This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [patch] tuning gcc for Intel Core2


> On Tue, Nov 14, 2006 at 07:49:56AM -0800, H. J. Lu wrote:
> > 
> > Can you turn on x86_rep_movl_optimal for -mtune=generic to see what
> > it does for SPEC CPU 2K on both 32bit and 64bit?
> > 
> 
> Here are the 64bit SPEC CPU 2K results of -O2 + x86_rep_movl_optimal
> vs. -O2.  There are little changes in SPEC CPU 2K FP.  For SPEC CPU 2K
> INT:
> 
> On Core 2 Duo
> 			-O2 + x86_rep_movl_optimal vs -O2
> 164.gzip                         -0.0724113%
> 175.vpr                          -0.175234%
> 176.gcc                          14.6647%
> 181.mcf                          -0.475687%
> 186.crafty                       0.0384911%
> 197.parser                       0.322841%
> 252.eon                          0.039604%
> 253.perlbmk                      1.15354%
> 254.gap                          0.319781%
> 255.vortex                       0.0425532%
> 256.bzip2                        -0.14918%
> 300.twolf                        -0.0391543%
> Est. SPECint_base2000            1.23335%
> 
> On Nocona
> 
> 164.gzip                         0.0954198%
> 175.vpr                          -0.286533%
> 176.gcc                          10.2731%
> 181.mcf                          0.871731%
> 186.crafty                       -0.519818%
> 197.parser                       -0.205761%
> 252.eon                          -0.285225%
> 253.perlbmk                      0.306185%
> 254.gap                          -0.119474%
> 255.vortex                       0.406977%
> 256.bzip2                        -0.541377%
> 300.twolf                        -3.44411%
> Est. SPECint_base2000            0.519673%
> 
> On Core 2 Duo
> 
> 			-O2 + x86_rep_movl_optimal vs -O2 -mtune=core2
> 164.gzip                         0.729927%
> 175.vpr                          -1.09954%
> 176.gcc                          -0.174292%
> 181.mcf                          -0.94687%
> 186.crafty                       0.932039%
> 197.parser                       0.566343%
> 252.eon                          1.1209%
> 253.perlbmk                      -0.625244%
> 254.gap                          -0.27248%
> 255.vortex                       1.42364%
> 256.bzip2                        0.500501%
> 300.twolf                        0%
> Est. SPECint_base2000            0.146413%
> 
> So for 64bit, -O2 + x86_rep_movl_optimal is better than -O2 and
> -O2 -mtune=core2 on Core 2 Duo and Nocona. I suspect that we will
> see similar trend on Opteron.
> 
> Jan, given what we have now, should we turn on x86_rep_movl_optimal
> for m_GENERIC64 if it will also improve Opteron? We can always turn it
> off when your memory optimization is in.

Actually on Opteron the gcc benchmark also has problems with library
calls to string functions (but other benchmarks improve), those are not
as high as in your scores.  I am not quite sure if this is not an
annomaly. SPECint is not really good benchmark for a string functions as
only benchmark that do depend on them is GCC and VORTEX. In SPECfp
mesa/aspi.  By effectivly making compiler to optimize out any library
calls you throw away any posibility for optimized library version
handling nicely the large blocks, so programs that do use them (such as
X) slows down.

Do you have any understanding what really happens to GCC to slow down as
much?  (sorry I didn't had chance to read the full thread, but will do
so tomorrow, so I am sorry if this was already explained)

Honza
> 
> 
> H.J.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]