[PATCH] cgroup: move a check to parse_cgroupfs_options()

[PATCH] cgroup: move a check to parse_cgroupfs_options()

[Li Zefan]

Always check the validity of mount options in parse_cgroupfs_options().

No functional change.

Signed-off-by: Li Zefan <lizf-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org>
---
 kernel/cgroup.c |   15 +++++++--------
 1 files changed, 7 insertions(+), 8 deletions(-)

diff --git a/kernel/cgroup.c b/kernel/cgroup.c
index 4936d88..bdb7994 100644
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -1218,7 +1218,6 @@ static int parse_cgroupfs_options(char *data, struct cgroup_sb_opts *opts)
 	    (opts->subsys_bits & mask))
 		return -EINVAL;
 
-
 	/* Can't specify "none" and some subsystems */
 	if (opts->subsys_bits && opts->none)
 		return -EINVAL;
@@ -1231,6 +1230,13 @@ static int parse_cgroupfs_options(char *data, struct cgroup_sb_opts *opts)
 		return -EINVAL;
 
 	/*
+	 * This can happen only when none of the subsystems is enabled
+	 * in the system.
+	 */
+	if (!opts->subsys_bits && !opts->none)
+		return -EINVAL;
+
+	/*
 	 * Grab references on all the modules we'll need, so the subsystems
 	 * don't dance around before rebind_subsystems attaches them. This may
 	 * take duplicate reference counts on a subsystem that's already used,
@@ -1401,9 +1407,6 @@ static struct cgroupfs_root *cgroup_root_from_opts(struct cgroup_sb_opts *opts)
 {
 	struct cgroupfs_root *root;
 
-	if (!opts->subsys_bits && !opts->none)
-		return NULL;
-
 	root = kzalloc(sizeof(*root), GFP_KERNEL);
 	if (!root)
 		return ERR_PTR(-ENOMEM);
@@ -1442,10 +1445,6 @@ static int cgroup_set_super(struct super_block *sb, void *data)
 	int ret;
 	struct cgroup_sb_opts *opts = data;
 
-	/* If we don't have a new root, we can't set up a new sb */
-	if (!opts->new_root)
-		return -EINVAL;
-
 	BUG_ON(!opts->subsys_bits && !opts->none);
 
 	ret = set_anon_super(sb, NULL);
-- 
1.7.3.1

Re: [PATCH] cgroup: move a check to parse_cgroupfs_options()

[Tejun Heo]

Hello,

On Thu, Dec 22, 2011 at 04:43:19PM +0800, Li Zefan wrote:
> Always check the validity of mount options in parse_cgroupfs_options().
> 
> No functional change.
> 
> Signed-off-by: Li Zefan <lizf-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org>
> ---
>  kernel/cgroup.c |   15 +++++++--------
>  1 files changed, 7 insertions(+), 8 deletions(-)
> 
> diff --git a/kernel/cgroup.c b/kernel/cgroup.c
> index 4936d88..bdb7994 100644
> --- a/kernel/cgroup.c
> +++ b/kernel/cgroup.c
> @@ -1218,7 +1218,6 @@ static int parse_cgroupfs_options(char *data, struct cgroup_sb_opts *opts)
>  	    (opts->subsys_bits & mask))
>  		return -EINVAL;
>  
> -
>  	/* Can't specify "none" and some subsystems */
>  	if (opts->subsys_bits && opts->none)
>  		return -EINVAL;
> @@ -1231,6 +1230,13 @@ static int parse_cgroupfs_options(char *data, struct cgroup_sb_opts *opts)
>  		return -EINVAL;
>  
>  	/*
> +	 * This can happen only when none of the subsystems is enabled
> +	 * in the system.
> +	 */
> +	if (!opts->subsys_bits && !opts->none)
> +		return -EINVAL;

The three ifs here share a lot and I think sharing the common parts
would make the logic clearer.  e.g. The first and this can be easily
combined to (bool)opts->subsys_bits != (bool)opts->none which would
explain the logic better too.

> @@ -1442,10 +1445,6 @@ static int cgroup_set_super(struct super_block *sb, void *data)
>  	int ret;
>  	struct cgroup_sb_opts *opts = data;
>  
> -	/* If we don't have a new root, we can't set up a new sb */
> -	if (!opts->new_root)
> -		return -EINVAL;
> -

And where did this one go?

Thanks.

-- 
tejun

Re: [PATCH] cgroup: move a check to parse_cgroupfs_options()

[Li Zefan]

于 2011年12月22日 23:20, Tejun Heo 写道:
> Hello,
> 
> On Thu, Dec 22, 2011 at 04:43:19PM +0800, Li Zefan wrote:
>> Always check the validity of mount options in parse_cgroupfs_options().
>>
>> No functional change.
>>
>> Signed-off-by: Li Zefan <lizf-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org>
>> ---
>>  kernel/cgroup.c |   15 +++++++--------
>>  1 files changed, 7 insertions(+), 8 deletions(-)
>>
>> diff --git a/kernel/cgroup.c b/kernel/cgroup.c
>> index 4936d88..bdb7994 100644
>> --- a/kernel/cgroup.c
>> +++ b/kernel/cgroup.c
>> @@ -1218,7 +1218,6 @@ static int parse_cgroupfs_options(char *data, struct cgroup_sb_opts *opts)
>>  	    (opts->subsys_bits & mask))
>>  		return -EINVAL;
>>  
>> -
>>  	/* Can't specify "none" and some subsystems */
>>  	if (opts->subsys_bits && opts->none)
>>  		return -EINVAL;
>> @@ -1231,6 +1230,13 @@ static int parse_cgroupfs_options(char *data, struct cgroup_sb_opts *opts)
>>  		return -EINVAL;
>>  
>>  	/*
>> +	 * This can happen only when none of the subsystems is enabled
>> +	 * in the system.
>> +	 */
>> +	if (!opts->subsys_bits && !opts->none)
>> +		return -EINVAL;
> 
> The three ifs here share a lot and I think sharing the common parts
> would make the logic clearer.  e.g. The first and this can be easily
> combined to (bool)opts->subsys_bits != (bool)opts->none which would
> explain the logic better too.
> 

ok.

>> @@ -1442,10 +1445,6 @@ static int cgroup_set_super(struct super_block *sb, void *data)
>>  	int ret;
>>  	struct cgroup_sb_opts *opts = data;
>>  
>> -	/* If we don't have a new root, we can't set up a new sb */
>> -	if (!opts->new_root)
>> -		return -EINVAL;
>> -
> 
> And where did this one go?
> 

The other one I removed sets opts->new_root to NULL, and this one detects
the NULL ptr:

if (!opts->subsys_bits && !opts->none)
	return NULL;
...
opts->new_root = NULL;
...
if (!opts->new_root)
	return -EINVAL;

shortcut to:

if (!opts->subsys_bits && !opts->none)
	return -EINVAL;

Re: [PATCH] cgroup: move a check to parse_cgroupfs_options()

[Tejun Heo]

Hello,

On Thu, Dec 22, 2011 at 7:40 PM, Li Zefan <lizf-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org> wrote:
>> And where did this one go?
>
> The other one I removed sets opts->new_root to NULL, and this one detects
> the NULL ptr:
>
> if (!opts->subsys_bits && !opts->none)
>        return NULL;
> ...
> opts->new_root = NULL;
> ...
> if (!opts->new_root)
>        return -EINVAL;
>
> shortcut to:
>
> if (!opts->subsys_bits && !opts->none)
>        return -EINVAL;

Thanks for the explanation. Maybe explaining it briefly in the commit
message would be nice?

-- 
tejun

Re: [PATCH] cgroup: move a check to parse_cgroupfs_options()

[Li Zefan]

Tejun Heo wrote:
> Hello,
> 
> On Thu, Dec 22, 2011 at 7:40 PM, Li Zefan <lizf-BthXqXjhjHXQFUHtdCDX3A@public.gmane.org> wrote:
>>> And where did this one go?
>>
>> The other one I removed sets opts->new_root to NULL, and this one detects
>> the NULL ptr:
>>
>> if (!opts->subsys_bits && !opts->none)
>>        return NULL;
>> ...
>> opts->new_root = NULL;
>> ...
>> if (!opts->new_root)
>>        return -EINVAL;
>>
>> shortcut to:
>>
>> if (!opts->subsys_bits && !opts->none)
>>        return -EINVAL;
> 
> Thanks for the explanation. Maybe explaining it briefly in the commit
> message would be nice?
> 

I recalled why we have this check in this place, that is to allow mounting
by hierarchy name, but it has been broken for a long time. I'll send a
fix with detailed changelog.