This is the mail archive of the
fortran@gcc.gnu.org
mailing list for the GNU Fortran project.
Re: Polyhedron 2005 Benchmarks
- From: Steve Kargl <sgk at troutmask dot apl dot washington dot edu>
- To: Jack Howarth <howarth at bromo dot msbb dot uc dot edu>
- Cc: fortran at gcc dot gnu dot org
- Date: Sun, 27 Nov 2005 07:13:30 -0800
- Subject: Re: Polyhedron 2005 Benchmarks
- References: <20051127030101.40EB61DC06D@bromo.msbb.uc.edu>
On Sat, Nov 26, 2005 at 10:01:01PM -0500, Jack Howarth wrote:
> Do patches exist for running the harness of the Polyhedron 2005
> benchmarks under linux? I would like to get the harness working under
> Darwin. Currently I can run all of the benchmarks individually but
> it would be nice to be able to use the harness instead for benchmarking
> progress in the optimizations of gfortran.
> Jack
I cobbled together a Makefile file for my harness. I'm
use to BSD's make, so it may not work with GNU make.
It also uses a feature of /usr/bin/time that may not
be available on other systems. In playing around with
a few of these programs, I get the impression that some
are IO bound and the optimization level is unimportant.
Watch cut-n-paste tab corruption.
NAMES = ac.f90 aermod.f90 air.f90 capacita.f90 channel.f90 doduc.f90 \
gas_dyn.f90 induct.f90 linpk.f90 mdbx.f90 nf.f90 \
protein.f90 rnflow.f90 test_fpu.f90 tfft.f90 fatigue.f90
F95 = gfc41
FLG = -w -o z
all:
date > output
echo '-O' >> output
.for Q in ${NAMES}
${F95} ${FLG} -O $Q
echo $Q >> output
time -a -o output ./z > /dev/null
time -a -o output ./z > /dev/null
time -a -o output ./z > /dev/null
rm -f z
.endfor
--
Steve