[ltp] T41 - kernel 2.6.11 - suspending problem - "Unlink after no-IRQ?"

[Michael Perry]

On Mon, Mar 07, 2005 at 12:58:16PM -0600, Alejandro Bonilla wrote:
> 
> 
> François Fleuret wrote:
> 
> >Dear All,
> >
> >I have been using the 2.6.9 kernel on my T41p for month without a
> >glitch (Debian Sarge) and have been giving a try with the 2.6.11 for a
> >few days. Everything seems okay except one problem with the suspending
> >(S3). When I close the lid, it happened twice that the computer freeze
> >without going into suspend at all. Looking at the log it seems that
> >both times, I got the following line in /var/log/syslog:
> >
> >kernel: uhci_hcd 0000:00:1d.1: Unlink after no-IRQ?  Controller is 
> >probably using the wrong IRQ.
> >
> > 
> >
> Try to unload all USB related, check lsmod and modprobe -r it. Test 
> again and see if the same thing happens.
> 
> Also a nice thing would be to cat /proc/interrupts and see if you get 
> the same thing in 2.6.10 or 2.6.11 if you still have the previous kernel 
> loaded around...
> 

In the past, I've moved to a new acpi diff patch which has solved some
of this for me.  In one case, I had pcmcia cards now allowing my T23 to
do acpi suspend events correctly on 2.6.9.  I downloaded new acpi diff
patches against that kernel and things got much better.  On 2.6.10, I
had a few interesting moments with acpi suspends also so I patched it
with the latest diffs.  

I just checked 2.6.11 on acpi.sf.net and there are not cumulative
patches (yet) for that kernel.  I have had to play with what gets
unloaded in a suspend script and find the suspect thing or things and
then use the suspend script to remove them and then re-insert them
after.

You may want to look at how the /etc/acpi scripts are set and add in
lines to remove certain modules which will not suspend correctly.  One
tool is to use lspci and compare it against the report from the kernel.


-- 
Michael Perry | do or do not. There is no try. -Master Yoda
mperry@lnxpowered.org | http://www.lnxpowered.org