linux-kernel - [patch 11/20] posix-timers: Document common_clock

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

Message-ID: <20230425183313.409169321@linutronix.de>
Date:   Tue, 25 Apr 2023 20:49:12 +0200 (CEST)
From:   Thomas Gleixner <tglx@...utronix.de>
To:     LKML <linux-kernel@...r.kernel.org>
Cc:     Frederic Weisbecker <frederic@...nel.org>,
        Anna-Maria Behnsen <anna-maria@...utronix.de>,
        Peter Zijlstra <peterz@...radead.org>,
        Sebastian Siewior <bigeasy@...utronix.de>,
        syzbot+5c54bd3eb218bb595aa9@...kaller.appspotmail.com,
        Dmitry Vyukov <dvyukov@...gle.com>,
        Michael Kerrisk <mtk.manpages@...il.com>
Subject: [patch 11/20] posix-timers: Document common_clock_get() correctly

Replace another confusing and inaccurate set of comments.

Signed-off-by: Thomas Gleixner <tglx@...utronix.de>
---
 kernel/time/posix-timers.c |   50 +++++++++++++++++++++++++++------------------
 1 file changed, 30 insertions(+), 20 deletions(-)

--- a/kernel/time/posix-timers.c
+++ b/kernel/time/posix-timers.c
@@ -659,20 +659,16 @@ static s64 common_hrtimer_forward(struct
 }
 
 /*
- * Get the time remaining on a POSIX.1b interval timer.  This function
- * is ALWAYS called with spin_lock_irq on the timer, thus it must not
- * mess with irq.
+ * Get the time remaining on a POSIX.1b interval timer.
  *
- * We have a couple of messes to clean up here.  First there is the case
- * of a timer that has a requeue pending.  These timers should appear to
- * be in the timer list with an expiry as if we were to requeue them
- * now.
+ * Two issues to handle here:
  *
- * The second issue is the SIGEV_NONE timer which may be active but is
- * not really ever put in the timer list (to save system resources).
- * This timer may be expired, and if so, we will do it here.  Otherwise
- * it is the same as a requeue pending timer WRT to what we should
- * report.
+ *  1) The timer has a requeue pending. The return value must appear as
+ *     if the timer has been requeued right now.
+ *
+ *  2) The timer is a SIGEV_NONE timer. These timers are never enqueued
+ *     into the hrtimer queue and therefore never expired. Emulate expiry
+ *     here taking #1 into account.
  */
 void common_timer_get(struct k_itimer *timr, struct itimerspec64 *cur_setting)
 {
@@ -688,8 +684,12 @@ void common_timer_get(struct k_itimer *t
 		cur_setting->it_interval = ktime_to_timespec64(iv);
 	} else if (!timr->it_active) {
 		/*
-		 * SIGEV_NONE oneshot timers are never queued. Check them
-		 * below.
+		 * SIGEV_NONE oneshot timers are never queued and therefore
+		 * timr->it_active is always false. The check below
+		 * vs. remaining time will handle this case.
+		 *
+		 * For all other timers there is nothing to update here, so
+		 * return.
 		 */
 		if (!sig_none)
 			return;
@@ -698,18 +698,29 @@ void common_timer_get(struct k_itimer *t
 	now = kc->clock_get_ktime(timr->it_clock);
 
 	/*
-	 * When a requeue is pending or this is a SIGEV_NONE timer move the
-	 * expiry time forward by intervals, so expiry is > now.
+	 * If this is an interval timer and either has requeue pending or
+	 * is a SIGEV_NONE timer move the expiry time forward by intervals,
+	 * so expiry is > now.
 	 */
 	if (iv && (timr->it_requeue_pending & REQUEUE_PENDING || sig_none))
 		timr->it_overrun += kc->timer_forward(timr, now);
 
 	remaining = kc->timer_remaining(timr, now);
-	/* Return 0 only, when the timer is expired and not pending */
+	/*
+	 * As @now is retrieved before a possible timer_forward() and
+	 * cannot be reevaluated by the compiler @remaining is based on the
+	 * same @now value. Therefore @remaining is consistent vs. @now.
+	 *
+	 * Consequently all interval timers, i.e. @iv > 0, cannot have a
+	 * remaining time <= 0 because timer_forward() guarantees to move
+	 * them forward so that the next timer expiry is > @now.
+	 */
 	if (remaining <= 0) {
 		/*
-		 * A single shot SIGEV_NONE timer must return 0, when
-		 * it is expired !
+		 * A single shot SIGEV_NONE timer must return 0, when it is
+		 * expired! Timers which have a real signal delivery mode
+		 * must return a remaining time greater than 0 because the
+		 * signal has not yet been delivered.
 		 */
 		if (!sig_none)
 			cur_setting->it_value.tv_nsec = 1;
@@ -718,7 +729,6 @@ void common_timer_get(struct k_itimer *t
 	}
 }
 
-/* Get the time remaining on a POSIX.1b interval timer. */
 static int do_timer_gettime(timer_t timer_id,  struct itimerspec64 *setting)
 {
 	struct k_itimer *timr;