Fix runout

[GMagician]

May fix #26061

Needs a seriuous review before testing it

I decided to limit changes only when runout works with distance because it shouldn't be a problem in other cases

[thisiskeithb]

Converted to draft since the patch doesn’t seem to work: #26061 (comment)

[justvlade]

@GMagician Yes, looks like now it is counting runout mm correctly, at least with retract_test_ps_464_fw.gcode. But for some reason printer is reloading while printing. I checked that its behavior depends on MIN_STEPS_PER_SEGMENT value.
If I set MIN_STEPS_PER_SEGMENT=1 then it is counting runout mm right, and successfully detects runout and runs M600 command to change filament. But if MIN_STEPS_PER_SEGMENT=6 (as default for ender 2 pro config), then runout is calculated correctly, but for some reason printer reboots during print. I also changed current bugfix code with MIN_STEPS_PER_SEGMENT=6 and it completes the print as it cannot detect runout.
I am attaching those logs in zip. I have found a way to get full logs using tail -f directly from printer, so those logs are full from start to finish with maximum verbosity.

I have used this branch to test patch https://github.com/GMagician/Marlin/tree/850d848e942682c284cd92174b304ec7b82f220d (as I understand it is current bugfix with patch applied)

log_M1_LOG.txt - MIN_STEPS_PER_SEGMENT=1 with patch
log_M6_LOG.txt - MIN_STEPS_PER_SEGMENT=6 with patch
log_M6_LOG_NOP2.txt - MIN_STEPS_PER_SEGMENT=6 current bugfix

patch_850d848_logs.zip

[GMagician]

@justvlade I was sure not to reenter the calling block that may generate a loop but maybe I'm wrong. I approached in a different way this issue and if it runs it's a lot cleaner than this (and also more resources saver). May you try it? it's #26082

[GMagician]

Replaced by #26082