New to Log40M love it but it freezes often - Maybe a solution?

[KN6KIG]

I registered with HDR before trying LogM40 V2 which was hinted by a user on the TS890 forum.
Too bad as I am sticking with LogM40 which is way better than HDR.
I plan to donate to LogM40 as I am in awe about what you guys are doing and you need to be supported.

My setup:
I am running windows 10 on a parallel VM on a Mac with Catalina with 8 cores and 16GB of ram
I have allocated 6 cores to Windows and 10GB
Station is a Kenwood TS890 + SDRPlay Uno with RSPDx
VSPE - Virtual Serial Emulator to create a Splitter from the Radio Com Port into OmniRig 1.19
SDR Uno and Log40M tap into Omnirig which taps into the Virtual port created above with VSPE.
WSTJX tap also into the same virtual port - Everything works.
Proper frequencies on radio, SDR Uno, WSTJX and Log40M.

I have setup Log40M with MariaDB MySQL and it works.
All interfaces to external services works.
WSTJX and JTAlert works allow lookup before transmitting and update the QSO properly. Great!

Only issue is that from time to time Log40M freezes and L4ONG.exe process need to be killed and restarted.
This happens a lot. Like every 10mn or so.
While LogM40 is frozen, JTAlert will still successfully log a new contact into the database. So not too much damage on the logging front.
When Only Log40M works, it does not freeze.

Let me know if you have any idea about what could cause this freezing issue!

Many thanks

Eric

[G4POP]

Eric
If you search this forum you will find this is an issue with just a few users and sometimes caused by an outside program.

We are however activly tracking this down and you can be assured we will find the cause

[KN6KIG]

Thank you Terry

[IW3HMH]

i can't reproduce the issue so i'm adding some debug instructions in latest alpha and providing them to some users that, instead, have frequent locks.
Unfortunately, without the issue occurring on my system, i cannot debug it.
Deadlocks are internal program races on the same resource, when 2 different threads works on the same resource both waiting for the other one.
They're extremely difficult to track down without a dump or without a debugger.
I'm trying to track them by adding in-out logs trying to find when a function enters and doesn't exit anymore from his execution