LLVM: Add patch to fix bindc commonblocks in fortran compilation

[Crivella]

Applies patch from

• [flang] handle common block used as BIND(C) module variables llvm/llvm-project#145669

to LLVMs before 21 aimed at being the base of a toolchain.

Needed EG to compile HDF5 with LLVM and avoid the error described in
https://discourse.llvm.org/t/hlfir-declare-op-of-numeric-logical-or-assumed-type-entity-must-not-have-length-parameters/83629

[Crivella]

@boegelbot please test @ jsc-zen3
EB_ARGS="--installpath /tmp/$USER/pr-24495"

[boegelbot]

@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=24495 EB_ARGS="--installpath /tmp/$USER/pr-24495" EB_CONTAINER= EB_REPO=easybuild-easyconfigs EB_BRANCH=develop /opt/software/slurm/bin/sbatch --job-name test_PR_24495 --ntasks=8 ~/boegelbot/eb_from_pr_upload_jsc-zen3.sh' executed!

• exit code: 0
• output:

Submitted batch job 8701

Test results coming soon (I hope)...

- notification for comment with ID 3506560029 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]

Test report by @boegelbot
SUCCESS
Build succeeded for 3 out of 3 (total: 15 hours 38 mins 36 secs) (3 easyconfigs in total)
jsczen3c1.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/7b0fc53013222902a74fdcb95c349736 for a full test report.

[Thyre]

We should test if this is sufficient for HDF5, but looks good to me otherwise.

[Crivella]

We should test if this is sufficient for HDF5, but looks good to me otherwise.

Already did on both 2023b and 2025b, will open more PRs tomorrow from work.
One thing that i want to investigate though before we go all-in is the rpath wrappers, and making sure we are using ld.lld instead of ld from GCC

[Thyre]

Great! I also have a few more things in my local repository (HPL, HPCG with EasyBlock changes, GROMACS, MUST, OTF-CPT and Score-P). I'll probably look into other stuff at some point as well, but those had the highest priority for me first.

[Crivella]

Going to re-run a build without --installpath so we get the patched version of LLVM when starting to use the bot to make llvm-compilers builds

[Crivella]

@boegelbot please test @ jsc-zen3

[boegelbot]

@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=24495 EB_ARGS= EB_CONTAINER= EB_REPO=easybuild-easyconfigs EB_BRANCH=develop /opt/software/slurm/bin/sbatch --job-name test_PR_24495 --ntasks=8 ~/boegelbot/eb_from_pr_upload_jsc-zen3.sh' executed!

• exit code: 0
• output:

Submitted batch job 8708

Test results coming soon (I hope)...

- notification for comment with ID 3511084472 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]

Test report by @boegelbot
SUCCESS
Build succeeded for 3 out of 3 (total: 15 hours 28 mins 22 secs) (3 easyconfigs in total)
jsczen3c1.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/06dd8990c4801a60bc9753658190dc85 for a full test report.

[Micket]

lgtm, did you guys manage to verify that it fixed the issue with hdf5?

[Thyre]

lgtm, did you guys manage to verify that it fixed the issue with hdf5?

@Crivella already verified this, but I think forgot to upload the test report. I have two builds of LLVM running and will verify with HDF5 afterwards.

[Thyre]

Test report by @Thyre
SUCCESS
Build succeeded for 1 out of 1 (total: 4 hours 16 mins 1 secs) (1 easyconfigs in total)
datenlager - Linux Ubuntu 24.04, x86_64, AMD Ryzen 7 3700X 8-Core Processor, Python 3.12.3
See https://gist.github.com/Thyre/61520dde446e222b41f0c16d58ee9705 for a full test report.

[Thyre]

@Micket, see #24516 (comment) for a successful test report for HDF5 with this PR.

[Thyre]

Test report by @Thyre
SUCCESS
Build succeeded for 1 out of 1 (total: 5 hours 48 mins 37 secs) (1 easyconfigs in total)
ZAM054 - Linux Zorin OS 18, x86_64, 12th Gen Intel(R) Core(TM) i7-1260P (skylake), 1 x NVIDIA NVIDIA GeForce MX550, 580.95.05, Python 3.12.3
See https://gist.github.com/Thyre/86405fc44db2e28724a05ea209d35dc7 for a full test report.