This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug optimization/8474] -mcpu=i686 produces much faster code than -mcpu=pentium4 on a P4 machine
- From: "pinskia at gcc dot gnu dot org" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: 30 Dec 2003 03:06:46 -0000
- Subject: [Bug optimization/8474] -mcpu=i686 produces much faster code than -mcpu=pentium4 on a P4 machine
- References: <20021105233600.8474.davejf@starband.net>
- Reply-to: gcc-bugzilla at gcc dot gnu dot org
------- Additional Comments From pinskia at gcc dot gnu dot org 2003-12-30 03:06 -------
On my pentium 4 2 GHz on the mainline, it is fixed:
tin:~/src/gnu/gcctest/pr8474>./a.out
whetstone took: 0.94 secs for 1064 MFLOPS (w/ math lib)
whetstone took: 0.23 secs for 4348 MFLOPS (w/o math lib)
tin:~/src/gnu/gcctest/pr8474>gcc -O3 whetstone.cpp -mcpu=pentium4 -O3 -ffast-math -
fmove-all-movables -fomit-frame-pointer -funroll-loops -lm
`-mcpu=' is deprecated. Use `-mtune=' or '-march=' instead.
tin:~/src/gnu/gcctest/pr8474>./a.out
whetstone took: 0.95 secs for 1053 MFLOPS (w/ math lib)
whetstone took: 0.22 secs for 4545 MFLOPS (w/o math lib)
tin:~/src/gnu/gcctest/pr8474>./a.out
whetstone took: 0.94 secs for 1064 MFLOPS (w/ math lib)
whetstone took: 0.23 secs for 4348 MFLOPS (w/o math lib)
.01 is the error margin.
--
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |RESOLVED
Resolution| |FIXED
Target Milestone|--- |3.4.0
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=8474