Merge tag 'amd-drm-fixes-6.1-2022-10-19' of...

Merge tag 'amd-drm-fixes-6.1-2022-10-19' of https://gitlab.freedesktop.org/agd5f/linux into drm-fixes

amd-drm-fixes-6.1-2022-10-19:

amdgpu:
- Mode2 reset fixes for Sienna Cichlid
- Revert broken fan speed sensor fix
- SMU 13.x fixes
- GC 11.x fixes
- RAS fixes
- SR-IOV fixes
- Fix BO move breakage on SI
- Misc compiler fixes
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20221019191357.6208-1-alexander.deucher@amd.com

drivers/gpu/drm/amd/amdgpu/amdgpu.h

@@ -274,9 +274,6 @@ extern int amdgpu_vcnfw_log;
 #define AMDGPU_RESET_VCE			(1 << 13)
 #define AMDGPU_RESET_VCE1			(1 << 14)
 
-#define AMDGPU_RESET_LEVEL_SOFT_RECOVERY (1 << 0)
-#define AMDGPU_RESET_LEVEL_MODE2 (1 << 1)
-
 /* max cursor sizes (in pixels) */
 #define CIK_CURSOR_WIDTH 128
 #define CIK_CURSOR_HEIGHT 128
@@ -1065,7 +1062,6 @@ struct amdgpu_device {
 
 	struct work_struct		reset_work;
 
-	uint32_t						amdgpu_reset_level_mask;
 	bool                            job_hang;
 };
 

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c

@@ -134,7 +134,6 @@ static void amdgpu_amdkfd_reset_work(struct work_struct *work)
 	reset_context.method = AMD_RESET_METHOD_NONE;
 	reset_context.reset_req_dev = adev;
 	clear_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags);
-	clear_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context.flags);
 
 	amdgpu_device_gpu_recover(adev, NULL, &reset_context);
 }

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v11.c

@@ -111,7 +111,7 @@ static int init_interrupts_v11(struct amdgpu_device *adev, uint32_t pipe_id)
 
 	lock_srbm(adev, mec, pipe, 0, 0);
 
-	WREG32(SOC15_REG_OFFSET(GC, 0, regCPC_INT_CNTL),
+	WREG32_SOC15(GC, 0, regCPC_INT_CNTL,
 		CP_INT_CNTL_RING0__TIME_STAMP_INT_ENABLE_MASK |
 		CP_INT_CNTL_RING0__OPCODE_ERROR_INT_ENABLE_MASK);
 

drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c

@@ -1954,8 +1954,6 @@ int amdgpu_debugfs_init(struct amdgpu_device *adev)
 		return PTR_ERR(ent);
 	}
 
-	debugfs_create_u32("amdgpu_reset_level", 0600, root, &adev->amdgpu_reset_level_mask);
-
 	/* Register debugfs entries for amdgpu_ttm */
 	amdgpu_ttm_debugfs_init(adev);
 	amdgpu_debugfs_pm_init(adev);

drivers/gpu/drm/amd/amdgpu/amdgpu_device.c

@@ -2928,6 +2928,14 @@ static int amdgpu_device_ip_suspend_phase1(struct amdgpu_device *adev)
 	amdgpu_device_set_pg_state(adev, AMD_PG_STATE_UNGATE);
 	amdgpu_device_set_cg_state(adev, AMD_CG_STATE_UNGATE);
 
+	/*
+	 * Per PMFW team's suggestion, driver needs to handle gfxoff
+	 * and df cstate features disablement for gpu reset(e.g. Mode1Reset)
+	 * scenario. Add the missing df cstate disablement here.
+	 */
+	if (amdgpu_dpm_set_df_cstate(adev, DF_CSTATE_DISALLOW))
+		dev_warn(adev->dev, "Failed to disallow df cstate");
+
 	for (i = adev->num_ip_blocks - 1; i >= 0; i--) {
 		if (!adev->ip_blocks[i].status.valid)
 			continue;
@@ -5210,7 +5218,6 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
 
 	reset_context->job = job;
 	reset_context->hive = hive;
-
 	/*
 	 * Build list of devices to reset.
 	 * In case we are in XGMI hive mode, resort the device list
@@ -5337,11 +5344,8 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
 			amdgpu_ras_resume(adev);
 	} else {
 		r = amdgpu_do_asic_reset(device_list_handle, reset_context);
-		if (r && r == -EAGAIN) {
-			set_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context->flags);
-			adev->asic_reset_res = 0;
+		if (r && r == -EAGAIN)
 			goto retry;
-		}
 
 		if (!r && gpu_reset_for_dev_remove)
 			goto recover_end;
@@ -5777,7 +5781,6 @@ pci_ers_result_t amdgpu_pci_slot_reset(struct pci_dev *pdev)
 	reset_context.reset_req_dev = adev;
 	set_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags);
 	set_bit(AMDGPU_SKIP_HW_RESET, &reset_context.flags);
-	set_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context.flags);
 
 	adev->no_hw_access = true;
 	r = amdgpu_device_pre_asic_reset(adev, &reset_context);

drivers/gpu/drm/amd/amdgpu/amdgpu_job.c

@@ -72,7 +72,6 @@ static enum drm_gpu_sched_stat amdgpu_job_timedout(struct drm_sched_job *s_job)
 		reset_context.method = AMD_RESET_METHOD_NONE;
 		reset_context.reset_req_dev = adev;
 		clear_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags);
-		clear_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context.flags);
 
 		r = amdgpu_device_gpu_recover(ring->adev, job, &reset_context);
 		if (r)

drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c

@@ -1950,7 +1950,6 @@ static void amdgpu_ras_do_recovery(struct work_struct *work)
 		reset_context.method = AMD_RESET_METHOD_NONE;
 		reset_context.reset_req_dev = adev;
 		clear_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags);
-		clear_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context.flags);
 
 		amdgpu_device_gpu_recover(ras->adev, NULL, &reset_context);
 	}
@@ -2268,6 +2267,25 @@ static int amdgpu_ras_recovery_fini(struct amdgpu_device *adev)
 
 static bool amdgpu_ras_asic_supported(struct amdgpu_device *adev)
 {
+	if (amdgpu_sriov_vf(adev)) {
+		switch (adev->ip_versions[MP0_HWIP][0]) {
+		case IP_VERSION(13, 0, 2):
+			return true;
+		default:
+			return false;
+		}
+	}
+
+	if (adev->asic_type == CHIP_IP_DISCOVERY) {
+		switch (adev->ip_versions[MP0_HWIP][0]) {
+		case IP_VERSION(13, 0, 0):
+		case IP_VERSION(13, 0, 10):
+			return true;
+		default:
+			return false;
+		}
+	}
+
 	return adev->asic_type == CHIP_VEGA10 ||
 		adev->asic_type == CHIP_VEGA20 ||
 		adev->asic_type == CHIP_ARCTURUS ||
@@ -2311,11 +2329,6 @@ static void amdgpu_ras_check_supported(struct amdgpu_device *adev)
 	    !amdgpu_ras_asic_supported(adev))
 		return;
 
-	/* If driver run on sriov guest side, only enable ras for aldebaran */
-	if (amdgpu_sriov_vf(adev) &&
-		adev->ip_versions[MP1_HWIP][0] != IP_VERSION(13, 0, 2))
-		return;
-
 	if (!adev->gmc.xgmi.connected_to_cpu) {
 		if (amdgpu_atomfirmware_mem_ecc_supported(adev)) {
 			dev_info(adev->dev, "MEM ECC is active.\n");

drivers/gpu/drm/amd/amdgpu/amdgpu_reset.c

@@ -37,8 +37,6 @@ int amdgpu_reset_init(struct amdgpu_device *adev)
 {
 	int ret = 0;
 
-	adev->amdgpu_reset_level_mask = 0x1;
-
 	switch (adev->ip_versions[MP1_HWIP][0]) {
 	case IP_VERSION(13, 0, 2):
 		ret = aldebaran_reset_init(adev);
@@ -76,12 +74,6 @@ int amdgpu_reset_prepare_hwcontext(struct amdgpu_device *adev,
 {
 	struct amdgpu_reset_handler *reset_handler = NULL;
 
-	if (!(adev->amdgpu_reset_level_mask & AMDGPU_RESET_LEVEL_MODE2))
-		return -ENOSYS;
-
-	if (test_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context->flags))
-		return -ENOSYS;
-
 	if (adev->reset_cntl && adev->reset_cntl->get_reset_handler)
 		reset_handler = adev->reset_cntl->get_reset_handler(
 			adev->reset_cntl, reset_context);
@@ -98,12 +90,6 @@ int amdgpu_reset_perform_reset(struct amdgpu_device *adev,
 	int ret;
 	struct amdgpu_reset_handler *reset_handler = NULL;
 
-	if (!(adev->amdgpu_reset_level_mask & AMDGPU_RESET_LEVEL_MODE2))
-		return -ENOSYS;
-
-	if (test_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context->flags))
-		return -ENOSYS;
-
 	if (adev->reset_cntl)
 		reset_handler = adev->reset_cntl->get_reset_handler(
 			adev->reset_cntl, reset_context);

drivers/gpu/drm/amd/amdgpu/amdgpu_reset.h

@@ -30,8 +30,7 @@ enum AMDGPU_RESET_FLAGS {
 
 	AMDGPU_NEED_FULL_RESET = 0,
 	AMDGPU_SKIP_HW_RESET = 1,
-	AMDGPU_SKIP_MODE2_RESET = 2,
-	AMDGPU_RESET_FOR_DEVICE_REMOVE = 3,
+	AMDGPU_RESET_FOR_DEVICE_REMOVE = 2,
 };
 
 struct amdgpu_reset_context {

drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c

@@ -405,9 +405,6 @@ bool amdgpu_ring_soft_recovery(struct amdgpu_ring *ring, unsigned int vmid,
 {
 	ktime_t deadline = ktime_add_us(ktime_get(), 10000);
 
-	if (!(ring->adev->amdgpu_reset_level_mask & AMDGPU_RESET_LEVEL_SOFT_RECOVERY))
-		return false;
-
 	if (amdgpu_sriov_vf(ring->adev) || !ring->funcs->soft_recovery || !fence)
 		return false;
 

drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c

@@ -439,6 +439,9 @@ static bool amdgpu_mem_visible(struct amdgpu_device *adev,
 	while (cursor.remaining) {
 		amdgpu_res_next(&cursor, cursor.size);
 
+		if (!cursor.remaining)
+			break;
+
 		/* ttm_resource_ioremap only supports contiguous memory */
 		if (end != cursor.start)
 			return false;

drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c

@@ -726,6 +726,12 @@ void amdgpu_detect_virtualization(struct amdgpu_device *adev)
 			adev->virt.caps |= AMDGPU_PASSTHROUGH_MODE;
 	}
 
+	if (amdgpu_sriov_vf(adev) && adev->asic_type == CHIP_SIENNA_CICHLID)
+		/* VF MMIO access (except mailbox range) from CPU
+		 * will be blocked during sriov runtime
+		 */
+		adev->virt.caps |= AMDGPU_VF_MMIO_ACCESS_PROTECT;
+
 	/* we have the ability to check now */
 	if (amdgpu_sriov_vf(adev)) {
 		switch (adev->asic_type) {

drivers/gpu/drm/amd/amdgpu/amdgpu_virt.h

@@ -31,6 +31,7 @@
 #define AMDGPU_SRIOV_CAPS_IS_VF        (1 << 2) /* this GPU is a virtual function */
 #define AMDGPU_PASSTHROUGH_MODE        (1 << 3) /* thw whole GPU is pass through for VM */
 #define AMDGPU_SRIOV_CAPS_RUNTIME      (1 << 4) /* is out of full access mode */
+#define AMDGPU_VF_MMIO_ACCESS_PROTECT  (1 << 5) /* MMIO write access is not allowed in sriov runtime */
 
 /* flags for indirect register access path supported by rlcg for sriov */
 #define AMDGPU_RLCG_GC_WRITE_LEGACY    (0x8 << 28)
@@ -297,6 +298,9 @@ struct amdgpu_video_codec_info;
 #define amdgpu_passthrough(adev) \
 ((adev)->virt.caps & AMDGPU_PASSTHROUGH_MODE)
 
+#define amdgpu_sriov_vf_mmio_access_protection(adev) \
+((adev)->virt.caps & AMDGPU_VF_MMIO_ACCESS_PROTECT)
+
 static inline bool is_virtual_machine(void)
 {
 #if defined(CONFIG_X86)

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c

@@ -2338,7 +2338,11 @@ void amdgpu_vm_manager_init(struct amdgpu_device *adev)
 	 */
 #ifdef CONFIG_X86_64
 	if (amdgpu_vm_update_mode == -1) {
-		if (amdgpu_gmc_vram_full_visible(&adev->gmc))
+		/* For asic with VF MMIO access protection
+		 * avoid using CPU for VM table updates
+		 */
+		if (amdgpu_gmc_vram_full_visible(&adev->gmc) &&
+		    !amdgpu_sriov_vf_mmio_access_protection(adev))
 			adev->vm_manager.vm_update_mode =
 				AMDGPU_VM_USE_CPU_FOR_COMPUTE;
 		else

drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c

@@ -1571,7 +1571,7 @@ static void gfx_v11_0_init_compute_vmid(struct amdgpu_device *adev)
 		WREG32_SOC15(GC, 0, regSH_MEM_BASES, sh_mem_bases);
 
 		/* Enable trap for each kfd vmid. */
-		data = RREG32(SOC15_REG_OFFSET(GC, 0, regSPI_GDBG_PER_VMID_CNTL));
+		data = RREG32_SOC15(GC, 0, regSPI_GDBG_PER_VMID_CNTL);
 		data = REG_SET_FIELD(data, SPI_GDBG_PER_VMID_CNTL, TRAP_EN, 1);
 	}
 	soc21_grbm_select(adev, 0, 0, 0, 0);
@@ -5076,6 +5076,7 @@ static int gfx_v11_0_set_clockgating_state(void *handle,
 	case IP_VERSION(11, 0, 0):
 	case IP_VERSION(11, 0, 1):
 	case IP_VERSION(11, 0, 2):
+	case IP_VERSION(11, 0, 3):
 	        gfx_v11_0_update_gfx_clock_gating(adev,
 	                        state ==  AMD_CG_STATE_GATE);
 	        break;

drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c

@@ -186,6 +186,10 @@ static void gmc_v11_0_flush_vm_hub(struct amdgpu_device *adev, uint32_t vmid,
 	/* Use register 17 for GART */
 	const unsigned eng = 17;
 	unsigned int i;
+	unsigned char hub_ip = 0;
+
+	hub_ip = (vmhub == AMDGPU_GFXHUB_0) ?
+		   GC_HWIP : MMHUB_HWIP;
 
 	spin_lock(&adev->gmc.invalidate_lock);
 	/*
@@ -199,8 +203,8 @@ static void gmc_v11_0_flush_vm_hub(struct amdgpu_device *adev, uint32_t vmid,
 	if (use_semaphore) {
 		for (i = 0; i < adev->usec_timeout; i++) {
 			/* a read return value of 1 means semaphore acuqire */
-			tmp = RREG32_NO_KIQ(hub->vm_inv_eng0_sem +
-					    hub->eng_distance * eng);
+			tmp = RREG32_RLC_NO_KIQ(hub->vm_inv_eng0_sem +
+					    hub->eng_distance * eng, hub_ip);
 			if (tmp & 0x1)
 				break;
 			udelay(1);
@@ -210,12 +214,12 @@ static void gmc_v11_0_flush_vm_hub(struct amdgpu_device *adev, uint32_t vmid,
 			DRM_ERROR("Timeout waiting for sem acquire in VM flush!\n");
 	}
 
-	WREG32_NO_KIQ(hub->vm_inv_eng0_req + hub->eng_distance * eng, inv_req);
+	WREG32_RLC_NO_KIQ(hub->vm_inv_eng0_req + hub->eng_distance * eng, inv_req, hub_ip);
 
 	/* Wait for ACK with a delay.*/
 	for (i = 0; i < adev->usec_timeout; i++) {
-		tmp = RREG32_NO_KIQ(hub->vm_inv_eng0_ack +
-				    hub->eng_distance * eng);
+		tmp = RREG32_RLC_NO_KIQ(hub->vm_inv_eng0_ack +
+				    hub->eng_distance * eng, hub_ip);
 		tmp &= 1 << vmid;
 		if (tmp)
 			break;
@@ -229,8 +233,8 @@ static void gmc_v11_0_flush_vm_hub(struct amdgpu_device *adev, uint32_t vmid,
 		 * add semaphore release after invalidation,
 		 * write with 0 means semaphore release
 		 */
-		WREG32_NO_KIQ(hub->vm_inv_eng0_sem +
-			      hub->eng_distance * eng, 0);
+		WREG32_RLC_NO_KIQ(hub->vm_inv_eng0_sem +
+			      hub->eng_distance * eng, 0, hub_ip);
 
 	/* Issue additional private vm invalidation to MMHUB */
 	if ((vmhub != AMDGPU_GFXHUB_0) &&

drivers/gpu/drm/amd/amdgpu/mes_v11_0.c

@@ -1156,6 +1156,42 @@ static int mes_v11_0_sw_fini(void *handle)
 	return 0;
 }
 
+static void mes_v11_0_kiq_dequeue_sched(struct amdgpu_device *adev)
+{
+	uint32_t data;
+	int i;
+
+	mutex_lock(&adev->srbm_mutex);
+	soc21_grbm_select(adev, 3, AMDGPU_MES_SCHED_PIPE, 0, 0);
+
+	/* disable the queue if it's active */
+	if (RREG32_SOC15(GC, 0, regCP_HQD_ACTIVE) & 1) {
+		WREG32_SOC15(GC, 0, regCP_HQD_DEQUEUE_REQUEST, 1);
+		for (i = 0; i < adev->usec_timeout; i++) {
+			if (!(RREG32_SOC15(GC, 0, regCP_HQD_ACTIVE) & 1))
+				break;
+			udelay(1);
+		}
+	}
+	data = RREG32_SOC15(GC, 0, regCP_HQD_PQ_DOORBELL_CONTROL);
+	data = REG_SET_FIELD(data, CP_HQD_PQ_DOORBELL_CONTROL,
+				DOORBELL_EN, 0);
+	data = REG_SET_FIELD(data, CP_HQD_PQ_DOORBELL_CONTROL,
+				DOORBELL_HIT, 1);
+	WREG32_SOC15(GC, 0, regCP_HQD_PQ_DOORBELL_CONTROL, data);
+
+	WREG32_SOC15(GC, 0, regCP_HQD_PQ_DOORBELL_CONTROL, 0);
+
+	WREG32_SOC15(GC, 0, regCP_HQD_PQ_WPTR_LO, 0);
+	WREG32_SOC15(GC, 0, regCP_HQD_PQ_WPTR_HI, 0);
+	WREG32_SOC15(GC, 0, regCP_HQD_PQ_RPTR, 0);
+
+	soc21_grbm_select(adev, 0, 0, 0, 0);
+	mutex_unlock(&adev->srbm_mutex);
+
+	adev->mes.ring.sched.ready = false;
+}
+
 static void mes_v11_0_kiq_setting(struct amdgpu_ring *ring)
 {
 	uint32_t tmp;
@@ -1207,6 +1243,9 @@ static int mes_v11_0_kiq_hw_init(struct amdgpu_device *adev)
 
 static int mes_v11_0_kiq_hw_fini(struct amdgpu_device *adev)
 {
+	if (adev->mes.ring.sched.ready)
+		mes_v11_0_kiq_dequeue_sched(adev);
+
 	mes_v11_0_enable(adev, false);
 	return 0;
 }
@@ -1262,9 +1301,6 @@ static int mes_v11_0_hw_init(void *handle)
 
 static int mes_v11_0_hw_fini(void *handle)
 {
-	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-
-	adev->mes.ring.sched.ready = false;
 	return 0;
 }
 
@@ -1296,7 +1332,8 @@ static int mes_v11_0_late_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!amdgpu_in_reset(adev))
+	if (!amdgpu_in_reset(adev) &&
+	    (adev->ip_versions[GC_HWIP][0] != IP_VERSION(11, 0, 3)))
 		amdgpu_mes_self_test(adev);
 
 	return 0;

drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c

@@ -290,7 +290,6 @@ static void xgpu_ai_mailbox_flr_work(struct work_struct *work)
 		reset_context.method = AMD_RESET_METHOD_NONE;
 		reset_context.reset_req_dev = adev;
 		clear_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags);
-		clear_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context.flags);
 
 		amdgpu_device_gpu_recover(adev, NULL, &reset_context);
 	}

drivers/gpu/drm/amd/amdgpu/mxgpu_nv.c

@@ -317,7 +317,6 @@ static void xgpu_nv_mailbox_flr_work(struct work_struct *work)
 		reset_context.method = AMD_RESET_METHOD_NONE;
 		reset_context.reset_req_dev = adev;
 		clear_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags);
-		clear_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context.flags);
 
 		amdgpu_device_gpu_recover(adev, NULL, &reset_context);
 	}

drivers/gpu/drm/amd/amdgpu/mxgpu_vi.c

@@ -529,7 +529,6 @@ static void xgpu_vi_mailbox_flr_work(struct work_struct *work)
 		reset_context.method = AMD_RESET_METHOD_NONE;
 		reset_context.reset_req_dev = adev;
 		clear_bit(AMDGPU_NEED_FULL_RESET, &reset_context.flags);
-		clear_bit(AMDGPU_SKIP_MODE2_RESET, &reset_context.flags);
 
 		amdgpu_device_gpu_recover(adev, NULL, &reset_context);
 	}

drivers/gpu/drm/amd/amdgpu/sienna_cichlid.c

@@ -31,12 +31,23 @@
 #include "amdgpu_psp.h"
 #include "amdgpu_xgmi.h"
 
+static bool sienna_cichlid_is_mode2_default(struct amdgpu_reset_control *reset_ctl)
+{
+#if 0
+	struct amdgpu_device *adev = (struct amdgpu_device *)reset_ctl->handle;
+
+	if (adev->ip_versions[MP1_HWIP][0] == IP_VERSION(11, 0, 7) &&
+	    adev->pm.fw_version >= 0x3a5500 && !amdgpu_sriov_vf(adev))
+		return true;
+#endif
+	return false;
+}
+
 static struct amdgpu_reset_handler *
 sienna_cichlid_get_reset_handler(struct amdgpu_reset_control *reset_ctl,
 			    struct amdgpu_reset_context *reset_context)
 {
 	struct amdgpu_reset_handler *handler;
-	struct amdgpu_device *adev = (struct amdgpu_device *)reset_ctl->handle;
 
 	if (reset_context->method != AMD_RESET_METHOD_NONE) {
 		list_for_each_entry(handler, &reset_ctl->reset_handlers,
@@ -44,15 +55,13 @@ sienna_cichlid_get_reset_handler(struct amdgpu_reset_control *reset_ctl,
 			if (handler->reset_method == reset_context->method)
 				return handler;
 		}
-	} else {
-		list_for_each_entry(handler, &reset_ctl->reset_handlers,
+	}
+
+	if (sienna_cichlid_is_mode2_default(reset_ctl)) {
+		list_for_each_entry (handler, &reset_ctl->reset_handlers,
 				     handler_list) {
-			if (handler->reset_method == AMD_RESET_METHOD_MODE2 &&
-			    adev->pm.fw_version >= 0x3a5500 &&
-			    !amdgpu_sriov_vf(adev)) {
-				reset_context->method = AMD_RESET_METHOD_MODE2;
+			if (handler->reset_method == AMD_RESET_METHOD_MODE2)
 				return handler;
-			}
 		}
 	}
 

drivers/gpu/drm/amd/amdgpu/soc21.c

@@ -423,6 +423,7 @@ static bool soc21_need_full_reset(struct amdgpu_device *adev)
 	case IP_VERSION(11, 0, 0):
 		return amdgpu_ras_is_supported(adev, AMDGPU_RAS_BLOCK__UMC);
 	case IP_VERSION(11, 0, 2):
+	case IP_VERSION(11, 0, 3):
 		return false;
 	default:
 		return true;
@@ -636,7 +637,11 @@ static int soc21_common_early_init(void *handle)
 		break;
 	case IP_VERSION(11, 0, 3):
 		adev->cg_flags = AMD_CG_SUPPORT_VCN_MGCG |
-			AMD_CG_SUPPORT_JPEG_MGCG;
+			AMD_CG_SUPPORT_JPEG_MGCG |
+			AMD_CG_SUPPORT_GFX_CGCG |
+			AMD_CG_SUPPORT_GFX_CGLS |
+			AMD_CG_SUPPORT_REPEATER_FGCG |
+			AMD_CG_SUPPORT_GFX_MGCG;
 		adev->pg_flags = AMD_PG_SUPPORT_VCN |
 			AMD_PG_SUPPORT_VCN_DPG |
 			AMD_PG_SUPPORT_JPEG;

drivers/gpu/drm/amd/display/dc/dml/Makefile

@@ -77,7 +77,7 @@ CFLAGS_$(AMDDALPATH)/dc/dml/dcn30/dcn30_fpu.o := $(dml_ccflags)
 CFLAGS_$(AMDDALPATH)/dc/dml/dcn32/dcn32_fpu.o := $(dml_ccflags)
 CFLAGS_$(AMDDALPATH)/dc/dml/dcn32/display_mode_vba_32.o := $(dml_ccflags) $(frame_warn_flag)
 CFLAGS_$(AMDDALPATH)/dc/dml/dcn32/display_rq_dlg_calc_32.o := $(dml_ccflags)
-CFLAGS_$(AMDDALPATH)/dc/dml/dcn32/display_mode_vba_util_32.o := $(dml_ccflags)
+CFLAGS_$(AMDDALPATH)/dc/dml/dcn32/display_mode_vba_util_32.o := $(dml_ccflags) $(frame_warn_flag)
 CFLAGS_$(AMDDALPATH)/dc/dml/dcn321/dcn321_fpu.o := $(dml_ccflags)
 CFLAGS_$(AMDDALPATH)/dc/dml/dcn31/dcn31_fpu.o := $(dml_ccflags)
 CFLAGS_$(AMDDALPATH)/dc/dml/dcn301/dcn301_fpu.o := $(dml_ccflags)

drivers/gpu/drm/amd/include/kgd_kfd_interface.h

@@ -262,8 +262,9 @@ struct kfd2kgd_calls {
 				uint32_t queue_id);
 
 	int (*hqd_destroy)(struct amdgpu_device *adev, void *mqd,
-				uint32_t reset_type, unsigned int timeout,
-				uint32_t pipe_id, uint32_t queue_id);
+				enum kfd_preempt_type reset_type,
+				unsigned int timeout, uint32_t pipe_id,
+				uint32_t queue_id);
 
 	bool (*hqd_sdma_is_occupied)(struct amdgpu_device *adev, void *mqd);
 

drivers/gpu/drm/amd/pm/amdgpu_pm.c

@@ -3362,11 +3362,11 @@ int amdgpu_pm_sysfs_init(struct amdgpu_device *adev)
 	if (adev->pm.sysfs_initialized)
 		return 0;
 
+	INIT_LIST_HEAD(&adev->pm.pm_attr_list);
+
 	if (adev->pm.dpm_enabled == 0)
 		return 0;
 
-	INIT_LIST_HEAD(&adev->pm.pm_attr_list);
-
 	adev->pm.int_hwmon_dev = hwmon_device_register_with_groups(adev->dev,
 								   DRIVER_NAME, adev,
 								   hwmon_groups);

drivers/gpu/drm/amd/pm/powerplay/hwmgr/vega10_thermal.c

@@ -67,21 +67,22 @@ int vega10_fan_ctrl_get_fan_speed_info(struct pp_hwmgr *hwmgr,
 int vega10_fan_ctrl_get_fan_speed_pwm(struct pp_hwmgr *hwmgr,
 		uint32_t *speed)
 {
-	struct amdgpu_device *adev = hwmgr->adev;
-	uint32_t duty100, duty;
-	uint64_t tmp64;
+	uint32_t current_rpm;
+	uint32_t percent = 0;
 
-	duty100 = REG_GET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL1),
-				CG_FDO_CTRL1, FMAX_DUTY100);
-	duty = REG_GET_FIELD(RREG32_SOC15(THM, 0, mmCG_THERMAL_STATUS),
-				CG_THERMAL_STATUS, FDO_PWM_DUTY);
+	if (hwmgr->thermal_controller.fanInfo.bNoFan)
+		return 0;
 
-	if (!duty100)
-		return -EINVAL;
+	if (vega10_get_current_rpm(hwmgr, &current_rpm))
+		return -1;
+
+	if (hwmgr->thermal_controller.
+			advanceFanControlParameters.usMaxFanRPM != 0)
+		percent = current_rpm * 255 /
+			hwmgr->thermal_controller.
+			advanceFanControlParameters.usMaxFanRPM;
 
-	tmp64 = (uint64_t)duty * 255;
-	do_div(tmp64, duty100);
-	*speed = MIN((uint32_t)tmp64, 255);
+	*speed = MIN(percent, 255);
 
 	return 0;
 }

drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c

@@ -1314,8 +1314,8 @@ static int smu_smc_hw_setup(struct smu_context *smu)
 
 	ret = smu_enable_thermal_alert(smu);
 	if (ret) {
-		dev_err(adev->dev, "Failed to enable thermal alert!\n");
-		return ret;
+	  dev_err(adev->dev, "Failed to enable thermal alert!\n");
+	  return ret;
 	}
 
 	ret = smu_notify_display_change(smu);

drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu13_driver_if_v13_0_4.h

@@ -27,7 +27,7 @@
 // *** IMPORTANT ***
 // SMU TEAM: Always increment the interface version if
 // any structure is changed in this file
-#define PMFW_DRIVER_IF_VERSION 5
+#define PMFW_DRIVER_IF_VERSION 7
 
 typedef struct {
   int32_t value;
@@ -163,8 +163,8 @@ typedef struct {
   uint16_t DclkFrequency;               //[MHz]
   uint16_t MemclkFrequency;             //[MHz]
   uint16_t spare;                       //[centi]
-  uint16_t UvdActivity;                 //[centi]
   uint16_t GfxActivity;                 //[centi]
+  uint16_t UvdActivity;                 //[centi]
 
   uint16_t Voltage[2];                  //[mV] indices: VDDCR_VDD, VDDCR_SOC
   uint16_t Current[2];                  //[mA] indices: VDDCR_VDD, VDDCR_SOC
@@ -199,6 +199,19 @@ typedef struct {
   uint16_t DeviceState;
   uint16_t CurTemp;                     //[centi-Celsius]
   uint16_t spare2;
+
+  uint16_t AverageGfxclkFrequency;
+  uint16_t AverageFclkFrequency;
+  uint16_t AverageGfxActivity;
+  uint16_t AverageSocclkFrequency;
+  uint16_t AverageVclkFrequency;
+  uint16_t AverageVcnActivity;
+  uint16_t AverageDRAMReads;          //Filtered DF Bandwidth::DRAM Reads
+  uint16_t AverageDRAMWrites;         //Filtered DF Bandwidth::DRAM Writes
+  uint16_t AverageSocketPower;        //Filtered value of CurrentSocketPower
+  uint16_t AverageCorePower;          //Filtered of [sum of CorePower[8]])
+  uint16_t AverageCoreC0Residency[8]; //Filtered of [average C0 residency %  per core]
+  uint32_t MetricsCounter;            //Counts the # of metrics table parameter reads per update to the metrics table, i.e. if the metrics table update happens every 1 second, this value could be up to 1000 if the smu collected metrics data every cycle, or as low as 0 if the smu was asleep the whole time. Reset to 0 after writing.
 } SmuMetrics_t;
 
 typedef struct {

drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h

@@ -28,7 +28,7 @@
 #define SMU13_DRIVER_IF_VERSION_INV 0xFFFFFFFF
 #define SMU13_DRIVER_IF_VERSION_YELLOW_CARP 0x04
 #define SMU13_DRIVER_IF_VERSION_ALDE 0x08
-#define SMU13_DRIVER_IF_VERSION_SMU_V13_0_4 0x05
+#define SMU13_DRIVER_IF_VERSION_SMU_V13_0_4 0x07
 #define SMU13_DRIVER_IF_VERSION_SMU_V13_0_5 0x04
 #define SMU13_DRIVER_IF_VERSION_SMU_V13_0_0 0x30
 #define SMU13_DRIVER_IF_VERSION_SMU_V13_0_7 0x2C

drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c

@@ -2242,9 +2242,17 @@ static void arcturus_get_unique_id(struct smu_context *smu)
 static int arcturus_set_df_cstate(struct smu_context *smu,
 				  enum pp_df_cstate state)
 {
+	struct amdgpu_device *adev = smu->adev;
 	uint32_t smu_version;
 	int ret;
 
+	/*
+	 * Arcturus does not need the cstate disablement
+	 * prerequisite for gpu reset.
+	 */
+	if (amdgpu_in_reset(adev) || adev->in_suspend)
+		return 0;
+
 	ret = smu_cmn_get_smc_version(smu, NULL, &smu_version);
 	if (ret) {
 		dev_err(smu->adev->dev, "Failed to get smu version!\n");

drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c

@@ -1640,6 +1640,15 @@ static bool aldebaran_is_baco_supported(struct smu_context *smu)
 static int aldebaran_set_df_cstate(struct smu_context *smu,
 				   enum pp_df_cstate state)
 {
+	struct amdgpu_device *adev = smu->adev;
+
+	/*
+	 * Aldebaran does not need the cstate disablement
+	 * prerequisite for gpu reset.
+	 */
+	if (amdgpu_in_reset(adev) || adev->in_suspend)
+		return 0;
+
 	return smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_DFCstateControl, state, NULL);
 }
 

drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0.c

@@ -211,7 +211,8 @@ int smu_v13_0_init_pptable_microcode(struct smu_context *smu)
 		return 0;
 
 	if ((adev->ip_versions[MP1_HWIP][0] == IP_VERSION(13, 0, 7)) ||
-	    (adev->ip_versions[MP1_HWIP][0] == IP_VERSION(13, 0, 0)))
+	    (adev->ip_versions[MP1_HWIP][0] == IP_VERSION(13, 0, 0)) ||
+	    (adev->ip_versions[MP1_HWIP][0] == IP_VERSION(13, 0, 10)))
 		return 0;
 
 	/* override pptable_id from driver parameter */
@@ -454,9 +455,6 @@ int smu_v13_0_setup_pptable(struct smu_context *smu)
 		dev_info(adev->dev, "override pptable id %d\n", pptable_id);
 	} else {
 		pptable_id = smu->smu_table.boot_values.pp_table_id;
-
-		if (adev->ip_versions[MP1_HWIP][0] == IP_VERSION(13, 0, 10))
-			pptable_id = 6666;
 	}
 
 	/* force using vbios pptable in sriov mode */

drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c

@@ -119,6 +119,7 @@ static struct cmn2asic_msg_mapping smu_v13_0_0_message_map[SMU_MSG_MAX_COUNT] =
 	MSG_MAP(NotifyPowerSource,		PPSMC_MSG_NotifyPowerSource,           0),
 	MSG_MAP(Mode1Reset,			PPSMC_MSG_Mode1Reset,                  0),
 	MSG_MAP(PrepareMp1ForUnload,		PPSMC_MSG_PrepareMp1ForUnload,         0),
+	MSG_MAP(DFCstateControl,		PPSMC_MSG_SetExternalClientDfCstateAllow, 0),
 };
 
 static struct cmn2asic_mapping smu_v13_0_0_clk_map[SMU_CLK_COUNT] = {
@@ -1753,6 +1754,15 @@ static int smu_v13_0_0_set_mp1_state(struct smu_context *smu,
 	return ret;
 }
 
+static int smu_v13_0_0_set_df_cstate(struct smu_context *smu,
+				     enum pp_df_cstate state)
+{
+	return smu_cmn_send_smc_msg_with_param(smu,
+					       SMU_MSG_DFCstateControl,
+					       state,
+					       NULL);
+}
+
 static const struct pptable_funcs smu_v13_0_0_ppt_funcs = {
 	.get_allowed_feature_mask = smu_v13_0_0_get_allowed_feature_mask,
 	.set_default_dpm_table = smu_v13_0_0_set_default_dpm_table,
@@ -1822,6 +1832,7 @@ static const struct pptable_funcs smu_v13_0_0_ppt_funcs = {
 	.mode1_reset_is_support = smu_v13_0_0_is_mode1_reset_supported,
 	.mode1_reset = smu_v13_0_mode1_reset,
 	.set_mp1_state = smu_v13_0_0_set_mp1_state,
+	.set_df_cstate = smu_v13_0_0_set_df_cstate,
 };
 
 void smu_v13_0_0_set_ppt_funcs(struct smu_context *smu)

drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c

@@ -121,6 +121,7 @@ static struct cmn2asic_msg_mapping smu_v13_0_7_message_map[SMU_MSG_MAX_COUNT] =
 	MSG_MAP(Mode1Reset,             PPSMC_MSG_Mode1Reset,                  0),
 	MSG_MAP(PrepareMp1ForUnload,		PPSMC_MSG_PrepareMp1ForUnload,         0),
 	MSG_MAP(SetMGpuFanBoostLimitRpm,	PPSMC_MSG_SetMGpuFanBoostLimitRpm,     0),
+	MSG_MAP(DFCstateControl,		PPSMC_MSG_SetExternalClientDfCstateAllow, 0),
 };
 
 static struct cmn2asic_mapping smu_v13_0_7_clk_map[SMU_CLK_COUNT] = {
@@ -1587,6 +1588,16 @@ static bool smu_v13_0_7_is_mode1_reset_supported(struct smu_context *smu)
 
 	return true;
 }
+
+static int smu_v13_0_7_set_df_cstate(struct smu_context *smu,
+				     enum pp_df_cstate state)
+{
+	return smu_cmn_send_smc_msg_with_param(smu,
+					       SMU_MSG_DFCstateControl,
+					       state,
+					       NULL);
+}
+
 static const struct pptable_funcs smu_v13_0_7_ppt_funcs = {
 	.get_allowed_feature_mask = smu_v13_0_7_get_allowed_feature_mask,
 	.set_default_dpm_table = smu_v13_0_7_set_default_dpm_table,
@@ -1649,6 +1660,7 @@ static const struct pptable_funcs smu_v13_0_7_ppt_funcs = {
 	.mode1_reset_is_support = smu_v13_0_7_is_mode1_reset_supported,
 	.mode1_reset = smu_v13_0_mode1_reset,
 	.set_mp1_state = smu_v13_0_7_set_mp1_state,
+	.set_df_cstate = smu_v13_0_7_set_df_cstate,
 };
 
 void smu_v13_0_7_set_ppt_funcs(struct smu_context *smu)