> On Fri, Mar 03, 1995 at 06:24:14PM +0000, Philip Lardner wrote:
> > I then fired up the kernel, again following the instructions to the letter,
> > and tried to specify CD ROM as the install media. Unfortunatly It wouldn't
> > recognise the old Cumana drive and its patch board which I had in the
I've been working on this (but got somewhat stalled)
The drive is a Panasonic CR-562-J. Linux has a driver for this drive when
used with a soundblaster card (sbpcd.c), but the Cumana daughterboard sits
on the IDE cable, and uses the IDE bus to talk to the drive.
[backround - the drive is not ATAPI, and the Cumana card doesn't behave as
either an IDE drive or an ATAPI device. I know how the driver activates the
card]
I've reverse engineered the Cumana driver for RISC OS, and can see how it
works once it's got exclusive possession of the IDE bus. Neither the RISC OS
driver for the IDE card nor the Linux driver for the soundblaster card use
interrupts - as far as I can tell they both busy wait
What I don't know for sure is:
On Linux how does one claim use of the IDE bus to talk to a device on it?
Does it matter that it doesn't generate an interrupt when it's ready to supply
data?
(possibly other things)
My other problem is that I'd really rather prefer to acquire a small IDE drive
before I start attempting to write a driver, so that if I trash the harddisc
accidentally it doesn't actually matter - I just reinstall from my proper HD
Nick (digressed into gcc, perl and job hunting. Anyone got a linux job?)
unsubscribe: body of `unsubscribe linux-arm' to [EMAIL PROTECTED]
