[Gluster-devel] Segmentation fault with afr on dapper using fuse 2.6.3 kernel module

Anand Avati avati at zresearch.com
Mon Apr 30 07:48:50 UTC 2007


Bernhard,
 just committed a fix for it. strange the bug never showed up so far!
please update to patch-134 to get the fix. 
thanks!

avati

On Mon, Apr 30, 2007 at 09:38:00AM +0200, Bernhard J. M. Grün wrote:
> Hi!
> 
> Here is my first try of a back trace:
> #0  0x00002aaaab182954 in memcpy () from /lib/libc.so.6
> #1  0x00002aaaab45a943 in afr_opendir_cbk ()
>   from /opt/glusterfs/lib/glusterfs/1.3.0-pre3/xlator/cluster/afr.so
> #2  0x00002aaaab351897 in client_opendir_cbk ()
>   from /opt/glusterfs/lib/glusterfs/1.3.0-pre3/xlator/protocol/client.so
> #3  0x00002aaaab3534d0 in client_protocol_interpret ()
>   from /opt/glusterfs/lib/glusterfs/1.3.0-pre3/xlator/protocol/client.so
> #4  0x00002aaaab353143 in client_protocol_notify ()
>   from /opt/glusterfs/lib/glusterfs/1.3.0-pre3/xlator/protocol/client.so
> #5  0x00002aaaaabd0c66 in transport_notify (this=0x54e5f0, event=1)
>    at transport.c:148
> #6  0x00002aaaaabd13b5 in epoll_notify (eevent=1, data=0x54e5f0) at 
> epoll.c:53
> #7  0x00002aaaaabd16a0 in sys_epoll_iteration (ctx=0x7fffffaa9c40)
>    at epoll.c:145
> #8  0x00002aaaaabd0e24 in poll_iteration (ctx=0x7fffffaa9c40)
>    at transport.c:251
> #9  0x0000000000403792 in main ()
> 
> Does this already help you?
> I know it is without debugging symbols. I'll try to get them in - but how?.
> In the meantime I also updated to patch-132 and the error is the same.
> The coredump I used for that back trace is also from patch-132.
> I think the problem is not only present with patch-131. It is just the
> version I tried out.
> 
> Bernhard
> 
> 2007/4/30, Anand Avati <avati at zresearch.com>:
> >
> >> You can download the logs from http://h0t.de/dapper-logs.tar.gz
> >
> >I saw the logs, they only indicate that there was a coredump.
> >
> >> I already tried it with the latest tla checkout. I think it was
> >> patch-131 that I tried. And that version worked on feisty (that uses
> >> the same fuse version).
> >
> >Do you mean that patch-131 did NOT work on dapper? you should have a
> >coredump from the segfault, is it possible to get a backtrace from that?
> >
> >avati
> 

-- 
ultimate_answer_t
deep_thought (void)
{ 
  sleep (years2secs (7500000)); 
  return 42;
}





More information about the Gluster-devel mailing list