[Mesa-dev] [Mesa-announce] Mesa 17.1.10 release candidate
Nicolai Hähnle
nicolai.haehnle at amd.com
Wed Sep 27 13:49:37 UTC 2017
On 27.09.2017 15:44, Emil Velikov wrote:
> On 26 September 2017 at 05:25, Timothy Arceri <tarceri at itsqueeze.com> wrote:
>> On 22/09/17 19:41, Juan A. Suarez Romero wrote:
>>>
>>> On Fri, 2017-09-22 at 16:38 +1000, Timothy Arceri wrote:
>>>>
>>>> On 22/09/17 05:10, Juan A. Suarez Romero wrote:
>>>>>
>>>>> Hello list,
>>>>>
>>>>> The candidate for the Mesa 17.1.10 is now available. Currently we have:
>>>>> - 41 queued
>>>>> - 0 nominated (outstanding)
>>>>> - and 5 rejected patches
>>>>>
>>>>> This is the last release for the 17.1 series.
>>>>>
>>>>> In the current queue we have:
>>>>>
>>>>> In build and integration system, we add a dependency on libunwind when
>>>>> running make distcheck, as this is optional for libgallium but we want
>>>>> to catch any problem. As consequence, also force LLVM 3.3 in Travis
>>>>> when building Gallium ST Other, as this is the minimum required version
>>>>> we want to test. On the other hand, we link libmesautil into
>>>>> u_atomic_test, as this is required by platforms without particular
>>>>> atomic operations. In this sense, there's a patch to implement
>>>>> __sync_val_compare_and_swap_8, required by 32-bit PowerPC platforms.
>>>>>
>>>>> The state tracker received a couple of patches, one that fixes a dEQP-
>>>>> GLES31 test and another that fixes the usage of 64-bit unsigned
>>>>> integers when used for boolean comparisons.
>>>>>
>>>>> The platform part adds a missing __DRI_BACKGROUND_CALLABLE extension.
>>>>>
>>>>> Intel i965 driver gets a fix for a crash that can happen in Haswell
>>>>> when uploading a stencil texture with blorp.
>>>>>
>>>>> Intel ANV driver fixes a problem when copying descriptors.
>>>>>
>>>>> VC4 driver gets several use-after-free fixes.
>>>>>
>>>>> SWR driver fixes a use case that happens when attaching/dettaching/re-
>>>>> attaching surfaces, that leads to wrong behaviour.
>>>>>
>>>>> AMD drivers get a fix for various dEQP-GLES31 tests. Also, the radeonsi
>>>>> driver receives a couple of patches more that fixes different dEQP's
>>>>> OpenGL-ES tests. In the same way, RADV driver also gets a fix for
>>>>> several dEQP's Vulkan tests.
>>>>>
>>>>> 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 17.1.10 next Monday (25th of September), around or
>>>>> shortly after 15:00 GMT.
>>>>>
>>>>> If you have any questions or suggestions - be that about the current
>>>>> patch queue or otherwise, please go ahead.
>>>>>
>>>>>
>>>>> Trivial merge conflicts
>>>>> -----------------------
>>>>>
>>>>> commit f348cd1430b7ca436bd83043c3f0155c4a045c1c
>>>>> Author: Nicolai Hähnle <nicolai.haehnle at amd.com>
>>>>>
>>>>> radeonsi: apply a mask to gl_SampleMaskIn in the PS prolog
>>>>>
>>>>> (cherry picked from commit
>>>>> 92c4277990799641d4986ce66a62366228713945)
>>>>>
>>>>> commit 21b223ebac2aba61bcf35201792c03ffca415afd
>>>>> Author: Jason Ekstrand <jason.ekstrand at intel.com>
>>>>>
>>>>> i965/blorp: Set r8stencil_needs_update when writing stencil
>>>>>
>>>>> (cherry picked from commit
>>>>> a43d379000260485fc4b2b03b069aedc46879557)
>>>>>
>>>>> commit 5a1e8e29dfa9b8394914b55b23fd6f497411dbd4
>>>>> Author: Nicolai Hähnle <nicolai.haehnle at amd.com>
>>>>>
>>>>> amd/common: round cube array slice in ac_prepare_cube_coords
>>>>>
>>>>> (cherry picked from commit
>>>>> e0af3bed2cb52b5d8cf1da08b42cf28bae131c76)
>>>>>
>>>>> commit e88c22f645663d80051da9fa0dd4abf533a04eba
>>>>> Author: Matt Turner <mattst88 at gmail.com>
>>>>>
>>>>> util: Link libmesautil into u_atomic_test
>>>>>
>>>>> (cherry picked from commit
>>>>> d075a4089ec62c489c8a3423f82371bf85f2ea6c)
>>>>>
>>>>> commit 81d3b3ca168bf1ebe736bb994b20b4aeb2e3966e
>>>>> Author: Nicolai Hähnle <nicolai.haehnle at amd.com>
>>>>>
>>>>> radeonsi: fix array textures layer coordinate
>>>>>
>>>>> (cherry picked from commit
>>>>> 87f7c7bd6566dbd95214df9b058bca5e6573374e)
>>>>>
>>>>>
>>>>> Cheers,
>>>>> J.A.
>>>>>
>>>>>
>>>>> Mesa stable queue
>>>>> -----------------
>>>>>
>>>>> Nominated (0)
>>>>> =============
>>>>>
>>>>>
>>>>> Queued (41)
>>>>> ===========
>>>>>
>>>>> Alexandre Demers (1):
>>>>> osmesa: link with libunwind if enabled (v2)
>>>>>
>>>>> Andres Gomez (12):
>>>>> docs: add sha256 checksums for 17.1.9
>>>>> cherry-ignore: add "st/mesa: skip draw calls with
>>>>> pipe_draw_info::count == 0"
>>>>> cherry-ignore: add "radv: use amdgpu_bo_va_op_raw."
>>>>> cherry-ignore: add "radv: use simpler indirect packet 3 if
>>>>> possible."
>>>>> cherry-ignore: add "radeonsi: don't always apply the PrimID
>>>>> instancing bug workaround on SI"
>>>>> cherry-ignore: add "intel/eu/validate: Look up types on demand
>>>>> in execution_type()"
>>>>> cherry-ignore: add "radv: gfx9 fixes"
>>>>> cherry-ignore: add "radv/gfx9: set mip0-depth correctly for 2d
>>>>> arrays/3d images"
>>>>> cherry-ignore: add "radv/gfx9: fix image resource handling."
>>>>> cherry-ignore: add "docs/egl: remove reference to
>>>>> EGL_DRIVERS_PATH"
>>>>> cherry-ignore: add "radv: Disable multilayer & multilevel DCC."
>>>>> cherry-ignore: add "radv: Don't allocate CMASK for linear
>>>>> images."
>>>>>
>>>>> Dave Airlie (2):
>>>>> radv/ac: bump params array for image atomic comp swap
>>>>> st/glsl->tgsi: fix u64 to bool comparisons.
>>>>>
>>>>> Emil Velikov (2):
>>>>> egl/x11/dri3: adding missing __DRI_BACKGROUND_CALLABLE extension
>>>>> automake: enable libunwind in `make distcheck'
>>>>>
>>>>> Eric Anholt (3):
>>>>> broadcom/vc4: Fix use-after-free for flushing when writing to a
>>>>> texture.
>>>>> broadcom/vc4: Fix use-after-free trying to mix a quad and tile
>>>>> clear.
>>>>> broadcom/vc4: Fix use-after-free when deleting a program.
>>>>>
>>>>> George Kyriazis (1):
>>>>> swr: invalidate attachment on transition change
>>>>>
>>>>> Gert Wollny (2):
>>>>> travis: force llvm-3.3 for "make Gallium ST Other"
>>>>> travis: Add libunwind-dev to gallium/make builds
>>>>>
>>>>> Jason Ekstrand (1):
>>>>> i965/blorp: Set r8stencil_needs_update when writing stencil
>>>>>
>>>>> Juan A. Suarez Romero (8):
>>>>> cherry-ignore: add "ac/surface: match Z and stencil tile config"
>>>>> cherry-ignore: add "radv/nir: call opt_remove_phis after trivial
>>>>> continues."
>>>>> cherry-ignore: add "amd/common: add workaround for cube map
>>>>> array layer clamping"
>>>>> cherry-ignore: add "radeonsi: workaround for gather4 on integer
>>>>> cube maps"
>>>>> cherry-ignore: add "Scons: Add LLVM 5.0 support"
>>>>> cherry-ignore: add "ac/surface: handle S8 on gfx9"
>>>>> cherry-ignore: add "radv: Check for GFX9 for 1D arrays in
>>>>> image_size intrinsic."
>>>>> cherry-ignore: add "glsl/linker: fix output variable overlap
>>>>> check"
>>>>>
>>>>> Józef Kucia (1):
>>>>> anv: Fix descriptors copying
>>>>>
>>>>> Matt Turner (2):
>>>>> util: Link libmesautil into u_atomic_test
>>>>> util/u_atomic: Add implementation of
>>>>> __sync_val_compare_and_swap_8
>>>>>
>>>>> Nicolai Hähnle (5):
>>>>> st/glsl_to_tgsi: only the first (inner-most) array reference can
>>>>> be a 2D index
>>>>> amd/common: round cube array slice in ac_prepare_cube_coords
>>>>> radeonsi: apply a mask to gl_SampleMaskIn in the PS prolog
>>>>> radeonsi: set MIP_POINT_PRECLAMP to 0
>>>>> radeonsi: fix array textures layer coordinate
>>>>>
>>>>> Tapani Pälli (1):
>>>>> mesa: free current ComputeProgram state in
>>>>> _mesa_free_context_data
>>>>>
>>>>> Rejected (5)
>>>>> ============
>>>>>
>>>>> Jason Ekstrand (1):
>>>>> intel/eu/validate: Look up types on demand in execution_type()
>>>>> Depends on earlier commit 4fab67a4415 which did not land
>>>>> in branch.
>>>>>
>>>>> Nicolai Hähnle (4):
>>>>> ac/surface: match Z and stencil tile config
>>>>> Depends on earlier commit f187a4932 which did not land in
>>>>> branch.
>>>>> amd/common: add workaround for cube map array layer clamping
>>>>> Depends on earlier commit 13a28ff2368 that did not land in
>>>>> branch.
>>>>> radeonsi: workaround for gather4 on integer cube maps
>>>>> Depends on earlier commit 5cb2eee557 which did not land in
>>>>> branch.
>>>>> glsl/linker: fix output variable overlap check
>>>>> This commit is causing several regressions on i965 driver
>>>>
>>>>
>>>>
>>>> Hi,
>>>>
>>>> The last one of Nicolai's patches should work ok if you apply this patch
>>>> on top of it:
>>>>
>>>> commit df8767a14e3eae4dcb8241b731b34e9379706795
>>>>
>>>> glsl/linker: properly fix output variable overlap check
>>>>
>>>> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102904
>>>> Fixes: 15cae12804e ("glsl/linker: fix output variable overlap check")
>>>> Reviewed-by: Kenneth Graunke <kenneth at whitecape.org>
>>>>
>>>>
>>>
>>>
>>> Right. Actually, I was waiting for the patch to land in master before
>>> doing the pre-announcement. But arrived late.
>>>
>>>
>>> As we are doing releases quite often (one every two weeks), and next
>>> week we are doing a new release I think it is safer to keep the queue
>>> as it is, and include this fix in the following releases.
>>
>>
>> Since this was the last 17.1 release it seems odd to have rushed it out.
>> Releasing every two weeks is fine but most distros only update to these
>> minor bugfix releases. I'm not overly concerned that this particular patch
>> didn't make it in, it just seems odd to have skipped it considering it was
>> the final 17.1 release.
>>
> Have to agree, having fixes omitted from the stable release isn't perfect.
> Then again, there will always be bugs.
>
> I wouldn't worry too much ... that unless the issue is quite
> noticeable and spans across dozens of games/apps.
I agree. For what it's worth, this particular issue *will* cause CTS
failures related to missing linker errors. Since it only links stuff
that should fail, games and other applications shouldn't be affected. So
it's definitely not ideal, but missing this patch is also quite unlikely
to break any use cases.
Cheers,
Nicolai
>
> With the help of Valve (thanks Pierre-Loup) that should be a bit
> easier to track.
>
> Thanks
> Emil
>
More information about the mesa-dev
mailing list