This is the mail archive of the
mailing list for the GCC project.
Re: compatibility Clarification
- To: Mark Mitchell <mark at codesourcery dot com>
- Subject: Re: compatibility Clarification
- From: Rainer Orth <ro at TechFak dot Uni-Bielefeld dot DE>
- Date: Tue, 22 May 2001 22:05:49 +0200 (MEST)
- Cc: rth at redhat dot com, aoliva at redhat dot com, madhu dot srinivasan at oracle dot com, gcc at gcc dot gnu dot org, Stan Shebs <shebs at apple dot com>
- References: <yddititfvbr.fsf@xayide.TechFak.Uni-Bielefeld.DE><20010522124723M.firstname.lastname@example.org>
Mark Mitchell writes:
> Who is the Tru64 maintainer? Or is there one?
I don't think we have a specific Tru64 maintainer, at least none is listed
in MAINTAINERS. There `only' Richard Henderson as the maintainer of the
alpha port, but he may well have been somewhat overwhelmed by all his other
> I don't think the autoconf test is worth it here. I *would* just
> hard-code the OS numbers for which the -oldas flag is needed. Also,
> if we cannot build on V5.0a because of this issue, then we should have
> documentation in the installation document for that.
Ok, I could do this (or just hardcode the use of -oldas in alpha/osf5.h),
since it's needed in V5.1 (and up, probably). We can always provide a
specific test later if Compaq releases a fixed adu whose output mips-tfile
can deal with.
Richard had some reservations about using -oldas at all:
but unfortunately it's the only option right now.
I'll post a new patch once it's clear what route to follow here.
> It is certainly bogus to have an explicit declaration of `strlen' like
> that in libobjc. This patch is OK. Do you have write privileges so
> that you can apply this one?
No, I don't. I've contacted Ovidiu Predescu (and gcc-patches) when I
noticed this patch only went on the trunk, but no reply yet:
> Please send mail to email@example.com. I am sorry that you
> haven't gotten a reply yet. Please indicate that this is a regression
> from GCC 2.95 as you mention.
Ok, I'll try again. Unfortunately, there are serious problems with the
current dejagnu based testsuite which links all tests with -static (PR
> Revise the patch as you suggest and repost it, copying Alexandre Oliva
> <firstname.lastname@example.org>. Alexandre can approve configuration patches.
Ok, will do.
Rainer Orth, Faculty of Technology, Bielefeld University