Stop print timer with M105/M109

[jbrazio]

This is a proposed fix for #3061.
Three hooks have been set in order to stop the print timer:

1. M105
2. M109
3. disable_all_heaters()

Note: M105/M109 hooks only stop the print timer when the temperature of the last extruder is set to zero.

[thinkyhead]

There's a very remote possibility that in a multi-extruder system, one extruder may finish its part and shut off, while the other continues. If the order of operations is COOL 1, HEAT 2 instead of HEAT 2, COOL 1, then the timer would be stopped by this code. I don't really think this is a problem. In any real, working print system, there should never be a time during a print when all heaters are off.

Should other commands, like M81 also turn off the print timer?

[thinkyhead]

Should be float temp = 0; because it's comparing with 0 below…

[Blue-Marlin]

This clock is also used for sd-printing. Does it need changes there?
M109 is not a very reliable criterion for starting the job-timer. I never realized it's also working for USB-prints because I always wait only for M190. For the nozzles I use M104.
How can the clock be reset? Currently it seems to add the time of a second print?

This PR seems to improve the current behaviour. But the concept of starting/stopping it is questionable because most non-3d-printer uses of Marlin do not need any temperature. (Start-g-code? Buffer usage? State?)

[jbrazio]

This code is indeed solving half of the problem. In fact the print timer code is messy, for instance it is only available if SDCARD is defined, and this flag isn't checked in half of the places where Marlin deals with the timer.

There's is also another issue I found when working in this PR but I'm preparing a post about it and will ask for your opinion there.

Nevertheless there are here some loose ends and a couple more hooks have to be introduced into this PR.

[thinkyhead]

With this code repeating we might as well make a function and call it in both places:

bool finish_print_job() {
  if (print_job_start_ms) {
    for (int i = 0; i < EXTRUDERS; i++) if (degTargetHotend(i) > 0) return false;
    print_job_stop_ms = millis();
    return true;
  }
  return false;
}

In gcode_M104:

  finish_print_job();

In gcode_M109:

  if (finish_print_job()) LCD_MESSAGEPGM(WELCOME_MSG);

[jbrazio]

Then I'll take the opportunity to rework the overall code a bit a implement a start_print_job() function also.

[thinkyhead]

I agree with @Blue-Marlin that it's tricky to determine when a print starts or stops. But since we are already making these timers dependent on the wait-for-temperature routines, I don't think it's too far outside the scope to also turn off the timer in the case of all temperatures being turned off. We can also re-set the timer to zero in the case when all heaters are currently turned off, but one or more of them is turned on with a wait option.

Generally, host software will keep its own print timer anyway. This timer is especially for SD printing, but since it hooks into the wait-for-heating stuff it also comes in handy when you have a host that doesn't keep a job timer (if there are any).

[jbrazio]

This is indeed purely a cosmetic change for users who have SDCARD enabled and print from a host as the timer will keep running after everything is finished.

[jbrazio]

Fixed bug: If during the print the nozzle temp is changed by a M109 the timer would reset itself.
Improvement: I've done a rework of print timer in Marlin, now all actions related with the timer have to go through three functions: print_job_start(), print_job_timer() or print_job_stop(); implemented all the feedback received in this thread.

[AnHardt]

Currently M109 does not wait for cooling with 'R'.
Clock functions are ok with sd and temperature triggering.

[AnHardt]

Sorry. Works. My target-temp was below EXTRUDE_MINTEMP/2. So that is also tested as working. ;-)

[jbrazio]

Thanks !
@AnHardt Do you confirm this PR is working properly ?

[AnHardt]

Yes.