From: Douglas Anderson <dianders@chromium.org> To: Petr Mladek <pmladek@suse.com>, Andrew Morton <akpm@linux-foundation.org> Cc: kgdb-bugreport@lists.sourceforge.net, linux-kernel@vger.kernel.org, Nicholas Piggin <npiggin@gmail.com>, Michael Ellerman <mpe@ellerman.id.au>, linuxppc-dev@lists.ozlabs.org, Christophe Leroy <christophe.leroy@csgroup.eu>, sparclinux@vger.kernel.org, "David S . Miller" <davem@davemloft.net>, linux-perf-users@vger.kernel.org, Douglas Anderson <dianders@chromium.org> Subject: [PATCH 09/10] watchdog/hardlockup: Move SMP barriers from common code to buddy code Date: Fri, 26 May 2023 18:41:39 -0700 [thread overview] Message-ID: <20230526184139.9.I5ab0a0eeb0bd52fb23f901d298c72fa5c396e22b@changeid> (raw) In-Reply-To: <20230527014153.2793931-1-dianders@chromium.org> It's been suggested that since the SMP barriers are only potentially useful for the buddy hardlockup detector, not the perf hardlockup detector, that the barriers belong in the buddy code. Let's move them and add clearer comments about why they're needed. Suggested-by: Petr Mladek <pmladek@suse.com> Signed-off-by: Douglas Anderson <dianders@chromium.org> --- kernel/watchdog.c | 6 ------ kernel/watchdog_buddy.c | 21 +++++++++++++++++++++ 2 files changed, 21 insertions(+), 6 deletions(-) diff --git a/kernel/watchdog.c b/kernel/watchdog.c index 6cc46b8e3d07..a351ab0c35eb 100644 --- a/kernel/watchdog.c +++ b/kernel/watchdog.c @@ -109,9 +109,6 @@ EXPORT_SYMBOL(arch_touch_nmi_watchdog); void watchdog_hardlockup_touch_cpu(unsigned int cpu) { per_cpu(watchdog_hardlockup_touched, cpu) = true; - - /* Match with smp_rmb() in watchdog_hardlockup_check() */ - smp_wmb(); } static bool is_hardlockup(unsigned int cpu) @@ -141,9 +138,6 @@ static void watchdog_hardlockup_kick(void) void watchdog_hardlockup_check(unsigned int cpu, struct pt_regs *regs) { - /* Match with smp_wmb() in watchdog_hardlockup_touch_cpu() */ - smp_rmb(); - if (per_cpu(watchdog_hardlockup_touched, cpu)) { per_cpu(watchdog_hardlockup_touched, cpu) = false; return; diff --git a/kernel/watchdog_buddy.c b/kernel/watchdog_buddy.c index 2ef88722c5e7..34dbfe091f4b 100644 --- a/kernel/watchdog_buddy.c +++ b/kernel/watchdog_buddy.c @@ -51,6 +51,13 @@ void watchdog_hardlockup_enable(unsigned int cpu) if (next_cpu < nr_cpu_ids) watchdog_hardlockup_touch_cpu(next_cpu); + /* + * Makes sure that watchdog is touched on this CPU before + * other CPUs could see it in watchdog_cpus. The counter + * part is in watchdog_buddy_check_hardlockup(). + */ + smp_wmb(); + cpumask_set_cpu(cpu, &watchdog_cpus); } @@ -68,6 +75,13 @@ void watchdog_hardlockup_disable(unsigned int cpu) if (next_cpu < nr_cpu_ids) watchdog_hardlockup_touch_cpu(next_cpu); + /* + * Makes sure that watchdog is touched on the next CPU before + * this CPU disappear in watchdog_cpus. The counter part is in + * watchdog_buddy_check_hardlockup(). + */ + smp_wmb(); + cpumask_clear_cpu(cpu, &watchdog_cpus); } @@ -88,5 +102,12 @@ void watchdog_buddy_check_hardlockup(int hrtimer_interrupts) if (next_cpu >= nr_cpu_ids) return; + /* + * Make sure that the watchdog was touched on next CPU when + * watchdog_next_cpu() returned another one because of + * a change in watchdog_hardlockup_enable()/disable(). + */ + smp_rmb(); + watchdog_hardlockup_check(next_cpu, NULL); } -- 2.41.0.rc0.172.g3f132b7071-goog
WARNING: multiple messages have this Message-ID (diff)
From: Douglas Anderson <dianders@chromium.org> To: Petr Mladek <pmladek@suse.com>, Andrew Morton <akpm@linux-foundation.org> Cc: kgdb-bugreport@lists.sourceforge.net, linux-kernel@vger.kernel.org, Douglas Anderson <dianders@chromium.org>, linux-perf-users@vger.kernel.org, Nicholas Piggin <npiggin@gmail.com>, sparclinux@vger.kernel.org, linuxppc-dev@lists.ozlabs.org, "David S . Miller" <davem@davemloft.net> Subject: [PATCH 09/10] watchdog/hardlockup: Move SMP barriers from common code to buddy code Date: Fri, 26 May 2023 18:41:39 -0700 [thread overview] Message-ID: <20230526184139.9.I5ab0a0eeb0bd52fb23f901d298c72fa5c396e22b@changeid> (raw) In-Reply-To: <20230527014153.2793931-1-dianders@chromium.org> It's been suggested that since the SMP barriers are only potentially useful for the buddy hardlockup detector, not the perf hardlockup detector, that the barriers belong in the buddy code. Let's move them and add clearer comments about why they're needed. Suggested-by: Petr Mladek <pmladek@suse.com> Signed-off-by: Douglas Anderson <dianders@chromium.org> --- kernel/watchdog.c | 6 ------ kernel/watchdog_buddy.c | 21 +++++++++++++++++++++ 2 files changed, 21 insertions(+), 6 deletions(-) diff --git a/kernel/watchdog.c b/kernel/watchdog.c index 6cc46b8e3d07..a351ab0c35eb 100644 --- a/kernel/watchdog.c +++ b/kernel/watchdog.c @@ -109,9 +109,6 @@ EXPORT_SYMBOL(arch_touch_nmi_watchdog); void watchdog_hardlockup_touch_cpu(unsigned int cpu) { per_cpu(watchdog_hardlockup_touched, cpu) = true; - - /* Match with smp_rmb() in watchdog_hardlockup_check() */ - smp_wmb(); } static bool is_hardlockup(unsigned int cpu) @@ -141,9 +138,6 @@ static void watchdog_hardlockup_kick(void) void watchdog_hardlockup_check(unsigned int cpu, struct pt_regs *regs) { - /* Match with smp_wmb() in watchdog_hardlockup_touch_cpu() */ - smp_rmb(); - if (per_cpu(watchdog_hardlockup_touched, cpu)) { per_cpu(watchdog_hardlockup_touched, cpu) = false; return; diff --git a/kernel/watchdog_buddy.c b/kernel/watchdog_buddy.c index 2ef88722c5e7..34dbfe091f4b 100644 --- a/kernel/watchdog_buddy.c +++ b/kernel/watchdog_buddy.c @@ -51,6 +51,13 @@ void watchdog_hardlockup_enable(unsigned int cpu) if (next_cpu < nr_cpu_ids) watchdog_hardlockup_touch_cpu(next_cpu); + /* + * Makes sure that watchdog is touched on this CPU before + * other CPUs could see it in watchdog_cpus. The counter + * part is in watchdog_buddy_check_hardlockup(). + */ + smp_wmb(); + cpumask_set_cpu(cpu, &watchdog_cpus); } @@ -68,6 +75,13 @@ void watchdog_hardlockup_disable(unsigned int cpu) if (next_cpu < nr_cpu_ids) watchdog_hardlockup_touch_cpu(next_cpu); + /* + * Makes sure that watchdog is touched on the next CPU before + * this CPU disappear in watchdog_cpus. The counter part is in + * watchdog_buddy_check_hardlockup(). + */ + smp_wmb(); + cpumask_clear_cpu(cpu, &watchdog_cpus); } @@ -88,5 +102,12 @@ void watchdog_buddy_check_hardlockup(int hrtimer_interrupts) if (next_cpu >= nr_cpu_ids) return; + /* + * Make sure that the watchdog was touched on next CPU when + * watchdog_next_cpu() returned another one because of + * a change in watchdog_hardlockup_enable()/disable(). + */ + smp_rmb(); + watchdog_hardlockup_check(next_cpu, NULL); } -- 2.41.0.rc0.172.g3f132b7071-goog
next prev parent reply other threads:[~2023-05-27 1:43 UTC|newest] Thread overview: 42+ messages / expand[flat|nested] mbox.gz Atom feed top 2023-05-27 1:41 [PATCH 00/10] watchdog: Cleanup / fixes after buddy series v5 reviews Douglas Anderson 2023-05-27 1:41 ` Douglas Anderson 2023-05-27 1:41 ` [PATCH 01/10] watchdog/hardlockup: Keep kernel.nmi_watchdog sysctl as 0444 if probe fails Douglas Anderson 2023-05-27 1:41 ` Douglas Anderson 2023-05-30 14:15 ` Petr Mladek 2023-05-30 14:15 ` Petr Mladek 2023-05-27 1:41 ` [PATCH 02/10] watchdog/hardlockup: HAVE_NMI_WATCHDOG must implement watchdog_hardlockup_probe() Douglas Anderson 2023-05-27 1:41 ` Douglas Anderson 2023-05-30 14:38 ` Petr Mladek 2023-05-30 14:38 ` Petr Mladek 2023-05-27 1:41 ` [PATCH 03/10] watchdog/hardlockup: Don't use raw_cpu_ptr() in watchdog_hardlockup_kick() Douglas Anderson 2023-05-27 1:41 ` Douglas Anderson 2023-05-30 14:39 ` Petr Mladek 2023-05-30 14:39 ` Petr Mladek 2023-05-27 1:41 ` [PATCH 04/10] watchdog/hardlockup: In watchdog_hardlockup_check() use cpumask_copy() Douglas Anderson 2023-05-27 1:41 ` Douglas Anderson 2023-05-30 14:40 ` Petr Mladek 2023-05-30 14:40 ` Petr Mladek 2023-05-27 1:41 ` [PATCH 05/10] watchdog/hardlockup: remove softlockup comment in touch_nmi_watchdog() Douglas Anderson 2023-05-27 1:41 ` Douglas Anderson 2023-05-30 14:42 ` Petr Mladek 2023-05-30 14:42 ` Petr Mladek 2023-05-27 1:41 ` [PATCH 06/10] watchdog/buddy: Cleanup how watchdog_buddy_check_hardlockup() is called Douglas Anderson 2023-05-27 1:41 ` Douglas Anderson 2023-05-30 14:56 ` Petr Mladek 2023-05-30 14:56 ` Petr Mladek 2023-05-27 1:41 ` [PATCH 07/10] watchdog/buddy: Don't copy the cpumask in watchdog_next_cpu() Douglas Anderson 2023-05-27 1:41 ` Douglas Anderson 2023-05-30 14:57 ` Petr Mladek 2023-05-30 14:57 ` Petr Mladek 2023-05-27 1:41 ` [PATCH 08/10] watchdog/buddy: Simplify the dependency for HARDLOCKUP_DETECTOR_PREFER_BUDDY Douglas Anderson 2023-05-27 1:41 ` Douglas Anderson 2023-05-30 14:58 ` Petr Mladek 2023-05-30 14:58 ` Petr Mladek 2023-05-27 1:41 ` Douglas Anderson [this message] 2023-05-27 1:41 ` [PATCH 09/10] watchdog/hardlockup: Move SMP barriers from common code to buddy code Douglas Anderson 2023-05-30 15:00 ` Petr Mladek 2023-05-30 15:00 ` Petr Mladek 2023-05-27 1:41 ` [PATCH 10/10] watchdog/hardlockup: Rename HAVE_HARDLOCKUP_DETECTOR_NON_ARCH to ..._PERF_OR_BUDDY Douglas Anderson 2023-05-27 1:41 ` Douglas Anderson 2023-06-01 13:03 ` Petr Mladek 2023-06-01 13:03 ` Petr Mladek
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=20230526184139.9.I5ab0a0eeb0bd52fb23f901d298c72fa5c396e22b@changeid \ --to=dianders@chromium.org \ --cc=akpm@linux-foundation.org \ --cc=christophe.leroy@csgroup.eu \ --cc=davem@davemloft.net \ --cc=kgdb-bugreport@lists.sourceforge.net \ --cc=linux-kernel@vger.kernel.org \ --cc=linux-perf-users@vger.kernel.org \ --cc=linuxppc-dev@lists.ozlabs.org \ --cc=mpe@ellerman.id.au \ --cc=npiggin@gmail.com \ --cc=pmladek@suse.com \ --cc=sparclinux@vger.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: linkBe 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.