Re: rwlock(9) and mutex(9) definitions

--------
Gleb Smirnoff writes:

> Such wraps don't really behave as a true void. For example you
> can not tail call them:
>
> void
> smartass_lock(lock, clue)
> {
> 	if (clue)
> 		return (rw_rlock(lock));
> 	else
> 		return (rw_wlock(lock));
> }
>
> This will fail on rw_wlock, but not on rw_rlock. However, if you
> have WITNESS it will compile correctly :)

I'm not going to go full Bruce on you here, but...

In your example above, at the very least the "else" is surplus to requirements.

The "return" around rw_wlock also doesnt do nothing for its keep.

That leaves us with:

void
smartass_lock(lock, clue)
{
	if (clue)
		return (rw_rlock(lock));
	rw_wlock(lock);
}

That has pretty low aesthetic qualities, which is why I myself would
have written it as:

void
smartass_lock(lock, clue)
{
	if (clue)
		rw_rlock(lock);
	else
		rw_wlock(lock);
}

Finally "return(something_void)" is bad style in my book, because
"void" literally means there is no return value to begin with.

So all in all, I fail to see a problem that needs fixing ?

I realize that the above may be a simplified example of the real
problem you're trying to solve, but if so, it is too simplified
to carry your argument.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk@FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.