> 1. double lock on device queue
yes. here's a "fresh" exerpt from the syslog:
May 13 23:22:24 rincewind kernel: double lock on device queue!
May 13 23:56:32 rincewind kernel: double lock on device queue!
May 14 02:39:45 rincewind kernel: double lock on device queue!
May 14 06:21:37 rincewind kernel: double lock on device queue!
seem these messages coincide with the mail-flood from vger coming in.
> 2. repeated socket destroy messages every 10 seconds
they would go, as soon as i terminate xntpd. (seen them on every shutdown
for a long time).
[7|8:22] akohlmey@rincewind:~/compile> netstat -u
Active Internet connections
Proto Recv-Q Send-Q Local Address Foreign Address (State)
User
udp 0 0 rincewind.chemie.u:721 temptc.chemie.uni:2049 ESTABLISHED
udp 0 0 localhost:718 localhost:amd ESTABLISHED
udp 0 0 localhost:ntp *:*
udp 0 1260 rincewind.chemie.u:ntp *:*
^^^^
and here's the number that will come with it. btw: i can ping hosts, that
are actually down, get a few socket destroy delayed msgs, and then it like
before. this is how it _should_ be, right?
> 7. TCP performance problems.
> 8. bogus packet size/mismatched pointers reports on 8390 based ethernet
> drivers.
since the smc-driver update my networking is faster than ever.
> I want to see how bad these problems actually are and how many people are
> seeing them - also hopefully a pattern so try and include hardware info.
i486-33MHz, ISA-bus, SMC-Ultra, 32MB, Mitsumi FX001D, SoundBlaster 16,
1.28 Gig Quantum and 540Meg Maxtor
Axel.
=========================================================================
Axel Kohlmeyer email: axel.kohlmeyer@chemie.uni-ulm.de
Abteilung fuer Theoretische Chemie Universitaet Ulm D-89069 Ulm/Donau
=========================================================================