Date: Mon, 9 Jan 2006 09:20:47 -0800
From: "Paul E. McKenney" <paulmck@us.ibm.com>
To: Oleg Nesterov <oleg@tv-sign.ru>
Cc: linux-kernel@vger.kernel.org, Dipankar Sarma <dipankar@in.ibm.com>,
       Manfred Spraul <manfred@colorfullife.com>,
       Linus Torvalds <torvalds@osdl.org>, Andrew Morton <akpm@osdl.org>
Subject: Re: [PATCH 1/5] rcu: uninline __rcu_pending()
Message-ID: <20060109172047.GA15083@us.ibm.com>
Reply-To: paulmck@us.ibm.com
References: <43C165B4.2FF3B78@tv-sign.ru>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <43C165B4.2FF3B78@tv-sign.ru>
User-Agent: Mutt/1.4.1i
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 3249
Lines: 102

On Sun, Jan 08, 2006 at 10:19:16PM +0300, Oleg Nesterov wrote:
> __rcu_pending() is rather fat and called twice from rcu_pending().
> 
> rcu_pending() has multiple callers, and not that small too.
> 
> This patch uninlines both of them.

Looks good -- I must defer to Dipankar on any performance problems, but
I would not expect any this far off of the fastpaths.  Also makes
handling multiple different RCU implementations a bit easier on the
eyes.

							Thanx, Paul

Acked-by: Paul E. McKenney <paulmck@us.ibm.com>
> Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
> 
> --- 2.6.15/include/linux/rcupdate.h~1_OUTL	2006-01-03 21:15:58.000000000 +0300
> +++ 2.6.15/include/linux/rcupdate.h	2006-01-08 21:36:07.000000000 +0300
> @@ -125,36 +125,7 @@ static inline void rcu_bh_qsctr_inc(int 
>  	rdp->passed_quiesc = 1;
>  }
>  
> -static inline int __rcu_pending(struct rcu_ctrlblk *rcp,
> -						struct rcu_data *rdp)
> -{
> -	/* This cpu has pending rcu entries and the grace period
> -	 * for them has completed.
> -	 */
> -	if (rdp->curlist && !rcu_batch_before(rcp->completed, rdp->batch))
> -		return 1;
> -
> -	/* This cpu has no pending entries, but there are new entries */
> -	if (!rdp->curlist && rdp->nxtlist)
> -		return 1;
> -
> -	/* This cpu has finished callbacks to invoke */
> -	if (rdp->donelist)
> -		return 1;
> -
> -	/* The rcu core waits for a quiescent state from the cpu */
> -	if (rdp->quiescbatch != rcp->cur || rdp->qs_pending)
> -		return 1;
> -
> -	/* nothing to do */
> -	return 0;
> -}
> -
> -static inline int rcu_pending(int cpu)
> -{
> -	return __rcu_pending(&rcu_ctrlblk, &per_cpu(rcu_data, cpu)) ||
> -		__rcu_pending(&rcu_bh_ctrlblk, &per_cpu(rcu_bh_data, cpu));
> -}
> +extern int rcu_pending(int cpu);
>  
>  /**
>   * rcu_read_lock - mark the beginning of an RCU read-side critical section.
> --- 2.6.15/kernel/rcupdate.c~1_OUTL	2006-01-03 21:15:58.000000000 +0300
> +++ 2.6.15/kernel/rcupdate.c	2006-01-08 21:35:21.000000000 +0300
> @@ -442,6 +442,36 @@ static void rcu_process_callbacks(unsign
>  				&__get_cpu_var(rcu_bh_data));
>  }
>  
> +static int __rcu_pending(struct rcu_ctrlblk *rcp, struct rcu_data *rdp)
> +{
> +	/* This cpu has pending rcu entries and the grace period
> +	 * for them has completed.
> +	 */
> +	if (rdp->curlist && !rcu_batch_before(rcp->completed, rdp->batch))
> +		return 1;
> +
> +	/* This cpu has no pending entries, but there are new entries */
> +	if (!rdp->curlist && rdp->nxtlist)
> +		return 1;
> +
> +	/* This cpu has finished callbacks to invoke */
> +	if (rdp->donelist)
> +		return 1;
> +
> +	/* The rcu core waits for a quiescent state from the cpu */
> +	if (rdp->quiescbatch != rcp->cur || rdp->qs_pending)
> +		return 1;
> +
> +	/* nothing to do */
> +	return 0;
> +}
> +
> +int rcu_pending(int cpu)
> +{
> +	return __rcu_pending(&rcu_ctrlblk, &per_cpu(rcu_data, cpu)) ||
> +		__rcu_pending(&rcu_bh_ctrlblk, &per_cpu(rcu_bh_data, cpu));
> +}
> +
>  void rcu_check_callbacks(int cpu, int user)
>  {
>  	if (user ||
> 
-
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/