Fix bilinear levelling z offset

[benlye]

Since run_probe was altered to return the probe Z position rather than the nozzle Z position bilinear levelling has been broken because the Z-offset has been applied twice - once in the run_probe function, and then again in the G29 code for bilinear levelling. This results in Z0 being below the bed when the Z-offset is negative.

On the assumption that the change to run_probe is needed elsewhere, this PR removes the extra addition of the Z offset in the G29 code for bilinear levelling.

[benlye]

Fixes the issue described in #6233, caused by a5e085c, committed in #6181.

[MagoKimbra]

zoffset is an offset that can be given with the subcommand Z, in that case we PROBE_MANUALLY. If you are not from the command zoffset Z is 0 ...

zoffset = code_seen('Z') ? code_value_axis_units(Z_AXIS) : 0;
#if HAS_BED_PROBE
  zoffset += zprobe_zoffset;
#endif

Since we are with the PROBE_MANUALLY HAS_BED_PROBE is false then zoffset is zero or the value given by the command Z.

[benlye]

OK. If I understand you correctly, my change is going to 'fix' the Z probe offset being applied twice but only by accident, and is also going to break passing a Z subcommand to G29.

In that case, is the way to address the double-addition of the Z probe offset to the bilinear correction grid simply to remove these lines?

        #if HAS_BED_PROBE
          zoffset += zprobe_zoffset;
        #endif

[benlye]

I compiled and tested with my initial proposed change reverted and the zoffset += zprobe_zoffset; line commented instead and bilinear levelling is working correctly and giving the expected output and correction.

Let me know what to do with this PR - I'm happy to update/close/resubmit as appropriate.

[MagoKimbra]

you're right!

[benlye]

@MagoKimbra @thinkyhead - I'm a little confused now. This was merged as-is, with my original proposed change, which seemed to be flawed. Won't the Z parameter to G29 now be ignored?

[thinkyhead]

Thanks for #6258. That makes more sense.