[PATCH 0/3] ratelimit: patchset focus on return value of __ratelimit()

[PATCH 0/3] ratelimit: patchset focus on return value of __ratelimit()

[Yong Zhang]

When try to use the new API pr_*_ratelimited(), I got the opposite outcome.
After some checking, I find print_ratelimited() misunderstand the return
value of __ratelimit(). So I fix it in this patchset and also annotate
__ratelimit() for later user.

I mark patch 3 with RFC because I'm afraid I misunderstand the intention,
Ingo, can you please take a look at it? You know that change is introduced
by edaac8e3167501cda336231d00611bf59c164346.

Yong Zhang (3):
      ratelimit: annotate ___ratelimit()
      kernel.h: fix wrong usage of __ratelimit()
      ratelimit: Fix return value when fail to get lock

 include/linux/kernel.h |    2 +-
 lib/ratelimit.c        |   11 ++++++++---
 2 files changed, 9 insertions(+), 4 deletions(-)

[PATCH 1/3] ratelimit: annotate ___ratelimit()

[Yong Zhang]

Especially to prevent from wrong using of the return value.

Signed-off-by: Yong Zhang <yong.zhang@windriver.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Dave Young <hidave.darkstar@gmail.com>
---
 lib/ratelimit.c |    9 +++++++--
 1 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/lib/ratelimit.c b/lib/ratelimit.c
index 09f5ce1..29a10b3 100644
--- a/lib/ratelimit.c
+++ b/lib/ratelimit.c
@@ -16,9 +16,14 @@
 /*
  * __ratelimit - rate limiting
  * @rs: ratelimit_state data
+ * @func: name of calling function
  *
- * This enforces a rate limit: not more than @rs->ratelimit_burst callbacks
- * in every @rs->ratelimit_jiffies
+ * This enforces a rate limit: not more than @rs->burst callbacks
+ * in every @rs->interval
+ *
+ * RETURNS:
+ * 0 means callbacks will be suppuressed.
+ * 1 means go ahead and do it.
  */
 int ___ratelimit(struct ratelimit_state *rs, const char *func)
 {
-- 
1.6.3.3

[PATCH 2/3] kernel.h: fix wrong usage of __ratelimit()

[Yong Zhang]

__ratelimit() rerurn 1 means we can go ahead.

Signed-off-by: Yong Zhang <yong.zhang@windriver.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Joe Perches <joe@perches.com>
---
 include/linux/kernel.h |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/include/linux/kernel.h b/include/linux/kernel.h
index 7f07074..9365227 100644
--- a/include/linux/kernel.h
+++ b/include/linux/kernel.h
@@ -426,7 +426,7 @@ static inline char *pack_hex_byte(char *buf, u8 byte)
 		.burst = DEFAULT_RATELIMIT_BURST,       \
 	};                                              \
 							\
-	if (!__ratelimit(&_rs))                         \
+	if (__ratelimit(&_rs))                          \
 		printk(fmt, ##__VA_ARGS__);		\
 })
 #else
-- 
1.6.3.3

[RFC PATCH 3/3] ratelimit: Fix return value when fail to get lock

[Yong Zhang]

>From log of commit edaac8e3167501cda336231d00611bf59c164346,
It seems that we want to suppress the callback when trylock
fails.

Signed-off-by: Yong Zhang <yong.zhang@windriver.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
---
 lib/ratelimit.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/lib/ratelimit.c b/lib/ratelimit.c
index 29a10b3..358638f 100644
--- a/lib/ratelimit.c
+++ b/lib/ratelimit.c
@@ -40,7 +40,7 @@ int ___ratelimit(struct ratelimit_state *rs, const char *func)
 	 * the entity that is holding the lock already:
 	 */
 	if (!spin_trylock_irqsave(&rs->lock, flags))
-		return 1;
+		return 0;
 
 	if (!rs->begin)
 		rs->begin = jiffies;
-- 
1.6.3.3

Re: [RFC PATCH 3/3] ratelimit: Fix return value when fail to get lock

[Andrew Morton]

On Wed, 17 Mar 2010 10:58:44 +0800
Yong Zhang <yong.zhang@windriver.com> wrote:

> >From log of commit edaac8e3167501cda336231d00611bf59c164346,
> It seems that we want to suppress the callback when trylock
> fails.
> 
> Signed-off-by: Yong Zhang <yong.zhang@windriver.com>
> Cc: Ingo Molnar <mingo@elte.hu>
> Cc: Christian Borntraeger <borntraeger@de.ibm.com>
> ---
>  lib/ratelimit.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
> 
> diff --git a/lib/ratelimit.c b/lib/ratelimit.c
> index 29a10b3..358638f 100644
> --- a/lib/ratelimit.c
> +++ b/lib/ratelimit.c
> @@ -40,7 +40,7 @@ int ___ratelimit(struct ratelimit_state *rs, const char *func)
>  	 * the entity that is holding the lock already:
>  	 */
>  	if (!spin_trylock_irqsave(&rs->lock, flags))
> -		return 1;
> +		return 0;
>  
>  	if (!rs->begin)
>  		rs->begin = jiffies;

Looks right to me.  I queued all three patches for 2.6.34.