On Fri, 21 Dec 2007, Ilpo Järvinen wrote: > On Fri, 21 Dec 2007, Bill Fink wrote: > > > If so it seems like a lot of unnecessary > > work just to avoid a 1 in 4 billion event, since it's my understanding > > that the whole tcp_tso_should_defer function is just an optimization > > and not a criticality to the proper functioning of TCP, especially > > considering it hasn't even been executing at all up to now. > > It would still be good to not return 1 in that case we didn't flag the > deferral, how about adding one additional tick (+comment) in the zero > case making tso_deferred == 0 again unambiguously defined, e.g.: > > tp->tso_deferred = min_t(u32, jiffies, 1);
Blah, max_t of course :-). -- i.