Raimo Niskanen

I think some tests should be added, that we simply keep this ancient behaviour behaviour, and document it.

A few callbacks that uses `throw/1` in the test suites for `gen_server`, `gen_event` (and `gen_fsm`) would be sufficient, just to exercise the code path and to etch the feature into...

@bucko909: From what I remember, `catch Expression` has always caught all 3 classes of exceptions: ``` erlang catch throw(4711). % -> 4711. catch exit(4711). % -> {'EXIT', 4711}. catch error(4711)....

> By the way we have tested socket NIF version ktls dist, it adds 5% more global CPU usage to our prod node compare to the inet_drv version Is that...

We have managed to get it to work. Simple and maybe stupid beginner's problem - it may be so obvious to the Linux people that operate in this field that...

> Maybe we can remove the ServerTx/ClientTx check? I think that the check is important to verify that the cipher suite we set is accepted by the kernel. Otherwise it...

The only improvement of your `socket` based `inet_tls_dist` module I can see immediately is that `receiver_loop/6` can avoid flattening all data into a binary when parsing the length. Now, if...

When I run the `ssl_dist_SUITE:ktls_basic` test in CommonTest I get this Notice in the console: ``` =NOTICE REPORT==== 11-Aug-2022::10:42:00.556270 === TLS server: In state start at tls_record.erl:564 generated SERVER ALERT:...

> By the way we have tested socket NIF version ktls dist, it adds 5% more global CPU usage to our prod node compare to the inet_drv version What CPU...

@zzydxm: I am working on an update with my suggested changes above - it became many alterations to set_ktls. Will try to push a commit tomorrow (CEST)