[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: CVS



   Good news...bad news. The good news is that I can take the current
debian glibc-2.3 cvs, drop in the glibc-2.2.92.tar.bz2 and
glibc-linuxthreads-2.2.92.tar.bz2, build it on debian ppc sid
against the new gcc-3.2-3.2.1-0pre1 and it passes make check 
just as well as Franz Sirls builds do on entropy.crashing.org.
We have a set of math test failures we both fail but I believe
Ulrich knows about these and since I see conflicting type warnings
in the math section of the build I'm not surprised.
   Now for the bad news...the freaking broken regex in perl 5.80
is still messing up all the glibc builds...

(echo "INFO-DIR-SECTION GNU C library functions";		\
 echo "START-INFO-DIR-ENTRY";					\
 gawk -f xtract-typefun.awk intro.texi creature.texi errno.texi memory.texi ctype.texi string.texi stpcpy.c.texi strdupa.c.texi strncat.c.texi charset.texi locale.texi message.texi search.texi search.c.texi pattern.texi io.texi stdio.texi rprintf.c.texi m
 echo "END-INFO-DIR-ENTRY") > dir-add.info.new
mv -f dir-add.info.new dir-add.info
* Libc: (libc). C library.
install-info(/home/howarth/debian-glibccvs/glibc-2.3/powerpc-linux/install_root/usr/share/info/libc.info): creating new section `GNU libraries'
install-info(/home/howarth/debian-glibccvs/glibc-2.3/powerpc-linux/install_root/usr/share/info/libc.info): no sections yet, creating Miscellaneous section too.
make[3]: *** [install] Error 139
make[3]: Leaving directory `/home/howarth/debian-glibccvs/glibc-2.3/glibc-2.2.92/manual'
make[2]: *** [manual/subdir_install] Error 2
make[2]: Leaving directory `/home/howarth/debian-glibccvs/glibc-2.3/glibc-2.2.92

Is there anyway we can temporarily hack around this broken perl in the rules
files for glibc? I have no faith in upstream fixing this anytime soon.
I haven't even be able to find any traffic about the bug in the perl5-porters
mailing list on this topic. As it seems the sid maintainers don't mind
perl being this broken (I am told unstable is supposed to be unstable),
it doesn't look like we will have a regression to perl 5.6. Could someone
explain what the plan is then?
                                     Jack



Reply to: