]> git.ipfire.org Git - thirdparty/kernel/stable.git/commitdiff
posix-cpu-timers: Force next expiration recalc after itimer reset
authorFrederic Weisbecker <frederic@kernel.org>
Mon, 26 Jul 2021 12:55:10 +0000 (14:55 +0200)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 15 Sep 2021 08:00:28 +0000 (10:00 +0200)
[ Upstream commit 406dd42bd1ba0c01babf9cde169bb319e52f6147 ]

When an itimer deactivates a previously armed expiration, it simply doesn't
do anything. As a result the process wide cputime counter keeps running and
the tick dependency stays set until it reaches the old ghost expiration
value.

This can be reproduced with the following snippet:

void trigger_process_counter(void)
{
struct itimerval n = {};

n.it_value.tv_sec = 100;
setitimer(ITIMER_VIRTUAL, &n, NULL);
n.it_value.tv_sec = 0;
setitimer(ITIMER_VIRTUAL, &n, NULL);
}

Fix this with resetting the relevant base expiration. This is similar to
disarming a timer.

Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20210726125513.271824-4-frederic@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
kernel/time/posix-cpu-timers.c

index aa52fc85dbcbf481f74cba6b464da2e0fa5e718c..a9f8d25220b1a41d93299884d25f306d29b5a5a9 100644 (file)
@@ -1346,8 +1346,6 @@ void set_process_cpu_timer(struct task_struct *tsk, unsigned int clkid,
                        }
                }
 
-               if (!*newval)
-                       return;
                *newval += now;
        }