Re: [PATCH 7/7 v2] drm/radeon: Refuse to migrate a prime BO to VRAM. (v2)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



I also haven't tested it on hardware (all my hardware seems to be supported by amdgpu), but the only times we touch prime_shared_count are:
1) It's set to 1 in radeon_gem_prime_import_sg_table. This shouldn't be called in the self-import case (and isn't, as far as I can tell), and
2) It's incremented in radeon_gem_prime_pin, right after the bo is pinned to GTT, so cannot be migrated to VRAM anyway (unless there's an unmatched radeon_bo_unpin call somewhere else?).

On Tue, 4 Apr 2017 at 02:30 Christian König <deathsimple@xxxxxxxxxxx> wrote:
Am 03.04.2017 um 17:51 schrieb Alex Deucher:
> On Mon, Apr 3, 2017 at 4:24 AM, Christian König <deathsimple@xxxxxxxxxxx> wrote:
>> Am 03.04.2017 um 05:35 schrieb raof@xxxxxxxxxx:
>>> From: Christopher James Halse Rogers
>>> <christopher.halse.rogers@xxxxxxxxxxxxx>
>>>
>>> BOs shared via dma-buf, either imported or exported, cannot sensibly be
>>> migrated to VRAM
>>> without breaking the dma-buf sharing. Refuse userspace requests to migrate
>>> to VRAM,
>>> ensure such BOs are not migrated during command submission, and refuse to
>>> pin them
>>> to VRAM.
>>>
>>> v2: Don't pin BOs in GTT. Instead, refuse to migrate BOs to VRAM.
>>>
>>> Signed-off-by: Christopher James Halse Rogers
>>> <christopher.halse.rogers@xxxxxxxxxxxxx>
>>
>> Patches #4, #6 and this one (#7) are Reviewed-by: Christian König
>> <christian.koenig@xxxxxxx>.
> Will this cause any issues with self sharing?  E.g., the standard use
> case with a single GPU and DRI3?

No, the code from amdgpu to count the number of times we really shared a
BO with somebody was ported over to radeon as far as I can see.

So that should work fine, but I haven't double checked or tested it.

Christian.

>
> Alex
>
>>
>>> ---
>>>    drivers/gpu/drm/radeon/radeon_cs.c     | 10 ++++++++++
>>>    drivers/gpu/drm/radeon/radeon_gem.c    |  4 ++++
>>>    drivers/gpu/drm/radeon/radeon_object.c |  5 +++++
>>>    3 files changed, 19 insertions(+)
>>>
>>> diff --git a/drivers/gpu/drm/radeon/radeon_cs.c
>>> b/drivers/gpu/drm/radeon/radeon_cs.c
>>> index a8442f7196d6..df6b58c08544 100644
>>> --- a/drivers/gpu/drm/radeon/radeon_cs.c
>>> +++ b/drivers/gpu/drm/radeon/radeon_cs.c
>>> @@ -164,6 +164,16 @@ static int radeon_cs_parser_relocs(struct
>>> radeon_cs_parser *p)
>>>                          p->relocs[i].allowed_domains = domain;
>>>                  }
>>>    +             /* Objects shared as dma-bufs cannot be moved to VRAM */
>>> +               if (p->relocs[i].robj->prime_shared_count) {
>>> +                       p->relocs[i].allowed_domains &=
>>> ~RADEON_GEM_DOMAIN_VRAM;
>>> +                       if (!p->relocs[i].allowed_domains) {
>>> +                               DRM_ERROR("BO associated with dma-buf
>>> cannot "
>>> +                                         "be moved to VRAM\n");
>>> +                               return -EINVAL;
>>> +                       }
>>> +               }
>>> +
>>>                  p->relocs[i].tv.bo = &p->relocs[i].robj->tbo;
>>>                  p->relocs[i].tv.shared = !r->write_domain;
>>>    diff --git a/drivers/gpu/drm/radeon/radeon_gem.c
>>> b/drivers/gpu/drm/radeon/radeon_gem.c
>>> index 96683f5b2b1b..067ea7d7658a 100644
>>> --- a/drivers/gpu/drm/radeon/radeon_gem.c
>>> +++ b/drivers/gpu/drm/radeon/radeon_gem.c
>>> @@ -120,6 +120,10 @@ static int radeon_gem_set_domain(struct
>>> drm_gem_object *gobj,
>>>                          return r;
>>>                  }
>>>          }
>>> +       if (domain == RADEON_GEM_DOMAIN_VRAM && robj->prime_shared_count)
>>> {
>>> +               /* A BO that is associated with a dma-buf cannot be
>>> sensibly migrated to VRAM */
>>> +               return -EINVAL;
>>> +       }
>>>          return 0;
>>>    }
>>>    diff --git a/drivers/gpu/drm/radeon/radeon_object.c
>>> b/drivers/gpu/drm/radeon/radeon_object.c
>>> index 74b276060c20..bec2ec056de4 100644
>>> --- a/drivers/gpu/drm/radeon/radeon_object.c
>>> +++ b/drivers/gpu/drm/radeon/radeon_object.c
>>> @@ -352,6 +352,11 @@ int radeon_bo_pin_restricted(struct radeon_bo *bo,
>>> u32 domain, u64 max_offset,
>>>                  return 0;
>>>          }
>>> +       if (bo->prime_shared_count && domain == RADEON_GEM_DOMAIN_VRAM) {
>>> +               /* A BO shared as a dma-buf cannot be sensibly migrated to
>>> VRAM */
>>> +               return -EINVAL;
>>> +       }
>>> +
>>>          radeon_ttm_placement_from_domain(bo, domain);
>>>          for (i = 0; i < bo->placement.num_placement; i++) {
>>>                  /* force to pin into visible video ram */
>>
>>
>> _______________________________________________
>> amd-gfx mailing list
>> amd-gfx@xxxxxxxxxxxxxxxxxxxxx
>> https://lists.freedesktop.org/mailman/listinfo/amd-gfx
> _______________________________________________
> amd-gfx mailing list
> amd-gfx@xxxxxxxxxxxxxxxxxxxxx
> https://lists.freedesktop.org/mailman/listinfo/amd-gfx


_______________________________________________
dri-devel mailing list
dri-devel@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/dri-devel

[Index of Archives]     [Linux DRI Users]     [Linux Intel Graphics]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [XFree86]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [XFree86]
  Powered by Linux