drm/amdgpu: handle pipeline sync without a VM fence

If we end up emitting a VM fence keep pipeline sync
associated with that fence.  If not, emit them as
part of the IB fence.

v2: fix need_pipe_sync handling
v3: simplify the function

Cc: David Rosca <david.rosca@amd.com>
Fixes: cb1e657ccac8 ("drm/amdgpu: handle GDS and SPM without a VM fence")
Reviewed-by: David Rosca <david.rosca@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

3 files changed, 19 insertions, 16 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
index da4dc489e80b..360e6f00cb7c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
@@ -222,7 +222,7 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned int num_ibs,
 		vm_af = job->hw_vm_fence;
 		/* VM sequence */
 		vm_af->ib_wptr = ring->wptr;
-		amdgpu_vm_flush(ring, job, need_pipe_sync, &emit_spm_needed,
+		amdgpu_vm_flush(ring, job, &need_pipe_sync, &emit_spm_needed,
 				&emit_gds_needed);
 		vm_af->ib_dw_size =
 			amdgpu_ring_get_dw_distance(ring, vm_af->ib_wptr, ring->wptr);
@@ -235,6 +235,10 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned int num_ibs,
 	if (ring->funcs->insert_start)
 		ring->funcs->insert_start(ring);
 
+	/* this may have been handled by amdgpu_vm_flush */
+	if (need_pipe_sync)
+		amdgpu_ring_emit_pipeline_sync(ring);
+
 	if (emit_spm_needed)
 		adev->gfx.rlc.funcs->update_spm_vmid(adev, ring->xcc_id, ring, job->vmid);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 71050a86bcc3..f6c5de63eae6 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -772,7 +772,7 @@ bool amdgpu_vm_need_pipeline_sync(struct amdgpu_ring *ring,
  * Emit a VM flush when it is necessary.
  */
 void amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job,
-		     bool need_pipe_sync, bool *emit_spm_needed,
+		     bool *need_pipe_sync, bool *emit_spm_needed,
 		     bool *emit_gds_needed)
 {
 	struct amdgpu_device *adev = ring->adev;
@@ -827,8 +827,7 @@ void amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job,
 	if (gds_switch_needed && emit_fence)
 		*emit_gds_needed = false;
 
-	if (!vm_flush_needed && !gds_switch_needed && !need_pipe_sync &&
-	    !cleaner_shader_needed && !spm_update_needed)
+	if (!emit_fence)
 		return;
 
 	amdgpu_ring_ib_begin(ring);
@@ -847,8 +846,10 @@ void amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job,
 		patch = amdgpu_ring_init_cond_exec(ring,
 						   ring->cond_exe_gpu_addr);
 
-	if (need_pipe_sync)
+	if (*need_pipe_sync) {
 		amdgpu_ring_emit_pipeline_sync(ring);
+		*need_pipe_sync = false;
+	}
 
 	if (cleaner_shader_needed)
 		ring->funcs->emit_cleaner_shader(ring);
@@ -861,21 +862,19 @@ void amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job,
 	if (pasid_mapping_needed)
 		amdgpu_gmc_emit_pasid_mapping(ring, job->vmid, job->pasid);
 
-	if (emit_fence) {
-		if (spm_update_needed)
-			adev->gfx.rlc.funcs->update_spm_vmid(adev, ring->xcc_id, ring, job->vmid);
+	if (spm_update_needed)
+		adev->gfx.rlc.funcs->update_spm_vmid(adev, ring->xcc_id, ring, job->vmid);
 
-		if (gds_switch_needed)
-			amdgpu_ring_emit_gds_switch(ring, job->vmid, job->gds_base,
+	if (gds_switch_needed)
+		amdgpu_ring_emit_gds_switch(ring, job->vmid, job->gds_base,
 						    job->gds_size, job->gws_base,
 						    job->gws_size, job->oa_base,
 						    job->oa_size);
 
-		amdgpu_fence_emit(ring, job->hw_vm_fence, 0);
-		fence = &job->hw_vm_fence->base;
-		/* get a ref for the job */
-		dma_fence_get(fence);
-	}
+	amdgpu_fence_emit(ring, job->hw_vm_fence, 0);
+	fence = &job->hw_vm_fence->base;
+	/* get a ref for the job */
+	dma_fence_get(fence);
 
 	if (vm_flush_needed) {
 		mutex_lock(&id_mgr->lock);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
index 7f2ba728e3ed..d32183cd9e0f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
@@ -512,7 +512,7 @@ int amdgpu_vm_validate(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 		       int (*callback)(void *p, struct amdgpu_bo *bo),
 		       void *param);
 void amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job,
-		     bool need_pipe_sync, bool *emit_spm_needed,
+		     bool *need_pipe_sync, bool *emit_spm_needed,
 		     bool *emit_gds_needed);
 int amdgpu_vm_update_pdes(struct amdgpu_device *adev,
 			  struct amdgpu_vm *vm, bool immediate);