[Mesa-dev] Mesa 13.0.1 release candidate

Jason Ekstrand jason at jlekstrand.net
Fri Nov 11 22:23:23 UTC 2016


On Fri, Nov 11, 2016 at 10:13 AM, Emil Velikov <emil.l.velikov at gmail.com>
wrote:

> Hello list,
>
> The candidate for the Mesa 12.0.4 is now available. Currently we have:
>  - 51 queued
>  - 18 nominated (outstanding)
>  - and 0 (self-)rejected patches
>
> As one can imagine huge hunk of the commits are fixes in the Vulkan
> drivers.
>
> Up-to 30% performance increase using the Intel ANV Vulkan driver. The
> Radeon
> RADV Vulkan driver now ships with distinctly named json files, which
> include
> the full Vulkan driver path. Making it possible for people to install and
> use the drivers outside their /usr/lib directory.
>
> RADV correctly advertises support for the Xlib platform extension and one
> can
> use it alongside the Dolphin emulator. ANV and RADV have improved handling
> of
> Xlib/XCB connections and advertise the currect versions of the extensions
> they
> support.
>
> We also have restored interop between third party OpenCL implementations
> and
> Mesa's libGL/libEGL. The GLX GLVND driver has a few noticable fixes as
> well.
>
> Take a look at section "Mesa stable queue" for more information.
>
>
> Testing reports/general approval
> --------------------------------
> Any testing reports (or general approval of the state of the branch) will
> be
> greatly appreciated.
>
> The plan is to have 13.0.1 this Sunday (13th of November), around or
> shortly after
> 18:00 GMT.
>
> If you have any questions or suggestions - be that about the current patch
> queue or otherwise, please go ahead.
>
>
> Trivial merge conflicts
> -----------------------
> commit 990f395e007c3204639daa34efc3049f350ee819
> Author: Emil Velikov <emil.velikov at collabora.com>
>
>     anv: automake: cleanup the generated json file during make clean
>
>     (cherry picked from commit 8df581520a823564be0ab5af7dbb7d501b1c9670)
>
>
> commit 89cefe6325b2534c4e35ac0a4b85155a3be57936
> Author: Jason Ekstrand <jason.ekstrand at intel.com>
>
>     intel/blorp: Rework our usage of ralloc when compiling shaders
>
>     (cherry picked from commit 43dadb6edd5e3e3e10b1198184a9f75556edad49)
>
>
> commit 5be463694bfbeed3066601cbdb70825fe7e97859
> Author: Jason Ekstrand <jason.ekstrand at intel.com>
>
>     intel/blorp: Use wm_prog_data instead of hand-rolling our own
>
>     (cherry picked from commit 058304f0819c96fd18eb17911022a410540aa295)
>
>
> Cheers,
> Emil
>
>
> Mesa stable queue
> -----------------
>
> Nominated (18)
> ==============
>
> Dave Airlie (1):
>       2de85eb radv: fix texturesamples to handle single sample case
>
> Emil Velikov (1):
>       a39ad18 configure.ac: honour LLVM_LIBDIR when linking against LLVM
>
> Eric Anholt (2):
>       4d019bd vc4: Don't abort when a shader compile fails.
>       08d5148 vc4: Clamp the shadow comparison value.
>
> Ian Romanick (2):
>       e85a747 glcpp: Handle '#version 0' and other invalid values
>       c8c4664 glsl: Parse 0 as a preprocessor INTCONSTANT
>
> Jason Ekstrand (10):
>       7139704 anv/gen8: Stall when needed in Cmd(Set|Reset)Event
>       73701be anv/wsi: Set the fence to signaled in AcquireNextImageKHR
>       843775b anv: Rework fences
>       3b6abfc vulkan/wsi/wayland: Include pthread.h
>       302f641 vulkan/wsi/wayland: Clean up some error handling paths
>



>       1e3e347 anv/cmd_buffer: Take a command buffer instead of a batch
> in two helpers
>       a6c3d0f anv/cmd_buffer: Enable a CS stall workaround for Sky Lake gt4
>

Please make sure these two land in 13.1.  They fix some pretty bad hangs on
SKL gt4


>       4fa0ca8 vulkan/wsi: Report the correct min/maxImageCount
>       e73d136 vulkan/wsi/x11: Implement FIFO mode.
>       a5e88e6 i965/gs: Allow primitive id to be a system value
>
> Kenneth Graunke (1):
>       aaee3da mesa: Fix pixel shader scratch space allocation on Gen9+
> platforms.
>

And this one as well.


>
> Kevin Strasser (1):
>       932bb3f vulkan/wsi: Add a thread-safe queue implementation
>
>
> Queued (51)
> ===========
>
> Adam Jackson (2):
>       glx/glvnd: Don't modify the dummy slot in the dispatch table
>       glx/glvnd: Fix dispatch function names and indices
>
> Andreas Boll (1):
>       glx/windows: Add wgl.h to the sources list
>
> Anuj Phogat (1):
>       i965: Fix GPU hang related to multiple render targets and alpha
> testing
>
> Chih-Wei Huang (1):
>       android: avoid using libdrm with host modules
>
> Dave Airlie (8):
>       radv: expose xlib platform extension
>       radv: fix dual source blending
>       Revert "st/vdpau: use linear layout for output surfaces"
>       radv: emit correct last export when Z/stencil export is enabled
>       ac/nir: add support for discard_if intrinsic (v2)
>       nir: add conditional discard optimisation (v4)
>       radv: enable conditional discard optimisation on radv.
>       radv: fix GetFenceStatus for signaled fences
>
> Emil Velikov (5):
>       docs: add sha256 checksums for 13.0.0
>       amd/addrlib: limit fastcall/regparm to GCC i386
>       anv: use correct .specVersion for extensions
>       radv: use correct .specVersion for extensions
>       radv: Suffix the radeon_icd file with the host CPU
>
> Eric Anholt (1):
>       vc4: Use Newton-Raphson on the 1/W write to fix glmark2 terrain.
>
> Francisco Jerez (1):
>       nir: Flip gl_SamplePosition in nir_lower_wpos_ytransform().
>
> Fredrik Höglund (1):
>       radv: add support for anisotropic filtering on VI+
>
> Jason Ekstrand (19):
>       anv/device: Return DEVICE_LOST if execbuf2 fails
>       vulkan/wsi/x11: Better handle wsi_x11_connection_create failure
>       vulkan/wsi/x11: Clean up connections in finish_wsi
>       anv: Better handle return codes from anv_physical_device_init
>       intel/blorp: Use wm_prog_data instead of hand-rolling our own
>       intel/blorp: Pass a brw_stage_prog_data to upload_shader
>       anv/pipeline: Put actual pointers in anv_shader_bin
>       anv/pipeline: Properly cache prog_data::param
>       intel/blorp: Emit all the binding tables
>       anv/device: Add an execbuf wrapper
>       anv: Add a cmd_buffer_execbuf helper
>       anv: Don't presume to know what address is in a surface relocation
>       anv: Add a new bo_pool_init helper
>       anv/allocator: Simplify anv_scratch_pool
>       anv: Initialize anv_bo::offset to -1
>       anv/batch_chain: Improve write_reloc
>       anv: Add an anv_execbuf helper struct
>       anv/batch: Move last_ss_pool_bo_offset to the command buffer
>       anv: Move relocation handling from EndCommandBuffer to QueueSubmit
>
> Kenneth Graunke (1):
>       glsl: Update deref types when resizing implicitly sized arrays.
>
> Kristian Høgsberg (1):
>       anv: Do relocations in userspace before execbuf ioctl
>
> Marek Olšák (4):
>       egl: use util/macros.h
>       egl: make interop ABI visible again
>       glx: make interop ABI visible again
>       radeonsi: fix an assertion failure in si_decompress_sampler_color_
> textures
>
> Nicolai Hähnle (4):
>       radeonsi: fix BFE/BFI lowering for GLSL semantics
>       glsl: fix lowering of UBO references of named blocks
>       st/glsl_to_tgsi: fix dvec[34] loads from SSBO
>       st/mesa: fix the layer of VDPAU surface samplers
>
> Timothy Arceri (1):
>       mesa/glsl: delete previously linked shaders earlier when linking
>
>
> Rejected (0)
> ============
> _______________________________________________
> mesa-dev mailing list
> mesa-dev at lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/mesa-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/mesa-dev/attachments/20161111/a170ca91/attachment-0001.html>


More information about the mesa-dev mailing list