[BUG] Screen throttling can't handle fast displays

[jmz52]

Bug Description

Display with update time less than 1 ms suffers from screen throttling.
uint16_t max_display_update_time is 0 due to really fast display update time.
max_display_update_time-- set max_display_update_time to 65535, meaning next screen update will be in 18 hours.

My Configurations

MKS Robin with reworked LCD subsystem

Steps to Reproduce

None.

Expected behavior: [What you expect to happen]
UI is displayed when printing from SD card

Actual behavior: [What actually happens]
Blank screen when printing from SD card

Additional Information

Possible solutions:

• Replace max_display_update_time--; with if (max_display_update_time) max_display_update_time--; in ultralcd.cpp (tested)
• Change max_display_update_time type to int16_t (not tested)
• Add 1 ms delay to MarlinUI::draw_status_screen() (tested)

[GMagician]

Where do you see max_display_update_time--? I can't see it in bugfix 2.0.x sources

[jmz52]

https://github.com/MarlinFirmware/Marlin/blob/2.0.x/Marlin/src/lcd/ultralcd.cpp#L970

[GMagician]

Download bugfix 2.0...it has been changed and there is no more such line

[jmz52]

Sorry, wrong branch.
https://github.com/MarlinFirmware/Marlin/blob/bugfix-2.0.x/Marlin/src/lcd/ultralcd.cpp#L884

[GMagician]

I found it....I not analyzed how lcd works but it seems you did it and since you found that I think you may post a PR. If in trouble posting it I may do it for you

[AnHardt]

When i wrote the code i never assumed a screen update could be as short as below a ms.
if (max_display_update_time) max_display_update_time--; seems to be the right solution.
If a screen update does not last longer, the probability the planner-buffer runs dry in that time is extremely low. Wasting a ms by an artificial delay would be wrong.

The line exists to make the algorithm adaptive. A very long one time screen-update (for example interrupted by something) should not have influence forever.

[jmz52]

@GMagician, thanks for the PR

[github-actions]

This issue has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.