Hi!

> > I did... (And wifi is not too reliable for me).
> >
> > But it seems there's better candidate: when I use "wrong" cable for
> > charging, green LED starts blinking very rapidly, and then system
> > seems to be sad/broken.
>
> I think this problem has been around for a bit longer. It seems to be a
> poor interaction between ke-recv-extra (the daemon that acts on USB
> plugs and loads usb gadgets) and the kernel cable detection.

Well, I believe problem must be in kernel, because UI is frozen. (OTOH
power button still does something). Would have to debug more.

> In short, I am not sure if they are related. I've seen the devices
> restart without being plugged into USB. I have a Droid 4 here directly
> connected to the Lab PSU, and it never got plugged into USB for it's
> uptime, and it still restarted yesterday night (unfortunately I was
> working and couldn't debug).
>
> While the restarts are annoying, I think I'd prefer to focus on some
> other parts first before tackling this issue. Still restructuring some
> parts of my house, but I should soon have the Droid 4 on permanent power
> and serial monitoring, so that should make this a lot easier.

Aha, it is not huge problem really. I just wanted to follow up on my
initial report.

> Btw - I guess there is no way to have the serial/system idle, but also
> get kernel messages on serial now, on 5.7? We currently detach the
> serial with 5.7.

Not 100% sure, but printing while system is idle will be tricky. I
guess we could wake up the serial when the message is ready, but
that's ... tricky stuff to implement.

Best regards,

                                    Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html