[PATCH] nvme-tcp: add basic support for the C2HTermReq PDU

[Maurizio Lombardi]

On Mon Feb 17, 2025 at 9:14 AM CET, Sagi Grimberg wrote:
>
>> +	fes = le16_to_cpu(pdu->fes);
>> +	if (fes && fes < ARRAY_SIZE(msg_table))
>> +		msg = msg_table[fes];
>> +	else
>> +		msg = "N/A";
>
> msg = "unknown" ? we shouldn't suggest N/A because it should be very 
> much applicable.

Changed it to "Unknown".

>
>> +
>> +	dev_err(queue->ctrl->ctrl.device,
>> +		"Received C2HTermReq (FES = %s)\n", msg);
>> +}
>> +
>
> AFAIR, the goal with C2HTerm was to also log the offending PDU itself. 
> However I don't think
> its really necessary looking back.

Yes, it also provides this information, I just didn't want
to make it too complicated to start with, I can improve it later.

>
> Can you test this also with header digest enabled?

Indeed there is a small change to do.
C2HTermReq is the only PDU coming from the controller that never includes
Header or Data digests, so the driver must skip the checks.

Will submit a V2 today.

Thanks,
Maurizio