Re: [PATCH v2] powerpc/watchdog: Use hrtimers for per-CPU heartbeat

2019-05-03 Thread Michael Ellerman
On Tue, 2019-04-09 at 04:40:05 UTC, Nicholas Piggin wrote: > Using a jiffies timer creates a dependency on the tick_do_timer_cpu > incrementing jiffies. If that CPU has locked up and jiffies is not > incrementing, the watchdog heartbeat timer for all CPUs stops and > creates false positives and

Re: [PATCH v2] powerpc/watchdog: Use hrtimers for per-CPU heartbeat

2019-04-12 Thread Gautham R Shenoy
Hello Nicholas, On Tue, Apr 09, 2019 at 02:40:05PM +1000, Nicholas Piggin wrote: > Using a jiffies timer creates a dependency on the tick_do_timer_cpu > incrementing jiffies. If that CPU has locked up and jiffies is not > incrementing, the watchdog heartbeat timer for all CPUs stops and >

Re: [PATCH v2] powerpc/watchdog: Use hrtimers for per-CPU heartbeat

2019-04-10 Thread Ravi Bangoria
On 4/9/19 10:10 AM, Nicholas Piggin wrote: > Using a jiffies timer creates a dependency on the tick_do_timer_cpu > incrementing jiffies. If that CPU has locked up and jiffies is not > incrementing, the watchdog heartbeat timer for all CPUs stops and > creates false positives and confusing

[PATCH v2] powerpc/watchdog: Use hrtimers for per-CPU heartbeat

2019-04-08 Thread Nicholas Piggin
Using a jiffies timer creates a dependency on the tick_do_timer_cpu incrementing jiffies. If that CPU has locked up and jiffies is not incrementing, the watchdog heartbeat timer for all CPUs stops and creates false positives and confusing warnings on local CPUs, and also causes the SMP detector to