ICA breakage

Dan Nicholson dbn.lists at gmail.com
Sat Mar 24 14:34:50 PDT 2007

On 3/22/07, Jeremy Huntwork <jhuntwork at linuxfromscratch.org> wrote:
> Anyone else hit this on section two of ICA testing?
> # tail /mnt/build_dir/jhalfs/logs/097-iproute2-build_2 -n8
> install -m 0644 man/man8/ip.8 man/man8/lnstat.8 man/man8/rtmon.8
> man/man8/ss.8 man/man8/tc-bfifo.8 man/man8/tc-cbq-details.8
> man/man8/tc-cbq.8 man/man8/tc-htb.8 man/man8/tc-pfifo_fast.8
> man/man8/tc-prio.8 man/man8/tc-red.8 man/man8/tc-sfq.8 man/man8/tc-tbf.8
> man/man8/tc.8 /usr/share/man/man8
> install: cannot create regular file `/usr/share/man/man8/tc-bfifo.8':
> File exists
> make[1]: *** [install] Error 1
> make[1]: Leaving directory `/sources/iproute-2.6.20-070313'

Hey, I was looking at the new iproute today, and I don't think this is
an ICA problem. I think the Makefile is just screwed up. Or, they
forgot to distribute a file. The install target has this:

        install -m 0644 $(shell find man/man8 -maxdepth 1 -type f)
        ln -sf tc-pbfifo.8  $(DESTDIR)$(MANDIR)/man8/tc-bfifo.8
        ln -sf tc-pbfifo.8  $(DESTDIR)$(MANDIR)/man8/tc-pfifo.8

But, the man pages are distributed like this:

# ls -l man/man8/*fifo.8
-rw-r--r-- 1 root root 1976 Mar 13 21:50 man/man8/tc-bfifo.8
lrwxrwxrwx 1 root root   10 Mar 24 21:19 man/man8/tc-pfifo.8 -> tc-bfifo.8

So, I think it's just a mistake in the Makefile since the only reason
to create the symlinks would be if tc-bfifo.8 and tc-pfifo.8 weren't
distributed. It ends up overwriting the real tc-bfifo.8 manpage and
leaving two dangling symlinks. I'm just suppressing the symlinking in
the Makefile:

# sed -i '/tc-pbfifo.8/d' Makefile

The only problem is that tc-pfifo.8 ends up being a copy of tc-bfifo.8
instead of a symlink. That could be fixed with one more command, but
it's probably not a big deal.

# ln -svf tc-bfifo.8 /usr/share/man/man8/tc-pfifo.8


More information about the alfs-discuss mailing list