Bug 357400

This is something we never tried, unfortunately. Do you have a printer attached to it?

Shouldn't it be the parport_pc driver?
The parport_pc driver itself is also broken..., but you might have more luck with this one.
As a rule of thumb, you should always try without dma (dma=none param iirc).
Also yast afaik is playing with tricks like un- and reloading parport_pc.
Can you also do:
for x in `find /sys |grep pnp|grep resources`; do echo $x; cat $x;done
(or similar) and search for the PNP device exporting parport resources (should be 0x3F8 or 0x2F8 for IO ports?).

yes, the module is parport_pc.

suspend to ram in the docking station,
resume outside docking station,
--> /dev/parport0 still exists.
insert into docking station
->

open("/dev/parport0", O_RDWR)           = 3
ioctl(3, PPCLAIM, 0xb7e77140)           = 0
ioctl(0, PPRDATA, 0xbfeab4fb)           = -1 EINVAL (Invalid argument)
ioctl(0, PPRCONTROL, 0xbfeab4fb)        = -1 EINVAL (Invalid argument)
ioctl(0, PPRSTATUS, 0xbfeab4fb)         = -1 EINVAL (Invalid argument)
ioctl(3, PPRDATA, 0xbfeab507)           = 0
ioctl(3, PPRCONTROL, 0xbfeab507)        = 0

but parport hardware does not communicate.


(-> serial port on the docking station is also dead then.)
after a reboot in the docking station, parport is functional.
while still in the docking station: suspend to ram, resume
parport0 is still functional. Good.


/sys/devices/pnp0/00:0c/resources
state = active
io 0x2f8-0x2ff
irq 3
dma 1
/sys/devices/pnp0/00:0b/resources
state = active
io 0x3bc-0x3be
irq 7
/sys/devices/pnp0/00:0a/resources
state = active
io 0x3f8-0x3ff
irq 4
...


after a simple undock, redock, it is dead.
(I pressed the undock button and waited for the red LED to go off.)
parport_pc remains loaded, /dev/parport0 remains present.
# rmmod parport_pc
# modprobe parport_pc dma=none
brings the port back to life.

undock, reboot.
/dev/parport0 is not there, parport_pc is present.
dock.
/dev/parport0 is not there, parport_pc is present.
# rmmod parport_pc
# modprobe parport_pc dma=none
# dmesg
...
ACPI: Fatal opcode executed
pnp: Device 00:0b activated.
pnp: Device 00:0b disabled.
parport_pc: probe of 00:0b failed with error -22

/dev/parport0 is still not there.
/sys/devices/pnp0/00:0b/resources
state = disabled
io disabled
irq 15

This just sounds that the initialization code of parport_pc is only run on boot, and if dock ist present. So this maybe is just a lack of hotplugging functionality in the driver.

However, the ACPI errors look bad. Thomas, any idea what's going on here?

You mean this:
ACPI: Fatal opcode executed

This looks indeed bad.
Juergen, can you confirm that his happens after un- and reloading parport_pc, pls.
Also attach acpidump output (as text mime).

Maybe we should also continue on a more recent kernel.
This probably won't get fixed for 10.3 (unless it is a really easy one).
It should be enough to install a kotd on 10.3.
Even better would be a 11.0 installation if you still have a small partition left...
Hmm, not sure we can also continue on 10.3 for now..., if you wanted to give 11.0 a test anyway we should go on there :)

When you know when the "ACPI: Fatal opcode executed" happens (I expect when docking in?) you might also want to increase ACPI debug level:
before unload disturbing, polling modules:
thermal and ac
echo 0x1F >/sys/module/acpi/parameters/debug_level
Then trigger the message and set the debug level back to default (to not get unrelated stuff in there):
echo 0x3 >/sys/module/acpi/parameters/debug_level
You should now have much more messages in /var/log/messages, pls attach.

Yes, please test on current FACTORY. We have enough issues to get docking support working at all.

I'm still going to try the parport myself, so you don't need to do it if you would have to install a 11.0 first.

I cannot reproduce on 11.0a2
/dev/parport0 never appears. Reboot in docking station, docker dock, 
rmmod/ modprobe  -- nothing helps

There are no ACPI errors in dmesg, though

Oh mit 11.0? Da koennte evtl. pnpacpi=off helfen?

*ping*

*** Bug 363199 has been marked as a duplicate of this bug. ***

Too late for 11.0, sorry.

I no longer use the docking station during work.
I switched back to 10.3 after testing 11.0.
Cannot the needed info. Sorry.

see comment 16, resetting needinfo.

Test result with 11.0 was not provided. Resetting NEEDINFO ;-) Best would be to try 11.1 now.

I have e1000e on that machine. Will 11.1 shred my eth0?

Not sure what's the exact status here. Maybe you should wait a little more. Or make sure to store 'ethtool -e eth0' before installing to make sure to be able to recover it.

11.1 will not anymore shred your eth0, so go ahead and test it...

booted 11.1 beta5 without docking station.
no parport as expected.
inserted into dockingstation. Still no parport

lsmod | grep par
has nothing.

modprobe parport_pc
lsmod | grep par
parport_pc           51516  0
parport              50388  2, ppdev,parport_pc


undock, dock, still no device.

booted SLED11-RC1 without docking station
No parport as expected.
Inserted into dockingstation. Still no parport, but

$ lsmod | grep par
parport_pc             35048  0 
parport                33832  3 lp,parport_pc,ppdev
$ 

undock, dock, still no device.

Any progress with 11.2? Anyway, I'm sorry, but there are no resources for this left.