[TRTLLM-4279] test: Add a protection test for checking trtllm custom ops #6291
Conversation
📝 Walkthrough""" WalkthroughA new test module is introduced to validate custom PyTorch operators in the "trtllm" namespace. It provides a base test class for operator discovery and setup, a decorator for conditional test execution, and a main test class with methods to check for missing fake implementations and ensure proper registration of custom ops. Additionally, two function registrations for custom operators with output tensors are removed from the Changes
Estimated code review effort🎯 2 (Simple) | ⏱️ ~15 minutes Suggested reviewers
Poem
Note ⚡️ Unit Test Generation is now available in beta!Learn more here, or try it out under "Finishing Touches" below. 📜 Recent review detailsConfiguration used: .coderabbit.yaml 📒 Files selected for processing (2)
💤 Files with no reviewable changes (1)
🧰 Additional context used🧠 Learnings (2)📓 Common learningstests/unittest/_torch/test_custom_ops.py (1)Learnt from: moraxu 🔇 Additional comments (5)
✨ Finishing Touches
🧪 Generate unit tests
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. 🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
SupportNeed help? Create a ticket on our support page for assistance with any issues or questions. Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments. CodeRabbit Commands (Invoked using PR comments)
Other keywords and placeholders
Documentation and Community
|
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🧹 Nitpick comments (5)
tests/unittest/_torch/test_custom_ops.py (5)
4-4: Replace wildcard import with specific type importsUsing
from typing import *pollutes the namespace and makes it harder to track which types are actually used.-from typing import * # noqa: F403 +from typing import List, TupleNote: Add specific types as needed when they're actually used in the code.
14-18: Add docstring to explain decorator purposeConsider adding a docstring to document why this decorator exists and when it should be used.
def requires_compile(fun): + """Skip tests that require torch.compile on Windows where it's not supported.""" fun = unittest.skipIf(IS_WINDOWS, "torch.compile doesn't work with windows")(fun) return fun
36-36: Consider using logging instead of print statementsReplace print statements with proper logging for better control over test output verbosity.
Add at the top of the file:
import logging logger = logging.getLogger(__name__)Then replace print statements:
- print(f"Total {len(ops)} custom ops in namespace {namespace}") + logger.info(f"Total {len(ops)} custom ops in namespace {namespace}")- print("Discovering custom ops:") + logger.info("Discovering custom ops:")- print(f" {op._name}") + logger.info(f" {op._name}")Also applies to: 49-49, 55-55
136-182: Consider externalizing the exemption listsThe hardcoded waiver and to-fix lists are quite large and will require frequent updates. Consider moving these to a configuration file or module-level constants for easier maintenance.
# At module level FAKE_IMPL_WAIVERS = { "trtllm::record_stream", "trtllm::wait_event", "trtllm::record_event", "trtllm::set_stream", } # Operators that need fake implementations added FAKE_IMPL_TODO = { "trtllm::lora_grouped_gemm", # ... rest of the list }Then in the test:
- waivers = { - "trtllm::record_stream", - # ... - } - - to_fix = { - "trtllm::lora_grouped_gemm", - # ... - } + waivers = FAKE_IMPL_WAIVERS + to_fix = FAKE_IMPL_TODO
199-199: Add newline at end of filePython files should end with a newline character.
if __name__ == "__main__": unittest.main() +
📜 Review details
Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (1)
tests/unittest/_torch/test_custom_ops.py(1 hunks)
🔇 Additional comments (5)
tests/unittest/_torch/test_custom_ops.py (5)
11-11: Good use of module import for side effectsThe import of
tensorrt_llmis correctly used here to ensure custom operators are registered before discovery.
40-58: Well-designed custom op discovery mechanismThe approach of using JIT schemas for discovery is appropriate for lazy-loaded C++ custom ops. The implementation correctly filters by namespace and uses the proper lookup mechanism.
95-129: Well-designed test for detecting missing fake implementationsThis test effectively validates that operators without fake implementations are properly detected by both
torch.library.opcheckandlibrary_utils.has_fake_kernel.
183-195: Clear and effective test implementationThe test logic correctly identifies operators missing fake implementations and provides a helpful error message listing the specific operators that need attention.
65-72: No changes needed for internal API usage in test teardownThe direct access to
torch._custom_op.impl.global_registryis confined to your test suite. Masking anAttributeErrorwith a try/except would only hide a legit break in teardown logic if PyTorch’s internals change. It’s preferable for tests to fail loudly so you can update them in lockstep with any upstream API adjustments.Likely an incorrect or invalid review comment.
…e impl. Signed-off-by: Yang Li <[email protected]>
Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com> Signed-off-by: Yang Li <[email protected]>
yali-arch
force-pushed
the
test_custom_ops
branch
from
add13de to
a6bd25a
Compare
|
/bot run |
|
PR_Github #13015 [ run ] triggered by Bot |
|
PR_Github #13015 [ run ] completed with state |
|
/bot run |
|
PR_Github #13666 [ run ] triggered by Bot |
|
/bot kill |
|
PR_Github #13670 [ ] completed with state |
|
PR_Github #13666 [ run ] completed with state |
2 participants
…e impl.
Summary by CodeRabbit
Description
It has been observed that the problem of missing fake implementations for new custom operations occurs from time to time, and such issues are often identified only after a certain period.
As a result, a new test file tensorrt_llm/_torch/custom_ops/torch_custom_ops.py is added to check the presence of fake implementations for each custom operation.
There are some CPP custom ops that don't have fake impl. They are tracked in the test but NOT fixed in this PR:
Test Coverage
GitHub Bot Help
/bot [-h] ['run', 'kill', 'skip', 'reuse-pipeline'] ...Provide a user friendly way for developers to interact with a Jenkins server.
Run
/bot [-h|--help]to print this help message.See details below for each supported subcommand.
Details
run [--reuse-test (optional)pipeline-id --disable-fail-fast --skip-test --stage-list "A10-PyTorch-1, xxx" --gpu-type "A30, H100_PCIe" --test-backend "pytorch, cpp" --add-multi-gpu-test --only-multi-gpu-test --disable-multi-gpu-test --post-merge --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx" --detailed-log --debug(experimental)]Launch build/test pipelines. All previously running jobs will be killed.
--reuse-test (optional)pipeline-id(OPTIONAL) : Allow the new pipeline to reuse build artifacts and skip successful test stages from a specified pipeline or the last pipeline if no pipeline-id is indicated. If the Git commit ID has changed, this option will be always ignored. The DEFAULT behavior of the bot is to reuse build artifacts and successful test results from the last pipeline.--disable-reuse-test(OPTIONAL) : Explicitly prevent the pipeline from reusing build artifacts and skipping successful test stages from a previous pipeline. Ensure that all builds and tests are run regardless of previous successes.--disable-fail-fast(OPTIONAL) : Disable fail fast on build/tests/infra failures.--skip-test(OPTIONAL) : Skip all test stages, but still run build stages, package stages and sanity check stages. Note: Does NOT update GitHub check status.--stage-list "A10-PyTorch-1, xxx"(OPTIONAL) : Only run the specified test stages. Examples: "A10-PyTorch-1, xxx". Note: Does NOT update GitHub check status.--gpu-type "A30, H100_PCIe"(OPTIONAL) : Only run the test stages on the specified GPU types. Examples: "A30, H100_PCIe". Note: Does NOT update GitHub check status.--test-backend "pytorch, cpp"(OPTIONAL) : Skip test stages which don't match the specified backends. Only support [pytorch, cpp, tensorrt, triton]. Examples: "pytorch, cpp" (does not run test stages with tensorrt or triton backend). Note: Does NOT update GitHub pipeline status.--only-multi-gpu-test(OPTIONAL) : Only run the multi-GPU tests. Note: Does NOT update GitHub check status.--disable-multi-gpu-test(OPTIONAL) : Disable the multi-GPU tests. Note: Does NOT update GitHub check status.--add-multi-gpu-test(OPTIONAL) : Force run the multi-GPU tests in addition to running L0 pre-merge pipeline.--post-merge(OPTIONAL) : Run the L0 post-merge pipeline instead of the ordinary L0 pre-merge pipeline.--extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx"(OPTIONAL) : Run the ordinary L0 pre-merge pipeline and specified test stages. Examples: --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx".--detailed-log(OPTIONAL) : Enable flushing out all logs to the Jenkins console. This will significantly increase the log volume and may slow down the job.--debug(OPTIONAL) : Experimental feature. Enable access to the CI container for debugging purpose. Note: Specify exactly one stage in thestage-listparameter to access the appropriate container environment. Note: Does NOT update GitHub check status.For guidance on mapping tests to stage names, see
docs/source/reference/ci-overview.mdand the
scripts/test_to_stage_mapping.pyhelper.kill
killKill all running builds associated with pull request.
skip
skip --comment COMMENTSkip testing for latest commit on pull request.
--comment "Reason for skipping build/test"is required. IMPORTANT NOTE: This is dangerous since lack of user care and validation can cause top of tree to break.reuse-pipeline
reuse-pipelineReuse a previous pipeline to validate current commit. This action will also kill all currently running builds associated with the pull request. IMPORTANT NOTE: This is dangerous since lack of user care and validation can cause top of tree to break.