drm/amdgpu: retire gfx ras query_utcl2_poison_status

Replace it with related interface in gfxhub functions.

v2: replace node id with xcc id.
    get node id for query_utcl2_poison_status
Signed-off-by: Tao Zhou <tao.zhou1@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

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

@@ -769,10 +769,11 @@ int amdgpu_amdkfd_send_close_event_drain_irq(struct amdgpu_device *adev,
 	return 0;
 }
 
-bool amdgpu_amdkfd_ras_query_utcl2_poison_status(struct amdgpu_device *adev)
+bool amdgpu_amdkfd_ras_query_utcl2_poison_status(struct amdgpu_device *adev,
+			int xcc_id)
 {
-	if (adev->gfx.ras && adev->gfx.ras->query_utcl2_poison_status)
-		return adev->gfx.ras->query_utcl2_poison_status(adev);
+	if (adev->gfxhub.funcs->query_utcl2_poison_status)
+		return adev->gfxhub.funcs->query_utcl2_poison_status(adev, xcc_id);
 	else
 		return false;
 }

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

@@ -341,7 +341,8 @@ bool amdgpu_amdkfd_is_fed(struct amdgpu_device *adev);
 bool amdgpu_amdkfd_bo_mapped_to_dev(struct amdgpu_device *adev, struct kgd_mem *mem);
 void amdgpu_amdkfd_block_mmu_notifications(void *p);
 int amdgpu_amdkfd_criu_resume(void *p);
-bool amdgpu_amdkfd_ras_query_utcl2_poison_status(struct amdgpu_device *adev);
+bool amdgpu_amdkfd_ras_query_utcl2_poison_status(struct amdgpu_device *adev,
+			int xcc_id);
 int amdgpu_amdkfd_reserve_mem_limit(struct amdgpu_device *adev,
 		uint64_t size, u32 alloc_flag, int8_t xcp_id);
 void amdgpu_amdkfd_unreserve_mem_limit(struct amdgpu_device *adev,

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

@@ -259,7 +259,6 @@ struct amdgpu_cu_info {
 struct amdgpu_gfx_ras {
 	struct amdgpu_ras_block_object  ras_block;
 	void (*enable_watchdog_timer)(struct amdgpu_device *adev);
-	bool (*query_utcl2_poison_status)(struct amdgpu_device *adev);
 	int (*rlc_gc_fed_irq)(struct amdgpu_device *adev,
 				struct amdgpu_irq_src *source,
 				struct amdgpu_iv_entry *entry);

drivers/gpu/drm/amd/amdgpu/gfx_v9_4_2.c

@@ -1909,18 +1909,7 @@ static void gfx_v9_4_2_reset_sq_timeout_status(struct amdgpu_device *adev)
 	mutex_unlock(&adev->grbm_idx_mutex);
 }
 
-static bool gfx_v9_4_2_query_uctl2_poison_status(struct amdgpu_device *adev)
-{
-	u32 status = 0;
-	struct amdgpu_vmhub *hub;
-
-	hub = &adev->vmhub[AMDGPU_GFXHUB(0)];
-	status = RREG32(hub->vm_l2_pro_fault_status);
-	/* reset page fault status */
-	WREG32_P(hub->vm_l2_pro_fault_cntl, 1, ~1);
 
-	return REG_GET_FIELD(status, VM_L2_PROTECTION_FAULT_STATUS, FED);
-}
 
 struct amdgpu_ras_block_hw_ops  gfx_v9_4_2_ras_ops = {
 		.query_ras_error_count = &gfx_v9_4_2_query_ras_error_count,
@@ -1934,5 +1923,4 @@ struct amdgpu_gfx_ras gfx_v9_4_2_ras = {
 		.hw_ops = &gfx_v9_4_2_ras_ops,
 	},
 	.enable_watchdog_timer = &gfx_v9_4_2_enable_watchdog_timer,
-	.query_utcl2_poison_status = gfx_v9_4_2_query_uctl2_poison_status,
 };

drivers/gpu/drm/amd/amdkfd/kfd_int_process_v10.c

@@ -367,10 +367,20 @@ static void event_interrupt_wq_v10(struct kfd_node *dev,
 		   client_id == SOC15_IH_CLIENTID_UTCL2) {
 		struct kfd_vm_fault_info info = {0};
 		uint16_t ring_id = SOC15_RING_ID_FROM_IH_ENTRY(ih_ring_entry);
+		uint32_t node_id = SOC15_NODEID_FROM_IH_ENTRY(ih_ring_entry);
+		uint32_t vmid_type = SOC15_VMID_TYPE_FROM_IH_ENTRY(ih_ring_entry);
+		int xcc_id = 0;
 		struct kfd_hsa_memory_exception_data exception_data;
 
-		if (client_id == SOC15_IH_CLIENTID_UTCL2 &&
-				amdgpu_amdkfd_ras_query_utcl2_poison_status(dev->adev)) {
+		if (!vmid_type && dev->adev->gfx.funcs->ih_node_to_logical_xcc) {
+			xcc_id = dev->adev->gfx.funcs->ih_node_to_logical_xcc(dev->adev,
+				node_id);
+			if (xcc_id < 0)
+				xcc_id = 0;
+		}
+
+		if (client_id == SOC15_IH_CLIENTID_UTCL2 && !vmid_type &&
+		    amdgpu_amdkfd_ras_query_utcl2_poison_status(dev->adev, xcc_id)) {
 			event_interrupt_poison_consumption(dev, pasid, client_id);
 			return;
 		}

drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c

@@ -413,10 +413,20 @@ static void event_interrupt_wq_v9(struct kfd_node *dev,
 		   client_id == SOC15_IH_CLIENTID_UTCL2) {
 		struct kfd_vm_fault_info info = {0};
 		uint16_t ring_id = SOC15_RING_ID_FROM_IH_ENTRY(ih_ring_entry);
+		uint32_t node_id = SOC15_NODEID_FROM_IH_ENTRY(ih_ring_entry);
+		uint32_t vmid_type = SOC15_VMID_TYPE_FROM_IH_ENTRY(ih_ring_entry);
+		int xcc_id = 0;
 		struct kfd_hsa_memory_exception_data exception_data;
 
-		if (client_id == SOC15_IH_CLIENTID_UTCL2 &&
-		    amdgpu_amdkfd_ras_query_utcl2_poison_status(dev->adev)) {
+		if (!vmid_type && dev->adev->gfx.funcs->ih_node_to_logical_xcc) {
+			xcc_id = dev->adev->gfx.funcs->ih_node_to_logical_xcc(dev->adev,
+				node_id);
+			if (xcc_id < 0)
+				xcc_id = 0;
+		}
+
+		if (client_id == SOC15_IH_CLIENTID_UTCL2 && !vmid_type &&
+		    amdgpu_amdkfd_ras_query_utcl2_poison_status(dev->adev, xcc_id)) {
 			event_interrupt_poison_consumption_v9(dev, pasid, client_id);
 			return;
 		}