[libcxx-commits] [libcxx] [libc++] Add a thread-safe version of std::lgamma in the dylib (PR #153631)

[Nikolas Klauser via libcxx-commits]

================
@@ -55,6 +55,31 @@ inline _LIBCPP_HIDE_FROM_ABI double tgamma(_A1 __x) _NOEXCEPT {
   return __builtin_tgamma((double)__x);
 }
 
+// __lgamma_r
+
+struct __lgamma_result {
+  double __result;
+  int __sign;
+};
----------------
philnik777 wrote:

The underlying function we currently use. If we replace it with an llvm-libc variant in the future, we might be able to get a faster function if we don't provide that information. If it falls out as a byproduct of the actual calculation anyways this point is of course moot. That's why I'd like to know whether it makes a difference in the computation.

https://github.com/llvm/llvm-project/pull/153631