{mpi}[GCCcore/10.3.0] OpenMPI v4.1.1 by robert-mijakovic · Pull Request #12566 · easybuilders/easybuild-easyconfigs

robert-mijakovic · 2021-04-12T09:26:02Z

(created using eb --new-pr)
Depends on ~~#12521~~ (GCC, binutils), ~~#12639~~ (libpciaccess), ~~#12604~~ (pkg-config)

boegel · 2021-04-24T18:51:23Z

@robert-mijakovic Final OpenMPI v4.1.1 release is out now, see https://www.open-mpi.org/software/ompi/v4.1/

…asyconfigs into 20210412092552_new_pr_OpenMPI411rc2

boegel · 2021-04-25T09:51:11Z

@boegelbot please test @ generoso
CORE_COUNT=16

boegelbot · 2021-04-25T09:55:09Z

@boegel: Request for testing this PR well received on generoso

PR test command 'EB_PR=12566 EB_ARGS= /apps/slurm/default/bin/sbatch --job-name test_PR_12566 --ntasks=4 ~/boegelbot/eb_from_pr_upload_generoso.sh' executed!

exit code: 0
output:

Submitted batch job 16916

Test results coming soon (I hope)...

Details

- notification for comment with ID 826294077 processed

Message to humans: this is just bookkeeping information for me,
it is of no use to you (unless you think I have a bug, which I don't).

boegelbot · 2021-04-25T10:12:10Z

Test report by @boegelbot
SUCCESS
Build succeeded for 6 out of 6 (6 easyconfigs in total)
generoso-c1-s-1 - Linux centos linux 8.2.2004, x86_64, Intel(R) Xeon(R) CPU E5-2667 v3 @ 3.20GHz (haswell), Python 3.6.8
See https://gist.github.com/1e93ab7f641377478d4365a5c4a30db6 for a full test report.

boegel · 2021-04-25T10:16:43Z

Test report by @boegel
SUCCESS
Build succeeded for 6 out of 6 (6 easyconfigs in total)
node3512.doduo.os - Linux RHEL 8.2, x86_64, AMD EPYC 7552 48-Core Processor (zen2), Python 3.6.8
See https://gist.github.com/bc936e191da9f3f3ef9c67414ff52ba2 for a full test report.

boegel · 2021-04-25T10:29:19Z

Test report by @boegel
SUCCESS
Build succeeded for 6 out of 6 (6 easyconfigs in total)
node3169.skitty.os - Linux centos linux 7.9.2009, x86_64, Intel(R) Xeon(R) Gold 6140 CPU @ 2.30GHz, Python 3.6.8
See https://gist.github.com/dc31d657517cb17792bab4eb88720f10 for a full test report.

boegel · 2021-04-25T10:34:25Z

Test report by @boegel
SUCCESS
Build succeeded for 6 out of 6 (6 easyconfigs in total)
node2690.swalot.os - Linux centos linux 7.9.2009, x86_64, Intel(R) Xeon(R) CPU E5-2660 v3 @ 2.60GHz (haswell), Python 3.6.8
See https://gist.github.com/83f83eac7de2312407a08ab98deb94e7 for a full test report.

boegel · 2021-04-25T13:19:50Z

Test report by @boegel
SUCCESS
Build succeeded for 6 out of 6 (6 easyconfigs in total)
select-pika-c6gd-2xlarge-0001 - Linux centos linux 8.3.2011, AArch64, ARM UNKNOWN (graviton2), Python 3.6.8
See https://gist.github.com/43349d1c89c2e5ba13f00c299c86a861 for a full test report.

boegel · 2021-04-25T13:26:10Z

Test report by @boegel
SUCCESS
Build succeeded for 6 out of 6 (6 easyconfigs in total)
easybuild2.novalocal - Linux centos linux 8.3.2011, POWER, IBM pSeries (emulated by qemu) (power9le), Python 3.6.8
See https://gist.github.com/f62a22d55621b189b267b12c4a00c837 for a full test report.

branfosj · 2021-04-25T17:10:04Z

Running various OSU microbenchmarks (5.7) against OpenMPI 4.1.1 I'm getting:

--------------------------------------------------------------------------
Open MPI failed an OFI Libfabric library call (fi_endpoint).  This is highly
unusual; your job may behave unpredictably (and/or abort) after this.  Local host: bear-pg0211u17a
  Location: mtl_ofi_component.c:513
  Error: Invalid argument (22)
--------------------------------------------------------------------------

I've filed ofiwg/libfabric#6710

Building libfabric with --disable-psm3 removes the issue.

branfosj · 2021-04-28T17:26:35Z

For libfabric:

The bug I am seeing suggests we should add --disable-psm3 for this version and look at enabling psm3 support in a future version when we have a clearer understanding of the problem
OpenMPI-4.0.3-GCC-9.3.0 compilation error on rhel 7 #11939 is more complicated. Based on my testing we'd need a conditional OS dep - if you have libnl v1 then you'll likely also need libnl v3 as well

branfosj · 2021-05-04T17:17:49Z

Running various OSU microbenchmarks (5.7) against OpenMPI 4.1.1 I'm getting:

--------------------------------------------------------------------------
Open MPI failed an OFI Libfabric library call (fi_endpoint).  This is highly
unusual; your job may behave unpredictably (and/or abort) after this.  Local host: bear-pg0211u17a
  Location: mtl_ofi_component.c:513
  Error: Invalid argument (22)
--------------------------------------------------------------------------

I've filed ofiwg/libfabric#6710

Building libfabric with --disable-psm3 removes the issue.

@robert-mijakovic I've put the libfabric patch from upstream in robert-mijakovic#1

add libfabric psm3 init patch from upstream

branfosj · 2021-05-06T08:52:00Z

Test report by @branfosj
SUCCESS
Build succeeded for 6 out of 6 (6 easyconfigs in total)
bear-pg0211u03a.bear.cluster - Linux RHEL 8.3, x86_64, Intel(R) Xeon(R) Gold 6248 CPU @ 2.50GHz (cascadelake), Python 3.6.8
See https://gist.github.com/07ac010b2c0a28a958297199d6814223 for a full test report.

boegel · 2021-05-11T19:28:16Z

@boegelbot please test @ generoso

boegelbot · 2021-05-11T19:30:08Z

@boegel: Request for testing this PR well received on generoso

PR test command 'EB_PR=12566 EB_ARGS= /apps/slurm/default/bin/sbatch --job-name test_PR_12566 --ntasks=4 ~/boegelbot/eb_from_pr_upload_generoso.sh' executed!

exit code: 0
output:

Submitted batch job 17080

Test results coming soon (I hope)...

Details

- notification for comment with ID 839047074 processed

Message to humans: this is just bookkeeping information for me,
it is of no use to you (unless you think I have a bug, which I don't).

boegel · 2021-05-11T19:45:31Z

Test report by @boegel
SUCCESS
Build succeeded for 6 out of 6 (6 easyconfigs in total)
node3125.skitty.os - Linux centos linux 7.9.2009, x86_64, Intel(R) Xeon(R) Gold 6140 CPU @ 2.30GHz, Python 3.6.8
See https://gist.github.com/bdf89b63e20f7e0ff45812dca9eba697 for a full test report.

requested changes made

boegelbot · 2021-05-11T19:47:54Z

Test report by @boegelbot
SUCCESS
Build succeeded for 6 out of 6 (6 easyconfigs in total)
generoso-c1-s-1 - Linux centos linux 8.2.2004, x86_64, Intel(R) Xeon(R) CPU E5-2667 v3 @ 3.20GHz (haswell), Python 3.6.8
See https://gist.github.com/292eb0f51a61a1c0643ce236c89d966d for a full test report.

boegel · 2021-05-11T19:54:12Z

Test report by @boegel
SUCCESS
Build succeeded for 6 out of 6 (6 easyconfigs in total)
node3503.doduo.os - Linux RHEL 8.2, x86_64, AMD EPYC 7552 48-Core Processor (zen2), Python 3.6.8
See https://gist.github.com/cf169ac0940a818be861f08b41ac070d for a full test report.

boegel · 2021-05-11T20:39:14Z

Going in, thanks @robert-mijakovic!

boegel · 2021-05-11T20:54:29Z

Test report by @boegel
SUCCESS
Build succeeded for 6 out of 6 (6 easyconfigs in total)
select-pika-c6gd-2xlarge-0001 - Linux centos linux 8.3.2011, AArch64, ARM UNKNOWN (graviton2), Python 3.6.8
See https://gist.github.com/bf24faaaac40925ede14285a359e07f6 for a full test report.

{WIP} OpenMPI 4.1.1

afd4c7d

This comment has been minimized.

Sign in to view

Adds other dependencies

736000b

robert-mijakovic commented Apr 12, 2021

View reviewed changes

Comment thread easybuild/easyconfigs/p/Perl/Perl-5.32.1-GCCcore-10.3.0.eb

Micket added this to the next release (4.3.5?) milestone Apr 12, 2021

Micket added 2021a update labels Apr 12, 2021

Micket previously requested changes Apr 13, 2021

View reviewed changes

Comment thread easybuild/easyconfigs/o/OpenMPI/OpenMPI-4.1.1rc2-GCC-10.3.0.eb Outdated

Comment thread easybuild/easyconfigs/n/numactl/numactl-2.0.14-GCCcore-10.3.0.eb

Removes XZ since it is merged in another PR.

1b2964b

boegel mentioned this pull request Apr 23, 2021

{toolchain}[system/system] gompi v2021.04 #12571

Merged

Merge branch 'develop' into 20210412092552_new_pr_OpenMPI411rc2

74a7fee

boegel changed the title ~~{WIP} OpenMPI 4.1.1~~ {mpi}[GCCcore/10.3.0] OpenMPI v4.1.1 Apr 24, 2021

boegel reviewed Apr 24, 2021

View reviewed changes

Comment thread easybuild/easyconfigs/o/OpenMPI/OpenMPI-4.1.1rc2-GCC-10.3.0.eb Outdated

boegel requested changes Apr 24, 2021

View reviewed changes

Comment thread easybuild/easyconfigs/u/UCX/UCX-1.10.0-GCCcore-10.3.0.eb Outdated

Comment thread easybuild/easyconfigs/o/OpenMPI/OpenMPI-4.1.1rc2-GCC-10.3.0.eb Outdated

Robert Mijakovic and others added 4 commits April 24, 2021 21:38

OpenMPI 4.1.1 is released thus RC2 is removed

c4c8ea0

OpenMPI now uses PMIx 3.2.3, which is added and UCX homepage is fixed.

5557c26

Fixes PMIx 3.2.3 dependencies.

c6ad643

Merge branch 'develop' of https://github.com/easybuilders/easybuild-e…

aa44a1f

…asyconfigs into 20210412092552_new_pr_OpenMPI411rc2

easybuilders deleted a comment from boegelbot Apr 25, 2021

add libfabric psm3 init patch from upstream

d9fff6f

branfosj mentioned this pull request May 4, 2021

add libfabric psm3 init patch from upstream robert-mijakovic/easybuild-easyconfigs#1

Merged

Merge pull request #1 from branfosj/libfabric

885473c

add libfabric psm3 init patch from upstream

boegel approved these changes May 11, 2021

View reviewed changes

boegel merged commit 03aa9f9 into easybuilders:develop May 11, 2021

branfosj mentioned this pull request Aug 6, 2021

{lib}[GCCcore/11.2.0] libevent v2.1.12, numactl v2.0.14, UCX v1.11.0, ... #13634

Merged

sassy-crick mentioned this pull request Aug 7, 2021

libfabric-1.12.1-GCCcore-10.3.0 on non-avx fails to build #13655

Open

Conversation

robert-mijakovic commented Apr 12, 2021 • edited by boegel Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

This comment has been minimized.

Uh oh!

Uh oh!

Uh oh!

boegel commented Apr 24, 2021

Uh oh!

Uh oh!

Uh oh!

Uh oh!

boegel commented Apr 25, 2021

Uh oh!

boegelbot commented Apr 25, 2021

Uh oh!

boegelbot commented Apr 25, 2021

Uh oh!

boegel commented Apr 25, 2021

Uh oh!

boegel commented Apr 25, 2021

Uh oh!

boegel commented Apr 25, 2021

Uh oh!

boegel commented Apr 25, 2021

Uh oh!

boegel commented Apr 25, 2021

Uh oh!

branfosj commented Apr 25, 2021

Uh oh!

branfosj commented Apr 28, 2021

Uh oh!

branfosj commented May 4, 2021

Uh oh!

branfosj commented May 6, 2021

Uh oh!

boegel commented May 11, 2021

Uh oh!

boegelbot commented May 11, 2021

Uh oh!

boegel commented May 11, 2021

Uh oh!

boegelbot commented May 11, 2021

Uh oh!

boegel commented May 11, 2021

Uh oh!

boegel commented May 11, 2021

Uh oh!

boegel commented May 11, 2021

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

robert-mijakovic commented Apr 12, 2021 •

edited by boegel

Loading