new pcmcia

["Ian F. Darwin" <ian_(_at_)_DarwinSys_(_dot_)_COM>]

I am one of those poor souls who has a notebook with a PCIC at a
weird address. The config example LAP shows using it at 0x3000, but
mine is at 0xFCFC.  My config under the "old" (2.3) PCMCIA code
was something like this (this is the dmesg output):

pcicmaster0 at isa0 port 0xfcfc-0xfcfd
pcic0 at pcicmaster0: Intel 82365sl Rev. 0 slots 0-1 iomem d4000-d4fff irq 11
pcmcia0 at pcic0
pcmcia0 slot 0: empty
pcmcia0 slot 1: <3Com Corporation, 3C589D, TP/BNC LAN Card Ver. 2a, 000002>
ep0 at pcmcia0 port 0x300-0x30f irq 10: address 00:60:97:88:f1:35, utp/aui/bnc (default utp)
pcicmaster0: master controller not configured

We also had the 
"Cirrus Logic CL-PD6729" rev 0xfe at pci0 dev 4 function 0 not configured
earlier on, during the PCI walk.

This broke when I ran config with "current", so I changed my 
config file to read:
...
pcic0	at isa? port 0xFCFC size 4 iomem 0xd4000 iosiz 0x4000
...
pcmcia*	at pcic? controller ? socket ?

(the size 4 is just there because that's what it had when I first
discovered the port address, using Windoze/95's device manager;
it's always been ignored, but seems harmless).

Now when it boots it says
pcic0 at isa0 port 0xfcfc-0xfcfd iomem 0xd4000-0xd7fff: using irq 3
pcic0: controller 0 (Cirrus PD672X) has sockets A and B
pcic0: controller 1 (Intel 82365SL Revision 1) has sockets A and B
extent_alloc_region: extent `ioport' (0x0 - 0xffff)
extent_alloc_region: start 0x100fc, end 0x100fd
panic: etent_alloc_region: region lies outside extent

and falls into the debugger.

Have I done something wrong?

Is it "rounding" that FCFC up to 100FC (and similarly for FCFD),
or are those numbers in the "start" and "end" printfs mere
coincidence?

Thanks to anyone who can help

Ian Darwin
ian_(_at_)_darwinsys_(_dot_)_com