Fix raise by MIN_Z_HEIGHT_FOR_HOMING (first attempt)

[thinkyhead]

When doing the initial raise for homing (which occurs before homing X and Y) we don't need to include the probe offset. This lift is just always done over the current Z position. The Z axis will be lifted more, if needed later, to deploy the probe before homing Z.

Reference: #4208

• Sanity check the probe raise values. There should always be raises for probing.
• Also, it looks like X and Y re-home after quick-homing. This PR includes a patch for that.

[AnHardt]

Why does this work? do_blocking_move_to_z() is only defined if HAS_BED_PROBE is defined.

[AnHardt]

If MIN_Z_HEIGHT_FOR_HOMING is defined you get.

Marlin_main.cpp:2898: error: 'do_blocking_move_to_z' was not declared in this scope

       do_blocking_move_to_z(destination[Z_AXIS]);

Because do_blocking_move_to_z() is only defined if HAS_BED_PROBE

[Blue-Marlin]

If you'll stay with this crap - at least change the comment above.

[AnHardt]

QUICK_HOME is broken now. It does not home the axis what not arrived first.
I suppose it never worked, but we did not see that because the 'normal' routines for the axis did the work.

The fix could be to reactivate the normal routines and remove the attempt to home the second axis.

For me the initial rise is completely broken now. Consecutive homings of x or y rise and rise and rise and ... .

[thinkyhead]

QUICK_HOME is broken now

Yeah, sorry, I misconstrued the logic there. It's reverted by the above PR, but keeping the quick_home function for neatness.

the initial rise is completely broken now

If you're using MIN_Z_HEIGHT_FOR_HOMING then the referenced PR should make it better. No more continuous raises on every G28. And it won't do a Z raise if you're only homing Z.