Re: Unloading cx8802 results in crash of ir_core:ir_unregister_class

[Andy Walls <awalls@md.metrocast.net>]

On Wed, 2010-10-20 at 12:31 +0200, Hans-Peter Jansen wrote:
> Hi Mauro,
> 
> since you got your fingers dirty with ir, I would like to direct this 
> issue towards you. 


> Unfortunately, unloading all the dvb drivers is not possible, because
> unloading cx8802 leads to this oops:
> 
> Welcome to openSUSE 11.1 - Kernel 2.6.34.7-4-pae (ttyS0)
> [111047.940641] BUG: unable to handle kernel paging request at 313a3030
> [111047.953710] IP: [<c03654fe>] sysfs_remove_group+0x7e/0xd0
> [111047.964838] *pdpt = 0000000029f53001 *pde = 0000000000000000 
> [111047.976687] Oops: 0000 [#1] SMP 
> [111047.983453] last sysfs file: /sys/devices/system/cpu/cpu7/cache/index2/shared_cpu_map
> [111047.999559] Modules linked in: ip6t_LOG ipt_MASQUERADE xt_pkttype xt_TCPMSS xt_tcpudp ipt_LOG xt_limit iptable_nat ]
> [111048.204691] 
> [111048.207843] Pid: 10029, comm: rmmod Not tainted 2.6.34.7-4-pae #1 P7F-E/System Product Name
> [111048.224918] EIP: 0060:[<c03654fe>] EFLAGS: 00210206 CPU: 0
> [111048.236201] EIP is at sysfs_remove_group+0x7e/0xd0
> [111048.246007] EAX: ef4ff008 EBX: 313a3030 ECX: f0600000 EDX: 00000000
> [111048.258879] ESI: 00000000 EDI: efb7e71c EBP: ef4ff008 ESP: e3ee7e64
> [111048.271716]  DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
> [111048.282668] Process rmmod (pid: 10029, ti=e3ee6000 task=eaf00e00 task.ti=e3ee6000)
> [111048.298133] Stack:
> [111048.302319]  f02cabe0 00000041 ef5208c0 ef49ac00 00200286 efb7e700 ef3f0000 00000000
> [111048.318132] <0> ef329800 f8f84932 efb7e700 ef3f0000 f8f84044 00000000 f8ffffff f015f1c0
> [111048.334811] <0> ef329800 f917c655 00000000 f8000000 f9178dd9 01000000 00000000 f8ffffff
> [111048.351939] Call Trace:
> [111048.357213]  [<f8f84932>] ir_unregister_class+0x32/0x60 [ir_core]
> [111048.369821]  [<f8f84044>] ir_input_unregister+0x44/0x90 [ir_core]
> [111048.382338]  [<f917c655>] cx88_ir_fini+0x25/0x50 [cx88xx]
> [111048.393437]  [<f9178dd9>] cx88_core_put+0xb9/0x140 [cx88xx]
> [111048.404834]  [<f91e8795>] cx8802_remove+0x82/0x169 [cx8802]
> [111048.416319]  [<c041d966>] pci_device_remove+0x16/0x40
> [111048.426610]  [<c04b214d>] __device_release_driver+0x6d/0xd0
> [111048.437993]  [<c04b222f>] driver_detach+0x7f/0x90
> [111048.447872]  [<c04b11fa>] bus_remove_driver+0x7a/0x100
> [111048.458687]  [<c041db6e>] pci_unregister_driver+0x2e/0x80
> [111048.469658]  [<c0279669>] sys_delete_module+0x179/0x240
> [111048.480262]  [<c020320c>] sysenter_do_call+0x12/0x22
> [111048.490357]  [<ffffe424>] 0xffffe424
> [111048.497680] Code: f0 ff 0e 0f 94 c0 84 c0 75 0b 83 c4 14 5b 5e 5f 5d c3 8d 76 00 83 c4 14 89 f0 5b 5e 5f 5d e9 ca e 

These code bytes are truncated; there are only 32 here.  There are
supposed to be 64 of them, with the Oops-ing instruction being around
the 48th byte.  

Please review your logs and provide the full Ooops.


> [111048.537852] EIP: [<c03654fe>] sysfs_remove_group+0x7e/0xd0 SS:ESP 0068:e3ee7e64
> [111048.552782] CR2: 00000000313a3030
> [111048.559854] ---[ end trace d7231a8e672c4adc ]---
> 
> Any ideas, what's going wrong here?

I don't. Mauro might since he knows the code better.  I just decode
Oops'es every once in a while.

Regards,
Andy

> Pete