scx_flush_disable_work(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

✨ NAME ✨

okay so bestie, this is scx_flush_disable_work(9) from the Kernel API Manual, and its whole personality is: flush the disable work and wait for it to finish. that's the assignment. πŸ’…

🌸 SYNOPSIS 🌸

void scx_flush_disable_work (struct scx_sched *sch );

πŸŽ€ ARGUMENTS πŸŽ€

  • sch β€” the scheduler. literally the scheduler you're working with, no extra lore. 🫢

πŸ¦‹ DESCRIPTION πŸ¦‹

so here's the vibe: sch->disable_work might still not be queued yet β€” yes, the original wording is a little β€œEnglish is a construct” β€” and that can make kthread_flush_work a total noop. like, it's waiting on something that hasn't even shown up to the group chat. 😭

the fix is to sync the irq_work first. that's required to guarantee the kthread work has actually been queued before you wait for it. no cap, order matters here. πŸ”₯

πŸ‘‘ SEE ALSO πŸ‘‘

Kernel file: ./kernel/sched/ext/ext.c β€” and the whole posse is:


with love from ./kernel/sched/ext 🫢 β€” stamped October 2026, and this was the scx_flush_disable_work(9) page. stay hydrated, bestie πŸ’–βœ¨

πŸ“¦ where this page comes from