On Sun, 11 Jul 2004, Sven Luther wrote:
> On Sat, Jul 10, 2004 at 09:59:46PM +0200, Christoph Hellwig wrote:
> > On Sat, Jul 10, 2004 at 12:56:06PM -0700, Brad Boyer wrote:
> > > It seems bad to me for one of these scripts to be so sure that it can
> > > load a particular module, when the rest of the install is pretty
> > > careful to allow a user in expert mode to override module selection.
> > > Of course, loading a module shouldn't really crash the machine, but
> > > that's another story.
> >
> > The problem with floppy.o and a few other ISA-aera modules is that they
> > randomly poke into the I/O space - something that will crash lots of
> > machines, including macs. But AFAIK there's no power4 or ppc970-based
> > machine with a floppy driver, so shouldn't we just drop that driver?
>
> That is another possibility, but still, the fact that it doesn't break
> on other pmacs of the powerpc flavour is probably only luck, so fixing
> the kernel driver in the powerpc specific patch would be a better work
> around, until the legacy driver problem is solved in the 2.7+ kernel
> trees.
Indeed.
All these drivers should check with a platform-specific ISA presence checker
first. Things that come into my mind are an array of platform devices, or even
a simple array with I/O port number.
> Those would be the serial driver, which is already fixed, the parallel
> one, the floppy driver, the ps2 atkbd and mouse stuff, what else ?
(cat /proc/ioports on ia32 box)
RTC? -> Probably already sufficiently shielded by Kconfig
VGA? -> vgacon and vga16fb
IDE? -> Probably already fixed
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [EMAIL PROTECTED]
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
