[Bugfix] Fix missing annotations for default CallNode Visitor#1659
Merged
LeiWang1999 merged 1 commit intotile-ai:mainfrom Jan 12, 2026
Merged
[Bugfix] Fix missing annotations for default CallNode Visitor#1659LeiWang1999 merged 1 commit intotile-ai:mainfrom
LeiWang1999 merged 1 commit intotile-ai:mainfrom
Conversation
|
👋 Hi! Thank you for contributing to the TileLang project. Please remember to run We appreciate you taking this step! Our team will review your contribution, and we look forward to your awesome work! 🚀 |
Contributor
📝 WalkthroughWalkthroughThis PR updates the 3rdparty/tvm submodule pointer to reference a newer commit. The change involves a single line modification to the submodule reference without any functional or behavioral impact to the codebase. Changes
Estimated code review effort🎯 1 (Trivial) | ⏱️ ~1 minute Possibly related PRs
Poem
🚥 Pre-merge checks | ✅ 2 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. 📜 Recent review detailsConfiguration used: defaults Review profile: CHILL Plan: Pro 📒 Files selected for processing (1)
⏰ 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). (2)
🔇 Additional comments (1)
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 |
tzj-fxz
pushed a commit
to wfloveiu/tilelang
that referenced
this pull request
Jan 14, 2026
LeiWang1999
added a commit
that referenced
this pull request
Jan 26, 2026
* finish KDA algorithm in tilelang * fix pre-commit.ci * fix pre-commit.ci * fix pre-commit local * [Style] Fix some code styles * [Refactor] Remove redundant swizzle for they can be automatically done * [Refactor] remove chunk_bwd_intra.py and rename chunk_bwd_intra_op.py and do some fix form coderabbitai * update ruff * update pre-commit * [Enhancement] Improve unroll loop functionality for dynamic extent and corresponding test case (#1654) * Add unroll loop functionality and corresponding test case - Introduced a new `UnrollLoop` function in the transform module to unroll loops based on various configuration options. - Added a test case in `test_tilelang_language_unroll.py` to validate the behavior of `T.unroll` with only the extent parameter, ensuring correct kernel generation with unroll pragmas. * Refactor unroll kernel implementation and update test case - Changed the kernel function in `test_tilelang_language_unroll.py` to use a new `unroll_kernel` function that compiles and returns the output tensor, improving clarity and structure. - Updated the `OptimizeForTarget` function in `phase.py` to ensure the `UnrollLoop` transformation is applied correctly, maintaining consistency in optimization phases. * lint fix * lint fix * [Bugfix] Fix missing annotations for default CallNode Visitor (#1659) tvm fix * [Clean] Remove unnecessary debug print (#1661) remove unnecessary debug print * [Bugfix] Fix variable scoping issue in InjectSoftwarePipeline for transitive LetStmt dependencies (#1657) * [Enhancement] Update global load/store functions for CUDA compatibility (#1652) Refactor the `ld_global_256` and `st_global_256` functions to support both CUDA versions above 12.9 and earlier versions. This change ensures that 256-bit loads and stores are handled correctly across different CUDA versions, improving performance and compatibility. The implementation now uses two 128-bit loads/stores for older versions, enhancing the robustness of the codebase. * Update comments in global load/store functions for CUDA compatibility Clarified comments in `ld_global_256` and `st_global_256` functions to indicate that the fallback for CUDA versions below 12.9 may have performance regressions. This change enhances code readability and provides better context for developers working with different CUDA versions. * Update submodule and enhance LetStmt handling in inject_pipeline.cc - Updated the TVM submodule to the latest commit. - Improved the handling of LetStmt in the inject_pipeline.cc file to account for transitive dependencies on loop variables, ensuring correct variable substitution in rewritten blocks. - Adjusted test_tilelang_issue_1263.py to remove unnecessary jit decorator and updated the kernel compilation process with specific pass configurations. * lint fix * revert tvm * remove unused test * test fix * [Refactor] Improve CallNode handling to include annotations in various operations (#1663) * [Enhancement] Update CallNode handling to include annotations in various operations - Modified CallNode invocations in multiple files to ensure that annotations are passed correctly, enhancing the consistency and functionality of the codebase. - Removed the "use_tma" annotation from AtomicAddNode and adjusted related calls to maintain expected behavior. - Updated CUDA intrinsic dispatch functions to include annotations, improving compatibility and correctness in CUDA operations. * lint fix * [EagerJIT] Add Support for Parameter Only Kernel Compilation (#1664) * [Fix] Refactor type hint extraction logic in DSLMutator for better clarity and handling of annotations * [Refactor] Remove redundant tensor creation in loop layout tests and update kernel compilation parameters * [AutoDD] Add Tilelang AutoDD to Reduce Buggy Program (#1639) * [Feat] Add tilelang autodd for delta debugging * fix typos * fix lint error * fix typos * fix lint error * fix bugs * Apply suggestions from code review Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * fix codeview comments * [Refactor] Move AutoDD detection to env module and update import logic * Refactor: Relocate the _is_running_autodd function to the env module for better organization and encapsulation. * Update initialization logic to skip logger and heavy imports based on a new light import mode, enhancing flexibility in module usage. * Ensure consistent handling of environment variables across the package, improving overall code clarity and maintainability. * [Documentation] Add AutoDD section to debug_tools_for_tilelang.md * Introduced a comprehensive guide on AutoDD (Automatic Delta Debugging) for isolating bugs in TileLang programs. * Explained Delta Debugging methodology, usage, parameters, and provided examples for clarity. * Highlighted the benefits of using AutoDD for large codebases and hard-to-locate errors, emphasizing time-saving aspects. * Included tips for effective usage and a reference to a complete example in the documentation. --------- Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: kurisu6912 <227995639+kurisu6912@users.noreply.github.com> Co-authored-by: LeiWang1999 <leiwang1999@outlook.com> * rebase origin * [Feature] Support `cp.reduce.async.bulk.tensor` (#1667) * support cp.reduce.async.bulk.tensor and add test * Refactor flash attention example by removing unnecessary layout annotations * support swizzle layout for tma reduce * auto swizzle for non-1d tma atomic add * upd example and test * lint * typo * add constraint for test * Refactor CUDA data type mapping by moving the to_CUtensorMapDataType function to utils.cc and utils.h, while removing redundant definitions from atomic_add.cc and copy.cc. * lint * rename basename according to CI * Update submodule TVM and remove deprecated KDA example files - Updated the TVM submodule to commit 354eef9a. - Removed several outdated KDA example files and utility scripts that are no longer in use, including chunk_bwd_dqkwg.py, chunk_bwd_dv.py, chunk_bwd_gla_dA.py, chunk_bwd_intra.py, chunk_delta_bwd.py, chunk_delta_h_fwd.py, chunk_inter_solve_fused.py, chunk_intra_token_parallel.py, chunk_o.py, README.md, test_utils_kda.py, wy_fast_bwd.py, wy_fast.py, and various FLA_KDA implementations. * lint fix --------- Co-authored-by: wufang <wufang@MBP-MK6VR66Y2M-2329.local> Co-authored-by: tzj-fxz <tzjfxz@gmail.com> Co-authored-by: Lei Wang <34334180+LeiWang1999@users.noreply.github.com> Co-authored-by: Kuris <227995639+kurisu6912@users.noreply.github.com> Co-authored-by: Kexing Zhou <KEKE_046@pku.edu.cn> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: LeiWang1999 <leiwang1999@outlook.com> Co-authored-by: Zhengju Tang <97930865+tzj-fxz@users.noreply.github.com> Co-authored-by: Tong WU <109033598+Rachmanino@users.noreply.github.com>
3 tasks
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
1 participant
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.
This pull request updates the TVM submodule reference to a newer commit, ensuring that the project uses the latest version of TVM and any associated improvements or bug fixes.
3rdparty/tvmsubmodule to point to commitda7f19b6908045a1f9bf94cb7e044beaa32421b6instead of the previous commit, bringing in upstream changes from TVM.Summary by CodeRabbit
✏️ Tip: You can customize this high-level summary in your review settings.