fhahn created this revision.
fhahn added reviewers: aaron.ballman, scanon, craig.topper, rjmccall,
erichkeane.
fhahn requested review of this revision.
Herald added a project: clang.
This patch implements __builtin_reduce_max and __builtin_reduce_min as
specified in D111529 <https://reviews.llvm.org/D111529>.
The order of operations does not matter for min or max reductions and
they can be directly lowered to the corresponding
llvm.vector.reduce.{fmin,fmax,umin,umax,smin,smax} intrinsic calls.
Repository:
rG LLVM Github Monorepo
https://reviews.llvm.org/D112001
Files:
clang/include/clang/Basic/Builtins.def
clang/include/clang/Sema/Sema.h
clang/lib/CodeGen/CGBuiltin.cpp
clang/lib/Sema/SemaChecking.cpp
clang/test/CodeGen/builtins-elementwise-math.c
clang/test/Sema/builtins-elementwise-math.c
Index: clang/test/Sema/builtins-elementwise-math.c
===================================================================
--- clang/test/Sema/builtins-elementwise-math.c
+++ clang/test/Sema/builtins-elementwise-math.c
@@ -61,3 +61,31 @@
i = __builtin_elementwise_min(v, iv);
// expected-error@-1 {{argument types do not match, 'float4' (vector of 4 'float' values) != 'int3' (vector of 3 'int' values)}}
}
+
+void test_builtin_reduce_max(int i, float4 v, int3 iv) {
+ struct Foo s = __builtin_reduce_max(iv);
+ // expected-error@-1 {{initializing 'struct Foo' with an expression of incompatible type 'int'}}
+
+ i = __builtin_reduce_max(v, v);
+ // expected-error@-1 {{too many arguments to function call, expected 1, have 2}}
+
+ i = __builtin_reduce_max();
+ // expected-error@-1 {{too few arguments to function call, expected 1, have 0}}
+
+ i = __builtin_reduce_max(i);
+ // expected-error@-1 {{argument must have a vector type, but was 'int'}}
+}
+
+void test_builtin_reduce_min(int i, float4 v, int3 iv) {
+ struct Foo s = __builtin_reduce_min(iv);
+ // expected-error@-1 {{initializing 'struct Foo' with an expression of incompatible type 'int'}}
+
+ i = __builtin_reduce_min(v, v);
+ // expected-error@-1 {{too many arguments to function call, expected 1, have 2}}
+
+ i = __builtin_reduce_min();
+ // expected-error@-1 {{too few arguments to function call, expected 1, have 0}}
+
+ i = __builtin_reduce_min(i);
+ // expected-error@-1 {{argument must have a vector type, but was 'int'}}
+}
Index: clang/test/CodeGen/builtins-elementwise-math.c
===================================================================
--- clang/test/CodeGen/builtins-elementwise-math.c
+++ clang/test/CodeGen/builtins-elementwise-math.c
@@ -110,3 +110,33 @@
// CHECK-NEXT: call <4 x i32> @llvm.umin.v4i32(<4 x i32> [[VU1]], <4 x i32> [[VU2]])
vu1 = __builtin_elementwise_min(vu1, vu2);
}
+
+void test_builtin_reduce_max(float4 vf1, si8 vi1, u4 vu1) {
+ // CHECK-LABEL: define void @test_builtin_reduce_max(
+ // CHECK: [[VF1:%.+]] = load <4 x float>, <4 x float>* %vf1.addr, align 16
+ // CHECK-NEXT: call float @llvm.vector.reduce.fmax.v4f32(<4 x float> [[VF1]])
+ float r1 = __builtin_reduce_max(vf1);
+
+ // CHECK: [[VI1:%.+]] = load <8 x i16>, <8 x i16>* %vi1.addr, align 16
+ // CHECK-NEXT: call i16 @llvm.vector.reduce.smax.v8i16(<8 x i16> [[VI1]])
+ short r2 = __builtin_reduce_max(vi1);
+
+ // CHECK: [[VU1:%.+]] = load <4 x i32>, <4 x i32>* %vu1.addr, align 16
+ // CHECK-NEXT: call i32 @llvm.vector.reduce.umax.v4i32(<4 x i32> [[VU1]])
+ unsigned r3 = __builtin_reduce_max(vu1);
+}
+
+void test_builtin_reduce_min(float4 vf1, si8 vi1, u4 vu1) {
+ // CHECK-LABEL: define void @test_builtin_reduce_min(
+ // CHECK: [[VF1:%.+]] = load <4 x float>, <4 x float>* %vf1.addr, align 16
+ // CHECK-NEXT: call float @llvm.vector.reduce.fmin.v4f32(<4 x float> [[VF1]])
+ float r1 = __builtin_reduce_min(vf1);
+
+ // CHECK: [[VI1:%.+]] = load <8 x i16>, <8 x i16>* %vi1.addr, align 16
+ // CHECK-NEXT: call i16 @llvm.vector.reduce.smin.v8i16(<8 x i16> [[VI1]])
+ short r2 = __builtin_reduce_min(vi1);
+
+ // CHECK: [[VU1:%.+]] = load <4 x i32>, <4 x i32>* %vu1.addr, align 16
+ // CHECK-NEXT: call i32 @llvm.vector.reduce.umin.v4i32(<4 x i32> [[VU1]])
+ unsigned r3 = __builtin_reduce_min(vu1);
+}
Index: clang/lib/Sema/SemaChecking.cpp
===================================================================
--- clang/lib/Sema/SemaChecking.cpp
+++ clang/lib/Sema/SemaChecking.cpp
@@ -1978,11 +1978,15 @@
case Builtin::BI__builtin_elementwise_abs:
return SemaBuiltinElementwiseMathOneArg(TheCall, TheCallResult);
-
case Builtin::BI__builtin_elementwise_min:
case Builtin::BI__builtin_elementwise_max:
return SemaBuiltinElementwiseMathTwoArgs(TheCall, TheCallResult);
+ case Builtin::BI__builtin_reduce_max:
+ case Builtin::BI__builtin_reduce_min:
+ return SemaBuiltinReduceMath(TheCall, TheCallResult);
+
+
case Builtin::BI__builtin_matrix_transpose:
return SemaBuiltinMatrixTranspose(TheCall, TheCallResult);
case Builtin::BI__builtin_matrix_column_major_load:
@@ -16656,6 +16660,25 @@
_2, _3, _4));
}
+ExprResult Sema::SemaBuiltinReduceMath(CallExpr *TheCall,
+ ExprResult CallResult) {
+ if (checkArgCount(*this, TheCall, 1))
+ return ExprError();
+
+ Expr *A = TheCall->getArg(0);
+ QualType TyA = A->getType();
+
+ const VectorType *VecTy = TyA->getAs<VectorType>();
+ if (!VecTy)
+ if (!TyA->getAs<VectorType>())
+ return Diag(A->getBeginLoc(), diag::err_elementwise_math_invalid_arg_type_2)
+ << "vector"
+ << TyA;
+
+ TheCall->setType(VecTy->getElementType());
+ return CallResult;
+}
+
ExprResult Sema::SemaBuiltinElementwiseMathOneArg(CallExpr *TheCall,
ExprResult CallResult) {
if (checkArgCount(*this, TheCall, 1))
Index: clang/lib/CodeGen/CGBuiltin.cpp
===================================================================
--- clang/lib/CodeGen/CGBuiltin.cpp
+++ clang/lib/CodeGen/CGBuiltin.cpp
@@ -3145,6 +3145,43 @@
return RValue::get(Result);
}
+ case Builtin::BI__builtin_reduce_max: {
+ auto GetIntrinsicID = [](QualType QT, llvm::Type *IrTy) {
+ if (IrTy->isIntOrIntVectorTy()) {
+ if (auto *VecTy = QT->getAs<VectorType>())
+ QT = VecTy->getElementType();
+ if (QT->isSignedIntegerType())
+ return llvm::Intrinsic::vector_reduce_smax;
+ else
+ return llvm::Intrinsic::vector_reduce_umax;
+ }
+ return llvm::Intrinsic::vector_reduce_fmax;
+ };
+ Value *Op0 = EmitScalarExpr(E->getArg(0));
+ Value *Result = Builder.CreateUnaryIntrinsic(GetIntrinsicID(E->getArg(0)->getType(), Op0->getType()),
+ Op0, nullptr, "rdx.min");
+ return RValue::get(Result);
+ }
+
+ case Builtin::BI__builtin_reduce_min: {
+ auto GetIntrinsicID = [](QualType QT, llvm::Type *IrTy) {
+ if (IrTy->isIntOrIntVectorTy()) {
+ if (auto *VecTy = QT->getAs<VectorType>())
+ QT = VecTy->getElementType();
+ if (QT->isSignedIntegerType())
+ return llvm::Intrinsic::vector_reduce_smin;
+ else
+ return llvm::Intrinsic::vector_reduce_umin;
+ }
+ return llvm::Intrinsic::vector_reduce_fmin;
+ };
+ Value *Op0 = EmitScalarExpr(E->getArg(0));
+ Value *Result = Builder.CreateUnaryIntrinsic(GetIntrinsicID(E->getArg(0)->getType(), Op0->getType()),
+ Op0, nullptr, "rdx.min");
+ return RValue::get(Result);
+ }
+
+
case Builtin::BI__builtin_matrix_transpose: {
const auto *MatrixTy = E->getArg(0)->getType()->getAs<ConstantMatrixType>();
Value *MatValue = EmitScalarExpr(E->getArg(0));
Index: clang/include/clang/Sema/Sema.h
===================================================================
--- clang/include/clang/Sema/Sema.h
+++ clang/include/clang/Sema/Sema.h
@@ -12716,6 +12716,8 @@
ExprResult CallResult);
ExprResult SemaBuiltinElementwiseMathTwoArgs(CallExpr *TheCall,
ExprResult CallResult);
+ ExprResult SemaBuiltinReduceMath(CallExpr *TheCall,
+ ExprResult CallResult);
// Matrix builtin handling.
ExprResult SemaBuiltinMatrixTranspose(CallExpr *TheCall,
Index: clang/include/clang/Basic/Builtins.def
===================================================================
--- clang/include/clang/Basic/Builtins.def
+++ clang/include/clang/Basic/Builtins.def
@@ -646,6 +646,8 @@
BUILTIN(__builtin_elementwise_abs, "v.", "nct")
BUILTIN(__builtin_elementwise_max, "v.", "nct")
BUILTIN(__builtin_elementwise_min, "v.", "nct")
+BUILTIN(__builtin_reduce_max, "v.", "nct")
+BUILTIN(__builtin_reduce_min, "v.", "nct")
BUILTIN(__builtin_matrix_transpose, "v.", "nFt")
BUILTIN(__builtin_matrix_column_major_load, "v.", "nFt")
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
