Repository navigation
Stop a crashed QML fixture's children through its process group - #17392
omarchybot wants to merge 1 commit into
Conversation
Fixture cleanup found Quickshell's children by walking down from a live Quickshell, so once Quickshell had crashed (a compositor disconnect, say) the walk never ran and its children, such as the plugin directory's inotifywait, were left to init. Start each fixture's Quickshell in a process group of its own with setsid and signal the group, which works whether or not its leader is still alive. A second fixture in the same file stops the first the same way. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CY6YBoMt9VUxZ6fJqSXaj1
[Low impact] This PR appears safe to merge. SummaryEach QML fixture now starts Quickshell with
Reviews (1) · Last reviewed commit: "Stop a crashed QML fixture's children th..." · Reviewed by Greptile |
|
I read the diff, fixture cleanup and its callers, and the existing review and CI results. My classification is The code review so far is by GPT-6 in Codex (reasoning effort not exposed). No worker tests or technical second review ran; the classification gate stopped verification. The configured review panel answered its availability check, which does not resolve the missing classification. |
Follow-up to #17388, from Greptile's review of its backport (#17390).
#17388 made QML fixture cleanup stop Quickshell's children, but it found them by walking down from a live Quickshell. If Quickshell had already crashed (a compositor disconnect, say), the walk never ran, and children such as the plugin directory's
inotifywaitwere left to init.Each fixture's Quickshell now starts with
setsidand leads its own process group. Cleanup signals the group, so it works whether or not the leader is still alive, and a second fixture in the same file stops the first the same way. In a non-interactive script a background job isn't a group leader, sosetsiddoesn't fork:$!is still Quickshell, and its pid is the group id.Testing
inotifywaitwas still alive in the group afterwards, which the old cleanup would have skipped. The new cleanup stopped it../test/shellpasses 418/418 here and 403/403 onv4-5-0(Backport the shell test fixes to v4-5-0 #17390), with noinotifywaitleft behind.codex review --base origin/quattrofound no issues.🤖 Generated with Claude Code
https://claude.ai/code/session_01CY6YBoMt9VUxZ6fJqSXaj1