Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Pcmcia-cs bug?

0 views
Skip to first unread message

Emil Erlandsson

unread,
Nov 11, 2002, 11:18:51 AM11/11/02
to
This is a followup message to "Mysterious cardmgr error".
I fixed my previous error by downloading the latest pcmcia-cs package
and recompiled. Now, a more serious error has occured.

When I try to start PCMCIA, I get a null pointer dereference from the
kernel.

dmesg:

Linux PCMCIA Card Services 3.2.3
kernel build: 2.4.18 #7 SMP Wed Nov 6 10:43:37 CET 2002
options: [pci] [cardbus] [apm]
Intel ISA/PCI/CardBus PCIC probe:
PCI: Found IRQ 10 for device 00:11.0
Ricoh RL5C475 rev 80 PCI-to-CardBus at slot 00:11, mem 0xd9001000
host opts [0]: [pci only] [io 3/6/1] [mem 3/6/1] [pci irq 10] [lat
32/176] [bus 2/5]
PCI card interrupts, PCI status changes
cs: memory probe 0xa0000000-0xa0ffffff: clean.
Linux ORiNOCO driver, Variant 1, Version 6.10, TurboCell 6.23r1.04
cs: IO port probe 0x0100-0x04ff: excluding 0x3b8-0x3df 0x4d0-0x4d7
cs: IO port probe 0x03e0-0x04cf: clean.
cs: IO port probe 0x04d8-0x04ff: clean.
cs: IO port probe 0x0800-0x08ff: clean.
cs: IO port probe 0x0a00-0x0aff: clean.
cs: IO port probe 0x0c00-0x0cff: clean.
license: XXXXXXXXXXXXXXXXXXXX // (I removed my licencenumber)
Unable to handle kernel NULL pointer dereference at virtual address 000003b3
printing eip:
d8897bd7
*pde = 00000000
Oops: 0002
CPU: 0
EIP: 0010:[<d8897bd7>] Tainted: P
EFLAGS: 00010202
eax: 000003b3 ebx: 00000000 ecx: 00000000 edx: 00000100
esi: c1640c00 edi: d705e000 ebp: d65e3818 esp: d65e3810
ds: 0018 es: 0018 ss: 0018
Process cardmgr (pid: 291, stackpage=d65e3000)
Stack: d705e000 00000000 d65e3848 d8899b95 00000000 00000100 d705e412
c02f6c3e
00000000 c0284a48 d65e3b28 d705e44c c1640c00 d705e000 d65e3b28
d8892448
d705e000 00000100 d65e3b08 d65e3994 c1646170 c1640c00 00000004
d887a197
Call Trace: [<d8899b95>] [<d8892448>] [<d887a197>] [<d889d8a6>]
[<d88831e9>]
[<d88882c0>] [<d88882c0>] [<d8886b65>] [<d887ab00>] [<d888739f>]
[<d887aca7>]
[<d887a04c>] [<d888739f>] [<d887a000>] [<d8879cca>] [<d887c0a5>]
[<d887abbc>]
[<d887aca7>] [<d887c225>] [<d88831e9>] [<d88882c0>] [<d88882c0>]
[<d8888208>]
[<d88915bd>] [<d88882c0>] [<d8878aef>] [<d887a197>] [<d8879d7a>]
[<d8891333>]
[<d887b216>] [<d889f2c0>] [<d88914e8>] [<d888c6be>] [<c015a05b>]
[<d888d180>]
[<c01ea1b2>] [<c0225dce>] [<c01e7564>] [<c01e8442>] [<c012ca12>]
[<c0124922>]

Im using pcmcia-cs-3.2.3 together with a modified orinocodriver from
TurboCell called wavelan2_cs.

I have tried to boot up with "Linux PCMCIA=no" and modprobed i82365 and
wavelan2_cs (with the same options as in the config files) and that went
well, but when I started cardmgr, I got the null pointer in dmesg again.

Is this a bug in pcmcia-cs or in the kernel, or perhaps in the driver I
got from my ISP...?

Hope anyone can help me with this!

regards
-- Emil

dhinds

unread,
Nov 11, 2002, 3:42:39 PM11/11/02
to
Emil Erlandsson <em...@buglix.org> wrote:

> Im using pcmcia-cs-3.2.3 together with a modified orinocodriver from
> TurboCell called wavelan2_cs.

> I have tried to boot up with "Linux PCMCIA=no" and modprobed i82365 and
> wavelan2_cs (with the same options as in the config files) and that went
> well, but when I started cardmgr, I got the null pointer in dmesg again.

> Is this a bug in pcmcia-cs or in the kernel, or perhaps in the driver I
> got from my ISP...?

My bet is on the modified orinoco driver.

-- Dave

Emil Erlandsson

unread,
Nov 11, 2002, 3:47:09 PM11/11/02
to
dhinds wrote:
> My bet is on the modified orinoco driver.

But are there not any other things I could test?

/Emil

dhinds

unread,
Nov 11, 2002, 5:26:14 PM11/11/02
to

How about trying the un-modified orinoco driver?

Did you get any errors/warnings when building the modified driver?

-- Dave

Emil Erlandsson

unread,
Nov 11, 2002, 5:34:42 PM11/11/02
to
dhinds wrote:

> How about trying the un-modified orinoco driver?

I must use the modified driver to be able to conntact my ISP. I have
tested with orinoco_cs without giving it any options, and it seems to work.

/Emil

Linux Doctor

unread,
Nov 11, 2002, 6:21:54 PM11/11/02
to
On Mon, 11 Nov 2002 at 11:34pm, Emil Erlandsson <em...@buglix.org> scribbled:

> I must use the modified driver to be able to conntact my ISP. I have
> tested with orinoco_cs without giving it any options, and it seems to work.

I don't get it.

What's in the standard driver that your ISP can't deal with? It seems
like what's in the driver should only be an issue between your card and
your access point, and totally transparent to the ISP. Why would the ISP
care, or worse yet, require something nonstandard? How does that even
work, given that he has no control or interest in your LAN specifics?

Neall

--
** Remove NOSPAM from return address when making personal reply **

Linux Doctor

unread,
Nov 11, 2002, 6:32:40 PM11/11/02
to
On Mon, 11 Nov 2002 at 4:21pm, Linux Doctor <NOSPAM...@haughtmail.com>...:

> On Mon, 11 Nov 2002 at 11:34pm, Emil Erlandsson <em...@buglix.org> scribbled:
>
> > I must use the modified driver to be able to conntact my ISP. I have
> > tested with orinoco_cs without giving it any options, and it seems to work.
>
> I don't get it.
>

Ok, I've answered my own question. If you're using that Turbocell ISP
base unit, then their web page says it's "proprietary and not part of the
802.11b standard." So, you need their turbocell driver. If you can
install the standard orinoco_cs driver, but it won't connect to your base,
and their proprietary driver you've compiled doesn't load, sounds like you
need to hash that out with Turbocell :) Or buy an 802.11b base unit and
use all the standard drivers :) Neall

dhinds

unread,
Nov 11, 2002, 8:28:25 PM11/11/02
to
Emil Erlandsson <em...@buglix.org> wrote:
> dhinds wrote:

>> How about trying the un-modified orinoco driver?

> I must use the modified driver to be able to contact my ISP. I have

> tested with orinoco_cs without giving it any options, and it seems to work.

Well it seems like that pins the problem down pretty well.

-- Dave

0 new messages