[triton-mlir][BACKEND] Support masked load/store (#657)
This PR does - fix some bugs to support masked load/store, - refine frontend, and support the `and` and `or` syntax in mask(by extending the BoolOp in python ast.visitor), e.g. `tl.store(..., mask=offset<n and other_conditions)`, - add `arith.cmpI` and `arith.cmpF` op conversion in backend(required by mask), - add more test cases in vecadd.
This commit is contained in:
@@ -1210,6 +1210,8 @@ void init_triton_translation(py::module &m) {
|
||||
llvm::LLVMContext llvmContext;
|
||||
auto llvmModule =
|
||||
::mlir::triton::translateTritonGPUToLLVMIR(&llvmContext, op);
|
||||
if (!llvmModule)
|
||||
llvm::report_fatal_error("Failed to translate TritonGPU to LLVM IR.");
|
||||
|
||||
std::string str;
|
||||
llvm::raw_string_ostream os(str);
|
||||
|
Reference in New Issue
Block a user