Re: [PATCH RESEND v2 06/12] coco: host: arm64: Add RMM device communication helpers

[Aneesh Kumar K.V <aneesh.kumar@kernel.org>]

Jonathan Cameron <jonathan.cameron@huawei.com> writes:
...

>> +		/*
>> +		 * Some device communication error will transition the
>> +		 * device to error state. Report that.
>> +		 */
>> +		if (type == PDEV_COMMUNICATE)
>> +			ret = rmi_pdev_get_state(virt_to_phys(pf0_dsc->rmm_pdev),
>> +						 (enum rmi_pdev_state *)&state);
>> +		if (ret)
>> +			state = error_state;
> Whilst not strictly needed I'd do this as:
>
> 		if (type == PDEV_COMMUNICATE) {
> 			ret = rmi_pdev_get_state(virt_to_phys(pf0_dsc->rmm_pdev),
> 						 (enum rmi_pdev_state *)&state);
> 			if (ret)
> 				state = error_state;
> 		}
>
> Just to make it clear that reg check is just on the output of the call above.
> If we didn't make that call it is definitely zero but nice not to have
> to reason about it.
>

Some of this is because follow up patch adds more details there. In this case.

		/*
		 * Some device communication error will transition the
		 * device to error state. Report that.
		 */
		if (type == PDEV_COMMUNICATE)
			ret = rmi_pdev_get_state(virt_to_phys(pf0_dsc->rmm_pdev),
						 (enum rmi_pdev_state *)&state);
		else
			ret = rmi_vdev_get_state(virt_to_phys(host_tdi->rmm_vdev),
						 (enum rmi_vdev_state *)&state);
		if (ret)
			state = error_state;



>
>> +	}
>> +
>> +	if (state == error_state)
>> +		pci_err(tsm->pdev, "device communication error\n");
>> +
>> +	return state;
>> +}
>> +

-aneesh