[Mlir-commits] [mlir] db11fd8 - [MLIR][NVVM] Change WgmmaMmaSyncOp to WgmmaMmaAsyncOp (NFC)
Guray Ozen
llvmlistbot at llvm.org
Sat Aug 12 00:55:17 PDT 2023
Author: Guray Ozen
Date: 2023-08-12T09:55:11+02:00
New Revision: db11fd8bf4ce1b1ce08db6011ea7c22d030eb473
URL: https://github.com/llvm/llvm-project/commit/db11fd8bf4ce1b1ce08db6011ea7c22d030eb473
DIFF: https://github.com/llvm/llvm-project/commit/db11fd8bf4ce1b1ce08db6011ea7c22d030eb473.diff
LOG: [MLIR][NVVM] Change WgmmaMmaSyncOp to WgmmaMmaAsyncOp (NFC)
WgmmaMmaSyncOp is asynchronous operation. There was a typo named op. This work fixes that.
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D157697
Added:
Modified:
mlir/include/mlir/Dialect/LLVMIR/NVVMOps.td
mlir/lib/Dialect/LLVMIR/IR/NVVMDialect.cpp
Removed:
################################################################################
diff --git a/mlir/include/mlir/Dialect/LLVMIR/NVVMOps.td b/mlir/include/mlir/Dialect/LLVMIR/NVVMOps.td
index 4845a226ec9c17..7ddf68e0a5b1a4 100644
--- a/mlir/include/mlir/Dialect/LLVMIR/NVVMOps.td
+++ b/mlir/include/mlir/Dialect/LLVMIR/NVVMOps.td
@@ -1497,7 +1497,7 @@ def WGMMAScaleOutAttr : EnumAttr<NVVM_Dialect, WGMMAScaleOut, "wgmma_scale_out">
let assemblyFormat = "`<` $value `>`";
}
-def NVVM_WgmmaMmaSyncOp : NVVM_Op<"wgmma.mma_async",
+def NVVM_WgmmaMmaAsyncOp : NVVM_Op<"wgmma.mma_async",
[DeclareOpInterfaceMethods<BasicPtxBuilderOpInterface>,
PredOpTrait<"input struct and result struct must be the same type",
TCresIsSameAsOpBase<0, 0>>,]>
diff --git a/mlir/lib/Dialect/LLVMIR/IR/NVVMDialect.cpp b/mlir/lib/Dialect/LLVMIR/IR/NVVMDialect.cpp
index 25f837547b46f5..70c774233797bb 100644
--- a/mlir/lib/Dialect/LLVMIR/IR/NVVMDialect.cpp
+++ b/mlir/lib/Dialect/LLVMIR/IR/NVVMDialect.cpp
@@ -708,7 +708,7 @@ LogicalResult NVVM::LdMatrixOp::verify() {
return success();
}
-LogicalResult NVVM::WgmmaMmaSyncOp::verify() {
+LogicalResult NVVM::WgmmaMmaAsyncOp::verify() {
Value outValue = getResults();
auto stype = dyn_cast<LLVM::LLVMStructType>(outValue.getType());
if (!stype)
@@ -889,7 +889,7 @@ LogicalResult NVVM::WgmmaMmaSyncOp::verify() {
return success();
}
-std::string NVVM::WgmmaMmaSyncOp::getPtx() {
+std::string NVVM::WgmmaMmaAsyncOp::getPtx() {
int m = getShape().getM(), n = getShape().getN(), k = getShape().getK();
bool isF16 = getTypeA() == mlir::NVVM::MMATypes::f16 ||
@@ -952,7 +952,7 @@ std::string NVVM::WgmmaMmaSyncOp::getPtx() {
return ptx;
}
-void NVVM::WgmmaMmaSyncOp::getAsmValues(
+void NVVM::WgmmaMmaAsyncOp::getAsmValues(
RewriterBase &rewriter,
llvm::SmallVectorImpl<std::pair<mlir::Value, mlir::NVVM::PTXRegisterMod>>
&asmValues) {
More information about the Mlir-commits
mailing list