Re: [d-i] Testing daily image of d-i
On Mon, 4 Dec 2006, Michael Schmitz wrote:
> > The current d-i daily build (2.6.17-2) doesn't boot on my Amiga 3000 (spice)
> > with the WarpEngine accel card. The reason is a kernel panic in the NCR scsi
> > driver.
> >
> > Kars de Jonk ported the new version of the driver (or somewhat like this)
> > last year to the new ABI/API/whatever. His kernel 2.6.14 does work great on
> > that particular box and on my A4000T (which has the same chip onboard):
> >
> > Linux vivaldi 2.6.14-m68k-amiga #10 Thu Dec 1 21:40:17 CET 2005 m68k
> > GNU/Linux
> >
> > This means, that at least 3 m68k buildds aren't able to run newer kernels
> > with the stock Debian kernel. I think this is bad.
> >
> > Kars, Christian? Can you please work on intergrating the needed patches into
> > the Debian kernel source?
>
> The patch in question appears to be obsolete/520-53c7xx.diff in Geert's
That's the (old) one in Linux/m68k CVS.
> patch queue. Kars suggested to drop the 53c7xx driver in favor of the
> 53c700 one (and Richard Hirst submitted a patch to 53c700.c fixing some
> issues) but the switch was never made (at least not in CVS). I'm unsure
> what driver is actually used in the kernel you're running, can you check
> the log messages or look at /proc/scsi to help decide between 53c7xx and
> 53c700?
>
> The old 520-53c7xx.diff would have to be changed to reflect the new dma
> API. The 53c700 diff would have to be reconstructed from scratch, or
> resubmitted by Richard if the attached 53700 patch doesn't work anymore.
If Kars' 2.6.14 works fine, you may want to try:
m68k-generic-io.diff
m68k-mvme-scsi-rename.diff
m68k-53c700-scsi.diff
These are for 2.6.19, but with minor changes (re-adding the pt_regs * argument
to interrupt routines), it should apply to 2.6.18.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
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
Reply to: