So, should I submit a bug report? I'm fairly sure this is a LibreELEC bug, since others have reported solving this issue by moving to CoreELEC.
The link above already leads to a bug report. I would just rename the thread title by using the [BUG] prefix. If you're lucky, they will fix it this week with LE 9.0.2.
What, my log? I've generated and uploaded it yesterday.
Yes, you have a mix of old and new entries in there. This is impossible, and probably another bug:
2018-06-22 08:11:52.981 T:4086567712 NOTICE: Kodi compiled 2019-02-21 by GCC 8.2.0 for Linux ARM 32-bit version 3.14.29 (200221)
PS: It's a pretty old log from 2018.
This line of the log file could be the cause:
2018-06-22 08:11:53.706 T:4064064368 WARNING: Pulseaudio module module-allow-passthrough not loaded - opening PT devices might failLooks like an LE bug.
The DEV's can have mine to work on as long as it's sent back.
If you do long term tests, and provide error log files, it would help dev's like escalade a lot.
How do you start autofs? What's the content of the config files? If it doesn't run in the background (demon mode), we can't expect a re-mount.
Also looking into using AutoFS, as it seems people claim it can handle this sort of situation better.
Does it run as a demon? If yes, you should see it with the "ps -A" command. In that case there is a chance to configure it for re-mount.
You could write a python script, started from autostart.sh, which automatically mounts / re-mounts your CIFS.
Sounds like an add-on error, which does a wrong key re-direction at first press. Try what happens when you uninstall the add-on. You actually don't need an add-on for key mapping, because you can manually create custom keymap files.
2019-04-27 12:07:16.007 T:1937639824 NOTICE: load skin from: /storage/.kodi/addons/skin.aeon.nox.silvo (version: 7.0.1)
2019-04-27 12:07:17.759 T:1937639824 WARNING: Skin has invalid include: skinshortcuts-template-widget1
2019-04-27 12:07:17.759 T:1937639824 WARNING: Skin has invalid include: skinshortcuts-template-widget2Try the default skin.
I have a BRIX which uses Intel the addon HDhomerun freezes
I know. 
You already had issues with the add-on on RPi, so I don't think it will be a fix for you. Here is the quote about Intel API:
VAAPI on Intel has gained some corrections for interlaced content that toggled interlaced flags during playback, and therefore caused stutter by reconfiguring the decoder.
Are there issues with the stability of 9.0.1?
Just to reconsider upgrading from 8.2.5 or wait until 9.0.2 or next iteration considered enough stable.
Depends on your hardware. If you have an RPi like me, you can safely update to 9.0.1. If you have an Intel, at least one issue (video stuttering) will be fixed with 9.0.2, so you better wait for that version.
I have no idea what you're trying to say there. What;s possible?
It is possible that a BIOS does not allow to install an OS from an external (USB) source device to an external (USB) target device.
That was my thoughts KODI it self has always been solid its the addons which makes it unstable. Been with KODI since the beginning of XBMC had problems with Raspberry but the size of library always showed lack of memory is why I went with a NUC
Yes, it's sad, you bought new hardware to fix it. 
Maybe your library so large that HD HomeRun developers don't test this scenario very well. Hopefully they'll give you hints, how to solve it. Without a debug output of the app, we really can't do anything here.
Are you sure? Because I've done exactly that on plenty of pc's over the years (openelec and libbreelec) and not many of them are what I'd call remotely modern.
No, I only did external -> internal (IDE) installations on older BIOS PC's. But BIOS'es can be different, so it's possible. Let's wait for his answer.
The output is fine, no need to expand any memory. I think the app of your HD HomeRun could be bad. You should ask those guys (see forum link above).
Using Putty cat /etc/fstab shows empty
OK. It could help to see the output of "free -m" and "df -hT" shortly before it freezes (like after 50 min. or so). Maybe we can find out where most memory is needed.
