Rover: add NAV_GUIDED_ENABLED support

[rmackay9]

This PR adds support to Rover for the NAV_GUIDED_ENABLE and DO_GUIDED_LIMITS commands.

Supporting these commands allows "Guided within Auto" which means an external navigation source (like ROS) can temporarily take control of the vehicle during a mission.

This feature is very similar to what is already in Copter but there are some changes:

• AP_Mission's NAV_GUIDED_ENABLE command handler stores a user provided lat, lon and alt.
• If a NAV_GUIDED_ENABLE command has a position then we send this to the external navigation controller within a SET_POSITION_TARGET_GLOBAL_INT message at 1hz. We also use this position to "verify" that the command has been completed (i.e. when the vehicle is within WP_RADIUS of the position we move to the next command)

Similar to copter, the DO_GUIDED_LIMITS command allows the user to specify a timeout and a safe radius (i.e. a small fence) to help ensure the external controller doesn't do anything crazy

This has been tested in SITL and on a real vehicle. Here's a video of the test using ROS as the external path planner.

There are some related changes to other system that I will take care of separately from this PR:

• change the MAVLink spec for the NAV_GUIDED_ENABLE command to allow lat,lon,alt to be stored. As far as I know we are the only users of this command - I originally added it as part of the red-balloon-popper project
• get this mavros PR merged to allow ROS to accept the position target sent from ArduPilot
• Mission Planner should be changed to make the GUIDED_ENABLE and DO_GUIDED_LIMITS commands available to the user (issue raised here)

Ah, there are also two small changes to Copter that should be non-functional changes to remove a duplicate call and fix some formatting and comments.

[rmackay9]

@OXINARF, could you have a peek at this when you have time? It's not desperately urgent of course so whenever is fine.. I've tested it and I think it's OK but I'd like a 2nd opinion before merging 'cuz I pretty much always make a mistake :-)

[magicrub]

confirmed working on our hardware

[rmackay9]

I think I've addressed pretty much all the items raised during the review so I'll merge after it passes travis. thanks @peterbarker!

[magicrub]

doh! @jmachuca77 found something wrong...

[rmackay9]

@magicrub, that's OK, now it's a support issue is all..

[jmachuca77]

@magicrub yeah its not a real issue, that's why I did not comment. Works once you know how it should work haha.