This is the mail archive of the
mailing list for the GCC project.
Re: [patch, bootstrap, hppa] Patch for bootstrap/40338
- From: Steve Ellcey <sje at cup dot hp dot com>
- To: Ian Lance Taylor <iant at google dot com>
- Cc: gcc-patches at gcc dot gnu dot org, dave at hiauly1 dot hia dot nrc dot ca
- Date: Mon, 15 Jun 2009 12:07:23 -0700
- Subject: Re: [patch, bootstrap, hppa] Patch for bootstrap/40338
- References: <200906121704.n5CH41126053@lucas.cup.hp.com> <email@example.com>
- Reply-to: sje at cup dot hp dot com
On Fri, 2009-06-12 at 18:34 -0700, Ian Lance Taylor wrote:
> Personally I would prefer to see the workarounds implemented. I think
> that file comparisons are an excellent way to find problems in the
> compiler, and we should strive to make that possible rather than
> avoiding it.
> But I will let the build system maintainers make the final decision.
Making the objects match is the prefered option but those changes
require more modifications to the overall make infrastructure then
restructing the comparision and I am not sure that disturbing all the
builds to that extent for the sake of one platform is the best approach.
So far I haven't heard of any platforms other then 32 bit HPPA having
this problem. It might be better to split the difference, modify the
build to use -frandom-seed= for PA in order to fix most of the
comparision issues and then just skip the comparsions on the