FAQ Search Today's Posts Mark Forums Read
» Video Reviews

» Linux Archive

Linux-archive is a website aiming to archive linux email lists and to make them easily accessible for linux users/developers.


» Sponsor

» Partners

» Sponsor

Go Back   Linux Archive > Gentoo > Gentoo Development

 
 
LinkBack Thread Tools
 
Old 12-14-2007, 01:12 PM
Matthias Langer
 
Default gentoo-x86 commit in sys-cluster/openmpi: ChangeLog openmpi-1.1.1.ebuild openmpi-1.2.4.ebuild

On Fri, 2007-12-14 at 12:13 +0000, Sébastien Fabbro wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> On 14/12/07 10:24, Matthias Langer wrote:
> >> On 02:10 Thu 13 Dec , Justin Bronder (jsbronder) wrote:
> >>> 1.1 sys-cluster/openmpi/openmpi-1.2.4.ebuild
> >>>
> >>> file : http://sources.gentoo.org/viewcvs.py/gentoo-x86/sys-cluster/openmpi/openmpi-1.2.4.ebuild?rev=1.1&view=markup
> >>> plain: http://sources.gentoo.org/viewcvs.py/gentoo-x86/sys-cluster/openmpi/openmpi-1.2.4.ebuild?rev=1.1&content-type=text/plain
> >
> > Why have you removed the "ifc" USE flag (as well as a few others), that
> > was present in the ebuilds that can still be found at bug 166787?
>
> For icc and f90-typesafe, I advised him to do so on #gentoo-science.

Well, too bad that gfortran generates code that is often not even half
as fast as the code generated by ifc. Thus, there is a *very valid*
reason to prefer icc over gfortran when it comes to performance critical
software, which programs that use MPI almost always are. However, the
fortran eclass prefers gfortran over ifc even if both are installed
(which is fine by me as long as it can be overridden) and the
fortran-compiler-wrapper "mpif90/opal_wrapper" (which is a binary for
some reason) wrapps gfortran then.

For sure, something like
F77="ifort" FC="ifort" FFLAGS="-O3 -xO" emerge -av openmpi
still does the trick, but if i want that "f90-typesave" stuff back, this
line would have to be certainly a bit longer...

Maybe someone can explain to me what positive side effects the removal
of the ifc USE flag has - and why this flag is generally discouraged.

>
> ifc/icc use flags should be avoided (see bug #97929).
> Disabling f90-typesafe did not make much sense as a use flag once you
> have the fortran one enabled, but why --with-mpi-f90-size=medium"
> and "--with-f90-max-array-dim=4" disappeared I don't know.

The reason it disappeared is that it makes gfortran horribly slow when
compiling against mpi. This is not the case with ifc, and therefore the
old ebuild in bugzilla emitted a bold warning when emerging with
USE="-ifc f90-typesafe" but kept quiet if USE="ifc f90-typesave". Thus
it *did make sense* to control it with a USE flag, at least with the
"ifc" USE flag being around also.

> For the gm and gx flags, I don't know, and there is still a nocxx one.

I'm not qualified to talk about these flags as I've never used them.

Matthias
 
Old 12-14-2007, 03:16 PM
Sébastien Fabbro
 
Default gentoo-x86 commit in sys-cluster/openmpi: ChangeLog openmpi-1.1.1.ebuild openmpi-1.2.4.ebuild

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 14/12/07 14:12, Matthias Langer wrote:

> F77="ifort" FC="ifort" FFLAGS="-O3 -xO" emerge -av openmpi

This how it should be. To make it automatically reproducible, specify
environment variables in the configuration files.

> Maybe someone can explain to me what positive side effects the removal
> of the ifc USE flag has - and why this flag is generally discouraged.

Positive side effect: avoid cluttering the tree. Why icc/ifc are
discouraged: you can always try to compile every C/C++ package with
CC=icc and fortran packages with F77=ifort or FC=ifort. Packages which
do specify more options with e.g. --enable-icc and friends can be easily
worked out with the toolchain-funcs and fortran eclass, and most of the
time they do nothing more than specify the environment variables.

If we allow icc/ifc flags, at some point, we could allow a whole bunch
of other compiler flags such as "sunstudio". New keywords for compilers
could be a better idea, but I doubt we have the human resources to test
them.

> The reason it disappeared is that it makes gfortran horribly slow when
> compiling against mpi. This is not the case with ifc, and therefore the
> old ebuild in bugzilla emitted a bold warning when emerging with
> USE="-ifc f90-typesafe" but kept quiet if USE="ifc f90-typesave". Thus
> it *did make sense* to control it with a USE flag, at least with the
> "ifc" USE flag being around also.

If the f90-typesafe options always improve compilation time with
gfortran only, why not using something like this (modified from the
openmpi bump bug):

if use fortran; then
case ${FORTRANC} in
g77) myconf="${myconf} --disable-mpi-f90" ;;
gfortran) myconf="${myconf} --with-mpi-f90-size=medium"
myconf="${myconf} --with-f90-max-array-dim=4" ;;
if*) myconf="${myconf} blah" ;;
*) die "unsupported fortran compiler: ${FORTRANC}"
esac
else
myconf="${myconf} --disable-mpi-f90 --disable-mpi-f77"
fi

Let the ebuild make reasonable choices instead of making a user trying
to find out about undocumented use flags.

- --
Sébastien
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFHYqxv1ycZbhPLE2ARAlRRAJ9BySHhbxAzLOgJdG4I2L 3RpCPPNwCgi8aF
v3OgmxW4UZj1Gqf7Pg2vBWE=
=vYF+
-----END PGP SIGNATURE-----
--
gentoo-dev@gentoo.org mailing list
 

Thread Tools




All times are GMT. The time now is 09:55 PM.

VBulletin, Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO ©2007, Crawlability, Inc.
Copyright ©2007 - 2008, www.linux-archive.org