All of lore.kernel.org
 help / color / mirror / Atom feed
From: Li Zefan <lizf@cn.fujitsu.com>
To: jacob.jun.pan@linux.intel.com
Cc: LKML <linux-kernel@vger.kernel.org>,
	"Kirill A. Shutemov" <kirill@shutemov.name>,
	Arjan van de Ven <arjan@linux.intel.com>,
	Matt Helsley <matthltc@us.ibm.com>,
	container cgroup <containers@lists.linux-foundation.org>,
	Paul Menage <menage@google.com>,
	akpm@linux-foundation.org, rdunlap@xenotime.net,
	Cedric Le Goater <clg@fr.ibm.com>
Subject: Re: [PATCH 1/1, v6] cgroup/freezer: add per freezer duty ratio control
Date: Wed, 09 Feb 2011 11:07:59 +0800	[thread overview]
Message-ID: <4D52050F.3060907@cn.fujitsu.com> (raw)
In-Reply-To: <1297213541-18156-1-git-send-email-jacob.jun.pan@linux.intel.com>

jacob.jun.pan@linux.intel.com wrote:
> From: Jacob Pan <jacob.jun.pan@linux.intel.com>
> 
> Freezer subsystem is used to manage batch jobs which can start
> stop at the same time. However, sometime it is desirable to let
> the kernel manage the freezer state automatically with a given
> duty ratio.
> For example, if we want to reduce the time that backgroup apps
> are allowed to run we can put them into a freezer subsystem and
> set the kernel to turn them THAWED/FROZEN at given duty ratio.
> 
> This patch introduces two file nodes under cgroup
> freezer.duty_ratio_pct and freezer.period_sec
> 
> Usage example: set period to be 5 seconds and frozen duty ratio 90%
> [root@localhost aoa]# echo 90 > freezer.duty_ratio_pct
> [root@localhost aoa]# echo 5000 > freezer.period_ms
> 

So after:

# echo FROZEN > freezer.state
# echo 90 > freezer.duty_ratio_pct
# echo 5000 > freezer.period_ms
...
# echo 0 > freezer.duty_ratio_pct

All the tasks in this cgroup are in THAWED state, but the cgroup is
in FROZEN state. This should be fixed.

> Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
> ---
>  Documentation/cgroups/freezer-subsystem.txt |   23 +++++
>  kernel/cgroup_freezer.c                     |  132 ++++++++++++++++++++++++++-
>  2 files changed, 154 insertions(+), 1 deletions(-)
> 
> diff --git a/Documentation/cgroups/freezer-subsystem.txt b/Documentation/cgroups/freezer-subsystem.txt
> index 41f37fe..7f06f05 100644
> --- a/Documentation/cgroups/freezer-subsystem.txt
> +++ b/Documentation/cgroups/freezer-subsystem.txt
> @@ -100,3 +100,26 @@ things happens:
>  		and returns EINVAL)
>  	3) The tasks that blocked the cgroup from entering the "FROZEN"
>  		state disappear from the cgroup's set of tasks.
> +
> +In embedded systems, it is desirable to manage group of applications
> +for power saving. E.g. tasks that are not in the foreground may be
> +frozen unfrozen periodically to save power without affecting user
> +experience. In this case, user/management software can attach tasks
> +into freezer cgroup then specify duty ratio and period that the
> +managed tasks are allowed to run.
> +
> +Usage example:
> +Assuming freezer cgroup is already mounted, application being managed
> +are included the "tasks" file node of the given freezer cgroup.
> +To make the tasks frozen at 90% of the time every 5 seconds, do:
> +
> +[root@localhost ]# echo 90 > freezer.duty_ratio_pct
> +[root@localhost ]# echo 5000 > freezer.period_ms
> +
> +After that, the application in this freezer cgroup will only be
> +allowed to run at the following pattern.
> +     __                    __                    __
> +    |  |<-- 90% frozen -->|  |                  |  |
> +____|  |__________________|  |__________________|  |_____
> +
> +    |<---- 5 seconds ---->|
> diff --git a/kernel/cgroup_freezer.c b/kernel/cgroup_freezer.c
> index e7bebb7..5808f28 100644
> --- a/kernel/cgroup_freezer.c
> +++ b/kernel/cgroup_freezer.c
> @@ -21,6 +21,7 @@
>  #include <linux/uaccess.h>
>  #include <linux/freezer.h>
>  #include <linux/seq_file.h>
> +#include <linux/kthread.h>
>  
>  enum freezer_state {
>  	CGROUP_THAWED = 0,
> @@ -28,12 +29,28 @@ enum freezer_state {
>  	CGROUP_FROZEN,
>  };
>  
> +enum duty_ratio_params {
> +	FREEZER_DUTY_RATIO = 0,
> +	FREEZER_PERIOD,
> +};
> +
> +struct freezer_duty {
> +	u32 ratio; /* percentage of time frozen */
> +	u32 period_pct_ms; /* one percent of the period in miliseconds */
> +};
> +
>  struct freezer {
>  	struct cgroup_subsys_state css;
>  	enum freezer_state state;
> +	struct freezer_duty duty;
> +	struct task_struct *fkh;
>  	spinlock_t lock; /* protects _writes_ to state */
>  };
>  
> +static struct task_struct *freezer_task;
> +static int try_to_freeze_cgroup(struct cgroup *cgroup, struct freezer *freezer);
> +static void unfreeze_cgroup(struct cgroup *cgroup, struct freezer *freezer);
> +
>  static inline struct freezer *cgroup_freezer(
>  		struct cgroup *cgroup)
>  {
> @@ -63,6 +80,31 @@ int cgroup_freezing_or_frozen(struct task_struct *task)
>  	return result;
>  }
>  
> +static DECLARE_WAIT_QUEUE_HEAD(freezer_wait);
> +
> +static int freezer_kh(void *data)
> +{
> +	struct cgroup *cgroup = (struct cgroup *)data;
> +	struct freezer *freezer = cgroup_freezer(cgroup);
> +
> +	do {
> +		if (freezer->duty.ratio < 100 && freezer->duty.ratio > 0 &&
> +			freezer->duty.period_pct_ms) {
> +			if (try_to_freeze_cgroup(cgroup, freezer))
> +				pr_info("cannot freeze\n");
> +			msleep(freezer->duty.period_pct_ms *
> +				freezer->duty.ratio);
> +			unfreeze_cgroup(cgroup, freezer);
> +			msleep(freezer->duty.period_pct_ms *
> +				(100 - freezer->duty.ratio));
> +		} else {
> +			sleep_on(&freezer_wait);
> +			pr_debug("freezer thread wake up\n");
> +		}
> +	} while (!kthread_should_stop());
> +	return 0;
> +}
> +
>  /*
>   * cgroups_write_string() limits the size of freezer state strings to
>   * CGROUP_LOCAL_BUFFER_SIZE
> @@ -150,7 +192,12 @@ static struct cgroup_subsys_state *freezer_create(struct cgroup_subsys *ss,
>  static void freezer_destroy(struct cgroup_subsys *ss,
>  			    struct cgroup *cgroup)
>  {
> -	kfree(cgroup_freezer(cgroup));
> +	struct freezer *freezer;
> +
> +	freezer = cgroup_freezer(cgroup);
> +	if (freezer->fkh)
> +		kthread_stop(freezer->fkh);
> +	kfree(freezer);
>  }
>  
>  /*
> @@ -282,6 +329,16 @@ static int freezer_read(struct cgroup *cgroup, struct cftype *cft,
>  	return 0;
>  }
>  
> +static u64 freezer_read_duty_ratio(struct cgroup *cgroup, struct cftype *cft)
> +{
> +	return cgroup_freezer(cgroup)->duty.ratio;
> +}
> +
> +static u64 freezer_read_period(struct cgroup *cgroup, struct cftype *cft)
> +{
> +	return cgroup_freezer(cgroup)->duty.period_pct_ms * 100;
> +}
> +
>  static int try_to_freeze_cgroup(struct cgroup *cgroup, struct freezer *freezer)
>  {
>  	struct cgroup_iter it;
> @@ -368,12 +425,85 @@ static int freezer_write(struct cgroup *cgroup,
>  	return retval;
>  }
>  
> +#define FREEZER_KH_PREFIX  "freezer_"
> +static int freezer_write_param(struct cgroup *cgroup, struct cftype *cft,
> +		u64 val)
> +{
> +	struct freezer *freezer;
> +	char thread_name[32];
> +	int ret = 0;
> +
> +	freezer = cgroup_freezer(cgroup);
> +
> +	if (!cgroup_lock_live_group(cgroup))
> +		return -ENODEV;
> +
> +	switch (cft->private) {
> +	case FREEZER_DUTY_RATIO:
> +		if (val >= 100 || val < 0) {

val will never < 0.

> +			ret = -EINVAL;
> +			goto exit;
> +		}
> +		freezer->duty.ratio = val;
> +		break;
> +	case FREEZER_PERIOD:
> +		if (val)
> +			do_div(val, 100);

Is 0 an invalid value for do_div()?

> +			freezer->duty.period_pct_ms = val;
> +		break;
> +	default:
> +		BUG();
> +	}
> +
> +	/* start/stop management kthread as needed, the rule is that
> +	 * if both duty ratio and period values are zero, then no management
> +	 * kthread is created. when both are non-zero, we create a kthread
> +	 * for the cgroup. When user set zero to duty ratio and period again
> +	 * the kthread is stopped.
> +	 */
> +	if (freezer->duty.ratio && freezer->duty.period_pct_ms) {
> +		if (!freezer->fkh) {
> +			snprintf(thread_name, 32, "%s%s", FREEZER_KH_PREFIX,
> +				cgroup->dentry->d_name.name);
> +			freezer->fkh = kthread_run(freezer_kh, (void *)cgroup,
> +						thread_name);
> +			if (IS_ERR(freezer_task)) {

You mean IS_ERR(freezer->fkh)?

> +				pr_err("create %s failed\n", thread_name);
> +				ret = PTR_ERR(freezer_task);
> +				goto exit;
> +			}
> +		} else
> +			wake_up(&freezer_wait);
> +	} else if ((!freezer->duty.ratio || !freezer->duty.period_pct_ms) &&
> +		freezer->fkh) {
> +			kthread_stop(freezer->fkh);
> +			freezer->fkh = NULL;
> +	}
> +
> +exit:
> +	cgroup_unlock();
> +	return ret;
> +}
> +
>  static struct cftype files[] = {
>  	{
>  		.name = "state",
>  		.read_seq_string = freezer_read,
>  		.write_string = freezer_write,
>  	},
> +	{
> +		.name = "duty_ratio_pct",
> +		.private = FREEZER_DUTY_RATIO,
> +		.read_u64 = freezer_read_duty_ratio,
> +		.write_u64 = freezer_write_param,
> +	},
> +	{
> +		.name = "period_ms",
> +		.private = FREEZER_PERIOD,
> +		.read_u64 = freezer_read_period,
> +		.write_u64 = freezer_write_param,
> +	},
> +
>  };
>  
>  static int freezer_populate(struct cgroup_subsys *ss, struct cgroup *cgroup)

  reply	other threads:[~2011-02-09  3:06 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-02-09  1:05 [PATCH 1/1, v6] cgroup/freezer: add per freezer duty ratio control jacob.jun.pan
2011-02-09  3:07 ` Li Zefan [this message]
2011-02-09 18:16   ` jacob pan
2011-02-10  1:26     ` Li Zefan
2011-02-10  1:26     ` Li Zefan
2011-02-10  4:43       ` jacob pan
     [not found]       ` <4D533EB0.6060405-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org>
2011-02-10  4:43         ` jacob pan
     [not found]   ` <4D52050F.3060907-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org>
2011-02-09 18:16     ` jacob pan
2011-02-10  4:51     ` jacob pan
2011-02-10  4:51   ` jacob pan
2011-02-10  3:04 ` Matt Helsley
2011-02-10  3:06   ` Arjan van de Ven
2011-02-10 19:11     ` Matt Helsley
2011-02-10 22:22       ` jacob pan
2011-02-10 22:43         ` Matt Helsley
2011-02-10 22:43         ` Matt Helsley
     [not found]       ` <20110210191117.GI16432-52DBMbEzqgQ/wnmkkaCWp/UQ3DHhIser@public.gmane.org>
2011-02-10 22:22         ` jacob pan
     [not found]     ` <4D535627.9090606-VuQAYsv1563Yd54FQh9/CA@public.gmane.org>
2011-02-10 19:11       ` Matt Helsley
2011-02-14 18:03       ` Vaidyanathan Srinivasan
2011-02-14 18:03     ` Vaidyanathan Srinivasan
2011-02-10  9:15   ` Kirill A. Shutemov
2011-02-10 18:58     ` Matt Helsley
     [not found]     ` <20110210091522.GA28511-oKw7cIdHH8eLwutG50LtGA@public.gmane.org>
2011-02-10 18:58       ` Matt Helsley
2011-02-10 23:06   ` Jacob Pan
     [not found]   ` <20110210030442.GG16432-52DBMbEzqgQ/wnmkkaCWp/UQ3DHhIser@public.gmane.org>
2011-02-10  3:06     ` Arjan van de Ven
2011-02-10  9:15     ` Kirill A. Shutemov
2011-02-10 23:06     ` Jacob Pan
     [not found] ` <1297213541-18156-1-git-send-email-jacob.jun.pan-VuQAYsv1563Yd54FQh9/CA@public.gmane.org>
2011-02-09  3:07   ` Li Zefan
2011-02-10  3:04   ` Matt Helsley
2011-02-09  1:05 jacob.jun.pan-VuQAYsv1563Yd54FQh9/CA

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=4D52050F.3060907@cn.fujitsu.com \
    --to=lizf@cn.fujitsu.com \
    --cc=akpm@linux-foundation.org \
    --cc=arjan@linux.intel.com \
    --cc=clg@fr.ibm.com \
    --cc=containers@lists.linux-foundation.org \
    --cc=jacob.jun.pan@linux.intel.com \
    --cc=kirill@shutemov.name \
    --cc=linux-kernel@vger.kernel.org \
    --cc=matthltc@us.ibm.com \
    --cc=menage@google.com \
    --cc=rdunlap@xenotime.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.