Date: Sun, 17 Jun 2001 14:56:00 +1000 (EST) From: Bruce Evans <bde@zeta.org.au> To: Garrett Wollman <wollman@khavrinen.lcs.mit.edu> Cc: Jonathan Lemon <jlemon@flugsvamp.com>, Jonathan Lemon <jlemon@FreeBSD.org>, cvs-committers@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/usr.sbin/rarpd rarpd.c Message-ID: <Pine.BSF.4.21.0106171421060.97985-100000@besplex.bde.org> In-Reply-To: <200106161821.OAA38884@khavrinen.lcs.mit.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, 16 Jun 2001, Garrett Wollman wrote: > <<On Sat, 16 Jun 2001 11:09:40 -0500, Jonathan Lemon <jlemon@flugsvamp.com> said: > > > Yes - sizeof() appears to be a 'long int' on the alpha, but an 'int' > > on the i386. > > The correct way to print the result of the `sizeof' operator using > `printf()' in C89 is with the `%lu' format (and the argument cast to > `unsigned long'), and in C99 is with the `%ju' format (and the > argument cast to `uintmax_t'). This is true always, on every > architecture, and has been true since C89 introduced size_t to begin > with. These are not incorrect ways. They may be gratuitously inefficient. E.g., the value returned by sizeof() may be known to be smaller than INT_MAX, and then you can print it using %d after casting it to int. Of course, efficiency of printf is usually unimportant. Someone mentioned that `%zu' can be used in C99 for arguments of type size_t. Note that `%zu' can't be used directly in rarp_checkd(), because the argument is the sum of 2 size_t's so it might not have type size_t. `%zu' can be used after casting the sum to size_t only if it is known that the sum is smaller than SIZE_MAX. The efficiency of the printf in rarpd_checkd() is certainly not important enough to justify this optimization. Bruce To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.21.0106171421060.97985-100000>