Re: [REGRESSION] Brightness at max level after waking up from sleep on AMD Laptop

Re: [REGRESSION] Brightness at max level after waking up from sleep on AMD Laptop

[Linux regression tracking (Thorsten Leemhuis)]

[+amd-glx, +lkml, +dri-devel]

On 27.07.24 18:52, serg.partizan@gmail.com wrote:
> 
> After updating from 6.8.9 to 6.9.1 I noticed a bug on my HP Envy x360
> with AMD Ryzen 5 4500U.
>
> [...]
> After waking up from sleep brightness is set to max level, ignoring
> previous value.
> 
> With the help of Arch Linux team, we was able to track bad commit to
> this:
> https://gitlab.freedesktop.org/agd5f/linux/-/commit/63d0b87213a0ba241b3fcfba3fe7b0aed0cd1cc5

Hamza Mahfooz, in case you missed it, that is a patch of yours:
63d0b87213a0ba ("drm/amd/display: add panel_power_savings sysfs entry to
eDP connectors") [v6.9-rc1].

> I have tested this on latest mainline kernel:
> 
> Results after waking up:
> 
>> cat /sys/class/backlight/amdgpu_bl1/{brightness,actual_brightness}
> 12
> 252
> 
> Then, on exact this commit (63d0b87213a0ba241b3fcfba3fe7b0aed0cd1cc5),
> result is the same.
> 
> Then, on commit just before this one (aeaf3e6cf842):
> 
>> cat /sys/class/backlight/amdgpu_bl1/{brightness,actual_brightness}
> 12
> 12
> 
> I hope I included all relevant information, more info can be found here:
> 
> https://gitlab.archlinux.org/archlinux/packaging/packages/linux/-/issues/52

Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker' hat)
--
Everything you wanna know about Linux kernel regression tracking:
https://linux-regtracking.leemhuis.info/about/#tldr
If I did something stupid, please tell me, as explained on that page.

P.S.:

#regzbot introduced: 63d0b87213a0ba241

Re: [REGRESSION] Brightness at max level after waking up from sleep on AMD Laptop

[Hamza Mahfooz]

On 7/31/24 02:38, Linux regression tracking (Thorsten Leemhuis) wrote:
> [+amd-glx, +lkml, +dri-devel]
> 
> On 27.07.24 18:52, serg.partizan@gmail.com wrote:
>>
>> After updating from 6.8.9 to 6.9.1 I noticed a bug on my HP Envy x360
>> with AMD Ryzen 5 4500U.
>>
>> [...]
>> After waking up from sleep brightness is set to max level, ignoring
>> previous value.
>>
>> With the help of Arch Linux team, we was able to track bad commit to
>> this:
>> https://gitlab.freedesktop.org/agd5f/linux/-/commit/63d0b87213a0ba241b3fcfba3fe7b0aed0cd1cc5
> 
> Hamza Mahfooz, in case you missed it, that is a patch of yours:
> 63d0b87213a0ba ("drm/amd/display: add panel_power_savings sysfs entry to
> eDP connectors") [v6.9-rc1].

Um, for the time being you should be able to set `amdgpu.abmlevel=0` in
the kernel's cmdline to avoid the issue.

> 
>> I have tested this on latest mainline kernel:
>>
>> Results after waking up:
>>
>>> cat /sys/class/backlight/amdgpu_bl1/{brightness,actual_brightness}
>> 12
>> 252
>>
>> Then, on exact this commit (63d0b87213a0ba241b3fcfba3fe7b0aed0cd1cc5),
>> result is the same.
>>
>> Then, on commit just before this one (aeaf3e6cf842):
>>
>>> cat /sys/class/backlight/amdgpu_bl1/{brightness,actual_brightness}
>> 12
>> 12
>>
>> I hope I included all relevant information, more info can be found here:
>>
>> https://gitlab.archlinux.org/archlinux/packaging/packages/linux/-/issues/52
> 
> Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker' hat)
> --
> Everything you wanna know about Linux kernel regression tracking:
> https://linux-regtracking.leemhuis.info/about/#tldr
> If I did something stupid, please tell me, as explained on that page.
> 
> P.S.:
> 
> #regzbot introduced: 63d0b87213a0ba241
-- 
Hamza

Re: [REGRESSION] Brightness at max level after waking up from sleep on AMD Laptop

[Filip Hejsek]

Hello,

On Sat, 2024-07-27 at 19:52 +0300, serg.partizan@gmail.com wrote:
> Hello,
> 
> After updating from 6.8.9 to 6.9.1 I noticed a bug on my HP Envy x360 
> with AMD Ryzen 5 4500U.
> 
> #regzbot introduced: v6.8.9..v6.9.1
> 
> After waking up from sleep brightness is set to max level, ignoring 
> previous value.

This issue is tracked here:
#regzbot link: https://gitlab.freedesktop.org/drm/amd/-/issues/3474


> With the help of Arch Linux team, we was able to track bad commit to 
> this: 
> https://gitlab.freedesktop.org/agd5f/linux/-/commit/63d0b87213a0ba241b3fcfba3fe7b0aed0cd1cc5
> 
> [...]

This commit does not introduce the bug, it only makes it possible to
set the ABM level without requiring support from the compositor. It's
very likely that the bug has existed for a long time, it just hasn't
surfaced because no compositor enabled the ABM power savings featue.
Now that power-profiles-daemon can enable it through sysfs, the bug has
been exposed.

So I think this actually might not be a regression, just a new feature
that exposed an old bug.

--
Filip