mirror of
https://github.com/stablyai/orca.git
synced 2026-11-16 10:58:10 +00:00
* fix(linux): release orphaned processes when their owner exits * fix(linux): handle inhibitor errors until streams close --------- Co-authored-by: m4air <m4air@m4airs-MacBook-Air.local>
22 lines
1.2 KiB
Markdown
22 lines
1.2 KiB
Markdown
# Linux sleep inhibitor lifetime regression
|
|
|
|
```sh
|
|
ORCA_BACKGROUND_LAUNCH=1 node config/scripts/run-linux-lid-sleep-assertion-docker.mjs --baseline <git-ref>
|
|
```
|
|
|
|
The runner builds the selected baseline and current production `LinuxLidSleepAssertion`,
|
|
then runs both in a temporary Linux container with real systemd and logind. It checks
|
|
the actual sleep and lid-switch inhibitor, both helper processes, the application's
|
|
scope, and an unrelated canary. The baseline must leak after SIGKILL. The candidate
|
|
must release the lock and helpers after SIGKILL and after an explicit stop; stopping
|
|
the lock must preserve the application itself.
|
|
|
|
Docker must support privileged containers and writable cgroups. Docker Desktop runs
|
|
this inside its Linux VM. The runner removes its container and image afterward.
|
|
`--container <name>` reuses an existing systemd container without removing it.
|
|
`ORCA_DOCKER` and `ORCA_DOCKER_PLATFORM` override the executable and architecture.
|
|
|
|
This reproduces the orphaned `systemd-inhibit`/`sleep` pair in issue #19831's process
|
|
inventory. That pair keeps an otherwise empty scope and sleep lock alive; it does
|
|
not explain the separate multi-gigabyte workload peaks.
|