test(bindings): work around Orin NVML field-values bug

[rwgk]

Description

Seed the field status with an invalid sentinel so a successful native no-op cannot accidentally pass on recycled memory. Xfail when Orin leaves the sentinel unchanged, preserving normal checks on fixed drivers and other devices.

Label the workaround explicitly in the source comment and xfail reason.

Interactive testing on Jetson Orin:

jetson:~/wrk/forked/work_around_orin_nvml_field_values_bug $ grep_pytest_summary `nlog`
/home/rgrossekunst/wrk/logs/cuda-python_qa_bindings_linux_2026-10-18+234118+0000_tests_log.txt
rootdir: /home/rgrossekunst/wrk/forked/work_around_orin_nvml_field_values_bug/cuda_pathfinder
======================= 1618 passed, 2 skipped in 13.78s =======================
rootdir: /home/rgrossekunst/wrk/forked/work_around_orin_nvml_field_values_bug/cuda_bindings
====== 633 passed, 80 skipped, 1 xfailed, 17 warnings in 70.06s (0:01:10) ======
rootdir: /home/rgrossekunst/wrk/forked/work_around_orin_nvml_field_values_bug/cuda_bindings
====== 633 passed, 80 skipped, 1 xfailed, 17 warnings in 66.12s (0:01:06) ======
rootdir: /home/rgrossekunst/wrk/forked/work_around_orin_nvml_field_values_bug/cuda_bindings
============================== 9 passed in 1.01s ===============================
rootdir: /home/rgrossekunst/wrk/forked/work_around_orin_nvml_field_values_bug/cuda_core
= 1 failed, 3741 passed, 535 skipped, 3 xfailed, 3 warnings, 1 error, 9 subtests passed in 127.94s (0:02:07) =
rootdir: /home/rgrossekunst/wrk/forked/work_around_orin_nvml_field_values_bug/cuda_core
============================== 1 passed in 0.38s ===============================

xrefs:

• nvbug 6952048
• test: handle partial NVML support on Jetson Orin #2947 (comment)

[copy-pr-bot]

Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually.

Contributors can view more details about this message here.

[coderabbitai]

📝 Summary

Summary by CodeRabbit

• Tests
  • Updated NVLink checks to account for expected device-specific behavior on Orin systems. Successful queries and responses indicating that NVLink is unsupported continue to be validated as before. No user-facing functionality changed.

The NVLink test initializes the queried field's scope and return-value sentinel. When the sentinel remains and the device name contains "Orin", the test reports an expected failure. Other results continue to the existing assertion.

Changes: NVLink Test Handling

Layer / File(s)	Summary
NVLink query result handling cuda_bindings/tests/nvml/test_nvlink.py	The test initializes the field scope and return-value sentinel. It marks the specified Orin result as an expected failure and keeps the existing assertion for other results.

Suggested reviewers: mdboom

Priority: ⬇️ Low

Change: Bug fix

Merge Risk: 🔵 Low · up to 41265

The NVLink test may skip checks for other devices when it encounters an Orin. Deferring the expected failure would preserve those checks; the issue is limited to test coverage, so merge risk is low.

- [ ] Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Actionable comments posted: 1

---

ℹ️ Review info

⚙️ Run configuration

• Configuration used: Repository: NVIDIA/cuda-python/.coderabbit.yaml
• Review profile: CHILL
• Plan: Enterprise
• Run ID: 85c60964-6f71-40be-90e0-cf913a2bf0c1

📥 Commits

Reviewing files that changed from the base of the PR and between 576ef68 and 412659b.

📒 Files selected for processing: 1

• cuda_bindings/tests/nvml/test_nvlink.py

Included review availability: This review used your included allowance. Your plan provides up to 12 included reviews per hour; 11 remain after this review.

[coderabbitai]

🎯 Functional Correctness | 🟡 Minor | ⚡ Quick win

🔎 Supported by static analysis

🏁 Script executed:

set -eu
printf '%s\n' '--- reviewed head test ---'
git show 412659b856b8c1d7086bdef9351f66e3b4cecabd:cuda_bindings/tests/nvml/test_nvlink.py | nl -ba | sed -n '1,90p'
printf '%s\n' '--- base test ---'
git show 576ef68874f2fcbbb00f7f2302960c3b20597164:cuda_bindings/tests/nvml/test_nvlink.py | nl -ba | sed -n '1,90p'
printf '%s\n' '--- diff ---'
git diff --unified=30 576ef68874f2fcbbb00f7f2302960c3b20597164 412659b856b8c1d7086bdef9351f66e3b4cecabd -- cuda_bindings/tests/nvml/test_nvlink.py

Repository: NVIDIA/cuda-python

Length of output: 5440

---

important: Defer the Orin xfail only after bypassing its sentinel assertion.

When the Orin sentinel remains, moving pytest.xfail() after the loop still reaches the first return-code assertion, which rejects -1 before the alternate query and later devices run. Record the Orin condition, skip only that assertion, keep the alternate query active, and call pytest.xfail() after the loop.

Suggested fix

 def test_nvlink_get_link_count(all_devices):
     """
     Checks that the link count of the device is same.
     """
+    orin_nvml_bug = False
     for device in all_devices:
...
         value = nvml.device_get_field_values(device, fields)[0]
         if value.nvml_return == -1 and "Orin" in nvml.device_get_name(device):
-            pytest.xfail(
-                "Orin NVML nvbug 6952048: nvmlDeviceGetFieldValues returns success without populating the result"
-            )
-        assert value.nvml_return == nvml.Return.SUCCESS or value.nvml_return == nvml.Return.ERROR_NOT_SUPPORTED, (
-            f"Unexpected return {value.nvml_return} for link count field query"
-        )
+            orin_nvml_bug = True
+        else:
+            assert value.nvml_return == nvml.Return.SUCCESS or value.nvml_return == nvml.Return.ERROR_NOT_SUPPORTED, (
+                f"Unexpected return {value.nvml_return} for link count field query"
+            )
...
         assert value.nvml_return == nvml.Return.SUCCESS or value.nvml_return == nvml.Return.ERROR_NOT_SUPPORTED, (
             f"Unexpected return {value.nvml_return} for link count field query"
         )
+
+    if orin_nvml_bug:
+        pytest.xfail(
+            "Orin NVML nvbug 6952048: nvmlDeviceGetFieldValues returns success without populating the result"
+        )

[github-actions]

Doc Preview CI
🚀 View preview at https://nvidia.github.io/cuda-python/pr-preview/pr-3148/
https://nvidia.github.io/cuda-python/pr-preview/pr-3148/cuda-core/
https://nvidia.github.io/cuda-python/pr-preview/pr-3148/cuda-bindings/
https://nvidia.github.io/cuda-python/pr-preview/pr-3148/cuda-pathfinder/
Preview will be ready when the GitHub Pages deployment is complete.

[mdboom]

I don't think we should just workaround this in a test. Seems like a bug that we should protect our users from. That is, we should set an error code inside of the device_get_field_values function and populate the error back out.