diff options
| author | Marek Olšák <marek.olsak@amd.com> | 2020-04-27 22:04:47 +0200 |
|---|---|---|
| committer | Alex Deucher <alexander.deucher@amd.com> | 2020-05-18 17:24:21 +0200 |
| commit | d35745bbec0958ec09a6580dea537828a571c632 (patch) | |
| tree | 0141e201d095ccb8f2fc8f35698d15106fcb4444 /drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c | |
| parent | drm/amdgpu: Add mem_sync implementation for all the ASICs. (diff) | |
| download | linux-d35745bbec0958ec09a6580dea537828a571c632.tar.xz linux-d35745bbec0958ec09a6580dea537828a571c632.zip | |
drm/amdgpu: apply AMDGPU_IB_FLAG_EMIT_MEM_SYNC to compute IBs too (v3)
Compute IBs need this too.
v2: split out version bump
v3: squash in emit frame count fixes
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/gfx_v7_0.c')
| -rw-r--r-- | drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c | 18 |
1 files changed, 17 insertions, 1 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c index b2f10e39eff1..0cc011f9190d 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c @@ -5010,6 +5010,20 @@ static void gfx_v7_0_emit_mem_sync(struct amdgpu_ring *ring) amdgpu_ring_write(ring, 0x0000000A); /* poll interval */ } +static void gfx_v7_0_emit_mem_sync_compute(struct amdgpu_ring *ring) +{ + amdgpu_ring_write(ring, PACKET3(PACKET3_ACQUIRE_MEM, 5)); + amdgpu_ring_write(ring, PACKET3_TCL1_ACTION_ENA | + PACKET3_TC_ACTION_ENA | + PACKET3_SH_KCACHE_ACTION_ENA | + PACKET3_SH_ICACHE_ACTION_ENA); /* CP_COHER_CNTL */ + amdgpu_ring_write(ring, 0xffffffff); /* CP_COHER_SIZE */ + amdgpu_ring_write(ring, 0xff); /* CP_COHER_SIZE_HI */ + amdgpu_ring_write(ring, 0); /* CP_COHER_BASE */ + amdgpu_ring_write(ring, 0); /* CP_COHER_BASE_HI */ + amdgpu_ring_write(ring, 0x0000000A); /* poll interval */ +} + static const struct amd_ip_funcs gfx_v7_0_ip_funcs = { .name = "gfx_v7_0", .early_init = gfx_v7_0_early_init, @@ -5075,7 +5089,8 @@ static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_compute = { 5 + /* hdp invalidate */ 7 + /* gfx_v7_0_ring_emit_pipeline_sync */ CIK_FLUSH_GPU_TLB_NUM_WREG * 5 + 7 + /* gfx_v7_0_ring_emit_vm_flush */ - 7 + 7 + 7, /* gfx_v7_0_ring_emit_fence_compute x3 for user fence, vm fence */ + 7 + 7 + 7 + /* gfx_v7_0_ring_emit_fence_compute x3 for user fence, vm fence */ + 7, /* gfx_v7_0_emit_mem_sync_compute */ .emit_ib_size = 7, /* gfx_v7_0_ring_emit_ib_compute */ .emit_ib = gfx_v7_0_ring_emit_ib_compute, .emit_fence = gfx_v7_0_ring_emit_fence_compute, @@ -5088,6 +5103,7 @@ static const struct amdgpu_ring_funcs gfx_v7_0_ring_funcs_compute = { .insert_nop = amdgpu_ring_insert_nop, .pad_ib = amdgpu_ring_generic_pad_ib, .emit_wreg = gfx_v7_0_ring_emit_wreg, + .emit_mem_sync = gfx_v7_0_emit_mem_sync_compute, }; static void gfx_v7_0_set_ring_funcs(struct amdgpu_device *adev) |