blob: bcdfda2d5f21311badc2864b11ece33ddde85290 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
|
From af116ecf443ae48632323179dbb878c0aa24f656 Mon Sep 17 00:00:00 2001
From: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Date: Mon, 2 Nov 2020 14:14:24 +0100
Subject: [PATCH 067/323] timers: Don't block on ->expiry_lock for
TIMER_IRQSAFE
Origin: https://www.kernel.org/pub/linux/kernel/projects/rt/5.10/older/patches-5.10.215-rt107.tar.xz
PREEMPT_RT does not spin and wait until a running timer completes its
callback but instead it blocks on a sleeping lock to prevent a deadlock.
This blocking can not be done for workqueue's IRQ_SAFE timer which will
be canceled in an IRQ-off region. It has to happen to in IRQ-off region
because changing the PENDING bit and clearing the timer must not be
interrupted to avoid a busy-loop.
The callback invocation of IRQSAFE timer is not preempted on PREEMPT_RT
so there is no need to synchronize on timer_base::expiry_lock.
Don't acquire the timer_base::expiry_lock for TIMER_IRQSAFE flagged
timer.
Add a lockdep annotation to ensure that this function is always invoked
in preemptible context on PREEMPT_RT.
Reported-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: stable-rt@vger.kernel.org
---
kernel/time/timer.c | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/kernel/time/timer.c b/kernel/time/timer.c
index c135cefa44ac0..68246a8c31bc2 100644
--- a/kernel/time/timer.c
+++ b/kernel/time/timer.c
@@ -1316,7 +1316,7 @@ static void del_timer_wait_running(struct timer_list *timer)
u32 tf;
tf = READ_ONCE(timer->flags);
- if (!(tf & TIMER_MIGRATING)) {
+ if (!(tf & (TIMER_MIGRATING | TIMER_IRQSAFE))) {
struct timer_base *base = get_timer_base(tf);
/*
@@ -1402,6 +1402,13 @@ int timer_delete_sync(struct timer_list *timer)
*/
WARN_ON(in_irq() && !(timer->flags & TIMER_IRQSAFE));
+ /*
+ * Must be able to sleep on PREEMPT_RT because of the slowpath in
+ * del_timer_wait_running().
+ */
+ if (IS_ENABLED(CONFIG_PREEMPT_RT) && !(timer->flags & TIMER_IRQSAFE))
+ lockdep_assert_preemption_enabled();
+
do {
ret = try_to_del_timer_sync(timer);
--
2.44.0
|
