Subject: Re: uint32_t & u_int32_t
To: Masao Uebayashi <>
From: David Laight <>
List: current-users
Date: 12/17/2002 10:39:29
> 	uint16_t	htons __P((uint16_t)) __attribute__((__const__));

On a slightly separate note the above prototype would actually
be better if it were:

 	uint32_t	htons __P((uint32_t)) __attribute__((__const__));

Although the domain is [0..65535] there is no reason to force
the compiler to go through all the hoops that a 16bit function
parameter and result requires.

(OTOH the posix defn has uint16_t here - 'cos I didn't but a CR
in early enough)


David Laight: