PyPi Wheel Generation by amd-juwillia · Pull Request #3794 · ROCm/rocm-systems

amd-juwillia · 2026-03-05T16:59:31Z

No description provided.

Signed-off-by: Justin Williams <Justin.Williams@amd.com>

Signed-off-by: Maisam Arif <Maisam.Arif@amd.com>

- Link amdsminic static library into libamd_smi_python.so to resolve undefined NIC symbols - Prevent pip-context wrapper from falling back to system libamd_smi.so to avoid dual-loading segfault - Add trailing newline to setup.cfg.in

Signed-off-by: Justin Williams <Justin.Williams@amd.com>

adam360x · 2026-03-11T15:46:23Z

projects/amdsmi/amdsmi_cli/amdsmi_init.py

+    except Exception:
+        _log_version_and_path_diagnostics()
+        print("[amdsmi-cli] Failed to read amdsmi Python package version; aborting to avoid incompatibility.")
+        sys.exit(1)


This is where my code is failing to run in the container. Once we get this fixed I will approve as the rest looks great

adam360x · 2026-03-11T15:47:49Z

projects/amdsmi/py-interface/amdsmi_wrapper.py

 class struct_amdsmi_pcie_info_t(Structure):
    pass

+class struct_pcie_static_(Structure):


This is unrelated to this PR, did you have a stray commit in here? I think we'd want to have this separate.

Signed-off-by: Justin Williams <Justin.Williams@amd.com>

marbre

I took a brief look myself and furthermore let an agent scan through this. As pointed out earlier this will need a written-down RFC before going further with the implementation. Here are some of the comments my agent left:

Python Wheel Version Tagging

The package has no C extension modules — no ext_modules, no cython, no cffi. libamd_smi_python.so is a C++ shared library bundled as package_data, not as a Python extension. pyproject.toml.in and setup.cfg.in both confirm this. Setuptools builds packages with no C extensions as py3-none-linux_x86_64 — Python version agnostic.

The manylinux workflow loops over six interpreters (cp38–cp313), but because the wheel tag is py3-none-linux_x86_64 for all of them, each iteration produces a file with the same name. Successive builds overwrite each other in $RAW_WHEELS. The "six wheels" outcome advertised by the workflow is not what actually happens.

auditwheel was designed for ABI-tagged wheels (cpXX-cpXX-linux_x86_64). Its behavior on py3-none-linux_x86_64 wheels that bundle .so data files is not well-defined and may produce incorrect tags, refuse the input, or pass it through unchanged.

Metadata Duplication:

Both files setup.cfg.in and pyproject.toml.in exist on the branch and both declare the full package metadata (name, version, author, classifiers, etc.). With setuptools ≥ 61, pyproject.toml takes precedence, but having two authoritative metadata sources creates a maintenance hazard: a version bump that updates one but not the other silently produces stale metadata.

CI / Workflow

rpm-test now needs: [manylinux-build, debian-test]. This creates a cross-platform serialization: RPM tests cannot start until all Debian matrix legs complete (Ubuntu20, Ubuntu22, Ubuntu24, Debian10). Previously the RPM build was independent. With continue-on-error: true on Debian legs this may not block execution, but it increases total wall time.
Manylinux gate hardcodes /opt/python/cp38-cp38/bin/python3. If manylinux_2_28 drops cp38 support in a future image update, the gate permanently breaks.
chmod 777 "$WHEEL_OUT" in manylinux-build.yml is overly broad for a container build artifact directory.

marbre · 2026-03-16T21:27:57Z

.github/workflows/amdsmi-build.yml

          done

+
+      - name: Verify Wheel in Site-Packages


What is the reason for not moving this to a dedicated scripts?

marbre · 2026-03-16T21:28:56Z

.github/workflows/amdsmi-build.yml

          echo "Displaying Unit Test Results for ${{ matrix.os }}"
          cat /tmp/test-results-${{ matrix.os }}/unit_test_output.txt || echo "No unit test results found for ${{ matrix.os }}"

+      - name: Perf Test Results


Why cat perf test results here instead of uploading them to a more accessible location?

projects/amdsmi/py-interface/amdsmi_wrapper.py

projects/amdsmi/py-interface/CMakeLists.txt

projects/amdsmi/amdsmi_cli/CMakeLists.txt

projects/amdsmi/py-interface/_find_lib.py.in

marbre · 2026-03-16T22:26:10Z

projects/amdsmi/py-interface/amdsmi_wrapper.py

 class struct_amdsmi_pcie_info_t(Structure):
    pass

+class struct_pcie_static_(Structure):


.github/workflows/manylinux-build.yml

marbre · 2026-03-16T22:58:33Z

As an additional note, I don't quite follow why you try to leverage .pth files which should be limited to a minimum:

https://docs.python.org/3/library/site.html

Note An executable line in a .pth file is run at every Python startup, regardless of whether a particular module is actually going to be used. Its impact should thus be kept to a minimum. The primary intended purpose of executable lines is to make the corresponding module(s) importable (load 3rd-party import hooks, adjust PATH etc). Any other initialization is supposed to be done upon a module’s actual import, if and when it happens. Limiting a code chunk to a single line is a deliberate measure to discourage putting anything more complex here.

Rather see https://www.debian.org/doc/packaging-manuals/python-policy/#module-path. Debian use this even though it is not standard, check how RPM-based distros handle it. As above, this needs a written down plan.

Signed-off-by: Justin Williams <Justin.Williams@amd.com>

github-actions bot added project: amdsmi github actions Pull requests that update GitHub Actions code labels Mar 5, 2026

amd-juwillia force-pushed the juwillia/python-final branch from 859f2ef to 57593b8 Compare March 5, 2026 17:26

marifamd force-pushed the juwillia/python-final branch from 57593b8 to b4a0d5a Compare March 5, 2026 18:15

amd-juwillia force-pushed the juwillia/python-final branch from b4a0d5a to 4d9a212 Compare March 5, 2026 18:33

systems-assistant bot added the organization: ROCm label Mar 5, 2026

Added Python Build Steps

9639199

Signed-off-by: Justin Williams <Justin.Williams@amd.com>

marifamd force-pushed the juwillia/python-final branch from b549d6e to 9639199 Compare March 5, 2026 19:43

Fix amdsmi_init.py wrapper references

8f00256

Signed-off-by: Maisam Arif <Maisam.Arif@amd.com>

marifamd force-pushed the juwillia/python-final branch 2 times, most recently from bef95b3 to 5618d07 Compare March 6, 2026 15:54

Fix DEB10 wheel build

87e9051

- Link amdsminic static library into libamd_smi_python.so to resolve undefined NIC symbols - Prevent pip-context wrapper from falling back to system libamd_smi.so to avoid dual-loading segfault - Add trailing newline to setup.cfg.in

amd-juwillia force-pushed the juwillia/python-final branch from 5618d07 to 87e9051 Compare March 6, 2026 20:24

Added ManyLinux Workflow

78e8a39

Signed-off-by: Justin Williams <Justin.Williams@amd.com>

github-actions bot added the MINOR ABI BREAKAGE label Mar 6, 2026

amd-juwillia force-pushed the juwillia/python-final branch from ed79383 to 68aa824 Compare March 8, 2026 02:52

Python-Site Packages changes

b0dc79f

Signed-off-by: Justin Williams <Justin.Williams@amd.com>

amd-juwillia force-pushed the juwillia/python-final branch from 68aa824 to b0dc79f Compare March 8, 2026 03:12

adam360x requested changes Mar 11, 2026

View reviewed changes

github-actions bot removed the MINOR ABI BREAKAGE label Mar 13, 2026

Added ManyLinux Python Scripts & CI Changes

83dea0d

Signed-off-by: Justin Williams <Justin.Williams@amd.com>

amd-juwillia force-pushed the juwillia/python-final branch from 00aaf6a to 83dea0d Compare March 13, 2026 18:22

amd-juwillia added 4 commits March 13, 2026 14:27

Merge branch 'develop' into juwillia/python-final

618b3b1

Added Python Test Cases

139f7fa

Signed-off-by: Justin Williams <Justin.Williams@amd.com>

Merge branch 'develop' into juwillia/python-final

3de6600

Removed CLI Unit Tests

430a95c

Signed-off-by: Justin Williams <Justin.Williams@amd.com>

marbre reviewed Mar 16, 2026

View reviewed changes

Updated workflow

6d00c1a

Signed-off-by: Justin Williams <Justin.Williams@amd.com>

amd-juwillia force-pushed the juwillia/python-final branch from 379bccb to 6d00c1a Compare March 19, 2026 14:20

Added AMDSMI Build Python Scripts

4b1c266

Signed-off-by: Justin Williams <Justin.Williams@amd.com>

amd-juwillia force-pushed the juwillia/python-final branch from 0f80d99 to 4b1c266 Compare March 21, 2026 19:34

Merge branch 'develop' into juwillia/python-final

20356db

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

PyPi Wheel Generation#3794

PyPi Wheel Generation#3794
amd-juwillia wants to merge 13 commits intodevelopfrom
juwillia/python-final

amd-juwillia commented Mar 5, 2026

Uh oh!

adam360x Mar 11, 2026

Uh oh!

adam360x Mar 11, 2026

Uh oh!

marbre Mar 16, 2026

Uh oh!

marbre left a comment •

edited

Loading

Uh oh!

marbre Mar 16, 2026

Uh oh!

marbre Mar 16, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

marbre Mar 16, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

marbre commented Mar 16, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Conversation

amd-juwillia commented Mar 5, 2026

Uh oh!

adam360x Mar 11, 2026

Choose a reason for hiding this comment

Uh oh!

adam360x Mar 11, 2026

Choose a reason for hiding this comment

Uh oh!

marbre Mar 16, 2026

Choose a reason for hiding this comment

Uh oh!

marbre left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Python Wheel Version Tagging

Metadata Duplication:

CI / Workflow

Uh oh!

marbre Mar 16, 2026

Choose a reason for hiding this comment

Uh oh!

marbre Mar 16, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

marbre Mar 16, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

marbre commented Mar 16, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

marbre left a comment •

edited

Loading