dev_priv) || HAS_PCH_KBP(dev_priv) ||
+			 HAS_PCH_CNP(dev_priv))
 			dev_priv->display.hpd_irq_setup = spt_hpd_irq_setup;
 		else
 			dev_priv->display.hpd_irq_setup = ilk_hpd_irq_setup;
@@ -4360,6 +4332,20 @@ void intel_irq_init(struct drm_i915_private *dev_priv)
 }
 
 /**
+ * intel_irq_fini - deinitializes IRQ support
+ * @i915: i915 device instance
+ *
+ * This function deinitializes all the IRQ support.
+ */
+void intel_irq_fini(struct drm_i915_private *i915)
+{
+	int i;
+
+	for (i = 0; i < MAX_L3_SLICES; ++i)
+		kfree(i915->l3_parity.remap_info[i]);
+}
+
+/**
  * intel_irq_install - enables the hardware interrupt
  * @dev_priv: i915 device instance
  *
diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c
index 1a78363c7f4a..03b5fe3e3036 100644
--- a/drivers/gpu/drm/i915/i915_pci.c
+++ b/drivers/gpu/drm/i915/i915_pci.c
@@ -220,7 +220,6 @@ static const struct intel_device_info intel_ironlake_m_info = {
 	.has_rc6 = 1, \
 	.has_rc6p = 1, \
 	.has_gmbus_irq = 1, \
-	.has_hw_contexts = 1, \
 	.has_aliasing_ppgtt = 1, \
 	GEN_DEFAULT_PIPEOFFSETS, \
 	CURSOR_OFFSETS
@@ -245,7 +244,6 @@ static const struct intel_device_info intel_sandybridge_m_info = {
 	.has_rc6 = 1, \
 	.has_rc6p = 1, \
 	.has_gmbus_irq = 1, \
-	.has_hw_contexts = 1, \
 	.has_aliasing_ppgtt = 1, \
 	.has_full_ppgtt = 1, \
 	GEN_DEFAULT_PIPEOFFSETS, \
@@ -280,7 +278,6 @@ static const struct intel_device_info intel_valleyview_info = {
 	.has_runtime_pm = 1,
 	.has_rc6 = 1,
 	.has_gmbus_irq = 1,
-	.has_hw_contexts = 1,
 	.has_gmch_display = 1,
 	.has_hotplug = 1,
 	.has_aliasing_ppgtt = 1,
@@ -315,16 +312,17 @@ static const struct intel_device_info intel_haswell_info = {
 	.has_full_48bit_ppgtt = 1, \
 	.has_64bit_reloc = 1
 
+#define BDW_PLATFORM \
+	BDW_FEATURES, \
+	.gen = 8, \
+	.platform = INTEL_BROADWELL
+
 static const struct intel_device_info intel_broadwell_info = {
-	BDW_FEATURES,
-	.gen = 8,
-	.platform = INTEL_BROADWELL,
+	BDW_PLATFORM,
 };
 
 static const struct intel_device_info intel_broadwell_gt3_info = {
-	BDW_FEATURES,
-	.gen = 8,
-	.platform = INTEL_BROADWELL,
+	BDW_PLATFORM,
 	.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING | BSD2_RING,
 };
 
@@ -340,7 +338,6 @@ static const struct intel_device_info intel_cherryview_info = {
 	.has_resource_streamer = 1,
 	.has_rc6 = 1,
 	.has_gmbus_irq = 1,
-	.has_hw_contexts = 1,
 	.has_logical_ring_contexts = 1,
 	.has_gmch_display = 1,
 	.has_aliasing_ppgtt = 1,
@@ -351,22 +348,20 @@ static const struct intel_device_info intel_cherryview_info = {
 	CHV_COLORS,
 };
 
+#define SKL_PLATFORM \
+	BDW_FEATURES, \
+	.gen = 9, \
+	.platform = INTEL_SKYLAKE, \
+	.has_csr = 1, \
+	.has_guc = 1, \
+	.ddb_size = 896
+
 static const struct intel_device_info intel_skylake_info = {
-	BDW_FEATURES,
-	.platform = INTEL_SKYLAKE,
-	.gen = 9,
-	.has_csr = 1,
-	.has_guc = 1,
-	.ddb_size = 896,
+	SKL_PLATFORM,
 };
 
 static const struct intel_device_info intel_skylake_gt3_info = {
-	BDW_FEATURES,
-	.platform = INTEL_SKYLAKE,
-	.gen = 9,
-	.has_csr = 1,
-	.has_guc = 1,
-	.ddb_size = 896,
+	SKL_PLATFORM,
 	.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING | BSD2_RING,
 };
 
@@ -387,7 +382,6 @@ static const struct intel_device_info intel_skylake_gt3_info = {
 	.has_rc6 = 1, \
 	.has_dp_mst = 1, \
 	.has_gmbus_irq = 1, \
-	.has_hw_contexts = 1, \
 	.has_logical_ring_contexts = 1, \
 	.has_guc = 1, \
 	.has_aliasing_ppgtt = 1, \
@@ -406,30 +400,53 @@ static const struct intel_device_info intel_broxton_info = {
 static const struct intel_device_info intel_geminilake_info = {
 	GEN9_LP_FEATURES,
 	.platform = INTEL_GEMINILAKE,
-	.is_alpha_support = 1,
 	.ddb_size = 1024,
 	.color = { .degamma_lut_size = 0, .gamma_lut_size = 1024 }
 };
 
+#define KBL_PLATFORM \
+	BDW_FEATURES, \
+	.gen = 9, \
+	.platform = INTEL_KABYLAKE, \
+	.has_csr = 1, \
+	.has_guc = 1, \
+	.ddb_size = 896
+
 static const struct intel_device_info intel_kabylake_info = {
-	BDW_FEATURES,
-	.platform = INTEL_KABYLAKE,
-	.gen = 9,
-	.has_csr = 1,
-	.has_guc = 1,
-	.ddb_size = 896,
+	KBL_PLATFORM,
 };
 
 static const struct intel_device_info intel_kabylake_gt3_info = {
-	BDW_FEATURES,
-	.platform = INTEL_KABYLAKE,
-	.gen = 9,
-	.has_csr = 1,
-	.has_guc = 1,
-	.ddb_size = 896,
+	KBL_PLATFORM,
 	.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING | BSD2_RING,
 };
 
+#define CFL_PLATFORM \
+	.is_alpha_support = 1, \
+	BDW_FEATURES, \
+	.gen = 9, \
+	.platform = INTEL_COFFEELAKE, \
+	.has_csr = 1, \
+	.has_guc = 1, \
+	.ddb_size = 896
+
+static const struct intel_device_info intel_coffeelake_info = {
+	CFL_PLATFORM,
+};
+
+static const struct intel_device_info intel_coffeelake_gt3_info = {
+	CFL_PLATFORM,
+	.ring_mask = RENDER_RING | BSD_RING | BLT_RING | VEBOX_RING | BSD2_RING,
+};
+
+static const struct intel_device_info intel_cannonlake_info = {
+	BDW_FEATURES,
+	.is_alpha_support = 1,
+	.platform = INTEL_CANNONLAKE,
+	.gen = 10,
+	.ddb_size = 1024,
+};
+
 /*
  * Make sure any device matches here are from most specific to most
  * general.  For example, since the Quanta match is based on the subsystem
@@ -474,6 +491,10 @@ static const struct pci_device_id pciidlist[] = {
 	INTEL_KBL_GT2_IDS(&intel_kabylake_info),
 	INTEL_KBL_GT3_IDS(&intel_kabylake_gt3_info),
 	INTEL_KBL_GT4_IDS(&intel_kabylake_gt3_info),
+	INTEL_CFL_S_IDS(&intel_coffeelake_info),
+	INTEL_CFL_H_IDS(&intel_coffeelake_info),
+	INTEL_CFL_U_IDS(&intel_coffeelake_gt3_info),
+	INTEL_CNL_IDS(&intel_cannonlake_info),
 	{0, 0, 0}
 };
 MODULE_DEVICE_TABLE(pci, pciidlist);
diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c
index 060b171480d5..85269bcc8372 100644
--- a/drivers/gpu/drm/i915/i915_perf.c
+++ b/drivers/gpu/drm/i915/i915_perf.c
@@ -205,25 +205,49 @@
 
 #define OA_TAKEN(tail, head)	((tail - head) & (OA_BUFFER_SIZE - 1))
 
-/* There's a HW race condition between OA unit tail pointer register updates and
+/**
+ * DOC: OA Tail Pointer Race
+ *
+ * There's a HW race condition between OA unit tail pointer register updates and
  * writes to memory whereby the tail pointer can sometimes get ahead of what's
- * been written out to the OA buffer so far.
+ * been written out to the OA buffer so far (in terms of what's visible to the
+ * CPU).
+ *
+ * Although this can be observed explicitly while copying reports to userspace
+ * by checking for a zeroed report-id field in tail reports, we want to account
+ * for this earlier, as part of the _oa_buffer_check to avoid lots of redundant
+ * read() attempts.
+ *
+ * In effect we define a tail pointer for reading that lags the real tail
+ * pointer by at least %OA_TAIL_MARGIN_NSEC nanoseconds, which gives enough
+ * time for the corresponding reports to become visible to the CPU.
+ *
+ * To manage this we actually track two tail pointers:
+ *  1) An 'aging' tail with an associated timestamp that is tracked until we
+ *     can trust the corresponding data is visible to the CPU; at which point
+ *     it is considered 'aged'.
+ *  2) An 'aged' tail that can be used for read()ing.
+ *
+ * The two separate pointers let us decouple read()s from tail pointer aging.
+ *
+ * The tail pointers are checked and updated at a limited rate within a hrtimer
+ * callback (the same callback that is used for delivering POLLIN events)
  *
- * Although this can be observed explicitly by checking for a zeroed report-id
- * field in tail reports, it seems preferable to account for this earlier e.g.
- * as part of the _oa_buffer_is_empty checks to minimize -EAGAIN polling cycles
- * in this situation.
+ * Initially the tails are marked invalid with %INVALID_TAIL_PTR which
+ * indicates that an updated tail pointer is needed.
  *
- * To give time for the most recent reports to land before they may be copied to
- * userspace, the driver operates as if the tail pointer effectively lags behind
- * the HW tail pointer by 'tail_margin' bytes. The margin in bytes is calculated
- * based on this constant in nanoseconds, the current OA sampling exponent
- * and current report size.
+ * Most of the implementation details for this workaround are in
+ * gen7_oa_buffer_check_unlocked() and gen7_appand_oa_reports()
  *
- * There is also a fallback check while reading to simply skip over reports with
- * a zeroed report-id.
+ * Note for posterity: previously the driver used to define an effective tail
+ * pointer that lagged the real pointer by a 'tail margin' measured in bytes
+ * derived from %OA_TAIL_MARGIN_NSEC and the configured sampling frequency.
+ * This was flawed considering that the OA unit may also automatically generate
+ * non-periodic reports (such as on context switch) or the OA unit may be
+ * enabled without any periodic sampling.
  */
 #define OA_TAIL_MARGIN_NSEC	100000ULL
+#define INVALID_TAIL_PTR	0xffffffff
 
 /* frequency for checking whether the OA unit has written new reports to the
  * circular OA buffer...
@@ -308,27 +332,121 @@ struct perf_open_properties {
 	int oa_period_exponent;
 };
 
-/* NB: This is either called via fops or the poll check hrtimer (atomic ctx)
+/**
+ * gen7_oa_buffer_check_unlocked - check for data and update tail ptr state
+ * @dev_priv: i915 device instance
+ *
+ * This is either called via fops (for blocking reads in user ctx) or the poll
+ * check hrtimer (atomic ctx) to check the OA buffer tail pointer and check
+ * if there is data available for userspace to read.
  *
- * It's safe to read OA config state here unlocked, assuming that this is only
- * called while the stream is enabled, while the global OA configuration can't
- * be modified.
+ * This function is central to providing a workaround for the OA unit tail
+ * pointer having a race with respect to what data is visible to the CPU.
+ * It is responsible for reading tail pointers from the hardware and giving
+ * the pointers time to 'age' before they are made available for reading.
+ * (See description of OA_TAIL_MARGIN_NSEC above for further details.)
  *
- * Note: we don't lock around the head/tail reads even though there's the slim
- * possibility of read() fop errors forcing a re-init of the OA buffer
- * pointers.  A race here could result in a false positive !empty status which
- * is acceptable.
+ * Besides returning true when there is data available to read() this function
+ * also has the side effect of updating the oa_buffer.tails[], .aging_timestamp
+ * and .aged_tail_idx state used for reading.
+ *
+ * Note: It's safe to read OA config state here unlocked, assuming that this is
+ * only called while the stream is enabled, while the global OA configuration
+ * can't be modified.
+ *
+ * Returns: %true if the OA buffer contains data, else %false
  */
-static bool gen7_oa_buffer_is_empty_fop_unlocked(struct drm_i915_private *dev_priv)
+static bool gen7_oa_buffer_check_unlocked(struct drm_i915_private *dev_priv)
 {
 	int report_size = dev_priv->perf.oa.oa_buffer.format_size;
-	u32 oastatus2 = I915_READ(GEN7_OASTATUS2);
-	u32 oastatus1 = I915_READ(GEN7_OASTATUS1);
-	u32 head = oastatus2 & GEN7_OASTATUS2_HEAD_MASK;
-	u32 tail = oastatus1 & GEN7_OASTATUS1_TAIL_MASK;
+	unsigned long flags;
+	unsigned int aged_idx;
+	u32 oastatus1;
+	u32 head, hw_tail, aged_tail, aging_tail;
+	u64 now;
+
+	/* We have to consider the (unlikely) possibility that read() errors
+	 * could result in an OA buffer reset which might reset the head,
+	 * tails[] and aged_tail state.
+	 */
+	spin_lock_irqsave(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags);
+
+	/* NB: The head we observe here might effectively be a little out of
+	 * date (between head and tails[aged_idx].offset if there is currently
+	 * a read() in progress.
+	 */
+	head = dev_priv->perf.oa.oa_buffer.head;
+
+	aged_idx = dev_priv->perf.oa.oa_buffer.aged_tail_idx;
+	aged_tail = dev_priv->perf.oa.oa_buffer.tails[aged_idx].offset;
+	aging_tail = dev_priv->perf.oa.oa_buffer.tails[!aged_idx].offset;
 
-	return OA_TAKEN(tail, head) <
-		dev_priv->perf.oa.tail_margin + report_size;
+	oastatus1 = I915_READ(GEN7_OASTATUS1);
+	hw_tail = oastatus1 & GEN7_OASTATUS1_TAIL_MASK;
+
+	/* The tail pointer increases in 64 byte increments,
+	 * not in report_size steps...
+	 */
+	hw_tail &= ~(report_size - 1);
+
+	now = ktime_get_mono_fast_ns();
+
+	/* Update the aged tail
+	 *
+	 * Flip the tail pointer available for read()s once the aging tail is
+	 * old enough to trust that the corresponding data will be visible to
+	 * the CPU...
+	 *
+	 * Do this before updating the aging pointer in case we may be able to
+	 * immediately start aging a new pointer too (if new data has become
+	 * available) without needing to wait for a later hrtimer callback.
+	 */
+	if (aging_tail != INVALID_TAIL_PTR &&
+	    ((now - dev_priv->perf.oa.oa_buffer.aging_timestamp) >
+	     OA_TAIL_MARGIN_NSEC)) {
+		aged_idx ^= 1;
+		dev_priv->perf.oa.oa_buffer.aged_tail_idx = aged_idx;
+
+		aged_tail = aging_tail;
+
+		/* Mark that we need a new pointer to start aging... */
+		dev_priv->perf.oa.oa_buffer.tails[!aged_idx].offset = INVALID_TAIL_PTR;
+		aging_tail = INVALID_TAIL_PTR;
+	}
+
+	/* Update the aging tail
+	 *
+	 * We throttle aging tail updates until we have a new tail that
+	 * represents >= one report more data than is already available for
+	 * reading. This ensures there will be enough data for a successful
+	 * read once this new pointer has aged and ensures we will give the new
+	 * pointer time to age.
+	 */
+	if (aging_tail == INVALID_TAIL_PTR &&
+	    (aged_tail == INVALID_TAIL_PTR ||
+	     OA_TAKEN(hw_tail, aged_tail) >= report_size)) {
+		struct i915_vma *vma = dev_priv->perf.oa.oa_buffer.vma;
+		u32 gtt_offset = i915_ggtt_offset(vma);
+
+		/* Be paranoid and do a bounds check on the pointer read back
+		 * from hardware, just in case some spurious hardware condition
+		 * could put the tail out of bounds...
+		 */
+		if (hw_tail >= gtt_offset &&
+		    hw_tail < (gtt_offset + OA_BUFFER_SIZE)) {
+			dev_priv->perf.oa.oa_buffer.tails[!aged_idx].offset =
+				aging_tail = hw_tail;
+			dev_priv->perf.oa.oa_buffer.aging_timestamp = now;
+		} else {
+			DRM_ERROR("Ignoring spurious out of range OA buffer tail pointer = %u\n",
+				  hw_tail);
+		}
+	}
+
+	spin_unlock_irqrestore(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags);
+
+	return aged_tail == INVALID_TAIL_PTR ?
+		false : OA_TAKEN(aged_tail, head) >= report_size;
 }
 
 /**
@@ -421,8 +539,6 @@ static int append_oa_sample(struct i915_perf_stream *stream,
  * @buf: destination buffer given by userspace
  * @count: the number of bytes userspace wants to read
  * @offset: (inout): the current position for writing into @buf
- * @head_ptr: (inout): the current oa buffer cpu read position
- * @tail: the current oa buffer gpu write position
  *
  * Notably any error condition resulting in a short read (-%ENOSPC or
  * -%EFAULT) will be returned even though one or more records may
@@ -431,7 +547,7 @@ static int append_oa_sample(struct i915_perf_stream *stream,
  * userspace.
  *
  * Note: reports are consumed from the head, and appended to the
- * tail, so the head chases the tail?... If you think that's mad
+ * tail, so the tail chases the head?... If you think that's mad
  * and back-to-front you're not alone, but this follows the
  * Gen PRM naming convention.
  *
@@ -440,57 +556,55 @@ static int append_oa_sample(struct i915_perf_stream *stream,
 static int gen7_append_oa_reports(struct i915_perf_stream *stream,
 				  char __user *buf,
 				  size_t count,
-				  size_t *offset,
-				  u32 *head_ptr,
-				  u32 tail)
+				  size_t *offset)
 {
 	struct drm_i915_private *dev_priv = stream->dev_priv;
 	int report_size = dev_priv->perf.oa.oa_buffer.format_size;
 	u8 *oa_buf_base = dev_priv->perf.oa.oa_buffer.vaddr;
-	int tail_margin = dev_priv->perf.oa.tail_margin;
 	u32 gtt_offset = i915_ggtt_offset(dev_priv->perf.oa.oa_buffer.vma);
 	u32 mask = (OA_BUFFER_SIZE - 1);
-	u32 head;
+	size_t start_offset = *offset;
+	unsigned long flags;
+	unsigned int aged_tail_idx;
+	u32 head, tail;
 	u32 taken;
 	int ret = 0;
 
 	if (WARN_ON(!stream->enabled))
 		return -EIO;
 
-	head = *head_ptr - gtt_offset;
-	tail -= gtt_offset;
+	spin_lock_irqsave(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags);
 
-	/* The OA unit is expected to wrap the tail pointer according to the OA
-	 * buffer size and since we should never write a misaligned head
-	 * pointer we don't expect to read one back either...
-	 */
-	if (tail > OA_BUFFER_SIZE || head > OA_BUFFER_SIZE ||
-	    head % report_size) {
-		DRM_ERROR("Inconsistent OA buffer pointer (head = %u, tail = %u): force restart\n",
-			  head, tail);
-		dev_priv->perf.oa.ops.oa_disable(dev_priv);
-		dev_priv->perf.oa.ops.oa_enable(dev_priv);
-		*head_ptr = I915_READ(GEN7_OASTATUS2) &
-			GEN7_OASTATUS2_HEAD_MASK;
-		return -EIO;
-	}
+	head = dev_priv->perf.oa.oa_buffer.head;
+	aged_tail_idx = dev_priv->perf.oa.oa_buffer.aged_tail_idx;
+	tail = dev_priv->perf.oa.oa_buffer.tails[aged_tail_idx].offset;
 
+	spin_unlock_irqrestore(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags);
 
-	/* The tail pointer increases in 64 byte increments, not in report_size
-	 * steps...
+	/* An invalid tail pointer here means we're still waiting for the poll
+	 * hrtimer callback to give us a pointer
 	 */
-	tail &= ~(report_size - 1);
+	if (tail == INVALID_TAIL_PTR)
+		return -EAGAIN;
 
-	/* Move the tail pointer back by the current tail_margin to account for
-	 * the possibility that the latest reports may not have really landed
-	 * in memory yet...
+	/* NB: oa_buffer.head/tail include the gtt_offset which we don't want
+	 * while indexing relative to oa_buf_base.
 	 */
+	head -= gtt_offset;
+	tail -= gtt_offset;
 
-	if (OA_TAKEN(tail, head) < report_size + tail_margin)
-		return -EAGAIN;
+	/* An out of bounds or misaligned head or tail pointer implies a driver
+	 * bug since we validate + align the tail pointers we read from the
+	 * hardware and we are in full control of the head pointer which should
+	 * only be incremented by multiples of the report size (notably also
+	 * all a power of two).
+	 */
+	if (WARN_ONCE(head > OA_BUFFER_SIZE || head % report_size ||
+		      tail > OA_BUFFER_SIZE || tail % report_size,
+		      "Inconsistent OA buffer pointers: head = %u, tail = %u\n",
+		      head, tail))
+		return -EIO;
 
-	tail -= tail_margin;
-	tail &= mask;
 
 	for (/* none */;
 	     (taken = OA_TAKEN(tail, head));
@@ -518,7 +632,8 @@ static int gen7_append_oa_reports(struct i915_perf_stream *stream,
 		 * copying it to userspace...
 		 */
 		if (report32[0] == 0) {
-			DRM_NOTE("Skipping spurious, invalid OA report\n");
+			if (__ratelimit(&dev_priv->perf.oa.spurious_report_rs))
+				DRM_NOTE("Skipping spurious, invalid OA report\n");
 			continue;
 		}
 
@@ -535,7 +650,21 @@ static int gen7_append_oa_reports(struct i915_perf_stream *stream,
 		report32[0] = 0;
 	}
 
-	*head_ptr = gtt_offset + head;
+	if (start_offset != *offset) {
+		spin_lock_irqsave(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags);
+
+		/* We removed the gtt_offset for the copy loop above, indexing
+		 * relative to oa_buf_base so put back here...
+		 */
+		head += gtt_offset;
+
+		I915_WRITE(GEN7_OASTATUS2,
+			   ((head & GEN7_OASTATUS2_HEAD_MASK) |
+			    OA_MEM_SELECT_GGTT));
+		dev_priv->perf.oa.oa_buffer.head = head;
+
+		spin_unlock_irqrestore(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags);
+	}
 
 	return ret;
 }
@@ -562,22 +691,14 @@ static int gen7_oa_read(struct i915_perf_stream *stream,
 			size_t *offset)
 {
 	struct drm_i915_private *dev_priv = stream->dev_priv;
-	int report_size = dev_priv->perf.oa.oa_buffer.format_size;
-	u32 oastatus2;
 	u32 oastatus1;
-	u32 head;
-	u32 tail;
 	int ret;
 
 	if (WARN_ON(!dev_priv->perf.oa.oa_buffer.vaddr))
 		return -EIO;
 
-	oastatus2 = I915_READ(GEN7_OASTATUS2);
 	oastatus1 = I915_READ(GEN7_OASTATUS1);
 
-	head = oastatus2 & GEN7_OASTATUS2_HEAD_MASK;
-	tail = oastatus1 & GEN7_OASTATUS1_TAIL_MASK;
-
 	/* XXX: On Haswell we don't have a safe way to clear oastatus1
 	 * bits while the OA unit is enabled (while the tail pointer
 	 * may be updated asynchronously) so we ignore status bits
@@ -616,11 +737,7 @@ static int gen7_oa_read(struct i915_perf_stream *stream,
 		dev_priv->perf.oa.ops.oa_disable(dev_priv);
 		dev_priv->perf.oa.ops.oa_enable(dev_priv);
 
-		oastatus2 = I915_READ(GEN7_OASTATUS2);
 		oastatus1 = I915_READ(GEN7_OASTATUS1);
-
-		head = oastatus2 & GEN7_OASTATUS2_HEAD_MASK;
-		tail = oastatus1 & GEN7_OASTATUS1_TAIL_MASK;
 	}
 
 	if (unlikely(oastatus1 & GEN7_OASTATUS1_REPORT_LOST)) {
@@ -632,29 +749,7 @@ static int gen7_oa_read(struct i915_perf_stream *stream,
 			GEN7_OASTATUS1_REPORT_LOST;
 	}
 
-	ret = gen7_append_oa_reports(stream, buf, count, offset,
-				     &head, tail);
-
-	/* All the report sizes are a power of two and the
-	 * head should always be incremented by some multiple
-	 * of the report size.
-	 *
-	 * A warning here, but notably if we later read back a
-	 * misaligned pointer we will treat that as a bug since
-	 * it could lead to a buffer overrun.
-	 */
-	WARN_ONCE(head & (report_size - 1),
-		  "i915: Writing misaligned OA head pointer");
-
-	/* Note: we update the head pointer here even if an error
-	 * was returned since the error may represent a short read
-	 * where some some reports were successfully copied.
-	 */
-	I915_WRITE(GEN7_OASTATUS2,
-		   ((head & GEN7_OASTATUS2_HEAD_MASK) |
-		    OA_MEM_SELECT_GGTT));
-
-	return ret;
+	return gen7_append_oa_reports(stream, buf, count, offset);
 }
 
 /**
@@ -679,14 +774,8 @@ static int i915_oa_wait_unlocked(struct i915_perf_stream *stream)
 	if (!dev_priv->perf.oa.periodic)
 		return -EIO;
 
-	/* Note: the oa_buffer_is_empty() condition is ok to run unlocked as it
-	 * just performs mmio reads of the OA buffer head + tail pointers and
-	 * it's assumed we're handling some operation that implies the stream
-	 * can't be destroyed until completion (such as a read()) that ensures
-	 * the device + OA buffer can't disappear
-	 */
 	return wait_event_interruptible(dev_priv->perf.oa.poll_wq,
-					!dev_priv->perf.oa.ops.oa_buffer_is_empty(dev_priv));
+					dev_priv->perf.oa.ops.oa_buffer_check(dev_priv));
 }
 
 /**
@@ -744,6 +833,7 @@ static int oa_get_render_ctx_id(struct i915_perf_stream *stream)
 {
 	struct drm_i915_private *dev_priv = stream->dev_priv;
 	struct intel_engine_cs *engine = dev_priv->engine[RCS];
+	struct intel_ring *ring;
 	int ret;
 
 	ret = i915_mutex_lock_interruptible(&dev_priv->drm);
@@ -755,9 +845,10 @@ static int oa_get_render_ctx_id(struct i915_perf_stream *stream)
 	 *
 	 * NB: implied RCS engine...
 	 */
-	ret = engine->context_pin(engine, stream->ctx);
-	if (ret)
-		goto unlock;
+	ring = engine->context_pin(engine, stream->ctx);
+	mutex_unlock(&dev_priv->drm.struct_mutex);
+	if (IS_ERR(ring))
+		return PTR_ERR(ring);
 
 	/* Explicitly track the ID (instead of calling i915_ggtt_offset()
 	 * on the fly) considering the difference with gen8+ and
@@ -766,10 +857,7 @@ static int oa_get_render_ctx_id(struct i915_perf_stream *stream)
 	dev_priv->perf.oa.specific_ctx_id =
 		i915_ggtt_offset(stream->ctx->engine[engine->id].state);
 
-unlock:
-	mutex_unlock(&dev_priv->drm.struct_mutex);
-
-	return ret;
+	return 0;
 }
 
 /**
@@ -824,19 +912,36 @@ static void i915_oa_stream_destroy(struct i915_perf_stream *stream)
 		oa_put_render_ctx_id(stream);
 
 	dev_priv->perf.oa.exclusive_stream = NULL;
+
+	if (dev_priv->perf.oa.spurious_report_rs.missed) {
+		DRM_NOTE("%d spurious OA report notices suppressed due to ratelimiting\n",
+			 dev_priv->perf.oa.spurious_report_rs.missed);
+	}
 }
 
 static void gen7_init_oa_buffer(struct drm_i915_private *dev_priv)
 {
 	u32 gtt_offset = i915_ggtt_offset(dev_priv->perf.oa.oa_buffer.vma);
+	unsigned long flags;
+
+	spin_lock_irqsave(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags);
 
 	/* Pre-DevBDW: OABUFFER must be set with counters off,
 	 * before OASTATUS1, but after OASTATUS2
 	 */
 	I915_WRITE(GEN7_OASTATUS2, gtt_offset | OA_MEM_SELECT_GGTT); /* head */
+	dev_priv->perf.oa.oa_buffer.head = gtt_offset;
+
 	I915_WRITE(GEN7_OABUFFER, gtt_offset);
+
 	I915_WRITE(GEN7_OASTATUS1, gtt_offset | OABUFFER_SIZE_16M); /* tail */
 
+	/* Mark that we need updated tail pointers to read from... */
+	dev_priv->perf.oa.oa_buffer.tails[0].offset = INVALID_TAIL_PTR;
+	dev_priv->perf.oa.oa_buffer.tails[1].offset = INVALID_TAIL_PTR;
+
+	spin_unlock_irqrestore(&dev_priv->perf.oa.oa_buffer.ptr_lock, flags);
+
 	/* On Haswell we have to track which OASTATUS1 flags we've
 	 * already seen since they can't be cleared while periodic
 	 * sampling is enabled.
@@ -1094,12 +1199,6 @@ static void i915_oa_stream_disable(struct i915_perf_stream *stream)
 		hrtimer_cancel(&dev_priv->perf.oa.poll_check_timer);
 }
 
-static u64 oa_exponent_to_ns(struct drm_i915_private *dev_priv, int exponent)
-{
-	return div_u64(1000000000ULL * (2ULL << exponent),
-		       dev_priv->perf.oa.timestamp_frequency);
-}
-
 static const struct i915_perf_stream_ops i915_oa_stream_ops = {
 	.destroy = i915_oa_stream_destroy,
 	.enable = i915_oa_stream_enable,
@@ -1173,6 +1272,26 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream,
 		return -EINVAL;
 	}
 
+	/* We set up some ratelimit state to potentially throttle any _NOTES
+	 * about spurious, invalid OA reports which we don't forward to
+	 * userspace.
+	 *
+	 * The initialization is associated with opening the stream (not driver
+	 * init) considering we print a _NOTE about any throttling when closing
+	 * the stream instead of waiting until driver _fini which no one would
+	 * ever see.
+	 *
+	 * Using the same limiting factors as printk_ratelimit()
+	 */
+	ratelimit_state_init(&dev_priv->perf.oa.spurious_report_rs,
+			     5 * HZ, 10);
+	/* Since we use a DRM_NOTE for spurious reports it would be
+	 * inconsistent to let __ratelimit() automatically print a warning for
+	 * throttling.
+	 */
+	ratelimit_set_flags(&dev_priv->perf.oa.spurious_report_rs,
+			    RATELIMIT_MSG_ON_RELEASE);
+
 	stream->sample_size = sizeof(struct drm_i915_perf_record_header);
 
 	format_size = dev_priv->perf.oa.oa_formats[props->oa_format].size;
@@ -1190,20 +1309,9 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream,
 	dev_priv->perf.oa.metrics_set = props->metrics_set;
 
 	dev_priv->perf.oa.periodic = props->oa_periodic;
-	if (dev_priv->perf.oa.periodic) {
-		u32 tail;
-
+	if (dev_priv->perf.oa.periodic)
 		dev_priv->perf.oa.period_exponent = props->oa_period_exponent;
 
-		/* See comment for OA_TAIL_MARGIN_NSEC for details
-		 * about this tail_margin...
-		 */
-		tail = div64_u64(OA_TAIL_MARGIN_NSEC,
-				 oa_exponent_to_ns(dev_priv,
-						   props->oa_period_exponent));
-		dev_priv->perf.oa.tail_margin = (tail + 1) * format_size;
-	}
-
 	if (stream->ctx) {
 		ret = oa_get_render_ctx_id(stream);
 		if (ret)
@@ -1352,7 +1460,15 @@ static ssize_t i915_perf_read(struct file *file,
 		mutex_unlock(&dev_priv->perf.lock);
 	}
 
-	if (ret >= 0) {
+	/* We allow the poll checking to sometimes report false positive POLLIN
+	 * events where we might actually report EAGAIN on read() if there's
+	 * not really any data available. In this situation though we don't
+	 * want to enter a busy loop between poll() reporting a POLLIN event
+	 * and read() returning -EAGAIN. Clearing the oa.pollin state here
+	 * effectively ensures we back off until the next hrtimer callback
+	 * before reporting another POLLIN event.
+	 */
+	if (ret >= 0 || ret == -EAGAIN) {
 		/* Maybe make ->pollin per-stream state if we support multiple
 		 * concurrent streams in the future.
 		 */
@@ -1368,7 +1484,7 @@ static enum hrtimer_restart oa_poll_check_timer_cb(struct hrtimer *hrtimer)
 		container_of(hrtimer, typeof(*dev_priv),
 			     perf.oa.poll_check_timer);
 
-	if (!dev_priv->perf.oa.ops.oa_buffer_is_empty(dev_priv)) {
+	if (dev_priv->perf.oa.ops.oa_buffer_check(dev_priv)) {
 		dev_priv->perf.oa.pollin = true;
 		wake_up(&dev_priv->perf.oa.poll_wq);
 	}
@@ -1817,11 +1933,13 @@ static int read_properties_unlocked(struct drm_i915_private *dev_priv,
 			break;
 		case DRM_I915_PERF_PROP_OA_FORMAT:
 			if (value == 0 || value >= I915_OA_FORMAT_MAX) {
-				DRM_DEBUG("Invalid OA report format\n");
+				DRM_DEBUG("Out-of-range OA report format %llu\n",
+					  value);
 				return -EINVAL;
 			}
 			if (!dev_priv->perf.oa.oa_formats[value].size) {
-				DRM_DEBUG("Invalid OA report format\n");
+				DRM_DEBUG("Unsupported OA report format %llu\n",
+					  value);
 				return -EINVAL;
 			}
 			props->oa_format = value;
@@ -2063,6 +2181,7 @@ void i915_perf_init(struct drm_i915_private *dev_priv)
 	INIT_LIST_HEAD(&dev_priv->perf.streams);
 	mutex_init(&dev_priv->perf.lock);
 	spin_lock_init(&dev_priv->perf.hook_lock);
+	spin_lock_init(&dev_priv->perf.oa.oa_buffer.ptr_lock);
 
 	dev_priv->perf.oa.ops.init_oa_buffer = gen7_init_oa_buffer;
 	dev_priv->perf.oa.ops.enable_metric_set = hsw_enable_metric_set;
@@ -2070,10 +2189,8 @@ void i915_perf_init(struct drm_i915_private *dev_priv)
 	dev_priv->perf.oa.ops.oa_enable = gen7_oa_enable;
 	dev_priv->perf.oa.ops.oa_disable = gen7_oa_disable;
 	dev_priv->perf.oa.ops.read = gen7_oa_read;
-	dev_priv->perf.oa.ops.oa_buffer_is_empty =
-		gen7_oa_buffer_is_empty_fop_unlocked;
-
-	dev_priv->perf.oa.timestamp_frequency = 12500000;
+	dev_priv->perf.oa.ops.oa_buffer_check =
+		gen7_oa_buffer_check_unlocked;
 
 	dev_priv->perf.oa.oa_formats = hsw_oa_formats;
 
diff --git a/drivers/gpu/drm/i915/i915_pvinfo.h b/drivers/gpu/drm/i915/i915_pvinfo.h
index c0cb2974caac..2cfe96d3e5d1 100644
--- a/drivers/gpu/drm/i915/i915_pvinfo.h
+++ b/drivers/gpu/drm/i915/i915_pvinfo.h
@@ -36,10 +36,6 @@
 #define VGT_VERSION_MAJOR 1
 #define VGT_VERSION_MINOR 0
 
-#define INTEL_VGT_IF_VERSION_ENCODE(major, minor) ((major) << 16 | (minor))
-#define INTEL_VGT_IF_VERSION \
-	INTEL_VGT_IF_VERSION_ENCODE(VGT_VERSION_MAJOR, VGT_VERSION_MINOR)
-
 /*
  * notifications from guest to vgpu device model
  */
@@ -55,8 +51,8 @@ enum vgt_g2v_type {
 
 struct vgt_if {
 	u64 magic;		/* VGT_MAGIC */
-	uint16_t version_major;
-	uint16_t version_minor;
+	u16 version_major;
+	u16 version_minor;
 	u32 vgt_id;		/* ID of vGT instance */
 	u32 rsv1[12];		/* pad to offset 0x40 */
 	/*
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 65b837e96fe6..b6d69e289974 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -58,10 +58,8 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define _MMIO_TRANS(tran, a, b) _MMIO(_TRANS(tran, a, b))
 #define _PORT(port, a, b) ((a) + (port)*((b)-(a)))
 #define _MMIO_PORT(port, a, b) _MMIO(_PORT(port, a, b))
-#define _PIPE3(pipe, ...) _PICK(pipe, __VA_ARGS__)
-#define _MMIO_PIPE3(pipe, a, b, c) _MMIO(_PIPE3(pipe, a, b, c))
-#define _PORT3(port, ...) _PICK(port, __VA_ARGS__)
-#define _MMIO_PORT3(pipe, a, b, c) _MMIO(_PORT3(pipe, a, b, c))
+#define _MMIO_PIPE3(pipe, a, b, c) _MMIO(_PICK(pipe, a, b, c))
+#define _MMIO_PORT3(pipe, a, b, c) _MMIO(_PICK(pipe, a, b, c))
 #define _PHY3(phy, ...) _PICK(phy, __VA_ARGS__)
 #define _MMIO_PHY3(phy, a, b, c) _MMIO(_PHY3(phy, a, b, c))
 
@@ -85,6 +83,14 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define VECS_HW		3
 #define VCS2_HW		4
 
+/* Engine class */
+
+#define RENDER_CLASS		0
+#define VIDEO_DECODE_CLASS	1
+#define VIDEO_ENHANCEMENT_CLASS	2
+#define COPY_ENGINE_CLASS	3
+#define OTHER_CLASS		4
+
 /* PCI config space */
 
 #define MCHBAR_I915 0x44
@@ -1057,6 +1063,7 @@ enum skl_disp_power_wells {
 	SKL_DISP_PW_MISC_IO,
 	SKL_DISP_PW_DDI_A_E,
 	GLK_DISP_PW_DDI_A = SKL_DISP_PW_DDI_A_E,
+	CNL_DISP_PW_DDI_A = SKL_DISP_PW_DDI_A_E,
 	SKL_DISP_PW_DDI_B,
 	SKL_DISP_PW_DDI_C,
 	SKL_DISP_PW_DDI_D,
@@ -1064,6 +1071,10 @@ enum skl_disp_power_wells {
 	GLK_DISP_PW_AUX_A = 8,
 	GLK_DISP_PW_AUX_B,
 	GLK_DISP_PW_AUX_C,
+	CNL_DISP_PW_AUX_A = GLK_DISP_PW_AUX_A,
+	CNL_DISP_PW_AUX_B = GLK_DISP_PW_AUX_B,
+	CNL_DISP_PW_AUX_C = GLK_DISP_PW_AUX_C,
+	CNL_DISP_PW_AUX_D,
 
 	SKL_DISP_PW_1 = 14,
 	SKL_DISP_PW_2,
@@ -2499,10 +2510,6 @@ enum skl_disp_power_wells {
 #define FBC_FENCE_OFF		_MMIO(0x3218) /* BSpec typo has 321Bh */
 #define FBC_TAG(i)		_MMIO(0x3300 + (i) * 4)
 
-#define FBC_STATUS2			_MMIO(0x43214)
-#define  IVB_FBC_COMPRESSION_MASK	0x7ff
-#define  BDW_FBC_COMPRESSION_MASK	0xfff
-
 #define FBC_LL_SIZE		(1536)
 
 #define FBC_LLC_READ_CTRL	_MMIO(0x9044)
@@ -2531,7 +2538,7 @@ enum skl_disp_power_wells {
 #define   DPFC_INVAL_SEG_SHIFT  (16)
 #define   DPFC_INVAL_SEG_MASK	(0x07ff0000)
 #define   DPFC_COMP_SEG_SHIFT	(0)
-#define   DPFC_COMP_SEG_MASK	(0x000003ff)
+#define   DPFC_COMP_SEG_MASK	(0x000007ff)
 #define DPFC_STATUS2		_MMIO(0x3214)
 #define DPFC_FENCE_YOFF		_MMIO(0x3218)
 #define DPFC_CHICKEN		_MMIO(0x3224)
@@ -2545,6 +2552,10 @@ enum skl_disp_power_wells {
 #define   DPFC_RESERVED		(0x1FFFFF00)
 #define ILK_DPFC_RECOMP_CTL	_MMIO(0x4320c)
 #define ILK_DPFC_STATUS		_MMIO(0x43210)
+#define  ILK_DPFC_COMP_SEG_MASK	0x7ff
+#define IVB_FBC_STATUS2		_MMIO(0x43214)
+#define  IVB_FBC_COMP_SEG_MASK	0x7ff
+#define  BDW_FBC_COMP_SEG_MASK	0xfff
 #define ILK_DPFC_FENCE_YOFF	_MMIO(0x43218)
 #define ILK_DPFC_CHICKEN	_MMIO(0x43224)
 #define   ILK_DPFC_DISABLE_DUMMY0 (1<<8)
@@ -2618,9 +2629,10 @@ enum skl_disp_power_wells {
 #define   GMBUS_PIN_DPB		5 /* SDVO, HDMIB */
 #define   GMBUS_PIN_DPD		6 /* HDMID */
 #define   GMBUS_PIN_RESERVED	7 /* 7 reserved */
-#define   GMBUS_PIN_1_BXT	1
+#define   GMBUS_PIN_1_BXT	1 /* BXT+ (atom) and CNP+ (big core) */
 #define   GMBUS_PIN_2_BXT	2
 #define   GMBUS_PIN_3_BXT	3
+#define   GMBUS_PIN_4_CNP	4
 #define   GMBUS_NUM_PINS	7 /* including 0 */
 #define GMBUS1			_MMIO(dev_priv->gpio_mmio_base + 0x5104) /* command/status */
 #define   GMBUS_SW_CLR_INT	(1<<31)
@@ -3366,16 +3378,6 @@ enum skl_disp_power_wells {
 #define GEN7_CXT_VFSTATE_SIZE(ctx_reg)	(((ctx_reg) >> 0) & 0x3f)
 #define GEN7_CXT_TOTAL_SIZE(ctx_reg)	(GEN7_CXT_EXTENDED_SIZE(ctx_reg) + \
 					 GEN7_CXT_VFSTATE_SIZE(ctx_reg))
-/* Haswell does have the CXT_SIZE register however it does not appear to be
- * valid. Now, docs explain in dwords what is in the context object. The full
- * size is 70720 bytes, however, the power context and execlist context will
- * never be saved (power context is stored elsewhere, and execlists don't work
- * on HSW) - so the final size, including the extra state required for the
- * Resource Streamer, is 66944 bytes, which rounds to 17 pages.
- */
-#define HSW_CXT_TOTAL_SIZE		(17 * PAGE_SIZE)
-/* Same as Haswell, but 72064 bytes now. */
-#define GEN8_CXT_TOTAL_SIZE		(18 * PAGE_SIZE)
 
 enum {
 	INTEL_ADVANCED_CONTEXT = 0,
@@ -5441,9 +5443,7 @@ enum {
 #define   CURSOR_MODE_128_ARGB_AX ((1 << 5) | CURSOR_MODE_128_32B_AX)
 #define   CURSOR_MODE_256_ARGB_AX ((1 << 5) | CURSOR_MODE_256_32B_AX)
 #define   CURSOR_MODE_64_ARGB_AX ((1 << 5) | CURSOR_MODE_64_32B_AX)
-#define   MCURSOR_PIPE_SELECT	(1 << 28)
-#define   MCURSOR_PIPE_A	0x00
-#define   MCURSOR_PIPE_B	(1 << 28)
+#define   MCURSOR_PIPE_SELECT(pipe)	((pipe) << 28)
 #define   MCURSOR_GAMMA_ENABLE  (1 << 26)
 #define   CURSOR_ROTATE_180	(1<<15)
 #define   CURSOR_TRICKLE_FEED_DISABLE	(1 << 14)
@@ -5453,7 +5453,9 @@ enum {
 #define   CURSOR_POS_SIGN       0x8000
 #define   CURSOR_X_SHIFT        0
 #define   CURSOR_Y_SHIFT        16
-#define CURSIZE			_MMIO(0x700a0)
+#define CURSIZE			_MMIO(0x700a0) /* 845/865 */
+#define _CUR_FBC_CTL_A		0x700a0 /* ivb+ */
+#define   CUR_FBC_CTL_EN	(1 << 31)
 #define _CURBCNTR		0x700c0
 #define _CURBBASE		0x700c4
 #define _CURBPOS		0x700c8
@@ -5469,6 +5471,7 @@ enum {
 #define CURCNTR(pipe) _CURSOR2(pipe, _CURACNTR)
 #define CURBASE(pipe) _CURSOR2(pipe, _CURABASE)
 #define CURPOS(pipe) _CURSOR2(pipe, _CURAPOS)
+#define CUR_FBC_CTL(pipe) _CURSOR2(pipe, _CUR_FBC_CTL_A)
 
 #define CURSOR_A_OFFSET 0x70080
 #define CURSOR_B_OFFSET 0x700c0
@@ -5501,8 +5504,7 @@ enum {
 #define   DISPPLANE_PIPE_CSC_ENABLE		(1<<24)
 #define   DISPPLANE_SEL_PIPE_SHIFT		24
 #define   DISPPLANE_SEL_PIPE_MASK		(3<<DISPPLANE_SEL_PIPE_SHIFT)
-#define   DISPPLANE_SEL_PIPE_A			0
-#define   DISPPLANE_SEL_PIPE_B			(1<<DISPPLANE_SEL_PIPE_SHIFT)
+#define   DISPPLANE_SEL_PIPE(pipe)		((pipe)<<DISPPLANE_SEL_PIPE_SHIFT)
 #define   DISPPLANE_SRC_KEY_ENABLE		(1<<22)
 #define   DISPPLANE_SRC_KEY_DISABLE		0
 #define   DISPPLANE_LINE_DOUBLE			(1<<20)
@@ -6840,6 +6842,10 @@ enum {
 #define  FDL_TP2_TIMER_SHIFT    10
 #define  FDL_TP2_TIMER_MASK     (3<<10)
 #define  RAWCLK_FREQ_MASK       0x3ff
+#define  CNP_RAWCLK_DIV_MASK	(0x3ff << 16)
+#define  CNP_RAWCLK_DIV(div)	((div) << 16)
+#define  CNP_RAWCLK_FRAC_MASK	(0xf << 26)
+#define  CNP_RAWCLK_FRAC(frac)	((frac) << 26)
 
 #define PCH_DPLL_TMR_CFG        _MMIO(0xc6208)
 
@@ -7794,13 +7800,6 @@ enum {
 #define  SKL_FUSE_PG1_DIST_STATUS              (1<<26)
 #define  SKL_FUSE_PG2_DIST_STATUS              (1<<25)
 
-/* Decoupled MMIO register pair for kernel driver */
-#define GEN9_DECOUPLED_REG0_DW0			_MMIO(0xF00)
-#define GEN9_DECOUPLED_REG0_DW1			_MMIO(0xF04)
-#define GEN9_DECOUPLED_DW1_GO			(1<<31)
-#define GEN9_DECOUPLED_PD_SHIFT			28
-#define GEN9_DECOUPLED_OP_SHIFT			24
-
 /* Per-pipe DDI Function Control */
 #define _TRANS_DDI_FUNC_CTL_A		0x60400
 #define _TRANS_DDI_FUNC_CTL_B		0x61400
@@ -8150,6 +8149,7 @@ enum {
 /* SFUSE_STRAP */
 #define SFUSE_STRAP			_MMIO(0xc2014)
 #define  SFUSE_STRAP_FUSE_LOCK		(1<<13)
+#define  SFUSE_STRAP_RAW_FREQUENCY	(1<<8)
 #define  SFUSE_STRAP_DISPLAY_DISABLED	(1<<7)
 #define  SFUSE_STRAP_CRT_DISABLED	(1<<6)
 #define  SFUSE_STRAP_DDIB_DETECTED	(1<<2)
diff --git a/drivers/gpu/drm/i915/i915_sw_fence.c b/drivers/gpu/drm/i915/i915_sw_fence.c
index a277f8eb7beb..474d23c0c0ce 100644
--- a/drivers/gpu/drm/i915/i915_sw_fence.c
+++ b/drivers/gpu/drm/i915/i915_sw_fence.c
@@ -12,6 +12,7 @@
 #include <linux/reservation.h>
 
 #include "i915_sw_fence.h"
+#include "i915_selftest.h"
 
 #define I915_SW_FENCE_FLAG_ALLOC BIT(3) /* after WQ_FLAG_* for safety */
 
@@ -120,34 +121,6 @@ void i915_sw_fence_fini(struct i915_sw_fence *fence)
 }
 #endif
 
-static void i915_sw_fence_release(struct kref *kref)
-{
-	struct i915_sw_fence *fence = container_of(kref, typeof(*fence), kref);
-
-	WARN_ON(atomic_read(&fence->pending) > 0);
-	debug_fence_destroy(fence);
-
-	if (fence->flags & I915_SW_FENCE_MASK) {
-		__i915_sw_fence_notify(fence, FENCE_FREE);
-	} else {
-		i915_sw_fence_fini(fence);
-		kfree(fence);
-	}
-}
-
-static void i915_sw_fence_put(struct i915_sw_fence *fence)
-{
-	debug_fence_assert(fence);
-	kref_put(&fence->kref, i915_sw_fence_release);
-}
-
-static struct i915_sw_fence *i915_sw_fence_get(struct i915_sw_fence *fence)
-{
-	debug_fence_assert(fence);
-	kref_get(&fence->kref);
-	return fence;
-}
-
 static void __i915_sw_fence_wake_up_all(struct i915_sw_fence *fence,
 					struct list_head *continuation)
 {
@@ -202,13 +175,15 @@ static void __i915_sw_fence_complete(struct i915_sw_fence *fence,
 
 	debug_fence_set_state(fence, DEBUG_FENCE_IDLE, DEBUG_FENCE_NOTIFY);
 
-	if (fence->flags & I915_SW_FENCE_MASK &&
-	    __i915_sw_fence_notify(fence, FENCE_COMPLETE) != NOTIFY_DONE)
+	if (__i915_sw_fence_notify(fence, FENCE_COMPLETE) != NOTIFY_DONE)
 		return;
 
 	debug_fence_set_state(fence, DEBUG_FENCE_NOTIFY, DEBUG_FENCE_IDLE);
 
 	__i915_sw_fence_wake_up_all(fence, continuation);
+
+	debug_fence_destroy(fence);
+	__i915_sw_fence_notify(fence, FENCE_FREE);
 }
 
 static void i915_sw_fence_complete(struct i915_sw_fence *fence)
@@ -232,33 +207,26 @@ void __i915_sw_fence_init(struct i915_sw_fence *fence,
 			  const char *name,
 			  struct lock_class_key *key)
 {
-	BUG_ON((unsigned long)fn & ~I915_SW_FENCE_MASK);
+	BUG_ON(!fn || (unsigned long)fn & ~I915_SW_FENCE_MASK);
 
 	debug_fence_init(fence);
 
 	__init_waitqueue_head(&fence->wait, name, key);
-	kref_init(&fence->kref);
 	atomic_set(&fence->pending, 1);
 	fence->flags = (unsigned long)fn;
 }
 
-static void __i915_sw_fence_commit(struct i915_sw_fence *fence)
-{
-	i915_sw_fence_complete(fence);
-	i915_sw_fence_put(fence);
-}
-
 void i915_sw_fence_commit(struct i915_sw_fence *fence)
 {
 	debug_fence_activate(fence);
-	__i915_sw_fence_commit(fence);
+	i915_sw_fence_complete(fence);
 }
 
 static int i915_sw_fence_wake(wait_queue_t *wq, unsigned mode, int flags, void *key)
 {
 	list_del(&wq->task_list);
 	__i915_sw_fence_complete(wq->private, key);
-	i915_sw_fence_put(wq->private);
+
 	if (wq->flags & I915_SW_FENCE_FLAG_ALLOC)
 		kfree(wq);
 	return 0;
@@ -307,7 +275,7 @@ static bool i915_sw_fence_check_if_after(struct i915_sw_fence *fence,
 	unsigned long flags;
 	bool err;
 
-	if (!IS_ENABLED(CONFIG_I915_SW_FENCE_CHECK_DAG))
+	if (!IS_ENABLED(CONFIG_DRM_I915_SW_FENCE_CHECK_DAG))
 		return false;
 
 	spin_lock_irqsave(&i915_sw_fence_lock, flags);
@@ -353,7 +321,7 @@ static int __i915_sw_fence_await_sw_fence(struct i915_sw_fence *fence,
 	INIT_LIST_HEAD(&wq->task_list);
 	wq->flags = pending;
 	wq->func = i915_sw_fence_wake;
-	wq->private = i915_sw_fence_get(fence);
+	wq->private = fence;
 
 	i915_sw_fence_await(fence);
 
@@ -402,7 +370,7 @@ static void timer_i915_sw_fence_wake(unsigned long data)
 	dma_fence_put(cb->dma);
 	cb->dma = NULL;
 
-	__i915_sw_fence_commit(cb->fence);
+	i915_sw_fence_complete(cb->fence);
 	cb->timer.function = NULL;
 }
 
@@ -413,7 +381,7 @@ static void dma_i915_sw_fence_wake(struct dma_fence *dma,
 
 	del_timer_sync(&cb->timer);
 	if (cb->timer.function)
-		__i915_sw_fence_commit(cb->fence);
+		i915_sw_fence_complete(cb->fence);
 	dma_fence_put(cb->dma);
 
 	kfree(cb);
@@ -440,7 +408,7 @@ int i915_sw_fence_await_dma_fence(struct i915_sw_fence *fence,
 		return dma_fence_wait(dma, false);
 	}
 
-	cb->fence = i915_sw_fence_get(fence);
+	cb->fence = fence;
 	i915_sw_fence_await(fence);
 
 	cb->dma = NULL;
@@ -523,3 +491,7 @@ int i915_sw_fence_await_reservation(struct i915_sw_fence *fence,
 
 	return ret;
 }
+
+#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
+#include "selftests/i915_sw_fence.c"
+#endif
diff --git a/drivers/gpu/drm/i915/i915_sw_fence.h b/drivers/gpu/drm/i915/i915_sw_fence.h
index d31cefbbcc04..1d3b6051daaf 100644
--- a/drivers/gpu/drm/i915/i915_sw_fence.h
+++ b/drivers/gpu/drm/i915/i915_sw_fence.h
@@ -23,7 +23,6 @@ struct reservation_object;
 struct i915_sw_fence {
 	wait_queue_head_t wait;
 	unsigned long flags;
-	struct kref kref;
 	atomic_t pending;
 };
 
diff --git a/drivers/gpu/drm/i915/i915_syncmap.c b/drivers/gpu/drm/i915/i915_syncmap.c
new file mode 100644
index 000000000000..0087acf731a8
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_syncmap.c
@@ -0,0 +1,412 @@
+/*
+ * Copyright © 2017 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ */
+
+#include <linux/slab.h>
+
+#include "i915_syncmap.h"
+
+#include "i915_gem.h" /* GEM_BUG_ON() */
+#include "i915_selftest.h"
+
+#define SHIFT ilog2(KSYNCMAP)
+#define MASK (KSYNCMAP - 1)
+
+/*
+ * struct i915_syncmap is a layer of a radixtree that maps a u64 fence
+ * context id to the last u32 fence seqno waited upon from that context.
+ * Unlike lib/radixtree it uses a parent pointer that allows traversal back to
+ * the root. This allows us to access the whole tree via a single pointer
+ * to the most recently used layer. We expect fence contexts to be dense
+ * and most reuse to be on the same i915_gem_context but on neighbouring
+ * engines (i.e. on adjacent contexts) and reuse the same leaf, a very
+ * effective lookup cache. If the new lookup is not on the same leaf, we
+ * expect it to be on the neighbouring branch.
+ *
+ * A leaf holds an array of u32 seqno, and has height 0. The bitmap field
+ * allows us to store whether a particular seqno is valid (i.e. allows us
+ * to distinguish unset from 0).
+ *
+ * A branch holds an array of layer pointers, and has height > 0, and always
+ * has at least 2 layers (either branches or leaves) below it.
+ *
+ * For example,
+ *	for x in
+ *	  0 1 2 0x10 0x11 0x200 0x201
+ *	  0x500000 0x500001 0x503000 0x503001
+ *	  0xE<<60:
+ *		i915_syncmap_set(&sync, x, lower_32_bits(x));
+ * will build a tree like:
+ *	0xXXXXXXXXXXXXXXXX
+ *	0-> 0x0000000000XXXXXX
+ *	|   0-> 0x0000000000000XXX
+ *	|   |   0-> 0x00000000000000XX
+ *	|   |   |   0-> 0x000000000000000X 0:0, 1:1, 2:2
+ *	|   |   |   1-> 0x000000000000001X 0:10, 1:11
+ *	|   |   2-> 0x000000000000020X 0:200, 1:201
+ *	|   5-> 0x000000000050XXXX
+ *	|       0-> 0x000000000050000X 0:500000, 1:500001
+ *	|       3-> 0x000000000050300X 0:503000, 1:503001
+ *	e-> 0xe00000000000000X e:e
+ */
+
+struct i915_syncmap {
+	u64 prefix;
+	unsigned int height;
+	unsigned int bitmap;
+	struct i915_syncmap *parent;
+	/*
+	 * Following this header is an array of either seqno or child pointers:
+	 * union {
+	 *	u32 seqno[KSYNCMAP];
+	 *	struct i915_syncmap *child[KSYNCMAP];
+	 * };
+	 */
+};
+
+/**
+ * i915_syncmap_init -- initialise the #i915_syncmap
+ * @root - pointer to the #i915_syncmap
+ */
+void i915_syncmap_init(struct i915_syncmap **root)
+{
+	BUILD_BUG_ON_NOT_POWER_OF_2(KSYNCMAP);
+	BUILD_BUG_ON_NOT_POWER_OF_2(SHIFT);
+	BUILD_BUG_ON(KSYNCMAP > BITS_PER_BYTE * sizeof((*root)->bitmap));
+	*root = NULL;
+}
+
+static inline u32 *__sync_seqno(struct i915_syncmap *p)
+{
+	GEM_BUG_ON(p->height);
+	return (u32 *)(p + 1);
+}
+
+static inline struct i915_syncmap **__sync_child(struct i915_syncmap *p)
+{
+	GEM_BUG_ON(!p->height);
+	return (struct i915_syncmap **)(p + 1);
+}
+
+static inline unsigned int
+__sync_branch_idx(const struct i915_syncmap *p, u64 id)
+{
+	return (id >> p->height) & MASK;
+}
+
+static inline unsigned int
+__sync_leaf_idx(const struct i915_syncmap *p, u64 id)
+{
+	GEM_BUG_ON(p->height);
+	return id & MASK;
+}
+
+static inline u64 __sync_branch_prefix(const struct i915_syncmap *p, u64 id)
+{
+	return id >> p->height >> SHIFT;
+}
+
+static inline u64 __sync_leaf_prefix(const struct i915_syncmap *p, u64 id)
+{
+	GEM_BUG_ON(p->height);
+	return id >> SHIFT;
+}
+
+static inline bool seqno_later(u32 a, u32 b)
+{
+	return (s32)(a - b) >= 0;
+}
+
+/**
+ * i915_syncmap_is_later -- compare against the last know sync point
+ * @root - pointer to the #i915_syncmap
+ * @id - the context id (other timeline) we are synchronising to
+ * @seqno - the sequence number along the other timeline
+ *
+ * If we have already synchronised this @root timeline with another (@id) then
+ * we can omit any repeated or earlier synchronisation requests. If the two
+ * timelines are already coupled, we can also omit the dependency between the
+ * two as that is already known via the timeline.
+ *
+ * Returns true if the two timelines are already synchronised wrt to @seqno,
+ * false if not and the synchronisation must be emitted.
+ */
+bool i915_syncmap_is_later(struct i915_syncmap **root, u64 id, u32 seqno)
+{
+	struct i915_syncmap *p;
+	unsigned int idx;
+
+	p = *root;
+	if (!p)
+		return false;
+
+	if (likely(__sync_leaf_prefix(p, id) == p->prefix))
+		goto found;
+
+	/* First climb the tree back to a parent branch */
+	do {
+		p = p->parent;
+		if (!p)
+			return false;
+
+		if (__sync_branch_prefix(p, id) == p->prefix)
+			break;
+	} while (1);
+
+	/* And then descend again until we find our leaf */
+	do {
+		if (!p->height)
+			break;
+
+		p = __sync_child(p)[__sync_branch_idx(p, id)];
+		if (!p)
+			return false;
+
+		if (__sync_branch_prefix(p, id) != p->prefix)
+			return false;
+	} while (1);
+
+	*root = p;
+found:
+	idx = __sync_leaf_idx(p, id);
+	if (!(p->bitmap & BIT(idx)))
+		return false;
+
+	return seqno_later(__sync_seqno(p)[idx], seqno);
+}
+
+static struct i915_syncmap *
+__sync_alloc_leaf(struct i915_syncmap *parent, u64 id)
+{
+	struct i915_syncmap *p;
+
+	p = kmalloc(sizeof(*p) + KSYNCMAP * sizeof(u32), GFP_KERNEL);
+	if (unlikely(!p))
+		return NULL;
+
+	p->parent = parent;
+	p->height = 0;
+	p->bitmap = 0;
+	p->prefix = __sync_leaf_prefix(p, id);
+	return p;
+}
+
+static inline void __sync_set_seqno(struct i915_syncmap *p, u64 id, u32 seqno)
+{
+	unsigned int idx = __sync_leaf_idx(p, id);
+
+	p->bitmap |= BIT(idx);
+	__sync_seqno(p)[idx] = seqno;
+}
+
+static inline void __sync_set_child(struct i915_syncmap *p,
+				    unsigned int idx,
+				    struct i915_syncmap *child)
+{
+	p->bitmap |= BIT(idx);
+	__sync_child(p)[idx] = child;
+}
+
+static noinline int __sync_set(struct i915_syncmap **root, u64 id, u32 seqno)
+{
+	struct i915_syncmap *p = *root;
+	unsigned int idx;
+
+	if (!p) {
+		p = __sync_alloc_leaf(NULL, id);
+		if (unlikely(!p))
+			return -ENOMEM;
+
+		goto found;
+	}
+
+	/* Caller handled the likely cached case */
+	GEM_BUG_ON(__sync_leaf_prefix(p, id) == p->prefix);
+
+	/* Climb back up the tree until we find a common prefix */
+	do {
+		if (!p->parent)
+			break;
+
+		p = p->parent;
+
+		if (__sync_branch_prefix(p, id) == p->prefix)
+			break;
+	} while (1);
+
+	/*
+	 * No shortcut, we have to descend the tree to find the right layer
+	 * containing this fence.
+	 *
+	 * Each layer in the tree holds 16 (KSYNCMAP) pointers, either fences
+	 * or lower layers. Leaf nodes (height = 0) contain the fences, all
+	 * other nodes (height > 0) are internal layers that point to a lower
+	 * node. Each internal layer has at least 2 descendents.
+	 *
+	 * Starting at the top, we check whether the current prefix matches. If
+	 * it doesn't, we have gone past our target and need to insert a join
+	 * into the tree, and a new leaf node for the target as a descendent
+	 * of the join, as well as the original layer.
+	 *
+	 * The matching prefix means we are still following the right branch
+	 * of the tree. If it has height 0, we have found our leaf and just
+	 * need to replace the fence slot with ourselves. If the height is
+	 * not zero, our slot contains the next layer in the tree (unless
+	 * it is empty, in which case we can add ourselves as a new leaf).
+	 * As descend the tree the prefix grows (and height decreases).
+	 */
+	do {
+		struct i915_syncmap *next;
+
+		if (__sync_branch_prefix(p, id) != p->prefix) {
+			unsigned int above;
+
+			/* Insert a join above the current layer */
+			next = kzalloc(sizeof(*next) + KSYNCMAP * sizeof(next),
+				       GFP_KERNEL);
+			if (unlikely(!next))
+				return -ENOMEM;
+
+			/* Compute the height at which these two diverge */
+			above = fls64(__sync_branch_prefix(p, id) ^ p->prefix);
+			above = round_up(above, SHIFT);
+			next->height = above + p->height;
+			next->prefix = __sync_branch_prefix(next, id);
+
+			/* Insert the join into the parent */
+			if (p->parent) {
+				idx = __sync_branch_idx(p->parent, id);
+				__sync_child(p->parent)[idx] = next;
+				GEM_BUG_ON(!(p->parent->bitmap & BIT(idx)));
+			}
+			next->parent = p->parent;
+
+			/* Compute the idx of the other branch, not our id! */
+			idx = p->prefix >> (above - SHIFT) & MASK;
+			__sync_set_child(next, idx, p);
+			p->parent = next;
+
+			/* Ascend to the join */
+			p = next;
+		} else {
+			if (!p->height)
+				break;
+		}
+
+		/* Descend into the next layer */
+		GEM_BUG_ON(!p->height);
+		idx = __sync_branch_idx(p, id);
+		next = __sync_child(p)[idx];
+		if (!next) {
+			next = __sync_alloc_leaf(p, id);
+			if (unlikely(!next))
+				return -ENOMEM;
+
+			__sync_set_child(p, idx, next);
+			p = next;
+			break;
+		}
+
+		p = next;
+	} while (1);
+
+found:
+	GEM_BUG_ON(p->prefix != __sync_leaf_prefix(p, id));
+	__sync_set_seqno(p, id, seqno);
+	*root = p;
+	return 0;
+}
+
+/**
+ * i915_syncmap_set -- mark the most recent syncpoint between contexts
+ * @root - pointer to the #i915_syncmap
+ * @id - the context id (other timeline) we have synchronised to
+ * @seqno - the sequence number along the other timeline
+ *
+ * When we synchronise this @root timeline with another (@id), we also know
+ * that we have synchronized with all previous seqno along that timeline. If
+ * we then have a request to synchronise with the same seqno or older, we can
+ * omit it, see i915_syncmap_is_later()
+ *
+ * Returns 0 on success, or a negative error code.
+ */
+int i915_syncmap_set(struct i915_syncmap **root, u64 id, u32 seqno)
+{
+	struct i915_syncmap *p = *root;
+
+	/*
+	 * We expect to be called in sequence following is_later(id), which
+	 * should have preloaded the root for us.
+	 */
+	if (likely(p && __sync_leaf_prefix(p, id) == p->prefix)) {
+		__sync_set_seqno(p, id, seqno);
+		return 0;
+	}
+
+	return __sync_set(root, id, seqno);
+}
+
+static void __sync_free(struct i915_syncmap *p)
+{
+	if (p->height) {
+		unsigned int i;
+
+		while ((i = ffs(p->bitmap))) {
+			p->bitmap &= ~0u << i;
+			__sync_free(__sync_child(p)[i - 1]);
+		}
+	}
+
+	kfree(p);
+}
+
+/**
+ * i915_syncmap_free -- free all memory associated with the syncmap
+ * @root - pointer to the #i915_syncmap
+ *
+ * Either when the timeline is to be freed and we no longer need the sync
+ * point tracking, or when the fences are all known to be signaled and the
+ * sync point tracking is redundant, we can free the #i915_syncmap to recover
+ * its allocations.
+ *
+ * Will reinitialise the @root pointer so that the #i915_syncmap is ready for
+ * reuse.
+ */
+void i915_syncmap_free(struct i915_syncmap **root)
+{
+	struct i915_syncmap *p;
+
+	p = *root;
+	if (!p)
+		return;
+
+	while (p->parent)
+		p = p->parent;
+
+	__sync_free(p);
+	*root = NULL;
+}
+
+#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
+#include "selftests/i915_syncmap.c"
+#endif
diff --git a/drivers/gpu/drm/i915/i915_syncmap.h b/drivers/gpu/drm/i915/i915_syncmap.h
new file mode 100644
index 000000000000..0653f70bee82
--- /dev/null
+++ b/drivers/gpu/drm/i915/i915_syncmap.h
@@ -0,0 +1,38 @@
+/*
+ * Copyright © 2017 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __I915_SYNCMAP_H__
+#define __I915_SYNCMAP_H__
+
+#include <linux/types.h>
+
+struct i915_syncmap;
+#define KSYNCMAP 16 /* radix of the tree, how many slots in each layer */
+
+void i915_syncmap_init(struct i915_syncmap **root);
+int i915_syncmap_set(struct i915_syncmap **root, u64 id, u32 seqno);
+bool i915_syncmap_is_later(struct i915_syncmap **root, u64 id, u32 seqno);
+void i915_syncmap_free(struct i915_syncmap **root);
+
+#endif /* __I915_SYNCMAP_H__ */
diff --git a/drivers/gpu/drm/i915/i915_sysfs.c b/drivers/gpu/drm/i915/i915_sysfs.c
index f3fdfda5e558..1eef3fae4db3 100644
--- a/drivers/gpu/drm/i915/i915_sysfs.c
+++ b/drivers/gpu/drm/i915/i915_sysfs.c
@@ -181,13 +181,10 @@ i915_l3_write(struct file *filp, struct kobject *kobj,
 	struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
 	struct drm_device *dev = &dev_priv->drm;
 	struct i915_gem_context *ctx;
-	u32 *temp = NULL; /* Just here to make handling failures easy */
 	int slice = (int)(uintptr_t)attr->private;
+	u32 **remap_info;
 	int ret;
 
-	if (!HAS_HW_CONTEXTS(dev_priv))
-		return -ENXIO;
-
 	ret = l3_access_valid(dev_priv, offset);
 	if (ret)
 		return ret;
@@ -196,11 +193,12 @@ i915_l3_write(struct file *filp, struct kobject *kobj,
 	if (ret)
 		return ret;
 
-	if (!dev_priv->l3_parity.remap_info[slice]) {
-		temp = kzalloc(GEN7_L3LOG_SIZE, GFP_KERNEL);
-		if (!temp) {
-			mutex_unlock(&dev->struct_mutex);
-			return -ENOMEM;
+	remap_info = &dev_priv->l3_parity.remap_info[slice];
+	if (!*remap_info) {
+		*remap_info = kzalloc(GEN7_L3LOG_SIZE, GFP_KERNEL);
+		if (!*remap_info) {
+			ret = -ENOMEM;
+			goto out;
 		}
 	}
 
@@ -208,18 +206,18 @@ i915_l3_write(struct file *filp, struct kobject *kobj,
 	 * aren't propagated. Since I cannot find a stable way to reset the GPU
 	 * at this point it is left as a TODO.
 	*/
-	if (temp)
-		dev_priv->l3_parity.remap_info[slice] = temp;
-
-	memcpy(dev_priv->l3_parity.remap_info[slice] + (offset/4), buf, count);
+	memcpy(*remap_info + (offset/4), buf, count);
 
 	/* NB: We defer the remapping until we switch to the context */
 	list_for_each_entry(ctx, &dev_priv->context_list, link)
 		ctx->remap_slice |= (1<<slice);
 
+	ret = count;
+
+out:
 	mutex_unlock(&dev->struct_mutex);
 
-	return count;
+	return ret;
 }
 
 static struct bin_attribute dpf_attrs = {
diff --git a/drivers/gpu/drm/i915/i915_trace.h b/drivers/gpu/drm/i915/i915_trace.h
index 66404c5aee82..b24a83d43559 100644
--- a/drivers/gpu/drm/i915/i915_trace.h
+++ b/drivers/gpu/drm/i915/i915_trace.h
@@ -89,6 +89,55 @@ TRACE_EVENT(intel_memory_cxsr,
 		      __entry->frame[PIPE_C], __entry->scanline[PIPE_C])
 );
 
+TRACE_EVENT(g4x_wm,
+	    TP_PROTO(struct intel_crtc *crtc, const struct g4x_wm_values *wm),
+	    TP_ARGS(crtc, wm),
+
+	    TP_STRUCT__entry(
+			     __field(enum pipe, pipe)
+			     __field(u32, frame)
+			     __field(u32, scanline)
+			     __field(u16, primary)
+			     __field(u16, sprite)
+			     __field(u16, cursor)
+			     __field(u16, sr_plane)
+			     __field(u16, sr_cursor)
+			     __field(u16, sr_fbc)
+			     __field(u16, hpll_plane)
+			     __field(u16, hpll_cursor)
+			     __field(u16, hpll_fbc)
+			     __field(bool, cxsr)
+			     __field(bool, hpll)
+			     __field(bool, fbc)
+			     ),
+
+	    TP_fast_assign(
+			   __entry->pipe = crtc->pipe;
+			   __entry->frame = crtc->base.dev->driver->get_vblank_counter(crtc->base.dev,
+										       crtc->pipe);
+			   __entry->scanline = intel_get_crtc_scanline(crtc);
+			   __entry->primary = wm->pipe[crtc->pipe].plane[PLANE_PRIMARY];
+			   __entry->sprite = wm->pipe[crtc->pipe].plane[PLANE_SPRITE0];
+			   __entry->cursor = wm->pipe[crtc->pipe].plane[PLANE_CURSOR];
+			   __entry->sr_plane = wm->sr.plane;
+			   __entry->sr_cursor = wm->sr.cursor;
+			   __entry->sr_fbc = wm->sr.fbc;
+			   __entry->hpll_plane = wm->hpll.plane;
+			   __entry->hpll_cursor = wm->hpll.cursor;
+			   __entry->hpll_fbc = wm->hpll.fbc;
+			   __entry->cxsr = wm->cxsr;
+			   __entry->hpll = wm->hpll_en;
+			   __entry->fbc = wm->fbc_en;
+			   ),
+
+	    TP_printk("pipe %c, frame=%u, scanline=%u, wm %d/%d/%d, sr %s/%d/%d/%d, hpll %s/%d/%d/%d, fbc %s",
+		      pipe_name(__entry->pipe), __entry->frame, __entry->scanline,
+		      __entry->primary, __entry->sprite, __entry->cursor,
+		      yesno(__entry->cxsr), __entry->sr_plane, __entry->sr_cursor, __entry->sr_fbc,
+		      yesno(__entry->hpll), __entry->hpll_plane, __entry->hpll_cursor, __entry->hpll_fbc,
+		      yesno(__entry->fbc))
+);
+
 TRACE_EVENT(vlv_wm,
 	    TP_PROTO(struct intel_crtc *crtc, const struct vlv_wm_values *wm),
 	    TP_ARGS(crtc, wm),
diff --git a/drivers/gpu/drm/i915/i915_utils.h b/drivers/gpu/drm/i915/i915_utils.h
index c5455d36b617..16ecd1ab108d 100644
--- a/drivers/gpu/drm/i915/i915_utils.h
+++ b/drivers/gpu/drm/i915/i915_utils.h
@@ -70,20 +70,27 @@
 #define overflows_type(x, T) \
 	(sizeof(x) > sizeof(T) && (x) >> (sizeof(T) * BITS_PER_BYTE))
 
-#define ptr_mask_bits(ptr) ({						\
+#define ptr_mask_bits(ptr, n) ({					\
 	unsigned long __v = (unsigned long)(ptr);			\
-	(typeof(ptr))(__v & PAGE_MASK);					\
+	(typeof(ptr))(__v & -BIT(n));					\
 })
 
-#define ptr_unpack_bits(ptr, bits) ({					\
+#define ptr_unmask_bits(ptr, n) ((unsigned long)(ptr) & (BIT(n) - 1))
+
+#define ptr_unpack_bits(ptr, bits, n) ({				\
 	unsigned long __v = (unsigned long)(ptr);			\
-	(bits) = __v & ~PAGE_MASK;					\
-	(typeof(ptr))(__v & PAGE_MASK);					\
+	*(bits) = __v & (BIT(n) - 1);					\
+	(typeof(ptr))(__v & -BIT(n));					\
 })
 
-#define ptr_pack_bits(ptr, bits)					\
+#define ptr_pack_bits(ptr, bits, n)					\
 	((typeof(ptr))((unsigned long)(ptr) | (bits)))
 
+#define page_mask_bits(ptr) ptr_mask_bits(ptr, PAGE_SHIFT)
+#define page_unmask_bits(ptr) ptr_unmask_bits(ptr, PAGE_SHIFT)
+#define page_pack_bits(ptr, bits) ptr_pack_bits(ptr, bits, PAGE_SHIFT)
+#define page_unpack_bits(ptr, bits) ptr_unpack_bits(ptr, bits, PAGE_SHIFT)
+
 #define ptr_offset(ptr, member) offsetof(typeof(*(ptr)), member)
 
 #define fetch_and_zero(ptr) ({						\
@@ -92,4 +99,19 @@
 	__T;								\
 })
 
+#define __mask_next_bit(mask) ({					\
+	int __idx = ffs(mask) - 1;					\
+	mask &= ~BIT(__idx);						\
+	__idx;								\
+})
+
+#include <linux/list.h>
+
+static inline void __list_del_many(struct list_head *head,
+				   struct list_head *first)
+{
+	first->prev = head;
+	WRITE_ONCE(head->next, first);
+}
+
 #endif /* !__I915_UTILS_H */
diff --git a/drivers/gpu/drm/i915/i915_vgpu.c b/drivers/gpu/drm/i915/i915_vgpu.c
index 4ab8a973b61f..cf7a958e4d3c 100644
--- a/drivers/gpu/drm/i915/i915_vgpu.c
+++ b/drivers/gpu/drm/i915/i915_vgpu.c
@@ -60,8 +60,8 @@
  */
 void i915_check_vgpu(struct drm_i915_private *dev_priv)
 {
-	uint64_t magic;
-	uint32_t version;
+	u64 magic;
+	u16 version_major;
 
 	BUILD_BUG_ON(sizeof(struct vgt_if) != VGT_PVINFO_SIZE);
 
@@ -69,10 +69,8 @@ void i915_check_vgpu(struct drm_i915_private *dev_priv)
 	if (magic != VGT_MAGIC)
 		return;
 
-	version = INTEL_VGT_IF_VERSION_ENCODE(
-		__raw_i915_read16(dev_priv, vgtif_reg(version_major)),
-		__raw_i915_read16(dev_priv, vgtif_reg(version_minor)));
-	if (version != INTEL_VGT_IF_VERSION) {
+	version_major = __raw_i915_read16(dev_priv, vgtif_reg(version_major));
+	if (version_major < VGT_VERSION_MAJOR) {
 		DRM_INFO("VGT interface version mismatch!\n");
 		return;
 	}
@@ -92,6 +90,18 @@ struct _balloon_info_ {
 
 static struct _balloon_info_ bl_info;
 
+static void vgt_deballoon_space(struct i915_ggtt *ggtt,
+				struct drm_mm_node *node)
+{
+	DRM_DEBUG_DRIVER("deballoon space: range [0x%llx - 0x%llx] %llu KiB.\n",
+			 node->start,
+			 node->start + node->size,
+			 node->size / 1024);
+
+	ggtt->base.reserved -= node->size;
+	drm_mm_remove_node(node);
+}
+
 /**
  * intel_vgt_deballoon - deballoon reserved graphics address trunks
  * @dev_priv: i915 device private data
@@ -108,12 +118,8 @@ void intel_vgt_deballoon(struct drm_i915_private *dev_priv)
 
 	DRM_DEBUG("VGT deballoon.\n");
 
-	for (i = 0; i < 4; i++) {
-		if (bl_info.space[i].allocated)
-			drm_mm_remove_node(&bl_info.space[i]);
-	}
-
-	memset(&bl_info, 0, sizeof(bl_info));
+	for (i = 0; i < 4; i++)
+		vgt_deballoon_space(&dev_priv->ggtt, &bl_info.space[i]);
 }
 
 static int vgt_balloon_space(struct i915_ggtt *ggtt,
@@ -121,15 +127,20 @@ static int vgt_balloon_space(struct i915_ggtt *ggtt,
 			     unsigned long start, unsigned long end)
 {
 	unsigned long size = end - start;
+	int ret;
 
 	if (start >= end)
 		return -EINVAL;
 
 	DRM_INFO("balloon space: range [ 0x%lx - 0x%lx ] %lu KiB.\n",
 		 start, end, size / 1024);
-	return i915_gem_gtt_reserve(&ggtt->base, node,
-				    size, start, I915_COLOR_UNEVICTABLE,
-				    0);
+	ret = i915_gem_gtt_reserve(&ggtt->base, node,
+				   size, start, I915_COLOR_UNEVICTABLE,
+				   0);
+	if (!ret)
+		ggtt->base.reserved += size;
+
+	return ret;
 }
 
 /**
@@ -222,7 +233,7 @@ int intel_vgt_balloon(struct drm_i915_private *dev_priv)
 		ret = vgt_balloon_space(ggtt, &bl_info.space[3],
 					unmappable_end, ggtt_end);
 		if (ret)
-			goto err;
+			goto err_upon_mappable;
 	}
 
 	/* Mappable graphic memory ballooning */
@@ -231,7 +242,7 @@ int intel_vgt_balloon(struct drm_i915_private *dev_priv)
 					0, mappable_base);
 
 		if (ret)
-			goto err;
+			goto err_upon_unmappable;
 	}
 
 	if (mappable_end < ggtt->mappable_end) {
@@ -239,14 +250,19 @@ int intel_vgt_balloon(struct drm_i915_private *dev_priv)
 					mappable_end, ggtt->mappable_end);
 
 		if (ret)
-			goto err;
+			goto err_below_mappable;
 	}
 
 	DRM_INFO("VGT balloon successfully\n");
 	return 0;
 
+err_below_mappable:
+	vgt_deballoon_space(ggtt, &bl_info.space[0]);
+err_upon_unmappable:
+	vgt_deballoon_space(ggtt, &bl_info.space[3]);
+err_upon_mappable:
+	vgt_deballoon_space(ggtt, &bl_info.space[2]);
 err:
 	DRM_ERROR("VGT balloon fail\n");
-	intel_vgt_deballoon(dev_priv);
 	return ret;
 }
diff --git a/drivers/gpu/drm/i915/intel_atomic.c b/drivers/gpu/drm/i915/intel_atomic.c
index 50fb1f76cc5f..d791b3ef89b5 100644
--- a/drivers/gpu/drm/i915/intel_atomic.c
+++ b/drivers/gpu/drm/i915/intel_atomic.c
@@ -36,44 +36,121 @@
 #include "intel_drv.h"
 
 /**
- * intel_connector_atomic_get_property - fetch connector property value
- * @connector: connector to fetch property for
- * @state: state containing the property value
- * @property: property to look up
- * @val: pointer to write property value into
+ * intel_digital_connector_atomic_get_property - hook for connector->atomic_get_property.
+ * @connector: Connector to get the property for.
+ * @state: Connector state to retrieve the property from.
+ * @property: Property to retrieve.
+ * @val: Return value for the property.
  *
- * The DRM core does not store shadow copies of properties for
- * atomic-capable drivers.  This entrypoint is used to fetch
- * the current value of a driver-specific connector property.
+ * Returns the atomic property value for a digital connector.
  */
-int
-intel_connector_atomic_get_property(struct drm_connector *connector,
-				    const struct drm_connector_state *state,
-				    struct drm_property *property,
-				    uint64_t *val)
+int intel_digital_connector_atomic_get_property(struct drm_connector *connector,
+						const struct drm_connector_state *state,
+						struct drm_property *property,
+						uint64_t *val)
 {
-	int i;
+	struct drm_device *dev = connector->dev;
+	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct intel_digital_connector_state *intel_conn_state =
+		to_intel_digital_connector_state(state);
+
+	if (property == dev_priv->force_audio_property)
+		*val = intel_conn_state->force_audio;
+	else if (property == dev_priv->broadcast_rgb_property)
+		*val = intel_conn_state->broadcast_rgb;
+	else {
+		DRM_DEBUG_ATOMIC("Unknown property %s\n", property->name);
+		return -EINVAL;
+	}
 
-	/*
-	 * TODO: We only have atomic modeset for planes at the moment, so the
-	 * crtc/connector code isn't quite ready yet.  Until it's ready,
-	 * continue to look up all property values in the DRM's shadow copy
-	 * in obj->properties->values[].
-	 *
-	 * When the crtc/connector state work matures, this function should
-	 * be updated to read the values out of the state structure instead.
-	 */
-	for (i = 0; i < connector->base.properties->count; i++) {
-		if (connector->base.properties->properties[i] == property) {
-			*val = connector->base.properties->values[i];
-			return 0;
-		}
+	return 0;
+}
+
+/**
+ * intel_digital_connector_atomic_set_property - hook for connector->atomic_set_property.
+ * @connector: Connector to set the property for.
+ * @state: Connector state to set the property on.
+ * @property: Property to set.
+ * @val: New value for the property.
+ *
+ * Sets the atomic property value for a digital connector.
+ */
+int intel_digital_connector_atomic_set_property(struct drm_connector *connector,
+						struct drm_connector_state *state,
+						struct drm_property *property,
+						uint64_t val)
+{
+	struct drm_device *dev = connector->dev;
+	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct intel_digital_connector_state *intel_conn_state =
+		to_intel_digital_connector_state(state);
+
+	if (property == dev_priv->force_audio_property) {
+		intel_conn_state->force_audio = val;
+		return 0;
 	}
 
+	if (property == dev_priv->broadcast_rgb_property) {
+		intel_conn_state->broadcast_rgb = val;
+		return 0;
+	}
+
+	DRM_DEBUG_ATOMIC("Unknown property %s\n", property->name);
 	return -EINVAL;
 }
 
-/*
+int intel_digital_connector_atomic_check(struct drm_connector *conn,
+					 struct drm_connector_state *new_state)
+{
+	struct intel_digital_connector_state *new_conn_state =
+		to_intel_digital_connector_state(new_state);
+	struct drm_connector_state *old_state =
+		drm_atomic_get_old_connector_state(new_state->state, conn);
+	struct intel_digital_connector_state *old_conn_state =
+		to_intel_digital_connector_state(old_state);
+	struct drm_crtc_state *crtc_state;
+
+	if (!new_state->crtc)
+		return 0;
+
+	crtc_state = drm_atomic_get_new_crtc_state(new_state->state, new_state->crtc);
+
+	/*
+	 * These properties are handled by fastset, and might not end
+	 * up in a modeset.
+	 */
+	if (new_conn_state->force_audio != old_conn_state->force_audio ||
+	    new_conn_state->broadcast_rgb != old_conn_state->broadcast_rgb ||
+	    new_conn_state->base.picture_aspect_ratio != old_conn_state->base.picture_aspect_ratio ||
+	    new_conn_state->base.scaling_mode != old_conn_state->base.scaling_mode)
+		crtc_state->mode_changed = true;
+
+	return 0;
+}
+
+/**
+ * intel_digital_connector_duplicate_state - duplicate connector state
+ * @connector: digital connector
+ *
+ * Allocates and returns a copy of the connector state (both common and
+ * digital connector specific) for the specified connector.
+ *
+ * Returns: The newly allocated connector state, or NULL on failure.
+ */
+struct drm_connector_state *
+intel_digital_connector_duplicate_state(struct drm_connector *connector)
+{
+	struct intel_digital_connector_state *state;
+
+	state = kmemdup(connector->state, sizeof(*state), GFP_KERNEL);
+	if (!state)
+		return NULL;
+
+	__drm_atomic_helper_connector_duplicate_state(connector, &state->base);
+	return &state->base;
+}
+
+/**
  * intel_crtc_duplicate_state - duplicate crtc state
  * @crtc: drm crtc
  *
diff --git a/drivers/gpu/drm/i915/intel_atomic_plane.c b/drivers/gpu/drm/i915/intel_atomic_plane.c
index cfb47293fd53..4325cb0a04f5 100644
--- a/drivers/gpu/drm/i915/intel_atomic_plane.c
+++ b/drivers/gpu/drm/i915/intel_atomic_plane.c
@@ -55,7 +55,7 @@ intel_create_plane_state(struct drm_plane *plane)
 		return NULL;
 
 	state->base.plane = plane;
-	state->base.rotation = DRM_ROTATE_0;
+	state->base.rotation = DRM_MODE_ROTATE_0;
 	state->ckey.flags = I915_SET_COLORKEY_NONE;
 
 	return state;
@@ -102,23 +102,7 @@ void
 intel_plane_destroy_state(struct drm_plane *plane,
 			  struct drm_plane_state *state)
 {
-	struct i915_vma *vma;
-
-	vma = fetch_and_zero(&to_intel_plane_state(state)->vma);
-
-	/*
-	 * FIXME: Normally intel_cleanup_plane_fb handles destruction of vma.
-	 * We currently don't clear all planes during driver unload, so we have
-	 * to be able to unpin vma here for now.
-	 *
-	 * Normally this can only happen during unload when kmscon is disabled
-	 * and userspace doesn't attempt to set a framebuffer at all.
-	 */
-	if (vma) {
-		mutex_lock(&plane->dev->struct_mutex);
-		intel_unpin_fb_vma(vma);
-		mutex_unlock(&plane->dev->struct_mutex);
-	}
+	WARN_ON(to_intel_plane_state(state)->vma);
 
 	drm_atomic_helper_plane_destroy_state(plane, state);
 }
@@ -178,14 +162,14 @@ int intel_plane_atomic_check_with_state(struct intel_crtc_state *crtc_state,
 
 	/* CHV ignores the mirror bit when the rotate bit is set :( */
 	if (IS_CHERRYVIEW(dev_priv) &&
-	    state->rotation & DRM_ROTATE_180 &&
-	    state->rotation & DRM_REFLECT_X) {
+	    state->rotation & DRM_MODE_ROTATE_180 &&
+	    state->rotation & DRM_MODE_REFLECT_X) {
 		DRM_DEBUG_KMS("Cannot rotate and reflect at the same time\n");
 		return -EINVAL;
 	}
 
 	intel_state->base.visible = false;
-	ret = intel_plane->check_plane(plane, crtc_state, intel_state);
+	ret = intel_plane->check_plane(intel_plane, crtc_state, intel_state);
 	if (ret)
 		return ret;
 
@@ -235,14 +219,14 @@ static void intel_plane_atomic_update(struct drm_plane *plane,
 		trace_intel_update_plane(plane,
 					 to_intel_crtc(crtc));
 
-		intel_plane->update_plane(plane,
+		intel_plane->update_plane(intel_plane,
 					  to_intel_crtc_state(crtc->state),
 					  intel_state);
 	} else {
 		trace_intel_disable_plane(plane,
 					  to_intel_crtc(crtc));
 
-		intel_plane->disable_plane(plane, crtc);
+		intel_plane->disable_plane(intel_plane, to_intel_crtc(crtc));
 	}
 }
 
diff --git a/drivers/gpu/drm/i915/intel_audio.c b/drivers/gpu/drm/i915/intel_audio.c
index 52c207e81f41..d805b6e6fe71 100644
--- a/drivers/gpu/drm/i915/intel_audio.c
+++ b/drivers/gpu/drm/i915/intel_audio.c
@@ -632,20 +632,9 @@ void intel_audio_codec_enable(struct intel_encoder *intel_encoder,
 						 (int) port, (int) pipe);
 	}
 
-	switch (intel_encoder->type) {
-	case INTEL_OUTPUT_HDMI:
-		intel_lpe_audio_notify(dev_priv, connector->eld, port, pipe,
-				       crtc_state->port_clock,
-				       false, 0);
-		break;
-	case INTEL_OUTPUT_DP:
-		intel_lpe_audio_notify(dev_priv, connector->eld, port, pipe,
-				       adjusted_mode->crtc_clock,
-				       true, crtc_state->port_clock);
-		break;
-	default:
-		break;
-	}
+	intel_lpe_audio_notify(dev_priv, pipe, port, connector->eld,
+			       crtc_state->port_clock,
+			       intel_encoder->type == INTEL_OUTPUT_DP);
 }
 
 /**
@@ -680,7 +669,7 @@ void intel_audio_codec_disable(struct intel_encoder *intel_encoder)
 						 (int) port, (int) pipe);
 	}
 
-	intel_lpe_audio_notify(dev_priv, NULL, port, pipe, 0, false, 0);
+	intel_lpe_audio_notify(dev_priv, pipe, port, NULL, 0, false);
 }
 
 /**
diff --git a/drivers/gpu/drm/i915/intel_breadcrumbs.c b/drivers/gpu/drm/i915/intel_breadcrumbs.c
index 9ccbf26124c6..4e00e5cb9fa1 100644
--- a/drivers/gpu/drm/i915/intel_breadcrumbs.c
+++ b/drivers/gpu/drm/i915/intel_breadcrumbs.c
@@ -64,10 +64,12 @@ static unsigned long wait_timeout(void)
 
 static noinline void missed_breadcrumb(struct intel_engine_cs *engine)
 {
-	DRM_DEBUG_DRIVER("%s missed breadcrumb at %pF, irq posted? %s\n",
+	DRM_DEBUG_DRIVER("%s missed breadcrumb at %pF, irq posted? %s, current seqno=%x, last=%x\n",
 			 engine->name, __builtin_return_address(0),
 			 yesno(test_bit(ENGINE_IRQ_BREADCRUMB,
-					&engine->irq_posted)));
+					&engine->irq_posted)),
+			 intel_engine_get_seqno(engine),
+			 intel_engine_last_submit(engine));
 
 	set_bit(engine->id, &engine->i915->gpu_error.missed_irq_rings);
 }
@@ -232,7 +234,7 @@ static void enable_fake_irq(struct intel_breadcrumbs *b)
 		mod_timer(&b->hangcheck, wait_timeout());
 }
 
-static void __intel_breadcrumbs_enable_irq(struct intel_breadcrumbs *b)
+static bool __intel_breadcrumbs_enable_irq(struct intel_breadcrumbs *b)
 {
 	struct intel_engine_cs *engine =
 		container_of(b, struct intel_engine_cs, breadcrumbs);
@@ -240,7 +242,7 @@ static void __intel_breadcrumbs_enable_irq(struct intel_breadcrumbs *b)
 
 	lockdep_assert_held(&b->irq_lock);
 	if (b->irq_armed)
-		return;
+		return false;
 
 	/* The breadcrumb irq will be disarmed on the interrupt after the
 	 * waiters are signaled. This gives us a single interrupt window in
@@ -258,7 +260,7 @@ static void __intel_breadcrumbs_enable_irq(struct intel_breadcrumbs *b)
 		 * implementation to call intel_engine_wakeup()
 		 * itself when it wants to simulate a user interrupt,
 		 */
-		return;
+		return true;
 	}
 
 	/* Since we are waiting on a request, the GPU should be busy
@@ -276,6 +278,7 @@ static void __intel_breadcrumbs_enable_irq(struct intel_breadcrumbs *b)
 	}
 
 	enable_fake_irq(b);
+	return true;
 }
 
 static inline struct intel_wait *to_wait(struct rb_node *node)
@@ -327,7 +330,7 @@ static bool __intel_engine_add_wait(struct intel_engine_cs *engine,
 {
 	struct intel_breadcrumbs *b = &engine->breadcrumbs;
 	struct rb_node **p, *parent, *completed;
-	bool first;
+	bool first, armed;
 	u32 seqno;
 
 	/* Insert the request into the retirement ordered list
@@ -342,6 +345,7 @@ static bool __intel_engine_add_wait(struct intel_engine_cs *engine,
 	 * removing stale elements in the tree, we may be able to reduce the
 	 * ping-pong between the old bottom-half and ourselves as first-waiter.
 	 */
+	armed = false;
 	first = true;
 	parent = NULL;
 	completed = NULL;
@@ -397,7 +401,7 @@ static bool __intel_engine_add_wait(struct intel_engine_cs *engine,
 		 * in the unlocked read of b->irq_seqno_bh in the irq handler)
 		 * and so we miss the wake up.
 		 */
-		__intel_breadcrumbs_enable_irq(b);
+		armed = __intel_breadcrumbs_enable_irq(b);
 		spin_unlock(&b->irq_lock);
 	}
 
@@ -424,20 +428,24 @@ static bool __intel_engine_add_wait(struct intel_engine_cs *engine,
 	GEM_BUG_ON(!b->irq_armed);
 	GEM_BUG_ON(rb_first(&b->waiters) != &b->irq_wait->node);
 
-	return first;
+	return armed;
 }
 
 bool intel_engine_add_wait(struct intel_engine_cs *engine,
 			   struct intel_wait *wait)
 {
 	struct intel_breadcrumbs *b = &engine->breadcrumbs;
-	bool first;
+	bool armed;
 
 	spin_lock_irq(&b->rb_lock);
-	first = __intel_engine_add_wait(engine, wait);
+	armed = __intel_engine_add_wait(engine, wait);
 	spin_unlock_irq(&b->rb_lock);
+	if (armed)
+		return armed;
 
-	return first;
+	/* Make the caller recheck if its request has already started. */
+	return i915_seqno_passed(intel_engine_get_seqno(engine),
+				 wait->seqno - 1);
 }
 
 static inline bool chain_wakeup(struct rb_node *rb, int priority)
@@ -665,12 +673,11 @@ static int intel_breadcrumbs_signaler(void *arg)
 	return 0;
 }
 
-void intel_engine_enable_signaling(struct drm_i915_gem_request *request)
+void intel_engine_enable_signaling(struct drm_i915_gem_request *request,
+				   bool wakeup)
 {
 	struct intel_engine_cs *engine = request->engine;
 	struct intel_breadcrumbs *b = &engine->breadcrumbs;
-	struct rb_node *parent, **p;
-	bool first, wakeup;
 	u32 seqno;
 
 	/* Note that we may be called from an interrupt handler on another
@@ -703,29 +710,38 @@ void intel_engine_enable_signaling(struct drm_i915_gem_request *request)
 	 * If we are the oldest waiter, enable the irq (after which we
 	 * must double check that the seqno did not complete).
 	 */
-	wakeup = __intel_engine_add_wait(engine, &request->signaling.wait);
+	wakeup &= __intel_engine_add_wait(engine, &request->signaling.wait);
 
-	/* Now insert ourselves into the retirement ordered list of signals
-	 * on this engine. We track the oldest seqno as that will be the
-	 * first signal to complete.
-	 */
-	parent = NULL;
-	first = true;
-	p = &b->signals.rb_node;
-	while (*p) {
-		parent = *p;
-		if (i915_seqno_passed(seqno,
-				      to_signaler(parent)->signaling.wait.seqno)) {
-			p = &parent->rb_right;
-			first = false;
-		} else {
-			p = &parent->rb_left;
+	if (!__i915_gem_request_completed(request, seqno)) {
+		struct rb_node *parent, **p;
+		bool first;
+
+		/* Now insert ourselves into the retirement ordered list of
+		 * signals on this engine. We track the oldest seqno as that
+		 * will be the first signal to complete.
+		 */
+		parent = NULL;
+		first = true;
+		p = &b->signals.rb_node;
+		while (*p) {
+			parent = *p;
+			if (i915_seqno_passed(seqno,
+					      to_signaler(parent)->signaling.wait.seqno)) {
+				p = &parent->rb_right;
+				first = false;
+			} else {
+				p = &parent->rb_left;
+			}
 		}
+		rb_link_node(&request->signaling.node, parent, p);
+		rb_insert_color(&request->signaling.node, &b->signals);
+		if (first)
+			rcu_assign_pointer(b->first_signal, request);
+	} else {
+		__intel_engine_remove_wait(engine, &request->signaling.wait);
+		i915_gem_request_put(request);
+		wakeup = false;
 	}
-	rb_link_node(&request->signaling.node, parent, p);
-	rb_insert_color(&request->signaling.node, &b->signals);
-	if (first)
-		rcu_assign_pointer(b->first_signal, request);
 
 	spin_unlock(&b->rb_lock);
 
diff --git a/drivers/gpu/drm/i915/intel_cdclk.c b/drivers/gpu/drm/i915/intel_cdclk.c
index f29a226e24d8..634c89fe6377 100644
--- a/drivers/gpu/drm/i915/intel_cdclk.c
+++ b/drivers/gpu/drm/i915/intel_cdclk.c
@@ -1071,9 +1071,15 @@ static int bxt_calc_cdclk(int max_pixclk)
 
 static int glk_calc_cdclk(int max_pixclk)
 {
-	if (max_pixclk > 2 * 158400)
+	/*
+	 * FIXME: Avoid using a pixel clock that is more than 99% of the cdclk
+	 * as a temporary workaround. Use a higher cdclk instead. (Note that
+	 * intel_compute_max_dotclk() limits the max pixel clock to 99% of max
+	 * cdclk.)
+	 */
+	if (max_pixclk > DIV_ROUND_UP(2 * 158400 * 99, 100))
 		return 316800;
-	else if (max_pixclk > 2 * 79200)
+	else if (max_pixclk > DIV_ROUND_UP(2 * 79200 * 99, 100))
 		return 158400;
 	else
 		return 79200;
@@ -1664,7 +1670,11 @@ static int intel_compute_max_dotclk(struct drm_i915_private *dev_priv)
 	int max_cdclk_freq = dev_priv->max_cdclk_freq;
 
 	if (IS_GEMINILAKE(dev_priv))
-		return 2 * max_cdclk_freq;
+		/*
+		 * FIXME: Limiting to 99% as a temporary workaround. See
+		 * glk_calc_cdclk() for details.
+		 */
+		return 2 * max_cdclk_freq * 99 / 100;
 	else if (INTEL_INFO(dev_priv)->gen >= 9 ||
 		 IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
 		return max_cdclk_freq;
@@ -1770,6 +1780,30 @@ void intel_update_cdclk(struct drm_i915_private *dev_priv)
 			   DIV_ROUND_UP(dev_priv->cdclk.hw.cdclk, 1000));
 }
 
+static int cnp_rawclk(struct drm_i915_private *dev_priv)
+{
+	u32 rawclk;
+	int divider, fraction;
+
+	if (I915_READ(SFUSE_STRAP) & SFUSE_STRAP_RAW_FREQUENCY) {
+		/* 24 MHz */
+		divider = 24000;
+		fraction = 0;
+	} else {
+		/* 19.2 MHz */
+		divider = 19000;
+		fraction = 200;
+	}
+
+	rawclk = CNP_RAWCLK_DIV((divider / 1000) - 1);
+	if (fraction)
+		rawclk |= CNP_RAWCLK_FRAC(DIV_ROUND_CLOSEST(1000,
+							    fraction) - 1);
+
+	I915_WRITE(PCH_RAWCLK_FREQ, rawclk);
+	return divider + fraction;
+}
+
 static int pch_rawclk(struct drm_i915_private *dev_priv)
 {
 	return (I915_READ(PCH_RAWCLK_FREQ) & RAWCLK_FREQ_MASK) * 1000;
@@ -1817,7 +1851,10 @@ static int g4x_hrawclk(struct drm_i915_private *dev_priv)
  */
 void intel_update_rawclk(struct drm_i915_private *dev_priv)
 {
-	if (HAS_PCH_SPLIT(dev_priv))
+
+	if (HAS_PCH_CNP(dev_priv))
+		dev_priv->rawclk_freq = cnp_rawclk(dev_priv);
+	else if (HAS_PCH_SPLIT(dev_priv))
 		dev_priv->rawclk_freq = pch_rawclk(dev_priv);
 	else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
 		dev_priv->rawclk_freq = vlv_hrawclk(dev_priv);
diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c
index 2797bf37c3ac..84a1f5e85153 100644
--- a/drivers/gpu/drm/i915/intel_crt.c
+++ b/drivers/gpu/drm/i915/intel_crt.c
@@ -777,13 +777,6 @@ out:
 	return ret;
 }
 
-static int intel_crt_set_property(struct drm_connector *connector,
-				  struct drm_property *property,
-				  uint64_t value)
-{
-	return 0;
-}
-
 void intel_crt_reset(struct drm_encoder *encoder)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->dev);
@@ -814,10 +807,9 @@ static const struct drm_connector_funcs intel_crt_connector_funcs = {
 	.late_register = intel_connector_register,
 	.early_unregister = intel_connector_unregister,
 	.destroy = intel_crt_destroy,
-	.set_property = intel_crt_set_property,
+	.set_property = drm_atomic_helper_connector_set_property,
 	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
 	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
-	.atomic_get_property = intel_connector_atomic_get_property,
 };
 
 static const struct drm_connector_helper_funcs intel_crt_connector_helper_funcs = {
diff --git a/drivers/gpu/drm/i915/intel_csr.c b/drivers/gpu/drm/i915/intel_csr.c
index 1575bde0cf90..fb6af0bcdf8f 100644
--- a/drivers/gpu/drm/i915/intel_csr.c
+++ b/drivers/gpu/drm/i915/intel_csr.c
@@ -291,7 +291,7 @@ static uint32_t *parse_csr_fw(struct drm_i915_private *dev_priv,
 
 	if (IS_GEMINILAKE(dev_priv)) {
 		required_version = GLK_CSR_VERSION_REQUIRED;
-	} else if (IS_KABYLAKE(dev_priv)) {
+	} else if (IS_KABYLAKE(dev_priv) || IS_COFFEELAKE(dev_priv)) {
 		required_version = KBL_CSR_VERSION_REQUIRED;
 	} else if (IS_SKYLAKE(dev_priv)) {
 		required_version = SKL_CSR_VERSION_REQUIRED;
@@ -440,7 +440,7 @@ void intel_csr_ucode_init(struct drm_i915_private *dev_priv)
 
 	if (IS_GEMINILAKE(dev_priv))
 		csr->fw_path = I915_CSR_GLK;
-	else if (IS_KABYLAKE(dev_priv))
+	else if (IS_KABYLAKE(dev_priv) || IS_COFFEELAKE(dev_priv))
 		csr->fw_path = I915_CSR_KBL;
 	else if (IS_SKYLAKE(dev_priv))
 		csr->fw_path = I915_CSR_SKL;
diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c
index 0914ad96a71b..2d35d97d170e 100644
--- a/drivers/gpu/drm/i915/intel_ddi.c
+++ b/drivers/gpu/drm/i915/intel_ddi.c
@@ -1732,12 +1732,18 @@ static void intel_ddi_post_disable(struct intel_encoder *intel_encoder,
 	struct drm_i915_private *dev_priv = to_i915(encoder->dev);
 	enum port port = intel_ddi_get_encoder_port(intel_encoder);
 	struct intel_digital_port *dig_port = enc_to_dig_port(encoder);
+	struct intel_dp *intel_dp = NULL;
 	int type = intel_encoder->type;
 	uint32_t val;
 	bool wait = false;
 
 	/* old_crtc_state and old_conn_state are NULL when called from DP_MST */
 
+	if (type == INTEL_OUTPUT_DP || type == INTEL_OUTPUT_EDP) {
+		intel_dp = enc_to_intel_dp(encoder);
+		intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_OFF);
+	}
+
 	val = I915_READ(DDI_BUF_CTL(port));
 	if (val & DDI_BUF_CTL_ENABLE) {
 		val &= ~DDI_BUF_CTL_ENABLE;
@@ -1753,9 +1759,7 @@ static void intel_ddi_post_disable(struct intel_encoder *intel_encoder,
 	if (wait)
 		intel_wait_ddi_buf_idle(dev_priv, port);
 
-	if (type == INTEL_OUTPUT_DP || type == INTEL_OUTPUT_EDP) {
-		struct intel_dp *intel_dp = enc_to_intel_dp(encoder);
-		intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_OFF);
+	if (intel_dp) {
 		intel_edp_panel_vdd_on(intel_dp);
 		intel_edp_panel_off(intel_dp);
 	}
@@ -1841,7 +1845,7 @@ static void intel_enable_ddi(struct intel_encoder *intel_encoder,
 		if (port == PORT_A && INTEL_GEN(dev_priv) < 9)
 			intel_dp_stop_link_train(intel_dp);
 
-		intel_edp_backlight_on(intel_dp);
+		intel_edp_backlight_on(pipe_config, conn_state);
 		intel_psr_enable(intel_dp);
 		intel_edp_drrs_enable(intel_dp, pipe_config);
 	}
@@ -1871,7 +1875,7 @@ static void intel_disable_ddi(struct intel_encoder *intel_encoder,
 
 		intel_edp_drrs_disable(intel_dp, old_crtc_state);
 		intel_psr_disable(intel_dp);
-		intel_edp_backlight_off(intel_dp);
+		intel_edp_backlight_off(old_conn_state);
 	}
 }
 
diff --git a/drivers/gpu/drm/i915/intel_device_info.c b/drivers/gpu/drm/i915/intel_device_info.c
index 7d01dfe7faac..77d3214e1a77 100644
--- a/drivers/gpu/drm/i915/intel_device_info.c
+++ b/drivers/gpu/drm/i915/intel_device_info.c
@@ -51,6 +51,8 @@ static const char * const platform_names[] = {
 	PLATFORM_NAME(BROXTON),
 	PLATFORM_NAME(KABYLAKE),
 	PLATFORM_NAME(GEMINILAKE),
+	PLATFORM_NAME(COFFEELAKE),
+	PLATFORM_NAME(CANNONLAKE),
 };
 #undef PLATFORM_NAME
 
@@ -183,16 +185,15 @@ static void gen9_sseu_info_init(struct drm_i915_private *dev_priv)
 				DIV_ROUND_UP(sseu->eu_total,
 					     sseu_subslice_total(sseu)) : 0;
 	/*
-	 * SKL supports slice power gating on devices with more than
+	 * SKL+ supports slice power gating on devices with more than
 	 * one slice, and supports EU power gating on devices with
-	 * more than one EU pair per subslice. BXT supports subslice
+	 * more than one EU pair per subslice. BXT+ supports subslice
 	 * power gating on devices with more than one subslice, and
 	 * supports EU power gating on devices with more than one EU
 	 * pair per subslice.
 	*/
 	sseu->has_slice_pg =
-		(IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) &&
-		hweight8(sseu->slice_mask) > 1;
+		!IS_GEN9_LP(dev_priv) && hweight8(sseu->slice_mask) > 1;
 	sseu->has_subslice_pg =
 		IS_GEN9_LP(dev_priv) && sseu_subslice_total(sseu) > 1;
 	sseu->has_eu_pg = sseu->eu_per_subslice > 2;
@@ -327,7 +328,7 @@ void intel_device_info_runtime_init(struct drm_i915_private *dev_priv)
 	 * we don't expose the topmost plane at all to prevent ABI breakage
 	 * down the line.
 	 */
-	if (IS_GEMINILAKE(dev_priv))
+	if (IS_GEN10(dev_priv) || IS_GEMINILAKE(dev_priv))
 		for_each_pipe(dev_priv, pipe)
 			info->num_sprites[pipe] = 3;
 	else if (IS_BROXTON(dev_priv)) {
@@ -337,7 +338,7 @@ void intel_device_info_runtime_init(struct drm_i915_private *dev_priv)
 	} else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) {
 		for_each_pipe(dev_priv, pipe)
 			info->num_sprites[pipe] = 2;
-	} else if (INTEL_GEN(dev_priv) >= 5) {
+	} else if (INTEL_GEN(dev_priv) >= 5 || IS_G4X(dev_priv)) {
 		for_each_pipe(dev_priv, pipe)
 			info->num_sprites[pipe] = 1;
 	}
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 569717a12723..e09523bbcbea 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -1277,7 +1277,7 @@ static void assert_sprites_disabled(struct drm_i915_private *dev_priv,
 		I915_STATE_WARN(val & SPRITE_ENABLE,
 		     "sprite %c assertion failure, should be off on pipe %c but is still active\n",
 		     plane_name(pipe), pipe_name(pipe));
-	} else if (INTEL_GEN(dev_priv) >= 5) {
+	} else if (INTEL_GEN(dev_priv) >= 5 || IS_G4X(dev_priv)) {
 		u32 val = I915_READ(DVSCNTR(pipe));
 		I915_STATE_WARN(val & DVS_ENABLE,
 		     "sprite %c assertion failure, should be off on pipe %c but is still active\n",
@@ -2084,6 +2084,18 @@ intel_fill_fb_ggtt_view(struct i915_ggtt_view *view,
 	}
 }
 
+static unsigned int intel_cursor_alignment(const struct drm_i915_private *dev_priv)
+{
+	if (IS_I830(dev_priv))
+		return 16 * 1024;
+	else if (IS_I85X(dev_priv))
+		return 256;
+	else if (IS_I845G(dev_priv) || IS_I865G(dev_priv))
+		return 32;
+	else
+		return 4 * 1024;
+}
+
 static unsigned int intel_linear_alignment(const struct drm_i915_private *dev_priv)
 {
 	if (INTEL_INFO(dev_priv)->gen >= 9)
@@ -2386,11 +2398,17 @@ u32 intel_compute_tile_offset(int *x, int *y,
 			      const struct intel_plane_state *state,
 			      int plane)
 {
-	const struct drm_i915_private *dev_priv = to_i915(state->base.plane->dev);
+	struct intel_plane *intel_plane = to_intel_plane(state->base.plane);
+	struct drm_i915_private *dev_priv = to_i915(intel_plane->base.dev);
 	const struct drm_framebuffer *fb = state->base.fb;
 	unsigned int rotation = state->base.rotation;
 	int pitch = intel_fb_pitch(fb, plane, rotation);
-	u32 alignment = intel_surf_alignment(fb, plane);
+	u32 alignment;
+
+	if (intel_plane->id == PLANE_CURSOR)
+		alignment = intel_cursor_alignment(dev_priv);
+	else
+		alignment = intel_surf_alignment(fb, plane);
 
 	return _intel_compute_tile_offset(dev_priv, x, y, fb, plane, pitch,
 					  rotation, alignment);
@@ -2468,7 +2486,7 @@ intel_fill_fb_info(struct drm_i915_private *dev_priv,
 
 		offset = _intel_compute_tile_offset(dev_priv, &x, &y,
 						    fb, i, fb->pitches[i],
-						    DRM_ROTATE_0, tile_size);
+						    DRM_MODE_ROTATE_0, tile_size);
 		offset /= tile_size;
 
 		if (fb->modifier != DRM_FORMAT_MOD_LINEAR) {
@@ -2503,7 +2521,7 @@ intel_fill_fb_info(struct drm_i915_private *dev_priv,
 			drm_rect_rotate(&r,
 					rot_info->plane[i].width * tile_width,
 					rot_info->plane[i].height * tile_height,
-					DRM_ROTATE_270);
+					DRM_MODE_ROTATE_270);
 			x = r.x1;
 			y = r.y1;
 
@@ -2750,7 +2768,7 @@ intel_find_initial_plane_obj(struct intel_crtc *intel_crtc,
 				false);
 	intel_pre_disable_primary_noatomic(&intel_crtc->base);
 	trace_intel_disable_plane(primary, intel_crtc);
-	intel_plane->disable_plane(primary, &intel_crtc->base);
+	intel_plane->disable_plane(intel_plane, intel_crtc);
 
 	return;
 
@@ -2939,7 +2957,7 @@ int skl_check_plane_surface(struct intel_plane_state *plane_state)
 	if (drm_rotation_90_or_270(rotation))
 		drm_rect_rotate(&plane_state->base.src,
 				fb->width << 16, fb->height << 16,
-				DRM_ROTATE_270);
+				DRM_MODE_ROTATE_270);
 
 	/*
 	 * Handle the AUX surface first since
@@ -2981,10 +2999,8 @@ static u32 i9xx_plane_ctl(const struct intel_crtc_state *crtc_state,
 	if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
 		dspcntr |= DISPPLANE_PIPE_CSC_ENABLE;
 
-	if (INTEL_GEN(dev_priv) < 4) {
-		if (crtc->pipe == PIPE_B)
-			dspcntr |= DISPPLANE_SEL_PIPE_B;
-	}
+	if (INTEL_GEN(dev_priv) < 4)
+		dspcntr |= DISPPLANE_SEL_PIPE(crtc->pipe);
 
 	switch (fb->format->format) {
 	case DRM_FORMAT_C8:
@@ -3017,10 +3033,10 @@ static u32 i9xx_plane_ctl(const struct intel_crtc_state *crtc_state,
 	    fb->modifier == I915_FORMAT_MOD_X_TILED)
 		dspcntr |= DISPPLANE_TILED;
 
-	if (rotation & DRM_ROTATE_180)
+	if (rotation & DRM_MODE_ROTATE_180)
 		dspcntr |= DISPPLANE_ROTATE_180;
 
-	if (rotation & DRM_REFLECT_X)
+	if (rotation & DRM_MODE_REFLECT_X)
 		dspcntr |= DISPPLANE_MIRROR;
 
 	return dspcntr;
@@ -3048,10 +3064,10 @@ int i9xx_check_plane_surface(struct intel_plane_state *plane_state)
 		int src_w = drm_rect_width(&plane_state->base.src) >> 16;
 		int src_h = drm_rect_height(&plane_state->base.src) >> 16;
 
-		if (rotation & DRM_ROTATE_180) {
+		if (rotation & DRM_MODE_ROTATE_180) {
 			src_x += src_w - 1;
 			src_y += src_h - 1;
-		} else if (rotation & DRM_REFLECT_X) {
+		} else if (rotation & DRM_MODE_REFLECT_X) {
 			src_x += src_w - 1;
 		}
 	}
@@ -3063,14 +3079,14 @@ int i9xx_check_plane_surface(struct intel_plane_state *plane_state)
 	return 0;
 }
 
-static void i9xx_update_primary_plane(struct drm_plane *primary,
+static void i9xx_update_primary_plane(struct intel_plane *primary,
 				      const struct intel_crtc_state *crtc_state,
 				      const struct intel_plane_state *plane_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(primary->dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc);
-	struct drm_framebuffer *fb = plane_state->base.fb;
-	int plane = intel_crtc->plane;
+	struct drm_i915_private *dev_priv = to_i915(primary->base.dev);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+	enum plane plane = primary->plane;
 	u32 linear_offset;
 	u32 dspcntr = plane_state->ctl;
 	i915_reg_t reg = DSPCNTR(plane);
@@ -3081,12 +3097,12 @@ static void i9xx_update_primary_plane(struct drm_plane *primary,
 	linear_offset = intel_fb_xy_to_linear(x, y, plane_state, 0);
 
 	if (INTEL_GEN(dev_priv) >= 4)
-		intel_crtc->dspaddr_offset = plane_state->main.offset;
+		crtc->dspaddr_offset = plane_state->main.offset;
 	else
-		intel_crtc->dspaddr_offset = linear_offset;
+		crtc->dspaddr_offset = linear_offset;
 
-	intel_crtc->adjusted_x = x;
-	intel_crtc->adjusted_y = y;
+	crtc->adjusted_x = x;
+	crtc->adjusted_y = y;
 
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
 
@@ -3112,31 +3128,29 @@ static void i9xx_update_primary_plane(struct drm_plane *primary,
 	if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) {
 		I915_WRITE_FW(DSPSURF(plane),
 			      intel_plane_ggtt_offset(plane_state) +
-			      intel_crtc->dspaddr_offset);
+			      crtc->dspaddr_offset);
 		I915_WRITE_FW(DSPOFFSET(plane), (y << 16) | x);
 	} else if (INTEL_GEN(dev_priv) >= 4) {
 		I915_WRITE_FW(DSPSURF(plane),
 			      intel_plane_ggtt_offset(plane_state) +
-			      intel_crtc->dspaddr_offset);
+			      crtc->dspaddr_offset);
 		I915_WRITE_FW(DSPTILEOFF(plane), (y << 16) | x);
 		I915_WRITE_FW(DSPLINOFF(plane), linear_offset);
 	} else {
 		I915_WRITE_FW(DSPADDR(plane),
 			      intel_plane_ggtt_offset(plane_state) +
-			      intel_crtc->dspaddr_offset);
+			      crtc->dspaddr_offset);
 	}
 	POSTING_READ_FW(reg);
 
 	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
 }
 
-static void i9xx_disable_primary_plane(struct drm_plane *primary,
-				       struct drm_crtc *crtc)
+static void i9xx_disable_primary_plane(struct intel_plane *primary,
+				       struct intel_crtc *crtc)
 {
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int plane = intel_crtc->plane;
+	struct drm_i915_private *dev_priv = to_i915(primary->base.dev);
+	enum plane plane = primary->plane;
 	unsigned long irqflags;
 
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
@@ -3271,17 +3285,17 @@ static u32 skl_plane_ctl_tiling(uint64_t fb_modifier)
 static u32 skl_plane_ctl_rotation(unsigned int rotation)
 {
 	switch (rotation) {
-	case DRM_ROTATE_0:
+	case DRM_MODE_ROTATE_0:
 		break;
 	/*
-	 * DRM_ROTATE_ is counter clockwise to stay compatible with Xrandr
+	 * DRM_MODE_ROTATE_ is counter clockwise to stay compatible with Xrandr
 	 * while i915 HW rotation is clockwise, thats why this swapping.
 	 */
-	case DRM_ROTATE_90:
+	case DRM_MODE_ROTATE_90:
 		return PLANE_CTL_ROTATE_270;
-	case DRM_ROTATE_180:
+	case DRM_MODE_ROTATE_180:
 		return PLANE_CTL_ROTATE_180;
-	case DRM_ROTATE_270:
+	case DRM_MODE_ROTATE_270:
 		return PLANE_CTL_ROTATE_90;
 	default:
 		MISSING_CASE(rotation);
@@ -3321,16 +3335,15 @@ u32 skl_plane_ctl(const struct intel_crtc_state *crtc_state,
 	return plane_ctl;
 }
 
-static void skylake_update_primary_plane(struct drm_plane *plane,
+static void skylake_update_primary_plane(struct intel_plane *plane,
 					 const struct intel_crtc_state *crtc_state,
 					 const struct intel_plane_state *plane_state)
 {
-	struct drm_device *dev = plane->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc);
-	struct drm_framebuffer *fb = plane_state->base.fb;
-	enum plane_id plane_id = to_intel_plane(plane)->id;
-	enum pipe pipe = to_intel_plane(plane)->pipe;
+	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+	enum plane_id plane_id = plane->id;
+	enum pipe pipe = plane->pipe;
 	u32 plane_ctl = plane_state->ctl;
 	unsigned int rotation = plane_state->base.rotation;
 	u32 stride = skl_plane_stride(fb, 0, rotation);
@@ -3352,10 +3365,10 @@ static void skylake_update_primary_plane(struct drm_plane *plane,
 	dst_w--;
 	dst_h--;
 
-	intel_crtc->dspaddr_offset = surf_addr;
+	crtc->dspaddr_offset = surf_addr;
 
-	intel_crtc->adjusted_x = src_x;
-	intel_crtc->adjusted_y = src_y;
+	crtc->adjusted_x = src_x;
+	crtc->adjusted_y = src_y;
 
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
 
@@ -3394,13 +3407,12 @@ static void skylake_update_primary_plane(struct drm_plane *plane,
 	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
 }
 
-static void skylake_disable_primary_plane(struct drm_plane *primary,
-					  struct drm_crtc *crtc)
+static void skylake_disable_primary_plane(struct intel_plane *primary,
+					  struct intel_crtc *crtc)
 {
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	enum plane_id plane_id = to_intel_plane(primary)->id;
-	enum pipe pipe = to_intel_plane(primary)->pipe;
+	struct drm_i915_private *dev_priv = to_i915(primary->base.dev);
+	enum plane_id plane_id = primary->id;
+	enum pipe pipe = primary->pipe;
 	unsigned long irqflags;
 
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
@@ -3433,7 +3445,7 @@ static void intel_update_primary_planes(struct drm_device *dev)
 			trace_intel_update_plane(&plane->base,
 						 to_intel_crtc(crtc));
 
-			plane->update_plane(&plane->base,
+			plane->update_plane(plane,
 					    to_intel_crtc_state(crtc->state),
 					    plane_state);
 		}
@@ -4598,7 +4610,7 @@ static void cpt_verify_modeset(struct drm_device *dev, int pipe)
 
 static int
 skl_update_scaler(struct intel_crtc_state *crtc_state, bool force_detach,
-		  unsigned scaler_user, int *scaler_id, unsigned int rotation,
+		  unsigned int scaler_user, int *scaler_id,
 		  int src_w, int src_h, int dst_w, int dst_h)
 {
 	struct intel_crtc_scaler_state *scaler_state =
@@ -4607,9 +4619,12 @@ skl_update_scaler(struct intel_crtc_state *crtc_state, bool force_detach,
 		to_intel_crtc(crtc_state->base.crtc);
 	int need_scaling;
 
-	need_scaling = drm_rotation_90_or_270(rotation) ?
-		(src_h != dst_w || src_w != dst_h):
-		(src_w != dst_w || src_h != dst_h);
+	/*
+	 * Src coordinates are already rotated by 270 degrees for
+	 * the 90/270 degree plane rotation cases (to match the
+	 * GTT mapping), hence no need to account for rotation here.
+	 */
+	need_scaling = src_w != dst_w || src_h != dst_h;
 
 	/*
 	 * if plane is being disabled or scaler is no more required or force detach
@@ -4671,7 +4686,7 @@ int skl_update_scaler_crtc(struct intel_crtc_state *state)
 	const struct drm_display_mode *adjusted_mode = &state->base.adjusted_mode;
 
 	return skl_update_scaler(state, !state->base.active, SKL_CRTC_INDEX,
-		&state->scaler_state.scaler_id, DRM_ROTATE_0,
+		&state->scaler_state.scaler_id,
 		state->pipe_src_w, state->pipe_src_h,
 		adjusted_mode->crtc_hdisplay, adjusted_mode->crtc_vdisplay);
 }
@@ -4700,7 +4715,6 @@ static int skl_update_scaler_plane(struct intel_crtc_state *crtc_state,
 	ret = skl_update_scaler(crtc_state, force_detach,
 				drm_plane_index(&intel_plane->base),
 				&plane_state->scaler_id,
-				plane_state->base.rotation,
 				drm_rect_width(&plane_state->base.src) >> 16,
 				drm_rect_height(&plane_state->base.src) >> 16,
 				drm_rect_width(&plane_state->base.dst),
@@ -4861,12 +4875,9 @@ static void intel_crtc_dpms_overlay_disable(struct intel_crtc *intel_crtc)
 {
 	if (intel_crtc->overlay) {
 		struct drm_device *dev = intel_crtc->base.dev;
-		struct drm_i915_private *dev_priv = to_i915(dev);
 
 		mutex_lock(&dev->struct_mutex);
-		dev_priv->mm.interruptible = false;
 		(void) intel_overlay_switch_off(intel_crtc->overlay);
-		dev_priv->mm.interruptible = true;
 		mutex_unlock(&dev->struct_mutex);
 	}
 
@@ -5086,7 +5097,7 @@ static void intel_crtc_disable_planes(struct drm_crtc *crtc, unsigned plane_mask
 	intel_crtc_dpms_overlay_disable(intel_crtc);
 
 	drm_for_each_plane_mask(p, dev, plane_mask)
-		to_intel_plane(p)->disable_plane(p, crtc);
+		to_intel_plane(p)->disable_plane(to_intel_plane(p), intel_crtc);
 
 	/*
 	 * FIXME: Once we grow proper nuclear flip support out of this we need
@@ -5722,6 +5733,8 @@ static void i9xx_set_pll_dividers(struct intel_crtc *crtc)
 static void i9xx_crtc_enable(struct intel_crtc_state *pipe_config,
 			     struct drm_atomic_state *old_state)
 {
+	struct intel_atomic_state *old_intel_state =
+		to_intel_atomic_state(old_state);
 	struct drm_crtc *crtc = pipe_config->base.crtc;
 	struct drm_device *dev = crtc->dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
@@ -5754,7 +5767,11 @@ static void i9xx_crtc_enable(struct intel_crtc_state *pipe_config,
 
 	intel_color_load_luts(&pipe_config->base);
 
-	intel_update_watermarks(intel_crtc);
+	if (dev_priv->display.initial_watermarks != NULL)
+		dev_priv->display.initial_watermarks(old_intel_state,
+						     intel_crtc->config);
+	else
+		intel_update_watermarks(intel_crtc);
 	intel_enable_pipe(intel_crtc);
 
 	assert_vblank_disabled(crtc);
@@ -5920,9 +5937,10 @@ void intel_encoder_destroy(struct drm_encoder *encoder)
 
 /* Cross check the actual hw state with our own modeset state tracking (and it's
  * internal consistency). */
-static void intel_connector_verify_state(struct intel_connector *connector)
+static void intel_connector_verify_state(struct drm_crtc_state *crtc_state,
+					 struct drm_connector_state *conn_state)
 {
-	struct drm_crtc *crtc = connector->base.state->crtc;
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 
 	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n",
 		      connector->base.base.id,
@@ -5930,15 +5948,14 @@ static void intel_connector_verify_state(struct intel_connector *connector)
 
 	if (connector->get_hw_state(connector)) {
 		struct intel_encoder *encoder = connector->encoder;
-		struct drm_connector_state *conn_state = connector->base.state;
 
-		I915_STATE_WARN(!crtc,
+		I915_STATE_WARN(!crtc_state,
 			 "connector enabled without attached crtc\n");
 
-		if (!crtc)
+		if (!crtc_state)
 			return;
 
-		I915_STATE_WARN(!crtc->state->active,
+		I915_STATE_WARN(!crtc_state->active,
 		      "connector is active, but attached crtc isn't\n");
 
 		if (!encoder || encoder->type == INTEL_OUTPUT_DP_MST)
@@ -5950,20 +5967,30 @@ static void intel_connector_verify_state(struct intel_connector *connector)
 		I915_STATE_WARN(conn_state->crtc != encoder->base.crtc,
 			"attached encoder crtc differs from connector crtc\n");
 	} else {
-		I915_STATE_WARN(crtc && crtc->state->active,
+		I915_STATE_WARN(crtc_state && crtc_state->active,
 			"attached crtc is active, but connector isn't\n");
-		I915_STATE_WARN(!crtc && connector->base.state->best_encoder,
+		I915_STATE_WARN(!crtc_state && conn_state->best_encoder,
 			"best encoder set without crtc!\n");
 	}
 }
 
 int intel_connector_init(struct intel_connector *connector)
 {
-	drm_atomic_helper_connector_reset(&connector->base);
+	struct intel_digital_connector_state *conn_state;
 
-	if (!connector->base.state)
+	/*
+	 * Allocate enough memory to hold intel_digital_connector_state,
+	 * This might be a few bytes too many, but for connectors that don't
+	 * need it we'll free the state and allocate a smaller one on the first
+	 * succesful commit anyway.
+	 */
+	conn_state = kzalloc(sizeof(*conn_state), GFP_KERNEL);
+	if (!conn_state)
 		return -ENOMEM;
 
+	__drm_atomic_helper_connector_reset(&connector->base,
+					    &conn_state->base);
+
 	return 0;
 }
 
@@ -6378,8 +6405,8 @@ static void vlv_pllb_recal_opamp(struct drm_i915_private *dev_priv, enum pipe
 	vlv_dpio_write(dev_priv, pipe, VLV_PLL_DW9(1), reg_val);
 
 	reg_val = vlv_dpio_read(dev_priv, pipe, VLV_REF_DW13);
-	reg_val &= 0x8cffffff;
-	reg_val = 0x8c000000;
+	reg_val &= 0x00ffffff;
+	reg_val |= 0x8c000000;
 	vlv_dpio_write(dev_priv, pipe, VLV_REF_DW13, reg_val);
 
 	reg_val = vlv_dpio_read(dev_priv, pipe, VLV_PLL_DW9(1));
@@ -8183,9 +8210,6 @@ static int ironlake_crtc_compute_clock(struct intel_crtc *crtc,
 {
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct dpll reduced_clock;
-	bool has_reduced_clock = false;
-	struct intel_shared_dpll *pll;
 	const struct intel_limit *limit;
 	int refclk = 120000;
 
@@ -8227,20 +8251,14 @@ static int ironlake_crtc_compute_clock(struct intel_crtc *crtc,
 		return -EINVAL;
 	}
 
-	ironlake_compute_dpll(crtc, crtc_state,
-			      has_reduced_clock ? &reduced_clock : NULL);
+	ironlake_compute_dpll(crtc, crtc_state, NULL);
 
-	pll = intel_get_shared_dpll(crtc, crtc_state, NULL);
-	if (pll == NULL) {
+	if (!intel_get_shared_dpll(crtc, crtc_state, NULL)) {
 		DRM_DEBUG_DRIVER("failed to find PLL for pipe %c\n",
 				 pipe_name(crtc->pipe));
 		return -EINVAL;
 	}
 
-	if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_LVDS) &&
-	    has_reduced_clock)
-		crtc->lowfreq_avail = true;
-
 	return 0;
 }
 
@@ -9144,38 +9162,171 @@ out:
 	return active;
 }
 
+static u32 intel_cursor_base(const struct intel_plane_state *plane_state)
+{
+	struct drm_i915_private *dev_priv =
+		to_i915(plane_state->base.plane->dev);
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_i915_gem_object *obj = intel_fb_obj(fb);
+	u32 base;
+
+	if (INTEL_INFO(dev_priv)->cursor_needs_physical)
+		base = obj->phys_handle->busaddr;
+	else
+		base = intel_plane_ggtt_offset(plane_state);
+
+	base += plane_state->main.offset;
+
+	/* ILK+ do this automagically */
+	if (HAS_GMCH_DISPLAY(dev_priv) &&
+	    plane_state->base.rotation & DRM_MODE_ROTATE_180)
+		base += (plane_state->base.crtc_h *
+			 plane_state->base.crtc_w - 1) * fb->format->cpp[0];
+
+	return base;
+}
+
+static u32 intel_cursor_position(const struct intel_plane_state *plane_state)
+{
+	int x = plane_state->base.crtc_x;
+	int y = plane_state->base.crtc_y;
+	u32 pos = 0;
+
+	if (x < 0) {
+		pos |= CURSOR_POS_SIGN << CURSOR_X_SHIFT;
+		x = -x;
+	}
+	pos |= x << CURSOR_X_SHIFT;
+
+	if (y < 0) {
+		pos |= CURSOR_POS_SIGN << CURSOR_Y_SHIFT;
+		y = -y;
+	}
+	pos |= y << CURSOR_Y_SHIFT;
+
+	return pos;
+}
+
+static bool intel_cursor_size_ok(const struct intel_plane_state *plane_state)
+{
+	const struct drm_mode_config *config =
+		&plane_state->base.plane->dev->mode_config;
+	int width = plane_state->base.crtc_w;
+	int height = plane_state->base.crtc_h;
+
+	return width > 0 && width <= config->cursor_width &&
+		height > 0 && height <= config->cursor_height;
+}
+
+static int intel_check_cursor(struct intel_crtc_state *crtc_state,
+			      struct intel_plane_state *plane_state)
+{
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+	int src_x, src_y;
+	u32 offset;
+	int ret;
+
+	ret = drm_plane_helper_check_state(&plane_state->base,
+					   &plane_state->clip,
+					   DRM_PLANE_HELPER_NO_SCALING,
+					   DRM_PLANE_HELPER_NO_SCALING,
+					   true, true);
+	if (ret)
+		return ret;
+
+	if (!fb)
+		return 0;
+
+	if (fb->modifier != DRM_FORMAT_MOD_LINEAR) {
+		DRM_DEBUG_KMS("cursor cannot be tiled\n");
+		return -EINVAL;
+	}
+
+	src_x = plane_state->base.src_x >> 16;
+	src_y = plane_state->base.src_y >> 16;
+
+	intel_add_fb_offsets(&src_x, &src_y, plane_state, 0);
+	offset = intel_compute_tile_offset(&src_x, &src_y, plane_state, 0);
+
+	if (src_x != 0 || src_y != 0) {
+		DRM_DEBUG_KMS("Arbitrary cursor panning not supported\n");
+		return -EINVAL;
+	}
+
+	plane_state->main.offset = offset;
+
+	return 0;
+}
+
 static u32 i845_cursor_ctl(const struct intel_crtc_state *crtc_state,
 			   const struct intel_plane_state *plane_state)
 {
-	unsigned int width = plane_state->base.crtc_w;
-	unsigned int stride = roundup_pow_of_two(width) * 4;
+	const struct drm_framebuffer *fb = plane_state->base.fb;
 
-	switch (stride) {
-	default:
-		WARN_ONCE(1, "Invalid cursor width/stride, width=%u, stride=%u\n",
-			  width, stride);
-		stride = 256;
-		/* fallthrough */
+	return CURSOR_ENABLE |
+		CURSOR_GAMMA_ENABLE |
+		CURSOR_FORMAT_ARGB |
+		CURSOR_STRIDE(fb->pitches[0]);
+}
+
+static bool i845_cursor_size_ok(const struct intel_plane_state *plane_state)
+{
+	int width = plane_state->base.crtc_w;
+
+	/*
+	 * 845g/865g are only limited by the width of their cursors,
+	 * the height is arbitrary up to the precision of the register.
+	 */
+	return intel_cursor_size_ok(plane_state) && IS_ALIGNED(width, 64);
+}
+
+static int i845_check_cursor(struct intel_plane *plane,
+			     struct intel_crtc_state *crtc_state,
+			     struct intel_plane_state *plane_state)
+{
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+	int ret;
+
+	ret = intel_check_cursor(crtc_state, plane_state);
+	if (ret)
+		return ret;
+
+	/* if we want to turn off the cursor ignore width and height */
+	if (!fb)
+		return 0;
+
+	/* Check for which cursor types we support */
+	if (!i845_cursor_size_ok(plane_state)) {
+		DRM_DEBUG("Cursor dimension %dx%d not supported\n",
+			  plane_state->base.crtc_w,
+			  plane_state->base.crtc_h);
+		return -EINVAL;
+	}
+
+	switch (fb->pitches[0]) {
 	case 256:
 	case 512:
 	case 1024:
 	case 2048:
 		break;
+	default:
+		DRM_DEBUG_KMS("Invalid cursor stride (%u)\n",
+			      fb->pitches[0]);
+		return -EINVAL;
 	}
 
-	return CURSOR_ENABLE |
-		CURSOR_GAMMA_ENABLE |
-		CURSOR_FORMAT_ARGB |
-		CURSOR_STRIDE(stride);
+	plane_state->ctl = i845_cursor_ctl(crtc_state, plane_state);
+
+	return 0;
 }
 
-static void i845_update_cursor(struct drm_crtc *crtc, u32 base,
+static void i845_update_cursor(struct intel_plane *plane,
+			       const struct intel_crtc_state *crtc_state,
 			       const struct intel_plane_state *plane_state)
 {
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	uint32_t cntl = 0, size = 0;
+	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
+	u32 cntl = 0, base = 0, pos = 0, size = 0;
+	unsigned long irqflags;
 
 	if (plane_state && plane_state->base.visible) {
 		unsigned int width = plane_state->base.crtc_w;
@@ -9183,35 +9334,41 @@ static void i845_update_cursor(struct drm_crtc *crtc, u32 base,
 
 		cntl = plane_state->ctl;
 		size = (height << 12) | width;
-	}
 
-	if (intel_crtc->cursor_cntl != 0 &&
-	    (intel_crtc->cursor_base != base ||
-	     intel_crtc->cursor_size != size ||
-	     intel_crtc->cursor_cntl != cntl)) {
-		/* On these chipsets we can only modify the base/size/stride
-		 * whilst the cursor is disabled.
-		 */
-		I915_WRITE_FW(CURCNTR(PIPE_A), 0);
-		POSTING_READ_FW(CURCNTR(PIPE_A));
-		intel_crtc->cursor_cntl = 0;
+		base = intel_cursor_base(plane_state);
+		pos = intel_cursor_position(plane_state);
 	}
 
-	if (intel_crtc->cursor_base != base) {
-		I915_WRITE_FW(CURBASE(PIPE_A), base);
-		intel_crtc->cursor_base = base;
-	}
+	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
 
-	if (intel_crtc->cursor_size != size) {
+	/* On these chipsets we can only modify the base/size/stride
+	 * whilst the cursor is disabled.
+	 */
+	if (plane->cursor.base != base ||
+	    plane->cursor.size != size ||
+	    plane->cursor.cntl != cntl) {
+		I915_WRITE_FW(CURCNTR(PIPE_A), 0);
+		I915_WRITE_FW(CURBASE(PIPE_A), base);
 		I915_WRITE_FW(CURSIZE, size);
-		intel_crtc->cursor_size = size;
-	}
-
-	if (intel_crtc->cursor_cntl != cntl) {
+		I915_WRITE_FW(CURPOS(PIPE_A), pos);
 		I915_WRITE_FW(CURCNTR(PIPE_A), cntl);
-		POSTING_READ_FW(CURCNTR(PIPE_A));
-		intel_crtc->cursor_cntl = cntl;
+
+		plane->cursor.base = base;
+		plane->cursor.size = size;
+		plane->cursor.cntl = cntl;
+	} else {
+		I915_WRITE_FW(CURPOS(PIPE_A), pos);
 	}
+
+	POSTING_READ_FW(CURCNTR(PIPE_A));
+
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
+}
+
+static void i845_disable_cursor(struct intel_plane *plane,
+				struct intel_crtc *crtc)
+{
+	i845_update_cursor(plane, NULL, NULL);
 }
 
 static u32 i9xx_cursor_ctl(const struct intel_crtc_state *crtc_state,
@@ -9220,7 +9377,6 @@ static u32 i9xx_cursor_ctl(const struct intel_crtc_state *crtc_state,
 	struct drm_i915_private *dev_priv =
 		to_i915(plane_state->base.plane->dev);
 	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-	enum pipe pipe = crtc->pipe;
 	u32 cntl;
 
 	cntl = MCURSOR_GAMMA_ENABLE;
@@ -9228,7 +9384,7 @@ static u32 i9xx_cursor_ctl(const struct intel_crtc_state *crtc_state,
 	if (HAS_DDI(dev_priv))
 		cntl |= CURSOR_PIPE_CSC_ENABLE;
 
-	cntl |= pipe << 28; /* Connect to correct pipe */
+	cntl |= MCURSOR_PIPE_SELECT(crtc->pipe);
 
 	switch (plane_state->base.crtc_w) {
 	case 64:
@@ -9245,122 +9401,160 @@ static u32 i9xx_cursor_ctl(const struct intel_crtc_state *crtc_state,
 		return 0;
 	}
 
-	if (plane_state->base.rotation & DRM_ROTATE_180)
+	if (plane_state->base.rotation & DRM_MODE_ROTATE_180)
 		cntl |= CURSOR_ROTATE_180;
 
 	return cntl;
 }
 
-static void i9xx_update_cursor(struct drm_crtc *crtc, u32 base,
-			       const struct intel_plane_state *plane_state)
+static bool i9xx_cursor_size_ok(const struct intel_plane_state *plane_state)
 {
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
-	uint32_t cntl = 0;
+	struct drm_i915_private *dev_priv =
+		to_i915(plane_state->base.plane->dev);
+	int width = plane_state->base.crtc_w;
+	int height = plane_state->base.crtc_h;
 
-	if (plane_state && plane_state->base.visible)
-		cntl = plane_state->ctl;
+	if (!intel_cursor_size_ok(plane_state))
+		return false;
 
-	if (intel_crtc->cursor_cntl != cntl) {
-		I915_WRITE_FW(CURCNTR(pipe), cntl);
-		POSTING_READ_FW(CURCNTR(pipe));
-		intel_crtc->cursor_cntl = cntl;
+	/* Cursor width is limited to a few power-of-two sizes */
+	switch (width) {
+	case 256:
+	case 128:
+	case 64:
+		break;
+	default:
+		return false;
 	}
 
-	/* and commit changes on next vblank */
-	I915_WRITE_FW(CURBASE(pipe), base);
-	POSTING_READ_FW(CURBASE(pipe));
+	/*
+	 * IVB+ have CUR_FBC_CTL which allows an arbitrary cursor
+	 * height from 8 lines up to the cursor width, when the
+	 * cursor is not rotated. Everything else requires square
+	 * cursors.
+	 */
+	if (HAS_CUR_FBC(dev_priv) &&
+	    plane_state->base.rotation & DRM_MODE_ROTATE_0) {
+		if (height < 8 || height > width)
+			return false;
+	} else {
+		if (height != width)
+			return false;
+	}
 
-	intel_crtc->cursor_base = base;
+	return true;
 }
 
-/* If no-part of the cursor is visible on the framebuffer, then the GPU may hang... */
-static void intel_crtc_update_cursor(struct drm_crtc *crtc,
-				     const struct intel_plane_state *plane_state)
+static int i9xx_check_cursor(struct intel_plane *plane,
+			     struct intel_crtc_state *crtc_state,
+			     struct intel_plane_state *plane_state)
 {
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	int pipe = intel_crtc->pipe;
-	u32 base = intel_crtc->cursor_addr;
-	unsigned long irqflags;
-	u32 pos = 0;
-
-	if (plane_state) {
-		int x = plane_state->base.crtc_x;
-		int y = plane_state->base.crtc_y;
+	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+	enum pipe pipe = plane->pipe;
+	int ret;
 
-		if (x < 0) {
-			pos |= CURSOR_POS_SIGN << CURSOR_X_SHIFT;
-			x = -x;
-		}
-		pos |= x << CURSOR_X_SHIFT;
+	ret = intel_check_cursor(crtc_state, plane_state);
+	if (ret)
+		return ret;
 
-		if (y < 0) {
-			pos |= CURSOR_POS_SIGN << CURSOR_Y_SHIFT;
-			y = -y;
-		}
-		pos |= y << CURSOR_Y_SHIFT;
+	/* if we want to turn off the cursor ignore width and height */
+	if (!fb)
+		return 0;
 
-		/* ILK+ do this automagically */
-		if (HAS_GMCH_DISPLAY(dev_priv) &&
-		    plane_state->base.rotation & DRM_ROTATE_180) {
-			base += (plane_state->base.crtc_h *
-				 plane_state->base.crtc_w - 1) * 4;
-		}
+	/* Check for which cursor types we support */
+	if (!i9xx_cursor_size_ok(plane_state)) {
+		DRM_DEBUG("Cursor dimension %dx%d not supported\n",
+			  plane_state->base.crtc_w,
+			  plane_state->base.crtc_h);
+		return -EINVAL;
 	}
 
-	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
+	if (fb->pitches[0] != plane_state->base.crtc_w * fb->format->cpp[0]) {
+		DRM_DEBUG_KMS("Invalid cursor stride (%u) (cursor width %d)\n",
+			      fb->pitches[0], plane_state->base.crtc_w);
+		return -EINVAL;
+	}
 
-	I915_WRITE_FW(CURPOS(pipe), pos);
+	/*
+	 * There's something wrong with the cursor on CHV pipe C.
+	 * If it straddles the left edge of the screen then
+	 * moving it away from the edge or disabling it often
+	 * results in a pipe underrun, and often that can lead to
+	 * dead pipe (constant underrun reported, and it scans
+	 * out just a solid color). To recover from that, the
+	 * display power well must be turned off and on again.
+	 * Refuse the put the cursor into that compromised position.
+	 */
+	if (IS_CHERRYVIEW(dev_priv) && pipe == PIPE_C &&
+	    plane_state->base.visible && plane_state->base.crtc_x < 0) {
+		DRM_DEBUG_KMS("CHV cursor C not allowed to straddle the left screen edge\n");
+		return -EINVAL;
+	}
 
-	if (IS_I845G(dev_priv) || IS_I865G(dev_priv))
-		i845_update_cursor(crtc, base, plane_state);
-	else
-		i9xx_update_cursor(crtc, base, plane_state);
+	plane_state->ctl = i9xx_cursor_ctl(crtc_state, plane_state);
 
-	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
+	return 0;
 }
 
-static bool cursor_size_ok(struct drm_i915_private *dev_priv,
-			   uint32_t width, uint32_t height)
+static void i9xx_update_cursor(struct intel_plane *plane,
+			       const struct intel_crtc_state *crtc_state,
+			       const struct intel_plane_state *plane_state)
 {
-	if (width == 0 || height == 0)
-		return false;
+	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
+	enum pipe pipe = plane->pipe;
+	u32 cntl = 0, base = 0, pos = 0, fbc_ctl = 0;
+	unsigned long irqflags;
 
-	/*
-	 * 845g/865g are special in that they are only limited by
-	 * the width of their cursors, the height is arbitrary up to
-	 * the precision of the register. Everything else requires
-	 * square cursors, limited to a few power-of-two sizes.
-	 */
-	if (IS_I845G(dev_priv) || IS_I865G(dev_priv)) {
-		if ((width & 63) != 0)
-			return false;
+	if (plane_state && plane_state->base.visible) {
+		cntl = plane_state->ctl;
 
-		if (width > (IS_I845G(dev_priv) ? 64 : 512))
-			return false;
+		if (plane_state->base.crtc_h != plane_state->base.crtc_w)
+			fbc_ctl = CUR_FBC_CTL_EN | (plane_state->base.crtc_h - 1);
 
-		if (height > 1023)
-			return false;
+		base = intel_cursor_base(plane_state);
+		pos = intel_cursor_position(plane_state);
+	}
+
+	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
+
+	/*
+	 * On some platforms writing CURCNTR first will also
+	 * cause CURPOS to be armed by the CURBASE write.
+	 * Without the CURCNTR write the CURPOS write would
+	 * arm itself.
+	 *
+	 * CURCNTR and CUR_FBC_CTL are always
+	 * armed by the CURBASE write only.
+	 */
+	if (plane->cursor.base != base ||
+	    plane->cursor.size != fbc_ctl ||
+	    plane->cursor.cntl != cntl) {
+		I915_WRITE_FW(CURCNTR(pipe), cntl);
+		if (HAS_CUR_FBC(dev_priv))
+			I915_WRITE_FW(CUR_FBC_CTL(pipe), fbc_ctl);
+		I915_WRITE_FW(CURPOS(pipe), pos);
+		I915_WRITE_FW(CURBASE(pipe), base);
+
+		plane->cursor.base = base;
+		plane->cursor.size = fbc_ctl;
+		plane->cursor.cntl = cntl;
 	} else {
-		switch (width | height) {
-		case 256:
-		case 128:
-			if (IS_GEN2(dev_priv))
-				return false;
-		case 64:
-			break;
-		default:
-			return false;
-		}
+		I915_WRITE_FW(CURPOS(pipe), pos);
 	}
 
-	return true;
+	POSTING_READ_FW(CURBASE(pipe));
+
+	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
 }
 
+static void i9xx_disable_cursor(struct intel_plane *plane,
+				struct intel_crtc *crtc)
+{
+	i9xx_update_cursor(plane, NULL, NULL);
+}
+
+
 /* VESA 640x480x72Hz mode to set on the pipe */
 static struct drm_display_mode load_detect_mode = {
 	DRM_MODE("640x480", DRM_MODE_TYPE_DEFAULT, 31500, 640, 664,
@@ -9572,6 +9766,7 @@ int intel_get_load_detect_pipe(struct drm_connector *connector,
 	 */
 	if (!crtc) {
 		DRM_DEBUG_KMS("no pipe available for load-detect\n");
+		ret = -ENODEV;
 		goto fail;
 	}
 
@@ -9628,6 +9823,7 @@ found:
 		DRM_DEBUG_KMS("reusing fbdev for load-detection framebuffer\n");
 	if (IS_ERR(fb)) {
 		DRM_DEBUG_KMS("failed to allocate framebuffer for load-detection\n");
+		ret = PTR_ERR(fb);
 		goto fail;
 	}
 
@@ -10859,21 +11055,21 @@ int intel_plane_atomic_calc_changes(struct drm_crtc_state *crtc_state,
 			 turn_off, turn_on, mode_changed);
 
 	if (turn_on) {
-		if (INTEL_GEN(dev_priv) < 5)
+		if (INTEL_GEN(dev_priv) < 5 && !IS_G4X(dev_priv))
 			pipe_config->update_wm_pre = true;
 
 		/* must disable cxsr around plane enable/disable */
 		if (plane->id != PLANE_CURSOR)
 			pipe_config->disable_cxsr = true;
 	} else if (turn_off) {
-		if (INTEL_GEN(dev_priv) < 5)
+		if (INTEL_GEN(dev_priv) < 5 && !IS_G4X(dev_priv))
 			pipe_config->update_wm_post = true;
 
 		/* must disable cxsr around plane enable/disable */
 		if (plane->id != PLANE_CURSOR)
 			pipe_config->disable_cxsr = true;
 	} else if (intel_wm_need_update(&plane->base, plane_state)) {
-		if (INTEL_GEN(dev_priv) < 5) {
+		if (INTEL_GEN(dev_priv) < 5 && !IS_G4X(dev_priv)) {
 			/* FIXME bollocks */
 			pipe_config->update_wm_pre = true;
 			pipe_config->update_wm_post = true;
@@ -10999,6 +11195,9 @@ static int intel_crtc_atomic_check(struct drm_crtc *crtc,
 			ret = skl_update_scaler_crtc(pipe_config);
 
 		if (!ret)
+			ret = skl_check_pipe_max_pixel_rate(intel_crtc,
+							    pipe_config);
+		if (!ret)
 			ret = intel_atomic_setup_scalers(dev_priv, intel_crtc,
 							 pipe_config);
 	}
@@ -11297,7 +11496,8 @@ clear_intel_crtc_state(struct intel_crtc_state *crtc_state)
 	shared_dpll = crtc_state->shared_dpll;
 	dpll_hw_state = crtc_state->dpll_hw_state;
 	force_thru = crtc_state->pch_pfit.force_thru;
-	if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
+	if (IS_G4X(dev_priv) ||
+	    IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
 		wm_state = crtc_state->wm;
 
 	/* Keep base drm_crtc_state intact, only clear our extended struct */
@@ -11309,7 +11509,8 @@ clear_intel_crtc_state(struct intel_crtc_state *crtc_state)
 	crtc_state->shared_dpll = shared_dpll;
 	crtc_state->dpll_hw_state = dpll_hw_state;
 	crtc_state->pch_pfit.force_thru = force_thru;
-	if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
+	if (IS_G4X(dev_priv) ||
+	    IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
 		crtc_state->wm = wm_state;
 }
 
@@ -11450,12 +11651,6 @@ intel_modeset_update_crtc_state(struct drm_atomic_state *state)
 	for_each_new_crtc_in_state(state, crtc, new_crtc_state, i) {
 		to_intel_crtc(crtc)->config = to_intel_crtc_state(new_crtc_state);
 
-		/* Update hwmode for vblank functions */
-		if (new_crtc_state->active)
-			crtc->hwmode = new_crtc_state->adjusted_mode;
-		else
-			crtc->hwmode.crtc_clock = 0;
-
 		/*
 		 * Update legacy state to satisfy fbc code. This can
 		 * be removed when fbc uses the atomic state.
@@ -11877,7 +12072,7 @@ static void verify_wm_state(struct drm_crtc *crtc,
 	 * allocation. In that case since the ddb allocation will be updated
 	 * once the plane becomes visible, we can skip this check
 	 */
-	if (intel_crtc->cursor_addr) {
+	if (1) {
 		hw_plane_wm = &hw_wm.planes[PLANE_CURSOR];
 		sw_plane_wm = &sw_wm->planes[PLANE_CURSOR];
 
@@ -11933,11 +12128,15 @@ verify_connector_state(struct drm_device *dev,
 
 	for_each_new_connector_in_state(state, connector, new_conn_state, i) {
 		struct drm_encoder *encoder = connector->encoder;
+		struct drm_crtc_state *crtc_state = NULL;
 
 		if (new_conn_state->crtc != crtc)
 			continue;
 
-		intel_connector_verify_state(to_intel_connector(connector));
+		if (crtc)
+			crtc_state = drm_atomic_get_new_crtc_state(state, new_conn_state->crtc);
+
+		intel_connector_verify_state(crtc_state, new_conn_state);
 
 		I915_STATE_WARN(new_conn_state->best_encoder != encoder,
 		     "connector's atomic encoder doesn't match legacy encoder\n");
@@ -12055,7 +12254,7 @@ verify_crtc_state(struct drm_crtc *crtc,
 
 	intel_pipe_config_sanity_check(dev_priv, pipe_config);
 
-	sw_config = to_intel_crtc_state(crtc->state);
+	sw_config = to_intel_crtc_state(new_crtc_state);
 	if (!intel_pipe_config_compare(dev_priv, sw_config,
 				       pipe_config, false)) {
 		I915_STATE_WARN(1, "pipe state doesn't match!\n");
@@ -12928,8 +13127,16 @@ static void intel_atomic_commit_tail(struct drm_atomic_state *state)
 
 	drm_atomic_helper_commit_hw_done(state);
 
-	if (intel_state->modeset)
+	if (intel_state->modeset) {
+		/* As one of the primary mmio accessors, KMS has a high
+		 * likelihood of triggering bugs in unclaimed access. After we
+		 * finish modesetting, see if an error has been flagged, and if
+		 * so enable debugging for the next modeset - and hope we catch
+		 * the culprit.
+		 */
+		intel_uncore_arm_unclaimed_mmio_detection(dev_priv);
 		intel_display_power_put(dev_priv, POWER_DOMAIN_MODESET);
+	}
 
 	mutex_lock(&dev->struct_mutex);
 	drm_atomic_helper_cleanup_planes(dev, state);
@@ -12939,19 +13146,6 @@ static void intel_atomic_commit_tail(struct drm_atomic_state *state)
 
 	drm_atomic_state_put(state);
 
-	/* As one of the primary mmio accessors, KMS has a high likelihood
-	 * of triggering bugs in unclaimed access. After we finish
-	 * modesetting, see if an error has been flagged, and if so
-	 * enable debugging for the next modeset - and hope we catch
-	 * the culprit.
-	 *
-	 * XXX note that we assume display power is on at this point.
-	 * This might hold true now but we need to add pm helper to check
-	 * unclaimed only when the hardware is on, as atomic commits
-	 * can happen also when the device is completely off.
-	 */
-	intel_uncore_arm_unclaimed_mmio_detection(dev_priv);
-
 	intel_atomic_helper_free_state(dev_priv);
 }
 
@@ -13083,43 +13277,6 @@ static int intel_atomic_commit(struct drm_device *dev,
 	return 0;
 }
 
-void intel_crtc_restore_mode(struct drm_crtc *crtc)
-{
-	struct drm_device *dev = crtc->dev;
-	struct drm_atomic_state *state;
-	struct drm_crtc_state *crtc_state;
-	int ret;
-
-	state = drm_atomic_state_alloc(dev);
-	if (!state) {
-		DRM_DEBUG_KMS("[CRTC:%d:%s] crtc restore failed, out of memory",
-			      crtc->base.id, crtc->name);
-		return;
-	}
-
-	state->acquire_ctx = crtc->dev->mode_config.acquire_ctx;
-
-retry:
-	crtc_state = drm_atomic_get_crtc_state(state, crtc);
-	ret = PTR_ERR_OR_ZERO(crtc_state);
-	if (!ret) {
-		if (!crtc_state->active)
-			goto out;
-
-		crtc_state->mode_changed = true;
-		ret = drm_atomic_commit(state);
-	}
-
-	if (ret == -EDEADLK) {
-		drm_atomic_state_clear(state);
-		drm_modeset_backoff(state->acquire_ctx);
-		goto retry;
-	}
-
-out:
-	drm_atomic_state_put(state);
-}
-
 static const struct drm_crtc_funcs intel_crtc_funcs = {
 	.gamma_set = drm_atomic_helper_legacy_gamma_set,
 	.set_config = drm_atomic_helper_set_config,
@@ -13160,7 +13317,7 @@ intel_prepare_plane_fb(struct drm_plane *plane,
 	if (obj) {
 		if (plane->type == DRM_PLANE_TYPE_CURSOR &&
 		    INTEL_INFO(dev_priv)->cursor_needs_physical) {
-			const int align = IS_I830(dev_priv) ? 16 * 1024 : 256;
+			const int align = intel_cursor_alignment(dev_priv);
 
 			ret = i915_gem_object_attach_phys(obj, align);
 			if (ret) {
@@ -13290,11 +13447,11 @@ skl_max_scale(struct intel_crtc *intel_crtc, struct intel_crtc_state *crtc_state
 }
 
 static int
-intel_check_primary_plane(struct drm_plane *plane,
+intel_check_primary_plane(struct intel_plane *plane,
 			  struct intel_crtc_state *crtc_state,
 			  struct intel_plane_state *state)
 {
-	struct drm_i915_private *dev_priv = to_i915(plane->dev);
+	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
 	struct drm_crtc *crtc = state->base.crtc;
 	int min_scale = DRM_PLANE_HELPER_NO_SCALING;
 	int max_scale = DRM_PLANE_HELPER_NO_SCALING;
@@ -13473,7 +13630,7 @@ intel_legacy_cursor_update(struct drm_plane *plane,
 		goto out_free;
 
 	if (INTEL_INFO(dev_priv)->cursor_needs_physical) {
-		int align = IS_I830(dev_priv) ? 16 * 1024 : 256;
+		int align = intel_cursor_alignment(dev_priv);
 
 		ret = i915_gem_object_attach_phys(intel_fb_obj(fb), align);
 		if (ret) {
@@ -13509,12 +13666,12 @@ intel_legacy_cursor_update(struct drm_plane *plane,
 
 	if (plane->state->visible) {
 		trace_intel_update_plane(plane, to_intel_crtc(crtc));
-		intel_plane->update_plane(plane,
+		intel_plane->update_plane(intel_plane,
 					  to_intel_crtc_state(crtc->state),
 					  to_intel_plane_state(plane->state));
 	} else {
 		trace_intel_disable_plane(plane, to_intel_crtc(crtc));
-		intel_plane->disable_plane(plane, crtc);
+		intel_plane->disable_plane(intel_plane, to_intel_crtc(crtc));
 	}
 
 	intel_cleanup_plane_fb(plane, new_plane_state);
@@ -13628,22 +13785,22 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 
 	if (INTEL_GEN(dev_priv) >= 9) {
 		supported_rotations =
-			DRM_ROTATE_0 | DRM_ROTATE_90 |
-			DRM_ROTATE_180 | DRM_ROTATE_270;
+			DRM_MODE_ROTATE_0 | DRM_MODE_ROTATE_90 |
+			DRM_MODE_ROTATE_180 | DRM_MODE_ROTATE_270;
 	} else if (IS_CHERRYVIEW(dev_priv) && pipe == PIPE_B) {
 		supported_rotations =
-			DRM_ROTATE_0 | DRM_ROTATE_180 |
-			DRM_REFLECT_X;
+			DRM_MODE_ROTATE_0 | DRM_MODE_ROTATE_180 |
+			DRM_MODE_REFLECT_X;
 	} else if (INTEL_GEN(dev_priv) >= 4) {
 		supported_rotations =
-			DRM_ROTATE_0 | DRM_ROTATE_180;
+			DRM_MODE_ROTATE_0 | DRM_MODE_ROTATE_180;
 	} else {
-		supported_rotations = DRM_ROTATE_0;
+		supported_rotations = DRM_MODE_ROTATE_0;
 	}
 
 	if (INTEL_GEN(dev_priv) >= 4)
 		drm_plane_create_rotation_property(&primary->base,
-						   DRM_ROTATE_0,
+						   DRM_MODE_ROTATE_0,
 						   supported_rotations);
 
 	drm_plane_helper_add(&primary->base, &intel_plane_helper_funcs);
@@ -13657,107 +13814,9 @@ fail:
 	return ERR_PTR(ret);
 }
 
-static int
-intel_check_cursor_plane(struct drm_plane *plane,
-			 struct intel_crtc_state *crtc_state,
-			 struct intel_plane_state *state)
-{
-	struct drm_i915_private *dev_priv = to_i915(plane->dev);
-	struct drm_framebuffer *fb = state->base.fb;
-	struct drm_i915_gem_object *obj = intel_fb_obj(fb);
-	enum pipe pipe = to_intel_plane(plane)->pipe;
-	unsigned stride;
-	int ret;
-
-	ret = drm_plane_helper_check_state(&state->base,
-					   &state->clip,
-					   DRM_PLANE_HELPER_NO_SCALING,
-					   DRM_PLANE_HELPER_NO_SCALING,
-					   true, true);
-	if (ret)
-		return ret;
-
-	/* if we want to turn off the cursor ignore width and height */
-	if (!obj)
-		return 0;
-
-	/* Check for which cursor types we support */
-	if (!cursor_size_ok(dev_priv, state->base.crtc_w,
-			    state->base.crtc_h)) {
-		DRM_DEBUG("Cursor dimension %dx%d not supported\n",
-			  state->base.crtc_w, state->base.crtc_h);
-		return -EINVAL;
-	}
-
-	stride = roundup_pow_of_two(state->base.crtc_w) * 4;
-	if (obj->base.size < stride * state->base.crtc_h) {
-		DRM_DEBUG_KMS("buffer is too small\n");
-		return -ENOMEM;
-	}
-
-	if (fb->modifier != DRM_FORMAT_MOD_LINEAR) {
-		DRM_DEBUG_KMS("cursor cannot be tiled\n");
-		return -EINVAL;
-	}
-
-	/*
-	 * There's something wrong with the cursor on CHV pipe C.
-	 * If it straddles the left edge of the screen then
-	 * moving it away from the edge or disabling it often
-	 * results in a pipe underrun, and often that can lead to
-	 * dead pipe (constant underrun reported, and it scans
-	 * out just a solid color). To recover from that, the
-	 * display power well must be turned off and on again.
-	 * Refuse the put the cursor into that compromised position.
-	 */
-	if (IS_CHERRYVIEW(dev_priv) && pipe == PIPE_C &&
-	    state->base.visible && state->base.crtc_x < 0) {
-		DRM_DEBUG_KMS("CHV cursor C not allowed to straddle the left screen edge\n");
-		return -EINVAL;
-	}
-
-	if (IS_I845G(dev_priv) || IS_I865G(dev_priv))
-		state->ctl = i845_cursor_ctl(crtc_state, state);
-	else
-		state->ctl = i9xx_cursor_ctl(crtc_state, state);
-
-	return 0;
-}
-
-static void
-intel_disable_cursor_plane(struct drm_plane *plane,
-			   struct drm_crtc *crtc)
-{
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-
-	intel_crtc->cursor_addr = 0;
-	intel_crtc_update_cursor(crtc, NULL);
-}
-
-static void
-intel_update_cursor_plane(struct drm_plane *plane,
-			  const struct intel_crtc_state *crtc_state,
-			  const struct intel_plane_state *state)
-{
-	struct drm_crtc *crtc = crtc_state->base.crtc;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	struct drm_i915_private *dev_priv = to_i915(plane->dev);
-	struct drm_i915_gem_object *obj = intel_fb_obj(state->base.fb);
-	uint32_t addr;
-
-	if (!obj)
-		addr = 0;
-	else if (!INTEL_INFO(dev_priv)->cursor_needs_physical)
-		addr = intel_plane_ggtt_offset(state);
-	else
-		addr = obj->phys_handle->busaddr;
-
-	intel_crtc->cursor_addr = addr;
-	intel_crtc_update_cursor(crtc, state);
-}
-
 static struct intel_plane *
-intel_cursor_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
+intel_cursor_plane_create(struct drm_i915_private *dev_priv,
+			  enum pipe pipe)
 {
 	struct intel_plane *cursor = NULL;
 	struct intel_plane_state *state = NULL;
@@ -13783,9 +13842,22 @@ intel_cursor_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 	cursor->plane = pipe;
 	cursor->id = PLANE_CURSOR;
 	cursor->frontbuffer_bit = INTEL_FRONTBUFFER_CURSOR(pipe);
-	cursor->check_plane = intel_check_cursor_plane;
-	cursor->update_plane = intel_update_cursor_plane;
-	cursor->disable_plane = intel_disable_cursor_plane;
+
+	if (IS_I845G(dev_priv) || IS_I865G(dev_priv)) {
+		cursor->update_plane = i845_update_cursor;
+		cursor->disable_plane = i845_disable_cursor;
+		cursor->check_plane = i845_check_cursor;
+	} else {
+		cursor->update_plane = i9xx_update_cursor;
+		cursor->disable_plane = i9xx_disable_cursor;
+		cursor->check_plane = i9xx_check_cursor;
+	}
+
+	cursor->cursor.base = ~0;
+	cursor->cursor.cntl = ~0;
+
+	if (IS_I845G(dev_priv) || IS_I865G(dev_priv) || HAS_CUR_FBC(dev_priv))
+		cursor->cursor.size = ~0;
 
 	ret = drm_universal_plane_init(&dev_priv->drm, &cursor->base,
 				       0, &intel_cursor_plane_funcs,
@@ -13798,9 +13870,9 @@ intel_cursor_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 
 	if (INTEL_GEN(dev_priv) >= 4)
 		drm_plane_create_rotation_property(&cursor->base,
-						   DRM_ROTATE_0,
-						   DRM_ROTATE_0 |
-						   DRM_ROTATE_180);
+						   DRM_MODE_ROTATE_0,
+						   DRM_MODE_ROTATE_0 |
+						   DRM_MODE_ROTATE_180);
 
 	if (INTEL_GEN(dev_priv) >= 9)
 		state->scaler_id = -1;
@@ -13894,10 +13966,6 @@ static int intel_crtc_init(struct drm_i915_private *dev_priv, enum pipe pipe)
 	intel_crtc->pipe = pipe;
 	intel_crtc->plane = primary->plane;
 
-	intel_crtc->cursor_base = ~0;
-	intel_crtc->cursor_cntl = ~0;
-	intel_crtc->cursor_size = ~0;
-
 	/* initialize shared scalers */
 	intel_crtc_init_scalers(intel_crtc, crtc_state);
 
@@ -14437,7 +14505,7 @@ static int intel_framebuffer_init(struct intel_framebuffer *intel_fb,
 	case DRM_FORMAT_UYVY:
 	case DRM_FORMAT_YVYU:
 	case DRM_FORMAT_VYUY:
-		if (INTEL_GEN(dev_priv) < 5) {
+		if (INTEL_GEN(dev_priv) < 5 && !IS_G4X(dev_priv)) {
 			DRM_DEBUG_KMS("unsupported pixel format: %s\n",
 				      drm_get_format_name(mode_cmd->pixel_format, &format_name));
 			goto err;
@@ -14949,6 +15017,7 @@ int intel_modeset_init(struct drm_device *dev)
 
 	dev->mode_config.funcs = &intel_mode_funcs;
 
+	init_llist_head(&dev_priv->atomic_helper.free_list);
 	INIT_WORK(&dev_priv->atomic_helper.free_work,
 		  intel_atomic_helper_free_state_worker);
 
@@ -15170,7 +15239,7 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc)
 				continue;
 
 			trace_intel_disable_plane(&plane->base, crtc);
-			plane->disable_plane(&plane->base, &crtc->base);
+			plane->disable_plane(plane, crtc);
 		}
 	}
 
@@ -15440,8 +15509,6 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 			to_intel_crtc_state(crtc->base.state);
 		int pixclk = 0;
 
-		crtc->base.hwmode = crtc_state->base.adjusted_mode;
-
 		memset(&crtc->base.mode, 0, sizeof(crtc->base.mode));
 		if (crtc_state->base.active) {
 			intel_mode_from_pipe_config(&crtc->base.mode, crtc_state);
@@ -15471,7 +15538,8 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 			if (IS_BROADWELL(dev_priv) && crtc_state->ips_enabled)
 				pixclk = DIV_ROUND_UP(pixclk * 100, 95);
 
-			drm_calc_timestamping_constants(&crtc->base, &crtc->base.hwmode);
+			drm_calc_timestamping_constants(&crtc->base,
+							&crtc_state->base.adjusted_mode);
 			update_scanline_offset(crtc);
 		}
 
@@ -15542,7 +15610,10 @@ intel_modeset_setup_hw_state(struct drm_device *dev)
 		pll->on = false;
 	}
 
-	if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) {
+	if (IS_G4X(dev_priv)) {
+		g4x_wm_get_hw_state(dev);
+		g4x_wm_sanitize(dev_priv);
+	} else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) {
 		vlv_wm_get_hw_state(dev);
 		vlv_wm_sanitize(dev_priv);
 	} else if (IS_GEN9(dev_priv)) {
@@ -15576,13 +15647,6 @@ void intel_display_resume(struct drm_device *dev)
 	if (state)
 		state->acquire_ctx = &ctx;
 
-	/*
-	 * This is a cludge because with real atomic modeset mode_config.mutex
-	 * won't be taken. Unfortunately some probed state like
-	 * audio_codec_enable is still protected by mode_config.mutex, so lock
-	 * it here for now.
-	 */
-	mutex_lock(&dev->mode_config.mutex);
 	drm_modeset_acquire_init(&ctx, 0);
 
 	while (1) {
@@ -15598,7 +15662,6 @@ void intel_display_resume(struct drm_device *dev)
 
 	drm_modeset_drop_locks(&ctx);
 	drm_modeset_acquire_fini(&ctx);
-	mutex_unlock(&dev->mode_config.mutex);
 
 	if (ret)
 		DRM_ERROR("Restoring old state failed with %i\n", ret);
diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index fc691b8b317c..950646fab48c 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -133,36 +133,55 @@ static void vlv_steal_power_sequencer(struct drm_device *dev,
 				      enum pipe pipe);
 static void intel_dp_unset_edid(struct intel_dp *intel_dp);
 
-static int
-intel_dp_max_link_bw(struct intel_dp  *intel_dp)
+static int intel_dp_num_rates(u8 link_bw_code)
 {
-	int max_link_bw = intel_dp->dpcd[DP_MAX_LINK_RATE];
-
-	switch (max_link_bw) {
+	switch (link_bw_code) {
+	default:
+		WARN(1, "invalid max DP link bw val %x, using 1.62Gbps\n",
+		     link_bw_code);
 	case DP_LINK_BW_1_62:
+		return 1;
 	case DP_LINK_BW_2_7:
+		return 2;
 	case DP_LINK_BW_5_4:
-		break;
-	default:
-		WARN(1, "invalid max DP link bw val %x, using 1.62Gbps\n",
-		     max_link_bw);
-		max_link_bw = DP_LINK_BW_1_62;
-		break;
+		return 3;
 	}
-	return max_link_bw;
 }
 
-static u8 intel_dp_max_lane_count(struct intel_dp *intel_dp)
+/* update sink rates from dpcd */
+static void intel_dp_set_sink_rates(struct intel_dp *intel_dp)
 {
-	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
-	u8 source_max, sink_max;
+	int i, num_rates;
+
+	num_rates = intel_dp_num_rates(intel_dp->dpcd[DP_MAX_LINK_RATE]);
 
-	source_max = intel_dig_port->max_lanes;
-	sink_max = intel_dp->max_sink_lane_count;
+	for (i = 0; i < num_rates; i++)
+		intel_dp->sink_rates[i] = default_rates[i];
+
+	intel_dp->num_sink_rates = num_rates;
+}
+
+/* Theoretical max between source and sink */
+static int intel_dp_max_common_rate(struct intel_dp *intel_dp)
+{
+	return intel_dp->common_rates[intel_dp->num_common_rates - 1];
+}
+
+/* Theoretical max between source and sink */
+static int intel_dp_max_common_lane_count(struct intel_dp *intel_dp)
+{
+	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
+	int source_max = intel_dig_port->max_lanes;
+	int sink_max = drm_dp_max_lane_count(intel_dp->dpcd);
 
 	return min(source_max, sink_max);
 }
 
+int intel_dp_max_lane_count(struct intel_dp *intel_dp)
+{
+	return intel_dp->max_link_lane_count;
+}
+
 int
 intel_dp_link_required(int pixel_clock, int bpp)
 {
@@ -205,34 +224,25 @@ intel_dp_downstream_max_dotclock(struct intel_dp *intel_dp)
 	return max_dotclk;
 }
 
-static int
-intel_dp_sink_rates(struct intel_dp *intel_dp, const int **sink_rates)
-{
-	if (intel_dp->num_sink_rates) {
-		*sink_rates = intel_dp->sink_rates;
-		return intel_dp->num_sink_rates;
-	}
-
-	*sink_rates = default_rates;
-
-	return (intel_dp->max_sink_link_bw >> 3) + 1;
-}
-
-static int
-intel_dp_source_rates(struct intel_dp *intel_dp, const int **source_rates)
+static void
+intel_dp_set_source_rates(struct intel_dp *intel_dp)
 {
 	struct intel_digital_port *dig_port = dp_to_dig_port(intel_dp);
 	struct drm_i915_private *dev_priv = to_i915(dig_port->base.base.dev);
+	const int *source_rates;
 	int size;
 
+	/* This should only be done once */
+	WARN_ON(intel_dp->source_rates || intel_dp->num_source_rates);
+
 	if (IS_GEN9_LP(dev_priv)) {
-		*source_rates = bxt_rates;
+		source_rates = bxt_rates;
 		size = ARRAY_SIZE(bxt_rates);
 	} else if (IS_GEN9_BC(dev_priv)) {
-		*source_rates = skl_rates;
+		source_rates = skl_rates;
 		size = ARRAY_SIZE(skl_rates);
 	} else {
-		*source_rates = default_rates;
+		source_rates = default_rates;
 		size = ARRAY_SIZE(default_rates);
 	}
 
@@ -240,7 +250,8 @@ intel_dp_source_rates(struct intel_dp *intel_dp, const int **source_rates)
 	if (!intel_dp_source_supports_hbr2(intel_dp))
 		size--;
 
-	return size;
+	intel_dp->source_rates = source_rates;
+	intel_dp->num_source_rates = size;
 }
 
 static int intersect_rates(const int *source_rates, int source_len,
@@ -266,50 +277,83 @@ static int intersect_rates(const int *source_rates, int source_len,
 	return k;
 }
 
-static int intel_dp_common_rates(struct intel_dp *intel_dp,
-				 int *common_rates)
+/* return index of rate in rates array, or -1 if not found */
+static int intel_dp_rate_index(const int *rates, int len, int rate)
 {
-	const int *source_rates, *sink_rates;
-	int source_len, sink_len;
+	int i;
 
-	sink_len = intel_dp_sink_rates(intel_dp, &sink_rates);
-	source_len = intel_dp_source_rates(intel_dp, &source_rates);
+	for (i = 0; i < len; i++)
+		if (rate == rates[i])
+			return i;
 
-	return intersect_rates(source_rates, source_len,
-			       sink_rates, sink_len,
-			       common_rates);
+	return -1;
 }
 
-static int intel_dp_link_rate_index(struct intel_dp *intel_dp,
-				    int *common_rates, int link_rate)
+static void intel_dp_set_common_rates(struct intel_dp *intel_dp)
 {
-	int common_len;
-	int index;
+	WARN_ON(!intel_dp->num_source_rates || !intel_dp->num_sink_rates);
 
-	common_len = intel_dp_common_rates(intel_dp, common_rates);
-	for (index = 0; index < common_len; index++) {
-		if (link_rate == common_rates[common_len - index - 1])
-			return common_len - index - 1;
+	intel_dp->num_common_rates = intersect_rates(intel_dp->source_rates,
+						     intel_dp->num_source_rates,
+						     intel_dp->sink_rates,
+						     intel_dp->num_sink_rates,
+						     intel_dp->common_rates);
+
+	/* Paranoia, there should always be something in common. */
+	if (WARN_ON(intel_dp->num_common_rates == 0)) {
+		intel_dp->common_rates[0] = default_rates[0];
+		intel_dp->num_common_rates = 1;
 	}
+}
 
-	return -1;
+/* get length of common rates potentially limited by max_rate */
+static int intel_dp_common_len_rate_limit(struct intel_dp *intel_dp,
+					  int max_rate)
+{
+	const int *common_rates = intel_dp->common_rates;
+	int i, common_len = intel_dp->num_common_rates;
+
+	/* Limit results by potentially reduced max rate */
+	for (i = 0; i < common_len; i++) {
+		if (common_rates[common_len - i - 1] <= max_rate)
+			return common_len - i;
+	}
+
+	return 0;
+}
+
+static bool intel_dp_link_params_valid(struct intel_dp *intel_dp)
+{
+	/*
+	 * FIXME: we need to synchronize the current link parameters with
+	 * hardware readout. Currently fast link training doesn't work on
+	 * boot-up.
+	 */
+	if (intel_dp->link_rate == 0 ||
+	    intel_dp->link_rate > intel_dp->max_link_rate)
+		return false;
+
+	if (intel_dp->lane_count == 0 ||
+	    intel_dp->lane_count > intel_dp_max_lane_count(intel_dp))
+		return false;
+
+	return true;
 }
 
 int intel_dp_get_link_train_fallback_values(struct intel_dp *intel_dp,
 					    int link_rate, uint8_t lane_count)
 {
-	int common_rates[DP_MAX_SUPPORTED_RATES];
-	int link_rate_index;
+	int index;
 
-	link_rate_index = intel_dp_link_rate_index(intel_dp,
-						   common_rates,
-						   link_rate);
-	if (link_rate_index > 0) {
-		intel_dp->max_sink_link_bw = drm_dp_link_rate_to_bw_code(common_rates[link_rate_index - 1]);
-		intel_dp->max_sink_lane_count = lane_count;
+	index = intel_dp_rate_index(intel_dp->common_rates,
+				    intel_dp->num_common_rates,
+				    link_rate);
+	if (index > 0) {
+		intel_dp->max_link_rate = intel_dp->common_rates[index - 1];
+		intel_dp->max_link_lane_count = lane_count;
 	} else if (lane_count > 1) {
-		intel_dp->max_sink_link_bw = intel_dp_max_link_bw(intel_dp);
-		intel_dp->max_sink_lane_count = lane_count >> 1;
+		intel_dp->max_link_rate = intel_dp_max_common_rate(intel_dp);
+		intel_dp->max_link_lane_count = lane_count >> 1;
 	} else {
 		DRM_ERROR("Link Training Unsuccessful\n");
 		return -1;
@@ -754,7 +798,7 @@ static void intel_pps_get_registers(struct drm_i915_private *dev_priv,
 	regs->pp_stat = PP_STATUS(pps_idx);
 	regs->pp_on = PP_ON_DELAYS(pps_idx);
 	regs->pp_off = PP_OFF_DELAYS(pps_idx);
-	if (!IS_GEN9_LP(dev_priv))
+	if (!IS_GEN9_LP(dev_priv) && !HAS_PCH_CNP(dev_priv))
 		regs->pp_div = PP_DIVISOR(pps_idx);
 }
 
@@ -1486,60 +1530,52 @@ static void snprintf_int_array(char *str, size_t len,
 
 static void intel_dp_print_rates(struct intel_dp *intel_dp)
 {
-	const int *source_rates, *sink_rates;
-	int source_len, sink_len, common_len;
-	int common_rates[DP_MAX_SUPPORTED_RATES];
 	char str[128]; /* FIXME: too big for stack? */
 
 	if ((drm_debug & DRM_UT_KMS) == 0)
 		return;
 
-	source_len = intel_dp_source_rates(intel_dp, &source_rates);
-	snprintf_int_array(str, sizeof(str), source_rates, source_len);
+	snprintf_int_array(str, sizeof(str),
+			   intel_dp->source_rates, intel_dp->num_source_rates);
 	DRM_DEBUG_KMS("source rates: %s\n", str);
 
-	sink_len = intel_dp_sink_rates(intel_dp, &sink_rates);
-	snprintf_int_array(str, sizeof(str), sink_rates, sink_len);
+	snprintf_int_array(str, sizeof(str),
+			   intel_dp->sink_rates, intel_dp->num_sink_rates);
 	DRM_DEBUG_KMS("sink rates: %s\n", str);
 
-	common_len = intel_dp_common_rates(intel_dp, common_rates);
-	snprintf_int_array(str, sizeof(str), common_rates, common_len);
+	snprintf_int_array(str, sizeof(str),
+			   intel_dp->common_rates, intel_dp->num_common_rates);
 	DRM_DEBUG_KMS("common rates: %s\n", str);
 }
 
-static int rate_to_index(int find, const int *rates)
-{
-	int i = 0;
-
-	for (i = 0; i < DP_MAX_SUPPORTED_RATES; ++i)
-		if (find == rates[i])
-			break;
-
-	return i;
-}
-
 int
 intel_dp_max_link_rate(struct intel_dp *intel_dp)
 {
-	int rates[DP_MAX_SUPPORTED_RATES] = {};
 	int len;
 
-	len = intel_dp_common_rates(intel_dp, rates);
+	len = intel_dp_common_len_rate_limit(intel_dp, intel_dp->max_link_rate);
 	if (WARN_ON(len <= 0))
 		return 162000;
 
-	return rates[len - 1];
+	return intel_dp->common_rates[len - 1];
 }
 
 int intel_dp_rate_select(struct intel_dp *intel_dp, int rate)
 {
-	return rate_to_index(rate, intel_dp->sink_rates);
+	int i = intel_dp_rate_index(intel_dp->sink_rates,
+				    intel_dp->num_sink_rates, rate);
+
+	if (WARN_ON(i < 0))
+		i = 0;
+
+	return i;
 }
 
 void intel_dp_compute_rate(struct intel_dp *intel_dp, int port_clock,
 			   uint8_t *link_bw, uint8_t *rate_select)
 {
-	if (intel_dp->num_sink_rates) {
+	/* eDP 1.4 rate select method. */
+	if (intel_dp->use_rate_select) {
 		*link_bw = 0;
 		*rate_select =
 			intel_dp_rate_select(intel_dp, port_clock);
@@ -1581,22 +1617,23 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 	enum port port = dp_to_dig_port(intel_dp)->port;
 	struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->base.crtc);
 	struct intel_connector *intel_connector = intel_dp->attached_connector;
+	struct intel_digital_connector_state *intel_conn_state =
+		to_intel_digital_connector_state(conn_state);
 	int lane_count, clock;
 	int min_lane_count = 1;
 	int max_lane_count = intel_dp_max_lane_count(intel_dp);
 	/* Conveniently, the link BW constants become indices with a shift...*/
 	int min_clock = 0;
 	int max_clock;
-	int link_rate_index;
 	int bpp, mode_rate;
 	int link_avail, link_clock;
-	int common_rates[DP_MAX_SUPPORTED_RATES] = {};
 	int common_len;
 	uint8_t link_bw, rate_select;
 	bool reduce_m_n = drm_dp_has_quirk(&intel_dp->desc,
 					   DP_DPCD_QUIRK_LIMITED_M_N);
 
-	common_len = intel_dp_common_rates(intel_dp, common_rates);
+	common_len = intel_dp_common_len_rate_limit(intel_dp,
+						    intel_dp->max_link_rate);
 
 	/* No common link rates between source and sink */
 	WARN_ON(common_len <= 0);
@@ -1607,7 +1644,12 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 		pipe_config->has_pch_encoder = true;
 
 	pipe_config->has_drrs = false;
-	pipe_config->has_audio = intel_dp->has_audio && port != PORT_A;
+	if (port == PORT_A)
+		pipe_config->has_audio = false;
+	else if (intel_conn_state->force_audio == HDMI_AUDIO_AUTO)
+		pipe_config->has_audio = intel_dp->has_audio;
+	else
+		pipe_config->has_audio = intel_conn_state->force_audio == HDMI_AUDIO_ON;
 
 	if (is_edp(intel_dp) && intel_connector->panel.fixed_mode) {
 		intel_fixed_panel_mode(intel_connector->panel.fixed_mode,
@@ -1622,10 +1664,10 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 
 		if (HAS_GMCH_DISPLAY(dev_priv))
 			intel_gmch_panel_fitting(intel_crtc, pipe_config,
-						 intel_connector->panel.fitting_mode);
+						 conn_state->scaling_mode);
 		else
 			intel_pch_panel_fitting(intel_crtc, pipe_config,
-						intel_connector->panel.fitting_mode);
+						conn_state->scaling_mode);
 	}
 
 	if (adjusted_mode->flags & DRM_MODE_FLAG_DBLCLK)
@@ -1633,16 +1675,18 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 
 	/* Use values requested by Compliance Test Request */
 	if (intel_dp->compliance.test_type == DP_TEST_LINK_TRAINING) {
-		link_rate_index = intel_dp_link_rate_index(intel_dp,
-							   common_rates,
-							   intel_dp->compliance.test_link_rate);
-		if (link_rate_index >= 0)
-			min_clock = max_clock = link_rate_index;
+		int index;
+
+		index = intel_dp_rate_index(intel_dp->common_rates,
+					    intel_dp->num_common_rates,
+					    intel_dp->compliance.test_link_rate);
+		if (index >= 0)
+			min_clock = max_clock = index;
 		min_lane_count = max_lane_count = intel_dp->compliance.test_lane_count;
 	}
 	DRM_DEBUG_KMS("DP link computation with max lane count %i "
 		      "max bw %d pixel clock %iKHz\n",
-		      max_lane_count, common_rates[max_clock],
+		      max_lane_count, intel_dp->common_rates[max_clock],
 		      adjusted_mode->crtc_clock);
 
 	/* Walk through all bpp values. Luckily they're all nicely spaced with 2
@@ -1678,7 +1722,7 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 				lane_count <= max_lane_count;
 				lane_count <<= 1) {
 
-				link_clock = common_rates[clock];
+				link_clock = intel_dp->common_rates[clock];
 				link_avail = intel_dp_max_data_rate(link_clock,
 								    lane_count);
 
@@ -1692,7 +1736,7 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 	return false;
 
 found:
-	if (intel_dp->color_range_auto) {
+	if (intel_conn_state->broadcast_rgb == INTEL_BROADCAST_RGB_AUTO) {
 		/*
 		 * See:
 		 * CEA-861-E - 5.1 Default Encoding Parameters
@@ -1704,13 +1748,13 @@ found:
 			HDMI_QUANTIZATION_RANGE_LIMITED;
 	} else {
 		pipe_config->limited_color_range =
-			intel_dp->limited_color_range;
+			intel_conn_state->broadcast_rgb == INTEL_BROADCAST_RGB_LIMITED;
 	}
 
 	pipe_config->lane_count = lane_count;
 
 	pipe_config->pipe_bpp = bpp;
-	pipe_config->port_clock = common_rates[clock];
+	pipe_config->port_clock = intel_dp->common_rates[clock];
 
 	intel_dp_compute_rate(intel_dp, pipe_config->port_clock,
 			      &link_bw, &rate_select);
@@ -2267,14 +2311,17 @@ static void _intel_edp_backlight_on(struct intel_dp *intel_dp)
 }
 
 /* Enable backlight PWM and backlight PP control. */
-void intel_edp_backlight_on(struct intel_dp *intel_dp)
+void intel_edp_backlight_on(const struct intel_crtc_state *crtc_state,
+			    const struct drm_connector_state *conn_state)
 {
+	struct intel_dp *intel_dp = enc_to_intel_dp(conn_state->best_encoder);
+
 	if (!is_edp(intel_dp))
 		return;
 
 	DRM_DEBUG_KMS("\n");
 
-	intel_panel_enable_backlight(intel_dp->attached_connector);
+	intel_panel_enable_backlight(crtc_state, conn_state);
 	_intel_edp_backlight_on(intel_dp);
 }
 
@@ -2306,15 +2353,17 @@ static void _intel_edp_backlight_off(struct intel_dp *intel_dp)
 }
 
 /* Disable backlight PP control and backlight PWM. */
-void intel_edp_backlight_off(struct intel_dp *intel_dp)
+void intel_edp_backlight_off(const struct drm_connector_state *old_conn_state)
 {
+	struct intel_dp *intel_dp = enc_to_intel_dp(old_conn_state->best_encoder);
+
 	if (!is_edp(intel_dp))
 		return;
 
 	DRM_DEBUG_KMS("\n");
 
 	_intel_edp_backlight_off(intel_dp);
-	intel_panel_disable_backlight(intel_dp->attached_connector);
+	intel_panel_disable_backlight(old_conn_state);
 }
 
 /*
@@ -2610,7 +2659,7 @@ static void intel_disable_dp(struct intel_encoder *encoder,
 	/* Make sure the panel is off before trying to change the mode. But also
 	 * ensure that we have vdd while we switch off the panel. */
 	intel_edp_panel_vdd_on(intel_dp);
-	intel_edp_backlight_off(intel_dp);
+	intel_edp_backlight_off(old_conn_state);
 	intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_OFF);
 	intel_edp_panel_off(intel_dp);
 
@@ -2824,10 +2873,8 @@ static void g4x_enable_dp(struct intel_encoder *encoder,
 			  struct intel_crtc_state *pipe_config,
 			  struct drm_connector_state *conn_state)
 {
-	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
-
 	intel_enable_dp(encoder, pipe_config, conn_state);
-	intel_edp_backlight_on(intel_dp);
+	intel_edp_backlight_on(pipe_config, conn_state);
 }
 
 static void vlv_enable_dp(struct intel_encoder *encoder,
@@ -2836,7 +2883,7 @@ static void vlv_enable_dp(struct intel_encoder *encoder,
 {
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 
-	intel_edp_backlight_on(intel_dp);
+	intel_edp_backlight_on(pipe_config, conn_state);
 	intel_psr_enable(intel_dp);
 }
 
@@ -3024,7 +3071,8 @@ static bool intel_dp_get_y_cord_status(struct intel_dp *intel_dp)
 {
 	uint8_t psr_caps = 0;
 
-	drm_dp_dpcd_readb(&intel_dp->aux, DP_PSR_CAPS, &psr_caps);
+	if (drm_dp_dpcd_readb(&intel_dp->aux, DP_PSR_CAPS, &psr_caps) != 1)
+		return false;
 	return psr_caps & DP_PSR2_SU_Y_COORDINATE_REQUIRED;
 }
 
@@ -3032,9 +3080,9 @@ static bool intel_dp_get_colorimetry_status(struct intel_dp *intel_dp)
 {
 	uint8_t dprx = 0;
 
-	drm_dp_dpcd_readb(&intel_dp->aux,
-			DP_DPRX_FEATURE_ENUMERATION_LIST,
-			&dprx);
+	if (drm_dp_dpcd_readb(&intel_dp->aux, DP_DPRX_FEATURE_ENUMERATION_LIST,
+			      &dprx) != 1)
+		return false;
 	return dprx & DP_VSC_SDP_EXT_FOR_COLORIMETRY_SUPPORTED;
 }
 
@@ -3042,7 +3090,9 @@ static bool intel_dp_get_alpm_status(struct intel_dp *intel_dp)
 {
 	uint8_t alpm_caps = 0;
 
-	drm_dp_dpcd_readb(&intel_dp->aux, DP_RECEIVER_ALPM_CAP, &alpm_caps);
+	if (drm_dp_dpcd_readb(&intel_dp->aux, DP_RECEIVER_ALPM_CAP,
+			      &alpm_caps) != 1)
+		return false;
 	return alpm_caps & DP_ALPM_CAP;
 }
 
@@ -3616,9 +3666,10 @@ intel_edp_init_dpcd(struct intel_dp *intel_dp)
 		uint8_t frame_sync_cap;
 
 		dev_priv->psr.sink_support = true;
-		drm_dp_dpcd_read(&intel_dp->aux,
-				 DP_SINK_DEVICE_AUX_FRAME_SYNC_CAP,
-				 &frame_sync_cap, 1);
+		if (drm_dp_dpcd_readb(&intel_dp->aux,
+				      DP_SINK_DEVICE_AUX_FRAME_SYNC_CAP,
+				      &frame_sync_cap) != 1)
+			frame_sync_cap = 0;
 		dev_priv->psr.aux_frame_sync = frame_sync_cap ? true : false;
 		/* PSR2 needs frame sync as well */
 		dev_priv->psr.psr2_support = dev_priv->psr.aux_frame_sync;
@@ -3669,6 +3720,13 @@ intel_edp_init_dpcd(struct intel_dp *intel_dp)
 		intel_dp->num_sink_rates = i;
 	}
 
+	if (intel_dp->num_sink_rates)
+		intel_dp->use_rate_select = true;
+	else
+		intel_dp_set_sink_rates(intel_dp);
+
+	intel_dp_set_common_rates(intel_dp);
+
 	return true;
 }
 
@@ -3676,11 +3734,18 @@ intel_edp_init_dpcd(struct intel_dp *intel_dp)
 static bool
 intel_dp_get_dpcd(struct intel_dp *intel_dp)
 {
+	u8 sink_count;
+
 	if (!intel_dp_read_dpcd(intel_dp))
 		return false;
 
-	if (drm_dp_dpcd_read(&intel_dp->aux, DP_SINK_COUNT,
-			     &intel_dp->sink_count, 1) < 0)
+	/* Don't clobber cached eDP rates. */
+	if (!is_edp(intel_dp)) {
+		intel_dp_set_sink_rates(intel_dp);
+		intel_dp_set_common_rates(intel_dp);
+	}
+
+	if (drm_dp_dpcd_readb(&intel_dp->aux, DP_SINK_COUNT, &sink_count) <= 0)
 		return false;
 
 	/*
@@ -3688,7 +3753,7 @@ intel_dp_get_dpcd(struct intel_dp *intel_dp)
 	 * a member variable in intel_dp will track any changes
 	 * between short pulse interrupts.
 	 */
-	intel_dp->sink_count = DP_GET_SINK_COUNT(intel_dp->sink_count);
+	intel_dp->sink_count = DP_GET_SINK_COUNT(sink_count);
 
 	/*
 	 * SINK_COUNT == 0 and DOWNSTREAM_PORT_PRESENT == 1 implies that
@@ -3717,7 +3782,7 @@ intel_dp_get_dpcd(struct intel_dp *intel_dp)
 static bool
 intel_dp_can_mst(struct intel_dp *intel_dp)
 {
-	u8 buf[1];
+	u8 mstm_cap;
 
 	if (!i915.enable_dp_mst)
 		return false;
@@ -3728,10 +3793,10 @@ intel_dp_can_mst(struct intel_dp *intel_dp)
 	if (intel_dp->dpcd[DP_DPCD_REV] < 0x12)
 		return false;
 
-	if (drm_dp_dpcd_read(&intel_dp->aux, DP_MSTM_CAP, buf, 1) != 1)
+	if (drm_dp_dpcd_readb(&intel_dp->aux, DP_MSTM_CAP, &mstm_cap) != 1)
 		return false;
 
-	return buf[0] & DP_MST_CAP;
+	return mstm_cap & DP_MST_CAP;
 }
 
 static void
@@ -3877,9 +3942,8 @@ stop:
 static bool
 intel_dp_get_sink_irq(struct intel_dp *intel_dp, u8 *sink_irq_vector)
 {
-	return drm_dp_dpcd_read(&intel_dp->aux,
-				       DP_DEVICE_SERVICE_IRQ_VECTOR,
-				       sink_irq_vector, 1) == 1;
+	return drm_dp_dpcd_readb(&intel_dp->aux, DP_DEVICE_SERVICE_IRQ_VECTOR,
+				 sink_irq_vector) == 1;
 }
 
 static bool
@@ -3900,7 +3964,6 @@ static uint8_t intel_dp_autotest_link_training(struct intel_dp *intel_dp)
 {
 	int status = 0;
 	int min_lane_count = 1;
-	int common_rates[DP_MAX_SUPPORTED_RATES] = {};
 	int link_rate_index, test_link_rate;
 	uint8_t test_lane_count, test_link_bw;
 	/* (DP CTS 1.2)
@@ -3917,7 +3980,7 @@ static uint8_t intel_dp_autotest_link_training(struct intel_dp *intel_dp)
 	test_lane_count &= DP_MAX_LANE_COUNT_MASK;
 	/* Validate the requested lane count */
 	if (test_lane_count < min_lane_count ||
-	    test_lane_count > intel_dp->max_sink_lane_count)
+	    test_lane_count > intel_dp->max_link_lane_count)
 		return DP_TEST_NAK;
 
 	status = drm_dp_dpcd_readb(&intel_dp->aux, DP_TEST_LINK_RATE,
@@ -3928,9 +3991,9 @@ static uint8_t intel_dp_autotest_link_training(struct intel_dp *intel_dp)
 	}
 	/* Validate the requested link rate */
 	test_link_rate = drm_dp_bw_code_to_link_rate(test_link_bw);
-	link_rate_index = intel_dp_link_rate_index(intel_dp,
-						   common_rates,
-						   test_link_rate);
+	link_rate_index = intel_dp_rate_index(intel_dp->common_rates,
+					      intel_dp->num_common_rates,
+					      test_link_rate);
 	if (link_rate_index < 0)
 		return DP_TEST_NAK;
 
@@ -3943,13 +4006,13 @@ static uint8_t intel_dp_autotest_link_training(struct intel_dp *intel_dp)
 static uint8_t intel_dp_autotest_video_pattern(struct intel_dp *intel_dp)
 {
 	uint8_t test_pattern;
-	uint16_t test_misc;
+	uint8_t test_misc;
 	__be16 h_width, v_height;
 	int status = 0;
 
 	/* Read the TEST_PATTERN (DP CTS 3.1.5) */
-	status = drm_dp_dpcd_read(&intel_dp->aux, DP_TEST_PATTERN,
-				  &test_pattern, 1);
+	status = drm_dp_dpcd_readb(&intel_dp->aux, DP_TEST_PATTERN,
+				   &test_pattern);
 	if (status <= 0) {
 		DRM_DEBUG_KMS("Test pattern read failed\n");
 		return DP_TEST_NAK;
@@ -3971,8 +4034,8 @@ static uint8_t intel_dp_autotest_video_pattern(struct intel_dp *intel_dp)
 		return DP_TEST_NAK;
 	}
 
-	status = drm_dp_dpcd_read(&intel_dp->aux, DP_TEST_MISC0,
-				  &test_misc, 1);
+	status = drm_dp_dpcd_readb(&intel_dp->aux, DP_TEST_MISC0,
+				   &test_misc);
 	if (status <= 0) {
 		DRM_DEBUG_KMS("TEST MISC read failed\n");
 		return DP_TEST_NAK;
@@ -4031,10 +4094,8 @@ static uint8_t intel_dp_autotest_edid(struct intel_dp *intel_dp)
 		 */
 		block += intel_connector->detect_edid->extensions;
 
-		if (!drm_dp_dpcd_write(&intel_dp->aux,
-					DP_TEST_EDID_CHECKSUM,
-					&block->checksum,
-					1))
+		if (drm_dp_dpcd_writeb(&intel_dp->aux, DP_TEST_EDID_CHECKSUM,
+				       block->checksum) <= 0)
 			DRM_DEBUG_KMS("Failed to write EDID checksum\n");
 
 		test_result = DP_TEST_ACK | DP_TEST_EDID_CHECKSUM_WRITE;
@@ -4198,9 +4259,11 @@ intel_dp_check_link_status(struct intel_dp *intel_dp)
 	if (!to_intel_crtc(intel_encoder->base.crtc)->active)
 		return;
 
-	/* FIXME: we need to synchronize this sort of stuff with hardware
-	 * readout. Currently fast link training doesn't work on boot-up. */
-	if (!intel_dp->lane_count)
+	/*
+	 * Validate the cached values of intel_dp->link_rate and
+	 * intel_dp->lane_count before attempting to retrain.
+	 */
+	if (!intel_dp_link_params_valid(intel_dp))
 		return;
 
 	/* Retrain if Channel EQ or CR not ok */
@@ -4523,10 +4586,7 @@ intel_dp_set_edid(struct intel_dp *intel_dp)
 	edid = intel_dp_get_edid(intel_dp);
 	intel_connector->detect_edid = edid;
 
-	if (intel_dp->force_audio != HDMI_AUDIO_AUTO)
-		intel_dp->has_audio = intel_dp->force_audio == HDMI_AUDIO_ON;
-	else
-		intel_dp->has_audio = drm_detect_monitor_audio(edid);
+	intel_dp->has_audio = drm_detect_monitor_audio(edid);
 }
 
 static void
@@ -4587,11 +4647,11 @@ intel_dp_long_pulse(struct intel_connector *intel_connector)
 		      yesno(drm_dp_tps3_supported(intel_dp->dpcd)));
 
 	if (intel_dp->reset_link_params) {
-		/* Set the max lane count for sink */
-		intel_dp->max_sink_lane_count = drm_dp_max_lane_count(intel_dp->dpcd);
+		/* Initial max link lane count */
+		intel_dp->max_link_lane_count = intel_dp_max_common_lane_count(intel_dp);
 
-		/* Set the max link BW for sink */
-		intel_dp->max_sink_link_bw = intel_dp_max_link_bw(intel_dp);
+		/* Initial max link rate */
+		intel_dp->max_link_rate = intel_dp_max_common_rate(intel_dp);
 
 		intel_dp->reset_link_params = false;
 	}
@@ -4735,112 +4795,6 @@ static int intel_dp_get_modes(struct drm_connector *connector)
 	return 0;
 }
 
-static bool
-intel_dp_detect_audio(struct drm_connector *connector)
-{
-	bool has_audio = false;
-	struct edid *edid;
-
-	edid = to_intel_connector(connector)->detect_edid;
-	if (edid)
-		has_audio = drm_detect_monitor_audio(edid);
-
-	return has_audio;
-}
-
-static int
-intel_dp_set_property(struct drm_connector *connector,
-		      struct drm_property *property,
-		      uint64_t val)
-{
-	struct drm_i915_private *dev_priv = to_i915(connector->dev);
-	struct intel_connector *intel_connector = to_intel_connector(connector);
-	struct intel_encoder *intel_encoder = intel_attached_encoder(connector);
-	struct intel_dp *intel_dp = enc_to_intel_dp(&intel_encoder->base);
-	int ret;
-
-	ret = drm_object_property_set_value(&connector->base, property, val);
-	if (ret)
-		return ret;
-
-	if (property == dev_priv->force_audio_property) {
-		int i = val;
-		bool has_audio;
-
-		if (i == intel_dp->force_audio)
-			return 0;
-
-		intel_dp->force_audio = i;
-
-		if (i == HDMI_AUDIO_AUTO)
-			has_audio = intel_dp_detect_audio(connector);
-		else
-			has_audio = (i == HDMI_AUDIO_ON);
-
-		if (has_audio == intel_dp->has_audio)
-			return 0;
-
-		intel_dp->has_audio = has_audio;
-		goto done;
-	}
-
-	if (property == dev_priv->broadcast_rgb_property) {
-		bool old_auto = intel_dp->color_range_auto;
-		bool old_range = intel_dp->limited_color_range;
-
-		switch (val) {
-		case INTEL_BROADCAST_RGB_AUTO:
-			intel_dp->color_range_auto = true;
-			break;
-		case INTEL_BROADCAST_RGB_FULL:
-			intel_dp->color_range_auto = false;
-			intel_dp->limited_color_range = false;
-			break;
-		case INTEL_BROADCAST_RGB_LIMITED:
-			intel_dp->color_range_auto = false;
-			intel_dp->limited_color_range = true;
-			break;
-		default:
-			return -EINVAL;
-		}
-
-		if (old_auto == intel_dp->color_range_auto &&
-		    old_range == intel_dp->limited_color_range)
-			return 0;
-
-		goto done;
-	}
-
-	if (is_edp(intel_dp) &&
-	    property == connector->dev->mode_config.scaling_mode_property) {
-		if (val == DRM_MODE_SCALE_NONE) {
-			DRM_DEBUG_KMS("no scaling not supported\n");
-			return -EINVAL;
-		}
-		if (HAS_GMCH_DISPLAY(dev_priv) &&
-		    val == DRM_MODE_SCALE_CENTER) {
-			DRM_DEBUG_KMS("centering not supported\n");
-			return -EINVAL;
-		}
-
-		if (intel_connector->panel.fitting_mode == val) {
-			/* the eDP scaling property is not changed */
-			return 0;
-		}
-		intel_connector->panel.fitting_mode = val;
-
-		goto done;
-	}
-
-	return -EINVAL;
-
-done:
-	if (intel_encoder->base.crtc)
-		intel_crtc_restore_mode(intel_encoder->base.crtc);
-
-	return 0;
-}
-
 static int
 intel_dp_connector_register(struct drm_connector *connector)
 {
@@ -4999,19 +4953,21 @@ static const struct drm_connector_funcs intel_dp_connector_funcs = {
 	.dpms = drm_atomic_helper_connector_dpms,
 	.force = intel_dp_force,
 	.fill_modes = drm_helper_probe_single_connector_modes,
-	.set_property = intel_dp_set_property,
-	.atomic_get_property = intel_connector_atomic_get_property,
+	.set_property = drm_atomic_helper_connector_set_property,
+	.atomic_get_property = intel_digital_connector_atomic_get_property,
+	.atomic_set_property = intel_digital_connector_atomic_set_property,
 	.late_register = intel_dp_connector_register,
 	.early_unregister = intel_dp_connector_unregister,
 	.destroy = intel_dp_connector_destroy,
 	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
-	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+	.atomic_duplicate_state = intel_digital_connector_duplicate_state,
 };
 
 static const struct drm_connector_helper_funcs intel_dp_connector_helper_funcs = {
 	.detect_ctx = intel_dp_detect,
 	.get_modes = intel_dp_get_modes,
 	.mode_valid = intel_dp_mode_valid,
+	.atomic_check = intel_digital_connector_atomic_check,
 };
 
 static const struct drm_encoder_funcs intel_dp_enc_funcs = {
@@ -5102,22 +5058,25 @@ bool intel_dp_is_edp(struct drm_i915_private *dev_priv, enum port port)
 	return intel_bios_is_port_edp(dev_priv, port);
 }
 
-void
+static void
 intel_dp_add_properties(struct intel_dp *intel_dp, struct drm_connector *connector)
 {
-	struct intel_connector *intel_connector = to_intel_connector(connector);
+	struct drm_i915_private *dev_priv = to_i915(connector->dev);
 
 	intel_attach_force_audio_property(connector);
 	intel_attach_broadcast_rgb_property(connector);
-	intel_dp->color_range_auto = true;
 
 	if (is_edp(intel_dp)) {
-		drm_mode_create_scaling_mode_property(connector->dev);
-		drm_object_attach_property(
-			&connector->base,
-			connector->dev->mode_config.scaling_mode_property,
-			DRM_MODE_SCALE_ASPECT);
-		intel_connector->panel.fitting_mode = DRM_MODE_SCALE_ASPECT;
+		u32 allowed_scalers;
+
+		allowed_scalers = BIT(DRM_MODE_SCALE_ASPECT) | BIT(DRM_MODE_SCALE_FULLSCREEN);
+		if (!HAS_GMCH_DISPLAY(dev_priv))
+			allowed_scalers |= BIT(DRM_MODE_SCALE_CENTER);
+
+		drm_connector_attach_scaling_mode_property(connector, allowed_scalers);
+
+		connector->state->scaling_mode = DRM_MODE_SCALE_ASPECT;
+
 	}
 }
 
@@ -5143,7 +5102,7 @@ intel_pps_readout_hw_state(struct drm_i915_private *dev_priv,
 
 	pp_on = I915_READ(regs.pp_on);
 	pp_off = I915_READ(regs.pp_off);
-	if (!IS_GEN9_LP(dev_priv)) {
+	if (!IS_GEN9_LP(dev_priv) && !HAS_PCH_CNP(dev_priv)) {
 		I915_WRITE(regs.pp_ctrl, pp_ctl);
 		pp_div = I915_READ(regs.pp_div);
 	}
@@ -5161,7 +5120,7 @@ intel_pps_readout_hw_state(struct drm_i915_private *dev_priv,
 	seq->t10 = (pp_off & PANEL_POWER_DOWN_DELAY_MASK) >>
 		   PANEL_POWER_DOWN_DELAY_SHIFT;
 
-	if (IS_GEN9_LP(dev_priv)) {
+	if (IS_GEN9_LP(dev_priv) || HAS_PCH_CNP(dev_priv)) {
 		u16 tmp = (pp_ctl & BXT_POWER_CYCLE_DELAY_MASK) >>
 			BXT_POWER_CYCLE_DELAY_SHIFT;
 		if (tmp > 0)
@@ -5318,7 +5277,7 @@ intel_dp_init_panel_power_sequencer_registers(struct drm_device *dev,
 		 (seq->t10 << PANEL_POWER_DOWN_DELAY_SHIFT);
 	/* Compute the divisor for the pp clock, simply match the Bspec
 	 * formula. */
-	if (IS_GEN9_LP(dev_priv)) {
+	if (IS_GEN9_LP(dev_priv) || HAS_PCH_CNP(dev_priv)) {
 		pp_div = I915_READ(regs.pp_ctrl);
 		pp_div &= ~BXT_POWER_CYCLE_DELAY_MASK;
 		pp_div |= (DIV_ROUND_UP((seq->t11_t12 + 1), 1000)
@@ -5344,7 +5303,7 @@ intel_dp_init_panel_power_sequencer_registers(struct drm_device *dev,
 
 	I915_WRITE(regs.pp_on, pp_on);
 	I915_WRITE(regs.pp_off, pp_off);
-	if (IS_GEN9_LP(dev_priv))
+	if (IS_GEN9_LP(dev_priv) || HAS_PCH_CNP(dev_priv))
 		I915_WRITE(regs.pp_ctrl, pp_div);
 	else
 		I915_WRITE(regs.pp_div, pp_div);
@@ -5352,7 +5311,7 @@ intel_dp_init_panel_power_sequencer_registers(struct drm_device *dev,
 	DRM_DEBUG_KMS("panel power sequencer register settings: PP_ON %#x, PP_OFF %#x, PP_DIV %#x\n",
 		      I915_READ(regs.pp_on),
 		      I915_READ(regs.pp_off),
-		      IS_GEN9_LP(dev_priv) ?
+		      (IS_GEN9_LP(dev_priv) || HAS_PCH_CNP(dev_priv)) ?
 		      (I915_READ(regs.pp_ctrl) & BXT_POWER_CYCLE_DELAY_MASK) :
 		      I915_READ(regs.pp_div));
 }
@@ -5907,6 +5866,29 @@ intel_dp_init_connector_port_info(struct intel_digital_port *intel_dig_port)
 	}
 }
 
+static void intel_dp_modeset_retry_work_fn(struct work_struct *work)
+{
+	struct intel_connector *intel_connector;
+	struct drm_connector *connector;
+
+	intel_connector = container_of(work, typeof(*intel_connector),
+				       modeset_retry_work);
+	connector = &intel_connector->base;
+	DRM_DEBUG_KMS("[CONNECTOR:%d:%s]\n", connector->base.id,
+		      connector->name);
+
+	/* Grab the locks before changing connector property*/
+	mutex_lock(&connector->dev->mode_config.mutex);
+	/* Set connector link status to BAD and send a Uevent to notify
+	 * userspace to do a modeset.
+	 */
+	drm_mode_connector_set_link_status_property(connector,
+						    DRM_MODE_LINK_STATUS_BAD);
+	mutex_unlock(&connector->dev->mode_config.mutex);
+	/* Send Hotplug uevent so userspace can reprobe */
+	drm_kms_helper_hotplug_event(connector->dev);
+}
+
 bool
 intel_dp_init_connector(struct intel_digital_port *intel_dig_port,
 			struct intel_connector *intel_connector)
@@ -5919,11 +5901,17 @@ intel_dp_init_connector(struct intel_digital_port *intel_dig_port,
 	enum port port = intel_dig_port->port;
 	int type;
 
+	/* Initialize the work for modeset in case of link train failure */
+	INIT_WORK(&intel_connector->modeset_retry_work,
+		  intel_dp_modeset_retry_work_fn);
+
 	if (WARN(intel_dig_port->max_lanes < 1,
 		 "Not enough lanes (%d) for DP on port %c\n",
 		 intel_dig_port->max_lanes, port_name(port)))
 		return false;
 
+	intel_dp_set_source_rates(intel_dp);
+
 	intel_dp->reset_link_params = true;
 	intel_dp->pps_pipe = INVALID_PIPE;
 	intel_dp->active_pipe = INVALID_PIPE;
diff --git a/drivers/gpu/drm/i915/intel_dp_aux_backlight.c b/drivers/gpu/drm/i915/intel_dp_aux_backlight.c
index 6532e226db29..6cc62980d0da 100644
--- a/drivers/gpu/drm/i915/intel_dp_aux_backlight.c
+++ b/drivers/gpu/drm/i915/intel_dp_aux_backlight.c
@@ -28,6 +28,10 @@ static void set_aux_backlight_enable(struct intel_dp *intel_dp, bool enable)
 {
 	uint8_t reg_val = 0;
 
+	/* Early return when display use other mechanism to enable backlight. */
+	if (!(intel_dp->edp_dpcd[1] & DP_EDP_BACKLIGHT_AUX_ENABLE_CAP))
+		return;
+
 	if (drm_dp_dpcd_readb(&intel_dp->aux, DP_EDP_DISPLAY_CONTROL_REGISTER,
 			      &reg_val) < 0) {
 		DRM_DEBUG_KMS("Failed to read DPCD register 0x%x\n",
@@ -74,8 +78,13 @@ static uint32_t intel_dp_aux_get_backlight(struct intel_connector *connector)
  * 8-bit or 16 bit value (MSB and LSB)
  */
 static void
-intel_dp_aux_set_backlight(struct intel_connector *connector, u32 level)
+intel_dp_aux_set_backlight(const struct drm_connector_state *conn_state, u32 level)
 {
+	/*
+	 * conn_state->best_encoder is likely NULL when called from
+	 * intel_dp_aux_setup_backlight()
+	 */
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct intel_dp *intel_dp = enc_to_intel_dp(&connector->encoder->base);
 	uint8_t vals[2] = { 0x0 };
 
@@ -93,24 +102,52 @@ intel_dp_aux_set_backlight(struct intel_connector *connector, u32 level)
 	}
 }
 
-static void intel_dp_aux_enable_backlight(struct intel_connector *connector)
+static void intel_dp_aux_enable_backlight(const struct intel_crtc_state *crtc_state,
+					  const struct drm_connector_state *conn_state)
 {
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
+	/*
+	 * conn_state->best_encoder (and crtc_state) are NULL when called from
+	 * intel_dp_aux_setup_backlight()
+	 */
 	struct intel_dp *intel_dp = enc_to_intel_dp(&connector->encoder->base);
 	uint8_t dpcd_buf = 0;
+	uint8_t edp_backlight_mode = 0;
 
-	set_aux_backlight_enable(intel_dp, true);
+	if (drm_dp_dpcd_readb(&intel_dp->aux,
+			DP_EDP_BACKLIGHT_MODE_SET_REGISTER, &dpcd_buf) != 1) {
+		DRM_DEBUG_KMS("Failed to read DPCD register 0x%x\n",
+			      DP_EDP_BACKLIGHT_MODE_SET_REGISTER);
+		return;
+	}
 
-	if ((drm_dp_dpcd_readb(&intel_dp->aux,
-			       DP_EDP_BACKLIGHT_MODE_SET_REGISTER, &dpcd_buf) == 1) &&
-	    ((dpcd_buf & DP_EDP_BACKLIGHT_CONTROL_MODE_MASK) ==
-	     DP_EDP_BACKLIGHT_CONTROL_MODE_PRESET))
-		drm_dp_dpcd_writeb(&intel_dp->aux, DP_EDP_BACKLIGHT_MODE_SET_REGISTER,
-				   (dpcd_buf | DP_EDP_BACKLIGHT_CONTROL_MODE_DPCD));
+	edp_backlight_mode = dpcd_buf & DP_EDP_BACKLIGHT_CONTROL_MODE_MASK;
+
+	switch (edp_backlight_mode) {
+	case DP_EDP_BACKLIGHT_CONTROL_MODE_PWM:
+	case DP_EDP_BACKLIGHT_CONTROL_MODE_PRESET:
+	case DP_EDP_BACKLIGHT_CONTROL_MODE_PRODUCT:
+		dpcd_buf &= ~DP_EDP_BACKLIGHT_CONTROL_MODE_MASK;
+		dpcd_buf |= DP_EDP_BACKLIGHT_CONTROL_MODE_DPCD;
+		if (drm_dp_dpcd_writeb(&intel_dp->aux,
+			DP_EDP_BACKLIGHT_MODE_SET_REGISTER, dpcd_buf) < 0) {
+			DRM_DEBUG_KMS("Failed to write aux backlight mode\n");
+		}
+		break;
+
+	/* Do nothing when it is already DPCD mode */
+	case DP_EDP_BACKLIGHT_CONTROL_MODE_DPCD:
+	default:
+		break;
+	}
+
+	set_aux_backlight_enable(intel_dp, true);
+	intel_dp_aux_set_backlight(conn_state, connector->panel.backlight.level);
 }
 
-static void intel_dp_aux_disable_backlight(struct intel_connector *connector)
+static void intel_dp_aux_disable_backlight(const struct drm_connector_state *old_conn_state)
 {
-	set_aux_backlight_enable(enc_to_intel_dp(&connector->encoder->base), false);
+	set_aux_backlight_enable(enc_to_intel_dp(old_conn_state->best_encoder), false);
 }
 
 static int intel_dp_aux_setup_backlight(struct intel_connector *connector,
@@ -119,7 +156,7 @@ static int intel_dp_aux_setup_backlight(struct intel_connector *connector,
 	struct intel_dp *intel_dp = enc_to_intel_dp(&connector->encoder->base);
 	struct intel_panel *panel = &connector->panel;
 
-	intel_dp_aux_enable_backlight(connector);
+	intel_dp_aux_enable_backlight(NULL, connector->base.state);
 
 	if (intel_dp->edp_dpcd[2] & DP_EDP_BACKLIGHT_BRIGHTNESS_BYTE_COUNT)
 		panel->backlight.max = 0xFFFF;
@@ -139,13 +176,12 @@ intel_dp_aux_display_control_capable(struct intel_connector *connector)
 {
 	struct intel_dp *intel_dp = enc_to_intel_dp(&connector->encoder->base);
 
-	/* Check the  eDP Display control capabilities registers to determine if
+	/* Check the eDP Display control capabilities registers to determine if
 	 * the panel can support backlight control over the aux channel
 	 */
 	if (intel_dp->edp_dpcd[1] & DP_EDP_TCON_BACKLIGHT_ADJUSTMENT_CAP &&
-	    (intel_dp->edp_dpcd[1] & DP_EDP_BACKLIGHT_AUX_ENABLE_CAP) &&
-	    !((intel_dp->edp_dpcd[1] & DP_EDP_BACKLIGHT_PIN_ENABLE_CAP) ||
-	      (intel_dp->edp_dpcd[2] & DP_EDP_BACKLIGHT_BRIGHTNESS_PWM_PIN_CAP))) {
+	    (intel_dp->edp_dpcd[2] & DP_EDP_BACKLIGHT_BRIGHTNESS_AUX_SET_CAP) &&
+	    !(intel_dp->edp_dpcd[2] & DP_EDP_BACKLIGHT_BRIGHTNESS_PWM_PIN_CAP)) {
 		DRM_DEBUG_KMS("AUX Backlight Control Supported!\n");
 		return true;
 	}
diff --git a/drivers/gpu/drm/i915/intel_dp_link_training.c b/drivers/gpu/drm/i915/intel_dp_link_training.c
index 0048b520baf7..b79c1c0e404c 100644
--- a/drivers/gpu/drm/i915/intel_dp_link_training.c
+++ b/drivers/gpu/drm/i915/intel_dp_link_training.c
@@ -146,7 +146,8 @@ intel_dp_link_training_clock_recovery(struct intel_dp *intel_dp)
 		link_config[1] |= DP_LANE_COUNT_ENHANCED_FRAME_EN;
 	drm_dp_dpcd_write(&intel_dp->aux, DP_LINK_BW_SET, link_config, 2);
 
-	if (intel_dp->num_sink_rates)
+	/* eDP 1.4 rate select method. */
+	if (!link_bw)
 		drm_dp_dpcd_write(&intel_dp->aux, DP_LINK_RATE_SET,
 				  &rate_select, 1);
 
@@ -313,6 +314,24 @@ void intel_dp_stop_link_train(struct intel_dp *intel_dp)
 void
 intel_dp_start_link_train(struct intel_dp *intel_dp)
 {
-	intel_dp_link_training_clock_recovery(intel_dp);
-	intel_dp_link_training_channel_equalization(intel_dp);
+	struct intel_connector *intel_connector = intel_dp->attached_connector;
+
+	if (!intel_dp_link_training_clock_recovery(intel_dp))
+		goto failure_handling;
+	if (!intel_dp_link_training_channel_equalization(intel_dp))
+		goto failure_handling;
+
+	DRM_DEBUG_KMS("Link Training Passed at Link Rate = %d, Lane count = %d",
+		      intel_dp->link_rate, intel_dp->lane_count);
+	return;
+
+ failure_handling:
+	DRM_DEBUG_KMS("Link Training failed at link rate = %d, lane count = %d",
+		      intel_dp->link_rate, intel_dp->lane_count);
+	if (!intel_dp_get_link_train_fallback_values(intel_dp,
+						     intel_dp->link_rate,
+						     intel_dp->lane_count))
+		/* Schedule a Hotplug Uevent to userspace to start modeset */
+		schedule_work(&intel_connector->modeset_retry_work);
+	return;
 }
diff --git a/drivers/gpu/drm/i915/intel_dp_mst.c b/drivers/gpu/drm/i915/intel_dp_mst.c
index 989e25577ac0..2cf046beae0f 100644
--- a/drivers/gpu/drm/i915/intel_dp_mst.c
+++ b/drivers/gpu/drm/i915/intel_dp_mst.c
@@ -39,7 +39,7 @@ static bool intel_dp_mst_compute_config(struct intel_encoder *encoder,
 	struct intel_dp *intel_dp = &intel_dig_port->dp;
 	struct intel_connector *connector =
 		to_intel_connector(conn_state->connector);
-	struct drm_atomic_state *state;
+	struct drm_atomic_state *state = pipe_config->base.state;
 	int bpp;
 	int lane_count, slots;
 	const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
@@ -58,21 +58,26 @@ static bool intel_dp_mst_compute_config(struct intel_encoder *encoder,
 	 * for MST we always configure max link bw - the spec doesn't
 	 * seem to suggest we should do otherwise.
 	 */
-	lane_count = drm_dp_max_lane_count(intel_dp->dpcd);
+	lane_count = intel_dp_max_lane_count(intel_dp);
 
 	pipe_config->lane_count = lane_count;
 
 	pipe_config->pipe_bpp = bpp;
-	pipe_config->port_clock = intel_dp_max_link_rate(intel_dp);
 
-	state = pipe_config->base.state;
+	pipe_config->port_clock = intel_dp_max_link_rate(intel_dp);
 
 	if (drm_dp_mst_port_has_audio(&intel_dp->mst_mgr, connector->port))
 		pipe_config->has_audio = true;
-	mst_pbn = drm_dp_calc_pbn_mode(adjusted_mode->crtc_clock, bpp);
 
+	mst_pbn = drm_dp_calc_pbn_mode(adjusted_mode->crtc_clock, bpp);
 	pipe_config->pbn = mst_pbn;
-	slots = drm_dp_find_vcpi_slots(&intel_dp->mst_mgr, mst_pbn);
+
+	slots = drm_dp_atomic_find_vcpi_slots(state, &intel_dp->mst_mgr,
+					      connector->port, mst_pbn);
+	if (slots < 0) {
+		DRM_DEBUG_KMS("failed finding vcpi slots:%d\n", slots);
+		return false;
+	}
 
 	intel_link_compute_m_n(bpp, lane_count,
 			       adjusted_mode->crtc_clock,
@@ -83,7 +88,38 @@ static bool intel_dp_mst_compute_config(struct intel_encoder *encoder,
 	pipe_config->dp_m_n.tu = slots;
 
 	return true;
+}
+
+static int intel_dp_mst_atomic_check(struct drm_connector *connector,
+		struct drm_connector_state *new_conn_state)
+{
+	struct drm_atomic_state *state = new_conn_state->state;
+	struct drm_connector_state *old_conn_state;
+	struct drm_crtc *old_crtc;
+	struct drm_crtc_state *crtc_state;
+	int slots, ret = 0;
+
+	old_conn_state = drm_atomic_get_old_connector_state(state, connector);
+	old_crtc = old_conn_state->crtc;
+	if (!old_crtc)
+		return ret;
+
+	crtc_state = drm_atomic_get_new_crtc_state(state, old_crtc);
+	slots = to_intel_crtc_state(crtc_state)->dp_m_n.tu;
+	if (drm_atomic_crtc_needs_modeset(crtc_state) && slots > 0) {
+		struct drm_dp_mst_topology_mgr *mgr;
+		struct drm_encoder *old_encoder;
+
+		old_encoder = old_conn_state->best_encoder;
+		mgr = &enc_to_mst(old_encoder)->primary->dp.mst_mgr;
 
+		ret = drm_dp_atomic_release_vcpi_slots(state, mgr, slots);
+		if (ret)
+			DRM_DEBUG_KMS("failed releasing %d vcpi slots:%d\n", slots, ret);
+		else
+			to_intel_crtc_state(crtc_state)->dp_m_n.tu = 0;
+	}
+	return ret;
 }
 
 static void intel_mst_disable_dp(struct intel_encoder *encoder,
@@ -297,14 +333,6 @@ intel_dp_mst_detect(struct drm_connector *connector, bool force)
 	return drm_dp_mst_detect_port(connector, &intel_dp->mst_mgr, intel_connector->port);
 }
 
-static int
-intel_dp_mst_set_property(struct drm_connector *connector,
-			  struct drm_property *property,
-			  uint64_t val)
-{
-	return 0;
-}
-
 static void
 intel_dp_mst_connector_destroy(struct drm_connector *connector)
 {
@@ -321,8 +349,7 @@ static const struct drm_connector_funcs intel_dp_mst_connector_funcs = {
 	.dpms = drm_atomic_helper_connector_dpms,
 	.detect = intel_dp_mst_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
-	.set_property = intel_dp_mst_set_property,
-	.atomic_get_property = intel_connector_atomic_get_property,
+	.set_property = drm_atomic_helper_connector_set_property,
 	.late_register = intel_connector_register,
 	.early_unregister = intel_connector_unregister,
 	.destroy = intel_dp_mst_connector_destroy,
@@ -346,7 +373,7 @@ intel_dp_mst_mode_valid(struct drm_connector *connector,
 	int max_rate, mode_rate, max_lanes, max_link_clock;
 
 	max_link_clock = intel_dp_max_link_rate(intel_dp);
-	max_lanes = drm_dp_max_lane_count(intel_dp->dpcd);
+	max_lanes = intel_dp_max_lane_count(intel_dp);
 
 	max_rate = intel_dp_max_data_rate(max_link_clock, max_lanes);
 	mode_rate = intel_dp_link_required(mode->clock, bpp);
@@ -390,6 +417,7 @@ static const struct drm_connector_helper_funcs intel_dp_mst_connector_helper_fun
 	.mode_valid = intel_dp_mst_mode_valid,
 	.atomic_best_encoder = intel_mst_atomic_best_encoder,
 	.best_encoder = intel_mst_best_encoder,
+	.atomic_check = intel_dp_mst_atomic_check,
 };
 
 static void intel_dp_mst_encoder_destroy(struct drm_encoder *encoder)
@@ -462,7 +490,6 @@ static struct drm_connector *intel_dp_add_mst_connector(struct drm_dp_mst_topolo
 		drm_mode_connector_attach_encoder(&intel_connector->base,
 						  &intel_dp->mst_encoders[i]->base.base);
 	}
-	intel_dp_add_properties(intel_dp, connector);
 
 	drm_object_attach_property(&connector->base, dev->mode_config.path_property, 0);
 	drm_object_attach_property(&connector->base, dev->mode_config.tile_property, 0);
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index f630c7af5020..18528a477cf6 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -88,7 +88,6 @@
 	int cpu, ret, timeout = (US) * 1000; \
 	u64 base; \
 	_WAIT_FOR_ATOMIC_CHECK(ATOMIC); \
-	BUILD_BUG_ON((US) > 50000); \
 	if (!(ATOMIC)) { \
 		preempt_disable(); \
 		cpu = smp_processor_id(); \
@@ -130,8 +129,14 @@
 	ret__; \
 })
 
-#define wait_for_atomic(COND, MS)	_wait_for_atomic((COND), (MS) * 1000, 1)
-#define wait_for_atomic_us(COND, US)	_wait_for_atomic((COND), (US), 1)
+#define wait_for_atomic_us(COND, US) \
+({ \
+	BUILD_BUG_ON(!__builtin_constant_p(US)); \
+	BUILD_BUG_ON((US) > 50000); \
+	_wait_for_atomic((COND), (US), 1); \
+})
+
+#define wait_for_atomic(COND, MS) wait_for_atomic_us((COND), (MS) * 1000)
 
 #define KHz(x) (1000 * (x))
 #define MHz(x) KHz(1000 * (x))
@@ -261,7 +266,6 @@ struct intel_encoder {
 struct intel_panel {
 	struct drm_display_mode *fixed_mode;
 	struct drm_display_mode *downclock_mode;
-	int fitting_mode;
 
 	/* backlight */
 	struct {
@@ -284,9 +288,10 @@ struct intel_panel {
 		/* Connector and platform specific backlight functions */
 		int (*setup)(struct intel_connector *connector, enum pipe pipe);
 		uint32_t (*get)(struct intel_connector *connector);
-		void (*set)(struct intel_connector *connector, uint32_t level);
-		void (*disable)(struct intel_connector *connector);
-		void (*enable)(struct intel_connector *connector);
+		void (*set)(const struct drm_connector_state *conn_state, uint32_t level);
+		void (*disable)(const struct drm_connector_state *conn_state);
+		void (*enable)(const struct intel_crtc_state *crtc_state,
+			       const struct drm_connector_state *conn_state);
 		uint32_t (*hz_to_pwm)(struct intel_connector *connector,
 				      uint32_t hz);
 		void (*power)(struct intel_connector *, bool enable);
@@ -321,8 +326,20 @@ struct intel_connector {
 	void *port; /* store this opaque as its illegal to dereference it */
 
 	struct intel_dp *mst_port;
+
+	/* Work struct to schedule a uevent on link train failure */
+	struct work_struct modeset_retry_work;
 };
 
+struct intel_digital_connector_state {
+	struct drm_connector_state base;
+
+	enum hdmi_force_audio force_audio;
+	int broadcast_rgb;
+};
+
+#define to_intel_digital_connector_state(x) container_of(x, struct intel_digital_connector_state, base)
+
 struct dpll {
 	/* given values */
 	int n;
@@ -504,8 +521,8 @@ enum vlv_wm_level {
 };
 
 struct vlv_wm_state {
-	struct vlv_pipe_wm wm[NUM_VLV_WM_LEVELS];
-	struct vlv_sr_wm sr[NUM_VLV_WM_LEVELS];
+	struct g4x_pipe_wm wm[NUM_VLV_WM_LEVELS];
+	struct g4x_sr_wm sr[NUM_VLV_WM_LEVELS];
 	uint8_t num_levels;
 	bool cxsr;
 };
@@ -514,6 +531,22 @@ struct vlv_fifo_state {
 	u16 plane[I915_MAX_PLANES];
 };
 
+enum g4x_wm_level {
+	G4X_WM_LEVEL_NORMAL,
+	G4X_WM_LEVEL_SR,
+	G4X_WM_LEVEL_HPLL,
+	NUM_G4X_WM_LEVELS,
+};
+
+struct g4x_wm_state {
+	struct g4x_pipe_wm wm;
+	struct g4x_sr_wm sr;
+	struct g4x_sr_wm hpll;
+	bool cxsr;
+	bool hpll_en;
+	bool fbc_en;
+};
+
 struct intel_crtc_wm_state {
 	union {
 		struct {
@@ -541,7 +574,7 @@ struct intel_crtc_wm_state {
 
 		struct {
 			/* "raw" watermarks (not inverted) */
-			struct vlv_pipe_wm raw[NUM_VLV_WM_LEVELS];
+			struct g4x_pipe_wm raw[NUM_VLV_WM_LEVELS];
 			/* intermediate watermarks (inverted) */
 			struct vlv_wm_state intermediate;
 			/* optimal watermarks (inverted) */
@@ -549,6 +582,15 @@ struct intel_crtc_wm_state {
 			/* display FIFO split */
 			struct vlv_fifo_state fifo_state;
 		} vlv;
+
+		struct {
+			/* "raw" watermarks */
+			struct g4x_pipe_wm raw[NUM_G4X_WM_LEVELS];
+			/* intermediate watermarks */
+			struct g4x_wm_state intermediate;
+			/* optimal watermarks */
+			struct g4x_wm_state optimal;
+		} g4x;
 	};
 
 	/*
@@ -766,11 +808,6 @@ struct intel_crtc {
 	int adjusted_x;
 	int adjusted_y;
 
-	uint32_t cursor_addr;
-	uint32_t cursor_cntl;
-	uint32_t cursor_size;
-	uint32_t cursor_base;
-
 	struct intel_crtc_state *config;
 
 	/* global reset count when the last flip was submitted */
@@ -786,6 +823,7 @@ struct intel_crtc {
 		union {
 			struct intel_pipe_wm ilk;
 			struct vlv_wm_state vlv;
+			struct g4x_wm_state g4x;
 		} active;
 	} wm;
 
@@ -811,18 +849,22 @@ struct intel_plane {
 	int max_downscale;
 	uint32_t frontbuffer_bit;
 
+	struct {
+		u32 base, cntl, size;
+	} cursor;
+
 	/*
 	 * NOTE: Do not place new plane state fields here (e.g., when adding
 	 * new plane properties).  New runtime state should now be placed in
 	 * the intel_plane_state structure and accessed via plane_state.
 	 */
 
-	void (*update_plane)(struct drm_plane *plane,
+	void (*update_plane)(struct intel_plane *plane,
 			     const struct intel_crtc_state *crtc_state,
 			     const struct intel_plane_state *plane_state);
-	void (*disable_plane)(struct drm_plane *plane,
-			      struct drm_crtc *crtc);
-	int (*check_plane)(struct drm_plane *plane,
+	void (*disable_plane)(struct intel_plane *plane,
+			      struct intel_crtc *crtc);
+	int (*check_plane)(struct intel_plane *plane,
 			   struct intel_crtc_state *crtc_state,
 			   struct intel_plane_state *state);
 };
@@ -863,13 +905,9 @@ struct intel_hdmi {
 		enum drm_dp_dual_mode_type type;
 		int max_tmds_clock;
 	} dp_dual_mode;
-	bool limited_color_range;
-	bool color_range_auto;
 	bool has_hdmi_sink;
 	bool has_audio;
-	enum hdmi_force_audio force_audio;
 	bool rgb_quant_range_selectable;
-	enum hdmi_picture_aspect aspect_ratio;
 	struct intel_connector *attached_connector;
 	void (*write_infoframe)(struct drm_encoder *encoder,
 				const struct intel_crtc_state *crtc_state,
@@ -934,20 +972,24 @@ struct intel_dp {
 	bool detect_done;
 	bool channel_eq_status;
 	bool reset_link_params;
-	enum hdmi_force_audio force_audio;
-	bool limited_color_range;
-	bool color_range_auto;
 	uint8_t dpcd[DP_RECEIVER_CAP_SIZE];
 	uint8_t psr_dpcd[EDP_PSR_RECEIVER_CAP_SIZE];
 	uint8_t downstream_ports[DP_MAX_DOWNSTREAM_PORTS];
 	uint8_t edp_dpcd[EDP_DISPLAY_CTL_CAP_SIZE];
-	/* sink rates as reported by DP_SUPPORTED_LINK_RATES */
-	uint8_t num_sink_rates;
+	/* source rates */
+	int num_source_rates;
+	const int *source_rates;
+	/* sink rates as reported by DP_MAX_LINK_RATE/DP_SUPPORTED_LINK_RATES */
+	int num_sink_rates;
 	int sink_rates[DP_MAX_SUPPORTED_RATES];
-	/* Max lane count for the sink as per DPCD registers */
-	uint8_t max_sink_lane_count;
-	/* Max link BW for the sink as per DPCD registers */
-	int max_sink_link_bw;
+	bool use_rate_select;
+	/* intersection of source and sink rates */
+	int num_common_rates;
+	int common_rates[DP_MAX_SUPPORTED_RATES];
+	/* Max lane count for the current link */
+	int max_link_lane_count;
+	/* Max rate for the current link */
+	int max_link_rate;
 	/* sink or branch descriptor */
 	struct drm_dp_desc desc;
 	struct drm_dp_aux aux;
@@ -1300,7 +1342,6 @@ unsigned int intel_rotation_info_size(const struct intel_rotation_info *rot_info
 bool intel_has_pending_fb_unpin(struct drm_i915_private *dev_priv);
 void intel_mark_busy(struct drm_i915_private *dev_priv);
 void intel_mark_idle(struct drm_i915_private *dev_priv);
-void intel_crtc_restore_mode(struct drm_crtc *crtc);
 int intel_display_suspend(struct drm_device *dev);
 void intel_pps_unlock_regs_wa(struct drm_i915_private *dev_priv);
 void intel_encoder_destroy(struct drm_encoder *encoder);
@@ -1479,15 +1520,16 @@ bool intel_dp_compute_config(struct intel_encoder *encoder,
 bool intel_dp_is_edp(struct drm_i915_private *dev_priv, enum port port);
 enum irqreturn intel_dp_hpd_pulse(struct intel_digital_port *intel_dig_port,
 				  bool long_hpd);
-void intel_edp_backlight_on(struct intel_dp *intel_dp);
-void intel_edp_backlight_off(struct intel_dp *intel_dp);
+void intel_edp_backlight_on(const struct intel_crtc_state *crtc_state,
+			    const struct drm_connector_state *conn_state);
+void intel_edp_backlight_off(const struct drm_connector_state *conn_state);
 void intel_edp_panel_vdd_on(struct intel_dp *intel_dp);
 void intel_edp_panel_on(struct intel_dp *intel_dp);
 void intel_edp_panel_off(struct intel_dp *intel_dp);
-void intel_dp_add_properties(struct intel_dp *intel_dp, struct drm_connector *connector);
 void intel_dp_mst_suspend(struct drm_device *dev);
 void intel_dp_mst_resume(struct drm_device *dev);
 int intel_dp_max_link_rate(struct intel_dp *intel_dp);
+int intel_dp_max_lane_count(struct intel_dp *intel_dp);
 int intel_dp_rate_select(struct intel_dp *intel_dp, int rate);
 void intel_dp_hot_plug(struct intel_encoder *intel_encoder);
 void intel_power_sequencer_reset(struct drm_i915_private *dev_priv);
@@ -1660,12 +1702,13 @@ void intel_pch_panel_fitting(struct intel_crtc *crtc,
 void intel_gmch_panel_fitting(struct intel_crtc *crtc,
 			      struct intel_crtc_state *pipe_config,
 			      int fitting_mode);
-void intel_panel_set_backlight_acpi(struct intel_connector *connector,
+void intel_panel_set_backlight_acpi(const struct drm_connector_state *conn_state,
 				    u32 level, u32 max);
 int intel_panel_setup_backlight(struct drm_connector *connector,
 				enum pipe pipe);
-void intel_panel_enable_backlight(struct intel_connector *connector);
-void intel_panel_disable_backlight(struct intel_connector *connector);
+void intel_panel_enable_backlight(const struct intel_crtc_state *crtc_state,
+				  const struct drm_connector_state *conn_state);
+void intel_panel_disable_backlight(const struct drm_connector_state *old_conn_state);
 void intel_panel_destroy_backlight(struct drm_connector *connector);
 enum drm_connector_status intel_panel_detect(struct drm_i915_private *dev_priv);
 extern struct drm_display_mode *intel_find_panel_downclock(
@@ -1815,6 +1858,7 @@ void gen6_rps_boost(struct drm_i915_private *dev_priv,
 		    struct intel_rps_client *rps,
 		    unsigned long submitted);
 void intel_queue_rps_boost_for_request(struct drm_i915_gem_request *req);
+void g4x_wm_get_hw_state(struct drm_device *dev);
 void vlv_wm_get_hw_state(struct drm_device *dev);
 void ilk_wm_get_hw_state(struct drm_device *dev);
 void skl_wm_get_hw_state(struct drm_device *dev);
@@ -1822,6 +1866,7 @@ void skl_ddb_get_hw_state(struct drm_i915_private *dev_priv,
 			  struct skl_ddb_allocation *ddb /* out */);
 void skl_pipe_wm_get_hw_state(struct drm_crtc *crtc,
 			      struct skl_pipe_wm *out);
+void g4x_wm_sanitize(struct drm_i915_private *dev_priv);
 void vlv_wm_sanitize(struct drm_i915_private *dev_priv);
 bool intel_can_enable_sagv(struct drm_atomic_state *state);
 int intel_enable_sagv(struct drm_i915_private *dev_priv);
@@ -1833,6 +1878,8 @@ bool skl_ddb_allocation_overlaps(const struct skl_ddb_entry **entries,
 				 int ignore);
 bool ilk_disable_lp_wm(struct drm_device *dev);
 int sanitize_rc6_option(struct drm_i915_private *dev_priv, int enable_rc6);
+int skl_check_pipe_max_pixel_rate(struct intel_crtc *intel_crtc,
+				  struct intel_crtc_state *cstate);
 static inline int intel_enable_rc6(void)
 {
 	return i915.enable_rc6;
@@ -1857,10 +1904,19 @@ void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work
 void intel_tv_init(struct drm_i915_private *dev_priv);
 
 /* intel_atomic.c */
-int intel_connector_atomic_get_property(struct drm_connector *connector,
-					const struct drm_connector_state *state,
-					struct drm_property *property,
-					uint64_t *val);
+int intel_digital_connector_atomic_get_property(struct drm_connector *connector,
+						const struct drm_connector_state *state,
+						struct drm_property *property,
+						uint64_t *val);
+int intel_digital_connector_atomic_set_property(struct drm_connector *connector,
+						struct drm_connector_state *state,
+						struct drm_property *property,
+						uint64_t val);
+int intel_digital_connector_atomic_check(struct drm_connector *conn,
+					 struct drm_connector_state *new_state);
+struct drm_connector_state *
+intel_digital_connector_duplicate_state(struct drm_connector *connector);
+
 struct drm_crtc_state *intel_crtc_duplicate_state(struct drm_crtc *crtc);
 void intel_crtc_destroy_state(struct drm_crtc *crtc,
 			       struct drm_crtc_state *state);
diff --git a/drivers/gpu/drm/i915/intel_dsi.c b/drivers/gpu/drm/i915/intel_dsi.c
index fc0ef492252a..721f3f3adc1e 100644
--- a/drivers/gpu/drm/i915/intel_dsi.c
+++ b/drivers/gpu/drm/i915/intel_dsi.c
@@ -320,10 +320,10 @@ static bool intel_dsi_compute_config(struct intel_encoder *encoder,
 
 		if (HAS_GMCH_DISPLAY(dev_priv))
 			intel_gmch_panel_fitting(crtc, pipe_config,
-						 intel_connector->panel.fitting_mode);
+						 conn_state->scaling_mode);
 		else
 			intel_pch_panel_fitting(crtc, pipe_config,
-						intel_connector->panel.fitting_mode);
+						conn_state->scaling_mode);
 	}
 
 	/* DSI uses short packets for sync events, so clear mode flags for DSI */
@@ -835,7 +835,7 @@ static void intel_dsi_pre_enable(struct intel_encoder *encoder,
 		intel_dsi_port_enable(encoder);
 	}
 
-	intel_panel_enable_backlight(intel_dsi->attached_connector);
+	intel_panel_enable_backlight(pipe_config, conn_state);
 	intel_dsi_vbt_exec_sequence(intel_dsi, MIPI_SEQ_BACKLIGHT_ON);
 }
 
@@ -866,7 +866,7 @@ static void intel_dsi_disable(struct intel_encoder *encoder,
 	DRM_DEBUG_KMS("\n");
 
 	intel_dsi_vbt_exec_sequence(intel_dsi, MIPI_SEQ_BACKLIGHT_OFF);
-	intel_panel_disable_backlight(intel_dsi->attached_connector);
+	intel_panel_disable_backlight(old_conn_state);
 
 	/*
 	 * Disable Device ready before the port shutdown in order
@@ -1587,48 +1587,6 @@ static int intel_dsi_get_modes(struct drm_connector *connector)
 	return 1;
 }
 
-static int intel_dsi_set_property(struct drm_connector *connector,
-				  struct drm_property *property,
-				  uint64_t val)
-{
-	struct drm_device *dev = connector->dev;
-	struct intel_connector *intel_connector = to_intel_connector(connector);
-	struct drm_crtc *crtc;
-	int ret;
-
-	ret = drm_object_property_set_value(&connector->base, property, val);
-	if (ret)
-		return ret;
-
-	if (property == dev->mode_config.scaling_mode_property) {
-		if (val == DRM_MODE_SCALE_NONE) {
-			DRM_DEBUG_KMS("no scaling not supported\n");
-			return -EINVAL;
-		}
-		if (HAS_GMCH_DISPLAY(to_i915(dev)) &&
-		    val == DRM_MODE_SCALE_CENTER) {
-			DRM_DEBUG_KMS("centering not supported\n");
-			return -EINVAL;
-		}
-
-		if (intel_connector->panel.fitting_mode == val)
-			return 0;
-
-		intel_connector->panel.fitting_mode = val;
-	}
-
-	crtc = connector->state->crtc;
-	if (crtc && crtc->state->enable) {
-		/*
-		 * If the CRTC is enabled, the display will be changed
-		 * according to the new panel fitting mode.
-		 */
-		intel_crtc_restore_mode(crtc);
-	}
-
-	return 0;
-}
-
 static void intel_dsi_connector_destroy(struct drm_connector *connector)
 {
 	struct intel_connector *intel_connector = to_intel_connector(connector);
@@ -1657,6 +1615,7 @@ static const struct drm_encoder_funcs intel_dsi_funcs = {
 static const struct drm_connector_helper_funcs intel_dsi_connector_helper_funcs = {
 	.get_modes = intel_dsi_get_modes,
 	.mode_valid = intel_dsi_mode_valid,
+	.atomic_check = intel_digital_connector_atomic_check,
 };
 
 static const struct drm_connector_funcs intel_dsi_connector_funcs = {
@@ -1665,22 +1624,28 @@ static const struct drm_connector_funcs intel_dsi_connector_funcs = {
 	.early_unregister = intel_connector_unregister,
 	.destroy = intel_dsi_connector_destroy,
 	.fill_modes = drm_helper_probe_single_connector_modes,
-	.set_property = intel_dsi_set_property,
-	.atomic_get_property = intel_connector_atomic_get_property,
+	.set_property = drm_atomic_helper_connector_set_property,
+	.atomic_get_property = intel_digital_connector_atomic_get_property,
+	.atomic_set_property = intel_digital_connector_atomic_set_property,
 	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
-	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+	.atomic_duplicate_state = intel_digital_connector_duplicate_state,
 };
 
 static void intel_dsi_add_properties(struct intel_connector *connector)
 {
-	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 
 	if (connector->panel.fixed_mode) {
-		drm_mode_create_scaling_mode_property(dev);
-		drm_object_attach_property(&connector->base.base,
-					   dev->mode_config.scaling_mode_property,
-					   DRM_MODE_SCALE_ASPECT);
-		connector->panel.fitting_mode = DRM_MODE_SCALE_ASPECT;
+		u32 allowed_scalers;
+
+		allowed_scalers = BIT(DRM_MODE_SCALE_ASPECT) | BIT(DRM_MODE_SCALE_FULLSCREEN);
+		if (!HAS_GMCH_DISPLAY(dev_priv))
+			allowed_scalers |= BIT(DRM_MODE_SCALE_CENTER);
+
+		drm_connector_attach_scaling_mode_property(&connector->base,
+								allowed_scalers);
+
+		connector->base.state->scaling_mode = DRM_MODE_SCALE_ASPECT;
 	}
 }
 
diff --git a/drivers/gpu/drm/i915/intel_dsi_dcs_backlight.c b/drivers/gpu/drm/i915/intel_dsi_dcs_backlight.c
index ac7c6020c443..6e09ceb71500 100644
--- a/drivers/gpu/drm/i915/intel_dsi_dcs_backlight.c
+++ b/drivers/gpu/drm/i915/intel_dsi_dcs_backlight.c
@@ -60,10 +60,9 @@ static u32 dcs_get_backlight(struct intel_connector *connector)
 	return data;
 }
 
-static void dcs_set_backlight(struct intel_connector *connector, u32 level)
+static void dcs_set_backlight(const struct drm_connector_state *conn_state, u32 level)
 {
-	struct intel_encoder *encoder = connector->encoder;
-	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
+	struct intel_dsi *intel_dsi = enc_to_intel_dsi(conn_state->best_encoder);
 	struct mipi_dsi_device *dsi_device;
 	u8 data = level;
 	enum port port;
@@ -76,14 +75,13 @@ static void dcs_set_backlight(struct intel_connector *connector, u32 level)
 	}
 }
 
-static void dcs_disable_backlight(struct intel_connector *connector)
+static void dcs_disable_backlight(const struct drm_connector_state *conn_state)
 {
-	struct intel_encoder *encoder = connector->encoder;
-	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
+	struct intel_dsi *intel_dsi = enc_to_intel_dsi(conn_state->best_encoder);
 	struct mipi_dsi_device *dsi_device;
 	enum port port;
 
-	dcs_set_backlight(connector, 0);
+	dcs_set_backlight(conn_state, 0);
 
 	for_each_dsi_port(port, intel_dsi->dcs_cabc_ports) {
 		u8 cabc = POWER_SAVE_OFF;
@@ -110,11 +108,11 @@ static void dcs_disable_backlight(struct intel_connector *connector)
 	}
 }
 
-static void dcs_enable_backlight(struct intel_connector *connector)
+static void dcs_enable_backlight(const struct intel_crtc_state *crtc_state,
+				 const struct drm_connector_state *conn_state)
 {
-	struct intel_encoder *encoder = connector->encoder;
-	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
-	struct intel_panel *panel = &connector->panel;
+	struct intel_dsi *intel_dsi = enc_to_intel_dsi(conn_state->best_encoder);
+	struct intel_panel *panel = &to_intel_connector(conn_state->connector)->panel;
 	struct mipi_dsi_device *dsi_device;
 	enum port port;
 
@@ -142,7 +140,7 @@ static void dcs_enable_backlight(struct intel_connector *connector)
 				   &cabc, sizeof(cabc));
 	}
 
-	dcs_set_backlight(connector, panel->backlight.level);
+	dcs_set_backlight(conn_state, panel->backlight.level);
 }
 
 static int dcs_setup_backlight(struct intel_connector *connector,
diff --git a/drivers/gpu/drm/i915/intel_dsi_vbt.c b/drivers/gpu/drm/i915/intel_dsi_vbt.c
index 0dce7792643a..7158c7ce9c09 100644
--- a/drivers/gpu/drm/i915/intel_dsi_vbt.c
+++ b/drivers/gpu/drm/i915/intel_dsi_vbt.c
@@ -694,8 +694,8 @@ bool intel_dsi_vbt_init(struct intel_dsi *intel_dsi, u16 panel_id)
 						clk_zero_cnt << 8 | prepare_cnt;
 
 	/*
-	 * LP to HS switch count = 4TLPX + PREP_COUNT * 2 + EXIT_ZERO_COUNT * 2
-	 *					+ 10UI + Extra Byte Count
+	 * LP to HS switch count = 4TLPX + PREP_COUNT * mul + EXIT_ZERO_COUNT *
+	 *					mul + 10UI + Extra Byte Count
 	 *
 	 * HS to LP switch count = THS-TRAIL + 2TLPX + Extra Byte Count
 	 * Extra Byte Count is calculated according to number of lanes.
@@ -708,8 +708,8 @@ bool intel_dsi_vbt_init(struct intel_dsi *intel_dsi, u16 panel_id)
 	/* B044 */
 	/* FIXME:
 	 * The comment above does not match with the code */
-	lp_to_hs_switch = DIV_ROUND_UP(4 * tlpx_ui + prepare_cnt * 2 +
-						exit_zero_cnt * 2 + 10, 8);
+	lp_to_hs_switch = DIV_ROUND_UP(4 * tlpx_ui + prepare_cnt * mul +
+						exit_zero_cnt * mul + 10, 8);
 
 	hs_to_lp_switch = DIV_ROUND_UP(mipi_config->ths_trail + 2 * tlpx_ui, 8);
 
diff --git a/drivers/gpu/drm/i915/intel_dvo.c b/drivers/gpu/drm/i915/intel_dvo.c
index 6025839ed3b7..c1544a53095d 100644
--- a/drivers/gpu/drm/i915/intel_dvo.c
+++ b/drivers/gpu/drm/i915/intel_dvo.c
@@ -350,7 +350,7 @@ static const struct drm_connector_funcs intel_dvo_connector_funcs = {
 	.early_unregister = intel_connector_unregister,
 	.destroy = intel_dvo_destroy,
 	.fill_modes = drm_helper_probe_single_connector_modes,
-	.atomic_get_property = intel_connector_atomic_get_property,
+	.set_property = drm_atomic_helper_connector_set_property,
 	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
 	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
 };
diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c
index f94eacff196c..bc38bd128b76 100644
--- a/drivers/gpu/drm/i915/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/intel_engine_cs.c
@@ -26,69 +26,177 @@
 #include "intel_ringbuffer.h"
 #include "intel_lrc.h"
 
-static const struct engine_info {
+/* Haswell does have the CXT_SIZE register however it does not appear to be
+ * valid. Now, docs explain in dwords what is in the context object. The full
+ * size is 70720 bytes, however, the power context and execlist context will
+ * never be saved (power context is stored elsewhere, and execlists don't work
+ * on HSW) - so the final size, including the extra state required for the
+ * Resource Streamer, is 66944 bytes, which rounds to 17 pages.
+ */
+#define HSW_CXT_TOTAL_SIZE		(17 * PAGE_SIZE)
+/* Same as Haswell, but 72064 bytes now. */
+#define GEN8_CXT_TOTAL_SIZE		(18 * PAGE_SIZE)
+
+#define GEN8_LR_CONTEXT_RENDER_SIZE	(20 * PAGE_SIZE)
+#define GEN9_LR_CONTEXT_RENDER_SIZE	(22 * PAGE_SIZE)
+
+#define GEN8_LR_CONTEXT_OTHER_SIZE	( 2 * PAGE_SIZE)
+
+struct engine_class_info {
 	const char *name;
-	unsigned int exec_id;
+	int (*init_legacy)(struct intel_engine_cs *engine);
+	int (*init_execlists)(struct intel_engine_cs *engine);
+};
+
+static const struct engine_class_info intel_engine_classes[] = {
+	[RENDER_CLASS] = {
+		.name = "rcs",
+		.init_execlists = logical_render_ring_init,
+		.init_legacy = intel_init_render_ring_buffer,
+	},
+	[COPY_ENGINE_CLASS] = {
+		.name = "bcs",
+		.init_execlists = logical_xcs_ring_init,
+		.init_legacy = intel_init_blt_ring_buffer,
+	},
+	[VIDEO_DECODE_CLASS] = {
+		.name = "vcs",
+		.init_execlists = logical_xcs_ring_init,
+		.init_legacy = intel_init_bsd_ring_buffer,
+	},
+	[VIDEO_ENHANCEMENT_CLASS] = {
+		.name = "vecs",
+		.init_execlists = logical_xcs_ring_init,
+		.init_legacy = intel_init_vebox_ring_buffer,
+	},
+};
+
+struct engine_info {
 	unsigned int hw_id;
+	unsigned int uabi_id;
+	u8 class;
+	u8 instance;
 	u32 mmio_base;
 	unsigned irq_shift;
-	int (*init_legacy)(struct intel_engine_cs *engine);
-	int (*init_execlists)(struct intel_engine_cs *engine);
-} intel_engines[] = {
+};
+
+static const struct engine_info intel_engines[] = {
 	[RCS] = {
-		.name = "rcs",
 		.hw_id = RCS_HW,
-		.exec_id = I915_EXEC_RENDER,
+		.uabi_id = I915_EXEC_RENDER,
+		.class = RENDER_CLASS,
+		.instance = 0,
 		.mmio_base = RENDER_RING_BASE,
 		.irq_shift = GEN8_RCS_IRQ_SHIFT,
-		.init_execlists = logical_render_ring_init,
-		.init_legacy = intel_init_render_ring_buffer,
 	},
 	[BCS] = {
-		.name = "bcs",
 		.hw_id = BCS_HW,
-		.exec_id = I915_EXEC_BLT,
+		.uabi_id = I915_EXEC_BLT,
+		.class = COPY_ENGINE_CLASS,
+		.instance = 0,
 		.mmio_base = BLT_RING_BASE,
 		.irq_shift = GEN8_BCS_IRQ_SHIFT,
-		.init_execlists = logical_xcs_ring_init,
-		.init_legacy = intel_init_blt_ring_buffer,
 	},
 	[VCS] = {
-		.name = "vcs",
 		.hw_id = VCS_HW,
-		.exec_id = I915_EXEC_BSD,
+		.uabi_id = I915_EXEC_BSD,
+		.class = VIDEO_DECODE_CLASS,
+		.instance = 0,
 		.mmio_base = GEN6_BSD_RING_BASE,
 		.irq_shift = GEN8_VCS1_IRQ_SHIFT,
-		.init_execlists = logical_xcs_ring_init,
-		.init_legacy = intel_init_bsd_ring_buffer,
 	},
 	[VCS2] = {
-		.name = "vcs2",
 		.hw_id = VCS2_HW,
-		.exec_id = I915_EXEC_BSD,
+		.uabi_id = I915_EXEC_BSD,
+		.class = VIDEO_DECODE_CLASS,
+		.instance = 1,
 		.mmio_base = GEN8_BSD2_RING_BASE,
 		.irq_shift = GEN8_VCS2_IRQ_SHIFT,
-		.init_execlists = logical_xcs_ring_init,
-		.init_legacy = intel_init_bsd2_ring_buffer,
 	},
 	[VECS] = {
-		.name = "vecs",
 		.hw_id = VECS_HW,
-		.exec_id = I915_EXEC_VEBOX,
+		.uabi_id = I915_EXEC_VEBOX,
+		.class = VIDEO_ENHANCEMENT_CLASS,
+		.instance = 0,
 		.mmio_base = VEBOX_RING_BASE,
 		.irq_shift = GEN8_VECS_IRQ_SHIFT,
-		.init_execlists = logical_xcs_ring_init,
-		.init_legacy = intel_init_vebox_ring_buffer,
 	},
 };
 
+/**
+ * ___intel_engine_context_size() - return the size of the context for an engine
+ * @dev_priv: i915 device private
+ * @class: engine class
+ *
+ * Each engine class may require a different amount of space for a context
+ * image.
+ *
+ * Return: size (in bytes) of an engine class specific context image
+ *
+ * Note: this size includes the HWSP, which is part of the context image
+ * in LRC mode, but does not include the "shared data page" used with
+ * GuC submission. The caller should account for this if using the GuC.
+ */
+static u32
+__intel_engine_context_size(struct drm_i915_private *dev_priv, u8 class)
+{
+	u32 cxt_size;
+
+	BUILD_BUG_ON(I915_GTT_PAGE_SIZE != PAGE_SIZE);
+
+	switch (class) {
+	case RENDER_CLASS:
+		switch (INTEL_GEN(dev_priv)) {
+		default:
+			MISSING_CASE(INTEL_GEN(dev_priv));
+		case 9:
+			return GEN9_LR_CONTEXT_RENDER_SIZE;
+		case 8:
+			return i915.enable_execlists ?
+			       GEN8_LR_CONTEXT_RENDER_SIZE :
+			       GEN8_CXT_TOTAL_SIZE;
+		case 7:
+			if (IS_HASWELL(dev_priv))
+				return HSW_CXT_TOTAL_SIZE;
+
+			cxt_size = I915_READ(GEN7_CXT_SIZE);
+			return round_up(GEN7_CXT_TOTAL_SIZE(cxt_size) * 64,
+					PAGE_SIZE);
+		case 6:
+			cxt_size = I915_READ(CXT_SIZE);
+			return round_up(GEN6_CXT_TOTAL_SIZE(cxt_size) * 64,
+					PAGE_SIZE);
+		case 5:
+		case 4:
+		case 3:
+		case 2:
+		/* For the special day when i810 gets merged. */
+		case 1:
+			return 0;
+		}
+		break;
+	default:
+		MISSING_CASE(class);
+	case VIDEO_DECODE_CLASS:
+	case VIDEO_ENHANCEMENT_CLASS:
+	case COPY_ENGINE_CLASS:
+		if (INTEL_GEN(dev_priv) < 8)
+			return 0;
+		return GEN8_LR_CONTEXT_OTHER_SIZE;
+	}
+}
+
 static int
 intel_engine_setup(struct drm_i915_private *dev_priv,
 		   enum intel_engine_id id)
 {
 	const struct engine_info *info = &intel_engines[id];
+	const struct engine_class_info *class_info;
 	struct intel_engine_cs *engine;
 
+	GEM_BUG_ON(info->class >= ARRAY_SIZE(intel_engine_classes));
+	class_info = &intel_engine_classes[info->class];
+
 	GEM_BUG_ON(dev_priv->engine[id]);
 	engine = kzalloc(sizeof(*engine), GFP_KERNEL);
 	if (!engine)
@@ -96,11 +204,20 @@ intel_engine_setup(struct drm_i915_private *dev_priv,
 
 	engine->id = id;
 	engine->i915 = dev_priv;
-	engine->name = info->name;
-	engine->exec_id = info->exec_id;
+	WARN_ON(snprintf(engine->name, sizeof(engine->name), "%s%u",
+			 class_info->name, info->instance) >=
+		sizeof(engine->name));
+	engine->uabi_id = info->uabi_id;
 	engine->hw_id = engine->guc_id = info->hw_id;
 	engine->mmio_base = info->mmio_base;
 	engine->irq_shift = info->irq_shift;
+	engine->class = info->class;
+	engine->instance = info->instance;
+
+	engine->context_size = __intel_engine_context_size(dev_priv,
+							   engine->class);
+	if (WARN_ON(engine->context_size > BIT(20)))
+		engine->context_size = 0;
 
 	/* Nothing to do here, execute in order of dependencies */
 	engine->schedule = NULL;
@@ -112,18 +229,18 @@ intel_engine_setup(struct drm_i915_private *dev_priv,
 }
 
 /**
- * intel_engines_init_early() - allocate the Engine Command Streamers
+ * intel_engines_init_mmio() - allocate and prepare the Engine Command Streamers
  * @dev_priv: i915 device private
  *
  * Return: non-zero if the initialization failed.
  */
-int intel_engines_init_early(struct drm_i915_private *dev_priv)
+int intel_engines_init_mmio(struct drm_i915_private *dev_priv)
 {
 	struct intel_device_info *device_info = mkwrite_device_info(dev_priv);
-	unsigned int ring_mask = INTEL_INFO(dev_priv)->ring_mask;
-	unsigned int mask = 0;
+	const unsigned int ring_mask = INTEL_INFO(dev_priv)->ring_mask;
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
+	unsigned int mask = 0;
 	unsigned int i;
 	int err;
 
@@ -150,6 +267,12 @@ int intel_engines_init_early(struct drm_i915_private *dev_priv)
 	if (WARN_ON(mask != ring_mask))
 		device_info->ring_mask = mask;
 
+	/* We always presume we have at least RCS available for later probing */
+	if (WARN_ON(!HAS_ENGINE(dev_priv, RCS))) {
+		err = -ENODEV;
+		goto cleanup;
+	}
+
 	device_info->num_rings = hweight32(mask);
 
 	return 0;
@@ -161,7 +284,7 @@ cleanup:
 }
 
 /**
- * intel_engines_init() - allocate, populate and init the Engine Command Streamers
+ * intel_engines_init() - init the Engine Command Streamers
  * @dev_priv: i915 device private
  *
  * Return: non-zero if the initialization failed.
@@ -175,12 +298,14 @@ int intel_engines_init(struct drm_i915_private *dev_priv)
 	int err = 0;
 
 	for_each_engine(engine, dev_priv, id) {
+		const struct engine_class_info *class_info =
+			&intel_engine_classes[engine->class];
 		int (*init)(struct intel_engine_cs *engine);
 
 		if (i915.enable_execlists)
-			init = intel_engines[id].init_execlists;
+			init = class_info->init_execlists;
 		else
-			init = intel_engines[id].init_legacy;
+			init = class_info->init_legacy;
 		if (!init) {
 			kfree(engine);
 			dev_priv->engine[id] = NULL;
@@ -223,6 +348,9 @@ void intel_engine_init_global_seqno(struct intel_engine_cs *engine, u32 seqno)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
 
+	GEM_BUG_ON(!intel_engine_is_idle(engine));
+	GEM_BUG_ON(i915_gem_active_isset(&engine->timeline->last_request));
+
 	/* Our semaphore implementation is strictly monotonic (i.e. we proceed
 	 * so long as the semaphore value in the register/page is greater
 	 * than the sync value), so whenever we reset the seqno,
@@ -253,13 +381,12 @@ void intel_engine_init_global_seqno(struct intel_engine_cs *engine, u32 seqno)
 	intel_write_status_page(engine, I915_GEM_HWS_INDEX, seqno);
 	clear_bit(ENGINE_IRQ_BREADCRUMB, &engine->irq_posted);
 
-	GEM_BUG_ON(i915_gem_active_isset(&engine->timeline->last_request));
-	engine->hangcheck.seqno = seqno;
-
 	/* After manually advancing the seqno, fake the interrupt in case
 	 * there are any waiters for that seqno.
 	 */
 	intel_engine_wakeup(engine);
+
+	GEM_BUG_ON(intel_engine_get_seqno(engine) != seqno);
 }
 
 static void intel_engine_init_timeline(struct intel_engine_cs *engine)
@@ -342,6 +469,7 @@ static void intel_engine_cleanup_scratch(struct intel_engine_cs *engine)
  */
 int intel_engine_init_common(struct intel_engine_cs *engine)
 {
+	struct intel_ring *ring;
 	int ret;
 
 	engine->set_default_submission(engine);
@@ -353,9 +481,9 @@ int intel_engine_init_common(struct intel_engine_cs *engine)
 	 * be available. To avoid this we always pin the default
 	 * context.
 	 */
-	ret = engine->context_pin(engine, engine->i915->kernel_context);
-	if (ret)
-		return ret;
+	ring = engine->context_pin(engine, engine->i915->kernel_context);
+	if (IS_ERR(ring))
+		return PTR_ERR(ring);
 
 	ret = intel_engine_init_breadcrumbs(engine);
 	if (ret)
@@ -723,8 +851,10 @@ static int gen9_init_workarounds(struct intel_engine_cs *engine)
 		 */
 	}
 
+	/* WaEnableYV12BugFixInHalfSliceChicken7:skl,bxt,kbl,glk */
 	/* WaEnableSamplerGPGPUPreemptionSupport:skl,bxt,kbl */
 	WA_SET_BIT_MASKED(GEN9_HALF_SLICE_CHICKEN7,
+			  GEN9_ENABLE_YV12_BUGFIX |
 			  GEN9_ENABLE_GPGPU_PREEMPTION);
 
 	/* Wa4x4STCOptimizationDisable:skl,bxt,kbl,glk */
@@ -1082,6 +1212,11 @@ static bool ring_is_idle(struct intel_engine_cs *engine)
 
 	intel_runtime_pm_get(dev_priv);
 
+	/* First check that no commands are left in the ring */
+	if ((I915_READ_HEAD(engine) & HEAD_ADDR) !=
+	    (I915_READ_TAIL(engine) & TAIL_ADDR))
+		idle = false;
+
 	/* No bit for gen2, so assume the CS parser is idle */
 	if (INTEL_GEN(dev_priv) > 2 && !(I915_READ_MODE(engine) & MODE_IDLE))
 		idle = false;
@@ -1100,17 +1235,26 @@ static bool ring_is_idle(struct intel_engine_cs *engine)
  */
 bool intel_engine_is_idle(struct intel_engine_cs *engine)
 {
+	struct drm_i915_private *dev_priv = engine->i915;
+
+	/* More white lies, if wedged, hw state is inconsistent */
+	if (i915_terminally_wedged(&dev_priv->gpu_error))
+		return true;
+
 	/* Any inflight/incomplete requests? */
 	if (!i915_seqno_passed(intel_engine_get_seqno(engine),
 			       intel_engine_last_submit(engine)))
 		return false;
 
+	if (I915_SELFTEST_ONLY(engine->breadcrumbs.mock))
+		return true;
+
 	/* Interrupt/tasklet pending? */
 	if (test_bit(ENGINE_IRQ_EXECLIST, &engine->irq_posted))
 		return false;
 
 	/* Both ports drained, no more ELSP submission? */
-	if (engine->execlist_port[0].request)
+	if (port_request(&engine->execlist_port[0]))
 		return false;
 
 	/* Ring stopped? */
@@ -1151,6 +1295,18 @@ void intel_engines_reset_default_submission(struct drm_i915_private *i915)
 		engine->set_default_submission(engine);
 }
 
+void intel_engines_mark_idle(struct drm_i915_private *i915)
+{
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+
+	for_each_engine(engine, i915, id) {
+		intel_engine_disarm_breadcrumbs(engine);
+		i915_gem_batch_pool_fini(&engine->batch_pool);
+		engine->no_priolist = false;
+	}
+}
+
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
 #include "selftests/mock_engine.c"
 #endif
diff --git a/drivers/gpu/drm/i915/intel_fbc.c b/drivers/gpu/drm/i915/intel_fbc.c
index d93c58410bff..860b8c26d29b 100644
--- a/drivers/gpu/drm/i915/intel_fbc.c
+++ b/drivers/gpu/drm/i915/intel_fbc.c
@@ -796,7 +796,7 @@ static bool intel_fbc_can_activate(struct intel_crtc *crtc)
 		return false;
 	}
 	if (INTEL_GEN(dev_priv) <= 4 && !IS_G4X(dev_priv) &&
-	    cache->plane.rotation != DRM_ROTATE_0) {
+	    cache->plane.rotation != DRM_MODE_ROTATE_0) {
 		fbc->no_fbc_reason = "rotation unsupported";
 		return false;
 	}
@@ -1307,14 +1307,12 @@ static int intel_sanitize_fbc_option(struct drm_i915_private *dev_priv)
 
 static bool need_fbc_vtd_wa(struct drm_i915_private *dev_priv)
 {
-#ifdef CONFIG_INTEL_IOMMU
 	/* WaFbcTurnOffFbcWhenHyperVisorIsUsed:skl,bxt */
-	if (intel_iommu_gfx_mapped &&
+	if (intel_vtd_active() &&
 	    (IS_SKYLAKE(dev_priv) || IS_BROXTON(dev_priv))) {
 		DRM_INFO("Disabling framebuffer compression (FBC) to prevent screen flicker with VT-d enabled\n");
 		return true;
 	}
-#endif
 
 	return false;
 }
diff --git a/drivers/gpu/drm/i915/intel_fbdev.c b/drivers/gpu/drm/i915/intel_fbdev.c
index 332254a8eebe..03347c6ae599 100644
--- a/drivers/gpu/drm/i915/intel_fbdev.c
+++ b/drivers/gpu/drm/i915/intel_fbdev.c
@@ -211,7 +211,7 @@ static int intelfb_create(struct drm_fb_helper *helper,
 	 * This also validates that any existing fb inherited from the
 	 * BIOS is suitable for own access.
 	 */
-	vma = intel_pin_and_fence_fb_obj(&ifbdev->fb->base, DRM_ROTATE_0);
+	vma = intel_pin_and_fence_fb_obj(&ifbdev->fb->base, DRM_MODE_ROTATE_0);
 	if (IS_ERR(vma)) {
 		ret = PTR_ERR(vma);
 		goto out_unlock;
diff --git a/drivers/gpu/drm/i915/intel_guc_ct.c b/drivers/gpu/drm/i915/intel_guc_ct.c
new file mode 100644
index 000000000000..c4cbec140101
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_guc_ct.c
@@ -0,0 +1,461 @@
+/*
+ * Copyright © 2016-2017 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ */
+
+#include "i915_drv.h"
+#include "intel_guc_ct.h"
+
+enum { CTB_SEND = 0, CTB_RECV = 1 };
+
+enum { CTB_OWNER_HOST = 0 };
+
+void intel_guc_ct_init_early(struct intel_guc_ct *ct)
+{
+	/* we're using static channel owners */
+	ct->host_channel.owner = CTB_OWNER_HOST;
+}
+
+static inline const char *guc_ct_buffer_type_to_str(u32 type)
+{
+	switch (type) {
+	case INTEL_GUC_CT_BUFFER_TYPE_SEND:
+		return "SEND";
+	case INTEL_GUC_CT_BUFFER_TYPE_RECV:
+		return "RECV";
+	default:
+		return "<invalid>";
+	}
+}
+
+static void guc_ct_buffer_desc_init(struct guc_ct_buffer_desc *desc,
+				    u32 cmds_addr, u32 size, u32 owner)
+{
+	DRM_DEBUG_DRIVER("CT: desc %p init addr=%#x size=%u owner=%u\n",
+			 desc, cmds_addr, size, owner);
+	memset(desc, 0, sizeof(*desc));
+	desc->addr = cmds_addr;
+	desc->size = size;
+	desc->owner = owner;
+}
+
+static void guc_ct_buffer_desc_reset(struct guc_ct_buffer_desc *desc)
+{
+	DRM_DEBUG_DRIVER("CT: desc %p reset head=%u tail=%u\n",
+			 desc, desc->head, desc->tail);
+	desc->head = 0;
+	desc->tail = 0;
+	desc->is_in_error = 0;
+}
+
+static int guc_action_register_ct_buffer(struct intel_guc *guc,
+					 u32 desc_addr,
+					 u32 type)
+{
+	u32 action[] = {
+		INTEL_GUC_ACTION_REGISTER_COMMAND_TRANSPORT_BUFFER,
+		desc_addr,
+		sizeof(struct guc_ct_buffer_desc),
+		type
+	};
+	int err;
+
+	/* Can't use generic send(), CT registration must go over MMIO */
+	err = intel_guc_send_mmio(guc, action, ARRAY_SIZE(action));
+	if (err)
+		DRM_ERROR("CT: register %s buffer failed; err=%d\n",
+			  guc_ct_buffer_type_to_str(type), err);
+	return err;
+}
+
+static int guc_action_deregister_ct_buffer(struct intel_guc *guc,
+					   u32 owner,
+					   u32 type)
+{
+	u32 action[] = {
+		INTEL_GUC_ACTION_DEREGISTER_COMMAND_TRANSPORT_BUFFER,
+		owner,
+		type
+	};
+	int err;
+
+	/* Can't use generic send(), CT deregistration must go over MMIO */
+	err = intel_guc_send_mmio(guc, action, ARRAY_SIZE(action));
+	if (err)
+		DRM_ERROR("CT: deregister %s buffer failed; owner=%d err=%d\n",
+			  guc_ct_buffer_type_to_str(type), owner, err);
+	return err;
+}
+
+static bool ctch_is_open(struct intel_guc_ct_channel *ctch)
+{
+	return ctch->vma != NULL;
+}
+
+static int ctch_init(struct intel_guc *guc,
+		     struct intel_guc_ct_channel *ctch)
+{
+	struct i915_vma *vma;
+	void *blob;
+	int err;
+	int i;
+
+	GEM_BUG_ON(ctch->vma);
+
+	/* We allocate 1 page to hold both descriptors and both buffers.
+	 *       ___________.....................
+	 *      |desc (SEND)|                   :
+	 *      |___________|                   PAGE/4
+	 *      :___________....................:
+	 *      |desc (RECV)|                   :
+	 *      |___________|                   PAGE/4
+	 *      :_______________________________:
+	 *      |cmds (SEND)                    |
+	 *      |                               PAGE/4
+	 *      |_______________________________|
+	 *      |cmds (RECV)                    |
+	 *      |                               PAGE/4
+	 *      |_______________________________|
+	 *
+	 * Each message can use a maximum of 32 dwords and we don't expect to
+	 * have more than 1 in flight at any time, so we have enough space.
+	 * Some logic further ahead will rely on the fact that there is only 1
+	 * page and that it is always mapped, so if the size is changed the
+	 * other code will need updating as well.
+	 */
+
+	/* allocate vma */
+	vma = intel_guc_allocate_vma(guc, PAGE_SIZE);
+	if (IS_ERR(vma)) {
+		err = PTR_ERR(vma);
+		goto err_out;
+	}
+	ctch->vma = vma;
+
+	/* map first page */
+	blob = i915_gem_object_pin_map(vma->obj, I915_MAP_WB);
+	if (IS_ERR(blob)) {
+		err = PTR_ERR(blob);
+		goto err_vma;
+	}
+	DRM_DEBUG_DRIVER("CT: vma base=%#x\n", guc_ggtt_offset(ctch->vma));
+
+	/* store pointers to desc and cmds */
+	for (i = 0; i < ARRAY_SIZE(ctch->ctbs); i++) {
+		GEM_BUG_ON((i != CTB_SEND) && (i != CTB_RECV));
+		ctch->ctbs[i].desc = blob + PAGE_SIZE/4 * i;
+		ctch->ctbs[i].cmds = blob + PAGE_SIZE/4 * i + PAGE_SIZE/2;
+	}
+
+	return 0;
+
+err_vma:
+	i915_vma_unpin_and_release(&ctch->vma);
+err_out:
+	DRM_DEBUG_DRIVER("CT: channel %d initialization failed; err=%d\n",
+			 ctch->owner, err);
+	return err;
+}
+
+static void ctch_fini(struct intel_guc *guc,
+		      struct intel_guc_ct_channel *ctch)
+{
+	GEM_BUG_ON(!ctch->vma);
+
+	i915_gem_object_unpin_map(ctch->vma->obj);
+	i915_vma_unpin_and_release(&ctch->vma);
+}
+
+static int ctch_open(struct intel_guc *guc,
+		     struct intel_guc_ct_channel *ctch)
+{
+	u32 base;
+	int err;
+	int i;
+
+	DRM_DEBUG_DRIVER("CT: channel %d reopen=%s\n",
+			 ctch->owner, yesno(ctch_is_open(ctch)));
+
+	if (!ctch->vma) {
+		err = ctch_init(guc, ctch);
+		if (unlikely(err))
+			goto err_out;
+	}
+
+	/* vma should be already allocated and map'ed */
+	base = guc_ggtt_offset(ctch->vma);
+
+	/* (re)initialize descriptors
+	 * cmds buffers are in the second half of the blob page
+	 */
+	for (i = 0; i < ARRAY_SIZE(ctch->ctbs); i++) {
+		GEM_BUG_ON((i != CTB_SEND) && (i != CTB_RECV));
+		guc_ct_buffer_desc_init(ctch->ctbs[i].desc,
+					base + PAGE_SIZE/4 * i + PAGE_SIZE/2,
+					PAGE_SIZE/4,
+					ctch->owner);
+	}
+
+	/* register buffers, starting wirh RECV buffer
+	 * descriptors are in first half of the blob
+	 */
+	err = guc_action_register_ct_buffer(guc,
+					    base + PAGE_SIZE/4 * CTB_RECV,
+					    INTEL_GUC_CT_BUFFER_TYPE_RECV);
+	if (unlikely(err))
+		goto err_fini;
+
+	err = guc_action_register_ct_buffer(guc,
+					    base + PAGE_SIZE/4 * CTB_SEND,
+					    INTEL_GUC_CT_BUFFER_TYPE_SEND);
+	if (unlikely(err))
+		goto err_deregister;
+
+	return 0;
+
+err_deregister:
+	guc_action_deregister_ct_buffer(guc,
+					ctch->owner,
+					INTEL_GUC_CT_BUFFER_TYPE_RECV);
+err_fini:
+	ctch_fini(guc, ctch);
+err_out:
+	DRM_ERROR("CT: can't open channel %d; err=%d\n", ctch->owner, err);
+	return err;
+}
+
+static void ctch_close(struct intel_guc *guc,
+		       struct intel_guc_ct_channel *ctch)
+{
+	GEM_BUG_ON(!ctch_is_open(ctch));
+
+	guc_action_deregister_ct_buffer(guc,
+					ctch->owner,
+					INTEL_GUC_CT_BUFFER_TYPE_SEND);
+	guc_action_deregister_ct_buffer(guc,
+					ctch->owner,
+					INTEL_GUC_CT_BUFFER_TYPE_RECV);
+	ctch_fini(guc, ctch);
+}
+
+static u32 ctch_get_next_fence(struct intel_guc_ct_channel *ctch)
+{
+	/* For now it's trivial */
+	return ++ctch->next_fence;
+}
+
+static int ctb_write(struct intel_guc_ct_buffer *ctb,
+		     const u32 *action,
+		     u32 len /* in dwords */,
+		     u32 fence)
+{
+	struct guc_ct_buffer_desc *desc = ctb->desc;
+	u32 head = desc->head / 4;	/* in dwords */
+	u32 tail = desc->tail / 4;	/* in dwords */
+	u32 size = desc->size / 4;	/* in dwords */
+	u32 used;			/* in dwords */
+	u32 header;
+	u32 *cmds = ctb->cmds;
+	unsigned int i;
+
+	GEM_BUG_ON(desc->size % 4);
+	GEM_BUG_ON(desc->head % 4);
+	GEM_BUG_ON(desc->tail % 4);
+	GEM_BUG_ON(tail >= size);
+
+	/*
+	 * tail == head condition indicates empty. GuC FW does not support
+	 * using up the entire buffer to get tail == head meaning full.
+	 */
+	if (tail < head)
+		used = (size - head) + tail;
+	else
+		used = tail - head;
+
+	/* make sure there is a space including extra dw for the fence */
+	if (unlikely(used + len + 1 >= size))
+		return -ENOSPC;
+
+	/* Write the message. The format is the following:
+	 * DW0: header (including action code)
+	 * DW1: fence
+	 * DW2+: action data
+	 */
+	header = (len << GUC_CT_MSG_LEN_SHIFT) |
+		 (GUC_CT_MSG_WRITE_FENCE_TO_DESC) |
+		 (action[0] << GUC_CT_MSG_ACTION_SHIFT);
+
+	cmds[tail] = header;
+	tail = (tail + 1) % size;
+
+	cmds[tail] = fence;
+	tail = (tail + 1) % size;
+
+	for (i = 1; i < len; i++) {
+		cmds[tail] = action[i];
+		tail = (tail + 1) % size;
+	}
+
+	/* now update desc tail (back in bytes) */
+	desc->tail = tail * 4;
+	GEM_BUG_ON(desc->tail > desc->size);
+
+	return 0;
+}
+
+/* Wait for the response from the GuC.
+ * @fence:	response fence
+ * @status:	placeholder for status
+ * return:	0 response received (status is valid)
+ *		-ETIMEDOUT no response within hardcoded timeout
+ *		-EPROTO no response, ct buffer was in error
+ */
+static int wait_for_response(struct guc_ct_buffer_desc *desc,
+			     u32 fence,
+			     u32 *status)
+{
+	int err;
+
+	/*
+	 * Fast commands should complete in less than 10us, so sample quickly
+	 * up to that length of time, then switch to a slower sleep-wait loop.
+	 * No GuC command should ever take longer than 10ms.
+	 */
+#define done (READ_ONCE(desc->fence) == fence)
+	err = wait_for_us(done, 10);
+	if (err)
+		err = wait_for(done, 10);
+#undef done
+
+	if (unlikely(err)) {
+		DRM_ERROR("CT: fence %u failed; reported fence=%u\n",
+			  fence, desc->fence);
+
+		if (WARN_ON(desc->is_in_error)) {
+			/* Something went wrong with the messaging, try to reset
+			 * the buffer and hope for the best
+			 */
+			guc_ct_buffer_desc_reset(desc);
+			err = -EPROTO;
+		}
+	}
+
+	*status = desc->status;
+	return err;
+}
+
+static int ctch_send(struct intel_guc *guc,
+		     struct intel_guc_ct_channel *ctch,
+		     const u32 *action,
+		     u32 len,
+		     u32 *status)
+{
+	struct intel_guc_ct_buffer *ctb = &ctch->ctbs[CTB_SEND];
+	struct guc_ct_buffer_desc *desc = ctb->desc;
+	u32 fence;
+	int err;
+
+	GEM_BUG_ON(!ctch_is_open(ctch));
+	GEM_BUG_ON(!len);
+	GEM_BUG_ON(len & ~GUC_CT_MSG_LEN_MASK);
+
+	fence = ctch_get_next_fence(ctch);
+	err = ctb_write(ctb, action, len, fence);
+	if (unlikely(err))
+		return err;
+
+	intel_guc_notify(guc);
+
+	err = wait_for_response(desc, fence, status);
+	if (unlikely(err))
+		return err;
+	if (*status != INTEL_GUC_STATUS_SUCCESS)
+		return -EIO;
+	return 0;
+}
+
+/*
+ * Command Transport (CT) buffer based GuC send function.
+ */
+static int intel_guc_send_ct(struct intel_guc *guc, const u32 *action, u32 len)
+{
+	struct intel_guc_ct_channel *ctch = &guc->ct.host_channel;
+	u32 status = ~0; /* undefined */
+	int err;
+
+	mutex_lock(&guc->send_mutex);
+
+	err = ctch_send(guc, ctch, action, len, &status);
+	if (unlikely(err)) {
+		DRM_ERROR("CT: send action %#X failed; err=%d status=%#X\n",
+			  action[0], err, status);
+	}
+
+	mutex_unlock(&guc->send_mutex);
+	return err;
+}
+
+/**
+ * Enable buffer based command transport
+ * Shall only be called for platforms with HAS_GUC_CT.
+ * @guc:	the guc
+ * return:	0 on success
+ *		non-zero on failure
+ */
+int intel_guc_enable_ct(struct intel_guc *guc)
+{
+	struct drm_i915_private *dev_priv = guc_to_i915(guc);
+	struct intel_guc_ct_channel *ctch = &guc->ct.host_channel;
+	int err;
+
+	GEM_BUG_ON(!HAS_GUC_CT(dev_priv));
+
+	err = ctch_open(guc, ctch);
+	if (unlikely(err))
+		return err;
+
+	/* Switch into cmd transport buffer based send() */
+	guc->send = intel_guc_send_ct;
+	DRM_INFO("CT: %s\n", enableddisabled(true));
+	return 0;
+}
+
+/**
+ * Disable buffer based command transport.
+ * Shall only be called for platforms with HAS_GUC_CT.
+ * @guc: the guc
+ */
+void intel_guc_disable_ct(struct intel_guc *guc)
+{
+	struct drm_i915_private *dev_priv = guc_to_i915(guc);
+	struct intel_guc_ct_channel *ctch = &guc->ct.host_channel;
+
+	GEM_BUG_ON(!HAS_GUC_CT(dev_priv));
+
+	if (!ctch_is_open(ctch))
+		return;
+
+	ctch_close(guc, ctch);
+
+	/* Disable send */
+	guc->send = intel_guc_send_nop;
+	DRM_INFO("CT: %s\n", enableddisabled(false));
+}
diff --git a/drivers/gpu/drm/i915/intel_guc_ct.h b/drivers/gpu/drm/i915/intel_guc_ct.h
new file mode 100644
index 000000000000..6d97f36fcc62
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_guc_ct.h
@@ -0,0 +1,86 @@
+/*
+ * Copyright © 2016-2017 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ */
+
+#ifndef _INTEL_GUC_CT_H_
+#define _INTEL_GUC_CT_H_
+
+struct intel_guc;
+struct i915_vma;
+
+#include "intel_guc_fwif.h"
+
+/**
+ * DOC: Command Transport (CT).
+ *
+ * Buffer based command transport is a replacement for MMIO based mechanism.
+ * It can be used to perform both host-2-guc and guc-to-host communication.
+ */
+
+/** Represents single command transport buffer.
+ *
+ * A single command transport buffer consists of two parts, the header
+ * record (command transport buffer descriptor) and the actual buffer which
+ * holds the commands.
+ *
+ * @desc: pointer to the buffer descriptor
+ * @cmds: pointer to the commands buffer
+ */
+struct intel_guc_ct_buffer {
+	struct guc_ct_buffer_desc *desc;
+	u32 *cmds;
+};
+
+/** Represents pair of command transport buffers.
+ *
+ * Buffers go in pairs to allow bi-directional communication.
+ * To simplify the code we place both of them in the same vma.
+ * Buffers from the same pair must share unique owner id.
+ *
+ * @vma: pointer to the vma with pair of CT buffers
+ * @ctbs: buffers for sending(0) and receiving(1) commands
+ * @owner: unique identifier
+ * @next_fence: fence to be used with next send command
+ */
+struct intel_guc_ct_channel {
+	struct i915_vma *vma;
+	struct intel_guc_ct_buffer ctbs[2];
+	u32 owner;
+	u32 next_fence;
+};
+
+/** Holds all command transport channels.
+ *
+ * @host_channel: main channel used by the host
+ */
+struct intel_guc_ct {
+	struct intel_guc_ct_channel host_channel;
+	/* other channels are tbd */
+};
+
+void intel_guc_ct_init_early(struct intel_guc_ct *ct);
+
+/* XXX: move to intel_uc.h ? don't fit there either */
+int intel_guc_enable_ct(struct intel_guc *guc);
+void intel_guc_disable_ct(struct intel_guc *guc);
+
+#endif /* _INTEL_GUC_CT_H_ */
diff --git a/drivers/gpu/drm/i915/intel_guc_fwif.h b/drivers/gpu/drm/i915/intel_guc_fwif.h
index cb36cbf3818f..5fa286074811 100644
--- a/drivers/gpu/drm/i915/intel_guc_fwif.h
+++ b/drivers/gpu/drm/i915/intel_guc_fwif.h
@@ -23,8 +23,8 @@
 #ifndef _INTEL_GUC_FWIF_H
 #define _INTEL_GUC_FWIF_H
 
-#define GFXCORE_FAMILY_GEN9		12
-#define GFXCORE_FAMILY_UNKNOWN		0x7fffffff
+#define GUC_CORE_FAMILY_GEN9		12
+#define GUC_CORE_FAMILY_UNKNOWN		0x7fffffff
 
 #define GUC_CLIENT_PRIORITY_KMD_HIGH	0
 #define GUC_CLIENT_PRIORITY_HIGH	1
@@ -331,6 +331,47 @@ struct guc_stage_desc {
 	u64 desc_private;
 } __packed;
 
+/*
+ * Describes single command transport buffer.
+ * Used by both guc-master and clients.
+ */
+struct guc_ct_buffer_desc {
+	u32 addr;		/* gfx address */
+	u64 host_private;	/* host private data */
+	u32 size;		/* size in bytes */
+	u32 head;		/* offset updated by GuC*/
+	u32 tail;		/* offset updated by owner */
+	u32 is_in_error;	/* error indicator */
+	u32 fence;		/* fence updated by GuC */
+	u32 status;		/* status updated by GuC */
+	u32 owner;		/* id of the channel owner */
+	u32 owner_sub_id;	/* owner-defined field for extra tracking */
+	u32 reserved[5];
+} __packed;
+
+/* Type of command transport buffer */
+#define INTEL_GUC_CT_BUFFER_TYPE_SEND	0x0u
+#define INTEL_GUC_CT_BUFFER_TYPE_RECV	0x1u
+
+/*
+ * Definition of the command transport message header (DW0)
+ *
+ * bit[4..0]	message len (in dwords)
+ * bit[7..5]	reserved
+ * bit[8]	write fence to desc
+ * bit[9]	write status to H2G buff
+ * bit[10]	send status (via G2H)
+ * bit[15..11]	reserved
+ * bit[31..16]	action code
+ */
+#define GUC_CT_MSG_LEN_SHIFT			0
+#define GUC_CT_MSG_LEN_MASK			0x1F
+#define GUC_CT_MSG_WRITE_FENCE_TO_DESC		(1 << 8)
+#define GUC_CT_MSG_WRITE_STATUS_TO_BUFF		(1 << 9)
+#define GUC_CT_MSG_SEND_STATUS			(1 << 10)
+#define GUC_CT_MSG_ACTION_SHIFT			16
+#define GUC_CT_MSG_ACTION_MASK			0xFFFF
+
 #define GUC_FORCEWAKE_RENDER	(1 << 0)
 #define GUC_FORCEWAKE_MEDIA	(1 << 1)
 
@@ -515,6 +556,8 @@ enum intel_guc_action {
 	INTEL_GUC_ACTION_EXIT_S_STATE = 0x502,
 	INTEL_GUC_ACTION_SLPC_REQUEST = 0x3003,
 	INTEL_GUC_ACTION_AUTHENTICATE_HUC = 0x4000,
+	INTEL_GUC_ACTION_REGISTER_COMMAND_TRANSPORT_BUFFER = 0x4505,
+	INTEL_GUC_ACTION_DEREGISTER_COMMAND_TRANSPORT_BUFFER = 0x4506,
 	INTEL_GUC_ACTION_UK_LOG_ENABLE_LOGGING = 0x0E000,
 	INTEL_GUC_ACTION_LIMIT
 };
diff --git a/drivers/gpu/drm/i915/intel_guc_loader.c b/drivers/gpu/drm/i915/intel_guc_loader.c
index 8a1a023e48b2..8b0ae7fce7f2 100644
--- a/drivers/gpu/drm/i915/intel_guc_loader.c
+++ b/drivers/gpu/drm/i915/intel_guc_loader.c
@@ -61,6 +61,9 @@
 #define KBL_FW_MAJOR 9
 #define KBL_FW_MINOR 14
 
+#define GLK_FW_MAJOR 10
+#define GLK_FW_MINOR 56
+
 #define GUC_FW_PATH(platform, major, minor) \
        "i915/" __stringify(platform) "_guc_ver" __stringify(major) "_" __stringify(minor) ".bin"
 
@@ -73,6 +76,8 @@ MODULE_FIRMWARE(I915_BXT_GUC_UCODE);
 #define I915_KBL_GUC_UCODE GUC_FW_PATH(kbl, KBL_FW_MAJOR, KBL_FW_MINOR)
 MODULE_FIRMWARE(I915_KBL_GUC_UCODE);
 
+#define I915_GLK_GUC_UCODE GUC_FW_PATH(glk, GLK_FW_MAJOR, GLK_FW_MINOR)
+
 
 static u32 get_gttype(struct drm_i915_private *dev_priv)
 {
@@ -86,11 +91,11 @@ static u32 get_core_family(struct drm_i915_private *dev_priv)
 
 	switch (gen) {
 	case 9:
-		return GFXCORE_FAMILY_GEN9;
+		return GUC_CORE_FAMILY_GEN9;
 
 	default:
-		WARN(1, "GEN%d does not support GuC operation!\n", gen);
-		return GFXCORE_FAMILY_UNKNOWN;
+		MISSING_CASE(gen);
+		return GUC_CORE_FAMILY_UNKNOWN;
 	}
 }
 
@@ -280,10 +285,6 @@ static int guc_ucode_xfer(struct drm_i915_private *dev_priv)
 
 	intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL);
 
-	/* init WOPCM */
-	I915_WRITE(GUC_WOPCM_SIZE, intel_guc_wopcm_size(dev_priv));
-	I915_WRITE(DMA_GUC_WOPCM_OFFSET, GUC_WOPCM_OFFSET_VALUE);
-
 	/* Enable MIA caching. GuC clock gating is disabled. */
 	I915_WRITE(GUC_SHIM_CONTROL, GUC_SHIM_CONTROL_VALUE);
 
@@ -401,10 +402,14 @@ int intel_guc_select_fw(struct intel_guc *guc)
 		guc->fw.path = I915_BXT_GUC_UCODE;
 		guc->fw.major_ver_wanted = BXT_FW_MAJOR;
 		guc->fw.minor_ver_wanted = BXT_FW_MINOR;
-	} else if (IS_KABYLAKE(dev_priv)) {
+	} else if (IS_KABYLAKE(dev_priv) || IS_COFFEELAKE(dev_priv)) {
 		guc->fw.path = I915_KBL_GUC_UCODE;
 		guc->fw.major_ver_wanted = KBL_FW_MAJOR;
 		guc->fw.minor_ver_wanted = KBL_FW_MINOR;
+	} else if (IS_GEMINILAKE(dev_priv)) {
+		guc->fw.path = I915_GLK_GUC_UCODE;
+		guc->fw.major_ver_wanted = GLK_FW_MAJOR;
+		guc->fw.minor_ver_wanted = GLK_FW_MINOR;
 	} else {
 		DRM_ERROR("No GuC firmware known for platform with GuC!\n");
 		return -ENOENT;
diff --git a/drivers/gpu/drm/i915/intel_guc_log.c b/drivers/gpu/drm/i915/intel_guc_log.c
index 6fb63a3c65b0..16d3b8719cab 100644
--- a/drivers/gpu/drm/i915/intel_guc_log.c
+++ b/drivers/gpu/drm/i915/intel_guc_log.c
@@ -359,12 +359,16 @@ static int guc_log_runtime_create(struct intel_guc *guc)
 	void *vaddr;
 	struct rchan *guc_log_relay_chan;
 	size_t n_subbufs, subbuf_size;
-	int ret = 0;
+	int ret;
 
 	lockdep_assert_held(&dev_priv->drm.struct_mutex);
 
 	GEM_BUG_ON(guc_log_has_runtime(guc));
 
+	ret = i915_gem_object_set_to_wc_domain(guc->log.vma->obj, true);
+	if (ret)
+		return ret;
+
 	/* Create a WC (Uncached for read) vmalloc mapping of log
 	 * buffer pages, so that we can directly get the data
 	 * (up-to-date) from memory.
diff --git a/drivers/gpu/drm/i915/intel_gvt.c b/drivers/gpu/drm/i915/intel_gvt.c
index e1ab6432a914..52d5b82790d9 100644
--- a/drivers/gpu/drm/i915/intel_gvt.c
+++ b/drivers/gpu/drm/i915/intel_gvt.c
@@ -51,6 +51,32 @@ static bool is_supported_device(struct drm_i915_private *dev_priv)
 }
 
 /**
+ * intel_gvt_sanitize_options - sanitize GVT related options
+ * @dev_priv: drm i915 private data
+ *
+ * This function is called at the i915 options sanitize stage.
+ */
+void intel_gvt_sanitize_options(struct drm_i915_private *dev_priv)
+{
+	if (!i915.enable_gvt)
+		return;
+
+	if (intel_vgpu_active(dev_priv)) {
+		DRM_INFO("GVT-g is disabled for guest\n");
+		goto bail;
+	}
+
+	if (!is_supported_device(dev_priv)) {
+		DRM_INFO("Unsupported device. GVT-g is disabled\n");
+		goto bail;
+	}
+
+	return;
+bail:
+	i915.enable_gvt = 0;
+}
+
+/**
  * intel_gvt_init - initialize GVT components
  * @dev_priv: drm i915 private data
  *
@@ -69,19 +95,14 @@ int intel_gvt_init(struct drm_i915_private *dev_priv)
 		return 0;
 	}
 
-	if (intel_vgpu_active(dev_priv)) {
-		DRM_DEBUG_DRIVER("GVT-g is disabled for guest\n");
-		goto bail;
-	}
-
-	if (!is_supported_device(dev_priv)) {
-		DRM_DEBUG_DRIVER("Unsupported device. GVT-g is disabled\n");
-		goto bail;
+	if (!i915.enable_execlists) {
+		DRM_ERROR("i915 GVT-g loading failed due to disabled execlists mode\n");
+		return -EIO;
 	}
 
-	if (!i915.enable_execlists) {
-		DRM_INFO("GPU guest virtualisation [GVT-g] disabled due to disabled execlist submission [i915.enable_execlists module parameter]\n");
-		goto bail;
+	if (i915.enable_guc_submission) {
+		DRM_ERROR("i915 GVT-g loading failed due to Graphics virtualization is not yet supported with GuC submission\n");
+		return -EIO;
 	}
 
 	/*
diff --git a/drivers/gpu/drm/i915/intel_gvt.h b/drivers/gpu/drm/i915/intel_gvt.h
index 25df2d65b985..61b246470282 100644
--- a/drivers/gpu/drm/i915/intel_gvt.h
+++ b/drivers/gpu/drm/i915/intel_gvt.h
@@ -32,6 +32,7 @@ void intel_gvt_cleanup(struct drm_i915_private *dev_priv);
 int intel_gvt_init_device(struct drm_i915_private *dev_priv);
 void intel_gvt_clean_device(struct drm_i915_private *dev_priv);
 int intel_gvt_init_host(void);
+void intel_gvt_sanitize_options(struct drm_i915_private *dev_priv);
 #else
 static inline int intel_gvt_init(struct drm_i915_private *dev_priv)
 {
@@ -40,6 +41,10 @@ static inline int intel_gvt_init(struct drm_i915_private *dev_priv)
 static inline void intel_gvt_cleanup(struct drm_i915_private *dev_priv)
 {
 }
+
+static inline void intel_gvt_sanitize_options(struct drm_i915_private *dev_priv)
+{
+}
 #endif
 
 #endif /* _INTEL_GVT_H_ */
diff --git a/drivers/gpu/drm/i915/intel_hangcheck.c b/drivers/gpu/drm/i915/intel_hangcheck.c
index dce742243ba6..9b0ece427bdc 100644
--- a/drivers/gpu/drm/i915/intel_hangcheck.c
+++ b/drivers/gpu/drm/i915/intel_hangcheck.c
@@ -407,7 +407,7 @@ static void hangcheck_declare_hang(struct drm_i915_private *i915,
 				 "%s, ", engine->name);
 	msg[len-2] = '\0';
 
-	return i915_handle_error(i915, hung, msg);
+	return i915_handle_error(i915, hung, "%s", msg);
 }
 
 /*
diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c
index 1d623b5e09d6..ec0779a52d53 100644
--- a/drivers/gpu/drm/i915/intel_hdmi.c
+++ b/drivers/gpu/drm/i915/intel_hdmi.c
@@ -1218,7 +1218,8 @@ static int intel_hdmi_source_max_tmds_clock(struct drm_i915_private *dev_priv)
 }
 
 static int hdmi_port_clock_limit(struct intel_hdmi *hdmi,
-				 bool respect_downstream_limits)
+				 bool respect_downstream_limits,
+				 bool force_dvi)
 {
 	struct drm_device *dev = intel_hdmi_to_dev(hdmi);
 	int max_tmds_clock = intel_hdmi_source_max_tmds_clock(to_i915(dev));
@@ -1234,7 +1235,7 @@ static int hdmi_port_clock_limit(struct intel_hdmi *hdmi,
 		if (info->max_tmds_clock)
 			max_tmds_clock = min(max_tmds_clock,
 					     info->max_tmds_clock);
-		else if (!hdmi->has_hdmi_sink)
+		else if (!hdmi->has_hdmi_sink || force_dvi)
 			max_tmds_clock = min(max_tmds_clock, 165000);
 	}
 
@@ -1243,13 +1244,14 @@ static int hdmi_port_clock_limit(struct intel_hdmi *hdmi,
 
 static enum drm_mode_status
 hdmi_port_clock_valid(struct intel_hdmi *hdmi,
-		      int clock, bool respect_downstream_limits)
+		      int clock, bool respect_downstream_limits,
+		      bool force_dvi)
 {
 	struct drm_i915_private *dev_priv = to_i915(intel_hdmi_to_dev(hdmi));
 
 	if (clock < 25000)
 		return MODE_CLOCK_LOW;
-	if (clock > hdmi_port_clock_limit(hdmi, respect_downstream_limits))
+	if (clock > hdmi_port_clock_limit(hdmi, respect_downstream_limits, force_dvi))
 		return MODE_CLOCK_HIGH;
 
 	/* BXT DPLL can't generate 223-240 MHz */
@@ -1273,6 +1275,8 @@ intel_hdmi_mode_valid(struct drm_connector *connector,
 	enum drm_mode_status status;
 	int clock;
 	int max_dotclk = to_i915(connector->dev)->max_dotclk_freq;
+	bool force_dvi =
+		READ_ONCE(to_intel_digital_connector_state(connector->state)->force_audio) == HDMI_AUDIO_OFF_DVI;
 
 	if (mode->flags & DRM_MODE_FLAG_DBLSCAN)
 		return MODE_NO_DBLESCAN;
@@ -1289,11 +1293,11 @@ intel_hdmi_mode_valid(struct drm_connector *connector,
 		clock *= 2;
 
 	/* check if we can do 8bpc */
-	status = hdmi_port_clock_valid(hdmi, clock, true);
+	status = hdmi_port_clock_valid(hdmi, clock, true, force_dvi);
 
 	/* if we can't do 8bpc we may still be able to do 12bpc */
-	if (!HAS_GMCH_DISPLAY(dev_priv) && status != MODE_OK)
-		status = hdmi_port_clock_valid(hdmi, clock * 3 / 2, true);
+	if (!HAS_GMCH_DISPLAY(dev_priv) && status != MODE_OK && hdmi->has_hdmi_sink && !force_dvi)
+		status = hdmi_port_clock_valid(hdmi, clock * 3 / 2, true, force_dvi);
 
 	return status;
 }
@@ -1327,6 +1331,11 @@ static bool hdmi_12bpc_possible(struct intel_crtc_state *crtc_state)
 			return false;
 	}
 
+	/* Display Wa #1139 */
+	if (IS_GLK_REVID(dev_priv, 0, GLK_REVID_A1) &&
+	    crtc_state->base.adjusted_mode.htotal > 5460)
+		return false;
+
 	return true;
 }
 
@@ -1338,16 +1347,19 @@ bool intel_hdmi_compute_config(struct intel_encoder *encoder,
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
 	struct drm_scdc *scdc = &conn_state->connector->display_info.hdmi.scdc;
+	struct intel_digital_connector_state *intel_conn_state =
+		to_intel_digital_connector_state(conn_state);
 	int clock_8bpc = pipe_config->base.adjusted_mode.crtc_clock;
 	int clock_12bpc = clock_8bpc * 3 / 2;
 	int desired_bpp;
+	bool force_dvi = intel_conn_state->force_audio == HDMI_AUDIO_OFF_DVI;
 
-	pipe_config->has_hdmi_sink = intel_hdmi->has_hdmi_sink;
+	pipe_config->has_hdmi_sink = !force_dvi && intel_hdmi->has_hdmi_sink;
 
 	if (pipe_config->has_hdmi_sink)
 		pipe_config->has_infoframe = true;
 
-	if (intel_hdmi->color_range_auto) {
+	if (intel_conn_state->broadcast_rgb == INTEL_BROADCAST_RGB_AUTO) {
 		/* See CEA-861-E - 5.1 Default Encoding Parameters */
 		pipe_config->limited_color_range =
 			pipe_config->has_hdmi_sink &&
@@ -1355,7 +1367,7 @@ bool intel_hdmi_compute_config(struct intel_encoder *encoder,
 			HDMI_QUANTIZATION_RANGE_LIMITED;
 	} else {
 		pipe_config->limited_color_range =
-			intel_hdmi->limited_color_range;
+			intel_conn_state->broadcast_rgb == INTEL_BROADCAST_RGB_LIMITED;
 	}
 
 	if (adjusted_mode->flags & DRM_MODE_FLAG_DBLCLK) {
@@ -1367,8 +1379,13 @@ bool intel_hdmi_compute_config(struct intel_encoder *encoder,
 	if (HAS_PCH_SPLIT(dev_priv) && !HAS_DDI(dev_priv))
 		pipe_config->has_pch_encoder = true;
 
-	if (pipe_config->has_hdmi_sink && intel_hdmi->has_audio)
-		pipe_config->has_audio = true;
+	if (pipe_config->has_hdmi_sink) {
+		if (intel_conn_state->force_audio == HDMI_AUDIO_AUTO)
+			pipe_config->has_audio = intel_hdmi->has_audio;
+		else
+			pipe_config->has_audio =
+				intel_conn_state->force_audio == HDMI_AUDIO_ON;
+	}
 
 	/*
 	 * HDMI is either 12 or 8, so if the display lets 10bpc sneak
@@ -1376,8 +1393,8 @@ bool intel_hdmi_compute_config(struct intel_encoder *encoder,
 	 * outputs. We also need to check that the higher clock still fits
 	 * within limits.
 	 */
-	if (pipe_config->pipe_bpp > 8*3 && pipe_config->has_hdmi_sink &&
-	    hdmi_port_clock_valid(intel_hdmi, clock_12bpc, true) == MODE_OK &&
+	if (pipe_config->pipe_bpp > 8*3 && pipe_config->has_hdmi_sink && !force_dvi &&
+	    hdmi_port_clock_valid(intel_hdmi, clock_12bpc, true, force_dvi) == MODE_OK &&
 	    hdmi_12bpc_possible(pipe_config)) {
 		DRM_DEBUG_KMS("picking bpc to 12 for HDMI output\n");
 		desired_bpp = 12*3;
@@ -1392,18 +1409,18 @@ bool intel_hdmi_compute_config(struct intel_encoder *encoder,
 	}
 
 	if (!pipe_config->bw_constrained) {
-		DRM_DEBUG_KMS("forcing pipe bpc to %i for HDMI\n", desired_bpp);
+		DRM_DEBUG_KMS("forcing pipe bpp to %i for HDMI\n", desired_bpp);
 		pipe_config->pipe_bpp = desired_bpp;
 	}
 
 	if (hdmi_port_clock_valid(intel_hdmi, pipe_config->port_clock,
-				  false) != MODE_OK) {
+				  false, force_dvi) != MODE_OK) {
 		DRM_DEBUG_KMS("unsupported HDMI clock, rejecting mode\n");
 		return false;
 	}
 
 	/* Set user selected PAR to incoming mode's member */
-	adjusted_mode->picture_aspect_ratio = intel_hdmi->aspect_ratio;
+	adjusted_mode->picture_aspect_ratio = conn_state->picture_aspect_ratio;
 
 	pipe_config->lane_count = 4;
 
@@ -1504,13 +1521,7 @@ intel_hdmi_set_edid(struct drm_connector *connector)
 			drm_rgb_quant_range_selectable(edid);
 
 		intel_hdmi->has_audio = drm_detect_monitor_audio(edid);
-		if (intel_hdmi->force_audio != HDMI_AUDIO_AUTO)
-			intel_hdmi->has_audio =
-				intel_hdmi->force_audio == HDMI_AUDIO_ON;
-
-		if (intel_hdmi->force_audio != HDMI_AUDIO_OFF_DVI)
-			intel_hdmi->has_hdmi_sink =
-				drm_detect_hdmi_monitor(edid);
+		intel_hdmi->has_hdmi_sink = drm_detect_hdmi_monitor(edid);
 
 		connected = true;
 	}
@@ -1572,108 +1583,6 @@ static int intel_hdmi_get_modes(struct drm_connector *connector)
 	return intel_connector_update_modes(connector, edid);
 }
 
-static bool
-intel_hdmi_detect_audio(struct drm_connector *connector)
-{
-	bool has_audio = false;
-	struct edid *edid;
-
-	edid = to_intel_connector(connector)->detect_edid;
-	if (edid && edid->input & DRM_EDID_INPUT_DIGITAL)
-		has_audio = drm_detect_monitor_audio(edid);
-
-	return has_audio;
-}
-
-static int
-intel_hdmi_set_property(struct drm_connector *connector,
-			struct drm_property *property,
-			uint64_t val)
-{
-	struct intel_hdmi *intel_hdmi = intel_attached_hdmi(connector);
-	struct intel_digital_port *intel_dig_port =
-		hdmi_to_dig_port(intel_hdmi);
-	struct drm_i915_private *dev_priv = to_i915(connector->dev);
-	int ret;
-
-	ret = drm_object_property_set_value(&connector->base, property, val);
-	if (ret)
-		return ret;
-
-	if (property == dev_priv->force_audio_property) {
-		enum hdmi_force_audio i = val;
-		bool has_audio;
-
-		if (i == intel_hdmi->force_audio)
-			return 0;
-
-		intel_hdmi->force_audio = i;
-
-		if (i == HDMI_AUDIO_AUTO)
-			has_audio = intel_hdmi_detect_audio(connector);
-		else
-			has_audio = (i == HDMI_AUDIO_ON);
-
-		if (i == HDMI_AUDIO_OFF_DVI)
-			intel_hdmi->has_hdmi_sink = 0;
-
-		intel_hdmi->has_audio = has_audio;
-		goto done;
-	}
-
-	if (property == dev_priv->broadcast_rgb_property) {
-		bool old_auto = intel_hdmi->color_range_auto;
-		bool old_range = intel_hdmi->limited_color_range;
-
-		switch (val) {
-		case INTEL_BROADCAST_RGB_AUTO:
-			intel_hdmi->color_range_auto = true;
-			break;
-		case INTEL_BROADCAST_RGB_FULL:
-			intel_hdmi->color_range_auto = false;
-			intel_hdmi->limited_color_range = false;
-			break;
-		case INTEL_BROADCAST_RGB_LIMITED:
-			intel_hdmi->color_range_auto = false;
-			intel_hdmi->limited_color_range = true;
-			break;
-		default:
-			return -EINVAL;
-		}
-
-		if (old_auto == intel_hdmi->color_range_auto &&
-		    old_range == intel_hdmi->limited_color_range)
-			return 0;
-
-		goto done;
-	}
-
-	if (property == connector->dev->mode_config.aspect_ratio_property) {
-		switch (val) {
-		case DRM_MODE_PICTURE_ASPECT_NONE:
-			intel_hdmi->aspect_ratio = HDMI_PICTURE_ASPECT_NONE;
-			break;
-		case DRM_MODE_PICTURE_ASPECT_4_3:
-			intel_hdmi->aspect_ratio = HDMI_PICTURE_ASPECT_4_3;
-			break;
-		case DRM_MODE_PICTURE_ASPECT_16_9:
-			intel_hdmi->aspect_ratio = HDMI_PICTURE_ASPECT_16_9;
-			break;
-		default:
-			return -EINVAL;
-		}
-		goto done;
-	}
-
-	return -EINVAL;
-
-done:
-	if (intel_dig_port->base.base.crtc)
-		intel_crtc_restore_mode(intel_dig_port->base.base.crtc);
-
-	return 0;
-}
-
 static void intel_hdmi_pre_enable(struct intel_encoder *encoder,
 				  struct intel_crtc_state *pipe_config,
 				  struct drm_connector_state *conn_state)
@@ -1798,18 +1707,20 @@ static const struct drm_connector_funcs intel_hdmi_connector_funcs = {
 	.detect = intel_hdmi_detect,
 	.force = intel_hdmi_force,
 	.fill_modes = drm_helper_probe_single_connector_modes,
-	.set_property = intel_hdmi_set_property,
-	.atomic_get_property = intel_connector_atomic_get_property,
+	.set_property = drm_atomic_helper_connector_set_property,
+	.atomic_get_property = intel_digital_connector_atomic_get_property,
+	.atomic_set_property = intel_digital_connector_atomic_set_property,
 	.late_register = intel_connector_register,
 	.early_unregister = intel_connector_unregister,
 	.destroy = intel_hdmi_destroy,
 	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
-	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+	.atomic_duplicate_state = intel_digital_connector_duplicate_state,
 };
 
 static const struct drm_connector_helper_funcs intel_hdmi_connector_helper_funcs = {
 	.get_modes = intel_hdmi_get_modes,
 	.mode_valid = intel_hdmi_mode_valid,
+	.atomic_check = intel_digital_connector_atomic_check,
 };
 
 static const struct drm_encoder_funcs intel_hdmi_enc_funcs = {
@@ -1821,9 +1732,8 @@ intel_hdmi_add_properties(struct intel_hdmi *intel_hdmi, struct drm_connector *c
 {
 	intel_attach_force_audio_property(connector);
 	intel_attach_broadcast_rgb_property(connector);
-	intel_hdmi->color_range_auto = true;
 	intel_attach_aspect_ratio_property(connector);
-	intel_hdmi->aspect_ratio = HDMI_PICTURE_ASPECT_NONE;
+	connector->state->picture_aspect_ratio = HDMI_PICTURE_ASPECT_NONE;
 }
 
 /*
@@ -1892,19 +1802,21 @@ static u8 intel_hdmi_ddc_pin(struct drm_i915_private *dev_priv,
 
 	switch (port) {
 	case PORT_B:
-		if (IS_GEN9_LP(dev_priv))
+		if (IS_GEN9_LP(dev_priv) || HAS_PCH_CNP(dev_priv))
 			ddc_pin = GMBUS_PIN_1_BXT;
 		else
 			ddc_pin = GMBUS_PIN_DPB;
 		break;
 	case PORT_C:
-		if (IS_GEN9_LP(dev_priv))
+		if (IS_GEN9_LP(dev_priv) || HAS_PCH_CNP(dev_priv))
 			ddc_pin = GMBUS_PIN_2_BXT;
 		else
 			ddc_pin = GMBUS_PIN_DPC;
 		break;
 	case PORT_D:
-		if (IS_CHERRYVIEW(dev_priv))
+		if (HAS_PCH_CNP(dev_priv))
+			ddc_pin = GMBUS_PIN_4_CNP;
+		else if (IS_CHERRYVIEW(dev_priv))
 			ddc_pin = GMBUS_PIN_DPD_CHV;
 		else
 			ddc_pin = GMBUS_PIN_DPD;
diff --git a/drivers/gpu/drm/i915/intel_huc.c b/drivers/gpu/drm/i915/intel_huc.c
index 9ee819666a4c..6145fa0d6773 100644
--- a/drivers/gpu/drm/i915/intel_huc.c
+++ b/drivers/gpu/drm/i915/intel_huc.c
@@ -52,6 +52,10 @@
 #define KBL_HUC_FW_MINOR 00
 #define KBL_BLD_NUM 1810
 
+#define GLK_HUC_FW_MAJOR 02
+#define GLK_HUC_FW_MINOR 00
+#define GLK_BLD_NUM 1748
+
 #define HUC_FW_PATH(platform, major, minor, bld_num) \
 	"i915/" __stringify(platform) "_huc_ver" __stringify(major) "_" \
 	__stringify(minor) "_" __stringify(bld_num) ".bin"
@@ -68,6 +72,9 @@ MODULE_FIRMWARE(I915_BXT_HUC_UCODE);
 	KBL_HUC_FW_MINOR, KBL_BLD_NUM)
 MODULE_FIRMWARE(I915_KBL_HUC_UCODE);
 
+#define I915_GLK_HUC_UCODE HUC_FW_PATH(glk, GLK_HUC_FW_MAJOR, \
+	GLK_HUC_FW_MINOR, GLK_BLD_NUM)
+
 /**
  * huc_ucode_xfer() - DMA's the firmware
  * @dev_priv: the drm_i915_private device
@@ -99,11 +106,6 @@ static int huc_ucode_xfer(struct drm_i915_private *dev_priv)
 
 	intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL);
 
-	/* init WOPCM */
-	I915_WRITE(GUC_WOPCM_SIZE, intel_guc_wopcm_size(dev_priv));
-	I915_WRITE(DMA_GUC_WOPCM_OFFSET, GUC_WOPCM_OFFSET_VALUE |
-			HUC_LOADING_AGENT_GUC);
-
 	/* Set the source address for the uCode */
 	offset = guc_ggtt_offset(vma) + huc_fw->header_offset;
 	I915_WRITE(DMA_ADDR_0_LOW, lower_32_bits(offset));
@@ -165,10 +167,14 @@ void intel_huc_select_fw(struct intel_huc *huc)
 		huc->fw.path = I915_BXT_HUC_UCODE;
 		huc->fw.major_ver_wanted = BXT_HUC_FW_MAJOR;
 		huc->fw.minor_ver_wanted = BXT_HUC_FW_MINOR;
-	} else if (IS_KABYLAKE(dev_priv)) {
+	} else if (IS_KABYLAKE(dev_priv) || IS_COFFEELAKE(dev_priv)) {
 		huc->fw.path = I915_KBL_HUC_UCODE;
 		huc->fw.major_ver_wanted = KBL_HUC_FW_MAJOR;
 		huc->fw.minor_ver_wanted = KBL_HUC_FW_MINOR;
+	} else if (IS_GEMINILAKE(dev_priv)) {
+		huc->fw.path = I915_GLK_HUC_UCODE;
+		huc->fw.major_ver_wanted = GLK_HUC_FW_MAJOR;
+		huc->fw.minor_ver_wanted = GLK_HUC_FW_MINOR;
 	} else {
 		DRM_ERROR("No HuC firmware known for platform with HuC!\n");
 		return;
@@ -186,68 +192,36 @@ void intel_huc_select_fw(struct intel_huc *huc)
  * earlier call to intel_huc_init(), so here we need only check that
  * is succeeded, and then transfer the image to the h/w.
  *
- * Return:	non-zero code on error
  */
-int intel_huc_init_hw(struct intel_huc *huc)
+void intel_huc_init_hw(struct intel_huc *huc)
 {
 	struct drm_i915_private *dev_priv = huc_to_i915(huc);
 	int err;
 
-	if (huc->fw.fetch_status == INTEL_UC_FIRMWARE_NONE)
-		return 0;
-
 	DRM_DEBUG_DRIVER("%s fw status: fetch %s, load %s\n",
 		huc->fw.path,
 		intel_uc_fw_status_repr(huc->fw.fetch_status),
 		intel_uc_fw_status_repr(huc->fw.load_status));
 
-	if (huc->fw.fetch_status == INTEL_UC_FIRMWARE_SUCCESS &&
-	    huc->fw.load_status == INTEL_UC_FIRMWARE_FAIL)
-		return -ENOEXEC;
+	if (huc->fw.fetch_status != INTEL_UC_FIRMWARE_SUCCESS)
+		return;
 
 	huc->fw.load_status = INTEL_UC_FIRMWARE_PENDING;
 
-	switch (huc->fw.fetch_status) {
-	case INTEL_UC_FIRMWARE_FAIL:
-		/* something went wrong :( */
-		err = -EIO;
-		goto fail;
-
-	case INTEL_UC_FIRMWARE_NONE:
-	case INTEL_UC_FIRMWARE_PENDING:
-	default:
-		/* "can't happen" */
-		WARN_ONCE(1, "HuC fw %s invalid fetch_status %s [%d]\n",
-			huc->fw.path,
-			intel_uc_fw_status_repr(huc->fw.fetch_status),
-			huc->fw.fetch_status);
-		err = -ENXIO;
-		goto fail;
-
-	case INTEL_UC_FIRMWARE_SUCCESS:
-		break;
-	}
-
 	err = huc_ucode_xfer(dev_priv);
-	if (err)
-		goto fail;
 
-	huc->fw.load_status = INTEL_UC_FIRMWARE_SUCCESS;
+	huc->fw.load_status = err ?
+		INTEL_UC_FIRMWARE_FAIL : INTEL_UC_FIRMWARE_SUCCESS;
 
 	DRM_DEBUG_DRIVER("%s fw status: fetch %s, load %s\n",
 		huc->fw.path,
 		intel_uc_fw_status_repr(huc->fw.fetch_status),
 		intel_uc_fw_status_repr(huc->fw.load_status));
 
-	return 0;
-
-fail:
-	if (huc->fw.load_status == INTEL_UC_FIRMWARE_PENDING)
-		huc->fw.load_status = INTEL_UC_FIRMWARE_FAIL;
-
-	DRM_ERROR("Failed to complete HuC uCode load with ret %d\n", err);
+	if (huc->fw.load_status != INTEL_UC_FIRMWARE_SUCCESS)
+		DRM_ERROR("Failed to complete HuC uCode load with ret %d\n", err);
 
-	return err;
+	return;
 }
 
 /**
diff --git a/drivers/gpu/drm/i915/intel_i2c.c b/drivers/gpu/drm/i915/intel_i2c.c
index b6401e8f1bd6..3c9e00d4ba5a 100644
--- a/drivers/gpu/drm/i915/intel_i2c.c
+++ b/drivers/gpu/drm/i915/intel_i2c.c
@@ -68,11 +68,20 @@ static const struct gmbus_pin gmbus_pins_bxt[] = {
 	[GMBUS_PIN_3_BXT] = { "misc", GPIOD },
 };
 
+static const struct gmbus_pin gmbus_pins_cnp[] = {
+	[GMBUS_PIN_1_BXT] = { "dpb", GPIOB },
+	[GMBUS_PIN_2_BXT] = { "dpc", GPIOC },
+	[GMBUS_PIN_3_BXT] = { "misc", GPIOD },
+	[GMBUS_PIN_4_CNP] = { "dpd", GPIOE },
+};
+
 /* pin is expected to be valid */
 static const struct gmbus_pin *get_gmbus_pin(struct drm_i915_private *dev_priv,
 					     unsigned int pin)
 {
-	if (IS_GEN9_LP(dev_priv))
+	if (HAS_PCH_CNP(dev_priv))
+		return &gmbus_pins_cnp[pin];
+	else if (IS_GEN9_LP(dev_priv))
 		return &gmbus_pins_bxt[pin];
 	else if (IS_GEN9_BC(dev_priv))
 		return &gmbus_pins_skl[pin];
@@ -87,7 +96,9 @@ bool intel_gmbus_is_valid_pin(struct drm_i915_private *dev_priv,
 {
 	unsigned int size;
 
-	if (IS_GEN9_LP(dev_priv))
+	if (HAS_PCH_CNP(dev_priv))
+		size = ARRAY_SIZE(gmbus_pins_cnp);
+	else if (IS_GEN9_LP(dev_priv))
 		size = ARRAY_SIZE(gmbus_pins_bxt);
 	else if (IS_GEN9_BC(dev_priv))
 		size = ARRAY_SIZE(gmbus_pins_skl);
diff --git a/drivers/gpu/drm/i915/intel_lpe_audio.c b/drivers/gpu/drm/i915/intel_lpe_audio.c
index 292fedf30b00..3bf65288ffff 100644
--- a/drivers/gpu/drm/i915/intel_lpe_audio.c
+++ b/drivers/gpu/drm/i915/intel_lpe_audio.c
@@ -111,6 +111,11 @@ lpe_audio_platdev_create(struct drm_i915_private *dev_priv)
 	pinfo.size_data = sizeof(*pdata);
 	pinfo.dma_mask = DMA_BIT_MASK(32);
 
+	pdata->num_pipes = INTEL_INFO(dev_priv)->num_pipes;
+	pdata->num_ports = IS_CHERRYVIEW(dev_priv) ? 3 : 2; /* B,C,D or B,C */
+	pdata->port[0].pipe = -1;
+	pdata->port[1].pipe = -1;
+	pdata->port[2].pipe = -1;
 	spin_lock_init(&pdata->lpe_audio_slock);
 
 	platdev = platform_device_register_full(&pinfo);
@@ -306,53 +311,47 @@ void intel_lpe_audio_teardown(struct drm_i915_private *dev_priv)
  * intel_lpe_audio_notify() - notify lpe audio event
  * audio driver and i915
  * @dev_priv: the i915 drm device private data
+ * @pipe: pipe
+ * @port: port
  * @eld : ELD data
- * @pipe: pipe id
- * @port: port id
- * @tmds_clk_speed: tmds clock frequency in Hz
+ * @ls_clock: Link symbol clock in kHz
+ * @dp_output: Driving a DP output?
  *
  * Notify lpe audio driver of eld change.
  */
 void intel_lpe_audio_notify(struct drm_i915_private *dev_priv,
-			    void *eld, int port, int pipe, int tmds_clk_speed,
-			    bool dp_output, int link_rate)
+			    enum pipe pipe, enum port port,
+			    const void *eld, int ls_clock, bool dp_output)
 {
-	unsigned long irq_flags;
-	struct intel_hdmi_lpe_audio_pdata *pdata = NULL;
+	unsigned long irqflags;
+	struct intel_hdmi_lpe_audio_pdata *pdata;
+	struct intel_hdmi_lpe_audio_port_pdata *ppdata;
 	u32 audio_enable;
 
 	if (!HAS_LPE_AUDIO(dev_priv))
 		return;
 
-	pdata = dev_get_platdata(
-		&(dev_priv->lpe_audio.platdev->dev));
+	pdata = dev_get_platdata(&dev_priv->lpe_audio.platdev->dev);
+	ppdata = &pdata->port[port - PORT_B];
 
-	spin_lock_irqsave(&pdata->lpe_audio_slock, irq_flags);
+	spin_lock_irqsave(&pdata->lpe_audio_slock, irqflags);
 
 	audio_enable = I915_READ(VLV_AUD_PORT_EN_DBG(port));
 
 	if (eld != NULL) {
-		memcpy(pdata->eld.eld_data, eld,
-			HDMI_MAX_ELD_BYTES);
-		pdata->eld.port_id = port;
-		pdata->eld.pipe_id = pipe;
-		pdata->hdmi_connected = true;
-
-		pdata->dp_output = dp_output;
-		if (tmds_clk_speed)
-			pdata->tmds_clock_speed = tmds_clk_speed;
-		if (link_rate)
-			pdata->link_rate = link_rate;
+		memcpy(ppdata->eld, eld, HDMI_MAX_ELD_BYTES);
+		ppdata->pipe = pipe;
+		ppdata->ls_clock = ls_clock;
+		ppdata->dp_output = dp_output;
 
 		/* Unmute the amp for both DP and HDMI */
 		I915_WRITE(VLV_AUD_PORT_EN_DBG(port),
 			   audio_enable & ~VLV_AMP_MUTE);
-
 	} else {
-		memset(pdata->eld.eld_data, 0,
-			HDMI_MAX_ELD_BYTES);
-		pdata->hdmi_connected = false;
-		pdata->dp_output = false;
+		memset(ppdata->eld, 0, HDMI_MAX_ELD_BYTES);
+		ppdata->pipe = -1;
+		ppdata->ls_clock = 0;
+		ppdata->dp_output = false;
 
 		/* Mute the amp for both DP and HDMI */
 		I915_WRITE(VLV_AUD_PORT_EN_DBG(port),
@@ -360,10 +359,7 @@ void intel_lpe_audio_notify(struct drm_i915_private *dev_priv,
 	}
 
 	if (pdata->notify_audio_lpe)
-		pdata->notify_audio_lpe(dev_priv->lpe_audio.platdev);
-	else
-		pdata->notify_pending = true;
+		pdata->notify_audio_lpe(dev_priv->lpe_audio.platdev, port - PORT_B);
 
-	spin_unlock_irqrestore(&pdata->lpe_audio_slock,
-			irq_flags);
+	spin_unlock_irqrestore(&pdata->lpe_audio_slock, irqflags);
 }
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index dac4e003c1f3..d49dbaa931b5 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -138,10 +138,6 @@
 #include "i915_drv.h"
 #include "intel_mocs.h"
 
-#define GEN9_LR_CONTEXT_RENDER_SIZE (22 * PAGE_SIZE)
-#define GEN8_LR_CONTEXT_RENDER_SIZE (20 * PAGE_SIZE)
-#define GEN8_LR_CONTEXT_OTHER_SIZE (2 * PAGE_SIZE)
-
 #define RING_EXECLIST_QFULL		(1 << 0x2)
 #define RING_EXECLIST1_VALID		(1 << 0x3)
 #define RING_EXECLIST0_VALID		(1 << 0x4)
@@ -208,6 +204,7 @@
 
 #define GEN8_CTX_RCS_INDIRECT_CTX_OFFSET_DEFAULT	0x17
 #define GEN9_CTX_RCS_INDIRECT_CTX_OFFSET_DEFAULT	0x26
+#define GEN10_CTX_RCS_INDIRECT_CTX_OFFSET_DEFAULT	0x19
 
 /* Typical size of the average request (2 pipecontrols and a MI_BB) */
 #define EXECLISTS_REQUEST_SIZE 64 /* bytes */
@@ -326,8 +323,7 @@ static u64 execlists_update_context(struct drm_i915_gem_request *rq)
 		rq->ctx->ppgtt ?: rq->i915->mm.aliasing_ppgtt;
 	u32 *reg_state = ce->lrc_reg_state;
 
-	assert_ring_tail_valid(rq->ring, rq->tail);
-	reg_state[CTX_RING_TAIL+1] = rq->tail;
+	reg_state[CTX_RING_TAIL+1] = intel_ring_set_tail(rq->ring, rq->tail);
 
 	/* True 32b PPGTT with dynamic page allocation: update PDP
 	 * registers and point the unallocated PDPs to scratch page.
@@ -342,39 +338,32 @@ static u64 execlists_update_context(struct drm_i915_gem_request *rq)
 
 static void execlists_submit_ports(struct intel_engine_cs *engine)
 {
-	struct drm_i915_private *dev_priv = engine->i915;
 	struct execlist_port *port = engine->execlist_port;
 	u32 __iomem *elsp =
-		dev_priv->regs + i915_mmio_reg_offset(RING_ELSP(engine));
-	u64 desc[2];
-
-	GEM_BUG_ON(port[0].count > 1);
-	if (!port[0].count)
-		execlists_context_status_change(port[0].request,
-						INTEL_CONTEXT_SCHEDULE_IN);
-	desc[0] = execlists_update_context(port[0].request);
-	GEM_DEBUG_EXEC(port[0].context_id = upper_32_bits(desc[0]));
-	port[0].count++;
-
-	if (port[1].request) {
-		GEM_BUG_ON(port[1].count);
-		execlists_context_status_change(port[1].request,
-						INTEL_CONTEXT_SCHEDULE_IN);
-		desc[1] = execlists_update_context(port[1].request);
-		GEM_DEBUG_EXEC(port[1].context_id = upper_32_bits(desc[1]));
-		port[1].count = 1;
-	} else {
-		desc[1] = 0;
-	}
-	GEM_BUG_ON(desc[0] == desc[1]);
-
-	/* You must always write both descriptors in the order below. */
-	writel(upper_32_bits(desc[1]), elsp);
-	writel(lower_32_bits(desc[1]), elsp);
+		engine->i915->regs + i915_mmio_reg_offset(RING_ELSP(engine));
+	unsigned int n;
+
+	for (n = ARRAY_SIZE(engine->execlist_port); n--; ) {
+		struct drm_i915_gem_request *rq;
+		unsigned int count;
+		u64 desc;
+
+		rq = port_unpack(&port[n], &count);
+		if (rq) {
+			GEM_BUG_ON(count > !n);
+			if (!count++)
+				execlists_context_status_change(rq, INTEL_CONTEXT_SCHEDULE_IN);
+			port_set(&port[n], port_pack(rq, count));
+			desc = execlists_update_context(rq);
+			GEM_DEBUG_EXEC(port[n].context_id = upper_32_bits(desc));
+		} else {
+			GEM_BUG_ON(!n);
+			desc = 0;
+		}
 
-	writel(upper_32_bits(desc[0]), elsp);
-	/* The context is automatically loaded after the following */
-	writel(lower_32_bits(desc[0]), elsp);
+		writel(upper_32_bits(desc), elsp);
+		writel(lower_32_bits(desc), elsp);
+	}
 }
 
 static bool ctx_single_port_submission(const struct i915_gem_context *ctx)
@@ -395,6 +384,17 @@ static bool can_merge_ctx(const struct i915_gem_context *prev,
 	return true;
 }
 
+static void port_assign(struct execlist_port *port,
+			struct drm_i915_gem_request *rq)
+{
+	GEM_BUG_ON(rq == port_request(port));
+
+	if (port_isset(port))
+		i915_gem_request_put(port_request(port));
+
+	port_set(port, port_pack(i915_gem_request_get(rq), port_count(port)));
+}
+
 static void execlists_dequeue(struct intel_engine_cs *engine)
 {
 	struct drm_i915_gem_request *last;
@@ -402,7 +402,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 	struct rb_node *rb;
 	bool submit = false;
 
-	last = port->request;
+	last = port_request(port);
 	if (last)
 		/* WaIdleLiteRestore:bdw,skl
 		 * Apply the wa NOOPs to prevent ring:HEAD == req:TAIL
@@ -412,7 +412,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 		 */
 		last->tail = last->wa_tail;
 
-	GEM_BUG_ON(port[1].request);
+	GEM_BUG_ON(port_isset(&port[1]));
 
 	/* Hardware submission is through 2 ports. Conceptually each port
 	 * has a (RING_START, RING_HEAD, RING_TAIL) tuple. RING_START is
@@ -437,72 +437,86 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 
 	spin_lock_irq(&engine->timeline->lock);
 	rb = engine->execlist_first;
+	GEM_BUG_ON(rb_first(&engine->execlist_queue) != rb);
 	while (rb) {
-		struct drm_i915_gem_request *cursor =
-			rb_entry(rb, typeof(*cursor), priotree.node);
-
-		/* Can we combine this request with the current port? It has to
-		 * be the same context/ringbuffer and not have any exceptions
-		 * (e.g. GVT saying never to combine contexts).
-		 *
-		 * If we can combine the requests, we can execute both by
-		 * updating the RING_TAIL to point to the end of the second
-		 * request, and so we never need to tell the hardware about
-		 * the first.
-		 */
-		if (last && !can_merge_ctx(cursor->ctx, last->ctx)) {
-			/* If we are on the second port and cannot combine
-			 * this request with the last, then we are done.
-			 */
-			if (port != engine->execlist_port)
-				break;
-
-			/* If GVT overrides us we only ever submit port[0],
-			 * leaving port[1] empty. Note that we also have
-			 * to be careful that we don't queue the same
-			 * context (even though a different request) to
-			 * the second port.
+		struct i915_priolist *p = rb_entry(rb, typeof(*p), node);
+		struct drm_i915_gem_request *rq, *rn;
+
+		list_for_each_entry_safe(rq, rn, &p->requests, priotree.link) {
+			/*
+			 * Can we combine this request with the current port?
+			 * It has to be the same context/ringbuffer and not
+			 * have any exceptions (e.g. GVT saying never to
+			 * combine contexts).
+			 *
+			 * If we can combine the requests, we can execute both
+			 * by updating the RING_TAIL to point to the end of the
+			 * second request, and so we never need to tell the
+			 * hardware about the first.
 			 */
-			if (ctx_single_port_submission(last->ctx) ||
-			    ctx_single_port_submission(cursor->ctx))
-				break;
+			if (last && !can_merge_ctx(rq->ctx, last->ctx)) {
+				/*
+				 * If we are on the second port and cannot
+				 * combine this request with the last, then we
+				 * are done.
+				 */
+				if (port != engine->execlist_port) {
+					__list_del_many(&p->requests,
+							&rq->priotree.link);
+					goto done;
+				}
+
+				/*
+				 * If GVT overrides us we only ever submit
+				 * port[0], leaving port[1] empty. Note that we
+				 * also have to be careful that we don't queue
+				 * the same context (even though a different
+				 * request) to the second port.
+				 */
+				if (ctx_single_port_submission(last->ctx) ||
+				    ctx_single_port_submission(rq->ctx)) {
+					__list_del_many(&p->requests,
+							&rq->priotree.link);
+					goto done;
+				}
+
+				GEM_BUG_ON(last->ctx == rq->ctx);
+
+				if (submit)
+					port_assign(port, last);
+				port++;
+			}
 
-			GEM_BUG_ON(last->ctx == cursor->ctx);
+			INIT_LIST_HEAD(&rq->priotree.link);
+			rq->priotree.priority = INT_MAX;
 
-			i915_gem_request_assign(&port->request, last);
-			port++;
+			__i915_gem_request_submit(rq);
+			trace_i915_gem_request_in(rq, port_index(port, engine));
+			last = rq;
+			submit = true;
 		}
 
 		rb = rb_next(rb);
-		rb_erase(&cursor->priotree.node, &engine->execlist_queue);
-		RB_CLEAR_NODE(&cursor->priotree.node);
-		cursor->priotree.priority = INT_MAX;
-
-		__i915_gem_request_submit(cursor);
-		trace_i915_gem_request_in(cursor, port - engine->execlist_port);
-		last = cursor;
-		submit = true;
-	}
-	if (submit) {
-		i915_gem_request_assign(&port->request, last);
-		engine->execlist_first = rb;
+		rb_erase(&p->node, &engine->execlist_queue);
+		INIT_LIST_HEAD(&p->requests);
+		if (p->priority != I915_PRIORITY_NORMAL)
+			kmem_cache_free(engine->i915->priorities, p);
 	}
+done:
+	engine->execlist_first = rb;
+	if (submit)
+		port_assign(port, last);
 	spin_unlock_irq(&engine->timeline->lock);
 
 	if (submit)
 		execlists_submit_ports(engine);
 }
 
-static bool execlists_elsp_idle(struct intel_engine_cs *engine)
-{
-	return !engine->execlist_port[0].request;
-}
-
 static bool execlists_elsp_ready(const struct intel_engine_cs *engine)
 {
 	const struct execlist_port *port = engine->execlist_port;
 
-	return port[0].count + port[1].count < 2;
+	return port_count(&port[0]) + port_count(&port[1]) < 2;
 }
 
 /*
@@ -515,6 +529,15 @@ static void intel_lrc_irq_handler(unsigned long data)
 	struct execlist_port *port = engine->execlist_port;
 	struct drm_i915_private *dev_priv = engine->i915;
 
+	/* We can skip acquiring intel_runtime_pm_get() here as it was taken
+	 * on our behalf by the request (see i915_gem_mark_busy()) and it will
+	 * not be relinquished until the device is idle (see
+	 * i915_gem_idle_work_handler()). As a precaution, we make sure
+	 * that all ELSP are drained i.e. we have processed the CSB,
+	 * before allowing ourselves to idle and calling intel_runtime_pm_put().
+	 */
+	GEM_BUG_ON(!dev_priv->gt.awake);
+
 	intel_uncore_forcewake_get(dev_priv, engine->fw_domains);
 
 	/* Prefer doing test_and_clear_bit() as a two stage operation to avoid
@@ -543,7 +566,9 @@ static void intel_lrc_irq_handler(unsigned long data)
 		tail = GEN8_CSB_WRITE_PTR(head);
 		head = GEN8_CSB_READ_PTR(head);
 		while (head != tail) {
+			struct drm_i915_gem_request *rq;
 			unsigned int status;
+			unsigned int count;
 
 			if (++head == GEN8_CSB_ENTRIES)
 				head = 0;
@@ -571,22 +596,26 @@ static void intel_lrc_irq_handler(unsigned long data)
 
 			/* Check the context/desc id for this event matches */
 			GEM_DEBUG_BUG_ON(readl(buf + 2 * head + 1) !=
-					 port[0].context_id);
+					 port->context_id);
 
-			GEM_BUG_ON(port[0].count == 0);
-			if (--port[0].count == 0) {
+			rq = port_unpack(port, &count);
+			GEM_BUG_ON(count == 0);
+			if (--count == 0) {
 				GEM_BUG_ON(status & GEN8_CTX_STATUS_PREEMPTED);
-				GEM_BUG_ON(!i915_gem_request_completed(port[0].request));
-				execlists_context_status_change(port[0].request,
-								INTEL_CONTEXT_SCHEDULE_OUT);
+				GEM_BUG_ON(!i915_gem_request_completed(rq));
+				execlists_context_status_change(rq, INTEL_CONTEXT_SCHEDULE_OUT);
+
+				trace_i915_gem_request_out(rq);
+				i915_gem_request_put(rq);
 
-				trace_i915_gem_request_out(port[0].request);
-				i915_gem_request_put(port[0].request);
 				port[0] = port[1];
 				memset(&port[1], 0, sizeof(port[1]));
+			} else {
+				port_set(port, port_pack(rq, count));
 			}
 
-			GEM_BUG_ON(port[0].count == 0 &&
+			/* After the final element, the hw should be idle */
+			GEM_BUG_ON(port_count(port) == 0 &&
 				   !(status & GEN8_CTX_STATUS_ACTIVE_IDLE));
 		}
 
@@ -600,28 +629,66 @@ static void intel_lrc_irq_handler(unsigned long data)
 	intel_uncore_forcewake_put(dev_priv, engine->fw_domains);
 }
 
-static bool insert_request(struct i915_priotree *pt, struct rb_root *root)
+static bool
+insert_request(struct intel_engine_cs *engine,
+	       struct i915_priotree *pt,
+	       int prio)
 {
-	struct rb_node **p, *rb;
+	struct i915_priolist *p;
+	struct rb_node **parent, *rb;
 	bool first = true;
 
+	if (unlikely(engine->no_priolist))
+		prio = I915_PRIORITY_NORMAL;
+
+find_priolist:
 	/* most positive priority is scheduled first, equal priorities fifo */
 	rb = NULL;
-	p = &root->rb_node;
-	while (*p) {
-		struct i915_priotree *pos;
-
-		rb = *p;
-		pos = rb_entry(rb, typeof(*pos), node);
-		if (pt->priority > pos->priority) {
-			p = &rb->rb_left;
-		} else {
-			p = &rb->rb_right;
+	parent = &engine->execlist_queue.rb_node;
+	while (*parent) {
+		rb = *parent;
+		p = rb_entry(rb, typeof(*p), node);
+		if (prio > p->priority) {
+			parent = &rb->rb_left;
+		} else if (prio < p->priority) {
+			parent = &rb->rb_right;
 			first = false;
+		} else {
+			list_add_tail(&pt->link, &p->requests);
+			return false;
 		}
 	}
-	rb_link_node(&pt->node, rb, p);
-	rb_insert_color(&pt->node, root);
+
+	if (prio == I915_PRIORITY_NORMAL) {
+		p = &engine->default_priolist;
+	} else {
+		p = kmem_cache_alloc(engine->i915->priorities, GFP_ATOMIC);
+		/* Convert an allocation failure to a priority bump */
+		if (unlikely(!p)) {
+			prio = I915_PRIORITY_NORMAL; /* recurses just once */
+
+			/* To maintain ordering with all rendering, after an
+			 * allocation failure we have to disable all scheduling.
+			 * Requests will then be executed in fifo, and schedule
+			 * will ensure that dependencies are emitted in fifo.
+			 * There will be still some reordering with existing
+			 * requests, so if userspace lied about their
+			 * dependencies that reordering may be visible.
+			 */
+			engine->no_priolist = true;
+			goto find_priolist;
+		}
+	}
+
+	p->priority = prio;
+	rb_link_node(&p->node, rb, parent);
+	rb_insert_color(&p->node, &engine->execlist_queue);
+
+	INIT_LIST_HEAD(&p->requests);
+	list_add_tail(&pt->link, &p->requests);
+
+	if (first)
+		engine->execlist_first = &p->node;
 
 	return first;
 }
@@ -634,12 +701,16 @@ static void execlists_submit_request(struct drm_i915_gem_request *request)
 	/* Will be called from irq-context when using foreign fences. */
 	spin_lock_irqsave(&engine->timeline->lock, flags);
 
-	if (insert_request(&request->priotree, &engine->execlist_queue)) {
-		engine->execlist_first = &request->priotree.node;
+	if (insert_request(engine,
+			   &request->priotree,
+			   request->priotree.priority)) {
 		if (execlists_elsp_ready(engine))
 			tasklet_hi_schedule(&engine->irq_tasklet);
 	}
 
+	GEM_BUG_ON(!engine->execlist_first);
+	GEM_BUG_ON(list_empty(&request->priotree.link));
+
 	spin_unlock_irqrestore(&engine->timeline->lock, flags);
 }
 
@@ -709,6 +780,19 @@ static void execlists_schedule(struct drm_i915_gem_request *request, int prio)
 		list_safe_reset_next(dep, p, dfs_link);
 	}
 
+	/* If we didn't need to bump any existing priorities, and we haven't
+	 * yet submitted this request (i.e. there is no potential race with
+	 * execlists_submit_request()), we can set our own priority and skip
+	 * acquiring the engine locks.
+	 */
+	if (request->priotree.priority == INT_MIN) {
+		GEM_BUG_ON(!list_empty(&request->priotree.link));
+		request->priotree.priority = prio;
+		if (stack.dfs_link.next == stack.dfs_link.prev)
+			return;
+		__list_del_entry(&stack.dfs_link);
+	}
+
 	engine = request->engine;
 	spin_lock_irq(&engine->timeline->lock);
 
@@ -724,10 +808,9 @@ static void execlists_schedule(struct drm_i915_gem_request *request, int prio)
 			continue;
 
 		pt->priority = prio;
-		if (!RB_EMPTY_NODE(&pt->node)) {
-			rb_erase(&pt->node, &engine->execlist_queue);
-			if (insert_request(pt, &engine->execlist_queue))
-				engine->execlist_first = &pt->node;
+		if (!list_empty(&pt->link)) {
+			__list_del_entry(&pt->link);
+			insert_request(engine, pt, prio);
 		}
 	}
 
@@ -736,8 +819,9 @@ static void execlists_schedule(struct drm_i915_gem_request *request, int prio)
 	/* XXX Do we need to preempt to make room for us and our deps? */
 }
 
-static int execlists_context_pin(struct intel_engine_cs *engine,
-				 struct i915_gem_context *ctx)
+static struct intel_ring *
+execlists_context_pin(struct intel_engine_cs *engine,
+		      struct i915_gem_context *ctx)
 {
 	struct intel_context *ce = &ctx->engine[engine->id];
 	unsigned int flags;
@@ -746,8 +830,8 @@ static int execlists_context_pin(struct intel_engine_cs *engine,
 
 	lockdep_assert_held(&ctx->i915->drm.struct_mutex);
 
-	if (ce->pin_count++)
-		return 0;
+	if (likely(ce->pin_count++))
+		goto out;
 	GEM_BUG_ON(!ce->pin_count); /* no overflow please! */
 
 	if (!ce->state) {
@@ -771,7 +855,7 @@ static int execlists_context_pin(struct intel_engine_cs *engine,
 		goto unpin_vma;
 	}
 
-	ret = intel_ring_pin(ce->ring, ctx->ggtt_offset_bias);
+	ret = intel_ring_pin(ce->ring, ctx->i915, ctx->ggtt_offset_bias);
 	if (ret)
 		goto unpin_map;
 
@@ -784,7 +868,8 @@ static int execlists_context_pin(struct intel_engine_cs *engine,
 	ce->state->obj->mm.dirty = true;
 
 	i915_gem_context_get(ctx);
-	return 0;
+out:
+	return ce->ring;
 
 unpin_map:
 	i915_gem_object_unpin_map(ce->state->obj);
@@ -792,7 +877,7 @@ unpin_vma:
 	__i915_vma_unpin(ce->state);
 err:
 	ce->pin_count = 0;
-	return ret;
+	return ERR_PTR(ret);
 }
 
 static void execlists_context_unpin(struct intel_engine_cs *engine,
@@ -829,9 +914,6 @@ static int execlists_request_alloc(struct drm_i915_gem_request *request)
 	 */
 	request->reserved_space += EXECLISTS_REQUEST_SIZE;
 
-	GEM_BUG_ON(!ce->ring);
-	request->ring = ce->ring;
-
 	if (i915.enable_guc_submission) {
 		/*
 		 * Check that the GuC has space for the request before
@@ -1139,14 +1221,12 @@ static int intel_init_workaround_bb(struct intel_engine_cs *engine)
 	return ret;
 }
 
-static u32 port_seqno(struct execlist_port *port)
-{
-	return port->request ? port->request->global_seqno : 0;
-}
-
 static int gen8_init_common_ring(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
+	struct execlist_port *port = engine->execlist_port;
+	unsigned int n;
+	bool submit;
 	int ret;
 
 	ret = intel_mocs_init_engine(engine);
@@ -1167,16 +1247,24 @@ static int gen8_init_common_ring(struct intel_engine_cs *engine)
 
 	/* After a GPU reset, we may have requests to replay */
 	clear_bit(ENGINE_IRQ_EXECLIST, &engine->irq_posted);
-	if (!i915.enable_guc_submission && !execlists_elsp_idle(engine)) {
-		DRM_DEBUG_DRIVER("Restarting %s from requests [0x%x, 0x%x]\n",
-				 engine->name,
-				 port_seqno(&engine->execlist_port[0]),
-				 port_seqno(&engine->execlist_port[1]));
-		engine->execlist_port[0].count = 0;
-		engine->execlist_port[1].count = 0;
-		execlists_submit_ports(engine);
+
+	submit = false;
+	for (n = 0; n < ARRAY_SIZE(engine->execlist_port); n++) {
+		if (!port_isset(&port[n]))
+			break;
+
+		DRM_DEBUG_DRIVER("Restarting %s:%d from 0x%x\n",
+				 engine->name, n,
+				 port_request(&port[n])->global_seqno);
+
+		/* Discard the current inflight count */
+		port_set(&port[n], port_request(&port[n]));
+		submit = true;
 	}
 
+	if (submit && !i915.enable_guc_submission)
+		execlists_submit_ports(engine);
+
 	return 0;
 }
 
@@ -1252,13 +1340,13 @@ static void reset_common_ring(struct intel_engine_cs *engine,
 	intel_ring_update_space(request->ring);
 
 	/* Catch up with any missed context-switch interrupts */
-	if (request->ctx != port[0].request->ctx) {
-		i915_gem_request_put(port[0].request);
+	if (request->ctx != port_request(port)->ctx) {
+		i915_gem_request_put(port_request(port));
 		port[0] = port[1];
 		memset(&port[1], 0, sizeof(port[1]));
 	}
 
-	GEM_BUG_ON(request->ctx != port[0].request->ctx);
+	GEM_BUG_ON(request->ctx != port_request(port)->ctx);
 
 	/* Reset WaIdleLiteRestore:bdw,skl as well */
 	request->tail =
@@ -1774,6 +1862,10 @@ static u32 intel_lr_indirect_ctx_offset(struct intel_engine_cs *engine)
 	default:
 		MISSING_CASE(INTEL_GEN(engine->i915));
 		/* fall through */
+	case 10:
+		indirect_ctx_offset =
+			GEN10_CTX_RCS_INDIRECT_CTX_OFFSET_DEFAULT;
+		break;
 	case 9:
 		indirect_ctx_offset =
 			GEN9_CTX_RCS_INDIRECT_CTX_OFFSET_DEFAULT;
@@ -1907,44 +1999,6 @@ populate_lr_context(struct i915_gem_context *ctx,
 	return 0;
 }
 
-/**
- * intel_lr_context_size() - return the size of the context for an engine
- * @engine: which engine to find the context size for
- *
- * Each engine may require a different amount of space for a context image,
- * so when allocating (or copying) an image, this function can be used to
- * find the right size for the specific engine.
- *
- * Return: size (in bytes) of an engine-specific context image
- *
- * Note: this size includes the HWSP, which is part of the context image
- * in LRC mode, but does not include the "shared data page" used with
- * GuC submission. The caller should account for this if using the GuC.
- */
-uint32_t intel_lr_context_size(struct intel_engine_cs *engine)
-{
-	int ret = 0;
-
-	WARN_ON(INTEL_GEN(engine->i915) < 8);
-
-	switch (engine->id) {
-	case RCS:
-		if (INTEL_GEN(engine->i915) >= 9)
-			ret = GEN9_LR_CONTEXT_RENDER_SIZE;
-		else
-			ret = GEN8_LR_CONTEXT_RENDER_SIZE;
-		break;
-	case VCS:
-	case BCS:
-	case VECS:
-	case VCS2:
-		ret = GEN8_LR_CONTEXT_OTHER_SIZE;
-		break;
-	}
-
-	return ret;
-}
-
 static int execlists_context_deferred_alloc(struct i915_gem_context *ctx,
 					    struct intel_engine_cs *engine)
 {
@@ -1957,8 +2011,7 @@ static int execlists_context_deferred_alloc(struct i915_gem_context *ctx,
 
 	WARN_ON(ce->state);
 
-	context_size = round_up(intel_lr_context_size(engine),
-				I915_GTT_PAGE_SIZE);
+	context_size = round_up(engine->context_size, I915_GTT_PAGE_SIZE);
 
 	/* One extra page as the sharing data between driver and GuC */
 	context_size += PAGE_SIZE * LRC_PPHWSP_PN;
@@ -2036,8 +2089,7 @@ void intel_lr_context_resume(struct drm_i915_private *dev_priv)
 			ce->state->obj->mm.dirty = true;
 			i915_gem_object_unpin_map(ce->state->obj);
 
-			ce->ring->head = ce->ring->tail = 0;
-			intel_ring_update_space(ce->ring);
+			intel_ring_reset(ce->ring, 0);
 		}
 	}
 }
diff --git a/drivers/gpu/drm/i915/intel_lrc.h b/drivers/gpu/drm/i915/intel_lrc.h
index e8015e7bf4e9..52b3a1fd4059 100644
--- a/drivers/gpu/drm/i915/intel_lrc.h
+++ b/drivers/gpu/drm/i915/intel_lrc.h
@@ -78,8 +78,6 @@ int logical_xcs_ring_init(struct intel_engine_cs *engine);
 struct drm_i915_private;
 struct i915_gem_context;
 
-uint32_t intel_lr_context_size(struct intel_engine_cs *engine);
-
 void intel_lr_context_resume(struct drm_i915_private *dev_priv);
 uint64_t intel_lr_context_descriptor(struct i915_gem_context *ctx,
 				     struct intel_engine_cs *engine);
diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c
index 8b942ef2b3ec..6fe5d7c3bc23 100644
--- a/drivers/gpu/drm/i915/intel_lvds.c
+++ b/drivers/gpu/drm/i915/intel_lvds.c
@@ -311,8 +311,6 @@ static void intel_enable_lvds(struct intel_encoder *encoder,
 {
 	struct drm_device *dev = encoder->base.dev;
 	struct intel_lvds_encoder *lvds_encoder = to_lvds_encoder(&encoder->base);
-	struct intel_connector *intel_connector =
-		&lvds_encoder->attached_connector->base;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 
 	I915_WRITE(lvds_encoder->reg, I915_READ(lvds_encoder->reg) | LVDS_PORT_EN);
@@ -322,7 +320,7 @@ static void intel_enable_lvds(struct intel_encoder *encoder,
 	if (intel_wait_for_register(dev_priv, PP_STATUS(0), PP_ON, PP_ON, 1000))
 		DRM_ERROR("timed out waiting for panel to power on\n");
 
-	intel_panel_enable_backlight(intel_connector);
+	intel_panel_enable_backlight(pipe_config, conn_state);
 }
 
 static void intel_disable_lvds(struct intel_encoder *encoder,
@@ -345,11 +343,7 @@ static void gmch_disable_lvds(struct intel_encoder *encoder,
 			      struct drm_connector_state *old_conn_state)
 
 {
-	struct intel_lvds_encoder *lvds_encoder = to_lvds_encoder(&encoder->base);
-	struct intel_connector *intel_connector =
-		&lvds_encoder->attached_connector->base;
-
-	intel_panel_disable_backlight(intel_connector);
+	intel_panel_disable_backlight(old_conn_state);
 
 	intel_disable_lvds(encoder, old_crtc_state, old_conn_state);
 }
@@ -358,11 +352,7 @@ static void pch_disable_lvds(struct intel_encoder *encoder,
 			     struct intel_crtc_state *old_crtc_state,
 			     struct drm_connector_state *old_conn_state)
 {
-	struct intel_lvds_encoder *lvds_encoder = to_lvds_encoder(&encoder->base);
-	struct intel_connector *intel_connector =
-		&lvds_encoder->attached_connector->base;
-
-	intel_panel_disable_backlight(intel_connector);
+	intel_panel_disable_backlight(old_conn_state);
 }
 
 static void pch_post_disable_lvds(struct intel_encoder *encoder,
@@ -433,10 +423,10 @@ static bool intel_lvds_compute_config(struct intel_encoder *intel_encoder,
 		pipe_config->has_pch_encoder = true;
 
 		intel_pch_panel_fitting(intel_crtc, pipe_config,
-					intel_connector->panel.fitting_mode);
+					conn_state->scaling_mode);
 	} else {
 		intel_gmch_panel_fitting(intel_crtc, pipe_config,
-					 intel_connector->panel.fitting_mode);
+					 conn_state->scaling_mode);
 
 	}
 
@@ -598,56 +588,24 @@ static void intel_lvds_destroy(struct drm_connector *connector)
 	kfree(connector);
 }
 
-static int intel_lvds_set_property(struct drm_connector *connector,
-				   struct drm_property *property,
-				   uint64_t value)
-{
-	struct intel_connector *intel_connector = to_intel_connector(connector);
-	struct drm_device *dev = connector->dev;
-
-	if (property == dev->mode_config.scaling_mode_property) {
-		struct drm_crtc *crtc;
-
-		if (value == DRM_MODE_SCALE_NONE) {
-			DRM_DEBUG_KMS("no scaling not supported\n");
-			return -EINVAL;
-		}
-
-		if (intel_connector->panel.fitting_mode == value) {
-			/* the LVDS scaling property is not changed */
-			return 0;
-		}
-		intel_connector->panel.fitting_mode = value;
-
-		crtc = intel_attached_encoder(connector)->base.crtc;
-		if (crtc && crtc->state->enable) {
-			/*
-			 * If the CRTC is enabled, the display will be changed
-			 * according to the new panel fitting mode.
-			 */
-			intel_crtc_restore_mode(crtc);
-		}
-	}
-
-	return 0;
-}
-
 static const struct drm_connector_helper_funcs intel_lvds_connector_helper_funcs = {
 	.get_modes = intel_lvds_get_modes,
 	.mode_valid = intel_lvds_mode_valid,
+	.atomic_check = intel_digital_connector_atomic_check,
 };
 
 static const struct drm_connector_funcs intel_lvds_connector_funcs = {
 	.dpms = drm_atomic_helper_connector_dpms,
 	.detect = intel_lvds_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
-	.set_property = intel_lvds_set_property,
-	.atomic_get_property = intel_connector_atomic_get_property,
+	.set_property = drm_atomic_helper_connector_set_property,
+	.atomic_get_property = intel_digital_connector_atomic_get_property,
+	.atomic_set_property = intel_digital_connector_atomic_set_property,
 	.late_register = intel_connector_register,
 	.early_unregister = intel_connector_unregister,
 	.destroy = intel_lvds_destroy,
 	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
-	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+	.atomic_duplicate_state = intel_digital_connector_duplicate_state,
 };
 
 static const struct drm_encoder_funcs intel_lvds_enc_funcs = {
@@ -988,6 +946,7 @@ void intel_lvds_init(struct drm_i915_private *dev_priv)
 	u32 lvds;
 	int pipe;
 	u8 pin;
+	u32 allowed_scalers;
 
 	if (!intel_lvds_supported(dev_priv))
 		return;
@@ -1083,11 +1042,11 @@ void intel_lvds_init(struct drm_i915_private *dev_priv)
 	lvds_encoder->reg = lvds_reg;
 
 	/* create the scaling mode property */
-	drm_mode_create_scaling_mode_property(dev);
-	drm_object_attach_property(&connector->base,
-				      dev->mode_config.scaling_mode_property,
-				      DRM_MODE_SCALE_ASPECT);
-	intel_connector->panel.fitting_mode = DRM_MODE_SCALE_ASPECT;
+	allowed_scalers = BIT(DRM_MODE_SCALE_ASPECT);
+	allowed_scalers |= BIT(DRM_MODE_SCALE_FULLSCREEN);
+	allowed_scalers |= BIT(DRM_MODE_SCALE_CENTER);
+	drm_connector_attach_scaling_mode_property(connector, allowed_scalers);
+	connector->state->scaling_mode = DRM_MODE_SCALE_ASPECT;
 
 	intel_lvds_pps_get_hw_state(dev_priv, &lvds_encoder->init_pps);
 	lvds_encoder->init_lvds_val = lvds;
diff --git a/drivers/gpu/drm/i915/intel_mocs.c b/drivers/gpu/drm/i915/intel_mocs.c
index 92e461c68385..f4c46b0b8f0a 100644
--- a/drivers/gpu/drm/i915/intel_mocs.c
+++ b/drivers/gpu/drm/i915/intel_mocs.c
@@ -178,7 +178,7 @@ static bool get_mocs_settings(struct drm_i915_private *dev_priv,
 {
 	bool result = false;
 
-	if (IS_GEN9_BC(dev_priv)) {
+	if (IS_GEN9_BC(dev_priv) || IS_CANNONLAKE(dev_priv)) {
 		table->size  = ARRAY_SIZE(skylake_mocs_table);
 		table->table = skylake_mocs_table;
 		result = true;
diff --git a/drivers/gpu/drm/i915/intel_opregion.c b/drivers/gpu/drm/i915/intel_opregion.c
index d44465190dc1..2bd03001cc70 100644
--- a/drivers/gpu/drm/i915/intel_opregion.c
+++ b/drivers/gpu/drm/i915/intel_opregion.c
@@ -461,7 +461,7 @@ static u32 asle_set_backlight(struct drm_i915_private *dev_priv, u32 bclp)
 	DRM_DEBUG_KMS("updating opregion backlight %d/255\n", bclp);
 	drm_connector_list_iter_begin(dev, &conn_iter);
 	for_each_intel_connector_iter(connector, &conn_iter)
-		intel_panel_set_backlight_acpi(connector, bclp, 255);
+		intel_panel_set_backlight_acpi(connector->base.state, bclp, 255);
 	drm_connector_list_iter_end(&conn_iter);
 	asle->cblv = DIV_ROUND_UP(bclp * 100, 255) | ASLE_CBLV_VALID;
 
diff --git a/drivers/gpu/drm/i915/intel_panel.c b/drivers/gpu/drm/i915/intel_panel.c
index cb50c527401f..96c2cbd81869 100644
--- a/drivers/gpu/drm/i915/intel_panel.c
+++ b/drivers/gpu/drm/i915/intel_panel.c
@@ -561,15 +561,18 @@ static u32 intel_panel_get_backlight(struct intel_connector *connector)
 	return val;
 }
 
-static void lpt_set_backlight(struct intel_connector *connector, u32 level)
+static void lpt_set_backlight(const struct drm_connector_state *conn_state, u32 level)
 {
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
+
 	u32 val = I915_READ(BLC_PWM_PCH_CTL2) & ~BACKLIGHT_DUTY_CYCLE_MASK;
 	I915_WRITE(BLC_PWM_PCH_CTL2, val | level);
 }
 
-static void pch_set_backlight(struct intel_connector *connector, u32 level)
+static void pch_set_backlight(const struct drm_connector_state *conn_state, u32 level)
 {
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	u32 tmp;
 
@@ -577,8 +580,9 @@ static void pch_set_backlight(struct intel_connector *connector, u32 level)
 	I915_WRITE(BLC_PWM_CPU_CTL, tmp | level);
 }
 
-static void i9xx_set_backlight(struct intel_connector *connector, u32 level)
+static void i9xx_set_backlight(const struct drm_connector_state *conn_state, u32 level)
 {
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_panel *panel = &connector->panel;
 	u32 tmp, mask;
@@ -604,50 +608,51 @@ static void i9xx_set_backlight(struct intel_connector *connector, u32 level)
 	I915_WRITE(BLC_PWM_CTL, tmp | level);
 }
 
-static void vlv_set_backlight(struct intel_connector *connector, u32 level)
+static void vlv_set_backlight(const struct drm_connector_state *conn_state, u32 level)
 {
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
-	enum pipe pipe = intel_get_pipe_from_connector(connector);
+	enum pipe pipe = to_intel_crtc(conn_state->crtc)->pipe;
 	u32 tmp;
 
-	if (WARN_ON(pipe != PIPE_A && pipe != PIPE_B))
-		return;
-
 	tmp = I915_READ(VLV_BLC_PWM_CTL(pipe)) & ~BACKLIGHT_DUTY_CYCLE_MASK;
 	I915_WRITE(VLV_BLC_PWM_CTL(pipe), tmp | level);
 }
 
-static void bxt_set_backlight(struct intel_connector *connector, u32 level)
+static void bxt_set_backlight(const struct drm_connector_state *conn_state, u32 level)
 {
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_panel *panel = &connector->panel;
 
 	I915_WRITE(BXT_BLC_PWM_DUTY(panel->backlight.controller), level);
 }
 
-static void pwm_set_backlight(struct intel_connector *connector, u32 level)
+static void pwm_set_backlight(const struct drm_connector_state *conn_state, u32 level)
 {
-	struct intel_panel *panel = &connector->panel;
+	struct intel_panel *panel = &to_intel_connector(conn_state->connector)->panel;
 	int duty_ns = DIV_ROUND_UP(level * CRC_PMIC_PWM_PERIOD_NS, 100);
 
 	pwm_config(panel->backlight.pwm, duty_ns, CRC_PMIC_PWM_PERIOD_NS);
 }
 
 static void
-intel_panel_actually_set_backlight(struct intel_connector *connector, u32 level)
+intel_panel_actually_set_backlight(const struct drm_connector_state *conn_state, u32 level)
 {
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct intel_panel *panel = &connector->panel;
 
 	DRM_DEBUG_DRIVER("set backlight PWM = %d\n", level);
 
 	level = intel_panel_compute_brightness(connector, level);
-	panel->backlight.set(connector, level);
+	panel->backlight.set(conn_state, level);
 }
 
 /* set backlight brightness to level in range [0..max], scaling wrt hw min */
-static void intel_panel_set_backlight(struct intel_connector *connector,
+static void intel_panel_set_backlight(const struct drm_connector_state *conn_state,
 				      u32 user_level, u32 user_max)
 {
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_panel *panel = &connector->panel;
 	u32 hw_level;
@@ -663,7 +668,7 @@ static void intel_panel_set_backlight(struct intel_connector *connector,
 	panel->backlight.level = hw_level;
 
 	if (panel->backlight.enabled)
-		intel_panel_actually_set_backlight(connector, hw_level);
+		intel_panel_actually_set_backlight(conn_state, hw_level);
 
 	mutex_unlock(&dev_priv->backlight_lock);
 }
@@ -671,21 +676,21 @@ static void intel_panel_set_backlight(struct intel_connector *connector,
 /* set backlight brightness to level in range [0..max], assuming hw min is
  * respected.
  */
-void intel_panel_set_backlight_acpi(struct intel_connector *connector,
+void intel_panel_set_backlight_acpi(const struct drm_connector_state *conn_state,
 				    u32 user_level, u32 user_max)
 {
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_panel *panel = &connector->panel;
-	enum pipe pipe = intel_get_pipe_from_connector(connector);
 	u32 hw_level;
 
 	/*
-	 * INVALID_PIPE may occur during driver init because
+	 * Lack of crtc may occur during driver init because
 	 * connection_mutex isn't held across the entire backlight
 	 * setup + modeset readout, and the BIOS can issue the
 	 * requests at any time.
 	 */
-	if (!panel->backlight.present || pipe == INVALID_PIPE)
+	if (!panel->backlight.present || !conn_state->crtc)
 		return;
 
 	mutex_lock(&dev_priv->backlight_lock);
@@ -702,17 +707,18 @@ void intel_panel_set_backlight_acpi(struct intel_connector *connector,
 					 panel->backlight.device->props.max_brightness);
 
 	if (panel->backlight.enabled)
-		intel_panel_actually_set_backlight(connector, hw_level);
+		intel_panel_actually_set_backlight(conn_state, hw_level);
 
 	mutex_unlock(&dev_priv->backlight_lock);
 }
 
-static void lpt_disable_backlight(struct intel_connector *connector)
+static void lpt_disable_backlight(const struct drm_connector_state *old_conn_state)
 {
+	struct intel_connector *connector = to_intel_connector(old_conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	u32 tmp;
 
-	intel_panel_actually_set_backlight(connector, 0);
+	intel_panel_actually_set_backlight(old_conn_state, 0);
 
 	/*
 	 * Although we don't support or enable CPU PWM with LPT/SPT based
@@ -732,12 +738,13 @@ static void lpt_disable_backlight(struct intel_connector *connector)
 	I915_WRITE(BLC_PWM_PCH_CTL1, tmp & ~BLM_PCH_PWM_ENABLE);
 }
 
-static void pch_disable_backlight(struct intel_connector *connector)
+static void pch_disable_backlight(const struct drm_connector_state *old_conn_state)
 {
+	struct intel_connector *connector = to_intel_connector(old_conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	u32 tmp;
 
-	intel_panel_actually_set_backlight(connector, 0);
+	intel_panel_actually_set_backlight(old_conn_state, 0);
 
 	tmp = I915_READ(BLC_PWM_CPU_CTL2);
 	I915_WRITE(BLC_PWM_CPU_CTL2, tmp & ~BLM_PWM_ENABLE);
@@ -746,44 +753,43 @@ static void pch_disable_backlight(struct intel_connector *connector)
 	I915_WRITE(BLC_PWM_PCH_CTL1, tmp & ~BLM_PCH_PWM_ENABLE);
 }
 
-static void i9xx_disable_backlight(struct intel_connector *connector)
+static void i9xx_disable_backlight(const struct drm_connector_state *old_conn_state)
 {
-	intel_panel_actually_set_backlight(connector, 0);
+	intel_panel_actually_set_backlight(old_conn_state, 0);
 }
 
-static void i965_disable_backlight(struct intel_connector *connector)
+static void i965_disable_backlight(const struct drm_connector_state *old_conn_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
+	struct drm_i915_private *dev_priv = to_i915(old_conn_state->connector->dev);
 	u32 tmp;
 
-	intel_panel_actually_set_backlight(connector, 0);
+	intel_panel_actually_set_backlight(old_conn_state, 0);
 
 	tmp = I915_READ(BLC_PWM_CTL2);
 	I915_WRITE(BLC_PWM_CTL2, tmp & ~BLM_PWM_ENABLE);
 }
 
-static void vlv_disable_backlight(struct intel_connector *connector)
+static void vlv_disable_backlight(const struct drm_connector_state *old_conn_state)
 {
+	struct intel_connector *connector = to_intel_connector(old_conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
-	enum pipe pipe = intel_get_pipe_from_connector(connector);
+	enum pipe pipe = to_intel_crtc(old_conn_state->crtc)->pipe;
 	u32 tmp;
 
-	if (WARN_ON(pipe != PIPE_A && pipe != PIPE_B))
-		return;
-
-	intel_panel_actually_set_backlight(connector, 0);
+	intel_panel_actually_set_backlight(old_conn_state, 0);
 
 	tmp = I915_READ(VLV_BLC_PWM_CTL2(pipe));
 	I915_WRITE(VLV_BLC_PWM_CTL2(pipe), tmp & ~BLM_PWM_ENABLE);
 }
 
-static void bxt_disable_backlight(struct intel_connector *connector)
+static void bxt_disable_backlight(const struct drm_connector_state *old_conn_state)
 {
+	struct intel_connector *connector = to_intel_connector(old_conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_panel *panel = &connector->panel;
 	u32 tmp, val;
 
-	intel_panel_actually_set_backlight(connector, 0);
+	intel_panel_actually_set_backlight(old_conn_state, 0);
 
 	tmp = I915_READ(BXT_BLC_PWM_CTL(panel->backlight.controller));
 	I915_WRITE(BXT_BLC_PWM_CTL(panel->backlight.controller),
@@ -796,8 +802,23 @@ static void bxt_disable_backlight(struct intel_connector *connector)
 	}
 }
 
-static void pwm_disable_backlight(struct intel_connector *connector)
+static void cnp_disable_backlight(const struct drm_connector_state *old_conn_state)
 {
+	struct intel_connector *connector = to_intel_connector(old_conn_state->connector);
+	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
+	struct intel_panel *panel = &connector->panel;
+	u32 tmp;
+
+	intel_panel_actually_set_backlight(old_conn_state, 0);
+
+	tmp = I915_READ(BXT_BLC_PWM_CTL(panel->backlight.controller));
+	I915_WRITE(BXT_BLC_PWM_CTL(panel->backlight.controller),
+		   tmp & ~BXT_BLC_PWM_ENABLE);
+}
+
+static void pwm_disable_backlight(const struct drm_connector_state *old_conn_state)
+{
+	struct intel_connector *connector = to_intel_connector(old_conn_state->connector);
 	struct intel_panel *panel = &connector->panel;
 
 	/* Disable the backlight */
@@ -806,8 +827,9 @@ static void pwm_disable_backlight(struct intel_connector *connector)
 	pwm_disable(panel->backlight.pwm);
 }
 
-void intel_panel_disable_backlight(struct intel_connector *connector)
+void intel_panel_disable_backlight(const struct drm_connector_state *old_conn_state)
 {
+	struct intel_connector *connector = to_intel_connector(old_conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_panel *panel = &connector->panel;
 
@@ -830,13 +852,15 @@ void intel_panel_disable_backlight(struct intel_connector *connector)
 	if (panel->backlight.device)
 		panel->backlight.device->props.power = FB_BLANK_POWERDOWN;
 	panel->backlight.enabled = false;
-	panel->backlight.disable(connector);
+	panel->backlight.disable(old_conn_state);
 
 	mutex_unlock(&dev_priv->backlight_lock);
 }
 
-static void lpt_enable_backlight(struct intel_connector *connector)
+static void lpt_enable_backlight(const struct intel_crtc_state *crtc_state,
+				 const struct drm_connector_state *conn_state)
 {
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_panel *panel = &connector->panel;
 	u32 pch_ctl1, pch_ctl2, schicken;
@@ -880,16 +904,16 @@ static void lpt_enable_backlight(struct intel_connector *connector)
 	I915_WRITE(BLC_PWM_PCH_CTL1, pch_ctl1 | BLM_PCH_PWM_ENABLE);
 
 	/* This won't stick until the above enable. */
-	intel_panel_actually_set_backlight(connector, panel->backlight.level);
+	intel_panel_actually_set_backlight(conn_state, panel->backlight.level);
 }
 
-static void pch_enable_backlight(struct intel_connector *connector)
+static void pch_enable_backlight(const struct intel_crtc_state *crtc_state,
+				 const struct drm_connector_state *conn_state)
 {
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_panel *panel = &connector->panel;
-	enum pipe pipe = intel_get_pipe_from_connector(connector);
-	enum transcoder cpu_transcoder =
-		intel_pipe_to_cpu_transcoder(dev_priv, pipe);
+	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
 	u32 cpu_ctl2, pch_ctl1, pch_ctl2;
 
 	cpu_ctl2 = I915_READ(BLC_PWM_CPU_CTL2);
@@ -915,7 +939,7 @@ static void pch_enable_backlight(struct intel_connector *connector)
 	I915_WRITE(BLC_PWM_CPU_CTL2, cpu_ctl2 | BLM_PWM_ENABLE);
 
 	/* This won't stick until the above enable. */
-	intel_panel_actually_set_backlight(connector, panel->backlight.level);
+	intel_panel_actually_set_backlight(conn_state, panel->backlight.level);
 
 	pch_ctl2 = panel->backlight.max << 16;
 	I915_WRITE(BLC_PWM_PCH_CTL2, pch_ctl2);
@@ -929,8 +953,10 @@ static void pch_enable_backlight(struct intel_connector *connector)
 	I915_WRITE(BLC_PWM_PCH_CTL1, pch_ctl1 | BLM_PCH_PWM_ENABLE);
 }
 
-static void i9xx_enable_backlight(struct intel_connector *connector)
+static void i9xx_enable_backlight(const struct intel_crtc_state *crtc_state,
+				  const struct drm_connector_state *conn_state)
 {
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_panel *panel = &connector->panel;
 	u32 ctl, freq;
@@ -955,7 +981,7 @@ static void i9xx_enable_backlight(struct intel_connector *connector)
 	POSTING_READ(BLC_PWM_CTL);
 
 	/* XXX: combine this into above write? */
-	intel_panel_actually_set_backlight(connector, panel->backlight.level);
+	intel_panel_actually_set_backlight(conn_state, panel->backlight.level);
 
 	/*
 	 * Needed to enable backlight on some 855gm models. BLC_HIST_CTL is
@@ -966,11 +992,13 @@ static void i9xx_enable_backlight(struct intel_connector *connector)
 		I915_WRITE(BLC_HIST_CTL, BLM_HISTOGRAM_ENABLE);
 }
 
-static void i965_enable_backlight(struct intel_connector *connector)
+static void i965_enable_backlight(const struct intel_crtc_state *crtc_state,
+				  const struct drm_connector_state *conn_state)
 {
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_panel *panel = &connector->panel;
-	enum pipe pipe = intel_get_pipe_from_connector(connector);
+	enum pipe pipe = to_intel_crtc(conn_state->crtc)->pipe;
 	u32 ctl, ctl2, freq;
 
 	ctl2 = I915_READ(BLC_PWM_CTL2);
@@ -996,19 +1024,18 @@ static void i965_enable_backlight(struct intel_connector *connector)
 	POSTING_READ(BLC_PWM_CTL2);
 	I915_WRITE(BLC_PWM_CTL2, ctl2 | BLM_PWM_ENABLE);
 
-	intel_panel_actually_set_backlight(connector, panel->backlight.level);
+	intel_panel_actually_set_backlight(conn_state, panel->backlight.level);
 }
 
-static void vlv_enable_backlight(struct intel_connector *connector)
+static void vlv_enable_backlight(const struct intel_crtc_state *crtc_state,
+				 const struct drm_connector_state *conn_state)
 {
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_panel *panel = &connector->panel;
-	enum pipe pipe = intel_get_pipe_from_connector(connector);
+	enum pipe pipe = to_intel_crtc(crtc_state->base.crtc)->pipe;
 	u32 ctl, ctl2;
 
-	if (WARN_ON(pipe != PIPE_A && pipe != PIPE_B))
-		return;
-
 	ctl2 = I915_READ(VLV_BLC_PWM_CTL2(pipe));
 	if (ctl2 & BLM_PWM_ENABLE) {
 		DRM_DEBUG_KMS("backlight already enabled\n");
@@ -1020,7 +1047,7 @@ static void vlv_enable_backlight(struct intel_connector *connector)
 	I915_WRITE(VLV_BLC_PWM_CTL(pipe), ctl);
 
 	/* XXX: combine this into above write? */
-	intel_panel_actually_set_backlight(connector, panel->backlight.level);
+	intel_panel_actually_set_backlight(conn_state, panel->backlight.level);
 
 	ctl2 = 0;
 	if (panel->backlight.active_low_pwm)
@@ -1030,11 +1057,13 @@ static void vlv_enable_backlight(struct intel_connector *connector)
 	I915_WRITE(VLV_BLC_PWM_CTL2(pipe), ctl2 | BLM_PWM_ENABLE);
 }
 
-static void bxt_enable_backlight(struct intel_connector *connector)
+static void bxt_enable_backlight(const struct intel_crtc_state *crtc_state,
+				 const struct drm_connector_state *conn_state)
 {
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_panel *panel = &connector->panel;
-	enum pipe pipe = intel_get_pipe_from_connector(connector);
+	enum pipe pipe = to_intel_crtc(crtc_state->base.crtc)->pipe;
 	u32 pwm_ctl, val;
 
 	/* Controller 1 uses the utility pin. */
@@ -1064,7 +1093,7 @@ static void bxt_enable_backlight(struct intel_connector *connector)
 	I915_WRITE(BXT_BLC_PWM_FREQ(panel->backlight.controller),
 			panel->backlight.max);
 
-	intel_panel_actually_set_backlight(connector, panel->backlight.level);
+	intel_panel_actually_set_backlight(conn_state, panel->backlight.level);
 
 	pwm_ctl = 0;
 	if (panel->backlight.active_low_pwm)
@@ -1076,19 +1105,54 @@ static void bxt_enable_backlight(struct intel_connector *connector)
 			pwm_ctl | BXT_BLC_PWM_ENABLE);
 }
 
-static void pwm_enable_backlight(struct intel_connector *connector)
+static void cnp_enable_backlight(const struct intel_crtc_state *crtc_state,
+				 const struct drm_connector_state *conn_state)
 {
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
+	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
+	struct intel_panel *panel = &connector->panel;
+	u32 pwm_ctl;
+
+	pwm_ctl = I915_READ(BXT_BLC_PWM_CTL(panel->backlight.controller));
+	if (pwm_ctl & BXT_BLC_PWM_ENABLE) {
+		DRM_DEBUG_KMS("backlight already enabled\n");
+		pwm_ctl &= ~BXT_BLC_PWM_ENABLE;
+		I915_WRITE(BXT_BLC_PWM_CTL(panel->backlight.controller),
+			   pwm_ctl);
+	}
+
+	I915_WRITE(BXT_BLC_PWM_FREQ(panel->backlight.controller),
+		   panel->backlight.max);
+
+	intel_panel_actually_set_backlight(conn_state, panel->backlight.level);
+
+	pwm_ctl = 0;
+	if (panel->backlight.active_low_pwm)
+		pwm_ctl |= BXT_BLC_PWM_POLARITY;
+
+	I915_WRITE(BXT_BLC_PWM_CTL(panel->backlight.controller), pwm_ctl);
+	POSTING_READ(BXT_BLC_PWM_CTL(panel->backlight.controller));
+	I915_WRITE(BXT_BLC_PWM_CTL(panel->backlight.controller),
+		   pwm_ctl | BXT_BLC_PWM_ENABLE);
+}
+
+static void pwm_enable_backlight(const struct intel_crtc_state *crtc_state,
+				 const struct drm_connector_state *conn_state)
+{
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct intel_panel *panel = &connector->panel;
 
 	pwm_enable(panel->backlight.pwm);
-	intel_panel_actually_set_backlight(connector, panel->backlight.level);
+	intel_panel_actually_set_backlight(conn_state, panel->backlight.level);
 }
 
-void intel_panel_enable_backlight(struct intel_connector *connector)
+void intel_panel_enable_backlight(const struct intel_crtc_state *crtc_state,
+				  const struct drm_connector_state *conn_state)
 {
+	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_panel *panel = &connector->panel;
-	enum pipe pipe = intel_get_pipe_from_connector(connector);
+	enum pipe pipe = to_intel_crtc(crtc_state->base.crtc)->pipe;
 
 	if (!panel->backlight.present)
 		return;
@@ -1108,7 +1172,7 @@ void intel_panel_enable_backlight(struct intel_connector *connector)
 						 panel->backlight.device->props.max_brightness);
 	}
 
-	panel->backlight.enable(connector);
+	panel->backlight.enable(crtc_state, conn_state);
 	panel->backlight.enabled = true;
 	if (panel->backlight.device)
 		panel->backlight.device->props.power = FB_BLANK_UNBLANK;
@@ -1126,7 +1190,7 @@ static int intel_backlight_device_update_status(struct backlight_device *bd)
 	drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
 	DRM_DEBUG_KMS("updating intel_backlight, brightness=%d/%d\n",
 		      bd->props.brightness, bd->props.max_brightness);
-	intel_panel_set_backlight(connector, bd->props.brightness,
+	intel_panel_set_backlight(connector->base.state, bd->props.brightness,
 				  bd->props.max_brightness);
 
 	/*
@@ -1239,6 +1303,17 @@ void intel_backlight_device_unregister(struct intel_connector *connector)
 #endif /* CONFIG_BACKLIGHT_CLASS_DEVICE */
 
 /*
+ * CNP: PWM clock frequency is 19.2 MHz or 24 MHz.
+ *      PWM increment = 1
+ */
+static u32 cnp_hz_to_pwm(struct intel_connector *connector, u32 pwm_freq_hz)
+{
+	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
+
+	return DIV_ROUND_CLOSEST(KHz(dev_priv->rawclk_freq), pwm_freq_hz);
+}
+
+/*
  * BXT: PWM clock frequency = 19.2 MHz.
  */
 static u32 bxt_hz_to_pwm(struct intel_connector *connector, u32 pwm_freq_hz)
@@ -1633,6 +1708,42 @@ bxt_setup_backlight(struct intel_connector *connector, enum pipe unused)
 	return 0;
 }
 
+static int
+cnp_setup_backlight(struct intel_connector *connector, enum pipe unused)
+{
+	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
+	struct intel_panel *panel = &connector->panel;
+	u32 pwm_ctl, val;
+
+	/*
+	 * CNP has the BXT implementation of backlight, but with only
+	 * one controller. Future platforms could have multiple controllers
+	 * so let's make this extensible and prepared for the future.
+	 */
+	panel->backlight.controller = 0;
+
+	pwm_ctl = I915_READ(BXT_BLC_PWM_CTL(panel->backlight.controller));
+
+	panel->backlight.active_low_pwm = pwm_ctl & BXT_BLC_PWM_POLARITY;
+	panel->backlight.max =
+		I915_READ(BXT_BLC_PWM_FREQ(panel->backlight.controller));
+
+	if (!panel->backlight.max)
+		panel->backlight.max = get_backlight_max_vbt(connector);
+
+	if (!panel->backlight.max)
+		return -ENODEV;
+
+	val = bxt_get_backlight(connector);
+	val = intel_panel_compute_brightness(connector, val);
+	panel->backlight.level = clamp(val, panel->backlight.min,
+				       panel->backlight.max);
+
+	panel->backlight.enabled = pwm_ctl & BXT_BLC_PWM_ENABLE;
+
+	return 0;
+}
+
 static int pwm_setup_backlight(struct intel_connector *connector,
 			       enum pipe pipe)
 {
@@ -1749,6 +1860,13 @@ intel_panel_init_backlight_funcs(struct intel_panel *panel)
 		panel->backlight.set = bxt_set_backlight;
 		panel->backlight.get = bxt_get_backlight;
 		panel->backlight.hz_to_pwm = bxt_hz_to_pwm;
+	} else if (HAS_PCH_CNP(dev_priv)) {
+		panel->backlight.setup = cnp_setup_backlight;
+		panel->backlight.enable = cnp_enable_backlight;
+		panel->backlight.disable = cnp_disable_backlight;
+		panel->backlight.set = bxt_set_backlight;
+		panel->backlight.get = bxt_get_backlight;
+		panel->backlight.hz_to_pwm = cnp_hz_to_pwm;
 	} else if (HAS_PCH_LPT(dev_priv) || HAS_PCH_SPT(dev_priv) ||
 		   HAS_PCH_KBP(dev_priv)) {
 		panel->backlight.setup = lpt_setup_backlight;
diff --git a/drivers/gpu/drm/i915/intel_pipe_crc.c b/drivers/gpu/drm/i915/intel_pipe_crc.c
index 206ee4f0150e..8fbd2bd0877f 100644
--- a/drivers/gpu/drm/i915/intel_pipe_crc.c
+++ b/drivers/gpu/drm/i915/intel_pipe_crc.c
@@ -513,16 +513,20 @@ static void hsw_trans_edp_pipe_A_crc_wa(struct drm_i915_private *dev_priv,
 	struct intel_crtc *crtc = intel_get_crtc_for_pipe(dev_priv, PIPE_A);
 	struct intel_crtc_state *pipe_config;
 	struct drm_atomic_state *state;
+	struct drm_modeset_acquire_ctx ctx;
 	int ret = 0;
 
-	drm_modeset_lock_all(dev);
+	drm_modeset_acquire_init(&ctx, 0);
+
 	state = drm_atomic_state_alloc(dev);
 	if (!state) {
 		ret = -ENOMEM;
 		goto unlock;
 	}
 
-	state->acquire_ctx = crtc->base.dev->mode_config.acquire_ctx;
+	state->acquire_ctx = &ctx;
+
+retry:
 	pipe_config = intel_atomic_get_crtc_state(state, crtc);
 	if (IS_ERR(pipe_config)) {
 		ret = PTR_ERR(pipe_config);
@@ -537,10 +541,17 @@ static void hsw_trans_edp_pipe_A_crc_wa(struct drm_i915_private *dev_priv,
 	ret = drm_atomic_commit(state);
 
 put_state:
+	if (ret == -EDEADLK) {
+		drm_atomic_state_clear(state);
+		drm_modeset_backoff(&ctx);
+		goto retry;
+	}
+
 	drm_atomic_state_put(state);
 unlock:
 	WARN(ret, "Toggling workaround to %i returns %i\n", enable, ret);
-	drm_modeset_unlock_all(dev);
+	drm_modeset_drop_locks(&ctx);
+	drm_modeset_acquire_fini(&ctx);
 }
 
 static int ivb_pipe_crc_ctl_reg(struct drm_i915_private *dev_priv,
@@ -842,19 +853,12 @@ static ssize_t display_crc_ctl_write(struct file *file, const char __user *ubuf,
 		return -E2BIG;
 	}
 
-	tmpbuf = kmalloc(len + 1, GFP_KERNEL);
-	if (!tmpbuf)
-		return -ENOMEM;
-
-	if (copy_from_user(tmpbuf, ubuf, len)) {
-		ret = -EFAULT;
-		goto out;
-	}
-	tmpbuf[len] = '\0';
+	tmpbuf = memdup_user_nul(ubuf, len);
+	if (IS_ERR(tmpbuf))
+		return PTR_ERR(tmpbuf);
 
 	ret = display_crc_ctl_parse(dev_priv, tmpbuf, len);
 
-out:
 	kfree(tmpbuf);
 	if (ret < 0)
 		return ret;
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index 2ca481b5aa69..0aed13dcedf0 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -58,24 +58,24 @@
 
 static void gen9_init_clock_gating(struct drm_i915_private *dev_priv)
 {
-	/* See Bspec note for PSR2_CTL bit 31, Wa#828:skl,bxt,kbl */
+	/* See Bspec note for PSR2_CTL bit 31, Wa#828:skl,bxt,kbl,cfl */
 	I915_WRITE(CHICKEN_PAR1_1,
 		   I915_READ(CHICKEN_PAR1_1) | SKL_EDP_PSR_FIX_RDWRAP);
 
 	I915_WRITE(GEN8_CONFIG0,
 		   I915_READ(GEN8_CONFIG0) | GEN9_DEFAULT_FIXES);
 
-	/* WaEnableChickenDCPR:skl,bxt,kbl,glk */
+	/* WaEnableChickenDCPR:skl,bxt,kbl,glk,cfl */
 	I915_WRITE(GEN8_CHICKEN_DCPR_1,
 		   I915_READ(GEN8_CHICKEN_DCPR_1) | MASK_WAKEMEM);
 
-	/* WaFbcTurnOffFbcWatermark:skl,bxt,kbl */
-	/* WaFbcWakeMemOn:skl,bxt,kbl,glk */
+	/* WaFbcTurnOffFbcWatermark:skl,bxt,kbl,cfl */
+	/* WaFbcWakeMemOn:skl,bxt,kbl,glk,cfl */
 	I915_WRITE(DISP_ARB_CTL, I915_READ(DISP_ARB_CTL) |
 		   DISP_FBC_WM_DIS |
 		   DISP_FBC_MEMORY_WAKE);
 
-	/* WaFbcHighMemBwCorruptionAvoidance:skl,bxt,kbl */
+	/* WaFbcHighMemBwCorruptionAvoidance:skl,bxt,kbl,cfl */
 	I915_WRITE(ILK_DPFC_CHICKEN, I915_READ(ILK_DPFC_CHICKEN) |
 		   ILK_DPFC_DISABLE_DUMMY0);
 }
@@ -386,13 +386,53 @@ static bool _intel_set_memory_cxsr(struct drm_i915_private *dev_priv, bool enabl
 	return was_enabled;
 }
 
+/**
+ * intel_set_memory_cxsr - Configure CxSR state
+ * @dev_priv: i915 device
+ * @enable: Allow vs. disallow CxSR
+ *
+ * Allow or disallow the system to enter a special CxSR
+ * (C-state self refresh) state. What typically happens in CxSR mode
+ * is that several display FIFOs may get combined into a single larger
+ * FIFO for a particular plane (so called max FIFO mode) to allow the
+ * system to defer memory fetches longer, and the memory will enter
+ * self refresh.
+ *
+ * Note that enabling CxSR does not guarantee that the system enter
+ * this special mode, nor does it guarantee that the system stays
+ * in that mode once entered. So this just allows/disallows the system
+ * to autonomously utilize the CxSR mode. Other factors such as core
+ * C-states will affect when/if the system actually enters/exits the
+ * CxSR mode.
+ *
+ * Note that on VLV/CHV this actually only controls the max FIFO mode,
+ * and the system is free to enter/exit memory self refresh at any time
+ * even when the use of CxSR has been disallowed.
+ *
+ * While the system is actually in the CxSR/max FIFO mode, some plane
+ * control registers will not get latched on vblank. Thus in order to
+ * guarantee the system will respond to changes in the plane registers
+ * we must always disallow CxSR prior to making changes to those registers.
+ * Unfortunately the system will re-evaluate the CxSR conditions at
+ * frame start which happens after vblank start (which is when the plane
+ * registers would get latched), so we can't proceed with the plane update
+ * during the same frame where we disallowed CxSR.
+ *
+ * Certain platforms also have a deeper HPLL SR mode. Fortunately the
+ * HPLL SR mode depends on CxSR itself, so we don't have to hand hold
+ * the hardware w.r.t. HPLL SR when writing to plane registers.
+ * Disallowing just CxSR is sufficient.
+ */
 bool intel_set_memory_cxsr(struct drm_i915_private *dev_priv, bool enable)
 {
 	bool ret;
 
 	mutex_lock(&dev_priv->wm.wm_mutex);
 	ret = _intel_set_memory_cxsr(dev_priv, enable);
-	dev_priv->wm.vlv.cxsr = enable;
+	if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
+		dev_priv->wm.vlv.cxsr = enable;
+	else if (IS_G4X(dev_priv))
+		dev_priv->wm.g4x.cxsr = enable;
 	mutex_unlock(&dev_priv->wm.wm_mutex);
 
 	return ret;
@@ -454,13 +494,6 @@ static void vlv_get_fifo_size(struct intel_crtc_state *crtc_state)
 	fifo_state->plane[PLANE_SPRITE0] = sprite1_start - sprite0_start;
 	fifo_state->plane[PLANE_SPRITE1] = 511 - sprite1_start;
 	fifo_state->plane[PLANE_CURSOR] = 63;
-
-	DRM_DEBUG_KMS("Pipe %c FIFO size: %d/%d/%d/%d\n",
-		      pipe_name(pipe),
-		      fifo_state->plane[PLANE_PRIMARY],
-		      fifo_state->plane[PLANE_SPRITE0],
-		      fifo_state->plane[PLANE_SPRITE1],
-		      fifo_state->plane[PLANE_CURSOR]);
 }
 
 static int i9xx_get_fifo_size(struct drm_i915_private *dev_priv, int plane)
@@ -538,20 +571,6 @@ static const struct intel_watermark_params pineview_cursor_hplloff_wm = {
 	.guard_size = PINEVIEW_CURSOR_GUARD_WM,
 	.cacheline_size = PINEVIEW_FIFO_LINE_SIZE,
 };
-static const struct intel_watermark_params g4x_wm_info = {
-	.fifo_size = G4X_FIFO_SIZE,
-	.max_wm = G4X_MAX_WM,
-	.default_wm = G4X_MAX_WM,
-	.guard_size = 2,
-	.cacheline_size = G4X_FIFO_LINE_SIZE,
-};
-static const struct intel_watermark_params g4x_cursor_wm_info = {
-	.fifo_size = I965_CURSOR_FIFO,
-	.max_wm = I965_CURSOR_MAX_WM,
-	.default_wm = I965_CURSOR_DFT_WM,
-	.guard_size = 2,
-	.cacheline_size = G4X_FIFO_LINE_SIZE,
-};
 static const struct intel_watermark_params i965_cursor_wm_info = {
 	.fifo_size = I965_CURSOR_FIFO,
 	.max_wm = I965_CURSOR_MAX_WM,
@@ -596,8 +615,104 @@ static const struct intel_watermark_params i845_wm_info = {
 };
 
 /**
+ * intel_wm_method1 - Method 1 / "small buffer" watermark formula
+ * @pixel_rate: Pipe pixel rate in kHz
+ * @cpp: Plane bytes per pixel
+ * @latency: Memory wakeup latency in 0.1us units
+ *
+ * Compute the watermark using the method 1 or "small buffer"
+ * formula. The caller may additonally add extra cachelines
+ * to account for TLB misses and clock crossings.
+ *
+ * This method is concerned with the short term drain rate
+ * of the FIFO, ie. it does not account for blanking periods
+ * which would effectively reduce the average drain rate across
+ * a longer period. The name "small" refers to the fact the
+ * FIFO is relatively small compared to the amount of data
+ * fetched.
+ *
+ * The FIFO level vs. time graph might look something like:
+ *
+ *   |\   |\
+ *   | \  | \
+ * __---__---__ (- plane active, _ blanking)
+ * -> time
+ *
+ * or perhaps like this:
+ *
+ *   |\|\  |\|\
+ * __----__----__ (- plane active, _ blanking)
+ * -> time
+ *
+ * Returns:
+ * The watermark in bytes
+ */
+static unsigned int intel_wm_method1(unsigned int pixel_rate,
+				     unsigned int cpp,
+				     unsigned int latency)
+{
+	uint64_t ret;
+
+	ret = (uint64_t) pixel_rate * cpp * latency;
+	ret = DIV_ROUND_UP_ULL(ret, 10000);
+
+	return ret;
+}
+
+/**
+ * intel_wm_method2 - Method 2 / "large buffer" watermark formula
+ * @pixel_rate: Pipe pixel rate in kHz
+ * @htotal: Pipe horizontal total
+ * @width: Plane width in pixels
+ * @cpp: Plane bytes per pixel
+ * @latency: Memory wakeup latency in 0.1us units
+ *
+ * Compute the watermark using the method 2 or "large buffer"
+ * formula. The caller may additonally add extra cachelines
+ * to account for TLB misses and clock crossings.
+ *
+ * This method is concerned with the long term drain rate
+ * of the FIFO, ie. it does account for blanking periods
+ * which effectively reduce the average drain rate across
+ * a longer period. The name "large" refers to the fact the
+ * FIFO is relatively large compared to the amount of data
+ * fetched.
+ *
+ * The FIFO level vs. time graph might look something like:
+ *
+ *    |\___       |\___
+ *    |    \___   |    \___
+ *    |        \  |        \
+ * __ --__--__--__--__--__--__ (- plane active, _ blanking)
+ * -> time
+ *
+ * Returns:
+ * The watermark in bytes
+ */
+static unsigned int intel_wm_method2(unsigned int pixel_rate,
+				     unsigned int htotal,
+				     unsigned int width,
+				     unsigned int cpp,
+				     unsigned int latency)
+{
+	unsigned int ret;
+
+	/*
+	 * FIXME remove once all users are computing
+	 * watermarks in the correct place.
+	 */
+	if (WARN_ON_ONCE(htotal == 0))
+		htotal = 1;
+
+	ret = (latency * pixel_rate) / (htotal * 10000);
+	ret = (ret + 1) * width * cpp;
+
+	return ret;
+}
+
+/**
  * intel_calculate_wm - calculate watermark level
- * @clock_in_khz: pixel clock
+ * @pixel_rate: pixel clock
  * @wm: chip FIFO params
  * @cpp: bytes per pixel
  * @latency_ns: memory latency for the platform
@@ -613,12 +728,12 @@ static const struct intel_watermark_params i845_wm_info = {
  * past the watermark point.  If the FIFO drains completely, a FIFO underrun
  * will occur, and a display engine hang could result.
  */
-static unsigned long intel_calculate_wm(unsigned long clock_in_khz,
-					const struct intel_watermark_params *wm,
-					int fifo_size, int cpp,
-					unsigned long latency_ns)
+static unsigned int intel_calculate_wm(int pixel_rate,
+				       const struct intel_watermark_params *wm,
+				       int fifo_size, int cpp,
+				       unsigned int latency_ns)
 {
-	long entries_required, wm_size;
+	int entries, wm_size;
 
 	/*
 	 * Note: we need to make sure we don't overflow for various clock &
@@ -626,18 +741,17 @@ static unsigned long intel_calculate_wm(unsigned long clock_in_khz,
 	 * clocks go from a few thousand to several hundred thousand.
 	 * latency is usually a few thousand
 	 */
-	entries_required = ((clock_in_khz / 1000) * cpp * latency_ns) /
-		1000;
-	entries_required = DIV_ROUND_UP(entries_required, wm->cacheline_size);
+	entries = intel_wm_method1(pixel_rate, cpp,
+				   latency_ns / 100);
+	entries = DIV_ROUND_UP(entries, wm->cacheline_size) +
+		wm->guard_size;
+	DRM_DEBUG_KMS("FIFO entries required for mode: %d\n", entries);
 
-	DRM_DEBUG_KMS("FIFO entries required for mode: %ld\n", entries_required);
-
-	wm_size = fifo_size - (entries_required + wm->guard_size);
-
-	DRM_DEBUG_KMS("FIFO watermark level: %ld\n", wm_size);
+	wm_size = fifo_size - entries;
+	DRM_DEBUG_KMS("FIFO watermark level: %d\n", wm_size);
 
 	/* Don't promote wm_size to unsigned... */
-	if (wm_size > (long)wm->max_wm)
+	if (wm_size > wm->max_wm)
 		wm_size = wm->max_wm;
 	if (wm_size <= 0)
 		wm_size = wm->default_wm;
@@ -655,6 +769,21 @@ static unsigned long intel_calculate_wm(unsigned long clock_in_khz,
 	return wm_size;
 }
 
+static bool is_disabling(int old, int new, int threshold)
+{
+	return old >= threshold && new < threshold;
+}
+
+static bool is_enabling(int old, int new, int threshold)
+{
+	return old < threshold && new >= threshold;
+}
+
+static int intel_wm_num_levels(struct drm_i915_private *dev_priv)
+{
+	return dev_priv->wm.max_level + 1;
+}
+
 static bool intel_wm_plane_visible(const struct intel_crtc_state *crtc_state,
 				   const struct intel_plane_state *plane_state)
 {
@@ -699,7 +828,7 @@ static void pineview_update_wm(struct intel_crtc *unused_crtc)
 	struct intel_crtc *crtc;
 	const struct cxsr_latency *latency;
 	u32 reg;
-	unsigned long wm;
+	unsigned int wm;
 
 	latency = intel_get_cxsr_latency(IS_PINEVIEW_G(dev_priv),
 					 dev_priv->is_ddr3,
@@ -733,7 +862,7 @@ static void pineview_update_wm(struct intel_crtc *unused_crtc)
 		/* cursor SR */
 		wm = intel_calculate_wm(clock, &pineview_cursor_wm,
 					pineview_display_wm.fifo_size,
-					cpp, latency->cursor_sr);
+					4, latency->cursor_sr);
 		reg = I915_READ(DSPFW3);
 		reg &= ~DSPFW_CURSOR_SR_MASK;
 		reg |= FW_WM(wm, CURSOR_SR);
@@ -751,7 +880,7 @@ static void pineview_update_wm(struct intel_crtc *unused_crtc)
 		/* cursor HPLL off SR */
 		wm = intel_calculate_wm(clock, &pineview_cursor_hplloff_wm,
 					pineview_display_hplloff_wm.fifo_size,
-					cpp, latency->cursor_hpll_disable);
+					4, latency->cursor_hpll_disable);
 		reg = I915_READ(DSPFW3);
 		reg &= ~DSPFW_HPLL_CURSOR_MASK;
 		reg |= FW_WM(wm, HPLL_CURSOR);
@@ -764,144 +893,50 @@ static void pineview_update_wm(struct intel_crtc *unused_crtc)
 	}
 }
 
-static bool g4x_compute_wm0(struct drm_i915_private *dev_priv,
-			    int plane,
-			    const struct intel_watermark_params *display,
-			    int display_latency_ns,
-			    const struct intel_watermark_params *cursor,
-			    int cursor_latency_ns,
-			    int *plane_wm,
-			    int *cursor_wm)
-{
-	struct intel_crtc *crtc;
-	const struct drm_display_mode *adjusted_mode;
-	const struct drm_framebuffer *fb;
-	int htotal, hdisplay, clock, cpp;
-	int line_time_us, line_count;
-	int entries, tlb_miss;
-
-	crtc = intel_get_crtc_for_plane(dev_priv, plane);
-	if (!intel_crtc_active(crtc)) {
-		*cursor_wm = cursor->guard_size;
-		*plane_wm = display->guard_size;
-		return false;
-	}
-
-	adjusted_mode = &crtc->config->base.adjusted_mode;
-	fb = crtc->base.primary->state->fb;
-	clock = adjusted_mode->crtc_clock;
-	htotal = adjusted_mode->crtc_htotal;
-	hdisplay = crtc->config->pipe_src_w;
-	cpp = fb->format->cpp[0];
-
-	/* Use the small buffer method to calculate plane watermark */
-	entries = ((clock * cpp / 1000) * display_latency_ns) / 1000;
-	tlb_miss = display->fifo_size*display->cacheline_size - hdisplay * 8;
-	if (tlb_miss > 0)
-		entries += tlb_miss;
-	entries = DIV_ROUND_UP(entries, display->cacheline_size);
-	*plane_wm = entries + display->guard_size;
-	if (*plane_wm > (int)display->max_wm)
-		*plane_wm = display->max_wm;
-
-	/* Use the large buffer method to calculate cursor watermark */
-	line_time_us = max(htotal * 1000 / clock, 1);
-	line_count = (cursor_latency_ns / line_time_us + 1000) / 1000;
-	entries = line_count * crtc->base.cursor->state->crtc_w * cpp;
-	tlb_miss = cursor->fifo_size*cursor->cacheline_size - hdisplay * 8;
-	if (tlb_miss > 0)
-		entries += tlb_miss;
-	entries = DIV_ROUND_UP(entries, cursor->cacheline_size);
-	*cursor_wm = entries + cursor->guard_size;
-	if (*cursor_wm > (int)cursor->max_wm)
-		*cursor_wm = (int)cursor->max_wm;
-
-	return true;
-}
-
 /*
- * Check the wm result.
- *
- * If any calculated watermark values is larger than the maximum value that
- * can be programmed into the associated watermark register, that watermark
- * must be disabled.
+ * Documentation says:
+ * "If the line size is small, the TLB fetches can get in the way of the
+ *  data fetches, causing some lag in the pixel data return which is not
+ *  accounted for in the above formulas. The following adjustment only
+ *  needs to be applied if eight whole lines fit in the buffer at once.
+ *  The WM is adjusted upwards by the difference between the FIFO size
+ *  and the size of 8 whole lines. This adjustment is always performed
+ *  in the actual pixel depth regardless of whether FBC is enabled or not."
  */
-static bool g4x_check_srwm(struct drm_i915_private *dev_priv,
-			   int display_wm, int cursor_wm,
-			   const struct intel_watermark_params *display,
-			   const struct intel_watermark_params *cursor)
+static int g4x_tlb_miss_wa(int fifo_size, int width, int cpp)
 {
-	DRM_DEBUG_KMS("SR watermark: display plane %d, cursor %d\n",
-		      display_wm, cursor_wm);
-
-	if (display_wm > display->max_wm) {
-		DRM_DEBUG_KMS("display watermark is too large(%d/%u), disabling\n",
-			      display_wm, display->max_wm);
-		return false;
-	}
-
-	if (cursor_wm > cursor->max_wm) {
-		DRM_DEBUG_KMS("cursor watermark is too large(%d/%u), disabling\n",
-			      cursor_wm, cursor->max_wm);
-		return false;
-	}
-
-	if (!(display_wm || cursor_wm)) {
-		DRM_DEBUG_KMS("SR latency is 0, disabling\n");
-		return false;
-	}
+	int tlb_miss = fifo_size * 64 - width * cpp * 8;
 
-	return true;
+	return max(0, tlb_miss);
 }
 
-static bool g4x_compute_srwm(struct drm_i915_private *dev_priv,
-			     int plane,
-			     int latency_ns,
-			     const struct intel_watermark_params *display,
-			     const struct intel_watermark_params *cursor,
-			     int *display_wm, int *cursor_wm)
+static void g4x_write_wm_values(struct drm_i915_private *dev_priv,
+				const struct g4x_wm_values *wm)
 {
-	struct intel_crtc *crtc;
-	const struct drm_display_mode *adjusted_mode;
-	const struct drm_framebuffer *fb;
-	int hdisplay, htotal, cpp, clock;
-	unsigned long line_time_us;
-	int line_count, line_size;
-	int small, large;
-	int entries;
-
-	if (!latency_ns) {
-		*display_wm = *cursor_wm = 0;
-		return false;
-	}
-
-	crtc = intel_get_crtc_for_plane(dev_priv, plane);
-	adjusted_mode = &crtc->config->base.adjusted_mode;
-	fb = crtc->base.primary->state->fb;
-	clock = adjusted_mode->crtc_clock;
-	htotal = adjusted_mode->crtc_htotal;
-	hdisplay = crtc->config->pipe_src_w;
-	cpp = fb->format->cpp[0];
-
-	line_time_us = max(htotal * 1000 / clock, 1);
-	line_count = (latency_ns / line_time_us + 1000) / 1000;
-	line_size = hdisplay * cpp;
-
-	/* Use the minimum of the small and large buffer method for primary */
-	small = ((clock * cpp / 1000) * latency_ns) / 1000;
-	large = line_count * line_size;
+	enum pipe pipe;
 
-	entries = DIV_ROUND_UP(min(small, large), display->cacheline_size);
-	*display_wm = entries + display->guard_size;
+	for_each_pipe(dev_priv, pipe)
+		trace_g4x_wm(intel_get_crtc_for_pipe(dev_priv, pipe), wm);
 
-	/* calculate the self-refresh watermark for display cursor */
-	entries = line_count * cpp * crtc->base.cursor->state->crtc_w;
-	entries = DIV_ROUND_UP(entries, cursor->cacheline_size);
-	*cursor_wm = entries + cursor->guard_size;
+	I915_WRITE(DSPFW1,
+		   FW_WM(wm->sr.plane, SR) |
+		   FW_WM(wm->pipe[PIPE_B].plane[PLANE_CURSOR], CURSORB) |
+		   FW_WM(wm->pipe[PIPE_B].plane[PLANE_PRIMARY], PLANEB) |
+		   FW_WM(wm->pipe[PIPE_A].plane[PLANE_PRIMARY], PLANEA));
+	I915_WRITE(DSPFW2,
+		   (wm->fbc_en ? DSPFW_FBC_SR_EN : 0) |
+		   FW_WM(wm->sr.fbc, FBC_SR) |
+		   FW_WM(wm->hpll.fbc, FBC_HPLL_SR) |
+		   FW_WM(wm->pipe[PIPE_B].plane[PLANE_SPRITE0], SPRITEB) |
+		   FW_WM(wm->pipe[PIPE_A].plane[PLANE_CURSOR], CURSORA) |
+		   FW_WM(wm->pipe[PIPE_A].plane[PLANE_SPRITE0], SPRITEA));
+	I915_WRITE(DSPFW3,
+		   (wm->hpll_en ? DSPFW_HPLL_SR_EN : 0) |
+		   FW_WM(wm->sr.cursor, CURSOR_SR) |
+		   FW_WM(wm->hpll.cursor, HPLL_CURSOR) |
+		   FW_WM(wm->hpll.plane, HPLL_SR));
 
-	return g4x_check_srwm(dev_priv,
-			      *display_wm, *cursor_wm,
-			      display, cursor);
+	POSTING_READ(DSPFW1);
 }
 
 #define FW_WM_VLV(value, plane) \
@@ -985,17 +1020,535 @@ static void vlv_write_wm_values(struct drm_i915_private *dev_priv,
 
 #undef FW_WM_VLV
 
+static void g4x_setup_wm_latency(struct drm_i915_private *dev_priv)
+{
+	/* all latencies in usec */
+	dev_priv->wm.pri_latency[G4X_WM_LEVEL_NORMAL] = 5;
+	dev_priv->wm.pri_latency[G4X_WM_LEVEL_SR] = 12;
+	dev_priv->wm.pri_latency[G4X_WM_LEVEL_HPLL] = 35;
+
+	dev_priv->wm.max_level = G4X_WM_LEVEL_HPLL;
+}
+
+static int g4x_plane_fifo_size(enum plane_id plane_id, int level)
+{
+	/*
+	 * DSPCNTR[13] supposedly controls whether the
+	 * primary plane can use the FIFO space otherwise
+	 * reserved for the sprite plane. It's not 100% clear
+	 * what the actual FIFO size is, but it looks like we
+	 * can happily set both primary and sprite watermarks
+	 * up to 127 cachelines. So that would seem to mean
+	 * that either DSPCNTR[13] doesn't do anything, or that
+	 * the total FIFO is >= 256 cachelines in size. Either
+	 * way, we don't seem to have to worry about this
+	 * repartitioning as the maximum watermark value the
+	 * register can hold for each plane is lower than the
+	 * minimum FIFO size.
+	 */
+	switch (plane_id) {
+	case PLANE_CURSOR:
+		return 63;
+	case PLANE_PRIMARY:
+		return level == G4X_WM_LEVEL_NORMAL ? 127 : 511;
+	case PLANE_SPRITE0:
+		return level == G4X_WM_LEVEL_NORMAL ? 127 : 0;
+	default:
+		MISSING_CASE(plane_id);
+		return 0;
+	}
+}
+
+static int g4x_fbc_fifo_size(int level)
+{
+	switch (level) {
+	case G4X_WM_LEVEL_SR:
+		return 7;
+	case G4X_WM_LEVEL_HPLL:
+		return 15;
+	default:
+		MISSING_CASE(level);
+		return 0;
+	}
+}
+
+static uint16_t g4x_compute_wm(const struct intel_crtc_state *crtc_state,
+			       const struct intel_plane_state *plane_state,
+			       int level)
+{
+	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
+	const struct drm_display_mode *adjusted_mode =
+		&crtc_state->base.adjusted_mode;
+	int clock, htotal, cpp, width, wm;
+	int latency = dev_priv->wm.pri_latency[level] * 10;
+
+	if (latency == 0)
+		return USHRT_MAX;
+
+	if (!intel_wm_plane_visible(crtc_state, plane_state))
+		return 0;
+
+	/*
+	 * Not 100% sure which way ELK should go here as the
+	 * spec only says CL/CTG should assume 32bpp and BW
+	 * doesn't need to. But as these things followed the
+	 * mobile vs. desktop lines on gen3 as well, let's
+	 * assume ELK doesn't need this.
+	 *
+	 * The spec also fails to list such a restriction for
+	 * the HPLL watermark, which seems a little strange.
+	 * Let's use 32bpp for the HPLL watermark as well.
+	 */
+	if (IS_GM45(dev_priv) && plane->id == PLANE_PRIMARY &&
+	    level != G4X_WM_LEVEL_NORMAL)
+		cpp = 4;
+	else
+		cpp = plane_state->base.fb->format->cpp[0];
+
+	clock = adjusted_mode->crtc_clock;
+	htotal = adjusted_mode->crtc_htotal;
+
+	if (plane->id == PLANE_CURSOR)
+		width = plane_state->base.crtc_w;
+	else
+		width = drm_rect_width(&plane_state->base.dst);
+
+	if (plane->id == PLANE_CURSOR) {
+		wm = intel_wm_method2(clock, htotal, width, cpp, latency);
+	} else if (plane->id == PLANE_PRIMARY &&
+		   level == G4X_WM_LEVEL_NORMAL) {
+		wm = intel_wm_method1(clock, cpp, latency);
+	} else {
+		int small, large;
+
+		small = intel_wm_method1(clock, cpp, latency);
+		large = intel_wm_method2(clock, htotal, width, cpp, latency);
+
+		wm = min(small, large);
+	}
+
+	wm += g4x_tlb_miss_wa(g4x_plane_fifo_size(plane->id, level),
+			      width, cpp);
+
+	wm = DIV_ROUND_UP(wm, 64) + 2;
+
+	return min_t(int, wm, USHRT_MAX);
+}
+
+static bool g4x_raw_plane_wm_set(struct intel_crtc_state *crtc_state,
+				 int level, enum plane_id plane_id, u16 value)
+{
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	bool dirty = false;
+
+	for (; level < intel_wm_num_levels(dev_priv); level++) {
+		struct g4x_pipe_wm *raw = &crtc_state->wm.g4x.raw[level];
+
+		dirty |= raw->plane[plane_id] != value;
+		raw->plane[plane_id] = value;
+	}
+
+	return dirty;
+}
+
+static bool g4x_raw_fbc_wm_set(struct intel_crtc_state *crtc_state,
+			       int level, u16 value)
+{
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	bool dirty = false;
+
+	/* NORMAL level doesn't have an FBC watermark */
+	level = max(level, G4X_WM_LEVEL_SR);
+
+	for (; level < intel_wm_num_levels(dev_priv); level++) {
+		struct g4x_pipe_wm *raw = &crtc_state->wm.g4x.raw[level];
+
+		dirty |= raw->fbc != value;
+		raw->fbc = value;
+	}
+
+	return dirty;
+}
+
+static uint32_t ilk_compute_fbc_wm(const struct intel_crtc_state *cstate,
+				   const struct intel_plane_state *pstate,
+				   uint32_t pri_val);
+
+static bool g4x_raw_plane_wm_compute(struct intel_crtc_state *crtc_state,
+				     const struct intel_plane_state *plane_state)
+{
+	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	int num_levels = intel_wm_num_levels(to_i915(plane->base.dev));
+	enum plane_id plane_id = plane->id;
+	bool dirty = false;
+	int level;
+
+	if (!intel_wm_plane_visible(crtc_state, plane_state)) {
+		dirty |= g4x_raw_plane_wm_set(crtc_state, 0, plane_id, 0);
+		if (plane_id == PLANE_PRIMARY)
+			dirty |= g4x_raw_fbc_wm_set(crtc_state, 0, 0);
+		goto out;
+	}
+
+	for (level = 0; level < num_levels; level++) {
+		struct g4x_pipe_wm *raw = &crtc_state->wm.g4x.raw[level];
+		int wm, max_wm;
+
+		wm = g4x_compute_wm(crtc_state, plane_state, level);
+		max_wm = g4x_plane_fifo_size(plane_id, level);
+
+		if (wm > max_wm)
+			break;
+
+		dirty |= raw->plane[plane_id] != wm;
+		raw->plane[plane_id] = wm;
+
+		if (plane_id != PLANE_PRIMARY ||
+		    level == G4X_WM_LEVEL_NORMAL)
+			continue;
+
+		wm = ilk_compute_fbc_wm(crtc_state, plane_state,
+					raw->plane[plane_id]);
+		max_wm = g4x_fbc_fifo_size(level);
+
+		/*
+		 * FBC wm is not mandatory as we
+		 * can always just disable its use.
+		 */
+		if (wm > max_wm)
+			wm = USHRT_MAX;
+
+		dirty |= raw->fbc != wm;
+		raw->fbc = wm;
+	}
+
+	/* mark watermarks as invalid */
+	dirty |= g4x_raw_plane_wm_set(crtc_state, level, plane_id, USHRT_MAX);
+
+	if (plane_id == PLANE_PRIMARY)
+		dirty |= g4x_raw_fbc_wm_set(crtc_state, level, USHRT_MAX);
+
+ out:
+	if (dirty) {
+		DRM_DEBUG_KMS("%s watermarks: normal=%d, SR=%d, HPLL=%d\n",
+			      plane->base.name,
+			      crtc_state->wm.g4x.raw[G4X_WM_LEVEL_NORMAL].plane[plane_id],
+			      crtc_state->wm.g4x.raw[G4X_WM_LEVEL_SR].plane[plane_id],
+			      crtc_state->wm.g4x.raw[G4X_WM_LEVEL_HPLL].plane[plane_id]);
+
+		if (plane_id == PLANE_PRIMARY)
+			DRM_DEBUG_KMS("FBC watermarks: SR=%d, HPLL=%d\n",
+				      crtc_state->wm.g4x.raw[G4X_WM_LEVEL_SR].fbc,
+				      crtc_state->wm.g4x.raw[G4X_WM_LEVEL_HPLL].fbc);
+	}
+
+	return dirty;
+}
+
+static bool g4x_raw_plane_wm_is_valid(const struct intel_crtc_state *crtc_state,
+				      enum plane_id plane_id, int level)
+{
+	const struct g4x_pipe_wm *raw = &crtc_state->wm.g4x.raw[level];
+
+	return raw->plane[plane_id] <= g4x_plane_fifo_size(plane_id, level);
+}
+
+static bool g4x_raw_crtc_wm_is_valid(const struct intel_crtc_state *crtc_state,
+				     int level)
+{
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+
+	if (level > dev_priv->wm.max_level)
+		return false;
+
+	return g4x_raw_plane_wm_is_valid(crtc_state, PLANE_PRIMARY, level) &&
+		g4x_raw_plane_wm_is_valid(crtc_state, PLANE_SPRITE0, level) &&
+		g4x_raw_plane_wm_is_valid(crtc_state, PLANE_CURSOR, level);
+}
+
+/* mark all levels starting from 'level' as invalid */
+static void g4x_invalidate_wms(struct intel_crtc *crtc,
+			       struct g4x_wm_state *wm_state, int level)
+{
+	if (level <= G4X_WM_LEVEL_NORMAL) {
+		enum plane_id plane_id;
+
+		for_each_plane_id_on_crtc(crtc, plane_id)
+			wm_state->wm.plane[plane_id] = USHRT_MAX;
+	}
+
+	if (level <= G4X_WM_LEVEL_SR) {
+		wm_state->cxsr = false;
+		wm_state->sr.cursor = USHRT_MAX;
+		wm_state->sr.plane = USHRT_MAX;
+		wm_state->sr.fbc = USHRT_MAX;
+	}
+
+	if (level <= G4X_WM_LEVEL_HPLL) {
+		wm_state->hpll_en = false;
+		wm_state->hpll.cursor = USHRT_MAX;
+		wm_state->hpll.plane = USHRT_MAX;
+		wm_state->hpll.fbc = USHRT_MAX;
+	}
+}
+
+static int g4x_compute_pipe_wm(struct intel_crtc_state *crtc_state)
+{
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_atomic_state *state =
+		to_intel_atomic_state(crtc_state->base.state);
+	struct g4x_wm_state *wm_state = &crtc_state->wm.g4x.optimal;
+	int num_active_planes = hweight32(crtc_state->active_planes &
+					  ~BIT(PLANE_CURSOR));
+	const struct g4x_pipe_wm *raw;
+	struct intel_plane_state *plane_state;
+	struct intel_plane *plane;
+	enum plane_id plane_id;
+	int i, level;
+	unsigned int dirty = 0;
+
+	for_each_intel_plane_in_state(state, plane, plane_state, i) {
+		const struct intel_plane_state *old_plane_state =
+			to_intel_plane_state(plane->base.state);
+
+		if (plane_state->base.crtc != &crtc->base &&
+		    old_plane_state->base.crtc != &crtc->base)
+			continue;
+
+		if (g4x_raw_plane_wm_compute(crtc_state, plane_state))
+			dirty |= BIT(plane->id);
+	}
+
+	if (!dirty)
+		return 0;
+
+	level = G4X_WM_LEVEL_NORMAL;
+	if (!g4x_raw_crtc_wm_is_valid(crtc_state, level))
+		goto out;
+
+	raw = &crtc_state->wm.g4x.raw[level];
+	for_each_plane_id_on_crtc(crtc, plane_id)
+		wm_state->wm.plane[plane_id] = raw->plane[plane_id];
+
+	level = G4X_WM_LEVEL_SR;
+
+	if (!g4x_raw_crtc_wm_is_valid(crtc_state, level))
+		goto out;
+
+	raw = &crtc_state->wm.g4x.raw[level];
+	wm_state->sr.plane = raw->plane[PLANE_PRIMARY];
+	wm_state->sr.cursor = raw->plane[PLANE_CURSOR];
+	wm_state->sr.fbc = raw->fbc;
+
+	wm_state->cxsr = num_active_planes == BIT(PLANE_PRIMARY);
+
+	level = G4X_WM_LEVEL_HPLL;
+
+	if (!g4x_raw_crtc_wm_is_valid(crtc_state, level))
+		goto out;
+
+	raw = &crtc_state->wm.g4x.raw[level];
+	wm_state->hpll.plane = raw->plane[PLANE_PRIMARY];
+	wm_state->hpll.cursor = raw->plane[PLANE_CURSOR];
+	wm_state->hpll.fbc = raw->fbc;
+
+	wm_state->hpll_en = wm_state->cxsr;
+
+	level++;
+
+ out:
+	if (level == G4X_WM_LEVEL_NORMAL)
+		return -EINVAL;
+
+	/* invalidate the higher levels */
+	g4x_invalidate_wms(crtc, wm_state, level);
+
+	/*
+	 * Determine if the FBC watermark(s) can be used. IF
+	 * this isn't the case we prefer to disable the FBC
+	 ( watermark(s) rather than disable the SR/HPLL
+	 * level(s) entirely.
+	 */
+	wm_state->fbc_en = level > G4X_WM_LEVEL_NORMAL;
+
+	if (level >= G4X_WM_LEVEL_SR &&
+	    wm_state->sr.fbc > g4x_fbc_fifo_size(G4X_WM_LEVEL_SR))
+		wm_state->fbc_en = false;
+	else if (level >= G4X_WM_LEVEL_HPLL &&
+		 wm_state->hpll.fbc > g4x_fbc_fifo_size(G4X_WM_LEVEL_HPLL))
+		wm_state->fbc_en = false;
+
+	return 0;
+}
+
+static int g4x_compute_intermediate_wm(struct drm_device *dev,
+				       struct intel_crtc *crtc,
+				       struct intel_crtc_state *crtc_state)
+{
+	struct g4x_wm_state *intermediate = &crtc_state->wm.g4x.intermediate;
+	const struct g4x_wm_state *optimal = &crtc_state->wm.g4x.optimal;
+	const struct g4x_wm_state *active = &crtc->wm.active.g4x;
+	enum plane_id plane_id;
+
+	intermediate->cxsr = optimal->cxsr && active->cxsr &&
+		!crtc_state->disable_cxsr;
+	intermediate->hpll_en = optimal->hpll_en && active->hpll_en &&
+		!crtc_state->disable_cxsr;
+	intermediate->fbc_en = optimal->fbc_en && active->fbc_en;
+
+	for_each_plane_id_on_crtc(crtc, plane_id) {
+		intermediate->wm.plane[plane_id] =
+			max(optimal->wm.plane[plane_id],
+			    active->wm.plane[plane_id]);
+
+		WARN_ON(intermediate->wm.plane[plane_id] >
+			g4x_plane_fifo_size(plane_id, G4X_WM_LEVEL_NORMAL));
+	}
+
+	intermediate->sr.plane = max(optimal->sr.plane,
+				     active->sr.plane);
+	intermediate->sr.cursor = max(optimal->sr.cursor,
+				      active->sr.cursor);
+	intermediate->sr.fbc = max(optimal->sr.fbc,
+				   active->sr.fbc);
+
+	intermediate->hpll.plane = max(optimal->hpll.plane,
+				       active->hpll.plane);
+	intermediate->hpll.cursor = max(optimal->hpll.cursor,
+					active->hpll.cursor);
+	intermediate->hpll.fbc = max(optimal->hpll.fbc,
+				     active->hpll.fbc);
+
+	WARN_ON((intermediate->sr.plane >
+		 g4x_plane_fifo_size(PLANE_PRIMARY, G4X_WM_LEVEL_SR) ||
+		 intermediate->sr.cursor >
+		 g4x_plane_fifo_size(PLANE_CURSOR, G4X_WM_LEVEL_SR)) &&
+		intermediate->cxsr);
+	WARN_ON((intermediate->sr.plane >
+		 g4x_plane_fifo_size(PLANE_PRIMARY, G4X_WM_LEVEL_HPLL) ||
+		 intermediate->sr.cursor >
+		 g4x_plane_fifo_size(PLANE_CURSOR, G4X_WM_LEVEL_HPLL)) &&
+		intermediate->hpll_en);
+
+	WARN_ON(intermediate->sr.fbc > g4x_fbc_fifo_size(1) &&
+		intermediate->fbc_en && intermediate->cxsr);
+	WARN_ON(intermediate->hpll.fbc > g4x_fbc_fifo_size(2) &&
+		intermediate->fbc_en && intermediate->hpll_en);
+
+	/*
+	 * If our intermediate WM are identical to the final WM, then we can
+	 * omit the post-vblank programming; only update if it's different.
+	 */
+	if (memcmp(intermediate, optimal, sizeof(*intermediate)) != 0)
+		crtc_state->wm.need_postvbl_update = true;
+
+	return 0;
+}
+
+static void g4x_merge_wm(struct drm_i915_private *dev_priv,
+			 struct g4x_wm_values *wm)
+{
+	struct intel_crtc *crtc;
+	int num_active_crtcs = 0;
+
+	wm->cxsr = true;
+	wm->hpll_en = true;
+	wm->fbc_en = true;
+
+	for_each_intel_crtc(&dev_priv->drm, crtc) {
+		const struct g4x_wm_state *wm_state = &crtc->wm.active.g4x;
+
+		if (!crtc->active)
+			continue;
+
+		if (!wm_state->cxsr)
+			wm->cxsr = false;
+		if (!wm_state->hpll_en)
+			wm->hpll_en = false;
+		if (!wm_state->fbc_en)
+			wm->fbc_en = false;
+
+		num_active_crtcs++;
+	}
+
+	if (num_active_crtcs != 1) {
+		wm->cxsr = false;
+		wm->hpll_en = false;
+		wm->fbc_en = false;
+	}
+
+	for_each_intel_crtc(&dev_priv->drm, crtc) {
+		const struct g4x_wm_state *wm_state = &crtc->wm.active.g4x;
+		enum pipe pipe = crtc->pipe;
+
+		wm->pipe[pipe] = wm_state->wm;
+		if (crtc->active && wm->cxsr)
+			wm->sr = wm_state->sr;
+		if (crtc->active && wm->hpll_en)
+			wm->hpll = wm_state->hpll;
+	}
+}
+
+static void g4x_program_watermarks(struct drm_i915_private *dev_priv)
+{
+	struct g4x_wm_values *old_wm = &dev_priv->wm.g4x;
+	struct g4x_wm_values new_wm = {};
+
+	g4x_merge_wm(dev_priv, &new_wm);
+
+	if (memcmp(old_wm, &new_wm, sizeof(new_wm)) == 0)
+		return;
+
+	if (is_disabling(old_wm->cxsr, new_wm.cxsr, true))
+		_intel_set_memory_cxsr(dev_priv, false);
+
+	g4x_write_wm_values(dev_priv, &new_wm);
+
+	if (is_enabling(old_wm->cxsr, new_wm.cxsr, true))
+		_intel_set_memory_cxsr(dev_priv, true);
+
+	*old_wm = new_wm;
+}
+
+static void g4x_initial_watermarks(struct intel_atomic_state *state,
+				   struct intel_crtc_state *crtc_state)
+{
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+
+	mutex_lock(&dev_priv->wm.wm_mutex);
+	crtc->wm.active.g4x = crtc_state->wm.g4x.intermediate;
+	g4x_program_watermarks(dev_priv);
+	mutex_unlock(&dev_priv->wm.wm_mutex);
+}
+
+static void g4x_optimize_watermarks(struct intel_atomic_state *state,
+				    struct intel_crtc_state *crtc_state)
+{
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc);
+
+	if (!crtc_state->wm.need_postvbl_update)
+		return;
+
+	mutex_lock(&dev_priv->wm.wm_mutex);
+	intel_crtc->wm.active.g4x = crtc_state->wm.g4x.optimal;
+	g4x_program_watermarks(dev_priv);
+	mutex_unlock(&dev_priv->wm.wm_mutex);
+}
+
 /* latency must be in 0.1us units. */
 static unsigned int vlv_wm_method2(unsigned int pixel_rate,
-				   unsigned int pipe_htotal,
-				   unsigned int horiz_pixels,
+				   unsigned int htotal,
+				   unsigned int width,
 				   unsigned int cpp,
 				   unsigned int latency)
 {
 	unsigned int ret;
 
-	ret = (latency * pixel_rate) / (pipe_htotal * 10000);
-	ret = (ret + 1) * horiz_pixels * cpp;
+	ret = intel_wm_method2(pixel_rate, htotal,
+			       width, cpp, latency);
 	ret = DIV_ROUND_UP(ret, 64);
 
 	return ret;
@@ -1029,17 +1582,15 @@ static uint16_t vlv_compute_wm_level(const struct intel_crtc_state *crtc_state,
 	if (dev_priv->wm.pri_latency[level] == 0)
 		return USHRT_MAX;
 
-	if (!plane_state->base.visible)
+	if (!intel_wm_plane_visible(crtc_state, plane_state))
 		return 0;
 
 	cpp = plane_state->base.fb->format->cpp[0];
 	clock = adjusted_mode->crtc_clock;
 	htotal = adjusted_mode->crtc_htotal;
 	width = crtc_state->pipe_src_w;
-	if (WARN_ON(htotal == 0))
-		htotal = 1;
 
-	if (plane->base.type == DRM_PLANE_TYPE_CURSOR) {
+	if (plane->id == PLANE_CURSOR) {
 		/*
 		 * FIXME the formula gives values that are
 		 * too big for the cursor FIFO, and hence we
@@ -1064,7 +1615,7 @@ static bool vlv_need_sprite0_fifo_workaround(unsigned int active_planes)
 static int vlv_compute_fifo(struct intel_crtc_state *crtc_state)
 {
 	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-	const struct vlv_pipe_wm *raw =
+	const struct g4x_pipe_wm *raw =
 		&crtc_state->wm.vlv.raw[VLV_WM_LEVEL_PM2];
 	struct vlv_fifo_state *fifo_state = &crtc_state->wm.vlv.fifo_state;
 	unsigned int active_planes = crtc_state->active_planes & ~BIT(PLANE_CURSOR);
@@ -1143,18 +1694,13 @@ static int vlv_compute_fifo(struct intel_crtc_state *crtc_state)
 	return 0;
 }
 
-static int vlv_num_wm_levels(struct drm_i915_private *dev_priv)
-{
-	return dev_priv->wm.max_level + 1;
-}
-
 /* mark all levels starting from 'level' as invalid */
 static void vlv_invalidate_wms(struct intel_crtc *crtc,
 			       struct vlv_wm_state *wm_state, int level)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
-	for (; level < vlv_num_wm_levels(dev_priv); level++) {
+	for (; level < intel_wm_num_levels(dev_priv); level++) {
 		enum plane_id plane_id;
 
 		for_each_plane_id_on_crtc(crtc, plane_id)
@@ -1181,11 +1727,11 @@ static bool vlv_raw_plane_wm_set(struct intel_crtc_state *crtc_state,
 				 int level, enum plane_id plane_id, u16 value)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
-	int num_levels = vlv_num_wm_levels(dev_priv);
+	int num_levels = intel_wm_num_levels(dev_priv);
 	bool dirty = false;
 
 	for (; level < num_levels; level++) {
-		struct vlv_pipe_wm *raw = &crtc_state->wm.vlv.raw[level];
+		struct g4x_pipe_wm *raw = &crtc_state->wm.vlv.raw[level];
 
 		dirty |= raw->plane[plane_id] != value;
 		raw->plane[plane_id] = value;
@@ -1194,22 +1740,22 @@ static bool vlv_raw_plane_wm_set(struct intel_crtc_state *crtc_state,
 	return dirty;
 }
 
-static bool vlv_plane_wm_compute(struct intel_crtc_state *crtc_state,
-				 const struct intel_plane_state *plane_state)
+static bool vlv_raw_plane_wm_compute(struct intel_crtc_state *crtc_state,
+				     const struct intel_plane_state *plane_state)
 {
 	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
 	enum plane_id plane_id = plane->id;
-	int num_levels = vlv_num_wm_levels(to_i915(plane->base.dev));
+	int num_levels = intel_wm_num_levels(to_i915(plane->base.dev));
 	int level;
 	bool dirty = false;
 
-	if (!plane_state->base.visible) {
+	if (!intel_wm_plane_visible(crtc_state, plane_state)) {
 		dirty |= vlv_raw_plane_wm_set(crtc_state, 0, plane_id, 0);
 		goto out;
 	}
 
 	for (level = 0; level < num_levels; level++) {
-		struct vlv_pipe_wm *raw = &crtc_state->wm.vlv.raw[level];
+		struct g4x_pipe_wm *raw = &crtc_state->wm.vlv.raw[level];
 		int wm = vlv_compute_wm_level(crtc_state, plane_state, level);
 		int max_wm = plane_id == PLANE_CURSOR ? 63 : 511;
 
@@ -1225,7 +1771,7 @@ static bool vlv_plane_wm_compute(struct intel_crtc_state *crtc_state,
 
 out:
 	if (dirty)
-		DRM_DEBUG_KMS("%s wms: [0]=%d,[1]=%d,[2]=%d\n",
+		DRM_DEBUG_KMS("%s watermarks: PM2=%d, PM5=%d, DDR DVFS=%d\n",
 			      plane->base.name,
 			      crtc_state->wm.vlv.raw[VLV_WM_LEVEL_PM2].plane[plane_id],
 			      crtc_state->wm.vlv.raw[VLV_WM_LEVEL_PM5].plane[plane_id],
@@ -1234,10 +1780,10 @@ out:
 	return dirty;
 }
 
-static bool vlv_plane_wm_is_valid(const struct intel_crtc_state *crtc_state,
-				  enum plane_id plane_id, int level)
+static bool vlv_raw_plane_wm_is_valid(const struct intel_crtc_state *crtc_state,
+				      enum plane_id plane_id, int level)
 {
-	const struct vlv_pipe_wm *raw =
+	const struct g4x_pipe_wm *raw =
 		&crtc_state->wm.vlv.raw[level];
 	const struct vlv_fifo_state *fifo_state =
 		&crtc_state->wm.vlv.fifo_state;
@@ -1245,12 +1791,12 @@ static bool vlv_plane_wm_is_valid(const struct intel_crtc_state *crtc_state,
 	return raw->plane[plane_id] <= fifo_state->plane[plane_id];
 }
 
-static bool vlv_crtc_wm_is_valid(const struct intel_crtc_state *crtc_state, int level)
+static bool vlv_raw_crtc_wm_is_valid(const struct intel_crtc_state *crtc_state, int level)
 {
-	return vlv_plane_wm_is_valid(crtc_state, PLANE_PRIMARY, level) &&
-		vlv_plane_wm_is_valid(crtc_state, PLANE_SPRITE0, level) &&
-		vlv_plane_wm_is_valid(crtc_state, PLANE_SPRITE1, level) &&
-		vlv_plane_wm_is_valid(crtc_state, PLANE_CURSOR, level);
+	return vlv_raw_plane_wm_is_valid(crtc_state, PLANE_PRIMARY, level) &&
+		vlv_raw_plane_wm_is_valid(crtc_state, PLANE_SPRITE0, level) &&
+		vlv_raw_plane_wm_is_valid(crtc_state, PLANE_SPRITE1, level) &&
+		vlv_raw_plane_wm_is_valid(crtc_state, PLANE_CURSOR, level);
 }
 
 static int vlv_compute_pipe_wm(struct intel_crtc_state *crtc_state)
@@ -1279,7 +1825,7 @@ static int vlv_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 		    old_plane_state->base.crtc != &crtc->base)
 			continue;
 
-		if (vlv_plane_wm_compute(crtc_state, plane_state))
+		if (vlv_raw_plane_wm_compute(crtc_state, plane_state))
 			dirty |= BIT(plane->id);
 	}
 
@@ -1313,7 +1859,7 @@ static int vlv_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 	}
 
 	/* initially allow all levels */
-	wm_state->num_levels = vlv_num_wm_levels(dev_priv);
+	wm_state->num_levels = intel_wm_num_levels(dev_priv);
 	/*
 	 * Note that enabling cxsr with no primary/sprite planes
 	 * enabled can wedge the pipe. Hence we only allow cxsr
@@ -1322,10 +1868,10 @@ static int vlv_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 	wm_state->cxsr = crtc->pipe != PIPE_C && num_active_planes == 1;
 
 	for (level = 0; level < wm_state->num_levels; level++) {
-		const struct vlv_pipe_wm *raw = &crtc_state->wm.vlv.raw[level];
+		const struct g4x_pipe_wm *raw = &crtc_state->wm.vlv.raw[level];
 		const int sr_fifo_size = INTEL_INFO(dev_priv)->num_pipes * 512 - 1;
 
-		if (!vlv_crtc_wm_is_valid(crtc_state, level))
+		if (!vlv_raw_crtc_wm_is_valid(crtc_state, level))
 			break;
 
 		for_each_plane_id_on_crtc(crtc, plane_id) {
@@ -1539,16 +2085,6 @@ static void vlv_merge_wm(struct drm_i915_private *dev_priv,
 	}
 }
 
-static bool is_disabling(int old, int new, int threshold)
-{
-	return old >= threshold && new < threshold;
-}
-
-static bool is_enabling(int old, int new, int threshold)
-{
-	return old < threshold && new >= threshold;
-}
-
 static void vlv_program_watermarks(struct drm_i915_private *dev_priv)
 {
 	struct vlv_wm_values *old_wm = &dev_priv->wm.vlv;
@@ -1609,65 +2145,6 @@ static void vlv_optimize_watermarks(struct intel_atomic_state *state,
 	mutex_unlock(&dev_priv->wm.wm_mutex);
 }
 
-#define single_plane_enabled(mask) is_power_of_2(mask)
-
-static void g4x_update_wm(struct intel_crtc *crtc)
-{
-	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
-	static const int sr_latency_ns = 12000;
-	int planea_wm, planeb_wm, cursora_wm, cursorb_wm;
-	int plane_sr, cursor_sr;
-	unsigned int enabled = 0;
-	bool cxsr_enabled;
-
-	if (g4x_compute_wm0(dev_priv, PIPE_A,
-			    &g4x_wm_info, pessimal_latency_ns,
-			    &g4x_cursor_wm_info, pessimal_latency_ns,
-			    &planea_wm, &cursora_wm))
-		enabled |= 1 << PIPE_A;
-
-	if (g4x_compute_wm0(dev_priv, PIPE_B,
-			    &g4x_wm_info, pessimal_latency_ns,
-			    &g4x_cursor_wm_info, pessimal_latency_ns,
-			    &planeb_wm, &cursorb_wm))
-		enabled |= 1 << PIPE_B;
-
-	if (single_plane_enabled(enabled) &&
-	    g4x_compute_srwm(dev_priv, ffs(enabled) - 1,
-			     sr_latency_ns,
-			     &g4x_wm_info,
-			     &g4x_cursor_wm_info,
-			     &plane_sr, &cursor_sr)) {
-		cxsr_enabled = true;
-	} else {
-		cxsr_enabled = false;
-		intel_set_memory_cxsr(dev_priv, false);
-		plane_sr = cursor_sr = 0;
-	}
-
-	DRM_DEBUG_KMS("Setting FIFO watermarks - A: plane=%d, cursor=%d, "
-		      "B: plane=%d, cursor=%d, SR: plane=%d, cursor=%d\n",
-		      planea_wm, cursora_wm,
-		      planeb_wm, cursorb_wm,
-		      plane_sr, cursor_sr);
-
-	I915_WRITE(DSPFW1,
-		   FW_WM(plane_sr, SR) |
-		   FW_WM(cursorb_wm, CURSORB) |
-		   FW_WM(planeb_wm, PLANEB) |
-		   FW_WM(planea_wm, PLANEA));
-	I915_WRITE(DSPFW2,
-		   (I915_READ(DSPFW2) & ~DSPFW_CURSORA_MASK) |
-		   FW_WM(cursora_wm, CURSORA));
-	/* HPLL off in SR has some issues on G4x... disable it */
-	I915_WRITE(DSPFW3,
-		   (I915_READ(DSPFW3) & ~(DSPFW_HPLL_SR_EN | DSPFW_CURSOR_SR_MASK)) |
-		   FW_WM(cursor_sr, CURSOR_SR));
-
-	if (cxsr_enabled)
-		intel_set_memory_cxsr(dev_priv, true);
-}
-
 static void i965_update_wm(struct intel_crtc *unused_crtc)
 {
 	struct drm_i915_private *dev_priv = to_i915(unused_crtc->base.dev);
@@ -1689,14 +2166,10 @@ static void i965_update_wm(struct intel_crtc *unused_crtc)
 		int htotal = adjusted_mode->crtc_htotal;
 		int hdisplay = crtc->config->pipe_src_w;
 		int cpp = fb->format->cpp[0];
-		unsigned long line_time_us;
 		int entries;
 
-		line_time_us = max(htotal * 1000 / clock, 1);
-
-		/* Use ns/us then divide to preserve precision */
-		entries = (((sr_latency_ns / line_time_us) + 1000) / 1000) *
-			cpp * hdisplay;
+		entries = intel_wm_method2(clock, htotal,
+					   hdisplay, cpp, sr_latency_ns / 100);
 		entries = DIV_ROUND_UP(entries, I915_FIFO_LINE_SIZE);
 		srwm = I965_FIFO_SIZE - entries;
 		if (srwm < 0)
@@ -1705,13 +2178,14 @@ static void i965_update_wm(struct intel_crtc *unused_crtc)
 		DRM_DEBUG_KMS("self-refresh entries: %d, wm: %d\n",
 			      entries, srwm);
 
-		entries = (((sr_latency_ns / line_time_us) + 1000) / 1000) *
-			cpp * crtc->base.cursor->state->crtc_w;
+		entries = intel_wm_method2(clock, htotal,
+					   crtc->base.cursor->state->crtc_w, 4,
+					   sr_latency_ns / 100);
 		entries = DIV_ROUND_UP(entries,
-					  i965_cursor_wm_info.cacheline_size);
-		cursor_sr = i965_cursor_wm_info.fifo_size -
-			(entries + i965_cursor_wm_info.guard_size);
+				       i965_cursor_wm_info.cacheline_size) +
+			i965_cursor_wm_info.guard_size;
 
+		cursor_sr = i965_cursor_wm_info.fifo_size - entries;
 		if (cursor_sr > i965_cursor_wm_info.max_wm)
 			cursor_sr = i965_cursor_wm_info.max_wm;
 
@@ -1848,7 +2322,6 @@ static void i9xx_update_wm(struct intel_crtc *unused_crtc)
 		int htotal = adjusted_mode->crtc_htotal;
 		int hdisplay = enabled->config->pipe_src_w;
 		int cpp;
-		unsigned long line_time_us;
 		int entries;
 
 		if (IS_I915GM(dev_priv) || IS_I945GM(dev_priv))
@@ -1856,11 +2329,8 @@ static void i9xx_update_wm(struct intel_crtc *unused_crtc)
 		else
 			cpp = fb->format->cpp[0];
 
-		line_time_us = max(htotal * 1000 / clock, 1);
-
-		/* Use ns/us then divide to preserve precision */
-		entries = (((sr_latency_ns / line_time_us) + 1000) / 1000) *
-			cpp * hdisplay;
+		entries = intel_wm_method2(clock, htotal, hdisplay, cpp,
+					   sr_latency_ns / 100);
 		entries = DIV_ROUND_UP(entries, wm_info->cacheline_size);
 		DRM_DEBUG_KMS("self-refresh entries: %d\n", entries);
 		srwm = wm_info->fifo_size - entries;
@@ -1917,34 +2387,31 @@ static void i845_update_wm(struct intel_crtc *unused_crtc)
 }
 
 /* latency must be in 0.1us units. */
-static uint32_t ilk_wm_method1(uint32_t pixel_rate, uint8_t cpp, uint32_t latency)
+static unsigned int ilk_wm_method1(unsigned int pixel_rate,
+				   unsigned int cpp,
+				   unsigned int latency)
 {
-	uint64_t ret;
-
-	if (WARN(latency == 0, "Latency value missing\n"))
-		return UINT_MAX;
+	unsigned int ret;
 
-	ret = (uint64_t) pixel_rate * cpp * latency;
-	ret = DIV_ROUND_UP_ULL(ret, 64 * 10000) + 2;
+	ret = intel_wm_method1(pixel_rate, cpp, latency);
+	ret = DIV_ROUND_UP(ret, 64) + 2;
 
 	return ret;
 }
 
 /* latency must be in 0.1us units. */
-static uint32_t ilk_wm_method2(uint32_t pixel_rate, uint32_t pipe_htotal,
-			       uint32_t horiz_pixels, uint8_t cpp,
-			       uint32_t latency)
+static unsigned int ilk_wm_method2(unsigned int pixel_rate,
+				   unsigned int htotal,
+				   unsigned int width,
+				   unsigned int cpp,
+				   unsigned int latency)
 {
-	uint32_t ret;
-
-	if (WARN(latency == 0, "Latency value missing\n"))
-		return UINT_MAX;
-	if (WARN_ON(!pipe_htotal))
-		return UINT_MAX;
+	unsigned int ret;
 
-	ret = (latency * pixel_rate) / (pipe_htotal * 10000);
-	ret = (ret + 1) * horiz_pixels * cpp;
+	ret = intel_wm_method2(pixel_rate, htotal,
+			       width, cpp, latency);
 	ret = DIV_ROUND_UP(ret, 64) + 2;
+
 	return ret;
 }
 
@@ -3082,7 +3549,7 @@ static bool skl_needs_memory_bw_wa(struct intel_atomic_state *state)
 static bool
 intel_has_sagv(struct drm_i915_private *dev_priv)
 {
-	if (IS_KABYLAKE(dev_priv))
+	if (IS_KABYLAKE(dev_priv) || IS_COFFEELAKE(dev_priv))
 		return true;
 
 	if (IS_SKYLAKE(dev_priv) &&
@@ -3360,38 +3827,137 @@ void skl_ddb_get_hw_state(struct drm_i915_private *dev_priv,
  * Return value is provided in 16.16 fixed point form to retain fractional part.
  * Caller should take care of dividing & rounding off the value.
  */
-static uint32_t
+static uint_fixed_16_16_t
 skl_plane_downscale_amount(const struct intel_crtc_state *cstate,
 			   const struct intel_plane_state *pstate)
 {
 	struct intel_plane *plane = to_intel_plane(pstate->base.plane);
-	uint32_t downscale_h, downscale_w;
 	uint32_t src_w, src_h, dst_w, dst_h;
+	uint_fixed_16_16_t fp_w_ratio, fp_h_ratio;
+	uint_fixed_16_16_t downscale_h, downscale_w;
 
 	if (WARN_ON(!intel_wm_plane_visible(cstate, pstate)))
-		return DRM_PLANE_HELPER_NO_SCALING;
+		return u32_to_fixed_16_16(0);
 
 	/* n.b., src is 16.16 fixed point, dst is whole integer */
 	if (plane->id == PLANE_CURSOR) {
-		src_w = pstate->base.src_w;
-		src_h = pstate->base.src_h;
+		/*
+		 * Cursors only support 0/180 degree rotation,
+		 * hence no need to account for rotation here.
+		 */
+		src_w = pstate->base.src_w >> 16;
+		src_h = pstate->base.src_h >> 16;
 		dst_w = pstate->base.crtc_w;
 		dst_h = pstate->base.crtc_h;
 	} else {
-		src_w = drm_rect_width(&pstate->base.src);
-		src_h = drm_rect_height(&pstate->base.src);
+		/*
+		 * Src coordinates are already rotated by 270 degrees for
+		 * the 90/270 degree plane rotation cases (to match the
+		 * GTT mapping), hence no need to account for rotation here.
+		 */
+		src_w = drm_rect_width(&pstate->base.src) >> 16;
+		src_h = drm_rect_height(&pstate->base.src) >> 16;
 		dst_w = drm_rect_width(&pstate->base.dst);
 		dst_h = drm_rect_height(&pstate->base.dst);
 	}
 
-	if (drm_rotation_90_or_270(pstate->base.rotation))
-		swap(dst_w, dst_h);
+	fp_w_ratio = fixed_16_16_div(src_w, dst_w);
+	fp_h_ratio = fixed_16_16_div(src_h, dst_h);
+	downscale_w = max_fixed_16_16(fp_w_ratio, u32_to_fixed_16_16(1));
+	downscale_h = max_fixed_16_16(fp_h_ratio, u32_to_fixed_16_16(1));
+
+	return mul_fixed16(downscale_w, downscale_h);
+}
+
+static uint_fixed_16_16_t
+skl_pipe_downscale_amount(const struct intel_crtc_state *crtc_state)
+{
+	uint_fixed_16_16_t pipe_downscale = u32_to_fixed_16_16(1);
+
+	if (!crtc_state->base.enable)
+		return pipe_downscale;
 
-	downscale_h = max(src_h / dst_h, (uint32_t)DRM_PLANE_HELPER_NO_SCALING);
-	downscale_w = max(src_w / dst_w, (uint32_t)DRM_PLANE_HELPER_NO_SCALING);
+	if (crtc_state->pch_pfit.enabled) {
+		uint32_t src_w, src_h, dst_w, dst_h;
+		uint32_t pfit_size = crtc_state->pch_pfit.size;
+		uint_fixed_16_16_t fp_w_ratio, fp_h_ratio;
+		uint_fixed_16_16_t downscale_h, downscale_w;
 
-	/* Provide result in 16.16 fixed point */
-	return (uint64_t)downscale_w * downscale_h >> 16;
+		src_w = crtc_state->pipe_src_w;
+		src_h = crtc_state->pipe_src_h;
+		dst_w = pfit_size >> 16;
+		dst_h = pfit_size & 0xffff;
+
+		if (!dst_w || !dst_h)
+			return pipe_downscale;
+
+		fp_w_ratio = fixed_16_16_div(src_w, dst_w);
+		fp_h_ratio = fixed_16_16_div(src_h, dst_h);
+		downscale_w = max_fixed_16_16(fp_w_ratio, u32_to_fixed_16_16(1));
+		downscale_h = max_fixed_16_16(fp_h_ratio, u32_to_fixed_16_16(1));
+
+		pipe_downscale = mul_fixed16(downscale_w, downscale_h);
+	}
+
+	return pipe_downscale;
+}
+
+int skl_check_pipe_max_pixel_rate(struct intel_crtc *intel_crtc,
+				  struct intel_crtc_state *cstate)
+{
+	struct drm_crtc_state *crtc_state = &cstate->base;
+	struct drm_atomic_state *state = crtc_state->state;
+	struct drm_plane *plane;
+	const struct drm_plane_state *pstate;
+	struct intel_plane_state *intel_pstate;
+	int crtc_clock, dotclk;
+	uint32_t pipe_max_pixel_rate;
+	uint_fixed_16_16_t pipe_downscale;
+	uint_fixed_16_16_t max_downscale = u32_to_fixed_16_16(1);
+
+	if (!cstate->base.enable)
+		return 0;
+
+	drm_atomic_crtc_state_for_each_plane_state(plane, pstate, crtc_state) {
+		uint_fixed_16_16_t plane_downscale;
+		uint_fixed_16_16_t fp_9_div_8 = fixed_16_16_div(9, 8);
+		int bpp;
+
+		if (!intel_wm_plane_visible(cstate,
+					    to_intel_plane_state(pstate)))
+			continue;
+
+		if (WARN_ON(!pstate->fb))
+			return -EINVAL;
+
+		intel_pstate = to_intel_plane_state(pstate);
+		plane_downscale = skl_plane_downscale_amount(cstate,
+							     intel_pstate);
+		bpp = pstate->fb->format->cpp[0] * 8;
+		if (bpp == 64)
+			plane_downscale = mul_fixed16(plane_downscale,
+						      fp_9_div_8);
+
+		max_downscale = max_fixed_16_16(plane_downscale, max_downscale);
+	}
+	pipe_downscale = skl_pipe_downscale_amount(cstate);
+
+	pipe_downscale = mul_fixed16(pipe_downscale, max_downscale);
+
+	crtc_clock = crtc_state->adjusted_mode.crtc_clock;
+	dotclk = to_intel_atomic_state(state)->cdclk.logical.cdclk;
+
+	if (IS_GEMINILAKE(to_i915(intel_crtc->base.dev)))
+		dotclk *= 2;
+
+	pipe_max_pixel_rate = div_round_up_u32_fixed16(dotclk, pipe_downscale);
+
+	if (pipe_max_pixel_rate < crtc_clock) {
+		DRM_DEBUG_KMS("Max supported pixel clock with scaling exceeded\n");
+		return -EINVAL;
+	}
+
+	return 0;
 }
 
 static unsigned int
@@ -3401,10 +3967,11 @@ skl_plane_relative_data_rate(const struct intel_crtc_state *cstate,
 {
 	struct intel_plane *plane = to_intel_plane(pstate->plane);
 	struct intel_plane_state *intel_pstate = to_intel_plane_state(pstate);
-	uint32_t down_scale_amount, data_rate;
+	uint32_t data_rate;
 	uint32_t width = 0, height = 0;
 	struct drm_framebuffer *fb;
 	u32 format;
+	uint_fixed_16_16_t down_scale_amount;
 
 	if (!intel_pstate->base.visible)
 		return 0;
@@ -3417,12 +3984,14 @@ skl_plane_relative_data_rate(const struct intel_crtc_state *cstate,
 	if (y && format != DRM_FORMAT_NV12)
 		return 0;
 
+	/*
+	 * Src coordinates are already rotated by 270 degrees for
+	 * the 90/270 degree plane rotation cases (to match the
+	 * GTT mapping), hence no need to account for rotation here.
+	 */
 	width = drm_rect_width(&intel_pstate->base.src) >> 16;
 	height = drm_rect_height(&intel_pstate->base.src) >> 16;
 
-	if (drm_rotation_90_or_270(pstate->rotation))
-		swap(width, height);
-
 	/* for planar format */
 	if (format == DRM_FORMAT_NV12) {
 		if (y)  /* y-plane data rate */
@@ -3438,7 +4007,7 @@ skl_plane_relative_data_rate(const struct intel_crtc_state *cstate,
 
 	down_scale_amount = skl_plane_downscale_amount(cstate, intel_pstate);
 
-	return (uint64_t)data_rate * down_scale_amount >> 16;
+	return mul_round_up_u32_fixed16(data_rate, down_scale_amount);
 }
 
 /*
@@ -3505,12 +4074,14 @@ skl_ddb_min_alloc(const struct drm_plane_state *pstate,
 	    fb->modifier != I915_FORMAT_MOD_Yf_TILED)
 		return 8;
 
+	/*
+	 * Src coordinates are already rotated by 270 degrees for
+	 * the 90/270 degree plane rotation cases (to match the
+	 * GTT mapping), hence no need to account for rotation here.
+	 */
 	src_w = drm_rect_width(&intel_pstate->base.src) >> 16;
 	src_h = drm_rect_height(&intel_pstate->base.src) >> 16;
 
-	if (drm_rotation_90_or_270(pstate->rotation))
-		swap(src_w, src_h);
-
 	/* Halve UV plane width and height for NV12 */
 	if (fb->format->format == DRM_FORMAT_NV12 && !y) {
 		src_w /= 2;
@@ -3569,13 +4140,41 @@ skl_ddb_calc_min(const struct intel_crtc_state *cstate, int num_active,
 	minimum[PLANE_CURSOR] = skl_cursor_allocation(num_active);
 }
 
+static void
+skl_enable_plane_wm_levels(const struct drm_i915_private *dev_priv,
+			   uint16_t plane_ddb,
+			   uint16_t max_level,
+			   struct skl_plane_wm *wm)
+{
+	int level;
+	/*
+	 * Now enable all levels in WM structure which can be enabled
+	 * using current DDB allocation
+	 */
+	for (level = ilk_wm_max_level(dev_priv); level >= 0; level--) {
+		struct skl_wm_level *level_wm = &wm->wm[level];
+
+		if (level > max_level || level_wm->plane_res_b == 0
+				      || level_wm->plane_res_l >= 31
+				      || level_wm->plane_res_b >= plane_ddb) {
+			level_wm->plane_en = false;
+			level_wm->plane_res_b = 0;
+			level_wm->plane_res_l = 0;
+		} else {
+			level_wm->plane_en = true;
+		}
+	}
+}
+
 static int
 skl_allocate_pipe_ddb(struct intel_crtc_state *cstate,
+		      struct skl_pipe_wm *pipe_wm,
 		      struct skl_ddb_allocation *ddb /* out */)
 {
 	struct drm_atomic_state *state = cstate->base.state;
 	struct drm_crtc *crtc = cstate->base.crtc;
 	struct drm_device *dev = crtc->dev;
+	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	enum pipe pipe = intel_crtc->pipe;
 	struct skl_ddb_entry *alloc = &cstate->wm.skl.ddb;
@@ -3587,6 +4186,10 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *cstate,
 	int num_active;
 	unsigned plane_data_rate[I915_MAX_PLANES] = {};
 	unsigned plane_y_data_rate[I915_MAX_PLANES] = {};
+	uint16_t total_min_blocks = 0;
+	uint16_t total_level_ddb;
+	uint16_t plane_blocks = 0;
+	int max_level, level;
 
 	/* Clear the partitioning for disabled planes. */
 	memset(ddb->plane[pipe], 0, sizeof(ddb->plane[pipe]));
@@ -3602,10 +4205,8 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *cstate,
 
 	skl_ddb_get_pipe_allocation_limits(dev, cstate, alloc, &num_active);
 	alloc_size = skl_ddb_entry_size(alloc);
-	if (alloc_size == 0) {
-		memset(ddb->plane[pipe], 0, sizeof(ddb->plane[pipe]));
+	if (alloc_size == 0)
 		return 0;
-	}
 
 	skl_ddb_calc_min(cstate, num_active, minimum, y_minimum);
 
@@ -3616,13 +4217,59 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *cstate,
 	 */
 
 	for_each_plane_id_on_crtc(intel_crtc, plane_id) {
-		alloc_size -= minimum[plane_id];
-		alloc_size -= y_minimum[plane_id];
+		total_min_blocks += minimum[plane_id];
+		total_min_blocks += y_minimum[plane_id];
 	}
 
-	ddb->plane[pipe][PLANE_CURSOR].start = alloc->end - minimum[PLANE_CURSOR];
+	if (total_min_blocks > alloc_size) {
+		DRM_DEBUG_KMS("Requested display configuration exceeds system DDB limitations");
+		DRM_DEBUG_KMS("minimum required %d/%d\n", total_min_blocks,
+							alloc_size);
+		return -EINVAL;
+	}
+
+	alloc_size -= minimum[PLANE_CURSOR];
+	ddb->plane[pipe][PLANE_CURSOR].start = alloc->end -
+							minimum[PLANE_CURSOR];
 	ddb->plane[pipe][PLANE_CURSOR].end = alloc->end;
 
+	for (level = ilk_wm_max_level(dev_priv); level >= 0; level--) {
+		total_level_ddb = 0;
+		for_each_plane_id_on_crtc(intel_crtc, plane_id) {
+			/*
+			 * TODO: We should calculate watermark values for Y/UV
+			 * plane both in case of NV12 format and use both values
+			 * for ddb calculation. NV12 is disabled as of now, So
+			 * using only single/UV plane value here.
+			 */
+			struct skl_plane_wm *wm = &pipe_wm->planes[plane_id];
+			uint16_t plane_res_b = wm->wm[level].plane_res_b;
+			uint16_t min = minimum[plane_id] + y_minimum[plane_id];
+
+			if (plane_id == PLANE_CURSOR)
+				continue;
+
+			total_level_ddb += max(plane_res_b, min);
+		}
+
+		/*
+		 * If This level can successfully be enabled with the
+		 * pipe's current DDB allocation, then all lower levels are
+		 * guaranteed to succeed as well.
+		 */
+		if (total_level_ddb <= alloc_size)
+			break;
+	}
+
+	if ((level < 0) || (total_min_blocks > alloc_size)) {
+		DRM_DEBUG_KMS("Requested display configuration exceeds system DDB limitations");
+		DRM_DEBUG_KMS("minimum required %d/%d\n", (level < 0) ?
+				total_level_ddb : total_min_blocks, alloc_size);
+		return -EINVAL;
+	}
+	max_level = level;
+	alloc_size -= total_level_ddb;
+
 	/*
 	 * 2. Distribute the remaining space in proportion to the amount of
 	 * data each plane needs to fetch from memory.
@@ -3632,13 +4279,24 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *cstate,
 	total_data_rate = skl_get_total_relative_data_rate(cstate,
 							   plane_data_rate,
 							   plane_y_data_rate);
+	/*
+	 * PLANE_CURSOR data rate is not included in total_data_rate.
+	 * If only cursor plane is enabled we have to enable its WM levels
+	 * explicitly before returning. Cursor has fixed ddb allocation,
+	 * So it's ok to always check cursor WM enabling before return.
+	 */
+	plane_blocks = skl_ddb_entry_size(&ddb->plane[pipe][PLANE_CURSOR]);
+	skl_enable_plane_wm_levels(dev_priv, plane_blocks, max_level,
+				   &pipe_wm->planes[PLANE_CURSOR]);
 	if (total_data_rate == 0)
 		return 0;
 
 	start = alloc->start;
 	for_each_plane_id_on_crtc(intel_crtc, plane_id) {
 		unsigned int data_rate, y_data_rate;
-		uint16_t plane_blocks, y_plane_blocks = 0;
+		uint16_t plane_blocks = 0, y_plane_blocks = 0;
+		struct skl_plane_wm *wm = &pipe_wm->planes[plane_id];
+		uint16_t plane_res_b = wm->wm[max_level].plane_res_b;
 
 		if (plane_id == PLANE_CURSOR)
 			continue;
@@ -3650,33 +4308,36 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *cstate,
 		 * promote the expression to 64 bits to avoid overflowing, the
 		 * result is < available as data_rate / total_data_rate < 1
 		 */
-		plane_blocks = minimum[plane_id];
-		plane_blocks += div_u64((uint64_t)alloc_size * data_rate,
-					total_data_rate);
 
 		/* Leave disabled planes at (0,0) */
 		if (data_rate) {
+			plane_blocks = max(minimum[plane_id], plane_res_b);
+			plane_blocks += div_u64((uint64_t)alloc_size *
+					data_rate, total_data_rate);
 			ddb->plane[pipe][plane_id].start = start;
 			ddb->plane[pipe][plane_id].end = start + plane_blocks;
+			start += plane_blocks;
 		}
 
-		start += plane_blocks;
-
 		/*
 		 * allocation for y_plane part of planar format:
+		 * TODO: Once we start calculating watermark values for Y/UV
+		 * plane both consider it for initial allowed wm blocks.
 		 */
 		y_data_rate = plane_y_data_rate[plane_id];
 
-		y_plane_blocks = y_minimum[plane_id];
-		y_plane_blocks += div_u64((uint64_t)alloc_size * y_data_rate,
-					total_data_rate);
-
 		if (y_data_rate) {
+			y_plane_blocks = y_minimum[plane_id];
+			y_plane_blocks += div_u64((uint64_t)alloc_size *
+					y_data_rate, total_data_rate);
 			ddb->y_plane[pipe][plane_id].start = start;
 			ddb->y_plane[pipe][plane_id].end = start + y_plane_blocks;
+			start += y_plane_blocks;
 		}
-
-		start += y_plane_blocks;
+		skl_enable_plane_wm_levels(dev_priv,
+					   plane_blocks,
+					   max_level,
+					   wm);
 	}
 
 	return 0;
@@ -3698,7 +4359,7 @@ static uint_fixed_16_16_t skl_wm_method1(uint32_t pixel_rate, uint8_t cpp,
 		return FP_16_16_MAX;
 
 	wm_intermediate_val = latency * pixel_rate * cpp;
-	ret = fixed_16_16_div_round_up_u64(wm_intermediate_val, 1000 * 512);
+	ret = fixed_16_16_div_u64(wm_intermediate_val, 1000 * 512);
 	return ret;
 }
 
@@ -3720,12 +4381,33 @@ static uint_fixed_16_16_t skl_wm_method2(uint32_t pixel_rate,
 	return ret;
 }
 
-static uint32_t skl_adjusted_plane_pixel_rate(const struct intel_crtc_state *cstate,
-					      struct intel_plane_state *pstate)
+static uint_fixed_16_16_t
+intel_get_linetime_us(struct intel_crtc_state *cstate)
+{
+	uint32_t pixel_rate;
+	uint32_t crtc_htotal;
+	uint_fixed_16_16_t linetime_us;
+
+	if (!cstate->base.active)
+		return u32_to_fixed_16_16(0);
+
+	pixel_rate = cstate->pixel_rate;
+
+	if (WARN_ON(pixel_rate == 0))
+		return u32_to_fixed_16_16(0);
+
+	crtc_htotal = cstate->base.adjusted_mode.crtc_htotal;
+	linetime_us = fixed_16_16_div_u64(crtc_htotal * 1000, pixel_rate);
+
+	return linetime_us;
+}
+
+static uint32_t
+skl_adjusted_plane_pixel_rate(const struct intel_crtc_state *cstate,
+			      const struct intel_plane_state *pstate)
 {
 	uint64_t adjusted_pixel_rate;
-	uint64_t downscale_amount;
-	uint64_t pixel_rate;
+	uint_fixed_16_16_t downscale_amount;
 
 	/* Shouldn't reach here on disabled planes... */
 	if (WARN_ON(!intel_wm_plane_visible(cstate, pstate)))
@@ -3738,24 +4420,20 @@ static uint32_t skl_adjusted_plane_pixel_rate(const struct intel_crtc_state *cst
 	adjusted_pixel_rate = cstate->pixel_rate;
 	downscale_amount = skl_plane_downscale_amount(cstate, pstate);
 
-	pixel_rate = adjusted_pixel_rate * downscale_amount >> 16;
-	WARN_ON(pixel_rate != clamp_t(uint32_t, pixel_rate, 0, ~0));
-
-	return pixel_rate;
+	return mul_round_up_u32_fixed16(adjusted_pixel_rate,
+					    downscale_amount);
 }
 
 static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv,
 				struct intel_crtc_state *cstate,
-				struct intel_plane_state *intel_pstate,
-				uint16_t ddb_allocation,
+				const struct intel_plane_state *intel_pstate,
 				int level,
 				uint16_t *out_blocks, /* out */
-				uint8_t *out_lines, /* out */
-				bool *enabled /* out */)
+				uint8_t *out_lines /* out */)
 {
 	struct intel_plane *plane = to_intel_plane(intel_pstate->base.plane);
-	struct drm_plane_state *pstate = &intel_pstate->base;
-	struct drm_framebuffer *fb = pstate->fb;
+	const struct drm_plane_state *pstate = &intel_pstate->base;
+	const struct drm_framebuffer *fb = pstate->fb;
 	uint32_t latency = dev_priv->wm.skl_latency[level];
 	uint_fixed_16_16_t method1, method2;
 	uint_fixed_16_16_t plane_blocks_per_line;
@@ -3774,17 +4452,16 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv,
 	bool y_tiled, x_tiled;
 
 	if (latency == 0 ||
-	    !intel_wm_plane_visible(cstate, intel_pstate)) {
-		*enabled = false;
+	    !intel_wm_plane_visible(cstate, intel_pstate))
 		return 0;
-	}
 
 	y_tiled = fb->modifier == I915_FORMAT_MOD_Y_TILED ||
 		  fb->modifier == I915_FORMAT_MOD_Yf_TILED;
 	x_tiled = fb->modifier == I915_FORMAT_MOD_X_TILED;
 
-	/* Display WA #1141: kbl. */
-	if (IS_KABYLAKE(dev_priv) && dev_priv->ipc_enabled)
+	/* Display WA #1141: kbl,cfl */
+	if ((IS_KABYLAKE(dev_priv) || IS_COFFEELAKE(dev_priv)) &&
+	    dev_priv->ipc_enabled)
 		latency += 4;
 
 	if (apply_memory_bw_wa && x_tiled)
@@ -3794,13 +4471,15 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv,
 		width = intel_pstate->base.crtc_w;
 		height = intel_pstate->base.crtc_h;
 	} else {
+		/*
+		 * Src coordinates are already rotated by 270 degrees for
+		 * the 90/270 degree plane rotation cases (to match the
+		 * GTT mapping), hence no need to account for rotation here.
+		 */
 		width = drm_rect_width(&intel_pstate->base.src) >> 16;
 		height = drm_rect_height(&intel_pstate->base.src) >> 16;
 	}
 
-	if (drm_rotation_90_or_270(pstate->rotation))
-		swap(width, height);
-
 	cpp = fb->format->cpp[0];
 	plane_pixel_rate = skl_adjusted_plane_pixel_rate(cstate, intel_pstate);
 
@@ -3834,8 +4513,8 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv,
 	if (y_tiled) {
 		interm_pbpl = DIV_ROUND_UP(plane_bytes_per_line *
 					   y_min_scanlines, 512);
-		plane_blocks_per_line =
-		      fixed_16_16_div_round_up(interm_pbpl, y_min_scanlines);
+		plane_blocks_per_line = fixed_16_16_div(interm_pbpl,
+							y_min_scanlines);
 	} else if (x_tiled) {
 		interm_pbpl = DIV_ROUND_UP(plane_bytes_per_line, 512);
 		plane_blocks_per_line = u32_to_fixed_16_16(interm_pbpl);
@@ -3856,19 +4535,22 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv,
 	if (y_tiled) {
 		selected_result = max_fixed_16_16(method2, y_tile_minimum);
 	} else {
+		uint32_t linetime_us;
+
+		linetime_us = fixed_16_16_to_u32_round_up(
+				intel_get_linetime_us(cstate));
 		if ((cpp * cstate->base.adjusted_mode.crtc_htotal / 512 < 1) &&
 		    (plane_bytes_per_line / 512 < 1))
 			selected_result = method2;
-		else if ((ddb_allocation /
-			fixed_16_16_to_u32_round_up(plane_blocks_per_line)) >= 1)
+		else if (latency >= linetime_us)
 			selected_result = min_fixed_16_16(method1, method2);
 		else
 			selected_result = method1;
 	}
 
 	res_blocks = fixed_16_16_to_u32_round_up(selected_result) + 1;
-	res_lines = DIV_ROUND_UP(selected_result.val,
-				 plane_blocks_per_line.val);
+	res_lines = div_round_up_fixed16(selected_result,
+					 plane_blocks_per_line);
 
 	if (level >= 1 && level <= 7) {
 		if (y_tiled) {
@@ -3879,82 +4561,45 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv,
 		}
 	}
 
-	if (res_blocks >= ddb_allocation || res_lines > 31) {
-		*enabled = false;
-
-		/*
-		 * If there are no valid level 0 watermarks, then we can't
-		 * support this display configuration.
-		 */
-		if (level) {
-			return 0;
-		} else {
-			struct drm_plane *plane = pstate->plane;
+	if (res_lines >= 31 && level == 0) {
+		struct drm_plane *plane = pstate->plane;
 
-			DRM_DEBUG_KMS("Requested display configuration exceeds system watermark limitations\n");
-			DRM_DEBUG_KMS("[PLANE:%d:%s] blocks required = %u/%u, lines required = %u/31\n",
-				      plane->base.id, plane->name,
-				      res_blocks, ddb_allocation, res_lines);
-			return -EINVAL;
-		}
+		DRM_DEBUG_KMS("Requested display configuration exceeds system watermark limitations\n");
+		DRM_DEBUG_KMS("[PLANE:%d:%s] lines required = %u/31\n",
+				plane->base.id, plane->name, res_lines);
+		return -EINVAL;
 	}
 
 	*out_blocks = res_blocks;
 	*out_lines = res_lines;
-	*enabled = true;
 
 	return 0;
 }
 
 static int
-skl_compute_wm_level(const struct drm_i915_private *dev_priv,
-		     struct skl_ddb_allocation *ddb,
-		     struct intel_crtc_state *cstate,
-		     struct intel_plane *intel_plane,
-		     int level,
-		     struct skl_wm_level *result)
+skl_compute_wm_levels(const struct drm_i915_private *dev_priv,
+		      struct intel_crtc_state *cstate,
+		      const struct intel_plane_state *intel_pstate,
+		      struct skl_plane_wm *wm)
 {
-	struct drm_atomic_state *state = cstate->base.state;
-	struct intel_crtc *intel_crtc = to_intel_crtc(cstate->base.crtc);
-	struct drm_plane *plane = &intel_plane->base;
-	struct intel_plane_state *intel_pstate = NULL;
-	uint16_t ddb_blocks;
-	enum pipe pipe = intel_crtc->pipe;
+	int level, max_level = ilk_wm_max_level(dev_priv);
 	int ret;
 
-	if (state)
-		intel_pstate =
-			intel_atomic_get_existing_plane_state(state,
-							      intel_plane);
-
-	/*
-	 * Note: If we start supporting multiple pending atomic commits against
-	 * the same planes/CRTC's in the future, plane->state will no longer be
-	 * the correct pre-state to use for the calculations here and we'll
-	 * need to change where we get the 'unchanged' plane data from.
-	 *
-	 * For now this is fine because we only allow one queued commit against
-	 * a CRTC.  Even if the plane isn't modified by this transaction and we
-	 * don't have a plane lock, we still have the CRTC's lock, so we know
-	 * that no other transactions are racing with us to update it.
-	 */
-	if (!intel_pstate)
-		intel_pstate = to_intel_plane_state(plane->state);
-
-	WARN_ON(!intel_pstate->base.fb);
-
-	ddb_blocks = skl_ddb_entry_size(&ddb->plane[pipe][intel_plane->id]);
+	if (WARN_ON(!intel_pstate->base.fb))
+		return -EINVAL;
 
-	ret = skl_compute_plane_wm(dev_priv,
-				   cstate,
-				   intel_pstate,
-				   ddb_blocks,
-				   level,
-				   &result->plane_res_b,
-				   &result->plane_res_l,
-				   &result->plane_en);
-	if (ret)
-		return ret;
+	for (level = 0; level <= max_level; level++) {
+		struct skl_wm_level *result = &wm->wm[level];
+
+		ret = skl_compute_plane_wm(dev_priv,
+					   cstate,
+					   intel_pstate,
+					   level,
+					   &result->plane_res_b,
+					   &result->plane_res_l);
+		if (ret)
+			return ret;
+	}
 
 	return 0;
 }
@@ -3964,19 +4609,16 @@ skl_compute_linetime_wm(struct intel_crtc_state *cstate)
 {
 	struct drm_atomic_state *state = cstate->base.state;
 	struct drm_i915_private *dev_priv = to_i915(state->dev);
-	uint32_t pixel_rate;
+	uint_fixed_16_16_t linetime_us;
 	uint32_t linetime_wm;
 
-	if (!cstate->base.active)
-		return 0;
-
-	pixel_rate = cstate->pixel_rate;
+	linetime_us = intel_get_linetime_us(cstate);
 
-	if (WARN_ON(pixel_rate == 0))
+	if (is_fixed16_zero(linetime_us))
 		return 0;
 
-	linetime_wm = DIV_ROUND_UP(8 * cstate->base.adjusted_mode.crtc_htotal *
-				   1000, pixel_rate);
+	linetime_wm = fixed_16_16_to_u32_round_up(mul_u32_fixed_16_16(8,
+				linetime_us));
 
 	/* Display WA #1135: bxt. */
 	if (IS_BROXTON(dev_priv) && dev_priv->ipc_enabled)
@@ -4000,10 +4642,11 @@ static int skl_build_pipe_wm(struct intel_crtc_state *cstate,
 			     struct skl_pipe_wm *pipe_wm)
 {
 	struct drm_device *dev = cstate->base.crtc->dev;
+	struct drm_crtc_state *crtc_state = &cstate->base;
 	const struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_plane *intel_plane;
+	struct drm_plane *plane;
+	const struct drm_plane_state *pstate;
 	struct skl_plane_wm *wm;
-	int level, max_level = ilk_wm_max_level(dev_priv);
 	int ret;
 
 	/*
@@ -4012,18 +4655,16 @@ static int skl_build_pipe_wm(struct intel_crtc_state *cstate,
 	 */
 	memset(pipe_wm->planes, 0, sizeof(pipe_wm->planes));
 
-	for_each_intel_plane_mask(&dev_priv->drm,
-				  intel_plane,
-				  cstate->base.plane_mask) {
-		wm = &pipe_wm->planes[intel_plane->id];
+	drm_atomic_crtc_state_for_each_plane_state(plane, pstate, crtc_state) {
+		const struct intel_plane_state *intel_pstate =
+						to_intel_plane_state(pstate);
+		enum plane_id plane_id = to_intel_plane(plane)->id;
+
+		wm = &pipe_wm->planes[plane_id];
 
-		for (level = 0; level <= max_level; level++) {
-			ret = skl_compute_wm_level(dev_priv, ddb, cstate,
-						   intel_plane, level,
-						   &wm->wm[level]);
-			if (ret)
-				return ret;
-		}
+		ret = skl_compute_wm_levels(dev_priv, cstate, intel_pstate, wm);
+		if (ret)
+			return ret;
 		skl_compute_transition_wm(cstate, &wm->trans_wm);
 	}
 	pipe_wm->linetime = skl_compute_linetime_wm(cstate);
@@ -4134,6 +4775,45 @@ bool skl_ddb_allocation_overlaps(const struct skl_ddb_entry **entries,
 	return false;
 }
 
+static int
+skl_ddb_add_affected_planes(struct intel_crtc_state *cstate,
+			    const struct skl_pipe_wm *old_pipe_wm,
+			    const struct skl_pipe_wm *pipe_wm)
+{
+	struct drm_atomic_state *state = cstate->base.state;
+	struct drm_device *dev = state->dev;
+	struct drm_crtc *crtc = cstate->base.crtc;
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
+	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct intel_atomic_state *intel_state = to_intel_atomic_state(state);
+	struct skl_ddb_allocation *new_ddb = &intel_state->wm_results.ddb;
+	struct skl_ddb_allocation *cur_ddb = &dev_priv->wm.skl_hw.ddb;
+	struct drm_plane_state *plane_state;
+	struct drm_plane *plane;
+	enum pipe pipe = intel_crtc->pipe;
+
+	WARN_ON(!drm_atomic_get_existing_crtc_state(state, crtc));
+
+	drm_for_each_plane_mask(plane, dev, cstate->base.plane_mask) {
+		enum plane_id plane_id = to_intel_plane(plane)->id;
+		const struct skl_plane_wm *wm = &pipe_wm->planes[plane_id];
+		const struct skl_plane_wm *old_wm = &old_pipe_wm->planes[plane_id];
+
+		if ((skl_ddb_entry_equal(&cur_ddb->plane[pipe][plane_id],
+					&new_ddb->plane[pipe][plane_id]) &&
+		    skl_ddb_entry_equal(&cur_ddb->y_plane[pipe][plane_id],
+					&new_ddb->y_plane[pipe][plane_id])) &&
+		    !memcmp(wm, old_wm, sizeof(struct skl_plane_wm)))
+			continue;
+
+		plane_state = drm_atomic_get_plane_state(state, plane);
+		if (IS_ERR(plane_state))
+			return PTR_ERR(plane_state);
+	}
+
+	return 0;
+}
+
 static int skl_update_pipe_wm(struct drm_crtc_state *cstate,
 			      const struct skl_pipe_wm *old_pipe_wm,
 			      struct skl_pipe_wm *pipe_wm, /* out */
@@ -4147,6 +4827,17 @@ static int skl_update_pipe_wm(struct drm_crtc_state *cstate,
 	if (ret)
 		return ret;
 
+	ret = skl_allocate_pipe_ddb(intel_cstate, pipe_wm, ddb);
+	if (ret)
+		return ret;
+	/*
+	 * TODO: Planes are included in state to arm WM registers.
+	 * Scope to optimize further, by just rewriting plane surf register.
+	 */
+	ret = skl_ddb_add_affected_planes(intel_cstate, old_pipe_wm, pipe_wm);
+	if (ret)
+		return ret;
+
 	if (!memcmp(old_pipe_wm, pipe_wm, sizeof(*pipe_wm)))
 		*changed = false;
 	else
@@ -4169,41 +4860,7 @@ pipes_modified(struct drm_atomic_state *state)
 }
 
 static int
-skl_ddb_add_affected_planes(struct intel_crtc_state *cstate)
-{
-	struct drm_atomic_state *state = cstate->base.state;
-	struct drm_device *dev = state->dev;
-	struct drm_crtc *crtc = cstate->base.crtc;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_atomic_state *intel_state = to_intel_atomic_state(state);
-	struct skl_ddb_allocation *new_ddb = &intel_state->wm_results.ddb;
-	struct skl_ddb_allocation *cur_ddb = &dev_priv->wm.skl_hw.ddb;
-	struct drm_plane_state *plane_state;
-	struct drm_plane *plane;
-	enum pipe pipe = intel_crtc->pipe;
-
-	WARN_ON(!drm_atomic_get_existing_crtc_state(state, crtc));
-
-	drm_for_each_plane_mask(plane, dev, cstate->base.plane_mask) {
-		enum plane_id plane_id = to_intel_plane(plane)->id;
-
-		if (skl_ddb_entry_equal(&cur_ddb->plane[pipe][plane_id],
-					&new_ddb->plane[pipe][plane_id]) &&
-		    skl_ddb_entry_equal(&cur_ddb->y_plane[pipe][plane_id],
-					&new_ddb->y_plane[pipe][plane_id]))
-			continue;
-
-		plane_state = drm_atomic_get_plane_state(state, plane);
-		if (IS_ERR(plane_state))
-			return PTR_ERR(plane_state);
-	}
-
-	return 0;
-}
-
-static int
-skl_compute_ddb(struct drm_atomic_state *state)
+skl_include_affected_crtcs(struct drm_atomic_state *state)
 {
 	struct drm_device *dev = state->dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
@@ -4267,14 +4924,6 @@ skl_compute_ddb(struct drm_atomic_state *state)
 		cstate = intel_atomic_get_crtc_state(state, intel_crtc);
 		if (IS_ERR(cstate))
 			return PTR_ERR(cstate);
-
-		ret = skl_allocate_pipe_ddb(cstate, ddb);
-		if (ret)
-			return ret;
-
-		ret = skl_ddb_add_affected_planes(cstate);
-		if (ret)
-			return ret;
 	}
 
 	return 0;
@@ -4364,7 +5013,7 @@ skl_compute_wm(struct drm_atomic_state *state)
 	/* Clear all dirty flags */
 	results->dirty_pipes = 0;
 
-	ret = skl_compute_ddb(state);
+	ret = skl_include_affected_crtcs(state);
 	if (ret)
 		return ret;
 
@@ -4663,6 +5312,32 @@ static void ilk_pipe_wm_get_hw_state(struct drm_crtc *crtc)
 #define _FW_WM_VLV(value, plane) \
 	(((value) & DSPFW_ ## plane ## _MASK_VLV) >> DSPFW_ ## plane ## _SHIFT)
 
+static void g4x_read_wm_values(struct drm_i915_private *dev_priv,
+			       struct g4x_wm_values *wm)
+{
+	uint32_t tmp;
+
+	tmp = I915_READ(DSPFW1);
+	wm->sr.plane = _FW_WM(tmp, SR);
+	wm->pipe[PIPE_B].plane[PLANE_CURSOR] = _FW_WM(tmp, CURSORB);
+	wm->pipe[PIPE_B].plane[PLANE_PRIMARY] = _FW_WM(tmp, PLANEB);
+	wm->pipe[PIPE_A].plane[PLANE_PRIMARY] = _FW_WM(tmp, PLANEA);
+
+	tmp = I915_READ(DSPFW2);
+	wm->fbc_en = tmp & DSPFW_FBC_SR_EN;
+	wm->sr.fbc = _FW_WM(tmp, FBC_SR);
+	wm->hpll.fbc = _FW_WM(tmp, FBC_HPLL_SR);
+	wm->pipe[PIPE_B].plane[PLANE_SPRITE0] = _FW_WM(tmp, SPRITEB);
+	wm->pipe[PIPE_A].plane[PLANE_CURSOR] = _FW_WM(tmp, CURSORA);
+	wm->pipe[PIPE_A].plane[PLANE_SPRITE0] = _FW_WM(tmp, SPRITEA);
+
+	tmp = I915_READ(DSPFW3);
+	wm->hpll_en = tmp & DSPFW_HPLL_SR_EN;
+	wm->sr.cursor = _FW_WM(tmp, CURSOR_SR);
+	wm->hpll.cursor = _FW_WM(tmp, HPLL_CURSOR);
+	wm->hpll.plane = _FW_WM(tmp, HPLL_SR);
+}
+
 static void vlv_read_wm_values(struct drm_i915_private *dev_priv,
 			       struct vlv_wm_values *wm)
 {
@@ -4739,6 +5414,147 @@ static void vlv_read_wm_values(struct drm_i915_private *dev_priv,
 #undef _FW_WM
 #undef _FW_WM_VLV
 
+void g4x_wm_get_hw_state(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct g4x_wm_values *wm = &dev_priv->wm.g4x;
+	struct intel_crtc *crtc;
+
+	g4x_read_wm_values(dev_priv, wm);
+
+	wm->cxsr = I915_READ(FW_BLC_SELF) & FW_BLC_SELF_EN;
+
+	for_each_intel_crtc(dev, crtc) {
+		struct intel_crtc_state *crtc_state =
+			to_intel_crtc_state(crtc->base.state);
+		struct g4x_wm_state *active = &crtc->wm.active.g4x;
+		struct g4x_pipe_wm *raw;
+		enum pipe pipe = crtc->pipe;
+		enum plane_id plane_id;
+		int level, max_level;
+
+		active->cxsr = wm->cxsr;
+		active->hpll_en = wm->hpll_en;
+		active->fbc_en = wm->fbc_en;
+
+		active->sr = wm->sr;
+		active->hpll = wm->hpll;
+
+		for_each_plane_id_on_crtc(crtc, plane_id) {
+			active->wm.plane[plane_id] =
+				wm->pipe[pipe].plane[plane_id];
+		}
+
+		if (wm->cxsr && wm->hpll_en)
+			max_level = G4X_WM_LEVEL_HPLL;
+		else if (wm->cxsr)
+			max_level = G4X_WM_LEVEL_SR;
+		else
+			max_level = G4X_WM_LEVEL_NORMAL;
+
+		level = G4X_WM_LEVEL_NORMAL;
+		raw = &crtc_state->wm.g4x.raw[level];
+		for_each_plane_id_on_crtc(crtc, plane_id)
+			raw->plane[plane_id] = active->wm.plane[plane_id];
+
+		if (++level > max_level)
+			goto out;
+
+		raw = &crtc_state->wm.g4x.raw[level];
+		raw->plane[PLANE_PRIMARY] = active->sr.plane;
+		raw->plane[PLANE_CURSOR] = active->sr.cursor;
+		raw->plane[PLANE_SPRITE0] = 0;
+		raw->fbc = active->sr.fbc;
+
+		if (++level > max_level)
+			goto out;
+
+		raw = &crtc_state->wm.g4x.raw[level];
+		raw->plane[PLANE_PRIMARY] = active->hpll.plane;
+		raw->plane[PLANE_CURSOR] = active->hpll.cursor;
+		raw->plane[PLANE_SPRITE0] = 0;
+		raw->fbc = active->hpll.fbc;
+
+	out:
+		for_each_plane_id_on_crtc(crtc, plane_id)
+			g4x_raw_plane_wm_set(crtc_state, level,
+					     plane_id, USHRT_MAX);
+		g4x_raw_fbc_wm_set(crtc_state, level, USHRT_MAX);
+
+		crtc_state->wm.g4x.optimal = *active;
+		crtc_state->wm.g4x.intermediate = *active;
+
+		DRM_DEBUG_KMS("Initial watermarks: pipe %c, plane=%d, cursor=%d, sprite=%d\n",
+			      pipe_name(pipe),
+			      wm->pipe[pipe].plane[PLANE_PRIMARY],
+			      wm->pipe[pipe].plane[PLANE_CURSOR],
+			      wm->pipe[pipe].plane[PLANE_SPRITE0]);
+	}
+
+	DRM_DEBUG_KMS("Initial SR watermarks: plane=%d, cursor=%d fbc=%d\n",
+		      wm->sr.plane, wm->sr.cursor, wm->sr.fbc);
+	DRM_DEBUG_KMS("Initial HPLL watermarks: plane=%d, SR cursor=%d fbc=%d\n",
+		      wm->hpll.plane, wm->hpll.cursor, wm->hpll.fbc);
+	DRM_DEBUG_KMS("Initial SR=%s HPLL=%s FBC=%s\n",
+		      yesno(wm->cxsr), yesno(wm->hpll_en), yesno(wm->fbc_en));
+}
+
+void g4x_wm_sanitize(struct drm_i915_private *dev_priv)
+{
+	struct intel_plane *plane;
+	struct intel_crtc *crtc;
+
+	mutex_lock(&dev_priv->wm.wm_mutex);
+
+	for_each_intel_plane(&dev_priv->drm, plane) {
+		struct intel_crtc *crtc =
+			intel_get_crtc_for_pipe(dev_priv, plane->pipe);
+		struct intel_crtc_state *crtc_state =
+			to_intel_crtc_state(crtc->base.state);
+		struct intel_plane_state *plane_state =
+			to_intel_plane_state(plane->base.state);
+		struct g4x_wm_state *wm_state = &crtc_state->wm.g4x.optimal;
+		enum plane_id plane_id = plane->id;
+		int level;
+
+		if (plane_state->base.visible)
+			continue;
+
+		for (level = 0; level < 3; level++) {
+			struct g4x_pipe_wm *raw =
+				&crtc_state->wm.g4x.raw[level];
+
+			raw->plane[plane_id] = 0;
+			wm_state->wm.plane[plane_id] = 0;
+		}
+
+		if (plane_id == PLANE_PRIMARY) {
+			for (level = 0; level < 3; level++) {
+				struct g4x_pipe_wm *raw =
+					&crtc_state->wm.g4x.raw[level];
+				raw->fbc = 0;
+			}
+
+			wm_state->sr.fbc = 0;
+			wm_state->hpll.fbc = 0;
+			wm_state->fbc_en = false;
+		}
+	}
+
+	for_each_intel_crtc(&dev_priv->drm, crtc) {
+		struct intel_crtc_state *crtc_state =
+			to_intel_crtc_state(crtc->base.state);
+
+		crtc_state->wm.g4x.intermediate =
+			crtc_state->wm.g4x.optimal;
+		crtc->wm.active.g4x = crtc_state->wm.g4x.optimal;
+	}
+
+	g4x_program_watermarks(dev_priv);
+
+	mutex_unlock(&dev_priv->wm.wm_mutex);
+}
+
 void vlv_wm_get_hw_state(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
@@ -4801,7 +5617,7 @@ void vlv_wm_get_hw_state(struct drm_device *dev)
 		active->cxsr = wm->cxsr;
 
 		for (level = 0; level < active->num_levels; level++) {
-			struct vlv_pipe_wm *raw =
+			struct g4x_pipe_wm *raw =
 				&crtc_state->wm.vlv.raw[level];
 
 			active->sr[level].plane = wm->sr.plane;
@@ -4861,7 +5677,7 @@ void vlv_wm_sanitize(struct drm_i915_private *dev_priv)
 			continue;
 
 		for (level = 0; level < wm_state->num_levels; level++) {
-			struct vlv_pipe_wm *raw =
+			struct g4x_pipe_wm *raw =
 				&crtc_state->wm.vlv.raw[level];
 
 			raw->plane[plane_id] = 0;
@@ -7497,7 +8313,7 @@ static void kabylake_init_clock_gating(struct drm_i915_private *dev_priv)
 		I915_WRITE(GEN6_UCGCTL1, I915_READ(GEN6_UCGCTL1) |
 			   GEN6_GAMUNIT_CLOCK_GATE_DISABLE);
 
-	/* WaFbcNukeOnHostModify:kbl */
+	/* WaFbcNukeOnHostModify:kbl,cfl */
 	I915_WRITE(ILK_DPFC_CHICKEN, I915_READ(ILK_DPFC_CHICKEN) |
 		   ILK_DPFC_NUKE_ON_ANY_MODIFICATION);
 }
@@ -7965,7 +8781,7 @@ void intel_init_clock_gating_hooks(struct drm_i915_private *dev_priv)
 {
 	if (IS_SKYLAKE(dev_priv))
 		dev_priv->display.init_clock_gating = skylake_init_clock_gating;
-	else if (IS_KABYLAKE(dev_priv))
+	else if (IS_KABYLAKE(dev_priv) || IS_COFFEELAKE(dev_priv))
 		dev_priv->display.init_clock_gating = kabylake_init_clock_gating;
 	else if (IS_BROXTON(dev_priv))
 		dev_priv->display.init_clock_gating = bxt_init_clock_gating;
@@ -8045,6 +8861,12 @@ void intel_init_pm(struct drm_i915_private *dev_priv)
 		dev_priv->display.initial_watermarks = vlv_initial_watermarks;
 		dev_priv->display.optimize_watermarks = vlv_optimize_watermarks;
 		dev_priv->display.atomic_update_watermarks = vlv_atomic_update_fifo;
+	} else if (IS_G4X(dev_priv)) {
+		g4x_setup_wm_latency(dev_priv);
+		dev_priv->display.compute_pipe_wm = g4x_compute_pipe_wm;
+		dev_priv->display.compute_intermediate_wm = g4x_compute_intermediate_wm;
+		dev_priv->display.initial_watermarks = g4x_initial_watermarks;
+		dev_priv->display.optimize_watermarks = g4x_optimize_watermarks;
 	} else if (IS_PINEVIEW(dev_priv)) {
 		if (!intel_get_cxsr_latency(IS_PINEVIEW_G(dev_priv),
 					    dev_priv->is_ddr3,
@@ -8060,8 +8882,6 @@ void intel_init_pm(struct drm_i915_private *dev_priv)
 			dev_priv->display.update_wm = NULL;
 		} else
 			dev_priv->display.update_wm = pineview_update_wm;
-	} else if (IS_G4X(dev_priv)) {
-		dev_priv->display.update_wm = g4x_update_wm;
 	} else if (IS_GEN4(dev_priv)) {
 		dev_priv->display.update_wm = i965_update_wm;
 	} else if (IS_GEN3(dev_priv)) {
@@ -8144,9 +8964,9 @@ int sandybridge_pcode_read(struct drm_i915_private *dev_priv, u32 mbox, u32 *val
 	I915_WRITE_FW(GEN6_PCODE_DATA1, 0);
 	I915_WRITE_FW(GEN6_PCODE_MAILBOX, GEN6_PCODE_READY | mbox);
 
-	if (intel_wait_for_register_fw(dev_priv,
-				       GEN6_PCODE_MAILBOX, GEN6_PCODE_READY, 0,
-				       500)) {
+	if (__intel_wait_for_register_fw(dev_priv,
+					 GEN6_PCODE_MAILBOX, GEN6_PCODE_READY, 0,
+					 500, 0, NULL)) {
 		DRM_ERROR("timeout waiting for pcode read (%d) to finish\n", mbox);
 		return -ETIMEDOUT;
 	}
@@ -8189,9 +9009,9 @@ int sandybridge_pcode_write(struct drm_i915_private *dev_priv,
 	I915_WRITE_FW(GEN6_PCODE_DATA1, 0);
 	I915_WRITE_FW(GEN6_PCODE_MAILBOX, GEN6_PCODE_READY | mbox);
 
-	if (intel_wait_for_register_fw(dev_priv,
-				       GEN6_PCODE_MAILBOX, GEN6_PCODE_READY, 0,
-				       500)) {
+	if (__intel_wait_for_register_fw(dev_priv,
+					 GEN6_PCODE_MAILBOX, GEN6_PCODE_READY, 0,
+					 500, 0, NULL)) {
 		DRM_ERROR("timeout waiting for pcode write (%d) to finish\n", mbox);
 		return -ETIMEDOUT;
 	}
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 66a2b8b83972..acd1da9b62a3 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -39,17 +39,27 @@
  */
 #define LEGACY_REQUEST_SIZE 200
 
-static int __intel_ring_space(int head, int tail, int size)
+static unsigned int __intel_ring_space(unsigned int head,
+				       unsigned int tail,
+				       unsigned int size)
 {
-	int space = head - tail;
-	if (space <= 0)
-		space += size;
-	return space - I915_RING_FREE_SPACE;
+	/*
+	 * "If the Ring Buffer Head Pointer and the Tail Pointer are on the
+	 * same cacheline, the Head Pointer must not be greater than the Tail
+	 * Pointer."
+	 */
+	GEM_BUG_ON(!is_power_of_2(size));
+	return (head - tail - CACHELINE_BYTES) & (size - 1);
 }
 
-void intel_ring_update_space(struct intel_ring *ring)
+unsigned int intel_ring_update_space(struct intel_ring *ring)
 {
-	ring->space = __intel_ring_space(ring->head, ring->tail, ring->size);
+	unsigned int space;
+
+	space = __intel_ring_space(ring->head, ring->emit, ring->size);
+
+	ring->space = space;
+	return space;
 }
 
 static int
@@ -538,9 +548,9 @@ static int init_ring_common(struct intel_engine_cs *engine)
 	I915_WRITE_CTL(engine, RING_CTL_SIZE(ring->size) | RING_VALID);
 
 	/* If the head is still not zero, the ring is dead */
-	if (intel_wait_for_register_fw(dev_priv, RING_CTL(engine->mmio_base),
-				       RING_VALID, RING_VALID,
-				       50)) {
+	if (intel_wait_for_register(dev_priv, RING_CTL(engine->mmio_base),
+				    RING_VALID, RING_VALID,
+				    50)) {
 		DRM_ERROR("%s initialization failed "
 			  "ctl %08x (valid? %d) head %08x [%08x] tail %08x [%08x] start %08x [expected %08x]\n",
 			  engine->name,
@@ -774,8 +784,8 @@ static void i9xx_submit_request(struct drm_i915_gem_request *request)
 
 	i915_gem_request_submit(request);
 
-	assert_ring_tail_valid(request->ring, request->tail);
-	I915_WRITE_TAIL(request->engine, request->tail);
+	I915_WRITE_TAIL(request->engine,
+			intel_ring_set_tail(request->ring, request->tail));
 }
 
 static void i9xx_emit_breadcrumb(struct drm_i915_gem_request *req, u32 *cs)
@@ -1259,6 +1269,8 @@ static int init_phys_status_page(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
 
+	GEM_BUG_ON(engine->id != RCS);
+
 	dev_priv->status_page_dmah =
 		drm_pci_alloc(&dev_priv->drm, PAGE_SIZE, PAGE_SIZE);
 	if (!dev_priv->status_page_dmah)
@@ -1270,17 +1282,18 @@ static int init_phys_status_page(struct intel_engine_cs *engine)
 	return 0;
 }
 
-int intel_ring_pin(struct intel_ring *ring, unsigned int offset_bias)
+int intel_ring_pin(struct intel_ring *ring,
+		   struct drm_i915_private *i915,
+		   unsigned int offset_bias)
 {
-	unsigned int flags;
-	enum i915_map_type map;
+	enum i915_map_type map = HAS_LLC(i915) ? I915_MAP_WB : I915_MAP_WC;
 	struct i915_vma *vma = ring->vma;
+	unsigned int flags;
 	void *addr;
 	int ret;
 
 	GEM_BUG_ON(ring->vaddr);
 
-	map = HAS_LLC(ring->engine->i915) ? I915_MAP_WB : I915_MAP_WC;
 
 	flags = PIN_GLOBAL;
 	if (offset_bias)
@@ -1316,11 +1329,23 @@ err:
 	return PTR_ERR(addr);
 }
 
+void intel_ring_reset(struct intel_ring *ring, u32 tail)
+{
+	GEM_BUG_ON(!list_empty(&ring->request_list));
+	ring->tail = tail;
+	ring->head = tail;
+	ring->emit = tail;
+	intel_ring_update_space(ring);
+}
+
 void intel_ring_unpin(struct intel_ring *ring)
 {
 	GEM_BUG_ON(!ring->vma);
 	GEM_BUG_ON(!ring->vaddr);
 
+	/* Discard any unused bytes beyond that submitted to hw. */
+	intel_ring_reset(ring, ring->tail);
+
 	if (i915_vma_is_map_and_fenceable(ring->vma))
 		i915_vma_unpin_iomap(ring->vma);
 	else
@@ -1338,7 +1363,7 @@ intel_ring_create_vma(struct drm_i915_private *dev_priv, int size)
 
 	obj = i915_gem_object_create_stolen(dev_priv, size);
 	if (!obj)
-		obj = i915_gem_object_create(dev_priv, size);
+		obj = i915_gem_object_create_internal(dev_priv, size);
 	if (IS_ERR(obj))
 		return ERR_CAST(obj);
 
@@ -1369,8 +1394,6 @@ intel_engine_create_ring(struct intel_engine_cs *engine, int size)
 	if (!ring)
 		return ERR_PTR(-ENOMEM);
 
-	ring->engine = engine;
-
 	INIT_LIST_HEAD(&ring->request_list);
 
 	ring->size = size;
@@ -1424,22 +1447,73 @@ static int context_pin(struct i915_gem_context *ctx)
 			    PIN_GLOBAL | PIN_HIGH);
 }
 
-static int intel_ring_context_pin(struct intel_engine_cs *engine,
-				  struct i915_gem_context *ctx)
+static struct i915_vma *
+alloc_context_vma(struct intel_engine_cs *engine)
+{
+	struct drm_i915_private *i915 = engine->i915;
+	struct drm_i915_gem_object *obj;
+	struct i915_vma *vma;
+
+	obj = i915_gem_object_create(i915, engine->context_size);
+	if (IS_ERR(obj))
+		return ERR_CAST(obj);
+
+	/*
+	 * Try to make the context utilize L3 as well as LLC.
+	 *
+	 * On VLV we don't have L3 controls in the PTEs so we
+	 * shouldn't touch the cache level, especially as that
+	 * would make the object snooped which might have a
+	 * negative performance impact.
+	 *
+	 * Snooping is required on non-llc platforms in execlist
+	 * mode, but since all GGTT accesses use PAT entry 0 we
+	 * get snooping anyway regardless of cache_level.
+	 *
+	 * This is only applicable for Ivy Bridge devices since
+	 * later platforms don't have L3 control bits in the PTE.
+	 */
+	if (IS_IVYBRIDGE(i915)) {
+		/* Ignore any error, regard it as a simple optimisation */
+		i915_gem_object_set_cache_level(obj, I915_CACHE_L3_LLC);
+	}
+
+	vma = i915_vma_instance(obj, &i915->ggtt.base, NULL);
+	if (IS_ERR(vma))
+		i915_gem_object_put(obj);
+
+	return vma;
+}
+
+static struct intel_ring *
+intel_ring_context_pin(struct intel_engine_cs *engine,
+		       struct i915_gem_context *ctx)
 {
 	struct intel_context *ce = &ctx->engine[engine->id];
 	int ret;
 
 	lockdep_assert_held(&ctx->i915->drm.struct_mutex);
 
-	if (ce->pin_count++)
-		return 0;
+	if (likely(ce->pin_count++))
+		goto out;
 	GEM_BUG_ON(!ce->pin_count); /* no overflow please! */
 
+	if (!ce->state && engine->context_size) {
+		struct i915_vma *vma;
+
+		vma = alloc_context_vma(engine);
+		if (IS_ERR(vma)) {
+			ret = PTR_ERR(vma);
+			goto err;
+		}
+
+		ce->state = vma;
+	}
+
 	if (ce->state) {
 		ret = context_pin(ctx);
 		if (ret)
-			goto error;
+			goto err;
 
 		ce->state->obj->mm.dirty = true;
 	}
@@ -1455,11 +1529,14 @@ static int intel_ring_context_pin(struct intel_engine_cs *engine,
 		ce->initialised = true;
 
 	i915_gem_context_get(ctx);
-	return 0;
 
-error:
+out:
+	/* One ringbuffer to rule them all */
+	return engine->buffer;
+
+err:
 	ce->pin_count = 0;
-	return ret;
+	return ERR_PTR(ret);
 }
 
 static void intel_ring_context_unpin(struct intel_engine_cs *engine,
@@ -1481,78 +1558,70 @@ static void intel_ring_context_unpin(struct intel_engine_cs *engine,
 
 static int intel_init_ring_buffer(struct intel_engine_cs *engine)
 {
-	struct drm_i915_private *dev_priv = engine->i915;
 	struct intel_ring *ring;
-	int ret;
-
-	WARN_ON(engine->buffer);
+	int err;
 
 	intel_engine_setup_common(engine);
 
-	ret = intel_engine_init_common(engine);
-	if (ret)
-		goto error;
+	err = intel_engine_init_common(engine);
+	if (err)
+		goto err;
+
+	if (HWS_NEEDS_PHYSICAL(engine->i915))
+		err = init_phys_status_page(engine);
+	else
+		err = init_status_page(engine);
+	if (err)
+		goto err;
 
 	ring = intel_engine_create_ring(engine, 32 * PAGE_SIZE);
 	if (IS_ERR(ring)) {
-		ret = PTR_ERR(ring);
-		goto error;
-	}
-
-	if (HWS_NEEDS_PHYSICAL(dev_priv)) {
-		WARN_ON(engine->id != RCS);
-		ret = init_phys_status_page(engine);
-		if (ret)
-			goto error;
-	} else {
-		ret = init_status_page(engine);
-		if (ret)
-			goto error;
+		err = PTR_ERR(ring);
+		goto err_hws;
 	}
 
 	/* Ring wraparound at offset 0 sometimes hangs. No idea why. */
-	ret = intel_ring_pin(ring, I915_GTT_PAGE_SIZE);
-	if (ret) {
-		intel_ring_free(ring);
-		goto error;
-	}
+	err = intel_ring_pin(ring, engine->i915, I915_GTT_PAGE_SIZE);
+	if (err)
+		goto err_ring;
+
+	GEM_BUG_ON(engine->buffer);
 	engine->buffer = ring;
 
 	return 0;
 
-error:
-	intel_engine_cleanup(engine);
-	return ret;
+err_ring:
+	intel_ring_free(ring);
+err_hws:
+	if (HWS_NEEDS_PHYSICAL(engine->i915))
+		cleanup_phys_status_page(engine);
+	else
+		cleanup_status_page(engine);
+err:
+	intel_engine_cleanup_common(engine);
+	return err;
 }
 
 void intel_engine_cleanup(struct intel_engine_cs *engine)
 {
-	struct drm_i915_private *dev_priv;
-
-	dev_priv = engine->i915;
+	struct drm_i915_private *dev_priv = engine->i915;
 
-	if (engine->buffer) {
-		WARN_ON(INTEL_GEN(dev_priv) > 2 &&
-			(I915_READ_MODE(engine) & MODE_IDLE) == 0);
+	WARN_ON(INTEL_GEN(dev_priv) > 2 &&
+		(I915_READ_MODE(engine) & MODE_IDLE) == 0);
 
-		intel_ring_unpin(engine->buffer);
-		intel_ring_free(engine->buffer);
-		engine->buffer = NULL;
-	}
+	intel_ring_unpin(engine->buffer);
+	intel_ring_free(engine->buffer);
 
 	if (engine->cleanup)
 		engine->cleanup(engine);
 
-	if (HWS_NEEDS_PHYSICAL(dev_priv)) {
-		WARN_ON(engine->id != RCS);
+	if (HWS_NEEDS_PHYSICAL(dev_priv))
 		cleanup_phys_status_page(engine);
-	} else {
+	else
 		cleanup_status_page(engine);
-	}
 
 	intel_engine_cleanup_common(engine);
 
-	engine->i915 = NULL;
 	dev_priv->engine[engine->id] = NULL;
 	kfree(engine);
 }
@@ -1562,8 +1631,9 @@ void intel_legacy_submission_resume(struct drm_i915_private *dev_priv)
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 
+	/* Restart from the beginning of the rings for convenience */
 	for_each_engine(engine, dev_priv, id)
-		engine->buffer->head = engine->buffer->tail;
+		intel_ring_reset(engine->buffer, 0);
 }
 
 static int ring_request_alloc(struct drm_i915_gem_request *request)
@@ -1578,9 +1648,6 @@ static int ring_request_alloc(struct drm_i915_gem_request *request)
 	 */
 	request->reserved_space += LEGACY_REQUEST_SIZE;
 
-	GEM_BUG_ON(!request->engine->buffer);
-	request->ring = request->engine->buffer;
-
 	cs = intel_ring_begin(request, 0);
 	if (IS_ERR(cs))
 		return PTR_ERR(cs);
@@ -1589,7 +1656,8 @@ static int ring_request_alloc(struct drm_i915_gem_request *request)
 	return 0;
 }
 
-static int wait_for_space(struct drm_i915_gem_request *req, int bytes)
+static noinline int wait_for_space(struct drm_i915_gem_request *req,
+				   unsigned int bytes)
 {
 	struct intel_ring *ring = req->ring;
 	struct drm_i915_gem_request *target;
@@ -1597,8 +1665,7 @@ static int wait_for_space(struct drm_i915_gem_request *req, int bytes)
 
 	lockdep_assert_held(&req->i915->drm.struct_mutex);
 
-	intel_ring_update_space(ring);
-	if (ring->space >= bytes)
+	if (intel_ring_update_space(ring) >= bytes)
 		return 0;
 
 	/*
@@ -1613,12 +1680,9 @@ static int wait_for_space(struct drm_i915_gem_request *req, int bytes)
 	GEM_BUG_ON(!req->reserved_space);
 
 	list_for_each_entry(target, &ring->request_list, ring_link) {
-		unsigned space;
-
 		/* Would completion of this request free enough space? */
-		space = __intel_ring_space(target->postfix, ring->tail,
-					   ring->size);
-		if (space >= bytes)
+		if (bytes <= __intel_ring_space(target->postfix,
+						ring->emit, ring->size))
 			break;
 	}
 
@@ -1638,59 +1702,64 @@ static int wait_for_space(struct drm_i915_gem_request *req, int bytes)
 	return 0;
 }
 
-u32 *intel_ring_begin(struct drm_i915_gem_request *req, int num_dwords)
+u32 *intel_ring_begin(struct drm_i915_gem_request *req,
+		      unsigned int num_dwords)
 {
 	struct intel_ring *ring = req->ring;
-	int remain_actual = ring->size - ring->tail;
-	int remain_usable = ring->effective_size - ring->tail;
-	int bytes = num_dwords * sizeof(u32);
-	int total_bytes, wait_bytes;
-	bool need_wrap = false;
+	const unsigned int remain_usable = ring->effective_size - ring->emit;
+	const unsigned int bytes = num_dwords * sizeof(u32);
+	unsigned int need_wrap = 0;
+	unsigned int total_bytes;
 	u32 *cs;
 
 	total_bytes = bytes + req->reserved_space;
+	GEM_BUG_ON(total_bytes > ring->effective_size);
 
-	if (unlikely(bytes > remain_usable)) {
-		/*
-		 * Not enough space for the basic request. So need to flush
-		 * out the remainder and then wait for base + reserved.
-		 */
-		wait_bytes = remain_actual + total_bytes;
-		need_wrap = true;
-	} else if (unlikely(total_bytes > remain_usable)) {
-		/*
-		 * The base request will fit but the reserved space
-		 * falls off the end. So we don't need an immediate wrap
-		 * and only need to effectively wait for the reserved
-		 * size space from the start of ringbuffer.
-		 */
-		wait_bytes = remain_actual + req->reserved_space;
-	} else {
-		/* No wrapping required, just waiting. */
-		wait_bytes = total_bytes;
+	if (unlikely(total_bytes > remain_usable)) {
+		const int remain_actual = ring->size - ring->emit;
+
+		if (bytes > remain_usable) {
+			/*
+			 * Not enough space for the basic request. So need to
+			 * flush out the remainder and then wait for
+			 * base + reserved.
+			 */
+			total_bytes += remain_actual;
+			need_wrap = remain_actual | 1;
+		} else  {
+			/*
+			 * The base request will fit but the reserved space
+			 * falls off the end. So we don't need an immediate
+			 * wrap and only need to effectively wait for the
+			 * reserved size from the start of ringbuffer.
+			 */
+			total_bytes = req->reserved_space + remain_actual;
+		}
 	}
 
-	if (wait_bytes > ring->space) {
-		int ret = wait_for_space(req, wait_bytes);
+	if (unlikely(total_bytes > ring->space)) {
+		int ret = wait_for_space(req, total_bytes);
 		if (unlikely(ret))
 			return ERR_PTR(ret);
 	}
 
 	if (unlikely(need_wrap)) {
-		GEM_BUG_ON(remain_actual > ring->space);
-		GEM_BUG_ON(ring->tail + remain_actual > ring->size);
+		need_wrap &= ~1;
+		GEM_BUG_ON(need_wrap > ring->space);
+		GEM_BUG_ON(ring->emit + need_wrap > ring->size);
 
 		/* Fill the tail with MI_NOOP */
-		memset(ring->vaddr + ring->tail, 0, remain_actual);
-		ring->tail = 0;
-		ring->space -= remain_actual;
+		memset(ring->vaddr + ring->emit, 0, need_wrap);
+		ring->emit = 0;
+		ring->space -= need_wrap;
 	}
 
-	GEM_BUG_ON(ring->tail > ring->size - bytes);
-	cs = ring->vaddr + ring->tail;
-	ring->tail += bytes;
+	GEM_BUG_ON(ring->emit > ring->size - bytes);
+	GEM_BUG_ON(ring->space < bytes);
+	cs = ring->vaddr + ring->emit;
+	GEM_DEBUG_EXEC(memset(cs, POISON_INUSE, bytes));
+	ring->emit += bytes;
 	ring->space -= bytes;
-	GEM_BUG_ON(ring->space < 0);
 
 	return cs;
 }
@@ -1699,7 +1768,7 @@ u32 *intel_ring_begin(struct drm_i915_gem_request *req, int num_dwords)
 int intel_ring_cacheline_align(struct drm_i915_gem_request *req)
 {
 	int num_dwords =
-		(req->ring->tail & (CACHELINE_BYTES - 1)) / sizeof(uint32_t);
+		(req->ring->emit & (CACHELINE_BYTES - 1)) / sizeof(uint32_t);
 	u32 *cs;
 
 	if (num_dwords == 0)
@@ -1736,11 +1805,11 @@ static void gen6_bsd_submit_request(struct drm_i915_gem_request *request)
 	I915_WRITE64_FW(GEN6_BSD_RNCID, 0x0);
 
 	/* Wait for the ring not to be idle, i.e. for it to wake up. */
-	if (intel_wait_for_register_fw(dev_priv,
-				       GEN6_BSD_SLEEP_PSMI_CONTROL,
-				       GEN6_BSD_SLEEP_INDICATOR,
-				       0,
-				       50))
+	if (__intel_wait_for_register_fw(dev_priv,
+					 GEN6_BSD_SLEEP_PSMI_CONTROL,
+					 GEN6_BSD_SLEEP_INDICATOR,
+					 0,
+					 1000, 0, NULL))
 		DRM_ERROR("timed out waiting for the BSD ring to wake up\n");
 
 	/* Now that the ring is fully powered up, update the tail */
@@ -2182,20 +2251,6 @@ int intel_init_bsd_ring_buffer(struct intel_engine_cs *engine)
 	return intel_init_ring_buffer(engine);
 }
 
-/**
- * Initialize the second BSD ring (eg. Broadwell GT3, Skylake GT3)
- */
-int intel_init_bsd2_ring_buffer(struct intel_engine_cs *engine)
-{
-	struct drm_i915_private *dev_priv = engine->i915;
-
-	intel_ring_default_vfuncs(dev_priv, engine);
-
-	engine->emit_flush = gen6_bsd_ring_flush;
-
-	return intel_init_ring_buffer(engine);
-}
-
 int intel_init_blt_ring_buffer(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index a82a0807f64d..6aa20ac8cde3 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -17,17 +17,6 @@
 #define CACHELINE_BYTES 64
 #define CACHELINE_DWORDS (CACHELINE_BYTES / sizeof(uint32_t))
 
-/*
- * Gen2 BSpec "1. Programming Environment" / 1.4.4.6 "Ring Buffer Use"
- * Gen3 BSpec "vol1c Memory Interface Functions" / 2.3.4.5 "Ring Buffer Use"
- * Gen4+ BSpec "vol1c Memory Interface and Command Stream" / 5.3.4.5 "Ring Buffer Use"
- *
- * "If the Ring Buffer Head Pointer and the Tail Pointer are on the same
- * cacheline, the Head Pointer must not be greater than the Tail
- * Pointer."
- */
-#define I915_RING_FREE_SPACE 64
-
 struct intel_hw_status_page {
 	struct i915_vma *vma;
 	u32 *page_addr;
@@ -139,16 +128,15 @@ struct intel_ring {
 	struct i915_vma *vma;
 	void *vaddr;
 
-	struct intel_engine_cs *engine;
-
 	struct list_head request_list;
 
 	u32 head;
 	u32 tail;
+	u32 emit;
 
-	int space;
-	int size;
-	int effective_size;
+	u32 space;
+	u32 size;
+	u32 effective_size;
 };
 
 struct i915_gem_context;
@@ -189,15 +177,28 @@ enum intel_engine_id {
 	VECS
 };
 
+struct i915_priolist {
+	struct rb_node node;
+	struct list_head requests;
+	int priority;
+};
+
+#define INTEL_ENGINE_CS_MAX_NAME 8
+
 struct intel_engine_cs {
 	struct drm_i915_private *i915;
-	const char	*name;
+	char name[INTEL_ENGINE_CS_MAX_NAME];
 	enum intel_engine_id id;
-	unsigned int exec_id;
+	unsigned int uabi_id;
 	unsigned int hw_id;
 	unsigned int guc_id;
-	u32		mmio_base;
+
+	u8 class;
+	u8 instance;
+	u32 context_size;
+	u32 mmio_base;
 	unsigned int irq_shift;
+
 	struct intel_ring *buffer;
 	struct intel_timeline *timeline;
 
@@ -265,8 +266,8 @@ struct intel_engine_cs {
 
 	void		(*set_default_submission)(struct intel_engine_cs *engine);
 
-	int		(*context_pin)(struct intel_engine_cs *engine,
-				       struct i915_gem_context *ctx);
+	struct intel_ring *(*context_pin)(struct intel_engine_cs *engine,
+					  struct i915_gem_context *ctx);
 	void		(*context_unpin)(struct intel_engine_cs *engine,
 					 struct i915_gem_context *ctx);
 	int		(*request_alloc)(struct drm_i915_gem_request *req);
@@ -372,9 +373,18 @@ struct intel_engine_cs {
 
 	/* Execlists */
 	struct tasklet_struct irq_tasklet;
+	struct i915_priolist default_priolist;
+	bool no_priolist;
 	struct execlist_port {
-		struct drm_i915_gem_request *request;
-		unsigned int count;
+		struct drm_i915_gem_request *request_count;
+#define EXECLIST_COUNT_BITS 2
+#define port_request(p) ptr_mask_bits((p)->request_count, EXECLIST_COUNT_BITS)
+#define port_count(p) ptr_unmask_bits((p)->request_count, EXECLIST_COUNT_BITS)
+#define port_pack(rq, count) ptr_pack_bits(rq, count, EXECLIST_COUNT_BITS)
+#define port_unpack(p, count) ptr_unpack_bits((p)->request_count, count, EXECLIST_COUNT_BITS)
+#define port_set(p, packed) ((p)->request_count = (packed))
+#define port_isset(p) ((p)->request_count)
+#define port_index(p, e) ((p) - (e)->execlist_port)
 		GEM_DEBUG_DECL(u32 context_id);
 	} execlist_port[2];
 	struct rb_root execlist_queue;
@@ -487,7 +497,11 @@ intel_write_status_page(struct intel_engine_cs *engine, int reg, u32 value)
 
 struct intel_ring *
 intel_engine_create_ring(struct intel_engine_cs *engine, int size);
-int intel_ring_pin(struct intel_ring *ring, unsigned int offset_bias);
+int intel_ring_pin(struct intel_ring *ring,
+		   struct drm_i915_private *i915,
+		   unsigned int offset_bias);
+void intel_ring_reset(struct intel_ring *ring, u32 tail);
+unsigned int intel_ring_update_space(struct intel_ring *ring);
 void intel_ring_unpin(struct intel_ring *ring);
 void intel_ring_free(struct intel_ring *ring);
 
@@ -498,7 +512,8 @@ void intel_legacy_submission_resume(struct drm_i915_private *dev_priv);
 
 int __must_check intel_ring_cacheline_align(struct drm_i915_gem_request *req);
 
-u32 __must_check *intel_ring_begin(struct drm_i915_gem_request *req, int n);
+u32 __must_check *intel_ring_begin(struct drm_i915_gem_request *req,
+				   unsigned int n);
 
 static inline void
 intel_ring_advance(struct drm_i915_gem_request *req, u32 *cs)
@@ -511,7 +526,7 @@ intel_ring_advance(struct drm_i915_gem_request *req, u32 *cs)
 	 * reserved for the command packet (i.e. the value passed to
 	 * intel_ring_begin()).
 	 */
-	GEM_BUG_ON((req->ring->vaddr + req->ring->tail) != cs);
+	GEM_BUG_ON((req->ring->vaddr + req->ring->emit) != cs);
 }
 
 static inline u32
@@ -538,9 +553,40 @@ assert_ring_tail_valid(const struct intel_ring *ring, unsigned int tail)
 	 */
 	GEM_BUG_ON(!IS_ALIGNED(tail, 8));
 	GEM_BUG_ON(tail >= ring->size);
+
+	/*
+	 * "Ring Buffer Use"
+	 *	Gen2 BSpec "1. Programming Environment" / 1.4.4.6
+	 *	Gen3 BSpec "1c Memory Interface Functions" / 2.3.4.5
+	 *	Gen4+ BSpec "1c Memory Interface and Command Stream" / 5.3.4.5
+	 * "If the Ring Buffer Head Pointer and the Tail Pointer are on the
+	 * same cacheline, the Head Pointer must not be greater than the Tail
+	 * Pointer."
+	 *
+	 * We use ring->head as the last known location of the actual RING_HEAD,
+	 * it may have advanced but in the worst case it is equally the same
+	 * as ring->head and so we should never program RING_TAIL to advance
+	 * into the same cacheline as ring->head.
+	 */
+#define cacheline(a) round_down(a, CACHELINE_BYTES)
+	GEM_BUG_ON(cacheline(tail) == cacheline(ring->head) &&
+		   tail < ring->head);
+#undef cacheline
 }
 
-void intel_ring_update_space(struct intel_ring *ring);
+static inline unsigned int
+intel_ring_set_tail(struct intel_ring *ring, unsigned int tail)
+{
+	/* Whilst writes to the tail are strictly order, there is no
+	 * serialisation between readers and the writers. The tail may be
+	 * read by i915_gem_request_retire() just as it is being updated
+	 * by execlists, as although the breadcrumb is complete, the context
+	 * switch hasn't been seen.
+	 */
+	assert_ring_tail_valid(ring, tail);
+	ring->tail = tail;
+	return tail;
+}
 
 void intel_engine_init_global_seqno(struct intel_engine_cs *engine, u32 seqno);
 
@@ -551,7 +597,6 @@ void intel_engine_cleanup_common(struct intel_engine_cs *engine);
 
 int intel_init_render_ring_buffer(struct intel_engine_cs *engine);
 int intel_init_bsd_ring_buffer(struct intel_engine_cs *engine);
-int intel_init_bsd2_ring_buffer(struct intel_engine_cs *engine);
 int intel_init_blt_ring_buffer(struct intel_engine_cs *engine);
 int intel_init_vebox_ring_buffer(struct intel_engine_cs *engine);
 
@@ -652,7 +697,8 @@ bool intel_engine_add_wait(struct intel_engine_cs *engine,
 			   struct intel_wait *wait);
 void intel_engine_remove_wait(struct intel_engine_cs *engine,
 			      struct intel_wait *wait);
-void intel_engine_enable_signaling(struct drm_i915_gem_request *request);
+void intel_engine_enable_signaling(struct drm_i915_gem_request *request,
+				   bool wakeup);
 void intel_engine_cancel_signaling(struct drm_i915_gem_request *request);
 
 static inline bool intel_engine_has_waiter(const struct intel_engine_cs *engine)
@@ -685,6 +731,7 @@ static inline u32 *gen8_emit_pipe_control(u32 *batch, u32 flags, u32 offset)
 bool intel_engine_is_idle(struct intel_engine_cs *engine);
 bool intel_engines_are_idle(struct drm_i915_private *dev_priv);
 
+void intel_engines_mark_idle(struct drm_i915_private *i915);
 void intel_engines_reset_default_submission(struct drm_i915_private *i915);
 
 #endif /* _INTEL_RINGBUFFER_H_ */
diff --git a/drivers/gpu/drm/i915/intel_runtime_pm.c b/drivers/gpu/drm/i915/intel_runtime_pm.c
index f8a375f8dde6..8a6f287d225b 100644
--- a/drivers/gpu/drm/i915/intel_runtime_pm.c
+++ b/drivers/gpu/drm/i915/intel_runtime_pm.c
@@ -494,6 +494,55 @@ static void hsw_set_power_well(struct drm_i915_private *dev_priv,
 	BIT_ULL(POWER_DOMAIN_AUX_A) |			\
 	BIT_ULL(POWER_DOMAIN_INIT))
 
+#define CNL_DISPLAY_POWERWELL_2_POWER_DOMAINS (		\
+	BIT_ULL(POWER_DOMAIN_TRANSCODER_A) |		\
+	BIT_ULL(POWER_DOMAIN_PIPE_B) |			\
+	BIT_ULL(POWER_DOMAIN_TRANSCODER_B) |		\
+	BIT_ULL(POWER_DOMAIN_PIPE_C) |			\
+	BIT_ULL(POWER_DOMAIN_TRANSCODER_C) |		\
+	BIT_ULL(POWER_DOMAIN_PIPE_B_PANEL_FITTER) |		\
+	BIT_ULL(POWER_DOMAIN_PIPE_C_PANEL_FITTER) |		\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_B_LANES) |		\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_C_LANES) |		\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_D_LANES) |		\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_E_LANES) |		\
+	BIT_ULL(POWER_DOMAIN_AUX_B) |                       \
+	BIT_ULL(POWER_DOMAIN_AUX_C) |			\
+	BIT_ULL(POWER_DOMAIN_AUX_D) |			\
+	BIT_ULL(POWER_DOMAIN_AUDIO) |			\
+	BIT_ULL(POWER_DOMAIN_VGA) |				\
+	BIT_ULL(POWER_DOMAIN_INIT))
+#define CNL_DISPLAY_DDI_A_IO_POWER_DOMAINS (		\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_A_IO) |		\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_E_IO) |		\
+	BIT_ULL(POWER_DOMAIN_INIT))
+#define CNL_DISPLAY_DDI_B_IO_POWER_DOMAINS (		\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_B_IO) |		\
+	BIT_ULL(POWER_DOMAIN_INIT))
+#define CNL_DISPLAY_DDI_C_IO_POWER_DOMAINS (		\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_C_IO) |		\
+	BIT_ULL(POWER_DOMAIN_INIT))
+#define CNL_DISPLAY_DDI_D_IO_POWER_DOMAINS (		\
+	BIT_ULL(POWER_DOMAIN_PORT_DDI_D_IO) |		\
+	BIT_ULL(POWER_DOMAIN_INIT))
+#define CNL_DISPLAY_AUX_A_POWER_DOMAINS (		\
+	BIT_ULL(POWER_DOMAIN_AUX_A) |			\
+	BIT_ULL(POWER_DOMAIN_INIT))
+#define CNL_DISPLAY_AUX_B_POWER_DOMAINS (		\
+	BIT_ULL(POWER_DOMAIN_AUX_B) |			\
+	BIT_ULL(POWER_DOMAIN_INIT))
+#define CNL_DISPLAY_AUX_C_POWER_DOMAINS (		\
+	BIT_ULL(POWER_DOMAIN_AUX_C) |			\
+	BIT_ULL(POWER_DOMAIN_INIT))
+#define CNL_DISPLAY_AUX_D_POWER_DOMAINS (		\
+	BIT_ULL(POWER_DOMAIN_AUX_D) |			\
+	BIT_ULL(POWER_DOMAIN_INIT))
+#define CNL_DISPLAY_DC_OFF_POWER_DOMAINS (		\
+	CNL_DISPLAY_POWERWELL_2_POWER_DOMAINS |		\
+	BIT_ULL(POWER_DOMAIN_MODESET) |			\
+	BIT_ULL(POWER_DOMAIN_AUX_A) |			\
+	BIT_ULL(POWER_DOMAIN_INIT))
+
 static void assert_can_enable_dc9(struct drm_i915_private *dev_priv)
 {
 	WARN_ONCE((I915_READ(DC_STATE_EN) & DC_STATE_EN_DC9),
@@ -762,13 +811,14 @@ static void skl_set_power_well(struct drm_i915_private *dev_priv,
 		}
 		break;
 	case SKL_DISP_PW_MISC_IO:
-	case SKL_DISP_PW_DDI_A_E: /* GLK_DISP_PW_DDI_A */
+	case SKL_DISP_PW_DDI_A_E: /* GLK_DISP_PW_DDI_A, CNL_DISP_PW_DDI_A */
 	case SKL_DISP_PW_DDI_B:
 	case SKL_DISP_PW_DDI_C:
 	case SKL_DISP_PW_DDI_D:
-	case GLK_DISP_PW_AUX_A:
-	case GLK_DISP_PW_AUX_B:
-	case GLK_DISP_PW_AUX_C:
+	case GLK_DISP_PW_AUX_A: /* CNL_DISP_PW_AUX_A */
+	case GLK_DISP_PW_AUX_B: /* CNL_DISP_PW_AUX_B */
+	case GLK_DISP_PW_AUX_C: /* CNL_DISP_PW_AUX_C */
+	case CNL_DISP_PW_AUX_D:
 		break;
 	default:
 		WARN(1, "Unknown power well %lu\n", power_well->id);
@@ -803,8 +853,7 @@ static void skl_set_power_well(struct drm_i915_private *dev_priv,
 			DRM_DEBUG_KMS("Disabling %s\n", power_well->name);
 		}
 
-		if (IS_GEN9(dev_priv))
-			gen9_sanitize_power_well_requests(dev_priv, power_well);
+		gen9_sanitize_power_well_requests(dev_priv, power_well);
 	}
 
 	if (wait_for(!!(I915_READ(HSW_PWR_WELL_DRIVER) & state_mask) == enable,
@@ -2275,6 +2324,82 @@ static struct i915_power_well glk_power_wells[] = {
 	},
 };
 
+static struct i915_power_well cnl_power_wells[] = {
+	{
+		.name = "always-on",
+		.always_on = 1,
+		.domains = POWER_DOMAIN_MASK,
+		.ops = &i9xx_always_on_power_well_ops,
+	},
+	{
+		.name = "power well 1",
+		/* Handled by the DMC firmware */
+		.domains = 0,
+		.ops = &skl_power_well_ops,
+		.id = SKL_DISP_PW_1,
+	},
+	{
+		.name = "AUX A",
+		.domains = CNL_DISPLAY_AUX_A_POWER_DOMAINS,
+		.ops = &skl_power_well_ops,
+		.id = CNL_DISP_PW_AUX_A,
+	},
+	{
+		.name = "AUX B",
+		.domains = CNL_DISPLAY_AUX_B_POWER_DOMAINS,
+		.ops = &skl_power_well_ops,
+		.id = CNL_DISP_PW_AUX_B,
+	},
+	{
+		.name = "AUX C",
+		.domains = CNL_DISPLAY_AUX_C_POWER_DOMAINS,
+		.ops = &skl_power_well_ops,
+		.id = CNL_DISP_PW_AUX_C,
+	},
+	{
+		.name = "AUX D",
+		.domains = CNL_DISPLAY_AUX_D_POWER_DOMAINS,
+		.ops = &skl_power_well_ops,
+		.id = CNL_DISP_PW_AUX_D,
+	},
+	{
+		.name = "DC off",
+		.domains = CNL_DISPLAY_DC_OFF_POWER_DOMAINS,
+		.ops = &gen9_dc_off_power_well_ops,
+		.id = SKL_DISP_PW_DC_OFF,
+	},
+	{
+		.name = "power well 2",
+		.domains = CNL_DISPLAY_POWERWELL_2_POWER_DOMAINS,
+		.ops = &skl_power_well_ops,
+		.id = SKL_DISP_PW_2,
+	},
+	{
+		.name = "DDI A IO power well",
+		.domains = CNL_DISPLAY_DDI_A_IO_POWER_DOMAINS,
+		.ops = &skl_power_well_ops,
+		.id = CNL_DISP_PW_DDI_A,
+	},
+	{
+		.name = "DDI B IO power well",
+		.domains = CNL_DISPLAY_DDI_B_IO_POWER_DOMAINS,
+		.ops = &skl_power_well_ops,
+		.id = SKL_DISP_PW_DDI_B,
+	},
+	{
+		.name = "DDI C IO power well",
+		.domains = CNL_DISPLAY_DDI_C_IO_POWER_DOMAINS,
+		.ops = &skl_power_well_ops,
+		.id = SKL_DISP_PW_DDI_C,
+	},
+	{
+		.name = "DDI D IO power well",
+		.domains = CNL_DISPLAY_DDI_D_IO_POWER_DOMAINS,
+		.ops = &skl_power_well_ops,
+		.id = SKL_DISP_PW_DDI_D,
+	},
+};
+
 static int
 sanitize_disable_power_well_option(const struct drm_i915_private *dev_priv,
 				   int disable_power_well)
@@ -2369,6 +2494,8 @@ int intel_power_domains_init(struct drm_i915_private *dev_priv)
 		set_power_wells(power_domains, bdw_power_wells);
 	} else if (IS_GEN9_BC(dev_priv)) {
 		set_power_wells(power_domains, skl_power_wells);
+	} else if (IS_CANNONLAKE(dev_priv)) {
+		set_power_wells(power_domains, cnl_power_wells);
 	} else if (IS_BROXTON(dev_priv)) {
 		set_power_wells(power_domains, bxt_power_wells);
 	} else if (IS_GEMINILAKE(dev_priv)) {
diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c
index 816a6f5a3fd9..f4329d20b6f6 100644
--- a/drivers/gpu/drm/i915/intel_sdvo.c
+++ b/drivers/gpu/drm/i915/intel_sdvo.c
@@ -100,18 +100,6 @@ struct intel_sdvo {
 	uint16_t hotplug_active;
 
 	/**
-	 * This is used to select the color range of RBG outputs in HDMI mode.
-	 * It is only valid when using TMDS encoding and 8 bit per color mode.
-	 */
-	uint32_t color_range;
-	bool color_range_auto;
-
-	/**
-	 * HDMI user specified aspect ratio
-	 */
-	enum hdmi_picture_aspect aspect_ratio;
-
-	/**
 	 * This is set if we're going to treat the device as TV-out.
 	 *
 	 * While we have these nice friendly flags for output types that ought
@@ -122,9 +110,6 @@ struct intel_sdvo {
 
 	enum port port;
 
-	/* This is for current tv format name */
-	int tv_format_index;
-
 	/**
 	 * This is set if we treat the device as HDMI, instead of DVI.
 	 */
@@ -159,8 +144,6 @@ struct intel_sdvo_connector {
 	/* Mark the type of connector */
 	uint16_t output_flag;
 
-	enum hdmi_force_audio force_audio;
-
 	/* This contains all current supported TV format */
 	u8 tv_format_supported[TV_FORMAT_NUM];
 	int   format_supported_num;
@@ -187,24 +170,19 @@ struct intel_sdvo_connector {
 	/* add the property for the SDVO-TV/LVDS */
 	struct drm_property *brightness;
 
-	/* Add variable to record current setting for the above property */
-	u32	left_margin, right_margin, top_margin, bottom_margin;
-
 	/* this is to get the range of margin.*/
-	u32	max_hscan,  max_vscan;
-	u32	max_hpos, cur_hpos;
-	u32	max_vpos, cur_vpos;
-	u32	cur_brightness, max_brightness;
-	u32	cur_contrast,	max_contrast;
-	u32	cur_saturation, max_saturation;
-	u32	cur_hue,	max_hue;
-	u32	cur_sharpness,	max_sharpness;
-	u32	cur_flicker_filter,		max_flicker_filter;
-	u32	cur_flicker_filter_adaptive,	max_flicker_filter_adaptive;
-	u32	cur_flicker_filter_2d,		max_flicker_filter_2d;
-	u32	cur_tv_chroma_filter,	max_tv_chroma_filter;
-	u32	cur_tv_luma_filter,	max_tv_luma_filter;
-	u32	cur_dot_crawl,	max_dot_crawl;
+	u32 max_hscan, max_vscan;
+};
+
+struct intel_sdvo_connector_state {
+	/* base.base: tv.saturation/contrast/hue/brightness */
+	struct intel_digital_connector_state base;
+
+	struct {
+		unsigned overscan_h, overscan_v, hpos, vpos, sharpness;
+		unsigned flicker_filter, flicker_filter_2d, flicker_filter_adaptive;
+		unsigned chroma_filter, luma_filter, dot_crawl;
+	} tv;
 };
 
 static struct intel_sdvo *to_sdvo(struct intel_encoder *encoder)
@@ -217,9 +195,16 @@ static struct intel_sdvo *intel_attached_sdvo(struct drm_connector *connector)
 	return to_sdvo(intel_attached_encoder(connector));
 }
 
-static struct intel_sdvo_connector *to_intel_sdvo_connector(struct drm_connector *connector)
+static struct intel_sdvo_connector *
+to_intel_sdvo_connector(struct drm_connector *connector)
+{
+	return container_of(connector, struct intel_sdvo_connector, base.base);
+}
+
+static struct intel_sdvo_connector_state *
+to_intel_sdvo_connector_state(struct drm_connector_state *conn_state)
 {
-	return container_of(to_intel_connector(connector), struct intel_sdvo_connector, base);
+	return container_of(conn_state, struct intel_sdvo_connector_state, base.base);
 }
 
 static bool
@@ -1035,12 +1020,13 @@ static bool intel_sdvo_set_avi_infoframe(struct intel_sdvo *intel_sdvo,
 					  sdvo_data, sizeof(sdvo_data));
 }
 
-static bool intel_sdvo_set_tv_format(struct intel_sdvo *intel_sdvo)
+static bool intel_sdvo_set_tv_format(struct intel_sdvo *intel_sdvo,
+				     struct drm_connector_state *conn_state)
 {
 	struct intel_sdvo_tv_format format;
 	uint32_t format_map;
 
-	format_map = 1 << intel_sdvo->tv_format_index;
+	format_map = 1 << conn_state->tv.mode;
 	memset(&format, 0, sizeof(format));
 	memcpy(&format, &format_map, min(sizeof(format), sizeof(format_map)));
 
@@ -1127,6 +1113,8 @@ static bool intel_sdvo_compute_config(struct intel_encoder *encoder,
 				      struct drm_connector_state *conn_state)
 {
 	struct intel_sdvo *intel_sdvo = to_sdvo(encoder);
+	struct intel_sdvo_connector_state *intel_sdvo_state =
+		to_intel_sdvo_connector_state(conn_state);
 	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
 	struct drm_display_mode *mode = &pipe_config->base.mode;
 
@@ -1165,9 +1153,14 @@ static bool intel_sdvo_compute_config(struct intel_encoder *encoder,
 	pipe_config->pixel_multiplier =
 		intel_sdvo_get_pixel_multiplier(adjusted_mode);
 
-	pipe_config->has_hdmi_sink = intel_sdvo->has_hdmi_monitor;
+	if (intel_sdvo_state->base.force_audio != HDMI_AUDIO_OFF_DVI)
+		pipe_config->has_hdmi_sink = intel_sdvo->has_hdmi_monitor;
 
-	if (intel_sdvo->color_range_auto) {
+	if (intel_sdvo_state->base.force_audio == HDMI_AUDIO_ON ||
+	    (intel_sdvo_state->base.force_audio == HDMI_AUDIO_AUTO && intel_sdvo->has_hdmi_audio))
+		pipe_config->has_audio = true;
+
+	if (intel_sdvo_state->base.broadcast_rgb == INTEL_BROADCAST_RGB_AUTO) {
 		/* See CEA-861-E - 5.1 Default Encoding Parameters */
 		/* FIXME: This bit is only valid when using TMDS encoding and 8
 		 * bit per color mode. */
@@ -1176,7 +1169,7 @@ static bool intel_sdvo_compute_config(struct intel_encoder *encoder,
 			pipe_config->limited_color_range = true;
 	} else {
 		if (pipe_config->has_hdmi_sink &&
-		    intel_sdvo->color_range == HDMI_COLOR_RANGE_16_235)
+		    intel_sdvo_state->base.broadcast_rgb == INTEL_BROADCAST_RGB_LIMITED)
 			pipe_config->limited_color_range = true;
 	}
 
@@ -1186,11 +1179,73 @@ static bool intel_sdvo_compute_config(struct intel_encoder *encoder,
 
 	/* Set user selected PAR to incoming mode's member */
 	if (intel_sdvo->is_hdmi)
-		adjusted_mode->picture_aspect_ratio = intel_sdvo->aspect_ratio;
+		adjusted_mode->picture_aspect_ratio = conn_state->picture_aspect_ratio;
 
 	return true;
 }
 
+#define UPDATE_PROPERTY(input, NAME) \
+	do { \
+		val = input; \
+		intel_sdvo_set_value(intel_sdvo, SDVO_CMD_SET_##NAME, &val, sizeof(val)); \
+	} while (0)
+
+static void intel_sdvo_update_props(struct intel_sdvo *intel_sdvo,
+				    struct intel_sdvo_connector_state *sdvo_state)
+{
+	struct drm_connector_state *conn_state = &sdvo_state->base.base;
+	struct intel_sdvo_connector *intel_sdvo_conn =
+		to_intel_sdvo_connector(conn_state->connector);
+	uint16_t val;
+
+	if (intel_sdvo_conn->left)
+		UPDATE_PROPERTY(sdvo_state->tv.overscan_h, OVERSCAN_H);
+
+	if (intel_sdvo_conn->top)
+		UPDATE_PROPERTY(sdvo_state->tv.overscan_v, OVERSCAN_V);
+
+	if (intel_sdvo_conn->hpos)
+		UPDATE_PROPERTY(sdvo_state->tv.hpos, HPOS);
+
+	if (intel_sdvo_conn->vpos)
+		UPDATE_PROPERTY(sdvo_state->tv.vpos, VPOS);
+
+	if (intel_sdvo_conn->saturation)
+		UPDATE_PROPERTY(conn_state->tv.saturation, SATURATION);
+
+	if (intel_sdvo_conn->contrast)
+		UPDATE_PROPERTY(conn_state->tv.contrast, CONTRAST);
+
+	if (intel_sdvo_conn->hue)
+		UPDATE_PROPERTY(conn_state->tv.hue, HUE);
+
+	if (intel_sdvo_conn->brightness)
+		UPDATE_PROPERTY(conn_state->tv.brightness, BRIGHTNESS);
+
+	if (intel_sdvo_conn->sharpness)
+		UPDATE_PROPERTY(sdvo_state->tv.sharpness, SHARPNESS);
+
+	if (intel_sdvo_conn->flicker_filter)
+		UPDATE_PROPERTY(sdvo_state->tv.flicker_filter, FLICKER_FILTER);
+
+	if (intel_sdvo_conn->flicker_filter_2d)
+		UPDATE_PROPERTY(sdvo_state->tv.flicker_filter_2d, FLICKER_FILTER_2D);
+
+	if (intel_sdvo_conn->flicker_filter_adaptive)
+		UPDATE_PROPERTY(sdvo_state->tv.flicker_filter_adaptive, FLICKER_FILTER_ADAPTIVE);
+
+	if (intel_sdvo_conn->tv_chroma_filter)
+		UPDATE_PROPERTY(sdvo_state->tv.chroma_filter, TV_CHROMA_FILTER);
+
+	if (intel_sdvo_conn->tv_luma_filter)
+		UPDATE_PROPERTY(sdvo_state->tv.luma_filter, TV_LUMA_FILTER);
+
+	if (intel_sdvo_conn->dot_crawl)
+		UPDATE_PROPERTY(sdvo_state->tv.dot_crawl, DOT_CRAWL);
+
+#undef UPDATE_PROPERTY
+}
+
 static void intel_sdvo_pre_enable(struct intel_encoder *intel_encoder,
 				  struct intel_crtc_state *crtc_state,
 				  struct drm_connector_state *conn_state)
@@ -1198,6 +1253,7 @@ static void intel_sdvo_pre_enable(struct intel_encoder *intel_encoder,
 	struct drm_i915_private *dev_priv = to_i915(intel_encoder->base.dev);
 	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
 	const struct drm_display_mode *adjusted_mode = &crtc_state->base.adjusted_mode;
+	struct intel_sdvo_connector_state *sdvo_state = to_intel_sdvo_connector_state(conn_state);
 	struct drm_display_mode *mode = &crtc_state->base.mode;
 	struct intel_sdvo *intel_sdvo = to_sdvo(intel_encoder);
 	u32 sdvox;
@@ -1205,6 +1261,8 @@ static void intel_sdvo_pre_enable(struct intel_encoder *intel_encoder,
 	struct intel_sdvo_dtd input_dtd, output_dtd;
 	int rate;
 
+	intel_sdvo_update_props(intel_sdvo, sdvo_state);
+
 	/* First, set the input mapping for the first input to our controlled
 	 * output. This is only correct if we're a single-input device, in
 	 * which case the first input is the output from the appropriate SDVO
@@ -1246,7 +1304,7 @@ static void intel_sdvo_pre_enable(struct intel_encoder *intel_encoder,
 		intel_sdvo_set_encode(intel_sdvo, SDVO_ENCODE_DVI);
 
 	if (intel_sdvo->is_tv &&
-	    !intel_sdvo_set_tv_format(intel_sdvo))
+	    !intel_sdvo_set_tv_format(intel_sdvo, conn_state))
 		return;
 
 	intel_sdvo_get_dtd_from_mode(&input_dtd, adjusted_mode);
@@ -1290,7 +1348,7 @@ static void intel_sdvo_pre_enable(struct intel_encoder *intel_encoder,
 	else
 		sdvox |= SDVO_PIPE_SEL(crtc->pipe);
 
-	if (intel_sdvo->has_hdmi_audio)
+	if (crtc_state->has_audio)
 		sdvox |= SDVO_AUDIO_ENABLE;
 
 	if (INTEL_GEN(dev_priv) >= 4) {
@@ -1699,12 +1757,6 @@ intel_sdvo_tmds_sink_detect(struct drm_connector *connector)
 		kfree(edid);
 	}
 
-	if (status == connector_status_connected) {
-		struct intel_sdvo_connector *intel_sdvo_connector = to_intel_sdvo_connector(connector);
-		if (intel_sdvo_connector->force_audio != HDMI_AUDIO_AUTO)
-			intel_sdvo->has_hdmi_audio = (intel_sdvo_connector->force_audio == HDMI_AUDIO_ON);
-	}
-
 	return status;
 }
 
@@ -1884,6 +1936,7 @@ static const struct drm_display_mode sdvo_tv_modes[] = {
 static void intel_sdvo_get_tv_modes(struct drm_connector *connector)
 {
 	struct intel_sdvo *intel_sdvo = intel_attached_sdvo(connector);
+	const struct drm_connector_state *conn_state = connector->state;
 	struct intel_sdvo_sdtv_resolution_request tv_res;
 	uint32_t reply = 0, format_map = 0;
 	int i;
@@ -1894,7 +1947,7 @@ static void intel_sdvo_get_tv_modes(struct drm_connector *connector)
 	/* Read the list of supported input resolutions for the selected TV
 	 * format.
 	 */
-	format_map = 1 << intel_sdvo->tv_format_index;
+	format_map = 1 << conn_state->tv.mode;
 	memcpy(&tv_res, &format_map,
 	       min(sizeof(format_map), sizeof(struct intel_sdvo_sdtv_resolution_request)));
 
@@ -1983,204 +2036,121 @@ static void intel_sdvo_destroy(struct drm_connector *connector)
 	kfree(intel_sdvo_connector);
 }
 
-static bool intel_sdvo_detect_hdmi_audio(struct drm_connector *connector)
-{
-	struct intel_sdvo *intel_sdvo = intel_attached_sdvo(connector);
-	struct edid *edid;
-	bool has_audio = false;
-
-	if (!intel_sdvo->is_hdmi)
-		return false;
-
-	edid = intel_sdvo_get_edid(connector);
-	if (edid != NULL && edid->input & DRM_EDID_INPUT_DIGITAL)
-		has_audio = drm_detect_monitor_audio(edid);
-	kfree(edid);
-
-	return has_audio;
-}
-
 static int
-intel_sdvo_set_property(struct drm_connector *connector,
-			struct drm_property *property,
-			uint64_t val)
+intel_sdvo_connector_atomic_get_property(struct drm_connector *connector,
+					 const struct drm_connector_state *state,
+					 struct drm_property *property,
+					 uint64_t *val)
 {
-	struct intel_sdvo *intel_sdvo = intel_attached_sdvo(connector);
 	struct intel_sdvo_connector *intel_sdvo_connector = to_intel_sdvo_connector(connector);
-	struct drm_i915_private *dev_priv = to_i915(connector->dev);
-	uint16_t temp_value;
-	uint8_t cmd;
-	int ret;
-
-	ret = drm_object_property_set_value(&connector->base, property, val);
-	if (ret)
-		return ret;
-
-	if (property == dev_priv->force_audio_property) {
-		int i = val;
-		bool has_audio;
-
-		if (i == intel_sdvo_connector->force_audio)
-			return 0;
-
-		intel_sdvo_connector->force_audio = i;
-
-		if (i == HDMI_AUDIO_AUTO)
-			has_audio = intel_sdvo_detect_hdmi_audio(connector);
-		else
-			has_audio = (i == HDMI_AUDIO_ON);
-
-		if (has_audio == intel_sdvo->has_hdmi_audio)
-			return 0;
-
-		intel_sdvo->has_hdmi_audio = has_audio;
-		goto done;
-	}
-
-	if (property == dev_priv->broadcast_rgb_property) {
-		bool old_auto = intel_sdvo->color_range_auto;
-		uint32_t old_range = intel_sdvo->color_range;
-
-		switch (val) {
-		case INTEL_BROADCAST_RGB_AUTO:
-			intel_sdvo->color_range_auto = true;
-			break;
-		case INTEL_BROADCAST_RGB_FULL:
-			intel_sdvo->color_range_auto = false;
-			intel_sdvo->color_range = 0;
-			break;
-		case INTEL_BROADCAST_RGB_LIMITED:
-			intel_sdvo->color_range_auto = false;
-			/* FIXME: this bit is only valid when using TMDS
-			 * encoding and 8 bit per color mode. */
-			intel_sdvo->color_range = HDMI_COLOR_RANGE_16_235;
-			break;
-		default:
-			return -EINVAL;
-		}
-
-		if (old_auto == intel_sdvo->color_range_auto &&
-		    old_range == intel_sdvo->color_range)
-			return 0;
-
-		goto done;
-	}
-
-	if (property == connector->dev->mode_config.aspect_ratio_property) {
-		switch (val) {
-		case DRM_MODE_PICTURE_ASPECT_NONE:
-			intel_sdvo->aspect_ratio = HDMI_PICTURE_ASPECT_NONE;
-			break;
-		case DRM_MODE_PICTURE_ASPECT_4_3:
-			intel_sdvo->aspect_ratio = HDMI_PICTURE_ASPECT_4_3;
-			break;
-		case DRM_MODE_PICTURE_ASPECT_16_9:
-			intel_sdvo->aspect_ratio = HDMI_PICTURE_ASPECT_16_9;
-			break;
-		default:
-			return -EINVAL;
-		}
-		goto done;
-	}
-
-#define CHECK_PROPERTY(name, NAME) \
-	if (intel_sdvo_connector->name == property) { \
-		if (intel_sdvo_connector->cur_##name == temp_value) return 0; \
-		if (intel_sdvo_connector->max_##name < temp_value) return -EINVAL; \
-		cmd = SDVO_CMD_SET_##NAME; \
-		intel_sdvo_connector->cur_##name = temp_value; \
-		goto set_value; \
-	}
+	const struct intel_sdvo_connector_state *sdvo_state = to_intel_sdvo_connector_state((void *)state);
 
 	if (property == intel_sdvo_connector->tv_format) {
-		if (val >= TV_FORMAT_NUM)
-			return -EINVAL;
-
-		if (intel_sdvo->tv_format_index ==
-		    intel_sdvo_connector->tv_format_supported[val])
-			return 0;
-
-		intel_sdvo->tv_format_index = intel_sdvo_connector->tv_format_supported[val];
-		goto done;
-	} else if (IS_TV_OR_LVDS(intel_sdvo_connector)) {
-		temp_value = val;
-		if (intel_sdvo_connector->left == property) {
-			drm_object_property_set_value(&connector->base,
-							 intel_sdvo_connector->right, val);
-			if (intel_sdvo_connector->left_margin == temp_value)
-				return 0;
+		int i;
 
-			intel_sdvo_connector->left_margin = temp_value;
-			intel_sdvo_connector->right_margin = temp_value;
-			temp_value = intel_sdvo_connector->max_hscan -
-				intel_sdvo_connector->left_margin;
-			cmd = SDVO_CMD_SET_OVERSCAN_H;
-			goto set_value;
-		} else if (intel_sdvo_connector->right == property) {
-			drm_object_property_set_value(&connector->base,
-							 intel_sdvo_connector->left, val);
-			if (intel_sdvo_connector->right_margin == temp_value)
-				return 0;
+		for (i = 0; i < intel_sdvo_connector->format_supported_num; i++)
+			if (state->tv.mode == intel_sdvo_connector->tv_format_supported[i]) {
+				*val = i;
 
-			intel_sdvo_connector->left_margin = temp_value;
-			intel_sdvo_connector->right_margin = temp_value;
-			temp_value = intel_sdvo_connector->max_hscan -
-				intel_sdvo_connector->left_margin;
-			cmd = SDVO_CMD_SET_OVERSCAN_H;
-			goto set_value;
-		} else if (intel_sdvo_connector->top == property) {
-			drm_object_property_set_value(&connector->base,
-							 intel_sdvo_connector->bottom, val);
-			if (intel_sdvo_connector->top_margin == temp_value)
 				return 0;
+			}
 
-			intel_sdvo_connector->top_margin = temp_value;
-			intel_sdvo_connector->bottom_margin = temp_value;
-			temp_value = intel_sdvo_connector->max_vscan -
-				intel_sdvo_connector->top_margin;
-			cmd = SDVO_CMD_SET_OVERSCAN_V;
-			goto set_value;
-		} else if (intel_sdvo_connector->bottom == property) {
-			drm_object_property_set_value(&connector->base,
-							 intel_sdvo_connector->top, val);
-			if (intel_sdvo_connector->bottom_margin == temp_value)
-				return 0;
+		WARN_ON(1);
+		*val = 0;
+	} else if (property == intel_sdvo_connector->top ||
+		   property == intel_sdvo_connector->bottom)
+		*val = intel_sdvo_connector->max_vscan - sdvo_state->tv.overscan_v;
+	else if (property == intel_sdvo_connector->left ||
+		 property == intel_sdvo_connector->right)
+		*val = intel_sdvo_connector->max_hscan - sdvo_state->tv.overscan_h;
+	else if (property == intel_sdvo_connector->hpos)
+		*val = sdvo_state->tv.hpos;
+	else if (property == intel_sdvo_connector->vpos)
+		*val = sdvo_state->tv.vpos;
+	else if (property == intel_sdvo_connector->saturation)
+		*val = state->tv.saturation;
+	else if (property == intel_sdvo_connector->contrast)
+		*val = state->tv.contrast;
+	else if (property == intel_sdvo_connector->hue)
+		*val = state->tv.hue;
+	else if (property == intel_sdvo_connector->brightness)
+		*val = state->tv.brightness;
+	else if (property == intel_sdvo_connector->sharpness)
+		*val = sdvo_state->tv.sharpness;
+	else if (property == intel_sdvo_connector->flicker_filter)
+		*val = sdvo_state->tv.flicker_filter;
+	else if (property == intel_sdvo_connector->flicker_filter_2d)
+		*val = sdvo_state->tv.flicker_filter_2d;
+	else if (property == intel_sdvo_connector->flicker_filter_adaptive)
+		*val = sdvo_state->tv.flicker_filter_adaptive;
+	else if (property == intel_sdvo_connector->tv_chroma_filter)
+		*val = sdvo_state->tv.chroma_filter;
+	else if (property == intel_sdvo_connector->tv_luma_filter)
+		*val = sdvo_state->tv.luma_filter;
+	else if (property == intel_sdvo_connector->dot_crawl)
+		*val = sdvo_state->tv.dot_crawl;
+	else
+		return intel_digital_connector_atomic_get_property(connector, state, property, val);
 
-			intel_sdvo_connector->top_margin = temp_value;
-			intel_sdvo_connector->bottom_margin = temp_value;
-			temp_value = intel_sdvo_connector->max_vscan -
-				intel_sdvo_connector->top_margin;
-			cmd = SDVO_CMD_SET_OVERSCAN_V;
-			goto set_value;
-		}
-		CHECK_PROPERTY(hpos, HPOS)
-		CHECK_PROPERTY(vpos, VPOS)
-		CHECK_PROPERTY(saturation, SATURATION)
-		CHECK_PROPERTY(contrast, CONTRAST)
-		CHECK_PROPERTY(hue, HUE)
-		CHECK_PROPERTY(brightness, BRIGHTNESS)
-		CHECK_PROPERTY(sharpness, SHARPNESS)
-		CHECK_PROPERTY(flicker_filter, FLICKER_FILTER)
-		CHECK_PROPERTY(flicker_filter_2d, FLICKER_FILTER_2D)
-		CHECK_PROPERTY(flicker_filter_adaptive, FLICKER_FILTER_ADAPTIVE)
-		CHECK_PROPERTY(tv_chroma_filter, TV_CHROMA_FILTER)
-		CHECK_PROPERTY(tv_luma_filter, TV_LUMA_FILTER)
-		CHECK_PROPERTY(dot_crawl, DOT_CRAWL)
-	}
+	return 0;
+}
 
-	return -EINVAL; /* unknown property */
+static int
+intel_sdvo_connector_atomic_set_property(struct drm_connector *connector,
+					 struct drm_connector_state *state,
+					 struct drm_property *property,
+					 uint64_t val)
+{
+	struct intel_sdvo_connector *intel_sdvo_connector = to_intel_sdvo_connector(connector);
+	struct intel_sdvo_connector_state *sdvo_state = to_intel_sdvo_connector_state(state);
 
-set_value:
-	if (!intel_sdvo_set_value(intel_sdvo, cmd, &temp_value, 2))
-		return -EIO;
+	if (property == intel_sdvo_connector->tv_format) {
+		state->tv.mode = intel_sdvo_connector->tv_format_supported[val];
 
+		if (state->crtc) {
+			struct drm_crtc_state *crtc_state =
+				drm_atomic_get_new_crtc_state(state->state, state->crtc);
 
-done:
-	if (intel_sdvo->base.base.crtc)
-		intel_crtc_restore_mode(intel_sdvo->base.base.crtc);
+			crtc_state->connectors_changed = true;
+		}
+	} else if (property == intel_sdvo_connector->top ||
+		   property == intel_sdvo_connector->bottom)
+		/* Cannot set these independent from each other */
+		sdvo_state->tv.overscan_v = intel_sdvo_connector->max_vscan - val;
+	else if (property == intel_sdvo_connector->left ||
+		 property == intel_sdvo_connector->right)
+		/* Cannot set these independent from each other */
+		sdvo_state->tv.overscan_h = intel_sdvo_connector->max_hscan - val;
+	else if (property == intel_sdvo_connector->hpos)
+		sdvo_state->tv.hpos = val;
+	else if (property == intel_sdvo_connector->vpos)
+		sdvo_state->tv.vpos = val;
+	else if (property == intel_sdvo_connector->saturation)
+		state->tv.saturation = val;
+	else if (property == intel_sdvo_connector->contrast)
+		state->tv.contrast = val;
+	else if (property == intel_sdvo_connector->hue)
+		state->tv.hue = val;
+	else if (property == intel_sdvo_connector->brightness)
+		state->tv.brightness = val;
+	else if (property == intel_sdvo_connector->sharpness)
+		sdvo_state->tv.sharpness = val;
+	else if (property == intel_sdvo_connector->flicker_filter)
+		sdvo_state->tv.flicker_filter = val;
+	else if (property == intel_sdvo_connector->flicker_filter_2d)
+		sdvo_state->tv.flicker_filter_2d = val;
+	else if (property == intel_sdvo_connector->flicker_filter_adaptive)
+		sdvo_state->tv.flicker_filter_adaptive = val;
+	else if (property == intel_sdvo_connector->tv_chroma_filter)
+		sdvo_state->tv.chroma_filter = val;
+	else if (property == intel_sdvo_connector->tv_luma_filter)
+		sdvo_state->tv.luma_filter = val;
+	else if (property == intel_sdvo_connector->dot_crawl)
+		sdvo_state->tv.dot_crawl = val;
+	else
+		return intel_digital_connector_atomic_set_property(connector, state, property, val);
 
 	return 0;
-#undef CHECK_PROPERTY
 }
 
 static int
@@ -2208,22 +2178,61 @@ intel_sdvo_connector_unregister(struct drm_connector *connector)
 	intel_connector_unregister(connector);
 }
 
+static struct drm_connector_state *
+intel_sdvo_connector_duplicate_state(struct drm_connector *connector)
+{
+	struct intel_sdvo_connector_state *state;
+
+	state = kmemdup(connector->state, sizeof(*state), GFP_KERNEL);
+	if (!state)
+		return NULL;
+
+	__drm_atomic_helper_connector_duplicate_state(connector, &state->base.base);
+	return &state->base.base;
+}
+
 static const struct drm_connector_funcs intel_sdvo_connector_funcs = {
 	.dpms = drm_atomic_helper_connector_dpms,
 	.detect = intel_sdvo_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
-	.set_property = intel_sdvo_set_property,
-	.atomic_get_property = intel_connector_atomic_get_property,
+	.set_property = drm_atomic_helper_connector_set_property,
+	.atomic_get_property = intel_sdvo_connector_atomic_get_property,
+	.atomic_set_property = intel_sdvo_connector_atomic_set_property,
 	.late_register = intel_sdvo_connector_register,
 	.early_unregister = intel_sdvo_connector_unregister,
 	.destroy = intel_sdvo_destroy,
 	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
-	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+	.atomic_duplicate_state = intel_sdvo_connector_duplicate_state,
 };
 
+static int intel_sdvo_atomic_check(struct drm_connector *conn,
+				   struct drm_connector_state *new_conn_state)
+{
+	struct drm_atomic_state *state = new_conn_state->state;
+	struct drm_connector_state *old_conn_state =
+		drm_atomic_get_old_connector_state(state, conn);
+	struct intel_sdvo_connector_state *old_state =
+		to_intel_sdvo_connector_state(old_conn_state);
+	struct intel_sdvo_connector_state *new_state =
+		to_intel_sdvo_connector_state(new_conn_state);
+
+	if (new_conn_state->crtc &&
+	    (memcmp(&old_state->tv, &new_state->tv, sizeof(old_state->tv)) ||
+	     memcmp(&old_conn_state->tv, &new_conn_state->tv, sizeof(old_conn_state->tv)))) {
+		struct drm_crtc_state *crtc_state =
+			drm_atomic_get_new_crtc_state(new_conn_state->state,
+						      new_conn_state->crtc);
+
+		crtc_state->connectors_changed = true;
+	}
+
+	return intel_digital_connector_atomic_check(conn, new_conn_state);
+}
+
 static const struct drm_connector_helper_funcs intel_sdvo_connector_helper_funcs = {
 	.get_modes = intel_sdvo_get_modes,
 	.mode_valid = intel_sdvo_mode_valid,
+	.atomic_check = intel_sdvo_atomic_check,
 };
 
 static void intel_sdvo_enc_destroy(struct drm_encoder *encoder)
@@ -2415,25 +2424,29 @@ intel_sdvo_add_hdmi_properties(struct intel_sdvo *intel_sdvo,
 	intel_attach_force_audio_property(&connector->base.base);
 	if (INTEL_GEN(dev_priv) >= 4 && IS_MOBILE(dev_priv)) {
 		intel_attach_broadcast_rgb_property(&connector->base.base);
-		intel_sdvo->color_range_auto = true;
 	}
 	intel_attach_aspect_ratio_property(&connector->base.base);
-	intel_sdvo->aspect_ratio = HDMI_PICTURE_ASPECT_NONE;
+	connector->base.base.state->picture_aspect_ratio = HDMI_PICTURE_ASPECT_NONE;
 }
 
 static struct intel_sdvo_connector *intel_sdvo_connector_alloc(void)
 {
 	struct intel_sdvo_connector *sdvo_connector;
+	struct intel_sdvo_connector_state *conn_state;
 
 	sdvo_connector = kzalloc(sizeof(*sdvo_connector), GFP_KERNEL);
 	if (!sdvo_connector)
 		return NULL;
 
-	if (intel_connector_init(&sdvo_connector->base) < 0) {
+	conn_state = kzalloc(sizeof(*conn_state), GFP_KERNEL);
+	if (!conn_state) {
 		kfree(sdvo_connector);
 		return NULL;
 	}
 
+	__drm_atomic_helper_connector_reset(&sdvo_connector->base.base,
+					    &conn_state->base.base);
+
 	return sdvo_connector;
 }
 
@@ -2725,31 +2738,30 @@ static bool intel_sdvo_tv_create_property(struct intel_sdvo *intel_sdvo,
 				intel_sdvo_connector->tv_format, i,
 				i, tv_format_names[intel_sdvo_connector->tv_format_supported[i]]);
 
-	intel_sdvo->tv_format_index = intel_sdvo_connector->tv_format_supported[0];
-	drm_object_attach_property(&intel_sdvo_connector->base.base.base,
-				      intel_sdvo_connector->tv_format, 0);
+	intel_sdvo_connector->base.base.state->tv.mode = intel_sdvo_connector->tv_format_supported[0];
+	drm_object_attach_property(&intel_sdvo_connector->base.base.base, 0, 0);
 	return true;
 
 }
 
-#define ENHANCEMENT(name, NAME) do { \
+#define _ENHANCEMENT(state_assignment, name, NAME) do { \
 	if (enhancements.name) { \
 		if (!intel_sdvo_get_value(intel_sdvo, SDVO_CMD_GET_MAX_##NAME, &data_value, 4) || \
 		    !intel_sdvo_get_value(intel_sdvo, SDVO_CMD_GET_##NAME, &response, 2)) \
 			return false; \
-		intel_sdvo_connector->max_##name = data_value[0]; \
-		intel_sdvo_connector->cur_##name = response; \
 		intel_sdvo_connector->name = \
 			drm_property_create_range(dev, 0, #name, 0, data_value[0]); \
 		if (!intel_sdvo_connector->name) return false; \
+		state_assignment = response; \
 		drm_object_attach_property(&connector->base, \
-					      intel_sdvo_connector->name, \
-					      intel_sdvo_connector->cur_##name); \
+					   intel_sdvo_connector->name, 0); \
 		DRM_DEBUG_KMS(#name ": max %d, default %d, current %d\n", \
 			      data_value[0], data_value[1], response); \
 	} \
 } while (0)
 
+#define ENHANCEMENT(state, name, NAME) _ENHANCEMENT((state)->name, name, NAME)
+
 static bool
 intel_sdvo_create_enhance_property_tv(struct intel_sdvo *intel_sdvo,
 				      struct intel_sdvo_connector *intel_sdvo_connector,
@@ -2757,6 +2769,9 @@ intel_sdvo_create_enhance_property_tv(struct intel_sdvo *intel_sdvo,
 {
 	struct drm_device *dev = intel_sdvo->base.base.dev;
 	struct drm_connector *connector = &intel_sdvo_connector->base.base;
+	struct drm_connector_state *conn_state = connector->state;
+	struct intel_sdvo_connector_state *sdvo_state =
+		to_intel_sdvo_connector_state(conn_state);
 	uint16_t response, data_value[2];
 
 	/* when horizontal overscan is supported, Add the left/right  property */
@@ -2771,17 +2786,16 @@ intel_sdvo_create_enhance_property_tv(struct intel_sdvo *intel_sdvo,
 					  &response, 2))
 			return false;
 
+		sdvo_state->tv.overscan_h = response;
+
 		intel_sdvo_connector->max_hscan = data_value[0];
-		intel_sdvo_connector->left_margin = data_value[0] - response;
-		intel_sdvo_connector->right_margin = intel_sdvo_connector->left_margin;
 		intel_sdvo_connector->left =
 			drm_property_create_range(dev, 0, "left_margin", 0, data_value[0]);
 		if (!intel_sdvo_connector->left)
 			return false;
 
 		drm_object_attach_property(&connector->base,
-					      intel_sdvo_connector->left,
-					      intel_sdvo_connector->left_margin);
+					   intel_sdvo_connector->left, 0);
 
 		intel_sdvo_connector->right =
 			drm_property_create_range(dev, 0, "right_margin", 0, data_value[0]);
@@ -2789,8 +2803,7 @@ intel_sdvo_create_enhance_property_tv(struct intel_sdvo *intel_sdvo,
 			return false;
 
 		drm_object_attach_property(&connector->base,
-					      intel_sdvo_connector->right,
-					      intel_sdvo_connector->right_margin);
+					      intel_sdvo_connector->right, 0);
 		DRM_DEBUG_KMS("h_overscan: max %d, "
 			      "default %d, current %d\n",
 			      data_value[0], data_value[1], response);
@@ -2807,9 +2820,9 @@ intel_sdvo_create_enhance_property_tv(struct intel_sdvo *intel_sdvo,
 					  &response, 2))
 			return false;
 
+		sdvo_state->tv.overscan_v = response;
+
 		intel_sdvo_connector->max_vscan = data_value[0];
-		intel_sdvo_connector->top_margin = data_value[0] - response;
-		intel_sdvo_connector->bottom_margin = intel_sdvo_connector->top_margin;
 		intel_sdvo_connector->top =
 			drm_property_create_range(dev, 0,
 					    "top_margin", 0, data_value[0]);
@@ -2817,8 +2830,7 @@ intel_sdvo_create_enhance_property_tv(struct intel_sdvo *intel_sdvo,
 			return false;
 
 		drm_object_attach_property(&connector->base,
-					      intel_sdvo_connector->top,
-					      intel_sdvo_connector->top_margin);
+					   intel_sdvo_connector->top, 0);
 
 		intel_sdvo_connector->bottom =
 			drm_property_create_range(dev, 0,
@@ -2827,40 +2839,37 @@ intel_sdvo_create_enhance_property_tv(struct intel_sdvo *intel_sdvo,
 			return false;
 
 		drm_object_attach_property(&connector->base,
-					      intel_sdvo_connector->bottom,
-					      intel_sdvo_connector->bottom_margin);
+					      intel_sdvo_connector->bottom, 0);
 		DRM_DEBUG_KMS("v_overscan: max %d, "
 			      "default %d, current %d\n",
 			      data_value[0], data_value[1], response);
 	}
 
-	ENHANCEMENT(hpos, HPOS);
-	ENHANCEMENT(vpos, VPOS);
-	ENHANCEMENT(saturation, SATURATION);
-	ENHANCEMENT(contrast, CONTRAST);
-	ENHANCEMENT(hue, HUE);
-	ENHANCEMENT(sharpness, SHARPNESS);
-	ENHANCEMENT(brightness, BRIGHTNESS);
-	ENHANCEMENT(flicker_filter, FLICKER_FILTER);
-	ENHANCEMENT(flicker_filter_adaptive, FLICKER_FILTER_ADAPTIVE);
-	ENHANCEMENT(flicker_filter_2d, FLICKER_FILTER_2D);
-	ENHANCEMENT(tv_chroma_filter, TV_CHROMA_FILTER);
-	ENHANCEMENT(tv_luma_filter, TV_LUMA_FILTER);
+	ENHANCEMENT(&sdvo_state->tv, hpos, HPOS);
+	ENHANCEMENT(&sdvo_state->tv, vpos, VPOS);
+	ENHANCEMENT(&conn_state->tv, saturation, SATURATION);
+	ENHANCEMENT(&conn_state->tv, contrast, CONTRAST);
+	ENHANCEMENT(&conn_state->tv, hue, HUE);
+	ENHANCEMENT(&conn_state->tv, brightness, BRIGHTNESS);
+	ENHANCEMENT(&sdvo_state->tv, sharpness, SHARPNESS);
+	ENHANCEMENT(&sdvo_state->tv, flicker_filter, FLICKER_FILTER);
+	ENHANCEMENT(&sdvo_state->tv, flicker_filter_adaptive, FLICKER_FILTER_ADAPTIVE);
+	ENHANCEMENT(&sdvo_state->tv, flicker_filter_2d, FLICKER_FILTER_2D);
+	_ENHANCEMENT(sdvo_state->tv.chroma_filter, tv_chroma_filter, TV_CHROMA_FILTER);
+	_ENHANCEMENT(sdvo_state->tv.luma_filter, tv_luma_filter, TV_LUMA_FILTER);
 
 	if (enhancements.dot_crawl) {
 		if (!intel_sdvo_get_value(intel_sdvo, SDVO_CMD_GET_DOT_CRAWL, &response, 2))
 			return false;
 
-		intel_sdvo_connector->max_dot_crawl = 1;
-		intel_sdvo_connector->cur_dot_crawl = response & 0x1;
+		sdvo_state->tv.dot_crawl = response & 0x1;
 		intel_sdvo_connector->dot_crawl =
 			drm_property_create_range(dev, 0, "dot_crawl", 0, 1);
 		if (!intel_sdvo_connector->dot_crawl)
 			return false;
 
 		drm_object_attach_property(&connector->base,
-					      intel_sdvo_connector->dot_crawl,
-					      intel_sdvo_connector->cur_dot_crawl);
+					   intel_sdvo_connector->dot_crawl, 0);
 		DRM_DEBUG_KMS("dot crawl: current %d\n", response);
 	}
 
@@ -2876,11 +2885,12 @@ intel_sdvo_create_enhance_property_lvds(struct intel_sdvo *intel_sdvo,
 	struct drm_connector *connector = &intel_sdvo_connector->base.base;
 	uint16_t response, data_value[2];
 
-	ENHANCEMENT(brightness, BRIGHTNESS);
+	ENHANCEMENT(&connector->state->tv, brightness, BRIGHTNESS);
 
 	return true;
 }
 #undef ENHANCEMENT
+#undef _ENHANCEMENT
 
 static bool intel_sdvo_create_enhance_property(struct intel_sdvo *intel_sdvo,
 					       struct intel_sdvo_connector *intel_sdvo_connector)
@@ -2892,11 +2902,10 @@ static bool intel_sdvo_create_enhance_property(struct intel_sdvo *intel_sdvo,
 
 	BUILD_BUG_ON(sizeof(enhancements) != 2);
 
-	enhancements.response = 0;
-	intel_sdvo_get_value(intel_sdvo,
-			     SDVO_CMD_GET_SUPPORTED_ENHANCEMENTS,
-			     &enhancements, sizeof(enhancements));
-	if (enhancements.response == 0) {
+	if (!intel_sdvo_get_value(intel_sdvo,
+				  SDVO_CMD_GET_SUPPORTED_ENHANCEMENTS,
+				  &enhancements, sizeof(enhancements)) ||
+	    enhancements.response == 0) {
 		DRM_DEBUG_KMS("No enhancement is supported\n");
 		return true;
 	}
diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c
index e6517edcd16b..0c650c2cbca8 100644
--- a/drivers/gpu/drm/i915/intel_sprite.c
+++ b/drivers/gpu/drm/i915/intel_sprite.c
@@ -231,16 +231,14 @@ void intel_pipe_update_end(struct intel_crtc *crtc, struct intel_flip_work *work
 }
 
 static void
-skl_update_plane(struct drm_plane *drm_plane,
+skl_update_plane(struct intel_plane *plane,
 		 const struct intel_crtc_state *crtc_state,
 		 const struct intel_plane_state *plane_state)
 {
-	struct drm_device *dev = drm_plane->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_plane *intel_plane = to_intel_plane(drm_plane);
-	struct drm_framebuffer *fb = plane_state->base.fb;
-	enum plane_id plane_id = intel_plane->id;
-	enum pipe pipe = intel_plane->pipe;
+	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+	enum plane_id plane_id = plane->id;
+	enum pipe pipe = plane->pipe;
 	u32 plane_ctl = plane_state->ctl;
 	const struct drm_intel_sprite_colorkey *key = &plane_state->ckey;
 	u32 surf_addr = plane_state->main.offset;
@@ -309,13 +307,11 @@ skl_update_plane(struct drm_plane *drm_plane,
 }
 
 static void
-skl_disable_plane(struct drm_plane *dplane, struct drm_crtc *crtc)
+skl_disable_plane(struct intel_plane *plane, struct intel_crtc *crtc)
 {
-	struct drm_device *dev = dplane->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_plane *intel_plane = to_intel_plane(dplane);
-	enum plane_id plane_id = intel_plane->id;
-	enum pipe pipe = intel_plane->pipe;
+	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
+	enum plane_id plane_id = plane->id;
+	enum pipe pipe = plane->pipe;
 	unsigned long irqflags;
 
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
@@ -329,10 +325,10 @@ skl_disable_plane(struct drm_plane *dplane, struct drm_crtc *crtc)
 }
 
 static void
-chv_update_csc(struct intel_plane *intel_plane, uint32_t format)
+chv_update_csc(struct intel_plane *plane, uint32_t format)
 {
-	struct drm_i915_private *dev_priv = to_i915(intel_plane->base.dev);
-	enum plane_id plane_id = intel_plane->id;
+	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
+	enum plane_id plane_id = plane->id;
 
 	/* Seems RGB data bypasses the CSC always */
 	if (!format_is_yuv(format))
@@ -419,10 +415,10 @@ static u32 vlv_sprite_ctl(const struct intel_crtc_state *crtc_state,
 	if (fb->modifier == I915_FORMAT_MOD_X_TILED)
 		sprctl |= SP_TILED;
 
-	if (rotation & DRM_ROTATE_180)
+	if (rotation & DRM_MODE_ROTATE_180)
 		sprctl |= SP_ROTATE_180;
 
-	if (rotation & DRM_REFLECT_X)
+	if (rotation & DRM_MODE_REFLECT_X)
 		sprctl |= SP_MIRROR;
 
 	if (key->flags & I915_SET_COLORKEY_SOURCE)
@@ -432,16 +428,14 @@ static u32 vlv_sprite_ctl(const struct intel_crtc_state *crtc_state,
 }
 
 static void
-vlv_update_plane(struct drm_plane *dplane,
+vlv_update_plane(struct intel_plane *plane,
 		 const struct intel_crtc_state *crtc_state,
 		 const struct intel_plane_state *plane_state)
 {
-	struct drm_device *dev = dplane->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_plane *intel_plane = to_intel_plane(dplane);
-	struct drm_framebuffer *fb = plane_state->base.fb;
-	enum pipe pipe = intel_plane->pipe;
-	enum plane_id plane_id = intel_plane->id;
+	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+	enum pipe pipe = plane->pipe;
+	enum plane_id plane_id = plane->id;
 	u32 sprctl = plane_state->ctl;
 	u32 sprsurf_offset = plane_state->main.offset;
 	u32 linear_offset;
@@ -463,7 +457,7 @@ vlv_update_plane(struct drm_plane *dplane,
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
 
 	if (IS_CHERRYVIEW(dev_priv) && pipe == PIPE_B)
-		chv_update_csc(intel_plane, fb->format->format);
+		chv_update_csc(plane, fb->format->format);
 
 	if (key->flags) {
 		I915_WRITE_FW(SPKEYMINVAL(pipe, plane_id), key->min_value);
@@ -490,13 +484,11 @@ vlv_update_plane(struct drm_plane *dplane,
 }
 
 static void
-vlv_disable_plane(struct drm_plane *dplane, struct drm_crtc *crtc)
+vlv_disable_plane(struct intel_plane *plane, struct intel_crtc *crtc)
 {
-	struct drm_device *dev = dplane->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_plane *intel_plane = to_intel_plane(dplane);
-	enum pipe pipe = intel_plane->pipe;
-	enum plane_id plane_id = intel_plane->id;
+	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
+	enum pipe pipe = plane->pipe;
+	enum plane_id plane_id = plane->id;
 	unsigned long irqflags;
 
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
@@ -554,7 +546,7 @@ static u32 ivb_sprite_ctl(const struct intel_crtc_state *crtc_state,
 	if (fb->modifier == I915_FORMAT_MOD_X_TILED)
 		sprctl |= SPRITE_TILED;
 
-	if (rotation & DRM_ROTATE_180)
+	if (rotation & DRM_MODE_ROTATE_180)
 		sprctl |= SPRITE_ROTATE_180;
 
 	if (key->flags & I915_SET_COLORKEY_DESTINATION)
@@ -566,15 +558,13 @@ static u32 ivb_sprite_ctl(const struct intel_crtc_state *crtc_state,
 }
 
 static void
-ivb_update_plane(struct drm_plane *plane,
+ivb_update_plane(struct intel_plane *plane,
 		 const struct intel_crtc_state *crtc_state,
 		 const struct intel_plane_state *plane_state)
 {
-	struct drm_device *dev = plane->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_plane *intel_plane = to_intel_plane(plane);
-	struct drm_framebuffer *fb = plane_state->base.fb;
-	enum pipe pipe = intel_plane->pipe;
+	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+	enum pipe pipe = plane->pipe;
 	u32 sprctl = plane_state->ctl, sprscale = 0;
 	u32 sprsurf_offset = plane_state->main.offset;
 	u32 linear_offset;
@@ -621,7 +611,7 @@ ivb_update_plane(struct drm_plane *plane,
 		I915_WRITE_FW(SPRLINOFF(pipe), linear_offset);
 
 	I915_WRITE_FW(SPRSIZE(pipe), (crtc_h << 16) | crtc_w);
-	if (intel_plane->can_scale)
+	if (plane->can_scale)
 		I915_WRITE_FW(SPRSCALE(pipe), sprscale);
 	I915_WRITE_FW(SPRCTL(pipe), sprctl);
 	I915_WRITE_FW(SPRSURF(pipe),
@@ -632,19 +622,17 @@ ivb_update_plane(struct drm_plane *plane,
 }
 
 static void
-ivb_disable_plane(struct drm_plane *plane, struct drm_crtc *crtc)
+ivb_disable_plane(struct intel_plane *plane, struct intel_crtc *crtc)
 {
-	struct drm_device *dev = plane->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_plane *intel_plane = to_intel_plane(plane);
-	int pipe = intel_plane->pipe;
+	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
+	enum pipe pipe = plane->pipe;
 	unsigned long irqflags;
 
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
 
 	I915_WRITE_FW(SPRCTL(pipe), 0);
 	/* Can't leave the scaler enabled... */
-	if (intel_plane->can_scale)
+	if (plane->can_scale)
 		I915_WRITE_FW(SPRSCALE(pipe), 0);
 
 	I915_WRITE_FW(SPRSURF(pipe), 0);
@@ -653,7 +641,7 @@ ivb_disable_plane(struct drm_plane *plane, struct drm_crtc *crtc)
 	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
 }
 
-static u32 ilk_sprite_ctl(const struct intel_crtc_state *crtc_state,
+static u32 g4x_sprite_ctl(const struct intel_crtc_state *crtc_state,
 			  const struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv =
@@ -695,7 +683,7 @@ static u32 ilk_sprite_ctl(const struct intel_crtc_state *crtc_state,
 	if (fb->modifier == I915_FORMAT_MOD_X_TILED)
 		dvscntr |= DVS_TILED;
 
-	if (rotation & DRM_ROTATE_180)
+	if (rotation & DRM_MODE_ROTATE_180)
 		dvscntr |= DVS_ROTATE_180;
 
 	if (key->flags & I915_SET_COLORKEY_DESTINATION)
@@ -707,15 +695,13 @@ static u32 ilk_sprite_ctl(const struct intel_crtc_state *crtc_state,
 }
 
 static void
-ilk_update_plane(struct drm_plane *plane,
+g4x_update_plane(struct intel_plane *plane,
 		 const struct intel_crtc_state *crtc_state,
 		 const struct intel_plane_state *plane_state)
 {
-	struct drm_device *dev = plane->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_plane *intel_plane = to_intel_plane(plane);
-	struct drm_framebuffer *fb = plane_state->base.fb;
-	int pipe = intel_plane->pipe;
+	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
+	const struct drm_framebuffer *fb = plane_state->base.fb;
+	enum pipe pipe = plane->pipe;
 	u32 dvscntr = plane_state->ctl, dvsscale = 0;
 	u32 dvssurf_offset = plane_state->main.offset;
 	u32 linear_offset;
@@ -768,12 +754,10 @@ ilk_update_plane(struct drm_plane *plane,
 }
 
 static void
-ilk_disable_plane(struct drm_plane *plane, struct drm_crtc *crtc)
+g4x_disable_plane(struct intel_plane *plane, struct intel_crtc *crtc)
 {
-	struct drm_device *dev = plane->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_plane *intel_plane = to_intel_plane(plane);
-	int pipe = intel_plane->pipe;
+	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
+	enum pipe pipe = plane->pipe;
 	unsigned long irqflags;
 
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
@@ -789,14 +773,12 @@ ilk_disable_plane(struct drm_plane *plane, struct drm_crtc *crtc)
 }
 
 static int
-intel_check_sprite_plane(struct drm_plane *plane,
+intel_check_sprite_plane(struct intel_plane *plane,
 			 struct intel_crtc_state *crtc_state,
 			 struct intel_plane_state *state)
 {
-	struct drm_i915_private *dev_priv = to_i915(plane->dev);
-	struct drm_crtc *crtc = state->base.crtc;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	struct intel_plane *intel_plane = to_intel_plane(plane);
+	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
 	struct drm_framebuffer *fb = state->base.fb;
 	int crtc_x, crtc_y;
 	unsigned int crtc_w, crtc_h;
@@ -818,7 +800,7 @@ intel_check_sprite_plane(struct drm_plane *plane,
 	}
 
 	/* Don't modify another pipe's plane */
-	if (intel_plane->pipe != intel_crtc->pipe) {
+	if (plane->pipe != crtc->pipe) {
 		DRM_DEBUG_KMS("Wrong plane <-> crtc mapping\n");
 		return -EINVAL;
 	}
@@ -835,16 +817,16 @@ intel_check_sprite_plane(struct drm_plane *plane,
 		if (state->ckey.flags == I915_SET_COLORKEY_NONE) {
 			can_scale = 1;
 			min_scale = 1;
-			max_scale = skl_max_scale(intel_crtc, crtc_state);
+			max_scale = skl_max_scale(crtc, crtc_state);
 		} else {
 			can_scale = 0;
 			min_scale = DRM_PLANE_HELPER_NO_SCALING;
 			max_scale = DRM_PLANE_HELPER_NO_SCALING;
 		}
 	} else {
-		can_scale = intel_plane->can_scale;
-		max_scale = intel_plane->max_downscale << 16;
-		min_scale = intel_plane->can_scale ? 1 : (1 << 16);
+		can_scale = plane->can_scale;
+		max_scale = plane->max_downscale << 16;
+		min_scale = plane->can_scale ? 1 : (1 << 16);
 	}
 
 	/*
@@ -988,7 +970,7 @@ intel_check_sprite_plane(struct drm_plane *plane,
 		if (ret)
 			return ret;
 
-		state->ctl = ilk_sprite_ctl(crtc_state, state);
+		state->ctl = g4x_sprite_ctl(crtc_state, state);
 	}
 
 	return 0;
@@ -1048,7 +1030,7 @@ out:
 	return ret;
 }
 
-static const uint32_t ilk_plane_formats[] = {
+static const uint32_t g4x_plane_formats[] = {
 	DRM_FORMAT_XRGB8888,
 	DRM_FORMAT_YUYV,
 	DRM_FORMAT_YVYU,
@@ -1152,29 +1134,29 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv,
 		intel_plane->can_scale = true;
 		intel_plane->max_downscale = 16;
 
-		intel_plane->update_plane = ilk_update_plane;
-		intel_plane->disable_plane = ilk_disable_plane;
+		intel_plane->update_plane = g4x_update_plane;
+		intel_plane->disable_plane = g4x_disable_plane;
 
 		if (IS_GEN6(dev_priv)) {
 			plane_formats = snb_plane_formats;
 			num_plane_formats = ARRAY_SIZE(snb_plane_formats);
 		} else {
-			plane_formats = ilk_plane_formats;
-			num_plane_formats = ARRAY_SIZE(ilk_plane_formats);
+			plane_formats = g4x_plane_formats;
+			num_plane_formats = ARRAY_SIZE(g4x_plane_formats);
 		}
 	}
 
 	if (INTEL_GEN(dev_priv) >= 9) {
 		supported_rotations =
-			DRM_ROTATE_0 | DRM_ROTATE_90 |
-			DRM_ROTATE_180 | DRM_ROTATE_270;
+			DRM_MODE_ROTATE_0 | DRM_MODE_ROTATE_90 |
+			DRM_MODE_ROTATE_180 | DRM_MODE_ROTATE_270;
 	} else if (IS_CHERRYVIEW(dev_priv) && pipe == PIPE_B) {
 		supported_rotations =
-			DRM_ROTATE_0 | DRM_ROTATE_180 |
-			DRM_REFLECT_X;
+			DRM_MODE_ROTATE_0 | DRM_MODE_ROTATE_180 |
+			DRM_MODE_REFLECT_X;
 	} else {
 		supported_rotations =
-			DRM_ROTATE_0 | DRM_ROTATE_180;
+			DRM_MODE_ROTATE_0 | DRM_MODE_ROTATE_180;
 	}
 
 	intel_plane->pipe = pipe;
@@ -1201,7 +1183,7 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv,
 		goto fail;
 
 	drm_plane_create_rotation_property(&intel_plane->base,
-					   DRM_ROTATE_0,
+					   DRM_MODE_ROTATE_0,
 					   supported_rotations);
 
 	drm_plane_helper_add(&intel_plane->base, &intel_plane_helper_funcs);
diff --git a/drivers/gpu/drm/i915/intel_tv.c b/drivers/gpu/drm/i915/intel_tv.c
index e077c2a9e694..784df024e230 100644
--- a/drivers/gpu/drm/i915/intel_tv.c
+++ b/drivers/gpu/drm/i915/intel_tv.c
@@ -48,41 +48,6 @@ struct intel_tv {
 	struct intel_encoder base;
 
 	int type;
-	const char *tv_format;
-	int margin[4];
-	u32 save_TV_H_CTL_1;
-	u32 save_TV_H_CTL_2;
-	u32 save_TV_H_CTL_3;
-	u32 save_TV_V_CTL_1;
-	u32 save_TV_V_CTL_2;
-	u32 save_TV_V_CTL_3;
-	u32 save_TV_V_CTL_4;
-	u32 save_TV_V_CTL_5;
-	u32 save_TV_V_CTL_6;
-	u32 save_TV_V_CTL_7;
-	u32 save_TV_SC_CTL_1, save_TV_SC_CTL_2, save_TV_SC_CTL_3;
-
-	u32 save_TV_CSC_Y;
-	u32 save_TV_CSC_Y2;
-	u32 save_TV_CSC_U;
-	u32 save_TV_CSC_U2;
-	u32 save_TV_CSC_V;
-	u32 save_TV_CSC_V2;
-	u32 save_TV_CLR_KNOBS;
-	u32 save_TV_CLR_LEVEL;
-	u32 save_TV_WIN_POS;
-	u32 save_TV_WIN_SIZE;
-	u32 save_TV_FILTER_CTL_1;
-	u32 save_TV_FILTER_CTL_2;
-	u32 save_TV_FILTER_CTL_3;
-
-	u32 save_TV_H_LUMA[60];
-	u32 save_TV_H_CHROMA[60];
-	u32 save_TV_V_LUMA[43];
-	u32 save_TV_V_CHROMA[43];
-
-	u32 save_TV_DAC;
-	u32 save_TV_CTL;
 };
 
 struct video_levels {
@@ -873,32 +838,18 @@ intel_disable_tv(struct intel_encoder *encoder,
 	I915_WRITE(TV_CTL, I915_READ(TV_CTL) & ~TV_ENC_ENABLE);
 }
 
-static const struct tv_mode *
-intel_tv_mode_lookup(const char *tv_format)
+static const struct tv_mode *intel_tv_mode_find(struct drm_connector_state *conn_state)
 {
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(tv_modes); i++) {
-		const struct tv_mode *tv_mode = &tv_modes[i];
+	int format = conn_state->tv.mode;
 
-		if (!strcmp(tv_format, tv_mode->name))
-			return tv_mode;
-	}
-	return NULL;
-}
-
-static const struct tv_mode *
-intel_tv_mode_find(struct intel_tv *intel_tv)
-{
-	return intel_tv_mode_lookup(intel_tv->tv_format);
+	return &tv_modes[format];
 }
 
 static enum drm_mode_status
 intel_tv_mode_valid(struct drm_connector *connector,
 		    struct drm_display_mode *mode)
 {
-	struct intel_tv *intel_tv = intel_attached_tv(connector);
-	const struct tv_mode *tv_mode = intel_tv_mode_find(intel_tv);
+	const struct tv_mode *tv_mode = intel_tv_mode_find(connector->state);
 	int max_dotclk = to_i915(connector->dev)->max_dotclk_freq;
 
 	if (mode->clock > max_dotclk)
@@ -925,8 +876,7 @@ intel_tv_compute_config(struct intel_encoder *encoder,
 			struct intel_crtc_state *pipe_config,
 			struct drm_connector_state *conn_state)
 {
-	struct intel_tv *intel_tv = enc_to_tv(encoder);
-	const struct tv_mode *tv_mode = intel_tv_mode_find(intel_tv);
+	const struct tv_mode *tv_mode = intel_tv_mode_find(conn_state);
 
 	if (!tv_mode)
 		return false;
@@ -1032,7 +982,7 @@ static void intel_tv_pre_enable(struct intel_encoder *encoder,
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(encoder->base.crtc);
 	struct intel_tv *intel_tv = enc_to_tv(encoder);
-	const struct tv_mode *tv_mode = intel_tv_mode_find(intel_tv);
+	const struct tv_mode *tv_mode = intel_tv_mode_find(conn_state);
 	u32 tv_ctl;
 	u32 scctl1, scctl2, scctl3;
 	int i, j;
@@ -1135,12 +1085,12 @@ static void intel_tv_pre_enable(struct intel_encoder *encoder,
 	else
 		ysize = 2*tv_mode->nbr_end + 1;
 
-	xpos += intel_tv->margin[TV_MARGIN_LEFT];
-	ypos += intel_tv->margin[TV_MARGIN_TOP];
-	xsize -= (intel_tv->margin[TV_MARGIN_LEFT] +
-		  intel_tv->margin[TV_MARGIN_RIGHT]);
-	ysize -= (intel_tv->margin[TV_MARGIN_TOP] +
-		  intel_tv->margin[TV_MARGIN_BOTTOM]);
+	xpos += conn_state->tv.margins.left;
+	ypos += conn_state->tv.margins.top;
+	xsize -= (conn_state->tv.margins.left +
+		  conn_state->tv.margins.right);
+	ysize -= (conn_state->tv.margins.top +
+		  conn_state->tv.margins.bottom);
 	I915_WRITE(TV_WIN_POS, (xpos<<16)|ypos);
 	I915_WRITE(TV_WIN_SIZE, (xsize<<16)|ysize);
 
@@ -1288,7 +1238,7 @@ intel_tv_detect_type(struct intel_tv *intel_tv,
 static void intel_tv_find_better_format(struct drm_connector *connector)
 {
 	struct intel_tv *intel_tv = intel_attached_tv(connector);
-	const struct tv_mode *tv_mode = intel_tv_mode_find(intel_tv);
+	const struct tv_mode *tv_mode = intel_tv_mode_find(connector->state);
 	int i;
 
 	if ((intel_tv->type == DRM_MODE_CONNECTOR_Component) ==
@@ -1304,9 +1254,7 @@ static void intel_tv_find_better_format(struct drm_connector *connector)
 			break;
 	}
 
-	intel_tv->tv_format = tv_mode->name;
-	drm_object_property_set_value(&connector->base,
-		connector->dev->mode_config.tv_mode_property, i);
+	connector->state->tv.mode = i;
 }
 
 /**
@@ -1347,16 +1295,15 @@ intel_tv_detect(struct drm_connector *connector,
 				connector_status_connected;
 		} else
 			status = connector_status_unknown;
-	} else
-		return connector->status;
 
-	if (status != connector_status_connected)
-		return status;
-
-	intel_tv->type = type;
-	intel_tv_find_better_format(connector);
+		if (status == connector_status_connected) {
+			intel_tv->type = type;
+			intel_tv_find_better_format(connector);
+		}
 
-	return connector_status_connected;
+		return status;
+	} else
+		return connector->status;
 }
 
 static const struct input_res {
@@ -1376,12 +1323,9 @@ static const struct input_res {
  * Chose preferred mode  according to line number of TV format
  */
 static void
-intel_tv_chose_preferred_modes(struct drm_connector *connector,
+intel_tv_choose_preferred_modes(const struct tv_mode *tv_mode,
 			       struct drm_display_mode *mode_ptr)
 {
-	struct intel_tv *intel_tv = intel_attached_tv(connector);
-	const struct tv_mode *tv_mode = intel_tv_mode_find(intel_tv);
-
 	if (tv_mode->nbr_end < 480 && mode_ptr->vdisplay == 480)
 		mode_ptr->type |= DRM_MODE_TYPE_PREFERRED;
 	else if (tv_mode->nbr_end > 480) {
@@ -1404,8 +1348,7 @@ static int
 intel_tv_get_modes(struct drm_connector *connector)
 {
 	struct drm_display_mode *mode_ptr;
-	struct intel_tv *intel_tv = intel_attached_tv(connector);
-	const struct tv_mode *tv_mode = intel_tv_mode_find(intel_tv);
+	const struct tv_mode *tv_mode = intel_tv_mode_find(connector->state);
 	int j, count = 0;
 	u64 tmp;
 
@@ -1448,7 +1391,7 @@ intel_tv_get_modes(struct drm_connector *connector)
 		mode_ptr->clock = (int) tmp;
 
 		mode_ptr->type = DRM_MODE_TYPE_DRIVER;
-		intel_tv_chose_preferred_modes(connector, mode_ptr);
+		intel_tv_choose_preferred_modes(tv_mode, mode_ptr);
 		drm_mode_probed_add(connector, mode_ptr);
 		count++;
 	}
@@ -1463,74 +1406,47 @@ intel_tv_destroy(struct drm_connector *connector)
 	kfree(connector);
 }
 
-
-static int
-intel_tv_set_property(struct drm_connector *connector, struct drm_property *property,
-		      uint64_t val)
-{
-	struct drm_device *dev = connector->dev;
-	struct intel_tv *intel_tv = intel_attached_tv(connector);
-	struct drm_crtc *crtc = intel_tv->base.base.crtc;
-	int ret = 0;
-	bool changed = false;
-
-	ret = drm_object_property_set_value(&connector->base, property, val);
-	if (ret < 0)
-		goto out;
-
-	if (property == dev->mode_config.tv_left_margin_property &&
-		intel_tv->margin[TV_MARGIN_LEFT] != val) {
-		intel_tv->margin[TV_MARGIN_LEFT] = val;
-		changed = true;
-	} else if (property == dev->mode_config.tv_right_margin_property &&
-		intel_tv->margin[TV_MARGIN_RIGHT] != val) {
-		intel_tv->margin[TV_MARGIN_RIGHT] = val;
-		changed = true;
-	} else if (property == dev->mode_config.tv_top_margin_property &&
-		intel_tv->margin[TV_MARGIN_TOP] != val) {
-		intel_tv->margin[TV_MARGIN_TOP] = val;
-		changed = true;
-	} else if (property == dev->mode_config.tv_bottom_margin_property &&
-		intel_tv->margin[TV_MARGIN_BOTTOM] != val) {
-		intel_tv->margin[TV_MARGIN_BOTTOM] = val;
-		changed = true;
-	} else if (property == dev->mode_config.tv_mode_property) {
-		if (val >= ARRAY_SIZE(tv_modes)) {
-			ret = -EINVAL;
-			goto out;
-		}
-		if (!strcmp(intel_tv->tv_format, tv_modes[val].name))
-			goto out;
-
-		intel_tv->tv_format = tv_modes[val].name;
-		changed = true;
-	} else {
-		ret = -EINVAL;
-		goto out;
-	}
-
-	if (changed && crtc)
-		intel_crtc_restore_mode(crtc);
-out:
-	return ret;
-}
-
 static const struct drm_connector_funcs intel_tv_connector_funcs = {
 	.dpms = drm_atomic_helper_connector_dpms,
 	.late_register = intel_connector_register,
 	.early_unregister = intel_connector_unregister,
 	.destroy = intel_tv_destroy,
-	.set_property = intel_tv_set_property,
-	.atomic_get_property = intel_connector_atomic_get_property,
+	.set_property = drm_atomic_helper_connector_set_property,
 	.fill_modes = drm_helper_probe_single_connector_modes,
 	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
 	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
 };
 
+static int intel_tv_atomic_check(struct drm_connector *connector,
+				 struct drm_connector_state *new_state)
+{
+	struct drm_crtc_state *new_crtc_state;
+	struct drm_connector_state *old_state;
+
+	if (!new_state->crtc)
+		return 0;
+
+	old_state = drm_atomic_get_old_connector_state(new_state->state, connector);
+	new_crtc_state = drm_atomic_get_new_crtc_state(new_state->state, new_state->crtc);
+
+	if (old_state->tv.mode != new_state->tv.mode ||
+	    old_state->tv.margins.left != new_state->tv.margins.left ||
+	    old_state->tv.margins.right != new_state->tv.margins.right ||
+	    old_state->tv.margins.top != new_state->tv.margins.top ||
+	    old_state->tv.margins.bottom != new_state->tv.margins.bottom) {
+		/* Force a modeset. */
+
+		new_crtc_state->connectors_changed = true;
+	}
+
+	return 0;
+}
+
 static const struct drm_connector_helper_funcs intel_tv_connector_helper_funcs = {
 	.detect_ctx = intel_tv_detect,
 	.mode_valid = intel_tv_mode_valid,
 	.get_modes = intel_tv_get_modes,
+	.atomic_check = intel_tv_atomic_check,
 };
 
 static const struct drm_encoder_funcs intel_tv_enc_funcs = {
@@ -1548,6 +1464,7 @@ intel_tv_init(struct drm_i915_private *dev_priv)
 	u32 tv_dac_on, tv_dac_off, save_tv_dac;
 	const char *tv_format_names[ARRAY_SIZE(tv_modes)];
 	int i, initial_mode = 0;
+	struct drm_connector_state *state;
 
 	if ((I915_READ(TV_CTL) & TV_FUSE_STATE_MASK) == TV_FUSE_STATE_DISABLED)
 		return;
@@ -1593,6 +1510,7 @@ intel_tv_init(struct drm_i915_private *dev_priv)
 
 	intel_encoder = &intel_tv->base;
 	connector = &intel_connector->base;
+	state = connector->state;
 
 	/* The documentation, for the older chipsets at least, recommend
 	 * using a polling method rather than hotplug detection for TVs.
@@ -1630,12 +1548,12 @@ intel_tv_init(struct drm_i915_private *dev_priv)
 	intel_tv->type = DRM_MODE_CONNECTOR_Unknown;
 
 	/* BIOS margin values */
-	intel_tv->margin[TV_MARGIN_LEFT] = 54;
-	intel_tv->margin[TV_MARGIN_TOP] = 36;
-	intel_tv->margin[TV_MARGIN_RIGHT] = 46;
-	intel_tv->margin[TV_MARGIN_BOTTOM] = 37;
+	state->tv.margins.left = 54;
+	state->tv.margins.top = 36;
+	state->tv.margins.right = 46;
+	state->tv.margins.bottom = 37;
 
-	intel_tv->tv_format = tv_modes[initial_mode].name;
+	state->tv.mode = initial_mode;
 
 	drm_connector_helper_add(connector, &intel_tv_connector_helper_funcs);
 	connector->interlace_allowed = false;
@@ -1649,17 +1567,17 @@ intel_tv_init(struct drm_i915_private *dev_priv)
 				      tv_format_names);
 
 	drm_object_attach_property(&connector->base, dev->mode_config.tv_mode_property,
-				   initial_mode);
+				   state->tv.mode);
 	drm_object_attach_property(&connector->base,
 				   dev->mode_config.tv_left_margin_property,
-				   intel_tv->margin[TV_MARGIN_LEFT]);
+				   state->tv.margins.left);
 	drm_object_attach_property(&connector->base,
 				   dev->mode_config.tv_top_margin_property,
-				   intel_tv->margin[TV_MARGIN_TOP]);
+				   state->tv.margins.top);
 	drm_object_attach_property(&connector->base,
 				   dev->mode_config.tv_right_margin_property,
-				   intel_tv->margin[TV_MARGIN_RIGHT]);
+				   state->tv.margins.right);
 	drm_object_attach_property(&connector->base,
 				   dev->mode_config.tv_bottom_margin_property,
-				   intel_tv->margin[TV_MARGIN_BOTTOM]);
+				   state->tv.margins.bottom);
 }
diff --git a/drivers/gpu/drm/i915/intel_uc.c b/drivers/gpu/drm/i915/intel_uc.c
index c117424f1f50..27e072cc96eb 100644
--- a/drivers/gpu/drm/i915/intel_uc.c
+++ b/drivers/gpu/drm/i915/intel_uc.c
@@ -94,12 +94,22 @@ void intel_uc_sanitize_options(struct drm_i915_private *dev_priv)
 		i915.enable_guc_submission = HAS_GUC_SCHED(dev_priv);
 }
 
+static void guc_write_irq_trigger(struct intel_guc *guc)
+{
+	struct drm_i915_private *dev_priv = guc_to_i915(guc);
+
+	I915_WRITE(GUC_SEND_INTERRUPT, GUC_SEND_TRIGGER);
+}
+
 void intel_uc_init_early(struct drm_i915_private *dev_priv)
 {
 	struct intel_guc *guc = &dev_priv->guc;
 
+	intel_guc_ct_init_early(&guc->ct);
+
 	mutex_init(&guc->send_mutex);
-	guc->send = intel_guc_send_mmio;
+	guc->send = intel_guc_send_nop;
+	guc->notify = guc_write_irq_trigger;
 }
 
 static void fetch_uc_fw(struct drm_i915_private *dev_priv,
@@ -252,13 +262,81 @@ void intel_uc_fini_fw(struct drm_i915_private *dev_priv)
 	__intel_uc_fw_fini(&dev_priv->huc.fw);
 }
 
+static inline i915_reg_t guc_send_reg(struct intel_guc *guc, u32 i)
+{
+	GEM_BUG_ON(!guc->send_regs.base);
+	GEM_BUG_ON(!guc->send_regs.count);
+	GEM_BUG_ON(i >= guc->send_regs.count);
+
+	return _MMIO(guc->send_regs.base + 4 * i);
+}
+
+static void guc_init_send_regs(struct intel_guc *guc)
+{
+	struct drm_i915_private *dev_priv = guc_to_i915(guc);
+	enum forcewake_domains fw_domains = 0;
+	unsigned int i;
+
+	guc->send_regs.base = i915_mmio_reg_offset(SOFT_SCRATCH(0));
+	guc->send_regs.count = SOFT_SCRATCH_COUNT - 1;
+
+	for (i = 0; i < guc->send_regs.count; i++) {
+		fw_domains |= intel_uncore_forcewake_for_reg(dev_priv,
+					guc_send_reg(guc, i),
+					FW_REG_READ | FW_REG_WRITE);
+	}
+	guc->send_regs.fw_domains = fw_domains;
+}
+
+static void guc_capture_load_err_log(struct intel_guc *guc)
+{
+	if (!guc->log.vma || i915.guc_log_level < 0)
+		return;
+
+	if (!guc->load_err_log)
+		guc->load_err_log = i915_gem_object_get(guc->log.vma->obj);
+
+	return;
+}
+
+static void guc_free_load_err_log(struct intel_guc *guc)
+{
+	if (guc->load_err_log)
+		i915_gem_object_put(guc->load_err_log);
+}
+
+static int guc_enable_communication(struct intel_guc *guc)
+{
+	struct drm_i915_private *dev_priv = guc_to_i915(guc);
+
+	guc_init_send_regs(guc);
+
+	if (HAS_GUC_CT(dev_priv))
+		return intel_guc_enable_ct(guc);
+
+	guc->send = intel_guc_send_mmio;
+	return 0;
+}
+
+static void guc_disable_communication(struct intel_guc *guc)
+{
+	struct drm_i915_private *dev_priv = guc_to_i915(guc);
+
+	if (HAS_GUC_CT(dev_priv))
+		intel_guc_disable_ct(guc);
+
+	guc->send = intel_guc_send_nop;
+}
+
 int intel_uc_init_hw(struct drm_i915_private *dev_priv)
 {
+	struct intel_guc *guc = &dev_priv->guc;
 	int ret, attempts;
 
 	if (!i915.enable_guc_loading)
 		return 0;
 
+	guc_disable_communication(guc);
 	gen9_reset_guc_interrupts(dev_priv);
 
 	/* We need to notify the guc whenever we change the GGTT */
@@ -274,6 +352,11 @@ int intel_uc_init_hw(struct drm_i915_private *dev_priv)
 			goto err_guc;
 	}
 
+	/* init WOPCM */
+	I915_WRITE(GUC_WOPCM_SIZE, intel_guc_wopcm_size(dev_priv));
+	I915_WRITE(DMA_GUC_WOPCM_OFFSET,
+		   GUC_WOPCM_OFFSET_VALUE | HUC_LOADING_AGENT_GUC);
+
 	/* WaEnableuKernelHeaderValidFix:skl */
 	/* WaEnableGuCBootHashCheckNotSet:skl,bxt,kbl */
 	if (IS_GEN9(dev_priv))
@@ -301,7 +384,11 @@ int intel_uc_init_hw(struct drm_i915_private *dev_priv)
 
 	/* Did we succeded or run out of retries? */
 	if (ret)
-		goto err_submission;
+		goto err_log_capture;
+
+	ret = guc_enable_communication(guc);
+	if (ret)
+		goto err_log_capture;
 
 	intel_guc_auth_huc(dev_priv);
 	if (i915.enable_guc_submission) {
@@ -325,7 +412,10 @@ int intel_uc_init_hw(struct drm_i915_private *dev_priv)
 	 * marks the GPU as wedged until reset).
 	 */
 err_interrupts:
+	guc_disable_communication(guc);
 	gen9_disable_guc_interrupts(dev_priv);
+err_log_capture:
+	guc_capture_load_err_log(guc);
 err_submission:
 	if (i915.enable_guc_submission)
 		i915_guc_submission_fini(dev_priv);
@@ -343,33 +433,36 @@ err_guc:
 		DRM_NOTE("Falling back from GuC submission to execlist mode\n");
 	}
 
+	i915.enable_guc_loading = 0;
+	DRM_NOTE("GuC firmware loading disabled\n");
+
 	return ret;
 }
 
 void intel_uc_fini_hw(struct drm_i915_private *dev_priv)
 {
+	guc_free_load_err_log(&dev_priv->guc);
+
 	if (!i915.enable_guc_loading)
 		return;
 
-	if (i915.enable_guc_submission) {
+	if (i915.enable_guc_submission)
 		i915_guc_submission_disable(dev_priv);
+
+	guc_disable_communication(&dev_priv->guc);
+
+	if (i915.enable_guc_submission) {
 		gen9_disable_guc_interrupts(dev_priv);
 		i915_guc_submission_fini(dev_priv);
 	}
+
 	i915_ggtt_disable_guc(dev_priv);
 }
 
-/*
- * Read GuC command/status register (SOFT_SCRATCH_0)
- * Return true if it contains a response rather than a command
- */
-static bool guc_recv(struct intel_guc *guc, u32 *status)
+int intel_guc_send_nop(struct intel_guc *guc, const u32 *action, u32 len)
 {
-	struct drm_i915_private *dev_priv = guc_to_i915(guc);
-
-	u32 val = I915_READ(SOFT_SCRATCH(0));
-	*status = val;
-	return INTEL_GUC_RECV_IS_RESPONSE(val);
+	WARN(1, "Unexpected send: action=%#x\n", *action);
+	return -ENODEV;
 }
 
 /*
@@ -382,30 +475,33 @@ int intel_guc_send_mmio(struct intel_guc *guc, const u32 *action, u32 len)
 	int i;
 	int ret;
 
-	if (WARN_ON(len < 1 || len > 15))
-		return -EINVAL;
+	GEM_BUG_ON(!len);
+	GEM_BUG_ON(len > guc->send_regs.count);
 
-	mutex_lock(&guc->send_mutex);
-	intel_uncore_forcewake_get(dev_priv, FORCEWAKE_BLITTER);
+	/* If CT is available, we expect to use MMIO only during init/fini */
+	GEM_BUG_ON(HAS_GUC_CT(dev_priv) &&
+		*action != INTEL_GUC_ACTION_REGISTER_COMMAND_TRANSPORT_BUFFER &&
+		*action != INTEL_GUC_ACTION_DEREGISTER_COMMAND_TRANSPORT_BUFFER);
 
-	dev_priv->guc.action_count += 1;
-	dev_priv->guc.action_cmd = action[0];
+	mutex_lock(&guc->send_mutex);
+	intel_uncore_forcewake_get(dev_priv, guc->send_regs.fw_domains);
 
 	for (i = 0; i < len; i++)
-		I915_WRITE(SOFT_SCRATCH(i), action[i]);
+		I915_WRITE(guc_send_reg(guc, i), action[i]);
 
-	POSTING_READ(SOFT_SCRATCH(i - 1));
+	POSTING_READ(guc_send_reg(guc, i - 1));
 
-	I915_WRITE(GUC_SEND_INTERRUPT, GUC_SEND_TRIGGER);
+	intel_guc_notify(guc);
 
 	/*
-	 * Fast commands should complete in less than 10us, so sample quickly
-	 * up to that length of time, then switch to a slower sleep-wait loop.
-	 * No inte_guc_send command should ever take longer than 10ms.
+	 * No GuC command should ever take longer than 10ms.
+	 * Fast commands should still complete in 10us.
 	 */
-	ret = wait_for_us(guc_recv(guc, &status), 10);
-	if (ret)
-		ret = wait_for(guc_recv(guc, &status), 10);
+	ret = __intel_wait_for_register_fw(dev_priv,
+					   guc_send_reg(guc, 0),
+					   INTEL_GUC_RECV_MASK,
+					   INTEL_GUC_RECV_MASK,
+					   10, 10, &status);
 	if (status != INTEL_GUC_STATUS_SUCCESS) {
 		/*
 		 * Either the GuC explicitly returned an error (which
@@ -418,13 +514,9 @@ int intel_guc_send_mmio(struct intel_guc *guc, const u32 *action, u32 len)
 		DRM_WARN("INTEL_GUC_SEND: Action 0x%X failed;"
 			 " ret=%d status=0x%08X response=0x%08X\n",
 			 action[0], ret, status, I915_READ(SOFT_SCRATCH(15)));
-
-		dev_priv->guc.action_fail += 1;
-		dev_priv->guc.action_err = ret;
 	}
-	dev_priv->guc.action_status = status;
 
-	intel_uncore_forcewake_put(dev_priv, FORCEWAKE_BLITTER);
+	intel_uncore_forcewake_put(dev_priv, guc->send_regs.fw_domains);
 	mutex_unlock(&guc->send_mutex);
 
 	return ret;
diff --git a/drivers/gpu/drm/i915/intel_uc.h b/drivers/gpu/drm/i915/intel_uc.h
index f84115261ae7..69daf4c01cd0 100644
--- a/drivers/gpu/drm/i915/intel_uc.h
+++ b/drivers/gpu/drm/i915/intel_uc.h
@@ -27,7 +27,7 @@
 #include "intel_guc_fwif.h"
 #include "i915_guc_reg.h"
 #include "intel_ringbuffer.h"
-
+#include "intel_guc_ct.h"
 #include "i915_vma.h"
 
 struct drm_i915_gem_request;
@@ -59,10 +59,6 @@ struct drm_i915_gem_request;
  *                available in the work queue (note, the queue is shared,
  *                not per-engine). It is OK for this to be nonzero, but
  *                it should not be huge!
- *   b_fail: failed to ring the doorbell. This should never happen, unless
- *           somehow the hardware misbehaves, or maybe if the GuC firmware
- *           crashes? We probably need to reset the GPU to recover.
- *   retcode: errno from last guc_submit()
  */
 struct i915_guc_client {
 	struct i915_vma *vma;
@@ -85,8 +81,6 @@ struct i915_guc_client {
 	uint32_t wq_tail;
 	uint32_t wq_rsvd;
 	uint32_t no_wq_space;
-	uint32_t b_fail;
-	int retcode;
 
 	/* Per-engine counts of GuC submissions */
 	uint64_t submissions[I915_NUM_ENGINES];
@@ -179,6 +173,10 @@ struct intel_guc_log {
 struct intel_guc {
 	struct intel_uc_fw fw;
 	struct intel_guc_log log;
+	struct intel_guc_ct ct;
+
+	/* Log snapshot if GuC errors during load */
+	struct drm_i915_gem_object *load_err_log;
 
 	/* intel_guc_recv interrupt related state */
 	bool interrupts_enabled;
@@ -193,21 +191,21 @@ struct intel_guc {
 	DECLARE_BITMAP(doorbell_bitmap, GUC_NUM_DOORBELLS);
 	uint32_t db_cacheline;		/* Cyclic counter mod pagesize	*/
 
-	/* Action status & statistics */
-	uint64_t action_count;		/* Total commands issued	*/
-	uint32_t action_cmd;		/* Last command word		*/
-	uint32_t action_status;		/* Last return status		*/
-	uint32_t action_fail;		/* Total number of failures	*/
-	int32_t action_err;		/* Last error code		*/
-
-	uint64_t submissions[I915_NUM_ENGINES];
-	uint32_t last_seqno[I915_NUM_ENGINES];
+	/* GuC's FW specific registers used in MMIO send */
+	struct {
+		u32 base;
+		unsigned int count;
+		enum forcewake_domains fw_domains;
+	} send_regs;
 
 	/* To serialize the intel_guc_send actions */
 	struct mutex send_mutex;
 
 	/* GuC's FW specific send function */
 	int (*send)(struct intel_guc *guc, const u32 *data, u32 len);
+
+	/* GuC's FW specific notify function */
+	void (*notify)(struct intel_guc *guc);
 };
 
 struct intel_huc {
@@ -225,12 +223,19 @@ void intel_uc_fini_fw(struct drm_i915_private *dev_priv);
 int intel_uc_init_hw(struct drm_i915_private *dev_priv);
 void intel_uc_fini_hw(struct drm_i915_private *dev_priv);
 int intel_guc_sample_forcewake(struct intel_guc *guc);
+int intel_guc_send_nop(struct intel_guc *guc, const u32 *action, u32 len);
 int intel_guc_send_mmio(struct intel_guc *guc, const u32 *action, u32 len);
+
 static inline int intel_guc_send(struct intel_guc *guc, const u32 *action, u32 len)
 {
 	return guc->send(guc, action, len);
 }
 
+static inline void intel_guc_notify(struct intel_guc *guc)
+{
+	guc->notify(guc);
+}
+
 /* intel_guc_loader.c */
 int intel_guc_select_fw(struct intel_guc *guc);
 int intel_guc_init_hw(struct intel_guc *guc);
@@ -264,7 +269,7 @@ static inline u32 guc_ggtt_offset(struct i915_vma *vma)
 
 /* intel_huc.c */
 void intel_huc_select_fw(struct intel_huc *huc);
-int intel_huc_init_hw(struct intel_huc *huc);
+void intel_huc_init_hw(struct intel_huc *huc);
 void intel_guc_auth_huc(struct drm_i915_private *dev_priv);
 
 #endif
diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c
index 6d1ea26b2493..9882724bc2b6 100644
--- a/drivers/gpu/drm/i915/intel_uncore.c
+++ b/drivers/gpu/drm/i915/intel_uncore.c
@@ -29,6 +29,7 @@
 #include <linux/pm_runtime.h>
 
 #define FORCEWAKE_ACK_TIMEOUT_MS 50
+#define GT_FIFO_TIMEOUT_MS	 10
 
 #define __raw_posting_read(dev_priv__, reg__) (void)__raw_i915_read32((dev_priv__), (reg__))
 
@@ -172,22 +173,6 @@ static void fw_domains_get_with_thread_status(struct drm_i915_private *dev_priv,
 	__gen6_gt_wait_for_thread_c0(dev_priv);
 }
 
-static void gen6_gt_check_fifodbg(struct drm_i915_private *dev_priv)
-{
-	u32 gtfifodbg;
-
-	gtfifodbg = __raw_i915_read32(dev_priv, GTFIFODBG);
-	if (WARN(gtfifodbg, "GT wake FIFO error 0x%x\n", gtfifodbg))
-		__raw_i915_write32(dev_priv, GTFIFODBG, gtfifodbg);
-}
-
-static void fw_domains_put_with_fifo(struct drm_i915_private *dev_priv,
-				     enum forcewake_domains fw_domains)
-{
-	fw_domains_put(dev_priv, fw_domains);
-	gen6_gt_check_fifodbg(dev_priv);
-}
-
 static inline u32 fifo_free_entries(struct drm_i915_private *dev_priv)
 {
 	u32 count = __raw_i915_read32(dev_priv, GTFIFOCTL);
@@ -195,30 +180,27 @@ static inline u32 fifo_free_entries(struct drm_i915_private *dev_priv)
 	return count & GT_FIFO_FREE_ENTRIES_MASK;
 }
 
-static int __gen6_gt_wait_for_fifo(struct drm_i915_private *dev_priv)
+static void __gen6_gt_wait_for_fifo(struct drm_i915_private *dev_priv)
 {
-	int ret = 0;
+	u32 n;
 
 	/* On VLV, FIFO will be shared by both SW and HW.
 	 * So, we need to read the FREE_ENTRIES everytime */
 	if (IS_VALLEYVIEW(dev_priv))
-		dev_priv->uncore.fifo_count = fifo_free_entries(dev_priv);
-
-	if (dev_priv->uncore.fifo_count < GT_FIFO_NUM_RESERVED_ENTRIES) {
-		int loop = 500;
-		u32 fifo = fifo_free_entries(dev_priv);
-
-		while (fifo <= GT_FIFO_NUM_RESERVED_ENTRIES && loop--) {
-			udelay(10);
-			fifo = fifo_free_entries(dev_priv);
+		n = fifo_free_entries(dev_priv);
+	else
+		n = dev_priv->uncore.fifo_count;
+
+	if (n <= GT_FIFO_NUM_RESERVED_ENTRIES) {
+		if (wait_for_atomic((n = fifo_free_entries(dev_priv)) >
+				    GT_FIFO_NUM_RESERVED_ENTRIES,
+				    GT_FIFO_TIMEOUT_MS)) {
+			DRM_DEBUG("GT_FIFO timeout, entries: %u\n", n);
+			return;
 		}
-		if (WARN_ON(loop < 0 && fifo <= GT_FIFO_NUM_RESERVED_ENTRIES))
-			++ret;
-		dev_priv->uncore.fifo_count = fifo;
 	}
-	dev_priv->uncore.fifo_count--;
 
-	return ret;
+	dev_priv->uncore.fifo_count = n - 1;
 }
 
 static enum hrtimer_restart
@@ -232,6 +214,9 @@ intel_uncore_fw_release_timer(struct hrtimer *timer)
 
 	assert_rpm_device_not_suspended(dev_priv);
 
+	if (xchg(&domain->active, false))
+		return HRTIMER_RESTART;
+
 	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
 	if (WARN_ON(domain->wake_count == 0))
 		domain->wake_count++;
@@ -262,6 +247,7 @@ static void intel_uncore_forcewake_reset(struct drm_i915_private *dev_priv,
 		active_domains = 0;
 
 		for_each_fw_domain(domain, dev_priv, tmp) {
+			smp_store_mb(domain->active, false);
 			if (hrtimer_cancel(&domain->timer) == 0)
 				continue;
 
@@ -384,31 +370,44 @@ vlv_check_for_unclaimed_mmio(struct drm_i915_private *dev_priv)
 }
 
 static bool
+gen6_check_for_fifo_debug(struct drm_i915_private *dev_priv)
+{
+	u32 fifodbg;
+
+	fifodbg = __raw_i915_read32(dev_priv, GTFIFODBG);
+
+	if (unlikely(fifodbg)) {
+		DRM_DEBUG_DRIVER("GTFIFODBG = 0x08%x\n", fifodbg);
+		__raw_i915_write32(dev_priv, GTFIFODBG, fifodbg);
+	}
+
+	return fifodbg;
+}
+
+static bool
 check_for_unclaimed_mmio(struct drm_i915_private *dev_priv)
 {
+	bool ret = false;
+
 	if (HAS_FPGA_DBG_UNCLAIMED(dev_priv))
-		return fpga_check_for_unclaimed_mmio(dev_priv);
+		ret |= fpga_check_for_unclaimed_mmio(dev_priv);
 
 	if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
-		return vlv_check_for_unclaimed_mmio(dev_priv);
+		ret |= vlv_check_for_unclaimed_mmio(dev_priv);
 
-	return false;
+	if (IS_GEN6(dev_priv) || IS_GEN7(dev_priv))
+		ret |= gen6_check_for_fifo_debug(dev_priv);
+
+	return ret;
 }
 
 static void __intel_uncore_early_sanitize(struct drm_i915_private *dev_priv,
 					  bool restore_forcewake)
 {
-	struct intel_device_info *info = mkwrite_device_info(dev_priv);
-
 	/* clear out unclaimed reg detection bit */
 	if (check_for_unclaimed_mmio(dev_priv))
 		DRM_DEBUG("unclaimed mmio detected on uncore init, clearing\n");
 
-	/* clear out old GT FIFO errors */
-	if (IS_GEN6(dev_priv) || IS_GEN7(dev_priv))
-		__raw_i915_write32(dev_priv, GTFIFODBG,
-				   __raw_i915_read32(dev_priv, GTFIFODBG));
-
 	/* WaDisableShadowRegForCpd:chv */
 	if (IS_CHERRYVIEW(dev_priv)) {
 		__raw_i915_write32(dev_priv, GTFIFOCTL,
@@ -417,9 +416,6 @@ static void __intel_uncore_early_sanitize(struct drm_i915_private *dev_priv,
 				   GT_FIFO_CTL_RC6_POLICY_STALL);
 	}
 
-	if (IS_BXT_REVID(dev_priv, 0, BXT_REVID_B_LAST))
-		info->has_decoupled_mmio = false;
-
 	intel_uncore_forcewake_reset(dev_priv, restore_forcewake);
 }
 
@@ -454,9 +450,12 @@ static void __intel_uncore_forcewake_get(struct drm_i915_private *dev_priv,
 
 	fw_domains &= dev_priv->uncore.fw_domains;
 
-	for_each_fw_domain_masked(domain, fw_domains, dev_priv, tmp)
-		if (domain->wake_count++)
+	for_each_fw_domain_masked(domain, fw_domains, dev_priv, tmp) {
+		if (domain->wake_count++) {
 			fw_domains &= ~domain->mask;
+			domain->active = true;
+		}
+	}
 
 	if (fw_domains)
 		dev_priv->uncore.funcs.force_wake_get(dev_priv, fw_domains);
@@ -521,8 +520,10 @@ static void __intel_uncore_forcewake_put(struct drm_i915_private *dev_priv,
 		if (WARN_ON(domain->wake_count == 0))
 			continue;
 
-		if (--domain->wake_count)
+		if (--domain->wake_count) {
+			domain->active = true;
 			continue;
+		}
 
 		fw_domain_arm_timer(domain);
 	}
@@ -804,66 +805,6 @@ unclaimed_reg_debug(struct drm_i915_private *dev_priv,
 	__unclaimed_reg_debug(dev_priv, reg, read, before);
 }
 
-static const enum decoupled_power_domain fw2dpd_domain[] = {
-	GEN9_DECOUPLED_PD_RENDER,
-	GEN9_DECOUPLED_PD_BLITTER,
-	GEN9_DECOUPLED_PD_ALL,
-	GEN9_DECOUPLED_PD_MEDIA,
-	GEN9_DECOUPLED_PD_ALL,
-	GEN9_DECOUPLED_PD_ALL,
-	GEN9_DECOUPLED_PD_ALL
-};
-
-/*
- * Decoupled MMIO access for only 1 DWORD
- */
-static void __gen9_decoupled_mmio_access(struct drm_i915_private *dev_priv,
-					 u32 reg,
-					 enum forcewake_domains fw_domain,
-					 enum decoupled_ops operation)
-{
-	enum decoupled_power_domain dp_domain;
-	u32 ctrl_reg_data = 0;
-
-	dp_domain = fw2dpd_domain[fw_domain - 1];
-
-	ctrl_reg_data |= reg;
-	ctrl_reg_data |= (operation << GEN9_DECOUPLED_OP_SHIFT);
-	ctrl_reg_data |= (dp_domain << GEN9_DECOUPLED_PD_SHIFT);
-	ctrl_reg_data |= GEN9_DECOUPLED_DW1_GO;
-	__raw_i915_write32(dev_priv, GEN9_DECOUPLED_REG0_DW1, ctrl_reg_data);
-
-	if (wait_for_atomic((__raw_i915_read32(dev_priv,
-			    GEN9_DECOUPLED_REG0_DW1) &
-			    GEN9_DECOUPLED_DW1_GO) == 0,
-			    FORCEWAKE_ACK_TIMEOUT_MS))
-		DRM_ERROR("Decoupled MMIO wait timed out\n");
-}
-
-static inline u32
-__gen9_decoupled_mmio_read32(struct drm_i915_private *dev_priv,
-			     u32 reg,
-			     enum forcewake_domains fw_domain)
-{
-	__gen9_decoupled_mmio_access(dev_priv, reg, fw_domain,
-				     GEN9_DECOUPLED_OP_READ);
-
-	return __raw_i915_read32(dev_priv, GEN9_DECOUPLED_REG0_DW0);
-}
-
-static inline void
-__gen9_decoupled_mmio_write(struct drm_i915_private *dev_priv,
-			    u32 reg, u32 data,
-			    enum forcewake_domains fw_domain)
-{
-
-	__raw_i915_write32(dev_priv, GEN9_DECOUPLED_REG0_DW0, data);
-
-	__gen9_decoupled_mmio_access(dev_priv, reg, fw_domain,
-				     GEN9_DECOUPLED_OP_WRITE);
-}
-
-
 #define GEN2_READ_HEADER(x) \
 	u##x val = 0; \
 	assert_rpm_wakelock_held(dev_priv);
@@ -960,28 +901,6 @@ func##_read##x(struct drm_i915_private *dev_priv, i915_reg_t reg, bool trace) {
 #define __gen6_read(x) __gen_read(gen6, x)
 #define __fwtable_read(x) __gen_read(fwtable, x)
 
-#define __gen9_decoupled_read(x) \
-static u##x \
-gen9_decoupled_read##x(struct drm_i915_private *dev_priv, \
-		       i915_reg_t reg, bool trace) { \
-	enum forcewake_domains fw_engine; \
-	GEN6_READ_HEADER(x); \
-	fw_engine = __fwtable_reg_read_fw_domains(offset); \
-	if (fw_engine & ~dev_priv->uncore.fw_domains_active) { \
-		unsigned i; \
-		u32 *ptr_data = (u32 *) &val; \
-		for (i = 0; i < x/32; i++, offset += sizeof(u32), ptr_data++) \
-			*ptr_data = __gen9_decoupled_mmio_read32(dev_priv, \
-								 offset, \
-								 fw_engine); \
-	} else { \
-		val = __raw_i915_read##x(dev_priv, reg); \
-	} \
-	GEN6_READ_FOOTER; \
-}
-
-__gen9_decoupled_read(32)
-__gen9_decoupled_read(64)
 __fwtable_read(8)
 __fwtable_read(16)
 __fwtable_read(32)
@@ -1047,15 +966,10 @@ __gen2_write(32)
 #define __gen6_write(x) \
 static void \
 gen6_write##x(struct drm_i915_private *dev_priv, i915_reg_t reg, u##x val, bool trace) { \
-	u32 __fifo_ret = 0; \
 	GEN6_WRITE_HEADER; \
-	if (NEEDS_FORCE_WAKE(offset)) { \
-		__fifo_ret = __gen6_gt_wait_for_fifo(dev_priv); \
-	} \
+	if (NEEDS_FORCE_WAKE(offset)) \
+		__gen6_gt_wait_for_fifo(dev_priv); \
 	__raw_i915_write##x(dev_priv, reg, val); \
-	if (unlikely(__fifo_ret)) { \
-		gen6_gt_check_fifodbg(dev_priv); \
-	} \
 	GEN6_WRITE_FOOTER; \
 }
 
@@ -1073,25 +987,6 @@ func##_write##x(struct drm_i915_private *dev_priv, i915_reg_t reg, u##x val, boo
 #define __gen8_write(x) __gen_write(gen8, x)
 #define __fwtable_write(x) __gen_write(fwtable, x)
 
-#define __gen9_decoupled_write(x) \
-static void \
-gen9_decoupled_write##x(struct drm_i915_private *dev_priv, \
-			i915_reg_t reg, u##x val, \
-		bool trace) { \
-	enum forcewake_domains fw_engine; \
-	GEN6_WRITE_HEADER; \
-	fw_engine = __fwtable_reg_write_fw_domains(offset); \
-	if (fw_engine & ~dev_priv->uncore.fw_domains_active) \
-		__gen9_decoupled_mmio_write(dev_priv, \
-					    offset, \
-					    val, \
-					    fw_engine); \
-	else \
-		__raw_i915_write##x(dev_priv, reg, val); \
-	GEN6_WRITE_FOOTER; \
-}
-
-__gen9_decoupled_write(32)
 __fwtable_write(8)
 __fwtable_write(16)
 __fwtable_write(32)
@@ -1108,19 +1003,19 @@ __gen6_write(32)
 #undef GEN6_WRITE_FOOTER
 #undef GEN6_WRITE_HEADER
 
-#define ASSIGN_WRITE_MMIO_VFUNCS(x) \
+#define ASSIGN_WRITE_MMIO_VFUNCS(i915, x) \
 do { \
-	dev_priv->uncore.funcs.mmio_writeb = x##_write8; \
-	dev_priv->uncore.funcs.mmio_writew = x##_write16; \
-	dev_priv->uncore.funcs.mmio_writel = x##_write32; \
+	(i915)->uncore.funcs.mmio_writeb = x##_write8; \
+	(i915)->uncore.funcs.mmio_writew = x##_write16; \
+	(i915)->uncore.funcs.mmio_writel = x##_write32; \
 } while (0)
 
-#define ASSIGN_READ_MMIO_VFUNCS(x) \
+#define ASSIGN_READ_MMIO_VFUNCS(i915, x) \
 do { \
-	dev_priv->uncore.funcs.mmio_readb = x##_read8; \
-	dev_priv->uncore.funcs.mmio_readw = x##_read16; \
-	dev_priv->uncore.funcs.mmio_readl = x##_read32; \
-	dev_priv->uncore.funcs.mmio_readq = x##_read64; \
+	(i915)->uncore.funcs.mmio_readb = x##_read8; \
+	(i915)->uncore.funcs.mmio_readw = x##_read16; \
+	(i915)->uncore.funcs.mmio_readl = x##_read32; \
+	(i915)->uncore.funcs.mmio_readq = x##_read64; \
 } while (0)
 
 
@@ -1190,11 +1085,7 @@ static void intel_uncore_fw_domains_init(struct drm_i915_private *dev_priv)
 			       FORCEWAKE_MEDIA_GEN9, FORCEWAKE_ACK_MEDIA_GEN9);
 	} else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) {
 		dev_priv->uncore.funcs.force_wake_get = fw_domains_get;
-		if (!IS_CHERRYVIEW(dev_priv))
-			dev_priv->uncore.funcs.force_wake_put =
-				fw_domains_put_with_fifo;
-		else
-			dev_priv->uncore.funcs.force_wake_put = fw_domains_put;
+		dev_priv->uncore.funcs.force_wake_put = fw_domains_put;
 		fw_domain_init(dev_priv, FW_DOMAIN_ID_RENDER,
 			       FORCEWAKE_VLV, FORCEWAKE_ACK_VLV);
 		fw_domain_init(dev_priv, FW_DOMAIN_ID_MEDIA,
@@ -1202,11 +1093,7 @@ static void intel_uncore_fw_domains_init(struct drm_i915_private *dev_priv)
 	} else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) {
 		dev_priv->uncore.funcs.force_wake_get =
 			fw_domains_get_with_thread_status;
-		if (IS_HASWELL(dev_priv))
-			dev_priv->uncore.funcs.force_wake_put =
-				fw_domains_put_with_fifo;
-		else
-			dev_priv->uncore.funcs.force_wake_put = fw_domains_put;
+		dev_priv->uncore.funcs.force_wake_put = fw_domains_put;
 		fw_domain_init(dev_priv, FW_DOMAIN_ID_RENDER,
 			       FORCEWAKE_MT, FORCEWAKE_ACK_HSW);
 	} else if (IS_IVYBRIDGE(dev_priv)) {
@@ -1223,8 +1110,7 @@ static void intel_uncore_fw_domains_init(struct drm_i915_private *dev_priv)
 		 */
 		dev_priv->uncore.funcs.force_wake_get =
 			fw_domains_get_with_thread_status;
-		dev_priv->uncore.funcs.force_wake_put =
-			fw_domains_put_with_fifo;
+		dev_priv->uncore.funcs.force_wake_put = fw_domains_put;
 
 		/* We need to init first for ECOBUS access and then
 		 * determine later if we want to reinit, in case of MT access is
@@ -1242,7 +1128,7 @@ static void intel_uncore_fw_domains_init(struct drm_i915_private *dev_priv)
 		spin_lock_irq(&dev_priv->uncore.lock);
 		fw_domains_get_with_thread_status(dev_priv, FORCEWAKE_RENDER);
 		ecobus = __raw_i915_read32(dev_priv, ECOBUS);
-		fw_domains_put_with_fifo(dev_priv, FORCEWAKE_RENDER);
+		fw_domains_put(dev_priv, FORCEWAKE_RENDER);
 		spin_unlock_irq(&dev_priv->uncore.lock);
 
 		if (!(ecobus & FORCEWAKE_MT_ENABLE)) {
@@ -1254,8 +1140,7 @@ static void intel_uncore_fw_domains_init(struct drm_i915_private *dev_priv)
 	} else if (IS_GEN6(dev_priv)) {
 		dev_priv->uncore.funcs.force_wake_get =
 			fw_domains_get_with_thread_status;
-		dev_priv->uncore.funcs.force_wake_put =
-			fw_domains_put_with_fifo;
+		dev_priv->uncore.funcs.force_wake_put = fw_domains_put;
 		fw_domain_init(dev_priv, FW_DOMAIN_ID_RENDER,
 			       FORCEWAKE, FORCEWAKE_ACK);
 	}
@@ -1310,42 +1195,34 @@ void intel_uncore_init(struct drm_i915_private *dev_priv)
 		i915_pmic_bus_access_notifier;
 
 	if (IS_GEN(dev_priv, 2, 4) || intel_vgpu_active(dev_priv)) {
-		ASSIGN_WRITE_MMIO_VFUNCS(gen2);
-		ASSIGN_READ_MMIO_VFUNCS(gen2);
+		ASSIGN_WRITE_MMIO_VFUNCS(dev_priv, gen2);
+		ASSIGN_READ_MMIO_VFUNCS(dev_priv, gen2);
 	} else if (IS_GEN5(dev_priv)) {
-		ASSIGN_WRITE_MMIO_VFUNCS(gen5);
-		ASSIGN_READ_MMIO_VFUNCS(gen5);
+		ASSIGN_WRITE_MMIO_VFUNCS(dev_priv, gen5);
+		ASSIGN_READ_MMIO_VFUNCS(dev_priv, gen5);
 	} else if (IS_GEN(dev_priv, 6, 7)) {
-		ASSIGN_WRITE_MMIO_VFUNCS(gen6);
+		ASSIGN_WRITE_MMIO_VFUNCS(dev_priv, gen6);
 
 		if (IS_VALLEYVIEW(dev_priv)) {
 			ASSIGN_FW_DOMAINS_TABLE(__vlv_fw_ranges);
-			ASSIGN_READ_MMIO_VFUNCS(fwtable);
+			ASSIGN_READ_MMIO_VFUNCS(dev_priv, fwtable);
 		} else {
-			ASSIGN_READ_MMIO_VFUNCS(gen6);
+			ASSIGN_READ_MMIO_VFUNCS(dev_priv, gen6);
 		}
 	} else if (IS_GEN8(dev_priv)) {
 		if (IS_CHERRYVIEW(dev_priv)) {
 			ASSIGN_FW_DOMAINS_TABLE(__chv_fw_ranges);
-			ASSIGN_WRITE_MMIO_VFUNCS(fwtable);
-			ASSIGN_READ_MMIO_VFUNCS(fwtable);
+			ASSIGN_WRITE_MMIO_VFUNCS(dev_priv, fwtable);
+			ASSIGN_READ_MMIO_VFUNCS(dev_priv, fwtable);
 
 		} else {
-			ASSIGN_WRITE_MMIO_VFUNCS(gen8);
-			ASSIGN_READ_MMIO_VFUNCS(gen6);
+			ASSIGN_WRITE_MMIO_VFUNCS(dev_priv, gen8);
+			ASSIGN_READ_MMIO_VFUNCS(dev_priv, gen6);
 		}
 	} else {
 		ASSIGN_FW_DOMAINS_TABLE(__gen9_fw_ranges);
-		ASSIGN_WRITE_MMIO_VFUNCS(fwtable);
-		ASSIGN_READ_MMIO_VFUNCS(fwtable);
-		if (HAS_DECOUPLED_MMIO(dev_priv)) {
-			dev_priv->uncore.funcs.mmio_readl =
-						gen9_decoupled_read32;
-			dev_priv->uncore.funcs.mmio_readq =
-						gen9_decoupled_read64;
-			dev_priv->uncore.funcs.mmio_writel =
-						gen9_decoupled_write32;
-		}
+		ASSIGN_WRITE_MMIO_VFUNCS(dev_priv, fwtable);
+		ASSIGN_READ_MMIO_VFUNCS(dev_priv, fwtable);
 	}
 
 	iosf_mbi_register_pmic_bus_access_notifier(
@@ -1353,8 +1230,6 @@ void intel_uncore_init(struct drm_i915_private *dev_priv)
 
 	i915_check_and_clear_faults(dev_priv);
 }
-#undef ASSIGN_WRITE_MMIO_VFUNCS
-#undef ASSIGN_READ_MMIO_VFUNCS
 
 void intel_uncore_fini(struct drm_i915_private *dev_priv)
 {
@@ -1435,9 +1310,39 @@ out:
 	return ret;
 }
 
-static int i915_reset_complete(struct pci_dev *pdev)
+static void gen3_stop_rings(struct drm_i915_private *dev_priv)
+{
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+
+	for_each_engine(engine, dev_priv, id) {
+		const u32 base = engine->mmio_base;
+		const i915_reg_t mode = RING_MI_MODE(base);
+
+		I915_WRITE_FW(mode, _MASKED_BIT_ENABLE(STOP_RING));
+		if (intel_wait_for_register_fw(dev_priv,
+					       mode,
+					       MODE_IDLE,
+					       MODE_IDLE,
+					       500))
+			DRM_DEBUG_DRIVER("%s: timed out on STOP_RING\n",
+					 engine->name);
+
+		I915_WRITE_FW(RING_CTL(base), 0);
+		I915_WRITE_FW(RING_HEAD(base), 0);
+		I915_WRITE_FW(RING_TAIL(base), 0);
+
+		/* Check acts as a post */
+		if (I915_READ_FW(RING_HEAD(base)) != 0)
+			DRM_DEBUG_DRIVER("%s: ring head not parked\n",
+					 engine->name);
+	}
+}
+
+static bool i915_reset_complete(struct pci_dev *pdev)
 {
 	u8 gdrst;
+
 	pci_read_config_byte(pdev, I915_GDRST, &gdrst);
 	return (gdrst & GRDOM_RESET_STATUS) == 0;
 }
@@ -1448,15 +1353,16 @@ static int i915_do_reset(struct drm_i915_private *dev_priv, unsigned engine_mask
 
 	/* assert reset for at least 20 usec */
 	pci_write_config_byte(pdev, I915_GDRST, GRDOM_RESET_ENABLE);
-	udelay(20);
+	usleep_range(50, 200);
 	pci_write_config_byte(pdev, I915_GDRST, 0);
 
 	return wait_for(i915_reset_complete(pdev), 500);
 }
 
-static int g4x_reset_complete(struct pci_dev *pdev)
+static bool g4x_reset_complete(struct pci_dev *pdev)
 {
 	u8 gdrst;
+
 	pci_read_config_byte(pdev, I915_GDRST, &gdrst);
 	return (gdrst & GRDOM_RESET_ENABLE) == 0;
 }
@@ -1464,6 +1370,10 @@ static int g4x_reset_complete(struct pci_dev *pdev)
 static int g33_do_reset(struct drm_i915_private *dev_priv, unsigned engine_mask)
 {
 	struct pci_dev *pdev = dev_priv->drm.pdev;
+
+	/* Stop engines before we reset; see g4x_do_reset() below for why. */
+	gen3_stop_rings(dev_priv);
+
 	pci_write_config_byte(pdev, I915_GDRST, GRDOM_RESET_ENABLE);
 	return wait_for(g4x_reset_complete(pdev), 500);
 }
@@ -1473,29 +1383,41 @@ static int g4x_do_reset(struct drm_i915_private *dev_priv, unsigned engine_mask)
 	struct pci_dev *pdev = dev_priv->drm.pdev;
 	int ret;
 
-	pci_write_config_byte(pdev, I915_GDRST,
-			      GRDOM_RENDER | GRDOM_RESET_ENABLE);
-	ret =  wait_for(g4x_reset_complete(pdev), 500);
-	if (ret)
-		return ret;
-
 	/* WaVcpClkGateDisableForMediaReset:ctg,elk */
-	I915_WRITE(VDECCLK_GATE_D, I915_READ(VDECCLK_GATE_D) | VCP_UNIT_CLOCK_GATE_DISABLE);
+	I915_WRITE(VDECCLK_GATE_D,
+		   I915_READ(VDECCLK_GATE_D) | VCP_UNIT_CLOCK_GATE_DISABLE);
 	POSTING_READ(VDECCLK_GATE_D);
 
+	/* We stop engines, otherwise we might get failed reset and a
+	 * dead gpu (on elk).
+	 * WaMediaResetMainRingCleanup:ctg,elk (presumably)
+	 */
+	gen3_stop_rings(dev_priv);
+
 	pci_write_config_byte(pdev, I915_GDRST,
 			      GRDOM_MEDIA | GRDOM_RESET_ENABLE);
 	ret =  wait_for(g4x_reset_complete(pdev), 500);
-	if (ret)
-		return ret;
+	if (ret) {
+		DRM_DEBUG_DRIVER("Wait for media reset failed\n");
+		goto out;
+	}
 
-	/* WaVcpClkGateDisableForMediaReset:ctg,elk */
-	I915_WRITE(VDECCLK_GATE_D, I915_READ(VDECCLK_GATE_D) & ~VCP_UNIT_CLOCK_GATE_DISABLE);
-	POSTING_READ(VDECCLK_GATE_D);
+	pci_write_config_byte(pdev, I915_GDRST,
+			      GRDOM_RENDER | GRDOM_RESET_ENABLE);
+	ret =  wait_for(g4x_reset_complete(pdev), 500);
+	if (ret) {
+		DRM_DEBUG_DRIVER("Wait for render reset failed\n");
+		goto out;
+	}
 
+out:
 	pci_write_config_byte(pdev, I915_GDRST, 0);
 
-	return 0;
+	I915_WRITE(VDECCLK_GATE_D,
+		   I915_READ(VDECCLK_GATE_D) & ~VCP_UNIT_CLOCK_GATE_DISABLE);
+	POSTING_READ(VDECCLK_GATE_D);
+
+	return ret;
 }
 
 static int ironlake_do_reset(struct drm_i915_private *dev_priv,
@@ -1503,41 +1425,51 @@ static int ironlake_do_reset(struct drm_i915_private *dev_priv,
 {
 	int ret;
 
-	I915_WRITE(ILK_GDSR,
-		   ILK_GRDOM_RENDER | ILK_GRDOM_RESET_ENABLE);
+	I915_WRITE(ILK_GDSR, ILK_GRDOM_RENDER | ILK_GRDOM_RESET_ENABLE);
 	ret = intel_wait_for_register(dev_priv,
 				      ILK_GDSR, ILK_GRDOM_RESET_ENABLE, 0,
 				      500);
-	if (ret)
-		return ret;
+	if (ret) {
+		DRM_DEBUG_DRIVER("Wait for render reset failed\n");
+		goto out;
+	}
 
-	I915_WRITE(ILK_GDSR,
-		   ILK_GRDOM_MEDIA | ILK_GRDOM_RESET_ENABLE);
+	I915_WRITE(ILK_GDSR, ILK_GRDOM_MEDIA | ILK_GRDOM_RESET_ENABLE);
 	ret = intel_wait_for_register(dev_priv,
 				      ILK_GDSR, ILK_GRDOM_RESET_ENABLE, 0,
 				      500);
-	if (ret)
-		return ret;
+	if (ret) {
+		DRM_DEBUG_DRIVER("Wait for media reset failed\n");
+		goto out;
+	}
 
+out:
 	I915_WRITE(ILK_GDSR, 0);
-
-	return 0;
+	POSTING_READ(ILK_GDSR);
+	return ret;
 }
 
 /* Reset the hardware domains (GENX_GRDOM_*) specified by mask */
 static int gen6_hw_domain_reset(struct drm_i915_private *dev_priv,
 				u32 hw_domain_mask)
 {
+	int err;
+
 	/* GEN6_GDRST is not in the gt power well, no need to check
 	 * for fifo space for the write or forcewake the chip for
 	 * the read
 	 */
 	__raw_i915_write32(dev_priv, GEN6_GDRST, hw_domain_mask);
 
-	/* Spin waiting for the device to ack the reset requests */
-	return intel_wait_for_register_fw(dev_priv,
+	/* Wait for the device to ack the reset requests */
+	err = intel_wait_for_register_fw(dev_priv,
 					  GEN6_GDRST, hw_domain_mask, 0,
 					  500);
+	if (err)
+		DRM_DEBUG_DRIVER("Wait for 0x%08x engines reset failed\n",
+				 hw_domain_mask);
+
+	return err;
 }
 
 /**
@@ -1585,19 +1517,23 @@ static int gen6_reset_engines(struct drm_i915_private *dev_priv,
 }
 
 /**
- * intel_wait_for_register_fw - wait until register matches expected state
+ * __intel_wait_for_register_fw - wait until register matches expected state
  * @dev_priv: the i915 device
  * @reg: the register to read
  * @mask: mask to apply to register value
  * @value: expected value
- * @timeout_ms: timeout in millisecond
+ * @fast_timeout_us: fast timeout in microsecond for atomic/tight wait
+ * @slow_timeout_ms: slow timeout in millisecond
+ * @out_value: optional placeholder to hold registry value
  *
  * This routine waits until the target register @reg contains the expected
  * @value after applying the @mask, i.e. it waits until ::
  *
  *     (I915_READ_FW(reg) & mask) == value
  *
- * Otherwise, the wait will timeout after @timeout_ms milliseconds.
+ * Otherwise, the wait will timeout after @slow_timeout_ms milliseconds.
+ * For atomic context @slow_timeout_ms must be zero and @fast_timeout_us
+ * must be not larger than 20,0000 microseconds.
  *
  * Note that this routine assumes the caller holds forcewake asserted, it is
  * not suitable for very long waits. See intel_wait_for_register() if you
@@ -1606,16 +1542,31 @@ static int gen6_reset_engines(struct drm_i915_private *dev_priv,
  *
  * Returns 0 if the register matches the desired condition, or -ETIMEOUT.
  */
-int intel_wait_for_register_fw(struct drm_i915_private *dev_priv,
-			       i915_reg_t reg,
-			       const u32 mask,
-			       const u32 value,
-			       const unsigned long timeout_ms)
-{
-#define done ((I915_READ_FW(reg) & mask) == value)
-	int ret = wait_for_us(done, 2);
-	if (ret)
-		ret = wait_for(done, timeout_ms);
+int __intel_wait_for_register_fw(struct drm_i915_private *dev_priv,
+				 i915_reg_t reg,
+				 u32 mask,
+				 u32 value,
+				 unsigned int fast_timeout_us,
+				 unsigned int slow_timeout_ms,
+				 u32 *out_value)
+{
+	u32 uninitialized_var(reg_value);
+#define done (((reg_value = I915_READ_FW(reg)) & mask) == value)
+	int ret;
+
+	/* Catch any overuse of this function */
+	might_sleep_if(slow_timeout_ms);
+	GEM_BUG_ON(fast_timeout_us > 20000);
+
+	ret = -ETIMEDOUT;
+	if (fast_timeout_us && fast_timeout_us <= 20000)
+		ret = _wait_for_atomic(done, fast_timeout_us, 0);
+	if (ret && slow_timeout_ms)
+		ret = wait_for(done, slow_timeout_ms);
+
+	if (out_value)
+		*out_value = reg_value;
+
 	return ret;
 #undef done
 }
@@ -1639,18 +1590,26 @@ int intel_wait_for_register_fw(struct drm_i915_private *dev_priv,
  */
 int intel_wait_for_register(struct drm_i915_private *dev_priv,
 			    i915_reg_t reg,
-			    const u32 mask,
-			    const u32 value,
-			    const unsigned long timeout_ms)
+			    u32 mask,
+			    u32 value,
+			    unsigned int timeout_ms)
 {
-
 	unsigned fw =
 		intel_uncore_forcewake_for_reg(dev_priv, reg, FW_REG_READ);
 	int ret;
 
-	intel_uncore_forcewake_get(dev_priv, fw);
-	ret = wait_for_us((I915_READ_FW(reg) & mask) == value, 2);
-	intel_uncore_forcewake_put(dev_priv, fw);
+	might_sleep();
+
+	spin_lock_irq(&dev_priv->uncore.lock);
+	intel_uncore_forcewake_get__locked(dev_priv, fw);
+
+	ret = __intel_wait_for_register_fw(dev_priv,
+					   reg, mask, value,
+					   2, 0, NULL);
+
+	intel_uncore_forcewake_put__locked(dev_priv, fw);
+	spin_unlock_irq(&dev_priv->uncore.lock);
+
 	if (ret)
 		ret = wait_for((I915_READ_NOTRACE(reg) & mask) == value,
 			       timeout_ms);
@@ -1658,7 +1617,7 @@ int intel_wait_for_register(struct drm_i915_private *dev_priv,
 	return ret;
 }
 
-static int gen8_request_engine_reset(struct intel_engine_cs *engine)
+static int gen8_reset_engine_start(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
 	int ret;
@@ -1677,7 +1636,7 @@ static int gen8_request_engine_reset(struct intel_engine_cs *engine)
 	return ret;
 }
 
-static void gen8_unrequest_engine_reset(struct intel_engine_cs *engine)
+static void gen8_reset_engine_cancel(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
 
@@ -1692,14 +1651,14 @@ static int gen8_reset_engines(struct drm_i915_private *dev_priv,
 	unsigned int tmp;
 
 	for_each_engine_masked(engine, dev_priv, engine_mask, tmp)
-		if (gen8_request_engine_reset(engine))
+		if (gen8_reset_engine_start(engine))
 			goto not_ready;
 
 	return gen6_reset_engines(dev_priv, engine_mask);
 
 not_ready:
 	for_each_engine_masked(engine, dev_priv, engine_mask, tmp)
-		gen8_unrequest_engine_reset(engine);
+		gen8_reset_engine_cancel(engine);
 
 	return -EIO;
 }
@@ -1730,8 +1689,11 @@ static reset_func intel_get_gpu_reset(struct drm_i915_private *dev_priv)
 int intel_gpu_reset(struct drm_i915_private *dev_priv, unsigned engine_mask)
 {
 	reset_func reset;
+	int retry;
 	int ret;
 
+	might_sleep();
+
 	reset = intel_get_gpu_reset(dev_priv);
 	if (reset == NULL)
 		return -ENODEV;
@@ -1740,7 +1702,13 @@ int intel_gpu_reset(struct drm_i915_private *dev_priv, unsigned engine_mask)
 	 * request may be dropped and never completes (causing -EIO).
 	 */
 	intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL);
-	ret = reset(dev_priv, engine_mask);
+	for (retry = 0; retry < 3; retry++) {
+		ret = reset(dev_priv, engine_mask);
+		if (ret != -ETIMEDOUT)
+			break;
+
+		cond_resched();
+	}
 	intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
 
 	return ret;
@@ -1754,17 +1722,12 @@ bool intel_has_gpu_reset(struct drm_i915_private *dev_priv)
 int intel_guc_reset(struct drm_i915_private *dev_priv)
 {
 	int ret;
-	unsigned long irqflags;
 
 	if (!HAS_GUC(dev_priv))
 		return -EINVAL;
 
 	intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL);
-	spin_lock_irqsave(&dev_priv->uncore.lock, irqflags);
-
 	ret = gen6_hw_domain_reset(dev_priv, GEN9_GRDOM_GUC);
-
-	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
 	intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL);
 
 	return ret;
@@ -1873,5 +1836,6 @@ intel_uncore_forcewake_for_reg(struct drm_i915_private *dev_priv,
 }
 
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
+#include "selftests/mock_uncore.c"
 #include "selftests/intel_uncore.c"
 #endif
diff --git a/drivers/gpu/drm/i915/intel_uncore.h b/drivers/gpu/drm/i915/intel_uncore.h
new file mode 100644
index 000000000000..5f90278da461
--- /dev/null
+++ b/drivers/gpu/drm/i915/intel_uncore.h
@@ -0,0 +1,170 @@
+/*
+ * Copyright © 2017 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __INTEL_UNCORE_H__
+#define __INTEL_UNCORE_H__
+
+struct drm_i915_private;
+
+enum forcewake_domain_id {
+	FW_DOMAIN_ID_RENDER = 0,
+	FW_DOMAIN_ID_BLITTER,
+	FW_DOMAIN_ID_MEDIA,
+
+	FW_DOMAIN_ID_COUNT
+};
+
+enum forcewake_domains {
+	FORCEWAKE_RENDER = BIT(FW_DOMAIN_ID_RENDER),
+	FORCEWAKE_BLITTER = BIT(FW_DOMAIN_ID_BLITTER),
+	FORCEWAKE_MEDIA	= BIT(FW_DOMAIN_ID_MEDIA),
+	FORCEWAKE_ALL = (FORCEWAKE_RENDER |
+			 FORCEWAKE_BLITTER |
+			 FORCEWAKE_MEDIA)
+};
+
+struct intel_uncore_funcs {
+	void (*force_wake_get)(struct drm_i915_private *dev_priv,
+			       enum forcewake_domains domains);
+	void (*force_wake_put)(struct drm_i915_private *dev_priv,
+			       enum forcewake_domains domains);
+
+	uint8_t  (*mmio_readb)(struct drm_i915_private *dev_priv,
+			       i915_reg_t r, bool trace);
+	uint16_t (*mmio_readw)(struct drm_i915_private *dev_priv,
+			       i915_reg_t r, bool trace);
+	uint32_t (*mmio_readl)(struct drm_i915_private *dev_priv,
+			       i915_reg_t r, bool trace);
+	uint64_t (*mmio_readq)(struct drm_i915_private *dev_priv,
+			       i915_reg_t r, bool trace);
+
+	void (*mmio_writeb)(struct drm_i915_private *dev_priv,
+			    i915_reg_t r, uint8_t val, bool trace);
+	void (*mmio_writew)(struct drm_i915_private *dev_priv,
+			    i915_reg_t r, uint16_t val, bool trace);
+	void (*mmio_writel)(struct drm_i915_private *dev_priv,
+			    i915_reg_t r, uint32_t val, bool trace);
+};
+
+struct intel_forcewake_range {
+	u32 start;
+	u32 end;
+
+	enum forcewake_domains domains;
+};
+
+struct intel_uncore {
+	spinlock_t lock; /** lock is also taken in irq contexts. */
+
+	const struct intel_forcewake_range *fw_domains_table;
+	unsigned int fw_domains_table_entries;
+
+	struct notifier_block pmic_bus_access_nb;
+	struct intel_uncore_funcs funcs;
+
+	unsigned int fifo_count;
+
+	enum forcewake_domains fw_domains;
+	enum forcewake_domains fw_domains_active;
+
+	u32 fw_set;
+	u32 fw_clear;
+	u32 fw_reset;
+
+	struct intel_uncore_forcewake_domain {
+		enum forcewake_domain_id id;
+		enum forcewake_domains mask;
+		unsigned int wake_count;
+		bool active;
+		struct hrtimer timer;
+		i915_reg_t reg_set;
+		i915_reg_t reg_ack;
+	} fw_domain[FW_DOMAIN_ID_COUNT];
+
+	int unclaimed_mmio_check;
+};
+
+/* Iterate over initialised fw domains */
+#define for_each_fw_domain_masked(domain__, mask__, dev_priv__, tmp__) \
+	for (tmp__ = (mask__); \
+	     tmp__ ? (domain__ = &(dev_priv__)->uncore.fw_domain[__mask_next_bit(tmp__)]), 1 : 0;)
+
+#define for_each_fw_domain(domain__, dev_priv__, tmp__) \
+	for_each_fw_domain_masked(domain__, (dev_priv__)->uncore.fw_domains, dev_priv__, tmp__)
+
+
+void intel_uncore_sanitize(struct drm_i915_private *dev_priv);
+void intel_uncore_init(struct drm_i915_private *dev_priv);
+bool intel_uncore_unclaimed_mmio(struct drm_i915_private *dev_priv);
+bool intel_uncore_arm_unclaimed_mmio_detection(struct drm_i915_private *dev_priv);
+void intel_uncore_fini(struct drm_i915_private *dev_priv);
+void intel_uncore_suspend(struct drm_i915_private *dev_priv);
+void intel_uncore_resume_early(struct drm_i915_private *dev_priv);
+
+u64 intel_uncore_edram_size(struct drm_i915_private *dev_priv);
+void assert_forcewakes_inactive(struct drm_i915_private *dev_priv);
+const char *intel_uncore_forcewake_domain_to_str(const enum forcewake_domain_id id);
+
+enum forcewake_domains
+intel_uncore_forcewake_for_reg(struct drm_i915_private *dev_priv,
+			       i915_reg_t reg, unsigned int op);
+#define FW_REG_READ  (1)
+#define FW_REG_WRITE (2)
+
+void intel_uncore_forcewake_get(struct drm_i915_private *dev_priv,
+				enum forcewake_domains domains);
+void intel_uncore_forcewake_put(struct drm_i915_private *dev_priv,
+				enum forcewake_domains domains);
+/* Like above but the caller must manage the uncore.lock itself.
+ * Must be used with I915_READ_FW and friends.
+ */
+void intel_uncore_forcewake_get__locked(struct drm_i915_private *dev_priv,
+					enum forcewake_domains domains);
+void intel_uncore_forcewake_put__locked(struct drm_i915_private *dev_priv,
+					enum forcewake_domains domains);
+
+int intel_wait_for_register(struct drm_i915_private *dev_priv,
+			    i915_reg_t reg,
+			    u32 mask,
+			    u32 value,
+			    unsigned int timeout_ms);
+int __intel_wait_for_register_fw(struct drm_i915_private *dev_priv,
+				 i915_reg_t reg,
+				 u32 mask,
+				 u32 value,
+				 unsigned int fast_timeout_us,
+				 unsigned int slow_timeout_ms,
+				 u32 *out_value);
+static inline
+int intel_wait_for_register_fw(struct drm_i915_private *dev_priv,
+			       i915_reg_t reg,
+			       u32 mask,
+			       u32 value,
+			       unsigned int timeout_ms)
+{
+	return __intel_wait_for_register_fw(dev_priv, reg, mask, value,
+					    2, timeout_ms, NULL);
+}
+
+#endif /* !__INTEL_UNCORE_H__ */
diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_coherency.c b/drivers/gpu/drm/i915/selftests/i915_gem_coherency.c
index f08d0179b3df..95d4aebc0181 100644
--- a/drivers/gpu/drm/i915/selftests/i915_gem_coherency.c
+++ b/drivers/gpu/drm/i915/selftests/i915_gem_coherency.c
@@ -138,10 +138,7 @@ static int wc_set(struct drm_i915_gem_object *obj,
 	typeof(v) *map;
 	int err;
 
-	/* XXX GTT write followed by WC write go missing */
-	i915_gem_object_flush_gtt_write_domain(obj);
-
-	err = i915_gem_object_set_to_gtt_domain(obj, true);
+	err = i915_gem_object_set_to_wc_domain(obj, true);
 	if (err)
 		return err;
 
@@ -162,10 +159,7 @@ static int wc_get(struct drm_i915_gem_object *obj,
 	typeof(v) map;
 	int err;
 
-	/* XXX WC write followed by GTT write go missing */
-	i915_gem_object_flush_gtt_write_domain(obj);
-
-	err = i915_gem_object_set_to_gtt_domain(obj, false);
+	err = i915_gem_object_set_to_wc_domain(obj, false);
 	if (err)
 		return err;
 
diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_dmabuf.c b/drivers/gpu/drm/i915/selftests/i915_gem_dmabuf.c
index 817bef74bbcb..d15cc9d3a5cd 100644
--- a/drivers/gpu/drm/i915/selftests/i915_gem_dmabuf.c
+++ b/drivers/gpu/drm/i915/selftests/i915_gem_dmabuf.c
@@ -271,6 +271,105 @@ err_obj:
 	return err;
 }
 
+static int igt_dmabuf_export_kmap(void *arg)
+{
+	struct drm_i915_private *i915 = arg;
+	struct drm_i915_gem_object *obj;
+	struct dma_buf *dmabuf;
+	void *ptr;
+	int err;
+
+	obj = i915_gem_object_create(i915, 2*PAGE_SIZE);
+	if (IS_ERR(obj))
+		return PTR_ERR(obj);
+
+	dmabuf = i915_gem_prime_export(&i915->drm, &obj->base, 0);
+	i915_gem_object_put(obj);
+	if (IS_ERR(dmabuf)) {
+		err = PTR_ERR(dmabuf);
+		pr_err("i915_gem_prime_export failed with err=%d\n", err);
+		return err;
+	}
+
+	ptr = dma_buf_kmap(dmabuf, 0);
+	if (!ptr) {
+		pr_err("dma_buf_kmap failed\n");
+		err = -ENOMEM;
+		goto err;
+	}
+
+	if (memchr_inv(ptr, 0, PAGE_SIZE)) {
+		dma_buf_kunmap(dmabuf, 0, ptr);
+		pr_err("Exported page[0] not initialiased to zero!\n");
+		err = -EINVAL;
+		goto err;
+	}
+
+	memset(ptr, 0xc5, PAGE_SIZE);
+	dma_buf_kunmap(dmabuf, 0, ptr);
+
+	ptr = i915_gem_object_pin_map(obj, I915_MAP_WB);
+	if (IS_ERR(ptr)) {
+		err = PTR_ERR(ptr);
+		pr_err("i915_gem_object_pin_map failed with err=%d\n", err);
+		goto err;
+	}
+	memset(ptr + PAGE_SIZE, 0xaa, PAGE_SIZE);
+	i915_gem_object_unpin_map(obj);
+
+	ptr = dma_buf_kmap(dmabuf, 1);
+	if (!ptr) {
+		pr_err("dma_buf_kmap failed\n");
+		err = -ENOMEM;
+		goto err;
+	}
+
+	if (memchr_inv(ptr, 0xaa, PAGE_SIZE)) {
+		dma_buf_kunmap(dmabuf, 1, ptr);
+		pr_err("Exported page[1] not set to 0xaa!\n");
+		err = -EINVAL;
+		goto err;
+	}
+
+	memset(ptr, 0xc5, PAGE_SIZE);
+	dma_buf_kunmap(dmabuf, 1, ptr);
+
+	ptr = dma_buf_kmap(dmabuf, 0);
+	if (!ptr) {
+		pr_err("dma_buf_kmap failed\n");
+		err = -ENOMEM;
+		goto err;
+	}
+	if (memchr_inv(ptr, 0xc5, PAGE_SIZE)) {
+		dma_buf_kunmap(dmabuf, 0, ptr);
+		pr_err("Exported page[0] did not retain 0xc5!\n");
+		err = -EINVAL;
+		goto err;
+	}
+	dma_buf_kunmap(dmabuf, 0, ptr);
+
+	ptr = dma_buf_kmap(dmabuf, 2);
+	if (ptr) {
+		pr_err("Erroneously kmapped beyond the end of the object!\n");
+		dma_buf_kunmap(dmabuf, 2, ptr);
+		err = -EINVAL;
+		goto err;
+	}
+
+	ptr = dma_buf_kmap(dmabuf, -1);
+	if (ptr) {
+		pr_err("Erroneously kmapped before the start of the object!\n");
+		dma_buf_kunmap(dmabuf, -1, ptr);
+		err = -EINVAL;
+		goto err;
+	}
+
+	err = 0;
+err:
+	dma_buf_put(dmabuf);
+	return err;
+}
+
 int i915_gem_dmabuf_mock_selftests(void)
 {
 	static const struct i915_subtest tests[] = {
@@ -279,6 +378,7 @@ int i915_gem_dmabuf_mock_selftests(void)
 		SUBTEST(igt_dmabuf_import),
 		SUBTEST(igt_dmabuf_import_ownership),
 		SUBTEST(igt_dmabuf_export_vmap),
+		SUBTEST(igt_dmabuf_export_kmap),
 	};
 	struct drm_i915_private *i915;
 	int err;
diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_object.c b/drivers/gpu/drm/i915/selftests/i915_gem_object.c
index 67d82bf1407f..8f011c447e41 100644
--- a/drivers/gpu/drm/i915/selftests/i915_gem_object.c
+++ b/drivers/gpu/drm/i915/selftests/i915_gem_object.c
@@ -266,7 +266,7 @@ static int check_partial_mapping(struct drm_i915_gem_object *obj,
 		if (offset >= obj->base.size)
 			continue;
 
-		i915_gem_object_flush_gtt_write_domain(obj);
+		flush_write_domain(obj, ~I915_GEM_DOMAIN_CPU);
 
 		p = i915_gem_object_get_page(obj, offset >> PAGE_SHIFT);
 		cpu = kmap(p) + offset_in_page(offset);
@@ -545,7 +545,9 @@ static int igt_mmap_offset_exhaustion(void *arg)
 		}
 
 		mutex_lock(&i915->drm.struct_mutex);
+		intel_runtime_pm_get(i915);
 		err = make_obj_busy(obj);
+		intel_runtime_pm_put(i915);
 		mutex_unlock(&i915->drm.struct_mutex);
 		if (err) {
 			pr_err("[loop %d] Failed to busy the object\n", loop);
diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_request.c b/drivers/gpu/drm/i915/selftests/i915_gem_request.c
index 98b7aac41eec..6664cb2eb0b8 100644
--- a/drivers/gpu/drm/i915/selftests/i915_gem_request.c
+++ b/drivers/gpu/drm/i915/selftests/i915_gem_request.c
@@ -580,7 +580,7 @@ static struct i915_vma *recursive_batch(struct drm_i915_private *i915)
 	if (err)
 		goto err;
 
-	err = i915_gem_object_set_to_gtt_domain(obj, true);
+	err = i915_gem_object_set_to_wc_domain(obj, true);
 	if (err)
 		goto err;
 
diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_timeline.c b/drivers/gpu/drm/i915/selftests/i915_gem_timeline.c
new file mode 100644
index 000000000000..7a44dab631b8
--- /dev/null
+++ b/drivers/gpu/drm/i915/selftests/i915_gem_timeline.c
@@ -0,0 +1,299 @@
+/*
+ * Copyright © 2017 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ */
+
+#include "../i915_selftest.h"
+#include "i915_random.h"
+
+#include "mock_gem_device.h"
+#include "mock_timeline.h"
+
+struct __igt_sync {
+	const char *name;
+	u32 seqno;
+	bool expected;
+	bool set;
+};
+
+static int __igt_sync(struct intel_timeline *tl,
+		      u64 ctx,
+		      const struct __igt_sync *p,
+		      const char *name)
+{
+	int ret;
+
+	if (__intel_timeline_sync_is_later(tl, ctx, p->seqno) != p->expected) {
+		pr_err("%s: %s(ctx=%llu, seqno=%u) expected passed %s but failed\n",
+		       name, p->name, ctx, p->seqno, yesno(p->expected));
+		return -EINVAL;
+	}
+
+	if (p->set) {
+		ret = __intel_timeline_sync_set(tl, ctx, p->seqno);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+static int igt_sync(void *arg)
+{
+	const struct __igt_sync pass[] = {
+		{ "unset", 0, false, false },
+		{ "new", 0, false, true },
+		{ "0a", 0, true, true },
+		{ "1a", 1, false, true },
+		{ "1b", 1, true, true },
+		{ "0b", 0, true, false },
+		{ "2a", 2, false, true },
+		{ "4", 4, false, true },
+		{ "INT_MAX", INT_MAX, false, true },
+		{ "INT_MAX-1", INT_MAX-1, true, false },
+		{ "INT_MAX+1", (u32)INT_MAX+1, false, true },
+		{ "INT_MAX", INT_MAX, true, false },
+		{ "UINT_MAX", UINT_MAX, false, true },
+		{ "wrap", 0, false, true },
+		{ "unwrap", UINT_MAX, true, false },
+		{},
+	}, *p;
+	struct intel_timeline *tl;
+	int order, offset;
+	int ret;
+
+	tl = mock_timeline(0);
+	if (!tl)
+		return -ENOMEM;
+
+	for (p = pass; p->name; p++) {
+		for (order = 1; order < 64; order++) {
+			for (offset = -1; offset <= (order > 1); offset++) {
+				u64 ctx = BIT_ULL(order) + offset;
+
+				ret = __igt_sync(tl, ctx, p, "1");
+				if (ret)
+					goto out;
+			}
+		}
+	}
+	mock_timeline_destroy(tl);
+
+	tl = mock_timeline(0);
+	if (!tl)
+		return -ENOMEM;
+
+	for (order = 1; order < 64; order++) {
+		for (offset = -1; offset <= (order > 1); offset++) {
+			u64 ctx = BIT_ULL(order) + offset;
+
+			for (p = pass; p->name; p++) {
+				ret = __igt_sync(tl, ctx, p, "2");
+				if (ret)
+					goto out;
+			}
+		}
+	}
+
+out:
+	mock_timeline_destroy(tl);
+	return ret;
+}
+
+static unsigned int random_engine(struct rnd_state *rnd)
+{
+	return ((u64)prandom_u32_state(rnd) * I915_NUM_ENGINES) >> 32;
+}
+
+static int bench_sync(void *arg)
+{
+	struct rnd_state prng;
+	struct intel_timeline *tl;
+	unsigned long end_time, count;
+	u64 prng32_1M;
+	ktime_t kt;
+	int order, last_order;
+
+	tl = mock_timeline(0);
+	if (!tl)
+		return -ENOMEM;
+
+	/* Lookups from cache are very fast and so the random number generation
+	 * and the loop itself becomes a significant factor in the per-iteration
+	 * timings. We try to compensate the results by measuring the overhead
+	 * of the prng and subtract it from the reported results.
+	 */
+	prandom_seed_state(&prng, i915_selftest.random_seed);
+	count = 0;
+	kt = ktime_get();
+	end_time = jiffies + HZ/10;
+	do {
+		u32 x;
+
+		/* Make sure the compiler doesn't optimise away the prng call */
+		WRITE_ONCE(x, prandom_u32_state(&prng));
+
+		count++;
+	} while (!time_after(jiffies, end_time));
+	kt = ktime_sub(ktime_get(), kt);
+	pr_debug("%s: %lu random evaluations, %lluns/prng\n",
+		 __func__, count, (long long)div64_ul(ktime_to_ns(kt), count));
+	prng32_1M = div64_ul(ktime_to_ns(kt) << 20, count);
+
+	/* Benchmark (only) setting random context ids */
+	prandom_seed_state(&prng, i915_selftest.random_seed);
+	count = 0;
+	kt = ktime_get();
+	end_time = jiffies + HZ/10;
+	do {
+		u64 id = i915_prandom_u64_state(&prng);
+
+		__intel_timeline_sync_set(tl, id, 0);
+		count++;
+	} while (!time_after(jiffies, end_time));
+	kt = ktime_sub(ktime_get(), kt);
+	kt = ktime_sub_ns(kt, (count * prng32_1M * 2) >> 20);
+	pr_info("%s: %lu random insertions, %lluns/insert\n",
+		__func__, count, (long long)div64_ul(ktime_to_ns(kt), count));
+
+	/* Benchmark looking up the exact same context ids as we just set */
+	prandom_seed_state(&prng, i915_selftest.random_seed);
+	end_time = count;
+	kt = ktime_get();
+	while (end_time--) {
+		u64 id = i915_prandom_u64_state(&prng);
+
+		if (!__intel_timeline_sync_is_later(tl, id, 0)) {
+			mock_timeline_destroy(tl);
+			pr_err("Lookup of %llu failed\n", id);
+			return -EINVAL;
+		}
+	}
+	kt = ktime_sub(ktime_get(), kt);
+	kt = ktime_sub_ns(kt, (count * prng32_1M * 2) >> 20);
+	pr_info("%s: %lu random lookups, %lluns/lookup\n",
+		__func__, count, (long long)div64_ul(ktime_to_ns(kt), count));
+
+	mock_timeline_destroy(tl);
+	cond_resched();
+
+	tl = mock_timeline(0);
+	if (!tl)
+		return -ENOMEM;
+
+	/* Benchmark setting the first N (in order) contexts */
+	count = 0;
+	kt = ktime_get();
+	end_time = jiffies + HZ/10;
+	do {
+		__intel_timeline_sync_set(tl, count++, 0);
+	} while (!time_after(jiffies, end_time));
+	kt = ktime_sub(ktime_get(), kt);
+	pr_info("%s: %lu in-order insertions, %lluns/insert\n",
+		__func__, count, (long long)div64_ul(ktime_to_ns(kt), count));
+
+	/* Benchmark looking up the exact same context ids as we just set */
+	end_time = count;
+	kt = ktime_get();
+	while (end_time--) {
+		if (!__intel_timeline_sync_is_later(tl, end_time, 0)) {
+			pr_err("Lookup of %lu failed\n", end_time);
+			mock_timeline_destroy(tl);
+			return -EINVAL;
+		}
+	}
+	kt = ktime_sub(ktime_get(), kt);
+	pr_info("%s: %lu in-order lookups, %lluns/lookup\n",
+		__func__, count, (long long)div64_ul(ktime_to_ns(kt), count));
+
+	mock_timeline_destroy(tl);
+	cond_resched();
+
+	tl = mock_timeline(0);
+	if (!tl)
+		return -ENOMEM;
+
+	/* Benchmark searching for a random context id and maybe changing it */
+	prandom_seed_state(&prng, i915_selftest.random_seed);
+	count = 0;
+	kt = ktime_get();
+	end_time = jiffies + HZ/10;
+	do {
+		u32 id = random_engine(&prng);
+		u32 seqno = prandom_u32_state(&prng);
+
+		if (!__intel_timeline_sync_is_later(tl, id, seqno))
+			__intel_timeline_sync_set(tl, id, seqno);
+
+		count++;
+	} while (!time_after(jiffies, end_time));
+	kt = ktime_sub(ktime_get(), kt);
+	kt = ktime_sub_ns(kt, (count * prng32_1M * 2) >> 20);
+	pr_info("%s: %lu repeated insert/lookups, %lluns/op\n",
+		__func__, count, (long long)div64_ul(ktime_to_ns(kt), count));
+	mock_timeline_destroy(tl);
+	cond_resched();
+
+	/* Benchmark searching for a known context id and changing the seqno */
+	for (last_order = 1, order = 1; order < 32;
+	     ({ int tmp = last_order; last_order = order; order += tmp; })) {
+		unsigned int mask = BIT(order) - 1;
+
+		tl = mock_timeline(0);
+		if (!tl)
+			return -ENOMEM;
+
+		count = 0;
+		kt = ktime_get();
+		end_time = jiffies + HZ/10;
+		do {
+			/* Without assuming too many details of the underlying
+			 * implementation, try to identify its phase-changes
+			 * (if any)!
+			 */
+			u64 id = (u64)(count & mask) << order;
+
+			__intel_timeline_sync_is_later(tl, id, 0);
+			__intel_timeline_sync_set(tl, id, 0);
+
+			count++;
+		} while (!time_after(jiffies, end_time));
+		kt = ktime_sub(ktime_get(), kt);
+		pr_info("%s: %lu cyclic/%d insert/lookups, %lluns/op\n",
+			__func__, count, order,
+			(long long)div64_ul(ktime_to_ns(kt), count));
+		mock_timeline_destroy(tl);
+		cond_resched();
+	}
+
+	return 0;
+}
+
+int i915_gem_timeline_mock_selftests(void)
+{
+	static const struct i915_subtest tests[] = {
+		SUBTEST(igt_sync),
+		SUBTEST(bench_sync),
+	};
+
+	return i915_subtests(tests, NULL);
+}
diff --git a/drivers/gpu/drm/i915/selftests/i915_mock_selftests.h b/drivers/gpu/drm/i915/selftests/i915_mock_selftests.h
index be9a9ebf5692..fc74687501ba 100644
--- a/drivers/gpu/drm/i915/selftests/i915_mock_selftests.h
+++ b/drivers/gpu/drm/i915/selftests/i915_mock_selftests.h
@@ -9,9 +9,12 @@
  * Tests are executed in order by igt/drv_selftest
  */
 selftest(sanitycheck, i915_mock_sanitycheck) /* keep first (igt selfcheck) */
+selftest(fence, i915_sw_fence_mock_selftests)
 selftest(scatterlist, scatterlist_mock_selftests)
+selftest(syncmap, i915_syncmap_mock_selftests)
 selftest(uncore, intel_uncore_mock_selftests)
 selftest(breadcrumbs, intel_breadcrumbs_mock_selftests)
+selftest(timelines, i915_gem_timeline_mock_selftests)
 selftest(requests, i915_gem_request_mock_selftests)
 selftest(objects, i915_gem_object_mock_selftests)
 selftest(dmabuf, i915_gem_dmabuf_mock_selftests)
diff --git a/drivers/gpu/drm/i915/selftests/i915_random.c b/drivers/gpu/drm/i915/selftests/i915_random.c
index c17c83c30637..d044bf9a6feb 100644
--- a/drivers/gpu/drm/i915/selftests/i915_random.c
+++ b/drivers/gpu/drm/i915/selftests/i915_random.c
@@ -30,6 +30,17 @@
 
 #include "i915_random.h"
 
+u64 i915_prandom_u64_state(struct rnd_state *rnd)
+{
+	u64 x;
+
+	x = prandom_u32_state(rnd);
+	x <<= 32;
+	x |= prandom_u32_state(rnd);
+
+	return x;
+}
+
 static inline u32 i915_prandom_u32_max_state(u32 ep_ro, struct rnd_state *state)
 {
 	return upper_32_bits((u64)prandom_u32_state(state) * ep_ro);
diff --git a/drivers/gpu/drm/i915/selftests/i915_random.h b/drivers/gpu/drm/i915/selftests/i915_random.h
index b9c334ce6cd9..6c9379871384 100644
--- a/drivers/gpu/drm/i915/selftests/i915_random.h
+++ b/drivers/gpu/drm/i915/selftests/i915_random.h
@@ -41,6 +41,8 @@
 #define I915_RND_SUBSTATE(name__, parent__) \
 	struct rnd_state name__ = I915_RND_STATE_INITIALIZER(prandom_u32_state(&(parent__)))
 
+u64 i915_prandom_u64_state(struct rnd_state *rnd);
+
 unsigned int *i915_random_order(unsigned int count,
 				struct rnd_state *state);
 void i915_random_reorder(unsigned int *order,
diff --git a/drivers/gpu/drm/i915/selftests/i915_sw_fence.c b/drivers/gpu/drm/i915/selftests/i915_sw_fence.c
new file mode 100644
index 000000000000..19d145d6bf52
--- /dev/null
+++ b/drivers/gpu/drm/i915/selftests/i915_sw_fence.c
@@ -0,0 +1,582 @@
+/*
+ * Copyright © 2017 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ */
+
+#include <linux/completion.h>
+#include <linux/delay.h>
+
+#include "../i915_selftest.h"
+
+static int __i915_sw_fence_call
+fence_notify(struct i915_sw_fence *fence, enum i915_sw_fence_notify state)
+{
+	switch (state) {
+	case FENCE_COMPLETE:
+		break;
+
+	case FENCE_FREE:
+		/* Leave the fence for the caller to free it after testing */
+		break;
+	}
+
+	return NOTIFY_DONE;
+}
+
+static struct i915_sw_fence *alloc_fence(void)
+{
+	struct i915_sw_fence *fence;
+
+	fence = kmalloc(sizeof(*fence), GFP_KERNEL);
+	if (!fence)
+		return NULL;
+
+	i915_sw_fence_init(fence, fence_notify);
+	return fence;
+}
+
+static void free_fence(struct i915_sw_fence *fence)
+{
+	i915_sw_fence_fini(fence);
+	kfree(fence);
+}
+
+static int __test_self(struct i915_sw_fence *fence)
+{
+	if (i915_sw_fence_done(fence))
+		return -EINVAL;
+
+	i915_sw_fence_commit(fence);
+	if (!i915_sw_fence_done(fence))
+		return -EINVAL;
+
+	i915_sw_fence_wait(fence);
+	if (!i915_sw_fence_done(fence))
+		return -EINVAL;
+
+	return 0;
+}
+
+static int test_self(void *arg)
+{
+	struct i915_sw_fence *fence;
+	int ret;
+
+	/* Test i915_sw_fence signaling and completion testing */
+	fence = alloc_fence();
+	if (!fence)
+		return -ENOMEM;
+
+	ret = __test_self(fence);
+
+	free_fence(fence);
+	return ret;
+}
+
+static int test_dag(void *arg)
+{
+	struct i915_sw_fence *A, *B, *C;
+	int ret = -EINVAL;
+
+	/* Test detection of cycles within the i915_sw_fence graphs */
+	if (!IS_ENABLED(CONFIG_DRM_I915_SW_FENCE_CHECK_DAG))
+		return 0;
+
+	A = alloc_fence();
+	if (!A)
+		return -ENOMEM;
+
+	if (i915_sw_fence_await_sw_fence_gfp(A, A, GFP_KERNEL) != -EINVAL) {
+		pr_err("recursive cycle not detected (AA)\n");
+		goto err_A;
+	}
+
+	B = alloc_fence();
+	if (!B) {
+		ret = -ENOMEM;
+		goto err_A;
+	}
+
+	i915_sw_fence_await_sw_fence_gfp(A, B, GFP_KERNEL);
+	if (i915_sw_fence_await_sw_fence_gfp(B, A, GFP_KERNEL) != -EINVAL) {
+		pr_err("single depth cycle not detected (BAB)\n");
+		goto err_B;
+	}
+
+	C = alloc_fence();
+	if (!C) {
+		ret = -ENOMEM;
+		goto err_B;
+	}
+
+	if (i915_sw_fence_await_sw_fence_gfp(B, C, GFP_KERNEL) == -EINVAL) {
+		pr_err("invalid cycle detected\n");
+		goto err_C;
+	}
+	if (i915_sw_fence_await_sw_fence_gfp(C, B, GFP_KERNEL) != -EINVAL) {
+		pr_err("single depth cycle not detected (CBC)\n");
+		goto err_C;
+	}
+	if (i915_sw_fence_await_sw_fence_gfp(C, A, GFP_KERNEL) != -EINVAL) {
+		pr_err("cycle not detected (BA, CB, AC)\n");
+		goto err_C;
+	}
+	if (i915_sw_fence_await_sw_fence_gfp(A, C, GFP_KERNEL) == -EINVAL) {
+		pr_err("invalid cycle detected\n");
+		goto err_C;
+	}
+
+	i915_sw_fence_commit(A);
+	i915_sw_fence_commit(B);
+	i915_sw_fence_commit(C);
+
+	ret = 0;
+	if (!i915_sw_fence_done(C)) {
+		pr_err("fence C not done\n");
+		ret = -EINVAL;
+	}
+	if (!i915_sw_fence_done(B)) {
+		pr_err("fence B not done\n");
+		ret = -EINVAL;
+	}
+	if (!i915_sw_fence_done(A)) {
+		pr_err("fence A not done\n");
+		ret = -EINVAL;
+	}
+err_C:
+	free_fence(C);
+err_B:
+	free_fence(B);
+err_A:
+	free_fence(A);
+	return ret;
+}
+
+static int test_AB(void *arg)
+{
+	struct i915_sw_fence *A, *B;
+	int ret;
+
+	/* Test i915_sw_fence (A) waiting on an event source (B) */
+	A = alloc_fence();
+	if (!A)
+		return -ENOMEM;
+	B = alloc_fence();
+	if (!B) {
+		ret = -ENOMEM;
+		goto err_A;
+	}
+
+	ret = i915_sw_fence_await_sw_fence_gfp(A, B, GFP_KERNEL);
+	if (ret < 0)
+		goto err_B;
+	if (ret == 0) {
+		pr_err("Incorrectly reported fence A was complete before await\n");
+		ret = -EINVAL;
+		goto err_B;
+	}
+
+	ret = -EINVAL;
+	i915_sw_fence_commit(A);
+	if (i915_sw_fence_done(A))
+		goto err_B;
+
+	i915_sw_fence_commit(B);
+	if (!i915_sw_fence_done(B)) {
+		pr_err("Fence B is not done\n");
+		goto err_B;
+	}
+
+	if (!i915_sw_fence_done(A)) {
+		pr_err("Fence A is not done\n");
+		goto err_B;
+	}
+
+	ret = 0;
+err_B:
+	free_fence(B);
+err_A:
+	free_fence(A);
+	return ret;
+}
+
+static int test_ABC(void *arg)
+{
+	struct i915_sw_fence *A, *B, *C;
+	int ret;
+
+	/* Test a chain of fences, A waits on B who waits on C */
+	A = alloc_fence();
+	if (!A)
+		return -ENOMEM;
+
+	B = alloc_fence();
+	if (!B) {
+		ret = -ENOMEM;
+		goto err_A;
+	}
+
+	C = alloc_fence();
+	if (!C) {
+		ret = -ENOMEM;
+		goto err_B;
+	}
+
+	ret = i915_sw_fence_await_sw_fence_gfp(A, B, GFP_KERNEL);
+	if (ret < 0)
+		goto err_C;
+	if (ret == 0) {
+		pr_err("Incorrectly reported fence B was complete before await\n");
+		goto err_C;
+	}
+
+	ret = i915_sw_fence_await_sw_fence_gfp(B, C, GFP_KERNEL);
+	if (ret < 0)
+		goto err_C;
+	if (ret == 0) {
+		pr_err("Incorrectly reported fence C was complete before await\n");
+		goto err_C;
+	}
+
+	ret = -EINVAL;
+	i915_sw_fence_commit(A);
+	if (i915_sw_fence_done(A)) {
+		pr_err("Fence A completed early\n");
+		goto err_C;
+	}
+
+	i915_sw_fence_commit(B);
+	if (i915_sw_fence_done(B)) {
+		pr_err("Fence B completed early\n");
+		goto err_C;
+	}
+
+	if (i915_sw_fence_done(A)) {
+		pr_err("Fence A completed early (after signaling B)\n");
+		goto err_C;
+	}
+
+	i915_sw_fence_commit(C);
+
+	ret = 0;
+	if (!i915_sw_fence_done(C)) {
+		pr_err("Fence C not done\n");
+		ret = -EINVAL;
+	}
+	if (!i915_sw_fence_done(B)) {
+		pr_err("Fence B not done\n");
+		ret = -EINVAL;
+	}
+	if (!i915_sw_fence_done(A)) {
+		pr_err("Fence A not done\n");
+		ret = -EINVAL;
+	}
+err_C:
+	free_fence(C);
+err_B:
+	free_fence(B);
+err_A:
+	free_fence(A);
+	return ret;
+}
+
+static int test_AB_C(void *arg)
+{
+	struct i915_sw_fence *A, *B, *C;
+	int ret = -EINVAL;
+
+	/* Test multiple fences (AB) waiting on a single event (C) */
+	A = alloc_fence();
+	if (!A)
+		return -ENOMEM;
+
+	B = alloc_fence();
+	if (!B) {
+		ret = -ENOMEM;
+		goto err_A;
+	}
+
+	C = alloc_fence();
+	if (!C) {
+		ret = -ENOMEM;
+		goto err_B;
+	}
+
+	ret = i915_sw_fence_await_sw_fence_gfp(A, C, GFP_KERNEL);
+	if (ret < 0)
+		goto err_C;
+	if (ret == 0) {
+		ret = -EINVAL;
+		goto err_C;
+	}
+
+	ret = i915_sw_fence_await_sw_fence_gfp(B, C, GFP_KERNEL);
+	if (ret < 0)
+		goto err_C;
+	if (ret == 0) {
+		ret = -EINVAL;
+		goto err_C;
+	}
+
+	i915_sw_fence_commit(A);
+	i915_sw_fence_commit(B);
+
+	ret = 0;
+	if (i915_sw_fence_done(A)) {
+		pr_err("Fence A completed early\n");
+		ret = -EINVAL;
+	}
+
+	if (i915_sw_fence_done(B)) {
+		pr_err("Fence B completed early\n");
+		ret = -EINVAL;
+	}
+
+	i915_sw_fence_commit(C);
+	if (!i915_sw_fence_done(C)) {
+		pr_err("Fence C not done\n");
+		ret = -EINVAL;
+	}
+
+	if (!i915_sw_fence_done(B)) {
+		pr_err("Fence B not done\n");
+		ret = -EINVAL;
+	}
+
+	if (!i915_sw_fence_done(A)) {
+		pr_err("Fence A not done\n");
+		ret = -EINVAL;
+	}
+
+err_C:
+	free_fence(C);
+err_B:
+	free_fence(B);
+err_A:
+	free_fence(A);
+	return ret;
+}
+
+static int test_C_AB(void *arg)
+{
+	struct i915_sw_fence *A, *B, *C;
+	int ret;
+
+	/* Test multiple event sources (A,B) for a single fence (C) */
+	A = alloc_fence();
+	if (!A)
+		return -ENOMEM;
+
+	B = alloc_fence();
+	if (!B) {
+		ret = -ENOMEM;
+		goto err_A;
+	}
+
+	C = alloc_fence();
+	if (!C) {
+		ret = -ENOMEM;
+		goto err_B;
+	}
+
+	ret = i915_sw_fence_await_sw_fence_gfp(C, A, GFP_KERNEL);
+	if (ret < 0)
+		goto err_C;
+	if (ret == 0) {
+		ret = -EINVAL;
+		goto err_C;
+	}
+
+	ret = i915_sw_fence_await_sw_fence_gfp(C, B, GFP_KERNEL);
+	if (ret < 0)
+		goto err_C;
+	if (ret == 0) {
+		ret = -EINVAL;
+		goto err_C;
+	}
+
+	ret = 0;
+	i915_sw_fence_commit(C);
+	if (i915_sw_fence_done(C))
+		ret = -EINVAL;
+
+	i915_sw_fence_commit(A);
+	i915_sw_fence_commit(B);
+
+	if (!i915_sw_fence_done(A)) {
+		pr_err("Fence A not done\n");
+		ret = -EINVAL;
+	}
+
+	if (!i915_sw_fence_done(B)) {
+		pr_err("Fence B not done\n");
+		ret = -EINVAL;
+	}
+
+	if (!i915_sw_fence_done(C)) {
+		pr_err("Fence C not done\n");
+		ret = -EINVAL;
+	}
+
+err_C:
+	free_fence(C);
+err_B:
+	free_fence(B);
+err_A:
+	free_fence(A);
+	return ret;
+}
+
+static int test_chain(void *arg)
+{
+	int nfences = 4096;
+	struct i915_sw_fence **fences;
+	int ret, i;
+
+	/* Test a long chain of fences */
+	fences = kmalloc_array(nfences, sizeof(*fences), GFP_KERNEL);
+	if (!fences)
+		return -ENOMEM;
+
+	for (i = 0; i < nfences; i++) {
+		fences[i] = alloc_fence();
+		if (!fences[i]) {
+			nfences = i;
+			ret = -ENOMEM;
+			goto err;
+		}
+
+		if (i > 0) {
+			ret = i915_sw_fence_await_sw_fence_gfp(fences[i],
+							       fences[i - 1],
+							       GFP_KERNEL);
+			if (ret < 0) {
+				nfences = i + 1;
+				goto err;
+			}
+
+			i915_sw_fence_commit(fences[i]);
+		}
+	}
+
+	ret = 0;
+	for (i = nfences; --i; ) {
+		if (i915_sw_fence_done(fences[i])) {
+			if (ret == 0)
+				pr_err("Fence[%d] completed early\n", i);
+			ret = -EINVAL;
+		}
+	}
+	i915_sw_fence_commit(fences[0]);
+	for (i = 0; ret == 0 && i < nfences; i++) {
+		if (!i915_sw_fence_done(fences[i])) {
+			pr_err("Fence[%d] is not done\n", i);
+			ret = -EINVAL;
+		}
+	}
+
+err:
+	for (i = 0; i < nfences; i++)
+		free_fence(fences[i]);
+	kfree(fences);
+	return ret;
+}
+
+struct task_ipc {
+	struct work_struct work;
+	struct completion started;
+	struct i915_sw_fence *in, *out;
+	int value;
+};
+
+static void task_ipc(struct work_struct *work)
+{
+	struct task_ipc *ipc = container_of(work, typeof(*ipc), work);
+
+	complete(&ipc->started);
+
+	i915_sw_fence_wait(ipc->in);
+	smp_store_mb(ipc->value, 1);
+	i915_sw_fence_commit(ipc->out);
+}
+
+static int test_ipc(void *arg)
+{
+	struct task_ipc ipc;
+	int ret = 0;
+
+	/* Test use of i915_sw_fence as an interprocess signaling mechanism */
+	ipc.in = alloc_fence();
+	if (!ipc.in)
+		return -ENOMEM;
+	ipc.out = alloc_fence();
+	if (!ipc.out) {
+		ret = -ENOMEM;
+		goto err_in;
+	}
+
+	/* use a completion to avoid chicken-and-egg testing */
+	init_completion(&ipc.started);
+
+	ipc.value = 0;
+	INIT_WORK_ONSTACK(&ipc.work, task_ipc);
+	schedule_work(&ipc.work);
+
+	wait_for_completion(&ipc.started);
+
+	usleep_range(1000, 2000);
+	if (READ_ONCE(ipc.value)) {
+		pr_err("worker updated value before i915_sw_fence was signaled\n");
+		ret = -EINVAL;
+	}
+
+	i915_sw_fence_commit(ipc.in);
+	i915_sw_fence_wait(ipc.out);
+
+	if (!READ_ONCE(ipc.value)) {
+		pr_err("worker signaled i915_sw_fence before value was posted\n");
+		ret = -EINVAL;
+	}
+
+	flush_work(&ipc.work);
+	destroy_work_on_stack(&ipc.work);
+	free_fence(ipc.out);
+err_in:
+	free_fence(ipc.in);
+	return ret;
+}
+
+int i915_sw_fence_mock_selftests(void)
+{
+	static const struct i915_subtest tests[] = {
+		SUBTEST(test_self),
+		SUBTEST(test_dag),
+		SUBTEST(test_AB),
+		SUBTEST(test_ABC),
+		SUBTEST(test_AB_C),
+		SUBTEST(test_C_AB),
+		SUBTEST(test_chain),
+		SUBTEST(test_ipc),
+	};
+
+	return i915_subtests(tests, NULL);
+}
diff --git a/drivers/gpu/drm/i915/selftests/i915_syncmap.c b/drivers/gpu/drm/i915/selftests/i915_syncmap.c
new file mode 100644
index 000000000000..bcab3d00a785
--- /dev/null
+++ b/drivers/gpu/drm/i915/selftests/i915_syncmap.c
@@ -0,0 +1,616 @@
+/*
+ * Copyright © 2017 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ */
+
+#include "../i915_selftest.h"
+#include "i915_random.h"
+
+static char *
+__sync_print(struct i915_syncmap *p,
+	     char *buf, unsigned long *sz,
+	     unsigned int depth,
+	     unsigned int last,
+	     unsigned int idx)
+{
+	unsigned long len;
+	unsigned int i, X;
+
+	if (depth) {
+		unsigned int d;
+
+		for (d = 0; d < depth - 1; d++) {
+			if (last & BIT(depth - d - 1))
+				len = scnprintf(buf, *sz, "|   ");
+			else
+				len = scnprintf(buf, *sz, "    ");
+			buf += len;
+			*sz -= len;
+		}
+		len = scnprintf(buf, *sz, "%x-> ", idx);
+		buf += len;
+		*sz -= len;
+	}
+
+	/* We mark bits after the prefix as "X" */
+	len = scnprintf(buf, *sz, "0x%016llx", p->prefix << p->height << SHIFT);
+	buf += len;
+	*sz -= len;
+	X = (p->height + SHIFT) / 4;
+	scnprintf(buf - X, *sz + X, "%*s", X, "XXXXXXXXXXXXXXXXX");
+
+	if (!p->height) {
+		for_each_set_bit(i, (unsigned long *)&p->bitmap, KSYNCMAP) {
+			len = scnprintf(buf, *sz, " %x:%x,",
+					i, __sync_seqno(p)[i]);
+			buf += len;
+			*sz -= len;
+		}
+		buf -= 1;
+		*sz += 1;
+	}
+
+	len = scnprintf(buf, *sz, "\n");
+	buf += len;
+	*sz -= len;
+
+	if (p->height) {
+		for_each_set_bit(i, (unsigned long *)&p->bitmap, KSYNCMAP) {
+			buf = __sync_print(__sync_child(p)[i], buf, sz,
+					   depth + 1,
+					   last << 1 | !!(p->bitmap >> (i + 1)),
+					   i);
+		}
+	}
+
+	return buf;
+}
+
+static bool
+i915_syncmap_print_to_buf(struct i915_syncmap *p, char *buf, unsigned long sz)
+{
+	if (!p)
+		return false;
+
+	while (p->parent)
+		p = p->parent;
+
+	__sync_print(p, buf, &sz, 0, 1, 0);
+	return true;
+}
+
+static int check_syncmap_free(struct i915_syncmap **sync)
+{
+	i915_syncmap_free(sync);
+	if (*sync) {
+		pr_err("sync not cleared after free\n");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int dump_syncmap(struct i915_syncmap *sync, int err)
+{
+	char *buf;
+
+	if (!err)
+		return check_syncmap_free(&sync);
+
+	buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
+	if (!buf)
+		goto skip;
+
+	if (i915_syncmap_print_to_buf(sync, buf, PAGE_SIZE))
+		pr_err("%s", buf);
+
+	kfree(buf);
+
+skip:
+	i915_syncmap_free(&sync);
+	return err;
+}
+
+static int igt_syncmap_init(void *arg)
+{
+	struct i915_syncmap *sync = (void *)~0ul;
+
+	/*
+	 * Cursory check that we can initialise a random pointer and transform
+	 * it into the root pointer of a syncmap.
+	 */
+
+	i915_syncmap_init(&sync);
+	return check_syncmap_free(&sync);
+}
+
+static int check_seqno(struct i915_syncmap *leaf, unsigned int idx, u32 seqno)
+{
+	if (leaf->height) {
+		pr_err("%s: not a leaf, height is %d\n",
+		       __func__, leaf->height);
+		return -EINVAL;
+	}
+
+	if (__sync_seqno(leaf)[idx] != seqno) {
+		pr_err("%s: seqno[%d], found %x, expected %x\n",
+		       __func__, idx, __sync_seqno(leaf)[idx], seqno);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int check_one(struct i915_syncmap **sync, u64 context, u32 seqno)
+{
+	int err;
+
+	err = i915_syncmap_set(sync, context, seqno);
+	if (err)
+		return err;
+
+	if ((*sync)->height) {
+		pr_err("Inserting first context=%llx did not return leaf (height=%d, prefix=%llx\n",
+		       context, (*sync)->height, (*sync)->prefix);
+		return -EINVAL;
+	}
+
+	if ((*sync)->parent) {
+		pr_err("Inserting first context=%llx created branches!\n",
+		       context);
+		return -EINVAL;
+	}
+
+	if (hweight32((*sync)->bitmap) != 1) {
+		pr_err("First bitmap does not contain a single entry, found %x (count=%d)!\n",
+		       (*sync)->bitmap, hweight32((*sync)->bitmap));
+		return -EINVAL;
+	}
+
+	err = check_seqno((*sync), ilog2((*sync)->bitmap), seqno);
+	if (err)
+		return err;
+
+	if (!i915_syncmap_is_later(sync, context, seqno)) {
+		pr_err("Lookup of first context=%llx/seqno=%x failed!\n",
+		       context, seqno);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int igt_syncmap_one(void *arg)
+{
+	I915_RND_STATE(prng);
+	IGT_TIMEOUT(end_time);
+	struct i915_syncmap *sync;
+	unsigned long max = 1;
+	int err;
+
+	/*
+	 * Check that inserting a new id, creates a leaf and only that leaf.
+	 */
+
+	i915_syncmap_init(&sync);
+
+	do {
+		u64 context = i915_prandom_u64_state(&prng);
+		unsigned long loop;
+
+		err = check_syncmap_free(&sync);
+		if (err)
+			goto out;
+
+		for (loop = 0; loop <= max; loop++) {
+			err = check_one(&sync, context,
+					prandom_u32_state(&prng));
+			if (err)
+				goto out;
+		}
+		max++;
+	} while (!__igt_timeout(end_time, NULL));
+	pr_debug("%s: Completed %lu single insertions\n",
+		 __func__, max * (max - 1) / 2);
+out:
+	return dump_syncmap(sync, err);
+}
+
+static int check_leaf(struct i915_syncmap **sync, u64 context, u32 seqno)
+{
+	int err;
+
+	err = i915_syncmap_set(sync, context, seqno);
+	if (err)
+		return err;
+
+	if ((*sync)->height) {
+		pr_err("Inserting context=%llx did not return leaf (height=%d, prefix=%llx\n",
+		       context, (*sync)->height, (*sync)->prefix);
+		return -EINVAL;
+	}
+
+	if (hweight32((*sync)->bitmap) != 1) {
+		pr_err("First entry into leaf (context=%llx) does not contain a single entry, found %x (count=%d)!\n",
+		       context, (*sync)->bitmap, hweight32((*sync)->bitmap));
+		return -EINVAL;
+	}
+
+	err = check_seqno((*sync), ilog2((*sync)->bitmap), seqno);
+	if (err)
+		return err;
+
+	if (!i915_syncmap_is_later(sync, context, seqno)) {
+		pr_err("Lookup of first entry context=%llx/seqno=%x failed!\n",
+		       context, seqno);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int igt_syncmap_join_above(void *arg)
+{
+	struct i915_syncmap *sync;
+	unsigned int pass, order;
+	int err;
+
+	i915_syncmap_init(&sync);
+
+	/*
+	 * When we have a new id that doesn't fit inside the existing tree,
+	 * we need to add a new layer above.
+	 *
+	 * 1: 0x00000001
+	 * 2: 0x00000010
+	 * 3: 0x00000100
+	 * 4: 0x00001000
+	 * ...
+	 * Each pass the common prefix shrinks and we have to insert a join.
+	 * Each join will only contain two branches, the latest of which
+	 * is always a leaf.
+	 *
+	 * If we then reuse the same set of contexts, we expect to build an
+	 * identical tree.
+	 */
+	for (pass = 0; pass < 3; pass++) {
+		for (order = 0; order < 64; order += SHIFT) {
+			u64 context = BIT_ULL(order);
+			struct i915_syncmap *join;
+
+			err = check_leaf(&sync, context, 0);
+			if (err)
+				goto out;
+
+			join = sync->parent;
+			if (!join) /* very first insert will have no parents */
+				continue;
+
+			if (!join->height) {
+				pr_err("Parent with no height!\n");
+				err = -EINVAL;
+				goto out;
+			}
+
+			if (hweight32(join->bitmap) != 2) {
+				pr_err("Join does not have 2 children: %x (%d)\n",
+				       join->bitmap, hweight32(join->bitmap));
+				err = -EINVAL;
+				goto out;
+			}
+
+			if (__sync_child(join)[__sync_branch_idx(join, context)] != sync) {
+				pr_err("Leaf misplaced in parent!\n");
+				err = -EINVAL;
+				goto out;
+			}
+		}
+	}
+out:
+	return dump_syncmap(sync, err);
+}
+
+static int igt_syncmap_join_below(void *arg)
+{
+	struct i915_syncmap *sync;
+	unsigned int step, order, idx;
+	int err;
+
+	i915_syncmap_init(&sync);
+
+	/*
+	 * Check that we can split a compacted branch by replacing it with
+	 * a join.
+	 */
+	for (step = 0; step < KSYNCMAP; step++) {
+		for (order = 64 - SHIFT; order > 0; order -= SHIFT) {
+			u64 context = step * BIT_ULL(order);
+
+			err = i915_syncmap_set(&sync, context, 0);
+			if (err)
+				goto out;
+
+			if (sync->height) {
+				pr_err("Inserting context=%llx (order=%d, step=%d) did not return leaf (height=%d, prefix=%llx\n",
+				       context, order, step, sync->height, sync->prefix);
+				err = -EINVAL;
+				goto out;
+			}
+		}
+	}
+
+	for (step = 0; step < KSYNCMAP; step++) {
+		for (order = SHIFT; order < 64; order += SHIFT) {
+			u64 context = step * BIT_ULL(order);
+
+			if (!i915_syncmap_is_later(&sync, context, 0)) {
+				pr_err("1: context %llx (order=%d, step=%d) not found\n",
+				       context, order, step);
+				err = -EINVAL;
+				goto out;
+			}
+
+			for (idx = 1; idx < KSYNCMAP; idx++) {
+				if (i915_syncmap_is_later(&sync, context + idx, 0)) {
+					pr_err("1: context %llx (order=%d, step=%d) should not exist\n",
+					       context + idx, order, step);
+					err = -EINVAL;
+					goto out;
+				}
+			}
+		}
+	}
+
+	for (order = SHIFT; order < 64; order += SHIFT) {
+		for (step = 0; step < KSYNCMAP; step++) {
+			u64 context = step * BIT_ULL(order);
+
+			if (!i915_syncmap_is_later(&sync, context, 0)) {
+				pr_err("2: context %llx (order=%d, step=%d) not found\n",
+				       context, order, step);
+				err = -EINVAL;
+				goto out;
+			}
+		}
+	}
+
+out:
+	return dump_syncmap(sync, err);
+}
+
+static int igt_syncmap_neighbours(void *arg)
+{
+	I915_RND_STATE(prng);
+	IGT_TIMEOUT(end_time);
+	struct i915_syncmap *sync;
+	int err;
+
+	/*
+	 * Each leaf holds KSYNCMAP seqno. Check that when we create KSYNCMAP
+	 * neighbouring ids, they all fit into the same leaf.
+	 */
+
+	i915_syncmap_init(&sync);
+	do {
+		u64 context = i915_prandom_u64_state(&prng) & ~MASK;
+		unsigned int idx;
+
+		if (i915_syncmap_is_later(&sync, context, 0)) /* Skip repeats */
+			continue;
+
+		for (idx = 0; idx < KSYNCMAP; idx++) {
+			err = i915_syncmap_set(&sync, context + idx, 0);
+			if (err)
+				goto out;
+
+			if (sync->height) {
+				pr_err("Inserting context=%llx did not return leaf (height=%d, prefix=%llx\n",
+				       context, sync->height, sync->prefix);
+				err = -EINVAL;
+				goto out;
+			}
+
+			if (sync->bitmap != BIT(idx + 1) - 1) {
+				pr_err("Inserting neighbouring context=0x%llx+%d, did not fit into the same leaf bitmap=%x (%d), expected %lx (%d)\n",
+				       context, idx,
+				       sync->bitmap, hweight32(sync->bitmap),
+				       BIT(idx + 1) - 1, idx + 1);
+				err = -EINVAL;
+				goto out;
+			}
+		}
+	} while (!__igt_timeout(end_time, NULL));
+out:
+	return dump_syncmap(sync, err);
+}
+
+static int igt_syncmap_compact(void *arg)
+{
+	struct i915_syncmap *sync;
+	unsigned int idx, order;
+	int err;
+
+	i915_syncmap_init(&sync);
+
+	/*
+	 * The syncmap are "space efficient" compressed radix trees - any
+	 * branch with only one child is skipped and replaced by the child.
+	 *
+	 * If we construct a tree with ids that are neighbouring at a non-zero
+	 * height, we form a join but each child of that join is directly a
+	 * leaf holding the single id.
+	 */
+	for (order = SHIFT; order < 64; order += SHIFT) {
+		err = check_syncmap_free(&sync);
+		if (err)
+			goto out;
+
+		/* Create neighbours in the parent */
+		for (idx = 0; idx < KSYNCMAP; idx++) {
+			u64 context = idx * BIT_ULL(order) + idx;
+
+			err = i915_syncmap_set(&sync, context, 0);
+			if (err)
+				goto out;
+
+			if (sync->height) {
+				pr_err("Inserting context=%llx (order=%d, idx=%d) did not return leaf (height=%d, prefix=%llx\n",
+				       context, order, idx,
+				       sync->height, sync->prefix);
+				err = -EINVAL;
+				goto out;
+			}
+		}
+
+		sync = sync->parent;
+		if (sync->parent) {
+			pr_err("Parent (join) of last leaf was not the sync!\n");
+			err = -EINVAL;
+			goto out;
+		}
+
+		if (sync->height != order) {
+			pr_err("Join does not have the expected height, found %d, expected %d\n",
+			       sync->height, order);
+			err = -EINVAL;
+			goto out;
+		}
+
+		if (sync->bitmap != BIT(KSYNCMAP) - 1) {
+			pr_err("Join is not full!, found %x (%d) expected %lx (%d)\n",
+			       sync->bitmap, hweight32(sync->bitmap),
+			       BIT(KSYNCMAP) - 1, KSYNCMAP);
+			err = -EINVAL;
+			goto out;
+		}
+
+		/* Each of our children should be a leaf */
+		for (idx = 0; idx < KSYNCMAP; idx++) {
+			struct i915_syncmap *leaf = __sync_child(sync)[idx];
+
+			if (leaf->height) {
+				pr_err("Child %d is a not leaf!\n", idx);
+				err = -EINVAL;
+				goto out;
+			}
+
+			if (leaf->parent != sync) {
+				pr_err("Child %d is not attached to us!\n",
+				       idx);
+				err = -EINVAL;
+				goto out;
+			}
+
+			if (!is_power_of_2(leaf->bitmap)) {
+				pr_err("Child %d holds more than one id, found %x (%d)\n",
+				       idx, leaf->bitmap, hweight32(leaf->bitmap));
+				err = -EINVAL;
+				goto out;
+			}
+
+			if (leaf->bitmap != BIT(idx)) {
+				pr_err("Child %d has wrong seqno idx, found %d, expected %d\n",
+				       idx, ilog2(leaf->bitmap), idx);
+				err = -EINVAL;
+				goto out;
+			}
+		}
+	}
+out:
+	return dump_syncmap(sync, err);
+}
+
+static int igt_syncmap_random(void *arg)
+{
+	I915_RND_STATE(prng);
+	IGT_TIMEOUT(end_time);
+	struct i915_syncmap *sync;
+	unsigned long count, phase, i;
+	u32 seqno;
+	int err;
+
+	i915_syncmap_init(&sync);
+
+	/*
+	 * Having tried to test the individual operations within i915_syncmap,
+	 * run a smoketest exploring the entire u64 space with random
+	 * insertions.
+	 */
+
+	count = 0;
+	phase = jiffies + HZ/100 + 1;
+	do {
+		u64 context = i915_prandom_u64_state(&prng);
+
+		err = i915_syncmap_set(&sync, context, 0);
+		if (err)
+			goto out;
+
+		count++;
+	} while (!time_after(jiffies, phase));
+	seqno = 0;
+
+	phase = 0;
+	do {
+		I915_RND_STATE(ctx);
+		u32 last_seqno = seqno;
+		bool expect;
+
+		seqno = prandom_u32_state(&prng);
+		expect = seqno_later(last_seqno, seqno);
+
+		for (i = 0; i < count; i++) {
+			u64 context = i915_prandom_u64_state(&ctx);
+
+			if (i915_syncmap_is_later(&sync, context, seqno) != expect) {
+				pr_err("context=%llu, last=%u this=%u did not match expectation (%d)\n",
+				       context, last_seqno, seqno, expect);
+				err = -EINVAL;
+				goto out;
+			}
+
+			err = i915_syncmap_set(&sync, context, seqno);
+			if (err)
+				goto out;
+		}
+
+		phase++;
+	} while (!__igt_timeout(end_time, NULL));
+	pr_debug("Completed %lu passes, each of %lu contexts\n", phase, count);
+out:
+	return dump_syncmap(sync, err);
+}
+
+int i915_syncmap_mock_selftests(void)
+{
+	static const struct i915_subtest tests[] = {
+		SUBTEST(igt_syncmap_init),
+		SUBTEST(igt_syncmap_one),
+		SUBTEST(igt_syncmap_join_above),
+		SUBTEST(igt_syncmap_join_below),
+		SUBTEST(igt_syncmap_neighbours),
+		SUBTEST(igt_syncmap_compact),
+		SUBTEST(igt_syncmap_random),
+	};
+
+	return i915_subtests(tests, NULL);
+}
diff --git a/drivers/gpu/drm/i915/selftests/intel_breadcrumbs.c b/drivers/gpu/drm/i915/selftests/intel_breadcrumbs.c
index 19860a372d90..7276194c04f7 100644
--- a/drivers/gpu/drm/i915/selftests/intel_breadcrumbs.c
+++ b/drivers/gpu/drm/i915/selftests/intel_breadcrumbs.c
@@ -117,7 +117,7 @@ static int igt_random_insert_remove(void *arg)
 
 	mock_engine_reset(engine);
 
-	waiters = drm_malloc_gfp(count, sizeof(*waiters), GFP_TEMPORARY);
+	waiters = kvmalloc_array(count, sizeof(*waiters), GFP_TEMPORARY);
 	if (!waiters)
 		goto out_engines;
 
@@ -169,7 +169,7 @@ out_order:
 out_bitmap:
 	kfree(bitmap);
 out_waiters:
-	drm_free_large(waiters);
+	kvfree(waiters);
 out_engines:
 	mock_engine_flush(engine);
 	return err;
@@ -187,7 +187,7 @@ static int igt_insert_complete(void *arg)
 
 	mock_engine_reset(engine);
 
-	waiters = drm_malloc_gfp(count, sizeof(*waiters), GFP_TEMPORARY);
+	waiters = kvmalloc_array(count, sizeof(*waiters), GFP_TEMPORARY);
 	if (!waiters)
 		goto out_engines;
 
@@ -254,7 +254,7 @@ static int igt_insert_complete(void *arg)
 out_bitmap:
 	kfree(bitmap);
 out_waiters:
-	drm_free_large(waiters);
+	kvfree(waiters);
 out_engines:
 	mock_engine_flush(engine);
 	return err;
@@ -368,7 +368,7 @@ static int igt_wakeup(void *arg)
 
 	mock_engine_reset(engine);
 
-	waiters = drm_malloc_gfp(count, sizeof(*waiters), GFP_TEMPORARY);
+	waiters = kvmalloc_array(count, sizeof(*waiters), GFP_TEMPORARY);
 	if (!waiters)
 		goto out_engines;
 
@@ -454,7 +454,7 @@ out_waiters:
 		put_task_struct(waiters[n].tsk);
 	}
 
-	drm_free_large(waiters);
+	kvfree(waiters);
 out_engines:
 	mock_engine_flush(engine);
 	return err;
diff --git a/drivers/gpu/drm/i915/selftests/mock_engine.c b/drivers/gpu/drm/i915/selftests/mock_engine.c
index 0ad624a1db90..5b18a2dc19a8 100644
--- a/drivers/gpu/drm/i915/selftests/mock_engine.c
+++ b/drivers/gpu/drm/i915/selftests/mock_engine.c
@@ -52,11 +52,12 @@ static void hw_delay_complete(unsigned long data)
 	spin_unlock(&engine->hw_lock);
 }
 
-static int mock_context_pin(struct intel_engine_cs *engine,
-			    struct i915_gem_context *ctx)
+static struct intel_ring *
+mock_context_pin(struct intel_engine_cs *engine,
+		 struct i915_gem_context *ctx)
 {
 	i915_gem_context_get(ctx);
-	return 0;
+	return engine->buffer;
 }
 
 static void mock_context_unpin(struct intel_engine_cs *engine,
@@ -72,7 +73,6 @@ static int mock_request_alloc(struct drm_i915_gem_request *request)
 	INIT_LIST_HEAD(&mock->link);
 	mock->delay = 0;
 
-	request->ring = request->engine->buffer;
 	return 0;
 }
 
@@ -112,7 +112,6 @@ static struct intel_ring *mock_ring(struct intel_engine_cs *engine)
 	if (!ring)
 		return NULL;
 
-	ring->engine = engine;
 	ring->size = sz;
 	ring->effective_size = sz;
 	ring->vaddr = (void *)(ring + 1);
@@ -141,7 +140,7 @@ struct intel_engine_cs *mock_engine(struct drm_i915_private *i915,
 
 	/* minimal engine setup for requests */
 	engine->base.i915 = i915;
-	engine->base.name = name;
+	snprintf(engine->base.name, sizeof(engine->base.name), "%s", name);
 	engine->base.id = id++;
 	engine->base.status_page.page_addr = (void *)(engine + 1);
 
diff --git a/drivers/gpu/drm/i915/selftests/mock_gem_device.c b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
index 9f24c5da3f8d..627e2aa09766 100644
--- a/drivers/gpu/drm/i915/selftests/mock_gem_device.c
+++ b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
@@ -30,6 +30,7 @@
 #include "mock_gem_device.h"
 #include "mock_gem_object.h"
 #include "mock_gtt.h"
+#include "mock_uncore.h"
 
 void mock_device_flush(struct drm_i915_private *i915)
 {
@@ -73,6 +74,7 @@ static void mock_device_release(struct drm_device *dev)
 
 	destroy_workqueue(i915->wq);
 
+	kmem_cache_destroy(i915->priorities);
 	kmem_cache_destroy(i915->dependencies);
 	kmem_cache_destroy(i915->requests);
 	kmem_cache_destroy(i915->vmas);
@@ -119,6 +121,7 @@ struct drm_i915_private *mock_gem_device(void)
 		goto err;
 
 	device_initialize(&pdev->dev);
+	pdev->class = PCI_BASE_CLASS_DISPLAY << 16;
 	pdev->dev.release = release_dev;
 	dev_set_name(&pdev->dev, "mock");
 	dma_coerce_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32));
@@ -143,6 +146,7 @@ struct drm_i915_private *mock_gem_device(void)
 	mkwrite_device_info(i915)->gen = -1;
 
 	spin_lock_init(&i915->mm.object_stat_lock);
+	mock_uncore_init(i915);
 
 	init_waitqueue_head(&i915->gpu_error.wait_queue);
 	init_waitqueue_head(&i915->gpu_error.reset_queue);
@@ -184,12 +188,16 @@ struct drm_i915_private *mock_gem_device(void)
 	if (!i915->dependencies)
 		goto err_requests;
 
+	i915->priorities = KMEM_CACHE(i915_priolist, SLAB_HWCACHE_ALIGN);
+	if (!i915->priorities)
+		goto err_dependencies;
+
 	mutex_lock(&i915->drm.struct_mutex);
 	INIT_LIST_HEAD(&i915->gt.timelines);
 	err = i915_gem_timeline_init__global(i915);
 	if (err) {
 		mutex_unlock(&i915->drm.struct_mutex);
-		goto err_dependencies;
+		goto err_priorities;
 	}
 
 	mock_init_ggtt(i915);
@@ -209,6 +217,8 @@ struct drm_i915_private *mock_gem_device(void)
 err_engine:
 	for_each_engine(engine, i915, id)
 		mock_engine_free(engine);
+err_priorities:
+	kmem_cache_destroy(i915->priorities);
 err_dependencies:
 	kmem_cache_destroy(i915->dependencies);
 err_requests:
diff --git a/drivers/gpu/drm/i915/selftests/mock_timeline.c b/drivers/gpu/drm/i915/selftests/mock_timeline.c
new file mode 100644
index 000000000000..47b1f47c5812
--- /dev/null
+++ b/drivers/gpu/drm/i915/selftests/mock_timeline.c
@@ -0,0 +1,45 @@
+/*
+ * Copyright © 2017 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ */
+
+#include "mock_timeline.h"
+
+struct intel_timeline *mock_timeline(u64 context)
+{
+	static struct lock_class_key class;
+	struct intel_timeline *tl;
+
+	tl = kzalloc(sizeof(*tl), GFP_KERNEL);
+	if (!tl)
+		return NULL;
+
+	__intel_timeline_init(tl, NULL, context, &class, "mock");
+
+	return tl;
+}
+
+void mock_timeline_destroy(struct intel_timeline *tl)
+{
+	__intel_timeline_fini(tl);
+	kfree(tl);
+}
diff --git a/drivers/gpu/drm/i915/selftests/mock_timeline.h b/drivers/gpu/drm/i915/selftests/mock_timeline.h
new file mode 100644
index 000000000000..c27ff4639b8b
--- /dev/null
+++ b/drivers/gpu/drm/i915/selftests/mock_timeline.h
@@ -0,0 +1,33 @@
+/*
+ * Copyright © 2017 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __MOCK_TIMELINE__
+#define __MOCK_TIMELINE__
+
+#include "../i915_gem_timeline.h"
+
+struct intel_timeline *mock_timeline(u64 context);
+void mock_timeline_destroy(struct intel_timeline *tl);
+
+#endif /* !__MOCK_TIMELINE__ */
diff --git a/drivers/gpu/drm/i915/selftests/mock_uncore.c b/drivers/gpu/drm/i915/selftests/mock_uncore.c
new file mode 100644
index 000000000000..8ef14c7e5e38
--- /dev/null
+++ b/drivers/gpu/drm/i915/selftests/mock_uncore.c
@@ -0,0 +1,46 @@
+/*
+ * Copyright © 2017 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ */
+
+#include "mock_uncore.h"
+
+#define __nop_write(x) \
+static void \
+nop_write##x(struct drm_i915_private *dev_priv, i915_reg_t reg, u##x val, bool trace) { }
+__nop_write(8)
+__nop_write(16)
+__nop_write(32)
+
+#define __nop_read(x) \
+static u##x \
+nop_read##x(struct drm_i915_private *dev_priv, i915_reg_t reg, bool trace) { return 0; }
+__nop_read(8)
+__nop_read(16)
+__nop_read(32)
+__nop_read(64)
+
+void mock_uncore_init(struct drm_i915_private *i915)
+{
+	ASSIGN_WRITE_MMIO_VFUNCS(i915, nop);
+	ASSIGN_READ_MMIO_VFUNCS(i915, nop);
+}
diff --git a/drivers/gpu/drm/i915/selftests/mock_uncore.h b/drivers/gpu/drm/i915/selftests/mock_uncore.h
new file mode 100644
index 000000000000..d79aa3ca4d51
--- /dev/null
+++ b/drivers/gpu/drm/i915/selftests/mock_uncore.h
@@ -0,0 +1,30 @@
+/*
+ * Copyright © 2017 Intel Corporation
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ *
+ */
+
+#ifndef __MOCK_UNCORE_H
+#define __MOCK_UNCORE_H
+
+void mock_uncore_init(struct drm_i915_private *i915);
+
+#endif /* !__MOCK_UNCORE_H */
diff --git a/drivers/gpu/drm/imx/ipuv3-plane.c b/drivers/gpu/drm/imx/ipuv3-plane.c
index d63e853a0300..49546222c6d3 100644
--- a/drivers/gpu/drm/imx/ipuv3-plane.c
+++ b/drivers/gpu/drm/imx/ipuv3-plane.c
@@ -273,7 +273,7 @@ void ipu_plane_state_reset(struct drm_plane *plane)
 
 	if (ipu_state) {
 		ipu_state->base.plane = plane;
-		ipu_state->base.rotation = DRM_ROTATE_0;
+		ipu_state->base.rotation = DRM_MODE_ROTATE_0;
 	}
 
 	plane->state = &ipu_state->base;
diff --git a/drivers/gpu/drm/mga/Makefile b/drivers/gpu/drm/mga/Makefile
index 60684785c203..49e972c2f787 100644
--- a/drivers/gpu/drm/mga/Makefile
+++ b/drivers/gpu/drm/mga/Makefile
@@ -2,7 +2,6 @@
 # Makefile for the drm device driver.  This driver provides support for the
 # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.
 
-ccflags-y := -Iinclude/drm
 mga-y := mga_drv.o mga_dma.o mga_state.o mga_warp.o mga_irq.o
 
 mga-$(CONFIG_COMPAT) += mga_ioc32.o
diff --git a/drivers/gpu/drm/mgag200/Makefile b/drivers/gpu/drm/mgag200/Makefile
index a9a0300f09fc..3d91d1d6c45d 100644
--- a/drivers/gpu/drm/mgag200/Makefile
+++ b/drivers/gpu/drm/mgag200/Makefile
@@ -1,4 +1,3 @@
-ccflags-y := -Iinclude/drm
 mgag200-y   := mgag200_main.o mgag200_mode.o mgag200_cursor.o \
 	mgag200_drv.o mgag200_fb.o mgag200_i2c.o mgag200_ttm.o
 
diff --git a/drivers/gpu/drm/mgag200/mgag200_ttm.c b/drivers/gpu/drm/mgag200/mgag200_ttm.c
index 565a217b46f2..3e7e1cd31395 100644
--- a/drivers/gpu/drm/mgag200/mgag200_ttm.c
+++ b/drivers/gpu/drm/mgag200/mgag200_ttm.c
@@ -26,8 +26,9 @@
  * Authors: Dave Airlie <airlied@redhat.com>
  */
 #include <drm/drmP.h>
+#include <drm/ttm/ttm_page_alloc.h>
+
 #include "mgag200_drv.h"
-#include <ttm/ttm_page_alloc.h>
 
 static inline struct mga_device *
 mgag200_bdev(struct ttm_bo_device *bd)
diff --git a/drivers/gpu/drm/msm/Makefile b/drivers/gpu/drm/msm/Makefile
index 5241ac8803ba..33008fa1be9b 100644
--- a/drivers/gpu/drm/msm/Makefile
+++ b/drivers/gpu/drm/msm/Makefile
@@ -1,4 +1,4 @@
-ccflags-y := -Iinclude/drm -Idrivers/gpu/drm/msm
+ccflags-y := -Idrivers/gpu/drm/msm
 ccflags-$(CONFIG_DRM_MSM_DSI) += -Idrivers/gpu/drm/msm/dsi
 
 msm-y := \
diff --git a/drivers/gpu/drm/msm/dsi/dsi.h b/drivers/gpu/drm/msm/dsi/dsi.h
index 32369975d155..9e6017387efb 100644
--- a/drivers/gpu/drm/msm/dsi/dsi.h
+++ b/drivers/gpu/drm/msm/dsi/dsi.h
@@ -17,9 +17,9 @@
 #include <linux/of_platform.h>
 #include <linux/platform_device.h>
 
-#include "drm_crtc.h"
-#include "drm_mipi_dsi.h"
-#include "drm_panel.h"
+#include <drm/drm_crtc.h>
+#include <drm/drm_mipi_dsi.h>
+#include <drm/drm_panel.h>
 
 #include "msm_drv.h"
 
diff --git a/drivers/gpu/drm/msm/edp/edp.h b/drivers/gpu/drm/msm/edp/edp.h
index ba5bedde5241..e0f5818ec9ca 100644
--- a/drivers/gpu/drm/msm/edp/edp.h
+++ b/drivers/gpu/drm/msm/edp/edp.h
@@ -18,9 +18,9 @@
 #include <linux/interrupt.h>
 #include <linux/kernel.h>
 #include <linux/platform_device.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_dp_helper.h>
 
-#include "drm_crtc.h"
-#include "drm_dp_helper.h"
 #include "msm_drv.h"
 
 #define edp_read(offset) msm_readl((offset))
diff --git a/drivers/gpu/drm/msm/edp/edp_ctrl.c b/drivers/gpu/drm/msm/edp/edp_ctrl.c
index 149bfe7ddd82..e32a4a4f3797 100644
--- a/drivers/gpu/drm/msm/edp/edp_ctrl.c
+++ b/drivers/gpu/drm/msm/edp/edp_ctrl.c
@@ -14,10 +14,10 @@
 #include <linux/clk.h>
 #include <linux/gpio/consumer.h>
 #include <linux/regulator/consumer.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_dp_helper.h>
+#include <drm/drm_edid.h>
 
-#include "drm_crtc.h"
-#include "drm_dp_helper.h"
-#include "drm_edid.h"
 #include "edp.h"
 #include "edp.xml.h"
 
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c
index f29194a74a19..698e514203c6 100644
--- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_crtc.c
@@ -15,12 +15,12 @@
  * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
-#include "mdp4_kms.h"
-
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_flip_work.h>
 #include <drm/drm_mode.h>
-#include "drm_crtc.h"
-#include "drm_crtc_helper.h"
-#include "drm_flip_work.h"
+
+#include "mdp4_kms.h"
 
 struct mdp4_crtc {
 	struct drm_crtc base;
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dsi_encoder.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dsi_encoder.c
index 106f0e772595..6a1ebdace391 100644
--- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dsi_encoder.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dsi_encoder.c
@@ -17,10 +17,10 @@
  * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
-#include "mdp4_kms.h"
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
 
-#include "drm_crtc.h"
-#include "drm_crtc_helper.h"
+#include "mdp4_kms.h"
 
 struct mdp4_dsi_encoder {
 	struct drm_encoder base;
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c
index 24258e3025e3..ba8e587f734b 100644
--- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_dtv_encoder.c
@@ -15,11 +15,10 @@
  * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
-#include "mdp4_kms.h"
-
-#include "drm_crtc.h"
-#include "drm_crtc_helper.h"
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
 
+#include "mdp4_kms.h"
 
 struct mdp4_dtv_encoder {
 	struct drm_encoder base;
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h
index 62712ca164ee..c413779d488a 100644
--- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_kms.h
@@ -18,12 +18,14 @@
 #ifndef __MDP4_KMS_H__
 #define __MDP4_KMS_H__
 
+#include <drm/drm_panel.h>
+
 #include "msm_drv.h"
 #include "msm_kms.h"
 #include "mdp/mdp_kms.h"
 #include "mdp4.xml.h"
 
-#include "drm_panel.h"
+struct device_node;
 
 struct mdp4_kms {
 	struct mdp_kms base;
diff --git a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lcdc_encoder.c b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lcdc_encoder.c
index a06b064f86c1..4a645926edb7 100644
--- a/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lcdc_encoder.c
+++ b/drivers/gpu/drm/msm/mdp/mdp4/mdp4_lcdc_encoder.c
@@ -16,10 +16,10 @@
  * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
-#include "mdp4_kms.h"
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
 
-#include "drm_crtc.h"
-#include "drm_crtc_helper.h"
+#include "mdp4_kms.h"
 
 struct mdp4_lcdc_encoder {
 	struct drm_encoder base;
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_cmd_encoder.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_cmd_encoder.c
index 8dafc7bdba48..aa7402e03f67 100644
--- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_cmd_encoder.c
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_cmd_encoder.c
@@ -11,10 +11,10 @@
  * GNU General Public License for more details.
  */
 
-#include "mdp5_kms.h"
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
 
-#include "drm_crtc.h"
-#include "drm_crtc_helper.h"
+#include "mdp5_kms.h"
 
 static struct mdp5_kms *get_kms(struct drm_encoder *encoder)
 {
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c
index 9217e0d6e93e..0764a6498110 100644
--- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_crtc.c
@@ -16,13 +16,13 @@
  * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
-#include "mdp5_kms.h"
-
 #include <linux/sort.h>
 #include <drm/drm_mode.h>
-#include "drm_crtc.h"
-#include "drm_crtc_helper.h"
-#include "drm_flip_work.h"
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_flip_work.h>
+
+#include "mdp5_kms.h"
 
 #define CURSOR_WIDTH	64
 #define CURSOR_HEIGHT	64
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_encoder.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_encoder.c
index c2ab0f033031..97f3294fbfc6 100644
--- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_encoder.c
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_encoder.c
@@ -16,10 +16,10 @@
  * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
-#include "mdp5_kms.h"
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
 
-#include "drm_crtc.h"
-#include "drm_crtc_helper.h"
+#include "mdp5_kms.h"
 
 static struct mdp5_kms *get_kms(struct drm_encoder *encoder)
 {
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c
index d3d6b4cae1e6..e2b3346ead48 100644
--- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_kms.c
@@ -527,31 +527,28 @@ static struct drm_encoder *get_encoder_from_crtc(struct drm_crtc *crtc)
 	return NULL;
 }
 
-static int mdp5_get_scanoutpos(struct drm_device *dev, unsigned int pipe,
-			       unsigned int flags, int *vpos, int *hpos,
-			       ktime_t *stime, ktime_t *etime,
-			       const struct drm_display_mode *mode)
+static bool mdp5_get_scanoutpos(struct drm_device *dev, unsigned int pipe,
+				bool in_vblank_irq, int *vpos, int *hpos,
+				ktime_t *stime, ktime_t *etime,
+				const struct drm_display_mode *mode)
 {
 	struct msm_drm_private *priv = dev->dev_private;
 	struct drm_crtc *crtc;
 	struct drm_encoder *encoder;
 	int line, vsw, vbp, vactive_start, vactive_end, vfp_end;
-	int ret = 0;
 
 	crtc = priv->crtcs[pipe];
 	if (!crtc) {
 		DRM_ERROR("Invalid crtc %d\n", pipe);
-		return 0;
+		return false;
 	}
 
 	encoder = get_encoder_from_crtc(crtc);
 	if (!encoder) {
 		DRM_ERROR("no encoder found for crtc %d\n", pipe);
-		return 0;
+		return false;
 	}
 
-	ret |= DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_ACCURATE;
-
 	vsw = mode->crtc_vsync_end - mode->crtc_vsync_start;
 	vbp = mode->crtc_vtotal - mode->crtc_vsync_end;
 
@@ -575,10 +572,8 @@ static int mdp5_get_scanoutpos(struct drm_device *dev, unsigned int pipe,
 
 	if (line < vactive_start) {
 		line -= vactive_start;
-		ret |= DRM_SCANOUTPOS_IN_VBLANK;
 	} else if (line > vactive_end) {
 		line = line - vfp_end - vactive_start;
-		ret |= DRM_SCANOUTPOS_IN_VBLANK;
 	} else {
 		line -= vactive_start;
 	}
@@ -589,31 +584,7 @@ static int mdp5_get_scanoutpos(struct drm_device *dev, unsigned int pipe,
 	if (etime)
 		*etime = ktime_get();
 
-	return ret;
-}
-
-static int mdp5_get_vblank_timestamp(struct drm_device *dev, unsigned int pipe,
-				     int *max_error,
-				     struct timeval *vblank_time,
-				     unsigned flags)
-{
-	struct msm_drm_private *priv = dev->dev_private;
-	struct drm_crtc *crtc;
-
-	if (pipe < 0 || pipe >= priv->num_crtcs) {
-		DRM_ERROR("Invalid crtc %d\n", pipe);
-		return -EINVAL;
-	}
-
-	crtc = priv->crtcs[pipe];
-	if (!crtc) {
-		DRM_ERROR("Invalid crtc %d\n", pipe);
-		return -EINVAL;
-	}
-
-	return drm_calc_vbltimestamp_from_scanoutpos(dev, pipe, max_error,
-						     vblank_time, flags,
-						     &crtc->mode);
+	return true;
 }
 
 static u32 mdp5_get_vblank_counter(struct drm_device *dev, unsigned int pipe)
@@ -725,7 +696,7 @@ struct msm_kms *mdp5_kms_init(struct drm_device *dev)
 	dev->mode_config.max_width = 0xffff;
 	dev->mode_config.max_height = 0xffff;
 
-	dev->driver->get_vblank_timestamp = mdp5_get_vblank_timestamp;
+	dev->driver->get_vblank_timestamp = drm_calc_vbltimestamp_from_scanoutpos;
 	dev->driver->get_scanout_position = mdp5_get_scanoutpos;
 	dev->driver->get_vblank_counter = mdp5_get_vblank_counter;
 	dev->max_vblank_count = 0xffffffff;
diff --git a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c
index 7d3741215387..abaaac7fea1a 100644
--- a/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c
+++ b/drivers/gpu/drm/msm/mdp/mdp5/mdp5_plane.c
@@ -67,11 +67,11 @@ static void mdp5_plane_install_rotation_property(struct drm_device *dev,
 		struct drm_plane *plane)
 {
 	drm_plane_create_rotation_property(plane,
-					   DRM_ROTATE_0,
-					   DRM_ROTATE_0 |
-					   DRM_ROTATE_180 |
-					   DRM_REFLECT_X |
-					   DRM_REFLECT_Y);
+					   DRM_MODE_ROTATE_0,
+					   DRM_MODE_ROTATE_0 |
+					   DRM_MODE_ROTATE_180 |
+					   DRM_MODE_REFLECT_X |
+					   DRM_MODE_REFLECT_Y);
 }
 
 /* helper to install properties which are common to planes and crtcs */
@@ -370,14 +370,14 @@ static int mdp5_plane_atomic_check_with_state(struct drm_crtc_state *crtc_state,
 			caps |= MDP_PIPE_CAP_SCALE;
 
 		rotation = drm_rotation_simplify(state->rotation,
-						 DRM_ROTATE_0 |
-						 DRM_REFLECT_X |
-						 DRM_REFLECT_Y);
+						 DRM_MODE_ROTATE_0 |
+						 DRM_MODE_REFLECT_X |
+						 DRM_MODE_REFLECT_Y);
 
-		if (rotation & DRM_REFLECT_X)
+		if (rotation & DRM_MODE_REFLECT_X)
 			caps |= MDP_PIPE_CAP_HFLIP;
 
-		if (rotation & DRM_REFLECT_Y)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			caps |= MDP_PIPE_CAP_VFLIP;
 
 		if (plane->type == DRM_PLANE_TYPE_CURSOR)
@@ -975,11 +975,11 @@ static int mdp5_plane_mode_set(struct drm_plane *plane,
 	DBG("scale config = %x", config);
 
 	rotation = drm_rotation_simplify(pstate->rotation,
-					 DRM_ROTATE_0 |
-					 DRM_REFLECT_X |
-					 DRM_REFLECT_Y);
-	hflip = !!(rotation & DRM_REFLECT_X);
-	vflip = !!(rotation & DRM_REFLECT_Y);
+					 DRM_MODE_ROTATE_0 |
+					 DRM_MODE_REFLECT_X |
+					 DRM_MODE_REFLECT_Y);
+	hflip = !!(rotation & DRM_MODE_REFLECT_X);
+	vflip = !!(rotation & DRM_MODE_REFLECT_Y);
 
 	spin_lock_irqsave(&mdp5_plane->pipe_lock, flags);
 
diff --git a/drivers/gpu/drm/msm/msm_fb.c b/drivers/gpu/drm/msm/msm_fb.c
index 5cf165c9c3a9..ba2733a95a4f 100644
--- a/drivers/gpu/drm/msm/msm_fb.c
+++ b/drivers/gpu/drm/msm/msm_fb.c
@@ -15,12 +15,12 @@
  * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
+
 #include "msm_drv.h"
 #include "msm_kms.h"
 
-#include "drm_crtc.h"
-#include "drm_crtc_helper.h"
-
 struct msm_framebuffer {
 	struct drm_framebuffer base;
 	const struct msm_format *format;
diff --git a/drivers/gpu/drm/msm/msm_fbdev.c b/drivers/gpu/drm/msm/msm_fbdev.c
index 951e40faf6e8..feea8ba4e05b 100644
--- a/drivers/gpu/drm/msm/msm_fbdev.c
+++ b/drivers/gpu/drm/msm/msm_fbdev.c
@@ -15,10 +15,10 @@
  * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
-#include "msm_drv.h"
+#include <drm/drm_crtc.h>
+#include <drm/drm_fb_helper.h>
 
-#include "drm_crtc.h"
-#include "drm_fb_helper.h"
+#include "msm_drv.h"
 #include "msm_gem.h"
 
 extern int msm_gem_mmap_obj(struct drm_gem_object *obj,
diff --git a/drivers/gpu/drm/msm/msm_gem.c b/drivers/gpu/drm/msm/msm_gem.c
index 50289a23baf8..be77a35a7a8e 100644
--- a/drivers/gpu/drm/msm/msm_gem.c
+++ b/drivers/gpu/drm/msm/msm_gem.c
@@ -50,13 +50,13 @@ static struct page **get_pages_vram(struct drm_gem_object *obj,
 	struct page **p;
 	int ret, i;
 
-	p = drm_malloc_ab(npages, sizeof(struct page *));
+	p = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL);
 	if (!p)
 		return ERR_PTR(-ENOMEM);
 
 	ret = drm_mm_insert_node(&priv->vram.mm, msm_obj->vram_node, npages);
 	if (ret) {
-		drm_free_large(p);
+		kvfree(p);
 		return ERR_PTR(ret);
 	}
 
@@ -127,7 +127,7 @@ static void put_pages(struct drm_gem_object *obj)
 			drm_gem_put_pages(obj, msm_obj->pages, true, false);
 		else {
 			drm_mm_remove_node(msm_obj->vram_node);
-			drm_free_large(msm_obj->pages);
+			kvfree(msm_obj->pages);
 		}
 
 		msm_obj->pages = NULL;
@@ -707,7 +707,7 @@ void msm_gem_free_object(struct drm_gem_object *obj)
 		 * ours, just free the array we allocated:
 		 */
 		if (msm_obj->pages)
-			drm_free_large(msm_obj->pages);
+			kvfree(msm_obj->pages);
 
 		drm_prime_gem_destroy(obj, msm_obj->sgt);
 	} else {
@@ -869,7 +869,7 @@ struct drm_gem_object *msm_gem_import(struct drm_device *dev,
 
 	msm_obj = to_msm_bo(obj);
 	msm_obj->sgt = sgt;
-	msm_obj->pages = drm_malloc_ab(npages, sizeof(struct page *));
+	msm_obj->pages = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL);
 	if (!msm_obj->pages) {
 		ret = -ENOMEM;
 		goto fail;
diff --git a/drivers/gpu/drm/nouveau/Kbuild b/drivers/gpu/drm/nouveau/Kbuild
index fde6e3656636..2e9ce53ae3a8 100644
--- a/drivers/gpu/drm/nouveau/Kbuild
+++ b/drivers/gpu/drm/nouveau/Kbuild
@@ -1,4 +1,3 @@
-ccflags-y := -Iinclude/drm
 ccflags-y += -I$(src)/include
 ccflags-y += -I$(src)/include/nvkm
 ccflags-y += -I$(src)/nvkm
diff --git a/drivers/gpu/drm/nouveau/nouveau_display.c b/drivers/gpu/drm/nouveau/nouveau_display.c
index 549763f5e17d..8d1df5678eaa 100644
--- a/drivers/gpu/drm/nouveau/nouveau_display.c
+++ b/drivers/gpu/drm/nouveau/nouveau_display.c
@@ -98,7 +98,7 @@ calc(int blanks, int blanke, int total, int line)
 	return line;
 }
 
-static int
+static bool
 nouveau_display_scanoutpos_head(struct drm_crtc *crtc, int *vpos, int *hpos,
 				ktime_t *stime, ktime_t *etime)
 {
@@ -111,16 +111,16 @@ nouveau_display_scanoutpos_head(struct drm_crtc *crtc, int *vpos, int *hpos,
 	};
 	struct nouveau_display *disp = nouveau_display(crtc->dev);
 	struct drm_vblank_crtc *vblank = &crtc->dev->vblank[drm_crtc_index(crtc)];
-	int ret, retry = 20;
+	int retry = 20;
+	bool ret = false;
 
 	do {
 		ret = nvif_mthd(&disp->disp, 0, &args, sizeof(args));
 		if (ret != 0)
-			return 0;
+			return false;
 
 		if (args.scan.vline) {
-			ret |= DRM_SCANOUTPOS_ACCURATE;
-			ret |= DRM_SCANOUTPOS_VALID;
+			ret = true;
 			break;
 		}
 
@@ -133,14 +133,12 @@ nouveau_display_scanoutpos_head(struct drm_crtc *crtc, int *vpos, int *hpos,
 	if (stime) *stime = ns_to_ktime(args.scan.time[0]);
 	if (etime) *etime = ns_to_ktime(args.scan.time[1]);
 
-	if (*vpos < 0)
-		ret |= DRM_SCANOUTPOS_IN_VBLANK;
 	return ret;
 }
 
-int
+bool
 nouveau_display_scanoutpos(struct drm_device *dev, unsigned int pipe,
-			   unsigned int flags, int *vpos, int *hpos,
+			   bool in_vblank_irq, int *vpos, int *hpos,
 			   ktime_t *stime, ktime_t *etime,
 			   const struct drm_display_mode *mode)
 {
@@ -153,28 +151,7 @@ nouveau_display_scanoutpos(struct drm_device *dev, unsigned int pipe,
 		}
 	}
 
-	return 0;
-}
-
-int
-nouveau_display_vblstamp(struct drm_device *dev, unsigned int pipe,
-			 int *max_error, struct timeval *time, unsigned flags)
-{
-	struct drm_crtc *crtc;
-
-	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
-		if (nouveau_crtc(crtc)->index == pipe) {
-			struct drm_display_mode *mode;
-			if (drm_drv_uses_atomic_modeset(dev))
-				mode = &crtc->state->adjusted_mode;
-			else
-				mode = &crtc->hwmode;
-			return drm_calc_vbltimestamp_from_scanoutpos(dev,
-					pipe, max_error, time, flags, mode);
-		}
-	}
-
-	return -EINVAL;
+	return false;
 }
 
 static void
diff --git a/drivers/gpu/drm/nouveau/nouveau_display.h b/drivers/gpu/drm/nouveau/nouveau_display.h
index e1d772d39488..201aec2ea5b8 100644
--- a/drivers/gpu/drm/nouveau/nouveau_display.h
+++ b/drivers/gpu/drm/nouveau/nouveau_display.h
@@ -68,11 +68,9 @@ int  nouveau_display_suspend(struct drm_device *dev, bool runtime);
 void nouveau_display_resume(struct drm_device *dev, bool runtime);
 int  nouveau_display_vblank_enable(struct drm_device *, unsigned int);
 void nouveau_display_vblank_disable(struct drm_device *, unsigned int);
-int  nouveau_display_scanoutpos(struct drm_device *, unsigned int,
-				unsigned int, int *, int *, ktime_t *,
-				ktime_t *, const struct drm_display_mode *);
-int  nouveau_display_vblstamp(struct drm_device *, unsigned int, int *,
-			      struct timeval *, unsigned);
+bool  nouveau_display_scanoutpos(struct drm_device *, unsigned int,
+				 bool, int *, int *, ktime_t *,
+				 ktime_t *, const struct drm_display_mode *);
 
 int  nouveau_crtc_page_flip(struct drm_crtc *crtc, struct drm_framebuffer *fb,
 			    struct drm_pending_vblank_event *event,
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c
index 15a13d09d431..7acd1cf8c5a6 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.c
@@ -29,8 +29,8 @@
 #include <linux/pm_runtime.h>
 #include <linux/vga_switcheroo.h>
 
-#include "drmP.h"
-#include "drm_crtc_helper.h"
+#include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
 
 #include <core/gpuobj.h>
 #include <core/option.h>
@@ -877,7 +877,7 @@ done:
 }
 
 static void
-nouveau_drm_preclose(struct drm_device *dev, struct drm_file *fpriv)
+nouveau_drm_postclose(struct drm_device *dev, struct drm_file *fpriv)
 {
 	struct nouveau_cli *cli = nouveau_cli(fpriv);
 	struct nouveau_drm *drm = nouveau_drm(dev);
@@ -893,12 +893,6 @@ nouveau_drm_preclose(struct drm_device *dev, struct drm_file *fpriv)
 	list_del(&cli->head);
 	mutex_unlock(&drm->client.mutex);
 
-}
-
-static void
-nouveau_drm_postclose(struct drm_device *dev, struct drm_file *fpriv)
-{
-	struct nouveau_cli *cli = nouveau_cli(fpriv);
 	nouveau_cli_fini(cli);
 	kfree(cli);
 	pm_runtime_mark_last_busy(dev->dev);
@@ -970,7 +964,6 @@ driver_stub = {
 	.load = nouveau_drm_load,
 	.unload = nouveau_drm_unload,
 	.open = nouveau_drm_open,
-	.preclose = nouveau_drm_preclose,
 	.postclose = nouveau_drm_postclose,
 	.lastclose = nouveau_vga_lastclose,
 
@@ -981,7 +974,7 @@ driver_stub = {
 	.enable_vblank = nouveau_display_vblank_enable,
 	.disable_vblank = nouveau_display_vblank_disable,
 	.get_scanout_position = nouveau_display_scanoutpos,
-	.get_vblank_timestamp = nouveau_display_vblstamp,
+	.get_vblank_timestamp = drm_calc_vbltimestamp_from_scanoutpos,
 
 	.ioctls = nouveau_ioctls,
 	.num_ioctls = ARRAY_SIZE(nouveau_ioctls),
diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.h b/drivers/gpu/drm/nouveau/nouveau_drv.h
index a11b6aaed325..822fe1d4d35e 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drv.h
+++ b/drivers/gpu/drm/nouveau/nouveau_drv.h
@@ -43,7 +43,7 @@
 #include <nvif/device.h>
 #include <nvif/ioctl.h>
 
-#include <drmP.h>
+#include <drm/drmP.h>
 
 #include <drm/ttm/ttm_bo_api.h>
 #include <drm/ttm/ttm_bo_driver.h>
diff --git a/drivers/gpu/drm/nouveau/nouveau_ttm.c b/drivers/gpu/drm/nouveau/nouveau_ttm.c
index 13e5cc5f07fe..999c35a25498 100644
--- a/drivers/gpu/drm/nouveau/nouveau_ttm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_ttm.c
@@ -28,7 +28,7 @@
 #include "nouveau_ttm.h"
 #include "nouveau_gem.h"
 
-#include "drm_legacy.h"
+#include <drm/drm_legacy.h>
 
 #include <core/tegra.h>
 
diff --git a/drivers/gpu/drm/nouveau/nv50_display.c b/drivers/gpu/drm/nouveau/nv50_display.c
index 06e564a9ccb2..56e2eaaee6e6 100644
--- a/drivers/gpu/drm/nouveau/nv50_display.c
+++ b/drivers/gpu/drm/nouveau/nv50_display.c
@@ -1033,7 +1033,7 @@ nv50_wndw_reset(struct drm_plane *plane)
 		plane->funcs->atomic_destroy_state(plane, plane->state);
 	plane->state = &asyw->state;
 	plane->state->plane = plane;
-	plane->state->rotation = DRM_ROTATE_0;
+	plane->state->rotation = DRM_MODE_ROTATE_0;
 }
 
 static void
diff --git a/drivers/gpu/drm/omapdrm/Makefile b/drivers/gpu/drm/omapdrm/Makefile
index 48b7b750c05c..b391be7ecb6c 100644
--- a/drivers/gpu/drm/omapdrm/Makefile
+++ b/drivers/gpu/drm/omapdrm/Makefile
@@ -6,7 +6,6 @@
 obj-y += dss/
 obj-y += displays/
 
-ccflags-y := -Iinclude/drm
 omapdrm-y := omap_drv.o \
 	omap_irq.o \
 	omap_debugfs.o \
diff --git a/drivers/gpu/drm/omapdrm/omap_drv.c b/drivers/gpu/drm/omapdrm/omap_drv.c
index e1f47f0b3ccf..663e930a7b0f 100644
--- a/drivers/gpu/drm/omapdrm/omap_drv.c
+++ b/drivers/gpu/drm/omapdrm/omap_drv.c
@@ -577,7 +577,7 @@ static void dev_lastclose(struct drm_device *dev)
 
 		drm_object_property_set_value(&crtc->base,
 					      crtc->primary->rotation_property,
-					      DRM_ROTATE_0);
+					      DRM_MODE_ROTATE_0);
 	}
 
 	for (i = 0; i < priv->num_planes; i++) {
@@ -588,7 +588,7 @@ static void dev_lastclose(struct drm_device *dev)
 
 		drm_object_property_set_value(&plane->base,
 					      plane->rotation_property,
-					      DRM_ROTATE_0);
+					      DRM_MODE_ROTATE_0);
 	}
 
 	if (priv->fbdev) {
diff --git a/drivers/gpu/drm/omapdrm/omap_fb.c b/drivers/gpu/drm/omapdrm/omap_fb.c
index 29dc677dd4d3..5ca0537bb427 100644
--- a/drivers/gpu/drm/omapdrm/omap_fb.c
+++ b/drivers/gpu/drm/omapdrm/omap_fb.c
@@ -167,30 +167,30 @@ void omap_framebuffer_update_scanout(struct drm_framebuffer *fb,
 		uint32_t w = win->src_w;
 		uint32_t h = win->src_h;
 
-		switch (win->rotation & DRM_ROTATE_MASK) {
+		switch (win->rotation & DRM_MODE_ROTATE_MASK) {
 		default:
 			dev_err(fb->dev->dev, "invalid rotation: %02x",
 					(uint32_t)win->rotation);
 			/* fallthru to default to no rotation */
 		case 0:
-		case DRM_ROTATE_0:
+		case DRM_MODE_ROTATE_0:
 			orient = 0;
 			break;
-		case DRM_ROTATE_90:
+		case DRM_MODE_ROTATE_90:
 			orient = MASK_XY_FLIP | MASK_X_INVERT;
 			break;
-		case DRM_ROTATE_180:
+		case DRM_MODE_ROTATE_180:
 			orient = MASK_X_INVERT | MASK_Y_INVERT;
 			break;
-		case DRM_ROTATE_270:
+		case DRM_MODE_ROTATE_270:
 			orient = MASK_XY_FLIP | MASK_Y_INVERT;
 			break;
 		}
 
-		if (win->rotation & DRM_REFLECT_X)
+		if (win->rotation & DRM_MODE_REFLECT_X)
 			orient ^= MASK_X_INVERT;
 
-		if (win->rotation & DRM_REFLECT_Y)
+		if (win->rotation & DRM_MODE_REFLECT_Y)
 			orient ^= MASK_Y_INVERT;
 
 		/* adjust x,y offset for flip/invert: */
@@ -205,9 +205,9 @@ void omap_framebuffer_update_scanout(struct drm_framebuffer *fb,
 		info->rotation_type = OMAP_DSS_ROT_TILER;
 		info->screen_width  = omap_gem_tiled_stride(plane->bo, orient);
 	} else {
-		switch (win->rotation & DRM_ROTATE_MASK) {
+		switch (win->rotation & DRM_MODE_ROTATE_MASK) {
 		case 0:
-		case DRM_ROTATE_0:
+		case DRM_MODE_ROTATE_0:
 			/* OK */
 			break;
 
diff --git a/drivers/gpu/drm/omapdrm/omap_plane.c b/drivers/gpu/drm/omapdrm/omap_plane.c
index 9168154d749e..d3d6818c68f8 100644
--- a/drivers/gpu/drm/omapdrm/omap_plane.c
+++ b/drivers/gpu/drm/omapdrm/omap_plane.c
@@ -141,7 +141,7 @@ static void omap_plane_atomic_disable(struct drm_plane *plane,
 	struct omap_plane_state *omap_state = to_omap_plane_state(plane->state);
 	struct omap_plane *omap_plane = to_omap_plane(plane);
 
-	plane->state->rotation = DRM_ROTATE_0;
+	plane->state->rotation = DRM_MODE_ROTATE_0;
 	omap_state->zorder = plane->type == DRM_PLANE_TYPE_PRIMARY
 			   ? 0 : omap_plane->id;
 
@@ -177,7 +177,7 @@ static int omap_plane_atomic_check(struct drm_plane *plane,
 	if (state->crtc_y + state->crtc_h > crtc_state->adjusted_mode.vdisplay)
 		return -EINVAL;
 
-	if (state->rotation != DRM_ROTATE_0 &&
+	if (state->rotation != DRM_MODE_ROTATE_0 &&
 	    !omap_framebuffer_supports_rotation(state->fb))
 		return -EINVAL;
 
@@ -213,15 +213,15 @@ void omap_plane_install_properties(struct drm_plane *plane,
 	if (priv->has_dmm) {
 		if (!plane->rotation_property)
 			drm_plane_create_rotation_property(plane,
-							   DRM_ROTATE_0,
-							   DRM_ROTATE_0 | DRM_ROTATE_90 |
-							   DRM_ROTATE_180 | DRM_ROTATE_270 |
-							   DRM_REFLECT_X | DRM_REFLECT_Y);
+							   DRM_MODE_ROTATE_0,
+							   DRM_MODE_ROTATE_0 | DRM_MODE_ROTATE_90 |
+							   DRM_MODE_ROTATE_180 | DRM_MODE_ROTATE_270 |
+							   DRM_MODE_REFLECT_X | DRM_MODE_REFLECT_Y);
 
 		/* Attach the rotation property also to the crtc object */
 		if (plane->rotation_property && obj != &plane->base)
 			drm_object_attach_property(obj, plane->rotation_property,
-						   DRM_ROTATE_0);
+						   DRM_MODE_ROTATE_0);
 	}
 
 	drm_object_attach_property(obj, priv->zorder_prop, 0);
@@ -273,7 +273,7 @@ static void omap_plane_reset(struct drm_plane *plane)
 	 */
 	omap_state->zorder = plane->type == DRM_PLANE_TYPE_PRIMARY
 			   ? 0 : omap_plane->id;
-	omap_state->base.rotation = DRM_ROTATE_0;
+	omap_state->base.rotation = DRM_MODE_ROTATE_0;
 
 	plane->state = &omap_state->base;
 	plane->state->plane = plane;
diff --git a/drivers/gpu/drm/pl111/Kconfig b/drivers/gpu/drm/pl111/Kconfig
new file mode 100644
index 000000000000..309f4fd52de7
--- /dev/null
+++ b/drivers/gpu/drm/pl111/Kconfig
@@ -0,0 +1,13 @@
+config DRM_PL111
+	tristate "DRM Support for PL111 CLCD Controller"
+	depends on DRM
+	depends on ARM || ARM64 || COMPILE_TEST
+	depends on COMMON_CLK
+	select DRM_KMS_HELPER
+	select DRM_KMS_CMA_HELPER
+	select DRM_GEM_CMA_HELPER
+	select VT_HW_CONSOLE_BINDING if FRAMEBUFFER_CONSOLE
+	help
+	  Choose this option for DRM support for the PL111 CLCD controller.
+	  If M is selected the module will be called pl111_drm.
+
diff --git a/drivers/gpu/drm/pl111/Makefile b/drivers/gpu/drm/pl111/Makefile
new file mode 100644
index 000000000000..59483d610ef5
--- /dev/null
+++ b/drivers/gpu/drm/pl111/Makefile
@@ -0,0 +1,7 @@
+pl111_drm-y +=	pl111_connector.o \
+		pl111_display.o \
+		pl111_drv.o
+
+pl111_drm-$(CONFIG_DEBUG_FS) += pl111_debugfs.o
+
+obj-$(CONFIG_DRM_PL111) += pl111_drm.o
diff --git a/drivers/gpu/drm/pl111/pl111_connector.c b/drivers/gpu/drm/pl111/pl111_connector.c
new file mode 100644
index 000000000000..3f213d7e7692
--- /dev/null
+++ b/drivers/gpu/drm/pl111/pl111_connector.c
@@ -0,0 +1,127 @@
+/*
+ * (C) COPYRIGHT 2012-2013 ARM Limited. All rights reserved.
+ *
+ * Parts of this file were based on sources as follows:
+ *
+ * Copyright (c) 2006-2008 Intel Corporation
+ * Copyright (c) 2007 Dave Airlie <airlied@linux.ie>
+ * Copyright (C) 2011 Texas Instruments
+ *
+ * This program is free software and is provided to you under the terms of the
+ * GNU General Public License version 2 as published by the Free Software
+ * Foundation, and any use by you of this program is subject to the terms of
+ * such GNU licence.
+ *
+ */
+
+/**
+ * pl111_drm_connector.c
+ * Implementation of the connector functions for PL111 DRM
+ */
+#include <linux/amba/clcd-regs.h>
+#include <linux/version.h>
+#include <linux/shmem_fs.h>
+#include <linux/dma-buf.h>
+
+#include <drm/drmP.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_of.h>
+#include <drm/drm_panel.h>
+
+#include "pl111_drm.h"
+
+static void pl111_connector_destroy(struct drm_connector *connector)
+{
+	struct pl111_drm_connector *pl111_connector =
+		to_pl111_connector(connector);
+
+	if (pl111_connector->panel)
+		drm_panel_detach(pl111_connector->panel);
+
+	drm_connector_unregister(connector);
+	drm_connector_cleanup(connector);
+}
+
+static enum drm_connector_status pl111_connector_detect(struct drm_connector
+							*connector, bool force)
+{
+	struct pl111_drm_connector *pl111_connector =
+		to_pl111_connector(connector);
+
+	return (pl111_connector->panel ?
+		connector_status_connected :
+		connector_status_disconnected);
+}
+
+static int pl111_connector_helper_get_modes(struct drm_connector *connector)
+{
+	struct pl111_drm_connector *pl111_connector =
+		to_pl111_connector(connector);
+
+	if (!pl111_connector->panel)
+		return 0;
+
+	return drm_panel_get_modes(pl111_connector->panel);
+}
+
+const struct drm_connector_funcs connector_funcs = {
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.destroy = pl111_connector_destroy,
+	.detect = pl111_connector_detect,
+	.dpms = drm_atomic_helper_connector_dpms,
+	.reset = drm_atomic_helper_connector_reset,
+	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
+};
+
+const struct drm_connector_helper_funcs connector_helper_funcs = {
+	.get_modes = pl111_connector_helper_get_modes,
+};
+
+/* Walks the OF graph to find the panel node and then asks DRM to look
+ * up the panel.
+ */
+static struct drm_panel *pl111_get_panel(struct device *dev)
+{
+	struct device_node *endpoint, *panel_node;
+	struct device_node *np = dev->of_node;
+	struct drm_panel *panel;
+
+	endpoint = of_graph_get_next_endpoint(np, NULL);
+	if (!endpoint) {
+		dev_err(dev, "no endpoint to fetch panel\n");
+		return NULL;
+	}
+
+	/* don't proceed if we have an endpoint but no panel_node tied to it */
+	panel_node = of_graph_get_remote_port_parent(endpoint);
+	of_node_put(endpoint);
+	if (!panel_node) {
+		dev_err(dev, "no valid panel node\n");
+		return NULL;
+	}
+
+	panel = of_drm_find_panel(panel_node);
+	of_node_put(panel_node);
+
+	return panel;
+}
+
+int pl111_connector_init(struct drm_device *dev)
+{
+	struct pl111_drm_dev_private *priv = dev->dev_private;
+	struct pl111_drm_connector *pl111_connector = &priv->connector;
+	struct drm_connector *connector = &pl111_connector->connector;
+
+	drm_connector_init(dev, connector, &connector_funcs,
+			   DRM_MODE_CONNECTOR_DPI);
+	drm_connector_helper_add(connector, &connector_helper_funcs);
+
+	pl111_connector->panel = pl111_get_panel(dev->dev);
+	if (pl111_connector->panel)
+		drm_panel_attach(pl111_connector->panel, connector);
+
+	return 0;
+}
+
diff --git a/drivers/gpu/drm/pl111/pl111_debugfs.c b/drivers/gpu/drm/pl111/pl111_debugfs.c
new file mode 100644
index 000000000000..0d9dee199b2c
--- /dev/null
+++ b/drivers/gpu/drm/pl111/pl111_debugfs.c
@@ -0,0 +1,55 @@
+/*
+ *  Copyright © 2017 Broadcom
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/amba/clcd-regs.h>
+#include <linux/seq_file.h>
+#include <drm/drm_debugfs.h>
+#include <drm/drmP.h>
+#include "pl111_drm.h"
+
+#define REGDEF(reg) { reg, #reg }
+static const struct {
+	u32 reg;
+	const char *name;
+} pl111_reg_defs[] = {
+	REGDEF(CLCD_TIM0),
+	REGDEF(CLCD_TIM1),
+	REGDEF(CLCD_TIM2),
+	REGDEF(CLCD_TIM3),
+	REGDEF(CLCD_UBAS),
+	REGDEF(CLCD_PL111_CNTL),
+	REGDEF(CLCD_PL111_IENB),
+};
+
+int pl111_debugfs_regs(struct seq_file *m, void *unused)
+{
+	struct drm_info_node *node = (struct drm_info_node *)m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct pl111_drm_dev_private *priv = dev->dev_private;
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(pl111_reg_defs); i++) {
+		seq_printf(m, "%s (0x%04x): 0x%08x\n",
+			   pl111_reg_defs[i].name, pl111_reg_defs[i].reg,
+			   readl(priv->regs + pl111_reg_defs[i].reg));
+	}
+
+	return 0;
+}
+
+static const struct drm_info_list pl111_debugfs_list[] = {
+	{"regs", pl111_debugfs_regs, 0},
+};
+
+int
+pl111_debugfs_init(struct drm_minor *minor)
+{
+	return drm_debugfs_create_files(pl111_debugfs_list,
+					ARRAY_SIZE(pl111_debugfs_list),
+					minor->debugfs_root, minor);
+}
diff --git a/drivers/gpu/drm/pl111/pl111_display.c b/drivers/gpu/drm/pl111/pl111_display.c
new file mode 100644
index 000000000000..3e0a4fa73ddb
--- /dev/null
+++ b/drivers/gpu/drm/pl111/pl111_display.c
@@ -0,0 +1,476 @@
+/*
+ * (C) COPYRIGHT 2012-2013 ARM Limited. All rights reserved.
+ *
+ * Parts of this file were based on sources as follows:
+ *
+ * Copyright (c) 2006-2008 Intel Corporation
+ * Copyright (c) 2007 Dave Airlie <airlied@linux.ie>
+ * Copyright (C) 2011 Texas Instruments
+ *
+ * This program is free software and is provided to you under the terms of the
+ * GNU General Public License version 2 as published by the Free Software
+ * Foundation, and any use by you of this program is subject to the terms of
+ * such GNU licence.
+ *
+ */
+
+#include <linux/amba/clcd-regs.h>
+#include <linux/clk.h>
+#include <linux/version.h>
+#include <linux/dma-buf.h>
+#include <linux/of_graph.h>
+
+#include <drm/drmP.h>
+#include <drm/drm_panel.h>
+#include <drm/drm_gem_cma_helper.h>
+#include <drm/drm_fb_cma_helper.h>
+
+#include "pl111_drm.h"
+
+irqreturn_t pl111_irq(int irq, void *data)
+{
+	struct pl111_drm_dev_private *priv = data;
+	u32 irq_stat;
+	irqreturn_t status = IRQ_NONE;
+
+	irq_stat = readl(priv->regs + CLCD_PL111_MIS);
+
+	if (!irq_stat)
+		return IRQ_NONE;
+
+	if (irq_stat & CLCD_IRQ_NEXTBASE_UPDATE) {
+		drm_crtc_handle_vblank(&priv->pipe.crtc);
+
+		status = IRQ_HANDLED;
+	}
+
+	/* Clear the interrupt once done */
+	writel(irq_stat, priv->regs + CLCD_PL111_ICR);
+
+	return status;
+}
+
+static u32 pl111_get_fb_offset(struct drm_plane_state *pstate)
+{
+	struct drm_framebuffer *fb = pstate->fb;
+	struct drm_gem_cma_object *obj = drm_fb_cma_get_gem_obj(fb, 0);
+
+	return (obj->paddr +
+		fb->offsets[0] +
+		fb->format->cpp[0] * pstate->src_x +
+		fb->pitches[0] * pstate->src_y);
+}
+
+static int pl111_display_check(struct drm_simple_display_pipe *pipe,
+			       struct drm_plane_state *pstate,
+			       struct drm_crtc_state *cstate)
+{
+	const struct drm_display_mode *mode = &cstate->mode;
+	struct drm_framebuffer *old_fb = pipe->plane.state->fb;
+	struct drm_framebuffer *fb = pstate->fb;
+
+	if (mode->hdisplay % 16)
+		return -EINVAL;
+
+	if (fb) {
+		u32 offset = pl111_get_fb_offset(pstate);
+
+		/* FB base address must be dword aligned. */
+		if (offset & 3)
+			return -EINVAL;
+
+		/* There's no pitch register -- the mode's hdisplay
+		 * controls it.
+		 */
+		if (fb->pitches[0] != mode->hdisplay * fb->format->cpp[0])
+			return -EINVAL;
+
+		/* We can't change the FB format in a flicker-free
+		 * manner (and only update it during CRTC enable).
+		 */
+		if (old_fb && old_fb->format != fb->format)
+			cstate->mode_changed = true;
+	}
+
+	return 0;
+}
+
+static void pl111_display_enable(struct drm_simple_display_pipe *pipe,
+				 struct drm_crtc_state *cstate)
+{
+	struct drm_crtc *crtc = &pipe->crtc;
+	struct drm_plane *plane = &pipe->plane;
+	struct drm_device *drm = crtc->dev;
+	struct pl111_drm_dev_private *priv = drm->dev_private;
+	const struct drm_display_mode *mode = &cstate->mode;
+	struct drm_framebuffer *fb = plane->state->fb;
+	struct drm_connector *connector = &priv->connector.connector;
+	u32 cntl;
+	u32 ppl, hsw, hfp, hbp;
+	u32 lpp, vsw, vfp, vbp;
+	u32 cpl, tim2;
+	int ret;
+
+	ret = clk_set_rate(priv->clk, mode->clock * 1000);
+	if (ret) {
+		dev_err(drm->dev,
+			"Failed to set pixel clock rate to %d: %d\n",
+			mode->clock * 1000, ret);
+	}
+
+	clk_prepare_enable(priv->clk);
+
+	ppl = (mode->hdisplay / 16) - 1;
+	hsw = mode->hsync_end - mode->hsync_start - 1;
+	hfp = mode->hsync_start - mode->hdisplay - 1;
+	hbp = mode->htotal - mode->hsync_end - 1;
+
+	lpp = mode->vdisplay - 1;
+	vsw = mode->vsync_end - mode->vsync_start - 1;
+	vfp = mode->vsync_start - mode->vdisplay;
+	vbp = mode->vtotal - mode->vsync_end;
+
+	cpl = mode->hdisplay - 1;
+
+	writel((ppl << 2) |
+	       (hsw << 8) |
+	       (hfp << 16) |
+	       (hbp << 24),
+	       priv->regs + CLCD_TIM0);
+	writel(lpp |
+	       (vsw << 10) |
+	       (vfp << 16) |
+	       (vbp << 24),
+	       priv->regs + CLCD_TIM1);
+
+	spin_lock(&priv->tim2_lock);
+
+	tim2 = readl(priv->regs + CLCD_TIM2);
+	tim2 &= (TIM2_BCD | TIM2_PCD_LO_MASK | TIM2_PCD_HI_MASK);
+
+	if (mode->flags & DRM_MODE_FLAG_NHSYNC)
+		tim2 |= TIM2_IHS;
+
+	if (mode->flags & DRM_MODE_FLAG_NVSYNC)
+		tim2 |= TIM2_IVS;
+
+	if (connector->display_info.bus_flags & DRM_BUS_FLAG_DE_LOW)
+		tim2 |= TIM2_IOE;
+
+	if (connector->display_info.bus_flags & DRM_BUS_FLAG_PIXDATA_NEGEDGE)
+		tim2 |= TIM2_IPC;
+
+	tim2 |= cpl << 16;
+	writel(tim2, priv->regs + CLCD_TIM2);
+	spin_unlock(&priv->tim2_lock);
+
+	writel(0, priv->regs + CLCD_TIM3);
+
+	drm_panel_prepare(priv->connector.panel);
+
+	/* Enable and Power Up */
+	cntl = CNTL_LCDEN | CNTL_LCDTFT | CNTL_LCDPWR | CNTL_LCDVCOMP(1);
+
+	/* Note that the the hardware's format reader takes 'r' from
+	 * the low bit, while DRM formats list channels from high bit
+	 * to low bit as you read left to right.
+	 */
+	switch (fb->format->format) {
+	case DRM_FORMAT_ABGR8888:
+	case DRM_FORMAT_XBGR8888:
+		cntl |= CNTL_LCDBPP24;
+		break;
+	case DRM_FORMAT_ARGB8888:
+	case DRM_FORMAT_XRGB8888:
+		cntl |= CNTL_LCDBPP24 | CNTL_BGR;
+		break;
+	case DRM_FORMAT_BGR565:
+		cntl |= CNTL_LCDBPP16_565;
+		break;
+	case DRM_FORMAT_RGB565:
+		cntl |= CNTL_LCDBPP16_565 | CNTL_BGR;
+		break;
+	case DRM_FORMAT_ABGR1555:
+	case DRM_FORMAT_XBGR1555:
+		cntl |= CNTL_LCDBPP16;
+		break;
+	case DRM_FORMAT_ARGB1555:
+	case DRM_FORMAT_XRGB1555:
+		cntl |= CNTL_LCDBPP16 | CNTL_BGR;
+		break;
+	case DRM_FORMAT_ABGR4444:
+	case DRM_FORMAT_XBGR4444:
+		cntl |= CNTL_LCDBPP16_444;
+		break;
+	case DRM_FORMAT_ARGB4444:
+	case DRM_FORMAT_XRGB4444:
+		cntl |= CNTL_LCDBPP16_444 | CNTL_BGR;
+		break;
+	default:
+		WARN_ONCE(true, "Unknown FB format 0x%08x\n",
+			  fb->format->format);
+		break;
+	}
+
+	writel(cntl, priv->regs + CLCD_PL111_CNTL);
+
+	drm_panel_enable(priv->connector.panel);
+
+	drm_crtc_vblank_on(crtc);
+}
+
+void pl111_display_disable(struct drm_simple_display_pipe *pipe)
+{
+	struct drm_crtc *crtc = &pipe->crtc;
+	struct drm_device *drm = crtc->dev;
+	struct pl111_drm_dev_private *priv = drm->dev_private;
+
+	drm_crtc_vblank_off(crtc);
+
+	drm_panel_disable(priv->connector.panel);
+
+	/* Disable and Power Down */
+	writel(0, priv->regs + CLCD_PL111_CNTL);
+
+	drm_panel_unprepare(priv->connector.panel);
+
+	clk_disable_unprepare(priv->clk);
+}
+
+static void pl111_display_update(struct drm_simple_display_pipe *pipe,
+				 struct drm_plane_state *old_pstate)
+{
+	struct drm_crtc *crtc = &pipe->crtc;
+	struct drm_device *drm = crtc->dev;
+	struct pl111_drm_dev_private *priv = drm->dev_private;
+	struct drm_pending_vblank_event *event = crtc->state->event;
+	struct drm_plane *plane = &pipe->plane;
+	struct drm_plane_state *pstate = plane->state;
+	struct drm_framebuffer *fb = pstate->fb;
+
+	if (fb) {
+		u32 addr = pl111_get_fb_offset(pstate);
+
+		writel(addr, priv->regs + CLCD_UBAS);
+	}
+
+	if (event) {
+		crtc->state->event = NULL;
+
+		spin_lock_irq(&crtc->dev->event_lock);
+		if (crtc->state->active && drm_crtc_vblank_get(crtc) == 0)
+			drm_crtc_arm_vblank_event(crtc, event);
+		else
+			drm_crtc_send_vblank_event(crtc, event);
+		spin_unlock_irq(&crtc->dev->event_lock);
+	}
+}
+
+int pl111_enable_vblank(struct drm_device *drm, unsigned int crtc)
+{
+	struct pl111_drm_dev_private *priv = drm->dev_private;
+
+	writel(CLCD_IRQ_NEXTBASE_UPDATE, priv->regs + CLCD_PL111_IENB);
+
+	return 0;
+}
+
+void pl111_disable_vblank(struct drm_device *drm, unsigned int crtc)
+{
+	struct pl111_drm_dev_private *priv = drm->dev_private;
+
+	writel(0, priv->regs + CLCD_PL111_IENB);
+}
+
+static int pl111_display_prepare_fb(struct drm_simple_display_pipe *pipe,
+				    struct drm_plane_state *plane_state)
+{
+	return drm_fb_cma_prepare_fb(&pipe->plane, plane_state);
+}
+
+static const struct drm_simple_display_pipe_funcs pl111_display_funcs = {
+	.check = pl111_display_check,
+	.enable = pl111_display_enable,
+	.disable = pl111_display_disable,
+	.update = pl111_display_update,
+	.prepare_fb = pl111_display_prepare_fb,
+};
+
+static int pl111_clk_div_choose_div(struct clk_hw *hw, unsigned long rate,
+				    unsigned long *prate, bool set_parent)
+{
+	int best_div = 1, div;
+	struct clk_hw *parent = clk_hw_get_parent(hw);
+	unsigned long best_prate = 0;
+	unsigned long best_diff = ~0ul;
+	int max_div = (1 << (TIM2_PCD_LO_BITS + TIM2_PCD_HI_BITS)) - 1;
+
+	for (div = 1; div < max_div; div++) {
+		unsigned long this_prate, div_rate, diff;
+
+		if (set_parent)
+			this_prate = clk_hw_round_rate(parent, rate * div);
+		else
+			this_prate = *prate;
+		div_rate = DIV_ROUND_UP_ULL(this_prate, div);
+		diff = abs(rate - div_rate);
+
+		if (diff < best_diff) {
+			best_div = div;
+			best_diff = diff;
+			best_prate = this_prate;
+		}
+	}
+
+	*prate = best_prate;
+	return best_div;
+}
+
+static long pl111_clk_div_round_rate(struct clk_hw *hw, unsigned long rate,
+				     unsigned long *prate)
+{
+	int div = pl111_clk_div_choose_div(hw, rate, prate, true);
+
+	return DIV_ROUND_UP_ULL(*prate, div);
+}
+
+static unsigned long pl111_clk_div_recalc_rate(struct clk_hw *hw,
+					       unsigned long prate)
+{
+	struct pl111_drm_dev_private *priv =
+		container_of(hw, struct pl111_drm_dev_private, clk_div);
+	u32 tim2 = readl(priv->regs + CLCD_TIM2);
+	int div;
+
+	if (tim2 & TIM2_BCD)
+		return prate;
+
+	div = tim2 & TIM2_PCD_LO_MASK;
+	div |= (tim2 & TIM2_PCD_HI_MASK) >>
+		(TIM2_PCD_HI_SHIFT - TIM2_PCD_LO_BITS);
+	div += 2;
+
+	return DIV_ROUND_UP_ULL(prate, div);
+}
+
+static int pl111_clk_div_set_rate(struct clk_hw *hw, unsigned long rate,
+				  unsigned long prate)
+{
+	struct pl111_drm_dev_private *priv =
+		container_of(hw, struct pl111_drm_dev_private, clk_div);
+	int div = pl111_clk_div_choose_div(hw, rate, &prate, false);
+	u32 tim2;
+
+	spin_lock(&priv->tim2_lock);
+	tim2 = readl(priv->regs + CLCD_TIM2);
+	tim2 &= ~(TIM2_BCD | TIM2_PCD_LO_MASK | TIM2_PCD_HI_MASK);
+
+	if (div == 1) {
+		tim2 |= TIM2_BCD;
+	} else {
+		div -= 2;
+		tim2 |= div & TIM2_PCD_LO_MASK;
+		tim2 |= (div >> TIM2_PCD_LO_BITS) << TIM2_PCD_HI_SHIFT;
+	}
+
+	writel(tim2, priv->regs + CLCD_TIM2);
+	spin_unlock(&priv->tim2_lock);
+
+	return 0;
+}
+
+static const struct clk_ops pl111_clk_div_ops = {
+	.recalc_rate = pl111_clk_div_recalc_rate,
+	.round_rate = pl111_clk_div_round_rate,
+	.set_rate = pl111_clk_div_set_rate,
+};
+
+static int
+pl111_init_clock_divider(struct drm_device *drm)
+{
+	struct pl111_drm_dev_private *priv = drm->dev_private;
+	struct clk *parent = devm_clk_get(drm->dev, "clcdclk");
+	struct clk_hw *div = &priv->clk_div;
+	const char *parent_name;
+	struct clk_init_data init = {
+		.name = "pl111_div",
+		.ops = &pl111_clk_div_ops,
+		.parent_names = &parent_name,
+		.num_parents = 1,
+		.flags = CLK_SET_RATE_PARENT,
+	};
+	int ret;
+
+	if (IS_ERR(parent)) {
+		dev_err(drm->dev, "CLCD: unable to get clcdclk.\n");
+		return PTR_ERR(parent);
+	}
+	parent_name = __clk_get_name(parent);
+
+	spin_lock_init(&priv->tim2_lock);
+	div->init = &init;
+
+	ret = devm_clk_hw_register(drm->dev, div);
+
+	priv->clk = div->clk;
+	return ret;
+}
+
+int pl111_display_init(struct drm_device *drm)
+{
+	struct pl111_drm_dev_private *priv = drm->dev_private;
+	struct device *dev = drm->dev;
+	struct device_node *endpoint;
+	u32 tft_r0b0g0[3];
+	int ret;
+	static const u32 formats[] = {
+		DRM_FORMAT_ABGR8888,
+		DRM_FORMAT_XBGR8888,
+		DRM_FORMAT_ARGB8888,
+		DRM_FORMAT_XRGB8888,
+		DRM_FORMAT_BGR565,
+		DRM_FORMAT_RGB565,
+		DRM_FORMAT_ABGR1555,
+		DRM_FORMAT_XBGR1555,
+		DRM_FORMAT_ARGB1555,
+		DRM_FORMAT_XRGB1555,
+		DRM_FORMAT_ABGR4444,
+		DRM_FORMAT_XBGR4444,
+		DRM_FORMAT_ARGB4444,
+		DRM_FORMAT_XRGB4444,
+	};
+
+	endpoint = of_graph_get_next_endpoint(dev->of_node, NULL);
+	if (!endpoint)
+		return -ENODEV;
+
+	if (of_property_read_u32_array(endpoint,
+				       "arm,pl11x,tft-r0g0b0-pads",
+				       tft_r0b0g0,
+				       ARRAY_SIZE(tft_r0b0g0)) != 0) {
+		dev_err(dev, "arm,pl11x,tft-r0g0b0-pads should be 3 ints\n");
+		of_node_put(endpoint);
+		return -ENOENT;
+	}
+	of_node_put(endpoint);
+
+	if (tft_r0b0g0[0] != 0 ||
+	    tft_r0b0g0[1] != 8 ||
+	    tft_r0b0g0[2] != 16) {
+		dev_err(dev, "arm,pl11x,tft-r0g0b0-pads != [0,8,16] not yet supported\n");
+		return -EINVAL;
+	}
+
+	ret = pl111_init_clock_divider(drm);
+	if (ret)
+		return ret;
+
+	ret = drm_simple_display_pipe_init(drm, &priv->pipe,
+					   &pl111_display_funcs,
+					   formats, ARRAY_SIZE(formats),
+					   &priv->connector.connector);
+	if (ret)
+		return ret;
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/pl111/pl111_drm.h b/drivers/gpu/drm/pl111/pl111_drm.h
new file mode 100644
index 000000000000..5c685bfc8fdc
--- /dev/null
+++ b/drivers/gpu/drm/pl111/pl111_drm.h
@@ -0,0 +1,67 @@
+/*
+ *
+ * (C) COPYRIGHT 2012-2013 ARM Limited. All rights reserved.
+ *
+ *
+ * Parts of this file were based on sources as follows:
+ *
+ * Copyright (c) 2006-2008 Intel Corporation
+ * Copyright (c) 2007 Dave Airlie <airlied@linux.ie>
+ * Copyright (C) 2011 Texas Instruments
+ *
+ * This program is free software and is provided to you under the terms of the
+ * GNU General Public License version 2 as published by the Free Software
+ * Foundation, and any use by you of this program is subject to the terms of
+ * such GNU licence.
+ *
+ */
+
+#ifndef _PL111_DRM_H_
+#define _PL111_DRM_H_
+
+#include <drm/drm_gem.h>
+#include <drm/drm_simple_kms_helper.h>
+#include <linux/clk-provider.h>
+
+#define CLCD_IRQ_NEXTBASE_UPDATE BIT(2)
+
+struct drm_minor;
+
+struct pl111_drm_connector {
+	struct drm_connector connector;
+	struct drm_panel *panel;
+};
+
+struct pl111_drm_dev_private {
+	struct drm_device *drm;
+
+	struct pl111_drm_connector connector;
+	struct drm_simple_display_pipe pipe;
+	struct drm_fbdev_cma *fbdev;
+
+	void *regs;
+	/* The pixel clock (a reference to our clock divider off of CLCDCLK). */
+	struct clk *clk;
+	/* pl111's internal clock divider. */
+	struct clk_hw clk_div;
+	/* Lock to sync access to CLCD_TIM2 between the common clock
+	 * subsystem and pl111_display_enable().
+	 */
+	spinlock_t tim2_lock;
+};
+
+#define to_pl111_connector(x) \
+	container_of(x, struct pl111_drm_connector, connector)
+
+int pl111_display_init(struct drm_device *dev);
+int pl111_enable_vblank(struct drm_device *drm, unsigned int crtc);
+void pl111_disable_vblank(struct drm_device *drm, unsigned int crtc);
+irqreturn_t pl111_irq(int irq, void *data);
+int pl111_connector_init(struct drm_device *dev);
+int pl111_encoder_init(struct drm_device *dev);
+int pl111_dumb_create(struct drm_file *file_priv,
+		      struct drm_device *dev,
+		      struct drm_mode_create_dumb *args);
+int pl111_debugfs_init(struct drm_minor *minor);
+
+#endif /* _PL111_DRM_H_ */
diff --git a/drivers/gpu/drm/pl111/pl111_drv.c b/drivers/gpu/drm/pl111/pl111_drv.c
new file mode 100644
index 000000000000..e96efad37d27
--- /dev/null
+++ b/drivers/gpu/drm/pl111/pl111_drv.c
@@ -0,0 +1,269 @@
+/*
+ * (C) COPYRIGHT 2012-2013 ARM Limited. All rights reserved.
+ *
+ * Parts of this file were based on sources as follows:
+ *
+ * Copyright (c) 2006-2008 Intel Corporation
+ * Copyright (c) 2007 Dave Airlie <airlied@linux.ie>
+ * Copyright (C) 2011 Texas Instruments
+ *
+ * This program is free software and is provided to you under the terms of the
+ * GNU General Public License version 2 as published by the Free Software
+ * Foundation, and any use by you of this program is subject to the terms of
+ * such GNU licence.
+ *
+ */
+
+/**
+ * DOC: ARM PrimeCell PL111 CLCD Driver
+ *
+ * The PL111 is a simple LCD controller that can support TFT and STN
+ * displays.  This driver exposes a standard KMS interface for them.
+ *
+ * This driver uses the same Device Tree binding as the fbdev CLCD
+ * driver.  While the fbdev driver supports panels that may be
+ * connected to the CLCD internally to the CLCD driver, in DRM the
+ * panels get split out to drivers/gpu/drm/panels/.  This means that,
+ * in converting from using fbdev to using DRM, you also need to write
+ * a panel driver (which may be as simple as an entry in
+ * panel-simple.c).
+ *
+ * The driver currently doesn't expose the cursor.  The DRM API for
+ * cursors requires support for 64x64 ARGB8888 cursor images, while
+ * the hardware can only support 64x64 monochrome with masking
+ * cursors.  While one could imagine trying to hack something together
+ * to look at the ARGB8888 and program reasonable in monochrome, we
+ * just don't expose the cursor at all instead, and leave cursor
+ * support to the X11 software cursor layer.
+ *
+ * TODO:
+ *
+ * - Fix race between setting plane base address and getting IRQ for
+ *   vsync firing the pageflip completion.
+ *
+ * - Expose the correct set of formats we can support based on the
+ *   "arm,pl11x,tft-r0g0b0-pads" DT property.
+ *
+ * - Use the "max-memory-bandwidth" DT property to filter the
+ *   supported formats.
+ *
+ * - Read back hardware state at boot to skip reprogramming the
+ *   hardware when doing a no-op modeset.
+ *
+ * - Use the CLKSEL bit to support switching between the two external
+ *   clock parents.
+ */
+
+#include <linux/amba/bus.h>
+#include <linux/amba/clcd-regs.h>
+#include <linux/version.h>
+#include <linux/shmem_fs.h>
+#include <linux/dma-buf.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+
+#include <drm/drmP.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_gem_cma_helper.h>
+#include <drm/drm_fb_cma_helper.h>
+
+#include "pl111_drm.h"
+
+#define DRIVER_DESC      "DRM module for PL111"
+
+static struct drm_mode_config_funcs mode_config_funcs = {
+	.fb_create = drm_fb_cma_create,
+	.atomic_check = drm_atomic_helper_check,
+	.atomic_commit = drm_atomic_helper_commit,
+};
+
+static int pl111_modeset_init(struct drm_device *dev)
+{
+	struct drm_mode_config *mode_config;
+	struct pl111_drm_dev_private *priv = dev->dev_private;
+	int ret = 0;
+
+	drm_mode_config_init(dev);
+	mode_config = &dev->mode_config;
+	mode_config->funcs = &mode_config_funcs;
+	mode_config->min_width = 1;
+	mode_config->max_width = 1024;
+	mode_config->min_height = 1;
+	mode_config->max_height = 768;
+
+	ret = pl111_connector_init(dev);
+	if (ret) {
+		dev_err(dev->dev, "Failed to create pl111_drm_connector\n");
+		goto out_config;
+	}
+
+	/* Don't actually attach if we didn't find a drm_panel
+	 * attached to us.  This will allow a kernel to include both
+	 * the fbdev pl111 driver and this one, and choose between
+	 * them based on which subsystem has support for the panel.
+	 */
+	if (!priv->connector.panel) {
+		dev_info(dev->dev,
+			 "Disabling due to lack of DRM panel device.\n");
+		ret = -ENODEV;
+		goto out_config;
+	}
+
+	ret = pl111_display_init(dev);
+	if (ret != 0) {
+		dev_err(dev->dev, "Failed to init display\n");
+		goto out_config;
+	}
+
+	ret = drm_vblank_init(dev, 1);
+	if (ret != 0) {
+		dev_err(dev->dev, "Failed to init vblank\n");
+		goto out_config;
+	}
+
+	drm_mode_config_reset(dev);
+
+	priv->fbdev = drm_fbdev_cma_init(dev, 32,
+					 dev->mode_config.num_connector);
+
+	drm_kms_helper_poll_init(dev);
+
+	goto finish;
+
+out_config:
+	drm_mode_config_cleanup(dev);
+finish:
+	return ret;
+}
+
+DEFINE_DRM_GEM_CMA_FOPS(drm_fops);
+
+static void pl111_lastclose(struct drm_device *dev)
+{
+	struct pl111_drm_dev_private *priv = dev->dev_private;
+
+	drm_fbdev_cma_restore_mode(priv->fbdev);
+}
+
+static struct drm_driver pl111_drm_driver = {
+	.driver_features =
+		DRIVER_MODESET | DRIVER_GEM | DRIVER_PRIME | DRIVER_ATOMIC,
+	.lastclose = pl111_lastclose,
+	.ioctls = NULL,
+	.fops = &drm_fops,
+	.name = "pl111",
+	.desc = DRIVER_DESC,
+	.date = "20170317",
+	.major = 1,
+	.minor = 0,
+	.patchlevel = 0,
+	.dumb_create = drm_gem_cma_dumb_create,
+	.dumb_destroy = drm_gem_dumb_destroy,
+	.dumb_map_offset = drm_gem_cma_dumb_map_offset,
+	.gem_free_object = drm_gem_cma_free_object,
+	.gem_vm_ops = &drm_gem_cma_vm_ops,
+
+	.enable_vblank = pl111_enable_vblank,
+	.disable_vblank = pl111_disable_vblank,
+
+	.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
+	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
+	.gem_prime_import = drm_gem_prime_import,
+	.gem_prime_import_sg_table = drm_gem_cma_prime_import_sg_table,
+	.gem_prime_export = drm_gem_prime_export,
+	.gem_prime_get_sg_table	= drm_gem_cma_prime_get_sg_table,
+
+#if defined(CONFIG_DEBUG_FS)
+	.debugfs_init = pl111_debugfs_init,
+#endif
+};
+
+#ifdef CONFIG_ARM_AMBA
+static int pl111_amba_probe(struct amba_device *amba_dev,
+			    const struct amba_id *id)
+{
+	struct device *dev = &amba_dev->dev;
+	struct pl111_drm_dev_private *priv;
+	struct drm_device *drm;
+	int ret;
+
+	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	drm = drm_dev_alloc(&pl111_drm_driver, dev);
+	if (IS_ERR(drm))
+		return PTR_ERR(drm);
+	amba_set_drvdata(amba_dev, drm);
+	priv->drm = drm;
+	drm->dev_private = priv;
+
+	priv->regs = devm_ioremap_resource(dev, &amba_dev->res);
+	if (IS_ERR(priv->regs)) {
+		dev_err(dev, "%s failed mmio\n", __func__);
+		return PTR_ERR(priv->regs);
+	}
+
+	/* turn off interrupts before requesting the irq */
+	writel(0, priv->regs + CLCD_PL111_IENB);
+
+	ret = devm_request_irq(dev, amba_dev->irq[0], pl111_irq, 0,
+			       "pl111", priv);
+	if (ret != 0) {
+		dev_err(dev, "%s failed irq %d\n", __func__, ret);
+		return ret;
+	}
+
+	ret = pl111_modeset_init(drm);
+	if (ret != 0)
+		goto dev_unref;
+
+	ret = drm_dev_register(drm, 0);
+	if (ret < 0)
+		goto dev_unref;
+
+	return 0;
+
+dev_unref:
+	drm_dev_unref(drm);
+	return ret;
+}
+
+static int pl111_amba_remove(struct amba_device *amba_dev)
+{
+	struct drm_device *drm = amba_get_drvdata(amba_dev);
+	struct pl111_drm_dev_private *priv = drm->dev_private;
+
+	drm_dev_unregister(drm);
+	if (priv->fbdev)
+		drm_fbdev_cma_fini(priv->fbdev);
+	drm_mode_config_cleanup(drm);
+	drm_dev_unref(drm);
+
+	return 0;
+}
+
+static struct amba_id pl111_id_table[] = {
+	{
+		.id = 0x00041111,
+		.mask = 0x000fffff,
+	},
+	{0, 0},
+};
+
+static struct amba_driver pl111_amba_driver = {
+	.drv = {
+		.name = "drm-clcd-pl111",
+	},
+	.probe = pl111_amba_probe,
+	.remove = pl111_amba_remove,
+	.id_table = pl111_id_table,
+};
+
+module_amba_driver(pl111_amba_driver);
+#endif /* CONFIG_ARM_AMBA */
+
+MODULE_DESCRIPTION(DRIVER_DESC);
+MODULE_AUTHOR("ARM Ltd.");
+MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/qxl/Makefile b/drivers/gpu/drm/qxl/Makefile
index bacc4aff1201..33a7d0c434b7 100644
--- a/drivers/gpu/drm/qxl/Makefile
+++ b/drivers/gpu/drm/qxl/Makefile
@@ -2,8 +2,6 @@
 # Makefile for the drm device driver.  This driver provides support for the
 # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.
 
-ccflags-y := -Iinclude/drm
-
 qxl-y := qxl_drv.o qxl_kms.o qxl_display.o qxl_ttm.o qxl_fb.o qxl_object.o qxl_gem.o qxl_cmd.o qxl_image.o qxl_draw.o qxl_debugfs.o qxl_irq.o qxl_dumb.o qxl_ioctl.o qxl_release.o qxl_prime.o
 
 obj-$(CONFIG_DRM_QXL)+= qxl.o
diff --git a/drivers/gpu/drm/qxl/qxl_debugfs.c b/drivers/gpu/drm/qxl/qxl_debugfs.c
index ffe821b61f7d..15c84068d3fb 100644
--- a/drivers/gpu/drm/qxl/qxl_debugfs.c
+++ b/drivers/gpu/drm/qxl/qxl_debugfs.c
@@ -30,7 +30,7 @@
 
 #include <linux/debugfs.h>
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "qxl_drv.h"
 #include "qxl_object.h"
 
diff --git a/drivers/gpu/drm/qxl/qxl_display.c b/drivers/gpu/drm/qxl/qxl_display.c
index 4a340efd8ba6..03fe182203ce 100644
--- a/drivers/gpu/drm/qxl/qxl_display.c
+++ b/drivers/gpu/drm/qxl/qxl_display.c
@@ -23,16 +23,15 @@
  *          Alon Levy
  */
 
-
 #include <linux/crc32.h>
-
-#include "qxl_drv.h"
-#include "qxl_object.h"
-#include "drm_crtc_helper.h"
+#include <drm/drm_crtc_helper.h>
 #include <drm/drm_plane_helper.h>
 #include <drm/drm_atomic_helper.h>
 #include <drm/drm_atomic.h>
 
+#include "qxl_drv.h"
+#include "qxl_object.h"
+
 static bool qxl_head_enabled(struct qxl_head *head)
 {
 	return head->width && head->height;
diff --git a/drivers/gpu/drm/qxl/qxl_drv.c b/drivers/gpu/drm/qxl/qxl_drv.c
index abf7b8360361..c2fc201d9e1b 100644
--- a/drivers/gpu/drm/qxl/qxl_drv.c
+++ b/drivers/gpu/drm/qxl/qxl_drv.c
@@ -31,9 +31,9 @@
 #include <linux/module.h>
 #include <linux/console.h>
 
-#include "drmP.h"
-#include "drm/drm.h"
-#include "drm_crtc_helper.h"
+#include <drm/drmP.h>
+#include <drm/drm.h>
+#include <drm/drm_crtc_helper.h>
 #include "qxl_drv.h"
 #include "qxl_object.h"
 
diff --git a/drivers/gpu/drm/qxl/qxl_drv.h b/drivers/gpu/drm/qxl/qxl_drv.h
index 5ea290a33a68..3591d2330a09 100644
--- a/drivers/gpu/drm/qxl/qxl_drv.h
+++ b/drivers/gpu/drm/qxl/qxl_drv.h
@@ -36,20 +36,18 @@
 #include <linux/firmware.h>
 #include <linux/platform_device.h>
 
-#include "drmP.h"
-#include "drm_crtc.h"
-#include <ttm/ttm_bo_api.h>
-#include <ttm/ttm_bo_driver.h>
-#include <ttm/ttm_placement.h>
-#include <ttm/ttm_module.h>
-
+#include <drm/drm_crtc.h>
 #include <drm/drm_encoder.h>
 #include <drm/drm_gem.h>
-
+#include <drm/drmP.h>
+#include <drm/ttm/ttm_bo_api.h>
+#include <drm/ttm/ttm_bo_driver.h>
 /* just for ttm_validate_buffer */
-#include <ttm/ttm_execbuf_util.h>
-
+#include <drm/ttm/ttm_execbuf_util.h>
+#include <drm/ttm/ttm_module.h>
+#include <drm/ttm/ttm_placement.h>
 #include <drm/qxl_drm.h>
+
 #include "qxl_dev.h"
 
 #define DRIVER_AUTHOR		"Dave Airlie"
diff --git a/drivers/gpu/drm/qxl/qxl_fb.c b/drivers/gpu/drm/qxl/qxl_fb.c
index 14e2a49a4dcf..573e7e9a5f98 100644
--- a/drivers/gpu/drm/qxl/qxl_fb.c
+++ b/drivers/gpu/drm/qxl/qxl_fb.c
@@ -25,14 +25,15 @@
  */
 #include <linux/module.h>
 
-#include "drmP.h"
-#include "drm/drm.h"
-#include "drm/drm_crtc.h"
-#include "drm/drm_crtc_helper.h"
+#include <drm/drmP.h>
+#include <drm/drm.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_fb_helper.h>
+
 #include "qxl_drv.h"
 
 #include "qxl_object.h"
-#include "drm_fb_helper.h"
 
 #define QXL_DIRTY_DELAY (HZ / 30)
 
diff --git a/drivers/gpu/drm/qxl/qxl_gem.c b/drivers/gpu/drm/qxl/qxl_gem.c
index 3f185c4da5b7..85f546719adb 100644
--- a/drivers/gpu/drm/qxl/qxl_gem.c
+++ b/drivers/gpu/drm/qxl/qxl_gem.c
@@ -23,8 +23,9 @@
  *          Alon Levy
  */
 
-#include "drmP.h"
-#include "drm/drm.h"
+#include <drm/drmP.h>
+#include <drm/drm.h>
+
 #include "qxl_drv.h"
 #include "qxl_object.h"
 
diff --git a/drivers/gpu/drm/qxl/qxl_ttm.c b/drivers/gpu/drm/qxl/qxl_ttm.c
index 0fdedee4509d..87fc1dbd0a2f 100644
--- a/drivers/gpu/drm/qxl/qxl_ttm.c
+++ b/drivers/gpu/drm/qxl/qxl_ttm.c
@@ -23,11 +23,11 @@
  *          Alon Levy
  */
 
-#include <ttm/ttm_bo_api.h>
-#include <ttm/ttm_bo_driver.h>
-#include <ttm/ttm_placement.h>
-#include <ttm/ttm_page_alloc.h>
-#include <ttm/ttm_module.h>
+#include <drm/ttm/ttm_bo_api.h>
+#include <drm/ttm/ttm_bo_driver.h>
+#include <drm/ttm/ttm_placement.h>
+#include <drm/ttm/ttm_page_alloc.h>
+#include <drm/ttm/ttm_module.h>
 #include <drm/drmP.h>
 #include <drm/drm.h>
 #include <drm/qxl_drm.h>
diff --git a/drivers/gpu/drm/r128/Makefile b/drivers/gpu/drm/r128/Makefile
index 1cc72ae3a880..1a6700ebaf09 100644
--- a/drivers/gpu/drm/r128/Makefile
+++ b/drivers/gpu/drm/r128/Makefile
@@ -2,7 +2,6 @@
 # Makefile for the drm device driver.  This driver provides support for the
 # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.
 
-ccflags-y := -Iinclude/drm
 r128-y   := r128_drv.o r128_cce.o r128_state.o r128_irq.o
 
 r128-$(CONFIG_COMPAT)   += r128_ioc32.o
diff --git a/drivers/gpu/drm/radeon/Makefile b/drivers/gpu/drm/radeon/Makefile
index 08bd17d3925c..a5d3cd3ecb5f 100644
--- a/drivers/gpu/drm/radeon/Makefile
+++ b/drivers/gpu/drm/radeon/Makefile
@@ -2,7 +2,7 @@
 # Makefile for the drm device driver.  This driver provides support for the
 # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.
 
-ccflags-y := -Iinclude/drm -Idrivers/gpu/drm/amd/include
+ccflags-y := -Idrivers/gpu/drm/amd/include
 
 hostprogs-y := mkregtable
 clean-files := rn50_reg_safe.h r100_reg_safe.h r200_reg_safe.h rv515_reg_safe.h r300_reg_safe.h r420_reg_safe.h rs600_reg_safe.h r600_reg_safe.h evergreen_reg_safe.h cayman_reg_safe.h
diff --git a/drivers/gpu/drm/radeon/btc_dpm.c b/drivers/gpu/drm/radeon/btc_dpm.c
index 38e5123708e7..95652e643da1 100644
--- a/drivers/gpu/drm/radeon/btc_dpm.c
+++ b/drivers/gpu/drm/radeon/btc_dpm.c
@@ -22,7 +22,7 @@
  * Authors: Alex Deucher
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "btcd.h"
diff --git a/drivers/gpu/drm/radeon/ci_dpm.c b/drivers/gpu/drm/radeon/ci_dpm.c
index ea36dc4dd5d2..c97fbb2ab48b 100644
--- a/drivers/gpu/drm/radeon/ci_dpm.c
+++ b/drivers/gpu/drm/radeon/ci_dpm.c
@@ -22,7 +22,7 @@
  */
 
 #include <linux/firmware.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "radeon_ucode.h"
diff --git a/drivers/gpu/drm/radeon/ci_smc.c b/drivers/gpu/drm/radeon/ci_smc.c
index 24760ee3063e..3356a21d97ec 100644
--- a/drivers/gpu/drm/radeon/ci_smc.c
+++ b/drivers/gpu/drm/radeon/ci_smc.c
@@ -23,7 +23,7 @@
  */
 
 #include <linux/firmware.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "cikd.h"
 #include "ppsmc.h"
diff --git a/drivers/gpu/drm/radeon/cik.c b/drivers/gpu/drm/radeon/cik.c
index 008c145b7f29..258912132b62 100644
--- a/drivers/gpu/drm/radeon/cik.c
+++ b/drivers/gpu/drm/radeon/cik.c
@@ -24,7 +24,7 @@
 #include <linux/firmware.h>
 #include <linux/slab.h>
 #include <linux/module.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "radeon_audio.h"
diff --git a/drivers/gpu/drm/radeon/cypress_dpm.c b/drivers/gpu/drm/radeon/cypress_dpm.c
index a4edd0702718..3eb7899a4035 100644
--- a/drivers/gpu/drm/radeon/cypress_dpm.c
+++ b/drivers/gpu/drm/radeon/cypress_dpm.c
@@ -22,7 +22,7 @@
  * Authors: Alex Deucher
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "evergreend.h"
diff --git a/drivers/gpu/drm/radeon/kv_dpm.c b/drivers/gpu/drm/radeon/kv_dpm.c
index a7e978677937..ae1529b0ef6f 100644
--- a/drivers/gpu/drm/radeon/kv_dpm.c
+++ b/drivers/gpu/drm/radeon/kv_dpm.c
@@ -21,7 +21,7 @@
  *
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "cikd.h"
 #include "r600_dpm.h"
diff --git a/drivers/gpu/drm/radeon/kv_smc.c b/drivers/gpu/drm/radeon/kv_smc.c
index 0000b59a6d05..af60bd32a287 100644
--- a/drivers/gpu/drm/radeon/kv_smc.c
+++ b/drivers/gpu/drm/radeon/kv_smc.c
@@ -22,7 +22,7 @@
  * Authors: Alex Deucher
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "cikd.h"
 #include "kv_dpm.h"
diff --git a/drivers/gpu/drm/radeon/ni_dpm.c b/drivers/gpu/drm/radeon/ni_dpm.c
index 4a601f990562..9416e72f86aa 100644
--- a/drivers/gpu/drm/radeon/ni_dpm.c
+++ b/drivers/gpu/drm/radeon/ni_dpm.c
@@ -21,7 +21,7 @@
  *
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "nid.h"
diff --git a/drivers/gpu/drm/radeon/r600_dpm.c b/drivers/gpu/drm/radeon/r600_dpm.c
index c7fc1dbfd192..31d1b4710844 100644
--- a/drivers/gpu/drm/radeon/r600_dpm.c
+++ b/drivers/gpu/drm/radeon/r600_dpm.c
@@ -22,7 +22,7 @@
  * Authors: Alex Deucher
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "r600d.h"
diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h
index c1c8e2208a21..342e3b1fb9c7 100644
--- a/drivers/gpu/drm/radeon/radeon.h
+++ b/drivers/gpu/drm/radeon/radeon.h
@@ -68,11 +68,11 @@
 #include <linux/hashtable.h>
 #include <linux/dma-fence.h>
 
-#include <ttm/ttm_bo_api.h>
-#include <ttm/ttm_bo_driver.h>
-#include <ttm/ttm_placement.h>
-#include <ttm/ttm_module.h>
-#include <ttm/ttm_execbuf_util.h>
+#include <drm/ttm/ttm_bo_api.h>
+#include <drm/ttm/ttm_bo_driver.h>
+#include <drm/ttm/ttm_placement.h>
+#include <drm/ttm/ttm_module.h>
+#include <drm/ttm/ttm_execbuf_util.h>
 
 #include <drm/drm_gem.h>
 
diff --git a/drivers/gpu/drm/radeon/radeon_cs.c b/drivers/gpu/drm/radeon/radeon_cs.c
index 3ac671f6c8e1..00b22af70f5c 100644
--- a/drivers/gpu/drm/radeon/radeon_cs.c
+++ b/drivers/gpu/drm/radeon/radeon_cs.c
@@ -87,7 +87,8 @@ static int radeon_cs_parser_relocs(struct radeon_cs_parser *p)
 	p->dma_reloc_idx = 0;
 	/* FIXME: we assume that each relocs use 4 dwords */
 	p->nrelocs = chunk->length_dw / 4;
-	p->relocs = drm_calloc_large(p->nrelocs, sizeof(struct radeon_bo_list));
+	p->relocs = kvmalloc_array(p->nrelocs, sizeof(struct radeon_bo_list),
+			GFP_KERNEL | __GFP_ZERO);
 	if (p->relocs == NULL) {
 		return -ENOMEM;
 	}
@@ -341,7 +342,7 @@ int radeon_cs_parser_init(struct radeon_cs_parser *p, void *data)
 				continue;
 		}
 
-		p->chunks[i].kdata = drm_malloc_ab(size, sizeof(uint32_t));
+		p->chunks[i].kdata = kvmalloc_array(size, sizeof(uint32_t), GFP_KERNEL);
 		size *= sizeof(uint32_t);
 		if (p->chunks[i].kdata == NULL) {
 			return -ENOMEM;
@@ -440,10 +441,10 @@ static void radeon_cs_parser_fini(struct radeon_cs_parser *parser, int error, bo
 		}
 	}
 	kfree(parser->track);
-	drm_free_large(parser->relocs);
-	drm_free_large(parser->vm_bos);
+	kvfree(parser->relocs);
+	kvfree(parser->vm_bos);
 	for (i = 0; i < parser->nchunks; i++)
-		drm_free_large(parser->chunks[i].kdata);
+		kvfree(parser->chunks[i].kdata);
 	kfree(parser->chunks);
 	kfree(parser->chunks_array);
 	radeon_ib_free(parser->rdev, &parser->ib);
diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c
index 93d45aa5c3d4..6f906abd612b 100644
--- a/drivers/gpu/drm/radeon/radeon_drv.c
+++ b/drivers/gpu/drm/radeon/radeon_drv.c
@@ -41,7 +41,7 @@
 #include <drm/drm_gem.h>
 #include <drm/drm_fb_helper.h>
 
-#include "drm_crtc_helper.h"
+#include <drm/drm_crtc_helper.h>
 #include "radeon_kfd.h"
 
 /*
@@ -115,10 +115,6 @@ int radeon_resume_kms(struct drm_device *dev, bool resume, bool fbcon);
 u32 radeon_get_vblank_counter_kms(struct drm_device *dev, unsigned int pipe);
 int radeon_enable_vblank_kms(struct drm_device *dev, unsigned int pipe);
 void radeon_disable_vblank_kms(struct drm_device *dev, unsigned int pipe);
-int radeon_get_vblank_timestamp_kms(struct drm_device *dev, unsigned int pipe,
-				    int *max_error,
-				    struct timeval *vblank_time,
-				    unsigned flags);
 void radeon_driver_irq_preinstall_kms(struct drm_device *dev);
 int radeon_driver_irq_postinstall_kms(struct drm_device *dev);
 void radeon_driver_irq_uninstall_kms(struct drm_device *dev);
@@ -530,6 +526,16 @@ static const struct file_operations radeon_driver_kms_fops = {
 #endif
 };
 
+static bool
+radeon_get_crtc_scanout_position(struct drm_device *dev, unsigned int pipe,
+				 bool in_vblank_irq, int *vpos, int *hpos,
+				 ktime_t *stime, ktime_t *etime,
+				 const struct drm_display_mode *mode)
+{
+	return radeon_get_crtc_scanoutpos(dev, pipe, 0, vpos, hpos,
+					  stime, etime, mode);
+}
+
 static struct drm_driver kms_driver = {
 	.driver_features =
 	    DRIVER_USE_AGP |
@@ -544,8 +550,8 @@ static struct drm_driver kms_driver = {
 	.get_vblank_counter = radeon_get_vblank_counter_kms,
 	.enable_vblank = radeon_enable_vblank_kms,
 	.disable_vblank = radeon_disable_vblank_kms,
-	.get_vblank_timestamp = radeon_get_vblank_timestamp_kms,
-	.get_scanout_position = radeon_get_crtc_scanoutpos,
+	.get_vblank_timestamp = drm_calc_vbltimestamp_from_scanoutpos,
+	.get_scanout_position = radeon_get_crtc_scanout_position,
 	.irq_preinstall = radeon_driver_irq_preinstall_kms,
 	.irq_postinstall = radeon_driver_irq_postinstall_kms,
 	.irq_uninstall = radeon_driver_irq_uninstall_kms,
diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c
index dddb372de2b9..574bf7e6b118 100644
--- a/drivers/gpu/drm/radeon/radeon_gem.c
+++ b/drivers/gpu/drm/radeon/radeon_gem.c
@@ -587,7 +587,7 @@ error_unreserve:
 	ttm_eu_backoff_reservation(&ticket, &list);
 
 error_free:
-	drm_free_large(vm_bos);
+	kvfree(vm_bos);
 
 	if (r && r != -ERESTARTSYS)
 		DRM_ERROR("Couldn't update BO_VA (%d)\n", r);
diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c
index 4761f27f2ca2..d0ad03674250 100644
--- a/drivers/gpu/drm/radeon/radeon_kms.c
+++ b/drivers/gpu/drm/radeon/radeon_kms.c
@@ -858,43 +858,6 @@ void radeon_disable_vblank_kms(struct drm_device *dev, int crtc)
 	spin_unlock_irqrestore(&rdev->irq.lock, irqflags);
 }
 
-/**
- * radeon_get_vblank_timestamp_kms - get vblank timestamp
- *
- * @dev: drm dev pointer
- * @crtc: crtc to get the timestamp for
- * @max_error: max error
- * @vblank_time: time value
- * @flags: flags passed to the driver
- *
- * Gets the timestamp on the requested crtc based on the
- * scanout position.  (all asics).
- * Returns postive status flags on success, negative error on failure.
- */
-int radeon_get_vblank_timestamp_kms(struct drm_device *dev, int crtc,
-				    int *max_error,
-				    struct timeval *vblank_time,
-				    unsigned flags)
-{
-	struct drm_crtc *drmcrtc;
-	struct radeon_device *rdev = dev->dev_private;
-
-	if (crtc < 0 || crtc >= dev->num_crtcs) {
-		DRM_ERROR("Invalid crtc %d\n", crtc);
-		return -EINVAL;
-	}
-
-	/* Get associated drm_crtc: */
-	drmcrtc = &rdev->mode_info.crtcs[crtc]->base;
-	if (!drmcrtc)
-		return -EINVAL;
-
-	/* Helper routine in DRM core does all the work: */
-	return drm_calc_vbltimestamp_from_scanoutpos(dev, crtc, max_error,
-						     vblank_time, flags,
-						     &drmcrtc->hwmode);
-}
-
 const struct drm_ioctl_desc radeon_ioctls_kms[] = {
 	DRM_IOCTL_DEF_DRV(RADEON_CP_INIT, drm_invalid_op, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
 	DRM_IOCTL_DEF_DRV(RADEON_CP_START, drm_invalid_op, DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY),
diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h
index ad282648fc8b..00f5ec5c12c7 100644
--- a/drivers/gpu/drm/radeon/radeon_mode.h
+++ b/drivers/gpu/drm/radeon/radeon_mode.h
@@ -691,6 +691,9 @@ struct atom_voltage_table
 };
 
 /* Driver internal use only flags of radeon_get_crtc_scanoutpos() */
+#define DRM_SCANOUTPOS_VALID        (1 << 0)
+#define DRM_SCANOUTPOS_IN_VBLANK    (1 << 1)
+#define DRM_SCANOUTPOS_ACCURATE     (1 << 2)
 #define USE_REAL_VBLANKSTART 		(1 << 30)
 #define GET_DISTANCE_TO_VBLANKSTART	(1 << 31)
 
diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c
index 8c7872339c2a..84802b201bef 100644
--- a/drivers/gpu/drm/radeon/radeon_ring.c
+++ b/drivers/gpu/drm/radeon/radeon_ring.c
@@ -314,7 +314,7 @@ unsigned radeon_ring_backup(struct radeon_device *rdev, struct radeon_ring *ring
 	}
 
 	/* and then save the content of the ring */
-	*data = drm_malloc_ab(size, sizeof(uint32_t));
+	*data = kvmalloc_array(size, sizeof(uint32_t), GFP_KERNEL);
 	if (!*data) {
 		mutex_unlock(&rdev->ring_lock);
 		return 0;
@@ -356,7 +356,7 @@ int radeon_ring_restore(struct radeon_device *rdev, struct radeon_ring *ring,
 	}
 
 	radeon_ring_unlock_commit(rdev, ring, false);
-	drm_free_large(data);
+	kvfree(data);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c
index 8b7623b5a624..faa021396da3 100644
--- a/drivers/gpu/drm/radeon/radeon_ttm.c
+++ b/drivers/gpu/drm/radeon/radeon_ttm.c
@@ -29,11 +29,11 @@
  *    Thomas Hellstrom <thomas-at-tungstengraphics-dot-com>
  *    Dave Airlie
  */
-#include <ttm/ttm_bo_api.h>
-#include <ttm/ttm_bo_driver.h>
-#include <ttm/ttm_placement.h>
-#include <ttm/ttm_module.h>
-#include <ttm/ttm_page_alloc.h>
+#include <drm/ttm/ttm_bo_api.h>
+#include <drm/ttm/ttm_bo_driver.h>
+#include <drm/ttm/ttm_placement.h>
+#include <drm/ttm/ttm_module.h>
+#include <drm/ttm/ttm_page_alloc.h>
 #include <drm/drmP.h>
 #include <drm/radeon_drm.h>
 #include <linux/seq_file.h>
diff --git a/drivers/gpu/drm/radeon/radeon_vm.c b/drivers/gpu/drm/radeon/radeon_vm.c
index a1358748cea5..5f68245579a3 100644
--- a/drivers/gpu/drm/radeon/radeon_vm.c
+++ b/drivers/gpu/drm/radeon/radeon_vm.c
@@ -132,8 +132,8 @@ struct radeon_bo_list *radeon_vm_get_bos(struct radeon_device *rdev,
 	struct radeon_bo_list *list;
 	unsigned i, idx;
 
-	list = drm_malloc_ab(vm->max_pde_used + 2,
-			     sizeof(struct radeon_bo_list));
+	list = kvmalloc_array(vm->max_pde_used + 2,
+			     sizeof(struct radeon_bo_list), GFP_KERNEL);
 	if (!list)
 		return NULL;
 
diff --git a/drivers/gpu/drm/radeon/rs780_dpm.c b/drivers/gpu/drm/radeon/rs780_dpm.c
index 94b48fc1e266..b5e4e09a8996 100644
--- a/drivers/gpu/drm/radeon/rs780_dpm.c
+++ b/drivers/gpu/drm/radeon/rs780_dpm.c
@@ -22,7 +22,7 @@
  * Authors: Alex Deucher
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "rs780d.h"
diff --git a/drivers/gpu/drm/radeon/rv6xx_dpm.c b/drivers/gpu/drm/radeon/rv6xx_dpm.c
index 25e29303b119..d91aa3944593 100644
--- a/drivers/gpu/drm/radeon/rv6xx_dpm.c
+++ b/drivers/gpu/drm/radeon/rv6xx_dpm.c
@@ -22,7 +22,7 @@
  * Authors: Alex Deucher
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "rv6xxd.h"
diff --git a/drivers/gpu/drm/radeon/rv730_dpm.c b/drivers/gpu/drm/radeon/rv730_dpm.c
index d37ba2cb886e..38fdb4152e2a 100644
--- a/drivers/gpu/drm/radeon/rv730_dpm.c
+++ b/drivers/gpu/drm/radeon/rv730_dpm.c
@@ -22,7 +22,7 @@
  * Authors: Alex Deucher
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "rv730d.h"
 #include "r600_dpm.h"
diff --git a/drivers/gpu/drm/radeon/rv740_dpm.c b/drivers/gpu/drm/radeon/rv740_dpm.c
index 4b850824fe06..afd597ec5085 100644
--- a/drivers/gpu/drm/radeon/rv740_dpm.c
+++ b/drivers/gpu/drm/radeon/rv740_dpm.c
@@ -22,7 +22,7 @@
  * Authors: Alex Deucher
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "rv740d.h"
 #include "r600_dpm.h"
diff --git a/drivers/gpu/drm/radeon/rv770_dpm.c b/drivers/gpu/drm/radeon/rv770_dpm.c
index a010decf59af..cb2a7ec4e217 100644
--- a/drivers/gpu/drm/radeon/rv770_dpm.c
+++ b/drivers/gpu/drm/radeon/rv770_dpm.c
@@ -22,7 +22,7 @@
  * Authors: Alex Deucher
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "rv770d.h"
diff --git a/drivers/gpu/drm/radeon/rv770_smc.c b/drivers/gpu/drm/radeon/rv770_smc.c
index b2a224407365..2b7ddee3984c 100644
--- a/drivers/gpu/drm/radeon/rv770_smc.c
+++ b/drivers/gpu/drm/radeon/rv770_smc.c
@@ -23,7 +23,7 @@
  */
 
 #include <linux/firmware.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "rv770d.h"
 #include "rv770_dpm.h"
diff --git a/drivers/gpu/drm/radeon/si_dpm.c b/drivers/gpu/drm/radeon/si_dpm.c
index c7af9fdd20c7..ee3e74266a13 100644
--- a/drivers/gpu/drm/radeon/si_dpm.c
+++ b/drivers/gpu/drm/radeon/si_dpm.c
@@ -21,7 +21,7 @@
  *
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "sid.h"
diff --git a/drivers/gpu/drm/radeon/si_smc.c b/drivers/gpu/drm/radeon/si_smc.c
index e5bb92f16775..51155abda8d8 100644
--- a/drivers/gpu/drm/radeon/si_smc.c
+++ b/drivers/gpu/drm/radeon/si_smc.c
@@ -23,7 +23,7 @@
  */
 
 #include <linux/firmware.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "sid.h"
 #include "ppsmc.h"
diff --git a/drivers/gpu/drm/radeon/sumo_dpm.c b/drivers/gpu/drm/radeon/sumo_dpm.c
index f0d5c1724f55..fd4804829e46 100644
--- a/drivers/gpu/drm/radeon/sumo_dpm.c
+++ b/drivers/gpu/drm/radeon/sumo_dpm.c
@@ -21,7 +21,7 @@
  *
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "sumod.h"
diff --git a/drivers/gpu/drm/radeon/sumo_smc.c b/drivers/gpu/drm/radeon/sumo_smc.c
index fb081d2ae374..cc051be42362 100644
--- a/drivers/gpu/drm/radeon/sumo_smc.c
+++ b/drivers/gpu/drm/radeon/sumo_smc.c
@@ -21,7 +21,7 @@
  *
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "sumod.h"
 #include "sumo_dpm.h"
diff --git a/drivers/gpu/drm/radeon/trinity_dpm.c b/drivers/gpu/drm/radeon/trinity_dpm.c
index 6730367ac228..2ef7c4e5e495 100644
--- a/drivers/gpu/drm/radeon/trinity_dpm.c
+++ b/drivers/gpu/drm/radeon/trinity_dpm.c
@@ -21,7 +21,7 @@
  *
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "radeon_asic.h"
 #include "trinityd.h"
diff --git a/drivers/gpu/drm/radeon/trinity_smc.c b/drivers/gpu/drm/radeon/trinity_smc.c
index 99dd0455334d..0310e36e3159 100644
--- a/drivers/gpu/drm/radeon/trinity_smc.c
+++ b/drivers/gpu/drm/radeon/trinity_smc.c
@@ -21,7 +21,7 @@
  *
  */
 
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "radeon.h"
 #include "trinityd.h"
 #include "trinity_dpm.h"
diff --git a/drivers/gpu/drm/rockchip/analogix_dp-rockchip.c b/drivers/gpu/drm/rockchip/analogix_dp-rockchip.c
index ce5f2d1f9994..9606121fa185 100644
--- a/drivers/gpu/drm/rockchip/analogix_dp-rockchip.c
+++ b/drivers/gpu/drm/rockchip/analogix_dp-rockchip.c
@@ -104,26 +104,18 @@ static void analogix_dp_psr_work(struct work_struct *work)
 {
 	struct rockchip_dp_device *dp =
 				container_of(work, typeof(*dp), psr_work);
-	struct drm_crtc *crtc = dp->encoder.crtc;
-	int psr_state = dp->psr_state;
-	int vact_end;
 	int ret;
 	unsigned long flags;
 
-	if (!crtc)
-		return;
-
-	vact_end = crtc->mode.vtotal - crtc->mode.vsync_start + crtc->mode.vdisplay;
-
-	ret = rockchip_drm_wait_line_flag(dp->encoder.crtc, vact_end,
-					  PSR_WAIT_LINE_FLAG_TIMEOUT_MS);
+	ret = rockchip_drm_wait_vact_end(dp->encoder.crtc,
+					 PSR_WAIT_LINE_FLAG_TIMEOUT_MS);
 	if (ret) {
 		dev_err(dp->dev, "line flag interrupt did not arrive\n");
 		return;
 	}
 
 	spin_lock_irqsave(&dp->psr_lock, flags);
-	if (psr_state == EDP_VSC_PSR_STATE_ACTIVE)
+	if (dp->psr_state == EDP_VSC_PSR_STATE_ACTIVE)
 		analogix_dp_enable_psr(dp->dev);
 	else
 		analogix_dp_disable_psr(dp->dev);
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_drv.h b/drivers/gpu/drm/rockchip/rockchip_drm_drv.h
index a48fcce3f5f6..47905faf5586 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_drv.h
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_drv.h
@@ -62,8 +62,7 @@ int rockchip_drm_dma_attach_device(struct drm_device *drm_dev,
 				   struct device *dev);
 void rockchip_drm_dma_detach_device(struct drm_device *drm_dev,
 				    struct device *dev);
-int rockchip_drm_wait_line_flag(struct drm_crtc *crtc, unsigned int line_num,
-				unsigned int mstimeout);
+int rockchip_drm_wait_vact_end(struct drm_crtc *crtc, unsigned int mstimeout);
 
 extern struct platform_driver cdn_dp_driver;
 extern struct platform_driver dw_hdmi_rockchip_pltfm_driver;
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
index 45589d6ce65e..c83f481a21c8 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
@@ -468,7 +468,7 @@ static bool vop_line_flag_irq_is_enabled(struct vop *vop)
 	return !!line_flag_irq;
 }
 
-static void vop_line_flag_irq_enable(struct vop *vop, int line_num)
+static void vop_line_flag_irq_enable(struct vop *vop)
 {
 	unsigned long flags;
 
@@ -477,7 +477,6 @@ static void vop_line_flag_irq_enable(struct vop *vop, int line_num)
 
 	spin_lock_irqsave(&vop->irq_lock, flags);
 
-	VOP_CTRL_SET(vop, line_flag_num[0], line_num);
 	VOP_INTR_SET_TYPE(vop, clear, LINE_FLAG_INTR, 1);
 	VOP_INTR_SET_TYPE(vop, enable, LINE_FLAG_INTR, 1);
 
@@ -989,6 +988,8 @@ static void vop_crtc_enable(struct drm_crtc *crtc)
 	VOP_CTRL_SET(vop, vact_st_end, val);
 	VOP_CTRL_SET(vop, vpost_st_end, val);
 
+	VOP_CTRL_SET(vop, line_flag_num[0], vact_end);
+
 	clk_set_rate(vop->dclk, adjusted_mode->clock * 1000);
 
 	VOP_CTRL_SET(vop, standby, 0);
@@ -1515,19 +1516,16 @@ static void vop_win_init(struct vop *vop)
 }
 
 /**
- * rockchip_drm_wait_line_flag - acqiure the give line flag event
+ * rockchip_drm_wait_vact_end
  * @crtc: CRTC to enable line flag
- * @line_num: interested line number
  * @mstimeout: millisecond for timeout
  *
- * Driver would hold here until the interested line flag interrupt have
- * happened or timeout to wait.
+ * Wait for vact_end line flag irq or timeout.
  *
  * Returns:
  * Zero on success, negative errno on failure.
  */
-int rockchip_drm_wait_line_flag(struct drm_crtc *crtc, unsigned int line_num,
-				unsigned int mstimeout)
+int rockchip_drm_wait_vact_end(struct drm_crtc *crtc, unsigned int mstimeout)
 {
 	struct vop *vop = to_vop(crtc);
 	unsigned long jiffies_left;
@@ -1535,14 +1533,14 @@ int rockchip_drm_wait_line_flag(struct drm_crtc *crtc, unsigned int line_num,
 	if (!crtc || !vop->is_enabled)
 		return -ENODEV;
 
-	if (line_num > crtc->mode.vtotal || mstimeout <= 0)
+	if (mstimeout <= 0)
 		return -EINVAL;
 
 	if (vop_line_flag_irq_is_enabled(vop))
 		return -EBUSY;
 
 	reinit_completion(&vop->line_flag_completion);
-	vop_line_flag_irq_enable(vop, line_num);
+	vop_line_flag_irq_enable(vop);
 
 	jiffies_left = wait_for_completion_timeout(&vop->line_flag_completion,
 						   msecs_to_jiffies(mstimeout));
@@ -1555,7 +1553,7 @@ int rockchip_drm_wait_line_flag(struct drm_crtc *crtc, unsigned int line_num,
 
 	return 0;
 }
-EXPORT_SYMBOL(rockchip_drm_wait_line_flag);
+EXPORT_SYMBOL(rockchip_drm_wait_vact_end);
 
 static int vop_bind(struct device *dev, struct device *master, void *data)
 {
diff --git a/drivers/gpu/drm/savage/Makefile b/drivers/gpu/drm/savage/Makefile
index d8f84ac7bb26..cfd436bb28e4 100644
--- a/drivers/gpu/drm/savage/Makefile
+++ b/drivers/gpu/drm/savage/Makefile
@@ -2,7 +2,6 @@
 # Makefile for the drm device driver.  This driver provides support for the
 # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.
 
-ccflags-y = -Iinclude/drm
 savage-y := savage_drv.o savage_bci.o savage_state.o
 
 obj-$(CONFIG_DRM_SAVAGE)+= savage.o
diff --git a/drivers/gpu/drm/selftests/test-drm_mm.c b/drivers/gpu/drm/selftests/test-drm_mm.c
index fa356f5dae27..dfdd858eda0a 100644
--- a/drivers/gpu/drm/selftests/test-drm_mm.c
+++ b/drivers/gpu/drm/selftests/test-drm_mm.c
@@ -514,6 +514,8 @@ static int igt_reserve(void *ignored)
 		ret = __igt_reserve(count, size + 1);
 		if (ret)
 			return ret;
+
+		cond_resched();
 	}
 
 	return 0;
@@ -712,6 +714,10 @@ static int igt_insert(void *ignored)
 			return ret;
 
 		ret = __igt_insert(count, size + 1, false);
+		if (ret)
+			return ret;
+
+		cond_resched();
 	}
 
 	return 0;
@@ -741,6 +747,10 @@ static int igt_replace(void *ignored)
 			return ret;
 
 		ret = __igt_insert(count, size + 1, true);
+		if (ret)
+			return ret;
+
+		cond_resched();
 	}
 
 	return 0;
@@ -1011,6 +1021,8 @@ static int igt_insert_range(void *ignored)
 		ret = __igt_insert_range(count, size, max/4+1, 3*max/4-1);
 		if (ret)
 			return ret;
+
+		cond_resched();
 	}
 
 	return 0;
@@ -1056,6 +1068,7 @@ static int igt_align(void *ignored)
 		drm_mm_for_each_node_safe(node, next, &mm)
 			drm_mm_remove_node(node);
 		DRM_MM_BUG_ON(!drm_mm_clean(&mm));
+		cond_resched();
 	}
 
 	ret = 0;
@@ -1097,6 +1110,8 @@ static int igt_align_pot(int max)
 			       align, bit);
 			goto out;
 		}
+
+		cond_resched();
 	}
 
 	ret = 0;
@@ -1471,6 +1486,8 @@ static int igt_evict(void *ignored)
 				goto out;
 			}
 		}
+
+		cond_resched();
 	}
 
 	ret = 0;
@@ -1566,6 +1583,8 @@ static int igt_evict_range(void *ignored)
 				goto out;
 			}
 		}
+
+		cond_resched();
 	}
 
 	ret = 0;
@@ -1683,6 +1702,7 @@ static int igt_topdown(void *ignored)
 		drm_mm_for_each_node_safe(node, next, &mm)
 			drm_mm_remove_node(node);
 		DRM_MM_BUG_ON(!drm_mm_clean(&mm));
+		cond_resched();
 	}
 
 	ret = 0;
@@ -1783,6 +1803,7 @@ static int igt_bottomup(void *ignored)
 		drm_mm_for_each_node_safe(node, next, &mm)
 			drm_mm_remove_node(node);
 		DRM_MM_BUG_ON(!drm_mm_clean(&mm));
+		cond_resched();
 	}
 
 	ret = 0;
@@ -1970,6 +1991,8 @@ static int igt_color(void *ignored)
 			drm_mm_remove_node(node);
 			kfree(node);
 		}
+
+		cond_resched();
 	}
 
 	ret = 0;
@@ -2047,6 +2070,7 @@ static int evict_color(struct drm_mm *mm,
 		}
 	}
 
+	cond_resched();
 	return 0;
 }
 
@@ -2132,6 +2156,8 @@ static int igt_color_evict(void *ignored)
 				goto out;
 			}
 		}
+
+		cond_resched();
 	}
 
 	ret = 0;
@@ -2231,6 +2257,8 @@ static int igt_color_evict_range(void *ignored)
 				goto out;
 			}
 		}
+
+		cond_resched();
 	}
 
 	ret = 0;
diff --git a/drivers/gpu/drm/sis/Makefile b/drivers/gpu/drm/sis/Makefile
index 441c061c3ad0..7bf4c130c8fd 100644
--- a/drivers/gpu/drm/sis/Makefile
+++ b/drivers/gpu/drm/sis/Makefile
@@ -2,7 +2,6 @@
 # Makefile for the drm device driver.  This driver provides support for the
 # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.
 
-ccflags-y = -Iinclude/drm
 sis-y := sis_drv.o sis_mm.o
 
 obj-$(CONFIG_DRM_SIS)   += sis.o
diff --git a/drivers/gpu/drm/sti/sti_compositor.c b/drivers/gpu/drm/sti/sti_compositor.c
index 11d4e885893a..6e4bf68262db 100644
--- a/drivers/gpu/drm/sti/sti_compositor.c
+++ b/drivers/gpu/drm/sti/sti_compositor.c
@@ -129,7 +129,7 @@ static int sti_compositor_bind(struct device *dev,
 			}
 			break;
 		default:
-			DRM_ERROR("Unknown subdev compoment type\n");
+			DRM_ERROR("Unknown subdev component type\n");
 			return 1;
 		}
 
diff --git a/drivers/gpu/drm/sti/sti_cursor.c b/drivers/gpu/drm/sti/sti_cursor.c
index cca75bddb9ad..5b3a41f74f21 100644
--- a/drivers/gpu/drm/sti/sti_cursor.c
+++ b/drivers/gpu/drm/sti/sti_cursor.c
@@ -33,7 +33,7 @@
 #define STI_CURS_MAX_SIZE   128
 
 /*
- * pixmap dma buffer stucture
+ * pixmap dma buffer structure
  *
  * @paddr:  physical address
  * @size:   buffer size
@@ -121,8 +121,7 @@ static int cursor_dbg_show(struct seq_file *s, void *data)
 	cursor_dbg_cml(s, cursor, readl(cursor->regs + CUR_CML));
 	DBGFS_DUMP(CUR_AWS);
 	DBGFS_DUMP(CUR_AWE);
-	seq_puts(s, "\n");
-
+	seq_putc(s, '\n');
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/sti/sti_dvo.c b/drivers/gpu/drm/sti/sti_dvo.c
index bb23318a44b7..24ebc6b2f34d 100644
--- a/drivers/gpu/drm/sti/sti_dvo.c
+++ b/drivers/gpu/drm/sti/sti_dvo.c
@@ -186,8 +186,7 @@ static int dvo_dbg_show(struct seq_file *s, void *data)
 	DBGFS_DUMP(DVO_LUT_PROG_MID);
 	DBGFS_DUMP(DVO_LUT_PROG_HIGH);
 	dvo_dbg_awg_microcode(s, dvo->regs + DVO_DIGSYNC_INSTR_I);
-	seq_puts(s, "\n");
-
+	seq_putc(s, '\n');
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/sti/sti_gdp.c b/drivers/gpu/drm/sti/sti_gdp.c
index 88f16cdf6a4b..5ee0503945c8 100644
--- a/drivers/gpu/drm/sti/sti_gdp.c
+++ b/drivers/gpu/drm/sti/sti_gdp.c
@@ -149,7 +149,7 @@ static void gdp_dbg_ctl(struct seq_file *s, int val)
 	seq_puts(s, "\tColor:");
 	for (i = 0; i < ARRAY_SIZE(gdp_format_to_str); i++) {
 		if (gdp_format_to_str[i].format == (val & 0x1F)) {
-			seq_printf(s, gdp_format_to_str[i].name);
+			seq_puts(s, gdp_format_to_str[i].name);
 			break;
 		}
 	}
@@ -266,8 +266,7 @@ static void gdp_node_dump_node(struct seq_file *s, struct sti_gdp_node *node)
 	seq_printf(s, "\n\tKEY2 0x%08X", node->gam_gdp_key2);
 	seq_printf(s, "\n\tPPT  0x%08X", node->gam_gdp_ppt);
 	gdp_dbg_ppt(s, node->gam_gdp_ppt);
-	seq_printf(s, "\n\tCML  0x%08X", node->gam_gdp_cml);
-	seq_puts(s, "\n");
+	seq_printf(s, "\n\tCML  0x%08X\n", node->gam_gdp_cml);
 }
 
 static int gdp_node_dbg_show(struct seq_file *s, void *arg)
diff --git a/drivers/gpu/drm/sti/sti_hda.c b/drivers/gpu/drm/sti/sti_hda.c
index 0c0a75bc8bc3..d6ed909d9d75 100644
--- a/drivers/gpu/drm/sti/sti_hda.c
+++ b/drivers/gpu/drm/sti/sti_hda.c
@@ -320,8 +320,7 @@ static void hda_dbg_awg_microcode(struct seq_file *s, void __iomem *reg)
 {
 	unsigned int i;
 
-	seq_puts(s, "\n\n");
-	seq_puts(s, "  HDA AWG microcode:");
+	seq_puts(s, "\n\n  HDA AWG microcode:");
 	for (i = 0; i < AWG_MAX_INST; i++) {
 		if (i % 8 == 0)
 			seq_printf(s, "\n  %04X:", i);
@@ -333,8 +332,7 @@ static void hda_dbg_video_dacs_ctrl(struct seq_file *s, void __iomem *reg)
 {
 	u32 val = readl(reg);
 
-	seq_puts(s, "\n");
-	seq_printf(s, "\n  %-25s 0x%08X", "VIDEO_DACS_CONTROL", val);
+	seq_printf(s, "\n\n  %-25s 0x%08X", "VIDEO_DACS_CONTROL", val);
 	seq_puts(s, "\tHD DACs ");
 	seq_puts(s, val & DAC_CFG_HD_HZUVW_OFF_MASK ? "disabled" : "enabled");
 }
@@ -356,8 +354,7 @@ static int hda_dbg_show(struct seq_file *s, void *data)
 	hda_dbg_awg_microcode(s, hda->regs + HDA_SYNC_AWGI);
 	if (hda->video_dacs_ctrl)
 		hda_dbg_video_dacs_ctrl(s, hda->video_dacs_ctrl);
-	seq_puts(s, "\n");
-
+	seq_putc(s, '\n');
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/sti/sti_hdmi.c b/drivers/gpu/drm/sti/sti_hdmi.c
index 243905b6ae59..a59c95a8081b 100644
--- a/drivers/gpu/drm/sti/sti_hdmi.c
+++ b/drivers/gpu/drm/sti/sti_hdmi.c
@@ -592,7 +592,7 @@ static void hdmi_dbg_cfg(struct seq_file *s, int val)
 {
 	int tmp;
 
-	seq_puts(s, "\t");
+	seq_putc(s, '\t');
 	tmp = val & HDMI_CFG_HDMI_NOT_DVI;
 	DBGFS_PRINT_STR("mode:", tmp ? "HDMI" : "DVI");
 	seq_puts(s, "\t\t\t\t\t");
@@ -616,7 +616,7 @@ static void hdmi_dbg_sta(struct seq_file *s, int val)
 {
 	int tmp;
 
-	seq_puts(s, "\t");
+	seq_putc(s, '\t');
 	tmp = (val & HDMI_STA_DLL_LCK);
 	DBGFS_PRINT_STR("pll:", tmp ? "locked" : "not locked");
 	seq_puts(s, "\t\t\t\t\t");
@@ -632,7 +632,7 @@ static void hdmi_dbg_sw_di_cfg(struct seq_file *s, int val)
 			       "once every field",
 			       "once every frame"};
 
-	seq_puts(s, "\t");
+	seq_putc(s, '\t');
 	tmp = (val & HDMI_IFRAME_CFG_DI_N(HDMI_IFRAME_MASK, 1));
 	DBGFS_PRINT_STR("Data island 1:", en_di[tmp]);
 	seq_puts(s, "\t\t\t\t\t");
@@ -664,16 +664,16 @@ static int hdmi_dbg_show(struct seq_file *s, void *data)
 	DBGFS_DUMP("\n", HDMI_STA);
 	hdmi_dbg_sta(s, hdmi_read(hdmi, HDMI_STA));
 	DBGFS_DUMP("", HDMI_ACTIVE_VID_XMIN);
-	seq_puts(s, "\t");
+	seq_putc(s, '\t');
 	DBGFS_PRINT_INT("Xmin:", hdmi_read(hdmi, HDMI_ACTIVE_VID_XMIN));
 	DBGFS_DUMP("", HDMI_ACTIVE_VID_XMAX);
-	seq_puts(s, "\t");
+	seq_putc(s, '\t');
 	DBGFS_PRINT_INT("Xmax:", hdmi_read(hdmi, HDMI_ACTIVE_VID_XMAX));
 	DBGFS_DUMP("", HDMI_ACTIVE_VID_YMIN);
-	seq_puts(s, "\t");
+	seq_putc(s, '\t');
 	DBGFS_PRINT_INT("Ymin:", hdmi_read(hdmi, HDMI_ACTIVE_VID_YMIN));
 	DBGFS_DUMP("", HDMI_ACTIVE_VID_YMAX);
-	seq_puts(s, "\t");
+	seq_putc(s, '\t');
 	DBGFS_PRINT_INT("Ymax:", hdmi_read(hdmi, HDMI_ACTIVE_VID_YMAX));
 	DBGFS_DUMP("", HDMI_SW_DI_CFG);
 	hdmi_dbg_sw_di_cfg(s, hdmi_read(hdmi, HDMI_SW_DI_CFG));
@@ -692,8 +692,7 @@ static int hdmi_dbg_show(struct seq_file *s, void *data)
 	DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD4, HDMI_IFRAME_SLOT_AVI);
 	DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD5, HDMI_IFRAME_SLOT_AVI);
 	DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD6, HDMI_IFRAME_SLOT_AVI);
-	seq_puts(s, "\n");
-	seq_printf(s, "\n AUDIO Infoframe (Data Island slot N=%d):",
+	seq_printf(s, "\n\n AUDIO Infoframe (Data Island slot N=%d):",
 		   HDMI_IFRAME_SLOT_AUDIO);
 	DBGFS_DUMP_DI(HDMI_SW_DI_N_HEAD_WORD, HDMI_IFRAME_SLOT_AUDIO);
 	DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD0, HDMI_IFRAME_SLOT_AUDIO);
@@ -703,8 +702,7 @@ static int hdmi_dbg_show(struct seq_file *s, void *data)
 	DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD4, HDMI_IFRAME_SLOT_AUDIO);
 	DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD5, HDMI_IFRAME_SLOT_AUDIO);
 	DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD6, HDMI_IFRAME_SLOT_AUDIO);
-	seq_puts(s, "\n");
-	seq_printf(s, "\n VENDOR SPECIFIC Infoframe (Data Island slot N=%d):",
+	seq_printf(s, "\n\n VENDOR SPECIFIC Infoframe (Data Island slot N=%d):",
 		   HDMI_IFRAME_SLOT_VENDOR);
 	DBGFS_DUMP_DI(HDMI_SW_DI_N_HEAD_WORD, HDMI_IFRAME_SLOT_VENDOR);
 	DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD0, HDMI_IFRAME_SLOT_VENDOR);
@@ -714,8 +712,7 @@ static int hdmi_dbg_show(struct seq_file *s, void *data)
 	DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD4, HDMI_IFRAME_SLOT_VENDOR);
 	DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD5, HDMI_IFRAME_SLOT_VENDOR);
 	DBGFS_DUMP_DI(HDMI_SW_DI_N_PKT_WORD6, HDMI_IFRAME_SLOT_VENDOR);
-	seq_puts(s, "\n");
-
+	seq_putc(s, '\n');
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/sti/sti_hqvdp.c b/drivers/gpu/drm/sti/sti_hqvdp.c
index 66f843148ef7..a1c161f77804 100644
--- a/drivers/gpu/drm/sti/sti_hqvdp.c
+++ b/drivers/gpu/drm/sti/sti_hqvdp.c
@@ -625,8 +625,7 @@ static int hqvdp_dbg_show(struct seq_file *s, void *data)
 		hqvdp_dbg_dump_cmd(s, (struct sti_hqvdp_cmd *)virt);
 	}
 
-	seq_puts(s, "\n");
-
+	seq_putc(s, '\n');
 	return 0;
 }
 
@@ -1357,12 +1356,12 @@ static int sti_hqvdp_probe(struct platform_device *pdev)
 
 	/* Get Memory resources */
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	if (res == NULL) {
+	if (!res) {
 		DRM_ERROR("Get memory resource failed\n");
 		return -ENXIO;
 	}
 	hqvdp->regs = devm_ioremap(dev, res->start, resource_size(res));
-	if (hqvdp->regs == NULL) {
+	if (!hqvdp->regs) {
 		DRM_ERROR("Register mapping failed\n");
 		return -ENXIO;
 	}
diff --git a/drivers/gpu/drm/sti/sti_mixer.c b/drivers/gpu/drm/sti/sti_mixer.c
index 4ddc58f7fe2e..2bd1d46fe1cd 100644
--- a/drivers/gpu/drm/sti/sti_mixer.c
+++ b/drivers/gpu/drm/sti/sti_mixer.c
@@ -162,8 +162,7 @@ static int mixer_dbg_show(struct seq_file *s, void *arg)
 	DBGFS_DUMP(GAM_MIXER_MBP);
 	DBGFS_DUMP(GAM_MIXER_MX0);
 	mixer_dbg_mxn(s, mixer->regs + GAM_MIXER_MX0);
-	seq_puts(s, "\n");
-
+	seq_putc(s, '\n');
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/sti/sti_tvout.c b/drivers/gpu/drm/sti/sti_tvout.c
index 8b8ea717c121..8959fcc743a8 100644
--- a/drivers/gpu/drm/sti/sti_tvout.c
+++ b/drivers/gpu/drm/sti/sti_tvout.c
@@ -459,7 +459,7 @@ static void tvout_dbg_vip(struct seq_file *s, int val)
 				   "Aux (color matrix by-passed)",
 				   "", "", "", "", "", "Force value"};
 
-	seq_puts(s, "\t");
+	seq_putc(s, '\t');
 	mask = TVO_VIP_REORDER_MASK << TVO_VIP_REORDER_R_SHIFT;
 	r = (val & mask) >> TVO_VIP_REORDER_R_SHIFT;
 	mask = TVO_VIP_REORDER_MASK << TVO_VIP_REORDER_G_SHIFT;
@@ -558,8 +558,7 @@ static int tvout_dbg_show(struct seq_file *s, void *data)
 	DBGFS_DUMP(TVO_CSC_AUX_M6);
 	DBGFS_DUMP(TVO_CSC_AUX_M7);
 	DBGFS_DUMP(TVO_AUX_IN_VID_FORMAT);
-	seq_puts(s, "\n");
-
+	seq_putc(s, '\n');
 	return 0;
 }
 
@@ -847,7 +846,7 @@ static int sti_tvout_probe(struct platform_device *pdev)
 
 	tvout->dev = dev;
 
-	/* get Memory ressources */
+	/* get memory resources */
 	res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "tvout-reg");
 	if (!res) {
 		DRM_ERROR("Invalid glue resource\n");
diff --git a/drivers/gpu/drm/sti/sti_vid.c b/drivers/gpu/drm/sti/sti_vid.c
index 2ad59892b57e..577a3341d3c1 100644
--- a/drivers/gpu/drm/sti/sti_vid.c
+++ b/drivers/gpu/drm/sti/sti_vid.c
@@ -61,7 +61,7 @@
 static void vid_dbg_ctl(struct seq_file *s, int val)
 {
 	val = val >> 30;
-	seq_puts(s, "\t");
+	seq_putc(s, '\t');
 
 	if (!(val & 1))
 		seq_puts(s, "NOT ");
@@ -114,8 +114,7 @@ static int vid_dbg_show(struct seq_file *s, void *arg)
 	DBGFS_DUMP(VID_BC);
 	DBGFS_DUMP(VID_TINT);
 	DBGFS_DUMP(VID_CSAT);
-	seq_puts(s, "\n");
-
+	seq_putc(s, '\n');
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/stm/Kconfig b/drivers/gpu/drm/stm/Kconfig
new file mode 100644
index 000000000000..2c4817fb0890
--- /dev/null
+++ b/drivers/gpu/drm/stm/Kconfig
@@ -0,0 +1,16 @@
+config DRM_STM
+	tristate "DRM Support for STMicroelectronics SoC Series"
+	depends on DRM && (ARCH_STM32 || ARCH_MULTIPLATFORM)
+	select DRM_KMS_HELPER
+	select DRM_GEM_CMA_HELPER
+	select DRM_KMS_CMA_HELPER
+	select DRM_PANEL
+	select VIDEOMODE_HELPERS
+	select FB_PROVIDE_GET_FB_UNMAPPED_AREA
+	default y
+
+	help
+	  Enable support for the on-chip display controller on
+	  STMicroelectronics STM32 MCUs.
+	  To compile this driver as a module, choose M here: the module
+	  will be called stm-drm.
diff --git a/drivers/gpu/drm/stm/Makefile b/drivers/gpu/drm/stm/Makefile
new file mode 100644
index 000000000000..a09ecf450218
--- /dev/null
+++ b/drivers/gpu/drm/stm/Makefile
@@ -0,0 +1,5 @@
+stm-drm-y := \
+	drv.o \
+	ltdc.o
+
+obj-$(CONFIG_DRM_STM) += stm-drm.o
diff --git a/drivers/gpu/drm/stm/drv.c b/drivers/gpu/drm/stm/drv.c
new file mode 100644
index 000000000000..83ab48f1fd00
--- /dev/null
+++ b/drivers/gpu/drm/stm/drv.c
@@ -0,0 +1,221 @@
+/*
+ * Copyright (C) STMicroelectronics SA 2017
+ *
+ * Authors: Philippe Cornu <philippe.cornu@st.com>
+ *          Yannick Fertre <yannick.fertre@st.com>
+ *          Fabien Dessenne <fabien.dessenne@st.com>
+ *          Mickael Reulier <mickael.reulier@st.com>
+ *
+ * License terms:  GNU General Public License (GPL), version 2
+ */
+
+#include <linux/component.h>
+#include <linux/of_platform.h>
+
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_fb_cma_helper.h>
+#include <drm/drm_gem_cma_helper.h>
+
+#include "ltdc.h"
+
+#define DRIVER_NAME		"stm"
+#define DRIVER_DESC		"STMicroelectronics SoC DRM"
+#define DRIVER_DATE		"20170330"
+#define DRIVER_MAJOR		1
+#define DRIVER_MINOR		0
+#define DRIVER_PATCH_LEVEL	0
+
+#define STM_MAX_FB_WIDTH	2048
+#define STM_MAX_FB_HEIGHT	2048 /* same as width to handle orientation */
+
+static void drv_output_poll_changed(struct drm_device *ddev)
+{
+	struct ltdc_device *ldev = ddev->dev_private;
+
+	drm_fbdev_cma_hotplug_event(ldev->fbdev);
+}
+
+static const struct drm_mode_config_funcs drv_mode_config_funcs = {
+	.fb_create = drm_fb_cma_create,
+	.output_poll_changed = drv_output_poll_changed,
+	.atomic_check = drm_atomic_helper_check,
+	.atomic_commit = drm_atomic_helper_commit,
+};
+
+static void drv_lastclose(struct drm_device *ddev)
+{
+	struct ltdc_device *ldev = ddev->dev_private;
+
+	DRM_DEBUG("%s\n", __func__);
+
+	drm_fbdev_cma_restore_mode(ldev->fbdev);
+}
+
+DEFINE_DRM_GEM_CMA_FOPS(drv_driver_fops);
+
+static struct drm_driver drv_driver = {
+	.driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_PRIME |
+			   DRIVER_ATOMIC,
+	.lastclose = drv_lastclose,
+	.name = DRIVER_NAME,
+	.desc = DRIVER_DESC,
+	.date = DRIVER_DATE,
+	.major = DRIVER_MAJOR,
+	.minor = DRIVER_MINOR,
+	.patchlevel = DRIVER_PATCH_LEVEL,
+	.fops = &drv_driver_fops,
+	.dumb_create = drm_gem_cma_dumb_create,
+	.dumb_map_offset = drm_gem_cma_dumb_map_offset,
+	.dumb_destroy = drm_gem_dumb_destroy,
+	.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
+	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
+	.gem_free_object_unlocked = drm_gem_cma_free_object,
+	.gem_vm_ops = &drm_gem_cma_vm_ops,
+	.gem_prime_export = drm_gem_prime_export,
+	.gem_prime_import = drm_gem_prime_import,
+	.gem_prime_get_sg_table = drm_gem_cma_prime_get_sg_table,
+	.gem_prime_import_sg_table = drm_gem_cma_prime_import_sg_table,
+	.gem_prime_vmap = drm_gem_cma_prime_vmap,
+	.gem_prime_vunmap = drm_gem_cma_prime_vunmap,
+	.gem_prime_mmap = drm_gem_cma_prime_mmap,
+	.enable_vblank = ltdc_crtc_enable_vblank,
+	.disable_vblank = ltdc_crtc_disable_vblank,
+};
+
+static int drv_load(struct drm_device *ddev)
+{
+	struct platform_device *pdev = to_platform_device(ddev->dev);
+	struct drm_fbdev_cma *fbdev;
+	struct ltdc_device *ldev;
+	int ret;
+
+	DRM_DEBUG("%s\n", __func__);
+
+	ldev = devm_kzalloc(ddev->dev, sizeof(*ldev), GFP_KERNEL);
+	if (!ldev)
+		return -ENOMEM;
+
+	ddev->dev_private = (void *)ldev;
+
+	drm_mode_config_init(ddev);
+
+	/*
+	 * set max width and height as default value.
+	 * this value would be used to check framebuffer size limitation
+	 * at drm_mode_addfb().
+	 */
+	ddev->mode_config.min_width = 0;
+	ddev->mode_config.min_height = 0;
+	ddev->mode_config.max_width = STM_MAX_FB_WIDTH;
+	ddev->mode_config.max_height = STM_MAX_FB_HEIGHT;
+	ddev->mode_config.funcs = &drv_mode_config_funcs;
+
+	ret = ltdc_load(ddev);
+	if (ret)
+		goto err;
+
+	drm_mode_config_reset(ddev);
+	drm_kms_helper_poll_init(ddev);
+
+	if (ddev->mode_config.num_connector) {
+		ldev = ddev->dev_private;
+		fbdev = drm_fbdev_cma_init(ddev, 16,
+					   ddev->mode_config.num_connector);
+		if (IS_ERR(fbdev)) {
+			DRM_DEBUG("Warning: fails to create fbdev\n");
+			fbdev = NULL;
+		}
+		ldev->fbdev = fbdev;
+	}
+
+	platform_set_drvdata(pdev, ddev);
+
+	return 0;
+err:
+	drm_mode_config_cleanup(ddev);
+	return ret;
+}
+
+static void drv_unload(struct drm_device *ddev)
+{
+	struct ltdc_device *ldev = ddev->dev_private;
+
+	DRM_DEBUG("%s\n", __func__);
+
+	if (ldev->fbdev) {
+		drm_fbdev_cma_fini(ldev->fbdev);
+		ldev->fbdev = NULL;
+	}
+	drm_kms_helper_poll_fini(ddev);
+	ltdc_unload(ddev);
+	drm_mode_config_cleanup(ddev);
+}
+
+static int stm_drm_platform_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct drm_device *ddev;
+	int ret;
+
+	DRM_DEBUG("%s\n", __func__);
+
+	dma_set_coherent_mask(dev, DMA_BIT_MASK(32));
+
+	ddev = drm_dev_alloc(&drv_driver, dev);
+	if (IS_ERR(ddev))
+		return PTR_ERR(ddev);
+
+	ret = drv_load(ddev);
+	if (ret)
+		goto err_unref;
+
+	ret = drm_dev_register(ddev, 0);
+	if (ret)
+		goto err_unref;
+
+	return 0;
+
+err_unref:
+	drm_dev_unref(ddev);
+
+	return ret;
+}
+
+static int stm_drm_platform_remove(struct platform_device *pdev)
+{
+	struct drm_device *ddev = platform_get_drvdata(pdev);
+
+	DRM_DEBUG("%s\n", __func__);
+
+	drm_dev_unregister(ddev);
+	drv_unload(ddev);
+	drm_dev_unref(ddev);
+
+	return 0;
+}
+
+static const struct of_device_id drv_dt_ids[] = {
+	{ .compatible = "st,stm32-ltdc"},
+	{ /* end node */ },
+};
+MODULE_DEVICE_TABLE(of, drv_dt_ids);
+
+static struct platform_driver stm_drm_platform_driver = {
+	.probe = stm_drm_platform_probe,
+	.remove = stm_drm_platform_remove,
+	.driver = {
+		.name = DRIVER_NAME,
+		.of_match_table = drv_dt_ids,
+	},
+};
+
+module_platform_driver(stm_drm_platform_driver);
+
+MODULE_AUTHOR("Philippe Cornu <philippe.cornu@st.com>");
+MODULE_AUTHOR("Yannick Fertre <yannick.fertre@st.com>");
+MODULE_AUTHOR("Fabien Dessenne <fabien.dessenne@st.com>");
+MODULE_AUTHOR("Mickael Reulier <mickael.reulier@st.com>");
+MODULE_DESCRIPTION("STMicroelectronics ST DRM LTDC driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/gpu/drm/stm/ltdc.c b/drivers/gpu/drm/stm/ltdc.c
new file mode 100644
index 000000000000..700cc0800e51
--- /dev/null
+++ b/drivers/gpu/drm/stm/ltdc.c
@@ -0,0 +1,1160 @@
+/*
+ * Copyright (C) STMicroelectronics SA 2017
+ *
+ * Authors: Philippe Cornu <philippe.cornu@st.com>
+ *          Yannick Fertre <yannick.fertre@st.com>
+ *          Fabien Dessenne <fabien.dessenne@st.com>
+ *          Mickael Reulier <mickael.reulier@st.com>
+ *
+ * License terms:  GNU General Public License (GPL), version 2
+ */
+
+#include <linux/clk.h>
+#include <linux/component.h>
+#include <linux/of_address.h>
+#include <linux/of_graph.h>
+#include <linux/reset.h>
+
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_fb_cma_helper.h>
+#include <drm/drm_gem_cma_helper.h>
+#include <drm/drm_of.h>
+#include <drm/drm_panel.h>
+#include <drm/drm_plane_helper.h>
+
+#include <video/videomode.h>
+
+#include "ltdc.h"
+
+#define NB_CRTC 1
+#define CRTC_MASK GENMASK(NB_CRTC - 1, 0)
+
+#define MAX_IRQ 4
+
+#define HWVER_10200 0x010200
+#define HWVER_10300 0x010300
+#define HWVER_20101 0x020101
+
+/*
+ * The address of some registers depends on the HW version: such registers have
+ * an extra offset specified with reg_ofs.
+ */
+#define REG_OFS_NONE	0
+#define REG_OFS_4	4 /* Insertion of "Layer Configuration 2" reg */
+#define REG_OFS		(ldev->caps.reg_ofs)
+#define LAY_OFS		0x80	/* Register Offset between 2 layers */
+
+/* Global register offsets */
+#define LTDC_IDR	0x0000 /* IDentification */
+#define LTDC_LCR	0x0004 /* Layer Count */
+#define LTDC_SSCR	0x0008 /* Synchronization Size Configuration */
+#define LTDC_BPCR	0x000C /* Back Porch Configuration */
+#define LTDC_AWCR	0x0010 /* Active Width Configuration */
+#define LTDC_TWCR	0x0014 /* Total Width Configuration */
+#define LTDC_GCR	0x0018 /* Global Control */
+#define LTDC_GC1R	0x001C /* Global Configuration 1 */
+#define LTDC_GC2R	0x0020 /* Global Configuration 2 */
+#define LTDC_SRCR	0x0024 /* Shadow Reload Configuration */
+#define LTDC_GACR	0x0028 /* GAmma Correction */
+#define LTDC_BCCR	0x002C /* Background Color Configuration */
+#define LTDC_IER	0x0034 /* Interrupt Enable */
+#define LTDC_ISR	0x0038 /* Interrupt Status */
+#define LTDC_ICR	0x003C /* Interrupt Clear */
+#define LTDC_LIPCR	0x0040 /* Line Interrupt Position Configuration */
+#define LTDC_CPSR	0x0044 /* Current Position Status */
+#define LTDC_CDSR	0x0048 /* Current Display Status */
+
+/* Layer register offsets */
+#define LTDC_L1LC1R	(0x0080)	   /* L1 Layer Configuration 1 */
+#define LTDC_L1LC2R	(0x0084)	   /* L1 Layer Configuration 2 */
+#define LTDC_L1CR	(0x0084 + REG_OFS) /* L1 Control */
+#define LTDC_L1WHPCR	(0x0088 + REG_OFS) /* L1 Window Hor Position Config */
+#define LTDC_L1WVPCR	(0x008C + REG_OFS) /* L1 Window Vert Position Config */
+#define LTDC_L1CKCR	(0x0090 + REG_OFS) /* L1 Color Keying Configuration */
+#define LTDC_L1PFCR	(0x0094 + REG_OFS) /* L1 Pixel Format Configuration */
+#define LTDC_L1CACR	(0x0098 + REG_OFS) /* L1 Constant Alpha Config */
+#define LTDC_L1DCCR	(0x009C + REG_OFS) /* L1 Default Color Configuration */
+#define LTDC_L1BFCR	(0x00A0 + REG_OFS) /* L1 Blend Factors Configuration */
+#define LTDC_L1FBBCR	(0x00A4 + REG_OFS) /* L1 FrameBuffer Bus Control */
+#define LTDC_L1AFBCR	(0x00A8 + REG_OFS) /* L1 AuxFB Control */
+#define LTDC_L1CFBAR	(0x00AC + REG_OFS) /* L1 Color FrameBuffer Address */
+#define LTDC_L1CFBLR	(0x00B0 + REG_OFS) /* L1 Color FrameBuffer Length */
+#define LTDC_L1CFBLNR	(0x00B4 + REG_OFS) /* L1 Color FrameBuffer Line Nb */
+#define LTDC_L1AFBAR	(0x00B8 + REG_OFS) /* L1 AuxFB Address */
+#define LTDC_L1AFBLR	(0x00BC + REG_OFS) /* L1 AuxFB Length */
+#define LTDC_L1AFBLNR	(0x00C0 + REG_OFS) /* L1 AuxFB Line Number */
+#define LTDC_L1CLUTWR	(0x00C4 + REG_OFS) /* L1 CLUT Write */
+#define LTDC_L1YS1R	(0x00E0 + REG_OFS) /* L1 YCbCr Scale 1 */
+#define LTDC_L1YS2R	(0x00E4 + REG_OFS) /* L1 YCbCr Scale 2 */
+
+/* Bit definitions */
+#define SSCR_VSH	GENMASK(10, 0)	/* Vertical Synchronization Height */
+#define SSCR_HSW	GENMASK(27, 16)	/* Horizontal Synchronization Width */
+
+#define BPCR_AVBP	GENMASK(10, 0)	/* Accumulated Vertical Back Porch */
+#define BPCR_AHBP	GENMASK(27, 16)	/* Accumulated Horizontal Back Porch */
+
+#define AWCR_AAH	GENMASK(10, 0)	/* Accumulated Active Height */
+#define AWCR_AAW	GENMASK(27, 16)	/* Accumulated Active Width */
+
+#define TWCR_TOTALH	GENMASK(10, 0)	/* TOTAL Height */
+#define TWCR_TOTALW	GENMASK(27, 16)	/* TOTAL Width */
+
+#define GCR_LTDCEN	BIT(0)		/* LTDC ENable */
+#define GCR_DEN		BIT(16)		/* Dither ENable */
+#define GCR_PCPOL	BIT(28)		/* Pixel Clock POLarity */
+#define GCR_DEPOL	BIT(29)		/* Data Enable POLarity */
+#define GCR_VSPOL	BIT(30)		/* Vertical Synchro POLarity */
+#define GCR_HSPOL	BIT(31)		/* Horizontal Synchro POLarity */
+
+#define GC1R_WBCH	GENMASK(3, 0)	/* Width of Blue CHannel output */
+#define GC1R_WGCH	GENMASK(7, 4)	/* Width of Green Channel output */
+#define GC1R_WRCH	GENMASK(11, 8)	/* Width of Red Channel output */
+#define GC1R_PBEN	BIT(12)		/* Precise Blending ENable */
+#define GC1R_DT		GENMASK(15, 14)	/* Dithering Technique */
+#define GC1R_GCT	GENMASK(19, 17)	/* Gamma Correction Technique */
+#define GC1R_SHREN	BIT(21)		/* SHadow Registers ENabled */
+#define GC1R_BCP	BIT(22)		/* Background Colour Programmable */
+#define GC1R_BBEN	BIT(23)		/* Background Blending ENabled */
+#define GC1R_LNIP	BIT(24)		/* Line Number IRQ Position */
+#define GC1R_TP		BIT(25)		/* Timing Programmable */
+#define GC1R_IPP	BIT(26)		/* IRQ Polarity Programmable */
+#define GC1R_SPP	BIT(27)		/* Sync Polarity Programmable */
+#define GC1R_DWP	BIT(28)		/* Dither Width Programmable */
+#define GC1R_STREN	BIT(29)		/* STatus Registers ENabled */
+#define GC1R_BMEN	BIT(31)		/* Blind Mode ENabled */
+
+#define GC2R_EDCA	BIT(0)		/* External Display Control Ability  */
+#define GC2R_STSAEN	BIT(1)		/* Slave Timing Sync Ability ENabled */
+#define GC2R_DVAEN	BIT(2)		/* Dual-View Ability ENabled */
+#define GC2R_DPAEN	BIT(3)		/* Dual-Port Ability ENabled */
+#define GC2R_BW		GENMASK(6, 4)	/* Bus Width (log2 of nb of bytes) */
+#define GC2R_EDCEN	BIT(7)		/* External Display Control ENabled */
+
+#define SRCR_IMR	BIT(0)		/* IMmediate Reload */
+#define SRCR_VBR	BIT(1)		/* Vertical Blanking Reload */
+
+#define BCCR_BCBLACK	0x00		/* Background Color BLACK */
+#define BCCR_BCBLUE	GENMASK(7, 0)	/* Background Color BLUE */
+#define BCCR_BCGREEN	GENMASK(15, 8)	/* Background Color GREEN */
+#define BCCR_BCRED	GENMASK(23, 16)	/* Background Color RED */
+#define BCCR_BCWHITE	GENMASK(23, 0)	/* Background Color WHITE */
+
+#define IER_LIE		BIT(0)		/* Line Interrupt Enable */
+#define IER_FUIE	BIT(1)		/* Fifo Underrun Interrupt Enable */
+#define IER_TERRIE	BIT(2)		/* Transfer ERRor Interrupt Enable */
+#define IER_RRIE	BIT(3)		/* Register Reload Interrupt enable */
+
+#define ISR_LIF		BIT(0)		/* Line Interrupt Flag */
+#define ISR_FUIF	BIT(1)		/* Fifo Underrun Interrupt Flag */
+#define ISR_TERRIF	BIT(2)		/* Transfer ERRor Interrupt Flag */
+#define ISR_RRIF	BIT(3)		/* Register Reload Interrupt Flag */
+
+#define LXCR_LEN	BIT(0)		/* Layer ENable */
+#define LXCR_COLKEN	BIT(1)		/* Color Keying Enable */
+#define LXCR_CLUTEN	BIT(4)		/* Color Look-Up Table ENable */
+
+#define LXWHPCR_WHSTPOS	GENMASK(11, 0)	/* Window Horizontal StarT POSition */
+#define LXWHPCR_WHSPPOS	GENMASK(27, 16)	/* Window Horizontal StoP POSition */
+
+#define LXWVPCR_WVSTPOS	GENMASK(10, 0)	/* Window Vertical StarT POSition */
+#define LXWVPCR_WVSPPOS	GENMASK(26, 16)	/* Window Vertical StoP POSition */
+
+#define LXPFCR_PF	GENMASK(2, 0)	/* Pixel Format */
+
+#define LXCACR_CONSTA	GENMASK(7, 0)	/* CONSTant Alpha */
+
+#define LXBFCR_BF2	GENMASK(2, 0)	/* Blending Factor 2 */
+#define LXBFCR_BF1	GENMASK(10, 8)	/* Blending Factor 1 */
+
+#define LXCFBLR_CFBLL	GENMASK(12, 0)	/* Color Frame Buffer Line Length */
+#define LXCFBLR_CFBP	GENMASK(28, 16)	/* Color Frame Buffer Pitch in bytes */
+
+#define LXCFBLNR_CFBLN	GENMASK(10, 0)	 /* Color Frame Buffer Line Number */
+
+#define HSPOL_AL   0		/* Horizontal Sync POLarity Active Low */
+#define VSPOL_AL   0		/* Vertical Sync POLarity Active Low */
+#define DEPOL_AL   0		/* Data Enable POLarity Active Low */
+#define PCPOL_IPC  0		/* Input Pixel Clock */
+#define HSPOL_AH   GCR_HSPOL	/* Horizontal Sync POLarity Active High */
+#define VSPOL_AH   GCR_VSPOL	/* Vertical Sync POLarity Active High */
+#define DEPOL_AH   GCR_DEPOL	/* Data Enable POLarity Active High */
+#define PCPOL_IIPC GCR_PCPOL	/* Inverted Input Pixel Clock */
+#define CONSTA_MAX 0xFF		/* CONSTant Alpha MAX= 1.0 */
+#define BF1_PAXCA  0x600	/* Pixel Alpha x Constant Alpha */
+#define BF1_CA     0x400	/* Constant Alpha */
+#define BF2_1PAXCA 0x007	/* 1 - (Pixel Alpha x Constant Alpha) */
+#define BF2_1CA	   0x005	/* 1 - Constant Alpha */
+
+#define NB_PF           8       /* Max nb of HW pixel format */
+
+enum ltdc_pix_fmt {
+	PF_NONE,
+	/* RGB formats */
+	PF_ARGB8888,    /* ARGB [32 bits] */
+	PF_RGBA8888,    /* RGBA [32 bits] */
+	PF_RGB888,      /* RGB [24 bits] */
+	PF_RGB565,      /* RGB [16 bits] */
+	PF_ARGB1555,    /* ARGB A:1 bit RGB:15 bits [16 bits] */
+	PF_ARGB4444,    /* ARGB A:4 bits R/G/B: 4 bits each [16 bits] */
+	/* Indexed formats */
+	PF_L8,          /* Indexed 8 bits [8 bits] */
+	PF_AL44,        /* Alpha:4 bits + indexed 4 bits [8 bits] */
+	PF_AL88         /* Alpha:8 bits + indexed 8 bits [16 bits] */
+};
+
+/* The index gives the encoding of the pixel format for an HW version */
+static const enum ltdc_pix_fmt ltdc_pix_fmt_a0[NB_PF] = {
+	PF_ARGB8888,	/* 0x00 */
+	PF_RGB888,	/* 0x01 */
+	PF_RGB565,	/* 0x02 */
+	PF_ARGB1555,	/* 0x03 */
+	PF_ARGB4444,	/* 0x04 */
+	PF_L8,		/* 0x05 */
+	PF_AL44,	/* 0x06 */
+	PF_AL88		/* 0x07 */
+};
+
+static const enum ltdc_pix_fmt ltdc_pix_fmt_a1[NB_PF] = {
+	PF_ARGB8888,	/* 0x00 */
+	PF_RGB888,	/* 0x01 */
+	PF_RGB565,	/* 0x02 */
+	PF_RGBA8888,	/* 0x03 */
+	PF_AL44,	/* 0x04 */
+	PF_L8,		/* 0x05 */
+	PF_ARGB1555,	/* 0x06 */
+	PF_ARGB4444	/* 0x07 */
+};
+
+static inline u32 reg_read(void __iomem *base, u32 reg)
+{
+	return readl_relaxed(base + reg);
+}
+
+static inline void reg_write(void __iomem *base, u32 reg, u32 val)
+{
+	writel_relaxed(val, base + reg);
+}
+
+static inline void reg_set(void __iomem *base, u32 reg, u32 mask)
+{
+	reg_write(base, reg, reg_read(base, reg) | mask);
+}
+
+static inline void reg_clear(void __iomem *base, u32 reg, u32 mask)
+{
+	reg_write(base, reg, reg_read(base, reg) & ~mask);
+}
+
+static inline void reg_update_bits(void __iomem *base, u32 reg, u32 mask,
+				   u32 val)
+{
+	reg_write(base, reg, (reg_read(base, reg) & ~mask) | val);
+}
+
+static inline struct ltdc_device *crtc_to_ltdc(struct drm_crtc *crtc)
+{
+	return (struct ltdc_device *)crtc->dev->dev_private;
+}
+
+static inline struct ltdc_device *plane_to_ltdc(struct drm_plane *plane)
+{
+	return (struct ltdc_device *)plane->dev->dev_private;
+}
+
+static inline struct ltdc_device *encoder_to_ltdc(struct drm_encoder *enc)
+{
+	return (struct ltdc_device *)enc->dev->dev_private;
+}
+
+static inline struct ltdc_device *connector_to_ltdc(struct drm_connector *con)
+{
+	return (struct ltdc_device *)con->dev->dev_private;
+}
+
+static inline enum ltdc_pix_fmt to_ltdc_pixelformat(u32 drm_fmt)
+{
+	enum ltdc_pix_fmt pf;
+
+	switch (drm_fmt) {
+	case DRM_FORMAT_ARGB8888:
+	case DRM_FORMAT_XRGB8888:
+		pf = PF_ARGB8888;
+		break;
+	case DRM_FORMAT_RGBA8888:
+	case DRM_FORMAT_RGBX8888:
+		pf = PF_RGBA8888;
+		break;
+	case DRM_FORMAT_RGB888:
+		pf = PF_RGB888;
+		break;
+	case DRM_FORMAT_RGB565:
+		pf = PF_RGB565;
+		break;
+	case DRM_FORMAT_ARGB1555:
+	case DRM_FORMAT_XRGB1555:
+		pf = PF_ARGB1555;
+		break;
+	case DRM_FORMAT_ARGB4444:
+	case DRM_FORMAT_XRGB4444:
+		pf = PF_ARGB4444;
+		break;
+	case DRM_FORMAT_C8:
+		pf = PF_L8;
+		break;
+	default:
+		pf = PF_NONE;
+		break;
+	/* Note: There are no DRM_FORMAT for AL44 and AL88 */
+	}
+
+	return pf;
+}
+
+static inline u32 to_drm_pixelformat(enum ltdc_pix_fmt pf)
+{
+	switch (pf) {
+	case PF_ARGB8888:
+		return DRM_FORMAT_ARGB8888;
+	case PF_RGBA8888:
+		return DRM_FORMAT_RGBA8888;
+	case PF_RGB888:
+		return DRM_FORMAT_RGB888;
+	case PF_RGB565:
+		return DRM_FORMAT_RGB565;
+	case PF_ARGB1555:
+		return DRM_FORMAT_ARGB1555;
+	case PF_ARGB4444:
+		return DRM_FORMAT_ARGB4444;
+	case PF_L8:
+		return DRM_FORMAT_C8;
+	case PF_AL44: /* No DRM support */
+	case PF_AL88: /* No DRM support */
+	case PF_NONE:
+	default:
+		return 0;
+	}
+}
+
+static irqreturn_t ltdc_irq_thread(int irq, void *arg)
+{
+	struct drm_device *ddev = arg;
+	struct ltdc_device *ldev = ddev->dev_private;
+	struct drm_crtc *crtc = drm_crtc_from_index(ddev, 0);
+
+	/* Line IRQ : trigger the vblank event */
+	if (ldev->irq_status & ISR_LIF)
+		drm_crtc_handle_vblank(crtc);
+
+	/* Save FIFO Underrun & Transfer Error status */
+	mutex_lock(&ldev->err_lock);
+	if (ldev->irq_status & ISR_FUIF)
+		ldev->error_status |= ISR_FUIF;
+	if (ldev->irq_status & ISR_TERRIF)
+		ldev->error_status |= ISR_TERRIF;
+	mutex_unlock(&ldev->err_lock);
+
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t ltdc_irq(int irq, void *arg)
+{
+	struct drm_device *ddev = arg;
+	struct ltdc_device *ldev = ddev->dev_private;
+
+	/* Read & Clear the interrupt status */
+	ldev->irq_status = reg_read(ldev->regs, LTDC_ISR);
+	reg_write(ldev->regs, LTDC_ICR, ldev->irq_status);
+
+	return IRQ_WAKE_THREAD;
+}
+
+/*
+ * DRM_CRTC
+ */
+
+static void ltdc_crtc_load_lut(struct drm_crtc *crtc)
+{
+	struct ltdc_device *ldev = crtc_to_ltdc(crtc);
+	unsigned int i, lay;
+
+	for (lay = 0; lay < ldev->caps.nb_layers; lay++)
+		for (i = 0; i < 256; i++)
+			reg_write(ldev->regs, LTDC_L1CLUTWR + lay * LAY_OFS,
+				  ldev->clut[i]);
+}
+
+static void ltdc_crtc_enable(struct drm_crtc *crtc)
+{
+	struct ltdc_device *ldev = crtc_to_ltdc(crtc);
+
+	DRM_DEBUG_DRIVER("\n");
+
+	/* Sets the background color value */
+	reg_write(ldev->regs, LTDC_BCCR, BCCR_BCBLACK);
+
+	/* Enable IRQ */
+	reg_set(ldev->regs, LTDC_IER, IER_RRIE | IER_FUIE | IER_TERRIE);
+
+	/* Immediately commit the planes */
+	reg_set(ldev->regs, LTDC_SRCR, SRCR_IMR);
+
+	/* Enable LTDC */
+	reg_set(ldev->regs, LTDC_GCR, GCR_LTDCEN);
+
+	drm_crtc_vblank_on(crtc);
+}
+
+static void ltdc_crtc_disable(struct drm_crtc *crtc)
+{
+	struct ltdc_device *ldev = crtc_to_ltdc(crtc);
+
+	DRM_DEBUG_DRIVER("\n");
+
+	drm_crtc_vblank_off(crtc);
+
+	/* disable LTDC */
+	reg_clear(ldev->regs, LTDC_GCR, GCR_LTDCEN);
+
+	/* disable IRQ */
+	reg_clear(ldev->regs, LTDC_IER, IER_RRIE | IER_FUIE | IER_TERRIE);
+
+	/* immediately commit disable of layers before switching off LTDC */
+	reg_set(ldev->regs, LTDC_SRCR, SRCR_IMR);
+}
+
+static void ltdc_crtc_mode_set_nofb(struct drm_crtc *crtc)
+{
+	struct ltdc_device *ldev = crtc_to_ltdc(crtc);
+	struct drm_display_mode *mode = &crtc->state->adjusted_mode;
+	struct videomode vm;
+	int rate = mode->clock * 1000;
+	u32 hsync, vsync, accum_hbp, accum_vbp, accum_act_w, accum_act_h;
+	u32 total_width, total_height;
+	u32 val;
+
+	drm_display_mode_to_videomode(mode, &vm);
+
+	DRM_DEBUG_DRIVER("CRTC:%d mode:%s\n", crtc->base.id, mode->name);
+	DRM_DEBUG_DRIVER("Video mode: %dx%d", vm.hactive, vm.vactive);
+	DRM_DEBUG_DRIVER(" hfp %d hbp %d hsl %d vfp %d vbp %d vsl %d\n",
+			 vm.hfront_porch, vm.hback_porch, vm.hsync_len,
+			 vm.vfront_porch, vm.vback_porch, vm.vsync_len);
+
+	/* Convert video timings to ltdc timings */
+	hsync = vm.hsync_len - 1;
+	vsync = vm.vsync_len - 1;
+	accum_hbp = hsync + vm.hback_porch;
+	accum_vbp = vsync + vm.vback_porch;
+	accum_act_w = accum_hbp + vm.hactive;
+	accum_act_h = accum_vbp + vm.vactive;
+	total_width = accum_act_w + vm.hfront_porch;
+	total_height = accum_act_h + vm.vfront_porch;
+
+	clk_disable(ldev->pixel_clk);
+
+	if (clk_set_rate(ldev->pixel_clk, rate) < 0) {
+		DRM_ERROR("Cannot set rate (%dHz) for pixel clk\n", rate);
+		return;
+	}
+
+	clk_enable(ldev->pixel_clk);
+
+	/* Configures the HS, VS, DE and PC polarities. */
+	val = HSPOL_AL | VSPOL_AL | DEPOL_AL | PCPOL_IPC;
+
+	if (vm.flags & DISPLAY_FLAGS_HSYNC_HIGH)
+		val |= HSPOL_AH;
+
+	if (vm.flags & DISPLAY_FLAGS_VSYNC_HIGH)
+		val |= VSPOL_AH;
+
+	if (vm.flags & DISPLAY_FLAGS_DE_HIGH)
+		val |= DEPOL_AH;
+
+	if (vm.flags & DISPLAY_FLAGS_PIXDATA_NEGEDGE)
+		val |= PCPOL_IIPC;
+
+	reg_update_bits(ldev->regs, LTDC_GCR,
+			GCR_HSPOL | GCR_VSPOL | GCR_DEPOL | GCR_PCPOL, val);
+
+	/* Set Synchronization size */
+	val = (hsync << 16) | vsync;
+	reg_update_bits(ldev->regs, LTDC_SSCR, SSCR_VSH | SSCR_HSW, val);
+
+	/* Set Accumulated Back porch */
+	val = (accum_hbp << 16) | accum_vbp;
+	reg_update_bits(ldev->regs, LTDC_BPCR, BPCR_AVBP | BPCR_AHBP, val);
+
+	/* Set Accumulated Active Width */
+	val = (accum_act_w << 16) | accum_act_h;
+	reg_update_bits(ldev->regs, LTDC_AWCR, AWCR_AAW | AWCR_AAH, val);
+
+	/* Set total width & height */
+	val = (total_width << 16) | total_height;
+	reg_update_bits(ldev->regs, LTDC_TWCR, TWCR_TOTALH | TWCR_TOTALW, val);
+
+	reg_write(ldev->regs, LTDC_LIPCR, (accum_act_h + 1));
+}
+
+static void ltdc_crtc_atomic_flush(struct drm_crtc *crtc,
+				   struct drm_crtc_state *old_crtc_state)
+{
+	struct ltdc_device *ldev = crtc_to_ltdc(crtc);
+	struct drm_pending_vblank_event *event = crtc->state->event;
+
+	DRM_DEBUG_ATOMIC("\n");
+
+	/* Commit shadow registers = update planes at next vblank */
+	reg_set(ldev->regs, LTDC_SRCR, SRCR_VBR);
+
+	if (event) {
+		crtc->state->event = NULL;
+
+		spin_lock_irq(&crtc->dev->event_lock);
+		if (drm_crtc_vblank_get(crtc) == 0)
+			drm_crtc_arm_vblank_event(crtc, event);
+		else
+			drm_crtc_send_vblank_event(crtc, event);
+		spin_unlock_irq(&crtc->dev->event_lock);
+	}
+}
+
+static struct drm_crtc_helper_funcs ltdc_crtc_helper_funcs = {
+	.load_lut = ltdc_crtc_load_lut,
+	.enable = ltdc_crtc_enable,
+	.disable = ltdc_crtc_disable,
+	.mode_set_nofb = ltdc_crtc_mode_set_nofb,
+	.atomic_flush = ltdc_crtc_atomic_flush,
+};
+
+int ltdc_crtc_enable_vblank(struct drm_device *ddev, unsigned int pipe)
+{
+	struct ltdc_device *ldev = ddev->dev_private;
+
+	DRM_DEBUG_DRIVER("\n");
+	reg_set(ldev->regs, LTDC_IER, IER_LIE);
+
+	return 0;
+}
+
+void ltdc_crtc_disable_vblank(struct drm_device *ddev, unsigned int pipe)
+{
+	struct ltdc_device *ldev = ddev->dev_private;
+
+	DRM_DEBUG_DRIVER("\n");
+	reg_clear(ldev->regs, LTDC_IER, IER_LIE);
+}
+
+static struct drm_crtc_funcs ltdc_crtc_funcs = {
+	.destroy = drm_crtc_cleanup,
+	.set_config = drm_atomic_helper_set_config,
+	.page_flip = drm_atomic_helper_page_flip,
+	.reset = drm_atomic_helper_crtc_reset,
+	.atomic_duplicate_state = drm_atomic_helper_crtc_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_crtc_destroy_state,
+};
+
+/*
+ * DRM_PLANE
+ */
+
+static int ltdc_plane_atomic_check(struct drm_plane *plane,
+				   struct drm_plane_state *state)
+{
+	struct drm_framebuffer *fb = state->fb;
+	u32 src_x, src_y, src_w, src_h;
+
+	DRM_DEBUG_DRIVER("\n");
+
+	if (!fb)
+		return 0;
+
+	/* convert src_ from 16:16 format */
+	src_x = state->src_x >> 16;
+	src_y = state->src_y >> 16;
+	src_w = state->src_w >> 16;
+	src_h = state->src_h >> 16;
+
+	/* Reject scaling */
+	if ((src_w != state->crtc_w) || (src_h != state->crtc_h)) {
+		DRM_ERROR("Scaling is not supported");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static void ltdc_plane_atomic_update(struct drm_plane *plane,
+				     struct drm_plane_state *oldstate)
+{
+	struct ltdc_device *ldev = plane_to_ltdc(plane);
+	struct drm_plane_state *state = plane->state;
+	struct drm_framebuffer *fb = state->fb;
+	u32 lofs = plane->index * LAY_OFS;
+	u32 x0 = state->crtc_x;
+	u32 x1 = state->crtc_x + state->crtc_w - 1;
+	u32 y0 = state->crtc_y;
+	u32 y1 = state->crtc_y + state->crtc_h - 1;
+	u32 src_x, src_y, src_w, src_h;
+	u32 val, pitch_in_bytes, line_length, paddr, ahbp, avbp, bpcr;
+	enum ltdc_pix_fmt pf;
+
+	if (!state->crtc || !fb) {
+		DRM_DEBUG_DRIVER("fb or crtc NULL");
+		return;
+	}
+
+	/* convert src_ from 16:16 format */
+	src_x = state->src_x >> 16;
+	src_y = state->src_y >> 16;
+	src_w = state->src_w >> 16;
+	src_h = state->src_h >> 16;
+
+	DRM_DEBUG_DRIVER(
+		"plane:%d fb:%d (%dx%d)@(%d,%d) -> (%dx%d)@(%d,%d)\n",
+		plane->base.id, fb->base.id,
+		src_w, src_h, src_x, src_y,
+		state->crtc_w, state->crtc_h, state->crtc_x, state->crtc_y);
+
+	bpcr = reg_read(ldev->regs, LTDC_BPCR);
+	ahbp = (bpcr & BPCR_AHBP) >> 16;
+	avbp = bpcr & BPCR_AVBP;
+
+	/* Configures the horizontal start and stop position */
+	val = ((x1 + 1 + ahbp) << 16) + (x0 + 1 + ahbp);
+	reg_update_bits(ldev->regs, LTDC_L1WHPCR + lofs,
+			LXWHPCR_WHSTPOS | LXWHPCR_WHSPPOS, val);
+
+	/* Configures the vertical start and stop position */
+	val = ((y1 + 1 + avbp) << 16) + (y0 + 1 + avbp);
+	reg_update_bits(ldev->regs, LTDC_L1WVPCR + lofs,
+			LXWVPCR_WVSTPOS | LXWVPCR_WVSPPOS, val);
+
+	/* Specifies the pixel format */
+	pf = to_ltdc_pixelformat(fb->format->format);
+	for (val = 0; val < NB_PF; val++)
+		if (ldev->caps.pix_fmt_hw[val] == pf)
+			break;
+
+	if (val == NB_PF) {
+		DRM_ERROR("Pixel format %.4s not supported\n",
+			  (char *)&fb->format->format);
+		val = 0; /* set by default ARGB 32 bits */
+	}
+	reg_update_bits(ldev->regs, LTDC_L1PFCR + lofs, LXPFCR_PF, val);
+
+	/* Configures the color frame buffer pitch in bytes & line length */
+	pitch_in_bytes = fb->pitches[0];
+	line_length = drm_format_plane_cpp(fb->format->format, 0) *
+		      (x1 - x0 + 1) + (ldev->caps.bus_width >> 3) - 1;
+	val = ((pitch_in_bytes << 16) | line_length);
+	reg_update_bits(ldev->regs, LTDC_L1CFBLR + lofs,
+			LXCFBLR_CFBLL | LXCFBLR_CFBP, val);
+
+	/* Specifies the constant alpha value */
+	val = CONSTA_MAX;
+	reg_update_bits(ldev->regs, LTDC_L1CACR + lofs,
+			LXCACR_CONSTA, val);
+
+	/* Specifies the blending factors */
+	val = BF1_PAXCA | BF2_1PAXCA;
+	reg_update_bits(ldev->regs, LTDC_L1BFCR + lofs,
+			LXBFCR_BF2 | LXBFCR_BF1, val);
+
+	/* Configures the frame buffer line number */
+	val = y1 - y0 + 1;
+	reg_update_bits(ldev->regs, LTDC_L1CFBLNR + lofs,
+			LXCFBLNR_CFBLN, val);
+
+	/* Sets the FB address */
+	paddr = (u32)drm_fb_cma_get_gem_addr(fb, state, 0);
+
+	DRM_DEBUG_DRIVER("fb: phys 0x%08x", paddr);
+	reg_write(ldev->regs, LTDC_L1CFBAR + lofs, paddr);
+
+	/* Enable layer and CLUT if needed */
+	val = fb->format->format == DRM_FORMAT_C8 ? LXCR_CLUTEN : 0;
+	val |= LXCR_LEN;
+	reg_update_bits(ldev->regs, LTDC_L1CR + lofs,
+			LXCR_LEN | LXCR_CLUTEN, val);
+
+	mutex_lock(&ldev->err_lock);
+	if (ldev->error_status & ISR_FUIF) {
+		DRM_DEBUG_DRIVER("Fifo underrun\n");
+		ldev->error_status &= ~ISR_FUIF;
+	}
+	if (ldev->error_status & ISR_TERRIF) {
+		DRM_DEBUG_DRIVER("Transfer error\n");
+		ldev->error_status &= ~ISR_TERRIF;
+	}
+	mutex_unlock(&ldev->err_lock);
+}
+
+static void ltdc_plane_atomic_disable(struct drm_plane *plane,
+				      struct drm_plane_state *oldstate)
+{
+	struct ltdc_device *ldev = plane_to_ltdc(plane);
+	u32 lofs = plane->index * LAY_OFS;
+
+	/* disable layer */
+	reg_clear(ldev->regs, LTDC_L1CR + lofs, LXCR_LEN);
+
+	DRM_DEBUG_DRIVER("CRTC:%d plane:%d\n",
+			 oldstate->crtc->base.id, plane->base.id);
+}
+
+static struct drm_plane_funcs ltdc_plane_funcs = {
+	.update_plane = drm_atomic_helper_update_plane,
+	.disable_plane = drm_atomic_helper_disable_plane,
+	.destroy = drm_plane_cleanup,
+	.set_property = drm_atomic_helper_plane_set_property,
+	.reset = drm_atomic_helper_plane_reset,
+	.atomic_duplicate_state = drm_atomic_helper_plane_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_plane_destroy_state,
+};
+
+static const struct drm_plane_helper_funcs ltdc_plane_helper_funcs = {
+	.atomic_check = ltdc_plane_atomic_check,
+	.atomic_update = ltdc_plane_atomic_update,
+	.atomic_disable = ltdc_plane_atomic_disable,
+};
+
+static struct drm_plane *ltdc_plane_create(struct drm_device *ddev,
+					   enum drm_plane_type type)
+{
+	unsigned long possible_crtcs = CRTC_MASK;
+	struct ltdc_device *ldev = ddev->dev_private;
+	struct device *dev = ddev->dev;
+	struct drm_plane *plane;
+	unsigned int i, nb_fmt = 0;
+	u32 formats[NB_PF];
+	u32 drm_fmt;
+	int ret;
+
+	/* Get supported pixel formats */
+	for (i = 0; i < NB_PF; i++) {
+		drm_fmt = to_drm_pixelformat(ldev->caps.pix_fmt_hw[i]);
+		if (!drm_fmt)
+			continue;
+		formats[nb_fmt++] = drm_fmt;
+	}
+
+	plane = devm_kzalloc(dev, sizeof(*plane), GFP_KERNEL);
+	if (!plane)
+		return 0;
+
+	ret = drm_universal_plane_init(ddev, plane, possible_crtcs,
+				       &ltdc_plane_funcs, formats, nb_fmt,
+				       type, NULL);
+	if (ret < 0)
+		return 0;
+
+	drm_plane_helper_add(plane, &ltdc_plane_helper_funcs);
+
+	DRM_DEBUG_DRIVER("plane:%d created\n", plane->base.id);
+
+	return plane;
+}
+
+static void ltdc_plane_destroy_all(struct drm_device *ddev)
+{
+	struct drm_plane *plane, *plane_temp;
+
+	list_for_each_entry_safe(plane, plane_temp,
+				 &ddev->mode_config.plane_list, head)
+		drm_plane_cleanup(plane);
+}
+
+static int ltdc_crtc_init(struct drm_device *ddev, struct drm_crtc *crtc)
+{
+	struct ltdc_device *ldev = ddev->dev_private;
+	struct drm_plane *primary, *overlay;
+	unsigned int i;
+	int res;
+
+	primary = ltdc_plane_create(ddev, DRM_PLANE_TYPE_PRIMARY);
+	if (!primary) {
+		DRM_ERROR("Can not create primary plane\n");
+		return -EINVAL;
+	}
+
+	res = drm_crtc_init_with_planes(ddev, crtc, primary, NULL,
+					&ltdc_crtc_funcs, NULL);
+	if (res) {
+		DRM_ERROR("Can not initialize CRTC\n");
+		goto cleanup;
+	}
+
+	drm_crtc_helper_add(crtc, &ltdc_crtc_helper_funcs);
+
+	DRM_DEBUG_DRIVER("CRTC:%d created\n", crtc->base.id);
+
+	/* Add planes. Note : the first layer is used by primary plane */
+	for (i = 1; i < ldev->caps.nb_layers; i++) {
+		overlay = ltdc_plane_create(ddev, DRM_PLANE_TYPE_OVERLAY);
+		if (!overlay) {
+			res = -ENOMEM;
+			DRM_ERROR("Can not create overlay plane %d\n", i);
+			goto cleanup;
+		}
+	}
+
+	return 0;
+
+cleanup:
+	ltdc_plane_destroy_all(ddev);
+	return res;
+}
+
+/*
+ * DRM_ENCODER
+ */
+
+static void ltdc_rgb_encoder_enable(struct drm_encoder *encoder)
+{
+	struct ltdc_device *ldev = encoder_to_ltdc(encoder);
+
+	DRM_DEBUG_DRIVER("\n");
+
+	drm_panel_prepare(ldev->panel);
+	drm_panel_enable(ldev->panel);
+}
+
+static void ltdc_rgb_encoder_disable(struct drm_encoder *encoder)
+{
+	struct ltdc_device *ldev = encoder_to_ltdc(encoder);
+
+	DRM_DEBUG_DRIVER("\n");
+
+	drm_panel_disable(ldev->panel);
+	drm_panel_unprepare(ldev->panel);
+}
+
+static const struct drm_encoder_helper_funcs ltdc_rgb_encoder_helper_funcs = {
+	.enable = ltdc_rgb_encoder_enable,
+	.disable = ltdc_rgb_encoder_disable,
+};
+
+static const struct drm_encoder_funcs ltdc_rgb_encoder_funcs = {
+	.destroy = drm_encoder_cleanup,
+};
+
+static struct drm_encoder *ltdc_rgb_encoder_create(struct drm_device *ddev)
+{
+	struct drm_encoder *encoder;
+
+	encoder = devm_kzalloc(ddev->dev, sizeof(*encoder), GFP_KERNEL);
+	if (!encoder)
+		return NULL;
+
+	encoder->possible_crtcs = CRTC_MASK;
+	encoder->possible_clones = 0; /* No cloning support */
+
+	drm_encoder_init(ddev, encoder, &ltdc_rgb_encoder_funcs,
+			 DRM_MODE_ENCODER_DPI, NULL);
+
+	drm_encoder_helper_add(encoder, &ltdc_rgb_encoder_helper_funcs);
+
+	DRM_DEBUG_DRIVER("RGB encoder:%d created\n", encoder->base.id);
+
+	return encoder;
+}
+
+/*
+ * DRM_CONNECTOR
+ */
+
+static int ltdc_rgb_connector_get_modes(struct drm_connector *connector)
+{
+	struct drm_device *ddev = connector->dev;
+	struct ltdc_device *ldev = ddev->dev_private;
+	int ret = 0;
+
+	DRM_DEBUG_DRIVER("\n");
+
+	if (ldev->panel)
+		ret = drm_panel_get_modes(ldev->panel);
+
+	return ret < 0 ? 0 : ret;
+}
+
+static struct drm_connector_helper_funcs ltdc_rgb_connector_helper_funcs = {
+	.get_modes = ltdc_rgb_connector_get_modes,
+};
+
+static enum drm_connector_status
+ltdc_rgb_connector_detect(struct drm_connector *connector, bool force)
+{
+	struct ltdc_device *ldev = connector_to_ltdc(connector);
+
+	return ldev->panel ? connector_status_connected :
+	       connector_status_disconnected;
+}
+
+static void ltdc_rgb_connector_destroy(struct drm_connector *connector)
+{
+	DRM_DEBUG_DRIVER("\n");
+
+	drm_connector_unregister(connector);
+	drm_connector_cleanup(connector);
+}
+
+static const struct drm_connector_funcs ltdc_rgb_connector_funcs = {
+	.dpms = drm_atomic_helper_connector_dpms,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.detect = ltdc_rgb_connector_detect,
+	.destroy = ltdc_rgb_connector_destroy,
+	.reset = drm_atomic_helper_connector_reset,
+	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
+};
+
+struct drm_connector *ltdc_rgb_connector_create(struct drm_device *ddev)
+{
+	struct drm_connector *connector;
+	int err;
+
+	connector = devm_kzalloc(ddev->dev, sizeof(*connector), GFP_KERNEL);
+	if (!connector) {
+		DRM_ERROR("Failed to allocate connector\n");
+		return NULL;
+	}
+
+	connector->polled = DRM_CONNECTOR_POLL_HPD;
+
+	err = drm_connector_init(ddev, connector, &ltdc_rgb_connector_funcs,
+				 DRM_MODE_CONNECTOR_DPI);
+	if (err) {
+		DRM_ERROR("Failed to initialize connector\n");
+		return NULL;
+	}
+
+	drm_connector_helper_add(connector, &ltdc_rgb_connector_helper_funcs);
+
+	DRM_DEBUG_DRIVER("RGB connector:%d created\n", connector->base.id);
+
+	return connector;
+}
+
+static int ltdc_get_caps(struct drm_device *ddev)
+{
+	struct ltdc_device *ldev = ddev->dev_private;
+	u32 bus_width_log2, lcr, gc2r;
+
+	/* at least 1 layer must be managed */
+	lcr = reg_read(ldev->regs, LTDC_LCR);
+
+	ldev->caps.nb_layers = max_t(int, lcr, 1);
+
+	/* set data bus width */
+	gc2r = reg_read(ldev->regs, LTDC_GC2R);
+	bus_width_log2 = (gc2r & GC2R_BW) >> 4;
+	ldev->caps.bus_width = 8 << bus_width_log2;
+	ldev->caps.hw_version = reg_read(ldev->regs, LTDC_IDR);
+
+	switch (ldev->caps.hw_version) {
+	case HWVER_10200:
+	case HWVER_10300:
+		ldev->caps.reg_ofs = REG_OFS_NONE;
+		ldev->caps.pix_fmt_hw = ltdc_pix_fmt_a0;
+		break;
+	case HWVER_20101:
+		ldev->caps.reg_ofs = REG_OFS_4;
+		ldev->caps.pix_fmt_hw = ltdc_pix_fmt_a1;
+		break;
+	default:
+		return -ENODEV;
+	}
+
+	return 0;
+}
+
+static struct drm_panel *ltdc_get_panel(struct drm_device *ddev)
+{
+	struct device *dev = ddev->dev;
+	struct device_node *np = dev->of_node;
+	struct device_node *entity, *port = NULL;
+	struct drm_panel *panel = NULL;
+
+	DRM_DEBUG_DRIVER("\n");
+
+	/*
+	 * Parse ltdc node to get remote port and find RGB panel / HDMI slave
+	 * If a dsi or a bridge (hdmi, lvds...) is connected to ltdc,
+	 * a remote port & RGB panel will not be found.
+	 */
+	for_each_endpoint_of_node(np, entity) {
+		if (!of_device_is_available(entity))
+			continue;
+
+		port = of_graph_get_remote_port_parent(entity);
+		if (port) {
+			panel = of_drm_find_panel(port);
+			of_node_put(port);
+			if (panel) {
+				DRM_DEBUG_DRIVER("remote panel %s\n",
+						 port->full_name);
+			} else {
+				DRM_DEBUG_DRIVER("panel missing\n");
+				of_node_put(entity);
+			}
+		}
+	}
+
+	return panel;
+}
+
+int ltdc_load(struct drm_device *ddev)
+{
+	struct platform_device *pdev = to_platform_device(ddev->dev);
+	struct ltdc_device *ldev = ddev->dev_private;
+	struct device *dev = ddev->dev;
+	struct device_node *np = dev->of_node;
+	struct drm_encoder *encoder;
+	struct drm_connector *connector = NULL;
+	struct drm_crtc *crtc;
+	struct reset_control *rstc;
+	struct resource res;
+	int irq, ret, i;
+
+	DRM_DEBUG_DRIVER("\n");
+
+	ldev->panel = ltdc_get_panel(ddev);
+	if (!ldev->panel)
+		return -EPROBE_DEFER;
+
+	rstc = of_reset_control_get(np, NULL);
+
+	mutex_init(&ldev->err_lock);
+
+	ldev->pixel_clk = devm_clk_get(dev, "lcd");
+	if (IS_ERR(ldev->pixel_clk)) {
+		DRM_ERROR("Unable to get lcd clock\n");
+		return -ENODEV;
+	}
+
+	if (clk_prepare_enable(ldev->pixel_clk)) {
+		DRM_ERROR("Unable to prepare pixel clock\n");
+		return -ENODEV;
+	}
+
+	if (of_address_to_resource(np, 0, &res)) {
+		DRM_ERROR("Unable to get resource\n");
+		return -ENODEV;
+	}
+
+	ldev->regs = devm_ioremap_resource(dev, &res);
+	if (IS_ERR(ldev->regs)) {
+		DRM_ERROR("Unable to get ltdc registers\n");
+		return PTR_ERR(ldev->regs);
+	}
+
+	for (i = 0; i < MAX_IRQ; i++) {
+		irq = platform_get_irq(pdev, i);
+		if (irq < 0)
+			continue;
+
+		ret = devm_request_threaded_irq(dev, irq, ltdc_irq,
+						ltdc_irq_thread, IRQF_ONESHOT,
+						dev_name(dev), ddev);
+		if (ret) {
+			DRM_ERROR("Failed to register LTDC interrupt\n");
+			return ret;
+		}
+	}
+
+	if (!IS_ERR(rstc))
+		reset_control_deassert(rstc);
+
+	/* Disable interrupts */
+	reg_clear(ldev->regs, LTDC_IER,
+		  IER_LIE | IER_RRIE | IER_FUIE | IER_TERRIE);
+
+	ret = ltdc_get_caps(ddev);
+	if (ret) {
+		DRM_ERROR("hardware identifier (0x%08x) not supported!\n",
+			  ldev->caps.hw_version);
+		return ret;
+	}
+
+	DRM_INFO("ltdc hw version 0x%08x - ready\n", ldev->caps.hw_version);
+
+	if (ldev->panel) {
+		encoder = ltdc_rgb_encoder_create(ddev);
+		if (!encoder) {
+			DRM_ERROR("Failed to create RGB encoder\n");
+			ret = -EINVAL;
+			goto err;
+		}
+
+		connector = ltdc_rgb_connector_create(ddev);
+		if (!connector) {
+			DRM_ERROR("Failed to create RGB connector\n");
+			ret = -EINVAL;
+			goto err;
+		}
+
+		ret = drm_mode_connector_attach_encoder(connector, encoder);
+		if (ret) {
+			DRM_ERROR("Failed to attach connector to encoder\n");
+			goto err;
+		}
+
+		drm_panel_attach(ldev->panel, connector);
+	}
+
+	crtc = devm_kzalloc(dev, sizeof(*crtc), GFP_KERNEL);
+	if (!crtc) {
+		DRM_ERROR("Failed to allocate crtc\n");
+		ret = -ENOMEM;
+		goto err;
+	}
+
+	ret = ltdc_crtc_init(ddev, crtc);
+	if (ret) {
+		DRM_ERROR("Failed to init crtc\n");
+		goto err;
+	}
+
+	ret = drm_vblank_init(ddev, NB_CRTC);
+	if (ret) {
+		DRM_ERROR("Failed calling drm_vblank_init()\n");
+		goto err;
+	}
+
+	/* Allow usage of vblank without having to call drm_irq_install */
+	ddev->irq_enabled = 1;
+
+	return 0;
+err:
+	if (ldev->panel)
+		drm_panel_detach(ldev->panel);
+
+	clk_disable_unprepare(ldev->pixel_clk);
+
+	return ret;
+}
+
+void ltdc_unload(struct drm_device *ddev)
+{
+	struct ltdc_device *ldev = ddev->dev_private;
+
+	DRM_DEBUG_DRIVER("\n");
+
+	drm_vblank_cleanup(ddev);
+
+	if (ldev->panel)
+		drm_panel_detach(ldev->panel);
+
+	clk_disable_unprepare(ldev->pixel_clk);
+}
+
+MODULE_AUTHOR("Philippe Cornu <philippe.cornu@st.com>");
+MODULE_AUTHOR("Yannick Fertre <yannick.fertre@st.com>");
+MODULE_AUTHOR("Fabien Dessenne <fabien.dessenne@st.com>");
+MODULE_AUTHOR("Mickael Reulier <mickael.reulier@st.com>");
+MODULE_DESCRIPTION("STMicroelectronics ST DRM LTDC driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/gpu/drm/stm/ltdc.h b/drivers/gpu/drm/stm/ltdc.h
new file mode 100644
index 000000000000..d7a9c736ac1e
--- /dev/null
+++ b/drivers/gpu/drm/stm/ltdc.h
@@ -0,0 +1,40 @@
+/*
+ * Copyright (C) STMicroelectronics SA 2017
+ *
+ * Authors: Philippe Cornu <philippe.cornu@st.com>
+ *          Yannick Fertre <yannick.fertre@st.com>
+ *          Fabien Dessenne <fabien.dessenne@st.com>
+ *          Mickael Reulier <mickael.reulier@st.com>
+ *
+ * License terms:  GNU General Public License (GPL), version 2
+ */
+
+#ifndef _LTDC_H_
+#define _LTDC_H_
+
+struct ltdc_caps {
+	u32 hw_version;		/* hardware version */
+	u32 nb_layers;		/* number of supported layers */
+	u32 reg_ofs;		/* register offset for applicable regs */
+	u32 bus_width;		/* bus width (32 or 64 bits) */
+	const u32 *pix_fmt_hw;	/* supported pixel formats */
+};
+
+struct ltdc_device {
+	struct drm_fbdev_cma *fbdev;
+	void __iomem *regs;
+	struct clk *pixel_clk;	/* lcd pixel clock */
+	struct drm_panel *panel;
+	struct mutex err_lock;	/* protecting error_status */
+	struct ltdc_caps caps;
+	u32 clut[256];		/* color look up table */
+	u32 error_status;
+	u32 irq_status;
+};
+
+int ltdc_crtc_enable_vblank(struct drm_device *dev, unsigned int pipe);
+void ltdc_crtc_disable_vblank(struct drm_device *dev, unsigned int pipe);
+int ltdc_load(struct drm_device *ddev);
+void ltdc_unload(struct drm_device *ddev);
+
+#endif
diff --git a/drivers/gpu/drm/tdfx/Makefile b/drivers/gpu/drm/tdfx/Makefile
index 0379f294b32a..74bd4ae32348 100644
--- a/drivers/gpu/drm/tdfx/Makefile
+++ b/drivers/gpu/drm/tdfx/Makefile
@@ -2,7 +2,6 @@
 # Makefile for the drm device driver.  This driver provides support for the
 # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.
 
-ccflags-y := -Iinclude/drm
 tdfx-y := tdfx_drv.o
 
 obj-$(CONFIG_DRM_TDFX)	+= tdfx.o
diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c
index 9a1e34e48f64..51c48a8e00ec 100644
--- a/drivers/gpu/drm/tegra/drm.c
+++ b/drivers/gpu/drm/tegra/drm.c
@@ -892,7 +892,7 @@ static int tegra_drm_context_cleanup(int id, void *p, void *data)
 	return 0;
 }
 
-static void tegra_drm_preclose(struct drm_device *drm, struct drm_file *file)
+static void tegra_drm_postclose(struct drm_device *drm, struct drm_file *file)
 {
 	struct tegra_drm_file *fpriv = file->driver_priv;
 
@@ -960,7 +960,7 @@ static struct drm_driver tegra_drm_driver = {
 	.load = tegra_drm_load,
 	.unload = tegra_drm_unload,
 	.open = tegra_drm_open,
-	.preclose = tegra_drm_preclose,
+	.postclose = tegra_drm_postclose,
 	.lastclose = tegra_drm_lastclose,
 
 #if defined(CONFIG_DEBUG_FS)
diff --git a/drivers/gpu/drm/tilcdc/Makefile b/drivers/gpu/drm/tilcdc/Makefile
index 6f675175a9e5..55ebd516728f 100644
--- a/drivers/gpu/drm/tilcdc/Makefile
+++ b/drivers/gpu/drm/tilcdc/Makefile
@@ -1,4 +1,3 @@
-ccflags-y := -Iinclude/drm
 ifeq (, $(findstring -W,$(EXTRA_CFLAGS)))
 	ccflags-y += -Werror
 endif
diff --git a/drivers/gpu/drm/tilcdc/tilcdc_drv.c b/drivers/gpu/drm/tilcdc/tilcdc_drv.c
index d7ae5be56d12..d67e18983a7d 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_drv.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_drv.c
@@ -22,6 +22,7 @@
 #include <linux/suspend.h>
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_fb_helper.h>
 
 #include "tilcdc_drv.h"
 #include "tilcdc_regs.h"
@@ -29,8 +30,6 @@
 #include "tilcdc_panel.h"
 #include "tilcdc_external.h"
 
-#include "drm_fb_helper.h"
-
 static LIST_HEAD(module_list);
 
 static const u32 tilcdc_rev1_formats[] = { DRM_FORMAT_RGB565 };
diff --git a/drivers/gpu/drm/ttm/Makefile b/drivers/gpu/drm/ttm/Makefile
index f92325800f8a..4d0c938ff4b2 100644
--- a/drivers/gpu/drm/ttm/Makefile
+++ b/drivers/gpu/drm/ttm/Makefile
@@ -1,7 +1,6 @@
 #
 # Makefile for the drm device driver.  This driver provides support for the
 
-ccflags-y := -Iinclude/drm
 ttm-y := ttm_memory.o ttm_tt.o ttm_bo.o \
 	ttm_bo_util.o ttm_bo_vm.o ttm_module.o \
 	ttm_object.o ttm_lock.o ttm_execbuf_util.o ttm_page_alloc.o \
diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c
index 9f53df95f35c..b442d12f2f7d 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_vm.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c
@@ -30,9 +30,9 @@
 
 #define pr_fmt(fmt) "[TTM] " fmt
 
-#include <ttm/ttm_module.h>
-#include <ttm/ttm_bo_driver.h>
-#include <ttm/ttm_placement.h>
+#include <drm/ttm/ttm_module.h>
+#include <drm/ttm/ttm_bo_driver.h>
+#include <drm/ttm/ttm_placement.h>
 #include <drm/drm_vma_manager.h>
 #include <linux/mm.h>
 #include <linux/pfn_t.h>
diff --git a/drivers/gpu/drm/ttm/ttm_tt.c b/drivers/gpu/drm/ttm/ttm_tt.c
index 5260179d788a..8ebc8d3560c3 100644
--- a/drivers/gpu/drm/ttm/ttm_tt.c
+++ b/drivers/gpu/drm/ttm/ttm_tt.c
@@ -39,7 +39,6 @@
 #include <linux/slab.h>
 #include <linux/export.h>
 #include <drm/drm_cache.h>
-#include <drm/drm_mem_util.h>
 #include <drm/ttm/ttm_module.h>
 #include <drm/ttm/ttm_bo_driver.h>
 #include <drm/ttm/ttm_placement.h>
@@ -53,14 +52,16 @@
  */
 static void ttm_tt_alloc_page_directory(struct ttm_tt *ttm)
 {
-	ttm->pages = drm_calloc_large(ttm->num_pages, sizeof(void*));
+	ttm->pages = kvmalloc_array(ttm->num_pages, sizeof(void*),
+			GFP_KERNEL | __GFP_ZERO);
 }
 
 static void ttm_dma_tt_alloc_page_directory(struct ttm_dma_tt *ttm)
 {
-	ttm->ttm.pages = drm_calloc_large(ttm->ttm.num_pages,
+	ttm->ttm.pages = kvmalloc_array(ttm->ttm.num_pages,
 					  sizeof(*ttm->ttm.pages) +
-					  sizeof(*ttm->dma_address));
+					  sizeof(*ttm->dma_address),
+					  GFP_KERNEL | __GFP_ZERO);
 	ttm->dma_address = (void *) (ttm->ttm.pages + ttm->ttm.num_pages);
 }
 
@@ -208,7 +209,7 @@ EXPORT_SYMBOL(ttm_tt_init);
 
 void ttm_tt_fini(struct ttm_tt *ttm)
 {
-	drm_free_large(ttm->pages);
+	kvfree(ttm->pages);
 	ttm->pages = NULL;
 }
 EXPORT_SYMBOL(ttm_tt_fini);
@@ -243,7 +244,7 @@ void ttm_dma_tt_fini(struct ttm_dma_tt *ttm_dma)
 {
 	struct ttm_tt *ttm = &ttm_dma->ttm;
 
-	drm_free_large(ttm->pages);
+	kvfree(ttm->pages);
 	ttm->pages = NULL;
 	ttm_dma->dma_address = NULL;
 }
diff --git a/drivers/gpu/drm/udl/Makefile b/drivers/gpu/drm/udl/Makefile
index 195bcac0b6c8..36f2e825102b 100644
--- a/drivers/gpu/drm/udl/Makefile
+++ b/drivers/gpu/drm/udl/Makefile
@@ -1,6 +1,3 @@
-
-ccflags-y := -Iinclude/drm
-
 udl-y := udl_drv.o udl_modeset.o udl_connector.o udl_encoder.o udl_main.o udl_fb.o udl_transfer.o udl_gem.o udl_dmabuf.o
 
 obj-$(CONFIG_DRM_UDL) := udl.o
diff --git a/drivers/gpu/drm/udl/udl_dmabuf.c b/drivers/gpu/drm/udl/udl_dmabuf.c
index ed0e636243b2..2e031a894813 100644
--- a/drivers/gpu/drm/udl/udl_dmabuf.c
+++ b/drivers/gpu/drm/udl/udl_dmabuf.c
@@ -228,7 +228,7 @@ static int udl_prime_create(struct drm_device *dev,
 		return -ENOMEM;
 
 	obj->sg = sg;
-	obj->pages = drm_malloc_ab(npages, sizeof(struct page *));
+	obj->pages = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL);
 	if (obj->pages == NULL) {
 		DRM_ERROR("obj pages is NULL %d\n", npages);
 		return -ENOMEM;
diff --git a/drivers/gpu/drm/udl/udl_gem.c b/drivers/gpu/drm/udl/udl_gem.c
index 775c50e4f02c..db9ceceba30e 100644
--- a/drivers/gpu/drm/udl/udl_gem.c
+++ b/drivers/gpu/drm/udl/udl_gem.c
@@ -146,7 +146,7 @@ int udl_gem_get_pages(struct udl_gem_object *obj)
 void udl_gem_put_pages(struct udl_gem_object *obj)
 {
 	if (obj->base.import_attach) {
-		drm_free_large(obj->pages);
+		kvfree(obj->pages);
 		obj->pages = NULL;
 		return;
 	}
diff --git a/drivers/gpu/drm/vc4/Kconfig b/drivers/gpu/drm/vc4/Kconfig
index 973b4203c0b2..b16aefe4a8d3 100644
--- a/drivers/gpu/drm/vc4/Kconfig
+++ b/drivers/gpu/drm/vc4/Kconfig
@@ -1,6 +1,6 @@
 config DRM_VC4
 	tristate "Broadcom VC4 Graphics"
-	depends on ARCH_BCM2835 || COMPILE_TEST
+	depends on ARCH_BCM || ARCH_BCM2835 || COMPILE_TEST
 	depends on DRM
 	depends on SND && SND_SOC
 	depends on COMMON_CLK
diff --git a/drivers/gpu/drm/vc4/Makefile b/drivers/gpu/drm/vc4/Makefile
index 61f45d122bd0..25bd5d30415d 100644
--- a/drivers/gpu/drm/vc4/Makefile
+++ b/drivers/gpu/drm/vc4/Makefile
@@ -1,5 +1,3 @@
-ccflags-y := -Iinclude/drm
-
 # Please keep these build lists sorted!
 
 # core driver code
@@ -9,6 +7,7 @@ vc4-y := \
 	vc4_drv.o \
 	vc4_dpi.o \
 	vc4_dsi.o \
+	vc4_fence.o \
 	vc4_kms.o \
 	vc4_gem.o \
 	vc4_hdmi.o \
diff --git a/drivers/gpu/drm/vc4/vc4_bo.c b/drivers/gpu/drm/vc4/vc4_bo.c
index af29432a6471..80b2f9e55c5c 100644
--- a/drivers/gpu/drm/vc4/vc4_bo.c
+++ b/drivers/gpu/drm/vc4/vc4_bo.c
@@ -19,6 +19,8 @@
  * rendering can return quickly.
  */
 
+#include <linux/dma-buf.h>
+
 #include "vc4_drv.h"
 #include "uapi/drm/vc4_drm.h"
 
@@ -88,6 +90,10 @@ static void vc4_bo_destroy(struct vc4_bo *bo)
 
 	vc4->bo_stats.num_allocated--;
 	vc4->bo_stats.size_allocated -= obj->size;
+
+	if (bo->resv == &bo->_resv)
+		reservation_object_fini(bo->resv);
+
 	drm_gem_cma_free_object(obj);
 }
 
@@ -244,8 +250,12 @@ struct vc4_bo *vc4_bo_create(struct drm_device *dev, size_t unaligned_size,
 			return ERR_PTR(-ENOMEM);
 		}
 	}
+	bo = to_vc4_bo(&cma_obj->base);
 
-	return to_vc4_bo(&cma_obj->base);
+	bo->resv = &bo->_resv;
+	reservation_object_init(bo->resv);
+
+	return bo;
 }
 
 int vc4_dumb_create(struct drm_file *file_priv,
@@ -369,6 +379,13 @@ static void vc4_bo_cache_time_timer(unsigned long data)
 	schedule_work(&vc4->bo_cache.time_work);
 }
 
+struct reservation_object *vc4_prime_res_obj(struct drm_gem_object *obj)
+{
+	struct vc4_bo *bo = to_vc4_bo(obj);
+
+	return bo->resv;
+}
+
 struct dma_buf *
 vc4_prime_export(struct drm_device *dev, struct drm_gem_object *obj, int flags)
 {
@@ -440,6 +457,24 @@ void *vc4_prime_vmap(struct drm_gem_object *obj)
 	return drm_gem_cma_prime_vmap(obj);
 }
 
+struct drm_gem_object *
+vc4_prime_import_sg_table(struct drm_device *dev,
+			  struct dma_buf_attachment *attach,
+			  struct sg_table *sgt)
+{
+	struct drm_gem_object *obj;
+	struct vc4_bo *bo;
+
+	obj = drm_gem_cma_prime_import_sg_table(dev, attach, sgt);
+	if (IS_ERR(obj))
+		return obj;
+
+	bo = to_vc4_bo(obj);
+	bo->resv = attach->dmabuf->resv;
+
+	return obj;
+}
+
 int vc4_create_bo_ioctl(struct drm_device *dev, void *data,
 			struct drm_file *file_priv)
 {
diff --git a/drivers/gpu/drm/vc4/vc4_crtc.c b/drivers/gpu/drm/vc4/vc4_crtc.c
index d86c8cce3182..0bfc4d88e4c2 100644
--- a/drivers/gpu/drm/vc4/vc4_crtc.c
+++ b/drivers/gpu/drm/vc4/vc4_crtc.c
@@ -32,13 +32,13 @@
  * ones that set the clock.
  */
 
-#include "drm_atomic.h"
-#include "drm_atomic_helper.h"
-#include "drm_crtc_helper.h"
-#include "linux/clk.h"
-#include "drm_fb_cma_helper.h"
-#include "linux/component.h"
-#include "linux/of_device.h"
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <linux/clk.h>
+#include <drm/drm_fb_cma_helper.h>
+#include <linux/component.h>
+#include <linux/of_device.h>
 #include "vc4_drv.h"
 #include "vc4_regs.h"
 
@@ -151,10 +151,10 @@ int vc4_crtc_debugfs_regs(struct seq_file *m, void *unused)
 }
 #endif
 
-int vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id,
-			    unsigned int flags, int *vpos, int *hpos,
-			    ktime_t *stime, ktime_t *etime,
-			    const struct drm_display_mode *mode)
+bool vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id,
+			     bool in_vblank_irq, int *vpos, int *hpos,
+			     ktime_t *stime, ktime_t *etime,
+			     const struct drm_display_mode *mode)
 {
 	struct vc4_dev *vc4 = to_vc4_dev(dev);
 	struct drm_crtc *crtc = drm_crtc_from_index(dev, crtc_id);
@@ -162,7 +162,7 @@ int vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id,
 	u32 val;
 	int fifo_lines;
 	int vblank_lines;
-	int ret = 0;
+	bool ret = false;
 
 	/* preempt_disable_rt() should go right here in PREEMPT_RT patchset. */
 
@@ -198,7 +198,7 @@ int vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id,
 	fifo_lines = vc4_crtc->cob_size / mode->crtc_hdisplay;
 
 	if (fifo_lines > 0)
-		ret |= DRM_SCANOUTPOS_VALID;
+		ret = true;
 
 	/* HVS more than fifo_lines into frame for compositing? */
 	if (*vpos > fifo_lines) {
@@ -216,7 +216,6 @@ int vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id,
 		 */
 		*vpos -= fifo_lines + 1;
 
-		ret |= DRM_SCANOUTPOS_ACCURATE;
 		return ret;
 	}
 
@@ -229,10 +228,9 @@ int vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id,
 	 * We can't get meaningful readings wrt. scanline position of the PV
 	 * and need to make things up in a approximative but consistent way.
 	 */
-	ret |= DRM_SCANOUTPOS_IN_VBLANK;
 	vblank_lines = mode->vtotal - mode->vdisplay;
 
-	if (flags & DRM_CALLED_FROM_VBLIRQ) {
+	if (in_vblank_irq) {
 		/*
 		 * Assume the irq handler got called close to first
 		 * line of vblank, so PV has about a full vblank
@@ -254,9 +252,10 @@ int vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id,
 		 * we are at the very beginning of vblank, as the hvs just
 		 * started refilling, and the stime and etime timestamps
 		 * truly correspond to start of vblank.
+		 *
+		 * Unfortunately there's no way to report this to upper levels
+		 * and make it more useful.
 		 */
-		if ((val & SCALER_DISPSTATX_FULL) != SCALER_DISPSTATX_FULL)
-			ret |= DRM_SCANOUTPOS_ACCURATE;
 	} else {
 		/*
 		 * No clue where we are inside vblank. Return a vpos of zero,
@@ -270,19 +269,6 @@ int vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id,
 	return ret;
 }
 
-int vc4_crtc_get_vblank_timestamp(struct drm_device *dev, unsigned int crtc_id,
-				  int *max_error, struct timeval *vblank_time,
-				  unsigned flags)
-{
-	struct drm_crtc *crtc = drm_crtc_from_index(dev, crtc_id);
-	struct drm_crtc_state *state = crtc->state;
-
-	/* Helper routine in DRM core does all the work: */
-	return drm_calc_vbltimestamp_from_scanoutpos(dev, crtc_id, max_error,
-						     vblank_time, flags,
-						     &state->adjusted_mode);
-}
-
 static void vc4_crtc_destroy(struct drm_crtc *crtc)
 {
 	drm_crtc_cleanup(crtc);
diff --git a/drivers/gpu/drm/vc4/vc4_dpi.c b/drivers/gpu/drm/vc4/vc4_dpi.c
index c6d703903fd9..39d68080873c 100644
--- a/drivers/gpu/drm/vc4/vc4_dpi.c
+++ b/drivers/gpu/drm/vc4/vc4_dpi.c
@@ -22,14 +22,14 @@
  * ALT2 function.
  */
 
-#include "drm_atomic_helper.h"
-#include "drm_crtc_helper.h"
-#include "drm_edid.h"
-#include "drm_panel.h"
-#include "linux/clk.h"
-#include "linux/component.h"
-#include "linux/of_graph.h"
-#include "linux/of_platform.h"
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_edid.h>
+#include <drm/drm_panel.h>
+#include <linux/clk.h>
+#include <linux/component.h>
+#include <linux/of_graph.h>
+#include <linux/of_platform.h>
 #include "vc4_drv.h"
 #include "vc4_regs.h"
 
diff --git a/drivers/gpu/drm/vc4/vc4_drv.c b/drivers/gpu/drm/vc4/vc4_drv.c
index 61e674baf3a6..136bb4213dc0 100644
--- a/drivers/gpu/drm/vc4/vc4_drv.c
+++ b/drivers/gpu/drm/vc4/vc4_drv.c
@@ -31,7 +31,7 @@
 #include <linux/of_platform.h>
 #include <linux/platform_device.h>
 #include <linux/pm_runtime.h>
-#include "drm_fb_cma_helper.h"
+#include <drm/drm_fb_cma_helper.h>
 #include <drm/drm_fb_helper.h>
 
 #include "uapi/drm/vc4_drm.h"
@@ -154,7 +154,7 @@ static struct drm_driver vc4_drm_driver = {
 	.irq_uninstall = vc4_irq_uninstall,
 
 	.get_scanout_position = vc4_crtc_get_scanoutpos,
-	.get_vblank_timestamp = vc4_crtc_get_vblank_timestamp,
+	.get_vblank_timestamp = drm_calc_vbltimestamp_from_scanoutpos,
 
 #if defined(CONFIG_DEBUG_FS)
 	.debugfs_init = vc4_debugfs_init,
@@ -168,8 +168,9 @@ static struct drm_driver vc4_drm_driver = {
 	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
 	.gem_prime_import = drm_gem_prime_import,
 	.gem_prime_export = vc4_prime_export,
+	.gem_prime_res_obj = vc4_prime_res_obj,
 	.gem_prime_get_sg_table	= drm_gem_cma_prime_get_sg_table,
-	.gem_prime_import_sg_table = drm_gem_cma_prime_import_sg_table,
+	.gem_prime_import_sg_table = vc4_prime_import_sg_table,
 	.gem_prime_vmap = vc4_prime_vmap,
 	.gem_prime_vunmap = drm_gem_cma_prime_vunmap,
 	.gem_prime_mmap = vc4_prime_mmap,
@@ -334,6 +335,7 @@ static int vc4_platform_drm_remove(struct platform_device *pdev)
 
 static const struct of_device_id vc4_of_match[] = {
 	{ .compatible = "brcm,bcm2835-vc4", },
+	{ .compatible = "brcm,cygnus-vc4", },
 	{},
 };
 MODULE_DEVICE_TABLE(of, vc4_of_match);
diff --git a/drivers/gpu/drm/vc4/vc4_drv.h b/drivers/gpu/drm/vc4/vc4_drv.h
index dffce6293d87..a97556f7ccba 100644
--- a/drivers/gpu/drm/vc4/vc4_drv.h
+++ b/drivers/gpu/drm/vc4/vc4_drv.h
@@ -6,10 +6,10 @@
  * published by the Free Software Foundation.
  */
 
-#include "drmP.h"
-#include "drm_gem_cma_helper.h"
-
+#include <linux/reservation.h>
+#include <drm/drmP.h>
 #include <drm/drm_encoder.h>
+#include <drm/drm_gem_cma_helper.h>
 
 struct vc4_dev {
 	struct drm_device *dev;
@@ -56,6 +56,8 @@ struct vc4_dev {
 	/* Protects bo_cache and the BO stats. */
 	struct mutex bo_lock;
 
+	uint64_t dma_fence_context;
+
 	/* Sequence number for the last job queued in bin_job_list.
 	 * Starts at 0 (no jobs emitted).
 	 */
@@ -95,12 +97,23 @@ struct vc4_dev {
 	 */
 	struct list_head seqno_cb_list;
 
-	/* The binner overflow memory that's currently set up in
-	 * BPOA/BPOS registers.  When overflow occurs and a new one is
-	 * allocated, the previous one will be moved to
-	 * vc4->current_exec's free list.
+	/* The memory used for storing binner tile alloc, tile state,
+	 * and overflow memory allocations.  This is freed when V3D
+	 * powers down.
+	 */
+	struct vc4_bo *bin_bo;
+
+	/* Size of blocks allocated within bin_bo. */
+	uint32_t bin_alloc_size;
+
+	/* Bitmask of the bin_alloc_size chunks in bin_bo that are
+	 * used.
 	 */
-	struct vc4_bo *overflow_mem;
+	uint32_t bin_alloc_used;
+
+	/* Bitmask of the current bin_alloc used for overflow memory. */
+	uint32_t bin_alloc_overflow;
+
 	struct work_struct overflow_mem_work;
 
 	int power_refcount;
@@ -150,6 +163,10 @@ struct vc4_bo {
 	 * DRM_IOCTL_VC4_CREATE_SHADER_BO.
 	 */
 	struct vc4_validated_shader_info *validated_shader;
+
+	/* normally (resv == &_resv) except for imported bo's */
+	struct reservation_object *resv;
+	struct reservation_object _resv;
 };
 
 static inline struct vc4_bo *
@@ -158,6 +175,19 @@ to_vc4_bo(struct drm_gem_object *bo)
 	return (struct vc4_bo *)bo;
 }
 
+struct vc4_fence {
+	struct dma_fence base;
+	struct drm_device *dev;
+	/* vc4 seqno for signaled() test */
+	uint64_t seqno;
+};
+
+static inline struct vc4_fence *
+to_vc4_fence(struct dma_fence *fence)
+{
+	return (struct vc4_fence *)fence;
+}
+
 struct vc4_seqno_cb {
 	struct work_struct work;
 	uint64_t seqno;
@@ -168,6 +198,7 @@ struct vc4_v3d {
 	struct vc4_dev *vc4;
 	struct platform_device *pdev;
 	void __iomem *regs;
+	struct clk *clk;
 };
 
 struct vc4_hvs {
@@ -230,6 +261,8 @@ struct vc4_exec_info {
 	/* Latest write_seqno of any BO that binning depends on. */
 	uint64_t bin_dep_seqno;
 
+	struct dma_fence *fence;
+
 	/* Last current addresses the hardware was processing when the
 	 * hangcheck timer checked on us.
 	 */
@@ -293,8 +326,12 @@ struct vc4_exec_info {
 	bool found_increment_semaphore_packet;
 	bool found_flush;
 	uint8_t bin_tiles_x, bin_tiles_y;
-	struct drm_gem_cma_object *tile_bo;
+	/* Physical address of the start of the tile alloc array
+	 * (where each tile's binned CL will start)
+	 */
 	uint32_t tile_alloc_offset;
+	/* Bitmask of which binner slots are freed when this job completes. */
+	uint32_t bin_slots;
 
 	/**
 	 * Computed addresses pointing into exec_bo where we start the
@@ -436,7 +473,11 @@ int vc4_mmap_bo_ioctl(struct drm_device *dev, void *data,
 int vc4_get_hang_state_ioctl(struct drm_device *dev, void *data,
 			     struct drm_file *file_priv);
 int vc4_mmap(struct file *filp, struct vm_area_struct *vma);
+struct reservation_object *vc4_prime_res_obj(struct drm_gem_object *obj);
 int vc4_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma);
+struct drm_gem_object *vc4_prime_import_sg_table(struct drm_device *dev,
+						 struct dma_buf_attachment *attach,
+						 struct sg_table *sgt);
 void *vc4_prime_vmap(struct drm_gem_object *obj);
 void vc4_bo_cache_init(struct drm_device *dev);
 void vc4_bo_cache_destroy(struct drm_device *dev);
@@ -446,13 +487,10 @@ int vc4_bo_stats_debugfs(struct seq_file *m, void *arg);
 extern struct platform_driver vc4_crtc_driver;
 bool vc4_event_pending(struct drm_crtc *crtc);
 int vc4_crtc_debugfs_regs(struct seq_file *m, void *arg);
-int vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id,
-			    unsigned int flags, int *vpos, int *hpos,
-			    ktime_t *stime, ktime_t *etime,
-			    const struct drm_display_mode *mode);
-int vc4_crtc_get_vblank_timestamp(struct drm_device *dev, unsigned int crtc_id,
-				  int *max_error, struct timeval *vblank_time,
-				  unsigned flags);
+bool vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id,
+			     bool in_vblank_irq, int *vpos, int *hpos,
+			     ktime_t *stime, ktime_t *etime,
+			     const struct drm_display_mode *mode);
 
 /* vc4_debugfs.c */
 int vc4_debugfs_init(struct drm_minor *minor);
@@ -468,6 +506,9 @@ int vc4_dpi_debugfs_regs(struct seq_file *m, void *unused);
 extern struct platform_driver vc4_dsi_driver;
 int vc4_dsi_debugfs_regs(struct seq_file *m, void *unused);
 
+/* vc4_fence.c */
+extern const struct dma_fence_ops vc4_fence_ops;
+
 /* vc4_gem.c */
 void vc4_gem_init(struct drm_device *dev);
 void vc4_gem_destroy(struct drm_device *dev);
@@ -522,6 +563,7 @@ void vc4_plane_async_set_fb(struct drm_plane *plane,
 extern struct platform_driver vc4_v3d_driver;
 int vc4_v3d_debugfs_ident(struct seq_file *m, void *unused);
 int vc4_v3d_debugfs_regs(struct seq_file *m, void *unused);
+int vc4_v3d_get_bin_slot(struct vc4_dev *vc4);
 
 /* vc4_validate.c */
 int
diff --git a/drivers/gpu/drm/vc4/vc4_dsi.c b/drivers/gpu/drm/vc4/vc4_dsi.c
index 160f981d1cf4..0ef41df3915f 100644
--- a/drivers/gpu/drm/vc4/vc4_dsi.c
+++ b/drivers/gpu/drm/vc4/vc4_dsi.c
@@ -29,20 +29,20 @@
  * hopefully present.
  */
 
-#include "drm_atomic_helper.h"
-#include "drm_crtc_helper.h"
-#include "drm_edid.h"
-#include "drm_mipi_dsi.h"
-#include "drm_panel.h"
-#include "linux/clk.h"
-#include "linux/clk-provider.h"
-#include "linux/completion.h"
-#include "linux/component.h"
-#include "linux/dmaengine.h"
-#include "linux/i2c.h"
-#include "linux/of_address.h"
-#include "linux/of_platform.h"
-#include "linux/pm_runtime.h"
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_edid.h>
+#include <drm/drm_mipi_dsi.h>
+#include <drm/drm_panel.h>
+#include <linux/clk.h>
+#include <linux/clk-provider.h>
+#include <linux/completion.h>
+#include <linux/component.h>
+#include <linux/dmaengine.h>
+#include <linux/i2c.h>
+#include <linux/of_address.h>
+#include <linux/of_platform.h>
+#include <linux/pm_runtime.h>
 #include "vc4_drv.h"
 #include "vc4_regs.h"
 
@@ -519,7 +519,8 @@ struct vc4_dsi {
 	/* DSI channel for the panel we're connected to. */
 	u32 channel;
 	u32 lanes;
-	enum mipi_dsi_pixel_format format;
+	u32 format;
+	u32 divider;
 	u32 mode_flags;
 
 	/* Input clock from CPRMAN to the digital PHY, for the DSI
@@ -906,13 +907,67 @@ static void vc4_dsi_encoder_disable(struct drm_encoder *encoder)
 	pm_runtime_put(dev);
 }
 
+/* Extends the mode's blank intervals to handle BCM2835's integer-only
+ * DSI PLL divider.
+ *
+ * On 2835, PLLD is set to 2Ghz, and may not be changed by the display
+ * driver since most peripherals are hanging off of the PLLD_PER
+ * divider.  PLLD_DSI1, which drives our DSI bit clock (and therefore
+ * the pixel clock), only has an integer divider off of DSI.
+ *
+ * To get our panel mode to refresh at the expected 60Hz, we need to
+ * extend the horizontal blank time.  This means we drive a
+ * higher-than-expected clock rate to the panel, but that's what the
+ * firmware does too.
+ */
+static bool vc4_dsi_encoder_mode_fixup(struct drm_encoder *encoder,
+				       const struct drm_display_mode *mode,
+				       struct drm_display_mode *adjusted_mode)
+{
+	struct vc4_dsi_encoder *vc4_encoder = to_vc4_dsi_encoder(encoder);
+	struct vc4_dsi *dsi = vc4_encoder->dsi;
+	struct clk *phy_parent = clk_get_parent(dsi->pll_phy_clock);
+	unsigned long parent_rate = clk_get_rate(phy_parent);
+	unsigned long pixel_clock_hz = mode->clock * 1000;
+	unsigned long pll_clock = pixel_clock_hz * dsi->divider;
+	int divider;
+
+	/* Find what divider gets us a faster clock than the requested
+	 * pixel clock.
+	 */
+	for (divider = 1; divider < 8; divider++) {
+		if (parent_rate / divider < pll_clock) {
+			divider--;
+			break;
+		}
+	}
+
+	/* Now that we've picked a PLL divider, calculate back to its
+	 * pixel clock.
+	 */
+	pll_clock = parent_rate / divider;
+	pixel_clock_hz = pll_clock / dsi->divider;
+
+	/* Round up the clk_set_rate() request slightly, since
+	 * PLLD_DSI1 is an integer divider and its rate selection will
+	 * never round up.
+	 */
+	adjusted_mode->clock = pixel_clock_hz / 1000 + 1;
+
+	/* Given the new pixel clock, adjust HFP to keep vrefresh the same. */
+	adjusted_mode->htotal = pixel_clock_hz / (mode->vrefresh * mode->vtotal);
+	adjusted_mode->hsync_end += adjusted_mode->htotal - mode->htotal;
+	adjusted_mode->hsync_start += adjusted_mode->htotal - mode->htotal;
+
+	return true;
+}
+
 static void vc4_dsi_encoder_enable(struct drm_encoder *encoder)
 {
-	struct drm_display_mode *mode = &encoder->crtc->mode;
+	struct drm_display_mode *mode = &encoder->crtc->state->adjusted_mode;
 	struct vc4_dsi_encoder *vc4_encoder = to_vc4_dsi_encoder(encoder);
 	struct vc4_dsi *dsi = vc4_encoder->dsi;
 	struct device *dev = &dsi->pdev->dev;
-	u32 format = 0, divider = 0;
 	bool debug_dump_regs = false;
 	unsigned long hs_clock;
 	u32 ui_ns;
@@ -940,26 +995,7 @@ static void vc4_dsi_encoder_enable(struct drm_encoder *encoder)
 		vc4_dsi_dump_regs(dsi);
 	}
 
-	switch (dsi->format) {
-	case MIPI_DSI_FMT_RGB888:
-		format = DSI_PFORMAT_RGB888;
-		divider = 24 / dsi->lanes;
-		break;
-	case MIPI_DSI_FMT_RGB666:
-		format = DSI_PFORMAT_RGB666;
-		divider = 24 / dsi->lanes;
-		break;
-	case MIPI_DSI_FMT_RGB666_PACKED:
-		format = DSI_PFORMAT_RGB666_PACKED;
-		divider = 18 / dsi->lanes;
-		break;
-	case MIPI_DSI_FMT_RGB565:
-		format = DSI_PFORMAT_RGB565;
-		divider = 16 / dsi->lanes;
-		break;
-	}
-
-	phy_clock = pixel_clock_hz * divider;
+	phy_clock = pixel_clock_hz * dsi->divider;
 	ret = clk_set_rate(dsi->pll_phy_clock, phy_clock);
 	if (ret) {
 		dev_err(&dsi->pdev->dev,
@@ -1134,8 +1170,9 @@ static void vc4_dsi_encoder_enable(struct drm_encoder *encoder)
 
 	if (dsi->mode_flags & MIPI_DSI_MODE_VIDEO) {
 		DSI_PORT_WRITE(DISP0_CTRL,
-			       VC4_SET_FIELD(divider, DSI_DISP0_PIX_CLK_DIV) |
-			       VC4_SET_FIELD(format, DSI_DISP0_PFORMAT) |
+			       VC4_SET_FIELD(dsi->divider,
+					     DSI_DISP0_PIX_CLK_DIV) |
+			       VC4_SET_FIELD(dsi->format, DSI_DISP0_PFORMAT) |
 			       VC4_SET_FIELD(DSI_DISP0_LP_STOP_PERFRAME,
 					     DSI_DISP0_LP_STOP_CTRL) |
 			       DSI_DISP0_ST_END |
@@ -1347,9 +1384,31 @@ static int vc4_dsi_host_attach(struct mipi_dsi_host *host,
 
 	dsi->lanes = device->lanes;
 	dsi->channel = device->channel;
-	dsi->format = device->format;
 	dsi->mode_flags = device->mode_flags;
 
+	switch (device->format) {
+	case MIPI_DSI_FMT_RGB888:
+		dsi->format = DSI_PFORMAT_RGB888;
+		dsi->divider = 24 / dsi->lanes;
+		break;
+	case MIPI_DSI_FMT_RGB666:
+		dsi->format = DSI_PFORMAT_RGB666;
+		dsi->divider = 24 / dsi->lanes;
+		break;
+	case MIPI_DSI_FMT_RGB666_PACKED:
+		dsi->format = DSI_PFORMAT_RGB666_PACKED;
+		dsi->divider = 18 / dsi->lanes;
+		break;
+	case MIPI_DSI_FMT_RGB565:
+		dsi->format = DSI_PFORMAT_RGB565;
+		dsi->divider = 16 / dsi->lanes;
+		break;
+	default:
+		dev_err(&dsi->pdev->dev, "Unknown DSI format: %d.\n",
+			dsi->format);
+		return 0;
+	}
+
 	if (!(dsi->mode_flags & MIPI_DSI_MODE_VIDEO)) {
 		dev_err(&dsi->pdev->dev,
 			"Only VIDEO mode panels supported currently.\n");
@@ -1397,6 +1456,7 @@ static const struct mipi_dsi_host_ops vc4_dsi_host_ops = {
 static const struct drm_encoder_helper_funcs vc4_dsi_encoder_helper_funcs = {
 	.disable = vc4_dsi_encoder_disable,
 	.enable = vc4_dsi_encoder_enable,
+	.mode_fixup = vc4_dsi_encoder_mode_fixup,
 };
 
 static const struct of_device_id vc4_dsi_dt_match[] = {
diff --git a/drivers/gpu/drm/vc4/vc4_fence.c b/drivers/gpu/drm/vc4/vc4_fence.c
new file mode 100644
index 000000000000..dbf5a5a5d5f5
--- /dev/null
+++ b/drivers/gpu/drm/vc4/vc4_fence.c
@@ -0,0 +1,56 @@
+/*
+ * Copyright © 2017 Broadcom
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the next
+ * paragraph) shall be included in all copies or substantial portions of the
+ * Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ */
+
+#include "vc4_drv.h"
+
+static const char *vc4_fence_get_driver_name(struct dma_fence *fence)
+{
+	return "vc4";
+}
+
+static const char *vc4_fence_get_timeline_name(struct dma_fence *fence)
+{
+	return "vc4-v3d";
+}
+
+static bool vc4_fence_enable_signaling(struct dma_fence *fence)
+{
+	return true;
+}
+
+static bool vc4_fence_signaled(struct dma_fence *fence)
+{
+	struct vc4_fence *f = to_vc4_fence(fence);
+	struct vc4_dev *vc4 = to_vc4_dev(f->dev);
+
+	return vc4->finished_seqno >= f->seqno;
+}
+
+const struct dma_fence_ops vc4_fence_ops = {
+	.get_driver_name = vc4_fence_get_driver_name,
+	.get_timeline_name = vc4_fence_get_timeline_name,
+	.enable_signaling = vc4_fence_enable_signaling,
+	.signaled = vc4_fence_signaled,
+	.wait = dma_fence_default_wait,
+	.release = dma_fence_free,
+};
diff --git a/drivers/gpu/drm/vc4/vc4_gem.c b/drivers/gpu/drm/vc4/vc4_gem.c
index e9c381c42139..9dc7646d49ed 100644
--- a/drivers/gpu/drm/vc4/vc4_gem.c
+++ b/drivers/gpu/drm/vc4/vc4_gem.c
@@ -463,6 +463,8 @@ vc4_update_bo_seqnos(struct vc4_exec_info *exec, uint64_t seqno)
 	for (i = 0; i < exec->bo_count; i++) {
 		bo = to_vc4_bo(&exec->bo[i]->base);
 		bo->seqno = seqno;
+
+		reservation_object_add_shared_fence(bo->resv, exec->fence);
 	}
 
 	list_for_each_entry(bo, &exec->unref_list, unref_head) {
@@ -472,7 +474,103 @@ vc4_update_bo_seqnos(struct vc4_exec_info *exec, uint64_t seqno)
 	for (i = 0; i < exec->rcl_write_bo_count; i++) {
 		bo = to_vc4_bo(&exec->rcl_write_bo[i]->base);
 		bo->write_seqno = seqno;
+
+		reservation_object_add_excl_fence(bo->resv, exec->fence);
+	}
+}
+
+static void
+vc4_unlock_bo_reservations(struct drm_device *dev,
+			   struct vc4_exec_info *exec,
+			   struct ww_acquire_ctx *acquire_ctx)
+{
+	int i;
+
+	for (i = 0; i < exec->bo_count; i++) {
+		struct vc4_bo *bo = to_vc4_bo(&exec->bo[i]->base);
+
+		ww_mutex_unlock(&bo->resv->lock);
+	}
+
+	ww_acquire_fini(acquire_ctx);
+}
+
+/* Takes the reservation lock on all the BOs being referenced, so that
+ * at queue submit time we can update the reservations.
+ *
+ * We don't lock the RCL the tile alloc/state BOs, or overflow memory
+ * (all of which are on exec->unref_list).  They're entirely private
+ * to vc4, so we don't attach dma-buf fences to them.
+ */
+static int
+vc4_lock_bo_reservations(struct drm_device *dev,
+			 struct vc4_exec_info *exec,
+			 struct ww_acquire_ctx *acquire_ctx)
+{
+	int contended_lock = -1;
+	int i, ret;
+	struct vc4_bo *bo;
+
+	ww_acquire_init(acquire_ctx, &reservation_ww_class);
+
+retry:
+	if (contended_lock != -1) {
+		bo = to_vc4_bo(&exec->bo[contended_lock]->base);
+		ret = ww_mutex_lock_slow_interruptible(&bo->resv->lock,
+						       acquire_ctx);
+		if (ret) {
+			ww_acquire_done(acquire_ctx);
+			return ret;
+		}
+	}
+
+	for (i = 0; i < exec->bo_count; i++) {
+		if (i == contended_lock)
+			continue;
+
+		bo = to_vc4_bo(&exec->bo[i]->base);
+
+		ret = ww_mutex_lock_interruptible(&bo->resv->lock, acquire_ctx);
+		if (ret) {
+			int j;
+
+			for (j = 0; j < i; j++) {
+				bo = to_vc4_bo(&exec->bo[j]->base);
+				ww_mutex_unlock(&bo->resv->lock);
+			}
+
+			if (contended_lock != -1 && contended_lock >= i) {
+				bo = to_vc4_bo(&exec->bo[contended_lock]->base);
+
+				ww_mutex_unlock(&bo->resv->lock);
+			}
+
+			if (ret == -EDEADLK) {
+				contended_lock = i;
+				goto retry;
+			}
+
+			ww_acquire_done(acquire_ctx);
+			return ret;
+		}
 	}
+
+	ww_acquire_done(acquire_ctx);
+
+	/* Reserve space for our shared (read-only) fence references,
+	 * before we commit the CL to the hardware.
+	 */
+	for (i = 0; i < exec->bo_count; i++) {
+		bo = to_vc4_bo(&exec->bo[i]->base);
+
+		ret = reservation_object_reserve_shared(bo->resv);
+		if (ret) {
+			vc4_unlock_bo_reservations(dev, exec, acquire_ctx);
+			return ret;
+		}
+	}
+
+	return 0;
 }
 
 /* Queues a struct vc4_exec_info for execution.  If no job is
@@ -484,19 +582,34 @@ vc4_update_bo_seqnos(struct vc4_exec_info *exec, uint64_t seqno)
  * then bump the end address.  That's a change for a later date,
  * though.
  */
-static void
-vc4_queue_submit(struct drm_device *dev, struct vc4_exec_info *exec)
+static int
+vc4_queue_submit(struct drm_device *dev, struct vc4_exec_info *exec,
+		 struct ww_acquire_ctx *acquire_ctx)
 {
 	struct vc4_dev *vc4 = to_vc4_dev(dev);
 	uint64_t seqno;
 	unsigned long irqflags;
+	struct vc4_fence *fence;
+
+	fence = kzalloc(sizeof(*fence), GFP_KERNEL);
+	if (!fence)
+		return -ENOMEM;
+	fence->dev = dev;
 
 	spin_lock_irqsave(&vc4->job_lock, irqflags);
 
 	seqno = ++vc4->emit_seqno;
 	exec->seqno = seqno;
+
+	dma_fence_init(&fence->base, &vc4_fence_ops, &vc4->job_lock,
+		       vc4->dma_fence_context, exec->seqno);
+	fence->seqno = exec->seqno;
+	exec->fence = &fence->base;
+
 	vc4_update_bo_seqnos(exec, seqno);
 
+	vc4_unlock_bo_reservations(dev, exec, acquire_ctx);
+
 	list_add_tail(&exec->head, &vc4->bin_job_list);
 
 	/* If no job was executing, kick ours off.  Otherwise, it'll
@@ -509,6 +622,8 @@ vc4_queue_submit(struct drm_device *dev, struct vc4_exec_info *exec)
 	}
 
 	spin_unlock_irqrestore(&vc4->job_lock, irqflags);
+
+	return 0;
 }
 
 /**
@@ -545,14 +660,15 @@ vc4_cl_lookup_bos(struct drm_device *dev,
 		return -EINVAL;
 	}
 
-	exec->bo = drm_calloc_large(exec->bo_count,
-				    sizeof(struct drm_gem_cma_object *));
+	exec->bo = kvmalloc_array(exec->bo_count,
+				    sizeof(struct drm_gem_cma_object *),
+				    GFP_KERNEL | __GFP_ZERO);
 	if (!exec->bo) {
 		DRM_ERROR("Failed to allocate validated BO pointers\n");
 		return -ENOMEM;
 	}
 
-	handles = drm_malloc_ab(exec->bo_count, sizeof(uint32_t));
+	handles = kvmalloc_array(exec->bo_count, sizeof(uint32_t), GFP_KERNEL);
 	if (!handles) {
 		ret = -ENOMEM;
 		DRM_ERROR("Failed to allocate incoming GEM handles\n");
@@ -584,7 +700,7 @@ vc4_cl_lookup_bos(struct drm_device *dev,
 	spin_unlock(&file_priv->table_lock);
 
 fail:
-	drm_free_large(handles);
+	kvfree(handles);
 	return ret;
 }
 
@@ -622,7 +738,7 @@ vc4_get_bcl(struct drm_device *dev, struct vc4_exec_info *exec)
 	 * read the contents back for validation, and I think the
 	 * bo->vaddr is uncached access.
 	 */
-	temp = drm_malloc_ab(temp_size, 1);
+	temp = kvmalloc_array(temp_size, 1, GFP_KERNEL);
 	if (!temp) {
 		DRM_ERROR("Failed to allocate storage for copying "
 			  "in bin/render CLs.\n");
@@ -697,7 +813,7 @@ vc4_get_bcl(struct drm_device *dev, struct vc4_exec_info *exec)
 	ret = vc4_wait_for_seqno(dev, exec->bin_dep_seqno, ~0ull, true);
 
 fail:
-	drm_free_large(temp);
+	kvfree(temp);
 	return ret;
 }
 
@@ -705,12 +821,19 @@ static void
 vc4_complete_exec(struct drm_device *dev, struct vc4_exec_info *exec)
 {
 	struct vc4_dev *vc4 = to_vc4_dev(dev);
+	unsigned long irqflags;
 	unsigned i;
 
+	/* If we got force-completed because of GPU reset rather than
+	 * through our IRQ handler, signal the fence now.
+	 */
+	if (exec->fence)
+		dma_fence_signal(exec->fence);
+
 	if (exec->bo) {
 		for (i = 0; i < exec->bo_count; i++)
 			drm_gem_object_unreference_unlocked(&exec->bo[i]->base);
-		drm_free_large(exec->bo);
+		kvfree(exec->bo);
 	}
 
 	while (!list_empty(&exec->unref_list)) {
@@ -720,6 +843,11 @@ vc4_complete_exec(struct drm_device *dev, struct vc4_exec_info *exec)
 		drm_gem_object_unreference_unlocked(&bo->base.base);
 	}
 
+	/* Free up the allocation of any bin slots we used. */
+	spin_lock_irqsave(&vc4->job_lock, irqflags);
+	vc4->bin_alloc_used &= ~exec->bin_slots;
+	spin_unlock_irqrestore(&vc4->job_lock, irqflags);
+
 	mutex_lock(&vc4->power_lock);
 	if (--vc4->power_refcount == 0) {
 		pm_runtime_mark_last_busy(&vc4->v3d->pdev->dev);
@@ -874,6 +1002,7 @@ vc4_submit_cl_ioctl(struct drm_device *dev, void *data,
 	struct vc4_dev *vc4 = to_vc4_dev(dev);
 	struct drm_vc4_submit_cl *args = data;
 	struct vc4_exec_info *exec;
+	struct ww_acquire_ctx acquire_ctx;
 	int ret = 0;
 
 	if ((args->flags & ~VC4_SUBMIT_CL_USE_CLEAR_COLOR) != 0) {
@@ -888,13 +1017,16 @@ vc4_submit_cl_ioctl(struct drm_device *dev, void *data,
 	}
 
 	mutex_lock(&vc4->power_lock);
-	if (vc4->power_refcount++ == 0)
+	if (vc4->power_refcount++ == 0) {
 		ret = pm_runtime_get_sync(&vc4->v3d->pdev->dev);
-	mutex_unlock(&vc4->power_lock);
-	if (ret < 0) {
-		kfree(exec);
-		return ret;
+		if (ret < 0) {
+			mutex_unlock(&vc4->power_lock);
+			vc4->power_refcount--;
+			kfree(exec);
+			return ret;
+		}
 	}
+	mutex_unlock(&vc4->power_lock);
 
 	exec->args = args;
 	INIT_LIST_HEAD(&exec->unref_list);
@@ -916,12 +1048,18 @@ vc4_submit_cl_ioctl(struct drm_device *dev, void *data,
 	if (ret)
 		goto fail;
 
+	ret = vc4_lock_bo_reservations(dev, exec, &acquire_ctx);
+	if (ret)
+		goto fail;
+
 	/* Clear this out of the struct we'll be putting in the queue,
 	 * since it's part of our stack.
 	 */
 	exec->args = NULL;
 
-	vc4_queue_submit(dev, exec);
+	ret = vc4_queue_submit(dev, exec, &acquire_ctx);
+	if (ret)
+		goto fail;
 
 	/* Return the seqno for our job. */
 	args->seqno = vc4->emit_seqno;
@@ -939,6 +1077,8 @@ vc4_gem_init(struct drm_device *dev)
 {
 	struct vc4_dev *vc4 = to_vc4_dev(dev);
 
+	vc4->dma_fence_context = dma_fence_context_alloc(1);
+
 	INIT_LIST_HEAD(&vc4->bin_job_list);
 	INIT_LIST_HEAD(&vc4->render_job_list);
 	INIT_LIST_HEAD(&vc4->job_done_list);
@@ -968,9 +1108,9 @@ vc4_gem_destroy(struct drm_device *dev)
 	/* V3D should already have disabled its interrupt and cleared
 	 * the overflow allocation registers.  Now free the object.
 	 */
-	if (vc4->overflow_mem) {
-		drm_gem_object_unreference_unlocked(&vc4->overflow_mem->base.base);
-		vc4->overflow_mem = NULL;
+	if (vc4->bin_bo) {
+		drm_gem_object_put_unlocked(&vc4->bin_bo->base.base);
+		vc4->bin_bo = NULL;
 	}
 
 	if (vc4->hang_state)
diff --git a/drivers/gpu/drm/vc4/vc4_hdmi.c b/drivers/gpu/drm/vc4/vc4_hdmi.c
index e9cbe269710b..ed63d4e85762 100644
--- a/drivers/gpu/drm/vc4/vc4_hdmi.c
+++ b/drivers/gpu/drm/vc4/vc4_hdmi.c
@@ -42,20 +42,21 @@
  * encoder block has CEC support.
  */
 
-#include "drm_atomic_helper.h"
-#include "drm_crtc_helper.h"
-#include "drm_edid.h"
-#include "linux/clk.h"
-#include "linux/component.h"
-#include "linux/i2c.h"
-#include "linux/of_address.h"
-#include "linux/of_gpio.h"
-#include "linux/of_platform.h"
-#include "linux/rational.h"
-#include "sound/dmaengine_pcm.h"
-#include "sound/pcm_drm_eld.h"
-#include "sound/pcm_params.h"
-#include "sound/soc.h"
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_edid.h>
+#include <linux/clk.h>
+#include <linux/component.h>
+#include <linux/i2c.h>
+#include <linux/of_address.h>
+#include <linux/of_gpio.h>
+#include <linux/of_platform.h>
+#include <linux/pm_runtime.h>
+#include <linux/rational.h>
+#include <sound/dmaengine_pcm.h>
+#include <sound/pcm_drm_eld.h>
+#include <sound/pcm_params.h>
+#include <sound/soc.h>
 #include "vc4_drv.h"
 #include "vc4_regs.h"
 
@@ -449,13 +450,38 @@ static void vc4_hdmi_set_infoframes(struct drm_encoder *encoder)
 	vc4_hdmi_set_spd_infoframe(encoder);
 }
 
-static void vc4_hdmi_encoder_mode_set(struct drm_encoder *encoder,
-				      struct drm_display_mode *unadjusted_mode,
-				      struct drm_display_mode *mode)
+static void vc4_hdmi_encoder_disable(struct drm_encoder *encoder)
+{
+	struct drm_device *dev = encoder->dev;
+	struct vc4_dev *vc4 = to_vc4_dev(dev);
+	struct vc4_hdmi *hdmi = vc4->hdmi;
+	int ret;
+
+	HDMI_WRITE(VC4_HDMI_RAM_PACKET_CONFIG, 0);
+
+	HDMI_WRITE(VC4_HDMI_TX_PHY_RESET_CTL, 0xf << 16);
+	HD_WRITE(VC4_HD_VID_CTL,
+		 HD_READ(VC4_HD_VID_CTL) & ~VC4_HD_VID_CTL_ENABLE);
+
+	HD_WRITE(VC4_HD_M_CTL, VC4_HD_M_SW_RST);
+	udelay(1);
+	HD_WRITE(VC4_HD_M_CTL, 0);
+
+	clk_disable_unprepare(hdmi->hsm_clock);
+	clk_disable_unprepare(hdmi->pixel_clock);
+
+	ret = pm_runtime_put(&hdmi->pdev->dev);
+	if (ret < 0)
+		DRM_ERROR("Failed to release power domain: %d\n", ret);
+}
+
+static void vc4_hdmi_encoder_enable(struct drm_encoder *encoder)
 {
+	struct drm_display_mode *mode = &encoder->crtc->state->adjusted_mode;
 	struct vc4_hdmi_encoder *vc4_encoder = to_vc4_hdmi_encoder(encoder);
 	struct drm_device *dev = encoder->dev;
 	struct vc4_dev *vc4 = to_vc4_dev(dev);
+	struct vc4_hdmi *hdmi = vc4->hdmi;
 	bool debug_dump_regs = false;
 	bool hsync_pos = mode->flags & DRM_MODE_FLAG_PHSYNC;
 	bool vsync_pos = mode->flags & DRM_MODE_FLAG_PVSYNC;
@@ -475,6 +501,64 @@ static void vc4_hdmi_encoder_mode_set(struct drm_encoder *encoder,
 					interlaced,
 					VC4_HDMI_VERTB_VBP));
 	u32 csc_ctl;
+	int ret;
+
+	ret = pm_runtime_get_sync(&hdmi->pdev->dev);
+	if (ret < 0) {
+		DRM_ERROR("Failed to retain power domain: %d\n", ret);
+		return;
+	}
+
+	/* This is the rate that is set by the firmware.  The number
+	 * needs to be a bit higher than the pixel clock rate
+	 * (generally 148.5Mhz).
+	 */
+	ret = clk_set_rate(hdmi->hsm_clock, 163682864);
+	if (ret) {
+		DRM_ERROR("Failed to set HSM clock rate: %d\n", ret);
+		return;
+	}
+
+	ret = clk_set_rate(hdmi->pixel_clock,
+			   mode->clock * 1000 *
+			   ((mode->flags & DRM_MODE_FLAG_DBLCLK) ? 2 : 1));
+	if (ret) {
+		DRM_ERROR("Failed to set pixel clock rate: %d\n", ret);
+		return;
+	}
+
+	ret = clk_prepare_enable(hdmi->pixel_clock);
+	if (ret) {
+		DRM_ERROR("Failed to turn on pixel clock: %d\n", ret);
+		return;
+	}
+
+	ret = clk_prepare_enable(hdmi->hsm_clock);
+	if (ret) {
+		DRM_ERROR("Failed to turn on HDMI state machine clock: %d\n",
+			  ret);
+		clk_disable_unprepare(hdmi->pixel_clock);
+		return;
+	}
+
+	HD_WRITE(VC4_HD_M_CTL, VC4_HD_M_SW_RST);
+	udelay(1);
+	HD_WRITE(VC4_HD_M_CTL, 0);
+
+	HD_WRITE(VC4_HD_M_CTL, VC4_HD_M_ENABLE);
+
+	HDMI_WRITE(VC4_HDMI_SW_RESET_CONTROL,
+		   VC4_HDMI_SW_RESET_HDMI |
+		   VC4_HDMI_SW_RESET_FORMAT_DETECT);
+
+	HDMI_WRITE(VC4_HDMI_SW_RESET_CONTROL, 0);
+
+	/* PHY should be in reset, like
+	 * vc4_hdmi_encoder_disable() does.
+	 */
+	HDMI_WRITE(VC4_HDMI_TX_PHY_RESET_CTL, 0xf << 16);
+
+	HDMI_WRITE(VC4_HDMI_TX_PHY_RESET_CTL, 0);
 
 	if (debug_dump_regs) {
 		DRM_INFO("HDMI regs before:\n");
@@ -483,9 +567,6 @@ static void vc4_hdmi_encoder_mode_set(struct drm_encoder *encoder,
 
 	HD_WRITE(VC4_HD_VID_CTL, 0);
 
-	clk_set_rate(vc4->hdmi->pixel_clock, mode->clock * 1000 *
-		     ((mode->flags & DRM_MODE_FLAG_DBLCLK) ? 2 : 1));
-
 	HDMI_WRITE(VC4_HDMI_SCHEDULER_CONTROL,
 		   HDMI_READ(VC4_HDMI_SCHEDULER_CONTROL) |
 		   VC4_HDMI_SCHEDULER_CONTROL_MANUAL_FORMAT |
@@ -559,28 +640,6 @@ static void vc4_hdmi_encoder_mode_set(struct drm_encoder *encoder,
 		DRM_INFO("HDMI regs after:\n");
 		vc4_hdmi_dump_regs(dev);
 	}
-}
-
-static void vc4_hdmi_encoder_disable(struct drm_encoder *encoder)
-{
-	struct drm_device *dev = encoder->dev;
-	struct vc4_dev *vc4 = to_vc4_dev(dev);
-
-	HDMI_WRITE(VC4_HDMI_RAM_PACKET_CONFIG, 0);
-
-	HDMI_WRITE(VC4_HDMI_TX_PHY_RESET_CTL, 0xf << 16);
-	HD_WRITE(VC4_HD_VID_CTL,
-		 HD_READ(VC4_HD_VID_CTL) & ~VC4_HD_VID_CTL_ENABLE);
-}
-
-static void vc4_hdmi_encoder_enable(struct drm_encoder *encoder)
-{
-	struct vc4_hdmi_encoder *vc4_encoder = to_vc4_hdmi_encoder(encoder);
-	struct drm_device *dev = encoder->dev;
-	struct vc4_dev *vc4 = to_vc4_dev(dev);
-	int ret;
-
-	HDMI_WRITE(VC4_HDMI_TX_PHY_RESET_CTL, 0);
 
 	HD_WRITE(VC4_HD_VID_CTL,
 		 HD_READ(VC4_HD_VID_CTL) |
@@ -646,7 +705,6 @@ static void vc4_hdmi_encoder_enable(struct drm_encoder *encoder)
 }
 
 static const struct drm_encoder_helper_funcs vc4_hdmi_encoder_helper_funcs = {
-	.mode_set = vc4_hdmi_encoder_mode_set,
 	.disable = vc4_hdmi_encoder_disable,
 	.enable = vc4_hdmi_encoder_enable,
 };
@@ -1147,33 +1205,6 @@ static int vc4_hdmi_bind(struct device *dev, struct device *master, void *data)
 		return -EPROBE_DEFER;
 	}
 
-	/* Enable the clocks at startup.  We can't quite recover from
-	 * turning off the pixel clock during disable/enables yet, so
-	 * it's always running.
-	 */
-	ret = clk_prepare_enable(hdmi->pixel_clock);
-	if (ret) {
-		DRM_ERROR("Failed to turn on pixel clock: %d\n", ret);
-		goto err_put_i2c;
-	}
-
-	/* This is the rate that is set by the firmware.  The number
-	 * needs to be a bit higher than the pixel clock rate
-	 * (generally 148.5Mhz).
-	 */
-	ret = clk_set_rate(hdmi->hsm_clock, 163682864);
-	if (ret) {
-		DRM_ERROR("Failed to set HSM clock rate: %d\n", ret);
-		goto err_unprepare_pix;
-	}
-
-	ret = clk_prepare_enable(hdmi->hsm_clock);
-	if (ret) {
-		DRM_ERROR("Failed to turn on HDMI state machine clock: %d\n",
-			  ret);
-		goto err_unprepare_pix;
-	}
-
 	/* Only use the GPIO HPD pin if present in the DT, otherwise
 	 * we'll use the HDMI core's register.
 	 */
@@ -1185,7 +1216,7 @@ static int vc4_hdmi_bind(struct device *dev, struct device *master, void *data)
 							 &hpd_gpio_flags);
 		if (hdmi->hpd_gpio < 0) {
 			ret = hdmi->hpd_gpio;
-			goto err_unprepare_hsm;
+			goto err_put_i2c;
 		}
 
 		hdmi->hpd_active_low = hpd_gpio_flags & OF_GPIO_ACTIVE_LOW;
@@ -1193,25 +1224,7 @@ static int vc4_hdmi_bind(struct device *dev, struct device *master, void *data)
 
 	vc4->hdmi = hdmi;
 
-	/* HDMI core must be enabled. */
-	if (!(HD_READ(VC4_HD_M_CTL) & VC4_HD_M_ENABLE)) {
-		HD_WRITE(VC4_HD_M_CTL, VC4_HD_M_SW_RST);
-		udelay(1);
-		HD_WRITE(VC4_HD_M_CTL, 0);
-
-		HD_WRITE(VC4_HD_M_CTL, VC4_HD_M_ENABLE);
-
-		HDMI_WRITE(VC4_HDMI_SW_RESET_CONTROL,
-			   VC4_HDMI_SW_RESET_HDMI |
-			   VC4_HDMI_SW_RESET_FORMAT_DETECT);
-
-		HDMI_WRITE(VC4_HDMI_SW_RESET_CONTROL, 0);
-
-		/* PHY should be in reset, like
-		 * vc4_hdmi_encoder_disable() does.
-		 */
-		HDMI_WRITE(VC4_HDMI_TX_PHY_RESET_CTL, 0xf << 16);
-	}
+	pm_runtime_enable(dev);
 
 	drm_encoder_init(drm, hdmi->encoder, &vc4_hdmi_encoder_funcs,
 			 DRM_MODE_ENCODER_TMDS, NULL);
@@ -1231,10 +1244,7 @@ static int vc4_hdmi_bind(struct device *dev, struct device *master, void *data)
 
 err_destroy_encoder:
 	vc4_hdmi_encoder_destroy(hdmi->encoder);
-err_unprepare_hsm:
-	clk_disable_unprepare(hdmi->hsm_clock);
-err_unprepare_pix:
-	clk_disable_unprepare(hdmi->pixel_clock);
+	pm_runtime_disable(dev);
 err_put_i2c:
 	put_device(&hdmi->ddc->dev);
 
@@ -1253,8 +1263,8 @@ static void vc4_hdmi_unbind(struct device *dev, struct device *master,
 	vc4_hdmi_connector_destroy(hdmi->connector);
 	vc4_hdmi_encoder_destroy(hdmi->encoder);
 
-	clk_disable_unprepare(hdmi->pixel_clock);
-	clk_disable_unprepare(hdmi->hsm_clock);
+	pm_runtime_disable(dev);
+
 	put_device(&hdmi->ddc->dev);
 
 	vc4->hdmi = NULL;
diff --git a/drivers/gpu/drm/vc4/vc4_hvs.c b/drivers/gpu/drm/vc4/vc4_hvs.c
index fd421ba3c5d7..2b62fc5b8d85 100644
--- a/drivers/gpu/drm/vc4/vc4_hvs.c
+++ b/drivers/gpu/drm/vc4/vc4_hvs.c
@@ -22,7 +22,7 @@
  * each CRTC.
  */
 
-#include "linux/component.h"
+#include <linux/component.h>
 #include "vc4_drv.h"
 #include "vc4_regs.h"
 
diff --git a/drivers/gpu/drm/vc4/vc4_irq.c b/drivers/gpu/drm/vc4/vc4_irq.c
index cdc6e6760705..7d7af3a93d94 100644
--- a/drivers/gpu/drm/vc4/vc4_irq.c
+++ b/drivers/gpu/drm/vc4/vc4_irq.c
@@ -59,50 +59,45 @@ vc4_overflow_mem_work(struct work_struct *work)
 {
 	struct vc4_dev *vc4 =
 		container_of(work, struct vc4_dev, overflow_mem_work);
-	struct drm_device *dev = vc4->dev;
-	struct vc4_bo *bo;
+	struct vc4_bo *bo = vc4->bin_bo;
+	int bin_bo_slot;
+	struct vc4_exec_info *exec;
+	unsigned long irqflags;
 
-	bo = vc4_bo_create(dev, 256 * 1024, true);
-	if (IS_ERR(bo)) {
+	bin_bo_slot = vc4_v3d_get_bin_slot(vc4);
+	if (bin_bo_slot < 0) {
 		DRM_ERROR("Couldn't allocate binner overflow mem\n");
 		return;
 	}
 
-	/* If there's a job executing currently, then our previous
-	 * overflow allocation is getting used in that job and we need
-	 * to queue it to be released when the job is done.  But if no
-	 * job is executing at all, then we can free the old overflow
-	 * object direcctly.
-	 *
-	 * No lock necessary for this pointer since we're the only
-	 * ones that update the pointer, and our workqueue won't
-	 * reenter.
-	 */
-	if (vc4->overflow_mem) {
-		struct vc4_exec_info *current_exec;
-		unsigned long irqflags;
-
-		spin_lock_irqsave(&vc4->job_lock, irqflags);
-		current_exec = vc4_first_bin_job(vc4);
-		if (!current_exec)
-			current_exec = vc4_last_render_job(vc4);
-		if (current_exec) {
-			vc4->overflow_mem->seqno = current_exec->seqno;
-			list_add_tail(&vc4->overflow_mem->unref_head,
-				      &current_exec->unref_list);
-			vc4->overflow_mem = NULL;
+	spin_lock_irqsave(&vc4->job_lock, irqflags);
+
+	if (vc4->bin_alloc_overflow) {
+		/* If we had overflow memory allocated previously,
+		 * then that chunk will free when the current bin job
+		 * is done.  If we don't have a bin job running, then
+		 * the chunk will be done whenever the list of render
+		 * jobs has drained.
+		 */
+		exec = vc4_first_bin_job(vc4);
+		if (!exec)
+			exec = vc4_last_render_job(vc4);
+		if (exec) {
+			exec->bin_slots |= vc4->bin_alloc_overflow;
+		} else {
+			/* There's nothing queued in the hardware, so
+			 * the old slot is free immediately.
+			 */
+			vc4->bin_alloc_used &= ~vc4->bin_alloc_overflow;
 		}
-		spin_unlock_irqrestore(&vc4->job_lock, irqflags);
 	}
+	vc4->bin_alloc_overflow = BIT(bin_bo_slot);
 
-	if (vc4->overflow_mem)
-		drm_gem_object_unreference_unlocked(&vc4->overflow_mem->base.base);
-	vc4->overflow_mem = bo;
-
-	V3D_WRITE(V3D_BPOA, bo->base.paddr);
+	V3D_WRITE(V3D_BPOA, bo->base.paddr + bin_bo_slot * vc4->bin_alloc_size);
 	V3D_WRITE(V3D_BPOS, bo->base.base.size);
 	V3D_WRITE(V3D_INTCTL, V3D_INT_OUTOMEM);
 	V3D_WRITE(V3D_INTENA, V3D_INT_OUTOMEM);
+	spin_unlock_irqrestore(&vc4->job_lock, irqflags);
 }
 
 static void
@@ -142,6 +137,10 @@ vc4_irq_finish_render_job(struct drm_device *dev)
 
 	vc4->finished_seqno++;
 	list_move_tail(&exec->head, &vc4->job_done_list);
+	if (exec->fence) {
+		dma_fence_signal_locked(exec->fence);
+		exec->fence = NULL;
+	}
 	vc4_submit_next_render_job(dev);
 
 	wake_up_all(&vc4->job_wait_queue);
diff --git a/drivers/gpu/drm/vc4/vc4_kms.c b/drivers/gpu/drm/vc4/vc4_kms.c
index ad7925a9e0ea..928d191ef90f 100644
--- a/drivers/gpu/drm/vc4/vc4_kms.c
+++ b/drivers/gpu/drm/vc4/vc4_kms.c
@@ -14,12 +14,12 @@
  * crtc, HDMI encoder).
  */
 
-#include "drm_crtc.h"
-#include "drm_atomic.h"
-#include "drm_atomic_helper.h"
-#include "drm_crtc_helper.h"
-#include "drm_plane_helper.h"
-#include "drm_fb_cma_helper.h"
+#include <drm/drm_crtc.h>
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_plane_helper.h>
+#include <drm/drm_fb_cma_helper.h>
 #include "vc4_drv.h"
 
 static void vc4_output_poll_changed(struct drm_device *dev)
@@ -230,10 +230,12 @@ int vc4_kms_load(struct drm_device *dev)
 
 	drm_mode_config_reset(dev);
 
-	vc4->fbdev = drm_fbdev_cma_init(dev, 32,
-					dev->mode_config.num_connector);
-	if (IS_ERR(vc4->fbdev))
-		vc4->fbdev = NULL;
+	if (dev->mode_config.num_connector) {
+		vc4->fbdev = drm_fbdev_cma_init(dev, 32,
+						dev->mode_config.num_connector);
+		if (IS_ERR(vc4->fbdev))
+			vc4->fbdev = NULL;
+	}
 
 	drm_kms_helper_poll_init(dev);
 
diff --git a/drivers/gpu/drm/vc4/vc4_plane.c b/drivers/gpu/drm/vc4/vc4_plane.c
index d34cd5393a9b..da18dec21696 100644
--- a/drivers/gpu/drm/vc4/vc4_plane.c
+++ b/drivers/gpu/drm/vc4/vc4_plane.c
@@ -18,12 +18,13 @@
  * into the region of the HVS that it has allocated for us.
  */
 
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_fb_cma_helper.h>
+#include <drm/drm_plane_helper.h>
+
 #include "vc4_drv.h"
 #include "vc4_regs.h"
-#include "drm_atomic.h"
-#include "drm_atomic_helper.h"
-#include "drm_fb_cma_helper.h"
-#include "drm_plane_helper.h"
 
 enum vc4_scaling_mode {
 	VC4_SCALING_NONE,
diff --git a/drivers/gpu/drm/vc4/vc4_render_cl.c b/drivers/gpu/drm/vc4/vc4_render_cl.c
index 4339471f517f..5dc19429d4ae 100644
--- a/drivers/gpu/drm/vc4/vc4_render_cl.c
+++ b/drivers/gpu/drm/vc4/vc4_render_cl.c
@@ -182,8 +182,7 @@ static void emit_tile(struct vc4_exec_info *exec,
 
 	if (has_bin) {
 		rcl_u8(setup, VC4_PACKET_BRANCH_TO_SUB_LIST);
-		rcl_u32(setup, (exec->tile_bo->paddr +
-				exec->tile_alloc_offset +
+		rcl_u32(setup, (exec->tile_alloc_offset +
 				(y * exec->bin_tiles_x + x) * 32));
 	}
 
diff --git a/drivers/gpu/drm/vc4/vc4_v3d.c b/drivers/gpu/drm/vc4/vc4_v3d.c
index 7cc346ad9b0b..5ae5518e605b 100644
--- a/drivers/gpu/drm/vc4/vc4_v3d.c
+++ b/drivers/gpu/drm/vc4/vc4_v3d.c
@@ -16,8 +16,9 @@
  * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
-#include "linux/component.h"
-#include "linux/pm_runtime.h"
+#include <linux/clk.h>
+#include <linux/component.h>
+#include <linux/pm_runtime.h>
 #include "vc4_drv.h"
 #include "vc4_regs.h"
 
@@ -156,6 +157,144 @@ static void vc4_v3d_init_hw(struct drm_device *dev)
 	V3D_WRITE(V3D_VPMBASE, 0);
 }
 
+int vc4_v3d_get_bin_slot(struct vc4_dev *vc4)
+{
+	struct drm_device *dev = vc4->dev;
+	unsigned long irqflags;
+	int slot;
+	uint64_t seqno = 0;
+	struct vc4_exec_info *exec;
+
+try_again:
+	spin_lock_irqsave(&vc4->job_lock, irqflags);
+	slot = ffs(~vc4->bin_alloc_used);
+	if (slot != 0) {
+		/* Switch from ffs() bit index to a 0-based index. */
+		slot--;
+		vc4->bin_alloc_used |= BIT(slot);
+		spin_unlock_irqrestore(&vc4->job_lock, irqflags);
+		return slot;
+	}
+
+	/* Couldn't find an open slot.  Wait for render to complete
+	 * and try again.
+	 */
+	exec = vc4_last_render_job(vc4);
+	if (exec)
+		seqno = exec->seqno;
+	spin_unlock_irqrestore(&vc4->job_lock, irqflags);
+
+	if (seqno) {
+		int ret = vc4_wait_for_seqno(dev, seqno, ~0ull, true);
+
+		if (ret == 0)
+			goto try_again;
+
+		return ret;
+	}
+
+	return -ENOMEM;
+}
+
+/**
+ * vc4_allocate_bin_bo() - allocates the memory that will be used for
+ * tile binning.
+ *
+ * The binner has a limitation that the addresses in the tile state
+ * buffer that point into the tile alloc buffer or binner overflow
+ * memory only have 28 bits (256MB), and the top 4 on the bus for
+ * tile alloc references end up coming from the tile state buffer's
+ * address.
+ *
+ * To work around this, we allocate a single large buffer while V3D is
+ * in use, make sure that it has the top 4 bits constant across its
+ * entire extent, and then put the tile state, tile alloc, and binner
+ * overflow memory inside that buffer.
+ *
+ * This creates a limitation where we may not be able to execute a job
+ * if it doesn't fit within the buffer that we allocated up front.
+ * However, it turns out that 16MB is "enough for anybody", and
+ * real-world applications run into allocation failures from the
+ * overall CMA pool before they make scenes complicated enough to run
+ * out of bin space.
+ */
+int
+vc4_allocate_bin_bo(struct drm_device *drm)
+{
+	struct vc4_dev *vc4 = to_vc4_dev(drm);
+	struct vc4_v3d *v3d = vc4->v3d;
+	uint32_t size = 16 * 1024 * 1024;
+	int ret = 0;
+	struct list_head list;
+
+	/* We may need to try allocating more than once to get a BO
+	 * that doesn't cross 256MB.  Track the ones we've allocated
+	 * that failed so far, so that we can free them when we've got
+	 * one that succeeded (if we freed them right away, our next
+	 * allocation would probably be the same chunk of memory).
+	 */
+	INIT_LIST_HEAD(&list);
+
+	while (true) {
+		struct vc4_bo *bo = vc4_bo_create(drm, size, true);
+
+		if (IS_ERR(bo)) {
+			ret = PTR_ERR(bo);
+
+			dev_err(&v3d->pdev->dev,
+				"Failed to allocate memory for tile binning: "
+				"%d. You may need to enable CMA or give it "
+				"more memory.",
+				ret);
+			break;
+		}
+
+		/* Check if this BO won't trigger the addressing bug. */
+		if ((bo->base.paddr & 0xf0000000) ==
+		    ((bo->base.paddr + bo->base.base.size - 1) & 0xf0000000)) {
+			vc4->bin_bo = bo;
+
+			/* Set up for allocating 512KB chunks of
+			 * binner memory.  The biggest allocation we
+			 * need to do is for the initial tile alloc +
+			 * tile state buffer.  We can render to a
+			 * maximum of ((2048*2048) / (32*32) = 4096
+			 * tiles in a frame (until we do floating
+			 * point rendering, at which point it would be
+			 * 8192).  Tile state is 48b/tile (rounded to
+			 * a page), and tile alloc is 32b/tile
+			 * (rounded to a page), plus a page of extra,
+			 * for a total of 320kb for our worst-case.
+			 * We choose 512kb so that it divides evenly
+			 * into our 16MB, and the rest of the 512kb
+			 * will be used as storage for the overflow
+			 * from the initial 32b CL per bin.
+			 */
+			vc4->bin_alloc_size = 512 * 1024;
+			vc4->bin_alloc_used = 0;
+			vc4->bin_alloc_overflow = 0;
+			WARN_ON_ONCE(sizeof(vc4->bin_alloc_used) * 8 !=
+				     bo->base.base.size / vc4->bin_alloc_size);
+
+			break;
+		}
+
+		/* Put it on the list to free later, and try again. */
+		list_add(&bo->unref_head, &list);
+	}
+
+	/* Free all the BOs we allocated but didn't choose. */
+	while (!list_empty(&list)) {
+		struct vc4_bo *bo = list_last_entry(&list,
+						    struct vc4_bo, unref_head);
+
+		list_del(&bo->unref_head);
+		drm_gem_object_put_unlocked(&bo->base.base);
+	}
+
+	return ret;
+}
+
 #ifdef CONFIG_PM
 static int vc4_v3d_runtime_suspend(struct device *dev)
 {
@@ -164,6 +303,11 @@ static int vc4_v3d_runtime_suspend(struct device *dev)
 
 	vc4_irq_uninstall(vc4->dev);
 
+	drm_gem_object_put_unlocked(&vc4->bin_bo->base.base);
+	vc4->bin_bo = NULL;
+
+	clk_disable_unprepare(v3d->clk);
+
 	return 0;
 }
 
@@ -171,6 +315,15 @@ static int vc4_v3d_runtime_resume(struct device *dev)
 {
 	struct vc4_v3d *v3d = dev_get_drvdata(dev);
 	struct vc4_dev *vc4 = v3d->vc4;
+	int ret;
+
+	ret = vc4_allocate_bin_bo(vc4->dev);
+	if (ret)
+		return ret;
+
+	ret = clk_prepare_enable(v3d->clk);
+	if (ret != 0)
+		return ret;
 
 	vc4_v3d_init_hw(vc4->dev);
 	vc4_irq_postinstall(vc4->dev);
@@ -202,12 +355,38 @@ static int vc4_v3d_bind(struct device *dev, struct device *master, void *data)
 	vc4->v3d = v3d;
 	v3d->vc4 = vc4;
 
+	v3d->clk = devm_clk_get(dev, NULL);
+	if (IS_ERR(v3d->clk)) {
+		int ret = PTR_ERR(v3d->clk);
+
+		if (ret == -ENOENT) {
+			/* bcm2835 didn't have a clock reference in the DT. */
+			ret = 0;
+			v3d->clk = NULL;
+		} else {
+			if (ret != -EPROBE_DEFER)
+				dev_err(dev, "Failed to get V3D clock: %d\n",
+					ret);
+			return ret;
+		}
+	}
+
 	if (V3D_READ(V3D_IDENT0) != V3D_EXPECTED_IDENT0) {
 		DRM_ERROR("V3D_IDENT0 read 0x%08x instead of 0x%08x\n",
 			  V3D_READ(V3D_IDENT0), V3D_EXPECTED_IDENT0);
 		return -EINVAL;
 	}
 
+	ret = clk_prepare_enable(v3d->clk);
+	if (ret != 0)
+		return ret;
+
+	ret = vc4_allocate_bin_bo(drm);
+	if (ret) {
+		clk_disable_unprepare(v3d->clk);
+		return ret;
+	}
+
 	/* Reset the binner overflow address/size at setup, to be sure
 	 * we don't reuse an old one.
 	 */
@@ -271,6 +450,7 @@ static int vc4_v3d_dev_remove(struct platform_device *pdev)
 
 static const struct of_device_id vc4_v3d_dt_match[] = {
 	{ .compatible = "brcm,bcm2835-v3d" },
+	{ .compatible = "brcm,cygnus-v3d" },
 	{ .compatible = "brcm,vc4-v3d" },
 	{}
 };
diff --git a/drivers/gpu/drm/vc4/vc4_validate.c b/drivers/gpu/drm/vc4/vc4_validate.c
index da6f1e138e8d..814b512c6b9a 100644
--- a/drivers/gpu/drm/vc4/vc4_validate.c
+++ b/drivers/gpu/drm/vc4/vc4_validate.c
@@ -172,7 +172,8 @@ vc4_check_tex_size(struct vc4_exec_info *exec, struct drm_gem_cma_object *fbo,
 	 * our math.
 	 */
 	if (width > 4096 || height > 4096) {
-		DRM_ERROR("Surface dimesions (%d,%d) too large", width, height);
+		DRM_ERROR("Surface dimensions (%d,%d) too large",
+			  width, height);
 		return false;
 	}
 
@@ -348,10 +349,11 @@ static int
 validate_tile_binning_config(VALIDATE_ARGS)
 {
 	struct drm_device *dev = exec->exec_bo->base.dev;
-	struct vc4_bo *tile_bo;
+	struct vc4_dev *vc4 = to_vc4_dev(dev);
 	uint8_t flags;
-	uint32_t tile_state_size, tile_alloc_size;
-	uint32_t tile_count;
+	uint32_t tile_state_size;
+	uint32_t tile_count, bin_addr;
+	int bin_slot;
 
 	if (exec->found_tile_binning_mode_config_packet) {
 		DRM_ERROR("Duplicate VC4_PACKET_TILE_BINNING_MODE_CONFIG\n");
@@ -377,13 +379,28 @@ validate_tile_binning_config(VALIDATE_ARGS)
 		return -EINVAL;
 	}
 
+	bin_slot = vc4_v3d_get_bin_slot(vc4);
+	if (bin_slot < 0) {
+		if (bin_slot != -EINTR && bin_slot != -ERESTARTSYS) {
+			DRM_ERROR("Failed to allocate binner memory: %d\n",
+				  bin_slot);
+		}
+		return bin_slot;
+	}
+
+	/* The slot we allocated will only be used by this job, and is
+	 * free when the job completes rendering.
+	 */
+	exec->bin_slots |= BIT(bin_slot);
+	bin_addr = vc4->bin_bo->base.paddr + bin_slot * vc4->bin_alloc_size;
+
 	/* The tile state data array is 48 bytes per tile, and we put it at
 	 * the start of a BO containing both it and the tile alloc.
 	 */
 	tile_state_size = 48 * tile_count;
 
 	/* Since the tile alloc array will follow us, align. */
-	exec->tile_alloc_offset = roundup(tile_state_size, 4096);
+	exec->tile_alloc_offset = bin_addr + roundup(tile_state_size, 4096);
 
 	*(uint8_t *)(validated + 14) =
 		((flags & ~(VC4_BIN_CONFIG_ALLOC_INIT_BLOCK_SIZE_MASK |
@@ -394,35 +411,13 @@ validate_tile_binning_config(VALIDATE_ARGS)
 		 VC4_SET_FIELD(VC4_BIN_CONFIG_ALLOC_BLOCK_SIZE_128,
 			       VC4_BIN_CONFIG_ALLOC_BLOCK_SIZE));
 
-	/* Initial block size. */
-	tile_alloc_size = 32 * tile_count;
-
-	/*
-	 * The initial allocation gets rounded to the next 256 bytes before
-	 * the hardware starts fulfilling further allocations.
-	 */
-	tile_alloc_size = roundup(tile_alloc_size, 256);
-
-	/* Add space for the extra allocations.  This is what gets used first,
-	 * before overflow memory.  It must have at least 4096 bytes, but we
-	 * want to avoid overflow memory usage if possible.
-	 */
-	tile_alloc_size += 1024 * 1024;
-
-	tile_bo = vc4_bo_create(dev, exec->tile_alloc_offset + tile_alloc_size,
-				true);
-	exec->tile_bo = &tile_bo->base;
-	if (IS_ERR(exec->tile_bo))
-		return PTR_ERR(exec->tile_bo);
-	list_add_tail(&tile_bo->unref_head, &exec->unref_list);
-
 	/* tile alloc address. */
-	*(uint32_t *)(validated + 0) = (exec->tile_bo->paddr +
-					exec->tile_alloc_offset);
+	*(uint32_t *)(validated + 0) = exec->tile_alloc_offset;
 	/* tile alloc size. */
-	*(uint32_t *)(validated + 4) = tile_alloc_size;
+	*(uint32_t *)(validated + 4) = (bin_addr + vc4->bin_alloc_size -
+					exec->tile_alloc_offset);
 	/* tile state address. */
-	*(uint32_t *)(validated + 8) = exec->tile_bo->paddr;
+	*(uint32_t *)(validated + 8) = bin_addr;
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/vgem/Makefile b/drivers/gpu/drm/vgem/Makefile
index bfcdea1330e6..cb5d413b9c93 100644
--- a/drivers/gpu/drm/vgem/Makefile
+++ b/drivers/gpu/drm/vgem/Makefile
@@ -1,4 +1,3 @@
-ccflags-y := -Iinclude/drm
 vgem-y := vgem_drv.o vgem_fence.o
 
 obj-$(CONFIG_DRM_VGEM)	+= vgem.o
diff --git a/drivers/gpu/drm/vgem/vgem_drv.c b/drivers/gpu/drm/vgem/vgem_drv.c
index 9fee38a942c4..18f401b442c2 100644
--- a/drivers/gpu/drm/vgem/vgem_drv.c
+++ b/drivers/gpu/drm/vgem/vgem_drv.c
@@ -42,10 +42,20 @@
 #define DRIVER_MAJOR	1
 #define DRIVER_MINOR	0
 
+static struct vgem_device {
+	struct drm_device drm;
+	struct platform_device *platform;
+} *vgem_device;
+
 static void vgem_gem_free_object(struct drm_gem_object *obj)
 {
 	struct drm_vgem_gem_object *vgem_obj = to_vgem_bo(obj);
 
+	kvfree(vgem_obj->pages);
+
+	if (obj->import_attach)
+		drm_prime_gem_destroy(obj, vgem_obj->table);
+
 	drm_gem_object_release(obj);
 	kfree(vgem_obj);
 }
@@ -56,26 +66,49 @@ static int vgem_gem_fault(struct vm_fault *vmf)
 	struct drm_vgem_gem_object *obj = vma->vm_private_data;
 	/* We don't use vmf->pgoff since that has the fake offset */
 	unsigned long vaddr = vmf->address;
-	struct page *page;
-
-	page = shmem_read_mapping_page(file_inode(obj->base.filp)->i_mapping,
-				       (vaddr - vma->vm_start) >> PAGE_SHIFT);
-	if (!IS_ERR(page)) {
-		vmf->page = page;
-		return 0;
-	} else switch (PTR_ERR(page)) {
-		case -ENOSPC:
-		case -ENOMEM:
-			return VM_FAULT_OOM;
-		case -EBUSY:
-			return VM_FAULT_RETRY;
-		case -EFAULT:
-		case -EINVAL:
-			return VM_FAULT_SIGBUS;
-		default:
-			WARN_ON_ONCE(PTR_ERR(page));
-			return VM_FAULT_SIGBUS;
+	int ret;
+	loff_t num_pages;
+	pgoff_t page_offset;
+	page_offset = (vaddr - vma->vm_start) >> PAGE_SHIFT;
+
+	num_pages = DIV_ROUND_UP(obj->base.size, PAGE_SIZE);
+
+	if (page_offset > num_pages)
+		return VM_FAULT_SIGBUS;
+
+	if (obj->pages) {
+		get_page(obj->pages[page_offset]);
+		vmf->page = obj->pages[page_offset];
+		ret = 0;
+	} else {
+		struct page *page;
+
+		page = shmem_read_mapping_page(
+					file_inode(obj->base.filp)->i_mapping,
+					page_offset);
+		if (!IS_ERR(page)) {
+			vmf->page = page;
+			ret = 0;
+		} else switch (PTR_ERR(page)) {
+			case -ENOSPC:
+			case -ENOMEM:
+				ret = VM_FAULT_OOM;
+				break;
+			case -EBUSY:
+				ret = VM_FAULT_RETRY;
+				break;
+			case -EFAULT:
+			case -EINVAL:
+				ret = VM_FAULT_SIGBUS;
+				break;
+			default:
+				WARN_ON(PTR_ERR(page));
+				ret = VM_FAULT_SIGBUS;
+				break;
+		}
+
 	}
+	return ret;
 }
 
 static const struct vm_operations_struct vgem_gem_vm_ops = {
@@ -112,12 +145,8 @@ static void vgem_postclose(struct drm_device *dev, struct drm_file *file)
 	kfree(vfile);
 }
 
-/* ioctls */
-
-static struct drm_gem_object *vgem_gem_create(struct drm_device *dev,
-					      struct drm_file *file,
-					      unsigned int *handle,
-					      unsigned long size)
+static struct drm_vgem_gem_object *__vgem_gem_create(struct drm_device *dev,
+						unsigned long size)
 {
 	struct drm_vgem_gem_object *obj;
 	int ret;
@@ -127,8 +156,31 @@ static struct drm_gem_object *vgem_gem_create(struct drm_device *dev,
 		return ERR_PTR(-ENOMEM);
 
 	ret = drm_gem_object_init(dev, &obj->base, roundup(size, PAGE_SIZE));
-	if (ret)
-		goto err_free;
+	if (ret) {
+		kfree(obj);
+		return ERR_PTR(ret);
+	}
+
+	return obj;
+}
+
+static void __vgem_gem_destroy(struct drm_vgem_gem_object *obj)
+{
+	drm_gem_object_release(&obj->base);
+	kfree(obj);
+}
+
+static struct drm_gem_object *vgem_gem_create(struct drm_device *dev,
+					      struct drm_file *file,
+					      unsigned int *handle,
+					      unsigned long size)
+{
+	struct drm_vgem_gem_object *obj;
+	int ret;
+
+	obj = __vgem_gem_create(dev, size);
+	if (IS_ERR(obj))
+		return ERR_CAST(obj);
 
 	ret = drm_gem_handle_create(file, &obj->base, handle);
 	drm_gem_object_unreference_unlocked(&obj->base);
@@ -137,9 +189,8 @@ static struct drm_gem_object *vgem_gem_create(struct drm_device *dev,
 
 	return &obj->base;
 
-err_free:
-	kfree(obj);
 err:
+	__vgem_gem_destroy(obj);
 	return ERR_PTR(ret);
 }
 
@@ -256,6 +307,37 @@ static struct sg_table *vgem_prime_get_sg_table(struct drm_gem_object *obj)
 	return st;
 }
 
+static struct drm_gem_object* vgem_prime_import(struct drm_device *dev,
+						struct dma_buf *dma_buf)
+{
+	struct vgem_device *vgem = container_of(dev, typeof(*vgem), drm);
+
+	return drm_gem_prime_import_dev(dev, dma_buf, &vgem->platform->dev);
+}
+
+static struct drm_gem_object *vgem_prime_import_sg_table(struct drm_device *dev,
+			struct dma_buf_attachment *attach, struct sg_table *sg)
+{
+	struct drm_vgem_gem_object *obj;
+	int npages;
+
+	obj = __vgem_gem_create(dev, attach->dmabuf->size);
+	if (IS_ERR(obj))
+		return ERR_CAST(obj);
+
+	npages = PAGE_ALIGN(attach->dmabuf->size) / PAGE_SIZE;
+
+	obj->table = sg;
+	obj->pages = kvmalloc_array(npages, sizeof(struct page *), GFP_KERNEL);
+	if (!obj->pages) {
+		__vgem_gem_destroy(obj);
+		return ERR_PTR(-ENOMEM);
+	}
+	drm_prime_sg_to_page_addr_arrays(obj->table, obj->pages, NULL,
+					npages);
+	return &obj->base;
+}
+
 static void *vgem_prime_vmap(struct drm_gem_object *obj)
 {
 	long n_pages = obj->size >> PAGE_SHIFT;
@@ -300,8 +382,19 @@ static int vgem_prime_mmap(struct drm_gem_object *obj,
 	return 0;
 }
 
+static void vgem_release(struct drm_device *dev)
+{
+	struct vgem_device *vgem = container_of(dev, typeof(*vgem), drm);
+
+	platform_device_unregister(vgem->platform);
+	drm_dev_fini(&vgem->drm);
+
+	kfree(vgem);
+}
+
 static struct drm_driver vgem_driver = {
 	.driver_features		= DRIVER_GEM | DRIVER_PRIME,
+	.release			= vgem_release,
 	.open				= vgem_open,
 	.postclose			= vgem_postclose,
 	.gem_free_object_unlocked	= vgem_gem_free_object,
@@ -314,8 +407,11 @@ static struct drm_driver vgem_driver = {
 	.dumb_map_offset		= vgem_gem_dumb_map,
 
 	.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
+	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
 	.gem_prime_pin = vgem_prime_pin,
+	.gem_prime_import = vgem_prime_import,
 	.gem_prime_export = drm_gem_prime_export,
+	.gem_prime_import_sg_table = vgem_prime_import_sg_table,
 	.gem_prime_get_sg_table = vgem_prime_get_sg_table,
 	.gem_prime_vmap = vgem_prime_vmap,
 	.gem_prime_vunmap = vgem_prime_vunmap,
@@ -328,34 +424,48 @@ static struct drm_driver vgem_driver = {
 	.minor	= DRIVER_MINOR,
 };
 
-static struct drm_device *vgem_device;
-
 static int __init vgem_init(void)
 {
 	int ret;
 
-	vgem_device = drm_dev_alloc(&vgem_driver, NULL);
-	if (IS_ERR(vgem_device)) {
-		ret = PTR_ERR(vgem_device);
-		goto out;
+	vgem_device = kzalloc(sizeof(*vgem_device), GFP_KERNEL);
+	if (!vgem_device)
+		return -ENOMEM;
+
+	ret = drm_dev_init(&vgem_device->drm, &vgem_driver, NULL);
+	if (ret)
+		goto out_free;
+
+	vgem_device->platform =
+		platform_device_register_simple("vgem", -1, NULL, 0);
+	if (IS_ERR(vgem_device->platform)) {
+		ret = PTR_ERR(vgem_device->platform);
+		goto out_fini;
 	}
 
-	ret  = drm_dev_register(vgem_device, 0);
+	dma_coerce_mask_and_coherent(&vgem_device->platform->dev,
+				     DMA_BIT_MASK(64));
+
+	/* Final step: expose the device/driver to userspace */
+	ret  = drm_dev_register(&vgem_device->drm, 0);
 	if (ret)
-		goto out_unref;
+		goto out_unregister;
 
 	return 0;
 
-out_unref:
-	drm_dev_unref(vgem_device);
-out:
+out_unregister:
+	platform_device_unregister(vgem_device->platform);
+out_fini:
+	drm_dev_fini(&vgem_device->drm);
+out_free:
+	kfree(vgem_device);
 	return ret;
 }
 
 static void __exit vgem_exit(void)
 {
-	drm_dev_unregister(vgem_device);
-	drm_dev_unref(vgem_device);
+	drm_dev_unregister(&vgem_device->drm);
+	drm_dev_unref(&vgem_device->drm);
 }
 
 module_init(vgem_init);
diff --git a/drivers/gpu/drm/vgem/vgem_drv.h b/drivers/gpu/drm/vgem/vgem_drv.h
index cb59c7ab98b9..1aae01419112 100644
--- a/drivers/gpu/drm/vgem/vgem_drv.h
+++ b/drivers/gpu/drm/vgem/vgem_drv.h
@@ -43,6 +43,8 @@ struct vgem_file {
 #define to_vgem_bo(x) container_of(x, struct drm_vgem_gem_object, base)
 struct drm_vgem_gem_object {
 	struct drm_gem_object base;
+	struct page **pages;
+	struct sg_table *table;
 };
 
 int vgem_fence_open(struct vgem_file *file);
diff --git a/drivers/gpu/drm/via/Makefile b/drivers/gpu/drm/via/Makefile
index d59e258e2c13..751fa8b8a014 100644
--- a/drivers/gpu/drm/via/Makefile
+++ b/drivers/gpu/drm/via/Makefile
@@ -2,7 +2,6 @@
 # Makefile for the drm device driver.  This driver provides support for the
 # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.
 
-ccflags-y := -Iinclude/drm
 via-y    := via_irq.o via_drv.o via_map.o via_mm.o via_dma.o via_verifier.o via_video.o via_dmablit.o
 
 obj-$(CONFIG_DRM_VIA)	+=via.o
diff --git a/drivers/gpu/drm/virtio/Makefile b/drivers/gpu/drm/virtio/Makefile
index 3fb8eac1084f..7684f613bdc3 100644
--- a/drivers/gpu/drm/virtio/Makefile
+++ b/drivers/gpu/drm/virtio/Makefile
@@ -2,8 +2,6 @@
 # Makefile for the drm device driver.  This driver provides support for the
 # Direct Rendering Infrastructure (DRI) in XFree86 4.1.0 and higher.
 
-ccflags-y := -Iinclude/drm
-
 virtio-gpu-y := virtgpu_drv.o virtgpu_kms.o virtgpu_drm_bus.o virtgpu_gem.o \
 	virtgpu_fb.o virtgpu_display.o virtgpu_vq.o virtgpu_ttm.o \
 	virtgpu_fence.o virtgpu_object.o virtgpu_debugfs.o virtgpu_plane.o \
diff --git a/drivers/gpu/drm/virtio/virtgpu_debugfs.c b/drivers/gpu/drm/virtio/virtgpu_debugfs.c
index f51240aa720d..73dc99046c43 100644
--- a/drivers/gpu/drm/virtio/virtgpu_debugfs.c
+++ b/drivers/gpu/drm/virtio/virtgpu_debugfs.c
@@ -24,8 +24,8 @@
  */
 
 #include <linux/debugfs.h>
+#include <drm/drmP.h>
 
-#include "drmP.h"
 #include "virtgpu_drv.h"
 
 static int
diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.c b/drivers/gpu/drm/virtio/virtgpu_drv.c
index 2d29b0141545..63d35c7e416c 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.c
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.c
@@ -29,8 +29,8 @@
 #include <linux/module.h>
 #include <linux/console.h>
 #include <linux/pci.h>
-#include "drmP.h"
-#include "drm/drm.h"
+#include <drm/drmP.h>
+#include <drm/drm.h>
 
 #include "virtgpu_drv.h"
 static struct drm_driver driver;
diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index 1328185bfd59..3a66abb8fd50 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -36,10 +36,10 @@
 #include <drm/drm_atomic.h>
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_encoder.h>
-#include <ttm/ttm_bo_api.h>
-#include <ttm/ttm_bo_driver.h>
-#include <ttm/ttm_placement.h>
-#include <ttm/ttm_module.h>
+#include <drm/ttm/ttm_bo_api.h>
+#include <drm/ttm/ttm_bo_driver.h>
+#include <drm/ttm/ttm_placement.h>
+#include <drm/ttm/ttm_module.h>
 
 #define DRIVER_NAME "virtio_gpu"
 #define DRIVER_DESC "virtio GPU"
diff --git a/drivers/gpu/drm/virtio/virtgpu_ioctl.c b/drivers/gpu/drm/virtio/virtgpu_ioctl.c
index 61f3a963af95..b94bd5440e57 100644
--- a/drivers/gpu/drm/virtio/virtgpu_ioctl.c
+++ b/drivers/gpu/drm/virtio/virtgpu_ioctl.c
@@ -26,9 +26,10 @@
  */
 
 #include <drm/drmP.h>
-#include "virtgpu_drv.h"
 #include <drm/virtgpu_drm.h>
-#include "ttm/ttm_execbuf_util.h"
+#include <drm/ttm/ttm_execbuf_util.h>
+
+#include "virtgpu_drv.h"
 
 static void convert_to_hw_box(struct virtio_gpu_box *dst,
 			      const struct drm_virtgpu_3d_box *src)
@@ -119,13 +120,14 @@ static int virtio_gpu_execbuffer_ioctl(struct drm_device *dev, void *data,
 	INIT_LIST_HEAD(&validate_list);
 	if (exbuf->num_bo_handles) {
 
-		bo_handles = drm_malloc_ab(exbuf->num_bo_handles,
-					   sizeof(uint32_t));
-		buflist = drm_calloc_large(exbuf->num_bo_handles,
-					   sizeof(struct ttm_validate_buffer));
+		bo_handles = kvmalloc_array(exbuf->num_bo_handles,
+					   sizeof(uint32_t), GFP_KERNEL);
+		buflist = kvmalloc_array(exbuf->num_bo_handles,
+					   sizeof(struct ttm_validate_buffer),
+					   GFP_KERNEL | __GFP_ZERO);
 		if (!bo_handles || !buflist) {
-			drm_free_large(bo_handles);
-			drm_free_large(buflist);
+			kvfree(bo_handles);
+			kvfree(buflist);
 			return -ENOMEM;
 		}
 
@@ -133,16 +135,16 @@ static int virtio_gpu_execbuffer_ioctl(struct drm_device *dev, void *data,
 		if (copy_from_user(bo_handles, user_bo_handles,
 				   exbuf->num_bo_handles * sizeof(uint32_t))) {
 			ret = -EFAULT;
-			drm_free_large(bo_handles);
-			drm_free_large(buflist);
+			kvfree(bo_handles);
+			kvfree(buflist);
 			return ret;
 		}
 
 		for (i = 0; i < exbuf->num_bo_handles; i++) {
 			gobj = drm_gem_object_lookup(drm_file, bo_handles[i]);
 			if (!gobj) {
-				drm_free_large(bo_handles);
-				drm_free_large(buflist);
+				kvfree(bo_handles);
+				kvfree(buflist);
 				return -ENOENT;
 			}
 
@@ -151,7 +153,7 @@ static int virtio_gpu_execbuffer_ioctl(struct drm_device *dev, void *data,
 
 			list_add(&buflist[i].head, &validate_list);
 		}
-		drm_free_large(bo_handles);
+		kvfree(bo_handles);
 	}
 
 	ret = virtio_gpu_object_list_validate(&ticket, &validate_list);
@@ -171,7 +173,7 @@ static int virtio_gpu_execbuffer_ioctl(struct drm_device *dev, void *data,
 
 	/* fence the command bo */
 	virtio_gpu_unref_list(&validate_list);
-	drm_free_large(buflist);
+	kvfree(buflist);
 	dma_fence_put(&fence->f);
 	return 0;
 
@@ -179,7 +181,7 @@ out_unresv:
 	ttm_eu_backoff_reservation(&ticket, &validate_list);
 out_free:
 	virtio_gpu_unref_list(&validate_list);
-	drm_free_large(buflist);
+	kvfree(buflist);
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/virtio/virtgpu_ttm.c b/drivers/gpu/drm/virtio/virtgpu_ttm.c
index 4e8e27d50922..c1f2af4ca4ca 100644
--- a/drivers/gpu/drm/virtio/virtgpu_ttm.c
+++ b/drivers/gpu/drm/virtio/virtgpu_ttm.c
@@ -25,11 +25,11 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 
-#include <ttm/ttm_bo_api.h>
-#include <ttm/ttm_bo_driver.h>
-#include <ttm/ttm_placement.h>
-#include <ttm/ttm_page_alloc.h>
-#include <ttm/ttm_module.h>
+#include <drm/ttm/ttm_bo_api.h>
+#include <drm/ttm/ttm_bo_driver.h>
+#include <drm/ttm/ttm_placement.h>
+#include <drm/ttm/ttm_page_alloc.h>
+#include <drm/ttm/ttm_module.h>
 #include <drm/drmP.h>
 #include <drm/drm.h>
 #include <drm/virtgpu_drm.h>
diff --git a/drivers/gpu/drm/vmwgfx/Makefile b/drivers/gpu/drm/vmwgfx/Makefile
index aac17a640cce..a365330bbb82 100644
--- a/drivers/gpu/drm/vmwgfx/Makefile
+++ b/drivers/gpu/drm/vmwgfx/Makefile
@@ -1,6 +1,3 @@
-
-ccflags-y := -Iinclude/drm
-
 vmwgfx-y := vmwgfx_execbuf.o vmwgfx_gmr.o vmwgfx_kms.o vmwgfx_drv.o \
 	    vmwgfx_fb.o vmwgfx_ioctl.o vmwgfx_resource.o vmwgfx_buffer.o \
 	    vmwgfx_fifo.o vmwgfx_irq.o vmwgfx_ldu.o vmwgfx_ttm_glue.o \
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf.c
index 77cb7c627e09..99a7f4ab7d97 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf.c
@@ -25,8 +25,9 @@
  *
  **************************************************************************/
 
+#include <drm/ttm/ttm_bo_api.h>
+
 #include "vmwgfx_drv.h"
-#include "ttm/ttm_bo_api.h"
 
 /*
  * Size of inline command buffers. Try to make sure that a page size is a
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_context.c b/drivers/gpu/drm/vmwgfx/vmwgfx_context.c
index 443d1ed00de7..bcc6d4136c87 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_context.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_context.c
@@ -25,10 +25,11 @@
  *
  **************************************************************************/
 
+#include <drm/ttm/ttm_placement.h>
+
 #include "vmwgfx_drv.h"
 #include "vmwgfx_resource_priv.h"
 #include "vmwgfx_binding.h"
-#include "ttm/ttm_placement.h"
 
 struct vmw_user_context {
 	struct ttm_base_object base;
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_cotable.c b/drivers/gpu/drm/vmwgfx/vmwgfx_cotable.c
index 265c81e6cf39..6c026d75c180 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_cotable.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_cotable.c
@@ -30,9 +30,10 @@
  * whenever the backing MOB is evicted.
  */
 
+#include <drm/ttm/ttm_placement.h>
+
 #include "vmwgfx_drv.h"
 #include "vmwgfx_resource_priv.h"
-#include <ttm/ttm_placement.h>
 #include "vmwgfx_so.h"
 
 /**
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
index 1d2db5d912b0..3d94ea67a825 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
@@ -452,7 +452,7 @@ int vmw_du_primary_plane_atomic_check(struct drm_plane *plane,
 
 	ret = drm_plane_helper_check_update(plane, state->crtc, new_fb,
 					    &src, &dest, &clip,
-					    DRM_ROTATE_0,
+					    DRM_MODE_ROTATE_0,
 					    DRM_PLANE_HELPER_NO_SCALING,
 					    DRM_PLANE_HELPER_NO_SCALING,
 					    false, true, &visible);
@@ -731,7 +731,7 @@ void vmw_du_plane_reset(struct drm_plane *plane)
 
 	plane->state = &vps->base;
 	plane->state->plane = plane;
-	plane->state->rotation = DRM_ROTATE_0;
+	plane->state->rotation = DRM_MODE_ROTATE_0;
 }
 
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c b/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c
index e57a0bad7a62..6063c9636d4a 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c
@@ -30,7 +30,7 @@
 #include <linux/kernel.h>
 #include <linux/frame.h>
 #include <asm/hypervisor.h>
-#include "drmP.h"
+#include <drm/drmP.h>
 #include "vmwgfx_msg.h"
 
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c b/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c
index 92f8b1d04f0f..68f135c5b0d8 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c
@@ -25,10 +25,11 @@
  *
  **************************************************************************/
 
+#include <drm/ttm/ttm_placement.h>
+
 #include "vmwgfx_drv.h"
 #include "vmwgfx_resource_priv.h"
 #include "vmwgfx_binding.h"
-#include "ttm/ttm_placement.h"
 
 struct vmw_shader {
 	struct vmw_resource res;
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c b/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c
index 6b70bd259953..a552e4ea5440 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c
@@ -25,11 +25,12 @@
  *
  **************************************************************************/
 
+#include <drm/ttm/ttm_placement.h>
+
 #include "vmwgfx_drv.h"
 #include "vmwgfx_resource_priv.h"
 #include "vmwgfx_so.h"
 #include "vmwgfx_binding.h"
-#include <ttm/ttm_placement.h>
 #include "device_include/svga3d_surfacedefs.h"
 
 
diff --git a/drivers/gpu/drm/zte/Makefile b/drivers/gpu/drm/zte/Makefile
index 01352b56c418..9df7766a7f9d 100644
--- a/drivers/gpu/drm/zte/Makefile
+++ b/drivers/gpu/drm/zte/Makefile
@@ -3,6 +3,7 @@ zxdrm-y := \
 	zx_hdmi.o \
 	zx_plane.o \
 	zx_tvenc.o \
+	zx_vga.o \
 	zx_vou.o
 
 obj-$(CONFIG_DRM_ZTE) += zxdrm.o
diff --git a/drivers/gpu/drm/zte/zx_common_regs.h b/drivers/gpu/drm/zte/zx_common_regs.h
new file mode 100644
index 000000000000..2afd80664c51
--- /dev/null
+++ b/drivers/gpu/drm/zte/zx_common_regs.h
@@ -0,0 +1,31 @@
+/*
+ * Copyright (C) 2017 Sanechips Technology Co., Ltd.
+ * Copyright 2017 Linaro Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef __ZX_COMMON_REGS_H__
+#define __ZX_COMMON_REGS_H__
+
+/* CSC registers */
+#define CSC_CTRL0			0x30
+#define CSC_COV_MODE_SHIFT		16
+#define CSC_COV_MODE_MASK		(0xffff << CSC_COV_MODE_SHIFT)
+#define CSC_BT601_IMAGE_RGB2YCBCR	0
+#define CSC_BT601_IMAGE_YCBCR2RGB	1
+#define CSC_BT601_VIDEO_RGB2YCBCR	2
+#define CSC_BT601_VIDEO_YCBCR2RGB	3
+#define CSC_BT709_IMAGE_RGB2YCBCR	4
+#define CSC_BT709_IMAGE_YCBCR2RGB	5
+#define CSC_BT709_VIDEO_RGB2YCBCR	6
+#define CSC_BT709_VIDEO_YCBCR2RGB	7
+#define CSC_BT2020_IMAGE_RGB2YCBCR	8
+#define CSC_BT2020_IMAGE_YCBCR2RGB	9
+#define CSC_BT2020_VIDEO_RGB2YCBCR	10
+#define CSC_BT2020_VIDEO_YCBCR2RGB	11
+#define CSC_WORK_ENABLE			BIT(0)
+
+#endif /* __ZX_COMMON_REGS_H__ */
diff --git a/drivers/gpu/drm/zte/zx_drm_drv.c b/drivers/gpu/drm/zte/zx_drm_drv.c
index 614e854f6be5..490aafc99610 100644
--- a/drivers/gpu/drm/zte/zx_drm_drv.c
+++ b/drivers/gpu/drm/zte/zx_drm_drv.c
@@ -233,6 +233,7 @@ static struct platform_driver *drivers[] = {
 	&zx_crtc_driver,
 	&zx_hdmi_driver,
 	&zx_tvenc_driver,
+	&zx_vga_driver,
 	&zx_drm_platform_driver,
 };
 
diff --git a/drivers/gpu/drm/zte/zx_drm_drv.h b/drivers/gpu/drm/zte/zx_drm_drv.h
index 5ca035b079c7..2a8cdc5f8be4 100644
--- a/drivers/gpu/drm/zte/zx_drm_drv.h
+++ b/drivers/gpu/drm/zte/zx_drm_drv.h
@@ -14,6 +14,7 @@
 extern struct platform_driver zx_crtc_driver;
 extern struct platform_driver zx_hdmi_driver;
 extern struct platform_driver zx_tvenc_driver;
+extern struct platform_driver zx_vga_driver;
 
 static inline u32 zx_readl(void __iomem *reg)
 {
diff --git a/drivers/gpu/drm/zte/zx_plane.c b/drivers/gpu/drm/zte/zx_plane.c
index d646ac931663..4a6252720c10 100644
--- a/drivers/gpu/drm/zte/zx_plane.c
+++ b/drivers/gpu/drm/zte/zx_plane.c
@@ -16,6 +16,7 @@
 #include <drm/drm_plane_helper.h>
 #include <drm/drmP.h>
 
+#include "zx_common_regs.h"
 #include "zx_drm_drv.h"
 #include "zx_plane.h"
 #include "zx_plane_regs.h"
diff --git a/drivers/gpu/drm/zte/zx_plane_regs.h b/drivers/gpu/drm/zte/zx_plane_regs.h
index 65f271aeabed..9c655f59f9f7 100644
--- a/drivers/gpu/drm/zte/zx_plane_regs.h
+++ b/drivers/gpu/drm/zte/zx_plane_regs.h
@@ -77,24 +77,6 @@
 #define LUMA_STRIDE(x)	 (((x) << LUMA_STRIDE_SHIFT) & LUMA_STRIDE_MASK)
 #define CHROMA_STRIDE(x) (((x) << CHROMA_STRIDE_SHIFT) & CHROMA_STRIDE_MASK)
 
-/* CSC registers */
-#define CSC_CTRL0			0x30
-#define CSC_COV_MODE_SHIFT		16
-#define CSC_COV_MODE_MASK		(0xffff << CSC_COV_MODE_SHIFT)
-#define CSC_BT601_IMAGE_RGB2YCBCR	0
-#define CSC_BT601_IMAGE_YCBCR2RGB	1
-#define CSC_BT601_VIDEO_RGB2YCBCR	2
-#define CSC_BT601_VIDEO_YCBCR2RGB	3
-#define CSC_BT709_IMAGE_RGB2YCBCR	4
-#define CSC_BT709_IMAGE_YCBCR2RGB	5
-#define CSC_BT709_VIDEO_RGB2YCBCR	6
-#define CSC_BT709_VIDEO_YCBCR2RGB	7
-#define CSC_BT2020_IMAGE_RGB2YCBCR	8
-#define CSC_BT2020_IMAGE_YCBCR2RGB	9
-#define CSC_BT2020_VIDEO_RGB2YCBCR	10
-#define CSC_BT2020_VIDEO_YCBCR2RGB	11
-#define CSC_WORK_ENABLE			BIT(0)
-
 /* RSZ registers */
 #define RSZ_SRC_CFG			0x00
 #define RSZ_DEST_CFG			0x04
diff --git a/drivers/gpu/drm/zte/zx_vga.c b/drivers/gpu/drm/zte/zx_vga.c
new file mode 100644
index 000000000000..1e0811f775cb
--- /dev/null
+++ b/drivers/gpu/drm/zte/zx_vga.c
@@ -0,0 +1,531 @@
+/*
+ * Copyright (C) 2017 Sanechips Technology Co., Ltd.
+ * Copyright 2017 Linaro Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/clk.h>
+#include <linux/component.h>
+#include <linux/mfd/syscon.h>
+#include <linux/regmap.h>
+
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drmP.h>
+
+#include "zx_drm_drv.h"
+#include "zx_vga_regs.h"
+#include "zx_vou.h"
+
+struct zx_vga_pwrctrl {
+	struct regmap *regmap;
+	u32 reg;
+	u32 mask;
+};
+
+struct zx_vga_i2c {
+	struct i2c_adapter adap;
+	struct mutex lock;
+};
+
+struct zx_vga {
+	struct drm_connector connector;
+	struct drm_encoder encoder;
+	struct zx_vga_i2c *ddc;
+	struct device *dev;
+	void __iomem *mmio;
+	struct clk *i2c_wclk;
+	struct zx_vga_pwrctrl pwrctrl;
+	struct completion complete;
+	bool connected;
+};
+
+#define to_zx_vga(x) container_of(x, struct zx_vga, x)
+
+static void zx_vga_encoder_enable(struct drm_encoder *encoder)
+{
+	struct zx_vga *vga = to_zx_vga(encoder);
+	struct zx_vga_pwrctrl *pwrctrl = &vga->pwrctrl;
+
+	/* Set bit to power up VGA DACs */
+	regmap_update_bits(pwrctrl->regmap, pwrctrl->reg, pwrctrl->mask,
+			   pwrctrl->mask);
+
+	vou_inf_enable(VOU_VGA, encoder->crtc);
+}
+
+static void zx_vga_encoder_disable(struct drm_encoder *encoder)
+{
+	struct zx_vga *vga = to_zx_vga(encoder);
+	struct zx_vga_pwrctrl *pwrctrl = &vga->pwrctrl;
+
+	vou_inf_disable(VOU_VGA, encoder->crtc);
+
+	/* Clear bit to power down VGA DACs */
+	regmap_update_bits(pwrctrl->regmap, pwrctrl->reg, pwrctrl->mask, 0);
+}
+
+static const struct drm_encoder_helper_funcs zx_vga_encoder_helper_funcs = {
+	.enable	= zx_vga_encoder_enable,
+	.disable = zx_vga_encoder_disable,
+};
+
+static const struct drm_encoder_funcs zx_vga_encoder_funcs = {
+	.destroy = drm_encoder_cleanup,
+};
+
+static int zx_vga_connector_get_modes(struct drm_connector *connector)
+{
+	struct zx_vga *vga = to_zx_vga(connector);
+	struct edid *edid;
+	int ret;
+
+	/*
+	 * Clear both detection bits to switch I2C bus from device
+	 * detecting to EDID reading.
+	 */
+	zx_writel(vga->mmio + VGA_AUTO_DETECT_SEL, 0);
+
+	edid = drm_get_edid(connector, &vga->ddc->adap);
+	if (!edid) {
+		/*
+		 * If EDID reading fails, we set the device state into
+		 * disconnected.  Locking is not required here, since the
+		 * VGA_AUTO_DETECT_SEL register write in irq handler cannot
+		 * be triggered when both detection bits are cleared as above.
+		 */
+		zx_writel(vga->mmio + VGA_AUTO_DETECT_SEL,
+			  VGA_DETECT_SEL_NO_DEVICE);
+		vga->connected = false;
+		return 0;
+	}
+
+	/*
+	 * As edid reading succeeds, device must be connected, so we set
+	 * up detection bit for unplug interrupt here.
+	 */
+	zx_writel(vga->mmio + VGA_AUTO_DETECT_SEL, VGA_DETECT_SEL_HAS_DEVICE);
+
+	drm_mode_connector_update_edid_property(connector, edid);
+	ret = drm_add_edid_modes(connector, edid);
+	kfree(edid);
+
+	return ret;
+}
+
+static enum drm_mode_status
+zx_vga_connector_mode_valid(struct drm_connector *connector,
+			    struct drm_display_mode *mode)
+{
+	return MODE_OK;
+}
+
+static struct drm_connector_helper_funcs zx_vga_connector_helper_funcs = {
+	.get_modes = zx_vga_connector_get_modes,
+	.mode_valid = zx_vga_connector_mode_valid,
+};
+
+static enum drm_connector_status
+zx_vga_connector_detect(struct drm_connector *connector, bool force)
+{
+	struct zx_vga *vga = to_zx_vga(connector);
+
+	return vga->connected ? connector_status_connected :
+				connector_status_disconnected;
+}
+
+static const struct drm_connector_funcs zx_vga_connector_funcs = {
+	.dpms = drm_atomic_helper_connector_dpms,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.detect = zx_vga_connector_detect,
+	.destroy = drm_connector_cleanup,
+	.reset = drm_atomic_helper_connector_reset,
+	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
+};
+
+static int zx_vga_register(struct drm_device *drm, struct zx_vga *vga)
+{
+	struct drm_encoder *encoder = &vga->encoder;
+	struct drm_connector *connector = &vga->connector;
+	struct device *dev = vga->dev;
+	int ret;
+
+	encoder->possible_crtcs = VOU_CRTC_MASK;
+
+	ret = drm_encoder_init(drm, encoder, &zx_vga_encoder_funcs,
+			       DRM_MODE_ENCODER_DAC, NULL);
+	if (ret) {
+		DRM_DEV_ERROR(dev, "failed to init encoder: %d\n", ret);
+		return ret;
+	};
+
+	drm_encoder_helper_add(encoder, &zx_vga_encoder_helper_funcs);
+
+	vga->connector.polled = DRM_CONNECTOR_POLL_HPD;
+
+	ret = drm_connector_init(drm, connector, &zx_vga_connector_funcs,
+				 DRM_MODE_CONNECTOR_VGA);
+	if (ret) {
+		DRM_DEV_ERROR(dev, "failed to init connector: %d\n", ret);
+		goto clean_encoder;
+	};
+
+	drm_connector_helper_add(connector, &zx_vga_connector_helper_funcs);
+
+	ret = drm_mode_connector_attach_encoder(connector, encoder);
+	if (ret) {
+		DRM_DEV_ERROR(dev, "failed to attach encoder: %d\n", ret);
+		goto clean_connector;
+	};
+
+	return 0;
+
+clean_connector:
+	drm_connector_cleanup(connector);
+clean_encoder:
+	drm_encoder_cleanup(encoder);
+	return ret;
+}
+
+static int zx_vga_pwrctrl_init(struct zx_vga *vga)
+{
+	struct zx_vga_pwrctrl *pwrctrl = &vga->pwrctrl;
+	struct device *dev = vga->dev;
+	struct of_phandle_args out_args;
+	struct regmap *regmap;
+	int ret;
+
+	ret = of_parse_phandle_with_fixed_args(dev->of_node,
+				"zte,vga-power-control", 2, 0, &out_args);
+	if (ret)
+		return ret;
+
+	regmap = syscon_node_to_regmap(out_args.np);
+	if (IS_ERR(regmap)) {
+		ret = PTR_ERR(regmap);
+		goto out;
+	}
+
+	pwrctrl->regmap = regmap;
+	pwrctrl->reg = out_args.args[0];
+	pwrctrl->mask = out_args.args[1];
+
+out:
+	of_node_put(out_args.np);
+	return ret;
+}
+
+static int zx_vga_i2c_read(struct zx_vga *vga, struct i2c_msg *msg)
+{
+	int len = msg->len;
+	u8 *buf = msg->buf;
+	u32 offset = 0;
+	int i;
+
+	reinit_completion(&vga->complete);
+
+	/* Select combo write */
+	zx_writel_mask(vga->mmio + VGA_CMD_CFG, VGA_CMD_COMBO, VGA_CMD_COMBO);
+	zx_writel_mask(vga->mmio + VGA_CMD_CFG, VGA_CMD_RW, 0);
+
+	while (len > 0) {
+		u32 cnt;
+
+		/* Clear RX FIFO */
+		zx_writel_mask(vga->mmio + VGA_RXF_CTRL, VGA_RX_FIFO_CLEAR,
+			       VGA_RX_FIFO_CLEAR);
+
+		/* Data offset to read from */
+		zx_writel(vga->mmio + VGA_SUB_ADDR, offset);
+
+		/* Kick off the transfer */
+		zx_writel_mask(vga->mmio + VGA_CMD_CFG, VGA_CMD_TRANS,
+			       VGA_CMD_TRANS);
+
+		if (!wait_for_completion_timeout(&vga->complete,
+						 msecs_to_jiffies(1000))) {
+			DRM_DEV_ERROR(vga->dev, "transfer timeout\n");
+			return -ETIMEDOUT;
+		}
+
+		cnt = zx_readl(vga->mmio + VGA_RXF_STATUS);
+		cnt = (cnt & VGA_RXF_COUNT_MASK) >> VGA_RXF_COUNT_SHIFT;
+		/* FIFO status may report more data than we need to read */
+		cnt = min_t(u32, len, cnt);
+
+		for (i = 0; i < cnt; i++)
+			*buf++ = zx_readl(vga->mmio + VGA_DATA);
+
+		len -= cnt;
+		offset += cnt;
+	}
+
+	return 0;
+}
+
+static int zx_vga_i2c_write(struct zx_vga *vga, struct i2c_msg *msg)
+{
+	/*
+	 * The DDC I2C adapter is only for reading EDID data, so we assume
+	 * that the write to this adapter must be the EDID data offset.
+	 */
+	if ((msg->len != 1) || ((msg->addr != DDC_ADDR)))
+		return -EINVAL;
+
+	/* Hardware will take care of the slave address shifting */
+	zx_writel(vga->mmio + VGA_DEVICE_ADDR, msg->addr);
+
+	return 0;
+}
+
+static int zx_vga_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg *msgs,
+			   int num)
+{
+	struct zx_vga *vga = i2c_get_adapdata(adap);
+	struct zx_vga_i2c *ddc = vga->ddc;
+	int ret = 0;
+	int i;
+
+	mutex_lock(&ddc->lock);
+
+	for (i = 0; i < num; i++) {
+		if (msgs[i].flags & I2C_M_RD)
+			ret = zx_vga_i2c_read(vga, &msgs[i]);
+		else
+			ret = zx_vga_i2c_write(vga, &msgs[i]);
+
+		if (ret < 0)
+			break;
+	}
+
+	if (!ret)
+		ret = num;
+
+	mutex_unlock(&ddc->lock);
+
+	return ret;
+}
+
+static u32 zx_vga_i2c_func(struct i2c_adapter *adapter)
+{
+	return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL;
+}
+
+static const struct i2c_algorithm zx_vga_algorithm = {
+	.master_xfer	= zx_vga_i2c_xfer,
+	.functionality	= zx_vga_i2c_func,
+};
+
+static int zx_vga_ddc_register(struct zx_vga *vga)
+{
+	struct device *dev = vga->dev;
+	struct i2c_adapter *adap;
+	struct zx_vga_i2c *ddc;
+	int ret;
+
+	ddc = devm_kzalloc(dev, sizeof(*ddc), GFP_KERNEL);
+	if (!ddc)
+		return -ENOMEM;
+
+	vga->ddc = ddc;
+	mutex_init(&ddc->lock);
+
+	adap = &ddc->adap;
+	adap->owner = THIS_MODULE;
+	adap->class = I2C_CLASS_DDC;
+	adap->dev.parent = dev;
+	adap->algo = &zx_vga_algorithm;
+	snprintf(adap->name, sizeof(adap->name), "zx vga i2c");
+
+	ret = i2c_add_adapter(adap);
+	if (ret) {
+		DRM_DEV_ERROR(dev, "failed to add I2C adapter: %d\n", ret);
+		return ret;
+	}
+
+	i2c_set_adapdata(adap, vga);
+
+	return 0;
+}
+
+static irqreturn_t zx_vga_irq_thread(int irq, void *dev_id)
+{
+	struct zx_vga *vga = dev_id;
+
+	drm_helper_hpd_irq_event(vga->connector.dev);
+
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t zx_vga_irq_handler(int irq, void *dev_id)
+{
+	struct zx_vga *vga = dev_id;
+	u32 status;
+
+	status = zx_readl(vga->mmio + VGA_I2C_STATUS);
+
+	/* Clear interrupt status */
+	zx_writel_mask(vga->mmio + VGA_I2C_STATUS, VGA_CLEAR_IRQ,
+		       VGA_CLEAR_IRQ);
+
+	if (status & VGA_DEVICE_CONNECTED) {
+		/*
+		 * Since VGA_DETECT_SEL bits need to be reset for switching DDC
+		 * bus from device detection to EDID read, rather than setting
+		 * up HAS_DEVICE bit here, we need to do that in .get_modes
+		 * hook for unplug detecting after EDID read succeeds.
+		 */
+		vga->connected = true;
+		return IRQ_WAKE_THREAD;
+	}
+
+	if (status & VGA_DEVICE_DISCONNECTED) {
+		zx_writel(vga->mmio + VGA_AUTO_DETECT_SEL,
+			  VGA_DETECT_SEL_NO_DEVICE);
+		vga->connected = false;
+		return IRQ_WAKE_THREAD;
+	}
+
+	if (status & VGA_TRANS_DONE) {
+		complete(&vga->complete);
+		return IRQ_HANDLED;
+	}
+
+	return IRQ_NONE;
+}
+
+static void zx_vga_hw_init(struct zx_vga *vga)
+{
+	unsigned long ref = clk_get_rate(vga->i2c_wclk);
+	int div;
+
+	/*
+	 * Set up I2C fast speed divider per formula below to get 400kHz.
+	 *   scl = ref / ((div + 1) * 4)
+	 */
+	div = DIV_ROUND_UP(ref / 1000, 400 * 4) - 1;
+	zx_writel(vga->mmio + VGA_CLK_DIV_FS, div);
+
+	/* Set up device detection */
+	zx_writel(vga->mmio + VGA_AUTO_DETECT_PARA, 0x80);
+	zx_writel(vga->mmio + VGA_AUTO_DETECT_SEL, VGA_DETECT_SEL_NO_DEVICE);
+
+	/*
+	 * We need to poke monitor via DDC bus to get connection irq
+	 * start working.
+	 */
+	zx_writel(vga->mmio + VGA_DEVICE_ADDR, DDC_ADDR);
+	zx_writel_mask(vga->mmio + VGA_CMD_CFG, VGA_CMD_TRANS, VGA_CMD_TRANS);
+}
+
+static int zx_vga_bind(struct device *dev, struct device *master, void *data)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct drm_device *drm = data;
+	struct resource *res;
+	struct zx_vga *vga;
+	int irq;
+	int ret;
+
+	vga = devm_kzalloc(dev, sizeof(*vga), GFP_KERNEL);
+	if (!vga)
+		return -ENOMEM;
+
+	vga->dev = dev;
+	dev_set_drvdata(dev, vga);
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	vga->mmio = devm_ioremap_resource(dev, res);
+	if (IS_ERR(vga->mmio))
+		return PTR_ERR(vga->mmio);
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0)
+		return irq;
+
+	vga->i2c_wclk = devm_clk_get(dev, "i2c_wclk");
+	if (IS_ERR(vga->i2c_wclk)) {
+		ret = PTR_ERR(vga->i2c_wclk);
+		DRM_DEV_ERROR(dev, "failed to get i2c_wclk: %d\n", ret);
+		return ret;
+	}
+
+	ret = zx_vga_pwrctrl_init(vga);
+	if (ret) {
+		DRM_DEV_ERROR(dev, "failed to init power control: %d\n", ret);
+		return ret;
+	}
+
+	ret = zx_vga_ddc_register(vga);
+	if (ret) {
+		DRM_DEV_ERROR(dev, "failed to register ddc: %d\n", ret);
+		return ret;
+	}
+
+	ret = zx_vga_register(drm, vga);
+	if (ret) {
+		DRM_DEV_ERROR(dev, "failed to register vga: %d\n", ret);
+		return ret;
+	}
+
+	init_completion(&vga->complete);
+
+	ret = devm_request_threaded_irq(dev, irq, zx_vga_irq_handler,
+					zx_vga_irq_thread, IRQF_SHARED,
+					dev_name(dev), vga);
+	if (ret) {
+		DRM_DEV_ERROR(dev, "failed to request threaded irq: %d\n", ret);
+		return ret;
+	}
+
+	ret = clk_prepare_enable(vga->i2c_wclk);
+	if (ret)
+		return ret;
+
+	zx_vga_hw_init(vga);
+
+	return 0;
+}
+
+static void zx_vga_unbind(struct device *dev, struct device *master,
+			  void *data)
+{
+	struct zx_vga *vga = dev_get_drvdata(dev);
+
+	clk_disable_unprepare(vga->i2c_wclk);
+}
+
+static const struct component_ops zx_vga_component_ops = {
+	.bind = zx_vga_bind,
+	.unbind = zx_vga_unbind,
+};
+
+static int zx_vga_probe(struct platform_device *pdev)
+{
+	return component_add(&pdev->dev, &zx_vga_component_ops);
+}
+
+static int zx_vga_remove(struct platform_device *pdev)
+{
+	component_del(&pdev->dev, &zx_vga_component_ops);
+	return 0;
+}
+
+static const struct of_device_id zx_vga_of_match[] = {
+	{ .compatible = "zte,zx296718-vga", },
+	{ /* end */ },
+};
+MODULE_DEVICE_TABLE(of, zx_vga_of_match);
+
+struct platform_driver zx_vga_driver = {
+	.probe = zx_vga_probe,
+	.remove = zx_vga_remove,
+	.driver	= {
+		.name = "zx-vga",
+		.of_match_table	= zx_vga_of_match,
+	},
+};
diff --git a/drivers/gpu/drm/zte/zx_vga_regs.h b/drivers/gpu/drm/zte/zx_vga_regs.h
new file mode 100644
index 000000000000..feaa345fe6a6
--- /dev/null
+++ b/drivers/gpu/drm/zte/zx_vga_regs.h
@@ -0,0 +1,36 @@
+/*
+ * Copyright (C) 2017 Sanechips Technology Co., Ltd.
+ * Copyright 2017 Linaro Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef __ZX_VGA_REGS_H__
+#define __ZX_VGA_REGS_H__
+
+#define VGA_CMD_CFG			0x04
+#define VGA_CMD_TRANS			BIT(6)
+#define VGA_CMD_COMBO			BIT(5)
+#define VGA_CMD_RW			BIT(4)
+#define VGA_SUB_ADDR			0x0c
+#define VGA_DEVICE_ADDR			0x10
+#define VGA_CLK_DIV_FS			0x14
+#define VGA_RXF_CTRL			0x20
+#define VGA_RX_FIFO_CLEAR		BIT(7)
+#define VGA_DATA			0x24
+#define VGA_I2C_STATUS			0x28
+#define VGA_DEVICE_DISCONNECTED		BIT(7)
+#define VGA_DEVICE_CONNECTED		BIT(6)
+#define VGA_CLEAR_IRQ			BIT(4)
+#define VGA_TRANS_DONE			BIT(0)
+#define VGA_RXF_STATUS			0x30
+#define VGA_RXF_COUNT_SHIFT		2
+#define VGA_RXF_COUNT_MASK		GENMASK(7, 2)
+#define VGA_AUTO_DETECT_PARA		0x34
+#define VGA_AUTO_DETECT_SEL		0x38
+#define VGA_DETECT_SEL_HAS_DEVICE	BIT(1)
+#define VGA_DETECT_SEL_NO_DEVICE	BIT(0)
+
+#endif /* __ZX_VGA_REGS_H__ */
diff --git a/drivers/gpu/drm/zte/zx_vou.c b/drivers/gpu/drm/zte/zx_vou.c
index b500c8dd0d9d..5fbd10b60ee5 100644
--- a/drivers/gpu/drm/zte/zx_vou.c
+++ b/drivers/gpu/drm/zte/zx_vou.c
@@ -23,6 +23,7 @@
 #include <drm/drm_plane_helper.h>
 #include <drm/drmP.h>
 
+#include "zx_common_regs.h"
 #include "zx_drm_drv.h"
 #include "zx_plane.h"
 #include "zx_vou.h"
@@ -122,6 +123,8 @@ struct zx_crtc {
 	struct drm_plane *primary;
 	struct zx_vou_hw *vou;
 	void __iomem *chnreg;
+	void __iomem *chncsc;
+	void __iomem *dither;
 	const struct zx_crtc_regs *regs;
 	const struct zx_crtc_bits *bits;
 	enum vou_chn_type chn_type;
@@ -204,6 +207,11 @@ static struct vou_inf vou_infs[] = {
 		.clocks_en_bits = BIT(15),
 		.clocks_sel_bits = BIT(11) | BIT(0),
 	},
+	[VOU_VGA] = {
+		.data_sel = VOU_RGB_888,
+		.clocks_en_bits = BIT(1),
+		.clocks_sel_bits = BIT(10),
+	},
 };
 
 static inline struct zx_vou_hw *crtc_to_vou(struct drm_crtc *crtc)
@@ -227,9 +235,26 @@ void vou_inf_enable(enum vou_inf_id id, struct drm_crtc *crtc)
 	struct zx_crtc *zcrtc = to_zx_crtc(crtc);
 	struct zx_vou_hw *vou = zcrtc->vou;
 	struct vou_inf *inf = &vou_infs[id];
+	void __iomem *dither = zcrtc->dither;
+	void __iomem *csc = zcrtc->chncsc;
 	bool is_main = zcrtc->chn_type == VOU_CHN_MAIN;
 	u32 data_sel_shift = id << 1;
 
+	if (inf->data_sel != VOU_YUV444) {
+		/* Enable channel CSC for RGB output */
+		zx_writel_mask(csc + CSC_CTRL0, CSC_COV_MODE_MASK,
+			       CSC_BT709_IMAGE_YCBCR2RGB << CSC_COV_MODE_SHIFT);
+		zx_writel_mask(csc + CSC_CTRL0, CSC_WORK_ENABLE,
+			       CSC_WORK_ENABLE);
+
+		/* Bypass Dither block for RGB output */
+		zx_writel_mask(dither + OSD_DITHER_CTRL0, DITHER_BYSPASS,
+			       DITHER_BYSPASS);
+	} else {
+		zx_writel_mask(csc + CSC_CTRL0, CSC_WORK_ENABLE, 0);
+		zx_writel_mask(dither + OSD_DITHER_CTRL0, DITHER_BYSPASS, 0);
+	}
+
 	/* Select data format */
 	zx_writel_mask(vou->vouctl + VOU_INF_DATA_SEL, 0x3 << data_sel_shift,
 		       inf->data_sel << data_sel_shift);
@@ -525,20 +550,24 @@ static int zx_crtc_init(struct drm_device *drm, struct zx_vou_hw *vou,
 
 	if (chn_type == VOU_CHN_MAIN) {
 		zplane->layer = vou->osd + MAIN_GL_OFFSET;
-		zplane->csc = vou->osd + MAIN_CSC_OFFSET;
+		zplane->csc = vou->osd + MAIN_GL_CSC_OFFSET;
 		zplane->hbsc = vou->osd + MAIN_HBSC_OFFSET;
 		zplane->rsz = vou->otfppu + MAIN_RSZ_OFFSET;
 		zplane->bits = &zx_gl_bits[0];
 		zcrtc->chnreg = vou->osd + OSD_MAIN_CHN;
+		zcrtc->chncsc = vou->osd + MAIN_CHN_CSC_OFFSET;
+		zcrtc->dither = vou->osd + MAIN_DITHER_OFFSET;
 		zcrtc->regs = &main_crtc_regs;
 		zcrtc->bits = &main_crtc_bits;
 	} else {
 		zplane->layer = vou->osd + AUX_GL_OFFSET;
-		zplane->csc = vou->osd + AUX_CSC_OFFSET;
+		zplane->csc = vou->osd + AUX_GL_CSC_OFFSET;
 		zplane->hbsc = vou->osd + AUX_HBSC_OFFSET;
 		zplane->rsz = vou->otfppu + AUX_RSZ_OFFSET;
 		zplane->bits = &zx_gl_bits[1];
 		zcrtc->chnreg = vou->osd + OSD_AUX_CHN;
+		zcrtc->chncsc = vou->osd + AUX_CHN_CSC_OFFSET;
+		zcrtc->dither = vou->osd + AUX_DITHER_OFFSET;
 		zcrtc->regs = &aux_crtc_regs;
 		zcrtc->bits = &aux_crtc_bits;
 	}
@@ -705,9 +734,6 @@ static void vou_hw_init(struct zx_vou_hw *vou)
 	/* Release reset for all VOU modules */
 	zx_writel(vou->vouctl + VOU_SOFT_RST, ~0);
 
-	/* Enable clock auto-gating for all VOU modules */
-	zx_writel(vou->vouctl + VOU_CLK_REQEN, ~0);
-
 	/* Enable all VOU module clocks */
 	zx_writel(vou->vouctl + VOU_CLK_EN, ~0);
 
diff --git a/drivers/gpu/drm/zte/zx_vou_regs.h b/drivers/gpu/drm/zte/zx_vou_regs.h
index c066ef123434..5a218351b497 100644
--- a/drivers/gpu/drm/zte/zx_vou_regs.h
+++ b/drivers/gpu/drm/zte/zx_vou_regs.h
@@ -13,13 +13,17 @@
 
 /* Sub-module offset */
 #define MAIN_GL_OFFSET			0x130
-#define MAIN_CSC_OFFSET			0x580
+#define MAIN_GL_CSC_OFFSET		0x580
+#define MAIN_CHN_CSC_OFFSET		0x6c0
 #define MAIN_HBSC_OFFSET		0x820
+#define MAIN_DITHER_OFFSET		0x960
 #define MAIN_RSZ_OFFSET			0x600 /* OTFPPU sub-module */
 
 #define AUX_GL_OFFSET			0x200
-#define AUX_CSC_OFFSET			0x5d0
+#define AUX_GL_CSC_OFFSET		0x5d0
+#define AUX_CHN_CSC_OFFSET		0x710
 #define AUX_HBSC_OFFSET			0x860
+#define AUX_DITHER_OFFSET		0x970
 #define AUX_RSZ_OFFSET			0x800
 
 #define OSD_VL0_OFFSET			0x040
@@ -78,6 +82,10 @@
 #define CHN_INTERLACE_BUF_CTRL		0x24
 #define CHN_INTERLACE_EN		BIT(2)
 
+/* Dither registers */
+#define OSD_DITHER_CTRL0		0x00
+#define DITHER_BYSPASS			BIT(31)
+
 /* TIMING_CTRL registers */
 #define TIMING_TC_ENABLE		0x04
 #define AUX_TC_EN			BIT(1)
diff --git a/include/drm/drmP.h b/include/drm/drmP.h
index e1daa4f343cd..b9b5566acfe6 100644
--- a/include/drm/drmP.h
+++ b/include/drm/drmP.h
@@ -70,7 +70,6 @@
 #include <drm/drm_fourcc.h>
 #include <drm/drm_global.h>
 #include <drm/drm_hashtab.h>
-#include <drm/drm_mem_util.h>
 #include <drm/drm_mm.h>
 #include <drm/drm_os_linux.h>
 #include <drm/drm_sarea.h>
@@ -320,15 +319,6 @@ struct pci_controller;
 #define DRM_IF_VERSION(maj, min) (maj << 16 | min)
 
 
-/* Flags and return codes for get_vblank_timestamp() driver function. */
-#define DRM_CALLED_FROM_VBLIRQ 1
-#define DRM_VBLANKTIME_SCANOUTPOS_METHOD (1 << 0)
-
-/* get_scanout_position() return flags */
-#define DRM_SCANOUTPOS_VALID        (1 << 0)
-#define DRM_SCANOUTPOS_IN_VBLANK    (1 << 1)
-#define DRM_SCANOUTPOS_ACCURATE     (1 << 2)
-
 /**
  * DRM device structure. This structure represent a complete card that
  * may contain multiple heads.
diff --git a/include/drm/drm_atomic.h b/include/drm/drm_atomic.h
index 788daf756f48..8645dcdef031 100644
--- a/include/drm/drm_atomic.h
+++ b/include/drm/drm_atomic.h
@@ -155,6 +155,53 @@ struct __drm_connnectors_state {
 };
 
 /**
+ * struct drm_private_state_funcs - atomic state functions for private objects
+ *
+ * These hooks are used by atomic helpers to create, swap and destroy states of
+ * private objects. The structure itself is used as a vtable to identify the
+ * associated private object type. Each private object type that needs to be
+ * added to the atomic states is expected to have an implementation of these
+ * hooks and pass a pointer to it's drm_private_state_funcs struct to
+ * drm_atomic_get_private_obj_state().
+ */
+struct drm_private_state_funcs {
+	/**
+	 * @duplicate_state:
+	 *
+	 * Duplicate the current state of the private object and return it. It
+	 * is an error to call this before obj->state has been initialized.
+	 *
+	 * RETURNS:
+	 *
+	 * Duplicated atomic state or NULL when obj->state is not
+	 * initialized or allocation failed.
+	 */
+	void *(*duplicate_state)(struct drm_atomic_state *state, void *obj);
+
+	/**
+	 * @swap_state:
+	 *
+	 * This function swaps the existing state of a private object @obj with
+	 * it's newly created state, the pointer to which is passed as
+	 * @obj_state_ptr.
+	 */
+	void (*swap_state)(void *obj, void **obj_state_ptr);
+
+	/**
+	 * @destroy_state:
+	 *
+	 * Frees the private object state created with @duplicate_state.
+	 */
+	void (*destroy_state)(void *obj_state);
+};
+
+struct __drm_private_objs_state {
+	void *obj;
+	void *obj_state;
+	const struct drm_private_state_funcs *funcs;
+};
+
+/**
  * struct drm_atomic_state - the global state object for atomic updates
  * @ref: count of all references to this state (will not be freed until zero)
  * @dev: parent DRM device
@@ -164,6 +211,8 @@ struct __drm_connnectors_state {
  * @crtcs: pointer to array of CRTC pointers
  * @num_connector: size of the @connectors and @connector_states arrays
  * @connectors: pointer to array of structures with per-connector data
+ * @num_private_objs: size of the @private_objs array
+ * @private_objs: pointer to array of private object pointers
  * @acquire_ctx: acquire context for this atomic modeset state update
  */
 struct drm_atomic_state {
@@ -176,6 +225,8 @@ struct drm_atomic_state {
 	struct __drm_crtcs_state *crtcs;
 	int num_connector;
 	struct __drm_connnectors_state *connectors;
+	int num_private_objs;
+	struct __drm_private_objs_state *private_objs;
 
 	struct drm_modeset_acquire_ctx *acquire_ctx;
 
@@ -268,6 +319,11 @@ int drm_atomic_connector_set_property(struct drm_connector *connector,
 		struct drm_connector_state *state, struct drm_property *property,
 		uint64_t val);
 
+void * __must_check
+drm_atomic_get_private_obj_state(struct drm_atomic_state *state,
+			      void *obj,
+			      const struct drm_private_state_funcs *funcs);
+
 /**
  * drm_atomic_get_existing_crtc_state - get crtc state, if it exists
  * @state: global atomic state object
@@ -753,6 +809,45 @@ void drm_state_dump(struct drm_device *dev, struct drm_printer *p);
 		for_each_if (plane)
 
 /**
+ * __for_each_private_obj - iterate over all private objects
+ * @__state: &struct drm_atomic_state pointer
+ * @obj: private object iteration cursor
+ * @obj_state: private object state iteration cursor
+ * @__i: int iteration cursor, for macro-internal use
+ * @__funcs: &struct drm_private_state_funcs iteration cursor
+ *
+ * This macro iterates over the array containing private object data in atomic
+ * state
+ */
+#define __for_each_private_obj(__state, obj, obj_state, __i, __funcs)	\
+	for ((__i) = 0;							\
+	     (__i) < (__state)->num_private_objs &&			\
+	     ((obj) = (__state)->private_objs[__i].obj,			\
+	      (__funcs) = (__state)->private_objs[__i].funcs,		\
+	      (obj_state) = (__state)->private_objs[__i].obj_state,	\
+	      1);							\
+	     (__i)++)							\
+
+/**
+ * for_each_private_obj - iterate over a specify type of private object
+ * @__state: &struct drm_atomic_state pointer
+ * @obj_funcs: &struct drm_private_state_funcs function table to filter
+ * 	private objects
+ * @obj: private object iteration cursor
+ * @obj_state: private object state iteration cursor
+ * @__i: int iteration cursor, for macro-internal use
+ * @__funcs: &struct drm_private_state_funcs iteration cursor
+ *
+ * This macro iterates over the private objects state array while filtering the
+ * objects based on the vfunc table that is passed as @obj_funcs. New macros
+ * can be created by passing in the vfunc table associated with a specific
+ * private object.
+ */
+#define for_each_private_obj(__state, obj_funcs, obj, obj_state, __i, __funcs)	\
+	__for_each_private_obj(__state, obj, obj_state, __i, __funcs)		\
+		for_each_if (__funcs == obj_funcs)
+
+/**
  * drm_atomic_crtc_needs_modeset - compute combined modeset need
  * @state: &drm_crtc_state for the CRTC
  *
diff --git a/include/drm/drm_blend.h b/include/drm/drm_blend.h
index 13221cf9b3eb..17606026590b 100644
--- a/include/drm/drm_blend.h
+++ b/include/drm/drm_blend.h
@@ -25,31 +25,15 @@
 
 #include <linux/list.h>
 #include <linux/ctype.h>
+#include <drm/drm_mode.h>
 
 struct drm_device;
 struct drm_atomic_state;
-
-/*
- * Rotation property bits. DRM_ROTATE_<degrees> rotates the image by the
- * specified amount in degrees in counter clockwise direction. DRM_REFLECT_X and
- * DRM_REFLECT_Y reflects the image along the specified axis prior to rotation
- *
- * WARNING: These defines are UABI since they're exposed in the rotation
- * property.
- */
-#define DRM_ROTATE_0	BIT(0)
-#define DRM_ROTATE_90	BIT(1)
-#define DRM_ROTATE_180	BIT(2)
-#define DRM_ROTATE_270	BIT(3)
-#define DRM_ROTATE_MASK (DRM_ROTATE_0   | DRM_ROTATE_90 | \
-			 DRM_ROTATE_180 | DRM_ROTATE_270)
-#define DRM_REFLECT_X	BIT(4)
-#define DRM_REFLECT_Y	BIT(5)
-#define DRM_REFLECT_MASK (DRM_REFLECT_X | DRM_REFLECT_Y)
+struct drm_plane;
 
 static inline bool drm_rotation_90_or_270(unsigned int rotation)
 {
-	return rotation & (DRM_ROTATE_90 | DRM_ROTATE_270);
+	return rotation & (DRM_MODE_ROTATE_90 | DRM_MODE_ROTATE_270);
 }
 
 int drm_plane_create_rotation_property(struct drm_plane *plane,
diff --git a/include/drm/drm_bridge.h b/include/drm/drm_bridge.h
index fdd82fcbf168..983054f2e86e 100644
--- a/include/drm/drm_bridge.h
+++ b/include/drm/drm_bridge.h
@@ -59,6 +59,40 @@ struct drm_bridge_funcs {
 	void (*detach)(struct drm_bridge *bridge);
 
 	/**
+	 * @mode_valid:
+	 *
+	 * This callback is used to check if a specific mode is valid in this
+	 * bridge. This should be implemented if the bridge has some sort of
+	 * restriction in the modes it can display. For example, a given bridge
+	 * may be responsible to set a clock value. If the clock can not
+	 * produce all the values for the available modes then this callback
+	 * can be used to restrict the number of modes to only the ones that
+	 * can be displayed.
+	 *
+	 * This hook is used by the probe helpers to filter the mode list in
+	 * drm_helper_probe_single_connector_modes(), and it is used by the
+	 * atomic helpers to validate modes supplied by userspace in
+	 * drm_atomic_helper_check_modeset().
+	 *
+	 * This function is optional.
+	 *
+	 * NOTE:
+	 *
+	 * Since this function is both called from the check phase of an atomic
+	 * commit, and the mode validation in the probe paths it is not allowed
+	 * to look at anything else but the passed-in mode, and validate it
+	 * against configuration-invariant hardward constraints. Any further
+	 * limits which depend upon the configuration can only be checked in
+	 * @mode_fixup.
+	 *
+	 * RETURNS:
+	 *
+	 * drm_mode_status Enum
+	 */
+	enum drm_mode_status (*mode_valid)(struct drm_bridge *crtc,
+					   const struct drm_display_mode *mode);
+
+	/**
 	 * @mode_fixup:
 	 *
 	 * This callback is used to validate and adjust a mode. The paramater
@@ -66,7 +100,7 @@ struct drm_bridge_funcs {
 	 * the display chain, either the final &drm_connector or the next
 	 * &drm_bridge. The parameter adjusted_mode is the input mode the bridge
 	 * requires. It can be modified by this callback and does not need to
-	 * match mode.
+	 * match mode. See also &drm_crtc_state.adjusted_mode for more details.
 	 *
 	 * This is the only hook that allows a bridge to reject a modeset. If
 	 * this function passes all other callbacks must succeed for this
@@ -82,6 +116,12 @@ struct drm_bridge_funcs {
 	 * NOT touch any persistent state (hardware or software) or data
 	 * structures except the passed in @state parameter.
 	 *
+	 * Also beware that userspace can request its own custom modes, neither
+	 * core nor helpers filter modes to the list of probe modes reported by
+	 * the GETCONNECTOR IOCTL and stored in &drm_connector.modes. To ensure
+	 * that modes are filtered consistently put any bridge constraints and
+	 * limits checks into @mode_valid.
+	 *
 	 * RETURNS:
 	 *
 	 * True if an acceptable configuration is possible, false if the modeset
diff --git a/include/drm/drm_color_mgmt.h b/include/drm/drm_color_mgmt.h
index bce4a532836d..03a59cbce621 100644
--- a/include/drm/drm_color_mgmt.h
+++ b/include/drm/drm_color_mgmt.h
@@ -25,6 +25,8 @@
 
 #include <linux/ctype.h>
 
+struct drm_crtc;
+
 uint32_t drm_color_lut_extract(uint32_t user_input, uint32_t bit_precision);
 
 void drm_crtc_enable_color_mgmt(struct drm_crtc *crtc,
diff --git a/include/drm/drm_connector.h b/include/drm/drm_connector.h
index 4eeda120e46d..2fe09c1ddfb8 100644
--- a/include/drm/drm_connector.h
+++ b/include/drm/drm_connector.h
@@ -25,6 +25,7 @@
 
 #include <linux/list.h>
 #include <linux/ctype.h>
+#include <linux/hdmi.h>
 #include <drm/drm_mode_object.h>
 
 #include <uapi/drm/drm_mode.h>
@@ -326,6 +327,21 @@ struct drm_connector_state {
 	struct drm_atomic_state *state;
 
 	struct drm_tv_connector_state tv;
+
+	/**
+	 * @picture_aspect_ratio: Connector property to control the
+	 * HDMI infoframe aspect ratio setting.
+	 *
+	 * The %DRM_MODE_PICTURE_ASPECT_\* values much match the
+	 * values for &enum hdmi_picture_aspect
+	 */
+	enum hdmi_picture_aspect picture_aspect_ratio;
+
+	/**
+	 * @scaling_mode: Connector property to control the
+	 * upscaling, mostly used for built-in panels.
+	 */
+	unsigned int scaling_mode;
 };
 
 /**
@@ -675,6 +691,7 @@ struct drm_cmdline_mode {
  * @tile_v_loc: vertical location of this tile
  * @tile_h_size: horizontal size of this tile.
  * @tile_v_size: vertical size of this tile.
+ * @scaling_mode_property:  Optional atomic property to control the upscaling.
  *
  * Each connector may be connected to one or more CRTCs, or may be clonable by
  * another connector if they can share a CRTC.  Each connector also has a specific
@@ -754,6 +771,8 @@ struct drm_connector {
 	struct drm_property_blob *edid_blob_ptr;
 	struct drm_object_properties properties;
 
+	struct drm_property *scaling_mode_property;
+
 	/**
 	 * @path_blob_ptr:
 	 *
@@ -953,6 +972,8 @@ int drm_mode_create_tv_properties(struct drm_device *dev,
 				  unsigned int num_modes,
 				  const char * const modes[]);
 int drm_mode_create_scaling_mode_property(struct drm_device *dev);
+int drm_connector_attach_scaling_mode_property(struct drm_connector *connector,
+					       u32 scaling_mode_mask);
 int drm_mode_create_aspect_ratio_property(struct drm_device *dev);
 int drm_mode_create_suggested_offset_properties(struct drm_device *dev);
 
@@ -1031,7 +1052,7 @@ void drm_connector_list_iter_end(struct drm_connector_list_iter *iter);
  *
  * Note that @connector is only valid within the list body, if you want to use
  * @connector after calling drm_connector_list_iter_end() then you need to grab
- * your own reference first using drm_connector_begin().
+ * your own reference first using drm_connector_get().
  */
 #define drm_for_each_connector_iter(connector, iter) \
 	while ((connector = drm_connector_list_iter_next(iter)))
diff --git a/include/drm/drm_crtc.h b/include/drm/drm_crtc.h
index a8176a836e25..b6e3713bd7a9 100644
--- a/include/drm/drm_crtc.h
+++ b/include/drm/drm_crtc.h
@@ -90,14 +90,7 @@ struct drm_plane_helper_funcs;
  * @plane_mask: bitmask of (1 << drm_plane_index(plane)) of attached planes
  * @connector_mask: bitmask of (1 << drm_connector_index(connector)) of attached connectors
  * @encoder_mask: bitmask of (1 << drm_encoder_index(encoder)) of attached encoders
- * @adjusted_mode: for use by helpers and drivers to compute adjusted mode timings
- * @mode: current mode timings
  * @mode_blob: &drm_property_blob for @mode
- * @degamma_lut: Lookup table for converting framebuffer pixel data
- *	before apply the conversion matrix
- * @ctm: Transformation matrix
- * @gamma_lut: Lookup table for converting pixel data after the
- *	conversion matrix
  * @state: backpointer to global drm_atomic_state
  *
  * Note that the distinction between @enable and @active is rather subtile:
@@ -136,17 +129,62 @@ struct drm_crtc_state {
 	u32 connector_mask;
 	u32 encoder_mask;
 
-	/* adjusted_mode: for use by helpers and drivers */
+	/**
+	 * @adjusted_mode:
+	 *
+	 * Internal display timings which can be used by the driver to handle
+	 * differences between the mode requested by userspace in @mode and what
+	 * is actually programmed into the hardware. It is purely driver
+	 * implementation defined what exactly this adjusted mode means. Usually
+	 * it is used to store the hardware display timings used between the
+	 * CRTC and encoder blocks.
+	 */
 	struct drm_display_mode adjusted_mode;
 
+	/**
+	 * @mode:
+	 *
+	 * Display timings requested by userspace. The driver should try to
+	 * match the refresh rate as close as possible (but note that it's
+	 * undefined what exactly is close enough, e.g. some of the HDMI modes
+	 * only differ in less than 1% of the refresh rate). The active width
+	 * and height as observed by userspace for positioning planes must match
+	 * exactly.
+	 *
+	 * For external connectors where the sink isn't fixed (like with a
+	 * built-in panel), this mode here should match the physical mode on the
+	 * wire to the last details (i.e. including sync polarities and
+	 * everything).
+	 */
 	struct drm_display_mode mode;
 
 	/* blob property to expose current mode to atomic userspace */
 	struct drm_property_blob *mode_blob;
 
-	/* blob property to expose color management to userspace */
+	/**
+	 * @degamma_lut:
+	 *
+	 * Lookup table for converting framebuffer pixel data before apply the
+	 * color conversion matrix @ctm. See drm_crtc_enable_color_mgmt(). The
+	 * blob (if not NULL) is an array of &struct drm_color_lut.
+	 */
 	struct drm_property_blob *degamma_lut;
+
+	/**
+	 * @ctm:
+	 *
+	 * Color transformation matrix. See drm_crtc_enable_color_mgmt(). The
+	 * blob (if not NULL) is a &struct drm_color_ctm.
+	 */
 	struct drm_property_blob *ctm;
+
+	/**
+	 * @gamma_lut:
+	 *
+	 * Lookup table for converting pixel data after the color conversion
+	 * matrix @ctm.  See drm_crtc_enable_color_mgmt(). The blob (if not
+	 * NULL) is an array of &struct drm_color_lut.
+	 */
 	struct drm_property_blob *gamma_lut;
 
 	/**
@@ -313,6 +351,12 @@ struct drm_crtc_funcs {
 	 *
 	 * This callback is optional.
 	 *
+	 * Atomic drivers who want to support gamma tables should implement the
+	 * atomic color management support, enabled by calling
+	 * drm_crtc_enable_color_mgmt(), which then supports the legacy gamma
+	 * interface through the drm_atomic_helper_legacy_gamma_set()
+	 * compatibility implementation.
+	 *
 	 * NOTE:
 	 *
 	 * Drivers that support gamma tables and also fbdev emulation through
diff --git a/include/drm/drm_dp_helper.h b/include/drm/drm_dp_helper.h
index bb837310c07e..b17476a6909c 100644
--- a/include/drm/drm_dp_helper.h
+++ b/include/drm/drm_dp_helper.h
@@ -179,6 +179,111 @@
 
 #define DP_GUID				    0x030   /* 1.2 */
 
+#define DP_DSC_SUPPORT                      0x060   /* DP 1.4 */
+# define DP_DSC_DECOMPRESSION_IS_SUPPORTED  (1 << 0)
+
+#define DP_DSC_REV                          0x061
+# define DP_DSC_MAJOR_MASK                  (0xf << 0)
+# define DP_DSC_MINOR_MASK                  (0xf << 4)
+# define DP_DSC_MAJOR_SHIFT                 0
+# define DP_DSC_MINOR_SHIFT                 4
+
+#define DP_DSC_RC_BUF_BLK_SIZE              0x062
+# define DP_DSC_RC_BUF_BLK_SIZE_1           0x0
+# define DP_DSC_RC_BUF_BLK_SIZE_4           0x1
+# define DP_DSC_RC_BUF_BLK_SIZE_16          0x2
+# define DP_DSC_RC_BUF_BLK_SIZE_64          0x3
+
+#define DP_DSC_RC_BUF_SIZE                  0x063
+
+#define DP_DSC_SLICE_CAP_1                  0x064
+# define DP_DSC_1_PER_DP_DSC_SINK           (1 << 0)
+# define DP_DSC_2_PER_DP_DSC_SINK           (1 << 1)
+# define DP_DSC_4_PER_DP_DSC_SINK           (1 << 3)
+# define DP_DSC_6_PER_DP_DSC_SINK           (1 << 4)
+# define DP_DSC_8_PER_DP_DSC_SINK           (1 << 5)
+# define DP_DSC_10_PER_DP_DSC_SINK          (1 << 6)
+# define DP_DSC_12_PER_DP_DSC_SINK          (1 << 7)
+
+#define DP_DSC_LINE_BUF_BIT_DEPTH           0x065
+# define DP_DSC_LINE_BUF_BIT_DEPTH_MASK     (0xf << 0)
+# define DP_DSC_LINE_BUF_BIT_DEPTH_9        0x0
+# define DP_DSC_LINE_BUF_BIT_DEPTH_10       0x1
+# define DP_DSC_LINE_BUF_BIT_DEPTH_11       0x2
+# define DP_DSC_LINE_BUF_BIT_DEPTH_12       0x3
+# define DP_DSC_LINE_BUF_BIT_DEPTH_13       0x4
+# define DP_DSC_LINE_BUF_BIT_DEPTH_14       0x5
+# define DP_DSC_LINE_BUF_BIT_DEPTH_15       0x6
+# define DP_DSC_LINE_BUF_BIT_DEPTH_16       0x7
+# define DP_DSC_LINE_BUF_BIT_DEPTH_8        0x8
+
+#define DP_DSC_BLK_PREDICTION_SUPPORT       0x066
+# define DP_DSC_BLK_PREDICTION_IS_SUPPORTED (1 << 0)
+
+#define DP_DSC_MAX_BITS_PER_PIXEL_LOW       0x067   /* eDP 1.4 */
+
+#define DP_DSC_MAX_BITS_PER_PIXEL_HI        0x068   /* eDP 1.4 */
+
+#define DP_DSC_DEC_COLOR_FORMAT_CAP         0x069
+# define DP_DSC_RGB                         (1 << 0)
+# define DP_DSC_YCbCr444                    (1 << 1)
+# define DP_DSC_YCbCr422_Simple             (1 << 2)
+# define DP_DSC_YCbCr422_Native             (1 << 3)
+# define DP_DSC_YCbCr420_Native             (1 << 4)
+
+#define DP_DSC_DEC_COLOR_DEPTH_CAP          0x06A
+# define DP_DSC_8_BPC                       (1 << 1)
+# define DP_DSC_10_BPC                      (1 << 2)
+# define DP_DSC_12_BPC                      (1 << 3)
+
+#define DP_DSC_PEAK_THROUGHPUT              0x06B
+# define DP_DSC_THROUGHPUT_MODE_0_MASK      (0xf << 0)
+# define DP_DSC_THROUGHPUT_MODE_0_SHIFT     0
+# define DP_DSC_THROUGHPUT_MODE_0_340       (1 << 0)
+# define DP_DSC_THROUGHPUT_MODE_0_400       (2 << 0)
+# define DP_DSC_THROUGHPUT_MODE_0_450       (3 << 0)
+# define DP_DSC_THROUGHPUT_MODE_0_500       (4 << 0)
+# define DP_DSC_THROUGHPUT_MODE_0_550       (5 << 0)
+# define DP_DSC_THROUGHPUT_MODE_0_600       (6 << 0)
+# define DP_DSC_THROUGHPUT_MODE_0_650       (7 << 0)
+# define DP_DSC_THROUGHPUT_MODE_0_700       (8 << 0)
+# define DP_DSC_THROUGHPUT_MODE_0_750       (9 << 0)
+# define DP_DSC_THROUGHPUT_MODE_0_800       (10 << 0)
+# define DP_DSC_THROUGHPUT_MODE_0_850       (11 << 0)
+# define DP_DSC_THROUGHPUT_MODE_0_900       (12 << 0)
+# define DP_DSC_THROUGHPUT_MODE_0_950       (13 << 0)
+# define DP_DSC_THROUGHPUT_MODE_0_1000      (14 << 0)
+# define DP_DSC_THROUGHPUT_MODE_1_MASK      (0xf << 4)
+# define DP_DSC_THROUGHPUT_MODE_1_SHIFT     4
+# define DP_DSC_THROUGHPUT_MODE_1_340       (1 << 4)
+# define DP_DSC_THROUGHPUT_MODE_1_400       (2 << 4)
+# define DP_DSC_THROUGHPUT_MODE_1_450       (3 << 4)
+# define DP_DSC_THROUGHPUT_MODE_1_500       (4 << 4)
+# define DP_DSC_THROUGHPUT_MODE_1_550       (5 << 4)
+# define DP_DSC_THROUGHPUT_MODE_1_600       (6 << 4)
+# define DP_DSC_THROUGHPUT_MODE_1_650       (7 << 4)
+# define DP_DSC_THROUGHPUT_MODE_1_700       (8 << 4)
+# define DP_DSC_THROUGHPUT_MODE_1_750       (9 << 4)
+# define DP_DSC_THROUGHPUT_MODE_1_800       (10 << 4)
+# define DP_DSC_THROUGHPUT_MODE_1_850       (11 << 4)
+# define DP_DSC_THROUGHPUT_MODE_1_900       (12 << 4)
+# define DP_DSC_THROUGHPUT_MODE_1_950       (13 << 4)
+# define DP_DSC_THROUGHPUT_MODE_1_1000      (14 << 4)
+
+#define DP_DSC_MAX_SLICE_WIDTH              0x06C
+
+#define DP_DSC_SLICE_CAP_2                  0x06D
+# define DP_DSC_16_PER_DP_DSC_SINK          (1 << 0)
+# define DP_DSC_20_PER_DP_DSC_SINK          (1 << 1)
+# define DP_DSC_24_PER_DP_DSC_SINK          (1 << 2)
+
+#define DP_DSC_BITS_PER_PIXEL_INC           0x06F
+# define DP_DSC_BITS_PER_PIXEL_1_16         0x0
+# define DP_DSC_BITS_PER_PIXEL_1_8          0x1
+# define DP_DSC_BITS_PER_PIXEL_1_4          0x2
+# define DP_DSC_BITS_PER_PIXEL_1_2          0x3
+# define DP_DSC_BITS_PER_PIXEL_1            0x4
+
 #define DP_PSR_SUPPORT                      0x070   /* XXX 1.2? */
 # define DP_PSR_IS_SUPPORTED                1
 # define DP_PSR2_IS_SUPPORTED		    2	    /* eDP 1.4 */
@@ -339,6 +444,8 @@
 #define DP_AUX_FRAME_SYNC_VALUE		    0x15c   /* eDP 1.4 */
 # define DP_AUX_FRAME_SYNC_VALID	    (1 << 0)
 
+#define DP_DSC_ENABLE                       0x160   /* DP 1.4 */
+
 #define DP_PSR_EN_CFG			    0x170   /* XXX 1.2? */
 # define DP_PSR_ENABLE			    (1 << 0)
 # define DP_PSR_MAIN_LINK_ACTIVE	    (1 << 1)
@@ -572,10 +679,12 @@
 #define DP_EDP_PWMGEN_BIT_COUNT             0x724
 #define DP_EDP_PWMGEN_BIT_COUNT_CAP_MIN     0x725
 #define DP_EDP_PWMGEN_BIT_COUNT_CAP_MAX     0x726
+# define DP_EDP_PWMGEN_BIT_COUNT_MASK       (0x1f << 0)
 
 #define DP_EDP_BACKLIGHT_CONTROL_STATUS     0x727
 
 #define DP_EDP_BACKLIGHT_FREQ_SET           0x728
+# define DP_EDP_BACKLIGHT_FREQ_BASE_KHZ     27000
 
 #define DP_EDP_BACKLIGHT_FREQ_CAP_MIN_MSB   0x72a
 #define DP_EDP_BACKLIGHT_FREQ_CAP_MIN_MID   0x72b
@@ -603,6 +712,9 @@
 #define DP_DEVICE_SERVICE_IRQ_VECTOR_ESI0   0x2003   /* 1.2 */
 
 #define DP_DEVICE_SERVICE_IRQ_VECTOR_ESI1   0x2004   /* 1.2 */
+# define DP_RX_GTC_MSTR_REQ_STATUS_CHANGE    (1 << 0)
+# define DP_LOCK_ACQUISITION_REQUEST         (1 << 1)
+# define DP_CEC_IRQ                          (1 << 2)
 
 #define DP_LINK_SERVICE_IRQ_VECTOR_ESI0     0x2005   /* 1.2 */
 
@@ -636,6 +748,62 @@
 # define DP_VSC_EXT_CEA_SDP_SUPPORTED			(1 << 6)  /* DP 1.4 */
 # define DP_VSC_EXT_CEA_SDP_CHAINING_SUPPORTED		(1 << 7)  /* DP 1.4 */
 
+/* HDMI CEC tunneling over AUX DP 1.3 section 5.3.3.3.1 DPCD 1.4+ */
+#define DP_CEC_TUNNELING_CAPABILITY            0x3000
+# define DP_CEC_TUNNELING_CAPABLE               (1 << 0)
+# define DP_CEC_SNOOPING_CAPABLE                (1 << 1)
+# define DP_CEC_MULTIPLE_LA_CAPABLE             (1 << 2)
+
+#define DP_CEC_TUNNELING_CONTROL               0x3001
+# define DP_CEC_TUNNELING_ENABLE                (1 << 0)
+# define DP_CEC_SNOOPING_ENABLE                 (1 << 1)
+
+#define DP_CEC_RX_MESSAGE_INFO                 0x3002
+# define DP_CEC_RX_MESSAGE_LEN_MASK             (0xf << 0)
+# define DP_CEC_RX_MESSAGE_LEN_SHIFT            0
+# define DP_CEC_RX_MESSAGE_HPD_STATE            (1 << 4)
+# define DP_CEC_RX_MESSAGE_HPD_LOST             (1 << 5)
+# define DP_CEC_RX_MESSAGE_ACKED                (1 << 6)
+# define DP_CEC_RX_MESSAGE_ENDED                (1 << 7)
+
+#define DP_CEC_TX_MESSAGE_INFO                 0x3003
+# define DP_CEC_TX_MESSAGE_LEN_MASK             (0xf << 0)
+# define DP_CEC_TX_MESSAGE_LEN_SHIFT            0
+# define DP_CEC_TX_RETRY_COUNT_MASK             (0x7 << 4)
+# define DP_CEC_TX_RETRY_COUNT_SHIFT            4
+# define DP_CEC_TX_MESSAGE_SEND                 (1 << 7)
+
+#define DP_CEC_TUNNELING_IRQ_FLAGS             0x3004
+# define DP_CEC_RX_MESSAGE_INFO_VALID           (1 << 0)
+# define DP_CEC_RX_MESSAGE_OVERFLOW             (1 << 1)
+# define DP_CEC_TX_MESSAGE_SENT                 (1 << 4)
+# define DP_CEC_TX_LINE_ERROR                   (1 << 5)
+# define DP_CEC_TX_ADDRESS_NACK_ERROR           (1 << 6)
+# define DP_CEC_TX_DATA_NACK_ERROR              (1 << 7)
+
+#define DP_CEC_LOGICAL_ADDRESS_MASK            0x300E /* 0x300F word */
+# define DP_CEC_LOGICAL_ADDRESS_0               (1 << 0)
+# define DP_CEC_LOGICAL_ADDRESS_1               (1 << 1)
+# define DP_CEC_LOGICAL_ADDRESS_2               (1 << 2)
+# define DP_CEC_LOGICAL_ADDRESS_3               (1 << 3)
+# define DP_CEC_LOGICAL_ADDRESS_4               (1 << 4)
+# define DP_CEC_LOGICAL_ADDRESS_5               (1 << 5)
+# define DP_CEC_LOGICAL_ADDRESS_6               (1 << 6)
+# define DP_CEC_LOGICAL_ADDRESS_7               (1 << 7)
+#define DP_CEC_LOGICAL_ADDRESS_MASK_2          0x300F /* 0x300E word */
+# define DP_CEC_LOGICAL_ADDRESS_8               (1 << 0)
+# define DP_CEC_LOGICAL_ADDRESS_9               (1 << 1)
+# define DP_CEC_LOGICAL_ADDRESS_10              (1 << 2)
+# define DP_CEC_LOGICAL_ADDRESS_11              (1 << 3)
+# define DP_CEC_LOGICAL_ADDRESS_12              (1 << 4)
+# define DP_CEC_LOGICAL_ADDRESS_13              (1 << 5)
+# define DP_CEC_LOGICAL_ADDRESS_14              (1 << 6)
+# define DP_CEC_LOGICAL_ADDRESS_15              (1 << 7)
+
+#define DP_CEC_RX_MESSAGE_BUFFER               0x3010
+#define DP_CEC_TX_MESSAGE_BUFFER               0x3020
+#define DP_CEC_MESSAGE_BUFFER_LENGTH             0x10
+
 /* DP 1.2 Sideband message defines */
 /* peer device type - DP 1.2a Table 2-92 */
 #define DP_PEER_DEVICE_NONE		0x0
diff --git a/include/drm/drm_dp_mst_helper.h b/include/drm/drm_dp_mst_helper.h
index 5b024764666c..177ab6f86855 100644
--- a/include/drm/drm_dp_mst_helper.h
+++ b/include/drm/drm_dp_mst_helper.h
@@ -24,6 +24,7 @@
 
 #include <linux/types.h>
 #include <drm/drm_dp_helper.h>
+#include <drm/drm_atomic.h>
 
 struct drm_dp_mst_branch;
 
@@ -403,6 +404,12 @@ struct drm_dp_payload {
 	int vcpi;
 };
 
+struct drm_dp_mst_topology_state {
+	int avail_slots;
+	struct drm_atomic_state *state;
+	struct drm_dp_mst_topology_mgr *mgr;
+};
+
 /**
  * struct drm_dp_mst_topology_mgr - DisplayPort MST manager
  *
@@ -481,6 +488,16 @@ struct drm_dp_mst_topology_mgr {
 	int pbn_div;
 
 	/**
+	 * @state: State information for topology manager
+	 */
+	struct drm_dp_mst_topology_state *state;
+
+	/**
+	 * @funcs: Atomic helper callbacks
+	 */
+	const struct drm_private_state_funcs *funcs;
+
+	/**
 	 * @qlock: protects @tx_msg_downq, the &drm_dp_mst_branch.txslost and
 	 * &drm_dp_sideband_msg_tx.state once they are queued
 	 */
@@ -596,4 +613,13 @@ void drm_dp_mst_dump_topology(struct seq_file *m,
 
 void drm_dp_mst_topology_mgr_suspend(struct drm_dp_mst_topology_mgr *mgr);
 int drm_dp_mst_topology_mgr_resume(struct drm_dp_mst_topology_mgr *mgr);
+struct drm_dp_mst_topology_state *drm_atomic_get_mst_topology_state(struct drm_atomic_state *state,
+								    struct drm_dp_mst_topology_mgr *mgr);
+int drm_dp_atomic_find_vcpi_slots(struct drm_atomic_state *state,
+				  struct drm_dp_mst_topology_mgr *mgr,
+				  struct drm_dp_mst_port *port, int pbn);
+int drm_dp_atomic_release_vcpi_slots(struct drm_atomic_state *state,
+				     struct drm_dp_mst_topology_mgr *mgr,
+				     int slots);
+
 #endif
diff --git a/include/drm/drm_drv.h b/include/drm/drm_drv.h
index 53b98321df9b..e64e33b9dd26 100644
--- a/include/drm/drm_drv.h
+++ b/include/drm/drm_drv.h
@@ -104,23 +104,6 @@ struct drm_driver {
 	int (*open) (struct drm_device *, struct drm_file *);
 
 	/**
-	 * @preclose:
-	 *
-	 * One of the driver callbacks when a new &struct drm_file is closed.
-	 * Useful for tearing down driver-private data structures allocated in
-	 * @open like buffer allocators, execution contexts or similar things.
-	 *
-	 * Since the display/modeset side of DRM can only be owned by exactly
-	 * one &struct drm_file (see &drm_file.is_master and &drm_device.master)
-	 * there should never be a need to tear down any modeset related
-	 * resources in this callback. Doing so would be a driver design bug.
-	 *
-	 * FIXME: It is not really clear why there's both @preclose and
-	 * @postclose. Without a really good reason, use @postclose only.
-	 */
-	void (*preclose) (struct drm_device *, struct drm_file *file_priv);
-
-	/**
 	 * @postclose:
 	 *
 	 * One of the driver callbacks when a new &struct drm_file is closed.
@@ -131,9 +114,6 @@ struct drm_driver {
 	 * one &struct drm_file (see &drm_file.is_master and &drm_device.master)
 	 * there should never be a need to tear down any modeset related
 	 * resources in this callback. Doing so would be a driver design bug.
-	 *
-	 * FIXME: It is not really clear why there's both @preclose and
-	 * @postclose. Without a really good reason, use @postclose only.
 	 */
 	void (*postclose) (struct drm_device *, struct drm_file *);
 
@@ -150,7 +130,7 @@ struct drm_driver {
 	 * state changes, e.g. in conjunction with the :ref:`vga_switcheroo`
 	 * infrastructure.
 	 *
-	 * This is called after @preclose and @postclose have been called.
+	 * This is called after @postclose hook has been called.
 	 *
 	 * NOTE:
 	 *
@@ -261,8 +241,10 @@ struct drm_driver {
 	 *     DRM device.
 	 * pipe:
 	 *     Id of the crtc to query.
-	 * flags:
-	 *     Flags from the caller (DRM_CALLED_FROM_VBLIRQ or 0).
+	 * in_vblank_irq:
+	 *     True when called from drm_crtc_handle_vblank().  Some drivers
+	 *     need to apply some workarounds for gpu-specific vblank irq quirks
+	 *     if flag is set.
 	 * vpos:
 	 *     Target location for current vertical scanout position.
 	 * hpos:
@@ -283,22 +265,19 @@ struct drm_driver {
 	 *
 	 * Returns:
 	 *
-	 * Flags, or'ed together as follows:
+	 * True on success, false if a reliable scanout position counter could
+	 * not be read out.
 	 *
-	 * DRM_SCANOUTPOS_VALID:
-	 *     Query successful.
-	 * DRM_SCANOUTPOS_INVBL:
-	 *     Inside vblank.
-	 * DRM_SCANOUTPOS_ACCURATE: Returned position is accurate. A lack of
-	 *     this flag means that returned position may be offset by a
-	 *     constant but unknown small number of scanlines wrt. real scanout
-	 *     position.
+	 * FIXME:
 	 *
+	 * Since this is a helper to implement @get_vblank_timestamp, we should
+	 * move it to &struct drm_crtc_helper_funcs, like all the other
+	 * helper-internal hooks.
 	 */
-	int (*get_scanout_position) (struct drm_device *dev, unsigned int pipe,
-				     unsigned int flags, int *vpos, int *hpos,
-				     ktime_t *stime, ktime_t *etime,
-				     const struct drm_display_mode *mode);
+	bool (*get_scanout_position) (struct drm_device *dev, unsigned int pipe,
+				      bool in_vblank_irq, int *vpos, int *hpos,
+				      ktime_t *stime, ktime_t *etime,
+				      const struct drm_display_mode *mode);
 
 	/**
 	 * @get_vblank_timestamp:
@@ -328,22 +307,25 @@ struct drm_driver {
 	 *     Returns true upper bound on error for timestamp.
 	 * vblank_time:
 	 *     Target location for returned vblank timestamp.
-	 * flags:
-	 *     0 = Defaults, no special treatment needed.
-	 *     DRM_CALLED_FROM_VBLIRQ = Function is called from vblank
-	 *     irq handler. Some drivers need to apply some workarounds
-	 *     for gpu-specific vblank irq quirks if flag is set.
+	 * in_vblank_irq:
+	 *     True when called from drm_crtc_handle_vblank().  Some drivers
+	 *     need to apply some workarounds for gpu-specific vblank irq quirks
+	 *     if flag is set.
 	 *
 	 * Returns:
 	 *
-	 * Zero if timestamping isn't supported in current display mode or a
-	 * negative number on failure. A positive status code on success,
-	 * which describes how the vblank_time timestamp was computed.
+	 * True on success, false on failure, which means the core should
+	 * fallback to a simple timestamp taken in drm_crtc_handle_vblank().
+	 *
+	 * FIXME:
+	 *
+	 * We should move this hook to &struct drm_crtc_funcs like all the other
+	 * vblank hooks.
 	 */
-	int (*get_vblank_timestamp) (struct drm_device *dev, unsigned int pipe,
+	bool (*get_vblank_timestamp) (struct drm_device *dev, unsigned int pipe,
 				     int *max_error,
 				     struct timeval *vblank_time,
-				     unsigned flags);
+				     bool in_vblank_irq);
 
 	/* these have to be filled in */
 
@@ -516,6 +498,7 @@ struct drm_driver {
 	/* List of devices hanging off this driver with stealth attach. */
 	struct list_head legacy_dev_list;
 	int (*firstopen) (struct drm_device *);
+	void (*preclose) (struct drm_device *, struct drm_file *file_priv);
 	int (*dma_ioctl) (struct drm_device *dev, void *data, struct drm_file *file_priv);
 	int (*dma_quiescent) (struct drm_device *);
 	int (*context_dtor) (struct drm_device *dev, int context);
diff --git a/include/drm/drm_fb_cma_helper.h b/include/drm/drm_fb_cma_helper.h
index a5ecc0a58260..199a63f48659 100644
--- a/include/drm/drm_fb_cma_helper.h
+++ b/include/drm/drm_fb_cma_helper.h
@@ -41,6 +41,10 @@ struct drm_framebuffer *drm_fb_cma_create(struct drm_device *dev,
 struct drm_gem_cma_object *drm_fb_cma_get_gem_obj(struct drm_framebuffer *fb,
 	unsigned int plane);
 
+dma_addr_t drm_fb_cma_get_gem_addr(struct drm_framebuffer *fb,
+				   struct drm_plane_state *state,
+				   unsigned int plane);
+
 int drm_fb_cma_prepare_fb(struct drm_plane *plane,
 			  struct drm_plane_state *state);
 
diff --git a/include/drm/drm_gem_cma_helper.h b/include/drm/drm_gem_cma_helper.h
index f962d33667cf..b42529e0fae0 100644
--- a/include/drm/drm_gem_cma_helper.h
+++ b/include/drm/drm_gem_cma_helper.h
@@ -26,6 +26,13 @@ to_drm_gem_cma_obj(struct drm_gem_object *gem_obj)
 	return container_of(gem_obj, struct drm_gem_cma_object, base);
 }
 
+#ifndef CONFIG_MMU
+#define DRM_GEM_CMA_UNMAPPED_AREA_FOPS \
+	.get_unmapped_area	= drm_gem_cma_get_unmapped_area,
+#else
+#define DRM_GEM_CMA_UNMAPPED_AREA_FOPS
+#endif
+
 /**
  * DEFINE_DRM_GEM_CMA_FOPS() - macro to generate file operations for CMA drivers
  * @name: name for the generated structure
@@ -50,6 +57,7 @@ to_drm_gem_cma_obj(struct drm_gem_object *gem_obj)
 		.read		= drm_read,\
 		.llseek		= noop_llseek,\
 		.mmap		= drm_gem_cma_mmap,\
+		DRM_GEM_CMA_UNMAPPED_AREA_FOPS \
 	}
 
 /* free GEM object */
@@ -85,15 +93,6 @@ unsigned long drm_gem_cma_get_unmapped_area(struct file *filp,
 					    unsigned long len,
 					    unsigned long pgoff,
 					    unsigned long flags);
-#else
-static inline unsigned long drm_gem_cma_get_unmapped_area(struct file *filp,
-							  unsigned long addr,
-							  unsigned long len,
-							  unsigned long pgoff,
-							  unsigned long flags)
-{
-	return -EINVAL;
-}
 #endif
 
 #ifdef CONFIG_DEBUG_FS
diff --git a/include/drm/drm_irq.h b/include/drm/drm_irq.h
index cf0be6594c8c..569ca86d4e1f 100644
--- a/include/drm/drm_irq.h
+++ b/include/drm/drm_irq.h
@@ -121,6 +121,18 @@ struct drm_vblank_crtc {
 	 * drm_calc_timestamping_constants().
 	 */
 	int linedur_ns;
+
+	/**
+	 * @hwmode:
+	 *
+	 * Cache of the current hardware display mode. Only valid when @enabled
+	 * is set. This is used by helpers like
+	 * drm_calc_vbltimestamp_from_scanoutpos(). We can't just access the
+	 * hardware mode by e.g. looking at &drm_crtc_state.adjusted_mode,
+	 * because that one is really hard to get from interrupt context.
+	 */
+	struct drm_display_mode hwmode;
+
 	/**
 	 * @enabled: Tracks the enabling state of the corresponding &drm_crtc to
 	 * avoid double-disabling and hence corrupting saved state. Needed by
@@ -153,11 +165,10 @@ void drm_crtc_vblank_on(struct drm_crtc *crtc);
 void drm_vblank_cleanup(struct drm_device *dev);
 u32 drm_accurate_vblank_count(struct drm_crtc *crtc);
 
-int drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev,
-					  unsigned int pipe, int *max_error,
-					  struct timeval *vblank_time,
-					  unsigned flags,
-					  const struct drm_display_mode *mode);
+bool drm_calc_vbltimestamp_from_scanoutpos(struct drm_device *dev,
+					   unsigned int pipe, int *max_error,
+					   struct timeval *vblank_time,
+					   bool in_vblank_irq);
 void drm_calc_timestamping_constants(struct drm_crtc *crtc,
 				     const struct drm_display_mode *mode);
 
diff --git a/include/drm/drm_mem_util.h b/include/drm/drm_mem_util.h
deleted file mode 100644
index d0f6cf2e5324..000000000000
--- a/include/drm/drm_mem_util.h
+++ /dev/null
@@ -1,78 +0,0 @@
-/*
- * Copyright © 2008 Intel Corporation
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice (including the next
- * paragraph) shall be included in all copies or substantial portions of the
- * Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
- * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
- * IN THE SOFTWARE.
- *
- * Authors:
- *     Jesse Barnes <jbarnes@virtuousgeek.org>
- *
- */
-#ifndef _DRM_MEM_UTIL_H_
-#define _DRM_MEM_UTIL_H_
-
-#include <linux/vmalloc.h>
-
-static __inline__ void *drm_calloc_large(size_t nmemb, size_t size)
-{
-	if (size != 0 && nmemb > SIZE_MAX / size)
-		return NULL;
-
-	if (size * nmemb <= PAGE_SIZE)
-	    return kcalloc(nmemb, size, GFP_KERNEL);
-
-	return vzalloc(size * nmemb);
-}
-
-/* Modeled after cairo's malloc_ab, it's like calloc but without the zeroing. */
-static __inline__ void *drm_malloc_ab(size_t nmemb, size_t size)
-{
-	if (size != 0 && nmemb > SIZE_MAX / size)
-		return NULL;
-
-	if (size * nmemb <= PAGE_SIZE)
-	    return kmalloc(nmemb * size, GFP_KERNEL);
-
-	return vmalloc(size * nmemb);
-}
-
-static __inline__ void *drm_malloc_gfp(size_t nmemb, size_t size, gfp_t gfp)
-{
-	if (size != 0 && nmemb > SIZE_MAX / size)
-		return NULL;
-
-	if (size * nmemb <= PAGE_SIZE)
-		return kmalloc(nmemb * size, gfp);
-
-	if (gfp & __GFP_RECLAIMABLE) {
-		void *ptr = kmalloc(nmemb * size,
-				    gfp | __GFP_NOWARN | __GFP_NORETRY);
-		if (ptr)
-			return ptr;
-	}
-
-	return __vmalloc(size * nmemb, gfp, PAGE_KERNEL);
-}
-
-static __inline void drm_free_large(void *ptr)
-{
-	kvfree(ptr);
-}
-
-#endif
diff --git a/include/drm/drm_modeset_helper_vtables.h b/include/drm/drm_modeset_helper_vtables.h
index c01c328f6cc8..85984b208218 100644
--- a/include/drm/drm_modeset_helper_vtables.h
+++ b/include/drm/drm_modeset_helper_vtables.h
@@ -106,6 +106,40 @@ struct drm_crtc_helper_funcs {
 	void (*commit)(struct drm_crtc *crtc);
 
 	/**
+	 * @mode_valid:
+	 *
+	 * This callback is used to check if a specific mode is valid in this
+	 * crtc. This should be implemented if the crtc has some sort of
+	 * restriction in the modes it can display. For example, a given crtc
+	 * may be responsible to set a clock value. If the clock can not
+	 * produce all the values for the available modes then this callback
+	 * can be used to restrict the number of modes to only the ones that
+	 * can be displayed.
+	 *
+	 * This hook is used by the probe helpers to filter the mode list in
+	 * drm_helper_probe_single_connector_modes(), and it is used by the
+	 * atomic helpers to validate modes supplied by userspace in
+	 * drm_atomic_helper_check_modeset().
+	 *
+	 * This function is optional.
+	 *
+	 * NOTE:
+	 *
+	 * Since this function is both called from the check phase of an atomic
+	 * commit, and the mode validation in the probe paths it is not allowed
+	 * to look at anything else but the passed-in mode, and validate it
+	 * against configuration-invariant hardward constraints. Any further
+	 * limits which depend upon the configuration can only be checked in
+	 * @mode_fixup or @atomic_check.
+	 *
+	 * RETURNS:
+	 *
+	 * drm_mode_status Enum
+	 */
+	enum drm_mode_status (*mode_valid)(struct drm_crtc *crtc,
+					   const struct drm_display_mode *mode);
+
+	/**
 	 * @mode_fixup:
 	 *
 	 * This callback is used to validate a mode. The parameter mode is the
@@ -113,7 +147,8 @@ struct drm_crtc_helper_funcs {
 	 * encoders need to be fed with. Note that this is the inverse semantics
 	 * of the meaning for the &drm_encoder and &drm_bridge_funcs.mode_fixup
 	 * vfunc. If the CRTC cannot support the requested conversion from mode
-	 * to adjusted_mode it should reject the modeset.
+	 * to adjusted_mode it should reject the modeset. See also
+	 * &drm_crtc_state.adjusted_mode for more details.
 	 *
 	 * This function is used by both legacy CRTC helpers and atomic helpers.
 	 * With atomic helpers it is optional.
@@ -130,22 +165,17 @@ struct drm_crtc_helper_funcs {
 	 * allowed.
 	 *
 	 * Atomic drivers which need to inspect and adjust more state should
-	 * instead use the @atomic_check callback.
-	 *
-	 * Also beware that neither core nor helpers filter modes before
-	 * passing them to the driver: While the list of modes that is
-	 * advertised to userspace is filtered using the
-	 * &drm_connector.mode_valid callback, neither the core nor the helpers
-	 * do any filtering on modes passed in from userspace when setting a
-	 * mode. It is therefore possible for userspace to pass in a mode that
-	 * was previously filtered out using &drm_connector.mode_valid or add a
-	 * custom mode that wasn't probed from EDID or similar to begin with.
-	 * Even though this is an advanced feature and rarely used nowadays,
-	 * some users rely on being able to specify modes manually so drivers
-	 * must be prepared to deal with it. Specifically this means that all
-	 * drivers need not only validate modes in &drm_connector.mode_valid but
-	 * also in this or in the &drm_encoder_helper_funcs.mode_fixup callback
-	 * to make sure invalid modes passed in from userspace are rejected.
+	 * instead use the @atomic_check callback, but note that they're not
+	 * perfectly equivalent: @mode_valid is called from
+	 * drm_atomic_helper_check_modeset(), but @atomic_check is called from
+	 * drm_atomic_helper_check_planes(), because originally it was meant for
+	 * plane update checks only.
+	 *
+	 * Also beware that userspace can request its own custom modes, neither
+	 * core nor helpers filter modes to the list of probe modes reported by
+	 * the GETCONNECTOR IOCTL and stored in &drm_connector.modes. To ensure
+	 * that modes are filtered consistently put any CRTC constraints and
+	 * limits checks into @mode_valid.
 	 *
 	 * RETURNS:
 	 *
@@ -341,6 +371,12 @@ struct drm_crtc_helper_funcs {
 	 * state objects passed-in or assembled in the overall &drm_atomic_state
 	 * update tracking structure.
 	 *
+	 * Also beware that userspace can request its own custom modes, neither
+	 * core nor helpers filter modes to the list of probe modes reported by
+	 * the GETCONNECTOR IOCTL and stored in &drm_connector.modes. To ensure
+	 * that modes are filtered consistently put any CRTC constraints and
+	 * limits checks into @mode_valid.
+	 *
 	 * RETURNS:
 	 *
 	 * 0 on success, -EINVAL if the state or the transition can't be
@@ -457,13 +493,48 @@ struct drm_encoder_helper_funcs {
 	void (*dpms)(struct drm_encoder *encoder, int mode);
 
 	/**
+	 * @mode_valid:
+	 *
+	 * This callback is used to check if a specific mode is valid in this
+	 * encoder. This should be implemented if the encoder has some sort
+	 * of restriction in the modes it can display. For example, a given
+	 * encoder may be responsible to set a clock value. If the clock can
+	 * not produce all the values for the available modes then this callback
+	 * can be used to restrict the number of modes to only the ones that
+	 * can be displayed.
+	 *
+	 * This hook is used by the probe helpers to filter the mode list in
+	 * drm_helper_probe_single_connector_modes(), and it is used by the
+	 * atomic helpers to validate modes supplied by userspace in
+	 * drm_atomic_helper_check_modeset().
+	 *
+	 * This function is optional.
+	 *
+	 * NOTE:
+	 *
+	 * Since this function is both called from the check phase of an atomic
+	 * commit, and the mode validation in the probe paths it is not allowed
+	 * to look at anything else but the passed-in mode, and validate it
+	 * against configuration-invariant hardward constraints. Any further
+	 * limits which depend upon the configuration can only be checked in
+	 * @mode_fixup or @atomic_check.
+	 *
+	 * RETURNS:
+	 *
+	 * drm_mode_status Enum
+	 */
+	enum drm_mode_status (*mode_valid)(struct drm_encoder *crtc,
+					   const struct drm_display_mode *mode);
+
+	/**
 	 * @mode_fixup:
 	 *
 	 * This callback is used to validate and adjust a mode. The parameter
 	 * mode is the display mode that should be fed to the next element in
 	 * the display chain, either the final &drm_connector or a &drm_bridge.
 	 * The parameter adjusted_mode is the input mode the encoder requires. It
-	 * can be modified by this callback and does not need to match mode.
+	 * can be modified by this callback and does not need to match mode. See
+	 * also &drm_crtc_state.adjusted_mode for more details.
 	 *
 	 * This function is used by both legacy CRTC helpers and atomic helpers.
 	 * This hook is optional.
@@ -480,23 +551,15 @@ struct drm_encoder_helper_funcs {
 	 * allowed.
 	 *
 	 * Atomic drivers which need to inspect and adjust more state should
-	 * instead use the @atomic_check callback.
-	 *
-	 * Also beware that neither core nor helpers filter modes before
-	 * passing them to the driver: While the list of modes that is
-	 * advertised to userspace is filtered using the connector's
-	 * &drm_connector_helper_funcs.mode_valid callback, neither the core nor
-	 * the helpers do any filtering on modes passed in from userspace when
-	 * setting a mode. It is therefore possible for userspace to pass in a
-	 * mode that was previously filtered out using
-	 * &drm_connector_helper_funcs.mode_valid or add a custom mode that
-	 * wasn't probed from EDID or similar to begin with.  Even though this
-	 * is an advanced feature and rarely used nowadays, some users rely on
-	 * being able to specify modes manually so drivers must be prepared to
-	 * deal with it. Specifically this means that all drivers need not only
-	 * validate modes in &drm_connector.mode_valid but also in this or in
-	 * the &drm_crtc_helper_funcs.mode_fixup callback to make sure
-	 * invalid modes passed in from userspace are rejected.
+	 * instead use the @atomic_check callback. If @atomic_check is used,
+	 * this hook isn't called since @atomic_check allows a strict superset
+	 * of the functionality of @mode_fixup.
+	 *
+	 * Also beware that userspace can request its own custom modes, neither
+	 * core nor helpers filter modes to the list of probe modes reported by
+	 * the GETCONNECTOR IOCTL and stored in &drm_connector.modes. To ensure
+	 * that modes are filtered consistently put any encoder constraints and
+	 * limits checks into @mode_valid.
 	 *
 	 * RETURNS:
 	 *
@@ -677,6 +740,11 @@ struct drm_encoder_helper_funcs {
 	 * update the CRTC to match what the encoder needs for the requested
 	 * connector.
 	 *
+	 * Since this provides a strict superset of the functionality of
+	 * @mode_fixup (the requested and adjusted modes are both available
+	 * through the passed in &struct drm_crtc_state) @mode_fixup is not
+	 * called when @atomic_check is implemented.
+	 *
 	 * This function is used by the atomic helpers, but it is optional.
 	 *
 	 * NOTE:
@@ -686,6 +754,12 @@ struct drm_encoder_helper_funcs {
 	 * state objects passed-in or assembled in the overall &drm_atomic_state
 	 * update tracking structure.
 	 *
+	 * Also beware that userspace can request its own custom modes, neither
+	 * core nor helpers filter modes to the list of probe modes reported by
+	 * the GETCONNECTOR IOCTL and stored in &drm_connector.modes. To ensure
+	 * that modes are filtered consistently put any encoder constraints and
+	 * limits checks into @mode_valid.
+	 *
 	 * RETURNS:
 	 *
 	 * 0 on success, -EINVAL if the state or the transition can't be
@@ -795,13 +869,20 @@ struct drm_connector_helper_funcs {
 	 * (which is usually derived from the EDID data block from the sink).
 	 * See e.g. drm_helper_probe_single_connector_modes().
 	 *
+	 * This function is optional.
+	 *
 	 * NOTE:
 	 *
 	 * This only filters the mode list supplied to userspace in the
-	 * GETCONNECOTR IOCTL. Userspace is free to create modes of its own and
-	 * ask the kernel to use them. It this case the atomic helpers or legacy
-	 * CRTC helpers will not call this function. Drivers therefore must
-	 * still fully validate any mode passed in in a modeset request.
+	 * GETCONNECTOR IOCTL. Compared to &drm_encoder_helper_funcs.mode_valid,
+	 * &drm_crtc_helper_funcs.mode_valid and &drm_bridge_funcs.mode_valid,
+	 * which are also called by the atomic helpers from
+	 * drm_atomic_helper_check_modeset(). This allows userspace to force and
+	 * ignore sink constraint (like the pixel clock limits in the screen's
+	 * EDID), which is useful for e.g. testing, or working around a broken
+	 * EDID. Any source hardware constraint (which always need to be
+	 * enforced) therefore should be checked in one of the above callbacks,
+	 * and not this one here.
 	 *
 	 * To avoid races with concurrent connector state updates, the helper
 	 * libraries always call this with the &drm_mode_config.connection_mutex
diff --git a/include/drm/drm_panel.h b/include/drm/drm_panel.h
index 1b364b0100f4..14ac240a1f64 100644
--- a/include/drm/drm_panel.h
+++ b/include/drm/drm_panel.h
@@ -24,8 +24,10 @@
 #ifndef __DRM_PANEL_H__
 #define __DRM_PANEL_H__
 
+#include <linux/errno.h>
 #include <linux/list.h>
 
+struct device_node;
 struct drm_connector;
 struct drm_device;
 struct drm_panel;
diff --git a/include/drm/drm_prime.h b/include/drm/drm_prime.h
index 0b2a235c4be0..59ccab402e85 100644
--- a/include/drm/drm_prime.h
+++ b/include/drm/drm_prime.h
@@ -50,6 +50,8 @@ struct drm_prime_file_private {
 	struct rb_root handles;
 };
 
+struct device;
+
 struct dma_buf_export_info;
 struct dma_buf;
 
@@ -65,6 +67,11 @@ int drm_gem_prime_handle_to_fd(struct drm_device *dev,
 			       int *prime_fd);
 struct drm_gem_object *drm_gem_prime_import(struct drm_device *dev,
 					    struct dma_buf *dma_buf);
+
+struct drm_gem_object *drm_gem_prime_import_dev(struct drm_device *dev,
+						struct dma_buf *dma_buf,
+						struct device *attach_dev);
+
 int drm_gem_prime_fd_to_handle(struct drm_device *dev,
 			       struct drm_file *file_priv, int prime_fd, uint32_t *handle);
 struct dma_buf *drm_gem_dmabuf_export(struct drm_device *dev,
diff --git a/include/drm/i915_pciids.h b/include/drm/i915_pciids.h
index 27e0dbaa6c0e..34c8f5600ce0 100644
--- a/include/drm/i915_pciids.h
+++ b/include/drm/i915_pciids.h
@@ -334,4 +334,44 @@
 	INTEL_KBL_GT3_IDS(info), \
 	INTEL_KBL_GT4_IDS(info)
 
+/* CFL S */
+#define INTEL_CFL_S_IDS(info) \
+	INTEL_VGA_DEVICE(0x3E90, info), /* SRV GT1 */ \
+	INTEL_VGA_DEVICE(0x3E93, info), /* SRV GT1 */ \
+	INTEL_VGA_DEVICE(0x3E91, info), /* SRV GT2 */ \
+	INTEL_VGA_DEVICE(0x3E92, info), /* SRV GT2 */ \
+	INTEL_VGA_DEVICE(0x3E96, info)  /* SRV GT2 */
+
+/* CFL H */
+#define INTEL_CFL_H_IDS(info) \
+	INTEL_VGA_DEVICE(0x3E9B, info), /* Halo GT2 */ \
+	INTEL_VGA_DEVICE(0x3E94, info)  /* Halo GT2 */
+
+/* CFL U */
+#define INTEL_CFL_U_IDS(info) \
+	INTEL_VGA_DEVICE(0x3EA6, info), /* ULT GT3 */ \
+	INTEL_VGA_DEVICE(0x3EA7, info), /* ULT GT3 */ \
+	INTEL_VGA_DEVICE(0x3EA8, info), /* ULT GT3 */ \
+	INTEL_VGA_DEVICE(0x3EA5, info)  /* ULT GT3 */
+
+/* CNL U 2+2 */
+#define INTEL_CNL_U_GT2_IDS(info) \
+	INTEL_VGA_DEVICE(0x5A52, info), \
+	INTEL_VGA_DEVICE(0x5A5A, info), \
+	INTEL_VGA_DEVICE(0x5A42, info), \
+	INTEL_VGA_DEVICE(0x5A4A, info)
+
+/* CNL Y 2+2 */
+#define INTEL_CNL_Y_GT2_IDS(info) \
+	INTEL_VGA_DEVICE(0x5A51, info), \
+	INTEL_VGA_DEVICE(0x5A59, info), \
+	INTEL_VGA_DEVICE(0x5A41, info), \
+	INTEL_VGA_DEVICE(0x5A49, info), \
+	INTEL_VGA_DEVICE(0x5A71, info), \
+	INTEL_VGA_DEVICE(0x5A79, info)
+
+#define INTEL_CNL_IDS(info) \
+	INTEL_CNL_U_GT2_IDS(info), \
+	INTEL_CNL_Y_GT2_IDS(info)
+
 #endif /* _I915_PCIIDS_H */
diff --git a/include/drm/intel_lpe_audio.h b/include/drm/intel_lpe_audio.h
index e9892b4c3af1..b6121c8fe539 100644
--- a/include/drm/intel_lpe_audio.h
+++ b/include/drm/intel_lpe_audio.h
@@ -31,20 +31,20 @@ struct platform_device;
 
 #define HDMI_MAX_ELD_BYTES	128
 
-struct intel_hdmi_lpe_audio_eld {
-	int port_id;
-	int pipe_id;
-	unsigned char eld_data[HDMI_MAX_ELD_BYTES];
+struct intel_hdmi_lpe_audio_port_pdata {
+	u8 eld[HDMI_MAX_ELD_BYTES];
+	int port;
+	int pipe;
+	int ls_clock;
+	bool dp_output;
 };
 
 struct intel_hdmi_lpe_audio_pdata {
-	bool notify_pending;
-	int tmds_clock_speed;
-	bool hdmi_connected;
-	bool dp_output;
-	int link_rate;
-	struct intel_hdmi_lpe_audio_eld eld;
-	void (*notify_audio_lpe)(struct platform_device *pdev);
+	struct intel_hdmi_lpe_audio_port_pdata port[3]; /* for ports B,C,D */
+	int num_ports;
+	int num_pipes;
+
+	void (*notify_audio_lpe)(struct platform_device *pdev, int port); /* port: 0==B,1==C,2==D */
 	spinlock_t lpe_audio_slock;
 };
 
diff --git a/include/drm/ttm/ttm_bo_driver.h b/include/drm/ttm/ttm_bo_driver.h
index 6bbd34d25a8d..990d529f823c 100644
--- a/include/drm/ttm/ttm_bo_driver.h
+++ b/include/drm/ttm/ttm_bo_driver.h
@@ -30,10 +30,6 @@
 #ifndef _TTM_BO_DRIVER_H_
 #define _TTM_BO_DRIVER_H_
 
-#include <ttm/ttm_bo_api.h>
-#include <ttm/ttm_memory.h>
-#include <ttm/ttm_module.h>
-#include <ttm/ttm_placement.h>
 #include <drm/drm_mm.h>
 #include <drm/drm_global.h>
 #include <drm/drm_vma_manager.h>
@@ -42,6 +38,11 @@
 #include <linux/spinlock.h>
 #include <linux/reservation.h>
 
+#include "ttm_bo_api.h"
+#include "ttm_memory.h"
+#include "ttm_module.h"
+#include "ttm_placement.h"
+
 #define TTM_MAX_BO_PRIORITY	4U
 
 struct ttm_backend_func {
diff --git a/include/drm/ttm/ttm_execbuf_util.h b/include/drm/ttm/ttm_execbuf_util.h
index 47f35b8e6d09..b0fdd1980034 100644
--- a/include/drm/ttm/ttm_execbuf_util.h
+++ b/include/drm/ttm/ttm_execbuf_util.h
@@ -31,9 +31,10 @@
 #ifndef _TTM_EXECBUF_UTIL_H_
 #define _TTM_EXECBUF_UTIL_H_
 
-#include <ttm/ttm_bo_api.h>
 #include <linux/list.h>
 
+#include "ttm_bo_api.h"
+
 /**
  * struct ttm_validate_buffer
  *
diff --git a/include/drm/ttm/ttm_lock.h b/include/drm/ttm/ttm_lock.h
index 2902beb5f689..0c3af9836863 100644
--- a/include/drm/ttm/ttm_lock.h
+++ b/include/drm/ttm/ttm_lock.h
@@ -49,10 +49,11 @@
 #ifndef _TTM_LOCK_H_
 #define _TTM_LOCK_H_
 
-#include <ttm/ttm_object.h>
 #include <linux/wait.h>
 #include <linux/atomic.h>
 
+#include "ttm_object.h"
+
 /**
  * struct ttm_lock
  *
diff --git a/include/drm/ttm/ttm_object.h b/include/drm/ttm/ttm_object.h
index 1487011fe057..a98bfeb4239e 100644
--- a/include/drm/ttm/ttm_object.h
+++ b/include/drm/ttm/ttm_object.h
@@ -42,7 +42,8 @@
 #include <linux/kref.h>
 #include <linux/rcupdate.h>
 #include <linux/dma-buf.h>
-#include <ttm/ttm_memory.h>
+
+#include "ttm_memory.h"
 
 /**
  * enum ttm_ref_type
diff --git a/include/linux/amba/clcd-regs.h b/include/linux/amba/clcd-regs.h
new file mode 100644
index 000000000000..516a6fda83c5
--- /dev/null
+++ b/include/linux/amba/clcd-regs.h
@@ -0,0 +1,86 @@
+/*
+ * David A Rusling
+ *
+ * Copyright (C) 2001 ARM Limited
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file COPYING in the main directory of this archive
+ * for more details.
+ */
+
+#ifndef AMBA_CLCD_REGS_H
+#define AMBA_CLCD_REGS_H
+
+/*
+ * CLCD Controller Internal Register addresses
+ */
+#define CLCD_TIM0		0x00000000
+#define CLCD_TIM1 		0x00000004
+#define CLCD_TIM2 		0x00000008
+#define CLCD_TIM3 		0x0000000c
+#define CLCD_UBAS 		0x00000010
+#define CLCD_LBAS 		0x00000014
+
+#define CLCD_PL110_IENB		0x00000018
+#define CLCD_PL110_CNTL		0x0000001c
+#define CLCD_PL110_STAT		0x00000020
+#define CLCD_PL110_INTR 	0x00000024
+#define CLCD_PL110_UCUR		0x00000028
+#define CLCD_PL110_LCUR		0x0000002C
+
+#define CLCD_PL111_CNTL		0x00000018
+#define CLCD_PL111_IENB		0x0000001c
+#define CLCD_PL111_RIS		0x00000020
+#define CLCD_PL111_MIS		0x00000024
+#define CLCD_PL111_ICR		0x00000028
+#define CLCD_PL111_UCUR		0x0000002c
+#define CLCD_PL111_LCUR		0x00000030
+
+#define CLCD_PALL 		0x00000200
+#define CLCD_PALETTE		0x00000200
+
+#define TIM2_PCD_LO_MASK	GENMASK(4, 0)
+#define TIM2_PCD_LO_BITS	5
+#define TIM2_CLKSEL		(1 << 5)
+#define TIM2_IVS		(1 << 11)
+#define TIM2_IHS		(1 << 12)
+#define TIM2_IPC		(1 << 13)
+#define TIM2_IOE		(1 << 14)
+#define TIM2_BCD		(1 << 26)
+#define TIM2_PCD_HI_MASK	GENMASK(31, 27)
+#define TIM2_PCD_HI_BITS	5
+#define TIM2_PCD_HI_SHIFT	27
+
+#define CNTL_LCDEN		(1 << 0)
+#define CNTL_LCDBPP1		(0 << 1)
+#define CNTL_LCDBPP2		(1 << 1)
+#define CNTL_LCDBPP4		(2 << 1)
+#define CNTL_LCDBPP8		(3 << 1)
+#define CNTL_LCDBPP16		(4 << 1)
+#define CNTL_LCDBPP16_565	(6 << 1)
+#define CNTL_LCDBPP16_444	(7 << 1)
+#define CNTL_LCDBPP24		(5 << 1)
+#define CNTL_LCDBW		(1 << 4)
+#define CNTL_LCDTFT		(1 << 5)
+#define CNTL_LCDMONO8		(1 << 6)
+#define CNTL_LCDDUAL		(1 << 7)
+#define CNTL_BGR		(1 << 8)
+#define CNTL_BEBO		(1 << 9)
+#define CNTL_BEPO		(1 << 10)
+#define CNTL_LCDPWR		(1 << 11)
+#define CNTL_LCDVCOMP(x)	((x) << 12)
+#define CNTL_LDMAFIFOTIME	(1 << 15)
+#define CNTL_WATERMARK		(1 << 16)
+
+/* ST Microelectronics variant bits */
+#define CNTL_ST_1XBPP_444	0x0
+#define CNTL_ST_1XBPP_5551	(1 << 17)
+#define CNTL_ST_1XBPP_565	(1 << 18)
+#define CNTL_ST_CDWID_12	0x0
+#define CNTL_ST_CDWID_16	(1 << 19)
+#define CNTL_ST_CDWID_18	(1 << 20)
+#define CNTL_ST_CDWID_24	((1 << 19)|(1 << 20))
+#define CNTL_ST_CEAEN		(1 << 21)
+#define CNTL_ST_LCDBPP24_PACKED	(6 << 1)
+
+#endif /* AMBA_CLCD_REGS_H */
diff --git a/include/linux/amba/clcd.h b/include/linux/amba/clcd.h
index 1035879b322c..d0c3be77c18e 100644
--- a/include/linux/amba/clcd.h
+++ b/include/linux/amba/clcd.h
@@ -10,73 +10,7 @@
  * for more details.
  */
 #include <linux/fb.h>
-
-/*
- * CLCD Controller Internal Register addresses
- */
-#define CLCD_TIM0		0x00000000
-#define CLCD_TIM1 		0x00000004
-#define CLCD_TIM2 		0x00000008
-#define CLCD_TIM3 		0x0000000c
-#define CLCD_UBAS 		0x00000010
-#define CLCD_LBAS 		0x00000014
-
-#define CLCD_PL110_IENB		0x00000018
-#define CLCD_PL110_CNTL		0x0000001c
-#define CLCD_PL110_STAT		0x00000020
-#define CLCD_PL110_INTR 	0x00000024
-#define CLCD_PL110_UCUR		0x00000028
-#define CLCD_PL110_LCUR		0x0000002C
-
-#define CLCD_PL111_CNTL		0x00000018
-#define CLCD_PL111_IENB		0x0000001c
-#define CLCD_PL111_RIS		0x00000020
-#define CLCD_PL111_MIS		0x00000024
-#define CLCD_PL111_ICR		0x00000028
-#define CLCD_PL111_UCUR		0x0000002c
-#define CLCD_PL111_LCUR		0x00000030
-
-#define CLCD_PALL 		0x00000200
-#define CLCD_PALETTE		0x00000200
-
-#define TIM2_CLKSEL		(1 << 5)
-#define TIM2_IVS		(1 << 11)
-#define TIM2_IHS		(1 << 12)
-#define TIM2_IPC		(1 << 13)
-#define TIM2_IOE		(1 << 14)
-#define TIM2_BCD		(1 << 26)
-
-#define CNTL_LCDEN		(1 << 0)
-#define CNTL_LCDBPP1		(0 << 1)
-#define CNTL_LCDBPP2		(1 << 1)
-#define CNTL_LCDBPP4		(2 << 1)
-#define CNTL_LCDBPP8		(3 << 1)
-#define CNTL_LCDBPP16		(4 << 1)
-#define CNTL_LCDBPP16_565	(6 << 1)
-#define CNTL_LCDBPP16_444	(7 << 1)
-#define CNTL_LCDBPP24		(5 << 1)
-#define CNTL_LCDBW		(1 << 4)
-#define CNTL_LCDTFT		(1 << 5)
-#define CNTL_LCDMONO8		(1 << 6)
-#define CNTL_LCDDUAL		(1 << 7)
-#define CNTL_BGR		(1 << 8)
-#define CNTL_BEBO		(1 << 9)
-#define CNTL_BEPO		(1 << 10)
-#define CNTL_LCDPWR		(1 << 11)
-#define CNTL_LCDVCOMP(x)	((x) << 12)
-#define CNTL_LDMAFIFOTIME	(1 << 15)
-#define CNTL_WATERMARK		(1 << 16)
-
-/* ST Microelectronics variant bits */
-#define CNTL_ST_1XBPP_444	0x0
-#define CNTL_ST_1XBPP_5551	(1 << 17)
-#define CNTL_ST_1XBPP_565	(1 << 18)
-#define CNTL_ST_CDWID_12	0x0
-#define CNTL_ST_CDWID_16	(1 << 19)
-#define CNTL_ST_CDWID_18	(1 << 20)
-#define CNTL_ST_CDWID_24	((1 << 19)|(1 << 20))
-#define CNTL_ST_CEAEN		(1 << 21)
-#define CNTL_ST_LCDBPP24_PACKED	(6 << 1)
+#include <linux/amba/clcd-regs.h>
 
 enum {
 	/* individual formats */
diff --git a/include/linux/sync_file.h b/include/linux/sync_file.h
index 3e3ab84fc4cd..5726107963b2 100644
--- a/include/linux/sync_file.h
+++ b/include/linux/sync_file.h
@@ -14,7 +14,6 @@
 #define _LINUX_SYNC_FILE_H
 
 #include <linux/types.h>
-#include <linux/kref.h>
 #include <linux/ktime.h>
 #include <linux/list.h>
 #include <linux/spinlock.h>
@@ -24,8 +23,6 @@
 /**
  * struct sync_file - sync file to export to the userspace
  * @file:		file representing this fence
- * @kref:		reference count on fence.
- * @name:		name of sync_file.  Useful for debugging
  * @sync_file_list:	membership in global file list
  * @wq:			wait queue for fence signaling
  * @fence:		fence with the fences in the sync_file
@@ -33,8 +30,14 @@
  */
 struct sync_file {
 	struct file		*file;
-	struct kref		kref;
-	char			name[32];
+	/**
+	 * @user_name:
+	 *
+	 * Name of the sync file provided by userspace, for merged fences.
+	 * Otherwise generated through driver callbacks (in which case the
+	 * entire array is 0).
+	 */
+	char			user_name[32];
 #ifdef CONFIG_DEBUG_FS
 	struct list_head	sync_file_list;
 #endif
@@ -49,5 +52,6 @@ struct sync_file {
 
 struct sync_file *sync_file_create(struct dma_fence *fence);
 struct dma_fence *sync_file_get_fence(int fd);
+char *sync_file_get_name(struct sync_file *sync_file, char *buf, int len);
 
 #endif /* _LINUX_SYNC_H */
diff --git a/include/uapi/drm/drm_mode.h b/include/uapi/drm/drm_mode.h
index 8c67fc03d53d..403339f98a92 100644
--- a/include/uapi/drm/drm_mode.h
+++ b/include/uapi/drm/drm_mode.h
@@ -75,7 +75,7 @@ extern "C" {
   * (define not exposed to user space).
   */
 #define DRM_MODE_FLAG_3D_MASK			(0x1f<<14)
-#define  DRM_MODE_FLAG_3D_NONE			(0<<14)
+#define  DRM_MODE_FLAG_3D_NONE		(0<<14)
 #define  DRM_MODE_FLAG_3D_FRAME_PACKING		(1<<14)
 #define  DRM_MODE_FLAG_3D_FIELD_ALTERNATIVE	(2<<14)
 #define  DRM_MODE_FLAG_3D_LINE_ALTERNATIVE	(3<<14)
@@ -127,6 +127,53 @@ extern "C" {
 #define DRM_MODE_LINK_STATUS_GOOD	0
 #define DRM_MODE_LINK_STATUS_BAD	1
 
+/*
+ * DRM_MODE_ROTATE_<degrees>
+ *
+ * Signals that a drm plane is been rotated <degrees> degrees in counter
+ * clockwise direction.
+ *
+ * This define is provided as a convenience, looking up the property id
+ * using the name->prop id lookup is the preferred method.
+ */
+#define DRM_MODE_ROTATE_0       (1<<0)
+#define DRM_MODE_ROTATE_90      (1<<1)
+#define DRM_MODE_ROTATE_180     (1<<2)
+#define DRM_MODE_ROTATE_270     (1<<3)
+
+/*
+ * DRM_MODE_ROTATE_MASK
+ *
+ * Bitmask used to look for drm plane rotations.
+ */
+#define DRM_MODE_ROTATE_MASK (\
+		DRM_MODE_ROTATE_0  | \
+		DRM_MODE_ROTATE_90  | \
+		DRM_MODE_ROTATE_180 | \
+		DRM_MODE_ROTATE_270)
+
+/*
+ * DRM_MODE_REFLECT_<axis>
+ *
+ * Signals that the contents of a drm plane is reflected in the <axis> axis,
+ * in the same way as mirroring.
+ *
+ * This define is provided as a convenience, looking up the property id
+ * using the name->prop id lookup is the preferred method.
+ */
+#define DRM_MODE_REFLECT_X      (1<<4)
+#define DRM_MODE_REFLECT_Y      (1<<5)
+
+/*
+ * DRM_MODE_REFLECT_MASK
+ *
+ * Bitmask used to look for drm plane reflections.
+ */
+#define DRM_MODE_REFLECT_MASK (\
+		DRM_MODE_REFLECT_X | \
+		DRM_MODE_REFLECT_Y)
+
+
 struct drm_mode_modeinfo {
 	__u32 clock;
 	__u16 hdisplay;
diff --git a/include/uapi/drm/i915_drm.h b/include/uapi/drm/i915_drm.h
index 3554495bef13..f24a80d2d42e 100644
--- a/include/uapi/drm/i915_drm.h
+++ b/include/uapi/drm/i915_drm.h
@@ -412,6 +412,12 @@ typedef struct drm_i915_irq_wait {
  */
 #define I915_PARAM_HAS_EXEC_FENCE	 44
 
+/* Query whether DRM_I915_GEM_EXECBUFFER2 supports the ability to capture
+ * user specified bufffers for post-mortem debugging of GPU hangs. See
+ * EXEC_OBJECT_CAPTURE.
+ */
+#define I915_PARAM_HAS_EXEC_CAPTURE	 45
+
 typedef struct drm_i915_getparam {
 	__s32 param;
 	/*
@@ -666,6 +672,8 @@ struct drm_i915_gem_relocation_entry {
 #define I915_GEM_DOMAIN_VERTEX		0x00000020
 /** GTT domain - aperture and scanout */
 #define I915_GEM_DOMAIN_GTT		0x00000040
+/** WC domain - uncached access */
+#define I915_GEM_DOMAIN_WC		0x00000080
 /** @} */
 
 struct drm_i915_gem_exec_object {
@@ -773,8 +781,15 @@ struct drm_i915_gem_exec_object2 {
  * I915_PARAM_HAS_EXEC_FENCE to order execbufs and execute them asynchronously.
  */
 #define EXEC_OBJECT_ASYNC		(1<<6)
+/* Request that the contents of this execobject be copied into the error
+ * state upon a GPU hang involving this batch for post-mortem debugging.
+ * These buffers are recorded in no particular order as "user" in
+ * /sys/class/drm/cardN/error. Query I915_PARAM_HAS_EXEC_CAPTURE to see
+ * if the kernel supports this flag.
+ */
+#define EXEC_OBJECT_CAPTURE		(1<<7)
 /* All remaining bits are MBZ and RESERVED FOR FUTURE USE */
-#define __EXEC_OBJECT_UNKNOWN_FLAGS -(EXEC_OBJECT_ASYNC<<1)
+#define __EXEC_OBJECT_UNKNOWN_FLAGS -(EXEC_OBJECT_CAPTURE<<1)
 	__u64 flags;
 
 	union {
diff --git a/sound/x86/intel_hdmi_audio.c b/sound/x86/intel_hdmi_audio.c
index b11d3920b9a5..c19efc9708d7 100644
--- a/sound/x86/intel_hdmi_audio.c
+++ b/sound/x86/intel_hdmi_audio.c
@@ -42,6 +42,11 @@
 #include <drm/intel_lpe_audio.h>
 #include "intel_hdmi_audio.h"
 
+#define for_each_pipe(card_ctx, pipe) \
+	for ((pipe) = 0; (pipe) < (card_ctx)->num_pipes; (pipe)++)
+#define for_each_port(card_ctx, port) \
+	for ((port) = 0; (port) < (card_ctx)->num_ports; (port)++)
+
 /*standard module options for ALSA. This module supports only one card*/
 static int hdmi_card_index = SNDRV_DEFAULT_IDX1;
 static char *hdmi_card_id = SNDRV_DEFAULT_STR1;
@@ -189,15 +194,30 @@ static void had_substream_put(struct snd_intelhad *intelhaddata)
 	spin_unlock_irqrestore(&intelhaddata->had_spinlock, flags);
 }
 
+static u32 had_config_offset(int pipe)
+{
+	switch (pipe) {
+	default:
+	case 0:
+		return AUDIO_HDMI_CONFIG_A;
+	case 1:
+		return AUDIO_HDMI_CONFIG_B;
+	case 2:
+		return AUDIO_HDMI_CONFIG_C;
+	}
+}
+
 /* Register access functions */
-static u32 had_read_register_raw(struct snd_intelhad *ctx, u32 reg)
+static u32 had_read_register_raw(struct snd_intelhad_card *card_ctx,
+				 int pipe, u32 reg)
 {
-	return ioread32(ctx->mmio_start + ctx->had_config_offset + reg);
+	return ioread32(card_ctx->mmio_start + had_config_offset(pipe) + reg);
 }
 
-static void had_write_register_raw(struct snd_intelhad *ctx, u32 reg, u32 val)
+static void had_write_register_raw(struct snd_intelhad_card *card_ctx,
+				   int pipe, u32 reg, u32 val)
 {
-	iowrite32(val, ctx->mmio_start + ctx->had_config_offset + reg);
+	iowrite32(val, card_ctx->mmio_start + had_config_offset(pipe) + reg);
 }
 
 static void had_read_register(struct snd_intelhad *ctx, u32 reg, u32 *val)
@@ -205,13 +225,13 @@ static void had_read_register(struct snd_intelhad *ctx, u32 reg, u32 *val)
 	if (!ctx->connected)
 		*val = 0;
 	else
-		*val = had_read_register_raw(ctx, reg);
+		*val = had_read_register_raw(ctx->card_ctx, ctx->pipe, reg);
 }
 
 static void had_write_register(struct snd_intelhad *ctx, u32 reg, u32 val)
 {
 	if (ctx->connected)
-		had_write_register_raw(ctx, reg, val);
+		had_write_register_raw(ctx->card_ctx, ctx->pipe, reg, val);
 }
 
 /*
@@ -1358,6 +1378,9 @@ static void had_process_hot_plug(struct snd_intelhad *intelhaddata)
 		return;
 	}
 
+	/* Disable Audio */
+	had_enable_audio(intelhaddata, false);
+
 	intelhaddata->connected = true;
 	dev_dbg(intelhaddata->dev,
 		"%s @ %d:DEBUG PLUG/UNPLUG : HAD_DRV_CONNECTED\n",
@@ -1519,22 +1542,32 @@ static const struct snd_kcontrol_new had_controls[] = {
  */
 static irqreturn_t display_pipe_interrupt_handler(int irq, void *dev_id)
 {
-	struct snd_intelhad *ctx = dev_id;
-	u32 audio_stat;
+	struct snd_intelhad_card *card_ctx = dev_id;
+	u32 audio_stat[3] = {};
+	int pipe, port;
+
+	for_each_pipe(card_ctx, pipe) {
+		/* use raw register access to ack IRQs even while disconnected */
+		audio_stat[pipe] = had_read_register_raw(card_ctx, pipe,
+							 AUD_HDMI_STATUS) &
+			(HDMI_AUDIO_UNDERRUN | HDMI_AUDIO_BUFFER_DONE);
+
+		if (audio_stat[pipe])
+			had_write_register_raw(card_ctx, pipe,
+					       AUD_HDMI_STATUS, audio_stat[pipe]);
+	}
 
-	/* use raw register access to ack IRQs even while disconnected */
-	audio_stat = had_read_register_raw(ctx, AUD_HDMI_STATUS);
+	for_each_port(card_ctx, port) {
+		struct snd_intelhad *ctx = &card_ctx->pcm_ctx[port];
+		int pipe = ctx->pipe;
 
-	if (audio_stat & HDMI_AUDIO_UNDERRUN) {
-		had_write_register_raw(ctx, AUD_HDMI_STATUS,
-				       HDMI_AUDIO_UNDERRUN);
-		had_process_buffer_underrun(ctx);
-	}
+		if (pipe < 0)
+			continue;
 
-	if (audio_stat & HDMI_AUDIO_BUFFER_DONE) {
-		had_write_register_raw(ctx, AUD_HDMI_STATUS,
-				       HDMI_AUDIO_BUFFER_DONE);
-		had_process_buffer_done(ctx);
+		if (audio_stat[pipe] & HDMI_AUDIO_BUFFER_DONE)
+			had_process_buffer_done(ctx);
+		if (audio_stat[pipe] & HDMI_AUDIO_UNDERRUN)
+			had_process_buffer_underrun(ctx);
 	}
 
 	return IRQ_HANDLED;
@@ -1543,9 +1576,10 @@ static irqreturn_t display_pipe_interrupt_handler(int irq, void *dev_id)
 /*
  * monitor plug/unplug notification from i915; just kick off the work
  */
-static void notify_audio_lpe(struct platform_device *pdev)
+static void notify_audio_lpe(struct platform_device *pdev, int port)
 {
-	struct snd_intelhad *ctx = platform_get_drvdata(pdev);
+	struct snd_intelhad_card *card_ctx = platform_get_drvdata(pdev);
+	struct snd_intelhad *ctx = &card_ctx->pcm_ctx[port];
 
 	schedule_work(&ctx->hdmi_audio_wq);
 }
@@ -1556,47 +1590,51 @@ static void had_audio_wq(struct work_struct *work)
 	struct snd_intelhad *ctx =
 		container_of(work, struct snd_intelhad, hdmi_audio_wq);
 	struct intel_hdmi_lpe_audio_pdata *pdata = ctx->dev->platform_data;
+	struct intel_hdmi_lpe_audio_port_pdata *ppdata = &pdata->port[ctx->port];
 
 	pm_runtime_get_sync(ctx->dev);
 	mutex_lock(&ctx->mutex);
-	if (!pdata->hdmi_connected) {
-		dev_dbg(ctx->dev, "%s: Event: HAD_NOTIFY_HOT_UNPLUG\n",
-			__func__);
+	if (ppdata->pipe < 0) {
+		dev_dbg(ctx->dev, "%s: Event: HAD_NOTIFY_HOT_UNPLUG : port = %d\n",
+			__func__, ctx->port);
+
 		memset(ctx->eld, 0, sizeof(ctx->eld)); /* clear the old ELD */
+
+		ctx->dp_output = false;
+		ctx->tmds_clock_speed = 0;
+		ctx->link_rate = 0;
+
+		/* Shut down the stream */
 		had_process_hot_unplug(ctx);
-	} else {
-		struct intel_hdmi_lpe_audio_eld *eld = &pdata->eld;
 
+		ctx->pipe = -1;
+	} else {
 		dev_dbg(ctx->dev, "%s: HAD_NOTIFY_ELD : port = %d, tmds = %d\n",
-			__func__, eld->port_id,	pdata->tmds_clock_speed);
-
-		switch (eld->pipe_id) {
-		case 0:
-			ctx->had_config_offset = AUDIO_HDMI_CONFIG_A;
-			break;
-		case 1:
-			ctx->had_config_offset = AUDIO_HDMI_CONFIG_B;
-			break;
-		case 2:
-			ctx->had_config_offset = AUDIO_HDMI_CONFIG_C;
-			break;
-		default:
-			dev_dbg(ctx->dev, "Invalid pipe %d\n",
-				eld->pipe_id);
-			break;
-		}
+			__func__, ctx->port, ppdata->ls_clock);
 
-		memcpy(ctx->eld, eld->eld_data, sizeof(ctx->eld));
+		memcpy(ctx->eld, ppdata->eld, sizeof(ctx->eld));
 
-		ctx->dp_output = pdata->dp_output;
-		ctx->tmds_clock_speed = pdata->tmds_clock_speed;
-		ctx->link_rate = pdata->link_rate;
+		ctx->dp_output = ppdata->dp_output;
+		if (ctx->dp_output) {
+			ctx->tmds_clock_speed = 0;
+			ctx->link_rate = ppdata->ls_clock;
+		} else {
+			ctx->tmds_clock_speed = ppdata->ls_clock;
+			ctx->link_rate = 0;
+		}
 
+		/*
+		 * Shut down the stream before we change
+		 * the pipe assignment for this pcm device
+		 */
 		had_process_hot_plug(ctx);
 
-		/* Process mode change if stream is active */
+		ctx->pipe = ppdata->pipe;
+
+		/* Restart the stream if necessary */
 		had_process_mode_change(ctx);
 	}
+
 	mutex_unlock(&ctx->mutex);
 	pm_runtime_mark_last_busy(ctx->dev);
 	pm_runtime_put_autosuspend(ctx->dev);
@@ -1605,11 +1643,17 @@ static void had_audio_wq(struct work_struct *work)
 /*
  * Jack interface
  */
-static int had_create_jack(struct snd_intelhad *ctx)
+static int had_create_jack(struct snd_intelhad *ctx,
+			   struct snd_pcm *pcm)
 {
+	char hdmi_str[32];
 	int err;
 
-	err = snd_jack_new(ctx->card, "HDMI/DP", SND_JACK_AVOUT, &ctx->jack,
+	snprintf(hdmi_str, sizeof(hdmi_str),
+		 "HDMI/DP,pcm=%d", pcm->device);
+
+	err = snd_jack_new(ctx->card_ctx->card, hdmi_str,
+			   SND_JACK_AVOUT, &ctx->jack,
 			   true, false);
 	if (err < 0)
 		return err;
@@ -1623,13 +1667,18 @@ static int had_create_jack(struct snd_intelhad *ctx)
 
 static int hdmi_lpe_audio_runtime_suspend(struct device *dev)
 {
-	struct snd_intelhad *ctx = dev_get_drvdata(dev);
-	struct snd_pcm_substream *substream;
+	struct snd_intelhad_card *card_ctx = dev_get_drvdata(dev);
+	int port;
 
-	substream = had_substream_get(ctx);
-	if (substream) {
-		snd_pcm_suspend(substream);
-		had_substream_put(ctx);
+	for_each_port(card_ctx, port) {
+		struct snd_intelhad *ctx = &card_ctx->pcm_ctx[port];
+		struct snd_pcm_substream *substream;
+
+		substream = had_substream_get(ctx);
+		if (substream) {
+			snd_pcm_suspend(substream);
+			had_substream_put(ctx);
+		}
 	}
 
 	return 0;
@@ -1637,12 +1686,12 @@ static int hdmi_lpe_audio_runtime_suspend(struct device *dev)
 
 static int __maybe_unused hdmi_lpe_audio_suspend(struct device *dev)
 {
-	struct snd_intelhad *ctx = dev_get_drvdata(dev);
+	struct snd_intelhad_card *card_ctx = dev_get_drvdata(dev);
 	int err;
 
 	err = hdmi_lpe_audio_runtime_suspend(dev);
 	if (!err)
-		snd_power_change_state(ctx->card, SNDRV_CTL_POWER_D3hot);
+		snd_power_change_state(card_ctx->card, SNDRV_CTL_POWER_D3hot);
 	return err;
 }
 
@@ -1654,24 +1703,34 @@ static int hdmi_lpe_audio_runtime_resume(struct device *dev)
 
 static int __maybe_unused hdmi_lpe_audio_resume(struct device *dev)
 {
-	struct snd_intelhad *ctx = dev_get_drvdata(dev);
+	struct snd_intelhad_card *card_ctx = dev_get_drvdata(dev);
 
 	hdmi_lpe_audio_runtime_resume(dev);
-	snd_power_change_state(ctx->card, SNDRV_CTL_POWER_D0);
+	snd_power_change_state(card_ctx->card, SNDRV_CTL_POWER_D0);
 	return 0;
 }
 
 /* release resources */
 static void hdmi_lpe_audio_free(struct snd_card *card)
 {
-	struct snd_intelhad *ctx = card->private_data;
+	struct snd_intelhad_card *card_ctx = card->private_data;
+	struct intel_hdmi_lpe_audio_pdata *pdata = card_ctx->dev->platform_data;
+	int port;
+
+	spin_lock_irq(&pdata->lpe_audio_slock);
+	pdata->notify_audio_lpe = NULL;
+	spin_unlock_irq(&pdata->lpe_audio_slock);
 
-	cancel_work_sync(&ctx->hdmi_audio_wq);
+	for_each_port(card_ctx, port) {
+		struct snd_intelhad *ctx = &card_ctx->pcm_ctx[port];
 
-	if (ctx->mmio_start)
-		iounmap(ctx->mmio_start);
-	if (ctx->irq >= 0)
-		free_irq(ctx->irq, ctx);
+		cancel_work_sync(&ctx->hdmi_audio_wq);
+	}
+
+	if (card_ctx->mmio_start)
+		iounmap(card_ctx->mmio_start);
+	if (card_ctx->irq >= 0)
+		free_irq(card_ctx->irq, card_ctx);
 }
 
 /*
@@ -1683,12 +1742,12 @@ static void hdmi_lpe_audio_free(struct snd_card *card)
 static int hdmi_lpe_audio_probe(struct platform_device *pdev)
 {
 	struct snd_card *card;
-	struct snd_intelhad *ctx;
+	struct snd_intelhad_card *card_ctx;
 	struct snd_pcm *pcm;
 	struct intel_hdmi_lpe_audio_pdata *pdata;
 	int irq;
 	struct resource *res_mmio;
-	int i, ret;
+	int port, ret;
 
 	pdata = pdev->dev.platform_data;
 	if (!pdata) {
@@ -1711,39 +1770,30 @@ static int hdmi_lpe_audio_probe(struct platform_device *pdev)
 
 	/* create a card instance with ALSA framework */
 	ret = snd_card_new(&pdev->dev, hdmi_card_index, hdmi_card_id,
-			   THIS_MODULE, sizeof(*ctx), &card);
+			   THIS_MODULE, sizeof(*card_ctx), &card);
 	if (ret)
 		return ret;
 
-	ctx = card->private_data;
-	spin_lock_init(&ctx->had_spinlock);
-	mutex_init(&ctx->mutex);
-	ctx->connected = false;
-	ctx->dev = &pdev->dev;
-	ctx->card = card;
-	ctx->aes_bits = SNDRV_PCM_DEFAULT_CON_SPDIF;
+	card_ctx = card->private_data;
+	card_ctx->dev = &pdev->dev;
+	card_ctx->card = card;
 	strcpy(card->driver, INTEL_HAD);
 	strcpy(card->shortname, "Intel HDMI/DP LPE Audio");
 	strcpy(card->longname, "Intel HDMI/DP LPE Audio");
 
-	ctx->irq = -1;
-	ctx->tmds_clock_speed = DIS_SAMPLE_RATE_148_5;
-	INIT_WORK(&ctx->hdmi_audio_wq, had_audio_wq);
+	card_ctx->irq = -1;
 
 	card->private_free = hdmi_lpe_audio_free;
 
-	/* assume pipe A as default */
-	ctx->had_config_offset = AUDIO_HDMI_CONFIG_A;
-
-	platform_set_drvdata(pdev, ctx);
+	platform_set_drvdata(pdev, card_ctx);
 
 	dev_dbg(&pdev->dev, "%s: mmio_start = 0x%x, mmio_end = 0x%x\n",
 		__func__, (unsigned int)res_mmio->start,
 		(unsigned int)res_mmio->end);
 
-	ctx->mmio_start = ioremap_nocache(res_mmio->start,
-					  (size_t)(resource_size(res_mmio)));
-	if (!ctx->mmio_start) {
+	card_ctx->mmio_start = ioremap_nocache(res_mmio->start,
+					       (size_t)(resource_size(res_mmio)));
+	if (!card_ctx->mmio_start) {
 		dev_err(&pdev->dev, "Could not get ioremap\n");
 		ret = -EACCES;
 		goto err;
@@ -1751,54 +1801,79 @@ static int hdmi_lpe_audio_probe(struct platform_device *pdev)
 
 	/* setup interrupt handler */
 	ret = request_irq(irq, display_pipe_interrupt_handler, 0,
-			  pdev->name, ctx);
+			  pdev->name, card_ctx);
 	if (ret < 0) {
 		dev_err(&pdev->dev, "request_irq failed\n");
 		goto err;
 	}
 
-	ctx->irq = irq;
-
-	ret = snd_pcm_new(card, INTEL_HAD, PCM_INDEX, MAX_PB_STREAMS,
-			  MAX_CAP_STREAMS, &pcm);
-	if (ret)
-		goto err;
-
-	/* setup private data which can be retrieved when required */
-	pcm->private_data = ctx;
-	pcm->info_flags = 0;
-	strncpy(pcm->name, card->shortname, strlen(card->shortname));
-	/* setup the ops for playabck */
-	snd_pcm_set_ops(pcm, SNDRV_PCM_STREAM_PLAYBACK, &had_pcm_ops);
+	card_ctx->irq = irq;
 
 	/* only 32bit addressable */
 	dma_set_mask(&pdev->dev, DMA_BIT_MASK(32));
 	dma_set_coherent_mask(&pdev->dev, DMA_BIT_MASK(32));
 
-	/* allocate dma pages;
-	 * try to allocate 600k buffer as default which is large enough
-	 */
-	snd_pcm_lib_preallocate_pages_for_all(pcm,
-			SNDRV_DMA_TYPE_DEV, NULL,
-			HAD_DEFAULT_BUFFER, HAD_MAX_BUFFER);
+	init_channel_allocations();
 
-	/* create controls */
-	for (i = 0; i < ARRAY_SIZE(had_controls); i++) {
-		ret = snd_ctl_add(card, snd_ctl_new1(&had_controls[i], ctx));
-		if (ret < 0)
+	card_ctx->num_pipes = pdata->num_pipes;
+	card_ctx->num_ports = pdata->num_ports;
+
+	for_each_port(card_ctx, port) {
+		struct snd_intelhad *ctx = &card_ctx->pcm_ctx[port];
+		int i;
+
+		ctx->card_ctx = card_ctx;
+		ctx->dev = card_ctx->dev;
+		ctx->port = port;
+		ctx->pipe = -1;
+
+		INIT_WORK(&ctx->hdmi_audio_wq, had_audio_wq);
+
+		ret = snd_pcm_new(card, INTEL_HAD, port, MAX_PB_STREAMS,
+				  MAX_CAP_STREAMS, &pcm);
+		if (ret)
 			goto err;
-	}
 
-	init_channel_allocations();
+		/* setup private data which can be retrieved when required */
+		pcm->private_data = ctx;
+		pcm->info_flags = 0;
+		strncpy(pcm->name, card->shortname, strlen(card->shortname));
+		/* setup the ops for playabck */
+		snd_pcm_set_ops(pcm, SNDRV_PCM_STREAM_PLAYBACK, &had_pcm_ops);
 
-	/* Register channel map controls */
-	ret = had_register_chmap_ctls(ctx, pcm);
-	if (ret < 0)
-		goto err;
+		/* allocate dma pages;
+		 * try to allocate 600k buffer as default which is large enough
+		 */
+		snd_pcm_lib_preallocate_pages_for_all(pcm,
+						      SNDRV_DMA_TYPE_DEV, NULL,
+						      HAD_DEFAULT_BUFFER, HAD_MAX_BUFFER);
+
+		/* create controls */
+		for (i = 0; i < ARRAY_SIZE(had_controls); i++) {
+			struct snd_kcontrol *kctl;
+
+			kctl = snd_ctl_new1(&had_controls[i], ctx);
+			if (!kctl) {
+				ret = -ENOMEM;
+				goto err;
+			}
 
-	ret = had_create_jack(ctx);
-	if (ret < 0)
-		goto err;
+			kctl->id.device = pcm->device;
+
+			ret = snd_ctl_add(card, kctl);
+			if (ret < 0)
+				goto err;
+		}
+
+		/* Register channel map controls */
+		ret = had_register_chmap_ctls(ctx, pcm);
+		if (ret < 0)
+			goto err;
+
+		ret = had_create_jack(ctx, pcm);
+		if (ret < 0)
+			goto err;
+	}
 
 	ret = snd_card_register(card);
 	if (ret)
@@ -1806,7 +1881,6 @@ static int hdmi_lpe_audio_probe(struct platform_device *pdev)
 
 	spin_lock_irq(&pdata->lpe_audio_slock);
 	pdata->notify_audio_lpe = notify_audio_lpe;
-	pdata->notify_pending = false;
 	spin_unlock_irq(&pdata->lpe_audio_slock);
 
 	pm_runtime_use_autosuspend(&pdev->dev);
@@ -1814,7 +1888,11 @@ static int hdmi_lpe_audio_probe(struct platform_device *pdev)
 	pm_runtime_set_active(&pdev->dev);
 
 	dev_dbg(&pdev->dev, "%s: handle pending notification\n", __func__);
-	schedule_work(&ctx->hdmi_audio_wq);
+	for_each_port(card_ctx, port) {
+		struct snd_intelhad *ctx = &card_ctx->pcm_ctx[port];
+
+		schedule_work(&ctx->hdmi_audio_wq);
+	}
 
 	return 0;
 
@@ -1830,9 +1908,9 @@ err:
  */
 static int hdmi_lpe_audio_remove(struct platform_device *pdev)
 {
-	struct snd_intelhad *ctx = platform_get_drvdata(pdev);
+	struct snd_intelhad_card *card_ctx = platform_get_drvdata(pdev);
 
-	snd_card_free(ctx->card);
+	snd_card_free(card_ctx->card);
 	return 0;
 }
 
diff --git a/sound/x86/intel_hdmi_audio.h b/sound/x86/intel_hdmi_audio.h
index 2d3e389f76b3..0d91bb5dbab7 100644
--- a/sound/x86/intel_hdmi_audio.h
+++ b/sound/x86/intel_hdmi_audio.h
@@ -32,7 +32,6 @@
 
 #include "intel_hdmi_lpe_audio.h"
 
-#define PCM_INDEX		0
 #define MAX_PB_STREAMS		1
 #define MAX_CAP_STREAMS		0
 #define BYTES_PER_WORD		0x4
@@ -101,7 +100,7 @@ struct pcm_stream_info {
  * @chmap: holds channel map info
  */
 struct snd_intelhad {
-	struct snd_card	*card;
+	struct snd_intelhad_card *card_ctx;
 	bool		connected;
 	struct		pcm_stream_info stream_info;
 	unsigned char	eld[HDMI_MAX_ELD_BYTES];
@@ -112,6 +111,8 @@ struct snd_intelhad {
 	struct snd_pcm_chmap *chmap;
 	int tmds_clock_speed;
 	int link_rate;
+	int port; /* fixed */
+	int pipe; /* can change dynamically */
 
 	/* ring buffer (BD) position index */
 	unsigned int bd_head;
@@ -123,9 +124,6 @@ struct snd_intelhad {
 	unsigned int period_bytes;	/* PCM period size in bytes */
 
 	/* internal stuff */
-	int irq;
-	void __iomem *mmio_start;
-	unsigned int had_config_offset;
 	union aud_cfg aud_config;	/* AUD_CONFIG reg value cache */
 	struct work_struct hdmi_audio_wq;
 	struct mutex mutex; /* for protecting chmap and eld */
@@ -133,4 +131,16 @@ struct snd_intelhad {
 	struct snd_jack *jack;
 };
 
+struct snd_intelhad_card {
+	struct snd_card	*card;
+	struct device *dev;
+
+	/* internal stuff */
+	int irq;
+	void __iomem *mmio_start;
+	int num_pipes;
+	int num_ports;
+	struct snd_intelhad pcm_ctx[3]; /* one for each port */
+};
+
 #endif /* _INTEL_HDMI_AUDIO_ */
author	Maarten Lankhorst <maarten.lankhorst@linux.intel.com>	2017-06-12 16:26:43 +0200
committer	Maarten Lankhorst <maarten.lankhorst@linux.intel.com>	2017-06-12 16:26:43 +0200
commit	6cb59a79097546ee0c1880a1d1cf4d6b852005f7 (patch)
tree	1c702e567e44d34b05442587f9654b82333d2287
parent	6e88007e224ce51969ccf9afeec645146c638816 (diff)
parent	7d025e0804f2cbeafeba2f5b8effa6361d7db8e4 (diff)