Mesa (staging/21.1): 28 new commits
GitLab Mirror
gitlab-mirror at kemper.freedesktop.org
Thu Jul 22 21:42:55 UTC 2021
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=4d1d06afe805813c3d6fc2440c0f50af183bc936
Author: Connor Abbott <cwabbott0 at gmail.com>
Date: Mon Jun 28 10:48:53 2021 +0200
freedreno/computerator: Fix local_size typo
Fixes: cbc68c79a51 ("freedreno: Add local_size to ir3_shader_variant")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11622>
(cherry picked from commit 56dc84b95cd80332c7e9377c2ad39ff41813d1ac)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=f268531e95d30b7838171d7c7cece3af8b172b85
Author: Rhys Perry <pendingchaos02 at gmail.com>
Date: Wed Jul 21 20:18:12 2021 +0100
aco: don't create v_madmk_f32/v_madak_f32 from v_fma_legacy_f16
Signed-off-by: Rhys Perry <pendingchaos02 at gmail.com>
Reviewed-by: Daniel Schürmann <daniel at schuermann.dev>
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5105
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12004>
(cherry picked from commit 211d1dfd34499d57709c15d10bdadc514e9e13c2)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=75e3883013a7045837c1e44759d6cfbfdaaef01c
Author: Timothy Arceri <tarceri at itsqueeze.com>
Date: Sat Jul 17 21:39:55 2021 +1000
util: add workaround for Full Bore
Unfortunately I contacted the dev about this issue years ago and he
made a fix, but it has never been released after all these years.
This stops the screen from being completely black in game.
CC: mesa-stable
Reviewed-by: Marek Olšák <marek.olsak at amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11941>
(cherry picked from commit f3ec4a934d6a9776e6f67cd44565180c098a6846)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=0230c00187ab8f66f314d81c165635e884b22ceb
Author: Dave Airlie <airlied at redhat.com>
Date: Mon Jul 19 11:57:12 2021 +1000
llvmpipe: add support for time elapsed queries.
It turns out for QBO you really need to explicitly support time
elapsed queries to avoid wierd interactions with the non-qbo
query paths.
Fixes: 506e51b8560f ("llvmpipe: initial query buffer object support. (v2)")
Reviewed-by: Michel Dänzer <mdaenzer at redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11946>
(cherry picked from commit 2f5cd08ede0b9d445535d54beffcad0e5c4a62bb)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=447b61cf71a2e71db15d9924e69e181cc4af8e17
Author: Chia-I Wu <olvaffe at gmail.com>
Date: Tue Jul 20 17:52:47 2021 -0700
meson: allow egl_native_platform to be specified
After commit f8dc22bf61c, it was no longer possible to have explicitly
enabled platforms with surfaceless being the EGL native platform. This
fixes that by adding -Degl-native-platform.
Fixes: f8dc22bf61c ("meson: drop deprecated EGL platform build options")
Signed-off-by: Chia-I Wu <olvaffe at gmail.com>
Reviewed-by: Adam Jackson <ajax at redhat.com>
Reviewed-by: Emil Velikov <emil.l.velikov at gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11992>
(cherry picked from commit 47946855f19cc7d082fd59e090bb83f5d3930871)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=2f19f92f391baff2d1875db4d902c19465b6faae
Author: Chia-I Wu <olvaffe at gmail.com>
Date: Tue Jul 20 14:31:46 2021 -0700
egl/surfaceless: try kms_swrast before swrast
Before commit f7e0cdcf1a5, we tried these in order
- if (!ForceSoftware) surfaceless_probe_device(disp, false);
- surfaceless_probe_device(disp, true);
- surfaceless_probe_device_sw(disp);
The commit changed it to
- surfaceless_probe_device(disp, ForceSoftware);
- surfaceless_probe_device_sw(disp);
and broke 2D virtio-gpu and vgem when ForceSoftware is false. This
commit restores the old behavior.
Fixes: f7e0cdcf1a5 ("egl/surfaceless: simplify dri2_initialize_surfaceless()")
Signed-off-by: Chia-I Wu <olvaffe at gmail.com>
Reviewed-by: Adam Jackson <ajax at redhat.com>
Reviewed-by: Emil Velikov <emil.l.velikov at gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11992>
(cherry picked from commit 384181921c23fc7630b4a7d373f5b6755d20c8af)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=6382db6e80cc382459ac7f85a93c369169f4da8b
Author: Ian Romanick <ian.d.romanick at intel.com>
Date: Tue Jul 20 13:37:31 2021 -0700
nir/gcm: Clear out pass_flags before starting
With this pass enabled in Intel drivers, running shader-db on
shaders/unity/38.shader_test resulted in
Program received signal SIGSEGV, Segmentation fault.
gcm_schedule_early_src (src=0x555555d45348, void_state=0x7fffffffba40) at ../../SOURCE/master/src/compiler/nir/nir_opt_gcm.c:297
297 if (info->early_block->index < src_info->early_block->index)
(gdb) print src_info->early_block
$1 = (nir_block *) 0x0
I tracked this down to an early exit from gcm_schedule_early_instr on
the parent instruction because instr->pass_flags was 0x1c. That
should be an impossible value for this pass, so I inferred that
pass_flags must have dirt left from some previous pass.
Fixes: 8dfe6f672f4 ("nir/GCM: Use pass_flags instead of bitsets for tracking visited/pinned")
Reviewed-by: Timothy Arceri <tarceri at itsqueeze.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/597>
(cherry picked from commit 436668874a1508edf127be16873c6811fe13a3e6)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=761b912788727159756a1627bfe47a0c4fdf326d
Author: Daniel Stone <daniels at collabora.com>
Date: Tue Jul 20 13:54:16 2021 +0100
egl/wayland: Allow EGLSurface to outlive wl_egl_window
According to the EGL spec, it is entirely valid for an EGLSurface to
outlive the native_window it was created from, provided that SwapBuffers
and MakeCurrent return EGL_BAD_NATIVE_WINDOW.
We don't have any facility to error on MakeCurrent, so just make sure we
can bundle on safely through rendering for now, then return
EGL_BAD_NATIVE_WINDOW from SwapBuffers.
Signed-off-by: Daniel Stone <daniels at collabora.com>
Reviewed-by: Simon Ser <contact at emersion.fr>
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/2251
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/4902
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11979>
(cherry picked from commit 6455ab6e5ac5b1ad815d3e67e72d9adef7ffd21c)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=61e28f0faea7ad29eeaa5b9606142c6478094207
Author: Daniel Stone <daniels at collabora.com>
Date: Tue Jul 20 13:40:00 2021 +0100
egl/wayland: Error on invalid native window
We unconditionally require a wl_egl_window to be passed as the native
window type, and do not permit a default window. The spec requires us to
return EGL_BAD_NATIVE_WINDOW when doing this, rather than crashing.
Further, if an EGLSurface has already been created for an existing
native window, we are required to return EGL_BAD_ALLOC.
Signed-off-by: Daniel Stone <daniels at collabora.com>
Reviewed-by: Simon Ser <contact at emersion.fr>
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/2251
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/4902
Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11979>
(cherry picked from commit 0e2464d26bace803aad575c117686412be82388a)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=8c5215e1c84d448e4b6a6f1e129576d4de0aeb75
Author: Juan A. Suarez Romero <jasuarez at igalia.com>
Date: Mon Jul 19 12:18:11 2021 +0200
broadcom/compiler: emit TMU flush before a jump
Like in the case of emitting a block, process pending TMU operations
before a jump is executed.
Fixes dEQP-VK.graphicsfuzz.stable-binarysearch-tree-nested-if-and-conditional.
Fixes: 197090a3fca ("broadcom/compiler: implement pipelining for general
TMU operations")
Reviewed-by: Iago Toral Quiroga <itoral at igalia.com>
Signed-off-by: Juan A. Suarez Romero <jasuarez at igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11971>
(cherry picked from commit dc40157888a8825fe5c411f20029466c7c2f1d47)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=f8dc338c28c76bfc7c8445fa2008673a5475bfb8
Author: Marek Olšák <marek.olsak at amd.com>
Date: Tue Jul 20 05:23:25 2021 -0400
st/mesa: always use PIPE_USAGE_STAGING for GL_MAP_READ_BIT usage
This fixes CPU read performance.
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5091
Cc: mesa-stable at lists.freedesktop.org
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer at amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11974>
(cherry picked from commit 54e1ec017d7852bfc4ba46b54c3ab516f96102f7)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=ba99f8f84e8448fa005b00fb6c45b08230c1d9d2
Author: Olivier Fourdan <ofourdan at redhat.com>
Date: Mon Jul 19 09:29:46 2021 +0200
radeonsi: Check aux_context on si_destroy_screen()
The function radeonsi_screen_create_impl() tries to create the
aux_context but doesn't actually check for the returned value from
si_create_context().
Then, on si_destroy_screen() the aux_context is used without actually
checking whether it's a thing or not.
As a result, if for any reason si_create_context() failed, we shall
crash in si_destroy_screen() with a NULL pointer dereference trying to
access ((struct si_context *)sscreen->aux_context)->log.
Simply check for aux_context not being NULL to avoid that crash.
Cc: mesa-stable
Signed-off-by: Olivier Fourdan <ofourdan at redhat.com>
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer at amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11948>
(cherry picked from commit 5bfd1a7e19ec1c510c2ba4adadad98d78f712eaf)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=b66db988737c8e8d10e7d03973b2f2162a4ce095
Author: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer at amd.com>
Date: Fri Jul 16 18:43:32 2021 +0200
mesa: fix bindless uniform samplers update
According to the comment below some extra magic is needed
for bindless samplers, so don't do an early return in this
case.
Fixes: 736f1f70ab8 ("mesa: skip redundant uniform updates for glUniform")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/4806
Reviewed-by: Marek Olšák <marek.olsak at amd.com>
Reviewed-by: Timothy Arceri <tarceri at itsqueeze.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11930>
(cherry picked from commit 8be61e8a9e49190593e0c90077a8ce100b5743de)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=a3476e3579bbb8b61140a099452d1f9914578a2d
Author: Daniel Schürmann <daniel at schuermann.dev>
Date: Mon Jul 19 15:01:09 2021 +0200
aco/optimizer: ensure to not erase high bits when propagating packed constants
Packed constants with non-zero values in the high half
might have been propagated as 16 bit, dropping the high half.
Cc: mesa-stable
Closes: #5070
Reviewed-by: Rhys Perry <pendingchaos02 at gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11954>
(cherry picked from commit 9b1a296172df38561d650abf98ee2e1c77ac6fc2)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=63a61c1a1436b55786482e239af5593715a074eb
Author: Jose Maria Casanova Crespo <jmcasanova at igalia.com>
Date: Fri Jul 16 14:30:51 2021 +0200
v3d/driconf: Expose non-MSAA texture limits for mutter and gnome-shell
To enable dual 4k displays on mutter or gnome-shell under X11 we need to
expose the non-MSAA texture limit as we did for Xorg at 60a64f028d75ad6
("v3d: Use driconf to expose non-MSAA texture limits for Xorg.")
https://gitlab.gnome.org/GNOME/mutter/-/issues/1874
Cc: 21.1 21.2 mesa-stable
Reviewed-by: Iago Toral Quiroga <itoral at igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11926>
(cherry picked from commit fd5fa73e6c8deeb8e09062e412d7d05fe57d0c59)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=0a2cd1fbb549cab12aea9617b64132615bd317f6
Author: Karol Herbst <kherbst at redhat.com>
Date: Thu May 27 18:59:37 2021 +0200
nv50/ir/nir: fix smem size for GL
Originally I fixed the case where the nir itself has a shared mem size of
0, but the frontend (e.g. clover) set it to some other value.
But st/mesa sets the shared mem size on the state object as well and we
end up actually doubling the value in the driver as we set smemSize to the
value from the state object before calling into the compiler.
So just max the value instead.
Fixes the compute_shader.shared-max CTS test.
Fixes: dc667b1f192 ("nv50/ir/nir: fix smem size")
Signed-off-by: Karol Herbst <kherbst at redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11047>
(cherry picked from commit ff55412f40a7d588e47e2c35d175aea0ac3cfe95)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=8e219e82adfa4a27c4e6f102324d8636172eb15c
Author: Dave Airlie <airlied at redhat.com>
Date: Fri Jul 16 12:28:57 2021 +1000
softpipe/aniso: move DDQ calculation to after scaling.
I've read the papers on EWA filters and it seems like the calculate
DDQ = 2 * A after the scaling of A happens. This seems to make
things less blurry and more like real aniso.
Fixes: 2135aba8 ("softpipe: Constify variables")
Reviewed-by: Timothy Arceri <tarceri at itsqueeze.com>
Reviewed-by: Roland Scheidegger <sroland at vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11917>
(cherry picked from commit 57dcfb4e55346ee2d7e3da79e93a7d36a8a5ca17)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=1ef376e3149200c177a76c4b56653a7efd231ed3
Author: Timothy Arceri <tarceri at itsqueeze.com>
Date: Thu Jul 15 22:28:00 2021 +1000
glsl: replace some C++ code with C
This replaces some new/delete uses with malloc/free.
This is more consistent with most of the other glsl IR code but
more importantly it allows the game "Battle Block Theater" to
start working on some mesa drivers. The game overrides new and
ends up throwing an assert and crashing when it sees this
function calling new [0].
Note: The game still crashes with radeonsi due to similar conflicts
with LLVM.
CC: mesa-stable
Reviewed-by: Ian Romanick <ian.d.romanick at intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11907>
(cherry picked from commit 749251391d967ce5450a6adfeb64cb773cce2508)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=c35cc4cac84806d12fec822010b9d983115ce946
Author: Mike Blumenkrantz <michael.blumenkrantz at gmail.com>
Date: Fri Jul 16 12:50:21 2021 -0400
util/foz: stop crashing on destroy if prepare hasn't been called
Fixes: eca6bb9540d ("util/fossilize_db: add basic fossilize db util to read/write shader caches")
Reviewed-by: Timothy Arceri <tarceri at itsqueeze.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11931>
(cherry picked from commit 2850db0a9ef3751b1b3429d9b7e75ed22dc4e59e)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=c1cac599a3000ee5084a2b1bc4e1f5c66899ffae
Author: Marek Olšák <marek.olsak at amd.com>
Date: Wed Jul 14 08:27:40 2021 -0400
radeonsi: don't expose no-attachment MSAA 16x on all 1 RB chips due to issues
Cc: mesa-stable at lists.freedesktop.org
Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer at amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11900>
(cherry picked from commit 86ff5b7ddb60ee82b04eea143b9cc7efa3ff8183)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=ef5bde7afa0b6085b266ba1c0a6e0f5e507026aa
Author: Jason Ekstrand <jason at jlekstrand.net>
Date: Wed Jul 7 14:21:58 2021 -0500
anv: Handle errors properly in anv_i915_query
DRM_IOCTL_I915_QUERY is a multi-query. The most egregious errors are
returned via the usual ioctl error mechanism but there are also
per-query errors that are indicated by item.length < 0. We need to
handle those as well. While we're at it, scrape errno so we can return
a proper integer error.
Fixes: c0d07c838a9f "anv: Support i915 query (DRM_IOCTL_I915_QUERY)..."
Reviewed-by: Paulo Zanoni <paulo.r.zanoni at intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11770>
(cherry picked from commit b664481ba91603d69b294ca5b1243e42c86a5863)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=a9faea77199600231de35f444cd405f8d3d69d01
Author: Jason Ekstrand <jason at jlekstrand.net>
Date: Tue May 4 20:24:54 2021 -0500
iris: Re-emit MEDIA_VFE_STATE for variable group size shaders
It implicitly contains the number of threads via the CURBE allocation
size field.
Fixes: 33c61eb2f105 "iris: Implement ARB_compute_variable_group_size"
Reviewed-by: Kenneth Graunke <kenneth at whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10640>
(cherry picked from commit e56d5db341eaf2112be366ee015a9956b2c3b704)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=18f5baef127f996f92b4bb30a619d91b2120ce1b
Author: Dave Airlie <airlied at redhat.com>
Date: Thu Jul 15 15:25:24 2021 +1000
draw: handle resetting draw_id between instances.
If we have a multi draw then the draw_id needs to get reset
between each instance.
Fixes the rest of the VK CTS multidraw tests
Fixes: ce0b1f23c30e ("aux/draw: track increment_draw_id value from draw info")
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz at gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11899>
(cherry picked from commit b3d2642ca4fe10faf05d51962339a598e622028e)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=38ebba486c3157cec11eab91bdbe01ff99ec8d78
Author: Samuel Pitoiset <samuel.pitoiset at gmail.com>
Date: Wed Jul 14 20:05:54 2021 +0200
radv: only init the TC-compat ZRANGE metadata for the depth aspect
With separate depth/stencil layouts, if the depth aspect is first
initialized and then cleared, the ZRANGE_PRECISION metadata might
be different than 0. Initializing it again for the stencil aspect
will overwrite the value.
Fixes rendering glitches with Scarlet Nexus on GFX8-9.
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5052
Cc: 21.1 21.2 mesa-stable
Signed-off-by: Samuel Pitoiset <samuel.pitoiset at gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas at basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11883>
(cherry picked from commit 1ea156f44c7ce8d6880217ef4d46692cc7ac0219)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=3fbb161e5bc3e0b2e43129ec551922b5766894c4
Author: Vasily Khoruzhick <anarsoul at gmail.com>
Date: Tue Jul 13 23:12:13 2021 -0700
lima: handle fp16 vertex formats
12128fb1351e marked fp16 vertex formats supported, but they aren't actually handled
by lima_pipe_format_to_attrib_type(). Fix it by handling it there.
FP16 seems to be the only missing index which is 0x3.
Fixes: 12128fb1351e ("lima: add natively supported vertex buffer formats")
Cc: 21.1 mesa-stable
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig at collabora.com>
Signed-off-by: Vasily Khoruzhick <anarsoul at gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11872>
(cherry picked from commit 4a3269dff65e198ac3d493b676800978ce0c01e5)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=463607c871803b84c82933566f5e03309f82d7bd
Author: Joshua Ashton <joshua at froggi.es>
Date: Wed Jul 14 03:21:11 2021 +0100
lavapipe: Use common Vulkan format helpers
Drops the vk_format_to_pipe (and it's outdated table) for vk_format_to_pipe_format, aswell as the duplicated vk_format_aspects function.
The old format table was missing USCALED and other values, causing incorrect rendering in many games.
Fixes rendering in Portal 1, Hat in Time, Half-Life 2 and pretty much every other D3D9 title with DXVK.
Fixes: b38879f8c5f ("vallium: initial import of the vulkan frontend")
Signed-off-by: Joshua Ashton <joshua at froggi.es>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz at gmail.com>
Reviewed-by: Dave Airlie <airlied at redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11863>
(cherry picked from commit 1744372714421d741f256f99148e792796888562)
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=d870091d8e33424be8ccacb792fca56bde6cad90
Author: Eric Engestrom <eric at engestrom.ch>
Date: Thu Jul 22 21:53:03 2021 +0100
.pick_status.json: Update to 1d6c790f9f6c373e667de955822b1048c7073ba3
URL: http://cgit.freedesktop.org/mesa/mesa/commit/?id=6f7aa6f8c846a59c43dac47bb8e2b78d11af8bc4
Author: Rob Clark <robdclark at chromium.org>
Date: Wed May 12 10:15:03 2021 -0700
freedreno/ir3: Don't force RTNE if rounding mode is undefined
Forcing round-to-nearest-even results in loss of opportunities for
conversion folding, causing a regression in gfxbench gl_alu2.
Fixes: de195671bdb ("ir3: nir_op_f2f16 should round to even")
Signed-off-by: Rob Clark <robdclark at chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10773>
(cherry picked from commit 6c530ebf407d2a165c4252b3af609fb426458ad0)
More information about the mesa-commit
mailing list