summaryrefslogtreecommitdiff
AgeCommit message (Expand)AuthorFilesLines
19 hourssna: Always flush the shadow scanoutChris Wilson1-1/+9
19 hourssna: Fix some DBG formatsChris Wilson1-5/+5
21 hourssna: Do not reuse physical cursors for the kernel is brokenChris Wilson1-1/+1
45 hourssna: Pre-emptively bind framebuffers into the GTTChris Wilson1-0/+10
4 dayssna: Pre-emptively move framebuffers into the DISPLAY cache domainChris Wilson1-0/+41
4 dayssna: Mark partial composite operations upfrontChris Wilson9-9/+10
4 dayssna: Discard TearFree readback when overdrawingChris Wilson3-10/+56
4 dayssna: Fix up cut'n'paste typo in assertionChris Wilson1-1/+1
4 dayssna: Add a cursor DBG for a failed pwrite updateChris Wilson1-1/+2
5 dayssna: Rearrange cursor reuse to avoid sharing rotationsChris Wilson1-1/+13
5 dayssna: Add some bounds checking asserts to damageChris Wilson1-10/+19
7 dayssna: Restrict WHOLE hint to PutImage on a complete PixmapChris Wilson1-3/+6
8 dayssna: Perform 2-color to ARGB cursor conversion inplaceChris Wilson1-49/+49
8 dayssna: Don't redraw cursor if the fg/bg colors do not changeChris Wilson1-5/+17
8 dayssna: Don't set cursor bits to an invalid pointerChris Wilson1-1/+1
8 dayssna: Only mark the transform_in_use when required for shadow fbChris Wilson1-7/+9
8 dayssna: Only reload an active HW cursorChris Wilson1-18/+32
8 dayssna: Don't create slave pixmaps for the virtual CRTCChris Wilson1-11/+0
9 dayssna: Validate the cursor everytime for hw supportChris Wilson1-16/+15
9 dayssna: Remove defunct Cursor interface on fake CRTCChris Wilson1-30/+0
9 dayssna: Prepare for an upcoming Cursor API changeChris Wilson1-0/+14
10 dayssna: Use pwrite to update physical cursorsChris Wilson1-11/+42
10 dayssna: Handle clipped PutImage uploads more carefullyChris Wilson1-5/+48
10 dayssna: Tweak application of WHOLE hint for uploadsChris Wilson2-23/+17
10 dayssna: Add some DBG for whole CPU pixmap migrationChris Wilson1-1/+4
10 dayssna: Restore empty region checkingChris Wilson1-4/+9
11 dayssna: Fix build for older Xorg without OsBlockSIGIO supportChris Wilson1-9/+32
11 dayssna: Refactor 3 similar CPU migration promotion checks into oneChris Wilson1-10/+5
11 dayssna: Add more assertions before dereferencing sna_crtcChris Wilson1-0/+6
11 dayssna: Continue to reprobe for ZaphodHeads with a mixed up configurationChris Wilson1-1/+1
12 dayssna: Simplify checking for singular damageChris Wilson1-2/+2
12 dayssna: Avoid double application of pixel widening for degenerate linesChris Wilson1-10/+15
14 dayssna: Avoid discarding damage when applying WHOLE hint to pixmap migrationChris Wilson1-15/+19
2014-04-04sna: Remove unitialized use of 'cursor'Chris Wilson1-16/+16
2014-04-04sna: Precompute OVER/ADD with solids to convert it into a BLT operationChris Wilson1-2/+125
2014-04-03sna: Short-circuit repeated clearing to the same colorChris Wilson1-2/+27
2014-04-02sna: Silence compiler warningChris Wilson1-3/+3
2014-04-02sna: Prevent signal re-entrancy into cursor update routinesChris Wilson2-48/+68
2014-03-31sna: Update tiling flags after reusing inactive VMAChris Wilson1-0/+2
2014-03-31sna: Fix predicate inversion for assertionChris Wilson1-2/+2
2014-03-31sna: Use a cheaper check for a replacement operationChris Wilson1-1/+1
2014-03-31sna: Allow reassignment of inactive VMA if not mapped into the GTTChris Wilson1-3/+7
2014-03-31sna: Discard damage tracking for operations to the whole pixmapChris Wilson1-0/+12
2014-03-31sna: Promote to the GPU operations that cover the whole pixmapChris Wilson3-4/+17
2014-03-31sna: Regularly check that damage does exceed the pixmapChris Wilson1-25/+29
2014-03-31sna: Search cursor cache firstChris Wilson1-7/+12
2014-03-28sna: Suppress a compiler warningChris Wilson1-2/+2
2014-03-28sna: Only transform the temporary cursor coordinatesChris Wilson1-31/+20
2014-03-28sna: Our cursors are always square.Chris Wilson2-17/+13
2014-03-28sna: Virtual CRTCs are last, so break loops earlyChris Wilson1-17/+36