Skip to content

[CIR][MLIR][LoweringThroughMLIR] CIRUnaryOpLowering on float values. #2032

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Dec 4, 2025
+73 −23
Merged

[CIR][MLIR][LoweringThroughMLIR] CIRUnaryOpLowering on float values. #2032

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
f3e89e6
[CIR][MLIR][LoweringThroughMLIR] CIRUnaryOpLowering on float values.
terapines-osc-cir Nov 28, 2025
75925e8
[CIR][MLIR][LoweringThroughMLIR] More type checks in UnaryOp Lowering.
terapines-osc-cir Dec 1, 2025
5604989
Refactor: Duplicated code got eliminated by a template method.
terapines-osc-cir Dec 3, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
48 changes: 35 additions & 13 deletions clang/lib/CIR/Lowering/ThroughMLIR/LowerCIRToMLIR.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -802,6 +802,33 @@ class CIRUnaryOpLowering : public mlir::OpConversionPattern<cir::UnaryOp> {
public:
using OpConversionPattern<cir::UnaryOp>::OpConversionPattern;

template <typename OpFloat, typename OpInt, bool rev>
mlir::Operation *
replaceImmediateOp(cir::UnaryOp op, mlir::Type type, mlir::Value input,
int64_t n,
mlir::ConversionPatternRewriter &rewriter) const {
if (type.isFloat()) {
auto imm = mlir::arith::ConstantOp::create(
rewriter, op.getLoc(),
mlir::FloatAttr::get(type, static_cast<double>(n)));
if constexpr (rev)
return rewriter.replaceOpWithNewOp<OpFloat>(op, type, imm, input);
else
return rewriter.replaceOpWithNewOp<OpFloat>(op, type, input, imm);
}
if (type.isInteger()) {
auto imm = mlir::arith::ConstantOp::create(
rewriter, op.getLoc(), mlir::IntegerAttr::get(type, n));
if constexpr (rev)
return rewriter.replaceOpWithNewOp<OpInt>(op, type, imm, input);
else
return rewriter.replaceOpWithNewOp<OpInt>(op, type, input, imm);
}
op->emitError("Unsupported type: ") << type << " at " << op->getLoc();
llvm_unreachable("CIRUnaryOpLowering met unsupported type");
return nullptr;
}

mlir::LogicalResult
matchAndRewrite(cir::UnaryOp op, OpAdaptor adaptor,
mlir::ConversionPatternRewriter &rewriter) const override {
Expand All @@ -810,36 +837,31 @@ class CIRUnaryOpLowering : public mlir::OpConversionPattern<cir::UnaryOp> {

switch (op.getKind()) {
case cir::UnaryOpKind::Inc: {
auto One = mlir::arith::ConstantOp::create(
rewriter, op.getLoc(), mlir::IntegerAttr::get(type, 1));
rewriter.replaceOpWithNewOp<mlir::arith::AddIOp>(op, type, input, One);
replaceImmediateOp<mlir::arith::AddFOp, mlir::arith::AddIOp, false>(
op, type, input, 1, rewriter);
break;
}
case cir::UnaryOpKind::Dec: {
auto One = mlir::arith::ConstantOp::create(
rewriter, op.getLoc(), mlir::IntegerAttr::get(type, 1));
rewriter.replaceOpWithNewOp<mlir::arith::SubIOp>(op, type, input, One);
replaceImmediateOp<mlir::arith::AddFOp, mlir::arith::AddIOp, false>(
op, type, input, -1, rewriter);
break;
}
case cir::UnaryOpKind::Plus: {
rewriter.replaceOp(op, op.getInput());
break;
}
case cir::UnaryOpKind::Minus: {
auto Zero = mlir::arith::ConstantOp::create(
rewriter, op.getLoc(), mlir::IntegerAttr::get(type, 0));
rewriter.replaceOpWithNewOp<mlir::arith::SubIOp>(op, type, Zero, input);
replaceImmediateOp<mlir::arith::SubFOp, mlir::arith::SubIOp, true>(
op, type, input, 0, rewriter);
break;
}
case cir::UnaryOpKind::Not: {
auto MinusOne = mlir::arith::ConstantOp::create(
auto o = mlir::arith::ConstantOp::create(
rewriter, op.getLoc(), mlir::IntegerAttr::get(type, -1));
rewriter.replaceOpWithNewOp<mlir::arith::XOrIOp>(op, type, MinusOne,
input);
rewriter.replaceOpWithNewOp<mlir::arith::XOrIOp>(op, type, o, input);
break;
}
}

return mlir::LogicalResult::success();
}
};
Expand Down
8 changes: 4 additions & 4 deletions clang/test/CIR/Lowering/ThroughMLIR/if.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -29,8 +29,8 @@ void foo() {
//CHECK: memref.store %[[SEVEN]], %[[alloca_0]][] : memref<i32>
//CHECK: } else {
//CHECK: %[[SIX:.+]] = memref.load %[[alloca_0]][] : memref<i32>
//CHECK: %[[C1_I32:.+]] = arith.constant 1 : i32
//CHECK: %[[SEVEN:.+]] = arith.subi %[[SIX]], %[[C1_I32]] : i32
//CHECK: %[[C1_I32:.+]] = arith.constant -1 : i32
//CHECK: %[[SEVEN:.+]] = arith.addi %[[SIX]], %[[C1_I32]] : i32
//CHECK: memref.store %[[SEVEN]], %[[alloca_0]][] : memref<i32>
//CHECK: }
//CHECK: }
Expand Down Expand Up @@ -106,8 +106,8 @@ void foo3() {
//CHECK: memref.store %[[THIRTEEN]], %[[alloca_0]][] : memref<i32>
//CHECK: } else {
//CHECK: %[[TWELVE:.+]] = memref.load %[[alloca_0]][] : memref<i32>
//CHECK: %[[C1_I32_5:.+]] = arith.constant 1 : i32
//CHECK: %[[THIRTEEN:.+]] = arith.subi %[[TWELVE]], %[[C1_I32_5]] : i32
//CHECK: %[[C1_I32_5:.+]] = arith.constant -1 : i32
//CHECK: %[[THIRTEEN:.+]] = arith.addi %[[TWELVE]], %[[C1_I32_5]] : i32
//CHECK: memref.store %[[THIRTEEN]], %[[alloca_0]][] : memref<i32>
//CHECK: }
//CHECK: }
Expand Down
30 changes: 24 additions & 6 deletions clang/test/CIR/Lowering/ThroughMLIR/unary-inc-dec.cir
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
// RUN: cir-opt %s -cir-to-mlir -o - | FileCheck %s -check-prefix=MLIR
// RUN: cir-opt %s -cir-to-mlir -cir-mlir-to-llvm -o - | mlir-translate -mlir-to-llvmir | FileCheck %s -check-prefix=LLVM
// RUN: cir-opt %s --cir-to-mlir -o - | FileCheck %s -check-prefix=MLIR
// RUN: cir-opt %s --cir-to-mlir -cir-mlir-to-llvm -o - | mlir-translate -mlir-to-llvmir | FileCheck %s -check-prefix=LLVM

!s32i = !cir.int<s, 32>
module {
Expand All @@ -17,14 +17,32 @@ module {
%5 = cir.load %1 : !cir.ptr<!s32i>, !s32i
%6 = cir.unary(dec, %5) : !s32i, !s32i
cir.store %6, %1 : !s32i, !cir.ptr<!s32i>

// test float
%7 = cir.alloca !s32i, !cir.ptr<!s32i>, ["b", init] {alignment = 4 : i64}
cir.return
}
}

// MLIR: = arith.constant 1
// MLIR: = arith.addi
// MLIR: = arith.constant 1
// MLIR: = arith.subi
// MLIR: = arith.constant -1
// MLIR: = arith.addi

// LLVM: = add i32 %[[#]], 1
// LLVM: = sub i32 %[[#]], 1
// LLVM: = add i32 %[[#]], -1


cir.func @floatingPoints(%arg0: !cir.double) {
%0 = cir.alloca !cir.double, !cir.ptr<!cir.double>, ["X", init] {alignment = 8 : i64}
cir.store %arg0, %0 : !cir.double, !cir.ptr<!cir.double>
%1 = cir.load %0 : !cir.ptr<!cir.double>, !cir.double
%2 = cir.unary(inc, %1) : !cir.double, !cir.double
%3 = cir.load %0 : !cir.ptr<!cir.double>, !cir.double
%4 = cir.unary(dec, %3) : !cir.double, !cir.double
cir.return
}
// MLIR: = arith.constant 1.0
// MLIR: = arith.addf
// MLIR: = arith.constant -1.0
// MLIR: = arith.addf
}
10 changes: 10 additions & 0 deletions clang/test/CIR/Lowering/ThroughMLIR/unary-plus-minus.cir
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,6 +19,16 @@ module {
cir.store %6, %1 : !s32i, !cir.ptr<!s32i>
cir.return
}

cir.func @floatingPoints(%arg0: !cir.double) {
%0 = cir.alloca !cir.double, !cir.ptr<!cir.double>, ["X", init] {alignment = 8 : i64}
cir.store %arg0, %0 : !cir.double, !cir.ptr<!cir.double>
%1 = cir.load %0 : !cir.ptr<!cir.double>, !cir.double
%2 = cir.unary(plus, %1) : !cir.double, !cir.double
%3 = cir.load %0 : !cir.ptr<!cir.double>, !cir.double
%4 = cir.unary(minus, %3) : !cir.double, !cir.double
cir.return
}
}

// MLIR: %[[#INPUT_PLUS:]] = memref.load
Expand Down