Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Fri, 20 Nov 2020 15:36:24 +0100
From:   Peter Zijlstra <peterz@infradead.org>
To:     Waiman Long <longman@redhat.com>
Cc:     Ingo Molnar <mingo@redhat.com>, Will Deacon <will@kernel.org>,
        linux-kernel@vger.kernel.org, Davidlohr Bueso <dave@stgolabs.net>,
        Phil Auld <pauld@redhat.com>
Subject: Re: [PATCH 3/5] locking/rwsem: Enable reader optimistic lock stealing
Message-ID: <20201120143624.GD3040@hirez.programming.kicks-ass.net>
References: <20201118030429.23017-1-longman@redhat.com>
 <20201118030429.23017-4-longman@redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20201118030429.23017-4-longman@redhat.com>
Precedence: bulk

On Tue, Nov 17, 2020 at 10:04:27PM -0500, Waiman Long wrote:
> diff --git a/kernel/locking/rwsem.c b/kernel/locking/rwsem.c
> index ee374ae061c3..930dd4af3639 100644
> --- a/kernel/locking/rwsem.c
> +++ b/kernel/locking/rwsem.c
> @@ -957,6 +957,12 @@ static inline bool rwsem_reader_phase_trylock(struct rw_semaphore *sem,
>  	}
>  	return false;
>  }
> +
> +static inline bool osq_is_empty(struct rw_semaphore *sem)
> +{
> +	return !osq_is_locked(&sem->osq);
> +}
> +
>  #else
>  static inline bool rwsem_can_spin_on_owner(struct rw_semaphore *sem,
>  					   unsigned long nonspinnable)
> @@ -977,6 +983,10 @@ static inline bool rwsem_reader_phase_trylock(struct rw_semaphore *sem,
>  	return false;
>  }
>  
> +static inline bool osq_is_empty(sem)
> +{
> +	return false;
> +}

Hurph, the naming seems to suggest this ought to be in osq_lock.h, but
it really is part of rwsem, it captures the lack of osq member for this
configuration.

How about: rwsem_no_spinners() instead ?

>  static inline int
>  rwsem_spin_on_owner(struct rw_semaphore *sem, unsigned long nonspinnable)
>  {
> @@ -1007,6 +1017,22 @@ rwsem_down_read_slowpath(struct rw_semaphore *sem, int state, long count)
>  	   !(count & RWSEM_WRITER_LOCKED))
>  		goto queue;
>  
> +	/*
> +	 * Reader optimistic lock stealing
> +	 *
> +	 * We can take the read lock directly without doing
> +	 * rwsem_optimistic_spin() if the conditions are right.
> +	 * Also wake up other readers if it is the first reader.
> +	 */
> +	if (!(count & (RWSEM_WRITER_LOCKED | RWSEM_FLAG_HANDOFF)) &&
> +	    osq_is_empty(sem)) {
> +		rwsem_set_reader_owned(sem);
> +		lockevent_inc(rwsem_rlock_steal);
> +		if (rcnt == 1)
> +			goto wake_readers;
> +		return sem;
> +	}

AFAICT this saves at least 3 atomic ops; how common is this case
(you did add a counter but forgot to mention this).