Hi, David Miller <davem@xxxxxxxxxxxxx> writes: > From: Felipe Balbi <felipe.balbi@xxxxxxxxxxxxxxx> > Date: Thu, 03 Nov 2016 09:04:54 +0200 > >> What Dave Miller is saying is that it's ALWAYS a bug to delay >> completion of SKBs. The only thing you're doing with chipidea is >> delaying interrupt by up to 125us; which is still a bug from the >> point of view of the networking layer, but it's more difficult to >> perceive any problems because of the short time where interrupt is >> delayed. > > I didn't say delaying was illegal. > > I said that the SKB free must occur in a reasonable, finite, amount of > time. "reasonable" is rather subjective. Completions are, of course, finite. It just means that once a transfer completes *and* has interrupt enabled, then several other SKBs will be completed along with it. That doesn't mean *all* SKBs completed at the same time, it means once we get an interrupt, we give back all previous ones, but we don't know exactly when they completed. -- balbi
Attachment:
signature.asc
Description: PGP signature