On Monday 02 March 2009, Matthias Andree wrote: > Am 27.02.2009, 12:03 Uhr, schrieb Frans Pop <[email protected]>: > > On Friday 27 February 2009, you wrote: > >> > I should be able to catch this in a TCP dump if needed. > >> > >> That would be nice, please do so. > > > > Would an strace help as well? If so, for what system calls? > > please check the tcpdump for "URG"ent data - that's likely triggering > the messages. You could try to strace, the -e trace=network macro > should do.
I have an strace and tcpdump for a fetchmail run that produces the error. I don't see any URG data in the dump. The syslog shows (account name removed): Mar 3 20:38:01 elrond fetchmail[10659]: 3 messages (1 seen) for xxxxxxx at pop.planet.nl (29679 octets). Mar 3 20:38:01 elrond kernel: TCP(fetchmail:10659): Application bug, race in MSG_PEEK. Mar 3 20:38:01 elrond fetchmail[10659]: reading message [email protected]:2 of 3 (19564 octets) flushed Mar 3 20:38:01 elrond fetchmail[10659]: reading message [email protected]:3 of 3 (4644 octets) flushed As the traces contain private data, I don't want to attach them in the BR or publish them, but if people want to analyze them, please mail me privately and I'll send you the data. Cheers, FJP -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected]
