KonstantinSeurer
/
mesa
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

intel/perf: fix perf equation subslice mask generation for gfx12+ 

v2: Fix comment change (Marcin)

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Cc: <mesa-stable@lists.freedesktop.org>
Reviewed-by: Francisco Jerez <currojerez@riseup.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10015>
This commit is contained in:
Lionel Landwerlin 2021-04-03 13:34:11 +03:00 committed by Marge Bot
parent a543a94404
commit 67619d8153
1 changed files with 3 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
src/intel/perf/intel_perf.c
View File

@ -364,15 +364,15 @@ compute_topology_builtins(struct intel_perf_config *perf,
perf->sys_vars.eu_threads_count = devinfo->num_thread_per_eu;
/* The subslice mask builtin contains bits for all slices. Prior to Gfx11
* it had groups of 3bits for each slice, on Gfx11 it's 8bits for each
* slice.
* it had groups of 3bits for each slice, on Gfx11 and above it's 8bits for
* each slice.
*
* Ideally equations would be updated to have a slice/subslice query
* function/operator.
*/
perf->sys_vars.subslice_mask = 0;
int bits_per_subslice = devinfo->ver == 11 ? 8 : 3;
int bits_per_subslice = devinfo->ver >= 11 ? 8 : 3;
for (int s = 0; s < util_last_bit(devinfo->slice_masks); s++) {
for (int ss = 0; ss < (devinfo->subslice_slice_stride * 8); ss++) {