Use this page to discuss issues with modems and other serial devices. Before posting, be sure you've checked any relevant sections of the PCMCIA-HOWTO, such as: Particularly useful bits of information to post include your laptop's make and model; the type(s) of card(s) you are using; and, of course, any relevant system log messages.

Messages Inline: 0 1

Question wrong UART detected

Date: 1999, Nov 04
From: Jaime V Miro jaime

Hi there,

Is there any way to force the driver to use a particular UART? I have a "Socket" pcmcia serial card (not multi), which is properly identified with "cardctl ident", but the UART is wrongly detected as 8250 instead of 16550. This is happening in kernel 2.0.36 and also in 2.2.9 (Red Hat systems) and with the same results in the following pcmcia packages: 3.0.12, 3.0.13, 3.0.14, and 3.1.3

I am currently downloading kernel 2.2.13 to see if the problem is sorted, but somehow I don't expect it to happen.

I am confident the serial card UART uses 16550 because I had it running under 2.0.36 and pcmcia-cs-3.0.12 (although not always, sometimes it would also detect a 8250, and after a cold reboot it would then recognise 16550). However, I updated to 2.2.9 and pcmcia-cs-3.1.3 to get my pcmcia network card to work (3COM 3c575), as recomended by D. Hinds. Network card is now working, but I have not been able to recognise the 16550 UART with all possible combinations of the above kernels and pcmcia packages.

Any recommendations?

Thanks in advance,

Jaime

None Is this the real problem?

Re: Question wrong UART detected (Jaime V Miro)
Date: 1999, Nov 04
From: David Hinds <dhinds@pcmcia.sourceforge.org>

On Thu, Nov 04, 1999 at 04:56:27AM -0800, Jaime V Miro wrote:

> Is there any way to force the driver to use a particular UART?

You are certainly free to use the "setserial" command to set the UART
type (or to retry the UART autodetection).  However, I wonder if this
is fixing a symptom rather than the real problem.

An obvious question would be, regardless of whether it says it is a
16550 or an 8250 UART, does the port work?  (if the only problem is
the UART detection, then the card should still work, with somewhat
reduced performance)

You say that kernel 2.0.36 and PCMCIA 3.0.12 is the only combination
that ever detects a 16550 UART (but only sometimes)?  That's odd... I
can't think of a good reason for how that could be.  In my experience,
the most common cause of UART misdetection is an IO port conflict.
Which would not be kernel (or PCMCIA driver) dependent.

-- Dave

Question Multitech MT5634ZLX modem problems

Date: 1999, Nov 05
From: Ray Muno munorc

We are trying to use a Multitech MT5634ZLX PCMCIA modem on a Dell Inspiron 3500. The card is recognized as a serial device and installed but we cannot talk to the modem at all.

Is anyone using this device?

BTW, the modem operates fine under Windows 98. The device is not a Winmodem.

We are running Debian 2.1, Card Services 3.0.14, 2.2.7 kernel.

I have no problems running other PCMCIA cards, Ethernet and even CardBus Fast Ethernet.

None What have you tried, exactly?

Re: Question Multitech MT5634ZLX modem problems (Ray Muno)
Date: 1999, Nov 05
From: David Hinds <dhinds@pcmcia.sourceforge.org>

Have you worked through the troubleshooting information for serial
cards in the PCMCIA-HOWTO?  My first guess would be something along
the lines of an IO port conflict.

-- Dave

Question O2Micro OZ6836/OZ6860 PCI-to-CardBus

Date: 1999, Nov 18
From: Michael Larsen larsenm

Problem:
------------
Hard boot of an old kernel required to 
get PCMCIA Card Services working with a new kernel.

Description:
-------------
I have EPS Tech. MP-979 Laptop with an O2Micro OZ6836/OZ6860
CardBus controller and a 56K modem card (Lucent Tech. Venus
I think?). The BIOS setting for pnp is yes.

I recently upgraded my 2.0.35 kernel to 2.2.13-7mdk. 
The 56K modem card was working with PCMCIA Card Services 
version 3.0.4 compiled into the kernel. To get 
the card detected in the 2.0.35  kernel, I had to 
include memory 0xd8000-0xdffff in /etc/pcmcia/config.opts.

For the new kernel, as suggested by the PCMCIA-HOWTO, 
I added include memory 0xa0000000-0xa0ffffff to /etc/pcmcia/config.opts.
My problem is that the new kernel fails to load the PCMCIA
modules (based on Card Services 3.1) and generates an oops:

Linux PCMCIA Card Services 3.1.0
  kernel build: 2.2.13-7mdk #1 Wed Sep 15 18:02:18 CEST 1999
  options:  [pci] [cardbus] [apm] [pnp]
PnP: PNP BIOS installation structure at 0x800f6270
PnP: PNP BIOS version 1.0, entry at f0000:a4e3, dseg at 400
PCI routing table version 1.0 at 0xfdf80
  00:08.2 -> irq 0
  00:03.0 -> irq 0
  00:03.1 -> irq 0
Intel PCIC probe: <6>base 68000000: ff ff ff ff
base f8000000: ff ff ff ff

  Bridge register mapping failed: check cb_mem_base setting
Unable to handle kernel paging request at virtual address 84834800
current->tss.cr3 = 026d5000, %cr3 = 026d5000
*pde = 03c80063
*pte = 00000000
Oops: 0000
CPU:    0
EIP:    0010:[<8482d847>]

I tracked this done to the idnetify routine of the i82365.o 
module.

If however, I hard boot the old 2.0.35 kernel first, and then 
warm reboot the new kernel, the modules load and the
card is recognized. (Further warm reboots of the new
kernel result in modules loading correctly and the card 
being  recognized:

Linux PCMCIA Card Services 3.1.0
  kernel build: 2.2.13-7mdk #1 Wed Sep 15 18:02:18 CEST 1999
  options:  [pci] [cardbus] [apm] [pnp]
PnP: PNP BIOS installation structure at 0x800f6270
PnP: PNP BIOS version 1.0, entry at f0000:a4e3, dseg at 400
PCI routing table version 1.0 at 0xfdf80
  00:08.2 -> irq 0
  00:03.0 -> irq 0
  00:03.1 -> irq 0
Intel PCIC probe: 
  O2Micro OZ6836/OZ6860 PCI-to-CardBus at bus 0 slot 3, mem 0x68000000, 2 sockets
    host opts [0]: [a 44] [b 81] [c 00] [d 83] [mhpg 18] [fifo 08] [e 0a] [no pci irq] [lat 168/176] [bus 32/34]
    host opts [1]: [a 44] [b 8d] [c 50] [d 83] [mhpg 18] [fifo 08] [e 0a] [no pci irq] [lat 168/176] [bus 35/37]
    ISA irqs (default) = 3,4,5,7,9,10,11,12 polling interval = 1000 ms
cs: IO port probe 0x0100-0x03ff: excluding 0x200-0x207
cs: memory probe 0xa0000000-0xa0ffffff: excluding 0xa0000000-0xa00fffff
tty02 at 0x03e8 (irq = 3) is a 16550A

Any suggestions on how to get things working without
a hard reboot of the old kernel?  

/etc/sysconfig/pcmcia:
PCMCIA=yes
PCIC=i82365
CORE_OPTS= 
CARDMGR_OPTS=

/etc/pcmcia/config.opts
include port 0x100-0x3ff 
include memory 0xd8000-0xdffff 
include memory 0xa0000000-0xa0ffffff
include memory 0x44000000-0x44001fff
include memory 0x48000000-0x48ffffff
exclude irq 4




Some things I tried:
--------------------
I get similar pattern with kernel 2.3.24 and PCMCIA 
Card Services version 3.1.4. Instead of an oops it
generates a loop beep+high beep. The modules load, 
but the card is detected as anonymous memory. I tried
restarting cardmgr with a few different memory ranges 
in config.opts without success. 



Thanks in advance,

--- Michael

None I think this has already been fixed

Re: Question O2Micro OZ6836/OZ6860 PCI-to-CardBus (Michael Larsen)
Date: 1999, Nov 23
From: David Hinds <dhinds@pcmcia.sourceforge.org>

This is almost certainly fixed in the 3.1.4 driver package.

-- Dave

Question Request IO: No more items

Date: 1999, Nov 30
From: Brian Mays brian

I am trying to help someone who is having problems with his PCMCIA modem. He is using version 3.1.4, and when inserting his modem, he gets the message:

  serial_cs: RequestIO: No more items

and the module is not loaded. The card used to work with a previous version of PC Card Services.

Any help would be appreciated. Thanks.

Brian

None More info?

Re: Question Request IO: No more items (Brian Mays)
Date: 1999, Dec 02
From: David Hinds <dhinds@pcmcia.sourceforge.org>

Can you get "dump_cis" output for the card?

What was the version of PCMCIA that worked?

-- Dave

Question OPTION GSM card under Linux

Date: 1999, Dec 06
From: Bernhard Graf bgraf

I have bought an OPTION GSM-only PCMCIA card and am trying to get it to work under Linux, Kernel 2.2.5 (SuSE 6.1), PCMCIA driver 3.1.5.
It is recognized by cardmgr and cardinfo displays it as a serial or modem card.
Cardctl displays it as :
Socket 1:
  product info: "Option International", "GSM Data/Fax/SMS Modem", "V1.300", ""
  manfid: 0x0013, 0x0000
  function: 2 (serial) 

But unfortunately, it seems not to work ! When I start minicom and try to connect to /dev/ttyS1, which is what the serial start command displays as the port, the modem does not answer to AT commands.

Question : Is this one of the kinds of modems that tell they are one but in reality need a (Windows) driver to function like one ? Or is it just a configuration issue ?

Please help !

Bernhard Graf

Ok OPTION GSM card works fine now !

Re: Question OPTION GSM card under Linux (Bernhard Graf)
Date: 1999, Dec 08
From: Bernhard Graf bgraf

Hi Everybody,

I am sorry to say that, but it was a very simple error that kept the OPTION GSM-Ready card from going : The serial drivers were simply not enabled in the SuSE Linux Kernel ! After I did that, it worked, although somewhat funny. After enabling all serial functions and rebooting, the GSM card was announced by the PCMCIA system (i.e. cardinfo) as working as ttyS1. Using minicom, it also answered AT commands at ttyS1. Interestingly, wvdial, the TCL-based PPP-Dialer under SuSE Linux could not talk to the modem at ttyS1. Just for fun I tried ttyS0 and : BINGO ! it worked ! I bloody don't know why, but it now works perfect.

So I definitively can recommend to buy and use this PCMCIA GSM modem under Linux !!!

Greetings, Bernhard Graf

Question Quatech quad serial card

Date: 1999, Dec 09
From: Jeff jas

Hey, I bough a Quatech quad serial card after seeing it was on the supported card list. Does anyone know where I could possibly find some documentation about making the serial devices and getting the card to work properly? cardctl -ident finds and identifies the card just fine, only I don't know what to do after that.

Thanks in advanced.

Jeff

None Check your system log

Re: Question Quatech quad serial card (Jeff)
Date: 1999, Dec 09
From: David Hinds <dhinds@pcmcia.sourceforge.org>

Did you get two high beeps when you inserted the card?

What got reported in your system log?

How about /var/run/stab (or /var/lib/pcmcia/stab)?

-- Dave

None Nokia Cardphone anyone

Date: 1999, Dec 13
From: Madsen Wikholm madsen

Has anybody tested the Nokia Cardphone 2.0. I'm in the need of mobile computing and a friend suggested I'd get this card. The question is: does it work under Linux?
I'm actually not sure if this card exists yet...but it seems like it.
best regards,
madsen

Question MC221 COM One

Date: 1999, Dec 15
From: Tobi Oetiker tobi

I just got a Vaio N505x ... It comes with a COM One Platinium
Modem CardBus Card MC221.

When I insert this card it does not get recoginzed by the PCMCIA module. "No product info available".

On the PCI bus howver (/proc/pci) it does show up as

----------------
Bus 32, device 0, function 0:
Communication controller: Lucent Mocroelectronics L56xMF (rev 1).
Medium devsel. Fast back-to-back capable
----------------

Now I guss the problem is that this thing is a soft modem. But then again it might work with this new Driver just published by lucent ...

any info on this?

tobi

None Well then, just try it...

Re: Question MC221 COM One (Tobi Oetiker)
Date: 1999, Dec 16
From: David Hinds <dhinds@pcmcia.sourceforge.org>

You can try loading the lucent driver after inserting the card, and
see what happens.

You won't be able to hot swap this card; I've played a bit with the
lucent driver and it seems to do bad things if you try to unload it.

-- Dave

Question Lucent = wavelan?

Re: None Well then, just try it... (David Hinds)
Date: 2000, Jan 18
From: Armin Obersteiner fonkie

i just found lucent wavelan cards, would be of no use for a modem or am i wrong ...

i found out the man. ids:

card "Lucent (ex-AT&T) Microelectronics L56xMF"
  manfid 0x11c1, 0x0440
  bind "serial_cs"                   

this entry won't make any difference:

boot:

nel: PCI routing table version 1.0 at 0xfdf60
nel:   00:0a.0 -> irq 9
nel: Intel PCIC probe:
nel:   Ricoh RL5C478 PCI-to-CardBus at bus 0 slot 10, mem 0x68000000, 2 sockets
nel:     host opts [0]: [isa irq] [io 3/6/1] [mem 3/6/1] [pci irq 9] [lat 168/17
nel:     host opts [1]: [isa irq] [io 3/6/1] [mem 3/6/1] [no pci irq] [lat 168/1
nel:     ISA irqs (default) = 3,4,5,7,10,11,12 polling interval = 1000 ms
nel: cs: IO port probe 0x1000-0x17ff: clean.
nel: cs: IO port probe 0x0100-0x04ff: clean.
nel: cs: IO port probe 0x0a00-0x0aff: clean.
nel: cs: cb_alloc(bus 32): vendor 0x11c1, device 0x0440
nel: cs: cb_alloc(bus 35): vendor 0x10b7, device 0x5157  

/proc/pci:        
Bus 32, device   0, function  0:
    Communication controller: Lucent (ex-AT&T) Microelectronics L56xMF (rev 1).
      Medium devsel.  Fast back-to-back capable. 

Jan 18 01:37:20 sack cardmgr[61]: initializing socket 0
Jan 18 01:37:20 sack cardmgr[61]: unsupported card in socket 0
Jan 18 01:37:20 sack cardmgr[61]:   no product info available         

how do i make the cardmgr to recognize and try the serial module ... ???

Thanx

None Sorry...

Re: Question Lucent = wavelan? (Armin Obersteiner)
Date: 2000, Jan 18
From: David Hinds <dhinds@pcmcia.sourceforge.org>

You are currently out of luck: there is, at present, no way to use
this card.

-- Dave

Sad register_serial(): autoconfig failed (Inet Spider II card -- 16550 CDPD modem)

Date: 1999, Dec 24
From: Scott Petrack scott.petrack

Dell Inspiron 7000, RedHat 6.1, kernel-pcmcia-cs-2.2.12-20

I regularly use a generic Viking modem card and Xircom RealPort CardBus 10/100 card. Just got an Inet Spider II CDPD modem card, which seems either to be (or maybe just emulate) a generic 16550 modem card. Under Windows the card works fine with the generic serial/modem drivers, using SLIP and dial-up networking, so I am full of hope that I can get it to work under Linux as well. But when I insert the card, I get the following failure (and the high-beep/low-beep sounds):

[scott]# tail /var/log/messages

Dec 24 06:34:15 scott cardmgr[900]: initializing socket 1
Dec 24 06:34:15 scott cardmgr[900]: socket 1: Serial or Modem
Dec 24 06:34:15 scott cardmgr[900]: executing: 'insmod /lib/modules/2.2.12-20/pcmcia/serial_cs.o'
Dec 24 06:34:15 scott kernel: register_serial(): autoconfig failed 
Dec 24 06:34:15 scott kernel: serial_cs: register_serial() at 0x02f8, irq 3 failed 
Dec 24 06:34:16 scott cardmgr[900]: get dev info on socket 1 failed: Resource temporarily unavailable

[scott]# dump_cis
Socket 0:
  no CIS present

Socket 1:
  dev_info
    NULL 0ns, 512b
  vers_1 5.0, "Nextcell", "Spider II Wireless IP Modem", "V1.1"
  manfid 0x01e1, 0x0100
  funcid serial_port
  config base 0x0200 mask 0x001f last_index 0x05
  cftable_entry 0x01 [default]
    [rdybsy] [pwrdown]
    Vcc Vnom 5V Istatic 200mA Iavg 400mA Ipeak 500mA
    Idown 20mA
    io 0x02f8-0x02ff [lines=0] [8bit] [range]
    irq mask 0xfffc [level]
  cftable_entry 0x03
    io 0x03e8-0x03ef [lines=0] [8bit] [range]
  cftable_entry 0x05
    io 0x02e8-0x02ef [lines=0] [8bit] [range]

The HOWTOs suggest an interrupt or port conflict, but there seems to be no interrupt nor io port conflict. As proof, I ejected the Spider card and then inserted my modem PCMCIA card. It works fine: two high beeps and here is the successful output:

[scott]# tail /var/log/messages

Dec 24 06:33:55 scott cardmgr[900]: initializing socket 1
Dec 24 06:33:55 scott cardmgr[900]: socket 1: Serial or Modem
Dec 24 06:33:55 scott cardmgr[900]: executing: 'insmod /lib/modules/2.2.12-20/pcmcia/serial_cs.o'
Dec 24 06:33:55 scott kernel: tty01 at 0x02f8 (irq = 3) is a 16550A 
Dec 24 06:33:55 scott cardmgr[900]: executing: './serial start ttyS1'

[scott]#dump_cis
Socket 0:
  no CIS present

Socket 1:
  dev_info
    no_info
  vers_1 4.1, "Viking", "V.90 K56flex", "021", "A"
  manfid 0x0013, 0x0000
  funcid serial_port
  config base 0x0200 mask 0x0017 last_index 0x30
  cftable_entry 0x20 [default]
    [rdybsy] [audio] [pwrdown]
    Vcc Istatic 25mA Iavg 250mA Ipeak 300mA Idown 5mA
    timing ready 500ms
    io 0x03f8-0x03ff [lines=10] [8bit] [range]
    irq mask 0x86bc [level]
  cftable_entry 0x21
    io 0x02f8-0x02ff [lines=10] [8bit] [range]
  cftable_entry 0x22
    io 0x03e8-0x03ef [lines=10] [8bit] [range]
  cftable_entry 0x23
    io 0x02e8-0x02ef [lines=10] [8bit] [range]
  cftable_entry 0x30
    io 0x0000-0x0007 [lines=3] [8bit]

As a final note about the Spider card, I did call Spider technical support before I bought the card and I was told that the card really does act like an generic serial port card to the computer. The page http://www.nextcell.com/solution/s2specs.html mentions that the card does "16650 emulation". From what I can see on the Windows side, there is not a single new driver installed, so it really does use the generic serial pcmcia drivers under Windows.

I know that the next step is to compile sources to serial_cs, or kernel_pcmcia_cs, but I was hoping that before I start mucking about with sources that you might be able to help me.

Thanks very much indeed. Happy holiday if you're into that sort of thing,

Scott

Note Oops, forgot cardctl -V and uname -rv

Re: Sad register_serial(): autoconfig failed (Inet Spider II card -- 16550 CDPD modem) (Scott Petrack)
Date: 1999, Dec 24
From: Scott Petrack scott.petrack

[scott]# cardctl -V
cardctl version 3.0.14
[scott]# uname -rv
2.2.12-20 #1 Mon Sep 27 10:40:35 EDT 1999

More compiled kernel-cs with pc_debug: "odd IO request"

Re: Note Oops, forgot cardctl -V and uname -rv (Scott Petrack)
Date: 1999, Dec 27
From: Scott Petrack scott.petrack

Now I've compiled and installed the pcmcia-cs sources with the PCMCIA_DEBUG flag. I can use the Viking serial card and the Xircom network card, but the Spider gives me the following errors. The new information is the one about "Odd IO request". Can someone tell me what's odd about it, and how to make it less odd? Thanks, Scott

Dec 26 14:09:00 scott cardmgr[1129]: initializing socket 0
Dec 26 14:09:00 scott cardmgr[1129]: socket 0: Serial or Modem
Dec 26 14:09:00 scott cardmgr[1129]: executing: 'insmod /lib/modules/2.2.12-20/pcmcia/serial_cs.o'
Dec 26 14:09:00 scott kernel: odd IO request: num 0008 align 0001
Dec 26 14:09:00 scott kernel: register_serial(): autoconfig failed
Dec 26 14:09:00 scott kernel: serial_cs: register_serial() at 0x02f8, irq 3 failed
Dec 26 14:09:01 scott cardmgr[1129]: get dev info on socket 0 failed: Resource temporarily unavailable

More Could the problem be with the "lines=0" in the dump_cis output?

Re: More compiled kernel-cs with pc_debug: "odd IO request" (Scott Petrack)
Date: 1999, Dec 27
From: Scott Petrack scott.petrack

The only place in the source where the "Odd IO Request message" occurs is in cs.c, and without any understanding of what's going on, could it be due to the fact that lines=0 in the output of cis_dump?

I crassly changed the code to put lines=10 (since this is what the ordinary serial modem card had), but it seems that I"ll need more luck than that.

I think that I've gone as far as I can go without actually learning and/or thinking (heaven forbid).

If it is the "lines=0" in the cis of the card, is there some nice thing I can do to override this?

Thanks

Scott

None Re: More: compiled kernel-cs with pc_debug: "odd IO request"

Re: More compiled kernel-cs with pc_debug: "odd IO request" (Scott Petrack)
Date: 1999, Dec 27
From: David Hinds <dhinds@pcmcia.sourceforge.org>

The "Odd IO request" message is more for my benefit than for anyone
else.  You're right that it means the "lines" field in the CIS is
incorrect.  But fixing it won't do any good, because the drivers
already noticed it was odd and took appropriate action.

At this point, since you've got another serial device that works fine
in exactly the same context (same IO, same irq), and this card also
claims to be a perfectly ordinary serial device, I'm not sure what
could be wrong.  The kernel serial driver is not happy with this card.
Maybe there's some mild incompatibility that isn't noticed by the
Windows serial driver, but screws up the Linux serial device probe.
Or maybe it is a timing problem in the initialization code.  But I
think the "Odd IO request" is a red herring.

-- Dave

More Point me in the right direction please?

Re: None Re: More: compiled kernel-cs with pc_debug: "odd IO request" (David Hinds)
Date: 1999, Dec 27
From: Scott Petrack scott.petrack

Thanks for this message, but I hope that with a bit of help I can get this card to work. Of course you are correct that it is likely that I'll need some technical support from the card manufacturer as well. (Of course I'll call their technical support as well).

1. You wrote:

>You're right that it means the "lines" field in the CIS is
>incorrect.  But fixing it won't do any good, because the >drivers already noticed it was odd and took appropriate >action.

Could you tell me what drivers noticed it, and what appropriate action they took? What I would have to do to the code (either PCMCIA or serial port) that would make it seem as if the card had said lines=10 or something. I'm wondering if the problem might be an incompatibility in the "appropriate action" you mention and what happens in cs.c when the Odd Request is noticed.

While we talking about the cis, are there any other parameters in the dump_cis which look funny for a generic serial card?

(Actually, I have no idea what the "right" numbers of lines should be. I wouldn't mind a quick intro to what these "lines" are. Is the number of "lines" equal to the number of bits starting with the base io address that need to be allocated? So that lines=8 mean I need one byte after the base address of 0x2f8? Is there a way to guess what the right number of lines should be?)

If you could tell me where I go in the pcmcia or kernel code to just forcibly replace the wierd cis values with "good" ones, I would try that quickly. If that works, then it's just a question of doing it more pretty.

2. Where is the detection code that actually decides if it is 16550 or what? I could always try to just force the serial port to recognize the card as a 16550, and see what happens. If this works, then I'll ask you for a pretty way to do this.

3. You mention a "timing problem in the initialization code." Your docs mention this too, and at one point I mucked about with cis_speed parameter. Your docs mention that setting it slower might help. How slow is reasonable to see if this is the problem? 500? 1500? 2000000? Similarly, could the problem be with io_speed? Any indications of things to try before I go into the serial driver would be helpful.

I am hopeful that between forcing the cis values, and forcing the recognition of 16550, and mucking with timing, I can get it to work.

I looked at a lot of CDPD cards, and this is the only low-power Type II card that claims 16550 emulation. I'll try to get to the bottom of it - either make it work or find out why it will never work.

Thanks for your help,

Scott

None Re: More: Point me in the right direction please?

Re: More Point me in the right direction please? (Scott Petrack)
Date: 1999, Dec 28
From: David Hinds <dhinds@pcmcia.sourceforge.org>

On Mon, Dec 27, 1999 at 07:41:30AM -0800, Scott Petrack wrote:
> 
> Could you tell me what drivers noticed it, and what appropriate action
> they took?

The "lines" field is the number of IO address lines the card decodes.
The setting is odd if the card claims to not decode enough lines to
distinguish all the ports it uses (the card asks for 8 ports, which
requires a minimum of 3 lines).  It is also odd if the card requires a
base address that has lines set that it claims to ignore (the card
asks for address 0x2f8, but if it decodes no address lines, it would
not be able to care about what base address is used).

If the core PCMCIA module sees a "lines" value that doesn't make
sense, it treats it as lines=16, and just tries to give the card the
exact IO port range it asks for.  If lines < 16, it can also try
alternate addresses where the undecoded address lines are changed; so,
for instance, if the card asks for 0x2f8 with lines=10, the driver
will try 0x2f8, 0x6f8, 0xaf8, etc.

In short, this isn't your problem.  The drivers were able to allocate
the first IO range the card asked for, 0x2f8-0x2ff, so the "lines"
setting was moot.

> While we talking about the cis, are there any other parameters in the
> dump_cis which look funny for a generic serial card?

Not that I noticed.

> 2. Where is the detection code that actually decides if it is 16550 or
> what? I could always try to just force the serial port to recognize the
> card as a 16550, and see what happens. If this works, then I'll ask you
> for a pretty way to do this.

It is in the kernel serial driver, linux/drivers/char/serial.c.  I
doubt the problem is identification of what kind of UART this is.
More likely, the problem is that the drivers can't talk to the card,
and the first thing that fails is the UART test.

> 3. You mention a "timing problem in the initialization code." Your docs
> mention this too, and at one point I mucked about with cis_speed
> parameter.

cis_speed affects card recognition.  Your card is identified fine.

I would increase unreset_delay to, say, 400, which puts a 4 second
delay between when the card is powered up, and when the serial driver
is fired up.

-- Dave

Ok SUCCESS, thanks to you. But please check the workaround....

Re: None Re: More: Point me in the right direction please? (David Hinds)
Date: 2000, Jan 02
From: Scott Petrack scott.petrack

Well, you were exactly right -- it all worked as soon as I commented out the following probe code in /usr/src/linux/drivers/char/serial.c (I included the comments from the source as well):

        /*
         * Do a simple existence test first; if we fail this, there's
         * no point trying anything else.
         *
         * 0x80 is used as a nonsense port to prevent against false
         * positives due to ISA bus float.  The assumption is that
         * 0x80 is a non-existent port; which should be safe since
         * include/asm/io.h also makes this assumption.
         */

        scratch = serial_inp(info, UART_IER);
        serial_outp(info, UART_IER, 0);
        outb(0xff, 0x080);
        scratch2 = serial_inp(info, UART_IER);
        serial_outp(info, UART_IER, scratch);
        if (scratch2) {
                restore_flags(flags);
                return;          /*We failed; there's nothing here*/
        }

I had inserted a printk before the return to discover that this was the cause of the autoconfig failure. I suppose that the emulation of this card can deal with any port, including 0x80 (not that I quite know what this means).

Question: what are the bad consequences of leaving out this code? Do I have access to some value or string in this driver which I could use to turn off the probe for this card only? Is there something cleaner that I can do to use my card, knowing as I now do that I have to turn off this test?

I can now talk to the card using minicom, although I haven't yet got myself connected, I imagine that this is a matter of minutes. Is there something I should do so that others can know how to use the Inet Spider II PC Card under Linux?

Thanks for all your help.

Scott

Question modem doesn't work until setserial irq 0 -- but works after

Date: 1999, Dec 29
From: John Ruttenberg rutt

System brand & model: gateway solo 9100xl
PCMCIA card: telepath 56k modem
Linux version: 2.2.13
Pcmcia version: 3.1.7 (+cardbus +PnP)

This modem doesn't work at all until I do:

	setserial /dev/modem irq 0

But the strange thing is that once this has been done once,
the modem will aways work even after:

	setserial /dev/modem irq 11

(11 is the irq assigned by pcmcia in the first place.)
In fact, the modem will even work after stopping and restarting pcmcia  with:

	/etc/rc.d/init.d/pcmcia stop
	/etc/rc.d/init.d/pcmcia start

But after a reboot, the modem won't work until the irq has once been set to 0.

By the way, I have:

	PCIC_OPTS="pci_csc=1"

None Kernel bug, workaround in 3.1.10

Re: Question modem doesn't work until setserial irq 0 -- but works after (John Ruttenberg)
Date: 2000, Feb 04
From: David Hinds <dhinds@pcmcia.sourceforge.org>

This turned out to be a kernel bug, but a workaround is included in
the 3.1.10 driver package.

-- Dave

Question modem dead?

Date: 2000, Jan 03
From: Richard Sevenich rsevenic

I have a laptop whose modem has stopped working and I know not whether it is a software or hardware problem - being relatively unfamiliar with pcmcia.

Symptoms:

1. When card services start, the card is successfully identified.

2. cardctl status indicates on the 2nd line
	function 0: [ready], [bat low], [bat dead]
   (the latter two flags make me nervous)

3. At start, /dev/modem is created as a soft link to
   /dev/ttyS1, but I cannot examine /dev/modem with
   setserial - the attempt hangs the system

4. The card is a PCMCIA MAKER, 56K V.90 DataFax Modem

I would appreciate help/hints.

Regards, Richard

None Hard to tell without more testing

Re: Question modem dead? (Richard Sevenich)
Date: 2000, Jan 03
From: David Hinds <dhinds@pcmcia.sourceforge.org>

If the modem stopped working without any change in your system
configuration (you didn't upgrade your kernel, didn't upgrade the
PCMCIA drivers, didn't add new hardware or change BIOS settings), then
my guess would be a hardware problem.  But the only way to really tell
would be to try another modem, or try this modem under Windows, or try
this modem in another laptop.

I don't recall hearing of driver problems that would cause setserial
to fail.  Is the port detected properly as a 16550A uart?  (check your
system log)

-- Dave

Sad Modem of Xircom multifunction stalls at reproducible point

Date: 2000, Jan 13
From: Jon Sundquist <jsundqui@buffnet.net>

The ethernet part of my Xircom REM56G-10BT (it is the 16-bit realport card, combo 10BT ether and 56K modem) works great as an ethernet card. It also works in modem mode, but with troubles. I can reproducibly stall a file transfer.

 Setup:
 Dell Inspiron 3500
 Redhat 6.1, with pcmcia upgraded to 3.1.8 (upgrade didn't help)
 Card: Xircom REM56G-10BT

 inspiron:~$ uname -rv
 2.2.12-20 #1 Mon Sep 27 10:40:35 EDT 1999
 inspiron:~$ /sbin/cardctl -V
 cardctl version 3.1.8

More details of problem: I use kppp to initiate a modem ppp connection (a usenet post said that kppp may be a culprit, but that seems unlikely, and I have not yet crafted a straight pppd/chat script to confirm). Connection is made OK (sometimes not at too great a speed, but I chalk that up to poor lines). Things like web surfing are fine.

Now go to this page: http://www.vmware.com/download/downloadtools.html and try to download the vmware-tools package from any of the mirrors there. It always, reproducibly stalls at 5% (at about 77000 - 78000 bytes). It will stall from any of the mirrors, both from ftp and http. Interestingly, although it stalls the same place each time, some packets still come through. I.e., it transfers fine till about 5%, then maybe a packet or two every minute, but gradually getting even slower.

I mainly use the ethernet, so I havaen't done extensive testing for different sites/limitations. However, there were other downloads where I had similar problems (e.g. a TkDesk rpm from RedHat contrib IIRC). However, I have also had some downloads with no problems.

Again, loading web pages seems unaffected. I had some problems (long delays) loading/refreshing an IMAP mailbox, but that has not been quite so reproducible.

This is a dual boot machine. I had no problem downloading these files in windows via modem.

When swicthing between ethernet and modem, I first bring down the eth0 interface (ifconfig down) along with the associated routes.

Script changes: none serial-related. I added the network.opts as recommended for redhat. I also used redhats /etc/pcmcia/network. But then again, I am not having problems with the network, but with the modem/serial.

From the logs: This is from /var/log/messages from after I restarted cardmgr (via /etc/rc.d/init.d/pcmcia stop|start) following the installation of 3.1.8. Note, I first tested ethernet to see if it worked, then brought that down, and ran the ppp session.

Jan 12 18:16:36 inspiron cardmgr[6631]: starting, version is 3.1.8
Jan 12 18:16:36 inspiron cardmgr[6631]: watching 2 sockets
Jan 12 18:16:36 inspiron cardmgr[6631]: initializing socket 0
Jan 12 18:16:36 inspiron cardmgr[6631]: socket 0: Xircom RealPort REM10BT Ethernet/Modem
Jan 12 18:16:36 inspiron cardmgr[6631]: executing: 'insmod /lib/modules/2.2.12-20/pcmcia/xirc2ps_cs.o'
Jan 12 18:16:36 inspiron cardmgr[6631]: executing: 'insmod /lib/modules/2.2.12-20/pcmcia/serial_cs.o'
Jan 12 18:16:41 inspiron cardmgr[6631]: executing: './network start eth0'
Jan 12 18:16:45 inspiron cardmgr[6631]: executing: './serial start ttyS3'
Jan 12 18:21:50 inspiron modprobe: can't locate module char-major-108
Jan 12 18:21:50 inspiron pppd[6744]: pppd 2.3.10 started by root, uid 0
Jan 12 18:21:50 inspiron pppd[6744]: Using interface ppp0
Jan 12 18:21:50 inspiron pppd[6744]: Connect: ppp0 <--> /dev/ttyS3
Jan 12 18:21:56 inspiron pppd[6744]: Remote message: Login Succeeded
Jan 12 18:21:57 inspiron pppd[6744]: local  IP address 208.28.191.14
Jan 12 18:21:57 inspiron pppd[6744]: remote IP address 208.28.191.2
Jan 12 18:22:27 inspiron pppd[6744]: CCP: timeout sending Config-Requests 
Jan 12 18:28:51 inspiron pppd[6744]: Terminating on signal 15.
Jan 12 18:28:51 inspiron pppd[6744]: Connection terminated.
Jan 12 18:28:51 inspiron pppd[6744]: Connect time 7.1 minutes.
Jan 12 18:28:51 inspiron pppd[6744]: Sent 21300 bytes, received 480012 bytes.
Jan 12 18:28:51 inspiron pppd[6744]: Exit.

I will post this at the Xircom area as well, although not sure if that is for the network "half" of it only.

TIA for any suggestions

Sad Modem doesn't work with versions 3.1.4 or later.

Date: 2000, Jan 14
From: Dave Caswell davec

I can't get my PCMCIA serial port to work with versions of pcmcia-cs
3.1.4 and beyond.  I haven't changed any configuration options, I just
install the new versions over the old, and 3.1.4 on don't work.  

When I start pppd, syslog gets an error:
Jan 13 16:24:01 krypton pppd[15407]: tcgetattr: Input/output error(5)
and pppd exits.

Looking back in my logs, I see that sometimes tty01 gets ports at
0x13f8, and sometimes at 0x02f8.  But the card works at both addresses
with the earlier versions of pcmcia_cs.

My computer started life as a Debian distribution, but is now running
kernel 2.1.14.  The computer itself is a Toshiba Satellite Pro 445 CDX. 


Linux krypton 2.2.14 #2 Wed Jan 5 14:48:34 MST 2000 i586 unknown

Socket 0: ActionTek Datalink Modem:
  product info: "PCMCIA  ", "33.6K Fax/Modem  ", "FM336LK  "
  manfid: 0x0175, 0x0000
  function: 2 (serial)
Socket 1:
  product info: "SMC", "EtherEZ Ethernet 8020", "V1.00", ""
  manfid: 0x0108, 0x0001
  function: 6 (network)



The log file lines that relate to PCMCIA are:


Jan 13 16:23:56 krypton kernel: Linux PCMCIA Card Services 3.1.4 
Jan 13 16:23:56 krypton kernel:   kernel build: 2.2.14 #2 Wed Jan 5 14:48:34 MST 2000 
Jan 13 16:23:56 krypton kernel:   options:  [pci] [cardbus] [pnp] 
Jan 13 16:23:56 krypton kernel: PnP: PNP BIOS installation structure at 0xc00f8e30 
Jan 13 16:23:56 krypton kernel: PnP: PNP BIOS version 1.0, entry at f0000:9c9d, dseg at 0 
Jan 13 16:23:56 krypton kernel: PCI routing table version 1.0 at 0xf97e0 
Jan 13 16:23:56 krypton kernel: Intel PCIC probe:  
Jan 13 16:23:56 krypton kernel:   Intel i82365sl B step ISA-to-PCMCIA at port 0x3e0 ofs 0x00, 2 sockets 
Jan 13 16:23:56 krypton kernel:     host opts [0]: none 
Jan 13 16:23:56 krypton kernel:     host opts [1]: none 
Jan 13 16:23:56 krypton kernel:     ISA irqs (scanned) = 3,4,10 polling interval = 1000 ms 
Jan 13 16:23:56 krypton cardmgr[15388]: starting, version is 3.1.4
Jan 13 16:23:56 krypton cardmgr[15388]: watching 2 sockets
Jan 13 16:23:56 krypton kernel: cs: IO port probe 0x1000-0x17ff: clean. 
Jan 13 16:23:56 krypton kernel: cs: IO port probe 0x0100-0x04ff: clean. 
Jan 13 16:23:56 krypton kernel: cs: IO port probe 0x0a00-0x0aff: clean. 
Jan 13 16:23:57 krypton cardmgr[15388]: initializing socket 0
Jan 13 16:23:57 krypton kernel: cs: memory probe 0x0d0000-0x0dffff: clean. 
Jan 13 16:23:57 krypton cardmgr[15388]: socket 0: Serial or Modem
Jan 13 16:23:57 krypton cardmgr[15388]: executing: 'insmod /lib/modules/2.2.14/pcmcia/serial_cs.o'
Jan 13 16:23:57 krypton kernel: tty01 at 0x13f8 (irq = 3) is a 16550A 
Jan 13 16:23:57 krypton cardmgr[15388]: executing: './serial start ttyS1'
Jan 13 16:24:01 krypton pppd[15407]: pppd 2.3.5 started by davec, uid 1000
Jan 13 16:24:01 krypton pppd[15407]: tcgetattr: Input/output error(5)
Jan 13 16:24:02 krypton pppd[15407]: Exit.
Jan 13 16:24:41 krypton cardmgr[15388]: executing: './serial check ttyS1'
Jan 13 16:24:42 krypton cardmgr[15388]: shutting down socket 0
Jan 13 16:24:42 krypton cardmgr[15388]: executing: './serial stop ttyS1'
Jan 13 16:24:42 krypton kernel: tty01 unloaded 
Jan 13 16:24:42 krypton cardmgr[15388]: executing: 'rmmod serial_cs'
Jan 13 16:24:42 krypton cardmgr[15388]: exiting


Thanks for your help.
    -davec

Ok change in config.opts makes modem work again

Re: Sad Modem doesn't work with versions 3.1.4 or later. (Dave Caswell)
Date: 2000, Jan 15
From: Dave Caswell davec

Modifying this line in my /etc/pcmcia/config.opts file
like this has made my modem work again!  I'm now using 
version 3.1.8

#include port 0x100-0x4ff, port 0x1000-0x17ff
include port 0x100-0x4ff

-davec

Question serial_cs device not found & null pointer

Date: 2000, Jan 18
From: Mario Rossi ArmyBot

I'm having problem trying to install a PCMCIA modem...(linux, slackware 7.0)
etc/pcmcia/config reflects what HOWTOs says
device "serial_cs"
  class "serial" module "misc/serial", "serial_cs"

I specified the manfid got with cardctl status, NO way I even tried specifying "serial_cs" as module ( 'cause there's serial_cs.o in pcmcia where is supposed to stay the file if no path is specified)

NO WAY stab says socket 1 is empty ( not true)

I'll add the whole boot log I got a null pointer probiung the second socket ( the modem ) , in fact the eth0 works fine but if you do cardctl status you see : Socket 0: 5V 16-bit card present Function 0: ready Socket 1: Socket 0: 5V 16-bit card present Function 0: ready,write protect <--- ????

Also it seems not to find the module serial_cs ( the same for serial) Could you help me ?

Thanks a lot ArmyBot

Linux version 2.2.13 (root@RsQgC) (gcc version egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)) #4 Tue Jan 18 18:29:37 CET 2000
Detected 119993660 Hz processor.
Console: colour VGA+ 80x25
Calibrating delay loop... 47.82 BogoMIPS
Memory: 62996k/65536k available (1112k kernel code, 412k reserved, 956k data, 60k init)
CPU: Intel Pentium 75 - 200 stepping 0c
Checking 386/387 coupling... OK, FPU using exception 16 error reporting.
Checking 'hlt' instruction... OK.
Intel Pentium with F0 0F bug - workaround enabled.
POSIX conformance testing by UNIFIX
PCI: PCI BIOS revision 2.10 entry at 0xf7019
PCI: Using configuration type 1
PCI: Probing PCI hardware
Linux NET4.0 for Linux 2.2
Based upon Swansea University Computer Society NET3.039
NET4: Unix domain sockets 1.0 for Linux NET4.0.
NET4: Linux TCP/IP 1.0 for NET4.0
IP Protocols: ICMP, UDP, TCP, IGMP
Initializing RT netlink socket
Starting kswapd v 1.5 
Detected PS/2 Mouse Port.
Serial driver version 4.27 with HUB-6 MANY_PORTS MULTIPORT SHARE_IRQ enabled
ttyS00 at 0x03f8 (irq = 4) is a 16550A
ttyS01 at 0x02f8 (irq = 3) is a 16550A
pty: 512 Unix98 ptys configured
apm: BIOS version 1.2 Flags 0x03 (Driver version 1.9)
Real Time Clock Driver v1.09
RAM disk driver initialized:  16 RAM disks of 4096K size
loop: registered device at major 7
PCI_IDE: unknown IDE controller on PCI bus 00 device 11, VID=10b9, DID=5219
PCI_IDE: not 100% native mode: will probe irqs later
PCI_IDE: simplex device:  DMA disabled
ide0: PCI_IDE Bus-Master DMA disabled (BIOS)
PCI_IDE: simplex device:  DMA disabled
ide1: PCI_IDE Bus-Master DMA disabled (BIOS)
hda: HITACHI_DK223A-81, ATA DISK drive
hdc: UJDCD8730, ATAPI CDROM drive
ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
ide1 at 0x170-0x177,0x376 on irq 15
hda: HITACHI_DK223A-81, 775MB w/128kB Cache, CHS=787/32/63
hdc: ATAPI 8X CD-ROM drive, 128kB Cache
Uniform CDROM driver Revision: 2.56
Floppy drive(s): fd0 is 1.44M
FDC 0 is a National Semiconductor PC87306
md driver 0.36.6 MAX_MD_DEV=4, MAX_REAL=8
linear personality registered
raid0 personality registered
Partition check:
 hda: hda1 hda2
VFS: Mounted root (ext2 filesystem) readonly.
Freeing unused kernel memory: 60k freed
Adding Swap: 45356k swap-space (priority -1)
parport0: PC-style at 0x378 [SPP]
lp0: using parport0 (polling).
CSLIP: code copyright 1989 Regents of the University of California
SLIP: version 0.8.4-NET3.019-NEWTTY-MODULAR (dynamic channels, max=256).
SLIP linefill/keepalive option.
PPP: version 2.3.7 (demand dialling)
PPP line discipline registered.
PPP BSD Compression module registered
Soundblaster audio driver Copyright (C) by Hannu Savolainen 1993-1996
SB 3.01 detected OK (220)
ESS chip ES1688 detected
Linux PCMCIA Card Services 3.0.14
  kernel build: 2.2.13 #22 Wed Oct 20 17:32:52 CDT 1999
  options:  [pci] [cardbus]
Intel PCIC probe: 
  TI 1130 PCI-to-CardBus at bus 0 slot 4, mem 0x68000000, 2 sockets
    host opts [0]: [isa irq] [no pci irq] [lat 64/176] [bus 1/5]
    host opts [1]: [isa irq] [no pci irq] [lat 64/176] [bus 6/10]
    ISA irqs (scanned) = 7,9,10,11,12 polling interval = 1000 ms
cs: IO port probe 0x1000-0x17ff: clean.
cs: IO port probe 0x0100-0x04ff: excluding 0x178-0x17f 0x388-0x38f 0x408-0x40f 0x480-0x48f 0x4d0-0x4d7
cs: IO port probe 0x0a00-0x0aff: clean.
cs: memory probe 0xa0000000-0xa0ffffff: clean.
eth0: NE2000 Compatible: io 0x320, irq 9, hw_addr 00:80:C8:87:E5:2C
Unable to handle kernel NULL pointer dereference at virtual address 00000004
current->tss.cr3 = 03d39000, %cr3 = 03d39000
*pde = 00000000
Oops: 0000
CPU:    0
EIP:    0010:[<c484dda8>]
EFLAGS: 00010246
eax: 00000000   ebx: c0071280   ecx: c3c897e0   edx: 00000000
esi: 00000000   edi: c3d7b800   ebp: c3c7dad8   esp: c3c7dad0
ds: 0018   es: 0018   ss: 0018
Process cardmgr (pid: 72, process nr: 9, stackpage=c3c7d000)
Stack: c3c897e0 c3d7b800 c3c7daec c484e6f3 c3c897e0 00000000 00000000 c3c7db6c 
       c486f2e9 0000002d c3c897e0 00000000 c3c89900 00000000 c486f6e0 4d000000 
       c486f6e0 c3c7db00 00000000 00000032 00000032 00000032 49434d43 41432041 
Call Trace: [<c484e6f3>] [<c486f2e9>] [<c486f6e0>] [<c486f6e0>] [<c01acd1b>] [<c486f4ee>] [<c486f450>] 
       [<c484d589>] [<c4857135>] [<c485b9a6>] [<c486d082>] [<c484e8f5>] [<c484e693>] [<c486f14e>] [<c486f6e4>] 
       [<c486f450>] [<c484f8a8>] [<c486f056>] [<c48625f3>] [<c01c1d89>] [<c4862f89>] [<c01b45af>] [<c01b5ccf>] 
       [<c01b705e>] [<c01b82e3>] [<c01bc8e1>] [<c01bbe63>] [<c01bc053>] [<c011c8ff>] [<c011ca2b>] [<c011c2ef>] 
       [<c011c5a1>] [<c012dc45>] [<c0109ef0>] 
Code: 8b 56 04 f6 c2 05 75 0d 66 83 7e 0a 00 74 0d f6 46 0c 05 74 
ISO 9660 Extensions: RRIP_1991A

Question MultiTech MT5634ZLX

Date: 2000, Jan 19
From: Aaron Kaplan kaplan

I can't get my MultiTech MT5634ZLX modem to work. The card is recognized and assigned an appropriate IRQ and I/O port, but when I try to talk to it using minicom, nothing happens. I'm quite sure this is not a winmodem--the Windows disk that came with it contains only a .INF file and a DOS point enabler, no special driver.

I'm using a Toshiba Satellite Pro 490XCDT, running RedHat 6.1. I tried first with the pcmcia package that was included in the distribution (3.0.14), and then upgraded to the latest 3.1.9, to no avail.

This modem works fine under Windows. I have another modem card, a 14.4 from Fujitsu, that works just fine under Linux, in the same slot and using the same IRQ and port.

MultiTech tech support won't help me.

-Aaron

None Interrupt conflict?

Re: Question MultiTech MT5634ZLX (Aaron Kaplan)
Date: 2000, Jan 19
From: David Hinds <dhinds@pcmcia.sourceforge.org>

The most likely problem is an interrupt conflict.  The easiest way to
tell if this is the problem is to do:

    setserial /dev/modem irq 0

and see if that seems to improve the situation.  If so, then you'll
need to start excluding interrupts in /etc/pcmcia/config.opts until
the problem is resolved.

-- Dave

None Re: Interrupt conflict?

Re: None Interrupt conflict? (David Hinds)
Date: 2000, Jan 20
From: Aaron Kaplan kaplan

> The most likely problem is an interrupt conflict.  The easiest way to
> tell if this is the problem is to do:
> 
>     setserial /dev/modem irq 0

I did that (as suggested in the FAQ), but it didn't change anything.
Also, as I said, a different modem card works fine when assigned the
same IRQ.

-Aaron

None Uh-oh

Re: None Re: Interrupt conflict? (Aaron Kaplan)
Date: 2000, Jan 20
From: David Hinds <dhinds@pcmcia.sourceforge.org>

Oh.  Now I remember that there is another report of this same modem
not working under Linux, and I've had a report of a similar problem
with another Multitech modem (the MT3400ZLX).  

I think this confirms that there's something specifically wrong with
these cards.

Could you send me the driver files (the .INF and the point enabler)
and the output of "dump_cis"?

-- Dave

Ok A fix will be in the next 3.1.9 beta

Re: Question MultiTech MT5634ZLX (Aaron Kaplan)
Date: 2000, Jan 25
From: David Hinds <dhinds@pcmcia.sourceforge.org>

The problem was that the modem needs more power than its configuration information indicates.

The next 3.1.9 beta will include a fix.

-- Dave

Sad Sierra Wireless Modem

Date: 2000, Jan 25
From: Roy Murphy Royster

I just upgraded my system from RedHat 5.2 to Mandrake 7.0 and my Sierra Wireless Modem/Cell Modem card has stopped working. I applied Jeff Schoby's patches for the aircard to a version 3.1.9 PCMCIA source and recompiled/reinstalled. I just mention this for completeness. The patch dosn't touch the serial_cs driver which is causing me the problem. I've also checked against an unpatched 3.1.9 install.

The problem is that the modem does not respond. If I load minicom, I can't get it to return an OK to ATZ and minicom hangs and will not quit if I try to shut it down until I remove the card. The card works in a Windows laptop (I can get an OK back from a terminal program) and should appear to be a generic modem. A Megahertz modem card works fine in this installation,

From /var/log/messages:

Jan 25 09:13:19 agape cardmgr[276]: initializing socket 0
Jan 25 09:13:19 agape cardmgr[276]: socket 0: Serial or Modem
Jan 25 09:13:20 agape cardmgr[276]: executing: 'insmod /lib/modules/2.2.14-15mdksecure/pcmcia/serial_cs.o'
Jan 25 09:13:20 agape kernel: tty02 at 0x13f8 (irq = 3) is a 16450
Jan 25 09:13:20 agape cardmgr[276]: executing: './serial start ttyS2'          

Two beeps on install. So far, so good.

[root@agape linux]# dump_cis
Socket 0:
  dev_info
    NULL 0ns, 512b
  vers_1 4.1, "Sierra Wireless", "Aircard Modem", "Version 1.0"
  manfid 0x0192, 0x0201
  funcid serial_port [post]
  serial_data_services
    data_rate 115200
    modulation [Bell103] [V.21] [V.23] [V.22] [Bell212A]
      [V.22bis] [V.27bis] [V.29] [V.32] [V.32bis]
    error_control [MNP2-4] [V.42/LAPM]
    compression [V.42bis] [MNP5]
    cmd_protocol [AT1] [AT2] [AT3] [MNP_AT] [V.25bis]
  serial_fax_services [class=1]
    data_rate 115200
    modulation [V.21-C2] [V.27ter] [V.29] [V.17]
    features [T.4]            
  config base 0x1000 mask 0x0007 last_index 0x23
  cftable_entry 0x20 [default]
    [rdybsy] [audio]
    Vcc Vnom 5V Istatic 180mA Ipeak 220mA
    timing ready 3705032704ns
    io 0x03f8-0x03ff [lines=10] [8bit] [range]
    irq mask 0x86bc [level]
  cftable_entry 0x21
    io 0x02f8-0x02ff [lines=10] [8bit] [range]
  cftable_entry 0x22
    io 0x03e8-0x03ef [lines=10] [8bit] [range]
  cftable_entry 0x23                                   
   io 0x02e8-0x02ef [lines=10] [8bit] [range]

everything looks cozy.

From the startup of PCMCIA services:

Jan 25 09:00:03 agape pcmcia: Starting PCMCIA services:
Jan 25 09:00:03 agape pcmcia:  modules
Jan 25 09:00:04 agape kernel: Linux PCMCIA Card Services 3.1.9
Jan 25 09:00:04 agape kernel:   kernel build: 2.2.14-15mdksecure #1 SMP Tue Jan 4 21:15:44 CET 2000
Jan 25 09:00:04 agape kernel:   options:  [pci] [cardbus]
Jan 25 09:00:04 agape kernel: Intel PCIC probe:
Jan 25 09:00:04 agape kernel:   TI 1131 PCI-to-CardBus at bus 0 slot 6, mem 0x68000000, 2 sockets
Jan 25 09:00:04 agape kernel:     host opts [0]: [ring] [clkrun irq 10] [pci + serial irq] [no pci irq] [lat 64/176] [bus 1/1]
Jan 25 09:00:04 agape kernel:     host opts [1]: [ring] [clkrun irq 10] [pci + serial irq] [no pci irq] [lat 64/176] [bus 2/2]
Jan 25 09:00:04 agape kernel:     ISA irqs (scanned) = 3,4,7,10,15 polling interval = 1000 ms
Jan 25 09:00:04 agape pcmcia:  cardmgr.
Jan 25 09:00:05 agape cardmgr[276]: starting, version is 3.1.9
Jan 25 09:00:05 agape rc: Starting pcmcia succeeded
Jan 25 09:00:05 agape cardmgr[276]: watching 2 sockets
Jan 25 09:00:05 agape kernel: cs: IO port probe 0x1000-0x17ff: clean.
Jan 25 09:00:05 agape kernel: cs: IO port probe 0x0100-0x04ff: excluding 0x170-0x177 0x200-0x207 0x220-0x22f 0x330-0x337 0x370-0x37f 0x388-0x38f 0x398-0x39f 0x4d0-0x4d7
Jan 25 09:00:05 agape kernel: cs: IO port probe 0x0a00-0x0aff: clean.

None It's an IO port issue

Re: Sad Sierra Wireless Modem (Roy Murphy)
Date: 2000, Jan 25
From: David Hinds <dhinds@pcmcia.sourceforge.org>

Exclude the IO port range 0x1000-0x17ff in /etc/pcmcia/config.opts and
all should be well.

-- Dave

None Very strange situation - difficult to find a good title!

Date: 2000, Feb 03
From: Alessandro Russo <russo@ian.pv.cnr.it>

I'm having different troubles with 2 modem cards. 

Configuration:

System brand & model: Toshiba 4010CDT
Linux version: 2.2.14 (mandrake 7.0)
Pcmcia version: 3.1.9

------------------------------------------------------------

1) card: 56000bps UltraSpeed

What happens is precisely the same as John Ruttenberg
described in message 10. Everything is fine from the logs,
but the modem doesn't work. But if I do (assuming that
5 was the irq assigned by pcmcia, but it works with other
irqs)

setserial /dev/modem irq 0
setserial /dev/modem irq 5

then everything is ok. Tha card can also be swapped.

This card used to work perfectly with Suse 5.3 on the
same computer.

It works perfectly on a Toshiba 4030CDT
with redhat 6.1 (kernel 2.2.12, PCMCIA 3.0.14)

-----------------------------------------------------------

2) card: 3COM 3CXM756 Global GSM and Modem PC Card

What happens is similar to what described by Jaime V Miro
in message 1. The UART detected is wrong, it is 8250 
instead of 16550A (for the other card the UART detection
is fine) and the modem doesn't work. I have not been able
to make it working by changing UART and irq with setserial.

Also this card works perfectly on my Toshiba 4030CDT
with redhat 6.1 (kernel 2.2.12, PCMCIA 3.0.14)

----------------------------------------------------------

I also have a 3C598D ethernet CARD that works well in both
computers and in all slots.


Thanks in advance for your attention.

Ciao

Alessandro Russo
IAN-CNR, Pavia (ITALY)
----------------------------------------------------
"Microsoft is learning a hard lesson with Linux, 
namely that there is more to life than product marketing"

russo@ian.pv.cnr.it

None At least the first part is fixed in 3.1.10

Re: None Very strange situation - difficult to find a good title! (Alessandro Russo)
Date: 2000, Feb 03
From: David Hinds <dhinds@pcmcia.sourceforge.org>

Try installing 3.1.10, and/or just copy the etc/serial file in 3.1.10
to /etc/pcmcia/serial.

The updated serial script just does the setserial irq 0 trick.  This
is a workaround for a bug in the kernel serial device driver.

For the second card, I'm not sure what the problem might be.  I might
want to see system log messages for that.

-- Dave

Question serial modem on Inspiron 3500

Date: 2000, Jan 03
From: Dirk Arnold dirk

  Hello,

  I have an Inspiron 3500 running RH6.1 (kernel 2.2.12;
pcmcia-3.0.14) and a Xircom RBM56G CardBus modem. probe
yields

PCI bridge probe: TI 1220 found, 2 sockets.


The card is identified correctly but cannot be configured.
Here's the relevant part of the system log:

cardmgr[289]: initializing socket 0
cardmgr[289]: socket 0: Serial or Modem
cardmgr[289]: executing: 'insmod /lib/modules/2.2.12-20/pcmcia/serial_cs.o'
kernel: serial_cs: ParseTuple: No more items
cardmgr[289]: get dev info on socket 0 failed: Resource temporarily unavaliable

Is this fixed by a newer version of the pcmcia package? Any
help very much appreciated.

  Dirk

Question Trouble with FM560CB modem card

Date: 2000, Feb 22
From: Martin Buck <mb@netwings.ch>

I've got a few problems with a FM560CB PCMCIA modem card
(probably made by Abocom - http://www.abocom.com - although
the package doesn't say anything about the manufacturer).

If I plug it in (into a HP Omnibook 4150) with a unmodified
PCMCIA config, I get the following:

kernel: Linux PCMCIA Card Services 3.1.8
kernel:   kernel build: 2.2.10 #8 Thu Oct 28 09:49:56 CEST 1999
kernel:   options:  [pci] [cardbus] [apm]
kernel: Intel PCIC probe:
kernel:   TI 1220 PCI-to-CardBus at bus 0 slot 4, mem 0x68000000, 2 sockets
kernel:     host opts [0]: [pwr save] [serial pci & irq] [no pci irq] [lat 64/32] [bus 32/34]
kernel:     host opts [1]: [pwr save] [serial pci & irq] [no pci irq] [lat 64/32] [bus 35/37]
kernel:     ISA irqs (scanned) = 3,4,5,7,11 polling interval = 1000 ms
cardmgr[1151]: starting, version is 3.1.8
cardmgr[1151]: watching 2 sockets
kernel: cs: IO port probe 0x1000-0x17ff: clean.
kernel: cs: IO port probe 0x0100-0x04ff: excluding 0x120-0x127 0x220-0x22f 0x378-0x37f 0x388-0x38f
kernel: cs: IO port probe 0x0a00-0x0aff: clean.
cardmgr[1151]: initializing socket 1
kernel: cs: memory probe 0xa0000000-0xa0ffffff: clean.
cardmgr[1151]: error parsing CIS on socket 1: Invalid argument
cardmgr[1151]: unsupported card in socket 1
cardmgr[1151]:   product info: "V90&K56Flex PCMCIA FAX MODEM", "", "", ""
cardmgr[1151]: exiting

So I added this to /etc/pcmcia/config:

card "V90&K56Flex PCMCIA FAX MODEM"
  version "V90&K56Flex PCMCIA FAX MODEM", "", "", ""
  bind "serial_cs"

Now the card is recognized, but configuration fails:

kernel: Linux PCMCIA Card Services 3.1.8
kernel:   kernel build: 2.2.10 #8 Thu Oct 28 09:49:56 CEST 1999
kernel:   options:  [pci] [cardbus] [apm]
kernel: Intel PCIC probe:
kernel:   TI 1220 PCI-to-CardBus at bus 0 slot 4, mem 0x68000000, 2 sockets
kernel:     host opts [0]: [pwr save] [serial pci & irq] [no pci irq] [lat 64/32] [bus 32/34]
kernel:     host opts [1]: [pwr save] [serial pci & irq] [no pci irq] [lat 64/32] [bus 35/37]
kernel:     ISA irqs (scanned) = 3,4,5,7,11 polling interval = 1000 ms
cardmgr[1354]: starting, version is 3.1.8
cardmgr[1354]: watching 2 sockets
kernel: cs: IO port probe 0x1000-0x17ff: clean.
kernel: cs: IO port probe 0x0100-0x04ff: excluding 0x120-0x127 0x220-0x22f 0x378-0x37f 0x388-0x38f
kernel: cs: IO port probe 0x0a00-0x0aff: clean.
cardmgr[1354]: initializing socket 1
kernel: cs: memory probe 0xa0000000-0xa0ffffff: clean.
cardmgr[1354]: error parsing CIS on socket 1: Invalid argument
cardmgr[1354]: socket 1: V90&K56Flex PCMCIA FAX MODEM
cardmgr[1354]: executing: 'insmod /lib/modules/2.2.10/pcmcia/serial_cs.o'
kernel: register_serial(): autoconfig failed
kernel: serial_cs: register_serial() at 0x02f8, irq 3 failed
cardmgr[1354]: get dev info on socket 1 failed: Resource temporarily unavailable

The "error parsing CIS" looks suspicious to me. Do I need a
magic file that I can drop into /etc/pcmcia/cis to make the
card work?

BTW, this is what cis_dump -v has to say:

Socket 0:
  no CIS present
Socket 1:
  offset 0x02, tuple 0x01, link 0x01
    d1
  parse error: Invalid argument

  offset 0x05, tuple 0x15, link 0x2b
    04 00 56 39 30 26 4b 35 36 46 6c 65 78 20 50 43
    4d 43 49 41 20 46 41 58 20 4d 4f 44 45 4d 00 00
    00 00 00 00 00 00 00 00 00 00 00
  vers_1 4.0, "V90&K56Flex PCMCIA FAX MODEM", "", "", ""

  offset 0x32, tuple 0x20, link 0x02
    02 00
  parse error: Invalid argument

  offset 0x36, tuple 0x1a, link 0x05
    01 23 00 01 03
  config base 0x0100 mask 0x0003 last_index 0x23

  offset 0x3d, tuple 0x1b, link 0x14
    e1 81 9d 68 86 23 86 50 35 fc 15 aa 60 f8 02 07
    30 b8 86 68
  cftable_entry 0x21 [default]
    [mwait] [audio] [pwrdown]
    Vcc Istatic 135mA Ipeak 180mA Idown 30mA
    timing wait 120us
    io 0x02f8-0x02ff [lines=10] [8bit] [range]
    irq mask 0x86b8 [level]

  offset 0x53, tuple 0x1b, link 0x07
    20 08 aa 60 f8 03 07
  cftable_entry 0x20
    io 0x03f8-0x03ff [lines=10] [8bit] [range]

  offset 0x5c, tuple 0x1b, link 0x07
    22 08 aa 60 e8 03 07
  cftable_entry 0x22
    io 0x03e8-0x03ef [lines=10] [8bit] [range]

  offset 0x65, tuple 0x1b, link 0x07
    23 08 aa 60 e8 02 07
  cftable_entry 0x23
    io 0x02e8-0x02ef [lines=10] [8bit] [range]

Any hints?

Thanks,
Martin

None Maybe a resource conflict

Re: Question Trouble with FM560CB modem card (Martin Buck)
Date: 2000, Feb 22
From: David Hinds <dhinds@pcmcia.sourceforge.org>

> cardmgr[1151]: unsupported card in socket 1
> cardmgr[1151]:   product info: "V90&K56Flex PCMCIA FAX MODEM", "", "", ""

That's a little bit disturbing, but not a big deal.

> kernel: serial_cs: register_serial() at 0x02f8, irq 3 failed

The simple thing to try would be to add:

  exclude port 0x2f8-0x2ff, irq 3

to /etc/pcmcia/config.opts and see if the card works better with a
different set of resources.

-- Dave

Sad Don't think so

Re: None Maybe a resource conflict (David Hinds)
Date: 2000, Feb 23
From: Martin Buck <mb@netwings.ch>

> That's a little bit disturbing, but not a big deal.

Well, what do you expect of cheap PC-hardware? :-|

> [...] and see if the card works better with a different
> set of resources.

I haven't tried your exclude statement, but I'm using a
network card in the same notebook (not plugged in when I did
the modem experiments) that normally gets IRQ 3 and works just
fine with it.

At some point, I also tried plugging in both cards at the same
time. Then the modem got IO 0xaf8 and IRQ 5, but it didn't
work there, either (same message from serial_cs, just with
different numbers). :-(

Martin

None More info

Re: None Maybe a resource conflict (David Hinds)
Date: 2000, Feb 24
From: Martin Buck <mb@netwings.ch>

I now tried the exclude statement you suggested, but that didn't help. As in the case when I used both the modem and the Ethernet card at the same time, the modem got 0xaf8/5 and didn't work. I also tried some other exclude statements to force the modem to other addresses like 0x2e8 etc., but I never got anything else than "autoconfig failed".

What does serial_cs do if autoconfig fails? Does it leave the card at the address it assigned before doing the autoconfig, or does it disable the card afterwards? Because I also tried using setserial to configure the card manually, but I only got I/O errors when trying to open the serial device. Of course, if serial_cs disables the card when autoconfig failes, this cannot work.

The parse errors from dump_cis still worry me. Could they cause these problems?

Martin

None Re: More info

Re: None More info (Martin Buck)
Date: 2000, Feb 24
From: David Hinds <dhinds@pcmcia.sourceforge.org>

If the autoconfig fails, then serial_cs un-configures the card.

The specific errors you get from dump_cis are not important.

-- Dave

Angry It's a WinModem

Re: None Re: More info (David Hinds)
Date: 2000, Feb 25
From: Martin Buck <mb@netwings.ch>

OK, I've got one more entry for the list of WinModems in your (un)supported devices list: FM560CB from Abocom

Abocom's webpage lists the FM560 as supported by Linux, but they don't mention the FM560CB anywhere, so I assumed they were the same. An e-mail from Abocom confirmed that they're not, because FM560 is a real modem and FM560CB is a WinModem.

Sorry for wasting your time, Martin

Sad Wasted time

Re: Angry It's a WinModem (Martin Buck)
Date: 2000, May 06
From: Massimiliano Perantoni maxper

:-(. Guess that's my same problem. I had an Hamlet modem that was working great, got broken and so I needed to get a substitute, but seems that PCMCIA has changed idea about modems and has produced the same modem with some sort of WinModem chipset: it gives the same tipology of errors:
unsupported card, once added to /etc/pcmcia/config.opts it Hi beeps but then it low beeps ending with an autoconfig failed as in this case. After that I tried to do many things, but nothing seems to work, and I think that it's one more modem to be added to the unsupported ones :-(.
Greetings Massimiliano

Sad CIS parse errors on Lucent modem

Date: 2000, Mar 02
From: Mark Neil holman

I have been trying to get an unbranded modem running on a Dell Latitude CPiA running Redhat 6.1 but without success.

OK so it doesn't say WinModem on the box but the logged messages (see below) point to it being so. However there is also a complaint from the drivers "parse error: Invalid argument" when reading the CIS. dump_cis gives similar errors.

I have tried upgrading from the standard 3.0.14 to 3.1.11 (in view of a brief mention of "Lucent Win Modems" in the CHANGES file) but the behaviour is the same. I have also messed with the /etc/pcmcia/config file in order to at least get it to try the serial module.

Does anyone have any idea as to what is going on, or whether I should cut my losses in view of the Winmodem bit

thanks

Mark Neil

ps there is also an ethernet card mentioned which works fine - the modem behaviour is the same whether or not the ethernet is inserted.

from /var/log/messages:

 Mar  2 14:56:50 localhost kernel: Linux PCMCIA Card Services 3.1.11 
 Mar  2 14:56:50 localhost kernel:   kernel build: 2.2.12-20 #1 Mon Sep 27 10:40:35 EDT 1999 
 Mar  2 14:56:50 localhost kernel:   options:  [pci] [cardbus] [apm] [pnp] 
 Mar  2 14:56:50 localhost kernel: PCI routing table version 1.0 at 0xfbda0 
 Mar  2 14:56:50 localhost kernel: PnP: PNP BIOS installation structure at 0xc00fe2d0 
 Mar  2 14:56:50 localhost kernel: PnP: PNP BIOS version 1.0, entry at f0000:e2f4, dseg at 40 
 Mar  2 14:56:50 localhost kernel: Intel PCIC probe:  
 Mar  2 14:56:50 localhost kernel:   TI 1225 PCI-to-CardBus at slot 00:03, mem 0x68000000 
 Mar  2 14:56:50 localhost kernel:     host opts [0]: [ring] [serial pci & irq] [pci irq 11] [lat 32/32] [bus 32/34] 
 Mar  2 14:56:50 localhost kernel:     host opts [1]: [ring] [serial pci & irq] [pci irq 11] [lat 32/32] [bus 35/37] 
 Mar  2 14:56:50 localhost kernel:     ISA irqs (scanned) = 3,9,10 PCI status changes 
 Mar  2 14:56:50 localhost cardmgr[5916]: starting, version is 3.1.11
 Mar  2 14:56:50 localhost cardmgr[5916]: watching 2 sockets
 Mar  2 14:56:50 localhost kernel: cs: IO port probe 0x1000-0x17ff: clean. 
 Mar  2 14:56:50 localhost kernel: cs: IO port probe 0x0100-0x04ff: clean. 
 Mar  2 14:56:50 localhost kernel: cs: IO port probe 0x0a00-0x0aff: clean. 
 Mar  2 14:56:50 localhost cardmgr[5916]: initializing socket 0
 Mar  2 14:56:50 localhost kernel: cs: memory probe 0xa0000000-0xa0ffffff: clean. 
 Mar  2 14:56:50 localhost cardmgr[5916]: error parsing CIS on socket 0: Invalid argument
 Mar  2 14:56:50 localhost cardmgr[5916]: socket 0: LT WIN MODEM GLOBAL
 Mar  2 14:56:50 localhost cardmgr[5916]: executing: 'insmod /lib/modules/2.2.12-20/pcmcia/serial_cs.o'
 Mar  2 14:56:51 localhost kernel: register_serial(): autoconfig failed 
 Mar  2 14:56:51 localhost kernel: serial_cs: register_serial() at 0x02f8, irq 3 failed 
 Mar  2 14:56:52 localhost cardmgr[5916]: get dev info on socket 0 failed: Resource temporarily unavailable
 Mar  2 14:56:52 localhost cardmgr[5916]: initializing socket 1
 Mar  2 14:56:52 localhost cardmgr[5916]: socket 1: NE2000 Compatible Ethernet
 Mar  2 14:56:52 localhost cardmgr[5916]: executing: 'insmod /lib/modules/2.2.12-20/net/8390.o'
 Mar  2 14:56:52 localhost cardmgr[5916]: executing: 'insmod /lib/modules/2.2.12-20/pcmcia/pcnet_cs.o'
 Mar  2 14:56:52 localhost kernel: eth0: NE2000 Compatible: io 0x300, irq 3, hw_addr 00:E0:98:1A:1E:16 
 Mar  2 14:56:52 localhost cardmgr[5916]: executing: './network start eth0'

None The CIS problem is not really important

Re: Sad CIS parse errors on Lucent modem (Mark Neil)
Date: 2000, Mar 02
From: David Hinds <dhinds@pcmcia.sourceforge.org>

As it seems to clearly be a WinModem, the fact that it also has some
sort of badness in the CIS is not too important.

-- Dave

Question How to configure Com1 V.90 MC221 Discovery on Sony Vaio F305 ?

Date: 2000, Jan 31
From: Axel Stockburger red

hi there, as newbie I definitively need help setting up 2 Pcmcia Cards running on a Sony Vaio F-305 under Suse Linux.

card 1 = Modemcard Com1 V.90 MC221 Discovery 56K (it is not supported by the package - but i think it should work if I could find out the right interrupts..)- but how ?

card 2 = Ethernetcard Argosy EN 235 10/100 ( I think it is supported, but cardctl and cardmanager always say there is nothing loaded...) also I don`t know what to write into the config file...

here is a boot.message printout:

ttyS00 at 0x03f8 (irq = 4) is a 16550A

ttyS02 at 0x03e8 (irq = 4) is a 16550A

Linux PCMCIA Card Services 3.1.3
  kernel build: 2.2.13 #1 Mon Nov 8 18:28:35 GMT 1999

  options:  [pci] [cardbus] [pnp]

PnP: PNP BIOS installation structure at 0xc00f6e00

PnP: PNP BIOS version 1.0, entry at f0000:b54f, dseg at 400

PCI routing table version 1.0 at 0xfdf60

00:0a.0 -> irq 9

Intel PCIC probe:

Ricoh RL5C478 PCI-to-CardBus at bus 0 slot 10, mem 0x68000000, 2 sockets

host opts [0]: [isa irq] [io 3/6/1] [mem 3/6/1] [pci irq 9] [lat 168/176] [bus 32/34]

host opts [1]: [isa irq] [io 3/6/1] [mem 3/6/1] [no pci irq] [lat 168/176] [bus 35/37]

ISA irqs (default) = 3,4,5,7,10,11,12 polling interval = 1000 ms cs: IO port probe 0x1000-0x17ff: clean.

cs: IO port probe 0x0100-0x04ff: clean.

cs: IO port probe 0x0a00-0x0aff: clean.

cs: cb_alloc(bus 32): vendor 0x10ec, device 0x8138

cs: cb_alloc(bus 35): vendor 0xffff, device 0x0440

hdc: packet command error: status=0x51 { DriveReady SeekComplete Error }

hdc: packet command error: error=0x50

ATAPI device hdc:
  Error: Illegal request -- (Sense key=0x05)

  Invalid field in command packet -- (asc=0x24, ascq=0x00)
  The failed "Start Stop Unit" packet command was: 
  "1b 00 00 00 03 00 00 00 00 00 00 00 "
  Error in command packet byte 256 bit 0
cdrom: open failed.
VFS: Disk change detected on device ide1(22,0)

to be honest, I don`t really know what it means... pleeeeez help me ... thanx in advance

red

None Neither card is supported

Re: Question How to configure Com1 V.90 MC221 Discovery on Sony Vaio F305 ? (Axel Stockburger)
Date: 2000, Jan 31
From: David Hinds <dhinds@pcmcia.sourceforge.org>

The Com1 MC221 Discovery modem is a Winmodem.

The Argosy EN235 is not on the Linux supported cards list.

> hdc: packet command error: status=0x51 { DriveReady SeekComplete Error }
> hdc: packet command error: error=0x50

This might be fixed by upgrading to the 3.1.9 PCMCIA driver package.

-- Dave

None Modem only works with irq 0

Date: 2000, Feb 03
From: Bennett Link blink

I recently upgraded from Redhat 5.2 to 6.1 on a Sony N505VX
notebook. The PCMCIA modem, which worked fine under RH 5.2, now works
only if I manually set the irq of /dev/ttyS1 to 0. I've verified that
no other irq works by excluding different irq values in 
/etc/pcmcia/config.opts. My pcmcia lan card works fine. 

I am using pcmcia card services 3.1.8, which I recompiled using the
defaults after installing the new kernel. According to the system log,
cardmgr is correctly identifying the card as a modem and loading 
the serial_cs module:

....

Feb  3 09:01:02 beatrice cardmgr[331]: initializing socket 0
Feb  3 09:01:02 beatrice kernel: cs: memory probe 0xa0000000-0xa0ffffff: clean. 
Feb  3 09:01:02 beatrice cardmgr[331]: socket 0: Serial or Modem
Feb  3 09:01:02 beatrice cardmgr[331]: module /lib/modules/preferred/misc/serial.o not available
Feb  3 09:01:02 beatrice cardmgr[331]: executing: 'insmod /lib/modules/preferred/pcmcia/serial_cs.o'
Feb  3 09:01:02 beatrice kernel: tty01 at 0x13f8 (irq = 4) is a 16550A 
Feb  3 09:01:02 beatrice cardmgr[331]: executing: './serial start ttyS1'

....

lsmod gives:


Module                  Size  Used by
serial_cs               5168   0  (unused)
ds                      6472   1  [serial_cs]
i82365                 29536   1 
pcmcia_core            43936   0  [serial_cs ds i82365]
opl3                   11208   0  (unused)
sb                     33620   0 
uart401                 5968   0  [sb]
sound                  57240   0  [opl3 sb uart401]
soundlow                 300   0  [sound]
soundcore               2372   6  [sb sound]


And setserial on /dev/modem gives:

/dev/modem, UART: 16550A, Port: 0x13f8, IRQ: 4


/proc/interrupts shows:

           CPU0       
  0:     366286          XT-PIC  timer
  1:       4193          XT-PIC  keyboard
  2:          0          XT-PIC  cascade
  5:          1          XT-PIC  soundblaster
  8:          1          XT-PIC  rtc
 12:      61070          XT-PIC  PS/2 Mouse
 13:          1          XT-PIC  fpu
 14:     131017          XT-PIC  ide0
NMI:          0


Thanks for any help, 

Bennett
blink@dante.physics.montana.edu

None There is a workaround in the 3.1.10 driver package

Re: None Modem only works with irq 0 (Bennett Link)
Date: 2000, Feb 03
From: David Hinds <dhinds@pcmcia.sourceforge.org>

It is a kernel bug, but there's a workaround in the 3.1.10 driver
package.

You really only need to get the 3.1.10 drivers and copy etc/serial to
/etc/pcmcia/serial.

-- Dave

Question Modem IO error

Date: 2000, Apr 16
From: Pawel Konieczny konieczp

This is a cross-post from comp.os.linux.portable (sorry David, I couldn't help it, maybe somebody here recognises the problem):

The card is properly recognised by the cardmgr, which assigns ttyS1 to it, but when I try to use /dev/ttyS1 I get 
an "input/output error" (if root) or "Operation not allowed for device" (if user).

Does anybody recognise the problem? Do you know the solution?
It seems a configuration problem to me, rather than a PCMCIA problem, but I really don't know. I am new to Linux (two weeks).

Pawel

Here are some details for those of you who are ready to help me:

Laptop:
DELL Latitude CPi

System:
Debian GNU/Linux 2.1 (try Linux because of the free software idea, not ease of use)
Linux kernel 2.2.12

PCMCIA:
original:
ii  pcmcia-cs       3.0.14-0slink1 PCMCIA Card Services for Linux.
upgraded to: 3.1.8

Tried two different cards:
RBEM56G-100 (Xircom RealPort CardBus Ethernet 10/100+Modem 56)
CM33 (Xircom GlobalAccess CreditCard Modem 33.6)
-- both recognised and configured smoothly by cardmgr

Here is the contents of /var/state/pcmcia/stab for the first card:
Socket 0: empty
Socket 1: Xircom CardBus 10/100 Ethernet + 56K Modem
1       network tulip_cb   0 eth0
1       serial  serial_cb  0 ttyS1 4 65

Here is the contents of /var/state/pcmcia/stab for the second card:
Socket 0: Serial or Modem
0       serial  serial_cs  0 ttyS1 4 65
Socket 1: empty

$ ls -l /dev/ttyS1
crw-rw-rw-   1 root     dialout    4,  65 Sep  9  1999 /dev/ttyS1


Things I tied (without success)
* read all serial-port or PCMCIA related HOWTOS
* scan for existing posts to newsgroups on this subject
* laptop istallation tips published on the web (selectively)
* disabled infrared ports in BIOS
* upgraded BIOS to the latest (two months old)
* upgraded pcmcia package to the latest stable release 3.1.8
* inserted and removed the card several times (recognition OK)
* tried an external modem on ttyS0 - works fine
* ifconfig eth0 down
* modified /etc/pcmcia/config not to load tulip_cb for
  RBEM56G-100(tulip_cb is tagged as "unreliable" in PCMCIA 
  HOWTO, but funny enough, my eth0 works fine with it!)
* Upgraded to PCMCIA 3.1.14
* Removed /dev/ttyS1 - cardmgr recreated it at insertion

None Fixed in 3.1.14

Re: Question Modem IO error (Pawel Konieczny)
Date: 2000, Apr 19
From: David Hinds <dhinds@pcmcia.sourceforge.org>

This is a kernel bug affecting serial devices configured for non-standard IO addresses... the 3.1.14 drivers include a workaround that avoids the issue.

-- Dave

Ok It is, indeed!

Re: None Fixed in 3.1.14 (David Hinds)
Date: 2000, Apr 25
From: Pawel Konieczny konieczp

After proper installing of 3.1.14, the problem is gone.

Question Hawking pcmcia installation trouble

Date: 2000, Apr 21
From: ed bailey erb

Hi,

I am trying to install a Hawking PN610 pcmcia modem on a Winbook XL2 running SUSE 6.4 with serial support compiled into the kernel. When I first inserted the card, the cardmanger listed the card as unsupported at bootup. Using cardctl ident I wrote a new config entry into /etc/pcmcia/config

card "Hawking V.90"
   manfid 0x018a, 0x0001
   bind "serial_cs"

and then rebooted

Then the cardmanger recognized my entry and executed and then gave the following error message - "probing for dev info failed - resource temporaily unavailible."

When I run dmesg I get the following

register_serial(): autoconfig failed serial_cs: register_serial() at 0x02f8, irq failed

Does any body have any ideas about what I should do - I have been messing with this for some time. Thank you in advance for any help you can provide.

Ed

None register_serial(): autoconfig failed

Re: Question Hawking pcmcia installation trouble (ed bailey)
Date: 2000, Apr 21
From: David Hinds <dhinds@pcmcia.sourceforge.org>

> 
> register_serial(): autoconfig failed
> serial_cs: register_serial() at 0x02f8, irq failed

If you're lucky, this simply means that there is some other device
interfering with configuration of this card at this IO port address
and/or this interrupt.  Try excluding these settings in
/etc/pcmcia/config.opts and see if that fixes the problem.

I checked Hawking's web site and it doesn't sound like this is a
WinModem, though they do claim only Win95/98/NT support.

-- Dave

Sad Probably a WinModem

Re: Question Hawking pcmcia installation trouble (ed bailey)
Date: 2000, Apr 21
From: David Hinds <dhinds@pcmcia.sourceforge.org>

The vendor apparently says it is "sort of" a WinModem.

Oh well.

-- Dave

Question error parsing CIS (Hamlet modem)

Date: 2000, May 18
From: Aram Gurekian aram

Hi all...
I'm a newbie in linux and hardware, so please be patient if my questions are silly! :)
I have an Hamlet FM560 56K - V.90 PCMCIA FAX MODEM.
It works fine on my windoz98 partition, but I can't make it work with linux! :(
I'm pretty sure that it isn't a winmodem (at least, I can't read 'winmodem' in any part of the documentation! :) )
I'd like to get rid of this 1GB vfat partition, but I can't until I can connect to the internet without windows!
I spent a few hours reading documentation under /usr/doc, but now I think that at this point I need some help...

I have also a PCMCIA Ethernet Card from Hamlet and it works fine with both windows and linux in the other pcmcia slot.

I still can't understand if my card is supported or not. I wouldn't be able (at the moment) to write a new driver, so I'm probably going to buy a supported modem (even an external one)... it would probably have been more convenient to do that from the beginning, but I wanted to try... in fact I learnt some interesting things, but now I'm done! :P

... I guess every answer is in the "Error parsing CIS on socket 1: Invalid argument

I tried to report as many information about my system (maybe too many?) as I could.

Any help or explanation would be appreciated.
Thank you in advance!

Ciao

  Aram Gurekian

Hardware
--------
Texas Instruments extensa 710DX
pentium 233 MMX
it has an internal modem that died some time ago :)

uname -a
--------
Linux spiff.weirdo 2.2.12-20 #1 Mon Sep 27 10:40:35 EDT 1999 i686 unknown

I use RedHat 6.1 and card services that came along with it

windows 98
----------

PCMCIA FAX MODEM 56K
HAMLET FM560 56K-V.90
PORT:COM3
IO: 0x03e8-0x03ef
interrupt: 10 (or sometimes 9)

CONTROLLER CARDBUS T.I. PCI-1250
interrupt:11
memory interval: 0x8c001000-0x8c001fff


/etc/rc.d/init.d/pcmcia stop - pcmcia start:
--------------------------------------------

May 18 17:21:57 spiff cardmgr[1888]: executing: './network check eth0'
May 18 17:21:57 spiff cardmgr[1888]: shutting down socket 0
May 18 17:21:57 spiff cardmgr[1888]: executing: './network stop eth0'
May 18 17:21:57 spiff cardmgr[1888]: executing: 'rmmod pcnet_cs'
May 18 17:21:57 spiff cardmgr[1888]: executing: 'rmmod 8390'
May 18 17:21:57 spiff cardmgr[1888]: shutting down socket 1
May 18 17:21:57 spiff cardmgr[1888]: executing: 'rmmod serial_cs'
May 18 17:21:57 spiff cardmgr[1888]: exiting
May 18 17:21:59 spiff kernel: unloading PCMCIA Card Services
May 18 17:22:01 spiff kernel: Linux PCMCIA Card Services 3.0.14
May 18 17:22:01 spiff kernel: kernel build: 2.2.12-20 #1 Mon Sep 27 10:40:35 E
DT 1999
May 18 17:22:01 spiff kernel: options: [pci] [cardbus] [apm]
May 18 17:22:01 spiff kernel: Intel PCIC probe:
May 18 17:22:01 spiff kernel: TI 1250A PCI-to-CardBus at bus 0 slot 6, mem 0x6
8000000, 2 sockets
May 18 17:22:01 spiff kernel: host opts [0]: [pwr save] [serial pci & irq] [
no pci irq] [lat 168/176] [bus 32/34]
May 18 17:22:01 spiff kernel: host opts [1]: [pwr save] [serial pci & irq] [
no pci irq] [lat 168/176] [bus 35/37]
May 18 17:22:01 spiff kernel: ISA irqs (scanned) = 3,4,7,9,10 status changeon irq 10
May 18 17:22:01 spiff cardmgr[2013]: starting, version is 3.0.14
May 18 17:22:01 spiff cardmgr[2013]: watching 2 sockets
                                                                              
May 18 17:22:01 spiff kernel: cs: IO port probe 0x1000-0x17ff: excluding 0x1000-
0x104f
May 18 17:22:01 spiff kernel: cs: IO port probe 0x0100-0x04ff: excluding 0x378-0
May 18 17:22:01 spiff cardmgr[2013]: initializing socket 0
May 18 17:22:01 spiff kernel: cs: memory probe 0xa0000000-0xa0ffffff: clean.
May 18 17:22:01 spiff cardmgr[2013]: socket 0: NE2000 Compatible Ethernet
May 18 17:22:02 spiff cardmgr[2013]: executing: 'insmod /lib/modules/2.2.12-20/net/8390.o'
May 18 17:22:02 spiff cardmgr[2013]: executing: 'insmod /lib/modules/2.2.12-20/pcmcia/pcnet_cs.o'
May 18 17:22:02 spiff kernel: eth0: NE2000 Compatible: io 0x300, irq 3, hw_addr
00:E0:98:19:2C:79
May 18 17:22:02 spiff cardmgr[2013]: executing: './network start eth0'
May 18 17:22:02 spiff cardmgr[2013]: initializing socket 1
May 18 17:22:02 spiff cardmgr[2013]: error parsing CIS on socket 1: Invalid argument
May 18 17:22:02 spiff cardmgr[2013]: socket 1: V90&K56Flex PCMCIA FAX MODEM
May 18 17:22:02 spiff cardmgr[2013]: executing: 'insmod /lib/modules/2.2.12-20/pcmcia/serial_cs.o irq_list=9,10'
May 18 17:22:03 spiff kernel: register_serial(): autoconfig failed
May 18 17:22:03 spiff kernel: serial_cs: register_serial() at 0x03e8, irq 9 failed
May 18 17:22:04 spiff cardmgr[2013]: get dev info on socket 1 failed: Resource temporarily unavailable



cardctl ident:
--------------

Socket 0:
  product info: "Ethernet", "Adapter", "2.0"
  manfid: 0x0149, 0xc1ab
  function: 6 (network)
Socket 1:
  product info: "V90&K56Flex PCMCIA FAX MODEM", "", "", ""



/etc/pcmcia/config
------------------

device "serial_cs"
  class "serial" module "serial_cs"
# tried even with -> opts "irq_list=9,10"

# I tried to add this lines to the file:
card "V90&K56Flex PCMCIA FAX MODEM"
  version "V90&K56Flex PCMCIA FAX MODEM"
  bind "serial_cs"


/etc/pcmcia/config.opts
-----------------------
#
# Local PCMCIA Configuration File
#
# System resources available for PCMCIA devices
#
include port 0x100-0x4ff, port 0x1000-0x17ff
# tried even to add those three rows (data taken from windows behaviour)
#include port 0x3e8-0x3ef
#include irq 11, irq 9
#include memory 0x8c001000-8c001fff
# end
include memory 0xc0000-0xfffff
include memory 0xa0000000-0xa0ffffff, memory 0x60000000-0x60ffffff
#
# Extra port range for IBM Token Ring
#
include port 0xa00-0xaff
#
# Resources we should not use, even if they appear to be available
#
# First built-in serial port
exclude irq 4
# Second built-in serial port
#exclude irq 3
# First built-in parallel port
exclude irq 7


/etc/sysconfig/pcmcia
---------------------

PCMCIA=yes
PCIC=i82365
PCIC_OPTS=
CORE_OPTS=


/proc/interrupts
----------------

           CPU0
  0: 556368 XT-PIC timer
  1: 14633 XT-PIC keyboard
  2: 0 XT-PIC cascade
  3: 62 XT-PIC pcnet_cs
  8: 1 XT-PIC rtc
 10: 69 XT-PIC i82365
 12: 82085 XT-PIC PS/2 Mouse
 13: 1 XT-PIC fpu
 14: 467685 XT-PIC ide0
 15: 47011 XT-PIC ide1
NMI: 0


/proc/ioports
-------------

0000-001f : dma1
0020-003f : pic1
0040-005f : timer
0060-006f : keyboard
0070-007f : rtc
0080-008f : dma page reg
00a0-00bf : pic2
00c0-00df : dma2
00f0-00ff : fpu
0170-0177 : ide1
01f0-01f7 : ide0
02f8-02ff : serial(auto)
0300-031f : pcnet_cs
0376-0376 : ide1
03c0-03df : vga+
03f6-03f6 : ide0
03f8-03ff : serial(auto)
fcd0-fcd7 : ide0
fcd8-fcdf : ide1

None The CIS error is harmless, not the problem

Re: Question error parsing CIS (Hamlet modem) (Aram Gurekian)
Date: 2000, May 23
From: David Hinds <dhinds@pcmcia.sourceforge.org>

I can't tell for sure what's going on here.  Certain modems seem to
fail some of the sanity checks in the linux kernel's serial device
configuration code.  Maybe that is the case here.

-- Dave

Question Nokia CardPhone 1.0 Problem

Date: 2000, May 18
From: Michael D. md

Hi everyone,

I`m trying to install a Nokia CardPhone 1.0 on a SuSE 6.2 System (Kernel 2.2.10) with the pcmcia package 3.1.14 newly compiled. Mashine is a Toshiba Satellite 2550CDT.

I managed to configure my D-Link 650TX Ethernet Card thanks to your lists here but I`m stuck with the Nokia card.

Here the output:

Serial driver version 4.27 with SHARE_IRQ enabled
ttyS00 at 0x03f8 (irq = 4) is a 16550A
Linux PCMCIA Card Services 3.1.14
  kernel build: 2.2.10 #7 Sun Aug 15 20:23:47 MEST 1999
  options:  [pci] [cardbus] [apm]
PCI routing table version 1.0 at 0xf0180
Intel PCIC probe: 
  Intel i82365sl B step rev 00 ISA-to-PCMCIA at port 0x3e0 ofs 0x00
    host opts [0]: none
    host opts [1]: none
    ISA irqs (scanned) = 3,4,5,7,10 polling interval = 1000 ms
cs: IO port probe 0x1000-0x17ff: clean.
cs: IO port probe 0x0100-0x04ff: excluding 0x378-0x37f 0x4d0-0x4d7
cs: IO port probe 0x0a00-0x0aff: clean.
cs: memory probe 0x0d0000-0x0dffff: clean.
eth0: NE2000 Compatible: io 0x300, irq 3, hw_addr 00:E0:98:77:95:07
tty01 at 0x0100 (irq = 5) is a 16450
register_serial(): autoconfig failed
serial_cs: register_serial() at 0x0108, irq 5 failed

What is the problem here? What means the last line (why 0x0108 and not 0x0100???) What can I try?

Thank you for every hint!

	Michael

None There is a specific Nokia driver for linux

Re: Question Nokia CardPhone 1.0 Problem (Michael D.)
Date: 2000, May 18
From: David Hinds <dhinds@pcmcia.sourceforge.org>

I think you need to get it from Nokia's web site.

The PCMCIA subsystem is trying to load the generic serial card driver,
and it is getting confused.

-- Dave

None Yes but it`s for pcmcia 2.9.12...

Re: None There is a specific Nokia driver for linux (David Hinds)
Date: 2000, May 19
From: Michael D. md

Hi Dave,

thank you for your quick answer!

I already downloaded the driver but it`s rather old (March 1998) and was developed with pcmcia version 2.9.12 and doesn`t compile. I don`t want to switch back to this old version.

I thought that because the driver is so old its now obsolete and the nokia is supported by the standard serial_cs???

I read in DejaNews, that the card is working under Linux but no details how.

Any hints?

Thank you

	Michael

None Re: Yes but it`s for pcmcia 2.9.12...

Re: None Yes but it`s for pcmcia 2.9.12... (Michael D.)
Date: 2000, May 19
From: David Hinds <dhinds@pcmcia.sourceforge.org>

I think you'll need to try to follow this up with Nokia; I can't
realistically provide support for cards I don't have, with drivers
written by the card vendor.

-- Dave

Question modem looks like it is working but I cannot communicate with it

Date: 2000, May 19
From: Randy Smith randysmith

I have a TI cardbus controller and a modem card. Here are the
log files...

pcmcia start

Linux PCMCIA Card Services 3.1.15
kernel build: 2.2.15 #1 Thu May 11 16:19:36 EDT 2000
options:  [pci] [cardbus] [pnp]
PCI routing table version 1.0 at 0xf82b0
PnP: PNP BIOS installation structure at 0xc00f7bd0
PnP: PNP BIOS version 1.0, entry at f0000:7344, dseg at f0000
Intel PCIC probe:
TI 1221 rev 00 PCI-to-CardBus at slot 00:10, mem 0x68000000
host opts [0]: [pci only] [pci irq 10] [lat 168/176] [bus 32/34]
host opts [1]: [pci only] [pci irq 10] [lat 168/176] [bus 35/37]
PCI card interrupts, PCI status changes
cardmgr[634]: starting, version is 3.1.15
cardmgr[634]: watching 2 sockets
cs: IO port probe 0x0c00-0x0cff: clean.
cs: IO port probe 0x0800-0x08ff: clean.
cs: IO port probe 0x0100-0x04ff: excluding 0x110-0x117
cs: IO port probe 0x0a00-0x0aff: clean.

insert the modem card:
(two high beeps heard)

cardmgr[634]: initializing socket 0
cardmgr[634]: socket 0: Cirrus Logic 14.4K Fax Modem
cs: memory probe 0xa0000000-0xa0ffffff: clean.
cardmgr[634]: executing: 'modprobe serial_cs'
tty02 at 0x0af8 (irq = 10) is a 16550A
cardmgr[634]: executing: './serial start ttyS2'

At this point it looks like it should be working, but I cannot 
communicate with the card.

randys:/etc/pcmcia # ditty /dev/ttyS2
TCGETA: I/O error

Now, I also have another pcmcia interface that plugs into the
ISA bus. This modem works with that interface just fine. The
above "ditty" will show the current config status of it.

Any suggestions?

Randy Smith

None A bug; fixed in the 19-May-00 beta drivers

Re: Question modem looks like it is working but I cannot communicate with it (Randy Smith)
Date: 2000, May 19
From: David Hinds <dhinds@pcmcia.sourceforge.org>

It is a bug in serial_cs; in this situation, it doesn't tell the
serial driver that the irq might need to be shared.  I guess this
situation (using a PCMCIA modem in a PCI card dock) hasn't come up
before.  Try with the 19-May-00 beta drivers on sourceforge.org in
/pcmcia/NEW.

-- Dave

More That fixed the modem

Re: None A bug; fixed in the 19-May-00 beta drivers (David Hinds)
Date: 2000, May 19
From: Randy Smith randysmith

Thanks! That fixed the modem card's problem.

The compact flash still won't work though. I have read everything
on the list regarding these things. I tried the two kernel patches
that were posted and no luck.
I am wondering now, if there is a similar problem with that ide module?

Thanks again!

Question Kingmax PCMCIA modem

Date: 2000, May 26
From: Grady Luffman Grady

...trying to get a Kingmax PCMCIA modem card to work on a laptop running RH Linux 6.1. Here's the dump:

The card is listed in the Linux Modem Compatibility Index (with no defining information):
         Kingmax Technology, MGA-FM336C-KF.  (not supported?)

The socket controller is the Cirrus Logic PD6729 PCI-to-PCMCIA bridge, located at bus 0 slot 24, port 0x3000, with 2 sockets:

1. The PCIC probe is picking up the controller correctly (as "Cirrus PD6729 PCI-to-PCMCIA bridge...) 2. /proc/pci shows device at bus 0, slot 24, I/O at 0x3000 [0x3001], slow devsel 3. /etc/sysconfig/hwconf shows:

          class:    other
            bus:    PCI
       detached:    0
        driver :    Card: Cirrus Logic GD6729
           desc:    "Cirrus Logic | CL6729"
       vendorID:    1013
       deviceID:    1100

4. 'lsmod' shows the i82365 and pcmcia_core modules loaded 5. /proc/ioports shows i82365 at 0x3000-0x3001

The controller appears to be functioning properly. The problem I think, is with the modem card.

Initially, when I first ran the startup script, 'cardmgr' mis-identified the modem card as an SVEC 10/100 Fast Ethernet Card which is bound to 'tulip_cb' (which caused memory problems). Since then, I've changed the entry in /pcmcia/config several times in an effort to get the card identified (without success). But at least now the cardmgr is beeping when you insert/remove the card. (Low beep on insertion). Funny though, different entries in the config file have resulted in 'cardctl' showing different 'manfid' numbers. Should the 'manfid' entries be the same as the 'vendor ID' and 'device ID' shown in 'hwconf'?

According to all I've read, I want to believe the card should work! I think I just need the correct entry to define the card in /pcmcia/config.

Any ideas? Thanks!

Note solved

Re: Question Kingmax PCMCIA modem (Grady Luffman)
Date: 2000, May 26
From: Grady Luffman Grady

Finally! I got the cardmgr to ID the card by using the simplest definition I could muster in /pcmcia/config:

   card "Serial port device"
      function serial_port
      bind "serial_cs"

Jeez! Now... on to the CompactFlash card!

Question PCMCIA modem unresponsive

Date: 2000, May 17
From: Johan Mellberg mellberg

Hi!

I'm having some trouble setting up pcmcia (I have only one modem so I don't now if it might be related to that specifically) - the modem just does not seem to exist. However it works just fine under Windows95 on the same machine (using it now). I know people are using pcmcia on this machine with Linux so it should not be impossible!

So what happens? When cardmgr starts up I hear two high beeps and I can just see the LEDs flash briefly. The card seems properly configured (see below). However, minicom says it is initialising the modem (/dev/ttyS1 - I have tried changing this with no change in behaviour) but nothing happens, seyon and kppp both complain that there is no modem. If I 'setserial /dev/modem irq 0' kppp _seems_ to work but the modem is completely inactive. Same thing if I do it with /dev/ttyS1. I can achieve the same if I exclude irq 9 and 11 in config.opts.

Any information on what I need to do to make this work would be appreciated - and it should be noted that I do not understand what the significance (if any?) of the I/O range info from windows is in relation to the settings in config.opts (see below).

Please find the following output enclosed:
- specifications/setup
- windows95 info
- cardctl config
- cardctl ident
- config.opts
- 'ls -l /dev/modem' and 'ls -l /dev/ttyS1'
- /var/state/pcmcia/stab
- relevant parts of /var/log/messages

* Specs
Omnibook 800CT (166MMX), 80MB RAM
BIOS is set to 'PCIC-compatible', other options are 'Disable' and 'Cardbus'
Slackware7/Windows95 dualboot with LILO
Kernel updated to 2.2.15
pcmcia 3.1.14 installed from scratch

* Windows95 pcmcia card/socket info
Socket input/output range: 03E0-03E1
pcmcia modem on COM2, irq3, I/O range: 02F8-02FF

* cardctl config
================
Socket 0:
  Vcc 5.0V Vpp1 0.0V Vpp2 0.0V
  interface type is "memory and I/O"
  irq 3 [exclusive] [level]
  Speaker output is enabled
  function 0:
    config base 0xff80
      option 0x57 status 0x08 pin 0x00
    io 0x02f8-0x02ff [8bit]
Socket 1:
  not configured

* cardctl ident
===============
Socket 0:
  product info: "Ericsson", "K56CC"
  manfid: 0x0200, 0x0001
  function: 2 (serial)
Socket 1:
  no product info available
Socket 0: Serial or Modem
0 serial serial_cs 0 ttyS1 4 65
Socket 1: empty

* devices
=========
lrwxrwxrwx 1 root root 10 May 17 09:49 /dev/modem -> /dev/ttyS1
crw-rw---- 1 root tty 4, 65 May 16 23:57 /dev/ttyS1


* /var/log/messages
===================
May 17 09:49:21 violator kernel: Linux PCMCIA Card Services 3.1.14
May 17 09:49:21 violator kernel: kernel build: 2.2.15 #1 Tue May 9 18:13:58 CEST 2000
May 17 09:49:21 violator kernel: options: [pci] [cardbus] [apm] [pnp]
May 17 09:49:21 violator kernel: PCI routing table version 1.0 at 0xf36e0
May 17 09:49:21 violator kernel: unknown PCI interrupt router 1004:0102
May 17 09:49:21 violator kernel: PnP: PNP BIOS installation structure at 0xc00ff000
May 17 09:49:21 violator kernel: PnP: PNP BIOS version 1.0, entry at e8000:33e1, dseg at e8000
May 17 09:49:21 violator kernel: Intel PCIC probe:
May 17 09:49:21 violator kernel: TI 1131 rev 01 PCI-to-CardBus at slot 00:04, mem 0x68000000
May 17 09:49:21 violator kernel: host opts [0]: [clkrun irq 10] [serial pci & irq] [no pci irq] [lat 64/32] [bus 32/34]
May 17 09:49:21 violator kernel: host opts [1]: [clkrun irq 10] [serial pci & irq] [no pci irq] [lat 64/32] [bus 35/37]
May 17 09:49:21 violator kernel: ISA irqs (scanned) = 3,9,11,15 status change on irq 15
May 17 09:49:21 violator cardmgr[143]: starting, version is 3.1.14
May 17 09:49:21 violator cardmgr[143]: watching 2 sockets
May 17 09:49:21 violator kernel: cs: IO port probe 0x0c00-0x0cff: clean.
May 17 09:49:21 violator kernel: cs: IO port probe 0x0800-0x08ff: clean.
May 17 09:49:21 violator kernel: cs: IO port probe 0x0100-0x04ff: clean.
May 17 09:49:22 violator cardmgr[143]: initializing socket 0
May 17 09:49:22 violator kernel: cs: memory probe 0xa0000000-0xa0ffffff: clean.
May 17 09:49:22 violator cardmgr[143]: socket 0: Serial or Modem
May 17 09:49:22 violator cardmgr[143]: executing: 'insmod /lib/modules/2.2.15/pcmcia/serial_cs.o'
May 17 09:49:22 violator kernel: tty01 at 0x02f8 (irq = 3) is a 16550A
May 17 09:49:22 violator cardmgr[143]: executing: './serial start ttyS1'


* config.opts
=============
#
# Local PCMCIA Configuration File
#
#----------------------------------------------------------------------

# System resources available for PCMCIA devices

include port 0x100-0x4ff, port 0x800-0x8ff, port 0xc00-0xcff
include memory 0xc0000-0xfffff
include memory 0xa0000000-0xa0ffffff, memory 0x60000000-0x60ffffff

# High port numbers do not always work...
#include port 0x1000-0x17ff

# Extra port range for IBM Token Ring
#include port 0xa00-0xaff

# Resources we should not use, even if they appear to be available

# First built-in serial port
exclude irq 4
# Second built-in serial port
#exclude irq 3
# First built-in parallel port
exclude irq 7

Question socketcom serial adaptor with ricochet modem

Date: 2000, May 31
From: Cengiz Alaettinoglu cengiz

Hi,

I am having difficulty with Socketcom serial adaptor S-I/O card when I use it with my ricochet cellular modem. The card is recognized, ttyS1 is configured, and a ppp connection is established, however pppd fails to configure the interface with an IP address. My ricochet setup works with the builtin serial port, i.e. ttyS0 (unfortunately this requires me to carry a port replicator).

I am including some diagnostics information. Did anyone seen similar problems or have any suggestions for me to try?

Thank you

Cengiz

# tail -f /var/log/messages

May 30 20:56:00 elf cardmgr[1995]: socket 0: Serial or Modem
May 30 20:56:00 elf cardmgr[1995]: executing: 'insmod /lib/modules/2.2.14/pcmcia/serial_cs.o'
May 30 20:56:00 elf kernel: tty01 at 0x02f8 (irq = 3) is a ST16650V2 
May 30 20:56:00 elf cardmgr[1995]: executing: './serial start ttyS1'

# cardctl ident
Socket 0:
  product info: "Socket", "Serial Port Card Rev 2.4"
  manfid: 0x0104, 0x0001
  function: 2 (serial)

# more /var/state/pcmcia/stab

Socket 0: Serial or Modem 0 serial serial_cs 0 ttyS1 4 65

#ifup ppp0
#tail -f /var/log/messages (on another terminal in parallel)
May 30 20:58:06 elf ifup-ppp: pppd started for ppp0 on /dev/ttyS1 at 38400
May 30 20:58:06 elf kernel: registered device ppp0 
May 30 20:58:06 elf pppd[2127]: pppd 2.3.11 started by root, uid 0
May 30 20:58:07 elf chat[2137]: abort on (BUSY)
May 30 20:58:07 elf chat[2137]: abort on (ERROR)
May 30 20:58:07 elf chat[2137]: abort on (NO CARRIER)
May 30 20:58:07 elf chat[2137]: abort on (NO DIALTONE)
May 30 20:58:07 elf chat[2137]: abort on (Invalid Login)
May 30 20:58:07 elf chat[2137]: abort on (Login incorrect)
May 30 20:58:07 elf chat[2137]: send (ATZ^M)
May 30 20:58:07 elf chat[2137]: expect (OK)
May 30 20:58:07 elf chat[2137]: ATZ^M^M
May 30 20:58:07 elf chat[2137]: OK
May 30 20:58:07 elf chat[2137]:  -- got it 
May 30 20:58:07 elf chat[2137]: send (ATDT777^M)
May 30 20:58:07 elf chat[2137]: expect (CONNECT)
May 30 20:58:07 elf chat[2137]: ^M
May 30 20:58:09 elf chat[2137]: ATDT777^M^M
May 30 20:58:09 elf chat[2137]: CONNECT
May 30 20:58:09 elf chat[2137]:  -- got it 
May 30 20:58:09 elf chat[2137]: send (^M)
May 30 20:58:09 elf pppd[2127]: Serial connection established.
May 30 20:58:09 elf pppd[2127]: Using interface ppp0
May 30 20:58:09 elf pppd[2127]: Connect: ppp0 <--> /dev/ttyS1

I tried to debug pppd manually. % minicom Welcome to minicom 1.83.0

OPTIONS: History Buffer, F-key Macros, Search History Buffer, I18n Compiled on Mar 7 2000, 06:12:31.

Press CTRL-A Z for help on special keys

AT S7=45 S0=0 L1 OK

atdt 777 CONNECT 38400 ~}#.!}!}"} }8}!}$}%}"}&} } } } }%}&..j}^}'}"}(}"/~~}#.!}!}#} }8}!}$}%}"}&}~

#pppd -d -detach /dev/ttyS1 38400 mru 1500 Using interface ppp0 Connect: ppp0 <--> /dev/ttyS1 sent [LCP ConfReq id=0x1 <asyncmap 0x0> <magic 0x1bd3379d> <pcomp> <accomp>] rcvd [LCP ConfReq id=0x6 <mru 1500> <asyncmap 0x0> <magic 0xb4886a7e> <pcomp> <a ccomp>] sent [LCP ConfAck id=0x6 <mru 1500> <asyncmap 0x0> <magic 0xb4886a7e> <pcomp> <a ccomp>] sent [LCP ConfReq id=0x1 <asyncmap 0x0> <magic 0x1bd3379d> <pcomp> <accomp>] rcvd [LCP ConfReq id=0x7 <mru 1500> <asyncmap 0x0> <magic 0xb4886a7e> <pcomp> <a ccomp>] sent [LCP ConfAck id=0x7 <mru 1500> <asyncmap 0x0> <magic 0xb4886a7e> <pcomp> <a ccomp>] sent [LCP ConfReq id=0x1 <asyncmap 0x0> <magic 0x1bd3379d> <pcomp> <accomp>] rcvd [LCP ConfReq id=0x8 <mru 1500> <asyncmap 0x0> <magic 0xb4886a7e> <pcomp> <a ccomp>] sent [LCP ConfAck id=0x8 <mru 1500> <asyncmap 0x0> <magic 0xb4886a7e> <pcomp> <a ccomp>] sent [LCP ConfReq id=0x1 <asyncmap 0x0> <magic 0x1bd3379d> <pcomp> <accomp>] rcvd [LCP ConfReq id=0x9 <mru 1500> <asyncmap 0x0> <magic 0xb4886a7e> <pcomp> <a ccomp>] sent [LCP ConfAck id=0x9 <mru 1500> <asyncmap 0x0> <magic 0xb4886a7e> <pcomp> <a ccomp>] sent [LCP ConfReq id=0x1 <asyncmap 0x0> <magic 0x1bd3379d> <pcomp> <accomp>] rcvd [LCP ConfReq id=0xa <mru 1500> <asyncmap 0x0> <magic 0xb4886a7e> <pcomp> <accomp>] sent [LCP ConfAck id=0xa <mru 1500> <asyncmap 0x0> <magic 0xb4886a7e> <pcomp> <accomp>] sent [LCP ConfReq id=0x1 <asyncmap 0x0> <magic 0x1bd3379d> <pcomp> <accomp>] rcvd [LCP ConfReq id=0xb <mru 1500> <asyncmap 0x0> <magic 0xb4886a7e> <pcomp> <accomp>] sent [LCP ConfAck id=0xb <mru 1500> <asyncmap 0x0> <magic 0xb4886a7e> <pcomp> <accomp>] sent [LCP ConfReq id=0x1 <asyncmap 0x0> <magic 0x1bd3379d> <pcomp> <accomp>] sent [LCP ConfReq id=0x1 <asyncmap 0x0> <magic 0x1bd3379d> <pcomp> <accomp>] sent [LCP ConfReq id=0x1 <asyncmap 0x0> <magic 0x1bd3379d> <pcomp> <accomp>] sent [LCP ConfReq id=0x1 <asyncmap 0x0> <magic 0x1bd3379d> <pcomp> <accomp>] LCP: timeout sending Config-Requests Connection terminated.

None It's strange that the connection starts up ok

Re: Question socketcom serial adaptor with ricochet modem (Cengiz Alaettinoglu)
Date: 2000, Jun 01
From: David Hinds <dhinds@pcmcia.sourceforge.org>

Normally, a serial device will either work or not; I haven't seen many
cases where a connection is established but then gets dropped like
this.  My first idea would be a PPP setup problem, but since the
built-in port works, that blows that idea.

A few things to try:

- you could try cutting down the modem speed just to see if that makes
  a difference.
- try playing with IO and interrupt resource settings: I'd exclude irq
  3 and port 0x2f8-0x2ff in /etc/pcmcia/config.opts, so the card will
  be configured somewhere else.
- I might try doing "setserial /dev/ttyS1 uart 16550A", just because I
  haven't seen cards reported as having ST16650V2 uart's before.

-- Dave

Ok "setserial /dev/ttyS1 uart 16550A" did the trick

Re: None It's strange that the connection starts up ok (David Hinds)
Date: 2000, Jun 01
From: Cengiz Alaettinoglu cengiz

The subject says it all. Thank you very much.

Cengiz

Sad EtherFast 10/100 + 56k problem

Date: 2000, Jun 13
From: David Arnold darnold78

Okay, a bit of mess but here it is.

RH 6.2, PCMCIA 3.1.15, EtherFast 10/100 +56K modem.

eth cable/connection works great.  Take the eth cable out,
plug it back in, no problems.  But the modem..  If you take
the eth cable out, then insert the phone cable into the modem
slot, test_modem will always return a poll error or such.  The
weird thing, test_modem will always work when the eth cable is plugged in!

I've tried configuring /etc/pcmcia/config.opts, commenting out
the various lines and restarting cardmgr, to no avail.

You have to eject the card, plug in the modem cable, then
insert back to get the modem recognised.  This is bearable but
there should be an easier way...

David

None You may have to live with it for now

Re: Sad EtherFast 10/100 + 56k problem (David Arnold)
Date: 2000, Jun 14
From: David Hinds <dhinds@pcmcia.sourceforge.org>

This sounds like the sort of problem that would be very specific to
this particular card.  It may have to wait until someone with this
card is sufficiently motivated to track it down; I don't have one of
these.

-- Dave

Feedback Semi-workaround

Re: None You may have to live with it for now (David Hinds)
Date: 2000, Jun 14
From: David Arnold darnold78

Hm.

After reading this board, I tried and succeeded with the following:
#setserial /dev/ttyS1 irq 0

So I take the eth cable out, put the modem in, then execute the above line. And it works flawlessly.

And no modifying /etc/pcmcia/config.opts does not fix the problem... I will setup a script similar to:

#!/bin/sh
setserial /dev/ttyS1 irq 0
/usr/sbin/pppd

To get the PPP connection working

David

Question Configuring Quatech serial ports for specific devices

Date: 2000, Jun 20
From: Jaime V Miro jaime

Hi, I've bought and installed a dual serial Quatech PCMCIA card (DSP-100) in my laptop. The laptop has got one serial port in /dev/ttyS0, and a secondary one for the infrared which I've diabled in BIOS. Linux (RH6.1) works fine and locates /dev/ttyS0, and cardmgr identifies the card with the 2 additional ports. However, when it configures the ports, these are not assigned to the next available devices, /dev/ttyS1 and /dev/ttyS2, which according to what I've read in the PCMCIA HOWTO documentation, I believe should happen.

It is a small inconvenience that would normally not bother me. So long as I know which dev's to address the ports to, no probs. However, I've got an application that runs on various Linux machines, and I'd like not to have to recompile it every time I run it in the laptop.

Q: is there a way to force cardmsg to assign the ports to specific devices? (I would only need one from the dual serial PCMCIA card, the other is not in use)

Thanks in advance

Jaime

None There are several options

Re: Question Configuring Quatech serial ports for specific devices (Jaime V Miro)
Date: 2000, Jun 20
From: David Hinds <dhinds@pcmcia.sourceforge.org>

> 
> Q: is there a way to force cardmsg to assign the ports to specific
> devices? (I would only need one from the dual serial PCMCIA card, the
> other is not in use)

The choice of device names is made by the kernel's serial driver and
can be somewhat arbitrary.  You should not rely on consistent naming
of the devices.

You have several options.  One is to parse /var/lib/pcmcia/stab on
your own, and extract the device names.  The other would be to set up
/etc/pcmcia/serial.opts to create links to the devices that would
always be the same, and then use those links instead of the actual
device files.  This would do it:

  case "$ADDRESS" in
  *,*,0)
      LINK="/dev/port0"
      ;;
  *,*,1)
      LINK="/dev/port1"
      ;;
  esac

-- Dave

Question 3COM 3CXM756 PCMCIA modem not working

Date: 2000, Jun 25
From: capkapak

Anyone please help. I have tried to get a 3Com MegaHertz 3CXM756 PCMCIA modem to work with Red Hat 6.2 with kernel 2.2.14-12 but with no luck at all. I tried a Psion DACOM Gold card and it works fine. When I insert the 3Com card I hear the two beeps indicating that the card has been recongnised. Check the messages log and it indicates that the card has been identified as a serial device but as a 16450 UART and not 16550. Have tried everything from using setserial to set the IRQ to 0 to re-configuring the "config" script in /etc/pcmcia/ but with no luck whatsoever.

Does anyone have been successful in utilising this particular PCMCIA modem in Linux?

Thanks.

None This might be an IO port conflict

Re: Question 3COM 3CXM756 PCMCIA modem not working
Date: 2000, Jul 08
From: David Hinds <dhinds@pcmcia.sourceforge.org>

The bad UART identification usually means an IO port issue.

If the device is being configured for a port above 0x1000, you could
also try removing this high port window from /etc/pcmcia/config.opts
since it seems to not work on some systems.  Look for and remove the
"port 0x1000-0x17ff".

-- Dave

Question same problem can you give me info ?

Re: Question 3COM 3CXM756 PCMCIA modem not working
Date: 2000, Aug 28
From: jean-raymond Chauviere jrc2

I have exactly the same problem with the same card. I have made the same test with the same PSION card ! (I didn't read your message before..) Can you give me every information you have on this subject and the /etc/pcmcia/config entry ? Thank JRC please reply to jr@chauviere.org

Question Socket dual rs-232 pcmcia... rev 2.1

Date: 2000, Jun 26
From: john gwynne <jsg@mrcday.com>

system: RH6.2 on winbook XP
PCMCIA card: Socket "Dual Serial I/O Card Rev 2.1"
kernel: redhat's 2.2.16-3 with pcmcia 3.1.17


I'm having problems trying to use a Socket "Dual Serial 
I/O Card" with the following identity:

[root@efi /root]# /sbin/cardctl ident 0
  product info: "Socket", "Dual I/O Card Rev 2.1"
  manfid: 0x0104, 0x0070
  function: 2 (serial)

This card is listed as supported; however, maybe the card
I have is a later version that fails????

The card is detected (positive beeps) with the following
kernel messages:

==============================================================
Jun 26 17:00:24 efi pcmcia:  cardmgr.
Jun 26 17:00:24 efi rc: Starting pcmcia succeeded
Jun 26 17:00:24 efi cardmgr[376]: starting, version is 3.1.17
Jun 26 17:00:24 efi cardmgr[376]: watching 2 sockets
Jun 26 17:00:24 efi kernel: cs: IO port probe 0x1000-0x17ff:
 excluding 0x1000-0x100f 0x1018-0x1027 0x1040-0x1047 
 0x1060-0x1067 0x1080-0x1097 0x10a0-0x10a7 0x10c0-0x10d7
 0x10f0-0x10f7 0x11f0-0x11f7 0x1268-0x126f 0x12f8-0x12ff
 0x1378-0x137f 0x13f0-0x140f 0x1418-0x1427 0x1440-0x1447
 0x1460-0x1467 0x1480-0x1497 0x14a0-0x14a7 0x14c0-0x14d7
 0x14f0-0x14f7 0x15f0-0x15f7 0x1668-0x166f 0x16f8-0x16ff
 0x1778-0x177f 0x17f0-0x17ff 
Jun 26 17:00:24 efi kernel: cs: IO port probe 0x0100-0x04ff:
 excluding 0x100-0x107 0x268-0x26f 0x378-0x37f 0x400-0x40f
 0x418-0x427 0x440-0x447 0x460-0x467 0x480-0x497 0x4a0-0x4a7
 0x4c0-0x4d7 0x4f0-0x4f7 
Jun 26 17:00:24 efi kernel: cs: IO port probe 0x0a00-0x0aff:
 excluding 0xa68-0xa6f 0xaf8-0xaff 
Jun 26 17:00:25 efi cardmgr[376]: initializing socket 0
Jun 26 17:00:25 efi kernel: cs: memory probe
 0x0d0000-0x0dffff: clean. 
Jun 26 17:00:25 efi cardmgr[376]: socket 0: Serial or Modem
Jun 26 17:00:25 efi cardmgr[376]: executing: 'modprobe
 serial_cs'
Jun 26 17:00:26 efi kernel: tty02 at 0x03e8 (irq = 5) is a
 8250 
Jun 26 17:00:26 efi cardmgr[376]: executing: './serial start
 ttyS2'
==============================================================


serial_cs identified only one channel and incorrectly shows a
8250 uart. This channel does not work nor do various settings
with setserial (including uart=16550a and irq=0).


the pcmcia bridge settings:

==============================================================
[root@efi /root]# dump_i365 
Intel PCIC probe: Vadem VG-468 found, 2 sockets
Socket 0:
  Identification and revision = 0x83
  Interface status = 0x7f
    [bvd1/stschg] [bvd2/spkr] [detect] [wrprot] [ready] [poweron]
  Power control = 0xd0
    [output] [Vcc=5v] [Vpp off]
  Interrupt and general control = 0x65
    [iocard] [irq=5]
  Card status change = 00
   
  Card status change interrupt control = 0xf8
    [detect] [irq = 15]
  Control register = 00
    [active low]
  Misc register = 0x80
    [unlock]
  Card detect and general control = 00
   
  Global control = 00
   
  Memory window 0: [ON] [WS1] [REG]
    start = 0x00d0, stop = 0x00d0, offset = 0x3f30
  Memory window 1: [OFF]
    start = 0000, stop = 0000, offset = 0000
  Memory window 2: [OFF]
    start = 0000, stop = 0000, offset = 0000
  Memory window 3: [OFF]
    start = 0000, stop = 0000, offset = 0000
  Memory window 4: [OFF]
    start = 0000, stop = 0000, offset = 0000
  I/O window 0: [ON]
    start = 0x03e8, stop = 0x03ef
  I/O window 1: [OFF]
    start = 0000, stop = 0x0001
==============================================================


a dump_cis follows...

==============================================================
[root@efi /root]# ./dump_cis    
Socket 0:
  dev_info
    NULL 0ns, 512b
  checksum 0x000c-0x0046 = 0xb3
  attr_dev_info
    fn_specific 200ns, 64kb
  mfc {
    manfid 0x0104, 0x0070
    vers_1 5.0, "Socket", "Dual I/O Card Rev 2.1"
    funcid serial_port [post]
    config base 0x0400 mask 0x0003 last_index 0x2d
    cftable_entry 0x0d [default]
      [rdybsy] [pwrdown]
      Vcc Vnom 5V Vmin 4500mV Vmax 5500mV Istatic 15mA
        Iavg 25mA Ipeak 2500uA Idown 20uA
      io 0x03f8-0x03ff [lines=16] [8bit] [range]
      irq mask 0xffff [level]
    cftable_entry 0x15 [default]
      [rdybsy] [pwrdown]
      Vcc Vnom 5V Vmin 4500mV Vmax 5500mV Istatic 15mA
        Iavg 25mA Ipeak 2500uA Idown 20uA
      io 0x02f8-0x02ff [lines=16] [8bit] [range]
      irq mask 0xffff [level]
    cftable_entry 0x1d [default]
      [rdybsy] [pwrdown]
      Vcc Vnom 5V Vmin 4500mV Vmax 5500mV Istatic 15mA
        Iavg 25mA Ipeak 2500uA Idown 20uA
      io 0x03e8-0x03ef [lines=16] [8bit] [range]
      irq mask 0xffff [level]
    cftable_entry 0x25 [default]
      [rdybsy] [pwrdown]
      Vcc Vnom 5V Vmin 4500mV Vmax 5500mV Istatic 15mA
        Iavg 25mA Ipeak 2500uA Idown 20uA
      io 0x02e8-0x02ef [lines=16] [8bit] [range]
      irq mask 0xffff [level]
    cftable_entry 0x2d [default]
      [rdybsy] [pwrdown]
      Vcc Vnom 5V Vmin 4500mV Vmax 5500mV Istatic 15mA
        Iavg 25mA Ipeak 2500uA Idown 20uA
      io 0x0000-0x0007 [lines=3] [8bit]
      irq mask 0xffff [level]
  }, {
    funcid serial_port [post]
    config base 0x0420 mask 0x0003 last_index 0x2d
    cftable_entry 0x0d [default]
      [rdybsy] [pwrdown]
      Vcc Vnom 5V Vmin 4500mV Vmax 5500mV Istatic 15mA
        Iavg 25mA Ipeak 2500uA Idown 20uA
      io 0x03f8-0x03ff [lines=16] [8bit] [range]
      irq mask 0xffff [level]
    cftable_entry 0x15 [default]
      [rdybsy] [pwrdown]
      Vcc Vnom 5V Vmin 4500mV Vmax 5500mV Istatic 15mA
        Iavg 25mA Ipeak 2500uA Idown 20uA
      io 0x02f8-0x02ff [lines=16] [8bit] [range]
      irq mask 0xffff [level]
    cftable_entry 0x1d [default]
      [rdybsy] [pwrdown]
      Vcc Vnom 5V Vmin 4500mV Vmax 5500mV Istatic 15mA
        Iavg 25mA Ipeak 2500uA Idown 20uA
      io 0x03e8-0x03ef [lines=16] [8bit] [range]
      irq mask 0xffff [level]
    cftable_entry 0x25 [default]
      [rdybsy] [pwrdown]
      Vcc Vnom 5V Vmin 4500mV Vmax 5500mV Istatic 15mA
        Iavg 25mA Ipeak 2500uA Idown 20uA
      io 0x02e8-0x02ef [lines=16] [8bit] [range]
      irq mask 0xffff [level]
    cftable_entry 0x2d [default]
      [rdybsy] [pwrdown]
      Vcc Vnom 5V Vmin 4500mV Vmax 5500mV Istatic 15mA
        Iavg 25mA Ipeak 2500uA Idown 20uA
      io 0x0000-0x0007 [lines=3] [8bit]
      irq mask 0xffff [level]
  }
==============================================================

and a dump_cisreg.... which has an i/o error???

==============================================================
[root@efi /root]# dump_cisreg 
Socket 0 function 0:
  Config register base = 0x0400, mask = 0x0003
  Configuration option register = 0x5f
    [level_req] [func_ena] [addr_decode] [ireq_ena] [index = 0x18]
  Card configuration and status register = 00
   

Socket 0 function 1:
  Config register base = 0x0420, mask = 0x0003
  read config register: Input/output error

Socket 1 function 0:
  Config register base = 0x0400, mask = 0x000b
  Configuration option register = 0x60
    [level_req] [index = 0x20]
  Card configuration and status register = 00
   
  Socket and copy register = 00
    [socket = 0] [copy = 0]
==============================================================


any help would be appreciated... 

TIA,
john gwynne
jsg@mrcday.com

None It seems to be a new revision

Re: Question Socket dual rs-232 pcmcia... rev 2.1 (john gwynne)
Date: 2000, Jul 11
From: David Hinds <dhinds@pcmcia.sourceforge.org>

Try adding to /etc/pcmcia/config:

  card "Socket Dual RS-232"
    manfid 0x0104, 0x0070
    bind "serial_cs" to 0, "serial_cs" to 1

which should get the driver to try to configure both ports.

> Jun 26 17:00:26 efi kernel: tty02 at 0x03e8 (irq = 5) is a 8250 

The bad UART makes me wonder if you might have an IO port conflict at
0x3e8-0x3ef.  It is at least worth excluding this port range in your
/etc/pcmcia/config.opts file to see if that makes any difference.

The dump_cis output doesn't look quite right to me.  You might also
try creating a file socket.cis containing:

  dev_info
    NULL 0ns, 512b
  mfc {
    manfid 0x0104, 0x0070
    vers_1 5.0, "Socket", "Dual I/O Card Rev 2.1"
    funcid serial_port [post]
    config base 0x0400 mask 0x0263 last_index 0x2d
    cftable_entry 0x2d [default]
      [rdybsy] [pwrdown]
      Vcc Vnom 5V Vmin 4500mV Vmax 5500mV Istatic 15mA
        Iavg 25mA Ipeak 2500uA Idown 20uA
      io 0x0000-0x0007 [lines=3] [8bit]
      irq mask 0xffff [level]
  }, {
    funcid serial_port [post]
    config base 0x0420 mask 0x0263 last_index 0x2d
    cftable_entry 0x2d [default]
      [rdybsy] [pwrdown]
      Vcc Vnom 5V Vmin 4500mV Vmax 5500mV Istatic 15mA
        Iavg 25mA Ipeak 2500uA Idown 20uA
      io 0x0000-0x0007 [lines=3] [8bit]
      irq mask 0xffff [level]
  }

and then do:

  pack_cis -o socket.dat socket.cis

then put the socket.dat file in /etc/pcmcia/cis, and change the config
file entry to:

  card "Socket Dual RS-232"
    manfid 0x0104, 0x0070
    cis "cis/socket.dat"
    bind "serial_cs" to 0, "serial_cs" to 1

then see how that behaves.

-- Dave

Feedback Re: It seems to be a new revision

Re: None It seems to be a new revision (David Hinds)
Date: 2000, Jul 11
From: john gwynne <jsg@mrcday.com>

| Try adding to /etc/pcmcia/config:
|
| card "Socket Dual RS-232"
| manfid 0x0104, 0x0070
| bind "serial_cs" to 0, "serial_cs" to 1
|
| which should get the driver to try to configure both ports.

fails...

  Jul 11 16:18:11 efi cardmgr[3800]: socket 0: Socket Dual RS-232
  Jul 11 16:18:12 efi cardmgr[3800]: executing: 'modprobe serial_cs'
  Jul 11 16:18:12 efi kernel: tty02 at 0x03e8 (irq = 5) is a 8250
  Jul 11 16:18:12 efi cardmgr[3800]: 'serial_cs' already bound to socket 0
  Jul 11 16:18:12 efi cardmgr[3800]: executing: './serial start ttyS2'
  Jul 11 16:18:13 efi cardmgr[3800]: executing: './serial start ttyS2'

as a WAG, i tried serial_cb...

  Jul 11 16:20:49 efi cardmgr[4184]: socket 0: Socket Dual RS-232
  Jul 11 16:20:49 efi cardmgr[4184]: executing: 'modprobe cb_enabler'
  Jul 11 16:20:50 efi cardmgr[4184]: executing: 'modprobe serial_cb'
  Jul 11 16:20:50 efi kernel: serial_cb: RequestIO: Unsupported mode
  Jul 11 16:20:51 efi cardmgr[4184]: get dev info on socket 0 failed: Resource temporarily unavailable


| The dump_cis output doesn't look quite right to me. You might also
| try creating a file socket.cis containing:
|
| dev_info
| NULL 0ns, 512b
| mfc {
| manfid 0x0104, 0x0070

<delete>

| irq mask 0xffff [level]
| }

still a no go... (using bind "serial_cs")

  Jul 11 16:33:55 efi cardmgr[4708]: socket 0: Socket Dual RS-232
  Jul 11 16:33:55 efi cardmgr[4708]: executing: 'modprobe serial_cs'
  Jul 11 16:33:56 efi kernel: tty02 at 0x03e8 (irq = 5) is a 8250
  Jul 11 16:33:56 efi cardmgr[4708]: executing: './serial start ttyS2'

BUT... this seems to be moving in the right
direction... I tried this description....

  dev_info
    NULL 0ns, 512b
  manfid 0x0104, 0x0070
  vers_1 5.0, "Socket", "Dual I/O Card Rev 2.1"
  funcid serial_port [post]
  config base 0x0400 mask 0x0003 last_index 0x2d
  cftable_entry 0x2d [default]
    [rdybsy] [pwrdown]
    Vcc Vnom 5V Vmin 4500mV Vmax 5500mV Istatic 15mA
      Iavg 25mA Ipeak 2500uA Idown 20uA
    io 0x03e8-0x03ef [lines=16] [8bit] [range]
    io 0x02e8-0x02ef [lines=16] [8bit] [range]
    irq mask 0xffff [level]

which gives...

  Jul 11 16:28:31 efi cardmgr[4513]: socket 0: Socket Dual RS-232
  Jul 11 16:28:31 efi cardmgr[4513]: executing: 'modprobe serial_cs'
  Jul 11 16:28:31 efi kernel: tty02 at 0x03e8 (irq = 5) is a ST16650V2
  Jul 11 16:28:31 efi kernel: tty03 at 0x02e8 (irq = 5) is a ST16650V2
  Jul 11 16:28:31 efi cardmgr[4513]: executing: './serial start ttyS2'
  Jul 11 16:28:32 efi cardmgr[4513]: executing: './serial start ttyS3'

At this point "port-A" responds as both ttyS2 and ttyS3... but at least it is working (small victory:). I tried several other mfc {},{} descriptions, but could not get any of those to work at all... only the above (with and without the second I/O window). How can I get both ports to work?

john gwynne

Question Xircom Realport2 ISDN Adapter (R2I)

Date: 2000, Jul 06
From: Mike Gerber mgerber

I'm not sure if that is the right forum (although there is 
none specific for ISDN adapters).

OK, I'm currently running kernel 2.2.14-12 (RedHat build) 
with pcmcia version 3.1.8 and I'm interested in getting 
the "Xircom Realport2 ISDN Adapter" to work with Linux (This
card is only ISDN, no eth or modem). Anyone had any luck?

cardmgr says:

Jun 29 10:04:59 thinkpad cardmgr[1701]: unsupported card in socket 0
Jun 29 10:04:59 thinkpad cardmgr[1701]:   product info: "Xircom", "RealPort2 ISDN Adapter", "R2I", "1.00"
Jun 29 10:04:59 thinkpad cardmgr[1701]:   manfid: 0x0105, 0x200a  function: 6 (network)

Question ADVICE: Configuring my PCMCIA GSM Modem

Date: 2000, Jul 09
From: F FrankC

Hi,

As a Linux newbie, I'm trying to configure an unsupported PCMCIA GSM Modem card using RedHat 6.1-Linux 2.2.12-20 and pcmcia-cs-3.1.17 on a Compaq Armada 1750.

But... I seem to be having a minor problem. When I modify the "test_modem" script to ATD<Phone Number> instead of ATI3, I get the following output:

"Checking modem at /dev/ttyS1:
  Settings: /dev/ttyS1, UART: 16550A, Port: 0x02f8, IRQ: 3
  Modem query timed out: trying polled mode.
  The modem is not working in polled mode."

I've tried all sort of things including setting the irq to 0 but no avail - what am I doing wrong????

I've included a listing of some useful outputs of my current configuration.

Any advice would be really welcome.

Cheers,

Frank

[ ***** LSMOD OUTPUT ***** ] Module Size Used by serial_cs 5072 0 (unused) ds 6088 2 [serial_cs] i82365 21404 2 pcmcia_core 43648 0 [serial_cs ds i82365] opl3 11208 0 (unused) sb 33620 0 uart401 5968 0 [sb] sound 57240 0 [opl3 sb uart401] soundlow 300 0 [sound] soundcore 2372 6 [sb sound]

[ ***** DMESG EXCERPT ***** ] ...

Serial driver version 4.27 with MANY_PORTS MULTIPORT SHARE_IRQ enabled ttyS00 at 0x03f8 (irq = 4) is a 16550A ttyS02 at 0x03e8 (irq = 4) is a 16550A

...

Linux PCMCIA Card Services 3.1.17
  kernel build: 2.2.12-20 #1 Mon Sep 27 10:40:35 EDT 1999
  options:  [pci] [cardbus] [apm]
PCI routing table version 1.0 at 0xf09a0
  00:11.0 -> irq 11
  00:11.1 -> irq 11
Intel PCIC probe: 
  TI 1225 rev 01 PCI-to-CardBus at slot 00:11, mem 0x7fffe000
    host opts [0]: [ring] [pci + serial irq] [pci irq 11] [lat 66/176] [bus 2/2]
    host opts [1]: [ring] [pci + serial irq] [pci irq 11] [lat 66/176] [bus 3/3]
    ISA irqs (scanned) = 3,4,7,9,10,15 PCI status changes
cs: IO port probe 0x02f8-0x02ff: clean.
cs: memory probe 0xa0000000-0xa0ffffff: clean.
tty01 at 0x02f8 (irq = 3) is a 16550A

[ *****  CARDCTL CONFIG  *****]
Socket 0:
  Vcc 5.0V  Vpp1 0.0V  Vpp2 0.0V
  interface type is "memory and I/O"
  irq 3 [exclusive] [level]
  Speaker output is enabled
  function 0:
    config base 0x0200
      option 0x61 status 0x08 pin 0x00 ext 0x00
    io 0x02f8-0x02ff [8bit]
Socket 1:
  not configured

[ *****  TEST_MODEM  *****]
Checking modem at /dev/ttyS1:
  Settings: /dev/ttyS1, UART: 16550A, Port: 0x02f8, IRQ: 3
  ATI3 = 'OK '
  The modem is operating normally.

[ ***** TEST_SETUP *****]
Current kernel: 2.2.12-20 #1 Mon Sep 27 10:40:35 EDT 1999
Module info from /lib/modules/2.2.12-20/pcmcia/pcmcia_core.o:
  Linux PCMCIA Card Services 3.1.17
  kernel build: 2.2.12-20 #1 Mon Sep 27 10:40:35 EDT 1999
  options:  [pci] [cardbus] [apm]

Startup options from /etc/sysconfig/pcmcia:
  PCMCIA=yes
  PCIC=i82365
  PCIC_OPTS=
  CORE_OPTS=

Checking current syslog files in /var/log:
  All PCMCIA messages are in /var/log/messages.

Module status:
  The PCMCIA kernel modules are loaded correctly.

Daemon status:
  cardmgr is running (process 366)

Current socket status from /var/lib/pcmcia/stab:
  Socket 0: Serial or Modem
  0	serial	serial_cs	0	ttyS1	4	65
  Socket 1: empty

None Why did you want to modify test_modem this way?

Re: Question ADVICE: Configuring my PCMCIA GSM Modem (F)
Date: 2000, Jul 10
From: David Hinds <dhinds@pcmcia.sourceforge.org>

The test_modem script is not meant to be used to do other things like
dial numbers.  From the un-altered test_modem output, it appears that
the modem is working normally.  Do you have other symptoms, like does
the modem not work when you use a communications program like minicom?

(also shouldn't the dial command be ATDTnnn-nnnn?)

-- Dave

Agree RE: Why did you want to modify test modem this way?

Re: None Why did you want to modify test_modem this way? (David Hinds)
Date: 2000, Jul 11
From: F FrankC

Yep good point...

Being a newbie, after posting this message - I did additional searches of other postings and news groups and discovered using 'minicom' was an alternative way of testing my configuration.

And... it works! DOH!

As my PCMCIA GSM modem also supports voice, my final challenge is to work out how accomplish voice XOR data communications. With data, I'm using the AT command set augmented by the ETSI command set of 07.05 and 07.07. With voice ... well I'm looking into it and again, it's probably something real simple.

Cheers,

Frank

Question Compex Linkport-Comm 56-E CTR21 V90 modem

Date: 2000, Aug 09
From: Mitja Pufic RAVEer

Hello!

I recently bought that modem with hope that it will work perfectly like the ethernet card EnetB from the same family.

I discovered that inside is rockwells RCDL56ACFL chipset, and i now i am little scared that this is winmodem, how can i know that?

well cardctl ident report: 
Socket 1:
  product info: "PC Card K56 Fax Modem", "RCDL56ACFL V.90 & K56flex", "021", "A"  
manfid: 0x0013, 0x0000
  function: 2 (serial)

and from /var/log/messages

Aug  9 13:40:01 raindrop cardmgr[588]: initializing socket 1
Aug  9 13:40:01 raindrop cardmgr[588]: socket 1: EZLink 4109 Ethernet
Aug  9 13:40:01 raindrop cardmgr[588]: executing: 'insmod
/lib/modules/2.2.14/net/8390.o' Aug 9 13:40:01 raindrop cardmgr[588]: executing: 'insmod /lib/modules/2.2.14/pcmcia/pcnet_cs.o'
Aug  9 13:40:01 raindrop kernel: eth0: NE2000 Compatible: io 0x300, irq 3, hw_addr 00:A0:0C:13:08:42
Aug  9 13:40:01 raindrop cardmgr[588]: executing: './network start eth0'
Aug  9 13:40:02 raindrop cardmgr[588]: + SIOCADDRT: Invalid argument
Aug  9 13:40:25 raindrop cardmgr[588]: initializing socket 0
Aug  9 13:40:25 raindrop cardmgr[588]: socket 0: Serial or Modem
Aug  9 13:40:25 raindrop cardmgr[588]: executing: 'insmod
/lib/modules/2.2.14/pcmcia/serial_cs.o port 0x03e8'

like i don't know how to recognize winmodem, but i think that this one is real, so what are the next steps to finaly get it working

like cardctl ident and status shows the card, but cat /var/stat/pcmcia/stab shows only ethernet card? any ideas?

l'd be really glad if someone could help me out, thanks in advance

mitja pufic raveer@kibla.org

None Bad serial_cs parameters

Re: Question Compex Linkport-Comm 56-E CTR21 V90 modem (Mitja Pufic)
Date: 2000, Aug 09
From: David Hinds <dhinds@pcmcia.sourceforge.org>

There is a more complete list of WinModems at:

  http://www.o2.net/~gromitkc/winmodem.html

> Aug  9 13:40:25 raindrop cardmgr[588]: initializing socket 0
> Aug  9 13:40:25 raindrop cardmgr[588]: socket 0: Serial or Modem
> Aug  9 13:40:25 raindrop cardmgr[588]: executing: 'insmod
> /lib/modules/2.2.14/pcmcia/serial_cs.o port 0x03e8'

And after that...?

I don't understand the "port 0x03e8".  Is this something you added to
your /etc/pcmcia/config.opts file?  Whatever it is, it will prevent
the serial_cs module from loading at all, since this is not a valid
parameter list.  I would expect the next few lines of your system log
to report that the module could not be loaded.

Undo whatever you did to add the "port 0x03e8".  What messages do you
get then?

-- Dave

More Compex Linkport-Comm 56-E CTR21 V90 modem 2

Re: None Bad serial_cs parameters (David Hinds)
Date: 2000, Aug 09
From: Mitja Pufic RAVEer

Hello, thanks for your help

yes i added that port in config.opts, now when i removed it the /var/log/messages output is:

Aug 10 00:45:43 raindrop cardmgr[292]: initializing socket 1
register_serial(): autoconfig failed
serial_cs: register_serial() at 0x13f8, irq 3 failed
Aug 10 00:45:43 raindrop cardmgr[292]: socket 1: Serial or Modem
Aug 10 00:45:43 raindrop kernel: cs: memory probe 0x0d0000-0x0dffff:
clean.
Aug 10 00:45:43 raindrop cardmgr[292]: executing: 'insmod
/lib/modules/2.2.14/pcmcia/serial_cs.o'
Aug 10 00:45:43 raindrop kernel: register_serial(): autoconfig failed
Aug 10 00:45:43 raindrop kernel: serial_cs: register_serial() at 0x13f8,
irq 3 failed
Aug 10 00:45:45 raindrop cardmgr[292]: get dev info on socket 1 failed:
Resource temporarily unavailable

when i inserted fisrt my eth card and then modem the output is:

Aug 10 00:46:42 raindrop cardmgr[292]: initializing socket 0
Aug 10 00:46:42 raindrop cardmgr[292]: socket 0: EZLink 4109 Ethernet
Aug 10 00:46:42 raindrop cardmgr[292]: executing: 'insmod
/lib/modules/2.2.14/net/8390.o'
Aug 10 00:46:42 raindrop cardmgr[292]: executing: 'insmod
/lib/modules/2.2.14/pcmcia/pcnet_cs.o'
Aug 10 00:46:42 raindrop kernel: eth0: NE2000 Compatible: io 0x300, irq 3,
hw_addr 00:A0:0C:13:08:42
Aug 10 00:46:42 raindrop cardmgr[292]: executing: './network start eth0'
register_serial(): autoconfig failed
serial_cs: register_serial() at 0x13f8, irq 9 failed
Aug 10 00:46:51 raindrop cardmgr[292]: initializing socket 1
Aug 10 00:46:51 raindrop cardmgr[292]: socket 1: Serial or Modem
Aug 10 00:46:51 raindrop cardmgr[292]: executing: 'insmod
/lib/modules/2.2.14/pcmcia/serial_cs.o'
Aug 10 00:46:51 raindrop kernel: register_serial(): autoconfig failed
Aug 10 00:46:51 raindrop kernel: serial_cs: register_serial() at 0x13f8,
irq 9 failed
Aug 10 00:46:52 raindrop cardmgr[292]: get dev info on socket 1 failed:
Resource temporarily unavailable

From this i logs i see that irq 3 is reserved for eth card, but why can't modem take up irq 9?

about the modem, i looked at database in www.o2.net/~gromitkc/winmodem.html but i didn't find my chipset listed, there were some similar chips e.g. RCDL56ACF 
without L at the and, and they had ? in cathegory box

thanks

mitja

None IO port issue

Re: More Compex Linkport-Comm 56-E CTR21 V90 modem 2 (Mitja Pufic)
Date: 2000, Aug 09
From: David Hinds <dhinds@pcmcia.sourceforge.org>

> serial_cs: register_serial() at 0x13f8, irq 9 failed

...

>From this i logs i see that irq 3 is reserved for eth card, but why
>can't modem take up irq 9?

The irq is fine.  Edit /etc/pcmcia/config.opts and comment out the
0x1000-0x17ff IO port window, and that may improve the situation.

-- Dave

Feedback Nokia Card Phone 2.0

Date: 2000, Aug 11
From: Jochen Scharrlach <js@mgm-net.de>

Finally I got it working. The trick on my system was, that
you must not talk to the second serial line at all - even
setting the irq results in a hanging card until you do a 
"cardctl reset". This effect seems to occur only on my
toshiba, on a Compaq everything works fine. My solution was
to set "DEVICE=ignore" in serial.opts if an "ADDRESS=*,*,1"
was requested and of course to exit the serial-script if
"$DEVICE==ignore".

Notebook: Toshiba 2180CDT with RedHat 6.2

dmesg:

Intel PCIC probe: 
  Toshiba ToPIC95-B rev 07 PCI-to-CardBus at slot 00:13, mem 0x68000000
    host opts [0]: [slot 0xf0] [ccr 0x10] [cdr 0x86] [rcr 0x00] [no pci irq] [lat 168/176] [bus 20/20]
    host opts [1]: [slot 0xf0] [ccr 0x20] [cdr 0x86] [rcr 0x00] [no pci irq] [lat 168/176] [bus 21/21]
    ISA irqs (default) = 3,5,9,10 polling interval = 1000 ms
cs: IO port probe 0x0100-0x04ff: clean.
cs: memory proe 0x0d0000-0x0dffff: clean.
[...]
tty16 at 0x0100 (irq = 5) is a 16550A
tty17 at 0x0108 (irq = 5) is a 8250

Card: Nokia Card Phone 2.0 (slot 1 above)

[root@ezri ~]# uname -rv
2.2.16-3 #1 Mon Jun 19 18:49:25 EDT 2000
[root@ezri ~]# cardctl -V
cardctl version 3.1.19

Question Diehl/DIVA ISDN card

Date: 2000, Aug 15
From: Rhys Morgan rmorgan

Hi,

I have come across an ISDN Terminal Adapter card manufactured by a company called "G.Diehl ISDN GmbH", in about 1996. I would like to know how to find out if one of the existing drivers would function (serial_cs doesn't), without going through them one-by-one. I am using a Toshiba Tecra8000 with a RedHat 2.2.9 kernel, and version 3.1.12 of the pcmcia card services.

For reference, the /var/log/messages file has the following relevent contents when I insert the card:

Aug 15 14:31:51 localhost cardmgr[391]: re-loading config file
Aug 15 14:31:51 localhost kernel: cs: IO port probe 0x1000-0x17ff: clean.
Aug 15 14:31:51 localhost kernel: cs: IO port probe 0x0100-0x04ff: excluding 0x220-0x22f 0x330-0x337 0x378-0x37f 0x388-0x38f 0x4d0-0x4d7
Aug 15 14:31:51 localhost kernel: cs: IO port probe 0x0a00-0x0aff: clean.
Aug 15 14:32:10 localhost cardmgr[391]: initializing socket 1
Aug 15 14:32:10 localhost cardmgr[391]: socket 1: Anonymous Memory
Aug 15 14:32:10 localhost cardmgr[391]: executing: 'insmod /lib/modules/2.2.9-17/pcmcia/sram_mtd.o'
Aug 15 14:32:10 localhost cardmgr[391]:   Common memory region at 0x0: Generic or SRAM
Aug 15 14:32:10 localhost kernel: sram_mtd: common at 0x0, 2 mb, 250 ns
Aug 15 14:32:10 localhost kernel: memory_cs: mem0: common 2 mb
Aug 15 14:32:10 localhost cardmgr[391]: executing: 'insmod /lib/modules/2.2.9-17/pcmcia/memory_cs.o'
Aug 15 14:32:10 localhost cardmgr[391]: executing: './memory start mem0'
Aug 15 14:32:28 localhost cardmgr[391]: executing: './memory check mem0'
Aug 15 14:32:28 localhost cardmgr[391]: shutting down socket 1
Aug 15 14:32:28 localhost cardmgr[391]: executing: './memory stop mem0'
Aug 15 14:32:29 localhost cardmgr[391]: executing: 'rmmod memory_cs'
Aug 15 14:32:29 localhost cardmgr[391]: executing: 'rmmod sram_mtd'

And the /etc/pcmcia/config file entry looks like (as copied from the "cardctl ident" output):

card "Diehl ISDN"
  version "ISDN-DIVA/PCM", "14.03.94", "1-1"
  bind "serial_cs"

Thank you,

Rhys Morgan

Question COMpad-32B 4 port card is recognized as parallel

Date: 2000, Aug 22
From: Felix Radensky felix

Hi,

I'm running pcmcia-cs-3.1.19 on 2.2.16 kernel. My distro is Slackware-7.1. I've got problems with COMpad-32B 4 port serial card. cardmgr reconizes it as parallel port card and tries loading parport_cs module. I know that COMpad-32/85 dual port cards are supported. Is there anything I can do with a 4 port one ?

Thanks in advance.

Felix.

Sad file transfer stalls at 115200 baud serial line speed

Date: 2000, Aug 27
From: Michael Feig meikkel

I am having similar problems as previously posted here and on the Xircom list with large file transfers (ftp over PPP) stalling reproducibly at the same point. While there was no mention of this in the earlier posts this only seems to happen in my system when the serial line speed is set to 115200 baud.

I have an Acer TravelMate 524TXV with an O2 Micro CardBus adapter running RedHat 6.2 with 2.2.16 kernel and PCMCIA version 3.1.19. I tried both a Xircom RealPort 56K as well as an ActionTec 56K Datalink PCMCIA card with the same problem.

Using PCIC_OPTS="pci_csc=1 pci_int=1" in the PCMCIA options file (/etc/sysconfig/pcmcia) helps a little bit for the Xircom card but does not seem to do much for the Actiontec. "A little bit" means that the file transfer does not stall completely but continues very sluggishly after the stalling point and eventually completes. The only fix so far is reducing the serial line speed to 57600 which is not desirable for a 56K modem using compression.

I have no idea at this point what else to try to get a 56K PCMCIA modem working with 115000 baud serial line speed and would be thankful for any suggestions.

I am also not sure how to isolate further what component is causing the problem (serial_cs, pcmcia_core, pppd?).

For reference various information from my system:

lspci:

00:00.0 Host bridge: Acer Laboratories Inc. [ALi] M1621 (rev 05)
00:01.0 PCI bridge: Acer Laboratories Inc. [ALi] M5247 (rev 01)
00:06.0 Multimedia audio controller: Acer Laboratories Inc. [ALi]: Unknown device 5451 (rev 01)
00:07.0 ISA bridge: Acer Laboratories Inc. [ALi] M1533 PCI to ISA Bridge [Aladdin IV]
00:0a.0 Ethernet controller: Intel Corporation 82557 [Ethernet Pro 100] (rev 08)00:10.0 IDE interface: Acer Laboratories Inc. [ALi] M5229 IDE (rev c3)
00:11.0 Bridge: Acer Laboratories Inc. [ALi] M7101 PMU
00:13.0 CardBus bridge: O2 Micro, Inc.: Unknown device 6933 (rev 01)
00:13.1 CardBus bridge: O2 Micro, Inc.: Unknown device 6933 (rev 01)
00:14.0 USB Controller: Acer Laboratories Inc. [ALi] M5237 USB (rev 03)
01:00.0 VGA compatible controller: ATI Technologies Inc 3D Rage P/M Mobility AGP 2x (Rev. 64)

dmesg:

...
Linux PCMCIA Card Services 3.1.19
  kernel build: 2.2.16-3 #11 Sat Aug 26 22:16:57 PDT 2000
  options:  [pci] [cardbus] [apm] [pnp]
PCI routing table version 1.0 at 0xfb9d0
  00:13.0 -> irq 11
  00:13.1 -> irq 11
PnP: PNP BIOS installation structure at 0xc00f65b0
PnP: PNP BIOS version 1.0, entry at fa000:0, dseg at f0000
Intel PCIC probe:
  Unknown [1217 6933] rev 01 PCI-to-CardBus at slot 00:13, mem 0x68000000
    host opts [0]: [pci irq 11] [lat 32/176] [bus 32/34]
    host opts [1]: [pci irq 11] [lat 32/176] [bus 35/37]
    ISA irqs (default) = 3,5,9 PCI status changes
cs: IO port probe 0x1000-0x17ff: clean.
cs: IO port probe 0x0100-0x04ff: clean.
cs: IO port probe 0x0a00-0x0aff: clean.
cs: memory probe 0xa0000000-0xa0ffffff: clean.
tty01 at 0x13f8 (irq = 3) is a 16550A
...                         

/etc/sysconfig/pcmcia:

PCMCIA=yes PCIC=i82365 PCIC_OPTS="pci_csc=1 pci_int=1" CORE_OPTS=

/etc/pcmcia/config.opts: (unmodified from default)

#
# Local PCMCIA Configuration File
#
# System resources available for PCMCIA devices
#
include port 0x100-0x4ff, port 0x1000-0x17ff include memory 0xc0000-0xfffff include memory 0xa0000000-0xa0ffffff, memory 0x60000000-0x60ffffff
#
# Extra port range for IBM Token Ring
#
include port 0xa00-0xaff
#
# Resources we should not use, even if they appear to be available
#
# First built-in serial port
exclude irq 4
# Second built-in serial port
#exclude irq 3
# First built-in parallel port
exclude irq 7
# PS/2 Mouse controller port, comment this out if you don't have a PS/2
# based mouse
exclude irq 12
#
# Examples of options for loadable modules
#
# To fix sluggish network with IBM ethernet adapter...
#module "pcnet_cs" opts "mem_speed=600"
#
# Options for Xircom Netwave driver...
#module "netwave_cs" opts "domain=0x100 scramble_key=0x0"
#
# Options for IBM Token Ring adapters
#module "ibmtr_cs" opts "mmiobase=0xd0000 srambase=0xd4000"     

Angry PCMCIA works very very very slowly

Date: 2000, Aug 28
From: Benson Fung bensonfung

Dear all,

I have installed a PCI-PCMCIA adaptor(Cirrus PD6729) in my destop to connect the Nokia Cardphone 2.0.

In fact, the card phone modem work fine under RH Linux 6.1. However, it works very very very slowly.

Can anyone provide me a solution how to fix the problem? Right now, I am using pcmcia-cs 3.1.8. Do I need to upgrade it?

Benson

None Interrupt problem; you'll need a driver upgrade

Re: Angry PCMCIA works very very very slowly (Benson Fung)
Date: 2000, Aug 28
From: David Hinds <dhinds@pcmcia.sourceforge.org>

This is an interrupt delivery problem.  You will need to upgrade to a
fairly recent PCMCIA package (3.1.19 is current), then specify:

  PCIC_OPTS="irq_mode=1"

which, for this adapter, forces it to use PCI interrupts.

-- Dave

Question 3CXM656 - What is exactly a WinModem ?

Date: 2000, Sep 26
From: Christian Gennerat c_gennerat

When I insert the modem, I get:

cardmgr[369]: initializing socket 0

cardmgr[369]: unsupported card in socket 0

cardmgr[369]: product info: "3Com", "Megahertz 3CXM656", "Global Modem For Windows"

cardmgr[369]: manfid: 0x0101, 0x003b function: 254 (null))

modprobe: can't locate module char-major-108

.

In the cis, I have noticed that:

-There is No funcid

-The I/O address are not usual serial port addresses

Can this be fixed with a hand-made cis ?

What are the fundamental differencies between a WinModem, a TuxModem, and a standard modem?

None You are out of luck...

Re: Question 3CXM656 - What is exactly a WinModem ? (Christian Gennerat)
Date: 2000, Sep 26
From: David Hinds <dhinds@pcmcia.sourceforge.org>

The "global modem for windows" is a hint.  Also this card is listed in
the supported.cards list as a WinModem.  You can't use it with Linux.

WinModems implement most of the modem functionality in software that
runs on your CPU, instead of using a standard serial port interface
and hardware on the modem card.  You need dedicated drivers that know
how to talk to your particular WinModem.  There are no WinModem
drivers for Linux.  (well, that's not strictly true, Lucent has a
Linux WinModem driver for one of their PCI WinModem chipsets, but it
does you no good)

-- Dave

Question IBM V.90 PC Card Modem with XJACK ... funcid: 254 (null)

Date: 2000, Nov 01
From: Tom Bradner wetclay

what's a fellow to do?

I have an IBM ThinkPad 760E.
Cardmgr fires up ok but didn't recognize the card.
So I inserted the card in /etc/pcmcia/config and it was happy.
After serial_cs is loaded it tells me, "serial_cs: IORequest = no more items".
That is then followed by "cardmgr: get dev info on socket 1 failed : Resource temporarily unavailable".

I have tried to glean what I can from some of the other postings. But alas, it may get down to either a different modem or some real work.

One last point. As mentioned in the title, the funcid returns a value of 254 (null). Perhaps I should get the big rubber stamp out and print 'LOSER' on my forehead. But I'll wait until someone confirms my suspicions.

Thanks

Tom
PCMCIA Serial and Modem Issues


Add Message to: "PCMCIA Serial and Modem Issues"

Members Subscribe Admin Mode Show Frames Help for HyperNews at pcmcia-cs.sourceforge.net 1.10