> Perhaps this would argue for, say, 10 possible IDE hardware interfaces
> (two supplied by the motherboard plus 4*2 from four additional PCI host
> adapters), for a total of 20 devices. My reason for playing with
> this is to evaluate IDE-based RAID; so far it's quite encouraging.
Are you asking for 10 possible IDE hardware interfaces
(reality beating my head into the wall) ???
The devfs overhead would be..........join the "devfs" debate, because the
implimentation of this would dominate the MAJOR device numbers.
With 20 Majors by 16 minors..........320 dev files..........
Have gone to take a cold shower............
The Promise FASTrax is a raid 0,1,stripping-0 to the best of my memory.
There is no support in 2.1.115 even with my future patch.........that I
know about..............
I have looked at it on the shelf, but at $129 each.........not out of
my pocket until (even though I can write it off as a business expense),
until there is positive movement in the IDE-DRIVER..........
I would drop the onboard PIIX controllers in order free wasted interrupts.
Four cards plus onboard IDE, requires a minimum of 6 interrupts.
While five cards would/should only require a minimum of 5 interrupts.
Under SMP, it is possible to shift this with edge/level interrupts.
But the thought is painful to dynamically adjust both PCI IO-Space and
interrupt number and type. This could allow for two cards to be assigned
(technically) the same interrupt, but determining which cards will
play nicely with an actual interrupt assignment greater than 15 is........
Hey, I WON THE LOTTERY................for that 256 Million.........argh...
Regards,
Andre Hedrick
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html