2.5.10: 2 OOPs - "BUG at usb.c" and "unable to handle kernel paging request"

2.5.10: 2 OOPs - "BUG at usb.c" and "unable to handle kernel paging request"

[Alex Walker]

[-- Attachment #1: Type: text/plain, Size: 545 bytes --]

Hi, I'm not subscribed - please CC me in any replies.

Two OOps when running 2.5.10, as attached. With attatched config.

First occurs on boot, but doesn't stop the whole system.  The second
occurs as I was rebooting - see the attached log to see where they
happen.

Any more info required?  Just ask.

Alex.

-- 
\\\\\\    You know it's Monday when you wake up and it's Tuesday.     //////
\\\\                             -- Garfield                            ////
\\                                                                        //

[-- Attachment #2: ksym-2.5.10-1 --]
[-- Type: text/plain, Size: 2256 bytes --]

ksymoops 2.4.5 on i686 2.4.18.  Options used
     -V (default)
     -K (specified)
     -L (specified)
     -o /lib/modules/2.5.10/ (specified)
     -m /boot/System.map-2.5.10 (specified)

No modules in ksyms, skipping objects
cpu: 0, clocks: 1328904, slice: 664452
kernel BUG at usb.c:856!
invalid operand: 0000
CPU:    0
EIP:    0010:[<c0202ace>]    Not tainted
Using defaults from ksymoops -t elf32-i386 -a i386
EFLAGS: 00010246
eax: 00000000   ebx: d3da1600   ecx: ffffe21a   edx: 00000002
esi: d3da1600   edi: 00000001   ebp: d3da1200   esp: d3e6df64
ds: 0018   es: 0018   ss: 0018
Stack: d3da173c c0205c5d d3da1600 00000001 00000002 d3da1200 d3d9b860 c02d0264 
       d3da13b8 00000004 d3da13b8 000000c8 c0205e09 d3d9b860 00000001 00000301 
       00000001 d3e6c000 d3e6c000 d3e6dfe0 0008e000 d3d9b888 d3e6dfc8 d3d9b888 
Call Trace: [<c0205c5d>] [<c0205e09>] [<c0205fd5>] [<c010560c>] 
Code: 0f 0b 58 03 f4 f9 29 c0 8b 83 cc 00 00 00 8b 40 1c 53 8b 40 


>>EIP; c0202ace <usb_free_dev+26/5c>   <=====

>>ebx; d3da1600 <END_OF_CODE+13a5a584/????>
>>ecx; ffffe21a <END_OF_CODE+3fcb719e/????>
>>esi; d3da1600 <END_OF_CODE+13a5a584/????>
>>ebp; d3da1200 <END_OF_CODE+13a5a184/????>
>>esp; d3e6df64 <END_OF_CODE+13b26ee8/????>

Trace; c0205c5d <usb_hub_port_connect_change+229/2bc>
Trace; c0205e09 <usb_hub_events+119/2a0>
Trace; c0205fd5 <usb_hub_thread+45/10c>
Trace; c010560c <kernel_thread+28/38>

Code;  c0202ace <usb_free_dev+26/5c>
00000000 <_EIP>:
Code;  c0202ace <usb_free_dev+26/5c>   <=====
   0:   0f 0b                     ud2a      <=====
Code;  c0202ad0 <usb_free_dev+28/5c>
   2:   58                        pop    %eax
Code;  c0202ad1 <usb_free_dev+29/5c>
   3:   03 f4                     add    %esp,%esi
Code;  c0202ad3 <usb_free_dev+2b/5c>
   5:   f9                        stc    
Code;  c0202ad4 <usb_free_dev+2c/5c>
   6:   29 c0                     sub    %eax,%eax
Code;  c0202ad6 <usb_free_dev+2e/5c>
   8:   8b 83 cc 00 00 00         mov    0xcc(%ebx),%eax
Code;  c0202adc <usb_free_dev+34/5c>
   e:   8b 40 1c                  mov    0x1c(%eax),%eax
Code;  c0202adf <usb_free_dev+37/5c>
  11:   53                        push   %ebx
Code;  c0202ae0 <usb_free_dev+38/5c>
  12:   8b 40 00                  mov    0x0(%eax),%eax

[-- Attachment #3: ksym-2.5.10-2 --]
[-- Type: text/plain, Size: 1587 bytes --]

ksymoops 2.4.5 on i686 2.4.18.  Options used
     -V (default)
     -K (specified)
     -L (specified)
     -o /lib/modules/2.5.10/ (specified)
     -m /boot/System.map-2.5.10 (specified)

Unable to handle kernel paging request at virtual address d4bedecc
d4bedecc
*pde = 13eb4067
Oops: 0000
CPU:    0
EIP:    0010:[<d4bedecc>]    Tainted: P  
EFLAGS: 00010286
eax: d4bedecc   ebx: d29de580   ecx: d4be2bd8   edx: d4be2bc4
esi: d29de660   edi: d29de580   ebp: d29de658   esp: d1c09f5c
ds: 0018   es: 0018   ss: 0018
Stack: d4be0d34 d29de580 d29db898 d4be0f3a d29de580 d29db898 fffffff0 d1caf000 
       d4be0000 d4be2bd8 d4be224c d29de580 d4be0000 c0118627 d4be0000 fffffff0 
       d1caf000 bfffed9c c01178a9 d4be0000 00000000 d1c08000 00000001 bfffed9c 
Call Trace: [<d4be0d34>] [<d4be0f3a>] [<d4be2bd8>] [<d4be224c>] [<c0118627>] 
   [<c01178a9>] [<c0106f1f>] 
Code:  Bad EIP value.


>>EIP; d4bedecc <END_OF_CODE+148a6e50/????>   <=====

>>eax; d4bedecc <END_OF_CODE+148a6e50/????>
>>ebx; d29de580 <END_OF_CODE+12697504/????>
>>ecx; d4be2bd8 <END_OF_CODE+1489bb5c/????>
>>edx; d4be2bc4 <END_OF_CODE+1489bb48/????>
>>esi; d29de660 <END_OF_CODE+126975e4/????>
>>edi; d29de580 <END_OF_CODE+12697504/????>
>>ebp; d29de658 <END_OF_CODE+126975dc/????>
>>esp; d1c09f5c <END_OF_CODE+118c2ee0/????>

Trace; d4be0d34 <END_OF_CODE+14899cb8/????>
Trace; d4be0f3a <END_OF_CODE+14899ebe/????>
Trace; d4be2bd8 <END_OF_CODE+1489bb5c/????>
Trace; d4be224c <END_OF_CODE+1489b1d0/????>
Trace; c0118627 <free_module+17/c0>
Trace; c01178a9 <sys_delete_module+12d/27c>
Trace; c0106f1f <syscall_call+7/b>


[-- Attachment #4: log-2.5.10.gz --]
[-- Type: application/octet-stream, Size: 4195 bytes --]

[-- Attachment #5: config-2.5.10.gz --]
[-- Type: application/octet-stream, Size: 5304 bytes --]

Re: 2.5.10: 2 OOPs - "BUG at usb.c" and "unable to handle kernel paging request"

[Greg KH]

On Wed, Apr 24, 2002 at 02:21:32PM +0100, Alex Walker wrote:
> Hi, I'm not subscribed - please CC me in any replies.
> 
> Two OOps when running 2.5.10, as attached. With attatched config.
> 
> First occurs on boot, but doesn't stop the whole system.  The second
> occurs as I was rebooting - see the attached log to see where they
> happen.
> 
> Any more info required?  Just ask.

Known problem with the uhci driver right now, just use usb-uhci instead
(not the ALT UHCI driver) for now until things get straightend out.

thanks,

greg k-h

Re: 2.5.10: 2 OOPs - "BUG at usb.c" and "unable to handle kernel paging request"

[Alex Walker]

'ello Greg

On Thu, Apr 25, 2002 at 12:46:51AM -0700, Greg KH wrote:
> > Hi, I'm not subscribed - please CC me in any replies.  Two OOps when
> > running 2.5.10, as attached. With attatched config.  First occurs on
> > boot, but doesn't stop the whole system.  The second occurs as I was
> > rebooting - see the attached log to see where they happen.  Any more
> > info required?  Just ask.
> Known problem with the uhci driver right now, just use usb-uhci instead
> (not the ALT UHCI driver) for now until things get straightend out.

Yes, that's solved the first oops.  Is the second one related?

Alex

-- 
Alex

Re: 2.5.10: 2 OOPs - "BUG at usb.c" and "unable to handle kernel paging request"

[Greg KH]

On Thu, Apr 25, 2002 at 01:37:03PM +0100, Alex Walker wrote:
> 'ello Greg
> 
> On Thu, Apr 25, 2002 at 12:46:51AM -0700, Greg KH wrote:
> > > Hi, I'm not subscribed - please CC me in any replies.  Two OOps when
> > > running 2.5.10, as attached. With attatched config.  First occurs on
> > > boot, but doesn't stop the whole system.  The second occurs as I was
> > > rebooting - see the attached log to see where they happen.  Any more
> > > info required?  Just ask.
> > Known problem with the uhci driver right now, just use usb-uhci instead
> > (not the ALT UHCI driver) for now until things get straightend out.
> 
> Yes, that's solved the first oops.  Is the second one related?

Probably, does the second one still happen now?

greg k-h

Re: 2.5.10: 2 OOPs - "BUG at usb.c" and "unable to handle kernel paging request"

[Greg KH]

On Thu, Apr 25, 2002 at 05:04:13PM +0100, Alex Walker wrote:
> Look! It's Greg!
> 
> On Thu, Apr 25, 2002 at 07:54:50AM -0700, Greg KH wrote:
> > On Thu, Apr 25, 2002 at 01:37:03PM +0100, Alex Walker wrote:
> > > 'ello Greg On Thu, Apr 25, 2002 at 12:46:51AM -0700, Greg KH wrote:
> > > > > Hi, I'm not subscribed - please CC me in any replies.  Two OOps
> > > > > when running 2.5.10, as attached. With attatched config.  First
> > > > > occurs on boot, but doesn't stop the whole system.  The second
> > > > > occurs as I was rebooting - see the attached log to see where
> > > > > they happen.  Any more info required?  Just ask.
> > > > Known problem with the uhci driver right now, just use usb-uhci
> > > > instead (not the ALT UHCI driver) for now until things get
> > > > straightend out.
> > > Yes, that's solved the first oops.  Is the second one related?
> > Probably, does the second one still happen now?
> 
> Sorry, that would have been useful for me to say wouldn't it?  yes it
> does still happen.

Ah, I have no idea.  The ksymoops report doesn't give very much to go
by, so I'll guess that it's not a USB problem :)

thanks,

greg k-h

Re: 2.5.10: 2 OOPs - "BUG at usb.c" and "unable to handle kernel paging request"

[Alex Walker]

Look! It's Greg!

On Thu, Apr 25, 2002 at 07:54:50AM -0700, Greg KH wrote:
> On Thu, Apr 25, 2002 at 01:37:03PM +0100, Alex Walker wrote:
> > 'ello Greg On Thu, Apr 25, 2002 at 12:46:51AM -0700, Greg KH wrote:
> > > > Hi, I'm not subscribed - please CC me in any replies.  Two OOps
> > > > when running 2.5.10, as attached. With attatched config.  First
> > > > occurs on boot, but doesn't stop the whole system.  The second
> > > > occurs as I was rebooting - see the attached log to see where
> > > > they happen.  Any more info required?  Just ask.
> > > Known problem with the uhci driver right now, just use usb-uhci
> > > instead (not the ALT UHCI driver) for now until things get
> > > straightend out.
> > Yes, that's solved the first oops.  Is the second one related?
> Probably, does the second one still happen now?

Sorry, that would have been useful for me to say wouldn't it?  yes it
does still happen.

Alex.

-- 
"I'm returning this note to you, instead of your paper, because it (your
paper)
presently occupies the bottom of my bird cage."
  -- English Professor, Providence College

Re: 2.5.10: 2 OOPs - "BUG at usb.c" and "unable to handle kernel paging request"

[Johannes Erdfelt]

On Thu, Apr 25, 2002, Greg KH <greg@kroah.com> wrote:
> On Wed, Apr 24, 2002 at 02:21:32PM +0100, Alex Walker wrote:
> > Hi, I'm not subscribed - please CC me in any replies.
> > 
> > Two OOps when running 2.5.10, as attached. With attatched config.
> > 
> > First occurs on boot, but doesn't stop the whole system.  The second
> > occurs as I was rebooting - see the attached log to see where they
> > happen.
> > 
> > Any more info required?  Just ask.
> 
> Known problem with the uhci driver right now, just use usb-uhci instead
> (not the ALT UHCI driver) for now until things get straightend out.

Just to clarify, it's a known problem with the USB core that uhci
happens to trigger.

I'm working on an approriate patch.

JE