llvmbot wrote:
<!--LLVM PR SUMMARY COMMENT--> @llvm/pr-subscribers-clangir @llvm/pr-subscribers-clang Author: Yeongu Choe (YeonguChoe) <details> <summary>Changes</summary> __builtin_signbit function checks if the sign bit of a floating-point number is set to 0 or 1. Reference - https://en.wikipedia.org/wiki/Sign_bit --- Full diff: https://github.com/llvm/llvm-project/pull/188433.diff 2 Files Affected: - (modified) clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp (+10-1) - (added) clang/test/CIR/CodeGenBuiltins/builtin-signbit.c (+158) ``````````diff diff --git a/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp b/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp index 1c62543d40bb3..bdae39b14dc88 100644 --- a/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp +++ b/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp @@ -1865,9 +1865,18 @@ RValue CIRGenFunction::emitBuiltinExpr(const GlobalDecl &gd, unsigned builtinID, return RValue::get(nullptr); } case Builtin::BI__scoped_atomic_thread_fence: + return errorBuiltinNYI(*this, e, builtinID); case Builtin::BI__builtin_signbit: case Builtin::BI__builtin_signbitf: - case Builtin::BI__builtin_signbitl: + case Builtin::BI__builtin_signbitl: { + CIRGenFunction::CIRGenFPOptionsRAII fPOptsRAII(*this, e); + mlir::Location loc = getLoc(e->getBeginLoc()); + mlir::Value value = emitScalarExpr(e->getArg(0)); + mlir::Operation *signBitOp = cir::SignBitOp::create(builder, loc, value); + mlir::Value result = builder.createBoolToInt(signBitOp->getResult(0), + convertType(e->getType())); + return RValue::get(result); + } case Builtin::BI__warn_memset_zero_len: case Builtin::BI__annotation: case Builtin::BI__builtin_annotation: diff --git a/clang/test/CIR/CodeGenBuiltins/builtin-signbit.c b/clang/test/CIR/CodeGenBuiltins/builtin-signbit.c new file mode 100644 index 0000000000000..2cf7579fcf077 --- /dev/null +++ b/clang/test/CIR/CodeGenBuiltins/builtin-signbit.c @@ -0,0 +1,158 @@ +// RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -fclangir -emit-cir %s -o %t.cir +// RUN: FileCheck %s --check-prefix=CIR --input-file %t.cir +// RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -fclangir -emit-llvm %s -o %t-cir.ll +// RUN: FileCheck %s --check-prefix=LLVM --input-file %t-cir.ll +// RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -emit-llvm %s -o %t.ll +// RUN: FileCheck %s --check-prefix=OGCG --input-file %t.ll + +void test_signbit_positive_zero(){ + double positiveZero = +0.0; + int result = __builtin_signbit(positiveZero); +// CIR: cir.alloca !cir.double, !cir.ptr<!cir.double>, ["positiveZero", init] +// CIR: cir.const #cir.fp<0.000000e+00> : !cir.double +// CIR: cir.signbit {{.*}} : !cir.double -> !cir.bool +// CIR: cir.cast bool_to_int {{.*}} : !cir.bool -> !s32i + +// LLVM: store double 0.000000e+00, ptr %{{.*}} +// LLVM: bitcast double %{{.*}} to i64 +// LLVM: icmp slt i64 %{{.*}}, 0 +// LLVM: zext i1 %{{.*}} to i32 + +// OGCG: store double 0.000000e+00, ptr %{{.*}} +// OGCG: bitcast double %{{.*}} to i64 +// OGCG: icmp slt i64 %{{.*}}, 0 +// OGCG: zext i1 %{{.*}} to i32 +} + +void test_signbit_negative_zero(){ + double negativeZero = -0.0; + int result = __builtin_signbit(negativeZero); +// CIR: cir.alloca !cir.double, !cir.ptr<!cir.double>, ["negativeZero", init] +// CIR: cir.const #cir.fp<-0.000000e+00> : !cir.double +// CIR: cir.signbit {{.*}} : !cir.double -> !cir.bool +// CIR: cir.cast bool_to_int {{.*}} : !cir.bool -> !s32i + +// LLVM: store double -0.000000e+00, ptr %{{.*}} +// LLVM: bitcast double %{{.*}} to i64 +// LLVM: icmp slt i64 %{{.*}}, 0 +// LLVM: zext i1 %{{.*}} to i32 + +// OGCG: store double -0.000000e+00, ptr %{{.*}} +// OGCG: bitcast double %{{.*}} to i64 +// OGCG: icmp slt i64 %{{.*}}, 0 +// OGCG: zext i1 %{{.*}} to i32 +} + +void test_signbit_positive_number(){ + double positiveNumber = 1.0; + int result = __builtin_signbit(positiveNumber); +// CIR: cir.alloca !cir.double, !cir.ptr<!cir.double>, ["positiveNumber", init] +// CIR: cir.const #cir.fp<1.000000e+00> : !cir.double +// CIR: cir.signbit {{.*}} : !cir.double -> !cir.bool +// CIR: cir.cast bool_to_int {{.*}} : !cir.bool -> !s32i + +// LLVM: store double 1.000000e+00, ptr %{{.*}} +// LLVM: bitcast double %{{.*}} to i64 +// LLVM: icmp slt i64 %{{.*}}, 0 +// LLVM: zext i1 %{{.*}} to i32 + +// OGCG: store double 1.000000e+00, ptr %{{.*}} +// OGCG: bitcast double %{{.*}} to i64 +// OGCG: icmp slt i64 %{{.*}}, 0 +// OGCG: zext i1 %{{.*}} to i32 +} + +void test_signbit_negative_number(){ + double negativeNumber = -1.0; + int result = __builtin_signbit(negativeNumber); +// CIR: cir.alloca !cir.double, !cir.ptr<!cir.double>, ["negativeNumber", init] +// CIR: cir.const #cir.fp<-1.000000e+00> : !cir.double +// CIR: cir.signbit {{.*}} : !cir.double -> !cir.bool +// CIR: cir.cast bool_to_int {{.*}} : !cir.bool -> !s32i + +// LLVM: store double -1.000000e+00, ptr %{{.*}} +// LLVM: bitcast double %{{.*}} to i64 +// LLVM: icmp slt i64 %{{.*}}, 0 +// LLVM: zext i1 %{{.*}} to i32 + +// OGCG: store double -1.000000e+00, ptr %{{.*}} +// OGCG: bitcast double %{{.*}} to i64 +// OGCG: icmp slt i64 %{{.*}}, 0 +// OGCG: zext i1 %{{.*}} to i32 +} + +void test_signbit_positive_nan(){ + double positiveNan = +__builtin_nan(""); + int result = __builtin_signbit(positiveNan); +// CIR: cir.alloca !cir.double, !cir.ptr<!cir.double>, ["positiveNan", init] +// CIR: cir.const #cir.fp<0x7FF8000000000000> : !cir.double +// CIR: cir.signbit {{.*}} : !cir.double -> !cir.bool +// CIR: cir.cast bool_to_int {{.*}} : !cir.bool -> !s32i + +// LLVM: store double 0x7FF8000000000000, ptr %{{.*}} +// LLVM: bitcast double %{{.*}} to i64 +// LLVM: icmp slt i64 %{{.*}}, 0 +// LLVM: zext i1 %{{.*}} to i32 + +// OGCG: store double 0x7FF8000000000000, ptr %{{.*}} +// OGCG: bitcast double %{{.*}} to i64 +// OGCG: icmp slt i64 %{{.*}}, 0 +// OGCG: zext i1 %{{.*}} to i32 +} + +void test_signbit_negative_nan(){ + double negativeNan = -__builtin_nan(""); + int result = __builtin_signbit(negativeNan); +// CIR: cir.alloca !cir.double, !cir.ptr<!cir.double>, ["negativeNan", init] +// CIR: cir.const #cir.fp<0xFFF8000000000000> : !cir.double +// CIR: cir.signbit {{.*}} : !cir.double -> !cir.bool +// CIR: cir.cast bool_to_int {{.*}} : !cir.bool -> !s32i + +// LLVM: store double 0xFFF8000000000000, ptr %{{.*}} +// LLVM: bitcast double %{{.*}} to i64 +// LLVM: icmp slt i64 %{{.*}}, 0 +// LLVM: zext i1 %{{.*}} to i32 + +// OGCG: store double 0xFFF8000000000000, ptr %{{.*}} +// OGCG: bitcast double %{{.*}} to i64 +// OGCG: icmp slt i64 %{{.*}}, 0 +// OGCG: zext i1 %{{.*}} to i32 +} + +void test_signbit_positive_infinity(){ + double positiveInfinity = +__builtin_inf(); + int result = __builtin_signbit(positiveInfinity); +// CIR: cir.alloca !cir.double, !cir.ptr<!cir.double>, ["positiveInfinity", init] +// CIR: cir.const #cir.fp<0x7FF0000000000000> : !cir.double +// CIR: cir.signbit {{.*}} : !cir.double -> !cir.bool +// CIR: cir.cast bool_to_int {{.*}} : !cir.bool -> !s32i + +// LLVM: store double 0x7FF0000000000000, ptr %{{.*}} +// LLVM: bitcast double %{{.*}} to i64 +// LLVM: icmp slt i64 %{{.*}}, 0 +// LLVM: zext i1 %{{.*}} to i32 + +// OGCG: store double 0x7FF0000000000000, ptr %{{.*}} +// OGCG: bitcast double %{{.*}} to i64 +// OGCG: icmp slt i64 %{{.*}}, 0 +// OGCG: zext i1 %{{.*}} to i32 +} + +void test_signbit_negative_infinity(){ + double negativeInfinity = -__builtin_inf(); + int result = __builtin_signbit(negativeInfinity); +// CIR: cir.alloca !cir.double, !cir.ptr<!cir.double>, ["negativeInfinity", init] +// CIR: cir.const #cir.fp<0xFFF0000000000000> : !cir.double +// CIR: cir.signbit {{.*}} : !cir.double -> !cir.bool +// CIR: cir.cast bool_to_int {{.*}} : !cir.bool -> !s32i + +// LLVM: store double 0xFFF0000000000000, ptr %{{.*}} +// LLVM: bitcast double %{{.*}} to i64 +// LLVM: icmp slt i64 %{{.*}}, 0 +// LLVM: zext i1 %{{.*}} to i32 + +// OGCG: store double 0xFFF0000000000000, ptr %{{.*}} +// OGCG: bitcast double %{{.*}} to i64 +// OGCG: icmp slt i64 %{{.*}}, 0 +// OGCG: zext i1 %{{.*}} to i32 +} `````````` </details> https://github.com/llvm/llvm-project/pull/188433 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
