Re: [PATCH evl] evl/thread: fix lost kick of a thread running on a remote CPU

[Philippe Gerum <rpm@xenomai.org>]

Tobias Schaffner <tobias.schaffner@siemens.com> writes:

> evl_kick_thread() skips the mayday if the kicked thread is current on
> its own runqueue instead of the local one. A thread running out-of-band
> on another CPU is therefore never kicked and keeps running with a
> pending signal.
>
> Check against the local runqueue again.
>
> Fixes: fecf5883ecd3 ("evl/sched: get rid of this_evl_rq_thread()")
> Signed-off-by: Tobias Schaffner <tobias.schaffner@siemens.com>
> ---
>  kernel/evl/thread.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/kernel/evl/thread.c b/kernel/evl/thread.c
> index 2aa07b494d40..94d073a1b7e7 100644
> --- a/kernel/evl/thread.c
> +++ b/kernel/evl/thread.c
> @@ -1323,7 +1323,7 @@ void evl_kick_thread(struct evl_thread *thread)
>  	 * stage demotion took place.
>  	 */
>  	if (thread->state & EVL_T_USER) {
> -		if (rq->curr != thread)
> +		if (this_evl_rq()->curr != thread)
>  			dovetail_send_mayday(p);
>  
>  		if (thread->state & EVL_THREAD_LAZY_MASK)
>
> base-commit: a3ad77d7ba3eac8a73f64026272747c01dd84fc6

Merged, thanks (upstreaming soon).

-- 
Philippe.