Re: mips branches broken between 2002-09-05 and 2002-09-06

[cgd at broadcom dot com]

At Tue, 10 Sep 2002 19:22:10 +0930, Alan Modra wrote:
> There's a good chance that I broke it.  Sorry, but this sort of
> cleanup is bound to have fallout.

Oh, I understand; i wasn't trying to assess blame, just report the
facts.  "If you're going to make an omelete..."  8-)


> You, or some other mips maintainer,
> should probably look at doing that.

FWIW, i see myself as "somebody who occasionally helps out with
MIPS-related stuff in binutils, and who may be the only person left
who cares (and whose users care 8-) about certain grotty bits of the
code."  8-)


> > 	(a) branches to global symbols (in the same file) are now
> >         relocated against the actual target symbols,
> 
> I believe that is correct.  For ELF shared libraries, ld.so must be
> able to override global symbols.

I don't actually care one way or another how the relocations
look, i was just noting that it was different.  I'm OK with your (and
H.J.'s) intention.

However, a question (for H.J. too):

So, when actually overriding the symbols being called via _branches_,
what result is expected other than "relocation overflow" or similar
for most of the relocations to symbols other than the one originally
used?  (or even that one, if it's not in the same file?)

(I don't think that's a problem.  In code where symbols can be
overridden, it's not smart, offset-wise, to use mips branch
instructions...  and it's better to get an error than to happily,
silentely invoke the non-overridden symbol.)

Is that the intent?


> > 	(c) as a result, the offsets in the final link end up being
> > 	incorrect.
> 
> Obviously, if a final link fails I need to help you in fixing it..

perhaps s/you/Eric/ 8-)




chris