Regression: Backlight control broken on HP Omen 15‑ax246nf with kernel 7.0.0‑28‑generic

Regression: Backlight control broken on HP Omen 15‑ax246nf with kernel 7.0.0‑28‑generic

[harryblanc]

Hello,

I would like to report a regression affecting backlight control on my HP 
Omen laptop when using kernel 7.0.0‑28‑generic (Ubuntu 24.04 base, Linux 
Mint 22.3 XFCE).

Summary of the issue:
With kernel 7.0, the screen backlight is stuck at maximum brightness.
Brightness keys show the OSD and the slider moves, but the actual 
backlight does not change.
With kernel 6.17, brightness works normally.

Hardware:
- HP Omen 15‑ax246nf
- Intel i5‑7300HQ
- Intel HD Graphics 630 (i915)
- Nvidia GTX 1050 Mobile (disabled via Optimus)
- Linux Mint 22.3 XFCE (Ubuntu 24.04 base)

Backlight controller:
/sys/class/backlight/ → intel_backlight
acpi_video0 is no longer present under kernel 7.0.

ACPI errors visible only with kernel 7.0:
ACPI BIOS Error (bug): Attempt to CreateField of length zero
ACPI Error: Aborting method \_SB.WMID.WQBK
ACPI BIOS Error (bug): AE_AML_BUFFER_LIMIT
ACPI Error: Aborting method \_SB.WMID.HWMC
ACPI: video: [Firmware Bug]: ACPI(PEGP) defines _DOD but not _DOS

Kernel version:
Linux 7.0.0-28-generic #28~24.04.1-Ubuntu SMP PREEMPT_DYNAMIC Wed Jul 1 
15:50:57 UTC 2024 x86_64 GNU/Linux

GPU information (inxi -Gxxx):
Device-1: Intel HD Graphics 630 driver: i915
Device-2: NVIDIA GTX 1050 Mobile driver: N/A (Optimus disabled)
Display: X11 driver: modesetting, iris
Monitor: eDP-1 1920x1080 LG Display

Hardware summary (lshw -short):
Intel(R) Core(TM) i5-7300HQ
HD Graphics 630
GeForce GTX 1050 Mobile (disabled)
16 GiB DDR4
Samsung NVMe SSD + 1TB HDD

Expected behavior:
Backlight should be adjustable via intel_backlight or acpi_video0, as in 
kernel 6.17.

Observed behavior:
Backlight remains at maximum brightness regardless of ACPI events or 
brightness keys.

Steps to reproduce:
1. Boot kernel 7.0.0‑28‑generic
2. Attempt to change brightness
3. Backlight does not change

This appears to be an ACPI/i915 regression introduced in kernel 7.0.

Thank you for your work.

Re: Regression: Backlight control broken on HP Omen 15‑ax246nf with kernel 7.0.0‑28‑generic

[Thorsten Leemhuis]

On 7/18/26 11:05, harryblanc wrote:
> 
> I would like to report a regression affecting backlight control on my HP
> Omen laptop when using kernel 7.0.0‑28‑generic (Ubuntu 24.04 base, Linux
> Mint 22.3 XFCE).

Hi! Thx for your report. But this is a heavily patched vendor kernel and
also a series that is EOL. Please test latest vanilla 7.2-rc or and
report back. Ciao, Thorsten
> Summary of the issue:
> With kernel 7.0, the screen backlight is stuck at maximum brightness.
> Brightness keys show the OSD and the slider moves, but the actual
> backlight does not change.
> With kernel 6.17, brightness works normally.
> 
> Hardware:
> - HP Omen 15‑ax246nf
> - Intel i5‑7300HQ
> - Intel HD Graphics 630 (i915)
> - Nvidia GTX 1050 Mobile (disabled via Optimus)
> - Linux Mint 22.3 XFCE (Ubuntu 24.04 base)
> 
> Backlight controller:
> /sys/class/backlight/ → intel_backlight
> acpi_video0 is no longer present under kernel 7.0.
> 
> ACPI errors visible only with kernel 7.0:
> ACPI BIOS Error (bug): Attempt to CreateField of length zero
> ACPI Error: Aborting method \_SB.WMID.WQBK
> ACPI BIOS Error (bug): AE_AML_BUFFER_LIMIT
> ACPI Error: Aborting method \_SB.WMID.HWMC
> ACPI: video: [Firmware Bug]: ACPI(PEGP) defines _DOD but not _DOS
> 
> Kernel version:
> Linux 7.0.0-28-generic #28~24.04.1-Ubuntu SMP PREEMPT_DYNAMIC Wed Jul 1
> 15:50:57 UTC 2024 x86_64 GNU/Linux
> 
> GPU information (inxi -Gxxx):
> Device-1: Intel HD Graphics 630 driver: i915
> Device-2: NVIDIA GTX 1050 Mobile driver: N/A (Optimus disabled)
> Display: X11 driver: modesetting, iris
> Monitor: eDP-1 1920x1080 LG Display
> 
> Hardware summary (lshw -short):
> Intel(R) Core(TM) i5-7300HQ
> HD Graphics 630
> GeForce GTX 1050 Mobile (disabled)
> 16 GiB DDR4
> Samsung NVMe SSD + 1TB HDD
> 
> Expected behavior:
> Backlight should be adjustable via intel_backlight or acpi_video0, as in
> kernel 6.17.
> 
> Observed behavior:
> Backlight remains at maximum brightness regardless of ACPI events or
> brightness keys.
> 
> Steps to reproduce:
> 1. Boot kernel 7.0.0‑28‑generic
> 2. Attempt to change brightness
> 3. Backlight does not change
> 
> This appears to be an ACPI/i915 regression introduced in kernel 7.0.
> 
> Thank you for your work.
> 

Re: Regression: Backlight control broken on HP Omen 15‑ax246nf with kernel 7.0.0‑28‑generic

[Rafael J. Wysocki (Intel)]

On Wed, Jul 22, 2026 at 7:26 AM Thorsten Leemhuis
<regressions@leemhuis.info> wrote:
>
> On 7/18/26 11:05, harryblanc wrote:
> >
> > I would like to report a regression affecting backlight control on my HP
> > Omen laptop when using kernel 7.0.0‑28‑generic (Ubuntu 24.04 base, Linux
> > Mint 22.3 XFCE).
>
> Hi! Thx for your report. But this is a heavily patched vendor kernel and
> also a series that is EOL. Please test latest vanilla 7.2-rc or and
> report back. Ciao, Thorsten

I was about to ask for that too, so thanks for mentioning it.

> > Summary of the issue:
> > With kernel 7.0, the screen backlight is stuck at maximum brightness.
> > Brightness keys show the OSD and the slider moves, but the actual
> > backlight does not change.
> > With kernel 6.17, brightness works normally.
> >
> > Hardware:
> > - HP Omen 15‑ax246nf
> > - Intel i5‑7300HQ
> > - Intel HD Graphics 630 (i915)
> > - Nvidia GTX 1050 Mobile (disabled via Optimus)
> > - Linux Mint 22.3 XFCE (Ubuntu 24.04 base)
> >
> > Backlight controller:
> > /sys/class/backlight/ → intel_backlight
> > acpi_video0 is no longer present under kernel 7.0.
> >
> > ACPI errors visible only with kernel 7.0:
> > ACPI BIOS Error (bug): Attempt to CreateField of length zero
> > ACPI Error: Aborting method \_SB.WMID.WQBK
> > ACPI BIOS Error (bug): AE_AML_BUFFER_LIMIT
> > ACPI Error: Aborting method \_SB.WMID.HWMC
> > ACPI: video: [Firmware Bug]: ACPI(PEGP) defines _DOD but not _DOS

So it looks like 7.0 uses a different ACPI object as the video bus
representation.

Please send the output of "ls -l /sys/bus/auxiliary/devices/*/" from
the failing kernel.

If you are still able to boot the 6.17-based kernel on the affected
machine, please send the output of "ls -l
/sys/bus/acpi/drivers/video/" obtained while it is running.

Re: Regression: Backlight control broken on HP Omen 15‑ax246nf with kernel 7.0.0‑28‑generic

[harryblanc]

*Subject:* Re: Regression: Backlight control broken on HP Omen 
15‑ax246nf with kernel 7.0.x

Hello Rafael, Thorsten,

Here are the outputs you requested, from both the working kernel 
(6.17.0‑42) and the failing one (7.0.0‑29).

For context:

  *

    Secure Boot is *disabled* on this machine.

  *

    GPU setup is *Optimus with NVIDIA GTX 1050 “on‑demand” mode* (Intel
    iGPU is primary).

  *

    These changes do not affect the issue: the backlight regression
    remains exactly the same on kernel 7.0.29.

  *

    Kernel 6.17.0‑42 (backlight working normally)

  *

    $ ls -l /sys/bus/acpi/drivers/video/
    total 0
    --w------- 1 root root 4096 août 18 14:50 bind
    lrwxrwxrwx 1 root root    0 août 18 14:50 LNXVIDEO:00 ->
    ../../../../devices/LNXSYSTM:00/LNXSYBUS:00/PNP0A08:00/device:00/LNXVIDEO:00
    lrwxrwxrwx 1 root root    0 août 18 14:50 LNXVIDEO:01 ->
    ../../../../devices/LNXSYSTM:00/LNXSYBUS:00/PNP0A08:00/LNXVIDEO:01
    lrwxrwxrwx 1 root root    0 août 18 14:43 module ->
    ../../../../module/video
    --w------- 1 root root 4096 août 18 14:50 uevent
    --w------- 1 root root 4096 août 18 14:50 unbind

  *

    The ACPI video driver attaches normally, and backlight control works
    as expected.

  *

    Kernel 7.0.0‑29 (backlight stuck at maximum)

  *

    $ ls -l /sys/bus/auxiliary/devices/*/

    /sys/bus/auxiliary/devices/acpi.video_bus.0/:
    driver -> ../../../../bus/auxiliary/drivers/video
    firmware_node ->
    ../../../LNXSYSTM:00/LNXSYBUS:00/PNP0A08:00/device:00/LNXVIDEO:00
    input/
    power/
    subsystem -> ../../../../bus/auxiliary
    uevent

    /sys/bus/auxiliary/devices/acpi.video_bus.1/:
    driver -> ../../../bus/auxiliary/drivers/video
    firmware_node -> ../../LNXSYSTM:00/LNXSYBUS:00/PNP0A08:00/LNXVIDEO:01
    input/
    power/
    subsystem -> ../../../bus/auxiliary
    uevent

  *

    On kernel 7.0.x, the ACPI video devices are no longer bound under
    |/sys/bus/acpi/drivers/video/|. Instead, they appear under the
    *auxiliary bus* (|acpi.video_bus.*|). As a result:

  *

    |acpi_video0| disappears

  *

    only |intel_backlight| remains

  *

    brightness keys move the slider, but the actual backlight does not
    change

  *

    ACPI errors appear in dmesg (CreateField length zero,
    AE_AML_BUFFER_LIMIT, WMID failures)

This matches the regression I initially reported.

If you need additional logs (ACPI, i915, auxiliary bus traces), I can 
provide them.

Best regards, Harry

  *


Le 18/07/2026 à 11:05, harryblanc a écrit :
> Hello,
>
> I would like to report a regression affecting backlight control on my 
> HP Omen laptop when using kernel 7.0.0‑28‑generic (Ubuntu 24.04 base, 
> Linux Mint 22.3 XFCE).
>
> Summary of the issue:
> With kernel 7.0, the screen backlight is stuck at maximum brightness.
> Brightness keys show the OSD and the slider moves, but the actual 
> backlight does not change.
> With kernel 6.17, brightness works normally.
>
> Hardware:
> - HP Omen 15‑ax246nf
> - Intel i5‑7300HQ
> - Intel HD Graphics 630 (i915)
> - Nvidia GTX 1050 Mobile (disabled via Optimus)
> - Linux Mint 22.3 XFCE (Ubuntu 24.04 base)
>
> Backlight controller:
> /sys/class/backlight/ → intel_backlight
> acpi_video0 is no longer present under kernel 7.0.
>
> ACPI errors visible only with kernel 7.0:
> ACPI BIOS Error (bug): Attempt to CreateField of length zero
> ACPI Error: Aborting method \_SB.WMID.WQBK
> ACPI BIOS Error (bug): AE_AML_BUFFER_LIMIT
> ACPI Error: Aborting method \_SB.WMID.HWMC
> ACPI: video: [Firmware Bug]: ACPI(PEGP) defines _DOD but not _DOS
>
> Kernel version:
> Linux 7.0.0-28-generic #28~24.04.1-Ubuntu SMP PREEMPT_DYNAMIC Wed Jul 
> 1 15:50:57 UTC 2024 x86_64 GNU/Linux
>
> GPU information (inxi -Gxxx):
> Device-1: Intel HD Graphics 630 driver: i915
> Device-2: NVIDIA GTX 1050 Mobile driver: N/A (Optimus disabled)
> Display: X11 driver: modesetting, iris
> Monitor: eDP-1 1920x1080 LG Display
>
> Hardware summary (lshw -short):
> Intel(R) Core(TM) i5-7300HQ
> HD Graphics 630
> GeForce GTX 1050 Mobile (disabled)
> 16 GiB DDR4
> Samsung NVMe SSD + 1TB HDD
>
> Expected behavior:
> Backlight should be adjustable via intel_backlight or acpi_video0, as 
> in kernel 6.17.
>
> Observed behavior:
> Backlight remains at maximum brightness regardless of ACPI events or 
> brightness keys.
>
> Steps to reproduce:
> 1. Boot kernel 7.0.0‑28‑generic
> 2. Attempt to change brightness
> 3. Backlight does not change
>
> This appears to be an ACPI/i915 regression introduced in kernel 7.0.
>
> Thank you for your work.
>

Re: Regression: Backlight control broken on HP Omen 15‑ax246nf with kernel 7.0.0‑28‑generic

[Rafael J. Wysocki (Intel)]

On Tue, Aug 18, 2026 at 3:03 PM harryblanc <garett26@gmail.com> wrote:
>
> *Subject:* Re: Regression: Backlight control broken on HP Omen
> 15‑ax246nf with kernel 7.0.x
>
> Hello Rafael, Thorsten,
>
> Here are the outputs you requested, from both the working kernel
> (6.17.0‑42) and the failing one (7.0.0‑29).
>
> For context:
>
>   *
>
>     Secure Boot is *disabled* on this machine.
>
>   *
>
>     GPU setup is *Optimus with NVIDIA GTX 1050 “on‑demand” mode* (Intel
>     iGPU is primary).
>
>   *
>
>     These changes do not affect the issue: the backlight regression
>     remains exactly the same on kernel 7.0.29.
>
>   *
>
>     Kernel 6.17.0‑42 (backlight working normally)
>
>   *
>
>     $ ls -l /sys/bus/acpi/drivers/video/
>     total 0
>     --w------- 1 root root 4096 août 18 14:50 bind
>     lrwxrwxrwx 1 root root    0 août 18 14:50 LNXVIDEO:00 ->
>     ../../../../devices/LNXSYSTM:00/LNXSYBUS:00/PNP0A08:00/device:00/LNXVIDEO:00
>     lrwxrwxrwx 1 root root    0 août 18 14:50 LNXVIDEO:01 ->
>     ../../../../devices/LNXSYSTM:00/LNXSYBUS:00/PNP0A08:00/LNXVIDEO:01
>     lrwxrwxrwx 1 root root    0 août 18 14:43 module ->
>     ../../../../module/video
>     --w------- 1 root root 4096 août 18 14:50 uevent
>     --w------- 1 root root 4096 août 18 14:50 unbind
>
>   *
>
>     The ACPI video driver attaches normally, and backlight control works
>     as expected.
>
>   *
>
>     Kernel 7.0.0‑29 (backlight stuck at maximum)
>
>   *
>
>     $ ls -l /sys/bus/auxiliary/devices/*/
>
>     /sys/bus/auxiliary/devices/acpi.video_bus.0/:
>     driver -> ../../../../bus/auxiliary/drivers/video
>     firmware_node ->
>     ../../../LNXSYSTM:00/LNXSYBUS:00/PNP0A08:00/device:00/LNXVIDEO:00
>     input/
>     power/
>     subsystem -> ../../../../bus/auxiliary
>     uevent
>
>     /sys/bus/auxiliary/devices/acpi.video_bus.1/:
>     driver -> ../../../bus/auxiliary/drivers/video
>     firmware_node -> ../../LNXSYSTM:00/LNXSYBUS:00/PNP0A08:00/LNXVIDEO:01
>     input/
>     power/
>     subsystem -> ../../../bus/auxiliary
>     uevent
>
>   *
>
>     On kernel 7.0.x, the ACPI video devices are no longer bound under
>     |/sys/bus/acpi/drivers/video/|. Instead, they appear under the
>     *auxiliary bus* (|acpi.video_bus.*|). As a result:
>
>   *
>
>     |acpi_video0| disappears

Well, this is not supposed to happen regardless.

The (auxiliary) video driver binds to both devices and it uses the
same platform firmware interfaces as before.

It should register the backlight interface for one of them, but that
doesn't happen for some reason, probably related to the ACPI error
messages you see when running 7.0.x.

>
>   *
>
>     only |intel_backlight| remains
>
>   *
>
>     brightness keys move the slider, but the actual backlight does not
>     change
>
>   *
>
>     ACPI errors appear in dmesg (CreateField length zero,
>     AE_AML_BUFFER_LIMIT, WMID failures)
>
> This matches the regression I initially reported.
>
> If you need additional logs (ACPI, i915, auxiliary bus traces), I can
> provide them.

Please boot the working 6.17.0‑42 kernel again, do

$ ls -l /sys/class/backlight

while running it and send the output of that command.

Also please send boot logs (the output of dmesg after a fresh boot)
from both the working (6.17.0‑42) and failing (7.0.0‑29) kernels.

Thanks!