This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: New build breakage on 3.0 branch
On Thu, Mar 01, 2001 at 03:59:27PM -0600, Robert Lipe wrote:
>
> #define _GLIBCPP_HAVE_MBSTATE_T 1
> /* #undef _GLIBCPP_HAVE_WCHAR_H */
Huh. I wonder what's defining mbstate_t... if <wchar.h> isn't there,
then the test for mbstate_t should fail, unless it's being magically
provided elsehow.
> Commenting out the MBSTATE_T thingy does allow compilation of the v3
> directory to complete. I have no idea if it's right, what I'm giving
> up, or if anything actually using libstdc++v3 works.
If the mbstate_t test fails, then we fake it, so the rest of everything
still happens correctly. I don't know how "fake" it is, really; it could
actually be complete. I keep falling behind on the multibyte chars issue.
Whether it works... well, "make check-install-target-libstdc++-v3" at the
top level after installation should give something.
Phil
--
pedwards at disaster dot jaj dot com | pme at sources dot redhat dot com
devphil at several other less interesting addresses in various dot domains
The gods do not protect fools. Fools are protected by more capable fools.