[OpenAFS-devel] Solaris 8 patch 108528-20 breaks afs module

[Douglas E. Engert]

Sergio Gelato wrote:
> 
> * Douglas E. Engert [2003-05-07 09:38:41 -0500]:
> > > Tried your fix. Got a BAD TRAP: type=9 after a few minutes of browsing
> > > a remote AFS cell. Stack trace says it was AFS-related. Will beat on
> > > it some more, to see if it's reproducible (it might have been a fluke).
> >
> > I would expect it to fail immediately if the patch had a problem,
> > as the tables it would have  touched where to find the syscall, and
> > lod module.
> 
> So would I, and indeed I have not been able to make it crash again.
> A fluke it must have been.
> 
> > > What's already clear is that the patch doesn't lead to a modload-time
> > > panic (with kernel 108528-19; will try the famous -20 later).
> 
> I have now tested it with kernel -20 as well, and it seems OK. (32-bit
> Solaris only, as I'm testing on the oldest hardware I've got. 64-bit
> tests will follow shortly.)

Send these coments to the developers as well.


-- 

 Douglas E. Engert  <DEEngert@anl.gov>
 Argonne National Laboratory
 9700 South Cass Avenue
 Argonne, Illinois  60439 
 (630) 252-5444