At 3:26 PM -0500 7/30/2000, David A. Gatwood wrote:
>Does anyone know if there's a way to disable the power-on memory check on
>Quadra/Centris machines?  I souped up a Centris 650 to 64 megs, and it
>takes... a while....  Just wondering.  :-)

In theory, yes. I think the MacOS manages to somehow do a quick reboot that
skips the memory check - atleast when warm-booted. I have no idea how it
does this, or how one would go about writing code to emulate the behaviour.

My C610's got 52 megs, so I sure know the pains of a slow boot. :)


>Also, if there's anyone on the list who is familiar with IDE on the Quadra
>630, from an interrupt standpoint, please drop me a line.  I'm working on
>the Performa 6214 and family for MkLinux, and the hardware is very
>similar.  As far as I can tell, if I trigger the interrupt handler
>manually (e.g. by hitting a key), things work (i.e. polling the F108
>register shows an IDE intr occurred), but the IDE hardware never actually
>starts the interrupt handler on its own, which makes me think that there's
>something about the interrupt levels on 68k that I just don't understand
>yet.  Anybody?

Mmm, triply cascated interrupts are always fun. IIRC the F108 should
cascade through a slot interrupt on VIA2 (slot E I think, but I could be
wrong), meaning that if you miss ANY slot interrupt, they'll ALL be gone
forever. I don't recall if the F108 has the same bogosity internally when
handling all the different interrupts it generates, but it might be worth
checking out.

I'd say check to make sure your slot ints aren't being hosed, (ie: make
sure you're never leaving the slot interrupt handler with any slot ints
still asserted) and you should be most of the way to getting all your
interrupts.


Cheers - Tony :)


--
Tony Mantler       Renaissance Nerd Extraordinaire     nicoya@apia.dhs.org
Winnipeg, Manitoba, Canada                     http://nicoya.feline.pp.se/