Hi all,
Recently I made the brave attempt to upgrade my PPP bought before 2023, with RK2AW.
The point with installing RK2AW is to get decent behavior in case of battery depletion - before RK2AW, if you deplete the battery, the phone would go into a battery-depleting perpetual loop and the only way to get the phone working again was to remove the USB and battery, push the Maskrom/RE PIN, and then boot the phone, and leave it like that for 10-20 hours.
For RK2AW distribution, I took this archive https://xff.cz/kernels/bootloaders-2024.04/ppp.tar.gz , as suggested by the Pine wiki https://pine64.org/documentation/PinePho...ion_image/ and also by Megi in his blog https://xnux.eu/log/098.html .
The installation steps were really only:
And the outcome of this was that RK2AW with the "RKBIN" closed-source memory initialization blob + a recent U-Boot had been installed.
After that, I powered off the PPP, and disconnected USB and battery for 30 seconds. Not sure if all those steps are really needed but, after that definitely RK2AW worked as expected.
The issue came when I booted up PostMarketOS Edge with this new RK2AW-flashed PinePhone Pro:
The phone kept behaving as if I had tapped the volume up or volume down button, while I had not.
And meanwhile, the phone when in PostMarketOS stopped responding to when I actually pushed the volume button up or down.
I'm not sure if the issue is that I stopped using Tow-Boot and use U-Boot now, or if the issue lies in RK2AW.
Googling and asking Perplexity AI, I see that other people have had the same issue.
The quick fix is to disable the OS' event logics for when you push the volume button in the first place.
But, I cannot find any concrete suggestions for this fix.
I have also seen PostMarketOS insert gibberish in both directions into the UART data. It is so bad that the UART at least periodically becomes unusuable, because new gibberish is pushed every second, and often the gibberish is interpreted as special key combinations, that make you get logged out.
I saw some suggestion that at the root of this is a driver issue, so it sounds like it's fixed by including a particular Linux kernel patch into PMOS.
I tried with the latest PMOS Edge dated 20250830 and the problem is there too.
Did anyone have this problem too, can you share exactly how you fixed it?
Do you know any Linux distribution for PPP which has fixed the issue?
Any insights appreciated.
Recently I made the brave attempt to upgrade my PPP bought before 2023, with RK2AW.
The point with installing RK2AW is to get decent behavior in case of battery depletion - before RK2AW, if you deplete the battery, the phone would go into a battery-depleting perpetual loop and the only way to get the phone working again was to remove the USB and battery, push the Maskrom/RE PIN, and then boot the phone, and leave it like that for 10-20 hours.
For RK2AW distribution, I took this archive https://xff.cz/kernels/bootloaders-2024.04/ppp.tar.gz , as suggested by the Pine wiki https://pine64.org/documentation/PinePho...ion_image/ and also by Megi in his blog https://xnux.eu/log/098.html .
The installation steps were really only:
Code:
sudo bash
cd /root
wget https://xff.cz/kernels/bootloaders-2024.04/ppp.tar.gz
sha256sum ppp.tar.gz
tar xvfz ppp.tar.gz
cd ppp
./spinor-flash-initial-setup.shAnd the outcome of this was that RK2AW with the "RKBIN" closed-source memory initialization blob + a recent U-Boot had been installed.
After that, I powered off the PPP, and disconnected USB and battery for 30 seconds. Not sure if all those steps are really needed but, after that definitely RK2AW worked as expected.
The issue came when I booted up PostMarketOS Edge with this new RK2AW-flashed PinePhone Pro:
The phone kept behaving as if I had tapped the volume up or volume down button, while I had not.
And meanwhile, the phone when in PostMarketOS stopped responding to when I actually pushed the volume button up or down.
I'm not sure if the issue is that I stopped using Tow-Boot and use U-Boot now, or if the issue lies in RK2AW.
Googling and asking Perplexity AI, I see that other people have had the same issue.
The quick fix is to disable the OS' event logics for when you push the volume button in the first place.
But, I cannot find any concrete suggestions for this fix.
I have also seen PostMarketOS insert gibberish in both directions into the UART data. It is so bad that the UART at least periodically becomes unusuable, because new gibberish is pushed every second, and often the gibberish is interpreted as special key combinations, that make you get logged out.
I saw some suggestion that at the root of this is a driver issue, so it sounds like it's fixed by including a particular Linux kernel patch into PMOS.
I tried with the latest PMOS Edge dated 20250830 and the problem is there too.
Did anyone have this problem too, can you share exactly how you fixed it?
Do you know any Linux distribution for PPP which has fixed the issue?
Any insights appreciated.
