This is the mail archive of the
mailing list for the GCC project.
Re: [patch] tuning gcc for Intel Core2
- From: Vladimir Makarov <vmakarov at redhat dot com>
- To: "H. J. Lu" <hjl at lucon dot org>
- Cc: "gcc-patches at gcc dot gnu dot org" <gcc-patches at gcc dot gnu dot org>, denis_nagorny at linux dot intel dot com, grigory_zagorodnev at linux dot intel dot com
- Date: Mon, 13 Nov 2006 14:16:03 -0500
- Subject: Re: [patch] tuning gcc for Intel Core2
- References: <4558A4DD.email@example.com> <20061113171610.GA6826@lucon.org>
H. J. Lu wrote:
On Mon, Nov 13, 2006 at 12:01:17PM -0500, Vladimir Makarov wrote:I'll try to get them tommorow. But as I remember vortex is broken. I
saw analogous bug when I worked on my rgister allocator branch. I can
guess that is because of wrong aliasing tag which permits scheduler to
move load before store.
Here is the patch for tuning gcc for Intel Core2 processor. I did
about 30 SPEC2000 runs to find good parameters which are practically
the same what Intel gave and recommended in their optimization guide
made public a few days ago.
The patch increases SPECINT2000 score to 1963 from 1925 (for
generic) or 1901 (for nocona). SPECFP2000 sore is the same as for
generic 1875 (nocona has 1856). One benchmark (gcc) did particular
well -- about 20% improvement (1788 for generic tuning vs 2210 for
core2). The size of code generated for Core2 is smaller (0.46% for
SPECInt and 0.54% SPECFp) than one for generic.
Thanks. Do you happen to have -m32 numbers? We will run 32bit on Core
and Core 2.