Date: Thu, 3 Dec 2015 16:37:26 +0000
From: Will Deacon <will.deacon@arm.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: mingo@kernel.org, oleg@redhat.com, linux-kernel@vger.kernel.org,
        paulmck@linux.vnet.ibm.com, boqun.feng@gmail.com, corbet@lwn.net,
        mhocko@kernel.org, dhowells@redhat.com, torvalds@linux-foundation.org,
        waiman.long@hpe.com, pjt@google.com
Subject: Re: [PATCH 3/4] locking: Introduce smp_cond_acquire()
Message-ID: <20151203163725.GJ11337@arm.com>
References: <20151203124010.627312076@infradead.org>
 <20151203124339.552838970@infradead.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20151203124339.552838970@infradead.org>
User-Agent: Mutt/1.5.23 (2014-03-12)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 2640
Lines: 72

Hi Peter,

On Thu, Dec 03, 2015 at 01:40:13PM +0100, Peter Zijlstra wrote:
> Introduce smp_cond_acquire() which combines a control dependency and a
> read barrier to form acquire semantics.
> 
> This primitive has two benefits:
>  - it documents control dependencies,
>  - its typically cheaper than using smp_load_acquire() in a loop.
> 
> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
> ---
>  include/linux/compiler.h   |   17 +++++++++++++++++
>  kernel/locking/qspinlock.c |    3 +--
>  kernel/sched/core.c        |    8 +-------
>  kernel/sched/sched.h       |    2 +-
>  4 files changed, 20 insertions(+), 10 deletions(-)
> 
> --- a/include/linux/compiler.h
> +++ b/include/linux/compiler.h
> @@ -299,6 +299,23 @@ static __always_inline void __write_once
>  	__u.__val;					\
>  })
>  
> +/**
> + * smp_cond_acquire() - Spin wait for cond with ACQUIRE ordering
> + * @cond: boolean expression to wait for
> + *
> + * Equivalent to using smp_load_acquire() on the condition variable but employs
> + * the control dependency of the wait to reduce the barrier on many platforms.
> + *
> + * The control dependency provides a LOAD->STORE order, the additional RMB
> + * provides LOAD->LOAD order, together they provide LOAD->{LOAD,STORE} order,
> + * aka. ACQUIRE.
> + */
> +#define smp_cond_acquire(cond)	do {		\
> +	while (!(cond))				\
> +		cpu_relax();			\
> +	smp_rmb(); /* ctrl + rmb := acquire */	\
> +} while (0)
> +
>  #endif /* __KERNEL__ */
>  
>  #endif /* __ASSEMBLY__ */
> --- a/kernel/locking/qspinlock.c
> +++ b/kernel/locking/qspinlock.c
> @@ -446,8 +446,7 @@ void queued_spin_lock_slowpath(struct qs
>  	if ((val = pv_wait_head_or_lock(lock, node)))
>  		goto locked;
>  
> -	while ((val = smp_load_acquire(&lock->val.counter)) & _Q_LOCKED_PENDING_MASK)
> -		cpu_relax();
> +	smp_cond_acquire(!((val = atomic_read(&lock->val)) & _Q_LOCKED_PENDING_MASK));

I think we spoke about this before, but what would work really well for
arm64 here is if we could override smp_cond_acquire in such a way that
the atomic_read could be performed explicitly in the macro. That would
allow us to use an LDXR to set the exclusive monitor, which in turn
means we can issue a WFE and get a cheap wakeup when lock->val is
actually modified.

With the current scheme, there's not enough information expressed in the
"cond" parameter to perform this optimisation.

Cheers,

Will
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/