st/mesa: remove random L3 pinning heuristic for glthread

This is not very effective. A better solution will be added to glthread.

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Acked-by: Jose Fonseca <jfonseca@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7054>
This commit is contained in:
Marek Olšák
2020-10-06 21:37:01 -04:00
committed by Marge Bot
parent 96d9f7761d
commit 3433d193e7
3 changed files with 0 additions and 62 deletions

View File

@@ -821,17 +821,6 @@ st_start_thread(struct st_context_iface *stctxi)
struct st_context *st = (struct st_context *) stctxi;
_mesa_glthread_init(st->ctx);
/* Pin all driver threads to one L3 cache for optimal performance
* on AMD Zen. This is only done if glthread is enabled.
*
* If glthread is disabled, st_draw.c re-pins driver threads regularly
* based on the location of the app thread.
*/
struct glthread_state *glthread = &st->ctx->GLThread;
if (glthread->enabled && st->pipe->set_context_param) {
util_pin_driver_threads_to_random_L3(st->pipe, &glthread->queue.threads[0]);
}
}