https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69881
--- Comment #15 from Bernd Edlinger <bernd.edlinger at hotmail dot de> ---
(In reply to Jonathan Wakely from comment #12)
> > Index: c_std/cstddef
> > ===================================================================
> > --- c_std/cstddef (revision 233581)
> > +++ c_std/cstddef (working copy)
> > @@ -41,6 +41,11 @@
> >
> > #pragma GCC system_header
> >
> > +#undef __need_wchar_t
> > +#undef __need_size_t
> > +#undef __need_ptrdiff_t
> > +#undef __need_NULL
> > +#undef __need_wint_t
> > #include <bits/c++config.h>
> > #include <stddef.h>
>
> How do you plan to test this?
I tried:
../gcc-trunk/configure --prefix=/home/ed/gnu/install --enable-languages=c,c++
--enable-cheaders=c_std --disable-multilib
But it does not look like the c_std headers are currently working.
I've got symlinks to non-existing c_std header files:
ccomplex
cfenv
cinttypes
cstdalign
cstdbool
cstdint
ctgmath
and c_std/cmath is missing the _GLIBCXX_USE_C99_FENV_TR1 stuff.
which I need, because my math.h supports that.
Once I fix these bugs the boot-strap seems to work.
Let's see if the libstdc++ testsuite has complaints.
Probably it will, there are many more differences
between c_global/cmath and c_std/cmath like:
-#ifndef __CORRECT_ISO_CPP_MATH_H_PROTO
- inline _GLIBCXX_CONSTEXPR double
+ inline double
abs(double __x)
{ return __builtin_fabs(__x); }
-#endif
but these dont make the boot-strap fail.
