2.6.30 kernel + powernow-k8 gives modprobe segmentation fault at boot

gsdefender · 2009-06-28 13:09:21

Dear all,
hi there, that's another (Archlinux) forum newbie.
I've been running the 64 bit flavour of Archlinux flawlessy on my asus l5d (1st generation amd athlon 64 3000+, 768 mb ram) until the recent kernel version bump (2.6.29.4 => 2.6.30).
From there on my system is experiencing segmentation fault while loading powernow-k8. Here you have the relevant dmesg extract

processor ACPI_CPU:00: registered as cooling_device0
powernow-k8: Found 1 AMD Athlon(tm) 64 Processor 3000+ processors (1 cpu cores) 
(version 2.20.00)
BUG: unable to handle kernel NULL pointer dereference at 000000000000001c
IP: [<ffffffffa041451d>] powernowk8_cpu_init+0x364/0xad2 [powernow_k8]
PGD 2ee22067 PUD 2e8dc067 PMD 0 
Oops: 0002 [#1] PREEMPT SMP 
last sysfs file: /sys/module/snd_pcm_oss/initstate
CPU 0 
Modules linked in: powernow_k8(+) freq_table processor vboxnetflt vboxdrv snd_se
q_dummy snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_intel8x0 snd_a
c97_codec snd_pcm_oss snd_pcm snd_timer snd_page_alloc snd_mixer_oss snd soundco
re ac97_bus slhc skge rtc_cmos rtc_core rtc_lib ext3 jbd mbcache ohci_hcd sr_mod
 ehci_hcd cdrom sd_mod usbcore ata_generic pata_amd pata_acpi libata scsi_mod
Pid: 975, comm: modprobe Not tainted 2.6.30-ARCH #1 L5D       
RIP: 0010:[<ffffffffa041451d>]  [<ffffffffa041451d>] powernowk8_cpu_init+0x364/0
xad2 [powernow_k8]
RSP: 0018:ffff88002eec3cc8  EFLAGS: 00010246
RAX: 0000000000000000 RBX: ffff88002eecd100 RCX: 0000000000000000
RDX: 0000000000000003 RSI: 00000000000c3500 RDI: 0000000000000012
RBP: ffff88002efe6300 R08: 0000000000000003 R09: ffff88002efe6600
R10: ffff88002ef032a0 R11: 0000000000000000 R12: ffff88002ebde640
R13: 0000000000000003 R14: ffff88002efe6340 R15: 0000000000000002
FS:  00007f00d10276f0(0000) GS:ffff88000100b000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 000000000000001c CR3: 000000002ee62000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process modprobe (pid: 975, threadinfo ffff88002eec2000, task ffff88002efab280)
Stack:
 ffffffff807c1e60 000000000d58bd2a ffff88002eec3d28 ffffffff803bde0f
 000000020000000e 000000000d58bd2a 000000000000e790 000000000d58bd2a
 0000000000000000 0000000000000000 ffff88002eecd100 ffff880001016428
Call Trace:
 [<ffffffff803bde0f>] ? vsnprintf+0x30f/0x5d0
 [<ffffffff8048059d>] ? cpufreq_add_dev+0x1cd/0x690
 [<ffffffff8054e8e8>] ? __mutex_lock_slowpath+0x238/0x340
 [<ffffffff80461906>] ? sysdev_driver_register+0xb6/0x150
 [<ffffffffa0414c8b>] ? powernowk8_init+0x0/0xca [powernow_k8]
 [<ffffffff8047f3c4>] ? cpufreq_register_driver+0xa4/0x190
 [<ffffffff8020907b>] ? do_one_initcall+0x4b/0x1b0
 [<ffffffff802727a9>] ? __blocking_notifier_call_chain+0x79/0xc0
 [<ffffffff80288680>] ? sys_init_module+0xd0/0x220
 [<ffffffff8020c442>] ? system_call_fastpath+0x16/0x1b
Code: 49 8b 04 00 48 69 d0 e8 03 00 00 48 39 d1 74 23 69 d0 e8 03 00 00 48 c7 c7
 b0 56 41 a0 31 c0 e8 2d 84 13 e0 48 8b 45 38 44 89 ea <c7> 44 d0 04 ff ff ff ff
 41 83 c5 01 44 3b 6d 68 0f 82 0c ff ff 
RIP  [<ffffffffa041451d>] powernowk8_cpu_init+0x364/0xad2 [powernow_k8]
 RSP <ffff88002eec3cc8>
CR2: 000000000000001c
---[ end trace 7086b00b1b998fa9 ]---

The system continues booting surprisingly well, but everything gets hotter and hotter (I used to have it running at 800 mhz to reduce fan activity, as I don't really need full speed).
I only wrote to know if anyone has got into similar troubles - I had a look over recent kernel ChangeLog and I've seen some activity about powernow-k8, but not really something
I might have thought would blow everything up. I'd like also to be pointed at custom patches someone may have come up with - I've no problem messing with kernel recompilation: if this doesn't get fixed in a new package, I'm thinking about reverting to 2.6.29.4 - or maybe some kernel parameters to cover the thing (at the moment and up until 2.6.29, II need "noapic nolapic").

Thank you in advance. Let me know if you need additional info.

Pierre · 2009-06-28 13:28:29

I can confirm that powernow_k8 is broken somehow. I just got this when trying to change the governour. But I cannot reproduce it.

INFO: task kondemand/0:3657 blocked for more than 120 seconds.                                                                         
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.                                                              
kondemand/0   D ffff88007e7992f0     0  3657      2                                                                                    
 ffff88007e7992f0 0000000000000046 000000007ef643c0 ffffffff8023f294                                                                   
 ffffffff80826130 ffff88007ea2eb50 ffff88007e799590 00000000000111c0                                                                   
 000000000000cb68 ffff88007e799590 ffff88007be8fdd0 ffff8800010211c0                                                                   
Call Trace:                                                                                                                            
 [<ffffffff8023f294>] ? task_rq_lock+0x64/0xc0                                                                                         
 [<ffffffff8054d9a9>] ? schedule+0x29/0x70                                                                                             
 [<ffffffff80550525>] ? __down_write_nested+0x95/0xe0                                                                                  
 [<ffffffff8048030b>] ? lock_policy_rwsem_write+0x5b/0xb0                                                                              
 [<ffffffffa0d8b8c9>] ? do_dbs_timer+0x69/0x370 [cpufreq_ondemand]
 [<ffffffffa0d8b860>] ? do_dbs_timer+0x0/0x370 [cpufreq_ondemand]
 [<ffffffff80266431>] ? worker_thread+0x161/0x300
 [<ffffffff8026c6d0>] ? autoremove_wake_function+0x0/0x60
 [<ffffffff802662d0>] ? worker_thread+0x0/0x300
 [<ffffffff8026c0b4>] ? kthread+0x64/0xc0
 [<ffffffff8024ae30>] ? schedule_tail+0x30/0x80
 [<ffffffff8020d4fa>] ? child_rip+0xa/0x20
 [<ffffffff8026c050>] ? kthread+0x0/0xc0
 [<ffffffff8020d4f0>] ? child_rip+0x0/0x20
INFO: task cpufreq-set:14658 blocked for more than 120 seconds.
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
cpufreq-set   D 000000007c9e91cb     0 14658      1
 ffff88006a0bf1a0 0000000000000082 0000000000000000 ffff880040d6ed80
 ffff88006a0bc570 ffff88007ee52c30 ffff88006a0bf440 00000000000111c0
 000000000000cb68 ffff88006a0bf440 ffff88006a0c7740 ffff880040d6ed80
Call Trace:
 [<ffffffff8054d2e7>] ? thread_return+0x4e/0x6e7
 [<ffffffff80214d05>] ? sched_clock+0x5/0x10
 [<ffffffff8054d9a9>] ? schedule+0x29/0x70
 [<ffffffff8054df25>] ? schedule_timeout+0x1a5/0x1f0
 [<ffffffff8054d9a9>] ? schedule+0x29/0x70
 [<ffffffff8054df25>] ? schedule_timeout+0x1a5/0x1f0
 [<ffffffff8054cd6d>] ? wait_for_common+0xdd/0x1b0
 [<ffffffff802494d0>] ? default_wake_function+0x0/0x40
 [<ffffffff80267210>] ? __cancel_work_timer+0x1c0/0x1d0
 [<ffffffff80266b50>] ? wq_barrier_func+0x0/0x40
 [<ffffffffa0d8bce7>] ? cpufreq_governor_dbs+0x117/0x37c [cpufreq_ondemand]
 [<ffffffff803ba272>] ? __up_read+0x32/0xe0
 [<ffffffff8047e5bc>] ? __cpufreq_governor+0xec/0x170
 [<ffffffff8047f2be>] ? __cpufreq_set_policy+0x15e/0x1c0
 [<ffffffff8047fb93>] ? store_scaling_governor+0x113/0x230
 [<ffffffff80480c10>] ? handle_update+0x0/0x40
 [<ffffffff8047fa80>] ? store_scaling_governor+0x0/0x230
 [<ffffffff80480ad1>] ? store+0x71/0xc0
 [<ffffffff8036b535>] ? sysfs_write_file+0xe5/0x190
 [<ffffffff802fc204>] ? vfs_write+0xe4/0x1c0
 [<ffffffff802fc3fe>] ? sys_write+0x5e/0xb0
 [<ffffffff802f8f2b>] ? do_sys_open+0xdb/0x130
 [<ffffffff8020c442>] ? system_call_fastpath+0x16/0x1b

arkham · 2009-06-28 14:14:59

Weird, I have the same processor and everything is working fine:

arkham ~ $ uname -a
Linux acerus 2.6.30-ARCH #1 SMP PREEMPT Fri Jun 19 21:25:17 UTC 2009 i686 AMD Athlon 64 Processor 3000+ AuthenticAMD GNU/Linux

arkham ~ $ lsmod | grep powernow_k8
powernow_k8 13992 0
freq_table 3716 2 powernow_k8,cpufreq_ondemand
processor 35864 2 powernow_k8

arkham ~ $ LANG=C cpufreq-info
cpufrequtils 005: cpufreq-info (C) Dominik Brodowski 2004-2006
Report errors and bugs to cpufreq@vger.kernel.org, please.
analyzing CPU 0:
driver: powernow-k8
CPUs which need to switch frequency at the same time: 0
hardware limits: 800 MHz - 1.80 GHz
available frequency steps: 1.80 GHz, 1.60 GHz, 800 MHz
available cpufreq governors: ondemand, conservative, powersave, userspace, performance
current policy: frequency should be within 800 MHz and 1.80 GHz.
The governor "powersave" may decide which speed to use
within this range.
current CPU frequency is 800 MHz.

Edit: OOOH, I'm using arch 32 bit though

Last edited by arkham (2009-06-28 14:15:56)

.:B:. · 2009-06-28 17:02:52

64 bit stuff > 64 bit forums. That's what we have them for .

As for the issue: I can confirm on 32 bit it works like a breeze too...

gsdefender · 2009-06-28 21:34:42

B wrote:

64 bit stuff > 64 bit forums. That's what we have them for .
.

I thought it could have been an architecture-independent issue, so I didn't even bother to think about posting in this 64-bit-only forum. Sorry. :-(

gsdefender · 2009-07-09 17:04:28

Has anyone tried to upgrade to 2.6.30.1 and see if that still happens?

Last edited by gsdefender (2009-07-09 17:05:31)

Arch Linux

#1 2009-06-28 13:09:21

2.6.30 kernel + powernow-k8 gives modprobe segmentation fault at boot

#2 2009-06-28 13:28:29

Re: 2.6.30 kernel + powernow-k8 gives modprobe segmentation fault at boot

#3 2009-06-28 14:14:59

Re: 2.6.30 kernel + powernow-k8 gives modprobe segmentation fault at boot

#4 2009-06-28 17:02:52

Re: 2.6.30 kernel + powernow-k8 gives modprobe segmentation fault at boot

#5 2009-06-28 21:34:42

Re: 2.6.30 kernel + powernow-k8 gives modprobe segmentation fault at boot

#6 2009-07-09 17:04:28

Re: 2.6.30 kernel + powernow-k8 gives modprobe segmentation fault at boot

Board footer