Bananian Linux - Bananian Linux
View Issue Details
0000055Bananian Linux[All Projects] Kernelpublic2014-10-11 12:042016-03-25 09:19
Matthias 
Nico 
normalminorhave not tried
resolvedfixed 
 
14.09 
16.0416.04 
0000055: Division by zero in kernel. (3.4.90+)
There are divisions by zero if my banana pi is a little bit busy. (For example transfers with rsync)

cat /proc/version
Linux version 3.4.90+ (root@nico-desktop) (gcc version 4.8.2 (Ubuntu/Linaro 4.8.2-16ubuntu4) ) #1 SMP PREEMPT Fri Sep 12 18:13:45 CEST 2014
dmesg output:

[Sa Okt 11 11:49:03 2014] Division by zero in kernel.
[Sa Okt 11 11:49:03 2014] [<c0015018>] (unwind_backtrace+0x0/0x134) from [<c02c7658>] (Ldiv0+0x8/0x10)
[Sa Okt 11 11:49:03 2014] [<c02c7658>] (Ldiv0+0x8/0x10) from [<c00e7944>] (vmpressure_work_fn+0x6c/0x180)
[Sa Okt 11 11:49:03 2014] [<c00e7944>] (vmpressure_work_fn+0x6c/0x180) from [<c004c74c>] (process_one_work+0x12c/0x3b0)
[Sa Okt 11 11:49:03 2014] [<c004c74c>] (process_one_work+0x12c/0x3b0) from [<c004cd24>] (worker_thread+0x158/0x394)
[Sa Okt 11 11:49:03 2014] [<c004cd24>] (worker_thread+0x158/0x394) from [<c0051300>] (kthread+0x90/0x94)
[Sa Okt 11 11:49:03 2014] [<c0051300>] (kthread+0x90/0x94) from [<c000f438>] (kernel_thread_exit+0x0/0x8)
[Sa Okt 11 11:49:03 2014] Division by zero in kernel.
[Sa Okt 11 11:49:03 2014] [<c0015018>] (unwind_backtrace+0x0/0x134) from [<c02c7658>] (Ldiv0+0x8/0x10)
[Sa Okt 11 11:49:03 2014] [<c02c7658>] (Ldiv0+0x8/0x10) from [<c00e7958>] (vmpressure_work_fn+0x80/0x180)
[Sa Okt 11 11:49:03 2014] [<c00e7958>] (vmpressure_work_fn+0x80/0x180) from [<c004c74c>] (process_one_work+0x12c/0x3b0)
[Sa Okt 11 11:49:03 2014] [<c004c74c>] (process_one_work+0x12c/0x3b0) from [<c004cd24>] (worker_thread+0x158/0x394)
[Sa Okt 11 11:49:03 2014] [<c004cd24>] (worker_thread+0x158/0x394) from [<c0051300>] (kthread+0x90/0x94)
[Sa Okt 11 11:49:03 2014] [<c0051300>] (kthread+0x90/0x94) from [<c000f438>] (kernel_thread_exit+0x0/0x8)
No tags attached.
Issue History
2014-10-11 12:04MatthiasNew Issue
2014-10-19 12:37NicoAssigned To => Nico
2014-10-19 12:37NicoStatusnew => assigned
2014-10-19 12:38NicoNote Added: 0000058
2014-10-21 18:39MaykNote Added: 0000060
2014-11-20 09:00NicoNote Added: 0000089
2014-12-17 22:07MatthiasNote Added: 0000110
2015-01-02 11:40NicoNote Added: 0000121
2015-01-02 11:40NicoStatusassigned => resolved
2015-01-02 11:40NicoResolutionopen => not fixable
2015-05-04 12:09MatthiasNote Added: 0000243
2015-05-04 12:09MatthiasStatusresolved => feedback
2015-05-04 12:09MatthiasResolutionnot fixable => reopened
2015-05-05 11:59NicoNote Added: 0000244
2015-07-12 20:42NicoNote Added: 0000256
2015-07-12 20:42NicoStatusfeedback => resolved
2015-07-12 20:42NicoResolutionreopened => not fixable
2016-03-25 09:19NicoStatusresolved => feedback
2016-03-25 09:19NicoResolutionnot fixable => reopened
2016-03-25 09:19NicoNote Added: 0000345
2016-03-25 09:19NicoStatusfeedback => resolved
2016-03-25 09:19NicoFixed in Version => 16.04
2016-03-25 09:19NicoResolutionreopened => fixed
2016-03-25 09:19NicoTarget Version => 16.04

Notes
(0000058)
Nico   
2014-10-19 12:38   
Can you reproduce the errors with Raspbian and/or Lubuntu?

I can't reproduce the error on my system.
(0000060)
Mayk   
2014-10-21 18:39   
Hello Matthias and Nico,

I confirm this bug is in current raspian:

[23635.931454] Division by zero in kernel.
[23635.943126] [<c0015018>] (unwind_backtrace+0x0/0x134) from [<c02c7658>] (Ldiv0+0x8/0x10)
[23635.951197] [<c02c7658>] (Ldiv0+0x8/0x10) from [<c00e7944>] (vmpressure_work_fn+0x6c/0x180)
[23635.959973] [<c00e7944>] (vmpressure_work_fn+0x6c/0x180) from [<c004c74c>] (process_one_work+0x12c/0x3b0)
[23635.967975] [<c004c74c>] (process_one_work+0x12c/0x3b0) from [<c004cd24>] (worker_thread+0x158/0x394)
[23635.974955] [<c004cd24>] (worker_thread+0x158/0x394) from [<c0051300>] (kthread+0x90/0x94)
[23635.982761] [<c0051300>] (kthread+0x90/0x94) from [<c000f438>] (kernel_thread_exit+0x0/0x8)
[23635.985780] Division by zero in kernel.
[23635.992973] [<c0015018>] (unwind_backtrace+0x0/0x134) from [<c02c7658>] (Ldiv0+0x8/0x10)
[23636.000425] [<c02c7658>] (Ldiv0+0x8/0x10) from [<c00e7958>] (vmpressure_work_fn+0x80/0x180)
[23636.009082] [<c00e7958>] (vmpressure_work_fn+0x80/0x180) from [<c004c74c>] (process_one_work+0x12c/0x3b0)
[23636.017467] [<c004c74c>] (process_one_work+0x12c/0x3b0) from [<c004cd24>] (worker_thread+0x158/0x394)
[23636.024636] [<c004cd24>] (worker_thread+0x158/0x394) from [<c0051300>] (kthread+0x90/0x94)
[23636.032114] [<c0051300>] (kthread+0x90/0x94) from [<c000f438>] (kernel_thread_exit+0x0/0x8)

It happened under medium to heavy load while binary-comparing two massive directory trees plus occasional mounting/unmounting of an nfs share. I'm not sure what triggered it, the filesystem I was running the compare from is btrfs,
(0000089)
Nico   
2014-11-20 09:00   
Can you please test with the upcoming 14.11 release?
(0000110)
Matthias   
2014-12-17 22:07   
Hello guys!

I can still confirm this bug in the release 14.11. This time I was copying files with rsync from A to B.

A = DD-Image (ext4) encrypted with dm-crypt on a real ext4 partition (extended USB drive)
B = DD-Image (btrfs) encrypted with dm-crypt on a real ext4 partition (extended USB drive)

cat /proc/version
Linux version 3.4.104+ (root@nico-desktop) (gcc version 4.8.2 (Ubuntu/Linaro 4.8.2-16ubuntu4) ) #3 SMP PREEMPT Wed Nov 19 08:28:34 CET 2014

cat /etc/bananian_version 
141101


dmesg output:

[Mi Dez 17 18:08:42 2014] Division by zero in kernel.
[Mi Dez 17 18:08:42 2014] [<c0015058>] (unwind_backtrace+0x0/0x134) from [<c02c8c18>] (Ldiv0+0x8/0x10)
[Mi Dez 17 18:08:42 2014] [<c02c8c18>] (Ldiv0+0x8/0x10) from [<c00e7fe0>] (vmpressure_work_fn+0x6c/0x180)
[Mi Dez 17 18:08:42 2014] [<c00e7fe0>] (vmpressure_work_fn+0x6c/0x180) from [<c004c7d4>] (process_one_work+0x12c/0x3b0)
[Mi Dez 17 18:08:42 2014] [<c004c7d4>] (process_one_work+0x12c/0x3b0) from [<c004cdac>] (worker_thread+0x158/0x394)
[Mi Dez 17 18:08:42 2014] [<c004cdac>] (worker_thread+0x158/0x394) from [<c0051388>] (kthread+0x90/0x94)
[Mi Dez 17 18:08:42 2014] [<c0051388>] (kthread+0x90/0x94) from [<c000f478>] (kernel_thread_exit+0x0/0x8)
[Mi Dez 17 18:08:42 2014] Division by zero in kernel.
[Mi Dez 17 18:08:42 2014] [<c0015058>] (unwind_backtrace+0x0/0x134) from [<c02c8c18>] (Ldiv0+0x8/0x10)
[Mi Dez 17 18:08:42 2014] [<c02c8c18>] (Ldiv0+0x8/0x10) from [<c00e7ff4>] (vmpressure_work_fn+0x80/0x180)
[Mi Dez 17 18:08:42 2014] [<c00e7ff4>] (vmpressure_work_fn+0x80/0x180) from [<c004c7d4>] (process_one_work+0x12c/0x3b0)
[Mi Dez 17 18:08:42 2014] [<c004c7d4>] (process_one_work+0x12c/0x3b0) from [<c004cdac>] (worker_thread+0x158/0x394)
[Mi Dez 17 18:08:42 2014] [<c004cdac>] (worker_thread+0x158/0x394) from [<c0051388>] (kthread+0x90/0x94)
[Mi Dez 17 18:08:42 2014] [<c0051388>] (kthread+0x90/0x94) from [<c000f478>] (kernel_thread_exit+0x0/0x8)
(0000121)
Nico   
2015-01-02 11:40   
This has to be fixed upstream. I am not a kernel developer, sorry.
(0000243)
Matthias   
2015-05-04 12:09   
Hello!

This bug is still annoying me. Where can I report it? Is it really a problem of the vanilla kernel (kernel.org)?


cat /proc/version
Linux version 3.4.104-bananian (root@bananapi) (gcc version 4.6.3 (Debian 4.6.3-14) ) #1 SMP PREEMPT Mon Apr 6 18:25:40 UTC 2015


cat /etc/bananian_version
150401


dmesg output:
[Sa Mai  2 22:58:13 2015] Division by zero in kernel.
[Sa Mai  2 22:58:13 2015] [<c0014e78>] (unwind_backtrace+0x0/0x138) from [<c02ccb58>] (Ldiv0+0x8/0x10)
[Sa Mai  2 22:58:13 2015] [<c02ccb58>] (Ldiv0+0x8/0x10) from [<c00e4c50>] (vmpressure_work_fn+0x60/0x174)
[Sa Mai  2 22:58:13 2015] [<c00e4c50>] (vmpressure_work_fn+0x60/0x174) from [<c004b034>] (process_one_work+0x118/0x388)
[Sa Mai  2 22:58:13 2015] [<c004b034>] (process_one_work+0x118/0x388) from [<c004b648>] (worker_thread+0x16c/0x36c)
[Sa Mai  2 22:58:13 2015] [<c004b648>] (worker_thread+0x16c/0x36c) from [<c004fc34>] (kthread+0x84/0x90)
[Sa Mai  2 22:58:13 2015] [<c004fc34>] (kthread+0x84/0x90) from [<c000f4dc>] (kernel_thread_exit+0x0/0x8)
[Sa Mai  2 22:58:13 2015] Division by zero in kernel.
[Sa Mai  2 22:58:13 2015] [<c0014e78>] (unwind_backtrace+0x0/0x138) from [<c02ccb58>] (Ldiv0+0x8/0x10)
[Sa Mai  2 22:58:13 2015] [<c02ccb58>] (Ldiv0+0x8/0x10) from [<c00e4c64>] (vmpressure_work_fn+0x74/0x174)
[Sa Mai  2 22:58:13 2015] [<c00e4c64>] (vmpressure_work_fn+0x74/0x174) from [<c004b034>] (process_one_work+0x118/0x388)
[Sa Mai  2 22:58:13 2015] [<c004b034>] (process_one_work+0x118/0x388) from [<c004b648>] (worker_thread+0x16c/0x36c)
[Sa Mai  2 22:58:13 2015] [<c004b648>] (worker_thread+0x16c/0x36c) from [<c004fc34>] (kthread+0x84/0x90)
[Sa Mai  2 22:58:13 2015] [<c004fc34>] (kthread+0x84/0x90) from [<c000f4dc>] (kernel_thread_exit+0x0/0x8)
[Mo Mai  4 03:04:55 2015] Division by zero in kernel.
[Mo Mai  4 03:04:55 2015] [<c0014e78>] (unwind_backtrace+0x0/0x138) from [<c02ccb58>] (Ldiv0+0x8/0x10)
[Mo Mai  4 03:04:55 2015] [<c02ccb58>] (Ldiv0+0x8/0x10) from [<c00e4c50>] (vmpressure_work_fn+0x60/0x174)
[Mo Mai  4 03:04:55 2015] [<c00e4c50>] (vmpressure_work_fn+0x60/0x174) from [<c004b034>] (process_one_work+0x118/0x388)
[Mo Mai  4 03:04:55 2015] [<c004b034>] (process_one_work+0x118/0x388) from [<c004b648>] (worker_thread+0x16c/0x36c)
[Mo Mai  4 03:04:55 2015] [<c004b648>] (worker_thread+0x16c/0x36c) from [<c004fc34>] (kthread+0x84/0x90)
[Mo Mai  4 03:04:55 2015] [<c004fc34>] (kthread+0x84/0x90) from [<c000f4dc>] (kernel_thread_exit+0x0/0x8)
[Mo Mai  4 03:04:55 2015] Division by zero in kernel.
[Mo Mai  4 03:04:55 2015] [<c0014e78>] (unwind_backtrace+0x0/0x138) from [<c02ccb58>] (Ldiv0+0x8/0x10)
[Mo Mai  4 03:04:55 2015] [<c02ccb58>] (Ldiv0+0x8/0x10) from [<c00e4c64>] (vmpressure_work_fn+0x74/0x174)
[Mo Mai  4 03:04:55 2015] [<c00e4c64>] (vmpressure_work_fn+0x74/0x174) from [<c004b034>] (process_one_work+0x118/0x388)
[Mo Mai  4 03:04:55 2015] [<c004b034>] (process_one_work+0x118/0x388) from [<c004b648>] (worker_thread+0x16c/0x36c)
[Mo Mai  4 03:04:55 2015] [<c004b648>] (worker_thread+0x16c/0x36c) from [<c004fc34>] (kthread+0x84/0x90)
[Mo Mai  4 03:04:55 2015] [<c004fc34>] (kthread+0x84/0x90) from [<c000f4dc>] (kernel_thread_exit+0x0/0x8)
(0000244)
Nico   
2015-05-05 11:59   
Raspbian seems to have the same problem so I would start testing to reproduce the problem with the mainline kernel 3.4.x from kernel.org

If there is a problem too, report it to the mainline kernel, otherwise it looks like a sunxi problem and you have to report it to linux-sunxi.org
(0000256)
Nico   
2015-07-12 20:42   
closed again because we can't fix it.
(0000345)
Nico   
2016-03-25 09:19   
https://github.com/Bananian/linux-bananapi/commit/4ed9b51fe95e00c882ba10659b1d828fee0f2ba5 [^]
https://github.com/Bananian/linux-bananapi/commit/4f1825f7c3753ccb7bde67967051b1f3bd15f71f [^]