[PATCH] cgroup: pretty format of /proc/cgoups stats showing

[PATCH] cgroup: pretty format of /proc/cgoups stats showing

[chai wen]

Just make /proc/cgroups have a better appearance, the original is a little
misleading as its print format.

Before:
#subsys_name	hierarchy	num_cgroups	enabled
cpuset	0	1	1
cpu	0	1	1
cpuacct	0	1	1
devices	0	1	1
freezer	0	1	1
net_cls	0	1	1
blkio	0	1	1
perf_event	0	1	1

After:
#subsys_name       hierarchy          num_groups         enabled           
cpuset             0                  1                  1                 
cpu                0                  1                  1                 
cpuacct            0                  1                  1                 
devices            0                  1                  1                 
freezer            0                  1                  1                 
net_cls            0                  1                  1                 
blkio              0                  1                  1                 
perf_event         0                  1                  1                 


Signed-off-by: chai wen <chaiw.fnst-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org>
---
 kernel/cgroup.c |    5 +++--
 1 files changed, 3 insertions(+), 2 deletions(-)

diff --git a/kernel/cgroup.c b/kernel/cgroup.c
index 7868fc3..54c839e 100644
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -4883,7 +4883,8 @@ static int proc_cgroupstats_show(struct seq_file *m, void *v)
 	struct cgroup_subsys *ss;
 	int i;
 
-	seq_puts(m, "#subsys_name\thierarchy\tnum_cgroups\tenabled\n");
+	seq_printf(m, "%-18s %-18s %-18s %-18s\n", "#subsys_name", "hierarchy",
+		   "num_groups", "enabled");
 	/*
 	 * ideally we don't want subsystems moving around while we do this.
 	 * cgroup_mutex is also necessary to guarantee an atomic snapshot of
@@ -4892,7 +4893,7 @@ static int proc_cgroupstats_show(struct seq_file *m, void *v)
 	mutex_lock(&cgroup_mutex);
 
 	for_each_subsys(ss, i)
-		seq_printf(m, "%s\t%d\t%d\t%d\n",
+		seq_printf(m, "%-18s %-18d %-18d %-18d\n",
 			   ss->name, ss->root->hierarchy_id,
 			   atomic_read(&ss->root->nr_cgrps), !ss->disabled);
 
-- 
1.7.1

Re: [PATCH] cgroup: pretty format of /proc/cgoups stats showing

[Li Zefan]

On 2014/7/2 16:24, chai wen wrote:
> Just make /proc/cgroups have a better appearance, the original is a little
> misleading as its print format.
> 
> Before:
> #subsys_name	hierarchy	num_cgroups	enabled
> cpuset	0	1	1
> cpu	0	1	1
> cpuacct	0	1	1
> devices	0	1	1
> freezer	0	1	1
> net_cls	0	1	1
> blkio	0	1	1
> perf_event	0	1	1
> 
> After:
> #subsys_name       hierarchy          num_groups         enabled           
> cpuset             0                  1                  1                 
> cpu                0                  1                  1                 
> cpuacct            0                  1                  1                 
> devices            0                  1                  1                 
> freezer            0                  1                  1                 
> net_cls            0                  1                  1                 
> blkio              0                  1                  1                 
> perf_event         0                  1                  1                 
> 
> 
> Signed-off-by: chai wen <chaiw.fnst-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org>

Someone posted a similar patch long ago, and it was rejected, because
there may be userspace programs/scripts depending on the current format,
and this change will break them.

> ---
>  kernel/cgroup.c |    5 +++--
>  1 files changed, 3 insertions(+), 2 deletions(-)
> 
> diff --git a/kernel/cgroup.c b/kernel/cgroup.c
> index 7868fc3..54c839e 100644
> --- a/kernel/cgroup.c
> +++ b/kernel/cgroup.c
> @@ -4883,7 +4883,8 @@ static int proc_cgroupstats_show(struct seq_file *m, void *v)
>  	struct cgroup_subsys *ss;
>  	int i;
>  
> -	seq_puts(m, "#subsys_name\thierarchy\tnum_cgroups\tenabled\n");
> +	seq_printf(m, "%-18s %-18s %-18s %-18s\n", "#subsys_name", "hierarchy",
> +		   "num_groups", "enabled");
>  	/*
>  	 * ideally we don't want subsystems moving around while we do this.
>  	 * cgroup_mutex is also necessary to guarantee an atomic snapshot of
> @@ -4892,7 +4893,7 @@ static int proc_cgroupstats_show(struct seq_file *m, void *v)
>  	mutex_lock(&cgroup_mutex);
>  
>  	for_each_subsys(ss, i)
> -		seq_printf(m, "%s\t%d\t%d\t%d\n",
> +		seq_printf(m, "%-18s %-18d %-18d %-18d\n",
>  			   ss->name, ss->root->hierarchy_id,
>  			   atomic_read(&ss->root->nr_cgrps), !ss->disabled);
>  
> 

Re: [PATCH] cgroup: pretty format of /proc/cgoups stats showing

[Chai Wen]

On 07/02/2014 04:47 PM, Li Zefan wrote:

> On 2014/7/2 16:24, chai wen wrote:
>> Just make /proc/cgroups have a better appearance, the original is a little
>> misleading as its print format.
>>
>> Before:
>> #subsys_name	hierarchy	num_cgroups	enabled
>> cpuset	0	1	1
>> cpu	0	1	1
>> cpuacct	0	1	1
>> devices	0	1	1
>> freezer	0	1	1
>> net_cls	0	1	1
>> blkio	0	1	1
>> perf_event	0	1	1
>>
>> After:
>> #subsys_name       hierarchy          num_groups         enabled           
>> cpuset             0                  1                  1                 
>> cpu                0                  1                  1                 
>> cpuacct            0                  1                  1                 
>> devices            0                  1                  1                 
>> freezer            0                  1                  1                 
>> net_cls            0                  1                  1                 
>> blkio              0                  1                  1                 
>> perf_event         0                  1                  1                 
>>
>>
>> Signed-off-by: chai wen <chaiw.fnst-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org>
> 
> Someone posted a similar patch long ago, and it was rejected, because
> there may be userspace programs/scripts depending on the current format,
> and this change will break them.


OK, it's reasonable at that point.

thanks
chai wen

> 
>> ---
>>  kernel/cgroup.c |    5 +++--
>>  1 files changed, 3 insertions(+), 2 deletions(-)
>>
>> diff --git a/kernel/cgroup.c b/kernel/cgroup.c
>> index 7868fc3..54c839e 100644
>> --- a/kernel/cgroup.c
>> +++ b/kernel/cgroup.c
>> @@ -4883,7 +4883,8 @@ static int proc_cgroupstats_show(struct seq_file *m, void *v)
>>  	struct cgroup_subsys *ss;
>>  	int i;
>>  
>> -	seq_puts(m, "#subsys_name\thierarchy\tnum_cgroups\tenabled\n");
>> +	seq_printf(m, "%-18s %-18s %-18s %-18s\n", "#subsys_name", "hierarchy",
>> +		   "num_groups", "enabled");
>>  	/*
>>  	 * ideally we don't want subsystems moving around while we do this.
>>  	 * cgroup_mutex is also necessary to guarantee an atomic snapshot of
>> @@ -4892,7 +4893,7 @@ static int proc_cgroupstats_show(struct seq_file *m, void *v)
>>  	mutex_lock(&cgroup_mutex);
>>  
>>  	for_each_subsys(ss, i)
>> -		seq_printf(m, "%s\t%d\t%d\t%d\n",
>> +		seq_printf(m, "%-18s %-18d %-18d %-18d\n",
>>  			   ss->name, ss->root->hierarchy_id,
>>  			   atomic_read(&ss->root->nr_cgrps), !ss->disabled);
>>  
>>
> 
> .
> 



-- 
Regards

Chai Wen