[OpenAFS-devel] RH 7.1 spec file: archlist

[Jeffrey Hutzelman]

On 25 Jul 2001, Derek Atkins wrote:

> Daniel Jacobowitz <dmj+afs@andrew.cmu.edu> writes:
> 
> > On Wed, Jul 25, 2001 at 05:22:18PM -0400, Derek Atkins wrote:
> > > Rudolph T Maceyko <rtm@cert.org> writes:
> > > 
> > > > The iget4 symbol is the same but the libafs modules themselves are NOT:
> > > 
> > > Oooh, that does not bode well.  I would be interested in knowing what
> > > the differences are.  This means that the SymTable is slightly broken;
> > > I'll need to apply the architecture to the SymTable as well.  Hmm..
> > > Not too hard for the x86 processors, but might be more challenging for
> > > the athlon.. Maybe...
> > 
> > I don't think it necessarily implies that - can't a change in config
> > affect the versions of some symbols and not others?
> 
> Perhaps, however _SO MUCH_ is encompassed in 'struct inode' that iget
> almost always changes if any other structure changes...  As I said,
> it would be interesting the note what is different.

Note that there's nothing that says you have to use iget4.  We used to use
iget back in the 2.0.x days because it was pretty much guaranteed to
change when anything we cared about did.  These days the matching code is
more general, so it can match on any symbol name.  If it turns out that
iget4 isn't doing the job, it's possible to use some other symbol.

-- Jeff