diff options
| author | Ritesh Raj Sarraf <rrs@debian.org> | 2016-11-18 13:28:15 (GMT) |
|---|---|---|
| committer | Ritesh Raj Sarraf <rrs@debian.org> | 2016-11-18 13:35:06 (GMT) |
| commit | 695afe51215a28a05f2ce0fba2656f32f2a0a421 (patch) | |
| tree | 179475f161118074fab7b6c7fcbadd0645b418ab | |
| parent | f4d6b93e33bb7a1fd030f073fb1bb48c0e60f4ba (diff) | |
It seems the systemd/udev folks have broken something, quietly, again. Earlier, it was told that udev/systemd would kill a script running too long in the foreground. Now, it looks like running a script, in the background, is quietly killed. Sigh!
Update with some comments
| -rwxr-xr-x | etc/rules/lmt-udev | 9 |
1 files changed, 5 insertions, 4 deletions
diff --git a/etc/rules/lmt-udev b/etc/rules/lmt-udev index a3273f7..04bab0c 100755 --- a/etc/rules/lmt-udev +++ b/etc/rules/lmt-udev @@ -46,13 +46,14 @@ else # # Also, if this happens during boot, we may want to wait until /usr is available # - # That said, we background the execution here, because, otherwise udevd will wait - # for this process which will block - + # On newer systemd/udev (232-3 over here), the behavior has changed in my tests. + # Now, it'll immediately and quietly kill the process that is backgrounded + # systemd is nice, but I hate when it tries to become a magician + # # This stanza will also be called for device plug/unplug events exec > /dev/null 2>/dev/null - non_systemd_way "$@" & + non_systemd_way "$@" [ -z $LMT_DEBUG ] || echo "regulard is $PPID $$ $@" > $LMT_DEBUG fi |