diff options
| author | Marek Olšák <marek.olsak@amd.com> | 2018-04-03 20:05:03 +0300 |
|---|---|---|
| committer | Alex Deucher <alexander.deucher@amd.com> | 2018-05-15 21:43:32 +0300 |
| commit | d240cd9eddd943dbe0267d081697195ff1e90b65 (patch) | |
| tree | ff9648c93738e223b8da42489f5d64aa8caa9895 /drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c | |
| parent | 3f188453faf7ba5b59e8064df4afffbc946e25ec (diff) | |
| download | linux-d240cd9eddd943dbe0267d081697195ff1e90b65.tar.xz | |
drm/amdgpu: optionally do a writeback but don't invalidate TC for IB fences
There is a new IB flag that enables this new behavior.
Full invalidation is unnecessary for RELEASE_MEM and doesn't make sense
when draw calls from two adjacent gfx IBs run in parallel. This will be
the new default for Mesa.
v2: bump the version
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Diffstat (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c')
| -rw-r--r-- | drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c | 8 |
1 files changed, 6 insertions, 2 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c index 311589e02d17..f70eeed9ed76 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c @@ -127,6 +127,7 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, struct amdgpu_vm *vm; uint64_t fence_ctx; uint32_t status = 0, alloc_size; + unsigned fence_flags = 0; unsigned i; int r = 0; @@ -227,7 +228,10 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, #endif amdgpu_asic_invalidate_hdp(adev, ring); - r = amdgpu_fence_emit(ring, f); + if (ib->flags & AMDGPU_IB_FLAG_TC_WB_NOT_INVALIDATE) + fence_flags |= AMDGPU_FENCE_FLAG_TC_WB_ONLY; + + r = amdgpu_fence_emit(ring, f, fence_flags); if (r) { dev_err(adev->dev, "failed to emit fence (%d)\n", r); if (job && job->vmid) @@ -242,7 +246,7 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, /* wrap the last IB with fence */ if (job && job->uf_addr) { amdgpu_ring_emit_fence(ring, job->uf_addr, job->uf_sequence, - AMDGPU_FENCE_FLAG_64BIT); + fence_flags | AMDGPU_FENCE_FLAG_64BIT); } if (patch_offset != ~0 && ring->funcs->patch_cond_exec) |