psd-suspend-sync
rstech209 opened this issue · comments
Description and solution of the problem initially proposed in PR #313 deleted by mistake
I think that's OK now...I rewrite the PR #317 to fix this problem...
History:
By testing psd-suspend-sync
function, I was able to verify that following the improvement made by smac89 (# 276) that inhibit lock is functional (by reproducing the problem with the previous version and see the expiration of the InhibitDelayMaxSec
timeout).
There is one more problem remaining:
A process gdbus monitor --system --dest org.freedesktop.login1
becomes orphan after each resume.
That's because killing the script psd-suspend-sync
does not stop this process.
In addition, I found that killing the script psd-suspend-sync
after each resume and restarting 2 gdbus
processes by cycle (one process initialized with the coproc
for suspend (when creating the inhibit lock) and one process for resume), is not optimal.
On the other hand, the release_inhibit_lock ()
function of the main script will try to kill the psd-suspend-sync
script regardless of the user (because of the pgrep -cf "psd-suspend-sync"
command) and will generate error messages. This situation is problematic in a multi-user environment.