Merge pull request #2494 from JacoboGuijar/patch-1

Update ardupilot_custom_mode.h misspelled name

core: fix param get regression (#2481)

It turns out single params are sent with index -1, so caught by the test
for the _HASH_CHECK param.

Merge pull request #2471 from mavlink/pr-v2-https-workaround

[v2 only] core: use http instead of https

Merge pull request #2458 from mavlink/pr-v2-param-get-caching

[BACKPORT v2] core: fix get_all_params caching

camera_server: set zoom flag (#2433)

This was missing before.

camera_server: prevent double ack+message (#2430)

It turns out we were sending the ack and message for storage information
as well as capture status twice, once directly in the request handler
callback, and once the MAVSDK user would call the respond function.

We should only call it in the respond function, not in the callback.

core: prevent invalid duplicate acks (#2424)

It turns out we respond with two acks to a MAV_REQUEST_MESSAGE when we
should only respond once with the successful one.

Alternatively, we could first grab all the results and then do some
magic to determine which one is the best ack but that's a bit too magic
and complex, so I'm not doing that yet.

plugins: fix install path (#2419)

core: prevent lockup on connection destruction (#2412)

When we destroy Mavsdk and clear the list of connections, we likely
end up in a deadlock.

What happens is that:
1. A connection wants to forward a message and is trying to acquire the
   connection mutex.
2. At the same time, the connection is being destroyed, so we are
   waiting for the connection receive thread to be joinable. While the
   connections are being destroyed, we have the connection mutex which
   is blocking 1.

The proposed solution is to:
1. Make it less likely by acquiring the connection mutex properly before
   checking _connections.size() and not for the individual connections.
2. Check the _should_exit flag before trying to acquire the mutex. I
   believe by the time the connections are being cleared, this flag is
   set, and hence the deadlock should not happen, fingers crossed.

Merge pull request #2407 from mavlink/pr-v2.12-system-debugging

[v2.12 BACKPORT] Add some system debugging