kernel/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c, branch linux-4.17.y

drm/amdgpu: Make amdgpu_vram_mgr_bo_invisible_size always accurate

2018-07-08T13:32:19Z

commit 7303b39e46b2f523334591f05fd9566cf929eb26 upstream. Even BOs with AMDGPU_GEM_CREATE_NO_CPU_ACCESS may end up at least partially in CPU visible VRAM, in particular when all VRAM is visible. v2: * Don't take VRAM mgr spinlock, not needed (Christian König) * Make loop logic simpler and clearer. Cc: stable@vger.kernel.org Signed-off-by: Michel Dänzer Reviewed-by: Christian König Signed-off-by: Alex Deucher Signed-off-by: Greg Kroah-Hartman

drm/amdgpu: Refactor amdgpu_vram_mgr_bo_invisible_size helper

2018-07-08T13:32:19Z

commit 5e9244ff585239630f15f8ad8e676bc91a94ca9e upstream. Preparation for the following fix, no functional change intended. Cc: stable@vger.kernel.org Signed-off-by: Michel Dänzer Reviewed-by: Christian König Signed-off-by: Alex Deucher Signed-off-by: Greg Kroah-Hartman

drm/amdgpu: Use kvmalloc_array for allocating VRAM manager nodes array

2018-07-08T13:32:19Z

commit 6fa39bc1e01dab8b4f54b23e95a181a2ed5a2d38 upstream. It can be quite big, and there's no need for it to be physically contiguous. This is less likely to fail under memory pressure (has actually happened while running piglit). Cc: stable@vger.kernel.org Signed-off-by: Michel Dänzer Reviewed-by: Christian König Signed-off-by: Alex Deucher Signed-off-by: Greg Kroah-Hartman

drm/amdgpu: move struct amdgpu_mc into amdgpu_gmc.h

2018-02-19T19:17:43Z

And rename it to amdgpu_gmc as well. Signed-off-by: Christian König Reviewed-by: Samuel Li Signed-off-by: Alex Deucher

drm/amdgpu:fix memleak in takedown

2017-11-16T18:35:48Z

this can fix the memory leak under the case that not all BO are freed during "takedown" stage, because originally it blocks following kfree on mgr. Signed-off-by: Monk Liu Reviewed-by: Christian König Signed-off-by: Alex Deucher

drm/amdgpu: move vram usage tracking into the vram manager v2

2017-08-17T19:46:03Z

Looks like a better place for this. v2: use atomic64_t members instead Signed-off-by: Christian König Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher

drm/amdgpu: move debug print into the MM managers

2017-08-17T19:46:01Z

Instead of the separate switch/case in the calling function. Signed-off-by: Christian König Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher

drm/ttm: make ttm_mem_type_manager_func debug more useful

2017-08-17T19:45:59Z

Provide the drm printer directly instead of just the callback. Signed-off-by: Christian König Reviewed-by: Alex Deucher Signed-off-by: Alex Deucher

drm/amdgpu: Use designated initializers

2017-05-28T17:23:00Z

The randstruct plugin requires structures that are entirely function pointers be initialized using designated initializers. Cc: Alex Deucher Cc: Christian König Signed-off-by: Kees Cook

drm/amdgpu: use TTM_PL_FLAG_CONTIGUOUS v2

2017-04-05T03:34:27Z

Implement AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS using TTM_PL_FLAG_CONTIGUOUS instead of a placement limit. That allows us to better handle CPU accessible placements. v2: prevent virtual BO start address from overflowing Signed-off-by: Christian König Acked-by: Michel Dänzer Reviewed-by: Nicolai Hähnle Signed-off-by: Alex Deucher