IDE Hacking

[Sam Mason]

Sam Mason wrote:
>I'll have a better play later on.  At the moment I'm hacking up Johan
>Rydberg's IDE code and attempting to access disks.

Just to prove I don't know what I'm doing I'll post some code.
It's heavily based on the IDE code found in Johan Rydberg's Nova
operating system and you can probably see the heritage.  Currently,
it just probes for two standard IDE controllers, then asks them to
describe themselves, prints out a few identifying numbers and then
reads in the first sector on the disk and prints out the first four
bytes of that.  It crashes for some reason, proving I need to get
a better understanding of how things work in Hurd-l4!

I'm not sure if things like the byte-swapping is correct, but it's
my best (uneducated) guess.  I'm not sure if I'm pulling in the
correct drive geometry information either - it'll work for small
drives, but once you get into half gig drives it's not going to work.

I guess I need to figure out how to handle IRQs and request resources
properly - rather than just assuming I have access to them.  Would
designing/implementing the RPCs that access this stuff come next?

Have fun,
  Sam

p.s. not sure what sort of legaleese is appropiate.  The header of
"deva-ide.c" is copied from "deva-class.c" and changed a bit.

deva-ide-0.0.1.patch
Description: Text document