------- Comment #4 from rguenther at suse dot de 2010-04-01 12:04 -------
Subject: Re: [4.5 regression] undesired operation when working
with mno-cygwin
On Thu, 1 Apr 2010, davek at gcc dot gnu dot org wrote:
> ------- Comment #3 from davek at gcc dot gnu dot org 2010-04-01 01:11 -------
> I figure this is worth fixing in the ~22-hour window remaining before 2nd
> April. The option although deprecated should not for preference be released
> in
> a broken state, and since it worked in all previous versions (i.e. before I
> added the libstdc++ wrappers) this is a regression. The fix should be really
> really trivial:
>
> Index: gcc/config/i386/cygwin.h
> ===================================================================
> --- gcc/config/i386/cygwin.h (revision 157789)
> +++ gcc/config/i386/cygwin.h (working copy)
> @@ -108,7 +108,7 @@ along with GCC; see the file COPYING3. If not see
> #define CXX_WRAP_SPEC_OPT "muse-libstdc-wrappers"
> #endif
>
> -#define CXX_WRAP_SPEC "%{" CXX_WRAP_SPEC_OPT ":" CXX_WRAP_SPEC_LIST "}"
> +#define CXX_WRAP_SPEC "%{!mno-cygwin:%{" CXX_WRAP_SPEC_OPT ":"
> CXX_WRAP_SPEC_LIST "}}"
>
> #else /* !defined (USE_CYGWIN_LIBSTDCXX_WRAPPERS) */
>
>
> I'll set the bootstrap going now. That'll leave me some time to test it this
> coming lunchtime/afternoon/evening (UK time) and check it in long before
> midnight.
>
> Richard G., please speak up if you have any objection to this plan, if I
> haven't heard anything from you by a few hours before the midnight deadline
> I'll assume you're ok with it.
It looks reasonable - well, I don't know nothing about cygwin. But
it doesn't touch other targets, so if a cygwin maintainer is ok with
it I'm fine with it as well.
Richard.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=42609
