fix: store state in deployment data #76
Conversation
danielskinstad
requested review from
larsewi,
lluiscampos,
vpodzime and
pasinskim
and removed request for
larsewi
| @@ -518,7 +518,6 @@ mender_client_initialization_work_function(void) { | |||
|
|
|||
| /* Delete pending deployment */ | |||
| mender_storage_delete_deployment_data(); | |||
There was a problem hiding this comment.
We discussed should call the data update data, it's no longer just a deployment info.
There was a problem hiding this comment.
Right, you think I should rename everything that's named deployment data to update data (including mender-deployment-data.c etc.?)
There was a problem hiding this comment.
We discussed should call the data update data (...)
To be fair with Daniel in the ticket we wrote "consider renaming..."
There was a problem hiding this comment.
Yeah, I but IIRC we discussed it on Slack or somewhere else too...
There was a problem hiding this comment.
I can do it in a separate PR
core/src/mender-client.c
Outdated
| @@ -1040,9 +1048,8 @@ mender_client_update_work_function(void) { | |||
| * verification should happen anyway, the callback in that | |||
| * state should be able to see if things went well or | |||
| * wrong. */ | |||
| mender_storage_save_update_state(MENDER_UPDATE_STATE_VERIFY_REBOOT, mender_update_module->artifact_type); | |||
| mender_deployment_data_set_state(mender_client_deployment_data, MENDER_UPDATE_STATE_VERIFY_REBOOT); | |||
There was a problem hiding this comment.
What's the point of setting the state here again?
There was a problem hiding this comment.
According to the comment above it's to ensure that reboot verification happens in case of a crash in the reboot-callback, but it's also set at the start of the code, so I guess this is unnecessary?
danielskinstad
force-pushed
the
update-state
branch
from
a5527fb
to
b8fad25
Compare
| * verification should happen anyway, the callback in that | ||
| * state should be able to see if things went well or | ||
| * wrong. */ | ||
| mender_storage_save_update_state(MENDER_UPDATE_STATE_VERIFY_REBOOT, mender_update_module->artifact_type); |
There was a problem hiding this comment.
I think we still want do do set_and_store_state() here iff there is a reboot callback. If there isn't, it will be called by NEXT_STATE below anyway. It just wouldn't be called twice in such a case.
There was a problem hiding this comment.
I moved it outside of the if-statement, but I moved it back now, because it's stored in NEXT_STATE as you said
danielskinstad
force-pushed
the
update-state
branch
from
b8fad25
to
b020054
Compare
They are passed in as NULL and populated later. Ticket: None Signed-off-by: Daniel Skinstad Drabitzius <[email protected]>
Removed the key and corresponding functions for storing the update_state and the artifact type. It is now a part of the deployment data. Changelog: Title Ticket: MEN-7515 Signed-off-by: Daniel Skinstad Drabitzius <[email protected]>
danielskinstad
force-pushed
the
update-state
branch
from
b020054
to
771da43
Compare
|
Merging these commits will result in the following changelog entries: Changelogsmender-mcu (update-state)New changes in mender-mcu since main: Bug Fixes
|
Removed the key and corresponding functions for storing the update_state
and the artifact type. It is now a part of the deployment data.
Changelog: Title
Ticket: MEN-7515