> > Your patch descriptions need improvement. They describe what you do, but > > this can be seen in the patch as well. What you need to describe is WHY you > > need this change. > > In case the slave nacks the transaction in the middle of the transfer The driver continues. > > > What was wrong before, > > The usual expectation is that in case of error / nack the transaction is halted. Yeah, this is going into the right direction. Please note, that I meant that all 9 patch descriptions need rework. I know this is extra work, but really needed to keep drivers maintainable. Please resend the series when you are done. > > Because this driver is long in use, we must be very careful about regressions > > and every change needs a good reason. > > I agree. Great! :)
Attachment:
signature.asc
Description: Digital signature
