文件最后提交记录最后更新时间
[BACKEND] Add verifiers for LocalAlloc and other memdesc ops (#4305) Add verifiers for LocalAlloc, LocalStore and CopyToLocal to make sure the mutable attribute is set consistently --------- Co-authored-by: Thomas Raoux <thomas.raoux@openai.com>1 年前
[ANALYSIS] Don't consider descending sequences as contiguous in AxisInfoAnalysis (#4871) Contiguity is used to issue wide load operations instead of multiple loads. This always assumes that the address of the first element in a sequence can be used to load the sequence of elements. If a sequence is diminishing, it leads to a wrong wide load operation. This patch fixes that by not preserving a contiguity of RHS for SubIOp operation. - [x] I am not making a trivial change, such as fixing a typo in a comment. - [x] I have written a PR description following these [rules](https://cbea.ms/git-commit/#why-not-how). - [ ] I have run pre-commit run --from-ref origin/main --to-ref HEAD. - Select one of the following. - [x] I have added tests. - /test for lit tests - /unittest for C++ tests - /python/test for end-to-end tests - [ ] This PR does not need a test because FILL THIS IN. - Select one of the following. - [x] I have not added any lit tests. - [ ] The lit tests I have added follow these [best practices](https://mlir.llvm.org/getting_started/TestingGuide/#filecheck-best-practices), including the "tests should be minimal" section. (Usually running Python code and using the instructions it generates is not minimal.) --------- Signed-off-by: Ilya Enkovich <ilya.enkovich@intel.com>1 年前
[Backend] Implement scaled_dot(mxfp4, fp8) (#4904) This PR includes https://github.com/triton-lang/triton/pull/4891 and https://github.com/triton-lang/triton/pull/4895. I will rebase once those have landed. It includes a number of hacks to work around bugs in DotOperandEncodingAttr. All these are marked as FIXME [Dot LL] to be easy to grep for. @Jokeren is working on a comprehensive revamp of DotOperandEncodingAttr which will get rid of all these. https://github.com/triton-lang/triton/pull/4895 is the first step in this direction.1 年前
[BACKEND] Fix potential bug in membar TMA rules (#4681) Also enable the special rules by default in lowering. Problem caught by @Jokeren during code review.1 年前