Re: RFA: libtool.m4

[JonY]

On 6/19/2009 17:01, Kai Tietz wrote:
> 2009/5/15 Ralf Wildenhues<address@hidden>:
> > Hello Kai,
> >
> > * Kai Tietz wrote on Fri, May 15, 2009 at 09:46:49AM CEST:
> > > 2009/5/1 Charles Wilson:
> > > > Cross-compilation environments involving windows variants (mingw32,
> > > > 64bit mingw, cygwin, etc) and libtool are a little flaky at the moment.
> > > > I've had a number of patches to address (some of) this but they've been
> > > > unreviewed for almost six months.  If we can get the ball rolling on
> > > > that, they *should* be easy to extend to 64bit mingw by someone with
> > > > access to a w64 system.
> >
> > > as JonY said, we have x64 windows machines and we are willing to test
> > > your patches. Could you provide to us a link to your patches, so we
> > > could verify for you. We can back post then the result of the test to
> > > this ML or to you directly.
> >
> > > Ralf, is this a way for you?
> >
> > Surely feedback is helpful.
> >
> > I'm not sure whether it's clear to y'all here, but quite a big part of
> > the concerns over the code in the pr-msvc-support branch is not just
> > about regressions for other w32-related systems, but also for completely
> > w32 unrelated systems.  Admittedly I don't remember whether that was the
> > case for the patches Charles mentioned.
> >
> > Cheers,
> > Ralf
>
> Hi Ralf,
>
> it would be good to see those patches and possible a mark to those
> concerns. I would be real happy if we finally could make those patches
> for libtool, so that for gcc the w64 can build shared libstdc++
> version. There are bugs within none-shared libstdc++ in string class
> (reasoned by lazy initialization IIUC), which can be solved by the
> shared version.
>
> Cheers,
> Kai

Hi,
I've just tested with git master and it works correctly for both win32 
and win64, thanks a bunch guys.

By the way, any idea when will this fix make it into GCC trunk?