[hwloc-devel] Create success (hwloc r1.1a1r1980)

2010-04-20 Thread MPI Team
Creating nightly hwloc snapshot SVN tarball was a success. Snapshot: hwloc 1.1a1r1980 Start time: Tue Apr 20 21:01:02 EDT 2010 End time: Tue Apr 20 21:03:00 EDT 2010 Your friendly daemon, Cyrador

Re: [hwloc-devel] [hwloc-svn] svn:hwloc r1976

2010-04-20 Thread Brice Goglin
Bert Wesarg wrote: > On Tue, Apr 20, 2010 at 21:45, wrote: > >> Author: bgoglin >> Date: 2010-04-20 15:45:15 EDT (Tue, 20 Apr 2010) >> New Revision: 1976 >> URL: https://svn.open-mpi.org/trac/hwloc/changeset/1976 >> >> Log: >> Backport trunk r1968-1970: >> * hwloc-calc: Add --single to singlif

Re: [hwloc-devel] [hwloc-svn] svn:hwloc r1976

2010-04-20 Thread Bert Wesarg
On Tue, Apr 20, 2010 at 21:45, wrote: > Author: bgoglin > Date: 2010-04-20 15:45:15 EDT (Tue, 20 Apr 2010) > New Revision: 1976 > URL: https://svn.open-mpi.org/trac/hwloc/changeset/1976 > > Log: > Backport trunk r1968-1970: > * hwloc-calc: Add --single to singlify the output > * hwloc-distrib: Ad

Re: [hwloc-devel] backward API compat or not?

2010-04-20 Thread Samuel Thibault
Brice Goglin, le Tue 20 Apr 2010 19:54:13 +0200, a écrit : > I tend to think that we should just drop the get_system_obj() and make > it clear that people have to fix everything, not everything except foo > and bar. I'd tend to agree. We should however mention the previous names in the documentat

[hwloc-devel] backward API compat or not?

2010-04-20 Thread Brice Goglin
We have a static inline in hwloc/helper.h to replace the old get_system_obj() into the new get_root_obj(). But there are other API changes in 1.0 which have no similar compat wrapper. For instance replacing HWLOC_OBJ_PU instead of HWLOC_OBJ_PROC probably broke every hwloc user on earth. So I tend t

Re: [hwloc-devel] dplace

2010-04-20 Thread Michael Raymond
dplace is a general purpose tool for placing applications and doesn't know anything about MPI. It does do internal accounting system wide to spread and pin processes and threads. For example if you run a job with dplace and tell it to just pin your job anywhere it can and another user does the

Re: [hwloc-devel] [hwloc-svn] svn:hwloc r1961

2010-04-20 Thread Samuel Thibault
Bert Wesarg, le Tue 20 Apr 2010 17:55:10 +0200, a écrit : > > +#define GNU_SOURCE > > That should be: _GNU_SOURCE. Oops, typo indeed, thanks for the proofread. Samuel

Re: [hwloc-devel] dplace

2010-04-20 Thread Brice Goglin
Does dplace do something clever like reading the MPI rank and communicator size and try to figure out how to distribute among the cores so as to maximize memory bandwidth or cache sharing ? Brice Michael Raymond wrote: > As of SGI ProPack 7 dplace uses hwloc internally to specify stride > pat

Re: [hwloc-devel] [hwloc-svn] svn:hwloc r1961

2010-04-20 Thread Bert Wesarg
Hi, On Tue, Apr 20, 2010 at 14:33, wrote: > Author: sthibaul > Date: 2010-04-20 08:33:33 EDT (Tue, 20 Apr 2010) > New Revision: 1961 > URL: https://svn.open-mpi.org/trac/hwloc/changeset/1961 > > Log: > properly include sched.h to get the sched_setaffinity prototype > Text files modified: >   tru

Re: [hwloc-devel] dplace

2010-04-20 Thread Michael Raymond
As of SGI ProPack 7 dplace uses hwloc internally to specify stride patterns. For example: mpirun -np 8 dplace -c SC a.out means to pin ranks to every core inside a socket before jumping to the next socket and doing the same. From the man page: For striding patterns any

[hwloc-devel] dplace

2010-04-20 Thread Brice Goglin
I discovered "dplace" today. I don't know how many people install/use it on their cluster, but it's something that looks interesting when you don't have advanced binding capabilities in the MPI implementation. For instance, you could do: $ mpirun -np 8 dplace 0,4,2,6,1,5,3,7 myprogram to bind pro

Re: [hwloc-devel] [hwloc-svn] svn:hwloc r1960

2010-04-20 Thread Jeff Squyres
On Apr 20, 2010, at 9:13 AM, Samuel Thibault wrote: > Jeff Squyres, le Tue 20 Apr 2010 09:09:15 -0400, a écrit : > > while ("$r" != "") > > echo merging in r$r... > > > > set r_minus_one=`expr $r - 1` > > set str="svn merge -r ${r_minus_one}:$r $otrunk ." > > You can replace this wi

Re: [hwloc-devel] [hwloc-svn] svn:hwloc r1960

2010-04-20 Thread Brice Goglin
I do git cherry-pick -e master and git svn dcommit :) Brice Jeff Squyres wrote: > Done. BTW, I use the attached helper script to merge stuff in to the > branches (can't remember if I've sent this before or not). It's a trivial > thing, but it reduces the amount of stuff I have to type on t

Re: [hwloc-devel] [hwloc-svn] svn:hwloc r1960

2010-04-20 Thread Samuel Thibault
Jeff Squyres, le Tue 20 Apr 2010 09:09:15 -0400, a écrit : > while ("$r" != "") > echo merging in r$r... > set r_minus_one=`expr $r - 1` > set str="svn merge -r ${r_minus_one}:$r $otrunk ." You can replace this with svn merge -c $r $otrunk . Samuel

Re: [hwloc-devel] [hwloc-svn] svn:hwloc r1960

2010-04-20 Thread Jeff Squyres
Done. BTW, I use the attached helper script to merge stuff in to the branches (can't remember if I've sent this before or not). It's a trivial thing, but it reduces the amount of stuff I have to type on the command line. On Apr 20, 2010, at 8:49 AM, Jeff Squyres (jsquyres) wrote: > Cool; I'

Re: [hwloc-devel] [hwloc-svn] svn:hwloc r1960

2010-04-20 Thread Jeff Squyres
Cool; I'll have a look right now. On Apr 20, 2010, at 8:43 AM, Brice Goglin wrote: > This needs to be backported to the 1.0 branch, but somebody might want > to fix/improve the text, so I am waiting a bit. > > Brice > > > > > bgog...@osl.iu.edu wrote: > > Author: bgoglin > > Date: 2010-04-20

Re: [hwloc-devel] [hwloc-svn] svn:hwloc r1960

2010-04-20 Thread Brice Goglin
This needs to be backported to the 1.0 branch, but somebody might want to fix/improve the text, so I am waiting a bit. Brice bgog...@osl.iu.edu wrote: > Author: bgoglin > Date: 2010-04-20 04:06:03 EDT (Tue, 20 Apr 2010) > New Revision: 1960 > URL: https://svn.open-mpi.org/trac/hwloc/changeset/