ming you are using a bash shell, you can simply do
CPPFLAGS="" configure ...
instead of
configure ... CPPFLAGS=""
Cheers,
Gilles
On 10/7/2015 4:42 PM, Siegmar Gross wrote:
Hi,
I tried to build openmpi-v2.x-dev-415-g5c9b192 and
openmpi-dev-2696-gd579a07 on my machines (Solaris 1
such as clustershell. it is written in python,
and an api is likely available.
Cheers,
Gilles
On Thursday, October 8, 2015, simona bellavista > wrote:
>
>
> 2015-10-07 14:59 GMT+02:00 Lisandro Dalcin :
>
>> On 7 October 2015 at 14:54, simona bellavista wrote:
>> >
Georg,
there won't be a 1.8.9
Cheers,
Gilles
On Wednesday, October 7, 2015, Georg Geiser wrote:
> Nathan,
>
> thanks for your rapid response. Do you consider to release 1.8.9?
> Actually, there is a bug tracking category for that version number. If so,
> please backpo
Jeff,
there are quite a lot of changes, I did not update master yet (need extra
pairs of eyes to review this...)
so unless you want to make rc2 today and rc3 a week later, it is imho way
safer to wait for v1.10.2
Ralph,
any thoughts ?
Cheers,
Gilles
On Wednesday, October 7, 2015, Jeff Squyres
Marcin,
here is a patch for the master, hopefully it fixes all the issues we
discussed
i will make sure it applies fine vs latest 1.10 tarball from tomorrow
Cheers,
Gilles
On 10/6/2015 7:22 PM, marcin.krotkiewski wrote:
Gilles,
Yes, it seemed that all was fine with binding in the patched
Hector,
numprocs and .false. are scalars and MPI_Cart_create expects one dimension
array.
can you fix this and try again ?
Cheers,
Gilles
On Wednesday, October 7, 2015, Hector E Barrios Molano
wrote:
> Hi Open MPI Experts!
>
> I'm using OpenMPI v1.10.0 and get this er
the required mapping policy
I will finalize them tomorrow hopefully
Cheers,
Gilles
On Tuesday, October 6, 2015, marcin.krotkiewski <
marcin.krotkiew...@gmail.com> wrote:
> Hi, Gilles
>
> you mentionned you had one failure with 1.10.1rc1 and -bind-to core
> could you please sen
--ntasks=2 --cpus-per-task=4 --cpu_bind=core,verbose -l grep
Cpus_allowed_list /proc/self/status
Cheers,
Gilles
On 10/6/2015 4:38 AM, marcin.krotkiewski wrote:
Yet another question about cpu binding under SLURM environment..
Short version: will OpenMPI support SLURM_CPUS_PER_TASK for the
,
Gilles
On Monday, October 5, 2015, marcin.krotkiewski
wrote:
>
> I have applied the patch to both 1.10.0 and 1.10.1rc1. For 1.10.0 it did
> not help - I am not sure how much (if) you want pursue this.
>
> For 1.10.1rc1 I was so far unable to reproduce any binding problems with
Ralph and Marcin,
here is a proof of concept for a fix (assert should be replaced with
proper error handling)
for v1.10 branch.
if you have any chance to test it, please let me know the results
Cheers,
Gilles
On 10/5/2015 1:08 PM, Gilles Gouaillardet wrote:
OK, i'll see what i c
OK, i'll see what i can do :-)
On 10/5/2015 12:39 PM, Ralph Castain wrote:
I would consider that a bug, myself - if there is some resource
available, we should use it
On Oct 4, 2015, at 5:42 PM, Gilles Gouaillardet <mailto:gil...@rist.or.jp>> wrote:
Marcin,
i ran a si
2 bound to socket 0[core 0[hwt 0-1]]:
[BB/../../../../../../..][../../../../../../../..]
Hello, world, I am 1 of 3, (Open MPI v1.10.1rc1, package: Open MPI
gilles@rapid Distribution, ident: 1.10.1rc1, repo rev:
v1.10.0-84-g15ae63f, Oct 03, 2015, 128)
Hello, world, I am 2 of 3, (Open MPI v1.10.1rc
investigate this from tomorrow
Cheers,
Gilles
On Sunday, October 4, 2015, Ralph Castain wrote:
> Thanks - please go ahead and release that allocation as I’m not going to
> get to this immediately. I’ve got several hot irons in the fire right now,
> and I’m not sure when I’ll get a chance
--oversubscribe --hetero-nodes, mpirun should not fail, and if it
still fails with v1.10.1rc1, I will ask some more details in order to fix
ompi
Cheers,
Gilles
On Saturday, October 3, 2015, Ralph Castain wrote:
> Thanks Marcin. Looking at this, I’m guessing that Slurm may be treating
>
and irecv with the second element, and then waitall the array
of size 2
note this is not equivalent to doing two MPI_Wait in a row, since that
would be prone to deadlock
Cheers,
Gilles
On Wednesday, September 30, 2015, Diego Avesani
wrote:
> Dear all,
> thank for the explanation, but som
MPi_Comm_rank, that makes no sense to me, and your
indexes should likely be hard coded from 1 to 3
Cheers,
Gilles
On Wednesday, September 30, 2015, Diego Avesani > wrote:
> Dear Gilles, Dear All,
>
> What do you mean that the array of requests has to be initialize via
> MPI_Isend or MPI_I
actually used, then you have to initialize
them with MPI_REQUEST_NULL
(it might be zero on ompi, but you cannot take this for granted)
Cheers,
Gilles
On Tuesday, September 29, 2015, Diego Avesani
wrote:
> dear Jeff, dear all,
> I have notice that if I initialize the variables, I do not ha
Harald,
thanks for the clarification, i clearly missed that !
i will fix it from now
Cheers,
Gilles
On 9/28/2015 4:49 PM, Harald Servat wrote:
Hello Gilles,
the webpages I pointed in the original mail and which are the
official open-mpi.org, miss the * in the declaration of MPI_Ibarrier
Harald,
could you be more specific ?
btw, do you check the www.open-mpi.org main site or a mirror ?
the man pages looks good to me, and the issue you described was fixed
one month ago.
Cheers,
Gilles
On 9/25/2015 8:07 PM, Harald Servat wrote:
Dear all,
I'd like to note you tha
Jeff,
I am not sure whether you made a typo or not ...
the issue only occuex with f90 bindings (aka use mpi)
f08 bindings (aka use mpi_f08) works fine
Cheers,
Gilles
On Thursday, September 24, 2015, Jeff Squyres (jsquyres)
wrote:
> I looked into the MPI_BCAST problem -- I think we (Open
int root,
MPI_Comm comm)
so you have
recvbuf = 0 (!)
recvcount = tmpVar[i*matrixSize]
i guess you meant to have recvcount = blockSize
that being said, tmpVar[i*matrixSize] is and int and it should likely be
an double *
Cheers,
Gilles
On 9/24/2015 8:13 AM, Surivinta Surivinta wrote:
program compiles and runs just
fine if you use mpi_f08 module (!)
Cheers,
Gilles
On 9/24/2015 1:00 AM, Fabrice Roy wrote:
program testmpi
use mpi
implicit none
integer :: pid
integer :: ierr
integer :: tok
call mpi_init(ierr)
call mpi_comm_rank(mpi_comm_world
remote start orted)
instead of this silent behaviour
Cheers,
Gilles
On Mon, Sep 21, 2015 at 11:43 PM, Patrick Begou
wrote:
> Hi Gilles,
>
> I've done a big mistake! Compiling the patched version of openMPI and
> creating a new module, I've forgotten to add the path to oar
r, there is a reachable framework. Could/should the tcp btl simply
use it ?
Cheers,
Gilles
On Saturday, September 19, 2015, Jeff Squyres (jsquyres)
wrote:
> Open MPI uses different heuristics depending on whether IP addresses are
> public or private.
>
> All your IP addresses are
you use a machine file (frog.txt) instead
of using $OAR_NODEFILE directly ?
/* not to mention I am surprised a French supercomputer is called "frog"
;-) */
Cheers,
Gilles
On Friday, September 18, 2015, Patrick Begou <
patrick.be...@legi.grenoble-inp.fr> wrote:
> Gilles
of ssh)
my concern is the remote orted might not run within the cpuset that was
created by OAR for this job,
so you might end up using all the cores on the remote nodes.
please let us know how that works for you
Cheers,
Gilles
On 9/18/2015 5:02 PM, Gilles Gouaillardet wrote:
Patrick,
i
Patrick,
i just filled PR 586 https://github.com/open-mpi/ompi-release/pull/586
for the v1.10 series
this is only a three line patch.
could you please give it a try ?
Cheers,
Gilles
On 9/18/2015 4:54 PM, Patrick Begou wrote:
Ralph Castain wrote:
As I said, if you don’t provide an explicit
same network
for example, on node 1, you can run
route add -host 23.0.0.2 gw 12.0.0.2
route add -host 23.0.0.3 gw 13.0.0.3
Cheers,
Gilles
On 9/18/2015 1:31 AM, Shang Li wrote:
Hi all,
I wanted to setup a 3-node ring network, each connects to the other 2
using 2 Ethernet ports directly without a
invokation is
correct ?
an other way to fix this could be to always set opal_hwloc_base_cpu_set
Cheers,
Gilles
On 9/16/2015 11:47 PM, Ralph Castain wrote:
As I said, if you don’t provide an explicit slot count in your
hostfile, we default to allowing oversubscription. We don’t have OAR
integration
pi or hwloc, or ompi should ask
the correct info to hwloc if it is available in hwloc.
makes sense ?
Brice, can you please comment on hwloc and cpuset ?
Cheers,
Gilles
On Wednesday, September 16, 2015, Ralph Castain wrote:
> Not precisely correct. It depends on the environment.
>
--disable-shared
that being said, i am not sure you do need --enable-static.
are mxm/fca/hcoll in your $LD_LIBRARY_PATH or ld.so.conf ?
if not, i recommend you set your LD_LIBRARY_PATH *before* you
configure and make openmpi
Cheers,
Gilles
On Tue, Sep 15, 2015 at 8:14 AM, Filippo Spiga wrote
please use mpijavac instead of javac
this will automagically set the classpath with the ompi java libraries.
if there is no javac, it is likely you did not configure ompi with
--enable-mpi-java
On Monday, September 14, 2015, Ibrahim Ikhlawi
wrote:
>
> Hi,
>
> I am beginner in OpenMPI. I want to
on linux, you can look at /proc/self/status and search allowed_cpus_list
or you can use the sched_getaffinity system call
note that in some (hopefully rare)cases, this will return different results
than hwloc
On Sunday, September 13, 2015, Saliya Ekanayake wrote:
> Thank you, I'll try this. Als
Thanks Siegmar,
at first glance, I suspect String.valueOf(buffer)
buffer is 256 chars, but the message you really want to print is only the
first num chars.
I will double check tomorrow, in the mean time, feel free to revamp the
test and see if it works better
Cheers,
Gilles
On Tuesday
Siegmar,
can you post your test program ?
did you try to run the very same test with ompi configure'd without
--enable-heterogeneous ?
did this help ?
can you reproduce the crash with the v2.x series ?
Cheers,
Gilles
On Tuesday, September 8, 2015, Siegmar Gross <
siegmar.gr...@infor
Diego,
did you update your code to check all MPI calls are successful ?
(e.g. test ierr is MPI_SUCCES after each MPI call)
can you write a short program that reproduce the same issue ?
if not, is your program and input data public ally available ?
Cheers,
Gilles
On Thursday, September 3, 2015
x intelmpi and openmpi, or use openmpi with a lib built with intelmpi.
Cheers,
Gilles
On Thursday, September 3, 2015, Diego Avesani
wrote:
> Dear Jeff, Dear all,
> I normaly use "USE MPI"
>
> This is the answar fro intel HPC forum:
>
> *If you are switching between intel an
recommend you first do this, so you can catch the error as soon it
happens, and hopefully understand why it occurs．
Cheers,
Gilles
On Wednesday, September 2, 2015, Diego Avesani
wrote:
> Dear all,
>
> I have notice small difference between OPEN-MPI and intel MPI.
> For example in M
are used in production.
Cheers,
Gilles
On 9/1/2015 1:57 PM, Saliya Ekanayake wrote:
One more question. I found this blog from Jeff [1] on vader and I got
the impression that it's used only for peer-to-peer communications and
not for collectives. Is this true or did I misunderstand?
[1]
understand what
the vader btl can do and how
Cheers,
Gilles
On 9/1/2015 1:28 PM, Saliya Ekanayake wrote:
Thank you Gilles. Is there some documentation on vader btl and how I
can check which (sm or vader) is being used?
On Tue, Sep 1, 2015 at 12:18 AM, Gilles Gouaillardet
mailto:gil
more optimized configurations.
Cheers,
Gilles
On 9/1/2015 5:59 AM, Saliya Ekanayake wrote:
Hi,
Just trying to see if there are any optimizations (or options) in
OpenMPI to improve communication between intra node processes. For
example do they use something like shared memory?
Thank you
Roy,
you can create your type without MPI_UB nor MPI_LB, and then use
MPI_Type_create_resized to set lower bound and extent
(note this sets extent and not upper bound)
Cheers,
Gilles
On Saturday, August 29, 2015, Roy Stogner wrote:
>
> From: George Bosilca
>>
>> First an
tweaking with hostnames, so
ssh node0198 ...
really do
ssh node0198.mako0 ...
under the hood
Cheers,
Gilles
On 8/27/2015 8:08 AM, Yong Qin wrote:
Yes all cross-node ssh works perfectly and this is our production
system which have been running for years. I've done all of these
testing an
both work ?
per your log, mpirun might remove the domain name from the ssh command
under the hood.
e.g.
ssh n0189.mako0 ssh n0198 echo ok
or
ssh n0198 ssh n0198.mako0 echo ok
if that is the case, then I have no idea why we are doing this ...
Cheers,
Gilles
On Thursday, August 27, 2015, Yong
here.
do you configure with --disable-dlopen on hopper ?
I wonder whether --mca mtl ^psm is effective if dlopen is disabled
Cheers,
Gilles
On Saturday, August 15, 2015, Howard Pritchard wrote:
> Hi Jeff,
>
> I don't know why Gilles keeps picking on the persistent request proble
data/
please run
mpirun --mca mtl ^psm -np 1 java MPITestBroke data/
that solved the issue for me
Cheers,
Gilles
On 8/13/2015 9:19 AM, Nate Chambers wrote:
*I appreciate you trying to help! I put the Java and its compiled
.class file on Dropbox. The directory contains the .java and .class
files
David,
i guess you do not want to use the ml coll module at all in openmpi 1.8.8
you can simply do
touch ompi/mca/coll/ml/.ompi_ignore
./autogen.pl
./configure ...
make && make install
so the ml component is not even built
Cheers,
Gilles
On 8/13/2015 7:30 AM, David Shrader
irectory ?
Cheers,
Gilles
On 8/13/2015 2:29 PM, Ehsan Moradi wrote:
hi,
my dear friends
i tried to upgrade my openmpi version from 1.2.8 to 1.8.8
but after installing it on different directory "/opt/openmpi-1.8.8/"
when i enter mpirun its version is 1.2.8
and after installing t
checked and ompi
transparently fall back to --heyero-nodes if needed
bottom line, on a heterogeneous cluster, it is required or safer to use the
--hetero-nodes option
Cheers,
Gilles
On Wednesday, August 12, 2015, Dave Love wrote:
> "Lane, William" > writes:
>
> >
Thanks David,
i made a PR for the v1.8 branch at
https://github.com/open-mpi/ompi-release/pull/492
the patch is attached (it required some back-porting)
Cheers,
Gilles
On 8/12/2015 4:01 AM, David Shrader wrote:
I have cloned Gilles' topic/hcoll_config branch and, after running
autog
nvoke
autogen.pl)
Cheers,
Gilles
On 8/11/2015 2:39 PM, Åke Sandgren wrote:
Please fix the hcoll test (and code) to be correct.
Any configure test that adds /usr/lib and/or /usr/include to any
compile flags is broken.
And if hcoll include files are under $HCOLL_HOME/include/hcoll (and
hcol
evamp the hcoll detection (e.g.
configure --with-hcoll) but you might
need to manually set CPPFLAGS='-I/usr/include/hcoll
-I/usr/include/hcoll/api'
if not, i guess i will simply update the configure help message ...
Cheers,
Gilles
On 8/11/2015 7:39 AM, David Shrader wrote:
H
on your report, we might recommend some tuning for the tuned module
(as you can guess, the basic coll module is not optimized)
Cheers,
Gilles
On Saturday, August 8, 2015, Ralph Castain wrote:
> My first suggestion would be to try using 1.8.8 instead to get all the bug
> fixes since 1.8
, since
i am not 100% convinced
the test is bulletproof :
if ofed version is different between compilation and runtime, let
openmpi crash
since this is a user-side problem.
Cheers,
Gilles
On 8/6/2015 8:43 AM, Ralph Castain wrote:
Yeah, I recall your earlier email on the subject. Sadly, I need
Nate,
a similar issue has already been reported at
https://github.com/open-mpi/ompi/issues/369, but we have
not yet been able to figure out what is going wrong.
right after MPI_Init(), can you add
Thread.sleep(5000);
and see if it helps ?
Cheers,
Gilles
On 8/4/2015 8:36 AM, Nate Chambers
That is a good sign, it means orted was started on both nodes
strictly speaking, you should confirm both nodes appear 16 times each in
the output,
do you can draw any firm conclusion
Cheers,
Gilles
On Monday, August 3, 2015, abhisek Mondal wrote:
> I wrote 2 new node names(which I had
simply replace nwchem with hostname
both hosts should be part of the output...
Cheers,
Gilles
On Sunday, August 2, 2015, abhisek Mondal wrote:
> Jeff, Gilles
>
> Here's my scenario again when I tried something different:
> I've interactively booked 2 nodes(cx1015 and
application
/.../bin/mpirun hostname
and then nwchem
if it still does not work, then run with verbose palm and post the output
Cheets,
Gilles
On Sunday, August 2, 2015, abhisek Mondal wrote:
> I'm on a HPC cluster. So, the openmpi-1.6.4 here installed as a module.
> In .pbs scr
wrong
last but not least, can you try to use full path for both mpirun and nwchem
?
Cheers,
Gilles
On Sunday, August 2, 2015, abhisek Mondal wrote:
> Yes, I have tried this and got following error:
>
> *mpirun was unable to launch the specified application as it could not
> find a
Can you try running invoking mpirun with its full path instead ?
e.g. /usr/local/bin/mpirun instead of mpirun
Cheers,
Gilles
On Sunday, August 2, 2015, abhisek Mondal wrote:
> Here is the other details,
>
> a. The Openmpi version is 1.6.4
>
> b. The error as being generated
Thomas,
can you please elaborate ?
I checked the code of opal_os_dirpath_create and could not find where such
a thing can happen
Thanks,
Gilles
On Wednesday, July 29, 2015, Thomas Jahns wrote:
> Hello,
>
> On 07/28/15 17:34, Schlottke-Lakemper, Michael wrote:
>
>> That’
Erik,
the OS X warning (which should not be OS X specific) is fixed in
https://github.com/open-mpi/ompi-release/pull/430
it will land into the v2.x series once reviewed
in the mean time, feel free to manually apply the patch on the tarball
Cheers,
Gilles
On 7/29/2015 10:35 AM, Erik
You are right and I misread your comment.
Michael is using ROMIO, which is independent of ompio.
Cheers,
Gilles
On Wednesday, July 29, 2015, Dave Love wrote:
> Gilles Gouaillardet > writes:
>
> > Dave,
> >
> > On 7/24/2015 1:53 AM, Dave Love wrote:
>
thanks for clarifying there is only one container per host.
do you always run 16 tasks per host/container ?
or do you always run 16 hosts/containers ?
also, do lxc sets iptables when you start a container ?
Cheers,
Gilles
On Tuesday, July 28, 2015, Cristian RUIZ wrote:
> Thank you
Cristian,
one more thing...
make sure tasks run on the same physical node with and without containers.
for example, if in native mode, tasks 0 to 15 run on node 0, then in
container mode, tasks 0 to 15 should run on 16 containers hosted by node 0
Cheers,
Gilles
On Tuesday, July 28, 2015
viour.
on the network, did you measure zero or few errors ?
few errors take some extra time to be fixed, and if your application is
communication intensive, these delays get propagated and you can end up
with huge performance hit.
Cheers,
Gilles
On Tuesday, July 28, 2015, Cristian RUIZ
d is mpifort ...
so you can run
mpifort -showme ...
to see the how gfortran is invoked.
it is likely mpifort simply run gfortran, and your PATH does not point
to gfortran 4.9.2
Cheers,
Gilles
On 7/28/2015 1:47 PM, Syed Ahsan Ali wrote:
I am getting this error during installation of an applic
have an OS X cluster ...)
Cheers,
Gilles
On Sunday, July 26, 2015, Erik Schnetter wrote:
> Mark
>
> No, it doesn't need to be 1.8.7.
>
> I just tried v2.x-dev-96-g918650a. This leads to run-time warnings on OS
> X; I see messages such as
>
> [warn] select: Bad file
.
Cheers,
Gilles
On Friday, July 24, 2015, Harald Servat wrote:
> Dear Cristian,
>
> according to your configuration:
>
> a) - 8 Linux containers on the same machine configured with 2 cores
> b) - 8 physical machines
> c) - 1 physical machine
>
> a) and c) hav
pport.)
on my system :
$ grep FILE_SYSTEM ./ompi/mca/io/romio/romio/config.status
S["FILE_SYSTEM"]="testfs ufs nfs"
unless i am misunderstanding, nfs is there
Cheers,
Gilles
Michael,
ROMIO is the default in the 1.8 series
you can run
ompi_info --all | grep io | grep priority
ROMIO priority should be 20 and ompio priority should be 10.
Cheers,
Gilles
On Thursday, July 23, 2015, Schlottke-Lakemper, Michael <
m.schlottke-lakem...@aia.rwth-aachen.de> wrote:
Michael,
are you running 1.8.7 or master ?
if not default, which io module are you running ?
(default is ROMIO with 1.8 but ompio with master)
by any chance, could you post a simple program that evidences this issue ?
Cheers,
Gilles
On Thursday, July 23, 2015, Schlottke-Lakemper, Michael
threads
per tasks
Cheers,
Gilles
On 7/22/2015 4:42 PM, Crisitan RUIZ wrote:
Sorry, I've just discovered that I was using the wrong command to run
on 8 machines. I have to get rid of the "-np 8"
So, I corrected the command and I used:
mpirun --machinefile machine_mpi_bug.txt --mca
Bill,
good news : 1.8.7 has been released yesterday
Cheers,
Gilles
On 7/16/2015 1:53 PM, Lane, William wrote:
Ralph,
I'd rather wait for the stable release of 1.8.7, but I'm willing to give
it a try if my supervisor is
collective modules (hierarch, ml, ?) implement
hierarchical collective, which means they should be optimized for multi
node / multi tasks per node.
that being said, ml is not production ready, and i am not sure wheter
hierarch is actively maintained)
i hope this helps
Gilles
On 7/9/2015 5:37
George,
I got confused since from bitbucket, ulfm was imported from svn r25756 and
this svn id comes from the v1.7 branch
Also, orte_job_map_t has a policy field in v1.6 but no more in v1.7
Anyway, you obviously know ulfm better than me, so I was wrong
Cheers,
Gilles
On Monday, July 6, 2015
Rafael,
At first glance, ulfm is based on openmpi 1.7
But the poe plm was written for openmpi 1.6
You'd better ask the ulfm folks about this issue
Cheers,
Gilles
On Monday, July 6, 2015, Rafael Lago wrote:
> Hello there!
> I'm trying to work in a fault-tolerance project, a
Nathan,
Shall I remove the --with-threads configure option ?
or make it dummy ?
Cheers,
Gilles
On Tuesday, June 30, 2015, Nathan Hjelm wrote:
>
> Ah, that would explain why I am not seeing it in master. Can you PR the
> changes to v1.10?
>
> -Nathan
>
> On Tue, Jun 3
Nathan,
I removed all of this (including the --with-threads configure option) on
master a while ago.
because this is a change in the configure command line, I never made a PR
for v1.8
Cheers,
Gilles
On Tuesday, June 30, 2015, Nathan Hjelm wrote:
>
> Open MPI has required posix threa
file to blacklist the coll_ml module to
ensure this is working.
Mike and Mellanox folks, could you please comment on that ?
Cheers,
Gilles
On 6/24/2015 5:23 PM, Daniel Letai wrote:
Gilles,
Attached the two output logs.
Thanks,
Daniel
On 06/22/2015 08:08 AM, Gilles Gouaillardet wrote
do not see how I can tackle the root cause without being able to
reproduce the issue :-(
can you try to reproduce the issue with the smallest hostfile, and then run
lstopo on all the nodes ?
btw, you are not mixing 32 bits and 64 bits OS, are you ?
Cheers,
Gilles
mca_btl_sm_add_procs(
int
Jeff,
it sounds like openmpI is not available on sone nodes !
an other possibility is it is installed but in an other directory
or mabye it is not in your path and you did not configure with
--enable-mpirun-prefix-by-default
Cheers,
Gilles
On Wednesday, June 24, 2015, Jeff Layton wrote
Daniel,
i double checked this and i cannot make any sense with these logs.
if coll_ml_priority is zero, then i do not any way how
ml_coll_hier_barrier_setup can be invoked.
could you please run again with --mca coll_base_verbose 100
with and without --mca coll ^ml
Cheers,
Gilles
On 6/22
Daniel,
ok, thanks
it seems that even if priority is zero, some code gets executed
I will confirm this tomorrow and send you a patch to work around the issue
if that if my guess is proven right
Cheers,
Gilles
On Sunday, June 21, 2015, Daniel Letai wrote:
> MCA coll: parame
1.8, so an
incorrect back port could be the root cause.
Cheers,
Gilles
On Friday, June 19, 2015, Ralph Castain wrote:
> Gilles
>
> I was fooled too, but that isn’t the issue. The problem is that
> ompi_free_list is segfaulting:
>
> [csclprd3-0-13:30901] *** Process received signal
?
Cheers,
Gilles
On Friday, June 19, 2015, Lane, William wrote:
> Ralph,
>
> I created a hostfile that just has the names of the hosts while
> specifying no slot information whatsoever (e.g. csclprd3-0-0)
> and received the following errors:
>
> mpirun -np 132 -report-b
This is really odd...
you can run
ompi_info --all
and search coll_ml_priority
it will display the current value and the origin
(e.g. default, system wide config, user config, cli, environment variable)
Cheers,
Gilles
On Thursday, June 18, 2015, Daniel Letai wrote:
> No, that's t
Daniel,
ML module is not ready for production and is disabled by default.
Did you explicitly enable this module ?
If yes, I encourage you to disable it
Cheers,
Gilles
On Thursday, June 18, 2015, Daniel Letai wrote:
> given a simple hello.c:
>
> #include
> #include
>
>
Siegmar,
these are just warnings, you can safely ignore them
Cheers,
Gilles
On Tuesday, June 16, 2015, Siegmar Gross <
siegmar.gr...@informatik.hs-fulda.de> wrote:
> Hi,
>
> today I tried to build openmpi-dev-1883-g7cce015 on my machines
> (Solaris 10 Sparc, Solaris 10 x8
Dave,
v1.6 is ok if configure'd with --with-devel-headers
Cheers,
Gilles
On Monday, June 15, 2015, Dave Love wrote:
> Gilles Gouaillardet > writes:
>
> > Dave,
> >
> > commit
> >
> https://github.com/nerscadmin/IPM/commit/8f628dadc502b3e0113d6ab307
* handle MPI_IN_PLACE in
Fortran with Open MPI > 1.6
Cheers,
Gilles
On 6/12/2015 1:09 AM, Dave Love wrote:
Filippo Spiga writes:
Dear OpenMPI experts,
I am rebuilding IPM (https://github.com/nerscadmin/ipm) based on OpenMPI 1.8.5. However, despite
OMPI is compiled with the "--wi
Ray,
one possibility is one of the loaded library was built with -rpath and this
causes the mess
an other option is you have to link _error.so with libmpi.so
Cheers,
Gilles
On Friday, June 12, 2015, Ray Sheppard wrote:
> Hi Gilles,
> Thanks for the reply. I completely forgot that
Ray,
this symbol is defined in libmpi.so.
can you run
ldd
/N/dc2/projects/ray/quarry/h5py/h5py-2.5.0/build/lib.linux-x86_64-2.7/h5py/_errors.so
and make sure this is linked with openmpi 1.8.4 ?
Cheers,
Gilles
On 6/12/2015 1:29 AM, Ray Sheppard wrote:
Hi List,
I know I saw this issue
On Jun 10, 2015, at 12:00 AM, Gilles Gouaillardet > wrote:
> >
> > that can happen indeed, in a complex but legitimate environment :
> >
> > mkdir ~/src
> > cd ~/src
> > tar xvfj openmpi-1.8.tar.bz2
> > mkdir ~/build/openmpi-v1.8
> > cd ~/build
Hi,
you can find some source rpms for various distros
for example at http://rpms.famillecollet.com/rpmphp/zoom.php?rpm=lam
Cheers,
Gilles
On 6/10/2015 6:07 PM, Cian Davis wrote:
Hi All,
While OpenMPI is the way forward, there is a fair amount of software
out there still compiled against
rm -rf openmpi-1.8.6-${SYSTEM_ENV}.${MACHINE_ENV}.64_gcc
mkdir openmpi-1.8.6-${SYSTEM_ENV}.${MACHINE_ENV}.64_gcc
cd openmpi-1.8.6-${SYSTEM_ENV}.${MACHINE_ENV}.64_gcc
Cheers,
Gilles
On 6/10/2015 2:42 PM, Siegmar Gross wrote:
Hi Gilles,
a simple workaround is you always run configure in an
hat caused one crash at least, and likely silently compiles old
sources most of the time)
Cheers,
Gilles
On 6/10/2015 10:01 AM, Jeff Squyres (jsquyres) wrote:
Siegmar --
I don't see any reason why this should be happening to you only sometimes; this
code has been unchanged in *forever*.
ls with an
understandable error message
when building for armv6
Cheers,
Gilles
On 6/10/2015 2:18 AM, Jeff Layton wrote:
Gilles,
I was looking in Raspbian a little and do you know what I found?
When I do "gcc -v" it says it was built with "--with-arch=armv6".
Since I'm t
Jeff,
btw, did you try a pI 1 before a pi2 ?
I checked some forums, and you will likely have to upgrade gcc to 4.8
a simpler option could be linaro
https://www.raspberrypi.org/forums/viewtopic.php?f=56&t=98997
Cheers,
Gilles
On Tuesday, June 9, 2015, Gilles Gouaillardet
wrote:
&g
Jeff,
can you
gcc -march=armv7-a foo.c
Cheers,
Gilles
On Tuesday, June 9, 2015, Jeff Layton wrote:
> Gilles,
>
> I'm not cross-compiling - I'm building on the Pi 2.
>
> I'm not sure how to check if gcc can generate armv7 code.
> I'm using Raspbian an
901 - 1000 of 1103 matches
Mail list logo
