Just to clarify, this is writing to /sys/power/wake_lock and /sys/power/wake_unlock respectively as root, if I understand correctly. Basically you echo some made up word and then you get no sleep
Iāve started experiencing this problem only recently, so it still needs more tests and observations to pinpoint some pattern and exact conditions. Please allow some more time - I will further test it and come back with as detailed feedback as possible.
One small correction for now is that I am not sure if it is the time (itself) since reboot what causes it or something else that triggers this problem. What I meant was that freshly after reboot zooming and switching between cameras always works perfectly fine and smooth, whereas after some time it starts freezing. Therefore, obviously, the longer the time that passed since reboot, the bigger the chance for something that triggers this problem to happen, so I didnāt mean just the time per se. As I said, it takes further observations so please allow a bit more time to further test it and Iāll certainly provide more information.
You might test if using other CPU governors and/or changing the min and max CPU frequencies of the three core groups makes any difference when it comes to decreasing heat and power consumption, which is actually horrible under heavy usage like web browsing (it often reaches 1.5 - 2.0 Amps, which is really a lot). Check /sys/bus/cpu/devices/cpuX/cpufreq/ for details. All you need to do is to write the desired values to the appropriate files and chances take place instantenously. There is no risk involved with playing with those values because they always get reset to defaults upon reboot. Iāve done some initial tests with e.g. conservative and schedutil governors as well as lowering the min frequency of 0-3 cores to 307 MHz and decreasing max frequencies of all core groups by even 30-40%, but I didnāt have enough time yet to get reliable results.
I spent a little time playing with the camera yesterday, as my uptime had crept up past 5d.
I couldnāt persuade it to show any issues after a few minutes of zooming in/out, shooting pictures/video, front/rear camera, changing filters, colour temp etc.
(5iv XQ-CQ54, pics stored in default location).
I did think at one point it was starting to act up, but then realised Iād accidentally enabled the 3s selftimer
For the pictures on SD card one fix could be to clear the cache from utilities, and maybe also deactivate images in dropbox as mentioned here - that worked for me for the last week at least.
About the camera: when it locks, it sometimes unfreezes after.. minutes!. It looks as if an android camera process is eating up all the memory. So it might be triggered just by more camera usage. It is hard to debug, I am looking at some logs after unfreeze happened but hard to spot the cause.
Just a quick information that I couldnāt reproduce the camera freeze issue anymore. Nearly 13 days of uptime and itās been working flawlessly. So, apparently, the problem must be triggered by something that I havenāt used recently - maybe Android camera apps in AAS⦠Iāll keep monitoring it.
Would be great to stick to bug reports without AAS. AAS is not tested nor supported and please at least highlight that you used it when issue happened. Its clear that using AAS camera can have side effects on SFOS.
I wasnāt reporting a bug in AAS. Iāve just re-checked what I wrote and it appears clear to me, so I donāt know how could I possibly rephrase it.
Thatās exactly what I intend to do, but it requires further testing. For now, I only know that I did not use it over the past couple of weeks, so maybe thatās why I could not reproduce the issue. So now I will start using it and see if it triggers the problem on the SFOS side.
I was suggesting to add (NOTE: I WAS USING AAS) in the bug reports for the port, as in the case of camera issues reported recently. Iff it is due to AAS, then we are getting maybe to some difference in camera bug as 1IV vs 5IV - those use slightly LOS different bases. But then it could be also due to differences in numbers of users between devices or vocal users
Anyway, thanks for getting back and testing it further.
In this specific case, I intentionally avoided using it recently, and the issue did not occur on the SFOS side, which may or may not have anything to do with it. A few moments ago I launched AAS and played with several camera apps in it (OpenCamera, stock AOSP camera app, etc.) but camera on SFOS side continues to work flawlessly. So itās quite hard to pinpoint and will require more testing, which I intend to do and report as soon as I discover anything useful.
18 days uptime and the camera still works flawlessly, including smooth zoom (accross ranges of the two lenses) without any issues. And that regardless of using various Android camera apps in AAS (to test if they have anything to do with it). So for now I have to assume that it was just a temporary glitch and itās gone. One last test will be to reboot the phone now and test it again over the next few days on a freshly restarted device.
I was just about to try this port, and I got an Xperia 5 iv off ebay USA, but I think Iāve got the same issue as Fellfrosh had late last year: it seems to be a model A204SO (Softbank Japan) thatās been reflashed with the global version of the firmware XQ-CQ54 but sill of course not bootloader unlockable. No vendor branding on the case to give it away either. (And I had to boot up a Windows 10 laptop to run Xperia Companion and I feel dirty.)
Oh well, supposedly itās returnable without hassle but Iām out the cost of postage and time.
Guess Iāll be back sometime later to try this again . . .
Sorry that happened man. A lot of sellers are doing that unfortunately. Ebay user meonov phone | eBay Stores sells legit Xperia 5 IV that are compatible. https://www.ebay.com/itm/387581193274 This isnāt me and isnāt sponsored. But I have bought 4 of them from this guy without issues.