RE: [hobbit] hobbitlaunch segfault / timewarp happend again

["Dugan, Darin D [EIT]" <dddugan (at) iastate.edu>]

I recently brought up a new client that has trouble keeping accurate
time...so I began encountering this time warp issue. As pointed out by
Henrik in January, it is definitely an infinite loop where errprintf()
calls getcurrenttime() to get its timestamp.

The attached patch modifies the functions in errormsg.c to use time()
instead of getcurrenttime(). That avoids any recursion-infinite loop
problems, and logs or prints errors with the system's actual time
instead of a Hobbit-adjusted-for-sanity time. In the absence of accurate
time, I think it would be best to log in the system's time so that you
can correlate Hobbit logs with other system logs.

Working for me so far, but use at your own risk. Comments?

Cheers.
Darin Dugan
dddugan (at) iastate.edu

-----Original Message-----
From: Alexander Keller [mailto:hobbit (at) alexkeller.de] 
Sent: Friday, March 21, 2008 10:22 AM
To: hobbit (at) hswn.dk
Subject: Re: [hobbit] hobbitlaunch segfault / timewarp happend again

Hi,

unfortunately nobody answered to my posting, so I did a quick'n dirty
hack to prevent timewarp segfaults in hobbitlaunch.

Just comment out the errprintf-statement in lib/timefunc.c:

  if (timewarphappened) {
  /*
   * Tell the world about it.
   * Must do this AFTER changing timewarp and lastresult,
   * or we will start an endless loop triggering a stack
   * overflow because errprintf() calls getcurrenttime().
   */
           /*
           * **** prevent segfault: do not log time warp. ****
           * errprintf("Time warp detected: Adjusting returned clock by
%d seconds\n", timewarp);
           */
   }

This is not a real solution, but it works for me. Maybe there is
somebody out, who can fix this issue properly :-)

Regards
 Alexander


> Hi Henrik,

> in january I reported a segfault with hobbitlaunch/timefunc.c. You
quickly
> provided a patch...

> Now I'm having a new error - see core dump:

> /opt/hobbit/client# gdb bin/hobbitlaunch core
> GNU gdb 6.4-debian
> Copyright 2005 Free Software Foundation, Inc.
> GDB is free software, covered by the GNU General Public License, and
you are
> welcome to change it and/or distribute copies of it under certain
conditions.
> Type "show copying" to see the conditions.
> There is absolutely no warranty for GDB.  Type "show warranty" for
details.
> This GDB was configured as "i486-linux-gnu"...Using host libthread_db
> library "/lib/tls/i686/cmov/libthread_db.so.1".

> Core was generated by `./bin/hobbitlaunch
--config=./etc/clientlaunch.cfg
> --log=./logs/clientlaunch.lo'.
> Program terminated with signal 11, Segmentation fault.

> warning: Can't read pathname for load map: Input/output error.
> Reading symbols from /usr/lib/libz.so.1...done.
> Loaded symbols for /usr/lib/libz.so.1
> Reading symbols from /lib/tls/i686/cmov/libc.so.6...done.
> Loaded symbols for /lib/tls/i686/cmov/libc.so.6
> Reading symbols from /lib/ld-linux.so.2...done.
> Loaded symbols for /lib/ld-linux.so.2
> #0  errprintf (fmt=0x6b <Address 0x6b out of bounds>) at errormsg.c:42
> 42              time_t now = getcurrenttime(NULL);
> (gdb) bt
> #0  errprintf (fmt=0x6b <Address 0x6b out of bounds>) at errormsg.c:42
> #1  0x0804f125 in getcurrenttime (retparm=0x0) at timefunc.c:73
> #2  0x0804b9e0 in errprintf (fmt=0x6b <Address 0x6b out of bounds>) at
> errormsg.c:42
> #3  0x0804f125 in getcurrenttime (retparm=0x0) at timefunc.c:73
> #4  0x0804b9e0 in errprintf (fmt=0x6b <Address 0x6b out of bounds>) at
> errormsg.c:42
> #5  0x0804f125 in getcurrenttime (retparm=0x0) at timefunc.c:73
> #6  0x0804b9e0 in errprintf (fmt=0x6b <Address 0x6b out of bounds>) at
> errormsg.c:42
> #7  0x0804f125 in getcurrenttime (retparm=0x0) at timefunc.c:73
> #8  0x0804b9e0 in errprintf (fmt=0x6b <Address 0x6b out of bounds>) at
> errormsg.c:42
> #9  0x0804f125 in getcurrenttime (retparm=0x0) at timefunc.c:73
> #10 0x0804b9e0 in errprintf (fmt=0x6b <Address 0x6b out of bounds>) at
> errormsg.c:42
> [...]

> I can reproduce the error with "ntpdate" using a misconfigured ntp
server
> (2 min in the past):

> 1. start hobbit client "runclient.sh start"
> 2. sync time with "ntpdate <misconfigured-time-server>"
> 3. get a core dump  


> Regards,
>  Alexander


To unsubscribe from the hobbit list, send an e-mail to
hobbit-unsubscribe (at) hswn.dk