> right, but even in this case, the message emission from the kernel
 > allows us to catch callers like routing daemons who issue the 
 > ndd ioctls directly, so this method did have that advantage. 

Maybe on a DEBUG system it has value, but I think it's incorrect to emit
them in a production environment, and will ultimately cause needless worry
and support calls.

-- 
meem

Reply via email to