On 05/23/2018 12:31 AM, Christian König wrote:
> Am 23.05.2018 um 00:31 schrieb Deepak Sharma:
>> When vram size <= THRESHOLD(256M) lets use GTT for dumb buffer
>> allocation. As SG will be enabled with vram size <= 256M
>> scan out will not be an issue.
>>
>> v2: Use amdgpu_display_supported_domains to get supported domain.
>>
>> Signed-off-by: Deepak Sharma <Deepak.Sharma at amd.com>
>> ---
>> Â drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 10 ++++++++--
>> Â 1 file changed, 8 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
>> b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
>> index 2c8e27370284..63758db5e2ea 100644
>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
>> @@ -30,6 +30,7 @@
>> Â #include <drm/drmP.h>
>> Â #include <drm/amdgpu_drm.h>
>> Â #include "amdgpu.h"
>> +#include "amdgpu_display.h"
>> Â void amdgpu_gem_object_free(struct drm_gem_object *gobj)
>> Â {
>> @@ -749,15 +750,20 @@ int amdgpu_mode_dumb_create(struct drm_file
>> *file_priv,
>> Â Â Â Â Â struct amdgpu_device *adev = dev->dev_private;
>> Â Â Â Â Â struct drm_gem_object *gobj;
>> Â Â Â Â Â uint32_t handle;
>> +Â Â Â u32 domain = amdgpu_display_supported_domains(adev);
>> Â Â Â Â Â int r;
>> Â Â Â Â Â args->pitch = amdgpu_align_pitch(adev, args->width,
>> Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â DIV_ROUND_UP(args->bpp, 8), 0);
>> Â Â Â Â Â args->size = (u64)args->pitch * args->height;
>> Â Â Â Â Â args->size = ALIGN(args->size, PAGE_SIZE);
>> +Â Â Â if (domain == (AMDGPU_GEM_DOMAIN_VRAM | AMDGPU_GEM_DOMAIN_GTT)) {
>> +Â Â Â Â Â Â Â domain = AMDGPU_GEM_DOMAIN_VRAM;
>> +Â Â Â Â Â Â Â if (adev->gmc.real_vram_size <= AMDGPU_SG_THRESHOLD)
>> +Â Â Â Â Â Â Â Â Â Â Â domain = AMDGPU_GEM_DOMAIN_GTT;
>> +Â Â Â }
>
> As Michel suggested as well better make that an helper in amdgpu_object.c.
>
> Apart from that looks good to me,
> Christian.
I can push another patch to create helper function and use that here as
well as in amdgpu_bo_pin_restricted.
otherwise I can push v3 if you want that change in this patch itself.
Thanks,
Deepak
>
>> -Â Â Â r = amdgpu_gem_object_create(adev, args->size, 0,
>> -Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â AMDGPU_GEM_DOMAIN_VRAM,
>> +Â Â Â r = amdgpu_gem_object_create(adev, args->size, 0, domain,
>> Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED,
>> Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â false, NULL, &gobj);
>> Â Â Â Â Â if (r)
>
