Emit constrained atan2 intrinsic for clang builtin

This change is part of this proposal: https://discourse.llvm.org/t/rfc-all-the-math-intrinsics/78294 - `Builtins.td` - Add f16 support for libm atan2 builtin - `CGBuiltin.cpp` - Emit constraint atan2 intrinsic for clang builtin Part of Implement the atan2 HLSL Function llvm#70096.
tex3d · Oct 21, 2024 · a78cf25 · a78cf25
1 parent 8c975b3
commit a78cf25
Show file tree

Hide file tree

Showing 8 changed files with 49 additions and 29 deletions.
diff --git a/clang/include/clang/Basic/Builtins.td b/clang/include/clang/Basic/Builtins.td
@@ -227,10 +227,10 @@ def FminimumNumF16F128 : Builtin, F16F128MathTemplate {
   let Prototype = "T(T, T)";
 }
 
-def Atan2F128 : Builtin {
-  let Spellings = ["__builtin_atan2f128"];
+def Atan2F16F128 : Builtin, F16F128MathTemplate {
+  let Spellings = ["__builtin_atan2"];
   let Attributes = [FunctionWithBuiltinPrefix, NoThrow, ConstIgnoringErrnoAndExceptions];
-  let Prototype = "__float128(__float128, __float128)";
+  let Prototype = "T(T, T)";
 }
 
 def CopysignF16 : Builtin {

diff --git a/clang/lib/CodeGen/CGBuiltin.cpp b/clang/lib/CodeGen/CGBuiltin.cpp
@@ -2724,6 +2724,17 @@ RValue CodeGenFunction::EmitBuiltinExpr(const GlobalDecl GD, unsigned BuiltinID,
       return RValue::get(emitUnaryMaybeConstrainedFPBuiltin(
           *this, E, Intrinsic::atan, Intrinsic::experimental_constrained_atan));
 
+    case Builtin::BIatan2:
+    case Builtin::BIatan2f:
+    case Builtin::BIatan2l:
+    case Builtin::BI__builtin_atan2:
+    case Builtin::BI__builtin_atan2f:
+    case Builtin::BI__builtin_atan2f16:
+    case Builtin::BI__builtin_atan2l:
+    case Builtin::BI__builtin_atan2f128:
+      return RValue::get(emitBinaryMaybeConstrainedFPBuiltin(
+          *this, E, Intrinsic::atan2, Intrinsic::experimental_constrained_atan2));
+
     case Builtin::BIceil:
     case Builtin::BIceilf:
     case Builtin::BIceill:

diff --git a/clang/test/CodeGen/X86/math-builtins.c b/clang/test/CodeGen/X86/math-builtins.c
@@ -45,18 +45,18 @@ void foo(double *d, float f, float *fp, long double *l, int *i, const char *c) {
 
   __builtin_atan2(f,f);    __builtin_atan2f(f,f) ;  __builtin_atan2l(f, f); __builtin_atan2f128(f,f);
 
-// NO__ERRNO: declare double @atan2(double noundef, double noundef) [[READNONE:#[0-9]+]]
-// NO__ERRNO: declare float @atan2f(float noundef, float noundef) [[READNONE]]
-// NO__ERRNO: declare x86_fp80 @atan2l(x86_fp80 noundef, x86_fp80 noundef) [[READNONE]]
-// NO__ERRNO: declare fp128 @atan2f128(fp128 noundef, fp128 noundef) [[READNONE]]
+// NO__ERRNO: declare double @llvm.atan2.f64(double, double) [[READNONE_INTRINSIC:#[0-9]+]]
+// NO__ERRNO: declare float @llvm.atan2.f32(float, float) [[READNONE_INTRINSIC]]
+// NO__ERRNO: declare x86_fp80 @llvm.atan2.f80(x86_fp80, x86_fp80) [[READNONE_INTRINSIC]]
+// NO__ERRNO: declare fp128 @llvm.atan2.f128(fp128, fp128) [[READNONE_INTRINSIC]]
 // HAS_ERRNO: declare double @atan2(double noundef, double noundef) [[NOT_READNONE]]
 // HAS_ERRNO: declare float @atan2f(float noundef, float noundef) [[NOT_READNONE]]
 // HAS_ERRNO: declare x86_fp80 @atan2l(x86_fp80 noundef, x86_fp80 noundef) [[NOT_READNONE]]
 // HAS_ERRNO: declare fp128 @atan2f128(fp128 noundef, fp128 noundef) [[NOT_READNONE]]
 
   __builtin_copysign(f,f); __builtin_copysignf(f,f); __builtin_copysignl(f,f); __builtin_copysignf128(f,f);
 
-// NO__ERRNO: declare double @llvm.copysign.f64(double, double) [[READNONE_INTRINSIC:#[0-9]+]]
+// NO__ERRNO: declare double @llvm.copysign.f64(double, double) [[READNONE_INTRINSIC]]
 // NO__ERRNO: declare float @llvm.copysign.f32(float, float) [[READNONE_INTRINSIC]]
 // NO__ERRNO: declare x86_fp80 @llvm.copysign.f80(x86_fp80, x86_fp80) [[READNONE_INTRINSIC]]
 // NO__ERRNO: declare fp128 @llvm.copysign.f128(fp128, fp128) [[READNONE_INTRINSIC]]
@@ -179,7 +179,7 @@ void foo(double *d, float f, float *fp, long double *l, int *i, const char *c) {
 
   __builtin_acosh(f);      __builtin_acoshf(f);     __builtin_acoshl(f);  __builtin_acoshf128(f);
 
-// NO__ERRNO: declare double @acosh(double noundef) [[READNONE]]
+// NO__ERRNO: declare double @acosh(double noundef) [[READNONE:#[0-9]+]]
 // NO__ERRNO: declare float @acoshf(float noundef) [[READNONE]]
 // NO__ERRNO: declare x86_fp80 @acoshl(x86_fp80 noundef) [[READNONE]]
 // NO__ERRNO: declare fp128 @acoshf128(fp128 noundef) [[READNONE]]
@@ -721,10 +721,10 @@ __builtin_trunc(f);      __builtin_truncf(f);     __builtin_truncl(f); __builtin
 // HAS_ERRNO: declare fp128 @llvm.trunc.f128(fp128) [[READNONE_INTRINSIC]]
 };
 
-// NO__ERRNO: attributes [[READNONE]] = { {{.*}}memory(none){{.*}} }
 // NO__ERRNO: attributes [[READNONE_INTRINSIC]] = { {{.*}}memory(none){{.*}} }
 // NO__ERRNO: attributes [[NOT_READNONE]] = { nounwind {{.*}} }
 // NO__ERRNO: attributes [[PURE]] = { {{.*}}memory(read){{.*}} }
+// NO__ERRNO: attributes [[READNONE]] = { {{.*}}memory(none){{.*}} }
 
 // HAS_ERRNO: attributes [[NOT_READNONE]] = { nounwind {{.*}} }
 // HAS_ERRNO: attributes [[READNONE_INTRINSIC]] = { {{.*}}memory(none){{.*}} }

diff --git a/clang/test/CodeGen/constrained-math-builtins.c b/clang/test/CodeGen/constrained-math-builtins.c
@@ -57,6 +57,13 @@ __builtin_atan(f);        __builtin_atanf(f);       __builtin_atanl(f); __builti
 // CHECK: call x86_fp80 @llvm.experimental.constrained.atan.f80(x86_fp80 %{{.*}}, metadata !"round.tonearest", metadata !"fpexcept.strict")
 // CHECK: call fp128 @llvm.experimental.constrained.atan.f128(fp128 %{{.*}}, metadata !"round.tonearest", metadata !"fpexcept.strict")
 
+__builtin_atan2(f,f);        __builtin_atan2f(f,f);       __builtin_atan2l(f,f); __builtin_atan2f128(f,f);
+
+// CHECK: call double @llvm.experimental.constrained.atan2.f64(double %{{.*}}, double %{{.*}}, metadata !"round.tonearest", metadata !"fpexcept.strict")
+// CHECK: call float @llvm.experimental.constrained.atan2.f32(float %{{.*}}, float %{{.*}}, metadata !"round.tonearest", metadata !"fpexcept.strict")
+// CHECK: call x86_fp80 @llvm.experimental.constrained.atan2.f80(x86_fp80 %{{.*}}, x86_fp80 %{{.*}}, metadata !"round.tonearest", metadata !"fpexcept.strict")
+// CHECK: call fp128 @llvm.experimental.constrained.atan2.f128(fp128 %{{.*}}, fp128 %{{.*}}, metadata !"round.tonearest", metadata !"fpexcept.strict")
+
   __builtin_ceil(f);       __builtin_ceilf(f);      __builtin_ceill(f); __builtin_ceilf128(f);
 
 // CHECK: call double @llvm.experimental.constrained.ceil.f64(double %{{.*}}, metadata !"fpexcept.strict")

diff --git a/clang/test/CodeGen/libcalls.c b/clang/test/CodeGen/libcalls.c
@@ -95,9 +95,9 @@ void test_builtins(double d, float f, long double ld) {
   double atan2_ = atan2(d, 2);
   long double atan2l_ = atan2l(ld, ld);
   float atan2f_ = atan2f(f, f);
-// CHECK-NO: declare double @atan2(double noundef, double noundef) [[NUW_RN:#[0-9]+]]
-// CHECK-NO: declare x86_fp80 @atan2l(x86_fp80 noundef, x86_fp80 noundef) [[NUW_RN]]
-// CHECK-NO: declare float @atan2f(float noundef, float noundef) [[NUW_RN]]
+// CHECK-NO: declare double @llvm.atan2.f64(double, double) [[NUW_RNI]]
+// CHECK-NO: declare x86_fp80 @llvm.atan2.f80(x86_fp80, x86_fp80) [[NUW_RNI]]
+// CHECK-NO: declare float @llvm.atan2.f32(float, float) [[NUW_RNI]]
 // CHECK-YES: declare double @atan2(double noundef, double noundef) [[NUW]]
 // CHECK-YES: declare x86_fp80 @atan2l(x86_fp80 noundef, x86_fp80 noundef) [[NUW]]
 // CHECK-YES: declare float @atan2f(float noundef, float noundef) [[NUW]]
@@ -124,5 +124,4 @@ void test_builtins(double d, float f, long double ld) {
 }
 
 // CHECK-YES: attributes [[NUW]] = { nounwind "no-trapping-math"="true" "stack-protector-buffer-size"="8" "target-features"="+cx8,+x87" }
-// CHECK-NO-DAG: attributes [[NUW_RN]] = { nounwind willreturn memory(none){{.*}} }
 // CHECK-NO-DAG: attributes [[NUW_RNI]] = { nocallback nofree nosync nounwind speculatable willreturn memory(none) }
diff --git a/clang/test/CodeGen/math-libcalls.c b/clang/test/CodeGen/math-libcalls.c
@@ -23,19 +23,19 @@ void foo(double *d, float f, float *fp, long double *l, int *i, const char *c) {
 
   atan2(f,f);    atan2f(f,f) ;  atan2l(f, f);
 
-  // NO__ERRNO: declare double @atan2(double noundef, double noundef) [[READNONE:#[0-9]+]]
-  // NO__ERRNO: declare float @atan2f(float noundef, float noundef) [[READNONE]]
-  // NO__ERRNO: declare x86_fp80 @atan2l(x86_fp80 noundef, x86_fp80 noundef) [[READNONE]]
+  // NO__ERRNO: declare double @llvm.atan2.f64(double, double) [[READNONE_INTRINSIC:#[0-9]+]]
+  // NO__ERRNO: declare float @llvm.atan2.f32(float, float) [[READNONE_INTRINSIC]]
+  // NO__ERRNO: declare x86_fp80 @llvm.atan2.f80(x86_fp80, x86_fp80) [[READNONE_INTRINSIC]]
   // HAS_ERRNO: declare double @atan2(double noundef, double noundef) [[NOT_READNONE]]
   // HAS_ERRNO: declare float @atan2f(float noundef, float noundef) [[NOT_READNONE]]
   // HAS_ERRNO: declare x86_fp80 @atan2l(x86_fp80 noundef, x86_fp80 noundef) [[NOT_READNONE]]
-  // HAS_MAYTRAP: declare double @atan2(double noundef, double noundef) [[NOT_READNONE:#[0-9]+]]
-  // HAS_MAYTRAP: declare float @atan2f(float noundef, float noundef) [[NOT_READNONE]]
-  // HAS_MAYTRAP: declare x86_fp80 @atan2l(x86_fp80 noundef, x86_fp80 noundef) [[NOT_READNONE]]
+  // HAS_MAYTRAP: declare double @llvm.experimental.constrained.atan2.f64(
+  // HAS_MAYTRAP: declare float @llvm.experimental.constrained.atan2.f32(
+  // HAS_MAYTRAP: declare x86_fp80 @llvm.experimental.constrained.atan2.f80(
 
   copysign(f,f); copysignf(f,f);copysignl(f,f);
 
-  // NO__ERRNO: declare double @llvm.copysign.f64(double, double) [[READNONE_INTRINSIC:#[0-9]+]]
+  // NO__ERRNO: declare double @llvm.copysign.f64(double, double) [[READNONE_INTRINSIC]]
   // NO__ERRNO: declare float @llvm.copysign.f32(float, float) [[READNONE_INTRINSIC]]
   // NO__ERRNO: declare x86_fp80 @llvm.copysign.f80(x86_fp80, x86_fp80) [[READNONE_INTRINSIC]]
   // HAS_ERRNO: declare double @llvm.copysign.f64(double, double) [[READNONE_INTRINSIC:#[0-9]+]]
@@ -65,13 +65,13 @@ void foo(double *d, float f, float *fp, long double *l, int *i, const char *c) {
   // HAS_ERRNO: declare double @frexp(double noundef, ptr noundef) [[NOT_READNONE]]
   // HAS_ERRNO: declare float @frexpf(float noundef, ptr noundef) [[NOT_READNONE]]
   // HAS_ERRNO: declare x86_fp80 @frexpl(x86_fp80 noundef, ptr noundef) [[NOT_READNONE]]
-  // HAS_MAYTRAP: declare double @frexp(double noundef, ptr noundef) [[NOT_READNONE]]
+  // HAS_MAYTRAP: declare double @frexp(double noundef, ptr noundef) [[NOT_READNONE:#[0-9]+]]
   // HAS_MAYTRAP: declare float @frexpf(float noundef, ptr noundef) [[NOT_READNONE]]
   // HAS_MAYTRAP: declare x86_fp80 @frexpl(x86_fp80 noundef, ptr noundef) [[NOT_READNONE]]
 
   ldexp(f,f);    ldexpf(f,f);   ldexpl(f,f);
 
-  // NO__ERRNO: declare double @ldexp(double noundef, i32 noundef) [[READNONE]]
+  // NO__ERRNO: declare double @ldexp(double noundef, i32 noundef) [[READNONE:#[0-9]+]]
   // NO__ERRNO: declare float @ldexpf(float noundef, i32 noundef) [[READNONE]]
   // NO__ERRNO: declare x86_fp80 @ldexpl(x86_fp80 noundef, i32 noundef) [[READNONE]]
   // HAS_ERRNO: declare double @ldexp(double noundef, i32 noundef) [[NOT_READNONE]]
@@ -719,9 +719,9 @@ void foo(double *d, float f, float *fp, long double *l, int *i, const char *c) {
 // HAS_ERRNO: declare x86_fp80 @llvm.trunc.f80(x86_fp80) [[READNONE_INTRINSIC]]
 };
 
-// NO__ERRNO: attributes [[READNONE]] = { {{.*}}memory(none){{.*}} }
 // NO__ERRNO: attributes [[READNONE_INTRINSIC]] = { {{.*}}memory(none){{.*}} }
 // NO__ERRNO: attributes [[NOT_READNONE]] = { nounwind {{.*}} }
+// NO__ERRNO: attributes [[READNONE]] = { {{.*}}memory(none){{.*}} }
 // NO__ERRNO: attributes [[READONLY]] = { {{.*}}memory(read){{.*}} }
 
 // HAS_ERRNO: attributes [[NOT_READNONE]] = { nounwind {{.*}} }

diff --git a/clang/test/CodeGenCXX/builtin-calling-conv.cpp b/clang/test/CodeGenCXX/builtin-calling-conv.cpp
@@ -1,8 +1,8 @@
-// RUN: %clang_cc1 -triple x86_64-linux-pc -DREDECL -emit-llvm %s -o - | FileCheck %s -check-prefix LINUX
-// RUN: %clang_cc1 -triple spir-unknown-unknown -DREDECL -DSPIR -emit-llvm %s -o - | FileCheck %s -check-prefix SPIR
-// RUN: %clang_cc1 -triple x86_64-linux-pc -emit-llvm %s -o - | FileCheck %s -check-prefix LINUX
-// RUN: %clang_cc1 -triple spir-unknown-unknown -DSPIR -emit-llvm %s -o - | FileCheck %s -check-prefix SPIR
-// RUN: %clang_cc1 -triple i386-windows-pc -fdefault-calling-conv=stdcall -emit-llvm %s -o - | FileCheck %s -check-prefix WIN32
+// RUN: %clang_cc1 -triple x86_64-linux-pc -DREDECL -emit-llvm -fmath-errno %s -o - | FileCheck %s -check-prefix LINUX
+// RUN: %clang_cc1 -triple spir-unknown-unknown -DREDECL -DSPIR -emit-llvm -fmath-errno %s -o - | FileCheck %s -check-prefix SPIR
+// RUN: %clang_cc1 -triple x86_64-linux-pc -emit-llvm -fmath-errno %s -o - | FileCheck %s -check-prefix LINUX
+// RUN: %clang_cc1 -triple spir-unknown-unknown -DSPIR -emit-llvm -fmath-errno %s -o - | FileCheck %s -check-prefix SPIR
+// RUN: %clang_cc1 -triple i386-windows-pc -fdefault-calling-conv=stdcall -emit-llvm -fmath-errno %s -o - | FileCheck %s -check-prefix WIN32
 
 #ifdef REDECL
 namespace std {

diff --git a/clang/test/CodeGenOpenCL/builtins-f16.cl b/clang/test/CodeGenOpenCL/builtins-f16.cl
@@ -15,6 +15,9 @@ void test_half_builtins(half h0, half h1, half h2, int i0) {
   // CHECK: call half @llvm.atan.f16(half %h0)
   res = __builtin_atanf16(h0);
 
+  // CHECK: call half @llvm.atan2.f16(half %h0, half %h1)
+  res = __builtin_atan2f16(h0, h1);
+
   // CHECK: call half @llvm.copysign.f16(half %h0, half %h1)
   res = __builtin_copysignf16(h0, h1);