[designing a keyboard controller to work with Linux]
>Whatever hardware you need, the best is really to use something that can
>be atached to the so called pcmcia bus on the SA1100. You might need some
>logic to decode the address and activate a chip-enable signal if you have
>more than one port to address. A GPIO, preferably between 0 and 10, can
>be used as interrupt line.
>
>With such a setup around the SA1100, current ISA drivers can be used with
>very minimal modifications as long as they aren't using DMA. The kernel
>code for the SA1100 already provides access to that bus through the
>in*()/out*() macros. You then have plenty of choice for an interface
>chip.
That looks like a good solution. The question remains then: what does an
8042 'look like', as seen from Linux ? I want to mimic the ISA side of a
regular keyboard controller (without actually cramming a 40pin DIL chip on
my 7.5x10cm PCB :-P). Does anyone have pointers to documents on this ?
A semi-related question: is it possible for multiple different devices on
an ISA-ish bus to share an interrupt *on the processor side* ? For example,
say I have a custom PIC which multiplexes (logic OR) the interrupt output
of a keyboard controller with the interrupt output of an IDE controller
onto one interrupt line for the processor. Does this work, or do I break
any hidden assumptions in either the keyboard or the IDE driver ?
Sincerely,
JDB.
--
Jan-Derk Bakker, [EMAIL PROTECTED]
The lazy man's proverb:
'There's no business like slow business !'
unsubscribe: body of `unsubscribe linux-arm' to [EMAIL PROTECTED]
