[TUTORIAL] Add the non-persistent softmax and make it for CPU #22
Conversation
This file contains 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
|
Thanks, Minjang! Interesting numbers. I'm surprised that TritonCPU is quite close to the native torch without any optimizations added for the CPU so far. Can we also have a torch-inductor column in addition to torch-native and torch-jit on CPU? For GPU it's probably not relevant because it's also Triton based but for CPU it's something we want to be compared to. |
minjang
force-pushed
the
add-fused-softmax-cpu
branch
from
6ceba1b to
2e92d57
Compare
|
(Sorry for the late update. I was busy for handling internal oncall issues.) Added We still see some big drops, but this looks a good starting baseline.
matmul-performance-fp32 (BLOCK_SIZE_M=32, BLOCK_SIZE_N=32, BLOCK_SIZE_K=32, GROUP_SIZE_M=8).png |
ienkovich
approved these changes
Jun 20, 2024
minjang
added a commit
to minjang/triton-cpu
that referenced
this pull request
Jun 22, 2024
…-lang#22) * [TUTORIAL] Add 02-fused-softmax with the previous non-persistent implementation * Add torch.compile cases * Preallocate output buffer for softmax tutorial
minjang
pushed a commit
that referenced
this pull request
Jun 24, 2024
When running [convert_blocked1d_to_slice0](https://github.com/triton-lang/triton/blob/0ba5f0c3cd029d5c3d1f01b9bf29dac32c27345e/test/Conversion/tritongpu_to_llvm.mlir#L924) Triton ends up computing a rank of a matrix with 0 columns during linear layout lowering, which trips up f2reduce, and causes undefined behavior, detectable through [UBSAN](https://clang.llvm.org/docs/UndefinedBehaviorSanitizer.html). Fix this by returning the rank (0) early in these cases, without calling f2reduce. <details><summary>Stack trace</summary> <p> ``` third_party/triton/third_party/f2reduce/f2reduce.cpp:421:30: runtime error: shift exponent 18446744073709551615 is too large for 64-bit type 'unsigned long long' #0 0x556ee2fea3be in inplace_rref_small third_party/triton/third_party/f2reduce/f2reduce.cpp:421:30 #1 0x556ee2fea3be in f2reduce::inplace_rref_strided(unsigned long*, unsigned long, unsigned long, unsigned long) third_party/triton/third_party/f2reduce/f2reduce.cpp:470:9 #2 0x556ee2ea70da in getMatrixRank third_party/triton/lib/Tools/LinearLayout.cpp:125:3 #3 0x556ee2ea70da in mlir::triton::LinearLayout::checkInvariants(bool) third_party/triton/lib/Tools/LinearLayout.cpp:299:7 #4 0x556ee2ea656d in mlir::triton::LinearLayout::tryCreate(llvm::MapVector<mlir::StringAttr, std::__u::vector<std::__u::vector<int, std::__u::allocator<int>>, std::__u::allocator<std::__u::vector<int, std::__u::allocator<int>>>>, llvm::DenseMap<mlir::StringAttr, unsigned int, llvm::DenseMapInfo<mlir::StringAttr, void>, llvm::detail::DenseMapPair<mlir::StringAttr, unsigned int>>, llvm::SmallVector<std::__u::pair<mlir::StringAttr, std::__u::vector<std::__u::vector<int, std::__u::allocator<int>>, std::__u::allocator<std::__u::vector<int, std::__u::allocator<int>>>>>, 0u>>, llvm::ArrayRef<std::__u::pair<mlir::StringAttr, int>>, bool) third_party/triton/lib/Tools/LinearLayout.cpp:190:41 #5 0x556ee2eb2150 in mlir::triton::LinearLayout::divideRight(mlir::triton::LinearLayout const&) third_party/triton/lib/Tools/LinearLayout.cpp:654:51 #6 0x556ee2ee1c39 in mlir::cvtNeedsSharedMemory(mlir::RankedTensorType, mlir::RankedTensorType) third_party/triton/lib/Analysis/Utility.cpp:652:14 #7 0x556ee2cf38fd in mlir::triton::getRepShapeForCvtLayout(mlir::triton::gpu::ConvertLayoutOp) third_party/triton/lib/Analysis/Allocation.cpp:66:8 #8 0x556ee2cf3efa in mlir::triton::getScratchConfigForCvtLayout(mlir::triton::gpu::ConvertLayoutOp, unsigned int&, unsigned int&) third_party/triton/lib/Analysis/Allocation.cpp:95:19 #9 0x556ee2cf6057 in mlir::triton::AllocationAnalysis::getScratchValueSize(mlir::Operation*) third_party/triton/lib/Analysis/Allocation.cpp:272:24 #10 0x556ee2cf5499 in operator() third_party/triton/lib/Analysis/Allocation.cpp:343:7 #11 0x556ee2cf5499 in void llvm::function_ref<void (mlir::Operation*)>::callback_fn<mlir::triton::AllocationAnalysis::getValuesAndSizes()::'lambda'(mlir::Operation*)>(long, mlir::Operation*) third_party/llvm/llvm-project/llvm/include/llvm/ADT/STLFunctionalExtras.h:45:12 #12 0x556edeeee7a9 in operator() third_party/llvm/llvm-project/llvm/include/llvm/ADT/STLFunctionalExtras.h:68:12 #13 0x556edeeee7a9 in void mlir::detail::walk<mlir::ForwardIterator>(mlir::Operation*, llvm::function_ref<void (mlir::Operation*)>, mlir::WalkOrder) third_party/llvm/llvm-project/mlir/include/mlir/IR/Visitors.h:174:5 #14 0x556edeeee87c in void mlir::detail::walk<mlir::ForwardIterator>(mlir::Operation*, llvm::function_ref<void (mlir::Operation*)>, mlir::WalkOrder) third_party/llvm/llvm-project/mlir/include/mlir/IR/Visitors.h:182:9 #15 0x556ee2cf49e7 in walk<(mlir::WalkOrder)0, mlir::ForwardIterator, (lambda at third_party/triton/lib/Analysis/Allocation.cpp:341:42), mlir::Operation *, void> third_party/llvm/llvm-project/mlir/include/mlir/IR/Visitors.h:313:10 #16 0x556ee2cf49e7 in walk<(mlir::WalkOrder)0, mlir::ForwardIterator, (lambda at third_party/triton/lib/Analysis/Allocation.cpp:341:42), void> third_party/llvm/llvm-project/mlir/include/mlir/IR/Operation.h:794:12 #17 0x556ee2cf49e7 in mlir::triton::AllocationAnalysis::getValuesAndSizes() third_party/triton/lib/Analysis/Allocation.cpp:341:16 #18 0x556ee2cf4852 in run third_party/triton/lib/Analysis/Allocation.cpp:182:5 #19 0x556ee2cf4852 in AllocationAnalysis third_party/triton/lib/Analysis/Allocation.cpp:169:5 #20 0x556ee2cf4852 in mlir::Allocation::run(llvm::DenseMap<mlir::FunctionOpInterface, mlir::Allocation, llvm::DenseMapInfo<mlir::FunctionOpInterface, void>, llvm::detail::DenseMapPair<mlir::FunctionOpInterface, mlir::Allocation>>&) third_party/triton/lib/Analysis/Allocation.cpp:627:3 #21 0x556ee1677402 in operator() third_party/triton/include/triton/Analysis/Allocation.h:227:26 #22 0x556ee1677402 in void mlir::CallGraph<mlir::Allocation>::doWalk<(mlir::WalkOrder)0, (mlir::WalkOrder)1, mlir::ModuleAllocation::ModuleAllocation(mlir::ModuleOp)::'lambda'(mlir::CallOpInterface, mlir::FunctionOpInterface), mlir::ModuleAllocation::ModuleAllocation(mlir::ModuleOp)::'lambda'(mlir::FunctionOpInterface)>(mlir::FunctionOpInterface, llvm::DenseSet<mlir::FunctionOpInterface, llvm::DenseMapInfo<mlir::FunctionOpInterface, void>>&, mlir::ModuleAllocation::ModuleAllocation(mlir::ModuleOp)::'lambda'(mlir::CallOpInterface, mlir::FunctionOpInterface), mlir::ModuleAllocation::ModuleAllocation(mlir::ModuleOp)::'lambda'(mlir::FunctionOpInterface)) third_party/triton/include/triton/Analysis/Utility.h:350:7 #23 0x556ee16756b3 in walk<(mlir::WalkOrder)0, (mlir::WalkOrder)1, (lambda at third_party/triton/include/triton/Analysis/Allocation.h:222:9), (lambda at third_party/triton/include/triton/Analysis/Allocation.h:224:9)> third_party/triton/include/triton/Analysis/Utility.h:242:7 #24 0x556ee16756b3 in mlir::ModuleAllocation::ModuleAllocation(mlir::ModuleOp) third_party/triton/include/triton/Analysis/Allocation.h:220:5 #25 0x556ee2c2bf18 in (anonymous namespace)::AllocateSharedMemory::runOnOperation() third_party/triton/lib/Conversion/TritonGPUToLLVM/AllocateSharedMemory.cpp:26:22 ... UndefinedBehaviorSanitizer: invalid-shift-exponent third_party/triton/third_party/f2reduce/f2reduce.cpp:421:30 ``` </p> </details>
minjang
added a commit
that referenced
this pull request
Jun 24, 2024
* [TUTORIAL] Add 02-fused-softmax with the previous non-persistent implementation * Add torch.compile cases * Preallocate output buffer for softmax tutorial
Devjiu
pushed a commit
to Devjiu/triton-cpu
that referenced
this pull request
Aug 13, 2024
…-lang#22) * [TUTORIAL] Add 02-fused-softmax with the previous non-persistent implementation * Add torch.compile cases * Preallocate output buffer for softmax tutorial
int3
pushed a commit
that referenced
this pull request
Aug 29, 2024
* [TUTORIAL] Add 02-fused-softmax with the previous non-persistent implementation * Add torch.compile cases * Preallocate output buffer for softmax tutorial
minjang
added a commit
that referenced
this pull request
Sep 22, 2024
* [TUTORIAL] Add 02-fused-softmax with the previous non-persistent implementation * Add torch.compile cases * Preallocate output buffer for softmax tutorial
minjang
added a commit
that referenced
this pull request
Oct 22, 2024
* [TUTORIAL] Add 02-fused-softmax with the previous non-persistent implementation * Add torch.compile cases * Preallocate output buffer for softmax tutorial
minjang
added a commit
that referenced
this pull request
Oct 24, 2024
* [TUTORIAL] Add 02-fused-softmax with the previous non-persistent implementation * Add torch.compile cases * Preallocate output buffer for softmax tutorial
int3
pushed a commit
that referenced
this pull request
Dec 6, 2024
* [TUTORIAL] Add 02-fused-softmax with the previous non-persistent implementation * Add torch.compile cases * Preallocate output buffer for softmax tutorial
ienkovich
pushed a commit
that referenced
this pull request
Dec 6, 2024
* [TUTORIAL] Add 02-fused-softmax with the previous non-persistent implementation * Add torch.compile cases * Preallocate output buffer for softmax tutorial
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.
The updated fused-softmax has only persistent thread approach, which isn't straightforward for CPU. So, at least for now, bring back the old version and make it for CPU.