joorava at pcu.helsinki.fi
Wed Jun 11 15:03:24 PDT 2003
Vidar Hoel wrote:
> Jouko Orava wrote:
> > Noticed that both the original coreutils-5.0-hostname.patch
> > and the version I sent earlier ignored the automake Makefiles,
> > src/Makefile.am and man/Makefile.am .
> > Regenerating the configuration files would remove the patch.
> > Probably nobody ever does that, but here is the updated patch anyway.
> If the idea was to patch with your patch, and not the original
> "coreutils-5.0-hostname.patch" as explained in the book (CVS), then I
> got the following error:
Did you try with a clean tree with the original patch at all?
I'd wager you get the same error then.
> make: Entering directory `/root/coreutils-5.0/src'
> ./wheel-gen.pl 5 \
> > wheel.h-t
> Can't locate POSIX.pm in @INC (@INC contains:
> /stage1/lib/perl5/5.8.0/i686-linux /stage1/lib/perl5/5.8.0
> /stage1/lib/perl5/site_perl .) at
> ./wheel-gen.pl line 12.
> BEGIN failed--compilation aborted at ./wheel-gen.pl line 12.
> make: *** [wheel.h] Error 2
> make: Leaving directory `/root/coreutils-5.0/src'
> make: *** [all-recursive] Error 1
> make: Leaving directory `/root/coreutils-5.0'
> make: *** [all] Error 2
I could not duplicate this error, no matter how hard I try.
wheel-gen.pl is a perl script; I think something
went wrong at Perl build at the end of Chapter 5.
Do you still have the Perl source tree?
In perl-5.8.0/hints/linux.sh, the last line should still be:
static_ext="IO re Fcntl"
but it must not be the ONLY line.
(This would happen if you only have a single > in
the echo command; the complaint about POSIX
would certainly fit this.
It should be possible to re-do the Perl step,
but you must do the '"Locking in" Glibc' from Chapter 5
before recompiling Perl, and immediately afterwards,
the 'Adjusting toolchain' from Chapter 6.)
Otherwise, you should untar a clean coreutils
tree, patch the sources, and try again.
Hope this helps,
And dammit, there is nothing wrong with the patch! ;)
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe lfs-dev' in the subject header of the message
More information about the lfs-dev