Bug 58991 - [BISECTED]Notebook keyboard not working after update to 3.9.4
[BISECTED]Notebook keyboard not working after update to 3.9.4
Status: RESOLVED CODE_FIX
Product: Drivers
Classification: Unclassified
Component: Input Devices
All Linux
: P1 blocking
Assigned To: drivers_input-devices
https://bugzilla.redhat.com/show_bug....
:
: 65791 (view as bug list)
Depends on:
Blocks:
  Show dependency treegraph
 
Reported: 2013-05-29 20:30 UTC by SP
Modified: 2014-10-14 09:40 UTC (History)
9 users (show)

See Also:
Kernel Version: 3.10.5-201.fc19.i686
Tree: Fedora
Regression: Yes


Attachments
attachment-24629-0.html (2.08 KB, text/html)
2014-04-19 05:02 UTC, Davor
Details
Result of command dmidecode on HP Probook 4730s (8.44 KB, text/plain)
2014-06-15 19:45 UTC, fhoudebert
Details
attachment-10563-0.html (1.32 KB, text/html)
2014-10-14 09:40 UTC, Davor
Details

Description SP 2013-05-29 20:30:41 UTC
After updating the kernel to 3.9.4-200 no input is accepted from the integrated notebook keyboard at any init level or in rescue mode.
The keyboard only works if I fall back to 3.8.11-200.fc18.i686.

Excerpt from /etc/default/grub:
GRUB_TIMEOUT=5
GRUB_DISTRIBUTOR="Fedora"
GRUB_DEFAULT=saved
GRUB_CMDLINE_LINUX="rd.md=0 rd.lvm.lv=vg_mercury/lv_swap rd.dm=0 rd.lvm.lv=vg_mercury/lv_root KEYMAP=us FONT=latarcyrheb-sun16 rhgb rd.luks=0 LANG=en_US.UTF-8"

from lshw:
description: Notebook
    product: LS50-AE9KY
    vendor: LG Electronics Inc.
    version: Not Applicable
    serial: 503KIDE223554
    width: 32 bits
    capabilities: smbios-2.31 dmi-2.31

dmesg|grep keyboard
[    2.426475] input: AT Translated Set 2 keyboard as /devices/platform/i8042/serio0/input/input4
Comment 1 SP 2013-08-12 21:35:46 UTC
As of kernel 3.10.5-201.fc19.i686 this has still not been resolved.  I refer to 
https://bugzilla.redhat.com/show_bug.cgi?id=969550
Where there are many reports of this problem.
Comment 2 Eli Wapniarski 2013-09-23 04:06:42 UTC
This bug has been around for awhile guys.

I have a hard time believing that nobody in kernel land thinks that this is not important enough to bother with.

Please fix it.

Thanks
Comment 3 Alan 2013-11-13 20:08:53 UTC
The kernel bugzilla is just used to track problem. You need to discuss fixing it with whoever your support partner or distribution is.
Comment 4 Eli Wapniarski 2013-11-17 09:44:48 UTC
It is being tracked at:

https://bugzilla.redhat.com/show_bug.cgi?id=969550

The problem, as you can see has been round since the end May this year with kernel 3.9.4 and the kernel has gone through 2 major revisions since then with no fix in sight. The problem still persists with 3.11.8

We lowly users (bow bow - just kidding) have no ability fix this ourselves. Distributors have no way to fix this.

Whose gonna fix this???
Comment 5 Alan 2013-11-17 11:41:49 UTC
Your distribution if you are only running distribution kernels.

If you can verify the workaround and also build some kernels to bisect the problem then it may get more direct involvement.

Right now there is insufficient info to even guess why some random laptop with a load of out of tree drivers in it is breaking
Comment 6 Eli Wapniarski 2013-11-17 12:55:03 UTC
Thanks for your postive response.

I realize that getting to every laptop in the whole world would be difficult, but it would seem that this is a whole family of laptops. The problem appeared at at a specific juncture.

Kernels prior to 3.9.4-200.fc18.i686 worked. That is kernel-3.9.3-201.fc18 and previous kernels worked just fine. After 3.9.4-200.fc18.i686 and consquently 3.9.4-200.fc19.i686 the problem appears and persists to the current kernel build available. And the dmesg output is available over several kernel build in the redhat bugzilla bug report mentioned above.

It would be difficult to go back to fc18, as such, if you really do need a bisection with this info I will do what I can to get it done, provided I have some hand holding on how to go about doing it.

I'm looking forward to your response :)
Comment 7 Eli Wapniarski 2013-11-17 12:57:04 UTC
Oh.... I will see if the workaround allows for the keyboard to work. With the limitations mentioned in the redhat bugzilla bug report and get back to you within a day or so.

Thanks again.
Comment 8 Eli Wapniarski 2013-11-17 13:05:00 UTC
Just tested the workaround.

Indeed i8042.dumbkbd=1 allows me type with the built in keyboard, but, as reported, the [fn] key is disfuntional
Comment 9 Alan 2013-11-17 14:52:55 UTC
Ok thats very helpful as a starting point and playing "guess where the bug might be"

You shouldn't need to drop a release to test different kernels built for bisecting the problem - especially as it seems you just need to boot to a command line and press a few keys to check.

As you are running Fedora I think the relevant document and guide on bisection is

https://fedoraproject.org/wiki/User:Ignatenkobrain/Kernel/Bisection

http://mm3test.fedoraproject.org/hyperkitty/list/devel@mm3test.fedoraproject.org/thread/MNFJI4P7DHJRDANFDPUTZ2WNWX3P2OBV/

also has some discussion on doing it in Fedora

http://landley.net/writing/git-bisect-howto.html

is a general howto with more detail


Roughly speaking the idea is that you find a good kernel and a bad kernel and git bisect then does the minimal number of in betwen kernel builds to find which specific change to the kernel caused the problem.

Alan
Comment 10 Alan 2013-11-26 21:25:18 UTC
*** Bug 65791 has been marked as a duplicate of this bug. ***
Comment 11 Joonas Saarinen 2013-11-27 10:55:56 UTC
> After updating the kernel to 3.9.4-200 no input is accepted from the integrated notebook keyboard at any init level or in rescue mode.

BTW as I noted in that duplicate bug, 3.9.0 seems to be the exact version where the bug appears, at least for my LG LW25 laptop.
Comment 12 Joonas Saarinen 2013-12-01 15:53:59 UTC
A kernel bisect suggests that the first bad commit is:

be2d7e4233a4fe439125b825470020cef83407d5
Input: atkbd - fix multi-byte scancode handling on reconnect
Comment 13 Davor 2014-02-05 05:16:50 UTC
it's not just LG.
HP ProBook 4730s has the same problem.
Debian Testing amd64 with 3.12.6-2 kernel

$ dmesg|grep keyboard
[    1.076507] input: AT Translated Set 2 keyboard as /devices/platform/i8042/serio0/input/input0
[ 4104.610482] atkbd serio0: Failed to enable keyboard on isa0060/serio0
[17396.052022] atkbd serio0: Failed to deactivate keyboard on isa0060/serio0
[17396.360274] atkbd serio0: Failed to enable keyboard on isa0060/serio0
[34176.156200] atkbd serio0: Failed to enable keyboard on isa0060/serio0
Comment 14 Shengliang Song 2014-04-16 00:40:31 UTC
Hi,

I am working on a fix on the issue. 
If you see a similar keyboard issue, 
could you collect the output of "sudo dmidecode" from your labtop/computer?

Thanks,
Shengliang
Comment 15 Shengliang Song 2014-04-18 21:32:52 UTC
Hi Davor,

I am not sure the issue you saw is the same issue that Joonas reported.

Could you cherry-pick changes from here and help me to verify it on HP ProBook 4730s? 

https://github.com/Shengliang/linux/commit/ef2471339bbb0aeb3820cd3aeea442dec9417858


Thanks,
Shengliang Song
Comment 16 Davor 2014-04-19 05:02:40 UTC
Created attachment 132981 [details]
attachment-24629-0.html

i'm also starting to think that it's not the same bug.

actually at the moment the onboard keyboard is working. since i'm using an
USB keyboard i didnt notice when it started working again.
so you can remove my laptop form the atkbd_deactivate_rst_fixup

ill get with you in touch if it happens again.

thanks



On Fri, Apr 18, 2014 at 11:32 PM, <bugzilla-daemon@bugzilla.kernel.org>wrote:

> https://bugzilla.kernel.org/show_bug.cgi?id=58991
>
> --- Comment #15 from Shengliang Song <ssl@chromium.org> ---
> Hi Davor,
>
> I am not sure the issue you saw is the same issue that Joonas reported.
>
> Could you cherry-pick changes from here and help me to verify it on HP
> ProBook
> 4730s?
>
>
> https://github.com/Shengliang/linux/commit/ef2471339bbb0aeb3820cd3aeea442dec9417858
>
>
> Thanks,
> Shengliang Song
>
> --
> You are receiving this mail because:
> You are on the CC list for the bug.
>
Comment 17 fhoudebert 2014-06-15 19:45:19 UTC
Created attachment 139841 [details]
Result of command dmidecode on HP Probook 4730s

I have a similar issue on my laptop HP Probook 4730s with mint 17.
Hope it helps
Comment 18 frank 2014-10-02 11:39:48 UTC
Hi,
I'm actually searching for a solution for the keyboard malfunction of my LG S1 Pro Express Dual. The problem occurred after system upgrade from OpenSuse 12.3 to 13.1, which included the Kernel upgrade from 3.7.10-1 to 3.11.10-7.

I found that Shengliang did a correction with Kernel 3.11.10.11 (Input: atkbd - fix keyboard not working on some LG laptops).

Unfortunately, even with the actual Kernel of OpenSuse 3.11.10-21 it doesn't work.
Is there still anybody working on this issue?

Thanks
Frank
Comment 19 frank 2014-10-02 20:56:22 UTC
Hallo,

I found now in other threads (Fedora, Ubuntu) a workaround
'quiet splash i8042.dumbkbd=1' which works for me as well.
Besides some functionality is still missing, I'd really appreciate to solve this problem.
Release of OpenSuse 13.2 is planned shortly (in 32 Days).


Thanks
Frank
Comment 20 Dmitry Torokhov 2014-10-14 00:54:29 UTC
The problem on LG laptops should be fixed with:

commit c01206796139e2b1feb7539bc72174fef1c6dc6e
Author: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Date:   Wed Sep 10 13:50:37 2014 -0700                                                                                 
                                                                                                                       
    Input: atkbd - do not try 'deactivate' keyboard on any LG laptops                                                  
                                                                                                                       
    We are getting more and more reports about LG laptops not having                                                   
    functioning keyboard if we try to deactivate keyboard during probe.                                                 
    Given that having keyboard deactivated is merely "nice to have"                                                     
    instead of a hard requirement for probing, let's disable it on all                                                  
    LG boxes instead of trying to hunt down particular models.                                                           
                                                                                                                         
    This change is prompted by patches trying to add "LG Electronics"/"ROCKY"                                              
    and "LG Electronics"/"LW60-F27B" to the DMI list.                                                                       
                                                                                                                            
    https://bugzilla.kernel.org/show_bug.cgi?id=77051                                                                       
                                                                                                                            
    Cc: stable@vger.kernel.org                                                                                              
    Reported-by: Jaime Velasco Juan <jsagarribay@gmail.com>
    Reported-by: Georgios Tsalikis <georgios@tsalikis.net>
    Tested-by: Jaime Velasco Juan <jsagarribay@gmail.com>
    Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>

for all others please open a new issue.
Comment 21 Davor 2014-10-14 09:40:46 UTC
Created attachment 153731 [details]
attachment-10563-0.html

Javi:: vrućeadd e f cd obii čovićk grbv.d p::vkbtv
On Apr 16, 2014 2:40 AM, <bugzilla-daemon@bugzilla.kernel.org> wrote:

> https://bugzilla.kernel.org/show_bug.cgi?id=58991
>
> Shengliang Song <ssl@chromium.org> changed:
>
>            What    |Removed                     |Added
>
> ----------------------------------------------------------------------------
>                  CC|                            |ssl@chromium.org
>
> --- Comment #14 from Shengliang Song <ssl@chromium.org> ---
> Hi,
>
> I am working on a fix on the issue.
> If you see a similar keyboard issue,
> could you collect the output of "sudo dmidecode" from your labtop/computer?
>
> Thanks,
> Shengliang
>
> --
> You are receiving this mail because:
> You are on the CC list for the bug.
>

Note You need to log in before you can comment on or make changes to this bug.