[PATCH] drm/amd/amdgpu: Update VCN initizalization behvaior

Bokun Zhang Bokun.Zhang at amd.com
Tue Nov 10 20:17:29 UTC 2020


- Issue:
  In the original vcn3.0 code, it assumes that the VCN's
  init_status is always 1, even after the MMSCH
  updates the header.

  This is incorrect since by default, it should be set to 0,
  and MMSCH will update it to 1 if VCN is ready.

- Fix:
  We need to read back the table header after send it to MMSCH.
  After that, if a VCN instance is not ready (host disabled it),
  we needs to disable the ring buffer as well.

Change-Id: I2c5dc4344c5919044370dfe606bf5980a202c662

Signed-off-by: Bokun Zhang <Bokun.Zhang at amd.com>
---
 drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c | 46 +++++++++++++++++++++------
 1 file changed, 37 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c
index 3970a0ccb6a5..c5e0a531caba 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c
@@ -294,17 +294,19 @@ static int vcn_v3_0_hw_init(void *handle)
 				continue;
 
 			ring = &adev->vcn.inst[i].ring_dec;
-			ring->wptr = 0;
-			ring->wptr_old = 0;
-			vcn_v3_0_dec_ring_set_wptr(ring);
-			ring->sched.ready = true;
+			if (ring->sched.ready) {
+				ring->wptr = 0;
+				ring->wptr_old = 0;
+				vcn_v3_0_dec_ring_set_wptr(ring);
+			}
 
 			for (j = 0; j < adev->vcn.num_enc_rings; ++j) {
 				ring = &adev->vcn.inst[i].ring_enc[j];
-				ring->wptr = 0;
-				ring->wptr_old = 0;
-				vcn_v3_0_enc_ring_set_wptr(ring);
-				ring->sched.ready = true;
+				if (ring->sched.ready) {
+					ring->wptr = 0;
+					ring->wptr_old = 0;
+					vcn_v3_0_enc_ring_set_wptr(ring);
+				}
 			}
 		}
 	} else {
@@ -1230,6 +1232,8 @@ static int vcn_v3_0_start_sriov(struct amdgpu_device *adev)
 	uint32_t table_size;
 	uint32_t size, size_dw;
 
+	bool is_vcn_ready;
+
 	struct mmsch_v3_0_cmd_direct_write
 		direct_wt = { {0} };
 	struct mmsch_v3_0_cmd_direct_read_modify_write
@@ -1367,7 +1371,7 @@ static int vcn_v3_0_start_sriov(struct amdgpu_device *adev)
 		MMSCH_V3_0_INSERT_END();
 
 		/* refine header */
-		header.inst[i].init_status = 1;
+		header.inst[i].init_status = 0;
 		header.inst[i].table_offset = header.total_size;
 		header.inst[i].table_size = table_size;
 		header.total_size += table_size;
@@ -1425,6 +1429,30 @@ static int vcn_v3_0_start_sriov(struct amdgpu_device *adev)
 		}
 	}
 
+	/* 6, check each VCN's init_status
+	 * if it remains as 0, then this VCN is not assigned to current VF
+	 * do not start ring for this VCN
+	 */
+	size = sizeof(struct mmsch_v3_0_init_header);
+	table_loc = (uint32_t *)table->cpu_addr;
+	memcpy(&header, (void *)table_loc, size);
+
+	for (i = 0; i < adev->vcn.num_vcn_inst; i++) {
+		if (adev->vcn.harvest_config & (1 << i))
+			continue;
+
+		is_vcn_ready = (header.inst[i].init_status == 1);
+		if (!is_vcn_ready)
+			DRM_INFO("VCN(%d) engine is disabled by hypervisor\n", i);
+
+		ring = &adev->vcn.inst[i].ring_dec;
+		ring->sched.ready = is_vcn_ready;
+		for (j = 0; j < adev->vcn.num_enc_rings; ++j) {
+			ring = &adev->vcn.inst[i].ring_enc[j];
+			ring->sched.ready = is_vcn_ready;
+		}
+	}
+
 	return 0;
 }
 
-- 
2.20.1



More information about the amd-gfx mailing list