Re: ra_svn tests stuck in diff_tests.py

On Tue, 2003-01-21 at 01:08, David Summers wrote:
 Here is a backtrace of one svnserve:
 #0 0x420e8132 in accept () from /lib/i686/libc.so.6

This is the master process waiting for a connection.

 Here is the backtrace of another svnserve:
 #0 0x420b48b9 in wait4 () from /lib/i686/libc.so.6
 #1 0x4213030c in __DTOR_END__ () from /lib/i686/libc.so.6
 #2 0x08050c9c in strcpy ()

This doesn't look like a very healthy stack trace. I have no idea
what's going on here.

 Here is the backtrace of the third svnserve:
 #0 0x420e19ee in select () from /lib/i686/libc.so.6
 #1 0x400db77c in __DTOR_END__ () from /usr/lib/libdb-4.0.so
 #2 0x400bd5e5 in __os_yield () from /usr/lib/libdb-4.0.so
 #8 0x40033457 in commit_trail (trail=0x8086aa8, fs=0x80795d0)
     at subversion/libsvn_fs/trail.c:100

And here we're modifying the DB during a report operation.

 Does any of that help determine why it hung? It seems to be sporadic,
 sometimes it hangs and sometimes it doesn't.

Looks like a race condition in the FS. Beyond that, I don't know.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sat Oct 14 02:03:29 2006