From Bugzilla Helper:
User-Agent: Mozilla/4.78 [en] (X11; U; Linux 2.4.9-7 i686)
Description of problem:
When I try to use my USB webcam (Philips' ToUcam pro) the Kernel crashes.
When loading the kernel module (drivers/usb/pwc.o) I get all the good
messages, but when trying to read from the device (/dev/video0) the kernel
panics leaving me reseting the system. I have added (the carefully written
down) oops as an attachment.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. modprobe pwc
2. cat /dev/video0 > /dev/null
or start a program that tries to read from the webcam
Actual Results: System freezes. When done at the console, system freezes
and displays the messages I added as an attachment (kernel panic, etc).
Expected Results: The system should have shown a nice live feed of me ;)
USB Controller: Intel Corporation 82371AB PIIX4 USB (rev 01)
When using the 2.2.19-7.0.1 kernel and the module version usb-pwc-6.3 it
Created attachment 36303 [details]
Kernel oops of crash.
This is with 2.4.9-7 ? We released an update kernel which has, amongst other
things, a null pointer derefence in usb cameras fixed.... The version is
Thanks for the oops trace, what about EIP? I cannot
find it in the trace. Is in process_iso?
Another thing that you can do is to download this:
It should work.
I installed the 2.4.9-13 kernel and that helped, thanks!
About the EIP, I can't find it. The kernel crashed real hard, so the only way
for me to get some info was by typing at the console
$cat /dev/video0 > /dev/null
Then the kernel messages started coming on screen, but only the end stays
visible. I think the EIP must have scrolled of screen. If you have any ideas to
get it in an other way, please tell me, I am willing to crash my system ones
It seems to be the case of infamouse list_del() in usb-uhci
- list_del (p);
+ p_tmp = p;
p = p->next;
+ list_del (p_tmp);
delete_desc (s, desc);
It went into Linus tree at 2.4.10, so I thought that 2.4.7-ac
would not be vulnerable. But apparently it was.
I was too hasty with the Rawhide, we do have an update
out which fixes the problem within the 7.2 trunk.
I just verified that it does contain the fix
(actually, I checked out the kernel_2_4_9-13 branch)