[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <alpine.LFD.0.98.0705011719520.3808@woody.linux-foundation.org>
Date: Tue, 1 May 2007 17:24:54 -0700 (PDT)
From: Linus Torvalds <torvalds@...ux-foundation.org>
To: Josh Triplett <josh@...edesktop.org>
cc: linux-sparse@...r.kernel.org, linux-kernel@...r.kernel.org,
Al Viro <viro@....linux.org.uk>
Subject: Re: sparse -Wptr-subtraction-blows: still needed?
On Tue, 1 May 2007, Josh Triplett wrote:
>
> Do you know whether the current version of GCC generates poor code for pointer
> subtraction?
You _cannot_ generate good code.
When you subtract two pointers, the C definition means that you first
subtract the values (cheap), and then you *divide* the result by the size
of the object the pointer points to (expensive!).
So pointer subtraction is by definition expensive, if the size of the
object is not some kind of nice power-of-two or similar.
Of course, modern CPU's are getting better at divides.
> Has anyone reported this poor code generation to the GCC bugzilla? If so, I
> can add a reference to the bug in any (hypothetical) documentation for
> -Wptr-subtraction-blows.
The only thing that was gcc-specific about it is that gcc goes to some
extreme lengths to turn the constant-sized division into a sequence of
shifts/multiples/subtracts, and can often turn a division into something
like ten cheaper operations instead.
But that optimization was also what made gcc take such a long time if the
constant division is very common (ie a header file with an inline
function, whether that function is actually _used_ or not apparently
didn't matter to gcc)
So gcc does pretty well on these divisions, and makes them cheaper (except
on CPU's where divides are really fast and possibly even cheaper than the
combination of shifts/subtracts, but afaik, that probably won't be until
the next-generation Intel Core 2 45nm "Penryn" thing)
Linus
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists