| To: | davem@xxxxxxxxxx (David S. Miller) |
|---|---|
| Subject: | Re: TCP hang with 2.2.14 <-> 2.2.15pre5 |
| From: | kuznet@xxxxxxxxxxxxx (Alexey Kuznetosv) |
| Date: | Sun, 26 Mar 2000 22:05:53 +0800 (SGT) |
| Cc: | kuznet@xxxxxxxxxxxxx, sfr@xxxxxxxxxxxxxxxx, netdev@xxxxxxxxxxx |
| In-reply-to: | <200003261127.DAA01525@pizda.ninka.net> from "David S. Miller" at Mar 26, 2000 03:27:41 AM |
| Sender: | owner-netdev@xxxxxxxxxxx |
Hello! > Kernel fix is in 2.2.15-pre* for some time now, look at this > change to tcp_timer.c:tcp_write_err() But I do not see how it could explain sending out-of-window segment. Look, instead of zero window probe it retransmits next segment. It is interesting that it is even correct behaviour, the result is the same as with sending zero window probe 8)8) But it was not expected yet! Actualy, it looks like receiver shrunk window (there was not enough information in tcpdump to check this), or like sender miscalculated it. Well, both sides are linuxes, so that all the bugs are ours. Alexey |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: TCP hang with 2.2.14 <-> 2.2.15pre5, David S. Miller |
|---|---|
| Next by Date: | Re: 3c59x.c, Andrew Morton |
| Previous by Thread: | Re: TCP hang with 2.2.14 <-> 2.2.15pre5, David S. Miller |
| Next by Thread: | Re: TCP hang with 2.2.14 <-> 2.2.15pre5, Alexey Kuznetsov |
| Indexes: | [Date] [Thread] [Top] [All Lists] |