interrupts [or the lack thereof]

[GoogleCodeExporter]

you half-assed it ,Mario is what it was.  you just didn't add-in a
'handler' for the keyboard, so anything that trips an interrupt on IRQ1,
just never gets handled. same with the mouse. there should be 256
interrupts available,according to what I'm reading.

everything just gets thrown into the general handler and causes the kernel
to panic.

9 is mapped to keyboard
33 mapped to mouse
21 needs to either be avoided, or  point somewhere else in the interrupt
table. this is because 'DOS'[the MS-DOS we know] isn't hooked in, and
probably never will be.

working on a fix, at least for the keyboard and mouse.

--Jazz

Original issue reported on code.google.com by jasm...@lavabit.com on 27 Feb 2009 at 1:53

[GoogleCodeExporter]

LOL - What is the progress on this - did it make it in the latest full release?

What Endian are those interrupts running at? Will they be portable?

Original comment by Jegas...@gmail.com on 8 Apr 2009 at 6:38

[GoogleCodeExporter]

How 'portable' do you need. Will it run on PPC arch, not yet. That would be 
another fork.

So far its like linux from scratch in pascal for x86.I am assuming little, I 
forget
which is which.

Latest SVN is at assembla[.com], you need a login,its free to view the boards 
there.I
put it there as Mario hadn't setup svn yet.Checkout should be available, but I
haven't posted my 'dramatic' changes yet.

Keyboard--9 and 16 , i think are handled so they don't fault. The ISR needed 
some help.

Mouse, I have no low level generic driver for yet, so I have to emulate with 
arrow
keys currently in use.[still at IRQ1 I believe]

21 and 80 are free and not needed.fileopts, more apropriately, 'DOS' unit has 
these
functions so to speak.

I need a few unit changes in the RTL before it will compile right now.Its a bit 
of work.

ACPI support should compile from direct C translation now, we have the classes 
we need.

Original comment by jasm...@lavabit.com on 9 Apr 2009 at 4:28