>From: Michael Graff <explorer@flame.org>
>Date: 08 Apr 1997 18:55:37 -0400
>
>> >16 fixed io ports: 10 address bits, alignment 1, min 0x802, max 0x802
>> >IRQ's supported: 3 L+
>> >
>> >16 fixed io ports: 10 address bits, alignment 1, min 0x803, max 0x803
>> >IRQ's supported: 4 5 7 L+
>
>My guess is that those entries are actually bogus.  Either that or I
>don't know what those ranges mean...

They're probably not bogus, although they may have been interpreted
improperly.  I will take my PnP spec home w/me tonight and take a look
at the PnP resource parsing code.

The first is probably supposed to be either DOS' COM2: or COM4:, and
the second is probably supposed to be COM1: or COM3:.  The final
option (which you prefer) is the device's way of throwing up its hands
and telling the PnP configuration code to put it anywhere it will fit.

>Are those possible configurations (select one of a set) or are they
>ranges where all the selections must be used?

The three lines are the three possible configurations.

>Also, do you know how to tell the isapnp portion to ignore them perhaps?
>The latter one (the 0x100 to 0x3f8 range) is correct for the modem, and
>is what windoze uses.

PnP calls the alternatives 'dependent functions'; we may wish to
consider adding another locator to isapnp (e.g. 'dep') which would
allow you to nail down which of the three alternatives is chosen.

Another possibility is to create a null ISA device (similiar to
the SCSI code's uk device), which does nothing but map I/O ports so
that isapnp can't use them.  Bogus, but it would get the job done.
-- 
Mike Long <mike.long@analog.com>     <URL:http://www.shore.net/~mikel>
VLSI Design Engineer         finger mikel@shore.net for PGP public key
Analog Devices, CPD Division          CCBF225E7D3F7ECB2C8F7ABB15D9BE7B
Norwood, MA 02062 USA       (eq (opinion 'ADI) (opinion 'mike)) -> nil