Hi,

Periodically, I retest to see if my Zoom modem card has started working with 
NetBSD and today, while examining debugging output, I ran across Johnny Lam's 
message from November of 2000:

	http://mail-index.netbsd.org/tech-kern/2000/11/18/0007.html

This message asks about allocation of i/o spaces for com devices and in 
particular about why 0x2f8-0x2ff is allocated to two separate devices.

I did not see any responses to his question and I wonder if the issue may be 
relevant to my problem.  Here is a partial dmesg output:

com3 at acpi0 (PNP0501)
com3: io 0x3f8-0x3ff irq 4
com3: ns16550a, working fifo
...
com4 at acpi0 (IBM0071)
com4: io 0x2f8-0x2ff irq 3 drq 3
com4: ns16550a, working fifo

With pcmcia_debug and pcmciacis_debug on, inserting the modem card produces:

pcmcia1: ++enabled_count = 1
cis mem map cc2af000
pcmcia1: CIS tuple chain:
01 02 00 ff
CISTPL_DEVICE type=null speed=null
15 21 04 01 5a 6f 6f 6d 2d 56 45 4e 55 53 00 50
43 4d 43 49 41 20 56 39 32 20 44 61 74 61 46 61
78 00 ff
CISTPL_VERS_1
20 04 00 02 01 00
CISTPL_MANFID
21 02 02 00
CISTPL_FUNCID
22 04 00 02 0f 5c
22 09 05 1f 1f 00 04 00 00 04 00
22 09 06 1f 1f 00 04 00 00 04 00
22 0c 02 06 00 3f 1c 03 03 0f 07 00 01 b5
22 08 13 06 00 1f 00 02 00 b5
22 08 23 06 00 1f 00 02 00 b5
1a 05 01 27 80 ff 67
CISTPL_CONFIG
1b 13 cf 41 99 79 55 3d 86 46 26 4c aa 60 f8 03
07 f0 bc 86 28
CISTPL_CFTABLE_ENTRY
1b 07 17 08 aa 60 f8 02 07
CISTPL_CFTABLE_ENTRY
1b 07 1f 08 aa 60 e8 03 07
CISTPL_CFTABLE_ENTRY
1b 07 27 08 aa 60 e8 02 07
CISTPL_CFTABLE_ENTRY
ff
CISTPL_END
cis mem map cc2af000
CISTPL_LINKTARGET expected, code ff observed
pcmcia1: CIS version PCMCIA 2.0 or 2.1
pcmcia1: CIS info: Zoom-VENUS, PCMCIA V92 DataFax
pcmcia1: Manufacturer code 0x200, product 0x1
pcmcia1: function 0: serial port, ccr addr ff80 mask 67
pcmcia1: function 0, config table entry 15: I/O card; irq mask 86bc; iomask a, 
iospace 3f8-3ff; rdybsy_active io8 irqshare irqpulse irqlevel powerdown audio
pcmcia1: function 0, config table entry 23: I/O card; irq mask 86bc; iomask a, 
iospace 2f8-2ff; rdybsy_active io8 irqshare irqpulse irqlevel powerdown audio
pcmcia1: function 0, config table entry 31: I/O card; irq mask 86bc; iomask a, 
iospace 3e8-3ef; rdybsy_active io8 irqshare irqpulse irqlevel powerdown audio
pcmcia1: function 0, config table entry 39: I/O card; irq mask 86bc; iomask a, 
iospace 2e8-2ef; rdybsy_active io8 irqshare irqpulse irqlevel powerdown audio
com5 at pcmcia1 function 0: <Zoom-VENUS, PCMCIA V92 DataFax>
pcmcia_config_alloc: io 0 start=3f8 length=8 align=400 skew=3f8
window iobase 4000 iomax 4007
function iobase 4000 iomax 4007
pcmcia1: ++enabled_count = 2
pcmcia1: function 0 CCR at 0 offset f80: 4f 88 ee 0, 0 40 ff ff, ff
com5: ns16550a, working fifo
pcmcia1: --enabled_count = 1
pcmcia1: --enabled_count = 0

The line "pcmcia_config_alloc: io 0 start=3f8 length=8 align=400 skew=3f8" 
seems to indicate that the first config table entry (#15) is being used -- in 
apparent conflict with com3.

Thanks for any help or insight that you can provide me on this.

Sverre