Emit constrained atan2 intrinsic for clang builtin #113636
Open
+53
−32
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
llvmbot
added
clang
|
@llvm/pr-subscribers-clang-codegen @llvm/pr-subscribers-backend-x86 Author: Tex Riddell (tex3d) ChangesThis change is part of this proposal: https://discourse.llvm.org/t/rfc-all-the-math-intrinsics/78294
Part of Implement the atan2 HLSL Function #70096. Full diff: https://github.com/llvm/llvm-project/pull/113636.diff 8 Files Affected:
diff --git a/clang/include/clang/Basic/Builtins.td b/clang/include/clang/Basic/Builtins.td
index 90475a361bb8f8..0d77f4105bb757 100644
--- 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
index e2d03eff8ab4a0..0bec8f32552110 100644
--- 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
index 48465df21cca19..bf107437fc63a3 100644
--- a/clang/test/CodeGen/X86/math-builtins.c
+++ b/clang/test/CodeGen/X86/math-builtins.c
@@ -45,10 +45,10 @@ 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]]
@@ -56,7 +56,7 @@ void foo(double *d, float f, float *fp, long double *l, int *i, const char *c) {
__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
index aa77620b445356..68b9e75283c547 100644
--- 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
index b1637121127c5b..1e4b06e34aaf92 100644
--- 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
index 2226212eca94ee..bcc61c8f046b43 100644
--- 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
index 7020d1e0a24144..92d698b43be0ab 100644
--- a/clang/test/CodeGenCXX/builtin-calling-conv.cpp
+++ b/clang/test/CodeGenCXX/builtin-calling-conv.cpp
@@ -13,7 +13,7 @@ using size_t = unsigned long;
#endif // SPIR
} // namespace std
-float __builtin_atan2f(float, float);
+float __builtin_erff(float);
void *operator new(std::size_t);
#endif // REDECL
@@ -22,32 +22,32 @@ void foo();
void user() {
int i;
::operator new(5);
- (void)__builtin_atan2f(1.1, 2.2);
+ (void)__builtin_erff(1.1);
foo();
}
// LINUX: define{{.*}} void @_Z4userv()
// LINUX: call noalias noundef nonnull ptr @_Znwm
-// LINUX: call float @atan2f
+// LINUX: call float @erff
// LINUX: call void @_Z3foov
// LINUX: declare noundef nonnull ptr @_Znwm(i64 noundef)
-// LINUX: declare float @atan2f(float noundef, float noundef)
+// LINUX: declare float @erff(float noundef)
// LINUX: declare void @_Z3foov()
// SPIR: define{{.*}} spir_func void @_Z4userv()
// SPIR: call spir_func noalias noundef nonnull ptr @_Znwj
-// SPIR: call spir_func float @atan2f
+// SPIR: call spir_func float @erff
// SPIR: call spir_func void @_Z3foov
// SPIR: declare spir_func noundef nonnull ptr @_Znwj(i32 noundef)
-// SPIR: declare spir_func float @atan2f(float noundef, float noundef)
+// SPIR: declare spir_func float @erff(float noundef)
// SPIR: declare spir_func void @_Z3foov()
// Note: Windows /G options should not change the platform default calling
// convention of builtins.
// WIN32: define dso_local x86_stdcallcc void @"?user@@YGXXZ"()
// WIN32: call noalias noundef nonnull ptr @"??2@YAPAXI@Z"
-// WIN32: call float @atan2f
+// WIN32: call float @erff
// WIN32: call x86_stdcallcc void @"?foo@@YGXXZ"
// WIN32: declare dso_local noundef nonnull ptr @"??2@YAPAXI@Z"(
-// WIN32: declare dso_local float @atan2f(float noundef, float noundef)
+// WIN32: declare dso_local float @erff(float noundef)
// WIN32: declare dso_local x86_stdcallcc void @"?foo@@YGXXZ"()
diff --git a/clang/test/CodeGenOpenCL/builtins-f16.cl b/clang/test/CodeGenOpenCL/builtins-f16.cl
index 8150bc1ac9e2d7..e8b62fe0830cdb 100644
--- 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);
|
|
✅ With the latest revision this PR passed the C/C++ code formatter. |
hekota
approved these changes
Nov 4, 2024
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.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
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 builtinCGBuiltin.cpp- Emit constraint atan2 intrinsic for clang builtinclang/test/CodeGenCXX/builtin-calling-conv.cpp- Use erff instead of atan2 for clang builtin to lib call calling convention check, now that atan2 maps to an intrinsic.Part of Implement the atan2 HLSL Function #70096.