[Bug 220722] New: Non existent battery reporting for a laptop with touchscreen MSI

[Bug 220722] New: Non existent battery reporting for a laptop with touchscreen MSI

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220722

            Bug ID: 220722
           Summary: Non existent battery reporting for a laptop with
                    touchscreen MSI
           Product: Drivers
           Version: 2.5
          Hardware: Intel
                OS: Linux
            Status: NEW
          Severity: normal
          Priority: P3
         Component: Platform_x86
          Assignee: drivers_platform_x86@kernel-bugs.osdl.org
          Reporter: andretiagob@protonmail.com
        Regression: No

Created attachment 308869
  --> https://bugzilla.kernel.org/attachment.cgi?id=308869&action=edit
Battery charge information

This issue is similar to this bug report: 
https://bugzilla.kernel.org/show_bug.cgi?id=216747

The same issue was filed in the upower gitlab page by another user:
https://gitlab.freedesktop.org/upower/upower/-/issues/197


SUMMARY
The laptop touch screen always reports a non existent battery with a charge
level of 0% (see attachment). I had this issue for a long time and it is always
present regardless of kernel versions, older and newer, and on different
distributions. 
The device in question is called: ELAN Touchscreen Stylus.


I first issued this bug on the upower gitlab page: 
https://gitlab.freedesktop.org/upower/upower/-/issues/329


From what I've been told this might be an issue with some driver in the kernel.
Here's some information that might be useful:


Device:/org/freedesktop/UPower/devices/battery_hid_0003o04F3o2FFAx0001_battery
  native-path: hid-0003:04F3:2FFA.0001-battery
  model:                ELAN Touchscreen Stylus
  power supply:         no
  updated:              qui 01 jan 1970 01:00:00 (1761601761 seconds ago)
  has history:          yes
  has statistics:       yes
  tablet
    warning-level:       none
    percentage:          0%
    icon-name:          'battery-missing-symbolic'

-------------------------------------------------------------------------------

andre@fedora:~$ upower -e
/org/freedesktop/UPower/devices/battery_BAT
/org/freedesktop/UPower/devices/battery_hid_0003o04F3o2FFAx0002_battery
/org/freedesktop/UPower/devices/line_power_ADP1
/org/freedesktop/UPower/devices/DisplayDevice

-------------------------------------------------------------------------------

andre@fedora:~$ ls
/sys/devices/pci0000:00/0000:00:14.0/usb3/3-8/3-8:1.0/0003:04F3:2FFA.0002/power_supply/hid-0003:04F3:2FFA.0002-battery 
capacity  extensions  model_name  power   present  status     type    wakeup48 
device    hwmon4      online      powers  scope    subsystem  uevent\

-------------------------------------------------------------------------------

andre@fedora:~$ udevadm info
/sys/devices/pci0000:00/0000:00:14.0/usb3/3-8/3-8:1.0/0003:04F3:2FFA.0002/power_s
upply/hid-0003:04F3:2FFA.0002-battery 
P:
/devices/pci0000:00/0000:00:14.0/usb3/3-8/3-8:1.0/0003:04F3:2FFA.0002/power_supply/hid-0003:04F3:2FFA.0002> 
M: hid-0003:04F3:2FFA.0002-battery 
J: +power_supply:hid-0003:04F3:2FFA.0002-battery 
U: power_supply 
T: power_supply 
E:
DEVPATH=/devices/pci0000:00/0000:00:14.0/usb3/3-8/3-8:1.0/0003:04F3:2FFA.0002/power_supply/hid-0003:04F3:2> 
E: SUBSYSTEM=power_supply 
E: DEVTYPE=power_supply 
E: POWER_SUPPLY_NAME=hid-0003:04F3:2FFA.0002-battery 
E: POWER_SUPPLY_TYPE=Battery 
E: POWER_SUPPLY_STATUS=Unknown 
E: POWER_SUPPLY_PRESENT=1 
E: POWER_SUPPLY_ONLINE=1 
E: POWER_SUPPLY_CAPACITY=0 
E: POWER_SUPPLY_SCOPE=Device 
E: POWER_SUPPLY_MODEL_NAME=ELAN Touchscreen


Currently I'm using KDE Linux with kernel version 6.17.5.
This is an MSI Summit E16 laptop with a convertible touch screen (A12UCT mode).
I don't have any stylus pen that could be sending the wrong battery
information.


Any help is appreciated,

Thank you!

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 220722] Non existent battery reporting for a laptop with touchscreen MSI

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220722

Hans de Goede (hansg@kernel.org) changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |hansg@kernel.org

--- Comment #1 from Hans de Goede (hansg@kernel.org) ---
This is weird, the hid-battery code is supposed to ignore ELAN touchscreens
exactly because of this problem, see:

https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/hid/hid-input.c#n401

What kernel version are you using? What is the output of "uname -a" ?

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 220722] Non existent battery reporting for a laptop with touchscreen MSI

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220722

--- Comment #2 from André Barata (andretiagob@protonmail.com) ---
I'm using 6.17.5:

Linux Andre-PC 6.17.5-arch1-1 #1 SMP PREEMPT_DYNAMIC Thu, 23 Oct 2025 18:49:03
+0000 x86_64 GNU/Linux

In Fedora 42 and 43 the same bug appears.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 220722] Non existent battery reporting for a laptop with touchscreen MSI

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220722

Mario Limonciello (AMD) (mario.limonciello@amd.com) changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mario.limonciello@amd.com

--- Comment #3 from Mario Limonciello (AMD) (mario.limonciello@amd.com) ---
AFAICT his touchscreen is USB, that quirk is for I2C.  We probably need another
one for USB interface.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 220722] Non existent battery reporting for a laptop with touchscreen MSI

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220722

--- Comment #4 from Mario Limonciello (AMD) (mario.limonciello@amd.com) ---
Created attachment 308872
  --> https://bugzilla.kernel.org/attachment.cgi?id=308872&action=edit
potential patch

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 220722] Non existent battery reporting for a laptop with touchscreen MSI

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220722

--- Comment #5 from Mario Limonciello (AMD) (mario.limonciello@amd.com) ---
@André:

Can you test this patch and see if it helps?

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 220722] Non existent battery reporting for a laptop with touchscreen MSI

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220722

--- Comment #6 from André Barata (andretiagob@protonmail.com) ---
(In reply to Mario Limonciello (AMD) from comment #5)
> @André:
> 
> Can you test this patch and see if it helps?

Can you tell me how to use this patch?

Thank you

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 220722] Non existent battery reporting for a laptop with touchscreen MSI

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220722

--- Comment #7 from Mario Limonciello (AMD) (mario.limonciello@amd.com) ---
This should help you get started:

https://wiki.archlinux.org/title/Patching_packages

If in doubt; use an LLM like ChatGPT or Gemini to help you.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 220722] Non existent battery reporting for a laptop with touchscreen MSI

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220722

--- Comment #8 from Hans de Goede (hansg@kernel.org) ---
Mario, good catch on the existing quirk only being for I2C devices, the
proposed patch looks good to me:

Reviewed-by: Hans de Goede <hansg@kernel.org>

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 220722] Non existent battery reporting for a laptop with touchscreen MSI

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220722

--- Comment #9 from André Barata (andretiagob@protonmail.com) ---
(In reply to Mario Limonciello (AMD) from comment #7)
> This should help you get started:
> 
> https://wiki.archlinux.org/title/Patching_packages
> 
> If in doubt; use an LLM like ChatGPT or Gemini to help you.

(In reply to Hans de Goede from comment #8)
> Mario, good catch on the existing quirk only being for I2C devices, the
> proposed patch looks good to me:
> 
> Reviewed-by: Hans de Goede <hansg@kernel.org>

I can report that the patch fixed the issue!

Thank you for your support!!

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 220722] Non existent battery reporting for a laptop with touchscreen MSI

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220722

André Barata (andretiagob@protonmail.com) changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |CODE_FIX

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 220722] Non existent battery reporting for a laptop with touchscreen MSI

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220722

ggrundik@gmail.com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |ggrundik@gmail.com

--- Comment #10 from ggrundik@gmail.com ---
As I stated in https://bugzilla.kernel.org/show_bug.cgi?id=221118, its not the
fake battery, its the pen battery. All you need is just use the pen, touch the
screen by it, and this battery level will update according to the pen battery.

So now you removed that feature, and pen battery is no longer showing.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 220722] Non existent battery reporting for a laptop with touchscreen MSI

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220722

--- Comment #11 from Hans de Goede (hansg@kernel.org) ---
(In reply to ggrundik from comment #10)
> As I stated in https://bugzilla.kernel.org/show_bug.cgi?id=221118, its not
> the fake battery, its the pen battery. All you need is just use the pen,
> touch the screen by it, and this battery level will update according to the
> pen battery.
> 
> So now you removed that feature, and pen battery is no longer showing.

The problem is that without pen nearby the HID battery reports 0 charge instead
of not-present. This results in GNOME showing an low battery warning every time
the user logs into a session which is very annoying.

So until we've some better fix we need to choose between many users getting the
annoying low battery warning all the time; or the few users who actually use a
pen with these touchscreens (which is quite rare) not having battery info.

Having the chose at all between this is not ideal, but until we've a better fix
just hiding the broken battery reporting is the best we can do.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 220722] Non existent battery reporting for a laptop with touchscreen MSI

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220722

--- Comment #12 from ggrundik@gmail.com ---
Isn't this a GNOME problem, not a kernel one?

Current PCs can have quite a lot of battery-powered devices, apart of PC
itself: wireless controllers, mices, keyboards, pens, VR headsets, headphones,
etc. Most of them can report their battery status. If a desktop environment
cant separate them apart, and dont understand which is PC itself and which is
not... Its clearly fault of desktop environment, since its quite a common case.
KDE, for example, has no problems with that. This bug should be reported in
GNOME, not in kernel.

Blocking some device metrics in a kernel, just because some specific userspace
program incorrectly handles that in a weird way does not seems like a wise
solution. It is a perfectly functional, correct and useful metric.

This touchscreen is has two sensor layers: one for fingers plus one for stylus,
and it does report battery status of stylus. Communications between stylus and
the drawing surface can happen only when stylus is in range of touchscreen, its
a hardware limitation. So touchscreen does not have any information about
stylus until this communication happens at least once, but this does not means
that there are no stylus at all, or it is never in use. But after this
communication happens, touchscreen remembers last known status, and its
correctly shown as a battery level.

Before this change was made, this touchscreen was working as good as hardware
could. Now there are no pen battery status not only in GNOME, but nowhere.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 220722] Non existent battery reporting for a laptop with touchscreen MSI

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220722

--- Comment #13 from ggrundik@gmail.com ---
Best solution could have been to hotplug the stylus battery device when stylus
comes in range first time, but that would be quite a trick for the driver.

Less convenient solution could be to hide this battery readings, if stylus is
not in use. This is a userspace solution.

But current solution, to hide this battery always and from everyone, seems like
a worst one. At least there should be some override of this behavior apart of
recompiling the module.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 220722] Non existent battery reporting for a laptop with touchscreen MSI

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220722

--- Comment #14 from Hans de Goede (hansg@kernel.org) ---
(In reply to ggrundik from comment #12)
> Isn't this a GNOME problem, not a kernel one?
> 
> Current PCs can have quite a lot of battery-powered devices, apart of PC
> itself: wireless controllers, mices, keyboards, pens, VR headsets,
> headphones, etc. Most of them can report their battery status. If a desktop
> environment cant separate them apart, and dont understand which is PC itself
> and which is not... Its clearly fault of desktop environment, since its
> quite a common case. KDE, for example, has no problems with that. This bug
> should be reported in GNOME, not in kernel.

GNOME already tells all these different batteries apart. e.g. if my wireless
Logitech mouse battery is getting empty it will throw a "Logitech mouse battery
is getting low warning", not a laptop-empty battery warning which you seem to
think.

And in case of these ELAN touchscreens it throws a "ELAN Touchscreen Stylus
battery low" warning and it does so *all* the time, because upower is telling
it that there is a "ELAN Touchscreen Stylus" battery at 0%.

This is really not a GNOME problem, this needs to be fixed either in upower or
in the kernel. GNOME is just forwarding the info it is receiving from upower
and this works for all those other wireless kbds, mice, game controllers and
headphones you mention.

(In reply to ggrundik from comment #13)
> Best solution could have been to hotplug the stylus battery device when
> stylus comes in range first time, but that would be quite a trick for the
> driver.

And probably also hot unplug it when the stylus goes out of range. I would say
patches welcome.

> Less convenient solution could be to hide this battery readings, if stylus
> is not in use. This is a userspace solution.

Yes, fixing this in upower might also be a solution, see:

https://gitlab.freedesktop.org/upower/upower/-/issues/197

for an existing upower issue.

> But current solution, to hide this battery always and from everyone, seems
> like a worst one.

Patches to properly solve this are welcome. But until there are fixes to upower
telling GNOME that their is a "ELAN Touchscreen Stylus" battery at 0%, the
filtering out of the HID battery reporting on these touchscreens is staying in
place, since it is the least bad option of the bad choices we have atm.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 220722] Non existent battery reporting for a laptop with touchscreen MSI

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220722

--- Comment #15 from ggrundik@gmail.com ---
> And probably also hot unplug it when the stylus goes out of range. I would
> say patches welcome.

That would not be possible: as I said, touchscreen just remembers last know
status, so if stylus goes out of range, panel continues to report that last
value. There are no event of "stylus is out of range", since "the range" is
about half an inch or less, and getting in and out of it is a regular thing. As
I can see in HID sniffing, the only event is "stylus current status is", with
position, orientation, buttons and battery status, all in the same message.

> Yes, fixing this in upower might also be a solution, see:
> https://gitlab.freedesktop.org/upower/upower/-/issues/197 for an existing
> upower issue.

I see. And that issue is closed wont move, since kernel "fixed" the problem.
Thank you for this wisest nuclear solution, oh dear.

> Patches to properly solve this are welcome

I can only repeat: this was a correctly working situation: kernel was doing
perfectly fine, forwarding device status reported by the device. Device is
working fine, reporting everything it has as best as it could. Handling all
this data with its quirks is not the kernel's business, but userspace.

So removing this metric is a regression, please revert it.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.

[Bug 220722] Non existent battery reporting for a laptop with touchscreen MSI

[bugzilla-daemon]

https://bugzilla.kernel.org/show_bug.cgi?id=220722

--- Comment #16 from ggrundik@gmail.com ---
> And in case of these ELAN touchscreens it throws a "ELAN Touchscreen Stylus
> battery low" warning and it does so *all* the time, because upower is telling
> it that there is a "ELAN Touchscreen Stylus" battery at 0%.

KDE uses the same upower, but does not freaks out. So its a GNOME problem.

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.