[PATCH] sysctl: remove impossible condition check

[PATCH] sysctl: remove impossible condition check

[Lucas De Marchi]

Signed-off-by: Lucas De Marchi <lucas.demarchi@profusion.mobi>
---
 fs/proc/proc_sysctl.c |    5 +----
 1 files changed, 1 insertions(+), 4 deletions(-)

diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c
index 2e5d3ec..98e82d4 100644
--- a/fs/proc/proc_sysctl.c
+++ b/fs/proc/proc_sysctl.c
@@ -52,11 +52,8 @@ static struct ctl_table *find_in_table(struct ctl_table *p, struct qstr *name)
 {
 	int len;
 	for ( ; p->procname; p++) {
-
-		if (!p->procname)
-			continue;
-
 		len = strlen(p->procname);
+
 		if (len != name->len)
 			continue;
 
-- 
1.7.5.2

Re: [PATCH] sysctl: remove impossible condition check

[Jesper Juhl]

On Wed, 1 Jun 2011, Lucas De Marchi wrote:

> Signed-off-by: Lucas De Marchi <lucas.demarchi@profusion.mobi>
> ---
>  fs/proc/proc_sysctl.c |    5 +----
>  1 files changed, 1 insertions(+), 4 deletions(-)
> 
> diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c
> index 2e5d3ec..98e82d4 100644
> --- a/fs/proc/proc_sysctl.c
> +++ b/fs/proc/proc_sysctl.c
> @@ -52,11 +52,8 @@ static struct ctl_table *find_in_table(struct ctl_table *p, struct qstr *name)
>  {
>  	int len;
>  	for ( ; p->procname; p++) {
> -
> -		if (!p->procname)
> -			continue;
> -
>  		len = strlen(p->procname);
> +
>  		if (len != name->len)
>  			continue;

How about compacting it even further by getting rid of the 'len' variable 
as well?
Like this:

Signed-off-by: Jesper Juhl <jj@chaosbits.net>
---
 proc_sysctl.c |   10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)

diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c
index f50133c..bd7f7af 100644
--- a/fs/proc/proc_sysctl.c
+++ b/fs/proc/proc_sysctl.c
@@ -49,17 +49,11 @@ out:
 
 static struct ctl_table *find_in_table(struct ctl_table *p, struct qstr *name)
 {
-	int len;
 	for ( ; p->procname; p++) {
-
-		if (!p->procname)
-			continue;
-
-		len = strlen(p->procname);
-		if (len != name->len)
+		if (strlen(p->procname) != name->len)
 			continue;
 
-		if (memcmp(p->procname, name->name, len) != 0)
+		if (memcmp(p->procname, name->name, name->len) != 0)
 			continue;
 
 		/* I have a match */


-- 
Jesper Juhl <jj@chaosbits.net>       http://www.chaosbits.net/
Don't top-post http://www.catb.org/jargon/html/T/top-post.html
Plain text mails only, please.

Re: [PATCH] sysctl: remove impossible condition check

[Lucas De Marchi]

[ CC'ing Al Viro ]

On Wed, Jun 1, 2011 at 10:25 AM, Jesper Juhl <jj@chaosbits.net> wrote:
> How about compacting it even further by getting rid of the 'len' variable
> as well?
> Like this:
>
> Signed-off-by: Jesper Juhl <jj@chaosbits.net>
> ---
>  proc_sysctl.c |   10 ++--------
>  1 file changed, 2 insertions(+), 8 deletions(-)
>
> diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c
> index f50133c..bd7f7af 100644
> --- a/fs/proc/proc_sysctl.c
> +++ b/fs/proc/proc_sysctl.c
> @@ -49,17 +49,11 @@ out:
>
>  static struct ctl_table *find_in_table(struct ctl_table *p, struct qstr *name)
>  {
> -       int len;
>        for ( ; p->procname; p++) {
> -
> -               if (!p->procname)
> -                       continue;
> -
> -               len = strlen(p->procname);
> -               if (len != name->len)
> +               if (strlen(p->procname) != name->len)
>                        continue;
>
> -               if (memcmp(p->procname, name->name, len) != 0)
> +               if (memcmp(p->procname, name->name, name->len) != 0)
>                        continue;
>
>                /* I have a match */
>


Looking again at the code, I'm wondering if this is not actually a
bug. There might be entries with procname == NULL, meaning they are
not mirrored in /proc. What seems wrong is the condition in the for().
It should stop when all fields are 0 (meaning the end of the table)
instead of stopping when procname is NULL.

>
> --
> Jesper Juhl <jj@chaosbits.net>       http://www.chaosbits.net/
> Don't top-post http://www.catb.org/jargon/html/T/top-post.html
> Plain text mails only, please.
>
>

Re: [PATCH] sysctl: remove impossible condition check

[Eric W. Biederman]

Lucas De Marchi <lucas.demarchi@profusion.mobi> writes:

> [ CC'ing Al Viro ]
>
> On Wed, Jun 1, 2011 at 10:25 AM, Jesper Juhl <jj@chaosbits.net> wrote:
>> How about compacting it even further by getting rid of the 'len' variable
>> as well?
>> Like this:
>>
>> Signed-off-by: Jesper Juhl <jj@chaosbits.net>
>> ---
>>  proc_sysctl.c |   10 ++--------
>>  1 file changed, 2 insertions(+), 8 deletions(-)
>>
>> diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c
>> index f50133c..bd7f7af 100644
>> --- a/fs/proc/proc_sysctl.c
>> +++ b/fs/proc/proc_sysctl.c
>> @@ -49,17 +49,11 @@ out:
>>
>>  static struct ctl_table *find_in_table(struct ctl_table *p, struct qstr *name)
>>  {
>> -       int len;
>>        for ( ; p->procname; p++) {
>> -
>> -               if (!p->procname)
>> -                       continue;
>> -
>> -               len = strlen(p->procname);
>> -               if (len != name->len)
>> +               if (strlen(p->procname) != name->len)
>>                        continue;
>>
>> -               if (memcmp(p->procname, name->name, len) != 0)
>> +               if (memcmp(p->procname, name->name, name->len) != 0)
>>                        continue;
>>
>>                /* I have a match */
>>
>
>
> Looking again at the code, I'm wondering if this is not actually a
> bug. There might be entries with procname == NULL, meaning they are
> not mirrored in /proc. What seems wrong is the condition in the for().
> It should stop when all fields are 0 (meaning the end of the table)
> instead of stopping when procname is NULL.

It is not a bug.  The condition was originally p->ctlname then
it became p->ctlname || p->procname and then finally I was able to
kill ctl_name.

What you see is a left over that didn't get removed.

This is also the second time in the last couple of weeks someone has
sent this patch. 

There is some ongoing work to make sysctl scale better that with
any luck should be ready for 3.1.  Decide which version of this
patch you like and please resend, and I will add this to my
sysctl tree.

Thank you,
Eric

Re: [PATCH] sysctl: remove impossible condition check

[Lucas De Marchi]

On Thu, Jun 2, 2011 at 10:40 AM, Eric W. Biederman
<ebiederm@xmission.com> wrote:
>> Looking again at the code, I'm wondering if this is not actually a
>> bug. There might be entries with procname == NULL, meaning they are
>> not mirrored in /proc. What seems wrong is the condition in the for().
>> It should stop when all fields are 0 (meaning the end of the table)
>> instead of stopping when procname is NULL.
>
> It is not a bug.  The condition was originally p->ctlname then
> it became p->ctlname || p->procname and then finally I was able to
> kill ctl_name.
>
> What you see is a left over that didn't get removed.
>

Alright then.

> This is also the second time in the last couple of weeks someone has
> sent this patch.
>
> There is some ongoing work to make sysctl scale better that with
> any luck should be ready for 3.1.  Decide which version of this
> patch you like and please resend, and I will add this to my
> sysctl tree.

There's also the same thing in this file, in scan() function. I'll add
it to the previous patch and resend.

thanks
Lucas De Marchi

[PATCH] sysctl: remove impossible condition check

[Lucas De Marchi]

Signed-off-by: Lucas De Marchi <lucas.demarchi@profusion.mobi>
---
 fs/proc/proc_sysctl.c |   14 ++------------
 1 files changed, 2 insertions(+), 12 deletions(-)

diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c
index f50133c..6cbcb1e 100644
--- a/fs/proc/proc_sysctl.c
+++ b/fs/proc/proc_sysctl.c
@@ -49,17 +49,11 @@ out:
 
 static struct ctl_table *find_in_table(struct ctl_table *p, struct qstr *name)
 {
-	int len;
 	for ( ; p->procname; p++) {
-
-		if (!p->procname)
-			continue;
-
-		len = strlen(p->procname);
-		if (len != name->len)
+		if (strlen(p->procname) != name->len)
 			continue;
 
-		if (memcmp(p->procname, name->name, len) != 0)
+		if (memcmp(p->procname, name->name, name->len) != 0)
 			continue;
 
 		/* I have a match */
@@ -223,10 +217,6 @@ static int scan(struct ctl_table_header *head, ctl_table *table,
 	for (; table->procname; table++, (*pos)++) {
 		int res;
 
-		/* Can't do anything without a proc name */
-		if (!table->procname)
-			continue;
-
 		if (*pos < file->f_pos)
 			continue;
 
-- 
1.7.5.2

Re: [PATCH] sysctl: remove impossible condition check

[Jesper Juhl]

On Fri, 10 Jun 2011, Lucas De Marchi wrote:

> Remove checks for conditions that will never happen. If procname is NULL
> the loop would already had bailed out, so there's no need to check it
> again.
> 
> At the same time this also compacts the function find_in_table() by
> refactoring it to be easier to read.
> 
> Signed-off-by: Lucas De Marchi <lucas.demarchi@profusion.mobi>

Looks good to me.

Reviewed-by: Jesper Juhl <jj@chaosbits.net>

-- 
Jesper Juhl <jj@chaosbits.net>       http://www.chaosbits.net/
Don't top-post http://www.catb.org/jargon/html/T/top-post.html
Plain text mails only, please.

Re: [PATCH] sysctl: remove impossible condition check

[Eric W. Biederman]

Jesper Juhl <jj@chaosbits.net> writes:

> On Fri, 10 Jun 2011, Lucas De Marchi wrote:
>
>> Remove checks for conditions that will never happen. If procname is NULL
>> the loop would already had bailed out, so there's no need to check it
>> again.
>> 
>> At the same time this also compacts the function find_in_table() by
>> refactoring it to be easier to read.
>> 
>> Signed-off-by: Lucas De Marchi <lucas.demarchi@profusion.mobi>
>
> Looks good to me.
>
> Reviewed-by: Jesper Juhl <jj@chaosbits.net>

Applied thanks all.

Eric