Merge tag 'amd-drm-next-6.7-2023-11-10' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-6.7-2023-11-10:

amdgpu:
- SR-IOV fixes
- DMCUB fixes
- DCN3.5 fixes
- DP2 fixes
- SubVP fixes
- SMU14 fixes
- SDMA4.x fixes
- Suspend/resume fixes
- AGP regression fix
- UAF fixes for some error cases
- SMU 13.0.6 fixes
- Documentation fixes
- RAS fixes
- Hotplug fixes
- Scheduling entity ordering fix
- GPUVM fixes

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231110190703.4741-1-alexander.deucher@amd.com

45 files changed, 877 insertions, 223 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index 8b4ca2576a3d..afec09930efa 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -1159,11 +1159,18 @@ uint32_t amdgpu_device_rreg(struct amdgpu_device *adev,
 			    uint32_t reg, uint32_t acc_flags);
 u32 amdgpu_device_indirect_rreg_ext(struct amdgpu_device *adev,
 				    u64 reg_addr);
+uint32_t amdgpu_device_xcc_rreg(struct amdgpu_device *adev,
+				uint32_t reg, uint32_t acc_flags,
+				uint32_t xcc_id);
 void amdgpu_device_wreg(struct amdgpu_device *adev,
 			uint32_t reg, uint32_t v,
 			uint32_t acc_flags);
 void amdgpu_device_indirect_wreg_ext(struct amdgpu_device *adev,
 				     u64 reg_addr, u32 reg_data);
+void amdgpu_device_xcc_wreg(struct amdgpu_device *adev,
+			    uint32_t reg, uint32_t v,
+			    uint32_t acc_flags,
+			    uint32_t xcc_id);
 void amdgpu_mm_wreg_mmio_rlc(struct amdgpu_device *adev,
 			     uint32_t reg, uint32_t v, uint32_t xcc_id);
 void amdgpu_mm_wreg8(struct amdgpu_device *adev, uint32_t offset, uint8_t value);
@@ -1204,8 +1211,8 @@ int emu_soc_asic_init(struct amdgpu_device *adev);
 #define RREG32_NO_KIQ(reg) amdgpu_device_rreg(adev, (reg), AMDGPU_REGS_NO_KIQ)
 #define WREG32_NO_KIQ(reg, v) amdgpu_device_wreg(adev, (reg), (v), AMDGPU_REGS_NO_KIQ)
 
-#define RREG32_KIQ(reg) amdgpu_kiq_rreg(adev, (reg))
-#define WREG32_KIQ(reg, v) amdgpu_kiq_wreg(adev, (reg), (v))
+#define RREG32_KIQ(reg) amdgpu_kiq_rreg(adev, (reg), 0)
+#define WREG32_KIQ(reg, v) amdgpu_kiq_wreg(adev, (reg), (v), 0)
 
 #define RREG8(reg) amdgpu_mm_rreg8(adev, (reg))
 #define WREG8(reg, v) amdgpu_mm_wreg8(adev, (reg), (v))
@@ -1215,6 +1222,8 @@ int emu_soc_asic_init(struct amdgpu_device *adev);
 #define WREG32(reg, v) amdgpu_device_wreg(adev, (reg), (v), 0)
 #define REG_SET(FIELD, v) (((v) << FIELD##_SHIFT) & FIELD##_MASK)
 #define REG_GET(FIELD, v) (((v) << FIELD##_SHIFT) & FIELD##_MASK)
+#define RREG32_XCC(reg, inst) amdgpu_device_xcc_rreg(adev, (reg), 0, inst)
+#define WREG32_XCC(reg, v, inst) amdgpu_device_xcc_wreg(adev, (reg), (v), 0, inst)
 #define RREG32_PCIE(reg) adev->pcie_rreg(adev, (reg))
 #define WREG32_PCIE(reg, v) adev->pcie_wreg(adev, (reg), (v))
 #define RREG32_PCIE_PORT(reg) adev->pciep_rreg(adev, (reg))
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c
index d62e49758635..e550067e5c5d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c
@@ -1497,6 +1497,9 @@ bool amdgpu_acpi_is_s0ix_active(struct amdgpu_device *adev)
 	if (adev->asic_type < CHIP_RAVEN)
 		return false;
 
+	if (!(adev->pm.pp_feature & PP_GFXOFF_MASK))
+		return false;
+
 	/*
 	 * If ACPI_FADT_LOW_POWER_S0 is not set in the FADT, it is generally
 	 * risky to do any special firmware-related preparations for entering
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gc_9_4_3.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gc_9_4_3.c
index 490c8f5ddb60..f6598b9e4faa 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gc_9_4_3.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gc_9_4_3.c
@@ -300,14 +300,13 @@ static int kgd_gfx_v9_4_3_hqd_load(struct amdgpu_device *adev, void *mqd,
 	hqd_end = SOC15_REG_OFFSET(GC, GET_INST(GC, inst), regCP_HQD_AQL_DISPATCH_ID_HI);
 
 	for (reg = hqd_base; reg <= hqd_end; reg++)
-		WREG32_RLC(reg, mqd_hqd[reg - hqd_base]);
+		WREG32_XCC(reg, mqd_hqd[reg - hqd_base], inst);
 
 
 	/* Activate doorbell logic before triggering WPTR poll. */
 	data = REG_SET_FIELD(m->cp_hqd_pq_doorbell_control,
 			     CP_HQD_PQ_DOORBELL_CONTROL, DOORBELL_EN, 1);
-	WREG32_RLC(SOC15_REG_OFFSET(GC, GET_INST(GC, inst), regCP_HQD_PQ_DOORBELL_CONTROL),
-				data);
+	WREG32_SOC15_RLC(GC, GET_INST(GC, inst), regCP_HQD_PQ_DOORBELL_CONTROL, data);
 
 	if (wptr) {
 		/* Don't read wptr with get_user because the user
@@ -336,27 +335,24 @@ static int kgd_gfx_v9_4_3_hqd_load(struct amdgpu_device *adev, void *mqd,
 		guessed_wptr += m->cp_hqd_pq_wptr_lo & ~(queue_size - 1);
 		guessed_wptr += (uint64_t)m->cp_hqd_pq_wptr_hi << 32;
 
-		WREG32_RLC(SOC15_REG_OFFSET(GC, GET_INST(GC, inst), regCP_HQD_PQ_WPTR_LO),
-		       lower_32_bits(guessed_wptr));
-		WREG32_RLC(SOC15_REG_OFFSET(GC, GET_INST(GC, inst), regCP_HQD_PQ_WPTR_HI),
-		       upper_32_bits(guessed_wptr));
-		WREG32_RLC(SOC15_REG_OFFSET(GC, GET_INST(GC, inst), regCP_HQD_PQ_WPTR_POLL_ADDR),
-		       lower_32_bits((uintptr_t)wptr));
-		WREG32_RLC(SOC15_REG_OFFSET(GC, GET_INST(GC, inst),
-			regCP_HQD_PQ_WPTR_POLL_ADDR_HI),
+		WREG32_SOC15_RLC(GC, GET_INST(GC, inst), regCP_HQD_PQ_WPTR_LO,
+			lower_32_bits(guessed_wptr));
+		WREG32_SOC15_RLC(GC, GET_INST(GC, inst), regCP_HQD_PQ_WPTR_HI,
+			upper_32_bits(guessed_wptr));
+		WREG32_SOC15_RLC(GC, GET_INST(GC, inst), regCP_HQD_PQ_WPTR_POLL_ADDR,
+			lower_32_bits((uintptr_t)wptr));
+		WREG32_SOC15_RLC(GC, GET_INST(GC, inst), regCP_HQD_PQ_WPTR_POLL_ADDR_HI,
 			upper_32_bits((uintptr_t)wptr));
-		WREG32(SOC15_REG_OFFSET(GC, GET_INST(GC, inst), regCP_PQ_WPTR_POLL_CNTL1),
-		       (uint32_t)kgd_gfx_v9_get_queue_mask(adev, pipe_id,
-			       queue_id));
+		WREG32_SOC15_RLC(GC, GET_INST(GC, inst), regCP_PQ_WPTR_POLL_CNTL1,
+			(uint32_t)kgd_gfx_v9_get_queue_mask(adev, pipe_id, queue_id));
 	}
 
 	/* Start the EOP fetcher */
-	WREG32_RLC(SOC15_REG_OFFSET(GC, GET_INST(GC, inst), regCP_HQD_EOP_RPTR),
-	       REG_SET_FIELD(m->cp_hqd_eop_rptr,
-			     CP_HQD_EOP_RPTR, INIT_FETCHER, 1));
+	WREG32_SOC15_RLC(GC, GET_INST(GC, inst), regCP_HQD_EOP_RPTR,
+	       REG_SET_FIELD(m->cp_hqd_eop_rptr, CP_HQD_EOP_RPTR, INIT_FETCHER, 1));
 
 	data = REG_SET_FIELD(m->cp_hqd_active, CP_HQD_ACTIVE, ACTIVE, 1);
-	WREG32_RLC(SOC15_REG_OFFSET(GC, GET_INST(GC, inst), regCP_HQD_ACTIVE), data);
+	WREG32_SOC15_RLC(GC, GET_INST(GC, inst), regCP_HQD_ACTIVE, data);
 
 	kgd_gfx_v9_release_queue(adev, inst);
 
@@ -494,15 +490,15 @@ static uint32_t kgd_gfx_v9_4_3_set_address_watch(
 			VALID,
 			1);
 
-	WREG32_RLC((SOC15_REG_OFFSET(GC, GET_INST(GC, inst),
+	WREG32_XCC((SOC15_REG_OFFSET(GC, GET_INST(GC, inst),
 			regTCP_WATCH0_ADDR_H) +
 			(watch_id * TCP_WATCH_STRIDE)),
-			watch_address_high);
+			watch_address_high, inst);
 
-	WREG32_RLC((SOC15_REG_OFFSET(GC, GET_INST(GC, inst),
+	WREG32_XCC((SOC15_REG_OFFSET(GC, GET_INST(GC, inst),
 			regTCP_WATCH0_ADDR_L) +
 			(watch_id * TCP_WATCH_STRIDE)),
-			watch_address_low);
+			watch_address_low, inst);
 
 	return watch_address_cntl;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
index 51011e8ee90d..00fbc0f44c92 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
@@ -91,8 +91,8 @@ void kgd_gfx_v9_program_sh_mem_settings(struct amdgpu_device *adev, uint32_t vmi
 {
 	kgd_gfx_v9_lock_srbm(adev, 0, 0, 0, vmid, inst);
 
-	WREG32_RLC(SOC15_REG_OFFSET(GC, GET_INST(GC, inst), mmSH_MEM_CONFIG), sh_mem_config);
-	WREG32_RLC(SOC15_REG_OFFSET(GC, GET_INST(GC, inst), mmSH_MEM_BASES), sh_mem_bases);
+	WREG32_SOC15_RLC(GC, GET_INST(GC, inst), mmSH_MEM_CONFIG, sh_mem_config);
+	WREG32_SOC15_RLC(GC, GET_INST(GC, inst), mmSH_MEM_BASES, sh_mem_bases);
 	/* APE1 no longer exists on GFX9 */
 
 	kgd_gfx_v9_unlock_srbm(adev, inst);
@@ -239,14 +239,13 @@ int kgd_gfx_v9_hqd_load(struct amdgpu_device *adev, void *mqd,
 
 	for (reg = hqd_base;
 	     reg <= SOC15_REG_OFFSET(GC, GET_INST(GC, inst), mmCP_HQD_PQ_WPTR_HI); reg++)
-		WREG32_RLC(reg, mqd_hqd[reg - hqd_base]);
+		WREG32_XCC(reg, mqd_hqd[reg - hqd_base], inst);
 
 
 	/* Activate doorbell logic before triggering WPTR poll. */
 	data = REG_SET_FIELD(m->cp_hqd_pq_doorbell_control,
 			     CP_HQD_PQ_DOORBELL_CONTROL, DOORBELL_EN, 1);
-	WREG32_RLC(SOC15_REG_OFFSET(GC, GET_INST(GC, inst), mmCP_HQD_PQ_DOORBELL_CONTROL),
-					data);
+	WREG32_SOC15_RLC(GC, GET_INST(GC, inst), mmCP_HQD_PQ_DOORBELL_CONTROL, data);
 
 	if (wptr) {
 		/* Don't read wptr with get_user because the user
@@ -275,25 +274,24 @@ int kgd_gfx_v9_hqd_load(struct amdgpu_device *adev, void *mqd,
 		guessed_wptr += m->cp_hqd_pq_wptr_lo & ~(queue_size - 1);
 		guessed_wptr += (uint64_t)m->cp_hqd_pq_wptr_hi << 32;
 
-		WREG32_RLC(SOC15_REG_OFFSET(GC, GET_INST(GC, inst), mmCP_HQD_PQ_WPTR_LO),
-		       lower_32_bits(guessed_wptr));
-		WREG32_RLC(SOC15_REG_OFFSET(GC, GET_INST(GC, inst), mmCP_HQD_PQ_WPTR_HI),
-		       upper_32_bits(guessed_wptr));
-		WREG32_RLC(SOC15_REG_OFFSET(GC, GET_INST(GC, inst), mmCP_HQD_PQ_WPTR_POLL_ADDR),
-		       lower_32_bits((uintptr_t)wptr));
-		WREG32_RLC(SOC15_REG_OFFSET(GC, GET_INST(GC, inst), mmCP_HQD_PQ_WPTR_POLL_ADDR_HI),
-		       upper_32_bits((uintptr_t)wptr));
-		WREG32_SOC15(GC, GET_INST(GC, inst), mmCP_PQ_WPTR_POLL_CNTL1,
-		       (uint32_t)kgd_gfx_v9_get_queue_mask(adev, pipe_id, queue_id));
+		WREG32_SOC15_RLC(GC, GET_INST(GC, inst), mmCP_HQD_PQ_WPTR_LO,
+			lower_32_bits(guessed_wptr));
+		WREG32_SOC15_RLC(GC, GET_INST(GC, inst), mmCP_HQD_PQ_WPTR_HI,
+			upper_32_bits(guessed_wptr));
+		WREG32_SOC15_RLC(GC, GET_INST(GC, inst), mmCP_HQD_PQ_WPTR_POLL_ADDR,
+			lower_32_bits((uintptr_t)wptr));
+		WREG32_SOC15_RLC(GC, GET_INST(GC, inst), mmCP_HQD_PQ_WPTR_POLL_ADDR_HI,
+			upper_32_bits((uintptr_t)wptr));
+		WREG32_SOC15_RLC(GC, GET_INST(GC, inst), mmCP_PQ_WPTR_POLL_CNTL1,
+			(uint32_t)kgd_gfx_v9_get_queue_mask(adev, pipe_id, queue_id));
 	}
 
 	/* Start the EOP fetcher */
-	WREG32_RLC(SOC15_REG_OFFSET(GC, GET_INST(GC, inst), mmCP_HQD_EOP_RPTR),
-	       REG_SET_FIELD(m->cp_hqd_eop_rptr,
-			     CP_HQD_EOP_RPTR, INIT_FETCHER, 1));
+	WREG32_SOC15_RLC(GC, GET_INST(GC, inst), mmCP_HQD_EOP_RPTR,
+	       REG_SET_FIELD(m->cp_hqd_eop_rptr, CP_HQD_EOP_RPTR, INIT_FETCHER, 1));
 
 	data = REG_SET_FIELD(m->cp_hqd_active, CP_HQD_ACTIVE, ACTIVE, 1);
-	WREG32_RLC(SOC15_REG_OFFSET(GC, GET_INST(GC, inst), mmCP_HQD_ACTIVE), data);
+	WREG32_SOC15_RLC(GC, GET_INST(GC, inst), mmCP_HQD_ACTIVE, data);
 
 	kgd_gfx_v9_release_queue(adev, inst);
 
@@ -556,7 +554,7 @@ int kgd_gfx_v9_hqd_destroy(struct amdgpu_device *adev, void *mqd,
 		break;
 	}
 
-	WREG32_RLC(SOC15_REG_OFFSET(GC, GET_INST(GC, inst), mmCP_HQD_DEQUEUE_REQUEST), type);
+	WREG32_SOC15_RLC(GC, GET_INST(GC, inst), mmCP_HQD_DEQUEUE_REQUEST, type);
 
 	end_jiffies = (utimeout * HZ / 1000) + jiffies;
 	while (true) {
@@ -908,8 +906,8 @@ void kgd_gfx_v9_get_iq_wait_times(struct amdgpu_device *adev,
 					uint32_t inst)
 
 {
-	*wait_times = RREG32(SOC15_REG_OFFSET(GC, GET_INST(GC, inst),
-			mmCP_IQ_WAIT_TIME2));
+	*wait_times = RREG32_SOC15_RLC(GC, GET_INST(GC, inst),
+			mmCP_IQ_WAIT_TIME2);
 }
 
 void kgd_gfx_v9_set_vm_context_page_table_base(struct amdgpu_device *adev,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c
index 781e5c5ce04d..702f6610d024 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c
@@ -172,6 +172,7 @@ int amdgpu_bo_list_get(struct amdgpu_fpriv *fpriv, int id,
 	}
 
 	rcu_read_unlock();
+	*result = NULL;
 	return -ENOENT;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index 86f88512eb7e..df3ecfa9e13f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -1415,7 +1415,7 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 		if (r == -ENOMEM)
 			DRM_ERROR("Not enough memory for command submission!\n");
 		else if (r != -ERESTARTSYS && r != -EAGAIN)
-			DRM_ERROR("Failed to process the buffer list %d!\n", r);
+			DRM_DEBUG("Failed to process the buffer list %d!\n", r);
 		goto error_fini;
 	}
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index fd8cd8e2d3f2..7eeaf0aa7f81 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -73,6 +73,7 @@
 #include "amdgpu_pmu.h"
 #include "amdgpu_fru_eeprom.h"
 #include "amdgpu_reset.h"
+#include "amdgpu_virt.h"
 
 #include <linux/suspend.h>
 #include <drm/task_barrier.h>
@@ -472,7 +473,7 @@ uint32_t amdgpu_device_rreg(struct amdgpu_device *adev,
 		if (!(acc_flags & AMDGPU_REGS_NO_KIQ) &&
 		    amdgpu_sriov_runtime(adev) &&
 		    down_read_trylock(&adev->reset_domain->sem)) {
-			ret = amdgpu_kiq_rreg(adev, reg);
+			ret = amdgpu_kiq_rreg(adev, reg, 0);
 			up_read(&adev->reset_domain->sem);
 		} else {
 			ret = readl(((void __iomem *)adev->rmmio) + (reg * 4));
@@ -509,6 +510,49 @@ uint8_t amdgpu_mm_rreg8(struct amdgpu_device *adev, uint32_t offset)
 	BUG();
 }
 
+
+/**
+ * amdgpu_device_xcc_rreg - read a memory mapped IO or indirect register with specific XCC
+ *
+ * @adev: amdgpu_device pointer
+ * @reg: dword aligned register offset
+ * @acc_flags: access flags which require special behavior
+ * @xcc_id: xcc accelerated compute core id
+ *
+ * Returns the 32 bit value from the offset specified.
+ */
+uint32_t amdgpu_device_xcc_rreg(struct amdgpu_device *adev,
+				uint32_t reg, uint32_t acc_flags,
+				uint32_t xcc_id)
+{
+	uint32_t ret, rlcg_flag;
+
+	if (amdgpu_device_skip_hw_access(adev))
+		return 0;
+
+	if ((reg * 4) < adev->rmmio_size) {
+		if (amdgpu_sriov_vf(adev) &&
+		    !amdgpu_sriov_runtime(adev) &&
+		    adev->gfx.rlc.rlcg_reg_access_supported &&
+		    amdgpu_virt_get_rlcg_reg_access_flag(adev, acc_flags,
+							 GC_HWIP, false,
+							 &rlcg_flag)) {
+			ret = amdgpu_virt_rlcg_reg_rw(adev, reg, 0, rlcg_flag, xcc_id);
+		} else if (!(acc_flags & AMDGPU_REGS_NO_KIQ) &&
+		    amdgpu_sriov_runtime(adev) &&
+		    down_read_trylock(&adev->reset_domain->sem)) {
+			ret = amdgpu_kiq_rreg(adev, reg, xcc_id);
+			up_read(&adev->reset_domain->sem);
+		} else {
+			ret = readl(((void __iomem *)adev->rmmio) + (reg * 4));
+		}
+	} else {
+		ret = adev->pcie_rreg(adev, reg * 4);
+	}
+
+	return ret;
+}
+
 /*
  * MMIO register write with bytes helper functions
  * @offset:bytes offset from MMIO start
@@ -556,7 +600,7 @@ void amdgpu_device_wreg(struct amdgpu_device *adev,
 		if (!(acc_flags & AMDGPU_REGS_NO_KIQ) &&
 		    amdgpu_sriov_runtime(adev) &&
 		    down_read_trylock(&adev->reset_domain->sem)) {
-			amdgpu_kiq_wreg(adev, reg, v);
+			amdgpu_kiq_wreg(adev, reg, v, 0);
 			up_read(&adev->reset_domain->sem);
 		} else {
 			writel(v, ((void __iomem *)adev->rmmio) + (reg * 4));
@@ -598,6 +642,47 @@ void amdgpu_mm_wreg_mmio_rlc(struct amdgpu_device *adev,
 }
 
 /**
+ * amdgpu_device_xcc_wreg - write to a memory mapped IO or indirect register with specific XCC
+ *
+ * @adev: amdgpu_device pointer
+ * @reg: dword aligned register offset
+ * @v: 32 bit value to write to the register
+ * @acc_flags: access flags which require special behavior
+ * @xcc_id: xcc accelerated compute core id
+ *
+ * Writes the value specified to the offset specified.
+ */
+void amdgpu_device_xcc_wreg(struct amdgpu_device *adev,
+			uint32_t reg, uint32_t v,
+			uint32_t acc_flags, uint32_t xcc_id)
+{
+	uint32_t rlcg_flag;
+
+	if (amdgpu_device_skip_hw_access(adev))
+		return;
+
+	if ((reg * 4) < adev->rmmio_size) {
+		if (amdgpu_sriov_vf(adev) &&
+		    !amdgpu_sriov_runtime(adev) &&
+		    adev->gfx.rlc.rlcg_reg_access_supported &&
+		    amdgpu_virt_get_rlcg_reg_access_flag(adev, acc_flags,
+							 GC_HWIP, true,
+							 &rlcg_flag)) {
+			amdgpu_virt_rlcg_reg_rw(adev, reg, v, rlcg_flag, xcc_id);
+		} else if (!(acc_flags & AMDGPU_REGS_NO_KIQ) &&
+		    amdgpu_sriov_runtime(adev) &&
+		    down_read_trylock(&adev->reset_domain->sem)) {
+			amdgpu_kiq_wreg(adev, reg, v, xcc_id);
+			up_read(&adev->reset_domain->sem);
+		} else {
+			writel(v, ((void __iomem *)adev->rmmio) + (reg * 4));
+		}
+	} else {
+		adev->pcie_wreg(adev, reg * 4, v);
+	}
+}
+
+/**
  * amdgpu_device_indirect_rreg - read an indirect register
  *
  * @adev: amdgpu_device pointer
@@ -2499,6 +2584,18 @@ static int amdgpu_device_init_schedulers(struct amdgpu_device *adev)
 				  ring->name);
 			return r;
 		}
+		r = amdgpu_uvd_entity_init(adev, ring);
+		if (r) {
+			DRM_ERROR("Failed to create UVD scheduling entity on ring %s.\n",
+				  ring->name);
+			return r;
+		}
+		r = amdgpu_vce_entity_init(adev, ring);
+		if (r) {
+			DRM_ERROR("Failed to create VCE scheduling entity on ring %s.\n",
+				  ring->name);
+			return r;
+		}
 	}
 
 	amdgpu_xcp_update_partition_sched_list(adev);
@@ -4486,19 +4583,18 @@ int amdgpu_device_resume(struct drm_device *dev, bool fbcon)
 	}
 	amdgpu_fence_driver_hw_init(adev);
 
-	r = amdgpu_device_ip_late_init(adev);
-	if (r)
-		goto exit;
-
-	queue_delayed_work(system_wq, &adev->delayed_init_work,
-			   msecs_to_jiffies(AMDGPU_RESUME_MS));
-
 	if (!adev->in_s0ix) {
 		r = amdgpu_amdkfd_resume(adev, adev->in_runpm);
 		if (r)
 			goto exit;
 	}
 
+	r = amdgpu_device_ip_late_init(adev);
+	if (r)
+		goto exit;
+
+	queue_delayed_work(system_wq, &adev->delayed_init_work,
+			   msecs_to_jiffies(AMDGPU_RESUME_MS));
 exit:
 	if (amdgpu_sriov_vf(adev)) {
 		amdgpu_virt_init_data_exchange(adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
index a2a29dcb2422..b9674c57c436 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
@@ -931,12 +931,12 @@ void amdgpu_gfx_ras_error_func(struct amdgpu_device *adev,
 		func(adev, ras_error_status, i);
 }
 
-uint32_t amdgpu_kiq_rreg(struct amdgpu_device *adev, uint32_t reg)
+uint32_t amdgpu_kiq_rreg(struct amdgpu_device *adev, uint32_t reg, uint32_t xcc_id)
 {
 	signed long r, cnt = 0;
 	unsigned long flags;
 	uint32_t seq, reg_val_offs = 0, value = 0;
-	struct amdgpu_kiq *kiq = &adev->gfx.kiq[0];
+	struct amdgpu_kiq *kiq = &adev->gfx.kiq[xcc_id];
 	struct amdgpu_ring *ring = &kiq->ring;
 
 	if (amdgpu_device_skip_hw_access(adev))
@@ -999,12 +999,12 @@ failed_kiq_read:
 	return ~0;
 }
 
-void amdgpu_kiq_wreg(struct amdgpu_device *adev, uint32_t reg, uint32_t v)
+void amdgpu_kiq_wreg(struct amdgpu_device *adev, uint32_t reg, uint32_t v, uint32_t xcc_id)
 {
 	signed long r, cnt = 0;
 	unsigned long flags;
 	uint32_t seq;
-	struct amdgpu_kiq *kiq = &adev->gfx.kiq[0];
+	struct amdgpu_kiq *kiq = &adev->gfx.kiq[xcc_id];
 	struct amdgpu_ring *ring = &kiq->ring;
 
 	BUG_ON(!ring->funcs->emit_wreg);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
index 7088c5015675..f23bafec71c5 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h
@@ -521,8 +521,8 @@ int amdgpu_gfx_process_ras_data_cb(struct amdgpu_device *adev,
 int amdgpu_gfx_cp_ecc_error_irq(struct amdgpu_device *adev,
 				  struct amdgpu_irq_src *source,
 				  struct amdgpu_iv_entry *entry);
-uint32_t amdgpu_kiq_rreg(struct amdgpu_device *adev, uint32_t reg);
-void amdgpu_kiq_wreg(struct amdgpu_device *adev, uint32_t reg, uint32_t v);
+uint32_t amdgpu_kiq_rreg(struct amdgpu_device *adev, uint32_t reg, uint32_t xcc_id);
+void amdgpu_kiq_wreg(struct amdgpu_device *adev, uint32_t reg, uint32_t v, uint32_t xcc_id);
 int amdgpu_gfx_get_num_kcq(struct amdgpu_device *adev);
 void amdgpu_gfx_cp_init_microcode(struct amdgpu_device *adev, uint32_t ucode_id);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
index 2dce338b0f1e..5f71414190e9 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.c
@@ -826,7 +826,10 @@ void amdgpu_gmc_noretry_set(struct amdgpu_device *adev)
 				gc_ver == IP_VERSION(9, 4, 3) ||
 				gc_ver >= IP_VERSION(10, 3, 0));
 
-	gmc->noretry = (amdgpu_noretry == -1) ? noretry_default : amdgpu_noretry;
+	if (!amdgpu_sriov_xnack_support(adev))
+		gmc->noretry = 1;
+	else
+		gmc->noretry = (amdgpu_noretry == -1) ? noretry_default : amdgpu_noretry;
 }
 
 void amdgpu_gmc_set_vm_fault_masks(struct amdgpu_device *adev, int hub_type,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c
index 5a828c175e3a..cf33eb219e25 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.c
@@ -143,6 +143,46 @@ int amdgpu_mca_mpio_ras_sw_init(struct amdgpu_device *adev)
 	return 0;
 }
 
+void amdgpu_mca_bank_set_init(struct mca_bank_set *mca_set)
+{
+	if (!mca_set)
+		return;
+
+	memset(mca_set, 0, sizeof(*mca_set));
+	INIT_LIST_HEAD(&mca_set->list);
+}
+
+int amdgpu_mca_bank_set_add_entry(struct mca_bank_set *mca_set, struct mca_bank_entry *entry)
+{
+	struct mca_bank_node *node;
+
+	if (!entry)
+		return -EINVAL;
+
+	node = kvzalloc(sizeof(*node), GFP_KERNEL);
+	if (!node)
+		return -ENOMEM;
+
+	memcpy(&node->entry, entry, sizeof(*entry));
+
+	INIT_LIST_HEAD(&node->node);
+	list_add_tail(&node->node, &mca_set->list);
+
+	mca_set->nr_entries++;
+
+	return 0;
+}
+
+void amdgpu_mca_bank_set_release(struct mca_bank_set *mca_set)
+{
+	struct mca_bank_node *node, *tmp;
+
+	list_for_each_entry_safe(node, tmp, &mca_set->list, node) {
+		list_del(&node->node);
+		kvfree(node);
+	}
+}
+
 void amdgpu_mca_smu_init_funcs(struct amdgpu_device *adev, const struct amdgpu_mca_smu_funcs *mca_funcs)
 {
 	struct amdgpu_mca *mca = &adev->mca;
@@ -160,6 +200,65 @@ int amdgpu_mca_smu_set_debug_mode(struct amdgpu_device *adev, bool enable)
 	return -EOPNOTSUPP;
 }
 
+static void amdgpu_mca_smu_mca_bank_dump(struct amdgpu_device *adev, int idx, struct mca_bank_entry *entry)
+{
+	dev_info(adev->dev, "[Hardware error] Accelerator Check Architecture events logged\n");
+	dev_info(adev->dev, "[Hardware error] aca entry[%02d].STATUS=0x%016llx\n",
+		 idx, entry->regs[MCA_REG_IDX_STATUS]);
+	dev_info(adev->dev, "[Hardware error] aca entry[%02d].ADDR=0x%016llx\n",
+		 idx, entry->regs[MCA_REG_IDX_ADDR]);
+	dev_info(adev->dev, "[Hardware error] aca entry[%02d].MISC0=0x%016llx\n",
+		 idx, entry->regs[MCA_REG_IDX_MISC0]);
+	dev_info(adev->dev, "[Hardware error] aca entry[%02d].IPID=0x%016llx\n",
+		 idx, entry->regs[MCA_REG_IDX_IPID]);
+	dev_info(adev->dev, "[Hardware error] aca entry[%02d].SYND=0x%016llx\n",
+		 idx, entry->regs[MCA_REG_IDX_SYND]);
+}
+
+int amdgpu_mca_smu_log_ras_error(struct amdgpu_device *adev, enum amdgpu_ras_block blk, enum amdgpu_mca_error_type type, struct ras_err_data *err_data)
+{
+	struct amdgpu_smuio_mcm_config_info mcm_info;
+	struct mca_bank_set mca_set;
+	struct mca_bank_node *node;
+	struct mca_bank_entry *entry;
+	uint32_t count;
+	int ret, i = 0;
+
+	amdgpu_mca_bank_set_init(&mca_set);
+
+	ret = amdgpu_mca_smu_get_mca_set(adev, blk, type, &mca_set);
+	if (ret)
+		goto out_mca_release;
+
+	list_for_each_entry(node, &mca_set.list, node) {
+		entry = &node->entry;
+
+		amdgpu_mca_smu_mca_bank_dump(adev, i++, entry);
+
+		count = 0;
+		ret = amdgpu_mca_smu_parse_mca_error_count(adev, blk, type, entry, &count);
+		if (ret)
+			goto out_mca_release;
+
+		if (!count)
+			continue;
+
+		mcm_info.socket_id = entry->info.socket_id;
+		mcm_info.die_id = entry->info.aid;
+
+		if (type == AMDGPU_MCA_ERROR_TYPE_UE)
+			amdgpu_ras_error_statistic_ue_count(err_data, &mcm_info, (uint64_t)count);
+		else
+			amdgpu_ras_error_statistic_ce_count(err_data, &mcm_info, (uint64_t)count);
+	}
+
+out_mca_release:
+	amdgpu_mca_bank_set_release(&mca_set);
+
+	return ret;
+}
+
+
 int amdgpu_mca_smu_get_valid_mca_count(struct amdgpu_device *adev, enum amdgpu_mca_error_type type, uint32_t *count)
 {
 	const struct amdgpu_mca_smu_funcs *mca_funcs = adev->mca.mca_funcs;
@@ -173,17 +272,77 @@ int amdgpu_mca_smu_get_valid_mca_count(struct amdgpu_device *adev, enum amdgpu_m
 	return -EOPNOTSUPP;
 }
 
-int amdgpu_mca_smu_get_error_count(struct amdgpu_device *adev, enum amdgpu_ras_block blk,
-				   enum amdgpu_mca_error_type type, uint32_t *count)
+int amdgpu_mca_smu_get_mca_set_error_count(struct amdgpu_device *adev, enum amdgpu_ras_block blk,
+					    enum amdgpu_mca_error_type type, uint32_t *total)
 {
 	const struct amdgpu_mca_smu_funcs *mca_funcs = adev->mca.mca_funcs;
-	if (!count)
+	struct mca_bank_set mca_set;
+	struct mca_bank_node *node;
+	struct mca_bank_entry *entry;
+	uint32_t count;
+	int ret;
+
+	if (!total)
 		return -EINVAL;
 
-	if (mca_funcs && mca_funcs->mca_get_error_count)
-		return mca_funcs->mca_get_error_count(adev, blk, type, count);
+	if (!mca_funcs)
+		return -EOPNOTSUPP;
 
-	return -EOPNOTSUPP;
+	if (!mca_funcs->mca_get_ras_mca_set || !mca_funcs->mca_get_valid_mca_count)
+		return -EOPNOTSUPP;
+
+	amdgpu_mca_bank_set_init(&mca_set);
+
+	ret = mca_funcs->mca_get_ras_mca_set(adev, blk, type, &mca_set);
+	if (ret)
+		goto err_mca_set_release;
+
+	*total = 0;
+	list_for_each_entry(node, &mca_set.list, node) {
+		entry = &node->entry;
+
+		count = 0;
+		ret = mca_funcs->mca_parse_mca_error_count(adev, blk, type, entry, &count);
+		if (ret)
+			goto err_mca_set_release;
+
+		*total += count;
+	}
+
+err_mca_set_release:
+	amdgpu_mca_bank_set_release(&mca_set);
+
+	return ret;
+}
+
+int amdgpu_mca_smu_parse_mca_error_count(struct amdgpu_device *adev, enum amdgpu_ras_block blk,
+					 enum amdgpu_mca_error_type type, struct mca_bank_entry *entry, uint32_t *count)
+{
+	const struct amdgpu_mca_smu_funcs *mca_funcs = adev->mca.mca_funcs;
+	if (!count || !entry)
+		return -EINVAL;
+
+	if (!mca_funcs || !mca_funcs->mca_parse_mca_error_count)
+		return -EOPNOTSUPP;
+
+
+	return mca_funcs->mca_parse_mca_error_count(adev, blk, type, entry, count);
+}
+
+int amdgpu_mca_smu_get_mca_set(struct amdgpu_device *adev, enum amdgpu_ras_block blk,
+			       enum amdgpu_mca_error_type type, struct mca_bank_set *mca_set)
+{
+	const struct amdgpu_mca_smu_funcs *mca_funcs = adev->mca.mca_funcs;
+
+	if (!mca_set)
+		return -EINVAL;
+
+	if (!mca_funcs || !mca_funcs->mca_get_ras_mca_set)
+		return -EOPNOTSUPP;
+
+	WARN_ON(!list_empty(&mca_set->list));
+
+	return mca_funcs->mca_get_ras_mca_set(adev, blk, type, mca_set);
 }
 
 int amdgpu_mca_smu_get_mca_entry(struct amdgpu_device *adev, enum amdgpu_mca_error_type type,
@@ -230,14 +389,21 @@ static int amdgpu_mca_smu_debug_mode_set(void *data, u64 val)
 static void mca_dump_entry(struct seq_file *m, struct mca_bank_entry *entry)
 {
 	int i, idx = entry->idx;
+	int reg_idx_array[] = {
+		MCA_REG_IDX_STATUS,
+		MCA_REG_IDX_ADDR,
+		MCA_REG_IDX_MISC0,
+		MCA_REG_IDX_IPID,
+		MCA_REG_IDX_SYND,
+	};
 
 	seq_printf(m, "mca entry[%d].type: %s\n", idx, entry->type == AMDGPU_MCA_ERROR_TYPE_UE ? "UE" : "CE");
 	seq_printf(m, "mca entry[%d].ip: %d\n", idx, entry->ip);
 	seq_printf(m, "mca entry[%d].info: socketid:%d aid:%d hwid:0x%03x mcatype:0x%04x\n",
 		   idx, entry->info.socket_id, entry->info.aid, entry->info.hwid, entry->info.mcatype);
 
-	for (i = 0; i < ARRAY_SIZE(entry->regs); i++)
-		seq_printf(m, "mca entry[%d].regs[%d]: 0x%016llx\n", idx, i, entry->regs[i]);
+	for (i = 0; i < ARRAY_SIZE(reg_idx_array); i++)
+		seq_printf(m, "mca entry[%d].regs[%d]: 0x%016llx\n", idx, reg_idx_array[i], entry->regs[reg_idx_array[i]]);
 }
 
 static int mca_dump_show(struct seq_file *m, enum amdgpu_mca_error_type type)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.h
index 28ad463cf5c9..2b488fcf2f95 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mca.h
@@ -25,6 +25,27 @@
 
 #define MCA_MAX_REGS_COUNT	(16)
 
+#define MCA_REG_FIELD(x, h, l)			(((x) & GENMASK_ULL(h, l)) >> l)
+#define MCA_REG__STATUS__VAL(x)			MCA_REG_FIELD(x, 63, 63)
+#define MCA_REG__STATUS__OVERFLOW(x)		MCA_REG_FIELD(x, 62, 62)
+#define MCA_REG__STATUS__UC(x)			MCA_REG_FIELD(x, 61, 61)
+#define MCA_REG__STATUS__EN(x)			MCA_REG_FIELD(x, 60, 60)
+#define MCA_REG__STATUS__MISCV(x)		MCA_REG_FIELD(x, 59, 59)
+#define MCA_REG__STATUS__ADDRV(x)		MCA_REG_FIELD(x, 58, 58)
+#define MCA_REG__STATUS__PCC(x)			MCA_REG_FIELD(x, 57, 57)
+#define MCA_REG__STATUS__ERRCOREIDVAL(x)	MCA_REG_FIELD(x, 56, 56)
+#define MCA_REG__STATUS__TCC(x)			MCA_REG_FIELD(x, 55, 55)
+#define MCA_REG__STATUS__SYNDV(x)		MCA_REG_FIELD(x, 53, 53)
+#define MCA_REG__STATUS__CECC(x)		MCA_REG_FIELD(x, 46, 46)
+#define MCA_REG__STATUS__UECC(x)		MCA_REG_FIELD(x, 45, 45)
+#define MCA_REG__STATUS__DEFERRED(x)		MCA_REG_FIELD(x, 44, 44)
+#define MCA_REG__STATUS__POISON(x)		MCA_REG_FIELD(x, 43, 43)
+#define MCA_REG__STATUS__SCRUB(x)		MCA_REG_FIELD(x, 40, 40)
+#define MCA_REG__STATUS__ERRCOREID(x)		MCA_REG_FIELD(x, 37, 32)
+#define MCA_REG__STATUS__ADDRLSB(x)		MCA_REG_FIELD(x, 29, 24)
+#define MCA_REG__STATUS__ERRORCODEEXT(x)	MCA_REG_FIELD(x, 21, 16)
+#define MCA_REG__STATUS__ERRORCODE(x)		MCA_REG_FIELD(x, 15, 0)
+
 enum amdgpu_mca_ip {
 	AMDGPU_MCA_IP_UNKNOW = -1,
 	AMDGPU_MCA_IP_PSP = 0,
@@ -33,6 +54,7 @@ enum amdgpu_mca_ip {
 	AMDGPU_MCA_IP_SMU,
 	AMDGPU_MCA_IP_MP5,
 	AMDGPU_MCA_IP_UMC,
+	AMDGPU_MCA_IP_PCS_XGMI,
 	AMDGPU_MCA_IP_COUNT,
 };
 
@@ -57,6 +79,15 @@ struct amdgpu_mca {
 	const struct amdgpu_mca_smu_funcs *mca_funcs;
 };
 
+enum mca_reg_idx {
+	MCA_REG_IDX_STATUS		= 1,
+	MCA_REG_IDX_ADDR		= 2,
+	MCA_REG_IDX_MISC0		= 3,
+	MCA_REG_IDX_IPID		= 5,
+	MCA_REG_IDX_SYND		= 6,
+	MCA_REG_IDX_COUNT		= 16,
+};
+
 struct mca_bank_info {
 	int socket_id;
 	int aid;
@@ -72,18 +103,28 @@ struct mca_bank_entry {
 	uint64_t regs[MCA_MAX_REGS_COUNT];
 };
 
+struct mca_bank_node {
+	struct mca_bank_entry entry;
+	struct list_head node;
+};
+
+struct mca_bank_set {
+	int nr_entries;
+	struct list_head list;
+};
+
 struct amdgpu_mca_smu_funcs {
 	int max_ue_count;
 	int max_ce_count;
 	int (*mca_set_debug_mode)(struct amdgpu_device *adev, bool enable);
-	int (*mca_get_error_count)(struct amdgpu_device *adev, enum amdgpu_ras_block blk,
-				   enum amdgpu_mca_error_type type, uint32_t *count);
+	int (*mca_get_ras_mca_set)(struct amdgpu_device *adev, enum amdgpu_ras_block blk, enum amdgpu_mca_error_type type,
+				   struct mca_bank_set *mca_set);
+	int (*mca_parse_mca_error_count)(struct amdgpu_device *adev, enum amdgpu_ras_block blk, enum amdgpu_mca_error_type type,
+					 struct mca_bank_entry *entry, uint32_t *count);
 	int (*mca_get_valid_mca_count)(struct amdgpu_device *adev, enum amdgpu_mca_error_type type,
 				       uint32_t *count);
 	int (*mca_get_mca_entry)(struct amdgpu_device *adev, enum amdgpu_mca_error_type type,
 				 int idx, struct mca_bank_entry *entry);
-	int (*mca_get_ras_mca_idx_array)(struct amdgpu_device *adev, enum amdgpu_ras_block blk,
-					 enum amdgpu_mca_error_type type, int *idx_array, int *idx_array_size);
 };
 
 void amdgpu_mca_query_correctable_error_count(struct amdgpu_device *adev,
@@ -107,11 +148,22 @@ int amdgpu_mca_mpio_ras_sw_init(struct amdgpu_device *adev);
 void amdgpu_mca_smu_init_funcs(struct amdgpu_device *adev, const struct amdgpu_mca_smu_funcs *mca_funcs);
 int amdgpu_mca_smu_set_debug_mode(struct amdgpu_device *adev, bool enable);
 int amdgpu_mca_smu_get_valid_mca_count(struct amdgpu_device *adev, enum amdgpu_mca_error_type type, uint32_t *count);
+int amdgpu_mca_smu_get_mca_set_error_count(struct amdgpu_device *adev, enum amdgpu_ras_block blk,
+					   enum amdgpu_mca_error_type type, uint32_t *total);
 int amdgpu_mca_smu_get_error_count(struct amdgpu_device *adev, enum amdgpu_ras_block blk,
 				   enum amdgpu_mca_error_type type, uint32_t *count);
+int amdgpu_mca_smu_parse_mca_error_count(struct amdgpu_device *adev, enum amdgpu_ras_block blk,
+					 enum amdgpu_mca_error_type type, struct mca_bank_entry *entry, uint32_t *count);
+int amdgpu_mca_smu_get_mca_set(struct amdgpu_device *adev, enum amdgpu_ras_block blk,
+			       enum amdgpu_mca_error_type type, struct mca_bank_set *mca_set);
 int amdgpu_mca_smu_get_mca_entry(struct amdgpu_device *adev, enum amdgpu_mca_error_type type,
 				 int idx, struct mca_bank_entry *entry);
 
 void amdgpu_mca_smu_debugfs_init(struct amdgpu_device *adev, struct dentry *root);
 
+void amdgpu_mca_bank_set_init(struct mca_bank_set *mca_set);
+int amdgpu_mca_bank_set_add_entry(struct mca_bank_set *mca_set, struct mca_bank_entry *entry);
+void amdgpu_mca_bank_set_release(struct mca_bank_set *mca_set);
+int amdgpu_mca_smu_log_ras_error(struct amdgpu_device *adev, enum amdgpu_ras_block blk, enum amdgpu_mca_error_type type, struct ras_err_data *err_data);
+
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index 0dcb6c36b02c..cef920a93924 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -1062,9 +1062,6 @@ static const char * const amdgpu_vram_names[] = {
  */
 int amdgpu_bo_init(struct amdgpu_device *adev)
 {
-	/* set the default AGP aperture state */
-	amdgpu_gmc_set_agp_default(adev, &adev->gmc);
-
 	/* On A+A platform, VRAM can be mapped as WB */
 	if (!adev->gmc.xgmi.connected_to_cpu && !adev->gmc.is_app_apu) {
 		/* reserve PAT memory space to WC for VRAM */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
index b7fe5951b166..84e5987b14e0 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
@@ -1165,13 +1165,53 @@ static void amdgpu_rasmgr_error_data_statistic_update(struct ras_manager *obj, s
 	}
 }
 
-/* query/inject/cure begin */
-int amdgpu_ras_query_error_status(struct amdgpu_device *adev,
-				  struct ras_query_if *info)
+static int amdgpu_ras_query_error_status_helper(struct amdgpu_device *adev,
+						struct ras_query_if *info,
+						struct ras_err_data *err_data,
+						unsigned int error_query_mode)
 {
+	enum amdgpu_ras_block blk = info ? info->head.block : AMDGPU_RAS_BLOCK_COUNT;
 	struct amdgpu_ras_block_object *block_obj = NULL;
+
+	if (error_query_mode == AMDGPU_RAS_INVALID_ERROR_QUERY)
+		return -EINVAL;
+
+	if (error_query_mode == AMDGPU_RAS_DIRECT_ERROR_QUERY) {
+		if (info->head.block == AMDGPU_RAS_BLOCK__UMC) {
+			amdgpu_ras_get_ecc_info(adev, err_data);
+		} else {
+			block_obj = amdgpu_ras_get_ras_block(adev, info->head.block, 0);
+			if (!block_obj || !block_obj->hw_ops) {
+				dev_dbg_once(adev->dev, "%s doesn't config RAS function\n",
+					     get_ras_block_str(&info->head));
+				return -EINVAL;
+			}
+
+			if (block_obj->hw_ops->query_ras_error_count)
+				block_obj->hw_ops->query_ras_error_count(adev, &err_data);
+
+			if ((info->head.block == AMDGPU_RAS_BLOCK__SDMA) ||
+			    (info->head.block == AMDGPU_RAS_BLOCK__GFX) ||
+			    (info->head.block == AMDGPU_RAS_BLOCK__MMHUB)) {
+				if (block_obj->hw_ops->query_ras_error_status)
+					block_obj->hw_ops->query_ras_error_status(adev);
+			}
+		}
+	} else {
+		/* FIXME: add code to check return value later */
+		amdgpu_mca_smu_log_ras_error(adev, blk, AMDGPU_MCA_ERROR_TYPE_UE, err_data);
+		amdgpu_mca_smu_log_ras_error(adev, blk, AMDGPU_MCA_ERROR_TYPE_CE, err_data);
+	}
+
+	return 0;
+}
+
+/* query/inject/cure begin */
+int amdgpu_ras_query_error_status(struct amdgpu_device *adev, struct ras_query_if *info)
+{
 	struct ras_manager *obj = amdgpu_ras_find_obj(adev, &info->head);
 	struct ras_err_data err_data;
+	unsigned int error_query_mode;
 	int ret;
 
 	if (!obj)
@@ -1181,27 +1221,14 @@ int amdgpu_ras_query_error_status(struct amdgpu_device *adev,
 	if (ret)
 		return ret;
 
-	if (info->head.block == AMDGPU_RAS_BLOCK__UMC) {
-		amdgpu_ras_get_ecc_info(adev, &err_data);
-	} else {
-		block_obj = amdgpu_ras_get_ras_block(adev, info->head.block, 0);
-		if (!block_obj || !block_obj->hw_ops)   {
-			dev_dbg_once(adev->dev, "%s doesn't config RAS function\n",
-				     get_ras_block_str(&info->head));
-			ret = -EINVAL;
-			goto out_fini_err_data;
-		}
-
-		if (block_obj->hw_ops->query_ras_error_count)
-			block_obj->hw_ops->query_ras_error_count(adev, &err_data);
+	if (!amdgpu_ras_get_error_query_mode(adev, &error_query_mode))
+		return -EINVAL;
 
-		if ((info->head.block == AMDGPU_RAS_BLOCK__SDMA) ||
-		    (info->head.block == AMDGPU_RAS_BLOCK__GFX) ||
-		    (info->head.block == AMDGPU_RAS_BLOCK__MMHUB)) {
-				if (block_obj->hw_ops->query_ras_error_status)
-					block_obj->hw_ops->query_ras_error_status(adev);
-			}
-	}
+	ret = amdgpu_ras_query_error_status_helper(adev, info,
+						   &err_data,
+						   error_query_mode);
+	if (ret)
+		goto out_fini_err_data;
 
 	amdgpu_rasmgr_error_data_statistic_update(obj, &err_data);
 
@@ -1537,7 +1564,8 @@ static void amdgpu_ras_sysfs_remove_bad_page_node(struct amdgpu_device *adev)
 {
 	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
 
-	sysfs_remove_file_from_group(&adev->dev->kobj,
+	if (adev->dev->kobj.sd)
+		sysfs_remove_file_from_group(&adev->dev->kobj,
 				&con->badpages_attr.attr,
 				RAS_FS_NAME);
 }
@@ -1556,7 +1584,8 @@ static int amdgpu_ras_sysfs_remove_dev_attr_node(struct amdgpu_device *adev)
 		.attrs = attrs,
 	};
 
-	sysfs_remove_group(&adev->dev->kobj, &group);
+	if (adev->dev->kobj.sd)
+		sysfs_remove_group(&adev->dev->kobj, &group);
 
 	return 0;
 }
@@ -1603,7 +1632,8 @@ int amdgpu_ras_sysfs_remove(struct amdgpu_device *adev,
 	if (!obj || !obj->attr_inuse)
 		return -EINVAL;
 
-	sysfs_remove_file_from_group(&adev->dev->kobj,
+	if (adev->dev->kobj.sd)
+		sysfs_remove_file_from_group(&adev->dev->kobj,
 				&obj->sysfs_attr.attr,
 				RAS_FS_NAME);
 	obj->attr_inuse = 0;
@@ -3397,6 +3427,26 @@ bool amdgpu_ras_get_mca_debug_mode(struct amdgpu_device *adev)
 		return true;
 }
 
+bool amdgpu_ras_get_error_query_mode(struct amdgpu_device *adev,
+				     unsigned int *error_query_mode)
+{
+	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
+	const struct amdgpu_mca_smu_funcs *mca_funcs = adev->mca.mca_funcs;
+
+	if (!con) {
+		*error_query_mode = AMDGPU_RAS_INVALID_ERROR_QUERY;
+		return false;
+	}
+
+	if (mca_funcs && mca_funcs->mca_set_debug_mode)
+		*error_query_mode =
+			(con->is_mca_debug_mode) ? AMDGPU_RAS_DIRECT_ERROR_QUERY : AMDGPU_RAS_FIRMWARE_ERROR_QUERY;
+	else
+		*error_query_mode = AMDGPU_RAS_DIRECT_ERROR_QUERY;
+
+	return true;
+}
+
 /* Register each ip ras block into amdgpu ras */
 int amdgpu_ras_register_ras_block(struct amdgpu_device *adev,
 		struct amdgpu_ras_block_object *ras_block_obj)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
index 665414c22ca9..19161916ac46 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h
@@ -320,6 +320,12 @@ enum amdgpu_ras_ret {
 	AMDGPU_RAS_PT,
 };
 
+enum amdgpu_ras_error_query_mode {
+	AMDGPU_RAS_INVALID_ERROR_QUERY		= 0,
+	AMDGPU_RAS_DIRECT_ERROR_QUERY		= 1,
+	AMDGPU_RAS_FIRMWARE_ERROR_QUERY		= 2,
+};
+
 /* ras error status reisger fields */
 #define ERR_STATUS_LO__ERR_STATUS_VALID_FLAG__SHIFT	0x0
 #define ERR_STATUS_LO__ERR_STATUS_VALID_FLAG_MASK	0x00000001L
@@ -769,6 +775,8 @@ int amdgpu_ras_set_context(struct amdgpu_device *adev, struct amdgpu_ras *ras_co
 
 void amdgpu_ras_set_mca_debug_mode(struct amdgpu_device *adev, bool enable);
 bool amdgpu_ras_get_mca_debug_mode(struct amdgpu_device *adev);
+bool amdgpu_ras_get_error_query_mode(struct amdgpu_device *adev,
+				     unsigned int *mode);
 
 int amdgpu_ras_register_ras_block(struct amdgpu_device *adev,
 				struct amdgpu_ras_block_object *ras_block_obj);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index 815b7c34ed33..65949cc7abb9 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -399,20 +399,20 @@ int amdgpu_uvd_sw_fini(struct amdgpu_device *adev)
  *
  * @adev: amdgpu_device pointer
  *
+ * Initialize the entity used for handle management in the kernel driver.
  */
-int amdgpu_uvd_entity_init(struct amdgpu_device *adev)
+int amdgpu_uvd_entity_init(struct amdgpu_device *adev, struct amdgpu_ring *ring)
 {
-	struct amdgpu_ring *ring;
-	struct drm_gpu_scheduler *sched;
-	int r;
+	if (ring == &adev->uvd.inst[0].ring) {
+		struct drm_gpu_scheduler *sched = &ring->sched;
+		int r;
 
-	ring = &adev->uvd.inst[0].ring;
-	sched = &ring->sched;
-	r = drm_sched_entity_init(&adev->uvd.entity, DRM_SCHED_PRIORITY_NORMAL,
-				  &sched, 1, NULL);
-	if (r) {
-		DRM_ERROR("Failed setting up UVD kernel entity.\n");
-		return r;
+		r = drm_sched_entity_init(&adev->uvd.entity, DRM_SCHED_PRIORITY_NORMAL,
+					  &sched, 1, NULL);
+		if (r) {
+			DRM_ERROR("Failed setting up UVD kernel entity.\n");
+			return r;
+		}
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h
index a9f342537c68..9dfad2f48ef4 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h
@@ -73,7 +73,7 @@ struct amdgpu_uvd {
 
 int amdgpu_uvd_sw_init(struct amdgpu_device *adev);
 int amdgpu_uvd_sw_fini(struct amdgpu_device *adev);
-int amdgpu_uvd_entity_init(struct amdgpu_device *adev);
+int amdgpu_uvd_entity_init(struct amdgpu_device *adev, struct amdgpu_ring *ring);
 int amdgpu_uvd_prepare_suspend(struct amdgpu_device *adev);
 int amdgpu_uvd_suspend(struct amdgpu_device *adev);
 int amdgpu_uvd_resume(struct amdgpu_device *adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
index 1904edf68407..0954447f689d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
@@ -231,20 +231,20 @@ int amdgpu_vce_sw_fini(struct amdgpu_device *adev)
  *
  * @adev: amdgpu_device pointer
  *
+ * Initialize the entity used for handle management in the kernel driver.
  */
-int amdgpu_vce_entity_init(struct amdgpu_device *adev)
+int amdgpu_vce_entity_init(struct amdgpu_device *adev, struct amdgpu_ring *ring)
 {
-	struct amdgpu_ring *ring;
-	struct drm_gpu_scheduler *sched;
-	int r;
-
-	ring = &adev->vce.ring[0];
-	sched = &ring->sched;
-	r = drm_sched_entity_init(&adev->vce.entity, DRM_SCHED_PRIORITY_NORMAL,
-				  &sched, 1, NULL);
-	if (r != 0) {
-		DRM_ERROR("Failed setting up VCE run queue.\n");
-		return r;
+	if (ring == &adev->vce.ring[0]) {
+		struct drm_gpu_scheduler *sched = &ring->sched;
+		int r;
+
+		r = drm_sched_entity_init(&adev->vce.entity, DRM_SCHED_PRIORITY_NORMAL,
+					  &sched, 1, NULL);
+		if (r != 0) {
+			DRM_ERROR("Failed setting up VCE run queue.\n");
+			return r;
+		}
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h
index ea680fc9a6c3..6e53f872d084 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.h
@@ -55,7 +55,7 @@ struct amdgpu_vce {
 
 int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size);
 int amdgpu_vce_sw_fini(struct amdgpu_device *adev);
-int amdgpu_vce_entity_init(struct amdgpu_device *adev);
+int amdgpu_vce_entity_init(struct amdgpu_device *adev, struct amdgpu_ring *ring);
 int amdgpu_vce_suspend(struct amdgpu_device *adev);
 int amdgpu_vce_resume(struct amdgpu_device *adev);
 void amdgpu_vce_free_handles(struct amdgpu_device *adev, struct drm_file *filp);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
index a0aa624f5a92..3a632c3b1a2c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
@@ -73,9 +73,10 @@ void amdgpu_virt_init_setting(struct amdgpu_device *adev)
 
 void amdgpu_virt_kiq_reg_write_reg_wait(struct amdgpu_device *adev,
 					uint32_t reg0, uint32_t reg1,
-					uint32_t ref, uint32_t mask)
+					uint32_t ref, uint32_t mask,
+					uint32_t xcc_inst)
 {
-	struct amdgpu_kiq *kiq = &adev->gfx.kiq[0];
+	struct amdgpu_kiq *kiq = &adev->gfx.kiq[xcc_inst];
 	struct amdgpu_ring *ring = &kiq->ring;
 	signed long r, cnt = 0;
 	unsigned long flags;
@@ -942,7 +943,7 @@ void amdgpu_virt_update_sriov_video_codec(struct amdgpu_device *adev,
 	}
 }
 
-static bool amdgpu_virt_get_rlcg_reg_access_flag(struct amdgpu_device *adev,
+bool amdgpu_virt_get_rlcg_reg_access_flag(struct amdgpu_device *adev,
 						 u32 acc_flags, u32 hwip,
 						 bool write, u32 *rlcg_flag)
 {
@@ -975,7 +976,7 @@ static bool amdgpu_virt_get_rlcg_reg_access_flag(struct amdgpu_device *adev,
 	return ret;
 }
 
-static u32 amdgpu_virt_rlcg_reg_rw(struct amdgpu_device *adev, u32 offset, u32 v, u32 flag, u32 xcc_id)
+u32 amdgpu_virt_rlcg_reg_rw(struct amdgpu_device *adev, u32 offset, u32 v, u32 flag, u32 xcc_id)
 {
 	struct amdgpu_rlcg_reg_access_ctrl *reg_access_ctrl;
 	uint32_t timeout = 50000;
@@ -1093,3 +1094,13 @@ u32 amdgpu_sriov_rreg(struct amdgpu_device *adev,
 	else
 		return RREG32(offset);
 }
+
+bool amdgpu_sriov_xnack_support(struct amdgpu_device *adev)
+{
+	bool xnack_mode = true;
+
+	if (amdgpu_sriov_vf(adev) && adev->ip_versions[GC_HWIP][0] == IP_VERSION(9, 4, 2))
+		xnack_mode = false;
+
+	return xnack_mode;
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h
index 858ef21ae515..d4207e44141f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h
@@ -334,7 +334,8 @@ bool amdgpu_virt_mmio_blocked(struct amdgpu_device *adev);
 void amdgpu_virt_init_setting(struct amdgpu_device *adev);
 void amdgpu_virt_kiq_reg_write_reg_wait(struct amdgpu_device *adev,
 					uint32_t reg0, uint32_t rreg1,
-					uint32_t ref, uint32_t mask);
+					uint32_t ref, uint32_t mask,
+					uint32_t xcc_inst);
 int amdgpu_virt_request_full_gpu(struct amdgpu_device *adev, bool init);
 int amdgpu_virt_release_full_gpu(struct amdgpu_device *adev, bool init);
 int amdgpu_virt_reset_gpu(struct amdgpu_device *adev);
@@ -365,4 +366,9 @@ u32 amdgpu_sriov_rreg(struct amdgpu_device *adev,
 bool amdgpu_virt_fw_load_skip_check(struct amdgpu_device *adev,
 			uint32_t ucode_id);
 void amdgpu_virt_post_reset(struct amdgpu_device *adev);
+bool amdgpu_sriov_xnack_support(struct amdgpu_device *adev);
+bool amdgpu_virt_get_rlcg_reg_access_flag(struct amdgpu_device *adev,
+					  u32 acc_flags, u32 hwip,
+					  bool write, u32 *rlcg_flag);
+u32 amdgpu_virt_rlcg_reg_rw(struct amdgpu_device *adev, u32 offset, u32 v, u32 flag, u32 xcc_id);
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index 904252456d25..d1b8afd105c9 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -1098,8 +1098,8 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va,
 				bo = gem_to_amdgpu_bo(gobj);
 		}
 		mem = bo->tbo.resource;
-		if (mem->mem_type == TTM_PL_TT ||
-		    mem->mem_type == AMDGPU_PL_PREEMPT)
+		if (mem && (mem->mem_type == TTM_PL_TT ||
+			    mem->mem_type == AMDGPU_PL_PREEMPT))
 			pages_addr = bo->tbo.ttm->dma_address;
 	}
 
@@ -2139,7 +2139,8 @@ long amdgpu_vm_wait_idle(struct amdgpu_vm *vm, long timeout)
  * Returns:
  * 0 for success, error for failure.
  */
-int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, int32_t xcp_id)
+int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
+		   int32_t xcp_id)
 {
 	struct amdgpu_bo *root_bo;
 	struct amdgpu_bo_vm *root;
@@ -2158,6 +2159,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, int32_t xcp
 	INIT_LIST_HEAD(&vm->done);
 	INIT_LIST_HEAD(&vm->pt_freed);
 	INIT_WORK(&vm->pt_free_work, amdgpu_vm_pt_free_work);
+	INIT_KFIFO(vm->faults);
 
 	r = amdgpu_vm_init_entities(adev, vm);
 	if (r)
@@ -2192,34 +2194,33 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, int32_t xcp
 				false, &root, xcp_id);
 	if (r)
 		goto error_free_delayed;
-	root_bo = &root->bo;
+
+	root_bo = amdgpu_bo_ref(&root->bo);
 	r = amdgpu_bo_reserve(root_bo, true);
-	if (r)
-		goto error_free_root;
+	if (r) {
+		amdgpu_bo_unref(&root->shadow);
+		amdgpu_bo_unref(&root_bo);
+		goto error_free_delayed;
+	}
 
+	amdgpu_vm_bo_base_init(&vm->root, vm, root_bo);
 	r = dma_resv_reserve_fences(root_bo->tbo.base.resv, 1);
 	if (r)
-		goto error_unreserve;
-
-	amdgpu_vm_bo_base_init(&vm->root, vm, root_bo);
+		goto error_free_root;
 
 	r = amdgpu_vm_pt_clear(adev, vm, root, false);
 	if (r)
-		goto error_unreserve;
+		goto error_free_root;
 
 	amdgpu_bo_unreserve(vm->root.bo);
-
-	INIT_KFIFO(vm->faults);
+	amdgpu_bo_unref(&root_bo);
 
 	return 0;
 
-error_unreserve:
-	amdgpu_bo_unreserve(vm->root.bo);
-
 error_free_root:
-	amdgpu_bo_unref(&root->shadow);
+	amdgpu_vm_pt_free_root(adev, vm);
+	amdgpu_bo_unreserve(vm->root.bo);
 	amdgpu_bo_unref(&root_bo);
-	vm->root.bo = NULL;
 
 error_free_delayed:
 	dma_fence_put(vm->last_tlb_flush);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
index 9d5d742ee9d3..bd20cb3b9819 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
@@ -103,6 +103,53 @@ static const int walf_pcs_err_noncorrectable_mask_reg_aldebaran[] = {
 	smnPCS_GOPX1_PCS_ERROR_NONCORRECTABLE_MASK + 0x100000
 };
 
+static const int xgmi3x16_pcs_err_status_reg_v6_4[] = {
+	smnPCS_XGMI3X16_PCS_ERROR_STATUS,
+	smnPCS_XGMI3X16_PCS_ERROR_STATUS + 0x100000
+};
+
+static const int xgmi3x16_pcs_err_noncorrectable_mask_reg_v6_4[] = {
+	smnPCS_XGMI3X16_PCS_ERROR_NONCORRECTABLE_MASK,
+	smnPCS_XGMI3X16_PCS_ERROR_NONCORRECTABLE_MASK + 0x100000
+};
+
+static const u64 xgmi_v6_4_0_mca_base_array[] = {
+	0x11a09200,
+	0x11b09200,
+};
+
+static const char *xgmi_v6_4_0_ras_error_code_ext[32] = {
+	[0x00] = "XGMI PCS DataLossErr",
+	[0x01] = "XGMI PCS TrainingErr",
+	[0x02] = "XGMI PCS FlowCtrlAckErr",
+	[0x03] = "XGMI PCS RxFifoUnderflowErr",
+	[0x04] = "XGMI PCS RxFifoOverflowErr",
+	[0x05] = "XGMI PCS CRCErr",
+	[0x06] = "XGMI PCS BERExceededErr",
+	[0x07] = "XGMI PCS TxMetaDataErr",
+	[0x08] = "XGMI PCS ReplayBufParityErr",
+	[0x09] = "XGMI PCS DataParityErr",
+	[0x0a] = "XGMI PCS ReplayFifoOverflowErr",
+	[0x0b] = "XGMI PCS ReplayFifoUnderflowErr",
+	[0x0c] = "XGMI PCS ElasticFifoOverflowErr",
+	[0x0d] = "XGMI PCS DeskewErr",
+	[0x0e] = "XGMI PCS FlowCtrlCRCErr",
+	[0x0f] = "XGMI PCS DataStartupLimitErr",
+	[0x10] = "XGMI PCS FCInitTimeoutErr",
+	[0x11] = "XGMI PCS RecoveryTimeoutErr",
+	[0x12] = "XGMI PCS ReadySerialTimeoutErr",
+	[0x13] = "XGMI PCS ReadySerialAttemptErr",
+	[0x14] = "XGMI PCS RecoveryAttemptErr",
+	[0x15] = "XGMI PCS RecoveryRelockAttemptErr",
+	[0x16] = "XGMI PCS ReplayAttemptErr",
+	[0x17] = "XGMI PCS SyncHdrErr",
+	[0x18] = "XGMI PCS TxReplayTimeoutErr",
+	[0x19] = "XGMI PCS RxReplayTimeoutErr",
+	[0x1a] = "XGMI PCS LinkSubTxTimeoutErr",
+	[0x1b] = "XGMI PCS LinkSubRxTimeoutErr",
+	[0x1c] = "XGMI PCS RxCMDPktErr",
+};
+
 static const struct amdgpu_pcs_ras_field xgmi_pcs_ras_fields[] = {
 	{"XGMI PCS DataLossErr",
 	 SOC15_REG_FIELD(XGMI0_PCS_GOPX16_PCS_ERROR_STATUS, DataLossErr)},
@@ -926,7 +973,7 @@ static void pcs_clear_status(struct amdgpu_device *adev, uint32_t pcs_status_reg
 	WREG32_PCIE(pcs_status_reg, 0);
 }
 
-static void amdgpu_xgmi_reset_ras_error_count(struct amdgpu_device *adev)
+static void amdgpu_xgmi_legacy_reset_ras_error_count(struct amdgpu_device *adev)
 {
 	uint32_t i;
 
@@ -952,6 +999,49 @@ static void amdgpu_xgmi_reset_ras_error_count(struct amdgpu_device *adev)
 	default:
 		break;
 	}
+
+	switch (amdgpu_ip_version(adev, XGMI_HWIP, 0)) {
+	case IP_VERSION(6, 4, 0):
+		for (i = 0; i < ARRAY_SIZE(xgmi3x16_pcs_err_status_reg_v6_4); i++)
+			pcs_clear_status(adev,
+					xgmi3x16_pcs_err_status_reg_v6_4[i]);
+		break;
+	default:
+		break;
+	}
+}
+
+static void __xgmi_v6_4_0_reset_error_count(struct amdgpu_device *adev, int xgmi_inst, u64 mca_base)
+{
+	WREG64_MCA(xgmi_inst, mca_base, MCA_REG_IDX_STATUS, 0ULL);
+}
+
+static void xgmi_v6_4_0_reset_error_count(struct amdgpu_device *adev, int xgmi_inst)
+{
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(xgmi_v6_4_0_mca_base_array); i++)
+		__xgmi_v6_4_0_reset_error_count(adev, xgmi_inst, xgmi_v6_4_0_mca_base_array[i]);
+}
+
+static void xgmi_v6_4_0_reset_ras_error_count(struct amdgpu_device *adev)
+{
+	int i;
+
+	for_each_inst(i, adev->aid_mask)
+		xgmi_v6_4_0_reset_error_count(adev, i);
+}
+
+static void amdgpu_xgmi_reset_ras_error_count(struct amdgpu_device *adev)
+{
+	switch (amdgpu_ip_version(adev, XGMI_HWIP, 0)) {
+	case IP_VERSION(6, 4, 0):
+		xgmi_v6_4_0_reset_ras_error_count(adev);
+		break;
+	default:
+		amdgpu_xgmi_legacy_reset_ras_error_count(adev);
+		break;
+	}
 }
 
 static int amdgpu_xgmi_query_pcs_error_status(struct amdgpu_device *adev,
@@ -969,7 +1059,9 @@ static int amdgpu_xgmi_query_pcs_error_status(struct amdgpu_device *adev,
 
 	if (is_xgmi_pcs) {
 		if (amdgpu_ip_version(adev, XGMI_HWIP, 0) ==
-		    IP_VERSION(6, 1, 0)) {
+		    IP_VERSION(6, 1, 0) ||
+		    amdgpu_ip_version(adev, XGMI_HWIP, 0) ==
+		    IP_VERSION(6, 4, 0)) {
 			pcs_ras_fields = &xgmi3x16_pcs_ras_fields[0];
 			field_array_size = ARRAY_SIZE(xgmi3x16_pcs_ras_fields);
 		} else {
@@ -1003,11 +1095,11 @@ static int amdgpu_xgmi_query_pcs_error_status(struct amdgpu_device *adev,
 	return 0;
 }
 
-static void amdgpu_xgmi_query_ras_error_count(struct amdgpu_device *adev,
-					     void *ras_error_status)
+static void amdgpu_xgmi_legacy_query_ras_error_count(struct amdgpu_device *adev,
+						     void *ras_error_status)
 {
 	struct ras_err_data *err_data = (struct ras_err_data *)ras_error_status;
-	int i;
+	int i, supported = 1;
 	uint32_t data, mask_data = 0;
 	uint32_t ue_cnt = 0, ce_cnt = 0;
 
@@ -1071,7 +1163,25 @@ static void amdgpu_xgmi_query_ras_error_count(struct amdgpu_device *adev,
 		}
 		break;
 	default:
-		dev_warn(adev->dev, "XGMI RAS error query not supported");
+		supported = 0;
+		break;
+	}
+
+	switch (amdgpu_ip_version(adev, XGMI_HWIP, 0)) {
+	case IP_VERSION(6, 4, 0):
+		/* check xgmi3x16 pcs error */
+		for (i = 0; i < ARRAY_SIZE(xgmi3x16_pcs_err_status_reg_v6_4); i++) {
+			data = RREG32_PCIE(xgmi3x16_pcs_err_status_reg_v6_4[i]);
+			mask_data =
+				RREG32_PCIE(xgmi3x16_pcs_err_noncorrectable_mask_reg_v6_4[i]);
+			if (data)
+				amdgpu_xgmi_query_pcs_error_status(adev, data,
+						mask_data, &ue_cnt, &ce_cnt, true, true);
+		}
+		break;
+	default:
+		if (!supported)
+			dev_warn(adev->dev, "XGMI RAS error query not supported");
 		break;
 	}
 
@@ -1081,32 +1191,116 @@ static void amdgpu_xgmi_query_ras_error_count(struct amdgpu_device *adev,
 	err_data->ce_count += ce_cnt;
 }
 
+static enum amdgpu_mca_error_type xgmi_v6_4_0_pcs_mca_get_error_type(struct amdgpu_device *adev, u64 status)
+{
+	const char *error_str;
+	int ext_error_code;
+
+	ext_error_code = MCA_REG__STATUS__ERRORCODEEXT(status);
+
+	error_str = ext_error_code < ARRAY_SIZE(xgmi_v6_4_0_ras_error_code_ext) ?
+		xgmi_v6_4_0_ras_error_code_ext[ext_error_code] : NULL;
+	if (error_str)
+		dev_info(adev->dev, "%s detected\n", error_str);
+
+	switch (ext_error_code) {
+	case 0:
+		return AMDGPU_MCA_ERROR_TYPE_UE;
+	case 6:
+		return AMDGPU_MCA_ERROR_TYPE_CE;
+	default:
+		return -EINVAL;
+	}
+
+	return -EINVAL;
+}
+
+static void __xgmi_v6_4_0_query_error_count(struct amdgpu_device *adev, struct amdgpu_smuio_mcm_config_info *mcm_info,
+					    u64 mca_base, struct ras_err_data *err_data)
+{
+	int xgmi_inst = mcm_info->die_id;
+	u64 status = 0;
+
+	status = RREG64_MCA(xgmi_inst, mca_base, MCA_REG_IDX_STATUS);
+	if (!MCA_REG__STATUS__VAL(status))
+		return;
+
+	switch (xgmi_v6_4_0_pcs_mca_get_error_type(adev, status)) {
+	case AMDGPU_MCA_ERROR_TYPE_UE:
+		amdgpu_ras_error_statistic_ue_count(err_data, mcm_info, 1ULL);
+		break;
+	case AMDGPU_MCA_ERROR_TYPE_CE:
+		amdgpu_ras_error_statistic_ce_count(err_data, mcm_info, 1ULL);
+		break;
+	default:
+		break;
+	}
+
+	WREG64_MCA(xgmi_inst, mca_base, MCA_REG_IDX_STATUS, 0ULL);
+}
+
+static void xgmi_v6_4_0_query_error_count(struct amdgpu_device *adev, int xgmi_inst, struct ras_err_data *err_data)
+{
+	struct amdgpu_smuio_mcm_config_info mcm_info = {
+		.socket_id = adev->smuio.funcs->get_socket_id(adev),
+		.die_id = xgmi_inst,
+	};
+	int i;
+
+	for (i = 0; i < ARRAY_SIZE(xgmi_v6_4_0_mca_base_array); i++)
+		__xgmi_v6_4_0_query_error_count(adev, &mcm_info, xgmi_v6_4_0_mca_base_array[i], err_data);
+}
+
+static void xgmi_v6_4_0_query_ras_error_count(struct amdgpu_device *adev, void *ras_error_status)
+{
+	struct ras_err_data *err_data = (struct ras_err_data *)ras_error_status;
+	int i;
+
+	for_each_inst(i, adev->aid_mask)
+		xgmi_v6_4_0_query_error_count(adev, i, err_data);
+}
+
+static void amdgpu_xgmi_query_ras_error_count(struct amdgpu_device *adev,
+					      void *ras_error_status)
+{
+	switch (amdgpu_ip_version(adev, XGMI_HWIP, 0)) {
+	case IP_VERSION(6, 4, 0):
+		xgmi_v6_4_0_query_ras_error_count(adev, ras_error_status);
+		break;
+	default:
+		amdgpu_xgmi_legacy_query_ras_error_count(adev, ras_error_status);
+		break;
+	}
+}
+
 /* Trigger XGMI/WAFL error */
 static int amdgpu_ras_error_inject_xgmi(struct amdgpu_device *adev,
 			void *inject_if, uint32_t instance_mask)
 {
-	int ret = 0;
+	int ret1, ret2;
 	struct ta_ras_trigger_error_input *block_info =
 				(struct ta_ras_trigger_error_input *)inject_if;
 
 	if (amdgpu_dpm_set_df_cstate(adev, DF_CSTATE_DISALLOW))
 		dev_warn(adev->dev, "Failed to disallow df cstate");
 
-	if (amdgpu_dpm_set_xgmi_plpd_mode(adev, XGMI_PLPD_DISALLOW))
+	ret1 = amdgpu_dpm_set_xgmi_plpd_mode(adev, XGMI_PLPD_DISALLOW);
+	if (ret1 && ret1 != -EOPNOTSUPP)
 		dev_warn(adev->dev, "Failed to disallow XGMI power down");
 
-	ret = psp_ras_trigger_error(&adev->psp, block_info, instance_mask);
+	ret2 = psp_ras_trigger_error(&adev->psp, block_info, instance_mask);
 
 	if (amdgpu_ras_intr_triggered())
-		return ret;
+		return ret2;
 
-	if (amdgpu_dpm_set_xgmi_plpd_mode(adev, XGMI_PLPD_DEFAULT))
+	ret1 = amdgpu_dpm_set_xgmi_plpd_mode(adev, XGMI_PLPD_DEFAULT);
+	if (ret1 && ret1 != -EOPNOTSUPP)
 		dev_warn(adev->dev, "Failed to allow XGMI power down");
 
 	if (amdgpu_dpm_set_df_cstate(adev, DF_CSTATE_ALLOW))
 		dev_warn(adev->dev, "Failed to allow df cstate");
 
-	return ret;
+	return ret2;
 }
 
 struct amdgpu_ras_block_hw_ops  xgmi_ras_hw_ops = {
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c
index 41bbabd9ad4d..40d06d32bb74 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_4_3.c
@@ -1102,6 +1102,7 @@ static void gfx_v9_4_3_init_rlcg_reg_access_ctrl(struct amdgpu_device *adev)
 		reg_access_ctrl->grbm_idx = SOC15_REG_OFFSET(GC, GET_INST(GC, xcc_id), regGRBM_GFX_INDEX);
 		reg_access_ctrl->spare_int = SOC15_REG_OFFSET(GC, GET_INST(GC, xcc_id), regRLC_SPARE_INT);
 	}
+	adev->gfx.rlc.rlcg_reg_access_supported = true;
 }
 
 static int gfx_v9_4_3_rlc_init(struct amdgpu_device *adev)
@@ -2738,16 +2739,16 @@ static void gfx_v9_4_3_xcc_set_compute_eop_interrupt_state(
 
 	switch (state) {
 	case AMDGPU_IRQ_STATE_DISABLE:
-		mec_int_cntl = RREG32(mec_int_cntl_reg);
+		mec_int_cntl = RREG32_XCC(mec_int_cntl_reg, xcc_id);
 		mec_int_cntl = REG_SET_FIELD(mec_int_cntl, CP_ME1_PIPE0_INT_CNTL,
 					     TIME_STAMP_INT_ENABLE, 0);
-		WREG32(mec_int_cntl_reg, mec_int_cntl);
+		WREG32_XCC(mec_int_cntl_reg, mec_int_cntl, xcc_id);
 		break;
 	case AMDGPU_IRQ_STATE_ENABLE:
-		mec_int_cntl = RREG32(mec_int_cntl_reg);
+		mec_int_cntl = RREG32_XCC(mec_int_cntl_reg, xcc_id);
 		mec_int_cntl = REG_SET_FIELD(mec_int_cntl, CP_ME1_PIPE0_INT_CNTL,
 					     TIME_STAMP_INT_ENABLE, 1);
-		WREG32(mec_int_cntl_reg, mec_int_cntl);
+		WREG32_XCC(mec_int_cntl_reg, mec_int_cntl, xcc_id);
 		break;
 	default:
 		break;
@@ -3799,6 +3800,27 @@ static void gfx_v9_4_3_inst_query_ras_err_count(struct amdgpu_device *adev,
 		}
 	}
 
+	/* handle extra register entries of UE */
+	for (; i < ARRAY_SIZE(gfx_v9_4_3_ue_reg_list); i++) {
+		for (j = 0; j < gfx_v9_4_3_ue_reg_list[i].se_num; j++) {
+			for (k = 0; k < gfx_v9_4_3_ue_reg_list[i].reg_entry.reg_inst; k++) {
+				/* no need to select if instance number is 1 */
+				if (gfx_v9_4_3_ue_reg_list[i].se_num > 1 ||
+					gfx_v9_4_3_ue_reg_list[i].reg_entry.reg_inst > 1)
+					gfx_v9_4_3_xcc_select_se_sh(adev, j, 0, k, xcc_id);
+
+				amdgpu_ras_inst_query_ras_error_count(adev,
+					&(gfx_v9_4_3_ue_reg_list[i].reg_entry),
+					1,
+					gfx_v9_4_3_ras_mem_list_array[gfx_v9_4_3_ue_reg_list[i].mem_id_type].mem_id_ent,
+					gfx_v9_4_3_ras_mem_list_array[gfx_v9_4_3_ue_reg_list[i].mem_id_type].size,
+					GET_INST(GC, xcc_id),
+					AMDGPU_RAS_ERROR__MULTI_UNCORRECTABLE,
+					&ue_count);
+			}
+		}
+	}
+
 	gfx_v9_4_3_xcc_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff,
 			xcc_id);
 	mutex_unlock(&adev->grbm_idx_mutex);
@@ -3838,6 +3860,23 @@ static void gfx_v9_4_3_inst_reset_ras_err_count(struct amdgpu_device *adev,
 		}
 	}
 
+	/* handle extra register entries of UE */
+	for (; i < ARRAY_SIZE(gfx_v9_4_3_ue_reg_list); i++) {
+		for (j = 0; j < gfx_v9_4_3_ue_reg_list[i].se_num; j++) {
+			for (k = 0; k < gfx_v9_4_3_ue_reg_list[i].reg_entry.reg_inst; k++) {
+				/* no need to select if instance number is 1 */
+				if (gfx_v9_4_3_ue_reg_list[i].se_num > 1 ||
+					gfx_v9_4_3_ue_reg_list[i].reg_entry.reg_inst > 1)
+					gfx_v9_4_3_xcc_select_se_sh(adev, j, 0, k, xcc_id);
+
+				amdgpu_ras_inst_reset_ras_error_count(adev,
+					&(gfx_v9_4_3_ue_reg_list[i].reg_entry),
+					1,
+					GET_INST(GC, xcc_id));
+			}
+		}
+	}
+
 	gfx_v9_4_3_xcc_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff,
 			xcc_id);
 	mutex_unlock(&adev->grbm_idx_mutex);
@@ -4300,7 +4339,7 @@ const struct amdgpu_ip_block_version gfx_v9_4_3_ip_block = {
 	.type = AMD_IP_BLOCK_TYPE_GFX,
 	.major = 9,
 	.minor = 4,
-	.rev = 0,
+	.rev = 3,
 	.funcs = &gfx_v9_4_3_ip_funcs,
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
index d8a4fddab9c1..0ec7b061d7c2 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c
@@ -268,7 +268,7 @@ static void gmc_v10_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
 	if (adev->gfx.kiq[0].ring.sched.ready && !adev->enable_mes &&
 	    (amdgpu_sriov_runtime(adev) || !amdgpu_sriov_vf(adev))) {
 		amdgpu_virt_kiq_reg_write_reg_wait(adev, req, ack, inv_req,
-				1 << vmid);
+				1 << vmid, GET_INST(GC, 0));
 		return;
 	}
 
@@ -672,6 +672,7 @@ static void gmc_v10_0_vram_gtt_location(struct amdgpu_device *adev,
 	/* add the xgmi offset of the physical node */
 	base += adev->gmc.xgmi.physical_node_id * adev->gmc.xgmi.node_segment_size;
 
+	amdgpu_gmc_set_agp_default(adev, mc);
 	amdgpu_gmc_vram_location(adev, &adev->gmc, base);
 	amdgpu_gmc_gart_location(adev, mc, AMDGPU_GART_PLACEMENT_BEST_FIT);
 	if (!amdgpu_sriov_vf(adev))
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
index 4713a62ad586..6dce9b29f675 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
@@ -229,7 +229,7 @@ static void gmc_v11_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
 	if ((adev->gfx.kiq[0].ring.sched.ready || adev->mes.ring.sched.ready) &&
 	    (amdgpu_sriov_runtime(adev) || !amdgpu_sriov_vf(adev))) {
 		amdgpu_virt_kiq_reg_write_reg_wait(adev, req, ack, inv_req,
-				1 << vmid);
+				1 << vmid, GET_INST(GC, 0));
 		return;
 	}
 
@@ -637,6 +637,7 @@ static void gmc_v11_0_vram_gtt_location(struct amdgpu_device *adev,
 
 	base = adev->mmhub.funcs->get_fb_location(adev);
 
+	amdgpu_gmc_set_agp_default(adev, mc);
 	amdgpu_gmc_vram_location(adev, &adev->gmc, base);
 	amdgpu_gmc_gart_location(adev, mc, AMDGPU_GART_PLACEMENT_HIGH);
 	if (!amdgpu_sriov_vf(adev) ||
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
index 7f66954fd302..42e103d7077d 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
@@ -211,6 +211,7 @@ static void gmc_v6_0_vram_gtt_location(struct amdgpu_device *adev,
 
 	base <<= 24;
 
+	amdgpu_gmc_set_agp_default(adev, mc);
 	amdgpu_gmc_vram_location(adev, mc, base);
 	amdgpu_gmc_gart_location(adev, mc, AMDGPU_GART_PLACEMENT_BEST_FIT);
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
index 61ca1a82b651..efc16e580f1e 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
@@ -239,6 +239,7 @@ static void gmc_v7_0_vram_gtt_location(struct amdgpu_device *adev,
 
 	base <<= 24;
 
+	amdgpu_gmc_set_agp_default(adev, mc);
 	amdgpu_gmc_vram_location(adev, mc, base);
 	amdgpu_gmc_gart_location(adev, mc, AMDGPU_GART_PLACEMENT_BEST_FIT);
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
index fa59749c2aef..ff4ae73d27ec 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
@@ -413,6 +413,7 @@ static void gmc_v8_0_vram_gtt_location(struct amdgpu_device *adev,
 		base = RREG32(mmMC_VM_FB_LOCATION) & 0xFFFF;
 	base <<= 24;
 
+	amdgpu_gmc_set_agp_default(adev, mc);
 	amdgpu_gmc_vram_location(adev, mc, base);
 	amdgpu_gmc_gart_location(adev, mc, AMDGPU_GART_PLACEMENT_BEST_FIT);
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
index b66c5f7e1c56..bde25eb4ed8e 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@@ -817,7 +817,7 @@ static void gmc_v9_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
 					uint32_t vmhub, uint32_t flush_type)
 {
 	bool use_semaphore = gmc_v9_0_use_invalidate_semaphore(adev, vmhub);
-	u32 j, inv_req, tmp, sem, req, ack;
+	u32 j, inv_req, tmp, sem, req, ack, inst;
 	const unsigned int eng = 17;
 	struct amdgpu_vmhub *hub;
 
@@ -832,13 +832,17 @@ static void gmc_v9_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
 	/* This is necessary for a HW workaround under SRIOV as well
 	 * as GFXOFF under bare metal
 	 */
-	if (adev->gfx.kiq[0].ring.sched.ready &&
+	if (vmhub >= AMDGPU_MMHUB0(0))
+		inst = GET_INST(GC, 0);
+	else
+		inst = vmhub;
+	if (adev->gfx.kiq[inst].ring.sched.ready &&
 	    (amdgpu_sriov_runtime(adev) || !amdgpu_sriov_vf(adev))) {
 		uint32_t req = hub->vm_inv_eng0_req + hub->eng_distance * eng;
 		uint32_t ack = hub->vm_inv_eng0_ack + hub->eng_distance * eng;
 
 		amdgpu_virt_kiq_reg_write_reg_wait(adev, req, ack, inv_req,
-						   1 << vmid);
+						   1 << vmid, inst);
 		return;
 	}
 
@@ -856,9 +860,9 @@ static void gmc_v9_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
 		for (j = 0; j < adev->usec_timeout; j++) {
 			/* a read return value of 1 means semaphore acquire */
 			if (vmhub >= AMDGPU_MMHUB0(0))
-				tmp = RREG32_SOC15_IP_NO_KIQ(MMHUB, sem);
+				tmp = RREG32_SOC15_IP_NO_KIQ(MMHUB, sem, inst);
 			else
-				tmp = RREG32_SOC15_IP_NO_KIQ(GC, sem);
+				tmp = RREG32_SOC15_IP_NO_KIQ(GC, sem, inst);
 			if (tmp & 0x1)
 				break;
 			udelay(1);
@@ -869,9 +873,9 @@ static void gmc_v9_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
 	}
 
 	if (vmhub >= AMDGPU_MMHUB0(0))
-		WREG32_SOC15_IP_NO_KIQ(MMHUB, req, inv_req);
+		WREG32_SOC15_IP_NO_KIQ(MMHUB, req, inv_req, inst);
 	else
-		WREG32_SOC15_IP_NO_KIQ(GC, req, inv_req);
+		WREG32_SOC15_IP_NO_KIQ(GC, req, inv_req, inst);
 
 	/*
 	 * Issue a dummy read to wait for the ACK register to
@@ -884,9 +888,9 @@ static void gmc_v9_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
 
 	for (j = 0; j < adev->usec_timeout; j++) {
 		if (vmhub >= AMDGPU_MMHUB0(0))
-			tmp = RREG32_SOC15_IP_NO_KIQ(MMHUB, ack);
+			tmp = RREG32_SOC15_IP_NO_KIQ(MMHUB, ack, inst);
 		else
-			tmp = RREG32_SOC15_IP_NO_KIQ(GC, ack);
+			tmp = RREG32_SOC15_IP_NO_KIQ(GC, ack, inst);
 		if (tmp & (1 << vmid))
 			break;
 		udelay(1);
@@ -899,9 +903,9 @@ static void gmc_v9_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
 		 * write with 0 means semaphore release
 		 */
 		if (vmhub >= AMDGPU_MMHUB0(0))
-			WREG32_SOC15_IP_NO_KIQ(MMHUB, sem, 0);
+			WREG32_SOC15_IP_NO_KIQ(MMHUB, sem, 0, inst);
 		else
-			WREG32_SOC15_IP_NO_KIQ(GC, sem, 0);
+			WREG32_SOC15_IP_NO_KIQ(GC, sem, 0, inst);
 	}
 
 	spin_unlock(&adev->gmc.invalidate_lock);
@@ -1176,7 +1180,10 @@ static void gmc_v9_0_get_coherence_flags(struct amdgpu_device *adev,
 		if (uncached) {
 			mtype = MTYPE_UC;
 		} else if (ext_coherent) {
-			mtype = is_local ? MTYPE_CC : MTYPE_UC;
+			if (adev->rev_id)
+				mtype = is_local ? MTYPE_CC : MTYPE_UC;
+			else
+				mtype = MTYPE_UC;
 		} else if (adev->flags & AMD_IS_APU) {
 			mtype = is_local ? mtype_local : MTYPE_NC;
 		} else {
@@ -1297,7 +1304,7 @@ static void gmc_v9_0_override_vm_pte_flags(struct amdgpu_device *adev,
 
 			*flags = (*flags & ~AMDGPU_PTE_MTYPE_VG10_MASK) |
 				 AMDGPU_PTE_MTYPE_VG10(mtype_local);
-		} else {
+		} else if (adev->rev_id) {
 			/* MTYPE_UC case */
 			*flags = (*flags & ~AMDGPU_PTE_MTYPE_VG10_MASK) |
 				 AMDGPU_PTE_MTYPE_VG10(MTYPE_CC);
@@ -1614,6 +1621,8 @@ static void gmc_v9_0_vram_gtt_location(struct amdgpu_device *adev,
 {
 	u64 base = adev->mmhub.funcs->get_fb_location(adev);
 
+	amdgpu_gmc_set_agp_default(adev, mc);
+
 	/* add the xgmi offset of the physical node */
 	base += adev->gmc.xgmi.physical_node_id * adev->gmc.xgmi.node_segment_size;
 	if (adev->gmc.xgmi.connected_to_cpu) {
diff --git a/drivers/gpu/drm/amd/amdgpu/hdp_v4_0.c b/drivers/gpu/drm/amd/amdgpu/hdp_v4_0.c
index 3f3a6445c006..49e934975719 100644
--- a/drivers/gpu/drm/amd/amdgpu/hdp_v4_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/hdp_v4_0.c
@@ -145,6 +145,10 @@ static void hdp_v4_0_init_registers(struct amdgpu_device *adev)
 		break;
 	}
 
+	/* Do not program registers if VF */
+	if (amdgpu_sriov_vf(adev))
+		return;
+
 	WREG32_FIELD15(HDP, 0, HDP_MISC_CNTL, FLUSH_INVALIDATE_CACHE, 1);
 
 	if (amdgpu_ip_version(adev, HDP_HWIP, 0) == IP_VERSION(4, 4, 0))
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 9a8ec4d7e333..82b6b62c170b 100644
--- a/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.c
+++ b/drivers/gpu/drm/amd/amdgpu/jpeg_v4_0_3.c
@@ -654,9 +654,11 @@ static void jpeg_v4_0_3_dec_ring_set_wptr(struct amdgpu_ring *ring)
  */
 static void jpeg_v4_0_3_dec_ring_insert_start(struct amdgpu_ring *ring)
 {
-	amdgpu_ring_write(ring, PACKETJ(regUVD_JRBC_EXTERNAL_REG_INTERNAL_OFFSET,
-		0, 0, PACKETJ_TYPE0));
-	amdgpu_ring_write(ring, 0x62a04); /* PCTL0_MMHUB_DEEPSLEEP_IB */
+	if (!amdgpu_sriov_vf(ring->adev)) {
+		amdgpu_ring_write(ring, PACKETJ(regUVD_JRBC_EXTERNAL_REG_INTERNAL_OFFSET,
+			0, 0, PACKETJ_TYPE0));
+		amdgpu_ring_write(ring, 0x62a04); /* PCTL0_MMHUB_DEEPSLEEP_IB */
+	}
 
 	amdgpu_ring_write(ring, PACKETJ(JRBC_DEC_EXTERNAL_REG_WRITE_ADDR,
 		0, 0, PACKETJ_TYPE0));
@@ -672,9 +674,11 @@ static void jpeg_v4_0_3_dec_ring_insert_start(struct amdgpu_ring *ring)
  */
 static void jpeg_v4_0_3_dec_ring_insert_end(struct amdgpu_ring *ring)
 {
-	amdgpu_ring_write(ring, PACKETJ(regUVD_JRBC_EXTERNAL_REG_INTERNAL_OFFSET,
-		0, 0, PACKETJ_TYPE0));
-	amdgpu_ring_write(ring, 0x62a04);
+	if (!amdgpu_sriov_vf(ring->adev)) {
+		amdgpu_ring_write(ring, PACKETJ(regUVD_JRBC_EXTERNAL_REG_INTERNAL_OFFSET,
+			0, 0, PACKETJ_TYPE0));
+		amdgpu_ring_write(ring, 0x62a04);
+	}
 
 	amdgpu_ring_write(ring, PACKETJ(JRBC_DEC_EXTERNAL_REG_WRITE_ADDR,
 		0, 0, PACKETJ_TYPE0));
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c
index c46bc6aa4f48..0f24af6f2810 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_4_2.c
@@ -427,6 +427,7 @@ static void sdma_v4_4_2_inst_gfx_stop(struct amdgpu_device *adev,
 				      uint32_t inst_mask)
 {
 	struct amdgpu_ring *sdma[AMDGPU_MAX_SDMA_INSTANCES];
+	u32 doorbell_offset, doorbell;
 	u32 rb_cntl, ib_cntl;
 	int i, unset = 0;
 
@@ -444,6 +445,18 @@ static void sdma_v4_4_2_inst_gfx_stop(struct amdgpu_device *adev,
 		ib_cntl = RREG32_SDMA(i, regSDMA_GFX_IB_CNTL);
 		ib_cntl = REG_SET_FIELD(ib_cntl, SDMA_GFX_IB_CNTL, IB_ENABLE, 0);
 		WREG32_SDMA(i, regSDMA_GFX_IB_CNTL, ib_cntl);
+
+		if (sdma[i]->use_doorbell) {
+			doorbell = RREG32_SDMA(i, regSDMA_GFX_DOORBELL);
+			doorbell_offset = RREG32_SDMA(i, regSDMA_GFX_DOORBELL_OFFSET);
+
+			doorbell = REG_SET_FIELD(doorbell, SDMA_GFX_DOORBELL, ENABLE, 0);
+			doorbell_offset = REG_SET_FIELD(doorbell_offset,
+					SDMA_GFX_DOORBELL_OFFSET,
+					OFFSET, 0);
+			WREG32_SDMA(i, regSDMA_GFX_DOORBELL, doorbell);
+			WREG32_SDMA(i, regSDMA_GFX_DOORBELL_OFFSET, doorbell_offset);
+		}
 	}
 }
 
@@ -631,12 +644,6 @@ static void sdma_v4_4_2_gfx_resume(struct amdgpu_device *adev, unsigned int i)
 	rb_cntl = sdma_v4_4_2_rb_cntl(ring, rb_cntl);
 	WREG32_SDMA(i, regSDMA_GFX_RB_CNTL, rb_cntl);
 
-	/* Initialize the ring buffer's read and write pointers */
-	WREG32_SDMA(i, regSDMA_GFX_RB_RPTR, 0);
-	WREG32_SDMA(i, regSDMA_GFX_RB_RPTR_HI, 0);
-	WREG32_SDMA(i, regSDMA_GFX_RB_WPTR, 0);
-	WREG32_SDMA(i, regSDMA_GFX_RB_WPTR_HI, 0);
-
 	/* set the wb address whether it's enabled or not */
 	WREG32_SDMA(i, regSDMA_GFX_RB_RPTR_ADDR_HI,
 	       upper_32_bits(adev->wb.gpu_addr + wb_offset) & 0xFFFFFFFF);
@@ -654,6 +661,12 @@ static void sdma_v4_4_2_gfx_resume(struct amdgpu_device *adev, unsigned int i)
 	/* before programing wptr to a less value, need set minor_ptr_update first */
 	WREG32_SDMA(i, regSDMA_GFX_MINOR_PTR_UPDATE, 1);
 
+	/* Initialize the ring buffer's read and write pointers */
+	WREG32_SDMA(i, regSDMA_GFX_RB_RPTR, 0);
+	WREG32_SDMA(i, regSDMA_GFX_RB_RPTR_HI, 0);
+	WREG32_SDMA(i, regSDMA_GFX_RB_WPTR, 0);
+	WREG32_SDMA(i, regSDMA_GFX_RB_WPTR_HI, 0);
+
 	doorbell = RREG32_SDMA(i, regSDMA_GFX_DOORBELL);
 	doorbell_offset = RREG32_SDMA(i, regSDMA_GFX_DOORBELL_OFFSET);
 
@@ -2048,7 +2061,7 @@ const struct amdgpu_ip_block_version sdma_v4_4_2_ip_block = {
 	.type = AMD_IP_BLOCK_TYPE_SDMA,
 	.major = 4,
 	.minor = 4,
-	.rev = 0,
+	.rev = 2,
 	.funcs = &sdma_v4_4_2_ip_funcs,
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15_common.h b/drivers/gpu/drm/amd/amdgpu/soc15_common.h
index da683afa0222..242b24f73c17 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15_common.h
+++ b/drivers/gpu/drm/amd/amdgpu/soc15_common.h
@@ -69,7 +69,7 @@
 
 #define RREG32_SOC15_IP(ip, reg) __RREG32_SOC15_RLC__(reg, 0, ip##_HWIP, 0)
 
-#define RREG32_SOC15_IP_NO_KIQ(ip, reg) __RREG32_SOC15_RLC__(reg, AMDGPU_REGS_NO_KIQ, ip##_HWIP, 0)
+#define RREG32_SOC15_IP_NO_KIQ(ip, reg, inst) __RREG32_SOC15_RLC__(reg, AMDGPU_REGS_NO_KIQ, ip##_HWIP, inst)
 
 #define RREG32_SOC15_NO_KIQ(ip, inst, reg) \
 	__RREG32_SOC15_RLC__(adev->reg_offset[ip##_HWIP][inst][reg##_BASE_IDX] + reg, \
@@ -86,8 +86,8 @@
 #define WREG32_SOC15_IP(ip, reg, value) \
 	 __WREG32_SOC15_RLC__(reg, value, 0, ip##_HWIP, 0)
 
-#define WREG32_SOC15_IP_NO_KIQ(ip, reg, value) \
-	 __WREG32_SOC15_RLC__(reg, value, AMDGPU_REGS_NO_KIQ, ip##_HWIP, 0)
+#define WREG32_SOC15_IP_NO_KIQ(ip, reg, value, inst) \
+	 __WREG32_SOC15_RLC__(reg, value, AMDGPU_REGS_NO_KIQ, ip##_HWIP, inst)
 
 #define WREG32_SOC15_NO_KIQ(ip, inst, reg, value) \
 	__WREG32_SOC15_RLC__(adev->reg_offset[ip##_HWIP][inst][reg##_BASE_IDX] + reg, \
@@ -140,7 +140,7 @@
 
 /* for GC only */
 #define RREG32_RLC(reg) \
-	__RREG32_SOC15_RLC__(reg, AMDGPU_REGS_RLC, GC_HWIP)
+	__RREG32_SOC15_RLC__(reg, AMDGPU_REGS_RLC, GC_HWIP, 0)
 
 #define WREG32_RLC_NO_KIQ(reg, value, hwip) \
 	__WREG32_SOC15_RLC__(reg, value, AMDGPU_REGS_NO_KIQ | AMDGPU_REGS_RLC, hwip, 0)
@@ -204,4 +204,10 @@
 			+ adev->asic_funcs->encode_ext_smn_addressing(ext), \
 			value) \
 
+#define RREG64_MCA(ext, mca_base, idx) \
+	RREG64_PCIE_EXT(adev->asic_funcs->encode_ext_smn_addressing(ext) + mca_base + (idx * 8))
+
+#define WREG64_MCA(ext, mca_base, idx, val) \
+	WREG64_PCIE_EXT(adev->asic_funcs->encode_ext_smn_addressing(ext) + mca_base + (idx * 8), val)
+
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/soc21.c b/drivers/gpu/drm/amd/amdgpu/soc21.c
index d5083c549330..48c6efcdeac9 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc21.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc21.c
@@ -381,6 +381,7 @@ soc21_asic_reset_method(struct amdgpu_device *adev)
 		return AMD_RESET_METHOD_MODE1;
 	case IP_VERSION(13, 0, 4):
 	case IP_VERSION(13, 0, 11):
+	case IP_VERSION(14, 0, 0):
 		return AMD_RESET_METHOD_MODE2;
 	default:
 		if (amdgpu_dpm_is_baco_supported(adev))
diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v12_0.c b/drivers/gpu/drm/amd/amdgpu/umc_v12_0.c
index 770b4b4e3138..e9c2ff74f0bc 100644
--- a/drivers/gpu/drm/amd/amdgpu/umc_v12_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/umc_v12_0.c
@@ -88,7 +88,7 @@ static void umc_v12_0_reset_error_count(struct amdgpu_device *adev)
 		umc_v12_0_reset_error_count_per_channel, NULL);
 }
 
-static bool umc_v12_0_is_uncorrectable_error(uint64_t mc_umc_status)
+bool umc_v12_0_is_uncorrectable_error(uint64_t mc_umc_status)
 {
 	return ((REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Val) == 1) &&
 		(REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, PCC) == 1 ||
@@ -96,7 +96,7 @@ static bool umc_v12_0_is_uncorrectable_error(uint64_t mc_umc_status)
 		REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, TCC) == 1));
 }
 
-static bool umc_v12_0_is_correctable_error(uint64_t mc_umc_status)
+bool umc_v12_0_is_correctable_error(uint64_t mc_umc_status)
 {
 	return (REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, Val) == 1 &&
 		(REG_GET_FIELD(mc_umc_status, MCA_UMC_UMC0_MCUMC_STATUST0, CECC) == 1 ||
diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v12_0.h b/drivers/gpu/drm/amd/amdgpu/umc_v12_0.h
index 4885b9fff272..b34b1e358f8b 100644
--- a/drivers/gpu/drm/amd/amdgpu/umc_v12_0.h
+++ b/drivers/gpu/drm/amd/amdgpu/umc_v12_0.h
@@ -117,6 +117,9 @@
 		(pa) |= (UMC_V12_0_CHANNEL_HASH_CH6(channel_idx, pa) << UMC_V12_0_PA_CH6_BIT); \
 	} while (0)
 
+bool umc_v12_0_is_uncorrectable_error(uint64_t mc_umc_status);
+bool umc_v12_0_is_correctable_error(uint64_t mc_umc_status);
+
 extern const uint32_t
 	umc_v12_0_channel_idx_tbl[]
 			[UMC_V12_0_UMC_INSTANCE_NUM]
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v3_1.c b/drivers/gpu/drm/amd/amdgpu/uvd_v3_1.c
index 58a8f78c003c..a6006f231c65 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v3_1.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v3_1.c
@@ -577,8 +577,6 @@ static int uvd_v3_1_sw_init(void *handle)
 	ptr += ucode_len;
 	memcpy(&adev->uvd.keyselect, ptr, 4);
 
-	r = amdgpu_uvd_entity_init(adev);
-
 	return r;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
index d3b1e31f5450..1aa09ad7bbe3 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
@@ -127,8 +127,6 @@ static int uvd_v4_2_sw_init(void *handle)
 	if (r)
 		return r;
 
-	r = amdgpu_uvd_entity_init(adev);
-
 	return r;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
index 5a8116437abf..f8b229b75435 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
@@ -125,8 +125,6 @@ static int uvd_v5_0_sw_init(void *handle)
 	if (r)
 		return r;
 
-	r = amdgpu_uvd_entity_init(adev);
-
 	return r;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
index 74c09230aeb3..a9a6880f44e3 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
@@ -432,8 +432,6 @@ static int uvd_v6_0_sw_init(void *handle)
 		}
 	}
 
-	r = amdgpu_uvd_entity_init(adev);
-
 	return r;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
index 1c42cf10cc29..6068b784dc69 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
@@ -480,10 +480,6 @@ static int uvd_v7_0_sw_init(void *handle)
 	if (r)
 		return r;
 
-	r = amdgpu_uvd_entity_init(adev);
-	if (r)
-		return r;
-
 	r = amdgpu_virt_alloc_mm_table(adev);
 	if (r)
 		return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c
index 67eb01fef789..a08e7abca423 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c
@@ -441,8 +441,6 @@ static int vce_v2_0_sw_init(void *handle)
 			return r;
 	}
 
-	r = amdgpu_vce_entity_init(adev);
-
 	return r;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
index 18f6e62af339..f4760748d349 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
@@ -450,8 +450,6 @@ static int vce_v3_0_sw_init(void *handle)
 			return r;
 	}
 
-	r = amdgpu_vce_entity_init(adev);
-
 	return r;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c
index e0b70cd3b697..06d787385ad4 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c
@@ -486,11 +486,6 @@ static int vce_v4_0_sw_init(void *handle)
 			return r;
 	}
 
-
-	r = amdgpu_vce_entity_init(adev);
-	if (r)
-		return r;
-
 	r = amdgpu_virt_alloc_mm_table(adev);
 	if (r)
 		return r;