How to analyze the boot up failures


#21

As best as I understand things, adopt a v4.14 kernel and reboot several times (rate of nonoccurence may be fairly low).


#22

Sometimes takes a few hundred reboots: Boot failure / crash during the boot?


#23

Hi @Loic, Why would you attribute the solution to “fuse a PBL config or modify the SBL code” here? Could you please share your thoughts?


#24

Note that Loic’s comment was about how to stop the system entering a debug mode after it had failed (it was not about stopping it from failing). In that case of debug mode no other code has been executed yet!


#25

Correct! my point was not about root cause resolution but how to prevent the bootloader to switch to ramdump mode after a crash.

I’ve tried to enable QCOM krait watchdog on Linux, and this also causes a reboot in ramdump mode when triggered. So I do not see any easy way to cleanly reboot after a crash/watchdog without modifying the second stage bootloader.