Re: crosstool-0.28-pre13: random walk towards release

[Stuart Hughes <seh at zee2 dot com>]

Hi Dan,

Regarding the --enable-clocale=gnu, did you figure out why LFS
recommended it (the referenced mail archive has gone), what was it that
made you think it was bad to do.

Also regarding locales, would it be true to say that this is doomed to
fail unless you are self-hosted (or building a cross on the same
architecture as the host) ?

Regards, Stuart

Daniel Kegel wrote:
> 
> 0.28-pre13
>       misc:
>       * for gcc-2.95.3-glibc-2.1.3.dat, use the same binutils Red Hat 6.2 used
>       * remove bogus patch gcc-2.95.3/c-parse-freeze.patch
>       * revert glibc configuration change "--host=$HOST" in pre12
>         Instead, tell glibc --build=$HOST, which here is the right way
>         to keep glibc from thinking it's a native configuration when we do a self-cross
>       * remove --enable-clocale=gnu from gcc configuration; this fails with glibc-2.1.3,
>         and doesn't seem like a good idea now that I've looked at libstdc++/aclocal.m4.
> 
> I tried installing the glibc locale data, but it didn't work for me,
> so I just added it as a comment in crosstool.sh:
> 
> # This doesn't seem to work when building a crosscompiler,
> # as it tries to execute localedef using the just-build ld.so!?
> #make localedata/install-locales install_root=${SYSROOT}

------
Want more information?  See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/
Want to unsubscribe? Send a note to crossgcc-unsubscribe@sources.redhat.com