I talked this over further with upstream.  It looks like there were two
separate problems related to that symbol, one of which was the one I
send you details on and was related to debugging, and the other of which
was fixed some time back and I had forgotten about.  Sorry about that.

"Brent S. Elmer Ph.D." <webe...@aim.com> writes:

> Yes it is a custom built kernel.  No I do not have kernel debugging
> on.  It is a problem with openafs.  I tried compiling openafs in the
> kernel directory for 2.6.28 that I compiled openafs in successfully
> earlier with apparently an older openafs and the newer openafs in
> squeeze fails the same as it did in the 2.6.30 kernel compile.

And as you mentioned later, works with the version in sid.

We can't figure out why the sid version would work and the squeeze
version wouldn't, since none of the changes between squeeze and sid
should have affected this.  The relevant fix for what we think you're
running into went into 1.4.7, so it should have been around for quite a
while.

How are you building the kernel modules?  Is it possible that you had an
old build in the same source tree that wasn't fully cleaned, or that
might have reused configure results from an earlier kernel?  (If you're
using module-assistant, it's particularly bad about this.  You have to
run module-assistant clean <module> before building against a new kernel
or you'll get unpredictable results.)

-- 
Russ Allbery (r...@debian.org)               <http://www.eyrie.org/~eagle/>



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Reply via email to