Let me see if I can get access to a Power 7 machine internally, and I will
try to replicate.
On Tue, May 3, 2016 at 9:54 AM, Paul Hargrove wrote:
> Josh,
>
> This is a Power7 (big-endian) system "BlueBiou" at Rice.
>
> Xlc is 13.1 with no updates:
> $ xlc -qversion
> IBM XL C/C++ for Linux, V13.
Josh,
This is a Power7 (big-endian) system "BlueBiou" at Rice.
Xlc is 13.1 with no updates:
$ xlc -qversion
IBM XL C/C++ for Linux, V13.1
Version: 13.01..
I don't have privileged to install updates.
FWIW: I too can build with xlc for little-endian Power8.
$ /opt/ibm/xlC/13.1.2/bin/xlc -
Fixed on master and PR'ed for v2.0.0. Thanks.
> On May 2, 2016, at 4:45 PM, Paul Hargrove wrote:
>
> It appears that 2.0.0rc2 contains the same sort of alloca()-related error I
> reported against 1.10.2rc1 in
> https://www.open-mpi.org/community/lists/devel/2015/12/18440.php
>
> This time th
Paul,
What generation of the power arch are you using?
We have successfully built (a few weeks ago) with the xlc compiler 13.1.3
on a Power 8 (pcc64le). It might be related the big vs. little endian, but
I wonder if it is something that was fixed in a point release of the xlc
compiler. Are you ab
Fixed in master and PR'ed for v2.0.1. Thanks.
> On May 2, 2016, at 11:19 PM, Paul Hargrove wrote:
>
> I can see why no competent OMPI developer ever encountered the following
> cut-and-paste error:
>
> diff --git a/config/opal_mca.m4 b/config/opal_mca.m4
> index 1ec342c..d3a7075 100644
> ---
well, actually the line that causes the trouble is unnecessary.
Detecting the file system has been abstracted out, and the struct statfs
is completely unused. I will create a one-line patch for that.
Edgar
On 5/2/2016 10:54 PM, Paul Hargrove wrote:
On NetNSD-7 I am testing PR1128 to get past
Thanks Brice; I filed the corresponding v2.0 PR here:
https://github.com/open-mpi/ompi-release/pull/1130
> On May 3, 2016, at 4:13 AM, Brice Goglin wrote:
>
> https://github.com/open-mpi/ompi/pull/1621 (against master, needs to go to
> 2.0 later)
>
>
> Le 03/05/2016 08:22, Brice Goglin
https://github.com/open-mpi/ompi/pull/1621 (against master, needs to go
to 2.0 later)
Le 03/05/2016 08:22, Brice Goglin a écrit :
> Yes we should backport this to OMPI master and v2.x.
> I am usually not the one doing the PR, I'd need to learn the exact
> procedure first :)
>
> Brice
>
>
>
> Le 0
For possible inclusion in README:
xlc-13.1.0 on Linux dies compiling the embedded hwloc in this rc (details
below).
The same is true of the released hwloc 1.11.3
-Paul
libtool: compile: xlc -DHAVE_CONFIG_H -I.
-I/gpfs-biou/phh1/OMPI/openmpi-2.0.0rc2-linux-ppc64-xlc-13.1/openmpi-2.0.0rc2/opal/mca
For the record:
When last we communicated on the issue of supporting FC=nagfor the
milestone was set at 1.10.3 (issue #1284).
While Gilles did make some progress, several of the same things reported
originally are still broken in 2.0.0rc2 (and presumably in master and
v1.10).
-Paul
--
Paul H. H
With the ekopath (aka pathscale) 5.0.5 and 6.0.527 compilers I get an ICE
(Internal Compiler Error) building this rc.
While this is clearly not an OMPI bug, it may be worth recording in the
README.
The (different) errors for these two compilers are shown below.
-Paul
libtool: compile: pathcc -D
Yes we should backport this to OMPI master and v2.x.
I am usually not the one doing the PR, I'd need to learn the exact
procedure first :)
Brice
Le 03/05/2016 08:15, Paul Hargrove a écrit :
> Thanks, Brice.
>
> Any plans to get this fix into Open MPI's embedded copy of hwloc
> 1.11.2, and into
Thanks, Brice.
Any plans to get this fix into Open MPI's embedded copy of hwloc 1.11.2,
and into v2.x in particular?
Or perhaps that is Jeff's job?
-Paul
On Mon, May 2, 2016 at 11:04 PM, Brice Goglin wrote:
> Should be fixed by
> https://github.com/open-mpi/hwloc/commit/9549fd59af04dca2e2340e1
Should be fixed by
https://github.com/open-mpi/hwloc/commit/9549fd59af04dca2e2340e17f0e685f8c552d818
Thanks for the report
Brice
Le 02/05/2016 21:53, Paul Hargrove a écrit :
> I have a linux/ppc64 host running Fedora 20.
> I have configured the 2.0.0rc2 tarball with
>
> --prefix=[] --ena
This is NOT a new issue, but I wanted to mention it explicitly once again
since no progress has been made since I first reported to problem in
1.8.6rc1 (about 1 year ago).
Though the directory name and line numbers have changed, the error reported
in https://www.open-mpi.org/community/lists/devel/
15 matches
Mail list logo