In 2014 (commit 02ca66fbc3e2b272), I changed the program cache code
to use a persistent mapping for the cache BO.
For some reason, I used drm_intel_gem_bo_unsynchronized. This is
rather pointless - we're mapping a freshly allocated, idle BO, so
a regular mapping won't stall. I have no idea why I did that.
drm_intel_gem_bo_unsynchronized does a GTT mapping, though, which
is also unnecessary - on LLC platforms, we can map linear buffers
via the CPU and the LLC ensures coherency.
Switch to regular drm_intel_bo_map().
Signed-off-by: Kenneth Graunke <kenn...@whitecape.org>
---
src/mesa/drivers/dri/i965/brw_program_cache.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/mesa/drivers/dri/i965/brw_program_cache.c
b/src/mesa/drivers/dri/i965/brw_program_cache.c
index 4d249ba6f93..0c1da172588 100644
--- a/src/mesa/drivers/dri/i965/brw_program_cache.c
+++ b/src/mesa/drivers/dri/i965/brw_program_cache.c
@@ -214,7 +214,7 @@ brw_cache_new_bo(struct brw_cache *cache, uint32_t new_size)
new_bo = drm_intel_bo_alloc(brw->bufmgr, "program cache", new_size, 64);
if (brw->has_llc)
- drm_intel_gem_bo_map_unsynchronized(new_bo);
+ drm_intel_bo_map(new_bo, true);
/* Copy any existing data that needs to be saved. */
if (cache->next_offset != 0) {
@@ -405,7 +405,7 @@ brw_init_caches(struct brw_context *brw)
cache->bo = drm_intel_bo_alloc(brw->bufmgr, "program cache", 4096, 64);
if (brw->has_llc)
- drm_intel_gem_bo_map_unsynchronized(cache->bo);
+ drm_intel_bo_map(cache->bo, true);
}
static void
--
2.11.0
_______________________________________________
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
