Skip to content

Conversation

@erman-gurses
Copy link
Contributor

@erman-gurses erman-gurses commented Dec 1, 2025

This PR removes the standalone amdsmi submodule under base/amdsmi and switches the amdsmi subproject to use the sources from ${THEROCK_ROCM_SYSTEMS_SOURCE_DIR}/projects/amdsmi.

This prepares the codebase for the follow-up PR that moves amdsmi from the base layer to the core layer without involving any submodule path changes.

Comment on lines -44 to -47
[submodule "amdsmi"]
path = base/amdsmi
url = https://github.com/ROCm/amdsmi.git
branch = amd-staging
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is failing the CI setup step: https://github.com/ROCm/TheRock/actions/runs/19814551871/job/56763040892?pr=2350#step:4:40

Enabling build jobs since a non-skippable path was modified
Traceback (most recent call last):
  File "/home/runner/work/TheRock/TheRock/./build_tools/github_actions/configure_ci.py", line 591, in <module>
    main(base_args, linux_families, windows_families)
  File "/home/runner/work/TheRock/TheRock/./build_tools/github_actions/configure_ci.py", line 522, in main
    submodule_paths = get_therock_submodule_paths()
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/runner/work/TheRock/TheRock/./build_tools/github_actions/configure_ci.py", line 96, in get_therock_submodule_paths
    response = subprocess.run(
               ^^^^^^^^^^^^^^^
  File "/usr/lib/python3.12/subprocess.py", line 571, in run
    raise CalledProcessError(retcode, process.args,
subprocess.CalledProcessError: Command '['git', 'submodule', 'status']' returned non-zero exit status 128.

You should also actually delete the submodule from https://github.com/ROCm/TheRock/tree/main/base.

I usually reference https://gist.github.com/myusuf3/7f645819ded92bda6677 or https://stackoverflow.com/a/1260982 whenever I delete a submodule

Copy link
Contributor Author

@erman-gurses erman-gurses Dec 2, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you! After the changes, I can see amd-smi was configured, built, and installed successfully from the last build.

Here are the logs:

Config Log: https://therock-ci-artifacts.s3.amazonaws.com/19841724039-linux/logs/gfx94X-dcgpu/amdsmi_configure.log

I see this in the config /__w/TheRock/TheRock/rocm-systems/projects/amdsmi that amd-smi is used from rocm-systems

Full configuration:

EXEC /__w/TheRock/TheRock/build/base/build /usr/local/therock-tools/bin/cmake -E env --unset=ROCM_PATH --unset=ROCM_DIR --unset=HIP_PATH --unset=HIP_DIR -- /usr/local/therock-tools/bin/cmake -GNinja -B/__w/TheRock/TheRock/build/base/build -S/__w/TheRock/TheRock/rocm-systems/projects/amdsmi -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=/__w/TheRock/TheRock/build/base/stage -DTHEROCK_STAGE_INSTALL_ROOT=/__w/TheRock/TheRock/build/base/stage -DCMAKE_TOOLCHAIN_FILE=/__w/TheRock/TheRock/build/base/_toolchain.cmake -DCMAKE_PROJECT_TOP_LEVEL_INCLUDES=/__w/TheRock/TheRock/build/base/_init.cmake -DCMAKE_VERBOSE_MAKEFILE=OFF

Build Log: https://therock-ci-artifacts.s3.amazonaws.com/19841724039-linux/logs/gfx94X-dcgpu/amdsmi_build.log

Install Log: https://therock-ci-artifacts.s3.amazonaws.com/19841724039-linux/logs/gfx94X-dcgpu/amdsmi_install.log
0.0 -- Installing: /__w/TheRock/TheRock/build/base/stage/bin/amd-smi

@erman-gurses erman-gurses marked this pull request as ready for review December 2, 2025 16:49
@erman-gurses erman-gurses merged commit 2f2710e into main Dec 4, 2025
61 of 64 checks passed
@erman-gurses erman-gurses deleted the user/erman-gurses/switch-amd-smi branch December 4, 2025 15:10
@github-project-automation github-project-automation bot moved this from TODO to Done in TheRock Triage Dec 4, 2025
erman-gurses added a commit that referenced this pull request Dec 5, 2025
jayhawk-commits pushed a commit that referenced this pull request Dec 7, 2025
…tems (#2350)

This PR removes the standalone amdsmi submodule under base/amdsmi and
switches the amdsmi subproject to use the sources from
`${THEROCK_ROCM_SYSTEMS_SOURCE_DIR}/projects/amdsmi`.

This prepares the codebase for the follow-up PR that moves amdsmi from
the base layer to the core layer without involving any submodule path
changes.
rponnuru5 pushed a commit that referenced this pull request Dec 9, 2025
…tems (#2350)

This PR removes the standalone amdsmi submodule under base/amdsmi and
switches the amdsmi subproject to use the sources from
`${THEROCK_ROCM_SYSTEMS_SOURCE_DIR}/projects/amdsmi`.

This prepares the codebase for the follow-up PR that moves amdsmi from
the base layer to the core layer without involving any submodule path
changes.
rponnuru5 pushed a commit that referenced this pull request Dec 9, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

Status: Done

Development

Successfully merging this pull request may close these issues.

3 participants