drm/amdgpu: add RAS poison handling for MCA

For MCA poison, if unmap queue fails, only gpu reset should be
triggered without page retirement handling, MCA notifier will do it.

v2: handle MCA poison consumption in umc_poison_handler directly.
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_umc.c

@@ -169,7 +169,9 @@ int amdgpu_umc_poison_handler(struct amdgpu_device *adev,
 		void *ras_error_status,
 		bool reset)
 {
-	int ret;
+	int ret = AMDGPU_RAS_SUCCESS;
+
+	if (!adev->gmc.xgmi.connected_to_cpu) {
 		struct ras_err_data *err_data = (struct ras_err_data *)ras_error_status;
 		struct ras_common_if head = {
 			.block = AMDGPU_RAS_BLOCK__UMC,
@@ -183,6 +185,13 @@ int amdgpu_umc_poison_handler(struct amdgpu_device *adev,
 			obj->err_data.ue_count += err_data->ue_count;
 			obj->err_data.ce_count += err_data->ce_count;
 		}
+	} else if (reset) {
+		/* MCA poison handler is only responsible for GPU reset,
+		 * let MCA notifier do page retirement.
+		 */
+		kgd2kfd_set_sram_ecc_flag(adev->kfd.dev);
+		amdgpu_ras_reset_gpu(adev);
+	}
 
 	return ret;
 }