linux-kernel - [PATCH 2/2] sched/fair: Fix vruntime_normalized() for TASK

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-Id: <20220924160429.71312-2-zhouchengming@bytedance.com>
Date:   Sun, 25 Sep 2022 00:04:29 +0800
From:   Chengming Zhou <zhouchengming@...edance.com>
To:     mingo@...hat.com, peterz@...radead.org, juri.lelli@...hat.com,
        vincent.guittot@...aro.org, dietmar.eggemann@....com,
        rostedt@...dmis.org, bsegall@...gle.com
Cc:     linux-kernel@...r.kernel.org,
        Chengming Zhou <zhouchengming@...edance.com>
Subject: [PATCH 2/2] sched/fair: Fix vruntime_normalized() for TASK_NEW

When !on_rq, vruntime of the task has usually NOT been normalized.
But there are two cases where it has already been normalized:
1. A forked child which is waiting for being woken up by
   wake_up_new_task().
2. A task which has been woken up by try_to_wake_up() and
   waiting for actually being woken up by sched_ttwu_pending().

The current code use !se->sum_exec_runtime to check the first case,
which is not always correct, like a !on_rq RT task switched_to_fair()
for the first time, its se->sum_exec_runtime == 0 but not TASK_NEW.

Although its vruntime is normalized indeed, we should add cfs_rq->min_vruntime
to be renormalized for this !on_rq !fair task when switched_to_fair().
Or later enqueue_entity() won't renorm it either, which cause abnormal
vruntime. OTOW, vruntime_normalized() should return false for this case.

Fix it by using the clearer TASK_NEW state to only consider a forked
child which is waiting for being woken up by wake_up_new_task().

Signed-off-by: Chengming Zhou <zhouchengming@...edance.com>
---
 kernel/sched/fair.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index dea86d8a6c02..36614bfd452f 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -11655,8 +11655,6 @@ prio_changed_fair(struct rq *rq, struct task_struct *p, int oldprio)

 static inline bool vruntime_normalized(struct task_struct *p)
 {
-	struct sched_entity *se = &p->se;
-
 	/*
 	 * In both the TASK_ON_RQ_QUEUED and TASK_ON_RQ_MIGRATING cases,
 	 * the dequeue_entity(.flags=0) will already have normalized the
@@ -11674,7 +11672,7 @@ static inline bool vruntime_normalized(struct task_struct *p)
 	 * - A task which has been woken up by try_to_wake_up() and
 	 *   waiting for actually being woken up by sched_ttwu_pending().
 	 */
-	if (!se->sum_exec_runtime ||
+	if (READ_ONCE(p->__state) == TASK_NEW ||
 	    (READ_ONCE(p->__state) == TASK_WAKING && p->sched_remote_wakeup))
 		return true;

-- 
2.37.2