]> git.baikalelectronics.ru Git - kernel.git/commit
drm/radeon: Override fpfn for all VRAM placements in radeon_evict_flags
authorMichel Dänzer <michel.daenzer@amd.com>
Fri, 24 Mar 2017 10:01:09 +0000 (19:01 +0900)
committerAlex Deucher <alexander.deucher@amd.com>
Mon, 27 Mar 2017 20:17:30 +0000 (16:17 -0400)
commit80fcc19b8e25745d817c6d506c25f75a1402b269
tree992a2ba61ad8a19f91314e09b3cfad6bd40690e2
parentb411336e6cc265065dee3385d5b3df547695c48d
drm/radeon: Override fpfn for all VRAM placements in radeon_evict_flags

We were accidentally only overriding the first VRAM placement. For BOs
with the RADEON_GEM_NO_CPU_ACCESS flag set,
radeon_ttm_placement_from_domain creates a second VRAM placment with
fpfn == 0. If VRAM is almost full, the first VRAM placement with
fpfn > 0 may not work, but the second one with fpfn == 0 always will
(the BO's current location trivially satisfies it). Because "moving"
the BO to its current location puts it back on the LRU list, this
results in an infinite loop.

Fixes: 0187914a6463 ("drm/radeon: Try evicting from CPU accessible to
                      inaccessible VRAM first")
Reported-by: Zachary Michaels <zmichaels@oblong.com>
Reported-and-Tested-by: Julien Isorce <jisorce@oblong.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
drivers/gpu/drm/radeon/radeon_ttm.c