On Wed, Oct 26, 2005 at 01:45:53PM +0200, Kars de Jong wrote:
> > I tried several prepackaged kernels from Debian, but all panic. I've put
> > some dmesg outputs on the net:
> Good, someone with a system other than MVME16x using an NCR53C710 :-)
Yes, it's the SCSI chip on A4091 SCSI card and A4000T onboard SCSI, but I
guess you already knew that. Perhaps there're not that much A4000T out there
running Debian or newer kernels. I assumed that an A4000T with an MKII is a
quite common hardware... ;)
> >From your dmesg:
> scsi-ncr53c7xx : NCR53c710 at memory 0xd0190040, io 0x0, irq 12
> This is the old ncr53c7xx driver, I believe it is broken on prepackaged
> Debian kernels because of a bogus virt_to_bus() call and/or the kernel
> being compiled with CONFIG_SINGLE_MEMORY_CHUNK=N. On early 2.4 kernels
> it did still work for me, but hasn't worked since 2.4.x (I don't
> remember x).
Well, it wouldn't make sense to download and test every kernel down to
2.4.0, I guess.... I would prefer to use a 2.6.x though... ;)
> I will try to rewrite the amiga7xx driver this evening to use the new
> 53c700 driver core. Should be pretty simple since it already works on
> MVE16x.
> It would be nice if you can test it, if it works I think we can safely
> get rid of the old ncr53c7xx driver.
No problem with testing. The only problem is that I'm just here (at parents
home) until Monday/Tuesday and intended to keep the machine there. So, when
we can solve that problem until then, that would be great! :)
BTW: in what state is amililo? That's something I could test as well in the
meantime... ;)
--
Ciao... // Fon: 0381-2744150
Ingo \X/ SIP: 2744150@sipgate.de
gpg pubkey: http://www.juergensmann.de/ij/public_key.asc
Attachment:
signature.asc
Description: Digital signature