fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2026-05-08 04:48:08 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

intel/compiler/gfx12.5+: Lower 64-bit cluster_broadcast with 32-bit ops

Browse source 
For MTL (verx10 == 125), float64 is supported, but int64 is not.
Therefore we need to lower cluster broadcast using 32-bit int ops.

For gfx12.5+ platforms that support int64, the register regions
used by cluster broadcast aren't supported by the 64-bit pipeline.

On MTL, dEQP-VK.subgroups.clustered.*_double* and
dEQP-VK.subgroups.clustered.*_dvec* were failing to validate the
compiled shader in debug mode, and reportedly gpu-hanging in release
mode.

With this change dEQP-VK.subgroups.clustered.*_double* passed all 48
tests and dEQP-VK.subgroups.clustered.*_dvec* passed all 140 tests on
MTL.

Rework:
 * Move from generator to brw_fs_lower_regioning.cpp. (Suggested by
   Francisco)
 * Apply to verx10 >= 125.. (Suggested by Francisco)

Cc: 23.1 <mesa-stable>
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Marcin Ślusarz <marcin.slusarz@intel.com> (v1)
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/22569>
(cherry picked from commit fcb72ffd0c)
This commit is contained in:
Jordan Justen 2023-04-18 20:11:41 -04:00 • committed by Eric Engestrom
parent 38da1ae68a
commit b8a7fba41f
2 changed files with 9 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
.pick_status.json
View file

@ -544,7 +544,7 @@
"description": "intel/compiler/gfx12.5+: Lower 64-bit cluster_broadcast with 32-bit ops",
"nominated": true,
"nomination_type": 0,
"resolution": 0,
"resolution": 1,
"main_sha": null,
"because_sha": null
},

9
src/intel/compiler/brw_fs_lower_regioning.cpp
View file

@ -174,10 +174,17 @@ namespace {
* integer DWord multiply, indirect addressing must not be
* used."
*
* For MTL (verx10 == 125), float64 is supported, but int64 is not.
* Therefore we need to lower cluster broadcast using 32-bit int ops.
*
* For gfx12.5+ platforms that support int64, the register regions
* used by cluster broadcast aren't supported by the 64-bit pipeline.
*
* Work around the above and handle platforms that don't
* support 64-bit types at all.
*/
if ((!has_64bit || devinfo->platform == INTEL_PLATFORM_CHV ||
if ((!has_64bit || devinfo->verx10 >= 125 ||
devinfo->platform == INTEL_PLATFORM_CHV ||
intel_device_info_is_9lp(devinfo)) && type_sz(t) > 4)
return BRW_REGISTER_TYPE_UD;
else