Re: anyone having problems with ssh (sshd) 1.2.27?

[http://dummy.us.eu.org/robert]

 > From: Stephen Pitts <http://www.midsouth.rr.com/~smpitts>
 > Date: Sat  Jul 24,  8:23pm
 >
 > > 
 > > The only reason I know it's a SEGV is that I ran the debugger (gdb) on it
 > > (via "attach") and it says "Program exited with signal segmentation violation"
 > > (or something like that).
 > If you haven't already, rebuild ssh with "-g" in the makefile and any
 > -O options omitted. Then, wait for it to crash in gdb, and use "where"
 > to locate what is happening. Reply with the stack trace. Personally,
 > I'm running the Debian-packaged version of ssh on a machine with a 17+
 > day uptime with no problems. Then again, its also running 2.2.10 and
 > potato with glibc 2.1.
 > 
 > -- 
 > Stephen Pitts
 > http://www.midsouth.rr.com/~smpitts
 > webmaster - http://www.mschess.org

Ah.  So, you are saying that it has nothing to do with libc6, then?

Anyway, I did compile it with -g and ran it through the debugger (one must
"attach" to it because it automatically backgrounds itself), and it did
not tell me where it crashed!  It just says "Program exited with signal
segmentation violation" and "where" says that "Program is not active."

As a workaround, I have a cron job (running under root) that just restarts
sshd every 15 minutes.  So, at the moment, while actually using it (via a
modem), I cannot work on this bug.  But, I'll try to do so tomorrow when
I'm not using it.

I assume that the segmentation violation is happening in a system call
and that's why I can't get a traceback.