cherry pick #2701, #2736, #2756, #3191#3418
Merged
ptillet merged 5 commits intotriton-lang:release/2.3.xfrom Mar 19, 2024
Merged
cherry pick #2701, #2736, #2756, #3191#3418ptillet merged 5 commits intotriton-lang:release/2.3.xfrom
ptillet merged 5 commits intotriton-lang:release/2.3.xfrom
Conversation
remove unused symbols and move driver functions to appropriate namespace
Summary: in triton.cc, enable_debug on PM will make sure debugging info is dumped when MLIR_ENABLE_DUMP is on. But for printing .ttir .ttgir, we are using the default OpPrintingFlags for Operation::print: const OpPrintingFlags &flags = std::nullopt With this patch, we pass in OpPrintingFlags with debug info enabled. Since the generation of debug info is controlled via env var: TRITON_DISABLE_LINE_INFO, it should be okay to print loc info as long as it exists. Co-authored-by: Manman Ren <mren@meta.com>
This commit cherry-pick PR triton-lang#3191 . It involves quite some manual work and is not a direct cherry pick. The reason is the original PR change ir.cc which was renamed from triton.cc by another PR triton-lang#2817 . Cherry pick PR triton-lang#2817 unfortunately cause a bunch of merge conflicts: https://gist.github.com/shunting314/bb84de294fa56d3f9e2b27fbf50f5f3b . With discussion with Oguz and Jason, we decide to do a manual cherry-pcik by applying the change to triton.cc directly. We are aware the ir module is under triton._C.libtriton.triton.ir while in latest version of triton it's under triton._C.libtriton.ir .
ptillet
added a commit
that referenced
this pull request
Apr 1, 2024
Cherry pick a few PRs to make PyTorch 2.3 work with Triton 2.2.x . Also this makes sure some python bindings needed by user defined triton kernel exists in the triton branch. --------- Co-authored-by: Philippe Tillet <phil@openai.com> Co-authored-by: Manman Ren <manman.ren@gmail.com> Co-authored-by: Manman Ren <mren@meta.com>
ptillet
added a commit
that referenced
this pull request
Apr 3, 2024
Cherry pick a few PRs to make PyTorch 2.3 work with Triton 2.2.x . Also this makes sure some python bindings needed by user defined triton kernel exists in the triton branch. --------- Co-authored-by: Philippe Tillet <phil@openai.com> Co-authored-by: Manman Ren <manman.ren@gmail.com> Co-authored-by: Manman Ren <mren@meta.com>
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.
Cherry pick a few PRs to make PyTorch 2.3 work with Triton 2.3.x . Also this makes sure some python bindings needed by user defined triton kernel exists in the triton branch.