Isn't it possible to determine the likely/unlikely ratio's? I've come up with this idea: Change the definition of likely and unlikely - with a DEBUG_LIKELINESS config option set - let the macro use __FILE__ and __LINE__ to determine where the (un/)likely occurred. Evaluate the condition, and log those. After a while you'll get a ratio for each (un/)likely, and it can be decided whether the use was valid. Is there a reason why this isn't usable? Does the measuring alter the behavior? Will this turn a supercomuter into a VAX? Comments welcome Roel -- To unsubscribe from this list: send an email with "unsubscribe kernelnewbies" to ecartis@xxxxxxxxxxxx Please read the FAQ at http://kernelnewbies.org/FAQ
