Random Screen dimming (while brightness slider stays at 100%) after A12 update

  • deleted for false information -

Edit: I falsely assumed I installed a new fairphone patches. Turns out, it was only CalyxOS patches. So actually nothing new on the Fairphone side of things.

2 Likes

Wait, are you sure that you have the July update? AFAIK, it hasnā€™t been publicly released yet. The topic you commented on is actually about the June update:

3 Likes

youā€™re right of course. I got an update, but it only was the calyxOS kernel update. I just assumed it included a fairphone update without actually checking. The vendor and firmware versions are still on May 5ths, so nothing has changed there. Well, this means there still is hope ā€¦ Thanks for pointing that out!

3 Likes

Now itā€™s gone really dark and the brightness isnā€™t returning to normal. The phone is quite cool and Iā€™m in the shade. Rebooting ainā€™t fixin it. Itā€™s been like this for 2 hours now.

Really quite annoying. Iā€™m willing to tolerate this and more because Iā€™ve internalized that there is sadly a price to pay for the right to repair oneā€™s property, and escape Googleā€™s fascist corporate surveillance a little in 2023. Had I not, Iā€™d be returning the phone at this point.

3 Likes

Fixed: I went to the barā€™s bathroom, put the phone in a plastic bag and ran cold water on it. It went back to fulll brightness in seconds.

If the problem is heat, it sure doesnā€™t take much.

Fix please Fairphone! Pretty pleaseā€¦ :slight_smile:

5 Likes

Hi everyone, Iā€™ve made some other tests and debug yesterday.
Can confirm that itā€™s related to the quiet-therm-usr sensor.
With an sh script like this:

for VARIABLE in $(seq 0 95)
do
	cat /sys/devices/virtual/thermal/thermal_zone$VARIABLE/type
  	cat /sys/devices/virtual/thermal/thermal_zone$VARIABLE/temp
  	cat /sys/devices/virtual/thermal/thermal_zone$VARIABLE/trip_point_0_temp
done

I have got all the name, actual temperature and threshold of the sensors.
The list is long, but quiet-therm-usr is:

quiet-therm-usr 39416 43000

So, 39,4 CĀ° actual temp and 43 CĀ° the threshold.
Then Iā€™ve tried to poll data while using the phone and the display dimmed when this sensor hit the threshold:

quiet-therm-usr
40046
43000

quiet-therm-usr
40178
43000

quiet-therm-usr
42570
43000

quiet-therm-usr
41695
43000

quiet-therm-usr
43105
43000

When the threshold is reached this sensor send an uevent and trigger a ā€œcooling_deviceā€ called:

panel0-backlight
255
168

When the display is at full brightness is like this:

panel0-backlight
255
0

The value 255 is the ā€œmax_stateā€ and 0 or 168 is the ā€œcur_stateā€.
To fix this there is a thermal-engine.conf in the system folder, that actually is empty.
Or probably there is some configuration in the ROM firmware or proprietary blob.

Anyway I have updated the issue in Gitlab and asked if it can be fixed/patched with the next beta release.
Hope they are not on vacation, because now with high weather temperatures itā€™s dimming even indoor, when the air conditioning is off.

26 Likes

Thank you!

Thank your for this analysis, thatā€™s awesome! If they donā€™t fix it by the next update, it will be completely inexcusable.

5 Likes

Yes, hope that it can help them to resolve the issue quickly.
Anyway itā€™s only an analysis of a user, probably thereā€™s is some other related issue under the hood.
Because from what I have read it started with Android 12.
So, I donā€™t know if something changed on the OS side.
E.g. : Android Dynamic Performance Framework

But for now, like Bon Jovi said, I will ā€œkeep the faithā€. :slight_smile:

1 Like

Same issue. So annoying in bright sunlight. Please fix asap

@anon64862762 , would be great to have an update. again many days have passed with your idealistic customers tolerating issues which are actually pretty unacceptable.

Thanks.

5 Likes

Hi You are not communicating with Fairphone really, so ā€˜Please fixā€™ goes over the heads of support, but Iā€™m sure they are on the case :slight_smile:

1 Like

I have a question for those who know - first time for me with an alternative ROM, so I donā€™t really know how this goes.

My Fairphone runs CalyxOS. Since Fairphone 4s running stock Android or /e/ are also affected, I presume the issue is either with one of the drivers or with AOSP.

Assuming itā€™s a driver: will Fairphone push the fix as some APK to install? Or is it done through a system OTA update in all cases? If the latter, do I have to wait for the CalyxOS folks to pick up the fix and push an update of their own my way?

1 Like

As far as I know yes and yes :wink:

4 Likes

Your last messages about debugging motivated me to do some stuff as well. I found this requesting nits when no mapping exists line interesting as well and when I Ducked/Googled it I got back to this topic :nerd_face:

I see the error pop up quite often. When the screen times out and when the screen is turned on again. But also indeed when the brightness error is triggered when e.g. calling. Which I guess relates more to the issue below. But might also be completely unrelated since the error line seems to happen all the time, but not one of the those bugs themselves occur.

So then I tried my old Google Pixel 3. Ran the same command and filtered only for the Error log level and above that with adb logcat -s *:E, as I also did with my Fairphone 4. And wow, the difference is day and night. On the Pixel there are hardly ANY errors. On the Fairphone 4 you get to see so many of the error tags itā€™s almost as if itā€™s a demo of logcat or as if I enabled a low level debug mode. But even worse, I excluded info level logs and suchā€¦ I knew Fairphone was bad, but jezzzzz.

Anyway, will dig into this more some other time. Maybe I have to setup an Elastic Search cluster to make sense of this amount of errors :crazy_face:

9 Likes

Okay thanks!

[quote=ā€œamoun, post:297, topic:93195ā€]ā€¦
but Iā€™m sure they are on the case :slight_smile:
[/quote]

Since when youā€™re sure about this? As now also for this bug several month have passed without FP finding a solution, I am pretty sure that FP has other priorities than customer satisfaction.

@facetto

Did you see the post from Fairphone?

1 Like

Managed on cloud or on prem? :smile:

Anyway Iā€™ve also got this requesting nits when no mapping exists a lot, not only before the screen dims, mostly after a bunch of seconds in loop. But didnā€™t know is related with the proximity sensor.

Actually I donā€™t have other Android phones to check, but from the log you have got the Pixel 3 is optimized, for sure.

1 Like

About this I can say that today a Fairphone employee replied to the Gitlab comments.
They have got the logs and info that I have updated yesterday and passed to the dev team.
Letā€™s see what happens next.

I am a user like you, so I am not taking sides or defending anyone.
I am optimistic, though, because I think the solution is close.

10 Likes