[PATCH] Move an assert under DEBUG_KERNEL.

[PATCH] Move an assert under DEBUG_KERNEL.

[Rob Landley]

From: Rob Landley <rlandley@parallels.com>

Move an assert under DEBUG_KERNEL.

Signed-off-by: Rob Landley <rlandley@parallels.com>
---

Saves about 3k from x86-64 defconfig according to scripts/bloat-o-meter.

  include/linux/rtnetlink.h |    4 ++++
  1 file changed, 4 insertions(+)

diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index bbad657..28c4025 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -782,6 +782,7 @@ extern struct netdev_queue 
*dev_ingress_queue_create(struct net_device *dev);
  extern void rtnetlink_init(void);
  extern void __rtnl_unlock(void);

+#ifdef CONFIG_DEBUG_KERNEL
  #define ASSERT_RTNL() do { \
  	if (unlikely(!rtnl_is_locked())) { \
  		printk(KERN_ERR "RTNL: assertion failed at %s (%d)\n", \
@@ -789,6 +790,9 @@ extern void __rtnl_unlock(void);
  		dump_stack(); \
  	} \
  } while(0)
+#else
+#define ASSERT_RTNL()
+#endif

  static inline u32 rtm_get_table(struct rtattr **rta, u8 table)
  {

Re: [PATCH] Move an assert under DEBUG_KERNEL.

[Andrew Morton]

On Thu, 6 Jan 2011 02:13:38 -0600
Rob Landley <rlandley@parallels.com> wrote:

> From: Rob Landley <rlandley@parallels.com>
> 
> Move an assert under DEBUG_KERNEL.
> 
> Signed-off-by: Rob Landley <rlandley@parallels.com>
> ---
> 
> Saves about 3k from x86-64 defconfig according to scripts/bloat-o-meter.
> 
>   include/linux/rtnetlink.h |    4 ++++
>   1 file changed, 4 insertions(+)
> 
> diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
> index bbad657..28c4025 100644
> --- a/include/linux/rtnetlink.h
> +++ b/include/linux/rtnetlink.h
> @@ -782,6 +782,7 @@ extern struct netdev_queue 
> *dev_ingress_queue_create(struct net_device *dev);
>   extern void rtnetlink_init(void);
>   extern void __rtnl_unlock(void);
> 
> +#ifdef CONFIG_DEBUG_KERNEL
>   #define ASSERT_RTNL() do { \
>   	if (unlikely(!rtnl_is_locked())) { \
>   		printk(KERN_ERR "RTNL: assertion failed at %s (%d)\n", \
> @@ -789,6 +790,9 @@ extern void __rtnl_unlock(void);
>   		dump_stack(); \
>   	} \
>   } while(0)
> +#else
> +#define ASSERT_RTNL()
> +#endif
> 
>   static inline u32 rtm_get_table(struct rtattr **rta, u8 table)
>   {

Probably a worthwhile thing to do, IMO.  If there's some net-specific
CONFIG_DEBUG_ setting then that wold be a better thing to use.

However the patch was a) wordwrapped, b) space-stuffed and c) not cc'ed
to the networking list.  So its prospects are dim.

Re: [PATCH] Move an assert under DEBUG_KERNEL.

[Rob Landley]

On 01/06/2011 05:41 PM, Andrew Morton wrote:
>> +#ifdef CONFIG_DEBUG_KERNEL
>>    #define ASSERT_RTNL() do { \
>>    	if (unlikely(!rtnl_is_locked())) { \
>>    		printk(KERN_ERR "RTNL: assertion failed at %s (%d)\n", \
>> @@ -789,6 +790,9 @@ extern void __rtnl_unlock(void);
>>    		dump_stack(); \
>>    	} \
>>    } while(0)
>> +#else
>> +#define ASSERT_RTNL()
>> +#endif
>>
>>    static inline u32 rtm_get_table(struct rtattr **rta, u8 table)
>>    {
>
> Probably a worthwhile thing to do, IMO.  If there's some net-specific
> CONFIG_DEBUG_ setting then that wold be a better thing to use.

I looked and didn't find one.  lib/Kconfig.debug has DEBUG_OBJECTS and 
PROVE_LOCKING and such but nothing quite on topic.  The only "DEBUG" in 
net/Kconfig is NETFLITER_DEBUG.  Nothing relevant in 
drivers/net/Kconfig, there isn't a Kconfig in net/core...

I thought about adding a new symbol, but CONFIG_DEBUG_KERNEL is already 
used in a few existing places:

   arch/powerpc/kernel/sysfs.c
   arch/parisc/mm/init.c
   arch/blackfin/include/asm/entry.h

So this isn't the first instance of it, but that doesn't mean those uses 
are correct. :)

> However the patch was a) wordwrapped, b) space-stuffed and c) not cc'ed
> to the networking list.  So its prospects are dim.

Sorry, finally gave up on kmail and set up thunderbird.  Still trying to 
beat the darn thing into submission.  (It looked right before I hit 
send.  And I cursored over the tabs to make sure. :)

I'll work out my email issues and then cc: the networking list on the 
resubmit.

Thanks,

Rob

[PATCH] Move an assert under DEBUG_KERNEL.

[Rob Landley]

From: Rob Landley <rlandley@parallels.com>

Move an assert under DEBUG_KERNEL.  (Minor cleanup, saves about 3k for me.)

Signed-off-by: Rob Landley <rlandley@parallels.com>
---

Note: I looked for a better config symbol than DEBUG_KERNEL but there
isn't currently a relevant DEBUG_NET variant and this doesn't seem
big enough to add one.  DEBUG_KERNEL is already used directly in
powerpc, parisc, and blackfin anyway.

 include/linux/rtnetlink.h |    4 ++++
 1 file changed, 4 insertions(+)

diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index bbad657..28c4025 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -782,6 +782,7 @@ extern struct netdev_queue *dev_ingress_queue_create(struct net_device *dev);
 extern void rtnetlink_init(void);
 extern void __rtnl_unlock(void);
 
+#ifdef CONFIG_DEBUG_KERNEL
 #define ASSERT_RTNL() do { \
 	if (unlikely(!rtnl_is_locked())) { \
 		printk(KERN_ERR "RTNL: assertion failed at %s (%d)\n", \
@@ -789,6 +790,9 @@ extern void __rtnl_unlock(void);
 		dump_stack(); \
 	} \
 } while(0)
+#else
+#define ASSERT_RTNL()
+#endif
 
 static inline u32 rtm_get_table(struct rtattr **rta, u8 table)
 {