This is the mail archive of the
mailing list for the GCC project.
Re: CC register for x86-64 asm statements. Was: glibc: syscalls for x86-64
- To: Robert Dewar <dewar at gnat dot com>
- Subject: Re: CC register for x86-64 asm statements. Was: glibc: syscalls for x86-64
- From: Richard Henderson <rth at redhat dot com>
- Date: Wed, 22 Nov 2000 18:19:30 -0800
- Cc: aj at suse dot de, discuss at x86-64 dot org, gcc at gcc dot gnu dot org, jh at suse dot cz
- References: <20001123021434.7143E34D80@nile.gnat.com>
On Wed, Nov 22, 2000 at 09:14:34PM -0500, Robert Dewar wrote:
> There are actually many uses of retaining flags over instructions that
> do not affect them...
I'll grant that there are a few such instances. But by far
the bulk of them concern instructions that affect _no_ flags
rather than affecting _some_ flags.
I'd be willing to bet beer that if you actually did the work
to track all of the flag bits individually on x86, you couldn't
measure a performance difference over the current scheme.