s6/src/libs6, branch v2.13.0.0The skarnet.org supervision suitehttps://git.skarnet.org/cgit/s6/atom/src/libs6?h=v2.13.0.02024-05-07T22:06:36Z Add process group support to s6-supervise2024-05-07T22:06:36ZLaurent Bercotska-skaware@skarnet.org2024-05-07T22:06:36Zurn:sha1:15e3e402c59089514f590bab8804f0ed60f0fee3
Signed-off-by: Laurent Bercot <ska@appnovation.com>
this is why I hate free's signature2024-05-05T22:28:02ZLaurent Bercotska-skaware@skarnet.org2024-05-05T22:28:02Zurn:sha1:b827410052a640f65650d1b72d5f63d9ada28933
Signed-off-by: Laurent Bercot <ska@appnovation.com>
s6-ftrigrd: don't have in-object pointers when reallocing2024-05-05T20:51:26ZLaurent Bercotska-skaware@skarnet.org2024-05-05T20:51:26Zurn:sha1:520110086dbb3cbe543fc1930fab3bbac2becc21
Make all pointers out-of-object, even if that means an
additional alloc.
Objects that move need to be flat, so in the absence of a
proper multi-object storage structure, deep allocs is the
only solution.
The alternative is to use gensetdyn so objects don't move,
but here we already have a ftrig1_t, a stralloc and a regex_t
making deep allocs, so who cares about one more - the overhead
of gensetdyn is probably more expensive.
Signed-off-by: Laurent Bercot <ska@appnovation.com>
Restore cleanups in s6-ftrigrd; fix an old bug and a new bug2024-05-03T20:43:37ZLaurent Bercotska-skaware@skarnet.org2024-05-03T20:43:37Zurn:sha1:6d0c114f57bcd554383a41c9ca791e25e95b81b5
ftrigw_clean() didn't clean properly.
ftrig1_make() didn't give enough perms to the fifo, notification
could fail in certain cases.
Signed-off-by: Laurent Bercot <ska@appnovation.com>
bugfix: bad length computation in s6-ftrigrd2024-04-25T12:59:03ZLaurent Bercotska-skaware@skarnet.org2024-04-25T12:59:03Zurn:sha1:a2f093bca2089bd6b3a7211e895a38e86ed8569d
Signed-off-by: Laurent Bercot <ska@appnovation.com>
Don't limit the amount of fifodirs s6-ftrigrd can listen to2024-04-18T01:23:31ZLaurent Bercotska-skaware@skarnet.org2024-04-18T01:23:31Zurn:sha1:d523885d0277ccfbbe5999a6891f481467223912
Signed-off-by: Laurent Bercot <ska@appnovation.com>
Remove s6_ucspiserver_spawn2023-09-12T04:11:07ZLaurent Bercotska-skaware@skarnet.org2023-09-12T04:11:07Zurn:sha1:fe3e7804a79f3e9e953bcd51f36f77787cc0ba11
Signed-off-by: Laurent Bercot <ska@appnovation.com>
Better s6-setlock; delete the s6lock subsystem2023-09-11T08:13:01ZLaurent Bercotska-skaware@skarnet.org2023-09-11T08:13:01Zurn:sha1:bc3863eaf3e4ae92eac8cd3ce0ca9dcb8915fc36
Signed-off-by: Laurent Bercot <ska@appnovation.com>
Adapt to skalibs-2.142023-09-09T04:35:20ZLaurent Bercotska-skaware@skarnet.org2023-09-09T04:35:20Zurn:sha1:1f592173a28907dadfb75132e1eb25dfe4fb3a94
Signed-off-by: Laurent Bercot <ska@appnovation.com>
child_spawn_workaround is public now2023-09-08T08:06:34ZLaurent Bercotska-skaware@skarnet.org2023-09-08T08:06:34Zurn:sha1:8f0980665c368c06adb46a1fddb6790e945c7c51
Signed-off-by: Laurent Bercot <ska@appnovation.com>