This way if someone wants a longer or shorter timeout that possible
by setting the env var SXMO_SUSPENDTIMEOUTS.
Also change the default to be 35s. This is roughly the duration of
ringing for an incoming phone call.
With the megi 5.9 kernel and other suspend-related merge requests in
postmarketos, disabling these drivers is no longer needed. All of this
should result in a more reliable suspending system.
Source: https://gitlab.com/postmarketOS/pmaports/-/merge_requests/1651
Signed-off-by: Miles Alan <m@milesalan.com>
Stop a compiler warning about overwriting the 100 char buffer with larger
string.
programs/sxmo_screenlock.c:61:3: note: ‘sprintf’ output between 39 and 294 bytes into a destination of size 100
E.g. so now you distinctively have to press down the powerbutton three times
rather than holding (which is easy to do inadvertently in pocket) to exit
screenlock.
Please apply to the crust feature branch. The rtc wakeup clock is useful cause it allows
users to script programs. If I put my phone into deep sleep, none of the scripts
I put in my cronjob will run.
For example, if I have an alarm script:
https://git.sr.ht/~anjan/sxmo-userscripts/tree/master/alarm.sh
It is supposed to run at 7:00 am but since I put my phone in deep sleep before going to sleep,
the script doesnt play sound to wake me up.
Enabling the rtc wakeup clock enables using programs like rtcwake as a wrapper around
your cronjobs to wake up the phone at a time and excecute the command you put in cron.
For info on how to use rtcwake, see:
https://wiki.postmarketos.org/wiki/Nokia_N900_(nokia-n900)#Setting_the_hardware_alarm_clock
Not sure if we should patch busybox cron to use rtcwake automatically or whether we should
trust the user to use rtcwake properly and document it. However, I am certain that the rtc
wakeup source should be enabled for this usecase.
Signed-off-by: Miles Alan <m@milesalan.com>
