[PATCH 3/3] memcgoup: allow memory.failcnt to be reset

[PATCH 3/3] memcgoup: allow memory.failcnt to be reset

[Li Zefan]

Allow memory.failcnt to be reset to 0:

        echo 0 > memory.failcnt

And '0' is the only valid value.

This is useful when testing or observing the memory resource
controller. Without this function, one will have to remember
the previous failcnt to decide whether memory reclaim has
happened *again*.

Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
---
 Documentation/controllers/memory.txt |    4 +++-
 mm/memcontrol.c                      |   15 +++++++++++++++
 2 files changed, 18 insertions(+), 1 deletions(-)

diff --git a/Documentation/controllers/memory.txt b/Documentation/controllers/memory.txt
index 866b9cd..28f80e3 100644
--- a/Documentation/controllers/memory.txt
+++ b/Documentation/controllers/memory.txt
@@ -194,7 +194,9 @@ this file after a write to guarantee the value committed by the kernel.
 4096
 
 The memory.failcnt field gives the number of times that the cgroup limit was
-exceeded.
+exceeded. It can be reset.
+
+# echo 0 > memory.failcnt
 
 The memory.stat file gives accounting information. Now, the number of
 caches, RSS and Active pages/Inactive pages are shown.
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 6145031..fd26dc2 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -883,6 +883,20 @@ static int mem_force_empty_write(struct cgroup *cont, struct cftype *cft,
 	return ret;
 }
 
+static int mem_failcnt_write(struct cgroup *cont, struct cftype *cft,
+			     u64 val)
+{
+	struct res_counter *counter;
+
+	if (val != 0)
+		return -EINVAL;
+
+	counter = &mem_cgroup_from_cont(cont)->res;
+	res_counter_write_u64(counter, cft->private, 0);
+
+	return 0;
+}
+
 static const struct mem_cgroup_stat_desc {
 	const char *msg;
 	u64 unit;
@@ -934,6 +948,7 @@ static struct cftype mem_cgroup_files[] = {
 	{
 		.name = "failcnt",
 		.private = RES_FAILCNT,
+		.write_u64 = mem_failcnt_write,
 		.read_u64 = mem_cgroup_read,
 	},
 	{
-- 
1.5.4.rc3

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH 3/3] memcgoup: allow memory.failcnt to be reset

[KAMEZAWA Hiroyuki]

On Tue, 11 Mar 2008 19:09:02 +0900
Li Zefan <lizf@cn.fujitsu.com> wrote:

> Allow memory.failcnt to be reset to 0:
> 
>         echo 0 > memory.failcnt
> 
> And '0' is the only valid value.
> 
Can't this be generic resource counter function ?

Thanks,
-Kame


> This is useful when testing or observing the memory resource
> controller. Without this function, one will have to remember
> the previous failcnt to decide whether memory reclaim has
> happened *again*.
> 
> Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
> ---
>  Documentation/controllers/memory.txt |    4 +++-
>  mm/memcontrol.c                      |   15 +++++++++++++++
>  2 files changed, 18 insertions(+), 1 deletions(-)
> 
> diff --git a/Documentation/controllers/memory.txt b/Documentation/controllers/memory.txt
> index 866b9cd..28f80e3 100644
> --- a/Documentation/controllers/memory.txt
> +++ b/Documentation/controllers/memory.txt
> @@ -194,7 +194,9 @@ this file after a write to guarantee the value committed by the kernel.
>  4096
>  
>  The memory.failcnt field gives the number of times that the cgroup limit was
> -exceeded.
> +exceeded. It can be reset.
> +
> +# echo 0 > memory.failcnt
>  
>  The memory.stat file gives accounting information. Now, the number of
>  caches, RSS and Active pages/Inactive pages are shown.
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index 6145031..fd26dc2 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -883,6 +883,20 @@ static int mem_force_empty_write(struct cgroup *cont, struct cftype *cft,
>  	return ret;
>  }
>  
> +static int mem_failcnt_write(struct cgroup *cont, struct cftype *cft,
> +			     u64 val)
> +{
> +	struct res_counter *counter;
> +
> +	if (val != 0)
> +		return -EINVAL;
> +
> +	counter = &mem_cgroup_from_cont(cont)->res;
> +	res_counter_write_u64(counter, cft->private, 0);
> +
> +	return 0;
> +}
> +
>  static const struct mem_cgroup_stat_desc {
>  	const char *msg;
>  	u64 unit;
> @@ -934,6 +948,7 @@ static struct cftype mem_cgroup_files[] = {
>  	{
>  		.name = "failcnt",
>  		.private = RES_FAILCNT,
> +		.write_u64 = mem_failcnt_write,
>  		.read_u64 = mem_cgroup_read,
>  	},
>  	{
> -- 
> 1.5.4.rc3
> 

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH 3/3] memcgoup: allow memory.failcnt to be reset

[Balbir Singh]

KAMEZAWA Hiroyuki wrote:
> On Tue, 11 Mar 2008 19:09:02 +0900
> Li Zefan <lizf@cn.fujitsu.com> wrote:
> 
>> Allow memory.failcnt to be reset to 0:
>>
>>         echo 0 > memory.failcnt
>>
>> And '0' is the only valid value.
>>
> Can't this be generic resource counter function ?
> 

I was about to suggest a generic cgroup option, since we do reset values even
for the cpu accounting subsystem.

-- 
	Warm Regards,
	Balbir Singh
	Linux Technology Center
	IBM, ISTL

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH 3/3] memcgoup: allow memory.failcnt to be reset

[Li Zefan]

Balbir Singh wrote:
> KAMEZAWA Hiroyuki wrote:
>> On Tue, 11 Mar 2008 19:09:02 +0900
>> Li Zefan <lizf@cn.fujitsu.com> wrote:
>>
>>> Allow memory.failcnt to be reset to 0:
>>>
>>>         echo 0 > memory.failcnt
>>>
>>> And '0' is the only valid value.
>>>
>> Can't this be generic resource counter function ?
>>
> 
> I was about to suggest a generic cgroup option, since we do reset values even
> for the cpu accounting subsystem.
> 

It won't help. You still have to write the write function, and you have to call
some res_counter routines to reset the value, and maybe also do some other
work.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>