diff options
Diffstat (limited to 'src/intel/isl/isl_format.c')
-rw-r--r-- | src/intel/isl/isl_format.c | 867 |
1 files changed, 479 insertions, 388 deletions
diff --git a/src/intel/isl/isl_format.c b/src/intel/isl/isl_format.c index 6b848c2c53c..59f6c551e3b 100644 --- a/src/intel/isl/isl_format.c +++ b/src/intel/isl/isl_format.c @@ -25,40 +25,44 @@ #include "isl.h" #include "isl_priv.h" -#include "dev/gen_device_info.h" +#include "dev/intel_device_info.h" -#include "main/macros.h" /* Needed for MAX3 and MAX2 for format_rgb9e5 */ -#include "util/format_srgb.h" -#include "util/format_rgb9e5.h" -#include "util/format_r11g11b10f.h" +#include "util/macros.h" /* Needed for MAX3 and MAX2 for format_rgb9e5 */ -/* Header-only format conversion include */ -#include "main/format_utils.h" +#include "util/format/format_utils.h" +#include "util/format_r11g11b10f.h" +#include "util/format_rgb9e5.h" +#include "util/format_srgb.h" +#include "util/half_float.h" +#include "util/rounding.h" +#include "util/u_math.h" struct surface_format_info { bool exists; - uint8_t sampling; - uint8_t filtering; - uint8_t shadow_compare; - uint8_t chroma_key; - uint8_t render_target; - uint8_t alpha_blend; - uint8_t input_vb; - uint8_t streamed_output_vb; - uint8_t color_processing; - uint8_t typed_write; - uint8_t typed_read; - uint8_t ccs_e; + /* These fields must fit the largest verx10 value. */ + uint16_t sampling; + uint16_t filtering; + uint16_t shadow_compare; + uint16_t chroma_key; + uint16_t render_target; + uint16_t alpha_blend; + uint16_t input_vb; + uint16_t streamed_output_vb; + uint16_t color_processing; + uint16_t typed_write; + uint16_t typed_read; + uint16_t typed_atomics; + uint16_t ccs_e; }; /* This macro allows us to write the table almost as it appears in the PRM, * while restructuring it to turn it into the C code we want. */ -#define SF(sampl, filt, shad, ck, rt, ab, vb, so, color, tw, tr, ccs_e, sf) \ - [ISL_FORMAT_##sf] = { true, sampl, filt, shad, ck, rt, ab, vb, so, color, tw, tr, ccs_e}, +#define SF(sampl, filt, shad, ck, rt, ab, vb, so, color, tw, tr, ccs_e, ta, sf) \ + [ISL_FORMAT_##sf] = { true, sampl, filt, shad, ck, rt, ab, vb, so, color, tw, tr, ta, ccs_e}, #define Y 0 -#define x 255 +#define x 0xFFFF /** * This is the table of support for surface (texture, renderbuffer, and vertex * buffer, but not depthbuffer) formats across the various hardware generations. @@ -69,9 +73,9 @@ struct surface_format_info { * * Y*: 45 * Y+: 45 (g45/gm45) - * Y~: 50 (gen5) - * Y^: 60 (gen6) - * Y#: 70 (gen7) + * Y~: 50 (gfx5) + * Y^: 60 (gfx6) + * Y#: 70 (gfx7) * * The abbreviations in the header below are: * smpl - Sampling Engine @@ -83,8 +87,11 @@ struct surface_format_info { * VB - Input Vertex Buffer * SO - Steamed Output Vertex Buffers (transform feedback) * color - Color Processing - * ccs_e - Lossless Compression Support (gen9+ only) + * TW - Typed Write + * TR - Typed Read + * ccs_e - Lossless Compression Support (gfx9+ only) * sf - Surface Format + * TA - Typed Atomics * * See page 88 of the Sandybridge PRM VOL4_Part1 PDF. * @@ -97,278 +104,278 @@ struct surface_format_info { * - Render Target Surface Types [SKL+] */ static const struct surface_format_info format_info[] = { -/* smpl filt shad CK RT AB VB SO color TW TR ccs_e */ - SF( Y, 50, x, x, Y, Y, Y, Y, x, 70, 90, 90, R32G32B32A32_FLOAT) - SF( Y, x, x, x, Y, x, Y, Y, x, 70, 90, 90, R32G32B32A32_SINT) - SF( Y, x, x, x, Y, x, Y, Y, x, 70, 90, 90, R32G32B32A32_UINT) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R32G32B32A32_UNORM) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R32G32B32A32_SNORM) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R64G64_FLOAT) - SF( Y, 50, x, x, 100, 100, x, x, x, x, x, 100, R32G32B32X32_FLOAT) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R32G32B32A32_SSCALED) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R32G32B32A32_USCALED) - SF( x, x, x, x, x, x, 75, x, x, x, x, x, R32G32B32A32_SFIXED) - SF( x, x, x, x, x, x, 80, x, x, x, x, x, R64G64_PASSTHRU) - SF( Y, 50, x, x, x, x, Y, Y, x, x, x, x, R32G32B32_FLOAT) - SF( Y, x, x, x, x, x, Y, Y, x, x, x, x, R32G32B32_SINT) - SF( Y, x, x, x, x, x, Y, Y, x, x, x, x, R32G32B32_UINT) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R32G32B32_UNORM) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R32G32B32_SNORM) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R32G32B32_SSCALED) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R32G32B32_USCALED) - SF( x, x, x, x, x, x, 75, x, x, x, x, x, R32G32B32_SFIXED) - SF( Y, Y, x, x, Y, 45, Y, x, 60, 70, 110, 90, R16G16B16A16_UNORM) - SF( Y, Y, x, x, Y, 60, Y, x, x, 70, 110, 90, R16G16B16A16_SNORM) - SF( Y, x, x, x, Y, x, Y, x, x, 70, 90, 90, R16G16B16A16_SINT) - SF( Y, x, x, x, Y, x, Y, x, x, 70, 75, 90, R16G16B16A16_UINT) - SF( Y, Y, x, x, Y, Y, Y, x, x, 70, 90, 90, R16G16B16A16_FLOAT) - SF( Y, 50, x, x, Y, Y, Y, Y, x, 70, 90, 90, R32G32_FLOAT) - SF( Y, 70, x, x, Y, Y, Y, Y, x, x, x, x, R32G32_FLOAT_LD) - SF( Y, x, x, x, Y, x, Y, Y, x, 70, 90, 90, R32G32_SINT) - SF( Y, x, x, x, Y, x, Y, Y, x, 70, 90, 90, R32G32_UINT) - SF( Y, 50, Y, x, x, x, x, x, x, x, x, x, R32_FLOAT_X8X24_TYPELESS) - SF( Y, x, x, x, x, x, x, x, x, x, x, x, X32_TYPELESS_G8X24_UINT) - SF( Y, 50, x, x, x, x, x, x, x, x, x, x, L32A32_FLOAT) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R32G32_UNORM) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R32G32_SNORM) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R64_FLOAT) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, R16G16B16X16_UNORM) - SF( Y, Y, x, x, 90, 90, x, x, x, x, x, 90, R16G16B16X16_FLOAT) - SF( Y, 50, x, x, x, x, x, x, x, x, x, x, A32X32_FLOAT) - SF( Y, 50, x, x, x, x, x, x, x, x, x, x, L32X32_FLOAT) - SF( Y, 50, x, x, x, x, x, x, x, x, x, x, I32X32_FLOAT) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R16G16B16A16_SSCALED) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R16G16B16A16_USCALED) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R32G32_SSCALED) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R32G32_USCALED) - SF( x, x, x, x, x, x, 75, x, x, x, x, x, R32G32_SFIXED) - SF( x, x, x, x, x, x, 80, x, x, x, x, x, R64_PASSTHRU) - SF( Y, Y, x, Y, Y, Y, Y, x, 60, 70, x, 90, B8G8R8A8_UNORM) - SF( Y, Y, x, x, Y, Y, x, x, x, x, x, 100, B8G8R8A8_UNORM_SRGB) -/* smpl filt shad CK RT AB VB SO color TW TR ccs_e */ - SF( Y, Y, x, x, Y, Y, Y, x, 60, 70, x, 100, R10G10B10A2_UNORM) - SF( Y, Y, x, x, x, x, x, x, 60, x, x, 120, R10G10B10A2_UNORM_SRGB) - SF( Y, x, x, x, Y, x, Y, x, x, 70, x, 100, R10G10B10A2_UINT) - SF( Y, Y, x, x, x, x, Y, x, x, x, x, x, R10G10B10_SNORM_A2_UNORM) - SF( Y, Y, x, x, Y, Y, Y, x, 60, 70, 110, 90, R8G8B8A8_UNORM) - SF( Y, Y, x, x, Y, Y, x, x, 60, x, x, 100, R8G8B8A8_UNORM_SRGB) - SF( Y, Y, x, x, Y, 60, Y, x, x, 70, 110, 90, R8G8B8A8_SNORM) - SF( Y, x, x, x, Y, x, Y, x, x, 70, 90, 90, R8G8B8A8_SINT) - SF( Y, x, x, x, Y, x, Y, x, x, 70, 75, 90, R8G8B8A8_UINT) - SF( Y, Y, x, x, Y, 45, Y, x, x, 70, 110, 90, R16G16_UNORM) - SF( Y, Y, x, x, Y, 60, Y, x, x, 70, 110, 90, R16G16_SNORM) - SF( Y, x, x, x, Y, x, Y, x, x, 70, 90, 90, R16G16_SINT) - SF( Y, x, x, x, Y, x, Y, x, x, 70, 75, 90, R16G16_UINT) - SF( Y, Y, x, x, Y, Y, Y, x, x, 70, 90, 90, R16G16_FLOAT) - SF( Y, Y, x, x, Y, Y, 75, x, 60, 70, x, 100, B10G10R10A2_UNORM) - SF( Y, Y, x, x, Y, Y, x, x, 60, x, x, 100, B10G10R10A2_UNORM_SRGB) - SF( Y, Y, x, x, Y, Y, Y, x, x, 70, x, 100, R11G11B10_FLOAT) - SF(120, 120, x, x, 120, 120, x, x, x, x, x, 120, R10G10B10_FLOAT_A2_UNORM) - SF( Y, x, x, x, Y, x, Y, Y, x, 70, 70, 90, R32_SINT) - SF( Y, x, x, x, Y, x, Y, Y, x, 70, 70, 90, R32_UINT) - SF( Y, 50, Y, x, Y, Y, Y, Y, x, 70, 70, 90, R32_FLOAT) - SF( Y, 50, Y, x, x, x, x, x, x, x, x, 120, R24_UNORM_X8_TYPELESS) - SF( Y, x, x, x, x, x, x, x, x, x, x, x, X24_TYPELESS_G8_UINT) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, L16A16_UNORM) - SF( Y, 50, Y, x, x, x, x, x, x, x, x, x, I24X8_UNORM) - SF( Y, 50, Y, x, x, x, x, x, x, x, x, x, L24X8_UNORM) - SF( Y, 50, Y, x, x, x, x, x, x, x, x, x, A24X8_UNORM) - SF( Y, 50, Y, x, x, x, x, x, x, x, x, x, I32_FLOAT) - SF( Y, 50, Y, x, x, x, x, x, x, x, x, x, L32_FLOAT) - SF( Y, 50, Y, x, x, x, x, x, x, x, x, x, A32_FLOAT) - SF( Y, Y, x, Y, 80, 80, x, x, 60, x, x, 90, B8G8R8X8_UNORM) - SF( Y, Y, x, x, 80, 80, x, x, x, x, x, 100, B8G8R8X8_UNORM_SRGB) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, R8G8B8X8_UNORM) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, R8G8B8X8_UNORM_SRGB) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, R9G9B9E5_SHAREDEXP) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, B10G10R10X2_UNORM) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, L16A16_FLOAT) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R32_UNORM) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R32_SNORM) -/* smpl filt shad CK RT AB VB SO color TW TR ccs_e */ - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R10G10B10X2_USCALED) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R8G8B8A8_SSCALED) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R8G8B8A8_USCALED) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R16G16_SSCALED) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R16G16_USCALED) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R32_SSCALED) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R32_USCALED) - SF( Y, Y, x, Y, Y, Y, x, x, x, 70, x, 120, B5G6R5_UNORM) - SF( Y, Y, x, x, Y, Y, x, x, x, x, x, 120, B5G6R5_UNORM_SRGB) - SF( Y, Y, x, Y, Y, Y, x, x, x, 70, x, 120, B5G5R5A1_UNORM) - SF( Y, Y, x, x, Y, Y, x, x, x, x, x, 120, B5G5R5A1_UNORM_SRGB) - SF( Y, Y, x, Y, Y, Y, x, x, x, 70, x, 120, B4G4R4A4_UNORM) - SF( Y, Y, x, x, Y, Y, x, x, x, x, x, 120, B4G4R4A4_UNORM_SRGB) - SF( Y, Y, x, x, Y, Y, Y, x, x, 70, 110, 120, R8G8_UNORM) - SF( Y, Y, x, Y, Y, 60, Y, x, x, 70, 110, 120, R8G8_SNORM) - SF( Y, x, x, x, Y, x, Y, x, x, 70, 90, 120, R8G8_SINT) - SF( Y, x, x, x, Y, x, Y, x, x, 70, 75, 120, R8G8_UINT) - SF( Y, Y, Y, x, Y, 45, Y, x, 70, 70, 110, 120, R16_UNORM) - SF( Y, Y, x, x, Y, 60, Y, x, x, 70, 110, 120, R16_SNORM) - SF( Y, x, x, x, Y, x, Y, x, x, 70, 90, 120, R16_SINT) - SF( Y, x, x, x, Y, x, Y, x, x, 70, 75, 120, R16_UINT) - SF( Y, Y, x, x, Y, Y, Y, x, x, 70, 90, 120, R16_FLOAT) - SF( 50, 50, x, x, x, x, x, x, x, x, x, x, A8P8_UNORM_PALETTE0) - SF( 50, 50, x, x, x, x, x, x, x, x, x, x, A8P8_UNORM_PALETTE1) - SF( Y, Y, Y, x, x, x, x, x, x, x, x, x, I16_UNORM) - SF( Y, Y, Y, x, x, x, x, x, x, x, x, x, L16_UNORM) - SF( Y, Y, Y, x, x, x, x, x, x, x, x, x, A16_UNORM) - SF( Y, Y, x, Y, x, x, x, x, x, x, x, x, L8A8_UNORM) - SF( Y, Y, Y, x, x, x, x, x, x, x, x, x, I16_FLOAT) - SF( Y, Y, Y, x, x, x, x, x, x, x, x, x, L16_FLOAT) - SF( Y, Y, Y, x, x, x, x, x, x, x, x, x, A16_FLOAT) - SF( 45, 45, x, x, x, x, x, x, x, x, x, x, L8A8_UNORM_SRGB) - SF( Y, Y, x, Y, x, x, x, x, x, x, x, x, R5G5_SNORM_B6_UNORM) - SF( x, x, x, x, Y, Y, x, x, x, 70, x, 120, B5G5R5X1_UNORM) - SF( x, x, x, x, Y, Y, x, x, x, x, x, 120, B5G5R5X1_UNORM_SRGB) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R8G8_SSCALED) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R8G8_USCALED) -/* smpl filt shad CK RT AB VB SO color TW TR ccs_e */ - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R16_SSCALED) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R16_USCALED) - SF( 50, 50, x, x, x, x, x, x, x, x, x, x, P8A8_UNORM_PALETTE0) - SF( 50, 50, x, x, x, x, x, x, x, x, x, x, P8A8_UNORM_PALETTE1) - SF(120, 120, x, x, 120, 120, x, x, x, x, x, 120, A1B5G5R5_UNORM) +/* smpl filt shad CK RT AB VB SO color TW TR ccs_e TA */ + SF( Y, 50, x, x, Y, Y, Y, Y, x, 70, 90, 90, x, R32G32B32A32_FLOAT) + SF( Y, x, x, x, Y, x, Y, Y, x, 70, 90, 90, x, R32G32B32A32_SINT) + SF( Y, x, x, x, Y, x, Y, Y, x, 70, 90, 90, x, R32G32B32A32_UINT) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R32G32B32A32_UNORM) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R32G32B32A32_SNORM) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R64G64_FLOAT) + SF( Y, 50, x, x, 110, 110, x, x, x, x, x, 110, x, R32G32B32X32_FLOAT) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R32G32B32A32_SSCALED) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R32G32B32A32_USCALED) + SF( x, x, x, x, x, x, 75, x, x, x, x, x, x, R32G32B32A32_SFIXED) + SF( x, x, x, x, x, x, 80, x, x, x, x, x, x, R64G64_PASSTHRU) + SF( Y, 50, x, x, x, x, Y, Y, x, x, x, x, x, R32G32B32_FLOAT) + SF( Y, x, x, x, x, x, Y, Y, x, x, x, x, x, R32G32B32_SINT) + SF( Y, x, x, x, x, x, Y, Y, x, x, x, x, x, R32G32B32_UINT) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R32G32B32_UNORM) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R32G32B32_SNORM) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R32G32B32_SSCALED) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R32G32B32_USCALED) + SF( x, x, x, x, x, x, 75, x, x, x, x, x, x, R32G32B32_SFIXED) + SF( Y, Y, x, x, Y, 45, Y, x, 60, 70, 110, 90, x, R16G16B16A16_UNORM) + SF( Y, Y, x, x, Y, 60, Y, x, x, 70, 110, 90, x, R16G16B16A16_SNORM) + SF( Y, x, x, x, Y, x, Y, x, x, 70, 90, 90, x, R16G16B16A16_SINT) + SF( Y, x, x, x, Y, x, Y, x, x, 70, 75, 90, x, R16G16B16A16_UINT) + SF( Y, Y, x, x, Y, Y, Y, x, x, 70, 90, 90, x, R16G16B16A16_FLOAT) + SF( Y, 50, x, x, Y, Y, Y, Y, x, 70, 90, 90, x, R32G32_FLOAT) + SF( Y, 70, x, x, Y, Y, Y, Y, x, x, x, x, x, R32G32_FLOAT_LD) + SF( Y, x, x, x, Y, x, Y, Y, x, 70, 90, 90, x, R32G32_SINT) + SF( Y, x, x, x, Y, x, Y, Y, x, 70, 90, 90, x, R32G32_UINT) + SF( Y, 50, Y, x, x, x, x, x, x, x, x, x, x, R32_FLOAT_X8X24_TYPELESS) + SF( Y, x, x, x, x, x, x, x, x, x, x, x, x, X32_TYPELESS_G8X24_UINT) + SF( Y, 50, x, x, x, x, x, x, x, x, x, x, x, L32A32_FLOAT) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R32G32_UNORM) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R32G32_SNORM) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, 90, R64_FLOAT) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, R16G16B16X16_UNORM) + SF( Y, Y, x, x, 90, 90, x, x, x, x, x, 90, x, R16G16B16X16_FLOAT) + SF( Y, 50, x, x, x, x, x, x, x, x, x, x, x, A32X32_FLOAT) + SF( Y, 50, x, x, x, x, x, x, x, x, x, x, x, L32X32_FLOAT) + SF( Y, 50, x, x, x, x, x, x, x, x, x, x, x, I32X32_FLOAT) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R16G16B16A16_SSCALED) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R16G16B16A16_USCALED) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R32G32_SSCALED) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R32G32_USCALED) + SF( x, x, x, x, x, x, 75, x, x, x, x, x, x, R32G32_SFIXED) + SF( x, x, x, x, x, x, 80, x, x, x, x, x, 90, R64_PASSTHRU) + SF( Y, Y, x, Y, Y, Y, Y, x, 60, 70, 125, 90, x, B8G8R8A8_UNORM) + SF( Y, Y, x, x, Y, Y, x, x, x, x, x, 110, x, B8G8R8A8_UNORM_SRGB) +/* smpl filt shad CK RT AB VB SO color TW TR ccs_e TA */ + SF( Y, Y, x, x, Y, Y, Y, x, 60, 70, 125, 110, x, R10G10B10A2_UNORM) + SF( Y, Y, x, x, x, x, x, x, 60, x, x, 120, x, R10G10B10A2_UNORM_SRGB) + SF( Y, x, x, x, Y, x, Y, x, x, 70, 125, 110, x, R10G10B10A2_UINT) + SF( Y, Y, x, x, x, x, Y, x, x, x, x, x, x, R10G10B10_SNORM_A2_UNORM) + SF( Y, Y, x, x, Y, Y, Y, x, 60, 70, 110, 90, x, R8G8B8A8_UNORM) + SF( Y, Y, x, x, Y, Y, x, x, 60, x, x, 110, x, R8G8B8A8_UNORM_SRGB) + SF( Y, Y, x, x, Y, 60, Y, x, x, 70, 110, 90, x, R8G8B8A8_SNORM) + SF( Y, x, x, x, Y, x, Y, x, x, 70, 90, 90, x, R8G8B8A8_SINT) + SF( Y, x, x, x, Y, x, Y, x, x, 70, 75, 90, x, R8G8B8A8_UINT) + SF( Y, Y, x, x, Y, 45, Y, x, x, 70, 110, 90, x, R16G16_UNORM) + SF( Y, Y, x, x, Y, 60, Y, x, x, 70, 110, 90, x, R16G16_SNORM) + SF( Y, x, x, x, Y, x, Y, x, x, 70, 90, 90, x, R16G16_SINT) + SF( Y, x, x, x, Y, x, Y, x, x, 70, 75, 90, x, R16G16_UINT) + SF( Y, Y, x, x, Y, Y, Y, x, x, 70, 90, 90, x, R16G16_FLOAT) + SF( Y, Y, x, x, Y, Y, 75, x, 60, 70, 125, 110, x, B10G10R10A2_UNORM) + SF( Y, Y, x, x, Y, Y, x, x, 60, x, x, 110, x, B10G10R10A2_UNORM_SRGB) + SF( Y, Y, x, x, Y, Y, Y, x, x, 70, 125, 110, x, R11G11B10_FLOAT) + SF(120, 120, x, x, 120, 120, x, x, x, 125, 125, 120, x, R10G10B10_FLOAT_A2_UNORM) + SF( Y, x, x, x, Y, x, Y, Y, x, 70, 70, 90, 70, R32_SINT) + SF( Y, x, x, x, Y, x, Y, Y, x, 70, 70, 90, 70, R32_UINT) + SF( Y, 50, Y, x, Y, Y, Y, Y, x, 70, 70, 90, 110, R32_FLOAT) + SF( Y, 50, Y, x, x, x, x, x, x, x, x, 120, x, R24_UNORM_X8_TYPELESS) + SF( Y, x, x, x, x, x, x, x, x, x, x, x, x, X24_TYPELESS_G8_UINT) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, L16A16_UNORM) + SF( Y, 50, Y, x, x, x, x, x, x, x, x, x, x, I24X8_UNORM) + SF( Y, 50, Y, x, x, x, x, x, x, x, x, x, x, L24X8_UNORM) + SF( Y, 50, Y, x, x, x, x, x, x, x, x, x, x, A24X8_UNORM) + SF( Y, 50, Y, x, x, x, x, x, x, x, x, x, x, I32_FLOAT) + SF( Y, 50, Y, x, x, x, x, x, x, x, x, x, x, L32_FLOAT) + SF( Y, 50, Y, x, x, x, x, x, x, x, x, x, x, A32_FLOAT) + SF( Y, Y, x, Y, 80, 80, x, x, 60, x, x, 90, x, B8G8R8X8_UNORM) + SF( Y, Y, x, x, 80, 80, x, x, x, x, x, 110, x, B8G8R8X8_UNORM_SRGB) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, R8G8B8X8_UNORM) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, R8G8B8X8_UNORM_SRGB) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, R9G9B9E5_SHAREDEXP) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, B10G10R10X2_UNORM) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, L16A16_FLOAT) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R32_UNORM) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R32_SNORM) +/* smpl filt shad CK RT AB VB SO color TW TR ccs_e TA */ + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R10G10B10X2_USCALED) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R8G8B8A8_SSCALED) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R8G8B8A8_USCALED) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R16G16_SSCALED) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R16G16_USCALED) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R32_SSCALED) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R32_USCALED) + SF( Y, Y, x, Y, Y, Y, x, x, x, 70, x, 120, x, B5G6R5_UNORM) + SF( Y, Y, x, x, Y, Y, x, x, x, x, x, 120, x, B5G6R5_UNORM_SRGB) + SF( Y, Y, x, Y, Y, Y, x, x, x, 70, x, 120, x, B5G5R5A1_UNORM) + SF( Y, Y, x, x, Y, Y, x, x, x, x, x, 120, x, B5G5R5A1_UNORM_SRGB) + SF( Y, Y, x, Y, Y, Y, x, x, x, 70, x, 120, x, B4G4R4A4_UNORM) + SF( Y, Y, x, x, Y, Y, x, x, x, x, x, 120, x, B4G4R4A4_UNORM_SRGB) + SF( Y, Y, x, x, Y, Y, Y, x, x, 70, 110, 120, x, R8G8_UNORM) + SF( Y, Y, x, Y, Y, 60, Y, x, x, 70, 110, 120, x, R8G8_SNORM) + SF( Y, x, x, x, Y, x, Y, x, x, 70, 90, 120, x, R8G8_SINT) + SF( Y, x, x, x, Y, x, Y, x, x, 70, 75, 120, x, R8G8_UINT) + SF( Y, Y, Y, x, Y, 45, Y, x, 70, 70, 110, 120, x, R16_UNORM) + SF( Y, Y, x, x, Y, 60, Y, x, x, 70, 110, 120, x, R16_SNORM) + SF( Y, x, x, x, Y, x, Y, x, x, 70, 90, 120, 120, R16_SINT) + SF( Y, x, x, x, Y, x, Y, x, x, 70, 75, 120, 120, R16_UINT) + SF( Y, Y, x, x, Y, Y, Y, x, x, 70, 90, 120, 120, R16_FLOAT) + SF( 50, 50, x, x, x, x, x, x, x, x, x, x, x, A8P8_UNORM_PALETTE0) + SF( 50, 50, x, x, x, x, x, x, x, x, x, x, x, A8P8_UNORM_PALETTE1) + SF( Y, Y, Y, x, x, x, x, x, x, x, x, x, x, I16_UNORM) + SF( Y, Y, Y, x, x, x, x, x, x, x, x, x, x, L16_UNORM) + SF( Y, Y, Y, x, x, x, x, x, x, x, x, x, x, A16_UNORM) + SF( Y, Y, x, Y, x, x, x, x, x, x, x, x, x, L8A8_UNORM) + SF( Y, Y, Y, x, x, x, x, x, x, x, x, x, x, I16_FLOAT) + SF( Y, Y, Y, x, x, x, x, x, x, x, x, x, x, L16_FLOAT) + SF( Y, Y, Y, x, x, x, x, x, x, x, x, x, x, A16_FLOAT) + SF( 45, 45, x, x, x, x, x, x, x, x, x, x, x, L8A8_UNORM_SRGB) + SF( Y, Y, x, Y, x, x, x, x, x, x, x, x, x, R5G5_SNORM_B6_UNORM) + SF( x, x, x, x, Y, Y, x, x, x, 70, x, 120, x, B5G5R5X1_UNORM) + SF( x, x, x, x, Y, Y, x, x, x, x, x, 120, x, B5G5R5X1_UNORM_SRGB) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R8G8_SSCALED) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R8G8_USCALED) +/* smpl filt shad CK RT AB VB SO color TW TR ccs_e TA */ + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R16_SSCALED) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R16_USCALED) + SF( 50, 50, x, x, x, x, x, x, x, x, x, x, x, P8A8_UNORM_PALETTE0) + SF( 50, 50, x, x, x, x, x, x, x, x, x, x, x, P8A8_UNORM_PALETTE1) + SF(120, 120, x, x, 120, 120, x, x, x, x, x, 120, x, A1B5G5R5_UNORM) /* According to the PRM, A4B4G4R4_UNORM isn't supported until Sky Lake * but empirical testing indicates that at least sampling works just fine * on Broadwell. */ - SF( 80, 80, x, x, 90, 120, x, x, x, x, x, 120, A4B4G4R4_UNORM) - SF( 90, x, x, x, x, x, x, x, x, x, x, x, L8A8_UINT) - SF( 90, x, x, x, x, x, x, x, x, x, x, x, L8A8_SINT) - SF( Y, Y, x, 45, Y, Y, Y, x, x, 70, 110, 120, R8_UNORM) - SF( Y, Y, x, x, Y, 60, Y, x, x, 70, 110, 120, R8_SNORM) - SF( Y, x, x, x, Y, x, Y, x, x, 70, 90, 120, R8_SINT) - SF( Y, x, x, x, Y, x, Y, x, x, 70, 75, 120, R8_UINT) - SF( Y, Y, x, Y, Y, Y, x, x, x, 70, 110, 120, A8_UNORM) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, I8_UNORM) - SF( Y, Y, x, Y, x, x, x, x, x, x, x, x, L8_UNORM) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, P4A4_UNORM_PALETTE0) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, A4P4_UNORM_PALETTE0) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R8_SSCALED) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R8_USCALED) - SF( 45, 45, x, x, x, x, x, x, x, x, x, x, P8_UNORM_PALETTE0) - SF( 45, 45, x, x, x, x, x, x, x, x, x, x, L8_UNORM_SRGB) - SF( 45, 45, x, x, x, x, x, x, x, x, x, x, P8_UNORM_PALETTE1) - SF( 45, 45, x, x, x, x, x, x, x, x, x, x, P4A4_UNORM_PALETTE1) - SF( 45, 45, x, x, x, x, x, x, x, x, x, x, A4P4_UNORM_PALETTE1) - SF( x, x, x, x, x, x, x, x, x, x, x, x, Y8_UNORM) - SF( 90, x, x, x, x, x, x, x, x, x, x, x, L8_UINT) - SF( 90, x, x, x, x, x, x, x, x, x, x, x, L8_SINT) - SF( 90, x, x, x, x, x, x, x, x, x, x, x, I8_UINT) - SF( 90, x, x, x, x, x, x, x, x, x, x, x, I8_SINT) - SF( 45, 45, x, x, x, x, x, x, x, x, x, x, DXT1_RGB_SRGB) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, R1_UNORM) - SF( Y, Y, x, Y, Y, x, x, x, 60, x, x, x, YCRCB_NORMAL) - SF( Y, Y, x, Y, Y, x, x, x, 60, x, x, x, YCRCB_SWAPUVY) - SF( 45, 45, x, x, x, x, x, x, x, x, x, x, P2_UNORM_PALETTE0) - SF( 45, 45, x, x, x, x, x, x, x, x, x, x, P2_UNORM_PALETTE1) - SF( Y, Y, x, Y, x, x, x, x, x, x, x, x, BC1_UNORM) - SF( Y, Y, x, Y, x, x, x, x, x, x, x, x, BC2_UNORM) - SF( Y, Y, x, Y, x, x, x, x, x, x, x, x, BC3_UNORM) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, BC4_UNORM) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, BC5_UNORM) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, BC1_UNORM_SRGB) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, BC2_UNORM_SRGB) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, BC3_UNORM_SRGB) - SF( Y, x, x, x, x, x, x, x, x, x, x, x, MONO8) - SF( Y, Y, x, x, Y, x, x, x, 60, x, x, x, YCRCB_SWAPUV) - SF( Y, Y, x, x, Y, x, x, x, 60, x, x, x, YCRCB_SWAPY) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, DXT1_RGB) -/* smpl filt shad CK RT AB VB SO color TW TR ccs_e */ - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, FXT1) - SF( 75, 75, x, x, x, x, Y, x, x, x, x, x, R8G8B8_UNORM) - SF( 75, 75, x, x, x, x, Y, x, x, x, x, x, R8G8B8_SNORM) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R8G8B8_SSCALED) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R8G8B8_USCALED) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R64G64B64A64_FLOAT) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R64G64B64_FLOAT) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, BC4_SNORM) - SF( Y, Y, x, x, x, x, x, x, x, x, x, x, BC5_SNORM) - SF( 50, 50, x, x, x, x, 60, x, x, x, x, x, R16G16B16_FLOAT) - SF( 75, 75, x, x, x, x, Y, x, x, x, x, x, R16G16B16_UNORM) - SF( 75, 75, x, x, x, x, Y, x, x, x, x, x, R16G16B16_SNORM) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R16G16B16_SSCALED) - SF( x, x, x, x, x, x, Y, x, x, x, x, x, R16G16B16_USCALED) - SF( 70, 70, x, x, x, x, x, x, x, x, x, x, BC6H_SF16) - SF( 70, 70, x, x, x, x, x, x, x, x, x, x, BC7_UNORM) - SF( 70, 70, x, x, x, x, x, x, x, x, x, x, BC7_UNORM_SRGB) - SF( 70, 70, x, x, x, x, x, x, x, x, x, x, BC6H_UF16) - SF( x, x, x, x, x, x, x, x, x, x, x, x, PLANAR_420_8) + SF( 80, 80, x, x, 90, 120, x, x, x, x, x, 120, x, A4B4G4R4_UNORM) + SF( 90, x, x, x, x, x, x, x, x, x, x, x, x, L8A8_UINT) + SF( 90, x, x, x, x, x, x, x, x, x, x, x, x, L8A8_SINT) + SF( Y, Y, x, 45, Y, Y, Y, x, x, 70, 110, 120, x, R8_UNORM) + SF( Y, Y, x, x, Y, 60, Y, x, x, 70, 110, 120, x, R8_SNORM) + SF( Y, x, x, x, Y, x, Y, x, x, 70, 90, 120, x, R8_SINT) + SF( Y, x, x, x, Y, x, Y, x, x, 70, 75, 120, x, R8_UINT) + SF( Y, Y, x, Y, Y, Y, x, x, x, 70, 110, 120, x, A8_UNORM) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, I8_UNORM) + SF( Y, Y, x, Y, x, x, x, x, x, x, x, x, x, L8_UNORM) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, P4A4_UNORM_PALETTE0) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, A4P4_UNORM_PALETTE0) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R8_SSCALED) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R8_USCALED) + SF( 45, 45, x, x, x, x, x, x, x, x, x, x, x, P8_UNORM_PALETTE0) + SF( 45, 45, x, x, x, x, x, x, x, x, x, x, x, L8_UNORM_SRGB) + SF( 45, 45, x, x, x, x, x, x, x, x, x, x, x, P8_UNORM_PALETTE1) + SF( 45, 45, x, x, x, x, x, x, x, x, x, x, x, P4A4_UNORM_PALETTE1) + SF( 45, 45, x, x, x, x, x, x, x, x, x, x, x, A4P4_UNORM_PALETTE1) + SF( x, x, x, x, x, x, x, x, x, x, x, x, x, Y8_UNORM) + SF( 90, x, x, x, x, x, x, x, x, x, x, x, x, L8_UINT) + SF( 90, x, x, x, x, x, x, x, x, x, x, x, x, L8_SINT) + SF( 90, x, x, x, x, x, x, x, x, x, x, x, x, I8_UINT) + SF( 90, x, x, x, x, x, x, x, x, x, x, x, x, I8_SINT) + SF( 45, 45, x, x, x, x, x, x, x, x, x, x, x, DXT1_RGB_SRGB) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, R1_UNORM) + SF( Y, Y, x, Y, Y, x, x, x, 60, x, x, 120, x, YCRCB_NORMAL) + SF( Y, Y, x, Y, Y, x, x, x, 60, x, x, x, x, YCRCB_SWAPUVY) + SF( 45, 45, x, x, x, x, x, x, x, x, x, x, x, P2_UNORM_PALETTE0) + SF( 45, 45, x, x, x, x, x, x, x, x, x, x, x, P2_UNORM_PALETTE1) + SF( Y, Y, x, Y, x, x, x, x, x, x, x, x, x, BC1_UNORM) + SF( Y, Y, x, Y, x, x, x, x, x, x, x, x, x, BC2_UNORM) + SF( Y, Y, x, Y, x, x, x, x, x, x, x, x, x, BC3_UNORM) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, BC4_UNORM) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, BC5_UNORM) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, BC1_UNORM_SRGB) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, BC2_UNORM_SRGB) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, BC3_UNORM_SRGB) + SF( Y, x, x, x, x, x, x, x, x, x, x, x, x, MONO8) + SF( Y, Y, x, x, Y, x, x, x, 60, x, x, x, x, YCRCB_SWAPUV) + SF( Y, Y, x, x, Y, x, x, x, 60, x, x, 120, x, YCRCB_SWAPY) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, DXT1_RGB) +/* smpl filt shad CK RT AB VB SO color TW TR ccs_e TA */ + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, FXT1) + SF( 75, 75, x, x, x, x, Y, x, x, x, x, x, x, R8G8B8_UNORM) + SF( 75, 75, x, x, x, x, Y, x, x, x, x, x, x, R8G8B8_SNORM) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R8G8B8_SSCALED) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R8G8B8_USCALED) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R64G64B64A64_FLOAT) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R64G64B64_FLOAT) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, BC4_SNORM) + SF( Y, Y, x, x, x, x, x, x, x, x, x, x, x, BC5_SNORM) + SF( 50, 50, x, x, x, x, 60, x, x, x, x, x, x, R16G16B16_FLOAT) + SF( 75, 75, x, x, x, x, Y, x, x, x, x, x, x, R16G16B16_UNORM) + SF( 75, 75, x, x, x, x, Y, x, x, x, x, x, x, R16G16B16_SNORM) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R16G16B16_SSCALED) + SF( x, x, x, x, x, x, Y, x, x, x, x, x, x, R16G16B16_USCALED) + SF( 70, 70, x, x, x, x, x, x, x, x, x, x, x, BC6H_SF16) + SF( 70, 70, x, x, x, x, x, x, x, x, x, x, x, BC7_UNORM) + SF( 70, 70, x, x, x, x, x, x, x, x, x, x, x, BC7_UNORM_SRGB) + SF( 70, 70, x, x, x, x, x, x, x, x, x, x, x, BC6H_UF16) + SF( x, x, x, x, x, x, x, x, x, x, x, x, x, PLANAR_420_8) /* The format enum for R8G8B8_UNORM_SRGB first shows up in the HSW PRM but * empirical testing indicates that it doesn't actually sRGB decode and - * acts identical to R8G8B8_UNORM. It does work on gen8+. + * acts identical to R8G8B8_UNORM. It does work on gfx8+. */ - SF( 80, 80, x, x, x, x, x, x, x, x, x, x, R8G8B8_UNORM_SRGB) - SF( 80, 80, x, x, x, x, x, x, x, x, x, x, ETC1_RGB8) - SF( 80, 80, x, x, x, x, x, x, x, x, x, x, ETC2_RGB8) - SF( 80, 80, x, x, x, x, x, x, x, x, x, x, EAC_R11) - SF( 80, 80, x, x, x, x, x, x, x, x, x, x, EAC_RG11) - SF( 80, 80, x, x, x, x, x, x, x, x, x, x, EAC_SIGNED_R11) - SF( 80, 80, x, x, x, x, x, x, x, x, x, x, EAC_SIGNED_RG11) - SF( 80, 80, x, x, x, x, x, x, x, x, x, x, ETC2_SRGB8) - SF( 90, x, x, x, x, x, 75, x, x, x, x, x, R16G16B16_UINT) - SF( 90, x, x, x, x, x, 75, x, x, x, x, x, R16G16B16_SINT) - SF( x, x, x, x, x, x, 75, x, x, x, x, x, R32_SFIXED) - SF( x, x, x, x, x, x, 75, x, x, x, x, x, R10G10B10A2_SNORM) - SF( x, x, x, x, x, x, 75, x, x, x, x, x, R10G10B10A2_USCALED) - SF( x, x, x, x, x, x, 75, x, x, x, x, x, R10G10B10A2_SSCALED) - SF( x, x, x, x, x, x, 75, x, x, x, x, x, R10G10B10A2_SINT) - SF( x, x, x, x, x, x, 75, x, x, x, x, x, B10G10R10A2_SNORM) - SF( x, x, x, x, x, x, 75, x, x, x, x, x, B10G10R10A2_USCALED) - SF( x, x, x, x, x, x, 75, x, x, x, x, x, B10G10R10A2_SSCALED) - SF( x, x, x, x, x, x, 75, x, x, x, x, x, B10G10R10A2_UINT) - SF( x, x, x, x, x, x, 75, x, x, x, x, x, B10G10R10A2_SINT) - SF( x, x, x, x, x, x, 80, x, x, x, x, x, R64G64B64A64_PASSTHRU) - SF( x, x, x, x, x, x, 80, x, x, x, x, x, R64G64B64_PASSTHRU) - SF( 80, 80, x, x, x, x, x, x, x, x, x, x, ETC2_RGB8_PTA) - SF( 80, 80, x, x, x, x, x, x, x, x, x, x, ETC2_SRGB8_PTA) - SF( 80, 80, x, x, x, x, x, x, x, x, x, x, ETC2_EAC_RGBA8) - SF( 80, 80, x, x, x, x, x, x, x, x, x, x, ETC2_EAC_SRGB8_A8) - SF( 90, x, x, x, x, x, 75, x, x, x, x, x, R8G8B8_UINT) - SF( 90, x, x, x, x, x, 75, x, x, x, x, x, R8G8B8_SINT) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_4X4_FLT16) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_5X4_FLT16) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_5X5_FLT16) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_6X5_FLT16) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_6X6_FLT16) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_8X5_FLT16) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_8X6_FLT16) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_8X8_FLT16) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_10X5_FLT16) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_10X6_FLT16) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_10X8_FLT16) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_10X10_FLT16) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_12X10_FLT16) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_12X12_FLT16) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_4X4_U8SRGB) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_5X4_U8SRGB) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_5X5_U8SRGB) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_6X5_U8SRGB) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_6X6_U8SRGB) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_8X5_U8SRGB) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_8X6_U8SRGB) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_8X8_U8SRGB) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_10X5_U8SRGB) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_10X6_U8SRGB) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_10X8_U8SRGB) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_10X10_U8SRGB) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_12X10_U8SRGB) - SF( 90, 90, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_12X12_U8SRGB) - SF(100, 100, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_4X4_FLT16) - SF(100, 100, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_5X4_FLT16) - SF(100, 100, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_5X5_FLT16) - SF(100, 100, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_6X5_FLT16) - SF(100, 100, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_6X6_FLT16) - SF(100, 100, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_8X5_FLT16) - SF(100, 100, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_8X6_FLT16) - SF(100, 100, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_8X8_FLT16) - SF(100, 100, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_10X5_FLT16) - SF(100, 100, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_10X6_FLT16) - SF(100, 100, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_10X8_FLT16) - SF(100, 100, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_10X10_FLT16) - SF(100, 100, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_12X10_FLT16) - SF(100, 100, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_12X12_FLT16) + SF( 80, 80, x, x, x, x, x, x, x, x, x, x, x, R8G8B8_UNORM_SRGB) + SF( 80, 80, x, x, x, x, x, x, x, x, x, x, x, ETC1_RGB8) + SF( 80, 80, x, x, x, x, x, x, x, x, x, x, x, ETC2_RGB8) + SF( 80, 80, x, x, x, x, x, x, x, x, x, x, x, EAC_R11) + SF( 80, 80, x, x, x, x, x, x, x, x, x, x, x, EAC_RG11) + SF( 80, 80, x, x, x, x, x, x, x, x, x, x, x, EAC_SIGNED_R11) + SF( 80, 80, x, x, x, x, x, x, x, x, x, x, x, EAC_SIGNED_RG11) + SF( 80, 80, x, x, x, x, x, x, x, x, x, x, x, ETC2_SRGB8) + SF( 90, x, x, x, x, x, 75, x, x, x, x, x, x, R16G16B16_UINT) + SF( 90, x, x, x, x, x, 75, x, x, x, x, x, x, R16G16B16_SINT) + SF( x, x, x, x, x, x, 75, x, x, x, x, x, x, R32_SFIXED) + SF( x, x, x, x, x, x, 75, x, x, x, x, x, x, R10G10B10A2_SNORM) + SF( x, x, x, x, x, x, 75, x, x, x, x, x, x, R10G10B10A2_USCALED) + SF( x, x, x, x, x, x, 75, x, x, x, x, x, x, R10G10B10A2_SSCALED) + SF( x, x, x, x, x, x, 75, x, x, x, x, x, x, R10G10B10A2_SINT) + SF( x, x, x, x, x, x, 75, x, x, x, x, x, x, B10G10R10A2_SNORM) + SF( x, x, x, x, x, x, 75, x, x, x, x, x, x, B10G10R10A2_USCALED) + SF( x, x, x, x, x, x, 75, x, x, x, x, x, x, B10G10R10A2_SSCALED) + SF( x, x, x, x, x, x, 75, x, x, x, x, x, x, B10G10R10A2_UINT) + SF( x, x, x, x, x, x, 75, x, x, x, x, x, x, B10G10R10A2_SINT) + SF( x, x, x, x, x, x, 80, x, x, x, x, x, x, R64G64B64A64_PASSTHRU) + SF( x, x, x, x, x, x, 80, x, x, x, x, x, x, R64G64B64_PASSTHRU) + SF( 80, 80, x, x, x, x, x, x, x, x, x, x, x, ETC2_RGB8_PTA) + SF( 80, 80, x, x, x, x, x, x, x, x, x, x, x, ETC2_SRGB8_PTA) + SF( 80, 80, x, x, x, x, x, x, x, x, x, x, x, ETC2_EAC_RGBA8) + SF( 80, 80, x, x, x, x, x, x, x, x, x, x, x, ETC2_EAC_SRGB8_A8) + SF( 90, x, x, x, x, x, 75, x, x, x, x, x, x, R8G8B8_UINT) + SF( 90, x, x, x, x, x, 75, x, x, x, x, x, x, R8G8B8_SINT) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_4X4_FLT16) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_5X4_FLT16) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_5X5_FLT16) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_6X5_FLT16) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_6X6_FLT16) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_8X5_FLT16) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_8X6_FLT16) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_8X8_FLT16) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_10X5_FLT16) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_10X6_FLT16) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_10X8_FLT16) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_10X10_FLT16) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_12X10_FLT16) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_12X12_FLT16) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_4X4_U8SRGB) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_5X4_U8SRGB) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_5X5_U8SRGB) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_6X5_U8SRGB) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_6X6_U8SRGB) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_8X5_U8SRGB) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_8X6_U8SRGB) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_8X8_U8SRGB) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_10X5_U8SRGB) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_10X6_U8SRGB) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_10X8_U8SRGB) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_10X10_U8SRGB) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_12X10_U8SRGB) + SF( 90, 90, x, x, x, x, x, x, x, x, x, x, x, ASTC_LDR_2D_12X12_U8SRGB) + SF(110, 110, x, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_4X4_FLT16) + SF(110, 110, x, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_5X4_FLT16) + SF(110, 110, x, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_5X5_FLT16) + SF(110, 110, x, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_6X5_FLT16) + SF(110, 110, x, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_6X6_FLT16) + SF(110, 110, x, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_8X5_FLT16) + SF(110, 110, x, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_8X6_FLT16) + SF(110, 110, x, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_8X8_FLT16) + SF(110, 110, x, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_10X5_FLT16) + SF(110, 110, x, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_10X6_FLT16) + SF(110, 110, x, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_10X8_FLT16) + SF(110, 110, x, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_10X10_FLT16) + SF(110, 110, x, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_12X10_FLT16) + SF(110, 110, x, x, x, x, x, x, x, x, x, x, x, ASTC_HDR_2D_12X12_FLT16) }; #undef x #undef Y @@ -597,6 +604,18 @@ isl_format_for_pipe_format(enum pipe_format pf) [PIPE_FORMAT_ETC1_RGB8] = ISL_FORMAT_ETC1_RGB8, + /* The formats say YCrCb, but there's no colorspace conversion. */ + [PIPE_FORMAT_R8G8_R8B8_UNORM] = ISL_FORMAT_YCRCB_NORMAL, + [PIPE_FORMAT_G8R8_B8R8_UNORM] = ISL_FORMAT_YCRCB_SWAPY, + + /* We map these formats to help configure media compression. */ + [PIPE_FORMAT_YUYV] = ISL_FORMAT_YCRCB_NORMAL, + [PIPE_FORMAT_UYVY] = ISL_FORMAT_YCRCB_SWAPY, + [PIPE_FORMAT_NV12] = ISL_FORMAT_PLANAR_420_8, + [PIPE_FORMAT_P010] = ISL_FORMAT_PLANAR_420_10, + [PIPE_FORMAT_P012] = ISL_FORMAT_PLANAR_420_12, + [PIPE_FORMAT_P016] = ISL_FORMAT_PLANAR_420_16, + [PIPE_FORMAT_R8G8B8X8_SRGB] = ISL_FORMAT_R8G8B8X8_UNORM_SRGB, [PIPE_FORMAT_B10G10R10X2_UNORM] = ISL_FORMAT_B10G10R10X2_UNORM, [PIPE_FORMAT_R16G16B16X16_UNORM] = ISL_FORMAT_R16G16B16X16_UNORM, @@ -672,12 +691,6 @@ isl_format_for_pipe_format(enum pipe_format pf) return table[pf]; } -static unsigned -format_gen(const struct gen_device_info *devinfo) -{ - return devinfo->gen * 10 + (devinfo->is_g4x || devinfo->is_haswell) * 5; -} - static bool format_info_exists(enum isl_format format) { @@ -687,93 +700,88 @@ format_info_exists(enum isl_format format) } bool -isl_format_supports_rendering(const struct gen_device_info *devinfo, +isl_format_supports_rendering(const struct intel_device_info *devinfo, enum isl_format format) { if (!format_info_exists(format)) return false; - return format_gen(devinfo) >= format_info[format].render_target; + /* If this fails, then we need to update struct surface_format_info */ + assert(devinfo->verx10 < + (1ul << (8 * sizeof(format_info[format].render_target)))); + return devinfo->verx10 >= format_info[format].render_target; } bool -isl_format_supports_alpha_blending(const struct gen_device_info *devinfo, +isl_format_supports_alpha_blending(const struct intel_device_info *devinfo, enum isl_format format) { if (!format_info_exists(format)) return false; - return format_gen(devinfo) >= format_info[format].alpha_blend; + return devinfo->verx10 >= format_info[format].alpha_blend; } bool -isl_format_supports_sampling(const struct gen_device_info *devinfo, +isl_format_supports_sampling(const struct intel_device_info *devinfo, enum isl_format format) { if (!format_info_exists(format)) return false; - if (devinfo->is_baytrail) { + if (devinfo->platform == INTEL_PLATFORM_BYT) { const struct isl_format_layout *fmtl = isl_format_get_layout(format); /* Support for ETC1 and ETC2 exists on Bay Trail even though big-core * GPUs didn't get it until Broadwell. */ if (fmtl->txc == ISL_TXC_ETC1 || fmtl->txc == ISL_TXC_ETC2) return true; - } else if (devinfo->is_cherryview) { - const struct isl_format_layout *fmtl = isl_format_get_layout(format); - /* Support for ASTC LDR exists on Cherry View even though big-core - * GPUs didn't get it until Skylake. + } else if (devinfo->platform == INTEL_PLATFORM_CHV) { + /* Support for ASTC LDR theoretically exists on Cherry View even though + * big-core GPUs didn't get it until Skylake. However, it's fairly + * badly broken and requires some nasty workarounds which no Mesa driver + * has ever implemented. */ - if (fmtl->txc == ISL_TXC_ASTC) - return format < ISL_FORMAT_ASTC_HDR_2D_4X4_FLT16; - } else if (gen_device_info_is_9lp(devinfo)) { + } else if (intel_device_info_is_9lp(devinfo)) { const struct isl_format_layout *fmtl = isl_format_get_layout(format); /* Support for ASTC HDR exists on Broxton even though big-core * GPUs didn't get it until Cannonlake. */ if (fmtl->txc == ISL_TXC_ASTC) return true; + } else if (devinfo->verx10 >= 125) { + const struct isl_format_layout *fmtl = isl_format_get_layout(format); + /* ASTC & FXT1 support was removed from the hardware on Gfx12.5. + * Annoyingly, our format_info table doesn't have a concept of things + * being removed so we handle it as yet another special case. + * + * See HSD 1408144932 (ASTC), 1407633611 (FXT1) + * + */ + if (fmtl->txc == ISL_TXC_ASTC || fmtl->txc == ISL_TXC_FXT1) + return false; } - return format_gen(devinfo) >= format_info[format].sampling; + return devinfo->verx10 >= format_info[format].sampling; } bool -isl_format_supports_filtering(const struct gen_device_info *devinfo, +isl_format_supports_filtering(const struct intel_device_info *devinfo, enum isl_format format) { if (!format_info_exists(format)) return false; - if (devinfo->is_baytrail) { - const struct isl_format_layout *fmtl = isl_format_get_layout(format); - /* Support for ETC1 and ETC2 exists on Bay Trail even though big-core - * GPUs didn't get it until Broadwell. - */ - if (fmtl->txc == ISL_TXC_ETC1 || fmtl->txc == ISL_TXC_ETC2) - return true; - } else if (devinfo->is_cherryview) { - const struct isl_format_layout *fmtl = isl_format_get_layout(format); - /* Support for ASTC LDR exists on Cherry View even though big-core - * GPUs didn't get it until Skylake. - */ - if (fmtl->txc == ISL_TXC_ASTC) - return format < ISL_FORMAT_ASTC_HDR_2D_4X4_FLT16; - } else if (gen_device_info_is_9lp(devinfo)) { - const struct isl_format_layout *fmtl = isl_format_get_layout(format); - /* Support for ASTC HDR exists on Broxton even though big-core - * GPUs didn't get it until Cannonlake. - */ - if (fmtl->txc == ISL_TXC_ASTC) - return true; + if (isl_format_is_compressed(format)) { + assert(format_info[format].filtering == format_info[format].sampling); + return isl_format_supports_sampling(devinfo, format); } - return format_gen(devinfo) >= format_info[format].filtering; + return devinfo->verx10 >= format_info[format].filtering; } bool -isl_format_supports_vertex_fetch(const struct gen_device_info *devinfo, +isl_format_supports_vertex_fetch(const struct intel_device_info *devinfo, enum isl_format format) { if (!format_info_exists(format)) @@ -782,25 +790,37 @@ isl_format_supports_vertex_fetch(const struct gen_device_info *devinfo, /* For vertex fetch, Bay Trail supports the same set of formats as Haswell * but is a superset of Ivy Bridge. */ - if (devinfo->is_baytrail) + if (devinfo->platform == INTEL_PLATFORM_BYT) return 75 >= format_info[format].input_vb; - return format_gen(devinfo) >= format_info[format].input_vb; + return devinfo->verx10 >= format_info[format].input_vb; } /** * Returns true if the given format can support typed writes. */ bool -isl_format_supports_typed_writes(const struct gen_device_info *devinfo, +isl_format_supports_typed_writes(const struct intel_device_info *devinfo, enum isl_format format) { if (!format_info_exists(format)) return false; - return format_gen(devinfo) >= format_info[format].typed_write; + return devinfo->verx10 >= format_info[format].typed_write; } +/** + * Returns true if the given format can support typed atomics. + */ +bool +isl_format_supports_typed_atomics(const struct intel_device_info *devinfo, + enum isl_format format) +{ + if (!format_info_exists(format)) + return false; + + return devinfo->verx10 >= format_info[format].typed_atomics; +} /** * Returns true if the given format can support typed reads with format @@ -813,13 +833,13 @@ isl_format_supports_typed_writes(const struct gen_device_info *devinfo, * occurrences. */ bool -isl_format_supports_typed_reads(const struct gen_device_info *devinfo, +isl_format_supports_typed_reads(const struct intel_device_info *devinfo, enum isl_format format) { if (!format_info_exists(format)) return false; - return format_gen(devinfo) >= format_info[format].typed_read; + return devinfo->verx10 >= format_info[format].typed_read; } /** @@ -829,13 +849,13 @@ isl_format_supports_typed_reads(const struct gen_device_info *devinfo, * and sample count. See isl_surf_get_ccs_surf for details. */ bool -isl_format_supports_ccs_d(const struct gen_device_info *devinfo, +isl_format_supports_ccs_d(const struct intel_device_info *devinfo, enum isl_format format) { /* Clear-only compression was first added on Ivy Bridge and was last * implemented on Ice lake (see BSpec: 43862). */ - if (devinfo->gen < 7 || devinfo->gen > 11) + if (devinfo->ver < 7 || devinfo->ver > 11) return false; if (!isl_format_supports_rendering(devinfo, format)) @@ -843,6 +863,12 @@ isl_format_supports_ccs_d(const struct gen_device_info *devinfo, const struct isl_format_layout *fmtl = isl_format_get_layout(format); + /* From the Ivy Bridge PRM, Vol2 Part1 11.7 "MCS Buffer for Render + * Target(s)", beneath the "Fast Color Clear" bullet (p326): + * + * - MCS buffer for non-MSRT is supported only for RT formats 32bpp, + * 64bpp, and 128bpp. + */ return fmtl->bpb == 32 || fmtl->bpb == 64 || fmtl->bpb == 128; } @@ -853,7 +879,7 @@ isl_format_supports_ccs_d(const struct gen_device_info *devinfo, * such as tiling and sample count. See isl_surf_get_ccs_surf for details. */ bool -isl_format_supports_ccs_e(const struct gen_device_info *devinfo, +isl_format_supports_ccs_e(const struct intel_device_info *devinfo, enum isl_format format) { if (!format_info_exists(format)) @@ -861,18 +887,19 @@ isl_format_supports_ccs_e(const struct gen_device_info *devinfo, /* For simplicity, only report that a format supports CCS_E if blorp can * perform bit-for-bit copies with an image of that format while compressed. - * Unfortunately, R11G11B10_FLOAT is in a compression class of its own and - * there is no way to copy to/from it which doesn't potentially loose data - * if one of the bit patterns being copied isn't valid finite floats. + * Unfortunately, R11G11B10_FLOAT is in a compression class of its own, and + * on ICL, there is no way to copy to/from it which doesn't potentially + * loose data if one of the bit patterns being copied isn't valid finite + * floats. */ - if (format == ISL_FORMAT_R11G11B10_FLOAT) + if (devinfo->ver == 11 && format == ISL_FORMAT_R11G11B10_FLOAT) return false; - return format_gen(devinfo) >= format_info[format].ccs_e; + return devinfo->verx10 >= format_info[format].ccs_e; } bool -isl_format_supports_multisampling(const struct gen_device_info *devinfo, +isl_format_supports_multisampling(const struct intel_device_info *devinfo, enum isl_format format) { /* From the Sandybridge PRM, Volume 4 Part 1 p72, SURFACE_STATE, Surface @@ -897,8 +924,23 @@ isl_format_supports_multisampling(const struct gen_device_info *devinfo, /* On SKL+, HiZ is always single-sampled even when the primary surface * is multisampled. See also isl_surf_get_hiz_surf(). */ - return devinfo->gen <= 8; - } else if (devinfo->gen < 7 && isl_format_get_layout(format)->bpb > 64) { + return devinfo->ver <= 8; + } else if (devinfo->ver == 7 && isl_format_has_sint_channel(format)) { + /* From the Ivy Bridge PRM, Vol4 Part1 p73 ("Number of Multisamples"): + * + * This field must be set to MULTISAMPLECOUNT_1 for SINT MSRTs when + * all RT channels are not written + * + * From the Ivy Bridge PRM, Vol4 Part1 p77 ("MCS Enable"): + * + * This field must be set to 0 for all SINT MSRTs when all RT channels + * are not written + * + * Disable multisampling support now as we don't handle the case when + * one of the render target channels is disabled. + */ + return false; + } else if (devinfo->ver < 7 && isl_format_get_layout(format)->bpb > 64) { return false; } else if (isl_format_is_compressed(format)) { return false; @@ -910,6 +952,30 @@ isl_format_supports_multisampling(const struct gen_device_info *devinfo, } /** + * Returns true if the two formats are component size compatible meaning that + * each component from one format has the same number of bits as the other + * format. + * + * This is useful to check whether an image used with 2 different formats can + * be fast cleared with a non 0 clear color. + */ +bool +isl_formats_have_same_bits_per_channel(enum isl_format format1, + enum isl_format format2) +{ + const struct isl_format_layout *fmtl1 = isl_format_get_layout(format1); + const struct isl_format_layout *fmtl2 = isl_format_get_layout(format2); + + return fmtl1->channels.r.bits == fmtl2->channels.r.bits && + fmtl1->channels.g.bits == fmtl2->channels.g.bits && + fmtl1->channels.b.bits == fmtl2->channels.b.bits && + fmtl1->channels.a.bits == fmtl2->channels.a.bits && + fmtl1->channels.l.bits == fmtl2->channels.l.bits && + fmtl1->channels.i.bits == fmtl2->channels.i.bits && + fmtl1->channels.p.bits == fmtl2->channels.p.bits; +} + +/** * Returns true if the two formats are "CCS_E compatible" meaning that you can * render in one format with CCS_E enabled and then texture using the other * format without needing a resolve. @@ -919,7 +985,7 @@ isl_format_supports_multisampling(const struct gen_device_info *devinfo, * format-dependent. */ bool -isl_formats_are_ccs_e_compatible(const struct gen_device_info *devinfo, +isl_formats_are_ccs_e_compatible(const struct intel_device_info *devinfo, enum isl_format format1, enum isl_format format2) { @@ -928,25 +994,17 @@ isl_formats_are_ccs_e_compatible(const struct gen_device_info *devinfo, !isl_format_supports_ccs_e(devinfo, format2)) return false; - /* Gen12 added CCS_E support for A8_UNORM, A8_UNORM and R8_UNORM share the - * same aux map format encoding so they are definitely compatible. + /* On TGL+, drivers may specify a compression format independently from the + * surface format. So, even if the surface format changes, hardware is + * still able to determine how to access the CCS. */ - if (format1 == ISL_FORMAT_A8_UNORM) - format1 = ISL_FORMAT_R8_UNORM; - - if (format2 == ISL_FORMAT_A8_UNORM) - format2 = ISL_FORMAT_R8_UNORM; - - const struct isl_format_layout *fmtl1 = isl_format_get_layout(format1); - const struct isl_format_layout *fmtl2 = isl_format_get_layout(format2); + if (devinfo->ver >= 12) + return true; /* The compression used by CCS is not dependent on the actual data encoding * of the format but only depends on the bit-layout of the channels. */ - return fmtl1->channels.r.bits == fmtl2->channels.r.bits && - fmtl1->channels.g.bits == fmtl2->channels.g.bits && - fmtl1->channels.b.bits == fmtl2->channels.b.bits && - fmtl1->channels.a.bits == fmtl2->channels.a.bits; + return isl_formats_have_same_bits_per_channel(format1, format2); } static bool @@ -1146,6 +1204,51 @@ isl_format_rgbx_to_rgba(enum isl_format rgbx) } } +/* + * Xe2 allows route of LD messages from Sampler to LSC to improve performance + * when some restrictions are met, here checking the format restrictions. + * + * RENDER_SURFACE_STATE::Enable Sampler Route to LSC: + * "The Surface Format is one of the following: + * + * R8_UNORM, R8G8_UNORM, R16_UNORM, R16G16_UNORM, R16G16B16A16_UNORM + * R16_FLOAT, R16G16_FLOAT, R16G16B16A16_FLOAT + * R32_FLOAT, R32G32_FLOAT, R32G32B32A32_FLOAT, R32_UINT, R32G32_UINT, R32G32B32A32_UINT + * R10G10B10A2_UNORM, R11G11B10_FLOAT + * " + */ +bool +isl_format_support_sampler_route_to_lsc(enum isl_format fmt) +{ + /* TODO/FIXME: even only enabling the optimization with formats below this + * is causing some tests to fail so completely disabling this optimization + * for now. + */ + return false; + + switch (fmt) { + case ISL_FORMAT_R8_UNORM: + case ISL_FORMAT_R8G8_UNORM: + case ISL_FORMAT_R16_UNORM: + case ISL_FORMAT_R16G16_UNORM: + case ISL_FORMAT_R16G16B16A16_UNORM: + case ISL_FORMAT_R16_FLOAT: + case ISL_FORMAT_R16G16_FLOAT: + case ISL_FORMAT_R16G16B16A16_FLOAT: + case ISL_FORMAT_R32_FLOAT: + case ISL_FORMAT_R32G32_FLOAT: + case ISL_FORMAT_R32G32B32A32_FLOAT: + case ISL_FORMAT_R32_UINT: + case ISL_FORMAT_R32G32_UINT: + case ISL_FORMAT_R32G32B32A32_UINT: + case ISL_FORMAT_R10G10B10A2_UNORM: + case ISL_FORMAT_R11G11B10_FLOAT: + return true; + default: + return false; + } +} + static inline void pack_channel(const union isl_color_value *value, unsigned i, const struct isl_channel_layout *layout, @@ -1184,11 +1287,11 @@ pack_channel(const union isl_color_value *value, unsigned i, } break; case ISL_UINT: - packed = MIN(value->u32[i], MAX_UINT(layout->bits)); + packed = MIN(value->u32[i], u_uintN_max(layout->bits)); break; case ISL_SINT: - packed = MIN(MAX(value->u32[i], MIN_INT(layout->bits)), - MAX_INT(layout->bits)); + packed = CLAMP(value->u32[i], u_intN_min(layout->bits), + u_intN_max(layout->bits)); break; default: @@ -1198,7 +1301,7 @@ pack_channel(const union isl_color_value *value, unsigned i, unsigned dword = layout->start_bit / 32; unsigned bit = layout->start_bit % 32; assert(bit + layout->bits <= 32); - data_out[dword] |= (packed & MAX_UINT(layout->bits)) << bit; + data_out[dword] |= (packed & u_uintN_max(layout->bits)) << bit; } /** @@ -1235,18 +1338,6 @@ isl_color_value_pack(const union isl_color_value *value, assert(fmtl->channels.p.bits == 0); } -/** Extend an N-bit signed integer to 32 bits */ -static inline int32_t -sign_extend(int32_t x, unsigned bits) -{ - if (bits < 32) { - unsigned shift = 32 - bits; - return (x << shift) >> shift; - } else { - return x; - } -} - static inline void unpack_channel(union isl_color_value *value, unsigned start, unsigned count, @@ -1260,7 +1351,7 @@ unpack_channel(union isl_color_value *value, unsigned dword = layout->start_bit / 32; unsigned bit = layout->start_bit % 32; assert(bit + layout->bits <= 32); - uint32_t packed = (data_in[dword] >> bit) & MAX_UINT(layout->bits); + uint32_t packed = (data_in[dword] >> bit) & u_uintN_max(layout->bits); union { uint32_t u32; @@ -1284,7 +1375,7 @@ unpack_channel(union isl_color_value *value, } break; case ISL_SNORM: - unpacked.f32 = _mesa_snorm_to_float(sign_extend(packed, layout->bits), + unpacked.f32 = _mesa_snorm_to_float(util_sign_extend(packed, layout->bits), layout->bits); break; case ISL_SFLOAT: @@ -1299,7 +1390,7 @@ unpack_channel(union isl_color_value *value, unpacked.u32 = packed; break; case ISL_SINT: - unpacked.u32 = sign_extend(packed, layout->bits); + unpacked.u32 = util_sign_extend(packed, layout->bits); break; default: @@ -1311,14 +1402,14 @@ unpack_channel(union isl_color_value *value, } /** - * Take unpack an isl_color_value from the actual bits as specified by + * Unpack an isl_color_value from the actual bits as specified by * the isl_format. This function is very slow for a format conversion * function but should be fine for a single pixel worth of data. */ void isl_color_value_unpack(union isl_color_value *value, enum isl_format format, - const uint32_t data_in[4]) + const uint32_t *data_in) { const struct isl_format_layout *fmtl = isl_format_get_layout(format); assert(fmtl->colorspace == ISL_COLORSPACE_LINEAR || |