All of lore.kernel.org
 help / color / mirror / Atom feed
From: Thomas Gleixner <tglx@linutronix.de>
To: LKML <linux-kernel@vger.kernel.org>
Cc: Frederic Weisbecker <frederic@kernel.org>,
	Anna-Maria Behnsen <anna-maria@linutronix.de>,
	John Stultz <jstultz@google.com>,
	Peter Zijlstra <peterz@infradead.org>,
	Ingo Molnar <mingo@kernel.org>, Stephen Boyd <sboyd@kernel.org>,
	Eric Biederman <ebiederm@xmission.com>,
	Oleg Nesterov <oleg@redhat.com>
Subject: [patch 29/45] posix-timers: Make signal overrun accounting sensible
Date: Tue,  6 Jun 2023 16:38:04 +0200 (CEST)	[thread overview]
Message-ID: <20230606142032.655376396@linutronix.de> (raw)
In-Reply-To: 20230606132949.068951363@linutronix.de

The handling of the timer overrun in the signal code is inconsistent as it
takes previous overruns into account. This is just wrong as after the
reprogramming of a timer the overrun count starts over from a clean state,
i.e. 0.

Make the accounting in send_sigqueue() consistent with that.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
---
 kernel/signal.c |   34 ++++++++++++++++++++++++++++------
 1 file changed, 28 insertions(+), 6 deletions(-)

--- a/kernel/signal.c
+++ b/kernel/signal.c
@@ -1957,6 +1957,34 @@ int send_sigqueue(struct sigqueue *q, st
 	 */
 	q->info.si_sys_private = si_private;
 
+	/*
+	 * Set the overrun count to zero unconditionally. The posix timer
+	 * code does not self rearm periodic timers. They are rearmed from
+	 * dequeue_signal().
+	 *
+	 * But there is a situation where @q is already enqueued:
+	 *
+	 * 1) timer_settime()
+	 *      arm_timer()
+	 * 2) timer_expires()
+	 *      send_sigqueue(@q)
+	 *        enqueue(@q)
+	 * 3) timer_settime()
+	 *      arm_timer()
+	 * 4) timer_expires()
+	 *      send_sigqueue(@q) <- Observes @q already queued
+	 *
+	 * In this case incrementing si_overrun does not make sense because
+	 * there is no relationship between timer_settime() #1 and #2.
+	 *
+	 * The POSIX specification is useful as always: "The effect of
+	 * disarming or resetting a timer with pending expiration
+	 * notifications is unspecified."
+	 *
+	 * Just do the sensible thing and reset the overrun.
+	 */
+	q->info.si_overrun = 0;
+
 	ret = 1; /* the signal is ignored */
 	result = TRACE_SIGNAL_IGNORED;
 	if (!prepare_signal(sig, t, false))
@@ -1964,15 +1992,9 @@ int send_sigqueue(struct sigqueue *q, st
 
 	ret = 0;
 	if (unlikely(!list_empty(&q->list))) {
-		/*
-		 * If an SI_TIMER entry is already queue just increment
-		 * the overrun count.
-		 */
-		q->info.si_overrun++;
 		result = TRACE_SIGNAL_ALREADY_PENDING;
 		goto out;
 	}
-	q->info.si_overrun = 0;
 
 	signalfd_notify(t, sig);
 	pending = (type != PIDTYPE_PID) ? &t->signal->shared_pending : &t->pending;


  parent reply	other threads:[~2023-06-06 14:40 UTC|newest]

Thread overview: 66+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-06-06 14:37 [patch 00/45] posix-timers: Cure inconsistencies and the SIG_IGN mess Thomas Gleixner
2023-06-06 14:37 ` [patch 01/45] selftests/timers/posix_timers: Make signal distribution test less fragile Thomas Gleixner
2023-06-06 14:37 ` [patch 02/45] selftests/timers/posix_timers: Use TAP reporting format Thomas Gleixner
2023-06-06 14:37 ` [patch 03/45] selftests/timers/posix_timers: Add SIG_IGN test Thomas Gleixner
2023-06-06 14:37 ` [patch 04/45] selftests/timers/posix_timers: Validate signal rules Thomas Gleixner
2023-06-06 14:37 ` [patch 05/45] selftests/timers/posix-timers: Validate SIGEV_NONE Thomas Gleixner
2023-06-06 14:37 ` [patch 06/45] selftests/timers/posix-timers: Validate timer_gettime() Thomas Gleixner
2023-06-06 14:37 ` [patch 07/45] selftests/timers/posix-timers: Validate overrun after unblock Thomas Gleixner
2023-06-06 14:37 ` [patch 08/45] posix-timers: Convert timer list to hlist Thomas Gleixner
2023-06-22 21:18   ` Frederic Weisbecker
2023-06-06 14:37 ` [patch 09/45] posix-cpu-timers: Fix posix_cpu_timer_get() behaviour Thomas Gleixner
2023-06-26 22:46   ` Frederic Weisbecker
2023-06-29 18:14     ` Thomas Gleixner
2023-06-06 14:37 ` [patch 10/45] posix-cpu-timers: Use @now instead of @val for clarity Thomas Gleixner
2023-06-27  9:53   ` Frederic Weisbecker
2023-06-06 14:37 ` [patch 11/45] posix-cpu-timers: Remove incorrect comment in posix_cpu_timer_set() Thomas Gleixner
2023-06-27 10:30   ` Frederic Weisbecker
2023-06-06 14:37 ` [patch 12/45] posix-cpu-timers: Simplify posix_cpu_timer_set() Thomas Gleixner
2023-06-27 10:51   ` Frederic Weisbecker
2023-06-29 18:43     ` Thomas Gleixner
2023-06-06 14:37 ` [patch 13/45] posix-cpu-timers: Replace old expiry retrieval in posix_cpu_timer_set() Thomas Gleixner
2023-06-27 11:32   ` Frederic Weisbecker
2023-06-06 14:37 ` [patch 14/45] posix-timers: Consolidate interval retrieval Thomas Gleixner
2023-06-28 13:08   ` Frederic Weisbecker
2023-06-29 18:47     ` Thomas Gleixner
2023-06-30 11:25       ` Frederic Weisbecker
2023-06-30 13:07         ` Thomas Gleixner
2023-06-30 14:04           ` Frederic Weisbecker
2023-07-01 18:01             ` Thomas Gleixner
2023-06-06 14:37 ` [patch 15/45] posix-timers: Clear overrun in common_timer_set() Thomas Gleixner
2023-06-30 21:40   ` Frederic Weisbecker
2023-06-06 14:37 ` [patch 16/45] posix-timers: Consolidate timer setup Thomas Gleixner
2023-07-03 21:12   ` Frederic Weisbecker
2023-06-06 14:37 ` [patch 17/45] posix-cpu-timers: Make k_itimer::it_active consistent Thomas Gleixner
2023-07-03 22:30   ` Frederic Weisbecker
2023-06-06 14:37 ` [patch 18/45] posix-timers: Consolidate signal queueing Thomas Gleixner
2023-07-03 22:51   ` Frederic Weisbecker
2023-06-06 14:37 ` [patch 19/45] signal: Remove task argument from dequeue_signal() Thomas Gleixner
2023-07-04 10:02   ` Frederic Weisbecker
2023-06-06 14:37 ` [patch 20/45] signal: Replace BUG_ON()s Thomas Gleixner
2023-07-04 10:24   ` Frederic Weisbecker
2023-06-06 14:37 ` [patch 21/45] signal: Confine POSIX_TIMERS properly Thomas Gleixner
2023-06-06 14:37 ` [patch 22/45] signal: Get rid of resched_timer logic Thomas Gleixner
2023-06-06 14:37 ` [patch 23/45] posix-timers: Cure si_sys_private race Thomas Gleixner
2023-06-06 14:37 ` [patch 24/45] signal: Allow POSIX timer signals to be dropped Thomas Gleixner
2023-06-06 14:37 ` [patch 25/45] posix-timers: Drop signal if timer has been deleted or reprogrammed Thomas Gleixner
2023-06-06 14:38 ` [patch 26/45] posix-timers: Rename k_itimer::it_requeue_pending Thomas Gleixner
2023-06-06 14:38 ` [patch 27/45] posix-timers: Add proper state tracking Thomas Gleixner
2023-06-06 14:38 ` [patch 28/45] posix-timers: Make signal delivery consistent Thomas Gleixner
2023-06-06 14:38 ` Thomas Gleixner [this message]
2023-06-06 14:38 ` [patch 30/45] posix-cpu-timers: Use dedicated flag for CPU timer nanosleep Thomas Gleixner
2023-06-06 14:38 ` [patch 31/45] posix-timers: Add a refcount to struct k_itimer Thomas Gleixner
2023-06-06 14:38 ` [patch 32/45] signal: Split up __sigqueue_alloc() Thomas Gleixner
2023-06-06 14:38 ` [patch 33/45] signal: Provide posixtimer_sigqueue_init() Thomas Gleixner
2023-06-06 14:38 ` [patch 34/45] signal: Add sys_private_ptr to siginfo::_sifields::_timer Thomas Gleixner
2023-06-06 14:38 ` [patch 35/45] signal: Refactor send_sigqueue() Thomas Gleixner
2023-06-06 14:38 ` [patch 36/45] posix-timers: Embed sigqueue in struct k_itimer Thomas Gleixner
2023-06-06 14:38 ` [patch 37/45] signal: Cleanup unused posix-timer leftovers Thomas Gleixner
2023-06-06 14:38 ` [patch 38/45] signal: Add task argument to flush_sigqueue_mask() Thomas Gleixner
2023-06-06 14:38 ` [patch 39/45] signal: Provide ignored_posix_timers list Thomas Gleixner
2023-06-06 14:38 ` [patch 40/45] posix-timers: Handle ignored list on delete and exit Thomas Gleixner
2023-06-06 14:38 ` [patch 41/45] signal: Handle ignored signals in do_sigaction(action != SIG_IGN) Thomas Gleixner
2023-06-06 14:38 ` [patch 42/45] signal: Queue ignored posixtimers on ignore list Thomas Gleixner
2023-06-06 14:38 ` [patch 43/45] posix-timers: Cleanup SIG_IGN workaround leftovers Thomas Gleixner
2023-06-06 14:38 ` [patch 44/45] alarmtimers: Remove the throttle mechanism from alarm_forward_now() Thomas Gleixner
2023-06-06 14:38 ` [patch 45/45] alarmtimers: Remove return value from alarm functions Thomas Gleixner

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20230606142032.655376396@linutronix.de \
    --to=tglx@linutronix.de \
    --cc=anna-maria@linutronix.de \
    --cc=ebiederm@xmission.com \
    --cc=frederic@kernel.org \
    --cc=jstultz@google.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@kernel.org \
    --cc=oleg@redhat.com \
    --cc=peterz@infradead.org \
    --cc=sboyd@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.