Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Fri, 25 Oct 2019 10:10:37 +0200
From:   Peter Zijlstra <peterz@infradead.org>
To:     Davidlohr Bueso <dave@stgolabs.net>
Cc:     mingo@kernel.org, linux-kernel@vger.kernel.org,
        Davidlohr Bueso <dbueso@suse.de>
Subject: Re: [PATCH -tip] locking/mutex: Complain upon api misuse wrt
 interrupt context
Message-ID: <20191025081037.GF4131@hirez.programming.kicks-ass.net>
References: <20191025033634.3330-1-dave@stgolabs.net>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20191025033634.3330-1-dave@stgolabs.net>
User-Agent: Mutt/1.10.1 (2018-07-13)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Thu, Oct 24, 2019 at 08:36:34PM -0700, Davidlohr Bueso wrote:
> Sprinkle warning checks, under CONFIG_DEBUG_MUTEXES. While the mutex
> rules and semantics are explicitly documented, this allows to expose
> any abusers and robustifies the whole thing. While trylock and unlock
> are non-blocking, calling from irq context is still forbidden (lock
> must be within the same context as unlock).
> 
> Signed-off-by: Davidlohr Bueso <dbueso@suse.de>
> ---
>  kernel/locking/mutex.c | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/kernel/locking/mutex.c b/kernel/locking/mutex.c
> index 468a9b8422e3..7e9c316c646c 100644
> --- a/kernel/locking/mutex.c
> +++ b/kernel/locking/mutex.c
> @@ -733,6 +733,9 @@ static noinline void __sched __mutex_unlock_slowpath(struct mutex *lock, unsigne
>   */
>  void __sched mutex_unlock(struct mutex *lock)
>  {
> +#ifdef CONFIG_DEBUG_MUTEXES
> +	WARN_ON(in_interrupt());
> +#endif
>  #ifndef CONFIG_DEBUG_LOCK_ALLOC
>  	if (__mutex_unlock_fast(lock))
>  		return;
> @@ -1413,6 +1416,7 @@ int __sched mutex_trylock(struct mutex *lock)
>  
>  #ifdef CONFIG_DEBUG_MUTEXES
>  	DEBUG_LOCKS_WARN_ON(lock->magic != lock);
> +	WARN_ON(in_interrupt());
>  #endif
>  
>  	locked = __mutex_trylock(lock);


No real objection, but should not lockdep already complain about this?
__mutex_unlock_slowpath() takes ->wait_lock irq-unsafe, so then using it
from an IRQ should generate an insta IRQ inversion report.