[RULE] PCMCIA problems in Red Hat?

C David Rigby cdrigby at 9online.fr
Sun May 30 10:51:42 EEST 2004


I ran into something very similar in setting up FC2 on the one notebook 
computer I have (Thinkpad 765L, P166MMX, 104MB RAM) that has sufficient 
RAM to run Anaconda (well, in text mode - graphics mode fails).  The 
issue for me was twofold: the 2.6.5 kernel's yenta_socket driver does 
not handle the IRQs well for this machine, and the init scripts start 
the network before pcmcia is started => the hardware for eth0 is not yet 
initialized at the time /etc/init.d/network is run.

For the first, what is required is to run modprobe against the 
yenta_socket driver twice!  The order of modprobe commands looks like this:

modprobe pcmcia
modprobe yenta_socket
modprobe yenta_socket pci=biosirq
modprobe ds

 From there, cardmgr can be run and pcmcia will work.  No combination of 
PCIC_OPTS in /etc/sysconfig/pcmcia or parmeters passed to the kernel 
allowed me to avoid running modprobe twice against yenta_socket.

What really needs to be done is to patch yenta_socket.c so that it works 
correctly.  However, the following ugly hack got the job done:

1) Edit /etc/sysconfig/pcmcia so that it looks like this:

PCMCIA=yes
PCIC=yenta_socket
PCIC_OPTS="pci=biosirq"
CORE_OPTS=

2) The kernel does recognize that it needs to use the yenta_socket.o 
module, but the script /etc/init.d/pcmcia only runs modprobe once with 
the options from /etc/sysconfig/pcmcia.  So, I editted 
/etc/init.d/pcmcia to call modprobe twice.  Once I learn how to use 
patch correctly, I will just include the diff results here in my emails. 
  For now, my /etc/init.d/pcmcia is attached to this post.

3) Reboot, and yenta_socket + cardmgr have the pcmcia slots under 
control.  But, we are now at the second problem, which is that the 
network is not operational.  In my case, I simply changed the order in 
which the scripts /etc/init.d/network and /etc/init.d/pcmcia are run at 
init time by changing the names of the links to the start up scripts 
that are found in /etc/rc2.d, /etc/rc3.d, /etc/rc4.d and /etc/rc5.d.  I 
did not keep a record of what the old links were (oops) but the new ones 
look like this:

[cdrigby at blackbox cdrigby]$ ls -l /etc/rc3.d/S22pcmcia
lrwxrwxrwx  1 root root 16 May 25 08:48 /etc/rc3.d/S22pcmcia -> 
../init.d/pcmcia
[cdrigby at blackbox cdrigby]$ ls -l /etc/rc3.d/S23network
lrwxrwxrwx  1 root root 17 May 25 08:48 /etc/rc3.d/S23network -> 
../init.d/network
[cdrigby at blackbox cdrigby]$

The series of commands that will do this are (where rcN.d is each of N = 
2 to 5 - i.e., carry out this operation four times; also XX and YY are 
whatever the numbers for the startup script links are after a clean 
installation for SXXnetwork and SYYpcmcia, I simply failed to record the 
initial names as I was beating on the configuration):

cd /etc/rcN.d/
rm SXXnetwork
rm SYYpcmcia
ln -s ../init.d/pcmcia S22pcmcia
ln -s ../init.d/network S23network

4) Reboot with your network card installed, and you should have eth0 
available and configured once you log in.

5) Let me know if this does not work, and we can try something else!

CDR

mfioretti wrote:

 >PCI: No IRQ known for interrupt pin A of device 00:03.0. Please try
 >>> using pci=biosirq.
 >>> PCI: No IRQ known for interrupt pin B of device 00:03.1. Please try
 >>> using pci=biosirq.
 >>> Yenta IRQ list 0a80, PCI irq0
 >>> Socket status: 30000006
 >>> Yenta IRQ list 0a80, PCI irq0
 >>> Socket status: 30000006

 >
 >The PCI errors are due to a bug/limitation in the kernel PCI subsystem
 >that is pretty much fatal for PCMCIA.  The pcibios suggestion is
 >rarely useful.

bugler at teuton.org wrote:
> Quoting "M. Fioretti" <mfioretti>:
> 
> 
>>Gabriel, what did you mean exactly by "not working"?
> 
> 
> working:  ifconfig shows eth0 configured to the PCMCIA device.
> non-working:  ifconfig only shows lo working.
> 
> In my case, eth0 (the PCMCIA NIC) isn't even configured on the first boot to RedHat.
> 
> I was able to get it up and running by hand using the "ideas" behind the scripts
>  (esp. with David's help).  (Namely PCIC=i365_base=0xfcfc in the pcmcia.sh
> script.)  Then I was able to ftp again to my desktop.
> 
> I haven't done the digging to see how to get RedHat to configure it every time
> on boot-up.  I know that it's not rocket science, though.
> 
> gotta go...
> 
> -Gabriel
> 
> 
> 
> _______________________________________________
> Original home page of the RULE project: www.rule-project.org
> Original Rule Development Site http://savannah.gnu.org/projects/rule/> 
Original RULE mailing list: Rule-list at nongnu.org, hosted at http://mail.nongnu.org/mailman/listinfo/rule-list
> http://lists.nongnu.org/mailman/listinfo/rule-list
> 
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: pcmcia
URL: <http://lists.hellug.gr/pipermail/rule-list/attachments/20040530/36cfb8bd/attachment.asc>
-------------- next part --------------
_______________________________________________
Original home page of the RULE project: www.rule-project.org
Original Rule Development Site http://savannah.gnu.org/projects/rule/
Original RULE mailing list: Rule-list at nongnu.org, hosted at http://mail.nongnu.org/mailman/listinfo/rule-list
http://lists.nongnu.org/mailman/listinfo/rule-list


This full static mirror of the Run Up to Date Linux Everywhere Project mailing list, originally hosted at http://lists.hellug.gr/mailman/listinfo/rule-list, is kept online by Free Software popularizer, researcher and trainer Marco Fioretti. To know how you can support this archive, and Marco's work in general, please click here