Make HPL easyblock intel-aware + allow running the `xhpl` benchmark as a test by Crivella · Pull Request #3917 · easybuilders/easybuild-easyblocks

Crivella · 2025-09-05T15:29:34Z

Potential fix for

HPL will not test correctly in intel/2022b and intel/2023a easybuild-easyconfigs#19322

I think we could also not run the benchmark as a a test instead of pinning everything to 1 core (if we do not wanna deal with pinning with every different type of MPI)

Crivella · 2025-09-05T15:37:25Z

@boegelbot please test @ jsc-zen3
EB_ARGS="--installpath /tmp/$USER/ebpr-3917 HPL-2.3-intel-2021a.eb HPL-2.3-intel-2022.00.eb HPL-2.3-intel-2022.11.eb HPL-2.3-intel-2022a.eb HPL-2.3-intel-2023a.eb HPL-2.3-intel-2024a.eb HPL-2.3-foss-2025a.eb HPL-2.3-foss-2024a.eb HPL-2.3-foss-2023a.eb HPL-2.3-foss-2022a.eb"

boegelbot · 2025-09-05T15:53:09Z

@Crivella: Request for testing this PR well received on jsczen3l1.int.jsc-zen3.fz-juelich.de

PR test command 'if [[ develop != 'develop' ]]; then EB_BRANCH=develop ./easybuild_develop.sh 2> /dev/null 1>&2; EB_PREFIX=/home/boegelbot/easybuild/develop source init_env_easybuild_develop.sh; fi; EB_PR=3917 EB_ARGS="--installpath /tmp/$USER/ebpr-3917 HPL-2.3-intel-2021a.eb HPL-2.3-intel-2022.00.eb HPL-2.3-intel-2022.11.eb HPL-2.3-intel-2022a.eb HPL-2.3-intel-2023a.eb HPL-2.3-intel-2024a.eb HPL-2.3-foss-2025a.eb HPL-2.3-foss-2024a.eb HPL-2.3-foss-2023a.eb HPL-2.3-foss-2022a.eb" EB_CONTAINER= EB_REPO=easybuild-easyblocks EB_BRANCH=develop /opt/software/slurm/bin/sbatch --job-name test_PR_3917 --ntasks=8 ~/boegelbot/eb_from_pr_upload_jsc-zen3.sh' executed!

exit code: 0
output:

Submitted batch job 7872

Test results coming soon (I hope)...

Details

- notification for comment with ID 3258815718 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 · 2025-09-05T16:13:17Z

Test report by @boegelbot

Overview of tested easyconfigs (in order)

SUCCESS HPL-2.3-intel-2021a.eb
SUCCESS HPL-2.3-intel-2022a.eb
SUCCESS HPL-2.3-intel-2023a.eb
SUCCESS HPL-2.3-intel-2024a.eb
SUCCESS HPL-2.3-foss-2025a.eb
SUCCESS HPL-2.3-foss-2024a.eb
SUCCESS HPL-2.3-foss-2023a.eb
SUCCESS HPL-2.3-foss-2022a.eb
SUCCESS impi-2021.5.0-intel-compilers-2022.0.1.eb
SUCCESS imkl-2022.0.1.eb
SUCCESS iimpi-2022.11.eb
SUCCESS imkl-FFTW-2022.2.1-iimpi-2022.11.eb
SUCCESS intel-2022.11.eb
SUCCESS HPL-2.3-intel-2022.11.eb
SUCCESS iimpi-2022.00.eb
SUCCESS imkl-FFTW-2022.0.1-iimpi-2022.00.eb
SUCCESS intel-2022.00.eb
SUCCESS HPL-2.3-intel-2022.00.eb

Build succeeded for 18 out of 18 (10 easyconfigs in total)
jsczen3c4.int.jsc-zen3.fz-juelich.de - Linux Rocky Linux 9.6, x86_64, AMD EPYC-Milan Processor (zen3), Python 3.9.21
See https://gist.github.com/boegelbot/d182309b85c1664162f79d5d440bbfef for a full test report.

Crivella · 2025-09-08T08:54:58Z

Test report by @Crivella

Overview of tested easyconfigs (in order)

SUCCESS HPL-2.3-intel-2021a.eb
SUCCESS HPL-2.3-intel-2022.00.eb
SUCCESS HPL-2.3-intel-2022.11.eb
SUCCESS HPL-2.3-intel-2022a.eb
SUCCESS HPL-2.3-intel-2023a.eb
SUCCESS HPL-2.3-intel-2024a.eb
SUCCESS HPL-2.3-intel-2025b.eb
SUCCESS HPL-2.3-foss-2025a.eb
SUCCESS HPL-2.3-foss-2024a.eb
SUCCESS HPL-2.3-foss-2023a.eb
SUCCESS HPL-2.3-foss-2022a.eb

Build succeeded for 11 out of 11 (11 easyconfigs in total)
crivella-desktop - Linux Ubuntu 22.04.5 LTS (Jammy Jellyfish), x86_64, 13th Gen Intel(R) Core(TM) i9-13900K (skylake), Python 3.11.13
See https://gist.github.com/Crivella/a947c3fdf94fdfd1d73f5206012236f7 for a full test report.

Thyre

LGTM.
Let me throw together an Intel + OpenMPI build real quick to also test that, but I would expect to see this succeed as well.

I guess you've tested the residual test check before implementing the configure option? It's hard to test this when we have CCNOOPT set 😄

Crivella · 2025-09-08T09:00:34Z

I guess you've tested the residual test check before implementing the configure option? It's hard to test this when we have CCNOOPT set

I had done a test locally by commenting that section and the intel ones would fail

I am also doing 2 tests on a single HPL with --max-parallel 2 for one and --disable-mpi-tests for the other to show that it still works (just takes a few minutes instead than a few seconds)

Thyre · 2025-09-08T09:01:33Z

Test report by @Thyre

Overview of tested easyconfigs (in order)

SUCCESS HPL-2.3-iomkl-2024a.eb

Build succeeded for 1 out of 1 (1 easyconfigs in total)
ZAM054 - Linux Zorin OS 17, x86_64, 12th Gen Intel(R) Core(TM) i7-1260P, 1 x NVIDIA NVIDIA GeForce MX550, 580.65.06, Python 3.10.12
See https://gist.github.com/Thyre/bfe9cadf88deacdd327742c432eb4efb for a full test report.

Thyre · 2025-09-08T09:04:04Z

Test report by @Thyre

Overview of tested easyconfigs (in order)

SUCCESS HPL-2.3-intel-2025b.eb
SUCCESS HPL-2.3-foss-2025b.eb
SUCCESS HPL-2.3-foss-2025a.eb

Build succeeded for 3 out of 3 (3 easyconfigs in total)
Linux - Linux Arch Linux UNKNOWN, x86_64, AMD Ryzen 7 7800X3D 8-Core Processor (zen4), 1 x AMD Navi 48 [Radeon RX 9070/9070 XT/9070 GRE] (device id: 0x7550, gfx: gfx1201, driver: 6.16.5-arch1-1), 1 x AMD Raphael (device id: 0x164e, gfx: gfx1036, driver: 6.16.5-arch1-1), Python 3.13.7
See https://gist.github.com/Thyre/571bcf3aad5617b61a64b3d259617601 for a full test report.

Thyre · 2025-09-08T09:04:40Z

Hm, we get this warning during the test step:

WARNING: Deprecated functionality, will no longer work in EasyBuild v6.0: Easyconfig parameter 'parallel' is deprecated, use 'max_parallel' or the parallel property instead.; see https://docs.easybuild.io/deprecated-functionality/ for more information

Should we use max_parallel instead?

Crivella · 2025-09-08T09:07:01Z

Ran with --max-parallel 2 forcing to oversubscribe to 1 cpu

Test report by @Crivella

Overview of tested easyconfigs (in order)

SUCCESS HPL-2.3-intel-2025a.eb

Build succeeded for 1 out of 1 (1 easyconfigs in total)
crivella-desktop - Linux Ubuntu 22.04.5 LTS (Jammy Jellyfish), x86_64, 13th Gen Intel(R) Core(TM) i9-13900K (skylake), Python 3.11.13
See https://gist.github.com/Crivella/3a507059c4528bcba6d13f3a9b3df04b for a full test report.

Crivella · 2025-09-08T09:07:08Z

Ran with --disable-mpi-tests forcing to oversubscribe to 1 cpu

Test report by @Crivella

Overview of tested easyconfigs (in order)

SUCCESS HPL-2.3-foss-2025a.eb

Build succeeded for 1 out of 1 (1 easyconfigs in total)
crivella-desktop - Linux Ubuntu 22.04.5 LTS (Jammy Jellyfish), x86_64, 13th Gen Intel(R) Core(TM) i9-13900K (skylake), Python 3.11.13
See https://gist.github.com/Crivella/b7ba38edf2c7ef875bbfc451ad13173e for a full test report.

Crivella · 2025-09-08T09:20:12Z

Thanks @ocaisa for the suggestion of using
https://github.com/easybuilders/easybuild-framework/blob/develop/easybuild/tools/toolchain/mpi.py#L273
directly and allow using srun or other command templates through --mpi-cmd-template

Gonna try and implement it here

Crivella · 2025-09-08T09:40:19Z

@boegelbot please test @ jsc-zen3
EB_ARGS="--installpath /tmp/$USER/ebpr-3917 HPL-2.3-intel-2021a.eb HPL-2.3-intel-2022.00.eb HPL-2.3-intel-2022.11.eb HPL-2.3-intel-2022a.eb HPL-2.3-intel-2023a.eb HPL-2.3-intel-2024a.eb HPL-2.3-foss-2025a.eb HPL-2.3-foss-2024a.eb HPL-2.3-foss-2023a.eb HPL-2.3-foss-2022a.eb"

Crivella · 2025-09-08T09:47:10Z

Concerning testing without the CCNOOPTS set i get the following

   * [SUCCESS] HPL/2.3-intel-2021a                                                                                                                                                                                                                                                              
   * [SUCCESS] HPL/2.3-intel-2022.00                                                                                                                                                                                                                                                            
   * [FAILED]  HPL/2.3-intel-2022.11                                                                                                                                                                                                                                                            
   * [SUCCESS] HPL/2.3-intel-2022a                                                                                                                                                                                                                                                              
   * [FAILED] HPL/2.3-intel-2023a                                                                                                                                                                                                                                                              
   * [FAILED] HPL/2.3-intel-2024a                                                                                                                                                                                                                                                              
   * [FAILED] HPL/2.3-intel-2025a

2025b succeeds because we have the flag in the EC right now

boegelbot · 2025-09-08T09:53:11Z

@Crivella: Request for testing this PR well received on jsczen3l1.int.jsc-zen3.fz-juelich.de

PR test command 'if [[ develop != 'develop' ]]; then EB_BRANCH=develop ./easybuild_develop.sh 2> /dev/null 1>&2; EB_PREFIX=/home/boegelbot/easybuild/develop source init_env_easybuild_develop.sh; fi; EB_PR=3917 EB_ARGS="--installpath /tmp/$USER/ebpr-3917 HPL-2.3-intel-2021a.eb HPL-2.3-intel-2022.00.eb HPL-2.3-intel-2022.11.eb HPL-2.3-intel-2022a.eb HPL-2.3-intel-2023a.eb HPL-2.3-intel-2024a.eb HPL-2.3-foss-2025a.eb HPL-2.3-foss-2024a.eb HPL-2.3-foss-2023a.eb HPL-2.3-foss-2022a.eb" EB_CONTAINER= EB_REPO=easybuild-easyblocks EB_BRANCH=develop /opt/software/slurm/bin/sbatch --job-name test_PR_3917 --ntasks=8 ~/boegelbot/eb_from_pr_upload_jsc-zen3.sh' executed!

exit code: 0
output:

Submitted batch job 7879

Test results coming soon (I hope)...

Details

- notification for comment with ID 3265449064 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 · 2025-09-08T10:18:15Z

Test report by @boegelbot

Overview of tested easyconfigs (in order)

SUCCESS HPL-2.3-intel-2021a.eb
SUCCESS HPL-2.3-intel-2022a.eb
SUCCESS HPL-2.3-intel-2023a.eb
SUCCESS HPL-2.3-intel-2024a.eb
SUCCESS HPL-2.3-foss-2025a.eb
SUCCESS HPL-2.3-foss-2024a.eb
SUCCESS HPL-2.3-foss-2023a.eb
SUCCESS HPL-2.3-foss-2022a.eb
SUCCESS impi-2021.5.0-intel-compilers-2022.0.1.eb
SUCCESS imkl-2022.0.1.eb
SUCCESS iimpi-2022.11.eb
SUCCESS imkl-FFTW-2022.2.1-iimpi-2022.11.eb
SUCCESS intel-2022.11.eb
SUCCESS HPL-2.3-intel-2022.11.eb
SUCCESS iimpi-2022.00.eb
SUCCESS imkl-FFTW-2022.0.1-iimpi-2022.00.eb
SUCCESS intel-2022.00.eb
SUCCESS HPL-2.3-intel-2022.00.eb

Build succeeded for 18 out of 18 (10 easyconfigs in total)
jsczen3c3.int.jsc-zen3.fz-juelich.de - Linux Rocky Linux 9.6, x86_64, AMD EPYC-Milan Processor (zen3), Python 3.9.21
See https://gist.github.com/boegelbot/e75b0579b7015a6700f409969b96e2e1 for a full test report.

boegel · 2025-09-10T14:26:37Z

Test report by @boegel

Overview of tested easyconfigs (in order)

SUCCESS HPL-2.3-foss-2023a.eb
SUCCESS HPL-2.3-foss-2023b.eb
SUCCESS HPL-2.3-foss-2024a.eb
SUCCESS HPL-2.3-foss-2025a.eb
SUCCESS HPL-2.3-foss-2025b.eb
SUCCESS HPL-2.3-intel-2023a.eb
SUCCESS HPL-2.3-intel-2023b.eb
SUCCESS HPL-2.3-intel-2024a.eb
SUCCESS HPL-2.3-intel-2025a.eb
SUCCESS HPL-2.3-intel-2025b.eb

Build succeeded for 10 out of 10 (10 easyconfigs in total)
node3522.doduo.os - Linux RHEL 9.4, x86_64, AMD EPYC 7552 48-Core Processor (zen2), Python 3.9.18
See https://gist.github.com/boegel/e088b290993430bb36fd09bfea750f17 for a full test report.

boegel

lgtm

intel-aware EB + running tests

0158fed

Crivella added enhancement bug fix labels Sep 5, 2025

Fix line length

674cf12

Crivella requested a review from Thyre September 5, 2025 15:39

Thyre reviewed Sep 5, 2025

View reviewed changes

Comment thread easybuild/easyblocks/h/hpl.py Outdated

Comment thread easybuild/easyblocks/h/hpl.py Outdated

Updated to set MPI flags based on mpifamily and not compiler family

0a2413d

Crivella marked this pull request as ready for review September 8, 2025 08:55

Thyre approved these changes Sep 8, 2025

View reviewed changes

Use parallel property to avoid deprecation

cd393fc

Crivella added 2 commits September 8, 2025 11:33

Use mpi_cmd_for instead of configopts to specify the mpi command

b056c2a

Remove unused import

71b61e6

Crivella mentioned this pull request Sep 8, 2025

Possibly enabling oversubscribe/cpu-pinning in framework easybuilders/easybuild-framework#4993

Open

boegel added this to the next release (5.1.2) milestone Sep 10, 2025

boegel changed the title ~~Make HPL EasyBlock intel-aware and allow running the xhpl benchmark as a tests~~ Make HPL awayblock intel-aware + allow running the xhpl benchmark as a tests Sep 10, 2025

boegel changed the title ~~Make HPL awayblock intel-aware + allow running the xhpl benchmark as a tests~~ Make HPL awayblock intel-aware + allow running the xhpl benchmark as a test Sep 10, 2025

boegel approved these changes Sep 10, 2025

View reviewed changes

boegel merged commit b6b4395 into easybuilders:develop Sep 10, 2025
17 checks passed

boegel mentioned this pull request Sep 10, 2025

HPL will not test correctly in intel/2022b and intel/2023a easybuilders/easybuild-easyconfigs#19322

Closed

Crivella deleted the fix-hpl branch September 10, 2025 14:45

Crivella mentioned this pull request Sep 10, 2025

remove now unneeded build option for HPL w/ intel/2025b easybuilders/easybuild-easyconfigs#23842

Merged

boegel changed the title ~~Make HPL awayblock intel-aware + allow running the xhpl benchmark as a test~~ Make HPL easyblock intel-aware + allow running the xhpl benchmark as a test Sep 24, 2025

Conversation

Crivella commented Sep 5, 2025

Uh oh!

Crivella commented Sep 5, 2025

Uh oh!

boegelbot commented Sep 5, 2025

Uh oh!

Uh oh!

Uh oh!

boegelbot commented Sep 5, 2025

Overview of tested easyconfigs (in order)

Uh oh!

Crivella commented Sep 8, 2025

Overview of tested easyconfigs (in order)

Uh oh!

Thyre left a comment

Choose a reason for hiding this comment

Uh oh!

Crivella commented Sep 8, 2025

Uh oh!

Thyre commented Sep 8, 2025

Overview of tested easyconfigs (in order)

Uh oh!

Thyre commented Sep 8, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Overview of tested easyconfigs (in order)

Uh oh!

Thyre commented Sep 8, 2025

Uh oh!

Crivella commented Sep 8, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Overview of tested easyconfigs (in order)

Uh oh!

Crivella commented Sep 8, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Overview of tested easyconfigs (in order)

Uh oh!

Crivella commented Sep 8, 2025

Uh oh!

Crivella commented Sep 8, 2025

Uh oh!

Crivella commented Sep 8, 2025

Uh oh!

boegelbot commented Sep 8, 2025

Uh oh!

boegelbot commented Sep 8, 2025

Overview of tested easyconfigs (in order)

Uh oh!

boegel commented Sep 10, 2025

Overview of tested easyconfigs (in order)

Uh oh!

boegel left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Thyre commented Sep 8, 2025 •

edited

Loading

Crivella commented Sep 8, 2025 •

edited

Loading

Crivella commented Sep 8, 2025 •

edited

Loading