Hello. I run a busy ftp server (about 35 000 xferlog lines a day). Since I've upgraded the system to RH7.0, I have, in /var/log/messages, lines like: Dec 7 19:24:26 colt1 ftp[29557]: exiting on signal 11: Segmentation fault I ran 'strace -fp <pid_of_inetd>' and traced faulty ftpd. I now have 4 log files (one per pid) showing wu-ftpd performing a SIGSEGV at the end of a transfert, just after producing message #426 ("426 Data connection: .."). The system has been updated with the latest patch and is managed by RedHatNetwork (colt1.cpio.com). I can give the 4 log files to RH tech staff upon request (about 10K after gzip, 150K before). Full log file of all process launched by inetd can be given also (55 megs before gzip, 4.3megs after :-)) Thanks.
Please add the log files - I can't reproduce the problem here (and looking at the code doesn't show something obviously wrong).
Created attachment 6293 [details] Trace of PID 29417 : this process dies after a transfert (server)
Created attachment 6294 [details] Yet another strace (pid 29485)
Created attachment 6295 [details] Yet another strace (pid 29558)
Created attachment 6296 [details] Yet another strace (29577)
So you get the strace of the 4 pids that crashed as I wrote. I can also send the full strace of inetd, but I do not want to post it in a public place like here since it may give information about the computer it was taken from. I can send it to a particular person however. Thanks.
I knew there was a bug like this and that it had to do with dropped connections. This information has helped me localize the problem. Give me a few days to work up a patch.
I am having exactly the same problem: Feb 6 19:17:20 ftp ftpd[1560]: exiting on signal 11: Segmentation fault The FTP server is running version wu-2.6.1(1) as shipped with RH7.0. Is the patch mentioned above available yet? Thanks.
Created attachment 12404 [details] Don't close already closed FILE*
*** Bug 14419 has been marked as a duplicate of this bug. ***
This is fixed in Red Hat Linux 7.1