Kenneth Graunke cb756ae8a2 brw: Don't rely on SIMD splitting in opt_combine_convergent_txfs ... The SIMD splitting pass does not handle wide force_writemask_all instructions correctly at the moment. For example, a SIMD32 TXF on pre-Xe2 would get split to a pair of SIMD16. But it will set the groups to operate on channels 15:0 and 31:16. That's not what we want for a NoMask instruction - both should be 15:0, i.e. bld.group(inst->exec_size, 0). We could (and perhaps should) fix the SIMD splitting pass to handle this, but the pass already has subtle complexity in which builders are used. Or we could alter fs_builder::group(), but that has broader implications. As a stop-gap, just make opt_combine_covergent_txfs stop relying on SIMD splitting. It's trivial to do and fixes the issue without risking other breakage. Fixes: 6341b3cd87 ("brw: Combine convergent texture buffer fetches into fewer loads") Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32714>		2024-12-19 23:16:12 +00:00
..
blorp	blorp: use 2D dimension for 1D tiled images	2024-12-12 17:10:45 +00:00
ci	ci/lava: Use CI_JOB_TIMEOUT instead of separate variable	2024-12-18 09:23:27 +00:00
common	intel/common: Enable mi_builder test for PTL	2024-11-20 01:47:04 +00:00
compiler	brw: Don't rely on SIMD splitting in opt_combine_convergent_txfs	2024-12-19 23:16:12 +00:00
decoder	intel/decoder: fix COMPUTE_WALKER handling	2024-12-12 07:35:18 +00:00
dev	intel/dev: update mesa_defs.json from internal database	2024-12-11 17:32:52 +00:00
ds	anv: add BVH building tracking through u_trace	2024-12-09 14:45:00 +00:00
executor	intel/executor: Add example using scalar register and send gather	2024-12-13 02:18:15 +00:00
genxml	intel/genxml/xe2: Add STATE_SYSTEM_MEM_FENCE_ADDRESS instruction	2024-12-18 17:16:05 +00:00
isl	intel/isl: Fix DecompressInL3 assignment on gfx12.5	2024-12-03 22:54:34 +00:00
nullhw-layer	build: pass licensing information in SPDX form	2024-06-29 12:42:49 -07:00
perf	anv: Protect memcpy/memset/qsort calls against NULL arguments	2024-11-14 01:05:01 +00:00
shaders	anv: implement Wa_16011107343/22018402687 for generated draws	2024-11-12 22:48:39 +00:00
tools	intel/hang_replay: remove EXEC_OBJECT_WRITE	2024-10-04 21:37:41 +00:00
vulkan	anv: use vk_descriptor_type_is_dynamic	2024-12-19 15:12:58 +00:00
vulkan_hasvk	hasvk: use vk_descriptor_type_is_dynamic	2024-12-19 15:12:58 +00:00
meson.build	intel: Add executor tool	2024-08-14 03:03:46 +00:00