diff options
| author | Dave Airlie <airlied@redhat.com> | 2020-09-23 09:25:17 +1000 |
|---|---|---|
| committer | Dave Airlie <airlied@redhat.com> | 2020-09-23 09:25:18 +1000 |
| commit | fc88fef916e8971eefeacc62241b7408b7e7939d (patch) | |
| tree | c2647e7388a752ffc1f9d12b5aaf9a83a69b5ba0 /drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | |
| parent | db29dc7d3346328b3fc61057d224a242f9928289 (diff) | |
| parent | 911d5bd5e7b8531b39301c2c27e5b90d7bd71b88 (diff) | |
| download | linux-fc88fef916e8971eefeacc62241b7408b7e7939d.tar.gz linux-fc88fef916e8971eefeacc62241b7408b7e7939d.tar.bz2 linux-fc88fef916e8971eefeacc62241b7408b7e7939d.zip | |
Merge tag 'amd-drm-next-5.10-2020-09-18' of git://people.freedesktop.org/~agd5f/linux into drm-next
amd-drm-next-5.10-2020-09-18:
amdgpu:
- Support for PCIe DPC recovery
- Sienna Cichlid updates
- Navy Flounder updates
- RAS fixes
- Refactor DC interrupt handling
- Display fixes
- Fix issues with OLED panels
- Mclk fixes for navi1x
- Watermark fixes for renoir and raven2
- Misc code cleanups
- Misc bug fixes
amdkfd:
- Fix a memory leak
- Fix a crach in GPU reset
- Add process eviction counters
radeon:
- expose sclk via sysfs hwmon interface
- Revert bad PLL fix
scheduler:
- Kernel doc fixes
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexdeucher@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200918204322.3931-1-alexander.deucher@amd.com
Diffstat (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c')
| -rw-r--r-- | drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c | 8 |
1 files changed, 7 insertions, 1 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index a18dc878339a..4a85f8cedd77 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -1021,6 +1021,7 @@ static int amdgpu_ttm_tt_pin_userptr(struct ttm_tt *ttm) release_sg: kfree(ttm->sg); + ttm->sg = NULL; return r; } @@ -1155,7 +1156,12 @@ static int amdgpu_ttm_backend_bind(struct ttm_tt *ttm, } /** - * amdgpu_ttm_alloc_gart - Allocate GART memory for buffer object + * amdgpu_ttm_alloc_gart - Make sure buffer object is accessible either + * through AGP or GART aperture. + * + * If bo is accessible through AGP aperture, then use AGP aperture + * to access bo; otherwise allocate logical space in GART aperture + * and map bo to GART aperture. */ int amdgpu_ttm_alloc_gart(struct ttm_buffer_object *bo) { |