A week in, grab the brown paper bags, for it is time to reveal a couple of critical bugs that spoilt the 2.20.0 release. Firstly we have the restoration of DRI for i810. I am sure that the solitary user will be overjoyed in a couple of years when a new xserver is forced upon it. That enjoyment will be short-lived as no actual acceleration remains, not even shadow, for the chipset. Perhaps a little more wildly felt, I hope!, will be that the SNA fallbacks were broken on 64-bit machines if they required clipping. One little misplaced cast of a pointer, and the screen is filled with corruption. Among the other tweaks this week: * A bug affecting gen4 handling of trapezoids was fixed, and CPU overhead reduced. https://bugs.freedesktop.org/show_bug.cgi?id=52158 * A fix for a bug causing corruption of a DRI2 unredirected client window that was resized whilst under a compositor. * Support for snoopable buffers on non-LLC architectures, coming to a future kernel. The aim to accelerate transfers between the CPU and the GPU, in particular to dramatically improve readback performance, and to further minimise clflushes. * Improvement to the composite performance on GT2 SandyBridge and IvyBridge devices, in particular the render copy is significantly improved. * Improved handling for when acceleration is disabled, including permitting DRI2 to remain supported even if the X server believes the GPU wedged. * Shadow support was dropped from UXA as it was neither complete nor correct, use SNA instead. -Chris Chris Wilson (83): uxa: Remove Shadow hack intel: Don't use stdbool without declaring it sna: Disable the scanout flush when switch off via DPMS sna: Add a few DBG to show when CPU bos are being used for xfer sna: Discard and recreate the CPU buffer when busy during move-to-cpu sna: Add a couple of DBG options to control accelerated up/downloads sna: Use set-cache-level to allocate snoopable upload buffers sna: Move the disabling of CPU bo for gen4 to the render unit sna/trapezoids: Add some DBG to unaligned fills sna/trapezoids: Fix inplace unaligned fills (on gen4) Wrap defines to avoid redefinition warnings sna: Fixup pixmap validation for sna_copy_area() sna: Disable snoopable uplaod buffers for gen4 sna: Disable snoopable bo for gen4 sna: Share the pixmap migration decision with the BLT composite routines sna: Promote an undamaged pixmap to use the full GPU sna: Fix glyph DBG to include clip extents and actual glyph origin sna: Only drop the clear flag when writing to the GPU pixmap sna: Limit the use of snoopable buffers to read/write uploads sna: Catch the short-circuit path for clearing clear on move-to-gpu as well sna: Avoid the CPU bo readback for render paths sna: Rebalance choice of GPU vs CPU bo sna/dri: Do not allow an exchange to take place on invalid buffers sna/gen7: Bump the number of pixel shader threads for IVB GT2 sna: prefer fbBlt over pixman_blt sna: Tweak fast blt path sna: Allow operation inplace to scanout whilst wedged sna: Allow inplace copies for wedged CopyArea sna: Allow wedged CopyPlane to operate inplace on the destination i810: Split xaa routines from common acceleration methods i810: Replace XAAGet.*ROP() with local tables sna: Maintain a short-lived cache of snoopable CPU bo for older gen sna: Reuse the snoopable cache more frequently for upload buffers sna: Add more DBG for fallback processing sna: Fix processing of the last fallback box sna/trapezoids: Use pixman from within the spans to reduce two-pass operations sna/trapezoids: Only reduce bounded operators to a single pass sna: Enable runtime detection of set-cacheing ioctl sna/gen7: Micro-optimise render copy emission sna/gen6: Micro-optimise render copy emission sna/dri: Allow DRI2 to be loaded even if we are wedged sna/gen4+: Drop unsupported source formats i810: DRI is not dependent upon XAA sna: Re-register the SHM funcs every server generation i810: Handle initialisation without the XAA module present at runtime i810: Correct the double negative and enable XAA when available sna: Tweak order of screen re-initialisation sna/gen4: Hookup composite spans sna: Handle mixed bo/buffers in assertions sna/gen6: Add a simple DBG option to limit usage of either BLT/RENDER sna/gen6: Bump the WM thread count to 80 sna: Remove unused scanout-is-dirty? flag sna: Replace 'sync' flag with equivalent 'flush' sna: Remove topmost unused 'flush' attribute sna: Allow the snoopable upload buffer to take pages from the CPU vma cache sna: Rename kgem_partial_bo to kgem_buffer sna: Update WIP userptr example usage sna/dri: Cleanup ring selection for SNB+ CopyRegion sna: Just use composite.box() when we only have one box sna/gen6: Prefer the more flexible render ring for large surfaces sna: Also check whether the first upload box can use the BLT sna: Only set the vmap flag after we make the bo snoopable sna: Tweak CPU bo promotion rules for CopyArea sna: Only update a buffer when it becomes dirty sna: Micro-optimise copying boxes with the blitter sna: Refresh experimental userptr vmap support sna: Fix role reversal of __kgem_bo_size() and kgem_bo_size()! sna: s/vmap/snoop/ since we use the flag more generically sna: Change the vmap interface name to userptr sna: Correct assertion for __kgem_bo_size() sna: Avoid marking io/snoop buffers as unreusable unnecessarily sna: Fix the reversed not SHM assertion sna: Enable snooping on the reused linear buffer sna: Also discard the last-was-cpu flag when overwriting cpu damage sna: Update assertion for cached io buffers sna: Expand the heuristic for predicting when to use CPU bo for readback sna: Adjust hints to prefer rendering convex polygon with the GPU sna/dri: We fail at predicting the flip frame sna: Use an upload buffer for large stipples sna: Promote tiled operations to the GPU if the tile is already on the GPU sna: Tweak the fallback hints for XYPixmap PutImage sna/glyphs: Also discard the glyph mask for bitmaps with an opaque source 2.20.1 release git tag: 2.20.1 http://xorg.freedesktop.org/archive/individual/driver/xf86-video-intel-2.20.1.tar.bz2 MD5: 9d86297b9b7b4aa41c01fc2108bbd706 xf86-video-intel-2.20.1.tar.bz2 SHA1: 19e9d2497f0070d6fd40ff74054d1ff75e55b694 xf86-video-intel-2.20.1.tar.bz2 SHA256: de340a527dfa26bba6f79b423c2fe2b1d70657e46fa3fb468613c047443ff424 xf86-video-intel-2.20.1.tar.bz2 http://xorg.freedesktop.org/archive/individual/driver/xf86-video-intel-2.20.1.tar.gz MD5: 02eb8ae38a6191d7baddb4b815b3f0ad xf86-video-intel-2.20.1.tar.gz SHA1: 28380e7589f124a7e2813d6a019659cf079da79d xf86-video-intel-2.20.1.tar.gz SHA256: 17adde739453499be9755d88d5d96982197e7baa0c0a2a7b69c81d6032d05d14 xf86-video-intel-2.20.1.tar.gz -- Chris Wilson, Intel Open Source Technology Centre -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 198 bytes Desc: Digital signature URL: <http://lists.freedesktop.org/archives/intel-gfx/attachments/20120723/cfd63240/attachment.pgp>
