cpufreq segfaults when network cable plugged in

cpufreq segfaults when network cable plugged in

[Simon Cooper]

right, this is very strange, cpufreq (compiled into the kernel) segfaults whenever the network cable on my laptop is plugged in on bootup but boots ok when it is plugged in after the kernel has loaded. i can post the entire segfault msg but it'll need me to write it all down. driver is 'r8139' (1000Mbps card), kernel version 2.6.5-mm2

Simon Cooper 

Re: cpufreq segfaults when network cable plugged in

[Dominik Brodowski]

On Mon, Apr 12, 2004 at 12:43:35AM +0100, Simon Cooper wrote:
> right, this is very strange, cpufreq (compiled into the kernel) segfaults
> whenever the network cable on my laptop is plugged in on bootup but boots
> ok when it is plugged in after the kernel has loaded. i can post the entire
> segfault msg but it'll need me to write it all down. driver is 'r8139'
> (1000Mbps card), kernel version 2.6.5-mm2

Hm, this is really very strange. 

Which cpufreq drivers do you use / are compiled into the kernel?
What are the last two to four functions called, where's the EIP at
the segfault? [the first few function names mentioned in the oops]

Thanks,
	Dominik

Re: cpufreq segfaults when network cable plugged in

[Simon Cooper]

On Mon, 12 Apr 2004 10:30:58 +0200
Dominik Brodowski <linux@dominikbrodowski.de> wrote:

> On Mon, Apr 12, 2004 at 12:43:35AM +0100, Simon Cooper wrote:
> > right, this is very strange, cpufreq (compiled into the kernel) segfaults
> > whenever the network cable on my laptop is plugged in on bootup but boots
> > ok when it is plugged in after the kernel has loaded. i can post the entire
> > segfault msg but it'll need me to write it all down. driver is 'r8139'
> > (1000Mbps card), kernel version 2.6.5-mm2
> 
> Hm, this is really very strange. 
> 
> Which cpufreq drivers do you use / are compiled into the kernel?
> What are the last two to four functions called, where's the EIP at
> the segfault? [the first few function names mentioned in the oops]
> 
> Thanks,
> 	Dominik
> 

well, technically its a kernel oops (attempted ot kill init). powersave and performance govenors are compiled in, using p4_clockmod driver. Sometimes it boots fine with the cable in, but it oops's most of the time:

Unable to handle kernel NULL pointer dereference at virtual address 00000018
...
EIP: 0060:[<c02f7c45>]  Not tainted VLI
EIP is at cpufreq_notify_transition+0x79/0x98
...
Call trace:
[<c010f543>] cpufreq_p4_setdc+0x1eb/0x306
[<c01706d3>] d_lookup+0x29/0x52
[<c010f6bb>] cpufreq_p4_target+0x5d/0x69
[<c02f757d>] __cpufreq_driver_target+0x22/0x26
...

Simon

Re: cpufreq segfaults when network cable plugged in

[Dominik Brodowski]

[-- Attachment #1.1: Type: text/plain, Size: 1769 bytes --]

On Mon, Apr 12, 2004 at 07:00:24PM +0100, Simon Cooper wrote:
> On Mon, 12 Apr 2004 10:30:58 +0200
> Dominik Brodowski <linux@dominikbrodowski.de> wrote:
> 
> > On Mon, Apr 12, 2004 at 12:43:35AM +0100, Simon Cooper wrote:
> > > right, this is very strange, cpufreq (compiled into the kernel) segfaults
> > > whenever the network cable on my laptop is plugged in on bootup but boots
> > > ok when it is plugged in after the kernel has loaded. i can post the entire
> > > segfault msg but it'll need me to write it all down. driver is 'r8139'
> > > (1000Mbps card), kernel version 2.6.5-mm2
> > 
> > Hm, this is really very strange. 
> > 
> > Which cpufreq drivers do you use / are compiled into the kernel?
> > What are the last two to four functions called, where's the EIP at
> > the segfault? [the first few function names mentioned in the oops]
> > 
> > Thanks,
> > 	Dominik
> > 
> 
> well, technically its a kernel oops (attempted ot kill init). powersave and performance govenors are compiled in, using p4_clockmod driver. Sometimes it boots fine with the cable in, but it oops's most of the time:
> 
> Unable to handle kernel NULL pointer dereference at virtual address 00000018
> ...
> EIP: 0060:[<c02f7c45>]  Not tainted VLI
> EIP is at cpufreq_notify_transition+0x79/0x98
> ...
> Call trace:
> [<c010f543>] cpufreq_p4_setdc+0x1eb/0x306
> [<c01706d3>] d_lookup+0x29/0x52
> [<c010f6bb>] cpufreq_p4_target+0x5d/0x69
> [<c02f757d>] __cpufreq_driver_target+0x22/0x26
> ...

This very strange. Especially as all such NULL pointer dereferences would
already happen _earlier_. It seems to me that this might be a bug in the
networking driver, which -randomly?- affects unrelated other
kernel/userspace components...

	Dominik

[-- Attachment #1.2: Type: application/pgp-signature, Size: 189 bytes --]

[-- Attachment #2: Type: text/plain, Size: 143 bytes --]

_______________________________________________
Cpufreq mailing list
Cpufreq@www.linux.org.uk
http://www.linux.org.uk/mailman/listinfo/cpufreq