Bug#207391: glibc: csu/Makefile patch
On Mon, Sep 15, 2003 at 11:02:54PM +0100, Colin Watson wrote:
> retitle 207391 glibc: fails to build if /bin/sh != bash
> thanks
>
> On Mon, Sep 15, 2003 at 01:58:18PM -0700, Eric Wong wrote:
> > Package: glibc
> > Version: 2.3.7-2; reported 2003-09-15
> > Followup-For: Bug #207391
> >
> > I've gotten this newline issue in version-info.h both while building from
> > non-modified sources for i386 and my own modified sources for i686.
> >
> > I'm not sure if it's a bug in make in the way it passes the argument, a bug in
> > echo in coreutils or a shell bug.
> >
> > The \\n\" sequence used in this section of the Makefile causes the newline
> > character to be interpreted, instead of the double backslash being interpreted
> > as a one. The below patch works around the problem. I'm using the latest
> > versions of all the build-dependencies from unstable.
> >
> > Below is a patch that works around the buggy \\n\" handling. Should be a
> > harmless fix even on systems that don't have the bug.
>
> This points straight at differences in the echo builtin between shells.
> POSIX notes: "It is not possible to use echo portably across all POSIX
> systems unless both -n (as the first argument) and escape sequences are
> omitted." I bet your /bin/sh isn't bash: for instance, dash's echo
> interprets its arguments in the way you describe. Zed, is the same true
> for you?
Yes, this is one of the big differences between dash and bash. Note
that autoconf falls afoul of this also, so I hope you don't need to
pass a backslash to any of 'configure's arguments if /bin/sh's echo
interprets backslashes.
> If you need any escape sequences in a portable shell script, then the
> only real option is to avoid echo altogether and use printf instead. I'd
> suggest the following modifications in place of yours:
That sounds like a good idea...
>
> > --- glibc-2.3.2.o/glibc-2.3.2/csu/Makefile 2003-09-15 13:35:41.000000000 -0700
> > +++ glibc-2.3.2/glibc-2.3.2/csu/Makefile 2003-09-15 13:34:21.000000000 -0700
> > @@ -231,13 +231,15 @@
> > if [ -z "$$os" ]; then \
> > os=Linux; \
> > fi; \
> > - echo "\"Compiled on a $$os $$version system" \
> > - "on `date +%Y-%m-%d`.\\n\"" ;; \
> > + echo -n "\"Compiled on a $$os $$version system" \
> > + "on `date +%Y-%m-%d`.\\" ; \
> > + echo "n\"";; \
>
> + printf '"Compiled on a %s %s system on %s.\\n"\n' \
> + "$os" "$version" "`date +%Y-%m-%d`" ;; \
>
> > *) ;; \
> > esac; \
> > files="$(all-Banner-files)"; \
> > if test -n "$$files"; then \
> > - echo "\"Available extensions:\\n\""; \
> > + echo -n "\"Available extensions:\\"; \
> > + echo "n\""; \
>
> + printf '"Available extensions:\\n"\n'; \
>
> (Sorry I haven't had time to construct a proper patch, but this should
> get the gist across.)
>
> Cheers,
>
> --
> Colin Watson [cjwatson@flatline.org.uk]
>
>
> --
> To UNSUBSCRIBE, email to debian-glibc-request@lists.debian.org
> with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
>
>
--
Daniel Jacobowitz
MontaVista Software Debian GNU/Linux Developer
Reply to: