linux-kernel - Re: [PATCH RFC v0 2/3] res_counter: implement thresholds

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <20091127092035.bbf2efdc.nishimura@mxp.nes.nec.co.jp>
Date:	Fri, 27 Nov 2009 09:20:35 +0900
From:	Daisuke Nishimura <nishimura@....nes.nec.co.jp>
To:	"Kirill A. Shutemov" <kirill@...temov.name>
Cc:	containers@...ts.linux-foundation.org, linux-mm@...ck.org,
	Paul Menage <menage@...gle.com>,
	Li Zefan <lizf@...fujitsu.com>,
	Andrew Morton <akpm@...ux-foundation.org>,
	KAMEZAWA Hiroyuki <kamezawa.hiroyu@...fujitsu.com>,
	Balbir Singh <balbir@...ux.vnet.ibm.com>,
	Pavel Emelyanov <xemul@...nvz.org>,
	linux-kernel@...r.kernel.org,
	Daisuke Nishimura <nishimura@....nes.nec.co.jp>
Subject: Re: [PATCH RFC v0 2/3] res_counter: implement thresholds

Hi.

On Thu, 26 Nov 2009 19:11:16 +0200, "Kirill A. Shutemov" <kirill@...temov.name> wrote:
> It allows to setup two thresholds: one above current usage and one
> below. Callback threshold_notifier() will be called if a threshold is
> crossed.
> 
> Signed-off-by: Kirill A. Shutemov <kirill@...temov.name>
> ---
>  include/linux/res_counter.h |   44 +++++++++++++++++++++++++++++++++++++++++++
>  kernel/res_counter.c        |    4 +++
>  2 files changed, 48 insertions(+), 0 deletions(-)
> 
> diff --git a/include/linux/res_counter.h b/include/linux/res_counter.h
> index fcb9884..bca99a5 100644
> --- a/include/linux/res_counter.h
> +++ b/include/linux/res_counter.h
> @@ -9,6 +9,10 @@
>   *
>   * Author: Pavel Emelianov <xemul@...nvz.org>
>   *
> + * Thresholds support
> + * Copyright (C) 2009 Nokia Corporation
> + * Author: Kirill A. Shutemov
> + *
>   * See Documentation/cgroups/resource_counter.txt for more
>   * info about what this counter is.
>   */
> @@ -42,6 +46,13 @@ struct res_counter {
>  	 * the number of unsuccessful attempts to consume the resource
>  	 */
>  	unsigned long long failcnt;
> +
> +	unsigned long long threshold_above;
> +	unsigned long long threshold_below;
> +	void (*threshold_notifier)(struct res_counter *counter,
> +			unsigned long long usage,
> +			unsigned long long threshold);
> +
>  	/*
>  	 * the lock to protect all of the above.
>  	 * the routines below consider this to be IRQ-safe
> @@ -145,6 +156,20 @@ static inline bool res_counter_soft_limit_check_locked(struct res_counter *cnt)
>  	return false;
>  }
>  
> +static inline void res_counter_threshold_notify_locked(struct res_counter *cnt)
> +{
> +	if (cnt->usage >= cnt->threshold_above) {
> +		cnt->threshold_notifier(cnt, cnt->usage, cnt->threshold_above);
> +		return;
> +	}
> +
> +	if (cnt->usage < cnt->threshold_below) {
> +		cnt->threshold_notifier(cnt, cnt->usage, cnt->threshold_below);
> +		return;
> +	}
> +}
> +
> +
>  /**
>   * Get the difference between the usage and the soft limit
>   * @cnt: The counter
> @@ -238,4 +263,23 @@ res_counter_set_soft_limit(struct res_counter *cnt,
>  	return 0;
>  }
>  
> +static inline int
> +res_counter_set_thresholds(struct res_counter *cnt,
> +		unsigned long long threshold_above,
> +		unsigned long long threshold_below)
> +{
> +	unsigned long flags;
> +	int ret = -EINVAL;
> +
> +	spin_lock_irqsave(&cnt->lock, flags);
> +	if ((cnt->usage < threshold_above) &&
> +			(cnt->usage >= threshold_below)) {
> +		cnt->threshold_above = threshold_above;
> +		cnt->threshold_below = threshold_below;
> +		ret = 0;
> +	}
> +	spin_unlock_irqrestore(&cnt->lock, flags);
> +	return ret;
> +}
> +
>  #endif
> diff --git a/kernel/res_counter.c b/kernel/res_counter.c
> index bcdabf3..646c29c 100644
> --- a/kernel/res_counter.c
> +++ b/kernel/res_counter.c
> @@ -20,6 +20,8 @@ void res_counter_init(struct res_counter *counter, struct res_counter *parent)
>  	spin_lock_init(&counter->lock);
>  	counter->limit = RESOURCE_MAX;
>  	counter->soft_limit = RESOURCE_MAX;
> +	counter->threshold_above = RESOURCE_MAX;
> +	counter->threshold_below = 0ULL;
>  	counter->parent = parent;
>  }
>  
> @@ -33,6 +35,7 @@ int res_counter_charge_locked(struct res_counter *counter, unsigned long val)
>  	counter->usage += val;
>  	if (counter->usage > counter->max_usage)
>  		counter->max_usage = counter->usage;
> +	res_counter_threshold_notify_locked(counter);
>  	return 0;
>  }
>  
> @@ -73,6 +76,7 @@ void res_counter_uncharge_locked(struct res_counter *counter, unsigned long val)
>  		val = counter->usage;
>  
>  	counter->usage -= val;
> +	res_counter_threshold_notify_locked(counter);
>  }
>  
hmm.. this adds new checks to hot-path of process life cycle.

Do you have any number on performance impact of these patches(w/o setting any threshold)?
IMHO, it might be small enough to be ignored because KAMEZAWA-san's coalesce charge/uncharge
patches have decreased charge/uncharge for res_counter itself, but I want to know just to make sure.


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