Re: cvs 1.11.5 pserver sig11 crash on FreeBSD 4.8-R

[Scott Mitchell]

Hi Mark,

On Sun, Jul 27, 2003 at 01:21:23PM -0700, Mark D. Baushke wrote:
> 
> FYI. My understanding is that FreeBSD 4.6-RELEASE provides the
> RELENG_4_6_0_RELEASE tag on src/contrib/cvs sources which is releated to
> the cvshome cvs version 1.11.1p1 plus a number of local FreeBSD patches.
> 
> FreeBSD 4.8-RELEASE provides the RELENG_4_8_0_RELEASE tag on
> src/contrib/cvs sources which is related to the cvshome cvs version
> 1.11.5 plus a number of local FreeBSD patches.

That's my understanding too.  Running the two binaries with -v gives me
'1.11.1p1-FreeBSD' and '1.11.5-FreeBSD' respectively.

> It would be useful if you could try using the cvs-1.11.6 or cvs-1.12.1
> release from cvshome.org to see if it has the same problems. It would
> also be useful to know if the clients that are having the problems are
> at the same basic release as the server or not.

A lot of the clients, especially the Windows ones, will still be running
1.10.  We don't need any of the FreeBSD extensions with this repository,
so I should be able to try 1.11.6 and 1.12.1 without breaking anything.
We're right at the end of a release cycle just now, so I can't touch
anything before next week, and I plan to run some more RAM diagnostics on
the server before I start messing with the software.

> It would be good to know at least one of the client side configurations
> that cause the core dump on the server. It might be useful to also do
> a correlation to see if the same client version caused all of the server
> core dumps.

The two clients that I'm certain have triggered server-side core dumps
are 1.10 on a Win2K box and 1.11.1p1 on a Debian (2.4.18 kernel) box.
All I get is a core and a UID in the logs, so unless I actually see it
happen and grab the user right then to see what they sere doing, I've no
way of knowing which client machine was the culprit :-(

Thanks for the feedback.  I'll see if I can get one or both of those newer
versions compiled and try them out next week.

Cheers,

        Scott