Hello!
I've found a kernel bug in linux-2.6-20-rc1 from kernel.org:
Dec 17 19:12:56 localhost kernel: kernel BUG at
drivers/cpufreq/cpufreq_userspace.c:140!
Dec 17 19:12:56 localhost kernel: invalid opcode: 0000 [#1]
Dec 17 19:12:56 localhost kernel: PREEMPT
Dec 17 19:12:56 localhost kernel: Modules linked in: acpi_cpufreq
freq_table processor sg scsi_mod usbhid hid joydev nsc_ircc pcmcia
snd_intel8x0 snd_ac97_codec ac97_bus snd_pcm_oss snd_mixer_oss
firmware_class snd_pcm snd_timer irda tifm_7xx1 tifm_sd mmc_block
mmc_core ide_cd crc_ccitt i2c_i801 snd soundcore snd_page_alloc
tifm_core psmouse ohci1394 ieee1394 ehci_hcd uhci_hcd yenta_socket
rsrc_nonstatic pcmcia_core rtc pcspkr cdrom usbcore intel_agp agpgart
evdev unix
Dec 17 19:12:56 localhost kernel: CPU: 0
Dec 17 19:12:56 localhost kernel: EIP: 0060:[<c022fe0d>] Not
tainted VLI
Dec 17 19:12:56 localhost kernel: EFLAGS: 00010246 (2.6.20-rc1-vanilla #1)
Dec 17 19:12:56 localhost kernel: EIP is at
cpufreq_governor_userspace+0x32/0x134
Dec 17 19:12:56 localhost kernel: eax: c1b85480 ebx: c1b85480 ecx:
c0308220 edx: 00000000
Dec 17 19:12:56 localhost kernel: esi: 00000000 edi: ffffffea ebp:
c1b85480 esp: f7f2bdb4
Dec 17 19:12:56 localhost kernel: ds: 007b es: 007b ss: 0068
Dec 17 19:12:56 localhost kernel: Process modprobe (pid: 2098,
ti=f7f2a000 task=c1938a70 task.ti=f7f2a000)
Dec 17 19:12:56 localhost kernel: Stack: c1b85480 00000000 00000001
c022ef25 f7f2be04 00000000 00000000 c1b85480
Dec 17 19:12:56 localhost kernel: c022f023 c1b854ac c1b85480
f7f2be04 c1b854ac c022f1e4 c1b85480 f9b1c918
Dec 17 19:12:56 localhost kernel: f7f2bea4 c022fb42 c03498e8
00000000 00000001 00000002 00000000 00186a00
Dec 17 19:12:56 localhost kernel: Call Trace:
Dec 17 19:12:56 localhost kernel: [<c022ef25>] __cpufreq_governor+0x56/0x87
Dec 17 19:12:56 localhost kernel: [<c022f023>]
__cpufreq_set_policy+0xcd/0x100
Dec 17 19:12:56 localhost kernel: [<c022f1e4>] cpufreq_set_policy+0x2a/0x5f
Dec 17 19:12:56 localhost kernel: [<c022fb42>] cpufreq_add_dev+0x1f1/0x269
Dec 17 19:12:56 localhost kernel: [<c022f1b2>] handle_update+0x0/0x8
Dec 17 19:12:56 localhost kernel: [<c020f7ff>]
sysdev_driver_register+0x4c/0x94
Dec 17 19:12:56 localhost kernel: [<c022fd5c>]
cpufreq_register_driver+0xa2/0x112
Dec 17 19:12:56 localhost kernel: [<f9b3504a>]
acpi_cpufreq_init+0x4a/0x4c [acpi_cpufreq]
Dec 17 19:12:56 localhost kernel: [<c012e538>]
sys_init_module+0x147a/0x15b3
Dec 17 19:12:56 localhost kernel: [<c0102c56>] sysenter_past_esp+0x5f/0x85
Dec 17 19:12:56 localhost kernel: =======================
Dec 17 19:12:56 localhost kernel: Code: 08 83 fa 02 74 7f 83 fa 03 0f 84
bb 00 00 00 31 ff 4a 0f 85 0f 01 00 00 bf ea ff ff ff 85 f6 0f 85 02 01
00 00 83 78 20 00 75 04 <0f> 0b eb fe b8 4c 82 30 c0 e8 dd 2c 05 00 8d
43 58 ba 58 82 30
Dec 17 19:12:56 localhost kernel: EIP: [<c022fe0d>]
cpufreq_governor_userspace+0x32/0x134 SS:ESP 0068:f7f2bdb4
Dec 17 19:12:56 localhost kernel: <7>ieee80211_crypt: registered
algorithm 'NULL'
...
My system is a Debian unstable/testing:
Gnu C 4.1.2
Gnu make 3.81
binutils 2.17
util-linux 2.12r
mount 2.12r
module-init-tools 3.3-pre2
e2fsprogs 1.40-WIP
reiserfsprogs 3.6.19
xfsprogs 2.8.18
pcmciautils 014
PPP 2.4.4
Linux C Library 2.3.6
Dynamic linker (ldd) 2.3.6
Procps 3.2.7
Net-tools 1.60
Console-tools 0.2.3
Sh-utils 5.97
udev 103
wireless-tools 28
Regards,
Berthold Cogel
In-Reply-To: <[email protected]>
On Sun, 17 Dec 2006 20:10:01 +0100, Berthold Cogel wrote:
> I've found a kernel bug in linux-2.6-20-rc1 from kernel.org:
>
> Dec 17 19:12:56 localhost kernel: kernel BUG at drivers/cpufreq/cpufreq_userspace.c:140!
Does this fix it?
Signed-off-by: Chuck Ebbert <[email protected]>
--- 2.6.20-rc1-32smp.orig/arch/i386/kernel/cpu/cpufreq/acpi-cpufreq.c
+++ 2.6.20-rc1-32smp/arch/i386/kernel/cpu/cpufreq/acpi-cpufreq.c
@@ -706,7 +706,7 @@ static int acpi_cpufreq_cpu_init(struct
break;
case ACPI_ADR_SPACE_FIXED_HARDWARE:
acpi_cpufreq_driver.get = get_cur_freq_on_cpu;
- get_cur_freq_on_cpu(cpu);
+ policy->cur = get_cur_freq_on_cpu(cpu);
break;
default:
break;
--
MBTI: IXTP
On Tue, Dec 19, 2006 at 04:18:13AM -0500, Chuck Ebbert wrote:
> In-Reply-To: <[email protected]>
>
> On Sun, 17 Dec 2006 20:10:01 +0100, Berthold Cogel wrote:
> > I've found a kernel bug in linux-2.6-20-rc1 from kernel.org:
> > Dec 17 19:12:56 localhost kernel: kernel BUG at drivers/cpufreq/cpufreq_userspace.c:140!
>
> Does this fix it?
>
> Signed-off-by: Chuck Ebbert <[email protected]>
>
> --- 2.6.20-rc1-32smp.orig/arch/i386/kernel/cpu/cpufreq/acpi-cpufreq.c
> +++ 2.6.20-rc1-32smp/arch/i386/kernel/cpu/cpufreq/acpi-cpufreq.c
> @@ -706,7 +706,7 @@ static int acpi_cpufreq_cpu_init(struct
> break;
> case ACPI_ADR_SPACE_FIXED_HARDWARE:
> acpi_cpufreq_driver.get = get_cur_freq_on_cpu;
> - get_cur_freq_on_cpu(cpu);
> + policy->cur = get_cur_freq_on_cpu(cpu);
> break;
> default:
> break;
A similar fix is in Linus' -git tree as
http://git.kernel.org/git/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=a507ac4b01ed379a74eca5060f3553c4a4e5854c
Hopefully one day the kernel.org scripts will get around to making a -git6 patch.
Dave
--
http://www.codemonkey.org.uk