Thanks, I applied most of this except for some minor personal style
things. For example I left out changes like the below :)
> - return sizeof (struct srp_cmd);
> + return sizeof(struct srp_cmd);
By the way, what's your feeling about upstream inclusion for the
driver in its current state?
> Will there be real error handling for srp one day?
Yes, it's on my TODO list. I need to figure out exactly what makes
sense. We definitely need to gracefully handle network failures and
handle it when the target logs us out. For the EH routines, I'm still
pondering...
I'm not sure there's anything sensible we can do in the abort routine,
since we don't know if the command has made it out of our local send
queue or not, and we can't stop it even if it hasn't been sent. The
reset routine should just tear down our connection and reconnect.
- R.
_______________________________________________
openib-general mailing list
[email protected]
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
