Re: [PATCH] Reduce the number of expensive division instructions doneby _parse_integer()

[Linus Torvalds]

On Thu, Feb 9, 2012 at 8:28 AM, Eric Dumazet <eric.dumazet@xxxxxxxxx> wrote:
>
> You could avoid the divide and have cleaner code I think.
>
>        unsigned long long next = *res * base + val;
>
>        if (next < *res)
>                overflow = 1;
>        *res = next;

No. That pattern only works for a single addition.

For multiplication, you can overflow *and* still be bigger than the
starting value.

You either need to do the multiplication in a wider type (which would
be the natural way to do it on many 64-bit architectures, since they
often support 64x64->128 bit multiplies, and then doing a carry-add is
trivial), or you need to check the value before the multiplication. Or
you need to have a multiply that gives you overflow information, which
most don't.

So David's patch looks ok, although if this really is
performance-critical (and I could imagine some crazy /proc or /sys
access loads where the divide really does show up very clearly) I do
think it could be even done more efficiently. But probably only if we
start doing some arch-specific stuff.

                  Linus
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/