amdgpu: fix multi-process hang issue

SWDEV-146499: hang during multi vulkan process testing

cause:
the second frame's PREAMBLE_IB have clear-state
and LOAD actions, those actions ruin the pipeline
that is still doing process in the previous frame's
work-load IB.

fix:
need insert pipeline sync if have context switch for
SRIOV (because only SRIOV will report PREEMPTION flag
to UMD)

Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Signed-off-by: Emily Deng <Emily.Deng@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This commit is contained in:
Emily Deng 2018-08-22 20:18:25 +08:00 committed by Alex Deucher
parent 9296435729
commit 4f0ecd36f2
1 changed files with 2 additions and 1 deletions

View File

@ -164,8 +164,10 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs,
return r; return r;
} }
need_ctx_switch = ring->current_ctx != fence_ctx;
if (ring->funcs->emit_pipeline_sync && job && if (ring->funcs->emit_pipeline_sync && job &&
((tmp = amdgpu_sync_get_fence(&job->sched_sync, NULL)) || ((tmp = amdgpu_sync_get_fence(&job->sched_sync, NULL)) ||
(amdgpu_sriov_vf(adev) && need_ctx_switch) ||
amdgpu_vm_need_pipeline_sync(ring, job))) { amdgpu_vm_need_pipeline_sync(ring, job))) {
need_pipe_sync = true; need_pipe_sync = true;
dma_fence_put(tmp); dma_fence_put(tmp);
@ -196,7 +198,6 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs,
} }
skip_preamble = ring->current_ctx == fence_ctx; skip_preamble = ring->current_ctx == fence_ctx;
need_ctx_switch = ring->current_ctx != fence_ctx;
if (job && ring->funcs->emit_cntxcntl) { if (job && ring->funcs->emit_cntxcntl) {
if (need_ctx_switch) if (need_ctx_switch)
status |= AMDGPU_HAVE_CTX_SWITCH; status |= AMDGPU_HAVE_CTX_SWITCH;