PCMCIA on Mandrake 9.1
 

I'm running Mandrake 9.1 on a Medion 5275 laptop.

Card services does not seem to be running at all. No beeps, no nothing.

I have installed the linux_wlan software, and followed the instructions to configure it.

When I restart the network (/etc/rc.d/init.d/pcmcia restart), I get an error:

PCIC module not defined in startup options!

Any thoughts?
Leon

I've been looking through the other threads (once I decided to search on the PCIC error instead of the problem) and I'm pre-emptively providing some info:

My assumption is that the problem is that card services is not recognized (output of dmesg), but I'm not sure how to proceed from here.

NOTE: For those who want to jump to the next output, search for "***==" and that will get you there.

Thanks again.
Leon

lspci output:
***============
00:00.0 Host bridge: Intel Corp. 82845 845 (Brookdale) Chipset Host Bridge (rev 04)
00:01.0 PCI bridge: Intel Corp. 82845 845 (Brookdale) Chipset AGP Bridge (rev 04)
00:1e.0 PCI bridge: Intel Corp. 82801BA/CA/DB PCI Bridge (rev 05)
00:1f.0 ISA bridge: Intel Corp. 82801BA ISA Bridge (LPC) (rev 05)
00:1f.1 IDE interface: Intel Corp. 82801BA IDE U100 (rev 05)
00:1f.2 USB Controller: Intel Corp. 82801BA/BAM USB (Hub #1) (rev 05)
00:1f.3 SMBus: Intel Corp. 82801BA/BAM SMBus (rev 05)
00:1f.4 USB Controller: Intel Corp. 82801BA/BAM USB (Hub #2) (rev 05)
00:1f.5 Multimedia audio controller: Intel Corp. 82801BA/BAM AC'97 Audio (rev 05)
01:00.0 VGA compatible controller: ATI Technologies Inc Radeon Mobility M6 LY
02:03.0 FireWire (IEEE 1394): Texas Instruments TSB43AB21 IEEE-1394a-2000 Controller (PHY/Link)
02:05.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139/8139C/8139C+ (rev 10)
02:07.0 Communication controller: Ambient Technologies Inc HaM controllerless modem (rev 02)
02:09.0 CardBus bridge: Texas Instruments PCI1250 PC card Cardbus Controller (rev 01)
02:09.1 CardBus bridge: Texas Instruments PCI1250 PC card Cardbus Controller (rev 01)

lsmod output
***==============
Module Size Used by Not tainted
isofs 27988 1 (autoclean)
zlib_inflate 21156 0 (autoclean) [isofs]
nls_iso8859-1 3516 1 (autoclean)
udf 90464 0 (autoclean)
sg 34636 0 (autoclean)
radeon 107428 1
agpgart 40896 3 (autoclean)
parport_pc 25096 1 (autoclean)
lp 8096 0 (autoclean)
parport 34176 1 (autoclean) [parport_pc lp]
i810_audio 26248 0
soundcore 6276 0 [i810_audio]
ac97_codec 12488 0 [i810_audio]
nfsd 74256 0 (autoclean)
af_packet 14952 1 (autoclean)
sr_mod 16920 2 (autoclean)
floppy 55132 0
8139too 17160 1 (autoclean)
mii 3832 0 (autoclean) [8139too]
ohci1394 19048 0 (unused)
ieee1394 45900 0 [ohci1394]
supermount 15296 2 (autoclean)
ide-cd 33856 0
cdrom 31648 0 [sr_mod ide-cd]
ide-scsi 11280 1
scsi_mod 103284 3 [sg sr_mod ide-scsi]
usb-uhci 24652 0 (unused)
usbcore 72992 1 [usb-uhci]
rtc 8060 0 (autoclean)
ext3 59916 2
jbd 38972 2 [ext3]

dmesg output
command is: dmesg|grep -A 20 'Linux Kernel Card Services'
***==============================================
Linux Kernel Card Services Kernel Version
options: [pci] [cardbus] [pm]
ds: no socket drivers loaded!
unloading Kernel Card Services
eth0: Setting 100mbps full-duplex based on auto-negotiated partner ability 41e1.
eth0: Setting 100mbps full-duplex based on auto-negotiated partner ability 41e1.

Contents of /etc/modules.conf
(note there is a blank line at the *start* of this file)
***=====================
probeall scsi_hostadapter ide-scsi
alias eth0 8139too
alias sound-slot-0 i810_audio
probeall usb-interface usb-uhci
alias ieee1394-controller ohci1394

Your pcmcia bridge is a pretty generic Texas Instruments one...

In /etc/rc.d/init.d/pcmcia there is a section for what modules to try and load, it'll look like (but not identical) to this:

Your bus should work with "yenta_socket", you may want to comment out the line that sets it to probe and uncomment the one that says yenta_socket.

That should get the bugger to load.

Cheers,

Finegan

Hmmmm.... It's not looking clear at the moment. What I have in /etc/rc.d/init.d/pcmcia that relates to "yenta" is:

101 if [ -x /sbin/modprobe ] ; then
102 /sbin/modprobe pcmcia_core $CORE_OPTS || break
103 /sbin/modprobe $PCIC $PCIC_OPTS >/dev/null 2>&1 ||
104 (/sbin/modprobe yenta_socket >/dev/null 2>&1 &&
105 echo "using yenta_socket instead of $PCIC") ||
106 /sbin/modprobe $PCIC $PCIC_OPTS || break
107 /sbin/modprobe ds || break
108 elif [ -d $PC ] ; then
109 /sbin/insmod $PC/pcmcia_core.o $CORE_OPTS
110 /sbin/insmod $PC/$PCIC.o $PCIC_OPTS
111 /sbin/insmod $PC/ds.o
112 else
113 echo "module directory $PC not found."
114 break
115 fi

.... and then lower down:
145 if grep -q "ds " /proc/modules ; then
146 /sbin/rmmod ds
147 /sbin/rmmod $PCIC 2>/dev/null || \
148 /sbin/rmmod yenta_socket 2>/dev/null
149 /sbin/rmmod pcmcia_core
150 fi

on mandrake the file he is referring to is /etc/sysconfig/pcmcia.

That file (/etc/sysconfig/pcmcia) only contains the line:

PCMCIA=yes

That might indicate that this file is either truncated beyond recognition (which would make sense), or it's not the droid we are looking for. Any way to pull the original from source?

I did a grep through every file on the entire file system for "yenta" and, aside from the files mentioned here (plus what appears to be some template pcmcia files - ie: k96pcmcia, s45pcmcia, etc).

I also found a reference to "disoriyenta" (when Aunt Sadie forgets if she is gossiping about the Berkowitzs or the Gluckmans), but I am sure that won't help us.

the command I used was:find . -exec grep -i "yenta" {} \; -ls >yentafind.txt



Thanks again for your assistance.

And of course all my laptops are running Slackware or a BSD right now...

Beavis got the right file, they just leave you with null values...

PCMCIA=yes
PCIC=probe
PCIC_OPTS=
CORE_OPTS=

or make PCIC=yenta_socket

Try them all, see what happens... i82365 is the last option...

This might take some OPTS calls too, they're documented at the pcmcia-cs sourceforge site, if none of this works I'll page through them too.

Cheers,

Finegan

Finegan, if your first name is "Bridget", "Kendra" or "Moira" then I could just kiss you. If, on the other hand, it's "Clancy", "Kyle" or "Shamus" then a handshake will have to do.

Either way I owe you a pint!

Cheers to akaBeavis for giving the right file and just assuming the stuff that was supposed to be there was there. Plus he's put up with a lot of the aggravation from a few other related posts (how do you log in as root again?).

Just to clarify, I put in the default settings (with PCIC=probe) and ran /etc/rc.d/init.d/pcmcia restart and the system was able to find the card. I can't test actual connectivity right now, since I'm not home. But the wg-lan stuff was pretty clear on how to set WEP, etc.

Thanks again.

I'm glad you got this working, just out of curiosity, when you do an lsmod, does yenta_socket show up? This info may be helpful to others in the future.

In response, yes, yenta_socket does indeed show up.

I failed to mention that, when I restarted PCMCIA, I got the following output:

Shutting down PCMCIA services: done.
Starting PCMCIA services: using yenta_socket instead of probe
cardmgr[26861]: watching 2 sockets
done.

So you can see that yenta (and who was Jewish writing these drivers?) was started from the beginning.

Thanks again.
Leon

Actually Linus himself handles that part of things, be reverent, be very, very reverent, :)

ps:
" I did a grep through every file on the entire file system for "yenta" and, aside from the files mentioned here (plus what appears to be some template pcmcia files - ie: k96pcmcia, s45pcmcia, etc)."

so aren't you wondering where in the heck yenta_socket is/got loaded from?

Excellent... hmm, if it says instead of probe then /etc/sysconfig/pcmcia must be overriding something set elsewhere, I need to install one of the sysV monkeys again before I lose all my skills.

Anyway, Linus actually wrote most of yenta_socket driver from scratch, pcmcia used to be handled as an out-of-kernel package, which still exists for user-side binaries and keeeerap, pcmcia-cs. I have no idea what its named after, knowing the usual tack, probably something arcane, technical and humourless.

Cheers,

Finegan

To akaBeavis, yes I am wondering where it comes from it there isn't a file or driver by that name.

To Finegan, I just assumed it was for the yiddish word "yenta", which is a neighborhood busybody.

So continuing on that tack, yenta is the driver that butts in if all the other drivers can't get loaded.

Just my two shekels.
Leon

Nor is there any file by that name (yenta_socket) on any of my systems either, regardless of distro, I'm gonna go out on a limb here and say it's compiled into the kernel itself. What do you think?

It can be, I've got it as a module:

I was just looking for what kernel option it takes to make yenta a modules, couldn't really find it that easily.

Cheers,

Finegan