This is the mail archive of the
mailing list for the glibc project.
Re: [PATCH] Define HAS_CPUID/HAS_I586/HAS_I686 from -march=
- From: "Carlos O'Donell" <carlos at redhat dot com>
- To: "H.J. Lu" <hjl dot tools at gmail dot com>, OndÅej BÃlka <neleai at seznam dot cz>
- Cc: GNU C Library <libc-alpha at sourceware dot org>
- Date: Tue, 18 Aug 2015 11:49:45 -0400
- Subject: Re: [PATCH] Define HAS_CPUID/HAS_I586/HAS_I686 from -march=
- Authentication-results: sourceware.org; auth=none
- References: <20150817231005 dot GA24205 at intel dot com> <20150818052442 dot GA26531 at domone> <CAMe9rOrqkqVJVrs9qy83u-vZTjOBk4Shd-H2zsBJTzc+sHa3Aw at mail dot gmail dot com>
On 08/18/2015 09:34 AM, H.J. Lu wrote:
> On Mon, Aug 17, 2015 at 10:24 PM, OndÅej BÃlka <firstname.lastname@example.org> wrote:
>> On Mon, Aug 17, 2015 at 04:10:05PM -0700, H.J. Lu wrote:
>>> cpuid, i586 and i686 instructions are available if the processor
>>> specified by -march= supports them. We can use this information
>>> to determine whether those instructions can be used safely.
>>> OK for master?
>> Patch itself looks ok. It fixes theoretical problem that if there would
>> be new processor without cpuid it would cause problem.
>> But could you explain rationale? Here you replaced one like with
>> equivalent longer macro and HAS_I586 is never used. Do you plan followup
>> patches that use that?
> I will use HAS_I586 and HAS_I686 in multi-arch for i486 and i586
> I am working on.
Does this include folding i486 up a directory and doing away with the
i386 support? Since we need the i486 instructions to implement NPTL.