This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: RFD, draft patch: IRA costs for reg_equiv_invariant regs
- From: Bernd Schmidt <bernds at codesourcery dot com>
- To: "Vladimir N. Makarov" <vmakarov at redhat dot com>
- Cc: GCC Patches <gcc-patches at gcc dot gnu dot org>
- Date: Wed, 02 Jun 2010 18:40:21 +0200
- Subject: Re: RFD, draft patch: IRA costs for reg_equiv_invariant regs
- References: <4BE46C12.6090206@codesourcery.com> <4BEC8539.5030804@redhat.com> <4C06851A.6040608@codesourcery.com>
On 06/02/2010 06:21 PM, Bernd Schmidt wrote:
> Can you redo these tests? I just ran tests with current gcc sources.
> For 32 bit SPEC2k on a Westmere Xeon (essentially Core i7) I'm seeing a
> minor perlbmk improvement with various compilation options including -O3
> -mpc64.
At one point I was even seeing a 10% perlbmk improvement with my patch,
but I didn't mention it since it went away. Now that I've thought about
it - maybe what we're both seeing are artefacts of the "Turbo Boost"
functionality of recent Intel chips, i.e. one set of benchmarks being
run at a different clock speed? It's probably a good idea to disable
that in the BIOS.
Bernd