[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20120512.155259.1178343836887150194.davem@davemloft.net>
Date: Sat, 12 May 2012 15:52:59 -0400 (EDT)
From: David Miller <davem@...emloft.net>
To: eric.dumazet@...il.com
Cc: dave.taht@...ferbloat.net, netdev@...r.kernel.org,
nichols@...lere.com, van@...lere.net, codel@...ts.bufferbloat.net,
ycheng@...gle.com, mattmathis@...gle.com, therbert@...gle.com,
shemminger@...tta.com, nanditad@...gle.com
Subject: Re: [PATCH net-next] codel: use Newton method instead of sqrt()
and divides
From: Eric Dumazet <eric.dumazet@...il.com>
Date: Sat, 12 May 2012 15:32:13 +0200
> From: Eric Dumazet <edumazet@...gle.com>
>
> As Van pointed out, interval/sqrt(count) can be implemented using
> multiplies only.
>
> http://en.wikipedia.org/wiki/Methods_of_computing_square_roots#Iterative_methods_for_reciprocal_square_roots
>
> This patch implements the Newton method and reciprocal divide.
>
> Total cost is 15 cycles instead of 120 on my Corei5 machine (64bit
> kernel).
>
> There is a small 'error' for count values < 5, but we don't really care.
>
> I reuse a hole in struct codel_vars :
> - pack the dropping boolean into one bit
> - use 31bit to store the reciprocal value of sqrt(count).
>
> Suggested-by: Van Jacobson <van@...lere.net>
> Signed-off-by: Eric Dumazet <edumazet@...gle.com>
Applied but I never like that bitfield sharing for real integers.
GCC makes a complete mess of it as it extracts and inserts the
integer value into that bit field. You are guarenteed to get
better code if you do this by hand in a full u32.
Either that or just bite the bullet and use a completely seperate
field, maybe we'll need more boolean states later.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists