History

sched_ext: Handle SCX_TASK_NONE in disable/switched_from paths

scx_fail_parent() leaves cgroup tasks at (state=NONE, sched=parent, sched_class=ext) until the parent itself is torn down by the scx_error() it raised. When the later root_disable iterates them, two paths trip on NONE. scx_disable_and_exit_task() re-enters the wrapper at NONE: the inner switch returns early but the trailing scx_set_task_sched(p, NULL) clobbers the parent sched left by scx_fail_parent(), and scx_set_task_state(p, NONE) wastes a write on an already-NONE task. switched_from_scx() then calls scx_disable_task(), which WARNs on non-ENABLED state and writes state=READY, producing a NONE -> READY transition the validation matrix rejects. Treat NONE as "nothing to do" in both paths. Add a NONE early-return at the top of scx_disable_and_exit_task() and a parallel NONE check in switched_from_scx() next to task_dead_and_done().

Tejun Heo committed 6 months ago

Commit d3e73a01 parent cd6aab71 file · +19 −0Browse files

Showing the changes to kernel (1 of 1 file). Show all

Unified · Split

kernel/sched/ext.c+19−0
@@ -3703,6 +3703,15 @@ static void scx_sub_init_cancel_task(struct scx_sched *sch, struct task_struct * · 3702 unmodified lines
37033703 static void scx_disable_and_exit_task(struct scx_sched *sch,
37043704 struct task_struct *p)
37053705 {
3706+ /*
3707+ * %NONE means @p is already detached at the SCX level (e.g. handed
3708+ * back to the parent by scx_fail_parent() with no init to undo).
3709+ * Skip to avoid clobbering scx_task_sched() and writing %NONE again
3710+ * on a state that's already %NONE.
3711+ */
3712+ if (scx_get_task_state(p) == SCX_TASK_NONE)
3713+ return;
3714+
37063715 __scx_disable_and_exit_task(sch, p);
37073716
37083717 /*
@@ -3921,6 +3930,16 @@ static void switched_from_scx(struct rq *rq, struct task_struct *p) · 212 unmodified lines
39213930 if (task_dead_and_done(p))
39223931 return;
39233932
3933+ /*
3934+ * %NONE means SCX is no longer tracking @p at the task level (e.g.
3935+ * scx_fail_parent() handed @p back to the parent at NONE pending the
3936+ * parent's own teardown). There is nothing to disable; calling
3937+ * scx_disable_task() would WARN on the non-%ENABLED state and trigger a
3938+ * NONE -> READY validation failure.
3939+ */
3940+ if (scx_get_task_state(p) == SCX_TASK_NONE)
3941+ return;
3942+
39243943 scx_disable_task(scx_task_sched(p), p);
39253944 }
39263945