Garrett Cooper yanefbsd at
Sun Dec 14 02:31:22 UTC 2008

Hi guys,
     I just came across this issue trying to update a repository to
push some patches upstream to another project I work with:

[gcooper at optimus /scratch/ltp]$ gdb cvs
Type "help" followed by command name for full documentation.
Command name abbreviations are allowed if unambiguous.
(gdb) set args up
(gdb) r
Starting program: /usr/bin/cvs up

Program received signal SIGPIPE, Broken pipe.
0x0000000800bdfeec in write () from /lib/
[gcooper at optimus /scratch/ltp]$ uname -a
Sun Dec  7 17:44:13 PST 2008
gcooper at  amd64

     It looks like a corrupted stack of some flavor because when I do
bt it prints out hundreds of ??? before it eventually fails.
     The sources I originally saw this issue with were based off mid
November (don't know when to look up the actual version date because
I've rebuilt the kernel a few times and cvsup'ed once or twice since
then), but my previous major sync (mid October 2008) didn't yield this
issue. I just recently rebuilt kernel + world (today) and I'm getting
the same result.
     Other apps requiring socket connections work (portsnap, fetch,
ssh, svn, etc).
     I'll provide more info if requested.
     For the time being I'll have to move to svn for the time being
because the project supports SVN as well as CVS.

More information about the freebsd-current mailing list