Issue in asus_ec_sensors in Fedora installations and other distributions

Issue in asus_ec_sensors in Fedora installations and other distributions

[Christopher Klooz]

Hi Eugene,

Currently, we have user reports of at least 4 Fedora installations that 
have issues with asus_ec_sensors since 5.19.4. They remain in 5.19.6.

All issues can be solved by blacklisting asus_ec_sensors.

https://bugzilla.redhat.com/show_bug.cgi?id=2121844

The bugzilla report contains kernel logs of two machines, one with 
nvidia-driver and one without a tainted kernel.

The report and the contained ask.fedora link contain some elaborations 
of how the issue manifests at the users' machines. The major issue is a 
wrong output of battery power percentage (in one case it gets stuck, in 
the other cases is becomes erratic). Also, in one case it has effects on 
suspending (see the report).

On the Internet, there are already reports from other Linux 
distributions noting the issue and the same solution (blacklisting).

Thanks for maintaining & regards,

Chris

Re: Issue in asus_ec_sensors in Fedora installations and other distributions

[Guenter Roeck]

On 9/3/22 04:00, Christopher Klooz wrote:
> Hi Eugene,
> 
> Currently, we have user reports of at least 4 Fedora installations that have issues with asus_ec_sensors since 5.19.4. They remain in 5.19.6.
> 
> All issues can be solved by blacklisting asus_ec_sensors.
> 
> https://bugzilla.redhat.com/show_bug.cgi?id=2121844
> 
> The bugzilla report contains kernel logs of two machines, one with nvidia-driver and one without a tainted kernel.
> 
> The report and the contained ask.fedora link contain some elaborations of how the issue manifests at the users' machines. The major issue is a wrong output of battery power percentage (in one case it gets stuck, in the other cases is becomes erratic). Also, in one case it has effects on suspending (see the report).
> 
> On the Internet, there are already reports from other Linux distributions noting the issue and the same solution (blacklisting).
> 

Eugene,

do you have an immediate idea other than reverting the patch pointed to in the bug report ?

Thanks,
Guenter

Re: Issue in asus_ec_sensors in Fedora installations and other distributions

[Eugene Shalygin]

Guenter,

yes, there is a patch that I don't really understand, but it fixes a
very similar problem:
https://bugzilla.kernel.org/show_bug.cgi?id=216412 I will test today
that change with a hardware where it needs to autoload and submit the
patch if it works.

Best regards,
Eugene

On Sat, 3 Sept 2022 at 17:09, Guenter Roeck <linux@roeck-us.net> wrote:
>
> On 9/3/22 04:00, Christopher Klooz wrote:
> > Hi Eugene,
> >
> > Currently, we have user reports of at least 4 Fedora installations that have issues with asus_ec_sensors since 5.19.4. They remain in 5.19.6.
> >
> > All issues can be solved by blacklisting asus_ec_sensors.
> >
> > https://bugzilla.redhat.com/show_bug.cgi?id=2121844
> >
> > The bugzilla report contains kernel logs of two machines, one with nvidia-driver and one without a tainted kernel.
> >
> > The report and the contained ask.fedora link contain some elaborations of how the issue manifests at the users' machines. The major issue is a wrong output of battery power percentage (in one case it gets stuck, in the other cases is becomes erratic). Also, in one case it has effects on suspending (see the report).
> >
> > On the Internet, there are already reports from other Linux distributions noting the issue and the same solution (blacklisting).
> >
>
> Eugene,
>
> do you have an immediate idea other than reverting the patch pointed to in the bug report ?
>
> Thanks,
> Guenter
>

Re: Issue in asus_ec_sensors in Fedora installations and other distributions

[Guenter Roeck]

On 9/3/22 08:13, Eugene Shalygin wrote:
> Guenter,
> 
> yes, there is a patch that I don't really understand, but it fixes a
> very similar problem:
> https://bugzilla.kernel.org/show_bug.cgi?id=216412 I will test today
> that change with a hardware where it needs to autoload and submit the
> patch if it works.
> 

I don't understand why that works either. Maybe it interferes with other
PNP0C09 devices and/or the code in drivers/acpi/ec.c.

Couple of comments:

- You'll need to drop __init.
- No need or reason to rename asus_ec_sensors_platform_driver
   to asus_ec_sensors_platform_driver_probe.

Thanks,
Guenter

> Best regards,
> Eugene
> 
> On Sat, 3 Sept 2022 at 17:09, Guenter Roeck <linux@roeck-us.net> wrote:
>>
>> On 9/3/22 04:00, Christopher Klooz wrote:
>>> Hi Eugene,
>>>
>>> Currently, we have user reports of at least 4 Fedora installations that have issues with asus_ec_sensors since 5.19.4. They remain in 5.19.6.
>>>
>>> All issues can be solved by blacklisting asus_ec_sensors.
>>>
>>> https://bugzilla.redhat.com/show_bug.cgi?id=2121844
>>>
>>> The bugzilla report contains kernel logs of two machines, one with nvidia-driver and one without a tainted kernel.
>>>
>>> The report and the contained ask.fedora link contain some elaborations of how the issue manifests at the users' machines. The major issue is a wrong output of battery power percentage (in one case it gets stuck, in the other cases is becomes erratic). Also, in one case it has effects on suspending (see the report).
>>>
>>> On the Internet, there are already reports from other Linux distributions noting the issue and the same solution (blacklisting).
>>>
>>
>> Eugene,
>>
>> do you have an immediate idea other than reverting the patch pointed to in the bug report ?
>>
>> Thanks,
>> Guenter
>>

Re: Issue in asus_ec_sensors in Fedora installations and other distributions

[Christopher Klooz]

On 03/09/2022 17:13, Eugene Shalygin wrote:
> Guenter,
>
> yes, there is a patch that I don't really understand, but it fixes a
> very similar problem:
> https://bugzilla.kernel.org/show_bug.cgi?id=216412 I will test today
> that change with a hardware where it needs to autoload and submit the
> patch if it works.
>
> Best regards,
> Eugene
I have forwarded the patch to the Fedora users experiencing the issue, 
one seems to currently try it. I will let you know if I get a feedback 
(https://ask.fedoraproject.org/t/battery-indicator-not-working-properly/25946/28).
>
> On Sat, 3 Sept 2022 at 17:09, Guenter Roeck <linux@roeck-us.net> wrote:
>> On 9/3/22 04:00, Christopher Klooz wrote:
>>> Hi Eugene,
>>>
>>> Currently, we have user reports of at least 4 Fedora installations that have issues with asus_ec_sensors since 5.19.4. They remain in 5.19.6.
>>>
>>> All issues can be solved by blacklisting asus_ec_sensors.
>>>
>>> https://bugzilla.redhat.com/show_bug.cgi?id=2121844
>>>
>>> The bugzilla report contains kernel logs of two machines, one with nvidia-driver and one without a tainted kernel.
>>>
>>> The report and the contained ask.fedora link contain some elaborations of how the issue manifests at the users' machines. The major issue is a wrong output of battery power percentage (in one case it gets stuck, in the other cases is becomes erratic). Also, in one case it has effects on suspending (see the report).
>>>
>>> On the Internet, there are already reports from other Linux distributions noting the issue and the same solution (blacklisting).
>>>
>> Eugene,
>>
>> do you have an immediate idea other than reverting the patch pointed to in the bug report ?
>>
>> Thanks,
>> Guenter
>>

Re: Issue in asus_ec_sensors in Fedora installations and other distributions

[Eugene Shalygin]

> I don't understand why that works either. Maybe it interferes with other
> PNP0C09 devices and/or the code in drivers/acpi/ec.c.

Turns out the patch breaks sensors on a system where it should provide
output. The module gets loaded but provides no output (which should
mean the probe code is not executed). I will try to investigate
tomorrow.

> Couple of comments:
>
> - You'll need to drop __init.
> - No need or reason to rename asus_ec_sensors_platform_driver
>    to asus_ec_sensors_platform_driver_probe.
That's in order to not drop __init.

Best regards,
Eugene

Re: Issue in asus_ec_sensors in Fedora installations and other distributions

[Guenter Roeck]

On 9/3/22 12:40, Eugene Shalygin wrote:
>> I don't understand why that works either. Maybe it interferes with other
>> PNP0C09 devices and/or the code in drivers/acpi/ec.c.
> 
> Turns out the patch breaks sensors on a system where it should provide
> output. The module gets loaded but provides no output (which should
> mean the probe code is not executed). I will try to investigate
> tomorrow.
> 
>> Couple of comments:
>>
>> - You'll need to drop __init.
>> - No need or reason to rename asus_ec_sensors_platform_driver
>>     to asus_ec_sensors_platform_driver_probe.
> That's in order to not drop __init.
> 

What does the name of the struct platform_driver data structure
have to do with __init ?

Thanks,
Guenter

Re: Issue in asus_ec_sensors in Fedora installations and other distributions

[Eugene Shalygin]

On Sun, 4 Sept 2022 at 00:00, Guenter Roeck <linux@roeck-us.net> wrote:
> What does the name of the struct platform_driver data structure
> have to do with __init ?

That's what the modpost warning message suggests:
WARNING: modpost:
/home/eugene/develop/asus-ec-sensors/asus-ec-sensors.o(.data+0x60):
Section mismatch in reference from the variable
asus_ec_sensors_platform_driver to the function
.init.text:asus_ec_probe()
The variable asus_ec_sensors_platform_driver references
the function __init asus_ec_probe()
If the reference is valid then annotate the
variable with __init* or __refdata (see linux/init.h) or name the variable:
*_template, *_timer, *_sht, *_ops, *_probe, *_probe_one, *_console

Regards,
Eugene

Re: Issue in asus_ec_sensors in Fedora installations and other distributions

[Guenter Roeck]

On 9/3/22 15:37, Eugene Shalygin wrote:
> On Sun, 4 Sept 2022 at 00:00, Guenter Roeck <linux@roeck-us.net> wrote:
>> What does the name of the struct platform_driver data structure
>> have to do with __init ?
> 
> That's what the modpost warning message suggests:
> WARNING: modpost:
> /home/eugene/develop/asus-ec-sensors/asus-ec-sensors.o(.data+0x60):
> Section mismatch in reference from the variable
> asus_ec_sensors_platform_driver to the function
> .init.text:asus_ec_probe()
> The variable asus_ec_sensors_platform_driver references
> the function __init asus_ec_probe()
> If the reference is valid then annotate the
> variable with __init* or __refdata (see linux/init.h) or name the variable:
> *_template, *_timer, *_sht, *_ops, *_probe, *_probe_one, *_console
> 

A driver with a probe function can be unloaded and reloaded manually.
I don't see how you can do that if the functions are marked __init
and the driver is built into the kernel.

Besides that, I _really_ dislike side effect programming.

Guenter

Re: Issue in asus_ec_sensors in Fedora installations and other distributions #forregzbot

[Thorsten Leemhuis]

TWIMC: this mail is primarily send for documentation purposes and for
regzbot, my Linux kernel regression tracking bot. These mails usually
contain '#forregzbot' in the subject, to make them easy to spot and filter.

[TLDR: I'm adding this regression report to the list of tracked
regressions; all text from me you find below is based on a few templates
paragraphs you might have encountered already already in similar form.]

Hi, this is your Linux kernel regression tracker.

On 03.09.22 13:00, Christopher Klooz wrote:
> Hi Eugene,
> 
> Currently, we have user reports of at least 4 Fedora installations that
> have issues with asus_ec_sensors since 5.19.4. They remain in 5.19.6.
> 
> All issues can be solved by blacklisting asus_ec_sensors.
> 
> https://bugzilla.redhat.com/show_bug.cgi?id=2121844
> 
> The bugzilla report contains kernel logs of two machines, one with
> nvidia-driver and one without a tainted kernel.
> 
> The report and the contained ask.fedora link contain some elaborations
> of how the issue manifests at the users' machines. The major issue is a
> wrong output of battery power percentage (in one case it gets stuck, in
> the other cases is becomes erratic). Also, in one case it has effects on
> suspending (see the report).
> 
> On the Internet, there are already reports from other Linux
> distributions noting the issue and the same solution (blacklisting).
> 
> Thanks for maintaining & regards,

Thanks for the report. To be sure below issue doesn't fall through the
cracks unnoticed, I'm adding it to regzbot, my Linux kernel regression
tracking bot:

#regzbot introduced 5cd29012028d ^
https://bugzilla.redhat.com/show_bug.cgi?id=2121844
#regzbot title asus_ec_sensors: causes misbehaviour; likely dup of
https://bugzilla.kernel.org/show_bug.cgi?id=216412
#regzbot ignore-activity

This isn't a regression? This issue or a fix for it are already
discussed somewhere else? It was fixed already? You want to clarify when
the regression started to happen? Or point out I got the title or
something else totally wrong? Then just reply -- ideally with also
telling regzbot about it, as explained here:
https://linux-regtracking.leemhuis.info/tracked-regression/

Reminder for developers: When fixing the issue, add 'Link:' tags
pointing to the report (the mail this one replies to), as explained for
in the Linux kernel's documentation; above webpage explains why this is
important for tracked regressions.

Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker' hat)

P.S.: As the Linux kernel's regression tracker I deal with a lot of
reports and sometimes miss something important when writing mails like
this. If that's the case here, don't hesitate to tell me in a public
reply, it's in everyone's interest to set the public record straight.

Re: Issue in asus_ec_sensors in Fedora installations and other distributions

[Eugene Shalygin]

On Sun, 4 Sept 2022 at 01:47, Guenter Roeck <linux@roeck-us.net> wrote:

> A driver with a probe function can be unloaded and reloaded manually.
> I don't see how you can do that if the functions are marked __init
> and the driver is built into the kernel.

Did that and a user tested those changes. Unfortunately, the ACPI EC
is still inoperational if a module uses the ACPI id of the EC device
and its probe function returns an error. Looks like I have to change
MODULE_DEVICE_TABLE completely.

Regards,
Eugene

Re: Issue in asus_ec_sensors in Fedora installations and other distributions

[Guenter Roeck]

On 9/4/22 06:19, Eugene Shalygin wrote:
> On Sun, 4 Sept 2022 at 01:47, Guenter Roeck <linux@roeck-us.net> wrote:
> 
>> A driver with a probe function can be unloaded and reloaded manually.
>> I don't see how you can do that if the functions are marked __init
>> and the driver is built into the kernel.
> 
> Did that and a user tested those changes. Unfortunately, the ACPI EC
> is still inoperational if a module uses the ACPI id of the EC device
> and its probe function returns an error. Looks like I have to change
> MODULE_DEVICE_TABLE completely.
> 

If you insist on the __probe name hack, please add a note to the code
explaining why that name is necessary. There is only one single driver
besides this using that trick (I'd call it hack), hardly anyone will
know about it, and thus it needs to be explained in the driver to
prevent that the structure is renamed.

Having said that, how did you de-instantiate the device when it was
built into the kernel ?

Thanks,
Guenter

Re: Issue in asus_ec_sensors in Fedora installations and other distributions

[Guenter Roeck]

On Sat, Sep 03, 2022 at 01:00:33PM +0200, Christopher Klooz wrote:
> Hi Eugene,
> 
> Currently, we have user reports of at least 4 Fedora installations that have
> issues with asus_ec_sensors since 5.19.4. They remain in 5.19.6.
> 
> All issues can be solved by blacklisting asus_ec_sensors.
> 
> https://bugzilla.redhat.com/show_bug.cgi?id=2121844
> 
> The bugzilla report contains kernel logs of two machines, one with
> nvidia-driver and one without a tainted kernel.
> 
> The report and the contained ask.fedora link contain some elaborations of
> how the issue manifests at the users' machines. The major issue is a wrong
> output of battery power percentage (in one case it gets stuck, in the other
> cases is becomes erratic). Also, in one case it has effects on suspending
> (see the report).
> 
> On the Internet, there are already reports from other Linux distributions
> noting the issue and the same solution (blacklisting).
> 
> Thanks for maintaining & regards,
> 

Are we getting any closer to fixing this problem ? If not, I'll have to start
reverting patches or even declare the driver as BROKEN.

Thanks,
Guenter

Re: Issue in asus_ec_sensors in Fedora installations and other distributions

[Eugene Shalygin]

On Tue, 6 Sept 2022 at 22:10, Guenter Roeck <linux@roeck-us.net> wrote:
> Are we getting any closer to fixing this problem ? If not, I'll have to start
> reverting patches or even declare the driver as BROKEN.

Out of a few workarounds I've tried, none worked. I rework the module
autoloading back to plain dmi data. Sorry, I'm totally busy until this
Friday and can't prepare a patch until the end of the week. If that is
too late, removing just the module autoloading
(MODULE_DEVICE_TABLE(acpi, ...)) will fix the problem.

Regards,
Eugene

Re: Issue in asus_ec_sensors in Fedora installations and other distributions

[Guenter Roeck]

On Tue, Sep 06, 2022 at 10:23:08PM +0200, Eugene Shalygin wrote:
> On Tue, 6 Sept 2022 at 22:10, Guenter Roeck <linux@roeck-us.net> wrote:
> > Are we getting any closer to fixing this problem ? If not, I'll have to start
> > reverting patches or even declare the driver as BROKEN.
> 
> Out of a few workarounds I've tried, none worked. I rework the module
> autoloading back to plain dmi data. Sorry, I'm totally busy until this
> Friday and can't prepare a patch until the end of the week. If that is
> too late, removing just the module autoloading
> (MODULE_DEVICE_TABLE(acpi, ...)) will fix the problem.
> 

Let's see if this works:

#regzbot fixed-by: 88700d1396ba

Guenter