Re: __attribute__((warn_unused_result)) for NetBSD?

On Tue, Sep 16, 2008 at 12:56:01AM +0300, Aleksey Cheusov wrote:
 > I already made my desicion years ago - in programming paranoia is good.

There's necessarily a tradeoff between paranoia and pragmatism.
Finding the right balance is essential.

 > The original question was... Do you like an idea of marking functions
 > (conditionally! e.g. if WARNS=4 etc.) with "warn_unused_result" IN
 > GENERAL?  And if yes, _what kind of functions_ you WILL mark with it.

Yes, but cautiously.

 > Variants:
 > 1) this mark is useless at all
 > 2) only for functions with no side effects
 > 3) for some functions that may fail (read, write etc.)
 > 4) 2 + 3

Setting warn-unused-result on that may fail is nearly useless. You
want warn-missing-failure-check, which is a lot harder to implement.
Warn-unused-result is not an adequate substitute; consider these code
fragments:

   (1)
	int *x = malloc(sizeof(*x));
	*x = 3;

   (2)
	int result = read(fd, buf, maxlen-1);
	buf[result] = 0;


-- 
David A. Holland
[email protected]