radeonsi: Don't use compute queue with regalloc hang bug

It already didn't use compute queues on GFX6, but some GFX7
chips are also affected by the same bug.

Compute queues may run compute dispatches in parallel with
the graphics queue, even from other processes/apps.
At the moment we don't have a way to restrict all workgroups
to 256 invocations, so instead let's make sure not to use the
compute queue.

Cc: mesa-stable
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/37885>
This commit is contained in:
Timur Kristóf 2025-10-15 02:43:09 +02:00 committed by Marge Bot
parent 15d375dc6e
commit 765a748840

View file

@ -526,7 +526,7 @@ static struct pipe_context *si_create_context(struct pipe_screen *screen, unsign
return NULL;
}
sctx->is_gfx_queue = sscreen->info.gfx_level == GFX6 ||
sctx->is_gfx_queue = sscreen->info.has_cs_regalloc_hang_bug ||
/* Compute queues hang on Raven and derivatives, see:
* https://gitlab.freedesktop.org/mesa/mesa/-/issues/12310 */
((sscreen->info.family == CHIP_RAVEN ||