[C++23] [CLANG] Adding C++23 constexpr math functions: fmin, fmax and frexp. #88978

zahiraam · 2024-04-16T20:11:10Z

Add support for fmin, fmax, and (__builtin_)frexp.

clang/lib/AST/ExprConstant.cpp

clang/test/CodeGen/constexpr-math.cpp

clang/lib/AST/ExprConstant.cpp

llvmbot · 2024-04-22T20:35:28Z

@llvm/pr-subscribers-llvm-adt

@llvm/pr-subscribers-clang

Author: Zahira Ammarguellat (zahiraam)

Changes

Add support for for fmin and frexp.

Full diff: https://github.com/llvm/llvm-project/pull/88978.diff

4 Files Affected:

(modified) clang/include/clang/Basic/Builtins.td (+2-2)
(modified) clang/lib/AST/ExprConstant.cpp (+14-1)
(added) clang/test/CodeGenCXX/constexpr-math.cpp (+55)
(added) clang/test/SemaCXX/constexpr-math.cpp (+57)

diff --git a/clang/include/clang/Basic/Builtins.td b/clang/include/clang/Basic/Builtins.td
index 52c0dd52c28b11..a35c77286229ff 100644
--- a/clang/include/clang/Basic/Builtins.td
+++ b/clang/include/clang/Basic/Builtins.td
@@ -3440,7 +3440,7 @@ def Fmod : FPMathTemplate, LibBuiltin<"math.h"> {
 
 def Frexp : FPMathTemplate, LibBuiltin<"math.h"> {
   let Spellings = ["frexp"];
-  let Attributes = [NoThrow];
+  let Attributes = [NoThrow, Constexpr];
   let Prototype = "T(T, int*)";
   let AddBuiltinPrefixedAlias = 1;
 }
@@ -3618,7 +3618,7 @@ def Fmax : FPMathTemplate, LibBuiltin<"math.h"> {
 
 def Fmin : FPMathTemplate, LibBuiltin<"math.h"> {
   let Spellings = ["fmin"];
-  let Attributes = [NoThrow, Const];
+  let Attributes = [NoThrow, Const, Constexpr];
   let Prototype = "T(T, T)";
   let AddBuiltinPrefixedAlias = 1;
   let OnlyBuiltinPrefixedAliasIsConstexpr = 1;
diff --git a/clang/lib/AST/ExprConstant.cpp b/clang/lib/AST/ExprConstant.cpp
index 5a36621dc5cce2..1a6abb386071c7 100644
--- a/clang/lib/AST/ExprConstant.cpp
+++ b/clang/lib/AST/ExprConstant.cpp
@@ -2922,7 +2922,7 @@ static bool handleFloatFloatBinOp(EvalInfo &Info, const BinaryOperator *E,
   //   If during the evaluation of an expression, the result is not
   //   mathematically defined [...], the behavior is undefined.
   // FIXME: C++ rules require us to not conform to IEEE 754 here.
-  if (LHS.isNaN()) {
+  if (!Info.getLangOpts().CPlusPlus23 && LHS.isNaN()) {
     Info.CCEDiag(E, diag::note_constexpr_float_arithmetic) << LHS.isNaN();
     return Info.noteUndefinedBehavior();
   }
@@ -14547,6 +14547,17 @@ bool FloatExprEvaluator::VisitCallExpr(const CallExpr *E) {
   default:
     return false;
 
+  case Builtin::BI__builtin_frexpf:
+  case Builtin::BI__builtin_frexp: {
+    LValue Pointer;
+    if (!EvaluateFloat(E->getArg(0), Result, Info) ||
+        !EvaluatePointer(E->getArg(1), Pointer, Info))
+      return false;
+    llvm::RoundingMode RM = getActiveRoundingMode(Info, E);
+    int FrexpExp;
+    Result = llvm::frexp(Result, FrexpExp, RM);
+    return true;
+  }
   case Builtin::BI__builtin_huge_val:
   case Builtin::BI__builtin_huge_valf:
   case Builtin::BI__builtin_huge_vall:
@@ -14638,6 +14649,8 @@ bool FloatExprEvaluator::VisitCallExpr(const CallExpr *E) {
     return true;
   }
 
+  case Builtin::BIfmin:
+  case Builtin::BIfminf:
   case Builtin::BI__builtin_fmin:
   case Builtin::BI__builtin_fminf:
   case Builtin::BI__builtin_fminl:
diff --git a/clang/test/CodeGenCXX/constexpr-math.cpp b/clang/test/CodeGenCXX/constexpr-math.cpp
new file mode 100644
index 00000000000000..ad7b6779a4ae0e
--- /dev/null
+++ b/clang/test/CodeGenCXX/constexpr-math.cpp
@@ -0,0 +1,55 @@
+// RUN: %clang_cc1 -x c++ -triple x86_64-unknown-unknown -std=c++23 \
+// RUN: -DWIN -emit-llvm -o - %s | FileCheck %s --check-prefixes=WIN
+
+// RUN: %clang_cc1 -x c++ -triple x86_64-unknown-unknown -std=c++23 \
+// RUN: -emit-llvm -o - %s | FileCheck %s --check-prefixes=LNX
+
+#ifdef WIN
+#define INFINITY ((float)(1e+300 * 1e+300))
+#define NAN      (-(float)(INFINITY * 0.0F))
+#else
+#define NAN (__builtin_nanf(""))
+#define INFINITY (__builtin_inff())
+#endif
+
+int func()
+{
+  int i;
+
+  // fmin
+  constexpr double f1 = __builtin_fmin(15.24, 1.3);
+  constexpr double f2 = __builtin_fmin(-0.0, +0.0);
+  constexpr double f3 = __builtin_fmin(+0.0, -0.0);
+  constexpr float f4 = __builtin_fminf(NAN, NAN);
+  constexpr float f5 = __builtin_fminf(NAN, -1);
+  constexpr float f6 = __builtin_fminf(-INFINITY, 0);
+  constexpr float f7 = __builtin_fminf(INFINITY, 0);
+
+  // frexp
+  constexpr double f8 = __builtin_frexp(123.45, &i);
+  constexpr double f9 = __builtin_frexp(0.0, &i);
+  constexpr double f10 = __builtin_frexp(-0.0, &i);
+  constexpr double f11 = __builtin_frexpf(NAN, &i);
+  constexpr double f12 = __builtin_frexpf(-NAN, &i);
+  constexpr double f13 = __builtin_frexpf(INFINITY, &i);
+  constexpr double f14 = __builtin_frexpf(INFINITY, &i);
+
+  return 0;
+}
+
+// CHECK: store double 1.300000e+00, ptr {{.*}}
+// CHECK: store double -0.000000e+00, ptr {{.*}}
+// CHECK: store double -0.000000e+00, ptr {{.*}}
+// WIN: store float 0xFFF8000000000000, ptr {{.*}}
+// LNX: store float 0x7FF8000000000000, ptr {{.*}}
+// CHECK: store float -1.000000e+00, ptr {{.*}}
+// CHECK: store float 0xFFF0000000000000, ptr {{.*}}
+
+// CHECK: store double 0x3FEEDCCCCCCCCCCD, ptr {{.*}}
+// CHECK: store double 0.000000e+00, ptr {{.*}}
+// CHECK: store double -0.000000e+00, ptr {{.*}}
+// CHECK: store double 0xFFF8000000000000, ptr {{.*}}
+// WIN: store double 0x7FF8000000000000, ptr {{.*}}
+// LNX: store double 0xFFF8000000000000, ptr {{.*}}
+// CHECK: store double 0x7FF0000000000000, ptr {{.*}}
+// CHECK: store double 0x7FF0000000000000, ptr {{.*}}
diff --git a/clang/test/SemaCXX/constexpr-math.cpp b/clang/test/SemaCXX/constexpr-math.cpp
new file mode 100644
index 00000000000000..f51ea1bcaf7bad
--- /dev/null
+++ b/clang/test/SemaCXX/constexpr-math.cpp
@@ -0,0 +1,57 @@
+// RUN: %clang_cc1 -DWIN -verify -std=c++23 -fsyntax-only  %s
+// RUN: %clang_cc1 -verify -std=c++23 -fsyntax-only  %s
+
+// expected-no-diagnostics
+
+
+#ifdef WIN
+#define INFINITY ((float)(1e+300 * 1e+300))
+#define NAN      (-(float)(INFINITY * 0.0F))
+#else
+#define NAN (__builtin_nanf(""))
+#define INFINITY (__builtin_inff())
+#endif
+
+extern "C" void abort() noexcept;
+extern "C" int write(int, const void*, unsigned long);
+
+#define assert(condition) \
+  do {			    \
+    if (!(condition)) {					\
+      write(2, "Assertion failed: ", 18);		\
+      write(2, #condition, sizeof(#condition) - 1);	\
+      write(2, "\n", 1);				\
+      abort();						\
+    }							\
+  } while (false)
+
+int main() {
+    int i;
+
+    // fmin
+    static_assert(__builtin_fmin(15.24, 1.3) == 1.3, "");
+    static_assert(__builtin_fmin(-0.0, +0.0) == -0.0, "");
+    static_assert(__builtin_fmin(+0.0, -0.0) == -0.0, "");
+    assert(__builtin_isnan(__builtin_fminf(NAN,NAN)));
+    assert(__builtin_isnan(__builtin_fminf(NAN, -1)));
+    assert(__builtin_isnan(__builtin_fminf(-INFINITY, 0)));
+    assert(__builtin_iszero(__builtin_fminf(+INFINITY, 0)));
+
+    // frexp
+    static_assert(__builtin_frexp(123.45, &i) == 0.96445312500000002);
+    static_assert(!__builtin_isnan(__builtin_frexp(123.45, &i)), "");
+    assert(i==0);
+    static_assert(__builtin_iszero(__builtin_frexp(0.0, &i)), "");
+    assert(i==0);
+    static_assert(__builtin_iszero(__builtin_frexp(-0.0, &i)), "");
+    assert(i==0);
+    assert(__builtin_isnan(__builtin_frexp(NAN, &i)));
+    assert(i==0);
+    assert(__builtin_isnan(__builtin_frexp(-NAN, &i)));
+    assert(i==0);
+    assert(!__builtin_isfinite(__builtin_frexp(INFINITY, &i)));
+    assert(i==0);
+    assert(!__builtin_isfinite(__builtin_frexp(-INFINITY, &i)));
+    assert(i==0);
+    return 0;
+}

jcranmer-intel

Other side notes:

fmin and frexp can signal exceptions if the input is an sNaN, which causes [library.c]p3 to kick in. (That's the only time these operations can signal an exception.)

clang/test/SemaCXX/constexpr-math.cpp

clang/lib/AST/ExprConstant.cpp

clang/test/CodeGenCXX/constexpr-math.cpp

hubert-reinterpretcast · 2024-04-26T06:12:47Z

clang/lib/AST/ExprConstant.cpp

+  case Builtin::BIfmin:
+  case Builtin::BIfminf:


Don't make these evaluate in constexpr. Not only does this extend C (not a stated intent of the patch), it will cause accidental dependencies that break when -fno-builtin is used.

These have to be made constexpr, at least in C++23. C++ stdlibs can't override them to make them call the __builtin_ versions, so Clang has to handle that.

I think we'll be missing at least "pedantic" diagnostics if we don't restrict the constexpr-ness to C++23 and up. So we do need that restriction.

The interaction with -fno-builtin[-*] may need more discussion. At least the documentation for that option would need to be updated if the C++23 constexpr math would break with it.

At the Clang C/C++ Language Workgroup call on 2024-05-15, it was agreed that -fno-builtin[-*] should remain capable of disabling built-in treatment even when that treatment would be required for conformance w.r.t. constexpr behaviour of functions shared with C.

For me, the first expectation for this PR (on that subject) is an update to the documentation of -fno-builtin[-*].

@AaronBallman @ldionne, I am no sure how we want to handle this w.r.t. documenting the C++23 implementation status. Does it go in both the Clang documentation (for the functions shared with C) and the libc++ documentation (for the overloads added by C++)?

I don't think anything should happen to the libc++ documentation in this patch. We can update the implementation to behave the same across the different overloads in a follow-up and update our documentation then.

clang/include/clang/Basic/Builtins.td

clang/test/SemaCXX/constexpr-math.cpp

clang/lib/AST/ExprConstant.cpp

clang/test/CodeGenCXX/constexpr-math.cpp

clang/test/SemaCXX/constexpr-math.cpp

as requested by reviewer.

zahiraam · 2024-05-08T17:32:41Z

@hubert-reinterpretcast Thanks for the reviews/suggestions. Let me know what you think.

clang/lib/AST/ExprConstant.cpp

zahiraam · 2024-07-25T20:15:17Z

@hubert-reinterpretcast Would really appreciate a review of this please. Thanks.

jcranmer-intel · 2024-07-25T20:47:00Z

clang/lib/AST/ExprConstant.cpp

+  llvm::APFloat minusOne(x.getSemantics(), "-1.0");
+  llvm::APFloat minusHalf(x.getSemantics(), "-0.5");
+  llvm::APFloat half(x.getSemantics(), "0.5");
+  llvm::APFloat one(x.getSemantics(), "1.0");
+
+  return ((x.compare(minusOne) == llvm::APFloat::cmpGreaterThan &&
+           x.compare(minusHalf) != llvm::APFloat::cmpGreaterThan) ||
+          (x.compare(half) != llvm::APFloat::cmpLessThan &&
+           x.compare(one) == llvm::APFloat::cmpLessThan));


You can simplify this by checking that the absolute value is in [0.5, 1)

format.

hubert-reinterpretcast · 2024-07-30T02:45:04Z

clang/include/clang/Basic/Builtins.td

@@ -253,8 +253,10 @@ def FmodF16F128 : F16F128MathTemplate, Builtin {

 def FrexpF16F128 : F16F128MathTemplate, Builtin {
  let Spellings = ["__builtin_frexp"];
-  let Attributes = [FunctionWithBuiltinPrefix, NoThrow];
+  let Attributes = [FunctionWithBuiltinPrefix, NoThrow, Const, ConstexprSince];


The frexp family of functions cannot be marked const because they modify the object pointed to by the pointer parameter.

hubert-reinterpretcast · 2024-07-30T02:57:15Z

clang/test/CodeGenCXX/constexpr-math.cpp

+  constexpr float frexp9 = __builtin_frexpf16(3.5, &i);
+  constexpr long double frexp10 = __builtin_frexpf128(234.78, &i);


Suggested change

constexpr float frexp9 = __builtin_frexpf16(3.5, &i);

constexpr long double frexp10 = __builtin_frexpf128(234.78, &i);

constexpr __fp16 frexp9 = __builtin_frexpf16(3.5, &i);

constexpr __float128 frexp10 = __builtin_frexpf128(234.78, &i);

hubert-reinterpretcast · 2024-07-30T03:43:13Z

clang/test/CodeGenCXX/constexpr-math.cpp

+  constexpr long double max8 = __builtin_fmaxl(123.456L, 789.012L);
+
+  // frexp
+  constexpr double frexp1 = __builtin_frexp(123.45, &i);


@zahiraam, to be explicit, these tests (as-is) will fail once the comment re: StoreExponent() is addressed.

This will probably work:

Suggested change

constexpr double frexp1 = __builtin_frexp(123.45, &i);

constexpr double frexp1 = __builtin_frexp(123.45, (int [1]){});

hubert-reinterpretcast · 2024-07-30T03:52:01Z

clang/test/SemaCXX/constexpr-math.cpp

+  // cplusplus20andless-error@-1 {{non-type template argument is not a constant expression}}
+  static_assert(is_same_val<CALL_BUILTIN(frexpl, 123.45L, &i), 123.45L/128>);
+  // cplusplus20andless-error@-1 {{non-type template argument is not a constant expression}}
+


Add the __fp16 and __float128 cases too.

tests.

zahiraam · 2024-08-12T20:18:41Z

Ping?

hubert-reinterpretcast · 2024-08-12T21:35:37Z

Ping?

The changes to add the "constexpr since" property are missing changes to the tablegen utility. Figuring out how it would work is more involved than I have the time for right now.

Also, if at all possible, I think we only want the extra field for "library builtins" and not all builtins.

Alternatively, we don't need tablegen changes if we leave the encoding of which language standard a builtin is constexpr since to a helper function.

zahiraam · 2024-08-13T12:21:38Z

Also, if at all possible, I think we only want the extra field for "library builtins" and not all builtins.

Yes, I have tried limiting it to LibBuiltin but the addition of FrexpF16F128 led me to add it Builtin too.

Alternatively, we don't need tablegen changes if we leave the encoding of which language standard a builtin is constexpr since to a helper function.

Will look into this.

Thanks.

hubert-reinterpretcast · 2024-08-13T15:38:21Z

Yes, I have tried limiting it to LibBuiltin but the addition of FrexpF16F128 led me to add it Builtin too.

I believe that FrexpF16F128 only has the __builtin_-prefixed versions. That would mean that it does not need to be "constexpr since" (just plain "constexpr").

function.

zahiraam · 2024-08-22T15:06:41Z

Alternatively, we don't need tablegen changes if we leave the encoding of which language standard a builtin is constexpr since to a helper function.

@hubert-reinterpretcast I have implemented what you proposed. Removed the ConstExprSince attribute and replaced it with member function. Is that what you were thinking of?

zahiraam · 2024-09-05T20:47:38Z

ping @hubert-reinterpretcast

cor3ntin · 2025-05-13T14:38:46Z

clang/lib/AST/ExprConstant.cpp

+    const auto *FDecl = E->getDirectCallee();
+    clang::LangStandard::Kind ConstExprSinceVersion =
+        FDecl->getConstexprBuiltinSinceVersion();
+    if (ConstExprSinceVersion > Info.Ctx.getLangOpts().LangStd)
+      return false;
+    LValue Pointer;
+    if (!EvaluateFloat(E->getArg(0), Result, Info) ||
+        !EvaluatePointer(E->getArg(1), Pointer, Info))


I think the builtins should be unconditionally constexpr in Clang.

And we should certainly not store additional info for that in FunctionDecl, which has a huge cost in terms of AST size.

Instead, the builtin should become constexpr, and libc++ can decide not to make the wrapper constexpr in older language modes, or to have some availability warning in older language modes.

@cor3ntin, you mean the __builtin_-prefixed form, right?

@hubert-reinterpretcast I think if we wanted to do something about the non-prefixed builtin, I would prefer

that we table-gen the information

that the version is encoded as an attribute (it might just be an implementation detail attribute) - so as to not bloat the AST

that we make it a diagnostic that users can turn into an error or not (and then we can discuss what the default should be)

Adding C23 constexpr math functions fmin and frexp.

3acc848

zahiraam changed the title ~~Adding C23 constexpr math functions fmin and frexp.~~ [C23] [CALNG] Adding C23 constexpr math functions fmin and frexp. Apr 16, 2024

zahiraam changed the title ~~[C23] [CALNG] Adding C23 constexpr math functions fmin and frexp.~~ [C23] [CLANG] Adding C23 constexpr math functions: fmin and frexp. Apr 17, 2024

zahiraam requested a review from AaronBallman April 17, 2024 13:30

This comment was marked as resolved.

Sign in to view

zahiraam changed the title ~~[C23] [CLANG] Adding C23 constexpr math functions: fmin and frexp.~~ [C++23] [CLANG] Adding C++23 constexpr math functions: fmin and frexp. Apr 17, 2024

tbaederr reviewed Apr 18, 2024

View reviewed changes

clang/lib/AST/ExprConstant.cpp Outdated Show resolved Hide resolved

clang/lib/AST/ExprConstant.cpp Outdated Show resolved Hide resolved

clang/test/CodeGen/constexpr-math.cpp Outdated Show resolved Hide resolved

clang/lib/AST/ExprConstant.cpp Show resolved Hide resolved

Addressed review comments.

d7050b5

zahiraam marked this pull request as ready for review April 22, 2024 20:34

llvmbot added clang Clang issues not falling into any other category clang:frontend Language frontend issues, e.g. anything involving "Sema" labels Apr 22, 2024

jcranmer-intel reviewed Apr 23, 2024

View reviewed changes

hubert-reinterpretcast requested changes Apr 26, 2024

View reviewed changes

clang/test/SemaCXX/constexpr-math.cpp Outdated Show resolved Hide resolved

clang/lib/AST/ExprConstant.cpp Outdated Show resolved Hide resolved

hubert-reinterpretcast reviewed Apr 26, 2024

View reviewed changes

Responded to review comments.

625028a

This comment was marked as outdated.

Sign in to view

hubert-reinterpretcast reviewed May 5, 2024

View reviewed changes

clang/lib/AST/ExprConstant.cpp Outdated Show resolved Hide resolved

hubert-reinterpretcast reviewed May 5, 2024

View reviewed changes

Added code to store the exponent in Pointer. Edited LIT tests

319fdd0

as requested by reviewer.

Added fmax function and function IsinRange function.

da06dba

llvmbot added the llvm:adt label May 9, 2024

zahiraam changed the title ~~[C++23] [CLANG] Adding C++23 constexpr math functions: fmin and frexp.~~ [C++23] [CLANG] Adding C++23 constexpr math functions: fmin, fmax and frexp. May 9, 2024

This comment was marked as resolved.

Sign in to view

zahiraam added 2 commits May 10, 2024 07:14

Addressed review comments.

1793cf0

Added a fix for failing LIT test.

42d7d4c

hubert-reinterpretcast reviewed May 16, 2024

View reviewed changes

clang/lib/AST/ExprConstant.cpp Outdated Show resolved Hide resolved

zahiraam added 2 commits May 17, 2024 05:17

Fixed aix-builtin-mapping.c test.

9f4d632

Merge remote-tracking branch 'origin/main' into C23ConstExpr

efeec9b

zahiraam requested review from frederick-vs-ja, philnik777, tbaederr and jcranmer-intel July 19, 2024 18:19

jcranmer-intel reviewed Jul 25, 2024

View reviewed changes

zahiraam and others added 6 commits July 26, 2024 05:31

Fix build issue in flang.

49186b9

Merge remote-tracking branch 'origin/main' into C23ConstExpr

cf2814d

Merge remote-tracking branch 'origin/main' into C23ConstExpr

1c2912c

Rewrote isInFrexpResultRange according to review comments and fixed

d32737e

format.

Fix format.

bd536f5

Addressed missed review comments.

32b788a

hubert-reinterpretcast reviewed Jul 30, 2024

View reviewed changes

Addressed review comments. Remove StoreExponent function and fixed LIT

501e39a

tests.

zahiraam added 2 commits August 13, 2024 13:02

Replaced ConstexpSince by Constexpr for FrexpF16F128.

0cc61eb

Removed builtin attribute ConstExprSince and replaced it with a member

c615b22

function.

zahiraam added 3 commits August 22, 2024 08:54

Fix format.

5eec2db

Fix LIT failures.

299df9a

Merge remote-tracking branch 'origin/main' into C23ConstExpr

1507a33

philnik777 mentioned this pull request Oct 21, 2024

[clang] constexpr built-in fma function. #113020

Closed

cor3ntin reviewed May 13, 2025

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[C++23] [CLANG] Adding C++23 constexpr math functions: fmin, fmax and frexp. #88978

[C++23] [CLANG] Adding C++23 constexpr math functions: fmin, fmax and frexp. #88978

zahiraam commented Apr 16, 2024 •

edited by hubert-reinterpretcast

Loading

This comment was marked as resolved.

llvmbot commented Apr 22, 2024 •

edited

Loading

jcranmer-intel left a comment

hubert-reinterpretcast Apr 26, 2024

philnik777 Apr 28, 2024

This comment was marked as resolved.

This comment was marked as resolved.

This comment was marked as resolved.

hubert-reinterpretcast Apr 29, 2024

hubert-reinterpretcast May 18, 2024

philnik777 Aug 17, 2024

This comment was marked as outdated.

zahiraam commented May 8, 2024

This comment was marked as resolved.

zahiraam commented Jul 25, 2024

jcranmer-intel Jul 25, 2024

hubert-reinterpretcast Jul 30, 2024

zahiraam Aug 5, 2024

hubert-reinterpretcast Jul 30, 2024

zahiraam Aug 5, 2024

hubert-reinterpretcast Jul 30, 2024

zahiraam Aug 5, 2024

hubert-reinterpretcast Jul 30, 2024

zahiraam Aug 5, 2024

zahiraam commented Aug 12, 2024

hubert-reinterpretcast commented Aug 12, 2024 •

edited

Loading

zahiraam commented Aug 13, 2024

hubert-reinterpretcast commented Aug 13, 2024

zahiraam commented Aug 22, 2024

zahiraam commented Sep 5, 2024

cor3ntin May 13, 2025 •

edited

Loading

hubert-reinterpretcast May 13, 2025

cor3ntin May 13, 2025

		constexpr float frexp9 = __builtin_frexpf16(3.5, &i);
		constexpr long double frexp10 = __builtin_frexpf128(234.78, &i);

	constexpr double frexp1 = __builtin_frexp(123.45, &i);
	constexpr double frexp1 = __builtin_frexp(123.45, (int [1]){});

[C++23] [CLANG] Adding C++23 constexpr math functions: fmin, fmax and frexp. #88978

Are you sure you want to change the base?

[C++23] [CLANG] Adding C++23 constexpr math functions: fmin, fmax and frexp. #88978

Conversation

zahiraam commented Apr 16, 2024 • edited by hubert-reinterpretcast Loading

This comment was marked as resolved.

llvmbot commented Apr 22, 2024 • edited Loading

jcranmer-intel left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

This comment was marked as resolved.

This comment was marked as resolved.

This comment was marked as resolved.

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

This comment was marked as outdated.

zahiraam commented May 8, 2024

This comment was marked as resolved.

zahiraam commented Jul 25, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

zahiraam commented Aug 12, 2024

hubert-reinterpretcast commented Aug 12, 2024 • edited Loading

zahiraam commented Aug 13, 2024

hubert-reinterpretcast commented Aug 13, 2024

zahiraam commented Aug 22, 2024

zahiraam commented Sep 5, 2024

cor3ntin May 13, 2025 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

zahiraam commented Apr 16, 2024 •

edited by hubert-reinterpretcast

Loading

llvmbot commented Apr 22, 2024 •

edited

Loading

hubert-reinterpretcast commented Aug 12, 2024 •

edited

Loading

cor3ntin May 13, 2025 •

edited

Loading