libstdc++: Fix up <complex> for extended floating point types [PR107649]

  Hi!

As filed by Jonathan in the PR, I've screwed up the requires syntax
in the extended floating point specialization:
-    requires(__complex_type<_Tp>::type)
+    requires requires { typename __complex_type<_Tp>::type; }
and doing this change resulted in lots of errors because __complex_whatever
overfloads from extended floating point types were declared after the
templates which used them.

The following patch fixes that.

Bootstrapped/regtested on x86_64-linux and i686-linux, additionally
I've tested that with _GLIBCXX_HAVE_FLOAT128_MATH not being defined
while __STDCPP_FLOAT128_T__ defined one can still use
std::complex<std::float128_t> for basic arithmetic etc., just one can't
expect std::sin etc. to work in that case (because we don't have any
implementation).

Ok for trunk?

2022-11-14  Jakub Jelinek  <jakub@redhat.com>
	    Jonathan Wakely  <jwakely@redhat.com>

	PR libstdc++/107649
	* include/std/complex (__complex_abs, __complex_arg, __complex_cos,
	__complex_cosh, __complex_exp, __complex_log, __complex_sin,
	__complex_sinh, __complex_sqrt, __complex_tan, __complex_tanh,
	__complex_pow): Move __complex__ _Float{16,32,64,128} and
	__complex__ decltype(0.0bf16) overloads earlier in the file.
	(complex): Fix up requires on the partial specialization for extended
	float types.
	(__complex_acos, __complex_asin, __complex_atan, __complex_acosh,
	__complex_asinh, __complex_atanh): Move
	__complex__ _Float{16,32,64,128} and __complex__ decltype(0.0bf16)
	overloads earlier in the file.


	Jakub
  

On Mon, 14 Nov 2022 at 13:57, Jakub Jelinek <jakub@redhat.com> wrote:

> Hi!
>
> As filed by Jonathan in the PR, I've screwed up the requires syntax
> in the extended floating point specialization:
> -    requires(__complex_type<_Tp>::type)
> +    requires requires { typename __complex_type<_Tp>::type; }
> and doing this change resulted in lots of errors because __complex_whatever
> overfloads from extended floating point types were declared after the
> templates which used them.
>
> The following patch fixes that.
>
> Bootstrapped/regtested on x86_64-linux and i686-linux, additionally
> I've tested that with _GLIBCXX_HAVE_FLOAT128_MATH not being defined
> while __STDCPP_FLOAT128_T__ defined one can still use
> std::complex<std::float128_t> for basic arithmetic etc., just one can't
> expect std::sin etc. to work in that case (because we don't have any
> implementation).
>
> Ok for trunk?
>

OK, thanks for the quick fix.



>
> 2022-11-14  Jakub Jelinek  <jakub@redhat.com>
>             Jonathan Wakely  <jwakely@redhat.com>
>
>         PR libstdc++/107649
>         * include/std/complex (__complex_abs, __complex_arg, __complex_cos,
>         __complex_cosh, __complex_exp, __complex_log, __complex_sin,
>         __complex_sinh, __complex_sqrt, __complex_tan, __complex_tanh,
>         __complex_pow): Move __complex__ _Float{16,32,64,128} and
>         __complex__ decltype(0.0bf16) overloads earlier in the file.
>         (complex): Fix up requires on the partial specialization for
> extended
>         float types.
>         (__complex_acos, __complex_asin, __complex_atan, __complex_acosh,
>         __complex_asinh, __complex_atanh): Move
>         __complex__ _Float{16,32,64,128} and __complex__ decltype(0.0bf16)
>         overloads earlier in the file.
>
> --- libstdc++-v3/include/std/complex.jj 2022-10-31 20:15:49.756552019 +0100
> +++ libstdc++-v3/include/std/complex    2022-11-12 01:56:42.970560123 +0100
> @@ -598,6 +598,264 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
>      { return __z.imag(); }
>  #endif
>
> +#if _GLIBCXX_USE_C99_COMPLEX
> +#if defined(__STDCPP_FLOAT16_T__) &&
> defined(_GLIBCXX_FLOAT_IS_IEEE_BINARY32)
> +  inline _Float16
> +  __complex_abs(__complex__ _Float16 __z)
> +  { return _Float16(__builtin_cabsf(__z)); }
> +
> +  inline _Float16
> +  __complex_arg(__complex__ _Float16 __z)
> +  { return _Float16(__builtin_cargf(__z)); }
> +
> +  inline __complex__ _Float16
> +  __complex_cos(__complex__ _Float16 __z)
> +  { return static_cast<__complex__ _Float16>(__builtin_ccosf(__z)); }
> +
> +  inline __complex__ _Float16
> +  __complex_cosh(__complex__ _Float16 __z)
> +  { return static_cast<__complex__ _Float16>(__builtin_ccoshf(__z)); }
> +
> +  inline __complex__ _Float16
> +  __complex_exp(__complex__ _Float16 __z)
> +  { return static_cast<__complex__ _Float16>(__builtin_cexpf(__z)); }
> +
> +  inline __complex__ _Float16
> +  __complex_log(__complex__ _Float16 __z)
> +  { return static_cast<__complex__ _Float16>(__builtin_clogf(__z)); }
> +
> +  inline __complex__ _Float16
> +  __complex_sin(__complex__ _Float16 __z)
> +  { return static_cast<__complex__ _Float16>(__builtin_csinf(__z)); }
> +
> +  inline __complex__ _Float16
> +  __complex_sinh(__complex__ _Float16 __z)
> +  { return static_cast<__complex__ _Float16>(__builtin_csinhf(__z)); }
> +
> +  inline __complex__ _Float16
> +  __complex_sqrt(__complex__ _Float16 __z)
> +  { return static_cast<__complex__ _Float16>(__builtin_csqrtf(__z)); }
> +
> +  inline __complex__ _Float16
> +  __complex_tan(__complex__ _Float16 __z)
> +  { return static_cast<__complex__ _Float16>(__builtin_ctanf(__z)); }
> +
> +  inline __complex__ _Float16
> +  __complex_tanh(__complex__ _Float16 __z)
> +  { return static_cast<__complex__ _Float16>(__builtin_ctanhf(__z)); }
> +
> +  inline __complex__ _Float16
> +  __complex_pow(__complex__ _Float16 __x, __complex__ _Float16 __y)
> +  { return static_cast<__complex__ _Float16>(__builtin_cpowf(__x, __y)); }
> +#endif
> +
> +#if defined(__STDCPP_FLOAT32_T__) &&
> defined(_GLIBCXX_FLOAT_IS_IEEE_BINARY32)
> +  inline _Float32
> +  __complex_abs(__complex__ _Float32 __z) { return __builtin_cabsf(__z); }
> +
> +  inline _Float32
> +  __complex_arg(__complex__ _Float32 __z) { return __builtin_cargf(__z); }
> +
> +  inline __complex__ _Float32
> +  __complex_cos(__complex__ _Float32 __z) { return __builtin_ccosf(__z); }
> +
> +  inline __complex__ _Float32
> +  __complex_cosh(__complex__ _Float32 __z) { return
> __builtin_ccoshf(__z); }
> +
> +  inline __complex__ _Float32
> +  __complex_exp(__complex__ _Float32 __z) { return __builtin_cexpf(__z); }
> +
> +  inline __complex__ _Float32
> +  __complex_log(__complex__ _Float32 __z) { return __builtin_clogf(__z); }
> +
> +  inline __complex__ _Float32
> +  __complex_sin(__complex__ _Float32 __z) { return __builtin_csinf(__z); }
> +
> +  inline __complex__ _Float32
> +  __complex_sinh(__complex__ _Float32 __z) { return
> __builtin_csinhf(__z); }
> +
> +  inline __complex__ _Float32
> +  __complex_sqrt(__complex__ _Float32 __z) { return
> __builtin_csqrtf(__z); }
> +
> +  inline __complex__ _Float32
> +  __complex_tan(__complex__ _Float32 __z) { return __builtin_ctanf(__z); }
> +
> +  inline __complex__ _Float32
> +  __complex_tanh(__complex__ _Float32 __z) { return
> __builtin_ctanhf(__z); }
> +
> +  inline __complex__ _Float32
> +  __complex_pow(__complex__ _Float32 __x, __complex__ _Float32 __y)
> +  { return __builtin_cpowf(__x, __y); }
> +#endif
> +
> +#if defined(__STDCPP_FLOAT64_T__) &&
> defined(_GLIBCXX_DOUBLE_IS_IEEE_BINARY64)
> +  inline _Float64
> +  __complex_abs(__complex__ _Float64 __z) { return __builtin_cabs(__z); }
> +
> +  inline _Float64
> +  __complex_arg(__complex__ _Float64 __z) { return __builtin_carg(__z); }
> +
> +  inline __complex__ _Float64
> +  __complex_cos(__complex__ _Float64 __z) { return __builtin_ccos(__z); }
> +
> +  inline __complex__ _Float64
> +  __complex_cosh(__complex__ _Float64 __z) { return __builtin_ccosh(__z);
> }
> +
> +  inline __complex__ _Float64
> +  __complex_exp(__complex__ _Float64 __z) { return __builtin_cexp(__z); }
> +
> +  inline __complex__ _Float64
> +  __complex_log(__complex__ _Float64 __z) { return __builtin_clog(__z); }
> +
> +  inline __complex__ _Float64
> +  __complex_sin(__complex__ _Float64 __z) { return __builtin_csin(__z); }
> +
> +  inline __complex__ _Float64
> +  __complex_sinh(__complex__ _Float64 __z) { return __builtin_csinh(__z);
> }
> +
> +  inline __complex__ _Float64
> +  __complex_sqrt(__complex__ _Float64 __z) { return __builtin_csqrt(__z);
> }
> +
> +  inline __complex__ _Float64
> +  __complex_tan(__complex__ _Float64 __z) { return __builtin_ctan(__z); }
> +
> +  inline __complex__ _Float64
> +  __complex_tanh(__complex__ _Float64 __z) { return __builtin_ctanh(__z);
> }
> +
> +  inline __complex__ _Float64
> +  __complex_pow(__complex__ _Float64 __x, __complex__ _Float64 __y)
> +  { return __builtin_cpow(__x, __y); }
> +#endif
> +
> +#if defined(__STDCPP_FLOAT128_T__) &&
> defined(_GLIBCXX_LDOUBLE_IS_IEEE_BINARY128)
> +  inline _Float128
> +  __complex_abs(__complex__ _Float128 __z) { return __builtin_cabsl(__z);
> }
> +
> +  inline _Float128
> +  __complex_arg(__complex__ _Float128 __z) { return __builtin_cargl(__z);
> }
> +
> +  inline __complex__ _Float128
> +  __complex_cos(__complex__ _Float128 __z) { return __builtin_ccosl(__z);
> }
> +
> +  inline __complex__ _Float128
> +  __complex_cosh(__complex__ _Float128 __z) { return
> __builtin_ccoshl(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_exp(__complex__ _Float128 __z) { return __builtin_cexpl(__z);
> }
> +
> +  inline __complex__ _Float128
> +  __complex_log(__complex__ _Float128 __z) { return __builtin_clogl(__z);
> }
> +
> +  inline __complex__ _Float128
> +  __complex_sin(__complex__ _Float128 __z) { return __builtin_csinl(__z);
> }
> +
> +  inline __complex__ _Float128
> +  __complex_sinh(__complex__ _Float128 __z) { return
> __builtin_csinhl(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_sqrt(__complex__ _Float128 __z) { return
> __builtin_csqrtl(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_tan(__complex__ _Float128 __z) { return __builtin_ctanl(__z);
> }
> +
> +  inline __complex__ _Float128
> +  __complex_tanh(__complex__ _Float128 __z) { return
> __builtin_ctanhl(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_pow(__complex__ _Float128 __x, __complex__ _Float128 __y)
> +  { return __builtin_cpowl(__x, __y); }
> +#elif defined(__STDCPP_FLOAT128_T__) &&
> defined(_GLIBCXX_HAVE_FLOAT128_MATH)
> +  inline _Float128
> +  __complex_abs(__complex__ _Float128 __z) { return
> __builtin_cabsf128(__z); }
> +
> +  inline _Float128
> +  __complex_arg(__complex__ _Float128 __z) { return
> __builtin_cargf128(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_cos(__complex__ _Float128 __z) { return
> __builtin_ccosf128(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_cosh(__complex__ _Float128 __z) { return
> __builtin_ccoshf128(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_exp(__complex__ _Float128 __z) { return
> __builtin_cexpf128(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_log(__complex__ _Float128 __z) { return
> __builtin_clogf128(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_sin(__complex__ _Float128 __z) { return
> __builtin_csinf128(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_sinh(__complex__ _Float128 __z) { return
> __builtin_csinhf128(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_sqrt(__complex__ _Float128 __z) { return
> __builtin_csqrtf128(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_tan(__complex__ _Float128 __z) { return
> __builtin_ctanf128(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_tanh(__complex__ _Float128 __z) { return
> __builtin_ctanhf128(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_pow(__complex__ _Float128 __x, __complex__ _Float128 __y)
> +  { return __builtin_cpowf128(__x, __y); }
> +#endif
> +
> +#if defined(__STDCPP_BFLOAT16_T__) &&
> defined(_GLIBCXX_FLOAT_IS_IEEE_BINARY32)
> +  inline __gnu_cxx::__bfloat16_t
> +  __complex_abs(__complex__ decltype(0.0bf16) __z)
> +  { return __gnu_cxx::__bfloat16_t(__builtin_cabsf(__z)); }
> +
> +  inline __gnu_cxx::__bfloat16_t
> +  __complex_arg(__complex__ decltype(0.0bf16) __z)
> +  { return __gnu_cxx::__bfloat16_t(__builtin_cargf(__z)); }
> +
> +  inline __complex__ decltype(0.0bf16)
> +  __complex_cos(__complex__ decltype(0.0bf16) __z)
> +  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_ccosf(__z)); }
> +
> +  inline __complex__ decltype(0.0bf16)
> +  __complex_cosh(__complex__ decltype(0.0bf16) __z)
> +  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_ccoshf(__z)); }
> +
> +  inline __complex__ decltype(0.0bf16)
> +  __complex_exp(__complex__ decltype(0.0bf16) __z)
> +  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_cexpf(__z)); }
> +
> +  inline __complex__ decltype(0.0bf16)
> +  __complex_log(__complex__ decltype(0.0bf16) __z)
> +  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_clogf(__z)); }
> +
> +  inline __complex__ decltype(0.0bf16)
> +  __complex_sin(__complex__ decltype(0.0bf16) __z)
> +  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_csinf(__z)); }
> +
> +  inline __complex__ decltype(0.0bf16)
> +  __complex_sinh(__complex__ decltype(0.0bf16) __z)
> +  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_csinhf(__z)); }
> +
> +  inline __complex__ decltype(0.0bf16)
> +  __complex_sqrt(__complex__ decltype(0.0bf16) __z)
> +  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_csqrtf(__z)); }
> +
> +  inline __complex__ decltype(0.0bf16)
> +  __complex_tan(__complex__ decltype(0.0bf16) __z)
> +  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_ctanf(__z)); }
> +
> +  inline __complex__ decltype(0.0bf16)
> +  __complex_tanh(__complex__ decltype(0.0bf16) __z)
> +  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_ctanhf(__z)); }
> +
> +  inline __complex__ decltype(0.0bf16)
> +  __complex_pow(__complex__ decltype(0.0bf16) __x,
> +               __complex__ decltype(0.0bf16) __y)
> +  { return static_cast<__complex__ decltype(0.0bf16)>(__builtin_cpowf(__x,
> +
>  __y)); }
> +#endif
> +#endif
> +
>    // 26.2.7/3 abs(__z):  Returns the magnitude of __z.
>    template<typename _Tp>
>      inline _Tp
> @@ -1080,264 +1338,6 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
>                          : std::pow(complex<_Tp>(__x), __y);
>      }
>
> -#if _GLIBCXX_USE_C99_COMPLEX
> -#if defined(__STDCPP_FLOAT16_T__) &&
> defined(_GLIBCXX_FLOAT_IS_IEEE_BINARY32)
> -  inline _Float16
> -  __complex_abs(__complex__ _Float16 __z)
> -  { return _Float16(__builtin_cabsf(__z)); }
> -
> -  inline _Float16
> -  __complex_arg(__complex__ _Float16 __z)
> -  { return _Float16(__builtin_cargf(__z)); }
> -
> -  inline __complex__ _Float16
> -  __complex_cos(__complex__ _Float16 __z)
> -  { return static_cast<__complex__ _Float16>(__builtin_ccosf(__z)); }
> -
> -  inline __complex__ _Float16
> -  __complex_cosh(__complex__ _Float16 __z)
> -  { return static_cast<__complex__ _Float16>(__builtin_ccoshf(__z)); }
> -
> -  inline __complex__ _Float16
> -  __complex_exp(__complex__ _Float16 __z)
> -  { return static_cast<__complex__ _Float16>(__builtin_cexpf(__z)); }
> -
> -  inline __complex__ _Float16
> -  __complex_log(__complex__ _Float16 __z)
> -  { return static_cast<__complex__ _Float16>(__builtin_clogf(__z)); }
> -
> -  inline __complex__ _Float16
> -  __complex_sin(__complex__ _Float16 __z)
> -  { return static_cast<__complex__ _Float16>(__builtin_csinf(__z)); }
> -
> -  inline __complex__ _Float16
> -  __complex_sinh(__complex__ _Float16 __z)
> -  { return static_cast<__complex__ _Float16>(__builtin_csinhf(__z)); }
> -
> -  inline __complex__ _Float16
> -  __complex_sqrt(__complex__ _Float16 __z)
> -  { return static_cast<__complex__ _Float16>(__builtin_csqrtf(__z)); }
> -
> -  inline __complex__ _Float16
> -  __complex_tan(__complex__ _Float16 __z)
> -  { return static_cast<__complex__ _Float16>(__builtin_ctanf(__z)); }
> -
> -  inline __complex__ _Float16
> -  __complex_tanh(__complex__ _Float16 __z)
> -  { return static_cast<__complex__ _Float16>(__builtin_ctanhf(__z)); }
> -
> -  inline __complex__ _Float16
> -  __complex_pow(__complex__ _Float16 __x, __complex__ _Float16 __y)
> -  { return static_cast<__complex__ _Float16>(__builtin_cpowf(__x, __y)); }
> -#endif
> -
> -#if defined(__STDCPP_FLOAT32_T__) &&
> defined(_GLIBCXX_FLOAT_IS_IEEE_BINARY32)
> -  inline _Float32
> -  __complex_abs(__complex__ _Float32 __z) { return __builtin_cabsf(__z); }
> -
> -  inline _Float32
> -  __complex_arg(__complex__ _Float32 __z) { return __builtin_cargf(__z); }
> -
> -  inline __complex__ _Float32
> -  __complex_cos(__complex__ _Float32 __z) { return __builtin_ccosf(__z); }
> -
> -  inline __complex__ _Float32
> -  __complex_cosh(__complex__ _Float32 __z) { return
> __builtin_ccoshf(__z); }
> -
> -  inline __complex__ _Float32
> -  __complex_exp(__complex__ _Float32 __z) { return __builtin_cexpf(__z); }
> -
> -  inline __complex__ _Float32
> -  __complex_log(__complex__ _Float32 __z) { return __builtin_clogf(__z); }
> -
> -  inline __complex__ _Float32
> -  __complex_sin(__complex__ _Float32 __z) { return __builtin_csinf(__z); }
> -
> -  inline __complex__ _Float32
> -  __complex_sinh(__complex__ _Float32 __z) { return
> __builtin_csinhf(__z); }
> -
> -  inline __complex__ _Float32
> -  __complex_sqrt(__complex__ _Float32 __z) { return
> __builtin_csqrtf(__z); }
> -
> -  inline __complex__ _Float32
> -  __complex_tan(__complex__ _Float32 __z) { return __builtin_ctanf(__z); }
> -
> -  inline __complex__ _Float32
> -  __complex_tanh(__complex__ _Float32 __z) { return
> __builtin_ctanhf(__z); }
> -
> -  inline __complex__ _Float32
> -  __complex_pow(__complex__ _Float32 __x, __complex__ _Float32 __y)
> -  { return __builtin_cpowf(__x, __y); }
> -#endif
> -
> -#if defined(__STDCPP_FLOAT64_T__) &&
> defined(_GLIBCXX_DOUBLE_IS_IEEE_BINARY64)
> -  inline _Float64
> -  __complex_abs(__complex__ _Float64 __z) { return __builtin_cabs(__z); }
> -
> -  inline _Float64
> -  __complex_arg(__complex__ _Float64 __z) { return __builtin_carg(__z); }
> -
> -  inline __complex__ _Float64
> -  __complex_cos(__complex__ _Float64 __z) { return __builtin_ccos(__z); }
> -
> -  inline __complex__ _Float64
> -  __complex_cosh(__complex__ _Float64 __z) { return __builtin_ccosh(__z);
> }
> -
> -  inline __complex__ _Float64
> -  __complex_exp(__complex__ _Float64 __z) { return __builtin_cexp(__z); }
> -
> -  inline __complex__ _Float64
> -  __complex_log(__complex__ _Float64 __z) { return __builtin_clog(__z); }
> -
> -  inline __complex__ _Float64
> -  __complex_sin(__complex__ _Float64 __z) { return __builtin_csin(__z); }
> -
> -  inline __complex__ _Float64
> -  __complex_sinh(__complex__ _Float64 __z) { return __builtin_csinh(__z);
> }
> -
> -  inline __complex__ _Float64
> -  __complex_sqrt(__complex__ _Float64 __z) { return __builtin_csqrt(__z);
> }
> -
> -  inline __complex__ _Float64
> -  __complex_tan(__complex__ _Float64 __z) { return __builtin_ctan(__z); }
> -
> -  inline __complex__ _Float64
> -  __complex_tanh(__complex__ _Float64 __z) { return __builtin_ctanh(__z);
> }
> -
> -  inline __complex__ _Float64
> -  __complex_pow(__complex__ _Float64 __x, __complex__ _Float64 __y)
> -  { return __builtin_cpow(__x, __y); }
> -#endif
> -
> -#if defined(__STDCPP_FLOAT128_T__) &&
> defined(_GLIBCXX_LDOUBLE_IS_IEEE_BINARY128)
> -  inline _Float128
> -  __complex_abs(__complex__ _Float128 __z) { return __builtin_cabsl(__z);
> }
> -
> -  inline _Float128
> -  __complex_arg(__complex__ _Float128 __z) { return __builtin_cargl(__z);
> }
> -
> -  inline __complex__ _Float128
> -  __complex_cos(__complex__ _Float128 __z) { return __builtin_ccosl(__z);
> }
> -
> -  inline __complex__ _Float128
> -  __complex_cosh(__complex__ _Float128 __z) { return
> __builtin_ccoshl(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_exp(__complex__ _Float128 __z) { return __builtin_cexpl(__z);
> }
> -
> -  inline __complex__ _Float128
> -  __complex_log(__complex__ _Float128 __z) { return __builtin_clogl(__z);
> }
> -
> -  inline __complex__ _Float128
> -  __complex_sin(__complex__ _Float128 __z) { return __builtin_csinl(__z);
> }
> -
> -  inline __complex__ _Float128
> -  __complex_sinh(__complex__ _Float128 __z) { return
> __builtin_csinhl(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_sqrt(__complex__ _Float128 __z) { return
> __builtin_csqrtl(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_tan(__complex__ _Float128 __z) { return __builtin_ctanl(__z);
> }
> -
> -  inline __complex__ _Float128
> -  __complex_tanh(__complex__ _Float128 __z) { return
> __builtin_ctanhl(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_pow(__complex__ _Float128 __x, __complex__ _Float128 __y)
> -  { return __builtin_cpowl(__x, __y); }
> -#elif defined(__STDCPP_FLOAT128_T__) &&
> defined(_GLIBCXX_HAVE_FLOAT128_MATH)
> -  inline _Float128
> -  __complex_abs(__complex__ _Float128 __z) { return
> __builtin_cabsf128(__z); }
> -
> -  inline _Float128
> -  __complex_arg(__complex__ _Float128 __z) { return
> __builtin_cargf128(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_cos(__complex__ _Float128 __z) { return
> __builtin_ccosf128(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_cosh(__complex__ _Float128 __z) { return
> __builtin_ccoshf128(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_exp(__complex__ _Float128 __z) { return
> __builtin_cexpf128(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_log(__complex__ _Float128 __z) { return
> __builtin_clogf128(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_sin(__complex__ _Float128 __z) { return
> __builtin_csinf128(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_sinh(__complex__ _Float128 __z) { return
> __builtin_csinhf128(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_sqrt(__complex__ _Float128 __z) { return
> __builtin_csqrtf128(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_tan(__complex__ _Float128 __z) { return
> __builtin_ctanf128(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_tanh(__complex__ _Float128 __z) { return
> __builtin_ctanhf128(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_pow(__complex__ _Float128 __x, __complex__ _Float128 __y)
> -  { return __builtin_cpowf128(__x, __y); }
> -#endif
> -
> -#if defined(__STDCPP_BFLOAT16_T__) &&
> defined(_GLIBCXX_FLOAT_IS_IEEE_BINARY32)
> -  inline __gnu_cxx::__bfloat16_t
> -  __complex_abs(__complex__ decltype(0.0bf16) __z)
> -  { return __gnu_cxx::__bfloat16_t(__builtin_cabsf(__z)); }
> -
> -  inline __gnu_cxx::__bfloat16_t
> -  __complex_arg(__complex__ decltype(0.0bf16) __z)
> -  { return __gnu_cxx::__bfloat16_t(__builtin_cargf(__z)); }
> -
> -  inline __complex__ decltype(0.0bf16)
> -  __complex_cos(__complex__ decltype(0.0bf16) __z)
> -  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_ccosf(__z)); }
> -
> -  inline __complex__ decltype(0.0bf16)
> -  __complex_cosh(__complex__ decltype(0.0bf16) __z)
> -  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_ccoshf(__z)); }
> -
> -  inline __complex__ decltype(0.0bf16)
> -  __complex_exp(__complex__ decltype(0.0bf16) __z)
> -  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_cexpf(__z)); }
> -
> -  inline __complex__ decltype(0.0bf16)
> -  __complex_log(__complex__ decltype(0.0bf16) __z)
> -  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_clogf(__z)); }
> -
> -  inline __complex__ decltype(0.0bf16)
> -  __complex_sin(__complex__ decltype(0.0bf16) __z)
> -  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_csinf(__z)); }
> -
> -  inline __complex__ decltype(0.0bf16)
> -  __complex_sinh(__complex__ decltype(0.0bf16) __z)
> -  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_csinhf(__z)); }
> -
> -  inline __complex__ decltype(0.0bf16)
> -  __complex_sqrt(__complex__ decltype(0.0bf16) __z)
> -  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_csqrtf(__z)); }
> -
> -  inline __complex__ decltype(0.0bf16)
> -  __complex_tan(__complex__ decltype(0.0bf16) __z)
> -  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_ctanf(__z)); }
> -
> -  inline __complex__ decltype(0.0bf16)
> -  __complex_tanh(__complex__ decltype(0.0bf16) __z)
> -  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_ctanhf(__z)); }
> -
> -  inline __complex__ decltype(0.0bf16)
> -  __complex_pow(__complex__ decltype(0.0bf16) __x,
> -               __complex__ decltype(0.0bf16) __y)
> -  { return static_cast<__complex__ decltype(0.0bf16)>(__builtin_cpowf(__x,
> -
>  __y)); }
> -#endif
> -#endif
> -
>    /// 26.2.3  complex specializations
>    /// complex<float> specialization
>    template<>
> @@ -1835,7 +1835,7 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
>  #endif
>
>    template<typename _Tp>
> -    requires(__complex_type<_Tp>::type)
> +    requires requires { typename __complex_type<_Tp>::type; }
>      class complex<_Tp>
>      {
>      public:
> @@ -2022,6 +2022,162 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
>      }
>
>  #if _GLIBCXX_USE_C99_COMPLEX_TR1
> +#if defined(__STDCPP_FLOAT16_T__) &&
> defined(_GLIBCXX_FLOAT_IS_IEEE_BINARY32)
> +  inline __complex__ _Float16
> +  __complex_acos(__complex__ _Float16 __z)
> +  { return static_cast<__complex__ _Float16>(__builtin_cacosf(__z)); }
> +
> +  inline __complex__ _Float16
> +  __complex_asin(__complex__ _Float16 __z)
> +  { return static_cast<__complex__ _Float16>(__builtin_casinf(__z)); }
> +
> +  inline __complex__ _Float16
> +  __complex_atan(__complex__ _Float16 __z)
> +  { return static_cast<__complex__ _Float16>(__builtin_catanf(__z)); }
> +
> +  inline __complex__ _Float16
> +  __complex_acosh(__complex__ _Float16 __z)
> +  { return static_cast<__complex__ _Float16>(__builtin_cacoshf(__z)); }
> +
> +  inline __complex__ _Float16
> +  __complex_asinh(__complex__ _Float16 __z)
> +  { return static_cast<__complex__ _Float16>(__builtin_casinhf(__z)); }
> +
> +  inline __complex__ _Float16
> +  __complex_atanh(__complex__ _Float16 __z)
> +  { return static_cast<__complex__ _Float16>(__builtin_catanhf(__z)); }
> +#endif
> +
> +#if defined(__STDCPP_FLOAT32_T__) &&
> defined(_GLIBCXX_FLOAT_IS_IEEE_BINARY32)
> +  inline __complex__ _Float32
> +  __complex_acos(__complex__ _Float32 __z)
> +  { return __builtin_cacosf(__z); }
> +
> +  inline __complex__ _Float32
> +  __complex_asin(__complex__ _Float32 __z)
> +  { return __builtin_casinf(__z); }
> +
> +  inline __complex__ _Float32
> +  __complex_atan(__complex__ _Float32 __z)
> +  { return __builtin_catanf(__z); }
> +
> +  inline __complex__ _Float32
> +  __complex_acosh(__complex__ _Float32 __z)
> +  { return __builtin_cacoshf(__z); }
> +
> +  inline __complex__ _Float32
> +  __complex_asinh(__complex__ _Float32 __z)
> +  { return __builtin_casinhf(__z); }
> +
> +  inline __complex__ _Float32
> +  __complex_atanh(__complex__ _Float32 __z)
> +  { return __builtin_catanhf(__z); }
> +#endif
> +
> +#if defined(__STDCPP_FLOAT64_T__) &&
> defined(_GLIBCXX_DOUBLE_IS_IEEE_BINARY64)
> +  inline __complex__ _Float64
> +  __complex_acos(__complex__ _Float64 __z)
> +  { return __builtin_cacos(__z); }
> +
> +  inline __complex__ _Float64
> +  __complex_asin(__complex__ _Float64 __z)
> +  { return __builtin_casin(__z); }
> +
> +  inline __complex__ _Float64
> +  __complex_atan(__complex__ _Float64 __z)
> +  { return __builtin_catan(__z); }
> +
> +  inline __complex__ _Float64
> +  __complex_acosh(__complex__ _Float64 __z)
> +  { return __builtin_cacosh(__z); }
> +
> +  inline __complex__ _Float64
> +  __complex_asinh(__complex__ _Float64 __z)
> +  { return __builtin_casinh(__z); }
> +
> +  inline __complex__ _Float64
> +  __complex_atanh(__complex__ _Float64 __z)
> +  { return __builtin_catanh(__z); }
> +#endif
> +
> +#if defined(__STDCPP_FLOAT128_T__) &&
> defined(_GLIBCXX_LDOUBLE_IS_IEEE_BINARY128)
> +  inline __complex__ _Float128
> +  __complex_acos(__complex__ _Float128 __z)
> +  { return __builtin_cacosl(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_asin(__complex__ _Float128 __z)
> +  { return __builtin_casinl(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_atan(__complex__ _Float128 __z)
> +  { return __builtin_catanl(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_acosh(__complex__ _Float128 __z)
> +  { return __builtin_cacoshl(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_asinh(__complex__ _Float128 __z)
> +  { return __builtin_casinhl(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_atanh(__complex__ _Float128 __z)
> +  { return __builtin_catanhl(__z); }
> +#elif defined(__STDCPP_FLOAT128_T__) &&
> defined(_GLIBCXX_HAVE_FLOAT128_MATH)
> +  inline __complex__ _Float128
> +  __complex_acos(__complex__ _Float128 __z)
> +  { return __builtin_cacosf128(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_asin(__complex__ _Float128 __z)
> +  { return __builtin_casinf128(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_atan(__complex__ _Float128 __z)
> +  { return __builtin_catanf128(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_acosh(__complex__ _Float128 __z)
> +  { return __builtin_cacoshf128(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_asinh(__complex__ _Float128 __z)
> +  { return __builtin_casinhf128(__z); }
> +
> +  inline __complex__ _Float128
> +  __complex_atanh(__complex__ _Float128 __z)
> +  { return __builtin_catanhf128(__z); }
> +#endif
> +
> +#if defined(__STDCPP_BFLOAT16_T__) &&
> defined(_GLIBCXX_FLOAT_IS_IEEE_BINARY32)
> +  inline __complex__ decltype(0.0bf16)
> +  __complex_acos(__complex__ decltype(0.0bf16) __z)
> +  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_cacosf(__z)); }
> +
> +  inline __complex__ decltype(0.0bf16)
> +  __complex_asin(__complex__ decltype(0.0bf16) __z)
> +  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_casinf(__z)); }
> +
> +  inline __complex__ decltype(0.0bf16)
> +  __complex_atan(__complex__ decltype(0.0bf16) __z)
> +  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_catanf(__z)); }
> +
> +  inline __complex__ decltype(0.0bf16)
> +  __complex_acosh(__complex__ decltype(0.0bf16) __z)
> +  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_cacoshf(__z)); }
> +
> +  inline __complex__ decltype(0.0bf16)
> +  __complex_asinh(__complex__ decltype(0.0bf16) __z)
> +  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_casinhf(__z)); }
> +
> +  inline __complex__ decltype(0.0bf16)
> +  __complex_atanh(__complex__ decltype(0.0bf16) __z)
> +  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_catanhf(__z)); }
> +#endif
> +#endif
> +
> +#if _GLIBCXX_USE_C99_COMPLEX_TR1
>    inline __complex__ float
>    __complex_acos(__complex__ float __z)
>    { return __builtin_cacosf(__z); }
> @@ -2247,162 +2403,6 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
>      { return __complex_atanh(__z); }
>  #endif
>
> -#if _GLIBCXX_USE_C99_COMPLEX_TR1
> -#if defined(__STDCPP_FLOAT16_T__) &&
> defined(_GLIBCXX_FLOAT_IS_IEEE_BINARY32)
> -  inline __complex__ _Float16
> -  __complex_acos(__complex__ _Float16 __z)
> -  { return static_cast<__complex__ _Float16>(__builtin_cacosf(__z)); }
> -
> -  inline __complex__ _Float16
> -  __complex_asin(__complex__ _Float16 __z)
> -  { return static_cast<__complex__ _Float16>(__builtin_casinf(__z)); }
> -
> -  inline __complex__ _Float16
> -  __complex_atan(__complex__ _Float16 __z)
> -  { return static_cast<__complex__ _Float16>(__builtin_catanf(__z)); }
> -
> -  inline __complex__ _Float16
> -  __complex_acosh(__complex__ _Float16 __z)
> -  { return static_cast<__complex__ _Float16>(__builtin_cacoshf(__z)); }
> -
> -  inline __complex__ _Float16
> -  __complex_asinh(__complex__ _Float16 __z)
> -  { return static_cast<__complex__ _Float16>(__builtin_casinhf(__z)); }
> -
> -  inline __complex__ _Float16
> -  __complex_atanh(__complex__ _Float16 __z)
> -  { return static_cast<__complex__ _Float16>(__builtin_catanhf(__z)); }
> -#endif
> -
> -#if defined(__STDCPP_FLOAT32_T__) &&
> defined(_GLIBCXX_FLOAT_IS_IEEE_BINARY32)
> -  inline __complex__ _Float32
> -  __complex_acos(__complex__ _Float32 __z)
> -  { return __builtin_cacosf(__z); }
> -
> -  inline __complex__ _Float32
> -  __complex_asin(__complex__ _Float32 __z)
> -  { return __builtin_casinf(__z); }
> -
> -  inline __complex__ _Float32
> -  __complex_atan(__complex__ _Float32 __z)
> -  { return __builtin_catanf(__z); }
> -
> -  inline __complex__ _Float32
> -  __complex_acosh(__complex__ _Float32 __z)
> -  { return __builtin_cacoshf(__z); }
> -
> -  inline __complex__ _Float32
> -  __complex_asinh(__complex__ _Float32 __z)
> -  { return __builtin_casinhf(__z); }
> -
> -  inline __complex__ _Float32
> -  __complex_atanh(__complex__ _Float32 __z)
> -  { return __builtin_catanhf(__z); }
> -#endif
> -
> -#if defined(__STDCPP_FLOAT64_T__) &&
> defined(_GLIBCXX_DOUBLE_IS_IEEE_BINARY64)
> -  inline __complex__ _Float64
> -  __complex_acos(__complex__ _Float64 __z)
> -  { return __builtin_cacos(__z); }
> -
> -  inline __complex__ _Float64
> -  __complex_asin(__complex__ _Float64 __z)
> -  { return __builtin_casin(__z); }
> -
> -  inline __complex__ _Float64
> -  __complex_atan(__complex__ _Float64 __z)
> -  { return __builtin_catan(__z); }
> -
> -  inline __complex__ _Float64
> -  __complex_acosh(__complex__ _Float64 __z)
> -  { return __builtin_cacosh(__z); }
> -
> -  inline __complex__ _Float64
> -  __complex_asinh(__complex__ _Float64 __z)
> -  { return __builtin_casinh(__z); }
> -
> -  inline __complex__ _Float64
> -  __complex_atanh(__complex__ _Float64 __z)
> -  { return __builtin_catanh(__z); }
> -#endif
> -
> -#if defined(__STDCPP_FLOAT128_T__) &&
> defined(_GLIBCXX_LDOUBLE_IS_IEEE_BINARY128)
> -  inline __complex__ _Float128
> -  __complex_acos(__complex__ _Float128 __z)
> -  { return __builtin_cacosl(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_asin(__complex__ _Float128 __z)
> -  { return __builtin_casinl(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_atan(__complex__ _Float128 __z)
> -  { return __builtin_catanl(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_acosh(__complex__ _Float128 __z)
> -  { return __builtin_cacoshl(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_asinh(__complex__ _Float128 __z)
> -  { return __builtin_casinhl(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_atanh(__complex__ _Float128 __z)
> -  { return __builtin_catanhl(__z); }
> -#elif defined(__STDCPP_FLOAT128_T__) &&
> defined(_GLIBCXX_HAVE_FLOAT128_MATH)
> -  inline __complex__ _Float128
> -  __complex_acos(__complex__ _Float128 __z)
> -  { return __builtin_cacosf128(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_asin(__complex__ _Float128 __z)
> -  { return __builtin_casinf128(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_atan(__complex__ _Float128 __z)
> -  { return __builtin_catanf128(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_acosh(__complex__ _Float128 __z)
> -  { return __builtin_cacoshf128(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_asinh(__complex__ _Float128 __z)
> -  { return __builtin_casinhf128(__z); }
> -
> -  inline __complex__ _Float128
> -  __complex_atanh(__complex__ _Float128 __z)
> -  { return __builtin_catanhf128(__z); }
> -#endif
> -
> -#if defined(__STDCPP_BFLOAT16_T__) &&
> defined(_GLIBCXX_FLOAT_IS_IEEE_BINARY32)
> -  inline __complex__ decltype(0.0bf16)
> -  __complex_acos(__complex__ decltype(0.0bf16) __z)
> -  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_cacosf(__z)); }
> -
> -  inline __complex__ decltype(0.0bf16)
> -  __complex_asin(__complex__ decltype(0.0bf16) __z)
> -  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_casinf(__z)); }
> -
> -  inline __complex__ decltype(0.0bf16)
> -  __complex_atan(__complex__ decltype(0.0bf16) __z)
> -  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_catanf(__z)); }
> -
> -  inline __complex__ decltype(0.0bf16)
> -  __complex_acosh(__complex__ decltype(0.0bf16) __z)
> -  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_cacoshf(__z)); }
> -
> -  inline __complex__ decltype(0.0bf16)
> -  __complex_asinh(__complex__ decltype(0.0bf16) __z)
> -  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_casinhf(__z)); }
> -
> -  inline __complex__ decltype(0.0bf16)
> -  __complex_atanh(__complex__ decltype(0.0bf16) __z)
> -  { return static_cast<__complex__
> decltype(0.0bf16)>(__builtin_catanhf(__z)); }
> -#endif
> -#endif
> -
>    template<typename _Tp>
>      inline _Tp
>      /// fabs(__z) [8.1.8].
>
>         Jakub
>
>
  

On Wed, 16 Nov 2022 at 13:42, Jonathan Wakely <jwakely@redhat.com> wrote:

>
>
> On Mon, 14 Nov 2022 at 13:57, Jakub Jelinek <jakub@redhat.com> wrote:
>
>> Hi!
>>
>> As filed by Jonathan in the PR, I've screwed up the requires syntax
>> in the extended floating point specialization:
>> -    requires(__complex_type<_Tp>::type)
>> +    requires requires { typename __complex_type<_Tp>::type; }
>> and doing this change resulted in lots of errors because
>> __complex_whatever
>> overfloads from extended floating point types were declared after the
>> templates which used them.
>>
>> The following patch fixes that.
>>
>> Bootstrapped/regtested on x86_64-linux and i686-linux, additionally
>> I've tested that with _GLIBCXX_HAVE_FLOAT128_MATH not being defined
>> while __STDCPP_FLOAT128_T__ defined one can still use
>> std::complex<std::float128_t> for basic arithmetic etc., just one can't
>> expect std::sin etc. to work in that case (because we don't have any
>> implementation).
>>
>> Ok for trunk?
>>
>
> OK, thanks for the quick fix.
>

Oh, also no released version of Clang can handle the C++20 rules about
'typename' yet, so we get this error:

complex:1843:15: error: no template named 'type' in '__complex_ty
pe<_Tp>'; did you mean 'ctype'?
     typedef std::__complex_type<_Tp>::type _ComplexT;
             ^~~~~~~~~~~~~~~~~~~~~~~~~~
It needs a 'typename' after 'typedef'.

I can add that if you want though.

(It's fixed on Clang trunk, so will be in 16.0.0)
  

On Wed, Nov 16, 2022 at 01:45:19PM +0000, Jonathan Wakely wrote:
> > On Mon, 14 Nov 2022 at 13:57, Jakub Jelinek <jakub@redhat.com> wrote:
> >
> >> Hi!
> >>
> >> As filed by Jonathan in the PR, I've screwed up the requires syntax
> >> in the extended floating point specialization:
> >> -    requires(__complex_type<_Tp>::type)
> >> +    requires requires { typename __complex_type<_Tp>::type; }
> >> and doing this change resulted in lots of errors because
> >> __complex_whatever
> >> overfloads from extended floating point types were declared after the
> >> templates which used them.
> >>
> >> The following patch fixes that.
> >>
> >> Bootstrapped/regtested on x86_64-linux and i686-linux, additionally
> >> I've tested that with _GLIBCXX_HAVE_FLOAT128_MATH not being defined
> >> while __STDCPP_FLOAT128_T__ defined one can still use
> >> std::complex<std::float128_t> for basic arithmetic etc., just one can't
> >> expect std::sin etc. to work in that case (because we don't have any
> >> implementation).
> >>
> >> Ok for trunk?
> >>
> >
> > OK, thanks for the quick fix.
> >
> 
> Oh, also no released version of Clang can handle the C++20 rules about
> 'typename' yet, so we get this error:
> 
> complex:1843:15: error: no template named 'type' in '__complex_ty
> pe<_Tp>'; did you mean 'ctype'?
>      typedef std::__complex_type<_Tp>::type _ComplexT;
>              ^~~~~~~~~~~~~~~~~~~~~~~~~~
> It needs a 'typename' after 'typedef'.
> 
> I can add that if you want though.

Can you?  Sorry, I have already committed the patch before noticing your
mail.

	Jakub
  

On Wed, 16 Nov 2022 at 13:47, Jakub Jelinek <jakub@redhat.com> wrote:

> On Wed, Nov 16, 2022 at 01:45:19PM +0000, Jonathan Wakely wrote:
> > > On Mon, 14 Nov 2022 at 13:57, Jakub Jelinek <jakub@redhat.com> wrote:
> > >
> > >> Hi!
> > >>
> > >> As filed by Jonathan in the PR, I've screwed up the requires syntax
> > >> in the extended floating point specialization:
> > >> -    requires(__complex_type<_Tp>::type)
> > >> +    requires requires { typename __complex_type<_Tp>::type; }
> > >> and doing this change resulted in lots of errors because
> > >> __complex_whatever
> > >> overfloads from extended floating point types were declared after the
> > >> templates which used them.
> > >>
> > >> The following patch fixes that.
> > >>
> > >> Bootstrapped/regtested on x86_64-linux and i686-linux, additionally
> > >> I've tested that with _GLIBCXX_HAVE_FLOAT128_MATH not being defined
> > >> while __STDCPP_FLOAT128_T__ defined one can still use
> > >> std::complex<std::float128_t> for basic arithmetic etc., just one
> can't
> > >> expect std::sin etc. to work in that case (because we don't have any
> > >> implementation).
> > >>
> > >> Ok for trunk?
> > >>
> > >
> > > OK, thanks for the quick fix.
> > >
> >
> > Oh, also no released version of Clang can handle the C++20 rules about
> > 'typename' yet, so we get this error:
> >
> > complex:1843:15: error: no template named 'type' in '__complex_ty
> > pe<_Tp>'; did you mean 'ctype'?
> >      typedef std::__complex_type<_Tp>::type _ComplexT;
> >              ^~~~~~~~~~~~~~~~~~~~~~~~~~
> > It needs a 'typename' after 'typedef'.
> >
> > I can add that if you want though.
>
> Can you?  Sorry, I have already committed the patch before noticing your
> mail.
>

Yes, no problem, I'll do it.