drm/amdgpu: Add msg handlers for SRIOV RAS Telemetry

Add message handlers for RAS telemetry.

Signed-off-by: Victor Skvortsov <[e-mail]>
Reviewed-by: Zhigang Luo <[e-mail]>
Signed-off-by: Alex Deucher <[e-mail]>

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

···
 	void (*ras_poison_handler)(struct amdgpu_device *adev,
 					enum amdgpu_ras_block block);
 	bool (*rcvd_ras_intr)(struct amdgpu_device *adev);
+	int (*req_ras_err_count)(struct amdgpu_device *adev);
 };
 
 /*

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

···
 static int xgpu_nv_mailbox_rcv_msg(struct amdgpu_device *adev,
 				   enum idh_event event)
 {
+	int r = 0;
 	u32 reg;
 
 	reg = RREG32_NO_KIQ(mmMAILBOX_MSGBUF_RCV_DW0);
-	if (reg != event)
+	if (reg == IDH_FAIL)
+		r = -EINVAL;
+	else if (reg != event)
 		return -ENOENT;
 
 	xgpu_nv_mailbox_send_ack(adev);
 
-	return 0;
+	return r;
 }
 
 static uint8_t xgpu_nv_peek_ack(struct amdgpu_device *adev)
···
 	case IDH_RAS_POISON:
 		if (data1 != 0)
 			event = IDH_RAS_POISON_READY;
+		break;
+	case IDH_REQ_RAS_ERROR_COUNT:
+		event = IDH_RAS_ERROR_COUNT_READY;
 		break;
 	default:
 		break;
···
 	return (msg == IDH_RAS_ERROR_DETECTED || msg == 0xFFFFFFFF);
 }
 
+static int xgpu_nv_req_ras_err_count(struct amdgpu_device *adev)
+{
+	return xgpu_nv_send_access_requests(adev, IDH_REQ_RAS_ERROR_COUNT);
+}
+
 const struct amdgpu_virt_ops xgpu_nv_virt_ops = {
 	.req_full_gpu	= xgpu_nv_request_full_gpu_access,
 	.rel_full_gpu	= xgpu_nv_release_full_gpu_access,
···
 	.trans_msg = xgpu_nv_mailbox_trans_msg,
 	.ras_poison_handler = xgpu_nv_ras_poison_handler,
 	.rcvd_ras_intr = xgpu_nv_rcvd_ras_intr,
+	.req_ras_err_count = xgpu_nv_req_ras_err_count,
 };