[hwloc-devel] Create success (hwloc r1.3.3a1r4531)

[MPI Team]

Creating nightly hwloc snapshot SVN tarball was a success.

Snapshot:   hwloc 1.3.3a1r4531
Start time: Tue May 29 21:07:15 EDT 2012
End time:   Tue May 29 21:10:15 EDT 2012

Your friendly daemon,
Cyrador

[hwloc-devel] Create success (hwloc r1.4.3a1r4532)

[MPI Team]

Creating nightly hwloc snapshot SVN tarball was a success.

Snapshot:   hwloc 1.4.3a1r4532
Start time: Tue May 29 21:04:26 EDT 2012
End time:   Tue May 29 21:07:15 EDT 2012

Your friendly daemon,
Cyrador

Re: [OMPI devel] Compile-time MPI_Datatype checking

[Jeff Squyres]

Sweet!

I downloaded clang 3.1 and added it to my nightly regression testing for the 
OMPI SVN trunk and v1.6 branch.

We'd only be able to accept this patch for the SVN trunk (i.e., what will 
become v1.7) -- the v1.6 series is closed for new features.

I'm pretty fried right now (end of a long day); I'll review your patch tomorrow.



On May 29, 2012, at 11:08 AM, Dmitri Gribenko wrote:

> Hello,
> 
> I've implemented a patch for clang that enables compile-time checking
> of  arguments to functions.  When applied to MPI,
> this enables the compiler to check that buffer type and MPI_Datatype
> match.
> 
> Latest version of clang patch can be found here. [1]  Please note that
> clang patch was not yet accepted to clang trunk.
> 
> On the OpenMPI side we need to:
> * add attributes to MPI functions to mark them as accepting pointers
> with type tags;
> * add attributes to ompi_mpi_* declarations to mark them as type tags.
> 
> All in all, the changes boil down to:
> 1. Annotate type tags:
> OMPI_DECLSPEC extern struct ompi_predefined_datatype_t
>ompi_mpi_float OMPI_ATTR_TYPE_TAG(float);
> 
> 2. Annotate functions:
> OMPI_DECLSPEC  int MPI_Send(void *buf, int count, MPI_Datatype
> datatype, int dest,
>int tag, MPI_Comm comm)
>OMPI_ATTR_POINTER_WITH_TYPE_TAG(1,3);
> 
> 
> where OMPI_ATTR* are macros that are defined to attributes when
> compiling under clang with this feature.
> 
> Attached is the OpenMPI patch I've arrived to.  Although I tried to be
> attentive, changes to mpi.h are very repetitive and error-prone, so
> please review them closely.
> 
> I've implemented a similar patch for MPICH2. [2]  OpenMPI developers
> might want to follow that discussion, too.
> 
> Dmitri
> 
> [1] 
> http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20120521/058137.html
> [2] http://lists.mcs.anl.gov/pipermail/mpich2-dev/2012-May/000938.html
> 
> -- 
> main(i,j){for(i=2;;i++){for(j=2;j (j){printf("%d\n",i);}}} /*Dmitri Gribenko */
> ___
> devel mailing list
> de...@open-mpi.org
> http://www.open-mpi.org/mailman/listinfo.cgi/devel


-- 
Jeff Squyres
jsquy...@cisco.com
For corporate legal information go to: 
http://www.cisco.com/web/about/doing_business/legal/cri/

Re: [OMPI devel] Warning in fcoll

[venkates]

I think this is one of those local functions I defined. I wanted to
generalize this for all collective operations, such that its a permanent
fix. I can do this today.

Regards,
Vish
> I'll look into this...
>
> Edgar
>
> On 5/29/2012 3:24 PM, Ralph Castain wrote:
>> Not entirely sure who this might belong to, but thought I should pass it
>> along - seen during an optimized build on Linux:
>>
>> fcoll_static_file_read_all.c: In function
>> ‘mca_fcoll_static_file_read_all’:
>> fcoll_static_file_read_all.c:74: warning: ‘sorted_file_offsets’ may be
>> used uninitialized in this function
>>
>>
>>
>>
>> ___
>> devel mailing list
>> de...@open-mpi.org
>> http://www.open-mpi.org/mailman/listinfo.cgi/devel
>
> --
> Edgar Gabriel
> Associate Professor
> Parallel Software Technologies Lab  http://pstl.cs.uh.edu
> Department of Computer Science  University of Houston
> Philip G. Hoffman Hall, Room 524Houston, TX-77204, USA
> Tel: +1 (713) 743-3857  Fax: +1 (713) 743-3335
>
> ___
> devel mailing list
> de...@open-mpi.org
> http://www.open-mpi.org/mailman/listinfo.cgi/devel

Re: [OMPI devel] Warning in fcoll

[Edgar Gabriel]

I'll look into this...

Edgar

On 5/29/2012 3:24 PM, Ralph Castain wrote:
> Not entirely sure who this might belong to, but thought I should pass it 
> along - seen during an optimized build on Linux:
> 
> fcoll_static_file_read_all.c: In function ‘mca_fcoll_static_file_read_all’:
> fcoll_static_file_read_all.c:74: warning: ‘sorted_file_offsets’ may be used 
> uninitialized in this function
> 
> 
> 
> 
> ___
> devel mailing list
> de...@open-mpi.org
> http://www.open-mpi.org/mailman/listinfo.cgi/devel

-- 
Edgar Gabriel
Associate Professor
Parallel Software Technologies Lab  http://pstl.cs.uh.edu
Department of Computer Science  University of Houston
Philip G. Hoffman Hall, Room 524Houston, TX-77204, USA
Tel: +1 (713) 743-3857  Fax: +1 (713) 743-3335



signature.asc
Description: OpenPGP digital signature

[OMPI devel] Warning in fcoll

[Ralph Castain]

Not entirely sure who this might belong to, but thought I should pass it along 
- seen during an optimized build on Linux:

fcoll_static_file_read_all.c: In function ‘mca_fcoll_static_file_read_all’:
fcoll_static_file_read_all.c:74: warning: ‘sorted_file_offsets’ may be used 
uninitialized in this function

[OMPI devel] OMPI 1.6 affinity fixes: PLEASE TEST

[Jeff Squyres]

Per ticket #3108, there were still some unfortunate bugs in the affinity code 
in 1.6.  :-(

These have now been fixed.  ...but since is the 2nd or 3rd time we have "fixed" 
the 1.5/1.6 series w.r.t. processor affinity, I'd really like people to test 
this stuff before it's committed and we ship 1.6.1.  I've put tarballs 
containing the fixes here:

http://www.open-mpi.org/~jsquyres/unofficial/

Can you please try mpirun options like --bind-to-core and --bind-to-socket and 
ensure that they still work for you?  (even on machines with hyperthreading 
enabled, if you have access to such things)

IBM: I'd particularly like to hear that we haven't made anything worse on POWER 
systems.  Thanks.

-- 
Jeff Squyres
jsquy...@cisco.com
For corporate legal information go to: 
http://www.cisco.com/web/about/doing_business/legal/cri/

Re: [OMPI devel] Open MPI meeting next week

[Josh Hursey]

FYI: I fixed the dates on the locations to match the meeting dates.

On Tue, May 29, 2012 at 11:52 AM, Jeff Squyres  wrote:
> I updated the wiki with the specific Cisco buildings that we'll be in next 
> week (we're in a different building each day):
>
>    https://svn.open-mpi.org/trac/ompi/wiki/June12Meeting
>
> I'll put a sign on the doors of each building with my cell phone number on 
> it, etc.
>
> *** PUT YOUR NAME ON THE WIKI BY 5pm US EASTERN TIME THIS FRIDAY (June 1, 
> 2012) IF YOU PLAN TO ATTEND.
>
> --
> Jeff Squyres
> jsquy...@cisco.com
> For corporate legal information go to: 
> http://www.cisco.com/web/about/doing_business/legal/cri/
>
>
> ___
> devel mailing list
> de...@open-mpi.org
> http://www.open-mpi.org/mailman/listinfo.cgi/devel



-- 
Joshua Hursey
Postdoctoral Research Associate
Oak Ridge National Laboratory
http://users.nccs.gov/~jjhursey

[OMPI devel] Open MPI meeting next week

[Jeff Squyres]

I updated the wiki with the specific Cisco buildings that we'll be in next week 
(we're in a different building each day):

https://svn.open-mpi.org/trac/ompi/wiki/June12Meeting

I'll put a sign on the doors of each building with my cell phone number on it, 
etc.

*** PUT YOUR NAME ON THE WIKI BY 5pm US EASTERN TIME THIS FRIDAY (June 1, 2012) 
IF YOU PLAN TO ATTEND.

-- 
Jeff Squyres
jsquy...@cisco.com
For corporate legal information go to: 
http://www.cisco.com/web/about/doing_business/legal/cri/

[OMPI devel] Compile-time MPI_Datatype checking

[Dmitri Gribenko]

Hello,

I've implemented a patch for clang that enables compile-time checking
of  arguments to functions.  When applied to MPI,
this enables the compiler to check that buffer type and MPI_Datatype
match.

Latest version of clang patch can be found here. [1]  Please note that
clang patch was not yet accepted to clang trunk.

On the OpenMPI side we need to:
* add attributes to MPI functions to mark them as accepting pointers
with type tags;
* add attributes to ompi_mpi_* declarations to mark them as type tags.

All in all, the changes boil down to:
1. Annotate type tags:
OMPI_DECLSPEC extern struct ompi_predefined_datatype_t
ompi_mpi_float OMPI_ATTR_TYPE_TAG(float);

2. Annotate functions:
 OMPI_DECLSPEC  int MPI_Send(void *buf, int count, MPI_Datatype
datatype, int dest,
int tag, MPI_Comm comm)
OMPI_ATTR_POINTER_WITH_TYPE_TAG(1,3);


where OMPI_ATTR* are macros that are defined to attributes when
compiling under clang with this feature.

Attached is the OpenMPI patch I've arrived to.  Although I tried to be
attentive, changes to mpi.h are very repetitive and error-prone, so
please review them closely.

I've implemented a similar patch for MPICH2. [2]  OpenMPI developers
might want to follow that discussion, too.

Dmitri

[1] 
http://lists.cs.uiuc.edu/pipermail/cfe-commits/Week-of-Mon-20120521/058137.html
[2] http://lists.mcs.anl.gov/pipermail/mpich2-dev/2012-May/000938.html

-- 
main(i,j){for(i=2;;i++){for(j=2;j*/


ompi-v2.patch
Description: Binary data

Re: [hwloc-devel] lstopo-gui

[Jeff Squyres]

On May 29, 2012, at 1:48 AM, Brice Goglin wrote:

> I implemented this in the branch lstopo-no-graphics. Additional changes
> to note:
> 1) lstopo-no-graphics is now text-only on Windows too
> 2) when cairo is disabled (and not running on windows), lstopo is a
> symlink on lstopo-no-graphics
> 3) item (2) is only true in the installdir. not sure we should do it in
> the builder too

Thank you!  That sounds ok to me.

-- 
Jeff Squyres
jsquy...@cisco.com
For corporate legal information go to: 
http://www.cisco.com/web/about/doing_business/legal/cri/

[OMPI devel] Migration of SVN/Trac/OpenGrok services

[Jeff Squyres]

Heads up: Indiana University will be migrating Open MPI's Subversion, Trac, and 
OpenGrok services to new servers sometime soon (likely later this week).  The 
purpose of this migration is to retire some (really) old server hardware.  
Delivered performance probably won't be significantly impacted.

I will let everyone know the exact timeframe of the outage and any other 
relevant details when the final time has been nailed down.

-- 
Jeff Squyres
jsquy...@cisco.com
For corporate legal information go to: 
http://www.cisco.com/web/about/doing_business/legal/cri/

Re: [hwloc-devel] lstopo-gui

[Brice Goglin]

Le 25/05/2012 12:34, Jeff Squyres a écrit :
> Hmm. I typed "lstopo-no-graphics" above, just to be descriptive, but
> is that a horrible name? If the main goal is for binary packagers who
> assumedly have /etc/alternative-type solutions such that users will
> rarely/never type that full name, how about just being descriptive
> with a lengthy suffix like that? 

I implemented this in the branch lstopo-no-graphics. Additional changes
to note:
1) lstopo-no-graphics is now text-only on Windows too
2) when cairo is disabled (and not running on windows), lstopo is a
symlink on lstopo-no-graphics
3) item (2) is only true in the installdir. not sure we should do it in
the builddir too

If people are ok with this renaming, I'll merge the branch in the next
days. And I think we'll start talking about doing v1.5 since most other
branches aren't ready and/or need plugins before they can get merged.

Brice