Merge tag 'drm-fixes-2024-08-16' of https://gitlab.freedesktop.org/drm/kernel

Pull drm fixes from Dave Airlie:
 "Weekly drm fixes, mostly amdgpu and xe. The larger amdgpu fix is for a
  new IP block introduced in rc1, so should be fine. The xe fixes
  contain some missed fixes from the end of the previous round along
  with some fixes which required precursor changes, but otherwise
  everything seems fine,

  mediatek:
   - fix cursor crash

  amdgpu:
   - Fix MES ring buffer overflow
   - DCN 3.5 fix
   - DCN 3.2.1 fix
   - DP MST fix
   - Cursor fixes
   - JPEG fixes
   - Context ops validation
   - MES 12 fixes
   - VCN 5.0 fix
   - HDP fix

  panel:
   - dt bindings style fix
   - orientation quirks

  rockchip:
   - inno-hdmi: fix infoframe upload

  v3d:
   - fix OOB access in v3d_csd_job_run()

  xe:
   - Validate user fence during creation
   - Fix use after free when client stats are captured
   - SRIOV fixes
   - Runtime PM fixes"

* tag 'drm-fixes-2024-08-16' of https://gitlab.freedesktop.org/drm/kernel: (37 commits)
  drm/xe: Hold a PM ref when GT TLB invalidations are inflight
  drm/xe: Drop xe_gt_tlb_invalidation_wait
  drm/xe: Add xe_gt_tlb_invalidation_fence_init helper
  drm/xe/pf: Fix VF config validation on multi-GT platforms
  drm/xe: Build PM into GuC CT layer
  drm/xe/vf: Fix register value lookup
  drm/xe: Fix use after free when client stats are captured
  drm/xe: Take a ref to xe file when user creates a VM
  drm/xe: Add ref counting for xe_file
  drm/xe: Move part of xe_file cleanup to a helper
  drm/xe: Validate user fence during creation
  drm/rockchip: inno-hdmi: Fix infoframe upload
  drm/amd/amdgpu: add HDP_SD support on gc 12.0.0/1
  drm/amdgpu: Update kmd_fw_shared for VCN5
  drm/amd/amdgpu: command submission parser for JPEG
  drm/amdgpu/mes12: fix suspend issue
  drm/amdgpu/mes12: sw/hw fini for unified mes
  drm/amdgpu/mes12: configure two pipes hardware resources
  drm/amdgpu/mes12: adjust mes12 sw/hw init for multiple pipes
  drm/amdgpu/mes12: add mes pipe switch support
  ...

20 files changed, 379 insertions, 240 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index 9aa952f258cf..6dfdff58bffd 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -1057,6 +1057,9 @@ static int amdgpu_cs_patch_ibs(struct amdgpu_cs_parser *p,
 			r = amdgpu_ring_parse_cs(ring, p, job, ib);
 			if (r)
 				return r;
+
+			if (ib->sa_bo)
+				ib->gpu_addr =  amdgpu_sa_bo_gpu_addr(ib->sa_bo);
 		} else {
 			ib->ptr = (uint32_t *)kptr;
 			r = amdgpu_ring_patch_cs_in_place(ring, p, job, ib);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
index 5cb33ac99f70..c43d1b6e5d66 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
@@ -685,16 +685,24 @@ int amdgpu_ctx_ioctl(struct drm_device *dev, void *data,
 
 	switch (args->in.op) {
 	case AMDGPU_CTX_OP_ALLOC_CTX:
+		if (args->in.flags)
+			return -EINVAL;
 		r = amdgpu_ctx_alloc(adev, fpriv, filp, priority, &id);
 		args->out.alloc.ctx_id = id;
 		break;
 	case AMDGPU_CTX_OP_FREE_CTX:
+		if (args->in.flags)
+			return -EINVAL;
 		r = amdgpu_ctx_free(fpriv, id);
 		break;
 	case AMDGPU_CTX_OP_QUERY_STATE:
+		if (args->in.flags)
+			return -EINVAL;
 		r = amdgpu_ctx_query(adev, fpriv, id, &args->out);
 		break;
 	case AMDGPU_CTX_OP_QUERY_STATE2:
+		if (args->in.flags)
+			return -EINVAL;
 		r = amdgpu_ctx_query2(adev, fpriv, id, &args->out);
 		break;
 	case AMDGPU_CTX_OP_GET_STABLE_PSTATE:
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
index 82452606ae6c..c770cb201e64 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
@@ -509,6 +509,16 @@ int amdgpu_gfx_disable_kcq(struct amdgpu_device *adev, int xcc_id)
 	int i, r = 0;
 	int j;
 
+	if (adev->enable_mes) {
+		for (i = 0; i < adev->gfx.num_compute_rings; i++) {
+			j = i + xcc_id * adev->gfx.num_compute_rings;
+			amdgpu_mes_unmap_legacy_queue(adev,
+						   &adev->gfx.compute_ring[j],
+						   RESET_QUEUES, 0, 0);
+		}
+		return 0;
+	}
+
 	if (!kiq->pmf || !kiq->pmf->kiq_unmap_queues)
 		return -EINVAL;
 
@@ -551,6 +561,18 @@ int amdgpu_gfx_disable_kgq(struct amdgpu_device *adev, int xcc_id)
 	int i, r = 0;
 	int j;
 
+	if (adev->enable_mes) {
+		if (amdgpu_gfx_is_master_xcc(adev, xcc_id)) {
+			for (i = 0; i < adev->gfx.num_gfx_rings; i++) {
+				j = i + xcc_id * adev->gfx.num_gfx_rings;
+				amdgpu_mes_unmap_legacy_queue(adev,
+						      &adev->gfx.gfx_ring[j],
+						      PREEMPT_QUEUES, 0, 0);
+			}
+		}
+		return 0;
+	}
+
 	if (!kiq->pmf || !kiq->pmf->kiq_unmap_queues)
 		return -EINVAL;
 
@@ -995,7 +1017,7 @@ uint32_t amdgpu_kiq_rreg(struct amdgpu_device *adev, uint32_t reg, uint32_t xcc_
 	if (amdgpu_device_skip_hw_access(adev))
 		return 0;
 
-	if (adev->mes.ring.sched.ready)
+	if (adev->mes.ring[0].sched.ready)
 		return amdgpu_mes_rreg(adev, reg);
 
 	BUG_ON(!ring->funcs->emit_rreg);
@@ -1065,7 +1087,7 @@ void amdgpu_kiq_wreg(struct amdgpu_device *adev, uint32_t reg, uint32_t v, uint3
 	if (amdgpu_device_skip_hw_access(adev))
 		return;
 
-	if (adev->mes.ring.sched.ready) {
+	if (adev->mes.ring[0].sched.ready) {
 		amdgpu_mes_wreg(adev, reg, v);
 		return;
 	}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
index c02659025656..b49b3650fd62 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
@@ -589,7 +589,8 @@ int amdgpu_gmc_allocate_vm_inv_eng(struct amdgpu_device *adev)
 		ring = adev->rings[i];
 		vmhub = ring->vm_hub;
 
-		if (ring == &adev->mes.ring ||
+		if (ring == &adev->mes.ring[0] ||
+		    ring == &adev->mes.ring[1] ||
 		    ring == &adev->umsch_mm.ring)
 			continue;
 
@@ -761,7 +762,7 @@ void amdgpu_gmc_fw_reg_write_reg_wait(struct amdgpu_device *adev,
 	unsigned long flags;
 	uint32_t seq;
 
-	if (adev->mes.ring.sched.ready) {
+	if (adev->mes.ring[0].sched.ready) {
 		amdgpu_mes_reg_write_reg_wait(adev, reg0, reg1,
 					      ref, mask);
 		return;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c
index dac88d2dd70d..1cb1ec7beefe 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c
@@ -135,9 +135,11 @@ int amdgpu_mes_init(struct amdgpu_device *adev)
 	idr_init(&adev->mes.queue_id_idr);
 	ida_init(&adev->mes.doorbell_ida);
 	spin_lock_init(&adev->mes.queue_id_lock);
-	spin_lock_init(&adev->mes.ring_lock);
 	mutex_init(&adev->mes.mutex_hidden);
 
+	for (i = 0; i < AMDGPU_MAX_MES_PIPES; i++)
+		spin_lock_init(&adev->mes.ring_lock[i]);
+
 	adev->mes.total_max_queue = AMDGPU_FENCE_MES_QUEUE_ID_MASK;
 	adev->mes.vmid_mask_mmhub = 0xffffff00;
 	adev->mes.vmid_mask_gfxhub = 0xffffff00;
@@ -163,36 +165,38 @@ int amdgpu_mes_init(struct amdgpu_device *adev)
 			adev->mes.sdma_hqd_mask[i] = 0xfc;
 	}
 
-	r = amdgpu_device_wb_get(adev, &adev->mes.sch_ctx_offs);
-	if (r) {
-		dev_err(adev->dev,
-			"(%d) ring trail_fence_offs wb alloc failed\n", r);
-		goto error_ids;
-	}
-	adev->mes.sch_ctx_gpu_addr =
-		adev->wb.gpu_addr + (adev->mes.sch_ctx_offs * 4);
-	adev->mes.sch_ctx_ptr =
-		(uint64_t *)&adev->wb.wb[adev->mes.sch_ctx_offs];
+	for (i = 0; i < AMDGPU_MAX_MES_PIPES; i++) {
+		r = amdgpu_device_wb_get(adev, &adev->mes.sch_ctx_offs[i]);
+		if (r) {
+			dev_err(adev->dev,
+				"(%d) ring trail_fence_offs wb alloc failed\n",
+				r);
+			goto error;
+		}
+		adev->mes.sch_ctx_gpu_addr[i] =
+			adev->wb.gpu_addr + (adev->mes.sch_ctx_offs[i] * 4);
+		adev->mes.sch_ctx_ptr[i] =
+			(uint64_t *)&adev->wb.wb[adev->mes.sch_ctx_offs[i]];
 
-	r = amdgpu_device_wb_get(adev, &adev->mes.query_status_fence_offs);
-	if (r) {
-		amdgpu_device_wb_free(adev, adev->mes.sch_ctx_offs);
-		dev_err(adev->dev,
-			"(%d) query_status_fence_offs wb alloc failed\n", r);
-		goto error_ids;
+		r = amdgpu_device_wb_get(adev,
+				 &adev->mes.query_status_fence_offs[i]);
+		if (r) {
+			dev_err(adev->dev,
+			      "(%d) query_status_fence_offs wb alloc failed\n",
+			      r);
+			goto error;
+		}
+		adev->mes.query_status_fence_gpu_addr[i] = adev->wb.gpu_addr +
+			(adev->mes.query_status_fence_offs[i] * 4);
+		adev->mes.query_status_fence_ptr[i] =
+			(uint64_t *)&adev->wb.wb[adev->mes.query_status_fence_offs[i]];
 	}
-	adev->mes.query_status_fence_gpu_addr =
-		adev->wb.gpu_addr + (adev->mes.query_status_fence_offs * 4);
-	adev->mes.query_status_fence_ptr =
-		(uint64_t *)&adev->wb.wb[adev->mes.query_status_fence_offs];
 
 	r = amdgpu_device_wb_get(adev, &adev->mes.read_val_offs);
 	if (r) {
-		amdgpu_device_wb_free(adev, adev->mes.sch_ctx_offs);
-		amdgpu_device_wb_free(adev, adev->mes.query_status_fence_offs);
 		dev_err(adev->dev,
 			"(%d) read_val_offs alloc failed\n", r);
-		goto error_ids;
+		goto error;
 	}
 	adev->mes.read_val_gpu_addr =
 		adev->wb.gpu_addr + (adev->mes.read_val_offs * 4);
@@ -212,10 +216,16 @@ int amdgpu_mes_init(struct amdgpu_device *adev)
 error_doorbell:
 	amdgpu_mes_doorbell_free(adev);
 error:
-	amdgpu_device_wb_free(adev, adev->mes.sch_ctx_offs);
-	amdgpu_device_wb_free(adev, adev->mes.query_status_fence_offs);
-	amdgpu_device_wb_free(adev, adev->mes.read_val_offs);
-error_ids:
+	for (i = 0; i < AMDGPU_MAX_MES_PIPES; i++) {
+		if (adev->mes.sch_ctx_ptr[i])
+			amdgpu_device_wb_free(adev, adev->mes.sch_ctx_offs[i]);
+		if (adev->mes.query_status_fence_ptr[i])
+			amdgpu_device_wb_free(adev,
+				      adev->mes.query_status_fence_offs[i]);
+	}
+	if (adev->mes.read_val_ptr)
+		amdgpu_device_wb_free(adev, adev->mes.read_val_offs);
+
 	idr_destroy(&adev->mes.pasid_idr);
 	idr_destroy(&adev->mes.gang_id_idr);
 	idr_destroy(&adev->mes.queue_id_idr);
@@ -226,13 +236,22 @@ error_ids:
 
 void amdgpu_mes_fini(struct amdgpu_device *adev)
 {
+	int i;
+
 	amdgpu_bo_free_kernel(&adev->mes.event_log_gpu_obj,
 			      &adev->mes.event_log_gpu_addr,
 			      &adev->mes.event_log_cpu_addr);
 
-	amdgpu_device_wb_free(adev, adev->mes.sch_ctx_offs);
-	amdgpu_device_wb_free(adev, adev->mes.query_status_fence_offs);
-	amdgpu_device_wb_free(adev, adev->mes.read_val_offs);
+	for (i = 0; i < AMDGPU_MAX_MES_PIPES; i++) {
+		if (adev->mes.sch_ctx_ptr[i])
+			amdgpu_device_wb_free(adev, adev->mes.sch_ctx_offs[i]);
+		if (adev->mes.query_status_fence_ptr[i])
+			amdgpu_device_wb_free(adev,
+				      adev->mes.query_status_fence_offs[i]);
+	}
+	if (adev->mes.read_val_ptr)
+		amdgpu_device_wb_free(adev, adev->mes.read_val_offs);
+
 	amdgpu_mes_doorbell_free(adev);
 
 	idr_destroy(&adev->mes.pasid_idr);
@@ -1499,7 +1518,7 @@ int amdgpu_mes_init_microcode(struct amdgpu_device *adev, int pipe)
 
 	amdgpu_ucode_ip_version_decode(adev, GC_HWIP, ucode_prefix,
 				       sizeof(ucode_prefix));
-	if (adev->enable_uni_mes && pipe == AMDGPU_MES_SCHED_PIPE) {
+	if (adev->enable_uni_mes) {
 		snprintf(fw_name, sizeof(fw_name),
 			 "amdgpu/%s_uni_mes.bin", ucode_prefix);
 	} else if (amdgpu_ip_version(adev, GC_HWIP, 0) >= IP_VERSION(11, 0, 0) &&
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h
index 2d659c612f03..0bc837dab578 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.h
@@ -82,8 +82,8 @@ struct amdgpu_mes {
 	uint64_t                        default_process_quantum;
 	uint64_t                        default_gang_quantum;
 
-	struct amdgpu_ring              ring;
-	spinlock_t                      ring_lock;
+	struct amdgpu_ring              ring[AMDGPU_MAX_MES_PIPES];
+	spinlock_t                      ring_lock[AMDGPU_MAX_MES_PIPES];
 
 	const struct firmware           *fw[AMDGPU_MAX_MES_PIPES];
 
@@ -112,12 +112,12 @@ struct amdgpu_mes {
 	uint32_t                        gfx_hqd_mask[AMDGPU_MES_MAX_GFX_PIPES];
 	uint32_t                        sdma_hqd_mask[AMDGPU_MES_MAX_SDMA_PIPES];
 	uint32_t                        aggregated_doorbells[AMDGPU_MES_PRIORITY_NUM_LEVELS];
-	uint32_t                        sch_ctx_offs;
-	uint64_t			sch_ctx_gpu_addr;
-	uint64_t			*sch_ctx_ptr;
-	uint32_t			query_status_fence_offs;
-	uint64_t			query_status_fence_gpu_addr;
-	uint64_t			*query_status_fence_ptr;
+	uint32_t                        sch_ctx_offs[AMDGPU_MAX_MES_PIPES];
+	uint64_t			sch_ctx_gpu_addr[AMDGPU_MAX_MES_PIPES];
+	uint64_t			*sch_ctx_ptr[AMDGPU_MAX_MES_PIPES];
+	uint32_t			query_status_fence_offs[AMDGPU_MAX_MES_PIPES];
+	uint64_t			query_status_fence_gpu_addr[AMDGPU_MAX_MES_PIPES];
+	uint64_t			*query_status_fence_ptr[AMDGPU_MAX_MES_PIPES];
 	uint32_t                        read_val_offs;
 	uint64_t			read_val_gpu_addr;
 	uint32_t			*read_val_ptr;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
index ad49cecb20b8..e6344a6b0a9f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
@@ -212,6 +212,8 @@ int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring,
 	 */
 	if (ring->funcs->type == AMDGPU_RING_TYPE_KIQ)
 		sched_hw_submission = max(sched_hw_submission, 256);
+	if (ring->funcs->type == AMDGPU_RING_TYPE_MES)
+		sched_hw_submission = 8;
 	else if (ring == &adev->sdma.instance[0].page)
 		sched_hw_submission = 256;
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
index 1a5439abd1a0..c87d68d4be53 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
@@ -461,8 +461,11 @@ struct amdgpu_vcn5_fw_shared {
 	struct amdgpu_fw_shared_unified_queue_struct sq;
 	uint8_t pad1[8];
 	struct amdgpu_fw_shared_fw_logging fw_log;
+	uint8_t pad2[20];
 	struct amdgpu_fw_shared_rb_setup rb_setup;
-	uint8_t pad2[4];
+	struct amdgpu_fw_shared_smu_interface_info smu_dpm_interface;
+	struct amdgpu_fw_shared_drm_key_wa drm_key_wa;
+	uint8_t pad3[9];
 };
 
 #define VCN_BLOCK_ENCODE_DISABLE_MASK 0x80
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
index 111c380f929b..b287a82e6177 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
@@ -858,7 +858,7 @@ void amdgpu_virt_post_reset(struct amdgpu_device *adev)
 		adev->gfx.is_poweron = false;
 	}
 
-	adev->mes.ring.sched.ready = false;
+	adev->mes.ring[0].sched.ready = false;
 }
 
 bool amdgpu_virt_fw_load_skip_check(struct amdgpu_device *adev, uint32_t ucode_id)
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c
index 506fa8003388..2c611b8577a7 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c
@@ -3546,33 +3546,9 @@ static int gfx_v12_0_hw_init(void *handle)
 	return r;
 }
 
-static int gfx_v12_0_kiq_disable_kgq(struct amdgpu_device *adev)
-{
-	struct amdgpu_kiq *kiq = &adev->gfx.kiq[0];
-	struct amdgpu_ring *kiq_ring = &kiq->ring;
-	int i, r = 0;
-
-	if (!kiq->pmf || !kiq->pmf->kiq_unmap_queues)
-		return -EINVAL;
-
-	if (amdgpu_ring_alloc(kiq_ring, kiq->pmf->unmap_queues_size *
-					adev->gfx.num_gfx_rings))
-		return -ENOMEM;
-
-	for (i = 0; i < adev->gfx.num_gfx_rings; i++)
-		kiq->pmf->kiq_unmap_queues(kiq_ring, &adev->gfx.gfx_ring[i],
-					   PREEMPT_QUEUES, 0, 0);
-
-	if (adev->gfx.kiq[0].ring.sched.ready)
-		r = amdgpu_ring_test_helper(kiq_ring);
-
-	return r;
-}
-
 static int gfx_v12_0_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	int r;
 	uint32_t tmp;
 
 	amdgpu_irq_put(adev, &adev->gfx.priv_reg_irq, 0);
@@ -3580,8 +3556,7 @@ static int gfx_v12_0_hw_fini(void *handle)
 
 	if (!adev->no_hw_access) {
 		if (amdgpu_async_gfx_ring) {
-			r = gfx_v12_0_kiq_disable_kgq(adev);
-			if (r)
+			if (amdgpu_gfx_disable_kgq(adev, 0))
 				DRM_ERROR("KGQ disable failed\n");
 		}
 
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
index b88a6fa173b3..2797fd84432b 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
@@ -231,7 +231,7 @@ static void gmc_v11_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
 	/* This is necessary for SRIOV as well as for GFXOFF to function
 	 * properly under bare metal
 	 */
-	if ((adev->gfx.kiq[0].ring.sched.ready || adev->mes.ring.sched.ready) &&
+	if ((adev->gfx.kiq[0].ring.sched.ready || adev->mes.ring[0].sched.ready) &&
 	    (amdgpu_sriov_runtime(adev) || !amdgpu_sriov_vf(adev))) {
 		amdgpu_gmc_fw_reg_write_reg_wait(adev, req, ack, inv_req,
 						 1 << vmid, GET_INST(GC, 0));
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v12_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v12_0.c
index 26efce9aa410..edcb5351f8cc 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v12_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v12_0.c
@@ -299,7 +299,7 @@ static void gmc_v12_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
 	/* This is necessary for SRIOV as well as for GFXOFF to function
 	 * properly under bare metal
 	 */
-	if ((adev->gfx.kiq[0].ring.sched.ready || adev->mes.ring.sched.ready) &&
+	if ((adev->gfx.kiq[0].ring.sched.ready || adev->mes.ring[0].sched.ready) &&
 	    (amdgpu_sriov_runtime(adev) || !amdgpu_sriov_vf(adev))) {
 		struct amdgpu_vmhub *hub = &adev->vmhub[vmhub];
 		const unsigned eng = 17;
diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v2_0.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v2_0.c
index 99adf3625657..98aa3ccd0d20 100644
--- a/drivers/gpu/drm/amd/amdgpu/jpeg_v2_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v2_0.c
@@ -538,11 +538,11 @@ void jpeg_v2_0_dec_ring_emit_ib(struct amdgpu_ring *ring,
 
 	amdgpu_ring_write(ring, PACKETJ(mmUVD_LMI_JRBC_IB_VMID_INTERNAL_OFFSET,
 		0, 0, PACKETJ_TYPE0));
-	amdgpu_ring_write(ring, (vmid | (vmid << 4)));
+	amdgpu_ring_write(ring, (vmid | (vmid << 4) | (vmid << 8)));
 
 	amdgpu_ring_write(ring, PACKETJ(mmUVD_LMI_JPEG_VMID_INTERNAL_OFFSET,
 		0, 0, PACKETJ_TYPE0));
-	amdgpu_ring_write(ring, (vmid | (vmid << 4)));
+	amdgpu_ring_write(ring, (vmid | (vmid << 4) | (vmid << 8)));
 
 	amdgpu_ring_write(ring,	PACKETJ(mmUVD_LMI_JRBC_IB_64BIT_BAR_LOW_INTERNAL_OFFSET,
 		0, 0, PACKETJ_TYPE0));
diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.c
index ad524ddc9760..6ae5a784e187 100644
--- a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.c
+++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.c
@@ -23,6 +23,7 @@
 
 #include "amdgpu.h"
 #include "amdgpu_jpeg.h"
+#include "amdgpu_cs.h"
 #include "soc15.h"
 #include "soc15d.h"
 #include "jpeg_v4_0_3.h"
@@ -782,11 +783,15 @@ void jpeg_v4_0_3_dec_ring_emit_ib(struct amdgpu_ring *ring,
 
 	amdgpu_ring_write(ring, PACKETJ(regUVD_LMI_JRBC_IB_VMID_INTERNAL_OFFSET,
 		0, 0, PACKETJ_TYPE0));
-	amdgpu_ring_write(ring, (vmid | (vmid << 4)));
+
+	if (ring->funcs->parse_cs)
+		amdgpu_ring_write(ring, 0);
+	else
+		amdgpu_ring_write(ring, (vmid | (vmid << 4) | (vmid << 8)));
 
 	amdgpu_ring_write(ring, PACKETJ(regUVD_LMI_JPEG_VMID_INTERNAL_OFFSET,
 		0, 0, PACKETJ_TYPE0));
-	amdgpu_ring_write(ring, (vmid | (vmid << 4)));
+	amdgpu_ring_write(ring, (vmid | (vmid << 4) | (vmid << 8)));
 
 	amdgpu_ring_write(ring,	PACKETJ(regUVD_LMI_JRBC_IB_64BIT_BAR_LOW_INTERNAL_OFFSET,
 		0, 0, PACKETJ_TYPE0));
@@ -1084,6 +1089,7 @@ static const struct amdgpu_ring_funcs jpeg_v4_0_3_dec_ring_vm_funcs = {
 	.get_rptr = jpeg_v4_0_3_dec_ring_get_rptr,
 	.get_wptr = jpeg_v4_0_3_dec_ring_get_wptr,
 	.set_wptr = jpeg_v4_0_3_dec_ring_set_wptr,
+	.parse_cs = jpeg_v4_0_3_dec_ring_parse_cs,
 	.emit_frame_size =
 		SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 +
 		SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 +
@@ -1248,3 +1254,56 @@ static void jpeg_v4_0_3_set_ras_funcs(struct amdgpu_device *adev)
 {
 	adev->jpeg.ras = &jpeg_v4_0_3_ras;
 }
+
+/**
+ * jpeg_v4_0_3_dec_ring_parse_cs - command submission parser
+ *
+ * @parser: Command submission parser context
+ * @job: the job to parse
+ * @ib: the IB to parse
+ *
+ * Parse the command stream, return -EINVAL for invalid packet,
+ * 0 otherwise
+ */
+int jpeg_v4_0_3_dec_ring_parse_cs(struct amdgpu_cs_parser *parser,
+			     struct amdgpu_job *job,
+			     struct amdgpu_ib *ib)
+{
+	uint32_t i, reg, res, cond, type;
+	struct amdgpu_device *adev = parser->adev;
+
+	for (i = 0; i < ib->length_dw ; i += 2) {
+		reg  = CP_PACKETJ_GET_REG(ib->ptr[i]);
+		res  = CP_PACKETJ_GET_RES(ib->ptr[i]);
+		cond = CP_PACKETJ_GET_COND(ib->ptr[i]);
+		type = CP_PACKETJ_GET_TYPE(ib->ptr[i]);
+
+		if (res) /* only support 0 at the moment */
+			return -EINVAL;
+
+		switch (type) {
+		case PACKETJ_TYPE0:
+			if (cond != PACKETJ_CONDITION_CHECK0 || reg < JPEG_REG_RANGE_START || reg > JPEG_REG_RANGE_END) {
+				dev_err(adev->dev, "Invalid packet [0x%08x]!\n", ib->ptr[i]);
+				return -EINVAL;
+			}
+			break;
+		case PACKETJ_TYPE3:
+			if (cond != PACKETJ_CONDITION_CHECK3 || reg < JPEG_REG_RANGE_START || reg > JPEG_REG_RANGE_END) {
+				dev_err(adev->dev, "Invalid packet [0x%08x]!\n", ib->ptr[i]);
+				return -EINVAL;
+			}
+			break;
+		case PACKETJ_TYPE6:
+			if (ib->ptr[i] == CP_PACKETJ_NOP)
+				continue;
+			dev_err(adev->dev, "Invalid packet [0x%08x]!\n", ib->ptr[i]);
+			return -EINVAL;
+		default:
+			dev_err(adev->dev, "Unknown packet type %d !\n", type);
+			return -EINVAL;
+		}
+	}
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.h b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.h
index 747a3e5f6856..71c54b294e15 100644
--- a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.h
+++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.h
@@ -46,6 +46,9 @@
 
 #define JRBC_DEC_EXTERNAL_REG_WRITE_ADDR				0x18000
 
+#define JPEG_REG_RANGE_START						0x4000
+#define JPEG_REG_RANGE_END						0x41c2
+
 extern const struct amdgpu_ip_block_version jpeg_v4_0_3_ip_block;
 
 void jpeg_v4_0_3_dec_ring_emit_ib(struct amdgpu_ring *ring,
@@ -62,5 +65,7 @@ void jpeg_v4_0_3_dec_ring_insert_end(struct amdgpu_ring *ring);
 void jpeg_v4_0_3_dec_ring_emit_wreg(struct amdgpu_ring *ring, uint32_t reg, uint32_t val);
 void jpeg_v4_0_3_dec_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg,
 					uint32_t val, uint32_t mask);
-
+int jpeg_v4_0_3_dec_ring_parse_cs(struct amdgpu_cs_parser *parser,
+				  struct amdgpu_job *job,
+				  struct amdgpu_ib *ib);
 #endif /* __JPEG_V4_0_3_H__ */
diff --git a/drivers/gpu/drm/amd/amdgpu/jpeg_v5_0_0.c b/drivers/gpu/drm/amd/amdgpu/jpeg_v5_0_0.c
index d694a276498a..f4daff90c770 100644
--- a/drivers/gpu/drm/amd/amdgpu/jpeg_v5_0_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v5_0_0.c
@@ -646,6 +646,7 @@ static const struct amdgpu_ring_funcs jpeg_v5_0_0_dec_ring_vm_funcs = {
 	.get_rptr = jpeg_v5_0_0_dec_ring_get_rptr,
 	.get_wptr = jpeg_v5_0_0_dec_ring_get_wptr,
 	.set_wptr = jpeg_v5_0_0_dec_ring_set_wptr,
+	.parse_cs = jpeg_v4_0_3_dec_ring_parse_cs,
 	.emit_frame_size =
 		SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 +
 		SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 +
diff --git a/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c b/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c
index f9343642ae7e..2ea8223eb969 100644
--- a/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c
@@ -162,13 +162,13 @@ static int mes_v11_0_submit_pkt_and_poll_completion(struct amdgpu_mes *mes,
 	union MESAPI__QUERY_MES_STATUS mes_status_pkt;
 	signed long timeout = 3000000; /* 3000 ms */
 	struct amdgpu_device *adev = mes->adev;
-	struct amdgpu_ring *ring = &mes->ring;
+	struct amdgpu_ring *ring = &mes->ring[0];
 	struct MES_API_STATUS *api_status;
 	union MESAPI__MISC *x_pkt = pkt;
 	const char *op_str, *misc_op_str;
 	unsigned long flags;
 	u64 status_gpu_addr;
-	u32 status_offset;
+	u32 seq, status_offset;
 	u64 *status_ptr;
 	signed long r;
 	int ret;
@@ -191,11 +191,18 @@ static int mes_v11_0_submit_pkt_and_poll_completion(struct amdgpu_mes *mes,
 	status_ptr = (u64 *)&adev->wb.wb[status_offset];
 	*status_ptr = 0;
 
-	spin_lock_irqsave(&mes->ring_lock, flags);
+	spin_lock_irqsave(&mes->ring_lock[0], flags);
 	r = amdgpu_ring_alloc(ring, (size + sizeof(mes_status_pkt)) / 4);
 	if (r)
 		goto error_unlock_free;
 
+	seq = ++ring->fence_drv.sync_seq;
+	r = amdgpu_fence_wait_polling(ring,
+				      seq - ring->fence_drv.num_fences_mask,
+				      timeout);
+	if (r < 1)
+		goto error_undo;
+
 	api_status = (struct MES_API_STATUS *)((char *)pkt + api_status_off);
 	api_status->api_completion_fence_addr = status_gpu_addr;
 	api_status->api_completion_fence_value = 1;
@@ -208,14 +215,13 @@ static int mes_v11_0_submit_pkt_and_poll_completion(struct amdgpu_mes *mes,
 	mes_status_pkt.header.dwsize = API_FRAME_SIZE_IN_DWORDS;
 	mes_status_pkt.api_status.api_completion_fence_addr =
 		ring->fence_drv.gpu_addr;
-	mes_status_pkt.api_status.api_completion_fence_value =
-		++ring->fence_drv.sync_seq;
+	mes_status_pkt.api_status.api_completion_fence_value = seq;
 
 	amdgpu_ring_write_multiple(ring, &mes_status_pkt,
 				   sizeof(mes_status_pkt) / 4);
 
 	amdgpu_ring_commit(ring);
-	spin_unlock_irqrestore(&mes->ring_lock, flags);
+	spin_unlock_irqrestore(&mes->ring_lock[0], flags);
 
 	op_str = mes_v11_0_get_op_string(x_pkt);
 	misc_op_str = mes_v11_0_get_misc_op_string(x_pkt);
@@ -229,7 +235,7 @@ static int mes_v11_0_submit_pkt_and_poll_completion(struct amdgpu_mes *mes,
 		dev_dbg(adev->dev, "MES msg=%d was emitted\n",
 			x_pkt->header.opcode);
 
-	r = amdgpu_fence_wait_polling(ring, ring->fence_drv.sync_seq, timeout);
+	r = amdgpu_fence_wait_polling(ring, seq, timeout);
 	if (r < 1 || !*status_ptr) {
 
 		if (misc_op_str)
@@ -252,8 +258,12 @@ static int mes_v11_0_submit_pkt_and_poll_completion(struct amdgpu_mes *mes,
 	amdgpu_device_wb_free(adev, status_offset);
 	return 0;
 
+error_undo:
+	dev_err(adev->dev, "MES ring buffer is full.\n");
+	amdgpu_ring_undo(ring);
+
 error_unlock_free:
-	spin_unlock_irqrestore(&mes->ring_lock, flags);
+	spin_unlock_irqrestore(&mes->ring_lock[0], flags);
 
 error_wb_free:
 	amdgpu_device_wb_free(adev, status_offset);
@@ -512,9 +522,9 @@ static int mes_v11_0_set_hw_resources(struct amdgpu_mes *mes)
 	mes_set_hw_res_pkt.vmid_mask_gfxhub = mes->vmid_mask_gfxhub;
 	mes_set_hw_res_pkt.gds_size = adev->gds.gds_size;
 	mes_set_hw_res_pkt.paging_vmid = 0;
-	mes_set_hw_res_pkt.g_sch_ctx_gpu_mc_ptr = mes->sch_ctx_gpu_addr;
+	mes_set_hw_res_pkt.g_sch_ctx_gpu_mc_ptr = mes->sch_ctx_gpu_addr[0];
 	mes_set_hw_res_pkt.query_status_fence_gpu_mc_ptr =
-		mes->query_status_fence_gpu_addr;
+		mes->query_status_fence_gpu_addr[0];
 
 	for (i = 0; i < MAX_COMPUTE_PIPES; i++)
 		mes_set_hw_res_pkt.compute_hqd_mask[i] =
@@ -1015,7 +1025,7 @@ static int mes_v11_0_kiq_enable_queue(struct amdgpu_device *adev)
 		return r;
 	}
 
-	kiq->pmf->kiq_map_queues(kiq_ring, &adev->mes.ring);
+	kiq->pmf->kiq_map_queues(kiq_ring, &adev->mes.ring[0]);
 
 	return amdgpu_ring_test_helper(kiq_ring);
 }
@@ -1029,7 +1039,7 @@ static int mes_v11_0_queue_init(struct amdgpu_device *adev,
 	if (pipe == AMDGPU_MES_KIQ_PIPE)
 		ring = &adev->gfx.kiq[0].ring;
 	else if (pipe == AMDGPU_MES_SCHED_PIPE)
-		ring = &adev->mes.ring;
+		ring = &adev->mes.ring[0];
 	else
 		BUG();
 
@@ -1071,7 +1081,7 @@ static int mes_v11_0_ring_init(struct amdgpu_device *adev)
 {
 	struct amdgpu_ring *ring;
 
-	ring = &adev->mes.ring;
+	ring = &adev->mes.ring[0];
 
 	ring->funcs = &mes_v11_0_ring_funcs;
 
@@ -1124,7 +1134,7 @@ static int mes_v11_0_mqd_sw_init(struct amdgpu_device *adev,
 	if (pipe == AMDGPU_MES_KIQ_PIPE)
 		ring = &adev->gfx.kiq[0].ring;
 	else if (pipe == AMDGPU_MES_SCHED_PIPE)
-		ring = &adev->mes.ring;
+		ring = &adev->mes.ring[0];
 	else
 		BUG();
 
@@ -1200,9 +1210,6 @@ static int mes_v11_0_sw_fini(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	int pipe;
 
-	amdgpu_device_wb_free(adev, adev->mes.sch_ctx_offs);
-	amdgpu_device_wb_free(adev, adev->mes.query_status_fence_offs);
-
 	for (pipe = 0; pipe < AMDGPU_MAX_MES_PIPES; pipe++) {
 		kfree(adev->mes.mqd_backup[pipe]);
 
@@ -1216,12 +1223,12 @@ static int mes_v11_0_sw_fini(void *handle)
 			      &adev->gfx.kiq[0].ring.mqd_gpu_addr,
 			      &adev->gfx.kiq[0].ring.mqd_ptr);
 
-	amdgpu_bo_free_kernel(&adev->mes.ring.mqd_obj,
-			      &adev->mes.ring.mqd_gpu_addr,
-			      &adev->mes.ring.mqd_ptr);
+	amdgpu_bo_free_kernel(&adev->mes.ring[0].mqd_obj,
+			      &adev->mes.ring[0].mqd_gpu_addr,
+			      &adev->mes.ring[0].mqd_ptr);
 
 	amdgpu_ring_fini(&adev->gfx.kiq[0].ring);
-	amdgpu_ring_fini(&adev->mes.ring);
+	amdgpu_ring_fini(&adev->mes.ring[0]);
 
 	if (adev->firmware.load_type == AMDGPU_FW_LOAD_DIRECT) {
 		mes_v11_0_free_ucode_buffers(adev, AMDGPU_MES_K