On Fri, 16 Jun 2000, Richard Gooch wrote:
> Keith Owens writes:
> > On Fri, 16 Jun 2000 16:38:50 -0700,
> > David Ford <david@kalifornia.com> wrote:
> > >"H. Peter Anvin" wrote:
> > >> It's somewhat unfortunate that this is in /boot, since it isn't a file
> > >> requiring for booting. /boot is often a very small filesystem.
> > >Agreed. /boot is used in most of the programs that use the map file tho and not
> > >too many of the programs look in the /lib/modules dir.
> > >
> > >I don't think it'd be too difficult to convince the program maintainers to try
> > >something new tho.
> >
> > <aol>Me Too</aol>. The amount of crud that gets stuck in /boot by
> > distributions is horrendous. /boot should contain boot loader stuff
> > only, i.e. second stage loaders, kernels and disk maps for the kernel
> > blocks. System.map and .config should go in /lib/modules/`uname -r`,
> > even on systems that have no modules. No need for a separate directory
> > like /lib/kernel, it would be just one more place to clean up.
> >
> > Alas this is a distribution problem, all l-k can do is recommend a
> > standard.
>
> And hard-code a test in init/main:init() to test for /boot/System.map
> and call BUG() if found. This would probably clean up most bad
> practice since many/most distribution installers are sufficiently
> braindamaged to put /boot on /.
>
> If they wise up and start making /boot a separate partition, we can
> add a similar test to do_mount().
Don't forget to patch the makefiles so that "make install" follows the new
policy. The current i386 install puts vmlinuz and System.map (no suffix)
in the same directory, wherever that is (and without editing the top
Makefile it'll be the root).
FSSTND probably needs touching up too.
-- Mark H. Wood, Lead System Programmer mwood@IUPUI.Edu 2000-05-05 13:27:15 GMT -- still no icebergs in the White River- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Fri Jun 23 2000 - 21:00:14 EST