hakurei/container/container.go, branch v0.3.7low-level userspace tooling for Rosa OShttp://src.rosa.moe/hakurei/atom/container/container.go?h=v0.3.72026-03-11T16:14:03Zcontainer: expose priority and SCHED_OTHER policy2026-03-11T16:14:03ZOphestracat@gensokyo.uk2026-03-11T16:14:03Zurn:sha1:196b200d0f19c41730db439c62db9b39e424f21f
The more explicit API removes the arbitrary limit preventing use of SCHED_OTHER (referred to as SCHED_NORMAL in the kernel). This change also exposes priority value to set.
Signed-off-by: Ophestra <cat@gensokyo.uk>
container: use policy name in log message2026-03-11T11:20:34ZOphestracat@gensokyo.uk2026-03-11T11:20:34Zurn:sha1:1e8ac5f68ef6911c278b4c48722a1924d996206c
This is more helpful than having the user resolve the integer.
Signed-off-by: Ophestra <cat@gensokyo.uk>
container: move scheduler policy constants to std2026-03-11T11:03:08ZOphestracat@gensokyo.uk2026-03-11T11:03:08Zurn:sha1:fd515badffc159b62c5140b8340164ccd29e41f5
This avoids depending on cgo.
Signed-off-by: Ophestra <cat@gensokyo.uk>
container: sched policy string representation2026-03-09T09:38:31ZOphestracat@gensokyo.uk2026-03-09T09:33:49Zurn:sha1:19a2737148609f0b40ea31cf5cd0cb809b60722f
This also uses priority obtained via sched_get_priority_min, and improves bounds checking.
Signed-off-by: Ophestra <cat@gensokyo.uk>
container: use /proc/self/exe directly2026-02-28T11:44:44ZOphestracat@gensokyo.uk2026-02-28T11:42:33Zurn:sha1:c74c269b66770f0eddde255d2805b79f274409e9
This is a more reliable form of pathname to self and also cheaper than os.Executable.
Signed-off-by: Ophestra <cat@gensokyo.uk>
container: improve documentation2026-02-28T11:18:30ZOphestracat@gensokyo.uk2026-02-28T11:18:30Zurn:sha1:cd9b534d6b1d432d3c997ccc5b14f630afb9dda6
This change removes inconsistencies collected over time in this package.
Signed-off-by: Ophestra <cat@gensokyo.uk>
container: set scheduling policy2026-02-26T07:29:47ZOphestracat@gensokyo.uk2026-02-26T07:29:47Zurn:sha1:a6160cd410b1def222018cd88951fcbbf1a65f5c
This is thread-directed so cannot be done externally. The glibc wrapper exposes this behaviour so most multithreaded programs using this is straight up incorrect.
Signed-off-by: Ophestra <cat@gensokyo.uk>
container/init: configure interface lo2026-01-10T18:36:20ZOphestracat@gensokyo.uk2026-01-10T18:32:21Zurn:sha1:2494ede1067d5071d3cc7d355cdc2ae3fac66dc5
This enables loopback networking when owning the net namespace.
Signed-off-by: Ophestra <cat@gensokyo.uk>
container: optionally allow orphan2025-12-27T00:12:02ZOphestracat@gensokyo.uk2025-12-27T00:12:02Zurn:sha1:3d720ada922def503977d6972aa953ffe924e714
This is required for the typical daemonise use case.
Signed-off-by: Ophestra <cat@gensokyo.uk>
container: set FD_CLOEXEC on all open files2025-11-11T15:18:29ZOphestracat@gensokyo.uk2025-11-11T15:18:11Zurn:sha1:ac3463589035394bedbedb15a323298251ed2c14
While fd created from this side always has the FD_CLOEXEC flag, the same is not true for files left open by the parent. This change prevents those files from leaking into the container.
Signed-off-by: Ophestra <cat@gensokyo.uk>