gaspersic/triton
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

[Triton-MLIR][Backend] Fix many problems to get the pipeline working (#809)

Browse Source 
1. Rewrite code generation of insert_slice_async.
2. Correct the wrong index passed to extract_slice in pipeline.
3. Add a prologue in pipeline to wait for dangling cp.asyncs.  
4. Move scf to cf conversion inside TritonGPUToLLVM because we need to
perform membar before scf to cf. It shouldn't be a technical limitation
and could be improved by a more general membar analysis.
5. Use an attribute to memoize the shared memory size and support
dynamic shared memory.
6. Prevent the combine pass to reorder insert_slice and extract_slice
across async_wait

Co-authored-by: Superjomn <yanchunwei@outlook.com>
This commit is contained in:
Keren Zhou
2022-10-27 22:09:06 -07:00
committed by GitHub
parent 42db3538e4
commit 3b80801dff
10 changed files with 122 additions and 70 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
lib/Target/LLVMIR/LLVMIRTranslation.cpp
View File

@@ -136,10 +136,12 @@ translateTritonGPUToLLVMIR(llvm::LLVMContext *llvmContext,
/*printAfterOnlyOnChange=*/true,
/*printAfterOnlyOnFailure*/ false, llvm::dbgs(), printingFlags);
pm.addPass(mlir::createLowerToCFGPass());
pm.addPass(createConvertTritonGPUToLLVMPass());
// Conanicalize to eliminate the remaining UnrealizedConversionCastOp
pm.addPass(mlir::createCanonicalizerPass());
pm.addPass(mlir::createCSEPass()); // Simplify the IR to improve readability.
pm.addPass(mlir::createSymbolDCEPass());
pm.addPass(mlir::createCanonicalizerPass());
if (failed(pm.run(module))) {
llvm::errs() << "Pass execution failed";