Phone seemingly bricked after being plugged in as it was dying

sargates

I have a Pixel 8 with GOS. It's worked very well for over a year but last night I saw my phone was at 1 percent and as I was reaching for my charger, it showed the "Shutting Down" dialogue and I ended up plugging it in as it was showing that. It ended up dying anyway; I didn't save it before it shut off.
But now the phone does not boot.

Last night I had been able to get it to start booting (into fastboot) by holding the power and volume down buttons for ₂ minutes. From this menu I could continue the boot process and pause it at the yellow-text part of the boot process (I don't know what it's called) and let the battery charge. I had used this trick before to combat the common issue of the phone dying immediately after booting because the battery is not charged enough.
The battery just being empty is not the issue (unless the battery is just fried). I have been able to see the charge reading of the battery (64% and 65%) for a few seconds (read below).

After letting the battery charge, the OS would boot, seemingly as normal, but after only a few seconds the device would completely shut off. The display goes black, nothing was shown to indicate shutting down. In the few seconds before the shutting down, I can see the battery reading, but the OS (just the lockscreen in this case) wouldn't respond at all. No attempts to unlock the device or check notifications would cause any change on the display.

All that was last night. I have left it plugged in overnight (I know it can be bad) and have not being able to get the phone to boot into fastboot since. I have no idea what the hell is going on. I cannot even tell if the device is on or not. I just want to figure out what is wrong. I am not familiar enough with probing USB devices in Linux to know whether plugging it into my laptop is even getting a response.

I do not know what the version of the OS I had, I had not checked in a while. I did not subscribe to any beta branches, but I think I was subscribed to security patches. I had been postponing a restart after getting the notification saying "Restart to install updates", maybe that would contribute.

Any guidance would be appreciated.

sargates

Found one on here as well, but he mentions that he was able to unlock the phone and such when I was not able to.
https://discuss.grapheneos.org/d/21814-help-needed-ran-out-of-battery-now-turns-off-after-startup

sargates

Update: Device boots with fastboot as described.
I am unable to edit the original post to amend that that I have not been "able to get the phone to boot into fastboot since". There's also a typo in the original.

sargates

Update: I was able to boot into the OS one more time. This time I was able to interact with the display and get movement. so the attached GOS forum post is the closest thing to the behavior I'm experiencing. Again, can't edit the original post to reflect that.

sargates

Booting with fastboot and selecting "Recovery Mode" (not "Rescue Mode") allows me to select "Apply update from ADB" which says in red "An update may already be in progress. If you proceed, the existing OS may not longer boot, and completing an update via ADB will be required." So it's possible that the cause is a fudged update.

However, I did not have USB debugging enabled before this issue occurred. Can anyone tell me if I will be able to apply the update via ADB if I did not enable this beforehand? GPT has mentioned that the ability to wipe cache from recovery mode was removed.
Am I SOL here?

userofgos

sargates However, I did not have USB debugging enabled before this issue occurred. Can anyone tell me if I will be able to apply the update via ADB if I did not enable this beforehand?

You DO NOT need USB debugging enabled if you are Side-loading the lasted update which should potentially resolve the issue. See https://grapheneos.org/usage#updates-sideloading

sargates GPT has mentioned that the ability to wipe cache from recovery mode was removed.

Please don't use ChatGPT or LLM's for issues like this. They are highly inaccurate.

sargates

userofgos It doesn't seem inaccurate because there was no option to wipe cache in recovery mode. I am aware that LLMs can be inaccurate. I asked a single pointed question about one specific thing to it.

sargates

I solved this. Somebody please mark this as solved since I don't even know if I have permissions to do it myself.

Selecting "Apply update with ADB" will start a unique ADB server on the device; enabling USB debugging is not required, which I only found out after I was already messing with ADB. If you try to run adb devices on a computer while the device is plugged in to that computer, nothing will show up unless USB debugging is enabled or without selecting "Apply update with ADB" in the recovery mode menu.

As @userofgos mentioned, I was able to solve this by sideloading an update. It took a few tries--the first time trying resulted in error code 1 (kError, see error_code.h). After the first failed update, rebooting and trying again worked and the update was successfully installed.

I believe what happened here was some weird ass race condition. The phone was in the process of shutting down and was also waiting for the next reboot to install an OS update. (Referring to whatever it uses to signal the device to install the update as "the flag") I assume that the device noticed the flag was in the process of being set while it was plugged in, and interpreted this as a post-reboot boot instead of a shutdown, thus starting the installation of an update but committing to the shutdown anyways and corrupting the OS. I think it would be safe to assume that this is upstream to AOSP and not a GOS issue.

Hopefully this will help someone in the future.