Segfault in MM 1.12.4 in port timeout detection logic

[Aleksander Morgado]

Hey

>
> After around 70 passes overnight, I didn't stumble upon a segfault, so for me, the patch seems solid. +1.
> The test was deliberately exploiting a bug in modem causing it's primary AT port to freeze and not respond to any commands, then waiting for MM to dispose of the modem object, to trigger modem reboot.
> Big thanks once again for very quick response!
>

Thanks for testing. Dan already merged the changes to master after
review, and I did the manual backport to 1.12 already now.

-- 
Aleksander
https://aleksander.es