From: akpm@linux-foundation.org Subject: [merged] watchdog-softlockup-enforce-that-timestamp-is-valid-on-boot.patch removed from -mm tree Date: Tue, 21 Jan 2020 12:32:31 -0800 Message-ID: <20200121203231.DBU7xMkDb%akpm@linux-foundation.org> Reply-To: linux-kernel@vger.kernel.org Return-path: Received: from mail.kernel.org ([198.145.29.99]:43918 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727383AbgAUUce (ORCPT ); Tue, 21 Jan 2020 15:32:34 -0500 Sender: mm-commits-owner@vger.kernel.org List-Id: mm-commits@vger.kernel.org To: catalin.marinas@arm.com, keescook@chromium.org, longman@redhat.com, mingo@kernel.org, mm-commits@vger.kernel.org, peterz@infradead.org, pmladek@suse.com, rppt@linux.ibm.com, rrichter@marvell.com, tglx@linutronix.de, will@kernel.org The patch titled Subject: watchdog/softlockup: enforce that timestamp is valid on boot has been removed from the -mm tree. Its filename was watchdog-softlockup-enforce-that-timestamp-is-valid-on-boot.patch This patch was dropped because it was merged into mainline or a subsystem tree ------------------------------------------------------ From: Thomas Gleixner Subject: watchdog/softlockup: enforce that timestamp is valid on boot Robert reported that during boot the watchdog timestamp is set to 0 for one second which is the indicator for a watchdog reset. The reason for this is that the timestamp is in seconds and the time is taken from sched clock and divided by ~1e9. sched clock starts at 0 which means that for the first second during boot the watchdog timestamp is 0, i.e. reset. Use ULONG_MAX as the reset indicator value so the watchdog works correctly right from the start. ULONG_MAX would only conflict with a real timestamp if the system reaches an uptime of 136 years on 32bit and almost eternity on 64bit. Link: http://lkml.kernel.org/r/87o8v3uuzl.fsf@nanos.tec.linutronix.de Signed-off-by: Thomas Gleixner Reported-by: Robert Richter Cc: Waiman Long Cc: Ingo Molnar Cc: Mike Rapoport Cc: Kees Cook Cc: Catalin Marinas Cc: Will Deacon Cc: Peter Zijlstra Cc: Petr Mladek Signed-off-by: Andrew Morton --- kernel/watchdog.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) --- a/kernel/watchdog.c~watchdog-softlockup-enforce-that-timestamp-is-valid-on-boot +++ a/kernel/watchdog.c @@ -161,6 +161,8 @@ static void lockup_detector_update_enabl #ifdef CONFIG_SOFTLOCKUP_DETECTOR +#define SOFTLOCKUP_RESET ULONG_MAX + /* Global variables, exported for sysctl */ unsigned int __read_mostly softlockup_panic = CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC_VALUE; @@ -274,7 +276,7 @@ notrace void touch_softlockup_watchdog_s * Preemption can be enabled. It doesn't matter which CPU's timestamp * gets zeroed here, so use the raw_ operation. */ - raw_cpu_write(watchdog_touch_ts, 0); + raw_cpu_write(watchdog_touch_ts, SOFTLOCKUP_RESET); } notrace void touch_softlockup_watchdog(void) @@ -298,14 +300,14 @@ void touch_all_softlockup_watchdogs(void * the softlockup check. */ for_each_cpu(cpu, &watchdog_allowed_mask) - per_cpu(watchdog_touch_ts, cpu) = 0; + per_cpu(watchdog_touch_ts, cpu) = SOFTLOCKUP_RESET; wq_watchdog_touch(-1); } void touch_softlockup_watchdog_sync(void) { __this_cpu_write(softlockup_touch_sync, true); - __this_cpu_write(watchdog_touch_ts, 0); + __this_cpu_write(watchdog_touch_ts, SOFTLOCKUP_RESET); } static int is_softlockup(unsigned long touch_ts) @@ -383,7 +385,7 @@ static enum hrtimer_restart watchdog_tim /* .. and repeat */ hrtimer_forward_now(hrtimer, ns_to_ktime(sample_period)); - if (touch_ts == 0) { + if (touch_ts == SOFTLOCKUP_RESET) { if (unlikely(__this_cpu_read(softlockup_touch_sync))) { /* * If the time stamp was touched atomically _ Patches currently in -mm which might be from tglx@linutronix.de are