Re: [PATCH 3.4 098/107] kernel/watchdog.c: touch_nmi_watchdog should only touch local cpu not every one

[Don Zickus]

On Wed, Mar 16, 2016 at 04:06:32PM +0800, lizf@xxxxxxxxxx wrote:
> From: Ben Zhang <benzh@xxxxxxxxxxxx>
> 
> 3.4.111-rc1 review patch.  If anyone has any objections, please let me know.

Just an FYI below, this patch won't work the way it was integrated..

comments below

> 
> ------------------
> 
> 
> commit 62572e29bc530b38921ef6059088b4788a9832a5 upstream.
> 
> I ran into a scenario where while one cpu was stuck and should have
> panic'd because of the NMI watchdog, it didn't.  The reason was another
> cpu was spewing stack dumps on to the console.  Upon investigation, I
> noticed that when writing to the console and also when dumping the
> stack, the watchdog is touched.
> 
> This causes all the cpus to reset their NMI watchdog flags and the
> 'stuck' cpu just spins forever.
> 
> This change causes the semantics of touch_nmi_watchdog to be changed
> slightly.  Previously, I accidentally changed the semantics and we
> noticed there was a codepath in which touch_nmi_watchdog could be
> touched from a preemtible area.  That caused a BUG() to happen when
> CONFIG_DEBUG_PREEMPT was enabled.  I believe it was the acpi code.
> 
> My attempt here re-introduces the change to have the
> touch_nmi_watchdog() code only touch the local cpu instead of all of the
> cpus.  But instead of using __get_cpu_var(), I use the
> __raw_get_cpu_var() version.
> 
> This avoids the preemption problem.  However my reasoning wasn't because
> I was trying to be lazy.  Instead I rationalized it as, well if
> preemption is enabled then interrupts should be enabled to and the NMI
> watchdog will have no reason to trigger.  So it won't matter if the
> wrong cpu is touched because the percpu interrupt counters the NMI
> watchdog uses should still be incrementing.
> 
> Don said:
> 
> : I'm ok with this patch, though it does alter the behaviour of how
> : touch_nmi_watchdog works.  For the most part I don't think most callers
> : need to touch all of the watchdogs (on each cpu).  Perhaps a corner case
> : will pop up (the scheduler??  to mimic touch_all_softlockup_watchdogs() ).
> :
> : But this does address an issue where if a system is locked up and one cpu
> : is spewing out useful debug messages (or error messages), the hard lockup
> : will fail to go off.  We have seen this on RHEL also.
> 
> Signed-off-by: Don Zickus <dzickus@xxxxxxxxxx>
> Signed-off-by: Ben Zhang <benzh@xxxxxxxxxxxx>
> Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> Signed-off-by: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
> [lizf: Backported to 3.4: adjust context]
> Signed-off-by: Zefan Li <lizefan@xxxxxxxxxx>
> ---
>  kernel/watchdog.c | 8 ++++++++
>  1 file changed, 8 insertions(+)
> 
> diff --git a/kernel/watchdog.c b/kernel/watchdog.c
> index 991aa93..7527c8c 100644
> --- a/kernel/watchdog.c
> +++ b/kernel/watchdog.c
> @@ -162,6 +162,14 @@ void touch_nmi_watchdog(void)
>  				per_cpu(watchdog_nmi_touch, cpu) = true;
>  		}
>  	}

The above for-loop was to be replaced by the non-for-loop below.

The above for-loop is the problem this patch was solving, so keeping it
around does not solve anything.  :-)


> +	/*
> +	 * Using __raw here because some code paths have
> +	 * preemption enabled.  If preemption is enabled
> +	 * then interrupts should be enabled too, in which
> +	 * case we shouldn't have to worry about the watchdog
> +	 * going off.
> +	 */
> +	__raw_get_cpu_var(watchdog_nmi_touch) = true;
>  	touch_softlockup_watchdog();
>  }
>  EXPORT_SYMBOL(touch_nmi_watchdog);

Cheers,
Don