So, I have determined that my Pi is dead.
This is what I did in the few days leading up to it:
I used I2C about once every 2 seconds, meaning I updated the display about that often.
I read the temperature of the CPU via
cat /sys/class/thermal/thermal_zone0/temp about once every two seconds.
Had my Pi overclocked [Note: I had it overclocked for months before the death of my Pi] to 1GHz and cooled with an externally powered fan.
The last few days
While reading the temperature, my Pi would occasionally crash, various different ways.
Way one: complete crash, all GPIOs forced to the ON state. PWR and ACT led as well.
Way two: kernel panic, resulting in hundreds of panics per seconds comming in on the console via SSH. At this point
sudo reboot and
sudo halt would not work, and I would have to unplug it.
Way three: system files become unresponsive, like
After formatting a new card, things were going ok, but my overclock settings would not apply. Not until I used
force_turbo=1. Then, I tried to install I2C again. I found that the file used to load modules at boot was not being edited by the install via
raspi-config. So I edited it myself. My Pi worked again for about 16 hours, and then crashed for good. No kernel panic, nothing. When plugged in with either SD card, the PWR light comes on, and the ACT light flashes like normal, but no wifi, no serial console, nothing. I am currently formatting another card right now.
Now for my question.
Was there something that I specifically did to break my Pi?