2002-11-02 18:37:32

by Jochen Friedrich

[permalink] [raw]
Subject: [BUG] USB Kernel bug in 2.5.45

Hi,

when running "rmmod hid" while a USB HID device is still connected (but
not used), i get this Oops messages:

drivers/usb/core/usb.c: deregistering driver hiddev
drivers/usb/core/usb.c: deregistering driver hid
devfs_put(fffffc0009393000): poisoned pointer
Forcing Oops
Kernel bug at fs/devfs/base.c:933

rmmod(1073): Kernel Bug 1
pc = [<fffffc00003d4ad4>] ra = [<fffffc00003d4ac8>] ps = 0000 Not
tainted
Using defaults from ksymoops -t elf64-alpha -a alpha
v0 = 000000000000000d t0 = 0000000000000001 t1 = fffffc000058d3a0
t2 = 0000000000000000 t3 = ffffffff00000000 t4 = 0000000000000001
t5 = 0000000000000000 t6 = fffffc0000630450 t7 = fffffc0006734000
a0 = 0000000000000000 a1 = 0000000000000001 a2 = 0000000000000000
a3 = 0000000000000000 a4 = 00000000000000a0 a5 = 000000000000000f
t8 = 0000000000000000 t9 = 0000000000003fff t10= 000000000000000f
t11= 0000000000000010 pv = fffffc000032d160 at = 0000000000000000
gp = fffffc00006281b0 sp = fffffc0006737d88
Trace:fffffc00003d63b8 fffffc00003f5ac0 fffffc00003f5bac fffffc00003f6018
fffffc00003f64ec fffffc00003f65f8 fffffc0000334e94 fffffc0000333db8
fffffc0000310d44
Code: a61dfe00 a77daf28 6b5b7556 27ba0025 23bd36e8 00000081
<000003a5> 0050f363
>>RA; fffffc00003d4ac8 <devfs_put+68/1e0>

>>PC; fffffc00003d4ad4 <devfs_put+74/1e0> <=====

Trace; fffffc00003d63b8 <devfs_unregister+38/60>
Trace; fffffc00003f5ac0 <detach+60/a0>
Trace; fffffc00003f5bac <driver_detach+6c/c0>
Trace; fffffc00003f6018 <bus_remove_driver+78/100>
Trace; fffffc00003f64ec <put_driver+6c/c0>
Trace; fffffc00003f65f8 <driver_unregister+18/40>
Trace; fffffc0000334e94 <free_module+34/160>
Trace; fffffc0000333db8 <sys_delete_module+1d8/340>
Trace; fffffc0000310d44 <entSys+a4/b8>

Code; fffffc00003d4abc <devfs_put+5c/1e0>
0000000000000000 <_PC>:
Code; fffffc00003d4abc <devfs_put+5c/1e0>
0: 00 fe 1d a6 ldq a0,-512(gp)
Code; fffffc00003d4ac0 <devfs_put+60/1e0>
4: 28 af 7d a7 ldq t12,-20696(gp)
Code; fffffc00003d4ac4 <devfs_put+64/1e0>
8: 56 75 5b 6b jsr ra,(t12),ffffffffffffd564
<_PC+0xffffffffffffd564> fffffc00003d2020 <ext2_rename+140/400>
Code; fffffc00003d4ac8 <devfs_put+68/1e0>
c: 25 00 ba 27 ldah gp,37(ra)
Code; fffffc00003d4acc <devfs_put+6c/1e0>
10: e8 36 bd 23 lda gp,14056(gp)
Code; fffffc00003d4ad0 <devfs_put+70/1e0>
14: 81 00 00 00 call_pal 0x81
Code; fffffc00003d4ad4 <devfs_put+74/1e0> <=====
18: a5 03 00 00 call_pal 0x3a5 <=====
Code; fffffc00003d4ad8 <devfs_put+78/1e0>
1c: 63 f3 50 00 call_pal 0x50f363


--jochen


2002-11-02 22:30:50

by Greg KH

[permalink] [raw]
Subject: Re: [BUG] USB Kernel bug in 2.5.45

On Sat, Nov 02, 2002 at 07:37:55PM +0100, Jochen Friedrich wrote:
> Hi,
>
> when running "rmmod hid" while a USB HID device is still connected (but
> not used), i get this Oops messages:
>
> drivers/usb/core/usb.c: deregistering driver hiddev
> drivers/usb/core/usb.c: deregistering driver hid
> devfs_put(fffffc0009393000): poisoned pointer

If you disable devfs does it work ok?

thanks,

greg k-h

2002-11-02 23:10:26

by Jochen Friedrich

[permalink] [raw]
Subject: Re: [BUG] USB Kernel bug in 2.5.45

Hi Greg,

> If you disable devfs does it work ok?

Yes, it does...

drivers/usb/core/usb.c: deregistering driver hiddev
drivers/usb/core/usb.c: deregistering driver hid

and the module is gone.

--jochen

2002-11-02 23:21:21

by Greg KH

[permalink] [raw]
Subject: Re: [BUG] USB Kernel bug in 2.5.45

On Sun, Nov 03, 2002 at 12:10:43AM +0100, Jochen Friedrich wrote:
> Hi Greg,
>
> > If you disable devfs does it work ok?
>
> Yes, it does...

Great, thanks for trying. I don't see any more problems here :)

thanks,

greg k-h