[None][fix] Let KV cache manager use single stream for cache block transfer (onboard/offload)#9560
[None][fix] Let KV cache manager use single stream for cache block transfer (onboard/offload)#9560eopXD wants to merge 3 commits intoNVIDIA:mainfrom
Conversation
eopXD
changed the title
eopXD
changed the title
📝 WalkthroughWalkthroughThese changes simplify KV cache transfer management by removing explicit onboard and offload buffer managers and their associated pending offload tracking. The implementation is refactored to use full-block copies through a single buffer manager, reuse existing CUDA streams instead of allocating new ones, and eliminate complex synchronization logic. Changes
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~20 minutes
Pre-merge checks and finishing touches❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✨ Finishing touches
🧪 Generate unit tests (beta)
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. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 0
🧹 Nitpick comments (1)
cpp/tensorrt_llm/batch_manager/kvCacheTransferManager.cpp (1)
98-100: Consider removing unusednumTokensToCopyparameter.After the simplification to full-block copies,
numTokensToCopyis no longer used in any code path (DRAM, POSIX, or GDS). Consider removing this parameter fromcopyBlock,onboard, andoffloadfunction signatures in a follow-up cleanup.
📜 Review details
Configuration used: Path: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (5)
cpp/include/tensorrt_llm/batch_manager/kvCacheTransferManager.h(0 hunks)cpp/tensorrt_llm/batch_manager/kvCacheTransferManager.cpp(2 hunks)tensorrt_llm/_torch/pyexecutor/resource_manager.py(1 hunks)tests/integration/defs/accuracy/test_llm_api_pytorch.py(0 hunks)tests/integration/test_lists/test-db/l0_h100.yml(1 hunks)
💤 Files with no reviewable changes (2)
- tests/integration/defs/accuracy/test_llm_api_pytorch.py
- cpp/include/tensorrt_llm/batch_manager/kvCacheTransferManager.h
🧰 Additional context used
📓 Path-based instructions (3)
**/*.{cpp,h,cu}
📄 CodeRabbit inference engine (CODING_GUIDELINES.md)
**/*.{cpp,h,cu}: Closing braces of namespaces should have a comment saying the namespace it closes (e.g.,} // namespace foo)
Preferconstorconstexprvariables over#definewhenever possible, as the latter are not visible to the compiler
A variable that is not modified after its initialization should be declared asconst
Except0(only used in comparison for checking signness/existence/emptiness) andnullptr,true,false, all other literals should only be used for variable initialization and should be replaced with named constants
Use Allman indentation style for braces in C++
Put the semicolon for an emptyfororwhileloop in a new line
The statement forming the body of aswitch,while,do .. whileorforstatement shall be a compound statement (use brace-delimited statements)
Ifandelseshould always be followed by brace-delimited statements, even if empty or a single statement
C++ filenames should use camel case with first letter lowercase (e.g.,thisIsASubDirandthisIsAFilename.cpp)
All filenames involved in compilation of a compilation target must have case-insensitive unique filenames
All types (including class names) should use camel case with uppercase first letter (e.g.,FooBarClass)
Local variables, methods and namespaces should use camel case with first letter lowercase (e.g.,localFooBar)
Non-magic-number global variables that are non-static and not defined in anonymous namespace should use camel case prefixed by a lower case 'g' (e.g.,gDontUseGlobalFoos)
Non-magic-number global variables that are static or defined in an anonymous namespace should use camel case prefixed by a lower case 's' (e.g.,sMutableStaticGlobal)
Locally visible static variables should use camel case with lowercase prefix 's' as the first letter of the name (e.g.,static std::once_flag sFlag;)
Public, private and protected class member variables should use camel case prefixed with 'm' (e.g.,mNbFooValues), though the 'm' pre...
Files:
cpp/tensorrt_llm/batch_manager/kvCacheTransferManager.cpp
**/*.{cpp,h,cu,py}
📄 CodeRabbit inference engine (CODING_GUIDELINES.md)
All TensorRT-LLM Open Source Software code files should contain an NVIDIA copyright header that includes the current year at the top
Files:
cpp/tensorrt_llm/batch_manager/kvCacheTransferManager.cpptensorrt_llm/_torch/pyexecutor/resource_manager.py
**/*.py
📄 CodeRabbit inference engine (CODING_GUIDELINES.md)
**/*.py: The code developed for TensorRT-LLM should conform to Python 3.8+
Indent Python code with 4 spaces; do not use tabs
Always maintain the namespace when importing in Python, even if only one class or function from a module is used (e.g., usefrom package.subpackage import fooand thenfoo.SomeClass()instead offrom package.subpackage.foo import SomeClass)
Python filenames should use snake_case (e.g.,some_file.py)
Python class names should use PascalCase (e.g.,class SomeClass)
Python function and method names should use snake_case (e.g.,def my_awesome_function():)
Python local variable names should use snake_case, with prefixkfor variable names that start with a number (e.g.,k_99th_percentile = ...)
Python global variables should use upper snake_case with prefixG(e.g.,G_MY_GLOBAL = ...)
Python constants should use upper snake_case (e.g.,MY_CONSTANT = ...)
Avoid shadowing variables declared in an outer scope in Python
Initialize all externally visible members of a Python class in the constructor
For Python interfaces that may be used outside a file, prefer docstrings over comments
Python comments should be reserved for code within a function, or interfaces that are local to a file
Use Google style docstrings for Python classes and functions, which can be parsed by Sphinx
Python attributes and variables can be documented inline with type and description (e.g.,self.x = 5followed by"""<type>: Description of 'x'""")
Avoid using reflection in Python when functionality can be easily achieved without reflection
When using try-except blocks in Python, limit the except clause to the smallest set of specific errors possible instead of catching all exceptions
When using try-except blocks in Python to handle multiple possible variable types (duck-typing), keep the body of the try as small as possible and use the else block to implement the logic
Files:
tensorrt_llm/_torch/pyexecutor/resource_manager.py
🧠 Learnings (16)
📓 Common learnings
Learnt from: eopXD
Repo: NVIDIA/TensorRT-LLM PR: 6768
File: cpp/include/tensorrt_llm/batch_manager/kvCacheManager.h:0-0
Timestamp: 2025-08-20T06:48:45.368Z
Learning: There is a planned refactoring to move cache block bookkeeping utilities from BlockManager/WindowBlockManager into the GenerationRequest class itself to improve code organization and make responsibilities clearer.
Learnt from: eopXD
Repo: NVIDIA/TensorRT-LLM PR: 6768
File: cpp/tensorrt_llm/batch_manager/kvCacheManager.cpp:2010-2045
Timestamp: 2025-08-21T09:41:49.347Z
Learning: In cpp/tensorrt_llm/batch_manager/kvCacheManager.cpp, updateSequenceCacheBlockOffsets is specifically for updating bookkeeping when blocks are added during the context phase, not for refreshing offsets after detach operations. During detach operations, GenerationRequest::removeFrontBlock handles the necessary cache block bookkeeping internally.
Learnt from: thorjohnsen
Repo: NVIDIA/TensorRT-LLM PR: 6910
File: cpp/tensorrt_llm/batch_manager/kvCacheManager.cpp:0-0
Timestamp: 2025-08-14T21:04:50.248Z
Learning: In KV cache onboarding logic during prefill in cpp/tensorrt_llm/batch_manager/kvCacheManager.cpp, when calculating which blocks fall within the attention window, use getTokensPerBlock() to advance token indices rather than block->getUniqueTokens().size(), because the calculation needs to consider the post-prefill state where blocks will be filled to capacity, not their current token count.
Learnt from: eopXD
Repo: NVIDIA/TensorRT-LLM PR: 6768
File: cpp/include/tensorrt_llm/batch_manager/kvCacheManager.h:0-0
Timestamp: 2025-08-20T06:48:45.368Z
Learning: In cpp/tensorrt_llm/batch_manager/kvCacheManager.cpp, updateSequenceCacheBlockOffsets is only called when adding a sequence, not during detach operations. During detach, the cache block bookkeeping is handled by GenerationRequest::removeFrontBlock.
📚 Learning: 2025-09-09T09:40:45.658Z
Learnt from: fredricz-20070104
Repo: NVIDIA/TensorRT-LLM PR: 7645
File: tests/integration/test_lists/qa/llm_function_core.txt:648-648
Timestamp: 2025-09-09T09:40:45.658Z
Learning: In TensorRT-LLM test lists, it's common and intentional for the same test to appear in multiple test list files when they serve different purposes (e.g., llm_function_core.txt for comprehensive core functionality testing and llm_function_core_sanity.txt for quick sanity checks). This duplication allows tests to be run in different testing contexts.
Applied to files:
tests/integration/test_lists/test-db/l0_h100.yml
📚 Learning: 2025-09-17T02:48:52.732Z
Learnt from: tongyuantongyu
Repo: NVIDIA/TensorRT-LLM PR: 7781
File: tests/integration/test_lists/waives.txt:313-313
Timestamp: 2025-09-17T02:48:52.732Z
Learning: In TensorRT-LLM, `tests/integration/test_lists/waives.txt` is specifically for waiving/skipping tests, while other test list files like those in `test-db/` and `qa/` directories are for different test execution contexts (pre-merge, post-merge, QA tests). The same test appearing in both waives.txt and execution list files is intentional - the test is part of test suites but will be skipped due to the waiver.
Applied to files:
tests/integration/test_lists/test-db/l0_h100.yml
📚 Learning: 2025-07-28T17:06:08.621Z
Learnt from: moraxu
Repo: NVIDIA/TensorRT-LLM PR: 6303
File: tests/integration/test_lists/qa/examples_test_list.txt:494-494
Timestamp: 2025-07-28T17:06:08.621Z
Learning: In TensorRT-LLM testing, it's common to have both CLI flow tests (test_cli_flow.py) and PyTorch API tests (test_llm_api_pytorch.py) for the same model. These serve different purposes: CLI flow tests validate the traditional command-line workflow, while PyTorch API tests validate the newer LLM API backend. Both are legitimate and should coexist.
Applied to files:
tests/integration/test_lists/test-db/l0_h100.yml
📚 Learning: 2025-08-26T09:49:04.956Z
Learnt from: pengbowang-nv
Repo: NVIDIA/TensorRT-LLM PR: 7192
File: tests/integration/test_lists/test-db/l0_dgx_b200.yml:56-72
Timestamp: 2025-08-26T09:49:04.956Z
Learning: In TensorRT-LLM test configuration files, the test scheduling system handles wildcard matching with special rules that prevent duplicate test execution even when the same tests appear in multiple yaml files with overlapping GPU wildcards (e.g., "*b200*" and "*gb200*").
Applied to files:
tests/integration/test_lists/test-db/l0_h100.yml
📚 Learning: 2025-08-13T11:07:11.772Z
Learnt from: Funatiq
Repo: NVIDIA/TensorRT-LLM PR: 6754
File: tests/integration/test_lists/test-db/l0_a30.yml:41-47
Timestamp: 2025-08-13T11:07:11.772Z
Learning: In TensorRT-LLM test configuration files like tests/integration/test_lists/test-db/l0_a30.yml, TIMEOUT values are specified in minutes, not seconds.
Applied to files:
tests/integration/test_lists/test-db/l0_h100.yml
📚 Learning: 2025-08-29T14:07:45.863Z
Learnt from: EmmaQiaoCh
Repo: NVIDIA/TensorRT-LLM PR: 7370
File: tests/unittest/trt/model_api/test_model_quantization.py:24-27
Timestamp: 2025-08-29T14:07:45.863Z
Learning: In TensorRT-LLM's CI infrastructure, pytest skip markers (pytest.mark.skip) are properly honored even when test files have __main__ blocks that call test functions directly. The testing system correctly skips tests without requiring modifications to the __main__ block execution pattern.
Applied to files:
tests/integration/test_lists/test-db/l0_h100.yml
📚 Learning: 2025-08-21T09:41:49.347Z
Learnt from: eopXD
Repo: NVIDIA/TensorRT-LLM PR: 6768
File: cpp/tensorrt_llm/batch_manager/kvCacheManager.cpp:2010-2045
Timestamp: 2025-08-21T09:41:49.347Z
Learning: In cpp/tensorrt_llm/batch_manager/kvCacheManager.cpp, updateSequenceCacheBlockOffsets is specifically for updating bookkeeping when blocks are added during the context phase, not for refreshing offsets after detach operations. During detach operations, GenerationRequest::removeFrontBlock handles the necessary cache block bookkeeping internally.
Applied to files:
cpp/tensorrt_llm/batch_manager/kvCacheTransferManager.cpp
📚 Learning: 2025-08-20T06:48:45.368Z
Learnt from: eopXD
Repo: NVIDIA/TensorRT-LLM PR: 6768
File: cpp/include/tensorrt_llm/batch_manager/kvCacheManager.h:0-0
Timestamp: 2025-08-20T06:48:45.368Z
Learning: In cpp/tensorrt_llm/batch_manager/kvCacheManager.cpp, updateSequenceCacheBlockOffsets is only called when adding a sequence, not during detach operations. During detach, the cache block bookkeeping is handled by GenerationRequest::removeFrontBlock.
Applied to files:
cpp/tensorrt_llm/batch_manager/kvCacheTransferManager.cpp
📚 Learning: 2025-08-15T06:46:54.897Z
Learnt from: eopXD
Repo: NVIDIA/TensorRT-LLM PR: 6767
File: cpp/tensorrt_llm/batch_manager/kvCacheManager.cpp:0-0
Timestamp: 2025-08-15T06:46:54.897Z
Learning: In cpp/tensorrt_llm/batch_manager/kvCacheManager.cpp addToken function, newly allocated blocks are unshared by design. The beam search path in addToken (when sequence.getNumTokens() > windowSize) is currently broken/non-functional with SWA, so the block allocation doesn't follow a shared-then-unshared pattern.
Applied to files:
cpp/tensorrt_llm/batch_manager/kvCacheTransferManager.cpp
📚 Learning: 2025-08-14T21:04:50.248Z
Learnt from: thorjohnsen
Repo: NVIDIA/TensorRT-LLM PR: 6910
File: cpp/tensorrt_llm/batch_manager/kvCacheManager.cpp:0-0
Timestamp: 2025-08-14T21:04:50.248Z
Learning: In KV cache onboarding logic during prefill in cpp/tensorrt_llm/batch_manager/kvCacheManager.cpp, when calculating which blocks fall within the attention window, use getTokensPerBlock() to advance token indices rather than block->getUniqueTokens().size(), because the calculation needs to consider the post-prefill state where blocks will be filled to capacity, not their current token count.
Applied to files:
cpp/tensorrt_llm/batch_manager/kvCacheTransferManager.cpp
📚 Learning: 2025-09-23T15:01:00.070Z
Learnt from: nv-lschneider
Repo: NVIDIA/TensorRT-LLM PR: 7910
File: cpp/tensorrt_llm/kernels/nccl_device/config.cu:15-17
Timestamp: 2025-09-23T15:01:00.070Z
Learning: In TensorRT-LLM NCCL device kernels, the <sstream> header is not needed as an explicit include in config.cu because it's provided transitively through other headers. Local compilation testing confirms this works without the explicit include.
Applied to files:
cpp/tensorrt_llm/batch_manager/kvCacheTransferManager.cpp
📚 Learning: 2025-08-06T08:18:28.669Z
Learnt from: zhengd-nv
Repo: NVIDIA/TensorRT-LLM PR: 6633
File: cpp/tensorrt_llm/batch_manager/dataTransceiverImpl.cpp:145-155
Timestamp: 2025-08-06T08:18:28.669Z
Learning: In cpp/tensorrt_llm/batch_manager/dataTransceiverImpl.cpp, the existing `mMtxForMap` mutex in DataSenderImpl is sufficient to synchronize measurement file operations in the `release` method, as all file operations occur within the same critical section that protects the `mRequestToSession` map access.
Applied to files:
cpp/tensorrt_llm/batch_manager/kvCacheTransferManager.cpp
📚 Learning: 2025-08-20T06:56:02.889Z
Learnt from: eopXD
Repo: NVIDIA/TensorRT-LLM PR: 6768
File: cpp/tensorrt_llm/batch_manager/kvCacheManager.cpp:577-579
Timestamp: 2025-08-20T06:56:02.889Z
Learning: In cpp/tensorrt_llm/batch_manager/kvCacheManager.cpp, maxSequenceLength is now enforced as a non-optional argument in the BlockManager constructor, so concerns about std::nullopt defaulting to 0 are not applicable. When windowSize > maxSequenceLength, a warning should be added instead of handling optional parameter cases.
Applied to files:
cpp/tensorrt_llm/batch_manager/kvCacheTransferManager.cpp
📚 Learning: 2025-09-23T14:58:05.372Z
Learnt from: nv-lschneider
Repo: NVIDIA/TensorRT-LLM PR: 7910
File: cpp/tensorrt_llm/kernels/nccl_device/config.cu:42-49
Timestamp: 2025-09-23T14:58:05.372Z
Learning: In TensorRT-LLM NCCL device kernels (cpp/tensorrt_llm/kernels/nccl_device/), the token partitioning intentionally uses ceil-like distribution (same token_per_rank for all ranks) to ensure all ranks launch the same number of blocks. This is required for optimal NCCL device API barrier performance, even though it may launch extra blocks for non-existent tokens on later ranks. Runtime bounds checking in the kernel (blockID validation) handles the overshoot cases.
Applied to files:
cpp/tensorrt_llm/batch_manager/kvCacheTransferManager.cpp
📚 Learning: 2025-08-17T15:07:01.420Z
Learnt from: amitz-nv
Repo: NVIDIA/TensorRT-LLM PR: 6968
File: cpp/tensorrt_llm/thop/loraOp.cpp:133-141
Timestamp: 2025-08-17T15:07:01.420Z
Learning: In TensorRT-LLM's LoRA implementation, the LoraImpl::run() method handles setStream() internally in _runGemm() (line 51 in lora.cpp), along with setWorkspace(). The stream parameter flows from loraOp.cpp through LoraImpl::run() to _runGemm() where setStream() is called appropriately. Adding setStream() in loraOp.cpp would be redundant and goes against the intended architectural design.
Applied to files:
tensorrt_llm/_torch/pyexecutor/resource_manager.py
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (1)
- GitHub Check: Pre-commit Check
🔇 Additional comments (4)
tests/integration/test_lists/test-db/l0_h100.yml (1)
49-49: LGTM!The test addition aligns with the PR objective to re-enable accuracy tests that previously failed. The placement is consistent with the existing VSWA test grouping, and the single onboard/offload stream changes should ensure this test passes.
tensorrt_llm/_torch/pyexecutor/resource_manager.py (1)
354-356: Simplification to use current CUDA stream aligns with PR objectives.The change from creating a new dedicated stream to reusing
torch.cuda.current_stream()supports the PR's goal of achieving functional correctness via a single default stream. Since the comment notes this stream is currently unused, the change has no immediate behavioral impact but correctly sets up for future host copy features.cpp/tensorrt_llm/batch_manager/kvCacheTransferManager.cpp (2)
109-116: Simplified full-block copy for DRAM path looks correct.The simplification to always copy the entire block via
mBufferManager.copy()aligns with the PR's objective of using a single stream for correctness. The buffer manager handles the copy operation on its associated stream, ensuring proper ordering.
192-192: EmptysyncTransfers()is correct for single-stream architecture.With the shift to using a single default CUDA stream for all onboard/offload operations, explicit synchronization between streams is no longer needed. GPU operations on the same stream are implicitly ordered, making this empty implementation correct.
|
/bot run --disable-fail-fast |
|
PR_Github #26307 [ run ] triggered by Bot. Commit: |
|
PR_Github #26307 [ run ] completed with state |
|
/bot run --disable-fail-fast |
|
PR_Github #26310 [ run ] triggered by Bot. Commit: |
|
PR_Github #26310 [ run ] completed with state |
Functionally, this PR looks good to me. But if we use the same stream for onboard/offload, that means the onboard/offload can no longer overlap. Are we worried about the performance drop caused by this? Or is this just a stopgap fix for functionality correctness and we will enable overlap some time in the future? @eopXD |
|
I don't think we are too worried about the performance drop on this. The secondary blocks were broken back then and we rarely see complaints, we don't have much performance user on onboard/offload for now. So I think its a good direction to take the first step by making it right first. |
|
I agree that we should pass the stream that is actually running the kernels to KV cache manager instead of an unused stream, that was always the intended use of the stream argument. However, I don't see why removing onboard and offload streams is necessary, since those two streams already wait for the buffer manager stream. Are there tests that pass with just one stream but fails otherwise? If we pass the right stream to KV cache manager (presumably torch.cuda.current_stream()), but otherwise leave everything as-is, do those tests pass? |
| # Note that this stream is unused for now. Will be used for copying to host | ||
| # when that feature is enabled. | ||
| self._stream = torch.cuda.Stream() | ||
| self._stream = torch.cuda.current_stream() |
There was a problem hiding this comment.
Good find. Passing an unused stream to KV cache manager will prevent any synchronization between kernels and block copies. Are you sure torch.cuda.current_stream() is the right stream? Presumably this method is called once during setup, torch.cuda.current_stream() is probably the default stream at that point. Is that what the kernels are run on, and how sure are we that this won't change?
There was a problem hiding this comment.
Yes, the sound solution is to be aware of all streams afterward, or the compiler itself has a main stream where all stream respects. Let me investigate more on this.
eopXD
force-pushed
the
single-onboard-offload-stream
branch
from
40453d9 to
e7ede33
Compare
Signed-off-by: eopXD <yuehtingc@nvidia.com>
…pend on Signed-off-by: eopXD <yuehtingc@nvidia.com>
Signed-off-by: eopXD <yuehtingc@nvidia.com>
eopXD
force-pushed
the
single-onboard-offload-stream
branch
from
e7ede33 to
9538d2f
Compare
Passing |
|
Dropping the MR, stream sychronization is indeed a problem. We have @SimengLiu-nv to work on that in #10060. Will create another MR that fixes accuracy drop issue. |
4 participants
Summary by CodeRabbit
Refactor
Tests
✏️ Tip: You can customize this high-level summary in your review settings.
Description
This MR aims to achieve functional correctness first for block copy. We leverage the single default stream for onboard/offload action.
The merge request has 3 parts of change:
current_streamas the stream for onboard/offload. Ideally we should now what are the streams that are going to be triggered ahead, or in another way, we should let onboard/offload register on default stream, and ideally future event execution that are kv cache block related should also respect events that are already registered to default stream.Test Coverage
Test coverage are enabled that was originally unable to pass due to accuracy drop. With correct block movement, the test shall pass now.
PR Checklist
Please review the following before submitting your PR:
PR description clearly explains what and why. If using CodeRabbit's summary, please make sure it makes sense.
PR Follows TRT-LLM CODING GUIDELINES to the best of your knowledge.
Test cases are provided for new code paths (see test instructions)
Any new dependencies have been scanned for license and vulnerabilities
CODEOWNERS updated if ownership changes
Documentation updated as needed
Update tava architecture diagram if there is a significant design change in PR.
The reviewers assigned automatically/manually are appropriate for the PR.
Please check this after reviewing the above items as appropriate for this PR.
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.