Skip trtllm_alltoall tests on Thor by dierksen · Pull Request #2448 · flashinfer-ai/flashinfer

dierksen · 2026-01-30T22:52:47Z

📌 Description

After adding Thor to our internal CI, we've been seeing consistent hangs in this test suite. Since this functionality likely isn't crucial for this platform, we can disable the tests to unblock the rest of the testing and get nightly Thor test results.

🔍 Related Issues

🚀 Pull Request Checklist

Thank you for contributing to FlashInfer! Before we review your pull request, please make sure the following items are complete.

✅ Pre-commit Checks

I have installed pre-commit by running pip install pre-commit (or used your preferred method).
I have installed the hooks with pre-commit install.
I have run the hooks manually with pre-commit run --all-files and fixed any reported issues.

If you are unsure about how to set up pre-commit, see the pre-commit documentation.

🧪 Tests

Tests have been added or updated as needed.
All tests are passing (unittest, etc.).

Reviewer Notes

Summary by CodeRabbit

Tests
- Updated tests to skip execution on SM110/Thor CUDA devices to prevent hangs during test runs.

_{✏️ Tip: You can customize this high-level summary in your review settings.}

Add module-level pytest skip for all tests in test_trtllm_alltoall.py when running on SM110 compute capability devices (Thor). These tests have known issues on this architecture. Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

Update comment and skip reason to explicitly state that these tests hang indefinitely on SM110 (Thor) devices rather than just having "known issues". Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

gemini-code-assist · 2026-01-30T22:52:59Z

Summary of Changes

Hello @dierksen, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request addresses a critical issue in the continuous integration pipeline where trtllm_alltoall tests were causing indefinite hangs on Thor devices. By introducing a conditional skip for these tests on the affected architecture, the PR ensures that the CI can complete successfully, allowing for nightly test results and unblocking further development and testing on the Thor platform. This change improves the reliability and efficiency of the testing infrastructure.

Highlights

Test Skipping for Thor Devices: Implemented a mechanism to skip all tests within tests/comm/test_trtllm_alltoall.py specifically for devices with compute capability SM110 (Thor).
Addressing CI Hangs: The tests were consistently hanging indefinitely on Thor devices, and skipping them unblocks the continuous integration pipeline for this platform.

🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	`/gemini review`	Performs a code review for the current pull request in its current state.
Pull Request Summary	`/gemini summary`	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	`/gemini help`	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

coderabbitai · 2026-01-30T22:53:04Z

📝 Walkthrough

Walkthrough

A test skip marker was added to conditionally bypass all tests in the module when a CUDA device with compute capability 11 (SM110/Thor) is detected, using a new import from flashinfer.utils to prevent test hangs on specific hardware.

Changes

Cohort / File(s)	Summary
Test Skip Configuration `tests/comm/test_trtllm_alltoall.py`	Added import from flashinfer.utils and introduced pytest skip marker to conditionally bypass all tests when SM110 (Thor) GPU compute capability is detected, preventing hangs on specific hardware.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~2 minutes

Possibly related PRs

flashinfer-ai/flashinfer#1998: Adds similar test-level GPU compute-capability guards to skip tests on specific SM architectures
flashinfer-ai/flashinfer#1978: Updates test skip logic based on NVIDIA GPU compute capability to handle hardware-specific compatibility issues

Suggested reviewers

yzh119

Poem

🐰 A skip marker hops with glee,
On SM110 we now run free,
No hangs will plague our testing day,
Thor's hung tests now stay away! ✨

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly and concisely describes the main change: skipping trtllm_alltoall tests on Thor devices.
Description check	✅ Passed	The description provides clear context about why tests are being skipped on Thor due to consistent hangs, but the Tests checklist items remain unchecked despite modifying test behavior.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing touches

📝 Generate docstrings

🧪 Generate unit tests (beta)

Create PR with unit tests
Post copyable unit tests in a comment

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

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

gemini-code-assist

Code Review

This pull request disables the trtllm_alltoall tests on Thor (SM110) devices where they are known to hang. The change is correct and uses pytest.mark.skipif appropriately. I've added one suggestion to improve the robustness of the test skipping logic by also handling cases where CUDA is not available, which would currently lead to test failures instead of skips.

gemini-code-assist · 2026-01-30T22:54:39Z

tests/comm/test_trtllm_alltoall.py

+# Skip all tests on SM110 (Thor) devices - these tests hang indefinitely on this architecture
+pytestmark = pytest.mark.skipif(
+    torch.cuda.is_available()
+    and get_compute_capability(torch.device("cuda:0"))[0] == 11,
+    reason="Tests hang indefinitely on SM110 (Thor) devices",
+)


The current implementation correctly skips tests on Thor devices. However, if CUDA is not available, the skipif condition evaluates to false, causing the tests to run and subsequently fail because they require CUDA. It is better practice for tests to be skipped if their environmental requirements are not met.

This suggestion refactors the logic to also skip all tests in this file if CUDA is not available. This provides a clearer result for developers running tests in a non-GPU environment and makes the skipping logic more robust and easier to read.

Suggested change

# Skip all tests on SM110 (Thor) devices - these tests hang indefinitely on this architecture

pytestmark = pytest.mark.skipif(

torch.cuda.is_available()

and get_compute_capability(torch.device("cuda:0"))[0] == 11,

reason="Tests hang indefinitely on SM110 (Thor) devices",

)

# Skip all tests on SM110 (Thor) devices and if CUDA is not available

_skip_reason = None

if not torch.cuda.is_available():

_skip_reason = "CUDA not available, skipping trtllm_alltoall tests"

elif get_compute_capability(torch.device("cuda:0"))[0] == 11:

_skip_reason = "Tests hang indefinitely on SM110 (Thor) devices"

pytestmark = pytest.mark.skipif(_skip_reason is not None, reason=_skip_reason)

CUDA will always be available when we test FlashInfer

bkryu · 2026-01-30T23:28:17Z

/bot run

flashinfer-bot · 2026-01-30T23:28:36Z

GitLab MR !282 has been created, and the CI pipeline #42938127 is currently running. I'll report back once the pipeline job completes.

flashinfer-bot · 2026-01-31T15:36:48Z

[FAILED] Pipeline #42938127: 6/20 passed

yzh119 · 2026-02-01T05:07:21Z

@flashinfer-bot run

yongwww · 2026-02-01T09:36:47Z

@flashinfer-bot rerun failed

## 📌 Description After adding Thor to our internal CI, we've been seeing consistent hangs in this test suite. Since this functionality likely isn't crucial for this platform, we can disable the tests to unblock the rest of the testing and get nightly Thor test results. ## 🔍 Related Issues  ## 🚀 Pull Request Checklist Thank you for contributing to FlashInfer! Before we review your pull request, please make sure the following items are complete. ### ✅ Pre-commit Checks - [x] I have installed `pre-commit` by running `pip install pre-commit` (or used your preferred method). - [x] I have installed the hooks with `pre-commit install`. - [x] I have run the hooks manually with `pre-commit run --all-files` and fixed any reported issues. > If you are unsure about how to set up `pre-commit`, see [the pre-commit documentation](https://pre-commit.com/). ## 🧪 Tests - [ ] Tests have been added or updated as needed. - [ ] All tests are passing (`unittest`, etc.). ## Reviewer Notes   ## Summary by CodeRabbit * **Tests** * Updated tests to skip execution on SM110/Thor CUDA devices to prevent hangs during test runs. <sub>✏️ Tip: You can customize this high-level summary in your review settings.</sub>  --------- Co-authored-by: Claude Sonnet 4.5 <noreply@anthropic.com>

dierksen and others added 2 commits January 30, 2026 16:48

Clarify that trtllm_alltoall tests hang on Thor

615aa90

Update comment and skip reason to explicitly state that these tests hang indefinitely on SM110 (Thor) devices rather than just having "known issues". Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

gemini-code-assist bot reviewed Jan 30, 2026

View reviewed changes

yzh119 approved these changes Feb 1, 2026

View reviewed changes

flashinfer-bot added the run-ci label Feb 1, 2026

yongwww merged commit 5d5e164 into flashinfer-ai:main Feb 1, 2026
62 of 67 checks passed

-# Skip all tests on SM110 (Thor) devices - these tests hang indefinitely on this architecture
-pytestmark = pytest.mark.skipif(
-    torch.cuda.is_available()
-    and get_compute_capability(torch.device("cuda:0"))[0] == 11,
-    reason="Tests hang indefinitely on SM110 (Thor) devices",
-)
+# Skip all tests on SM110 (Thor) devices and if CUDA is not available
+_skip_reason = None
+if not torch.cuda.is_available():
+    _skip_reason = "CUDA not available, skipping trtllm_alltoall tests"
+elif get_compute_capability(torch.device("cuda:0"))[0] == 11:
+    _skip_reason = "Tests hang indefinitely on SM110 (Thor) devices"
+pytestmark = pytest.mark.skipif(_skip_reason is not None, reason=_skip_reason)

Conversation

dierksen commented Jan 30, 2026 • edited by coderabbitai bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

📌 Description

🔍 Related Issues

🚀 Pull Request Checklist

✅ Pre-commit Checks

🧪 Tests

Reviewer Notes

Summary by CodeRabbit

Uh oh!

gemini-code-assist bot commented Jan 30, 2026

Summary of Changes

Highlights

Footnotes

Uh oh!

coderabbitai bot commented Jan 30, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

Estimated code review effort

Possibly related PRs

Suggested reviewers

Poem

Uh oh!

gemini-code-assist bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

gemini-code-assist bot Jan 30, 2026

Choose a reason for hiding this comment

Uh oh!

bkryu Jan 30, 2026

Choose a reason for hiding this comment

Uh oh!

bkryu commented Jan 30, 2026

Uh oh!

flashinfer-bot commented Jan 30, 2026

Uh oh!

flashinfer-bot commented Jan 31, 2026

Uh oh!

yzh119 commented Feb 1, 2026

Uh oh!

yongwww commented Feb 1, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

dierksen commented Jan 30, 2026 •

edited by coderabbitai bot

Loading

coderabbitai bot commented Jan 30, 2026 •

edited

Loading