Repository navigation
Conversation
The report-only process-group check on main lists four entries that the allowlist does not match: - NVIDIA#7394 added a second get_tensor_and_data_parallel_group() read in finalize_model_grads. - NVIDIA#7156 added a use_mpu_process_groups() fallback in SimplifiedDSGQAIndexer.__init__. - NVIDIA#4743 added a get_expert_tensor_and_model_parallel_group() fallback in DynamicInferenceContext.__init__. - NVIDIA#7047 removed the use_mpu_process_groups() call in _CudaGraphRunner._set_gtp_finalize_hook_plan, so its entry is stale. Record the three reads and drop the stale entry, so that the check only reports reads a pull request adds. The total goes from 194 to 196. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Signed-off-by: Yan Xu <yxu1@nvidia.com>
|
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. |
This was referenced Nov 1, 2026
Contributor
Author
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Why.
tools/check_process_group_usage.pyis report-only on main (#7986), so drift in its allowlist does not fail CI. On main it now reports 3 new reads and misses 1 stale entry. Every pull request that runs the check gets the same 3 lines in its log, whether or not it touches those files.get_tensor_and_data_parallel_group()read infinalize_model_grads.use_mpu_process_groups()fallback inSimplifiedDSGQAIndexer.__init__.get_expert_tensor_and_model_parallel_group()fallback inDynamicInferenceContext.__init__.use_mpu_process_groups()call in_CudaGraphRunner._set_gtp_finalize_hook_plan, so its entry is stale.What.
tools/process_group_usage_allowlist.jsonand drop the stale entry. The total goes from 194 to 196.Tests.
python tools/check_process_group_usage.py: "OK: no new global process-group reads (196 grandfathered)". On main it lists the 3 reads.python tests/unit_tests/test_check_process_group_usage.py: 29 tests pass.Part of #6307.
🤖 Generated with Claude Code