Filament loading state fixes #3564
Conversation
|
Is |
wavexx
force-pushed
the
fil_loading_state
branch
from
b0561a0 to
63ac928
Compare
I would be in favor to rename this as auto[_]loading_flag, but we need to this after the MK3_MMU2 branch is merged. |
wavexx
force-pushed
the
fil_loading_state
branch
from
63ac928 to
4b28c35
Compare
|
@gudnimg rereading the code it seems that now eFilamentAction does what we need, so I removed loading_flag completely. The old code was resetting the "autoload" flag for some reason. I changed the meaning so that the "AutoLoad" state is the moment when autoload has been triggered, but loading is still not taking place. AutoLoad can be cancelled (if you're still in the preheat menu), while "Load" is already starting to feed the filament and cannot be undone. IMHO this is a lot saner, and I hope I understood all existing scenarios. Can you review the MMU part? I added one case in lcd_wizard_load which I think was missing. @3d-gussner because eFilamentAction is now used everywhere this requires some retesting of #2651 but also retesting all scenarios with loading/unloading, autoloading+cancel, and using the menu load/load filament correctly clears the load flag, otherwise you won't see the "load/unload" menus anymore. |
wavexx
force-pushed
the
fil_loading_state
branch
from
4b28c35 to
e9e62fd
Compare
wavexx
force-pushed
the
fil_loading_state
branch
from
e9e62fd to
47db468
Compare
|
Should fix #3297 |
gudnimg
force-pushed
the
fil_loading_state
branch
from
47db468 to
37f05bc
Compare
|
@wavexx I rebased the PR to resolve the merge conflicts. Hopefully I didn't break anything. The PR was quite a bit behind MK3 branch :) |
All custom commands are transitory and eventually switch back to Idle state by themselves. It doesn't make any sense to explicitly check for Layer1Cal: any non-idle state is active by design. Fix this check in the main menu. This is probably incomplete (Layer1Cal is incorrectly used in several other places).
Do not blidnly clear the loading_flag, check for it! Just disallowing the SD menu while loading is being performed is not sufficient, since the menu can be entered also by inserting card while loading is taking place. This is also nicer in behavior, as we allow to navigate the SD card while loading.
Remove loading_flag and check for eFilamentAction instead which already flags both load/unload (in addition to mmu actions). Correctly transition from AutoLoad to Load as soon as the operation cannot be cancelled anymore as opposed to resetting it.
gudnimg
force-pushed
the
fil_loading_state
branch
from
37f05bc to
3485c20
Compare
|
Rebased again. For some reason I had to re-do the entire rebase (~600 commits) |
There was a problem hiding this comment.
Approved and fixes issues like #3595
Other tests with MK404 show that this PR is an improvement to previous version.
There are some issues in the management of "filament autoloading" when the printer is idle. Because the autoloading is performed asynchronously, we need to ensure incompatible actions are disabled while loading is taking place.
With this PR we make impossible to perform another autoload/unload/filament action when one is already taking place (the autoload/load/unload menus are now hidden while autoloading is taking place).
To do so we check for the loading_flag. This was previously done for loading only, but doing so for unload makes the behavior uniform (#2318 performs more than a single move and is also non-blocking - you shouldn't load while the old unload is still pending).
Using bFilamentState flag would have been a little bit cleaner here (removing the need for loading_flag), but it's currently used both as a flag for autoload and also for the autoload state...
We also now prevent to start an SD print while autoload is taking place (fixes #2651). You can still navigate the SD card when waiting, but clicking on the file will just beep (see my comments on b0561a0)
As a side note, in the first commit, we change PRINTER_ACTIVE to consider any custom command to be non-idle, which is logical. All custom commands are incompatible with each other, and will return the printer to Idle when complete (LongPause included). Checking for Layer1Cal in the main menu is simply incorrect.
Change in memory (MK3S+ Multilang):
Flash: -8 bytes
SRAM: -1 byte