[PATCH AUTOSEL 4.18 20/56] amdgpu: fix multi-process hang issue

[Sasha Levin]

From: Emily Deng <Emily.Deng@xxxxxxx>

[ Upstream commit 2f40c6eac74a2a60921cdec9e9a8a57e88e31434 ]

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@xxxxxxx>
Signed-off-by: Emily Deng <Emily.Deng@xxxxxxx>
Reviewed-by: Christian KÃnig <christian.koenig@xxxxxxx>
Signed-off-by: Alex Deucher <alexander.deucher@xxxxxxx>
Signed-off-by: Sasha Levin <alexander.levin@xxxxxxxxxxxxx>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
index 7aaa263ad8c7..6b5d4a20860d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
@@ -164,8 +164,10 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs,
 		return r;
 	}
 
+	need_ctx_switch = ring->current_ctx != fence_ctx;
 	if (ring->funcs->emit_pipeline_sync && job &&
 	    ((tmp = amdgpu_sync_get_fence(&job->sched_sync, NULL)) ||
+	     (amdgpu_sriov_vf(adev) && need_ctx_switch) ||
 	     amdgpu_vm_need_pipeline_sync(ring, job))) {
 		need_pipe_sync = true;
 		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;
-	need_ctx_switch = ring->current_ctx != fence_ctx;
 	if (job && ring->funcs->emit_cntxcntl) {
 		if (need_ctx_switch)
 			status |= AMDGPU_HAVE_CTX_SWITCH;
-- 
2.17.1