[Techtalk] Second try before reinstall--Kernel Oops

[Telsa Gwynne]

On Wed, Dec 31, 2003 at 08:53:09AM -0500 or thereabouts, Beth E. Johnson wrote:
> Second try: I might just wipe the drive anyway because of all the little
> prolems I'm having with Fedora (speaker applet doesn't control sound,
> printing's dead, clie's not syncing).
> 
> I'm getting these messages via tail -f /var/log/messages when my Clie
> disconnects.  I've been trying to get it syncing up with Evolution after
> doing a clean install of Fedora Core 1 (the upgrade from RH 9 worked
> fine as far as sync went).  Can someone explain to me what is really
> going on and how I can reset it, so to speak, without rebooting?  The
> sync did complete, but if I try hotsyncing again it times out and no
> further kernel messages are logged, at least related to the modules. 
> Which team should the bug report go to?  gpilot?  RH/Fedora on the
> kernel?

I asked my husband, who understands kernels. He says, "It is a
bug in the Fedora kernel, and is the fault of the linux 2.4 kernel. 
It is fixed in linux 2.6."

There isn't much you can do about this beyond reporting it to
Red Hat's bugzilla. If you have a serial cable for the thing
instead of a USB cable, that will (should) work. And if this
worked on RH 9.0, then you should be able to use the RH 9.0
kernel on Fedora.

His summary of the log is:

"Insert device, kernel spots it and tries to talk to it..."
> SLOBBIT kernel: hub.c: new USB device 00:1f.2-1, assigned address 3
> SLOBBIT kernel: usbserial.c: Handspring Visor / Treo / Palm 4.0 / Clie
> 4.x converter detected
> SLOBBIT kernel: usb-uhci.c: interrupt, status 2, frame# 738
> SLOBBIT kernel: usb_control/bulk_msg: timeout
> SLOBBIT kernel: visor.c: visor_startup - error -110 getting connection
> info

"Kernel setting up the devices.."
> SLOBBIT kernel: usbserial.c: Handspring Visor / Treo / Palm 4.0 / Clie
> 4.x converter now attached to ttyUSB0 (or usb/tts/0 for devfs)
> SLOBBIT kernel: usbserial.c: Handspring Visor / Treo / Palm 4.0 / Clie
> 4.x converter now attached to ttyUSB1 (or usb/tts/1 for devfs)

"Kernel having trouble talking to it"
> SLOBBIT kernel: usb-uhci.c: interrupt, status 3, frame# 251
> SLOBBIT usb.agent[6388]: missing kernel or user mode driver visor
> SLOBBIT kernel:  hdb: hdb4

(btw, "hdb is unrelated", he says.)

> SLOBBIT last message repeated 3 times
> SLOBBIT devlabel: devlabel service started/restarted
> SLOBBIT kernel: usb-uhci.c: interrupt, status 2, frame# 1102
> SLOBBIT kernel: usb_control/bulk_msg: timeout
> SLOBBIT kernel: usb_control/bulk_msg: timeout
> SLOBBIT kernel: visor.c: Bytes In = 2181  Bytes Out = 669
> SLOBBIT kernel: usb-uhci.c: interrupt, status 3, frame# 1830

"It disconnects.."
> SLOBBIT kernel: usb.c: USB disconnect on device 00:1f.2-1 address 3
> SLOBBIT kernel: visor.c: Bytes In = 42  Bytes Out = 0

"And the kernel crashes: this is the Linux equivalent of a blue screen" 
> SLOBBIT kernel: Unable to handle kernel NULL pointer dereference at
> virtual address 00000998
> SLOBBIT kernel:  printing eip:
> SLOBBIT kernel: e09a26bc
> SLOBBIT kernel: *pde = 1f55e067
> SLOBBIT kernel: *pte = 00000000
> SLOBBIT kernel: Oops: 0002
[snip snip munch]

Incidentally, printing not working: I had some printing trouble
with Fedora Core 1. The printer had been working fine on RH 9.0,
but it wasn't using CUPS. The upgrade to Fedora included a 
migration to CUPS and suddenly it wouldn't print my daily crossword
for me, although I am sure it did print the test page. 

Eventually, I just ran the redhat-setup-print* (whatever it's called:
I just hit [tab] to complete the command name!) and went through 
each section of it. That worked. 

Telsa