From: Frederic Weisbecker <frederic@kernel.org>
To: Anna-Maria Behnsen <anna-maria@linutronix.de>
Cc: linux-kernel@vger.kernel.org,
Peter Zijlstra <peterz@infradead.org>,
John Stultz <jstultz@google.com>,
Thomas Gleixner <tglx@linutronix.de>,
Eric Dumazet <edumazet@google.com>,
"Rafael J . Wysocki" <rafael.j.wysocki@intel.com>,
Arjan van de Ven <arjan@infradead.org>,
"Paul E . McKenney" <paulmck@kernel.org>,
Rik van Riel <riel@surriel.com>,
Steven Rostedt <rostedt@goodmis.org>,
Sebastian Siewior <bigeasy@linutronix.de>,
Giovanni Gherdovich <ggherdovich@suse.cz>,
Lukasz Luba <lukasz.luba@arm.com>,
"Gautham R . Shenoy" <gautham.shenoy@amd.com>,
Srinivas Pandruvada <srinivas.pandruvada@intel.com>,
K Prateek Nayak <kprateek.nayak@amd.com>
Subject: Re: [PATCH v10 18/20] timers: Implement the hierarchical pull model
Date: Fri, 26 Jan 2024 13:53:13 +0100 [thread overview]
Message-ID: <ZbOrOV8kWUd59h9q@lothringen> (raw)
In-Reply-To: <20240115143743.27827-19-anna-maria@linutronix.de>
On Mon, Jan 15, 2024 at 03:37:41PM +0100, Anna-Maria Behnsen wrote:
> + * Protection of the tmigr group state information:
> + * ------------------------------------------------
> + *
> + * The state information with the list of active children and migrator needs to
> + * be protected by a sequence counter. It prevents a race when updates in child
> + * groups are propagated in changed order. The state update is performed
> + * lockless and group wise. The following scenario describes what happens
> + * without updating the sequence counter:
> + *
> + * Therefore, let's take three groups and four CPUs (CPU2 and CPU3 as well
> + * as GRP0:1 will not change during the scenario):
> + *
> + * LVL 1 [GRP1:0]
> + * migrator = GRP0:1
> + * active = GRP0:0, GRP0:1
> + * / \
> + * LVL 0 [GRP0:0] [GRP0:1]
> + * migrator = CPU0 migrator = CPU2
> + * active = CPU0 active = CPU2
> + * / \ / \
> + * CPUs 0 1 2 3
> + * active idle active idle
> + *
> + *
> + * 1. CPU0 goes idle. As the update is performed group wise, in the first step
> + * only GRP0:0 is updated. The update of GRP1:0 is pending as CPU0 has to
> + * walk the hierarchy.
> + *
> + * LVL 1 [GRP1:0]
> + * migrator = GRP0:1
> + * active = GRP0:0, GRP0:1
> + * / \
> + * LVL 0 [GRP0:0] [GRP0:1]
> + * --> migrator = TMIGR_NONE migrator = CPU2
> + * --> active = active = CPU2
> + * / \ / \
> + * CPUs 0 1 2 3
> + * --> idle idle active idle
> + *
> + * 2. While CPU0 goes idle and continues to update the state, CPU1 comes out of
> + * idle. CPU1 updates GRP0:0. The update for GRP1:0 is pending as CPU1 also
> + * has to the hierarchy. Both CPUs (CPU0 and CPU1) now walk the hierarchy to
> + * perform the needed update from their point of view. The currently visible
> + * state looks the following:
> + *
> + * LVL 1 [GRP1:0]
> + * migrator = GRP0:1
> + * active = GRP0:0, GRP0:1
> + * / \
> + * LVL 0 [GRP0:0] [GRP0:1]
> + * --> migrator = CPU1 migrator = CPU2
> + * --> active = CPU1 active = CPU2
> + * / \ / \
> + * CPUs 0 1 2 3
> + * idle --> active active idle
> + *
> + * 3. Here is the race condition: CPU1 managed to propagate its changes (from
> + * step 2) through the hierarchy to GRP1:0 before CPU0 (step 1) did. The
> + * active members of GRP1:0 remain unchanged after the update since it is
> + * still valid from CPU1 current point of view:
> + *
> + * LVL 1 [GRP1:0]
> + * --> migrator = GRP0:1
> + * --> active = GRP0:0, GRP0:1
> + * / \
> + * LVL 0 [GRP0:0] [GRP0:1]
> + * migrator = CPU1 migrator = CPU2
> + * active = CPU1 active = CPU2
> + * / \ / \
> + * CPUs 0 1 2 3
> + * idle active active idle
So let's take this scenario and suppose we are at this stage. CPU 1
has propagated the state up to [GRP1:0] and CPU 0 is going to do it
but hasn't yet.
> +static bool tmigr_inactive_up(struct tmigr_group *group,
> + struct tmigr_group *child,
> + void *ptr)
> +{
> + union tmigr_state curstate, newstate, childstate;
> + struct tmigr_walk *data = ptr;
> + bool walk_done;
> + u8 childmask;
> +
> + childmask = data->childmask;
> + curstate.state = atomic_read(&group->migr_state);
And now suppose CPU 0 arrives here and sees the group->migr_state change
performed by CPU 1. So it's all good, right? The below atomic_cmpxchg()
will success on the first take.
> + childstate.state = 0;
> +
> + do {
> + if (child)
> + childstate.state = atomic_read(&child->migr_state);
But then how do you guarantee that CPU 0 will load here the version of
child->migr_state modified by CPU 1? What prevents from loading the stale value?
The one that was modified by CPU 0 instead? Nothing because the two above reads
are unordered. As a result, CPU 0 may ignore the fact that CPU 1 is up and
wrongly report GRP0:0 as active up to GRP1:0.
One way to solve this is to change the above atomic_read(&group->migr_state)
into atomic_read_acquire(&group->migr_state). It's cheap and pairs with the
order enforced by the upwards successful cmpxchg calls.
> +
> + newstate = curstate;
> + walk_done = true;
> +
> + /* Reset active bit when the child is no longer active */
> + if (!childstate.active)
> + newstate.active &= ~childmask;
> +
> + if (newstate.migrator == childmask) {
> + /*
> + * Find a new migrator for the group, because the child
> + * group is idle!
> + */
> + if (!childstate.active) {
> + unsigned long new_migr_bit, active = newstate.active;
> +
> + new_migr_bit = find_first_bit(&active, BIT_CNT);
> +
> + if (new_migr_bit != BIT_CNT) {
> + newstate.migrator = BIT(new_migr_bit);
> + } else {
> + newstate.migrator = TMIGR_NONE;
> +
> + /* Changes need to be propagated */
> + walk_done = false;
> + }
> + }
> + }
> +
> + newstate.seq++;
> +
> + WARN_ON_ONCE((newstate.migrator != TMIGR_NONE) && !(newstate.active));
> +
> + } while (!atomic_try_cmpxchg(&group->migr_state, &curstate.state, newstate.state));
Similarly, I seem to remember that a failing cmpxchg() doesn't imply a full
memory barrier. If it's the case, you may need to reload &group->migr_state
using an acquire barrier. But lemme check that...
Thanks.
next prev parent reply other threads:[~2024-01-26 12:53 UTC|newest]
Thread overview: 89+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-01-15 14:37 [PATCH v10 00/20] timers: Move from a push remote at enqueue to a pull at expiry model Anna-Maria Behnsen
2024-01-15 14:37 ` [PATCH v10 01/20] timers: Restructure get_next_timer_interrupt() Anna-Maria Behnsen
2024-01-17 15:01 ` Frederic Weisbecker
2024-01-15 14:37 ` [PATCH v10 02/20] timers: Split out get next timer interrupt Anna-Maria Behnsen
2024-01-17 15:06 ` Frederic Weisbecker
2024-01-15 14:37 ` [PATCH v10 03/20] timers: Move marking timer bases idle into tick_nohz_stop_tick() Anna-Maria Behnsen
2024-01-17 16:02 ` Frederic Weisbecker
2024-01-22 11:45 ` Anna-Maria Behnsen
2024-01-22 21:49 ` Frederic Weisbecker
2024-02-19 8:52 ` [PATCH v10a] " Anna-Maria Behnsen
2024-02-19 22:37 ` Frederic Weisbecker
2024-02-20 10:48 ` Anna-Maria Behnsen
2024-02-20 11:41 ` Frederic Weisbecker
2024-02-20 12:02 ` Anna-Maria Behnsen
2024-02-20 12:34 ` Frederic Weisbecker
2024-02-20 14:00 ` Anna-Maria Behnsen
2024-02-20 15:10 ` Frederic Weisbecker
2024-02-20 15:23 ` Anna-Maria Behnsen
2024-02-20 15:25 ` Frederic Weisbecker
2024-01-15 14:37 ` [PATCH v10 04/20] timers: Optimization for timer_base_try_to_set_idle() Anna-Maria Behnsen
2024-01-17 16:45 ` Frederic Weisbecker
2024-01-22 11:48 ` Anna-Maria Behnsen
2024-01-22 22:22 ` Frederic Weisbecker
2024-01-15 14:37 ` [PATCH v10 05/20] timers: Introduce add_timer() variants which modify timer flags Anna-Maria Behnsen
2024-01-17 17:01 ` Frederic Weisbecker
2024-01-22 11:50 ` Anna-Maria Behnsen
2024-01-15 14:37 ` [PATCH v10 06/20] workqueue: Use global variant for add_timer() Anna-Maria Behnsen
2024-01-15 14:37 ` [PATCH v10 07/20] timers: add_timer_on(): Make sure TIMER_PINNED flag is set Anna-Maria Behnsen
2024-01-15 14:37 ` [PATCH v10 08/20] timers: Ease code in run_local_timers() Anna-Maria Behnsen
2024-01-15 14:37 ` [PATCH v10 09/20] timers: Split next timer interrupt logic Anna-Maria Behnsen
2024-01-23 14:28 ` Frederic Weisbecker
2024-01-15 14:37 ` [PATCH v10 10/20] timers: Keep the pinned timers separate from the others Anna-Maria Behnsen
2024-01-15 14:37 ` [PATCH v10 11/20] timers: Retrieve next expiry of pinned/non-pinned timers separately Anna-Maria Behnsen
2024-01-15 14:37 ` [PATCH v10 12/20] timers: Split out "get next timer interrupt" functionality Anna-Maria Behnsen
2024-01-15 14:37 ` [PATCH v10 13/20] timers: Add get next timer interrupt functionality for remote CPUs Anna-Maria Behnsen
2024-02-19 16:04 ` Frederic Weisbecker
2024-02-19 16:57 ` Anna-Maria Behnsen
2024-01-15 14:37 ` [PATCH v10 14/20] timers: Restructure internal locking Anna-Maria Behnsen
2024-01-24 13:56 ` Frederic Weisbecker
2024-01-15 14:37 ` [PATCH v10 15/20] timers: Check if timers base is handled already Anna-Maria Behnsen
2024-01-24 14:22 ` Frederic Weisbecker
2024-01-15 14:37 ` [PATCH v10 16/20] tick/sched: Split out jiffies update helper function Anna-Maria Behnsen
2024-01-24 14:42 ` Frederic Weisbecker
2024-01-15 14:37 ` [PATCH v10 17/20] timers: Introduce function to check timer base is_idle flag Anna-Maria Behnsen
2024-01-24 14:52 ` Frederic Weisbecker
2024-01-15 14:37 ` [PATCH v10 18/20] timers: Implement the hierarchical pull model Anna-Maria Behnsen
2024-01-25 14:30 ` Frederic Weisbecker
2024-01-28 15:58 ` Anna-Maria Behnsen
2024-01-30 15:29 ` Frederic Weisbecker
2024-01-30 16:45 ` Anna-Maria Behnsen
2024-01-26 12:53 ` Frederic Weisbecker [this message]
2024-01-27 22:54 ` Frederic Weisbecker
2024-01-29 10:50 ` Anna-Maria Behnsen
2024-01-29 22:21 ` Frederic Weisbecker
2024-01-30 13:32 ` Anna-Maria Behnsen
2024-01-29 13:50 ` Paul E. McKenney
2024-01-29 1:04 ` Frederic Weisbecker
2024-01-30 17:56 ` Anna-Maria Behnsen
2024-01-30 21:13 ` Frederic Weisbecker
2024-01-31 11:19 ` Anna-Maria Behnsen
2024-01-30 15:37 ` Frederic Weisbecker
2024-02-01 14:59 ` Anna-Maria Behnsen
2024-02-01 15:05 ` Frederic Weisbecker
2024-02-01 16:15 ` Anna-Maria Behnsen
2024-02-01 17:43 ` Frederic Weisbecker
2024-02-01 20:52 ` Anna-Maria Behnsen
2024-02-05 13:29 ` Anna-Maria Behnsen
2024-02-05 20:30 ` Frederic Weisbecker
2024-02-06 10:06 ` Anna-Maria Behnsen
2024-02-06 10:29 ` Frederic Weisbecker
2024-02-01 16:33 ` Frederic Weisbecker
2024-02-05 15:59 ` Anna-Maria Behnsen
2024-02-05 20:28 ` Frederic Weisbecker
2024-02-04 22:02 ` Frederic Weisbecker
2024-02-06 11:03 ` Anna-Maria Behnsen
2024-02-06 11:11 ` Frederic Weisbecker
2024-02-04 22:32 ` Frederic Weisbecker
2024-02-06 11:36 ` Anna-Maria Behnsen
2024-02-06 13:21 ` Frederic Weisbecker
2024-02-06 14:13 ` Anna-Maria Behnsen
2024-02-06 14:21 ` Frederic Weisbecker
2024-01-15 14:37 ` [PATCH v10 19/20] timer_migration: Add tracepoints Anna-Maria Behnsen
2024-02-01 16:47 ` Frederic Weisbecker
2024-01-15 14:37 ` [PATCH v10 20/20] timers: Always queue timers on the local CPU Anna-Maria Behnsen
2024-02-01 17:36 ` Frederic Weisbecker
2024-02-01 20:58 ` Anna-Maria Behnsen
2024-02-02 11:57 ` Frederic Weisbecker
2024-01-30 22:07 ` [PATCH v10 00/20] timers: Move from a push remote at enqueue to a pull at expiry model Christian Loehle
2024-02-01 15:03 ` Anna-Maria Behnsen
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=ZbOrOV8kWUd59h9q@lothringen \
--to=frederic@kernel.org \
--cc=anna-maria@linutronix.de \
--cc=arjan@infradead.org \
--cc=bigeasy@linutronix.de \
--cc=edumazet@google.com \
--cc=gautham.shenoy@amd.com \
--cc=ggherdovich@suse.cz \
--cc=jstultz@google.com \
--cc=kprateek.nayak@amd.com \
--cc=linux-kernel@vger.kernel.org \
--cc=lukasz.luba@arm.com \
--cc=paulmck@kernel.org \
--cc=peterz@infradead.org \
--cc=rafael.j.wysocki@intel.com \
--cc=riel@surriel.com \
--cc=rostedt@goodmis.org \
--cc=srinivas.pandruvada@intel.com \
--cc=tglx@linutronix.de \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).