Open lilyinstarlight opened 1 year ago
SIGUSR1 is not intended for external processes to indicate success. It's a feature to unlock the session when it's stuck.
SIGUSR1 is not intended for external processes to indicate success. It's a feature to unlock the session when it's stuck.
Adding a SIGUSR2 was suggested in https://github.com/swaywm/swaylock/pull/283#issuecomment-1678615442. If this is not the correct approach, is there something else that can be done to support this use-case?
Thanks!
swaylock is not meant to be extensible through plugins. I don't think we want to support extensibility because of the complexity. Regardless, signals in particular are not a great API.
I would suggest forking swaylock or creating your own screen locker with fprintd support.
Okay, feel free to close this then. Thanks!
This adds support for indicating external authentication failures using
SIGUSR2
, in addition to the existing support for external authentication successes usingSIGUSR1
Since
SIGUSR*
defaults to termination, we ignore them in the forked pw backend children (which worked already becauseSIGUSR1
ends up in termination anyway, but we do not want that to happen forSIGUSR2
)This is intended to be usable with an external program that can handle fprintd communication as an alternative to #283
Supersedes and closes #311