[Intel-gfx] [PATCH 2/2] RFC: soft/hardlookup: taint kernel

[Laurence Oberman]

On Thu, 2019-05-02 at 21:42 +0200, Daniel Vetter wrote:
> There's the soft/hardlookup_panic sysctls, but that's a bit an
> extreme
> measure. As a fallback taint at least the machine.
> 
> Our CI uses this to decide when a reboot is necessary, plus to figure
> out whether the kernel is still happy.
> 
> Signed-off-by: Daniel Vetter <daniel.vetter at intel.com>
> Cc: Thomas Gleixner <tglx at linutronix.de>
> Cc: Ingo Molnar <mingo at kernel.org>
> Cc: Peter Zijlstra <peterz at infradead.org>
> Cc: Valdis Kletnieks <valdis.kletnieks at vt.edu>
> Cc: Laurence Oberman <loberman at redhat.com>
> Cc: Vincent Whitchurch <vincent.whitchurch at axis.com>
> Cc: Don Zickus <dzickus at redhat.com>
> Cc: Andrew Morton <akpm at linux-foundation.org>
> Cc: Sergey Senozhatsky <sergey.senozhatsky.work at gmail.com>
> Cc: Sinan Kaya <okaya at kernel.org>
> Cc: Daniel Vetter <daniel.vetter at ffwll.ch>
> ---
>  kernel/watchdog.c     | 2 ++
>  kernel/watchdog_hld.c | 2 ++
>  2 files changed, 4 insertions(+)
> 
> diff --git a/kernel/watchdog.c b/kernel/watchdog.c
> index 6a5787233113..de7a60503517 100644
> --- a/kernel/watchdog.c
> +++ b/kernel/watchdog.c
> @@ -469,6 +469,8 @@ static enum hrtimer_restart
> watchdog_timer_fn(struct hrtimer *hrtimer)
>  		add_taint(TAINT_SOFTLOCKUP, LOCKDEP_STILL_OK);
>  		if (softlockup_panic)
>  			panic("softlockup: hung tasks");
> +		else
> +			add_taint(TAINT_WARN, LOCKDEP_STILL_OK);
>  		__this_cpu_write(soft_watchdog_warn, true);
>  	} else
>  		__this_cpu_write(soft_watchdog_warn, false);
> diff --git a/kernel/watchdog_hld.c b/kernel/watchdog_hld.c
> index 247bf0b1582c..cce46cf75d76 100644
> --- a/kernel/watchdog_hld.c
> +++ b/kernel/watchdog_hld.c
> @@ -154,6 +154,8 @@ static void watchdog_overflow_callback(struct
> perf_event *event,
>  
>  		if (hardlockup_panic)
>  			nmi_panic(regs, "Hard LOCKUP");
> +		else
> +			add_taint(TAINT_WARN, LOCKDEP_STILL_OK);
>  
>  		__this_cpu_write(hard_watchdog_warn, true);
>  		return;

This looks OK to me, could be useful to know we would have triggered
had the flags been set.

Reviewed-by: Laurence Oberman <loberman at redhat.com>