On Thu, 11 Oct 2007, TAKANO Ryousei wrote: > From: "Ilpo Järvinen" <[EMAIL PROTECTED]> > Subject: [RFC PATCH] [TCP]: Fix lost_retrans loop vs fastpath problems > Date: Tue, 9 Oct 2007 15:20:01 +0300 > > Thanks Ilpo! I am trying to evaluate this patch.
There's a minor problem in this 2nd patch, it's just preventing the cnt == tp->retrans_out short-circuit from working, not a correctness problem though it could affect the performance. I'll post larger patch series among which is a fixed version (hopefully today). > But, I got > a kernel panic at net_rx_action() in our experimental setting. > I am using the net-2.6.24 kernel _without_ this patch. > (I will post a bug report separately). ...Please do. :-) > Anyway, I will report the result as soon as possible. Thanks. ...It's very interesting to see because it's not that clear cut how the extra processing that is necessary affects high-speed performance, it could add yet another source of RTOs due processing latency. -- i.