Messages in this thread Patch in this message | | | Date | Fri, 23 Aug 2019 02:12:15 -0000 | From | tip-bot2 for Julien Grall <> | Subject | [tip: timers/core] hrtimer: Protect lockless access to timer->base |
| |
The following commit has been merged into the timers/core branch of tip:
Commit-ID: dd2261ed45aaeddeb77768f291d604179bcab096 Gitweb: https://git.kernel.org/tip/dd2261ed45aaeddeb77768f291d604179bcab096 Author: Julien Grall <julien.grall@arm.com> AuthorDate: Wed, 21 Aug 2019 10:24:07 +01:00 Committer: Thomas Gleixner <tglx@linutronix.de> CommitterDate: Wed, 21 Aug 2019 16:10:01 +02:00
hrtimer: Protect lockless access to timer->base
The update to timer->base is protected by the base->cpu_base->lock(). However, hrtimer_cancel_wait_running() does access it lockless. So the compiler is allowed to refetch timer->base which can cause havoc when the timer base is changed concurrently.
Use READ_ONCE() to prevent this.
[ tglx: Adapted from a RT patch ]
Signed-off-by: Julien Grall <julien.grall@arm.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Link: https://lkml.kernel.org/r/20190821092409.13225-2-julien.grall@arm.com --- kernel/time/hrtimer.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/kernel/time/hrtimer.c b/kernel/time/hrtimer.c index 8333537..f48864e 100644 --- a/kernel/time/hrtimer.c +++ b/kernel/time/hrtimer.c @@ -1214,7 +1214,8 @@ static void hrtimer_sync_wait_running(struct hrtimer_cpu_base *cpu_base, */ void hrtimer_cancel_wait_running(const struct hrtimer *timer) { - struct hrtimer_clock_base *base = timer->base; + /* Lockless read. Prevent the compiler from reloading it below */ + struct hrtimer_clock_base *base = READ_ONCE(timer->base); if (!timer->is_soft || !base || !base->cpu_base) { cpu_relax();
| |