[gst-devel] Oops: kernel panic caused by gstv4lsrc

[Bastien Nocera]

Hi,

On 24 Apr 2001 16:45:17 -0700, Erik Walthinsen wrote:
> With an ov511 (usb webcam), I get the following oops on my 2.4.3 laptop:

It used to work on my ov511, on PPC !

> Apr 24 17:37:04 omicron kernel: Unable to handle kernel NULL pointer
> dereference at virtual address 00000004
> Apr 24 17:37:04 omicron kernel:  printing eip:
> Apr 24 17:37:04 omicron kernel: c0110e9b
> Apr 24 17:37:04 omicron kernel: *pde = 00000000
> Apr 24 17:37:04 omicron kernel: Oops: 0002
> Apr 24 17:37:04 omicron kernel: CPU:    0
> Apr 24 17:37:04 omicron kernel: EIP:    0010:[interruptible_sleep_on+43/96]
> Apr 24 17:37:04 omicron kernel: EFLAGS: 00210086
> Apr 24 17:37:04 omicron kernel: eax: cc8528b8   ebx: cc8528bc   ecx:
> cadadd6c   edx: 00000000
> Apr 24 17:37:04 omicron kernel: esi: 00200286   edi: cc852800   ebp:
> cadadd7c   esp: cadadd64
> Apr 24 17:37:04 omicron kernel: ds: 0018   es: 0018   ss: 0018
> Apr 24 17:37:04 omicron kernel: Process lt-gstreamer-la (pid: 1056,
> stackpage=cadad000)
> Apr 24 17:37:04 omicron kernel: Stack: 00000000 cadac000 c01a6cf9 cb383000
> ffffffff cadac000 ffffffe7 d091ad26
> Apr 24 17:37:04 omicron kernel:        00200282 cb383984 00000001 00200282
> 00000001 cb383980 cadaded8 cb26cf22
> Apr 24 17:37:04 omicron kernel:        c01a6cf9 cb383000 cadaddb4 cadaddb4
> 00000000 00000000 000071c6 cadac000
> Apr 24 17:37:04 omicron kernel: Call Trace: [n_tty_receive_buf+2745/2800]
> [<d091ad26>] [n_tty_receive_buf+2745/2800] [process_timeout+0/80]
> [__insert_vm_struct+227/384] [__alloc_pages+228/720]
> [remap_page_range+225/672]
> Apr 24 17:37:04 omicron kernel:        [pty_write+285/304] [<d0ae3000>]
> [<d091b3e4>] [insert_vm_struct+25/48] [do_mmap_pgoff+872/1056]
> [old_mmap+195/256] [old_mmap+232/256] [<d091323d>]
> Apr 24 17:37:04 omicron kernel:        [sys_ioctl+375/400] [system_call+51/56]
> Apr 24 17:37:04 omicron kernel:
> Apr 24 17:37:04 omicron kernel: Code: 89 4a 04 89 55 f0 89 5d f4 89 48 04
> e8 c4 fa ff ff fa 8b 55

Run it through ksymoops to get something useful.

> This is obviously something wrong with the way v4lsrc interacts with the
> kernel v4l subsystem...  If I have time I'll try to trace it down some,

Nope, v4lsrc just triggers a bug in the ov511 driver. Write to the kind
maintainer, he's been helpful with me (because we arrived to the
conclusion that the PPC modules were not very reliable), which is
obviously not the case.

> but I don't know the v4lsrc source well enough....

Cheers


-- 
/Bastien Nocera
http://hadess.net