Re: gcc/mozilla bug.
Melvin Jenklowiecz <jenklowiecz@yahoo.com> writes:
> We determined that yahoo was using som,e form of md5hash, and that
> the alpha was generating wildly different hashes from any other
> machine. A bug was filed on mozilla, and can be found here:
>
> http://bugzilla.mozilla.org/show_bug.cgi?id=220732
Unless I'm mistaken, this is gcc PR 9164 (http://gcc.gnu.org/PR9164),
which was fixed quite some time ago. Most gcc people don't devote a
lot of time to older gcc versions, so I'm not sure anybody is going to
do a backport. Since the patch was fairly small, it might be easy,
though. I suggest you try to apply it to stable's gcc and then open a
bug report with a patch.
> It also links to a gcc bug, which is the reason for RedHat's
> switching to the Compee/H-paq patched gcc 2.96.
I can't find the link, can you please show it to me?
> There is a good possibility that most of the random wierness with X,
> Afterstep, and xine among other things are caused by the gcc bug.
Unlikely, it's a pretty obscure bug.
--
Falk
Reply to: