netdev
[Top] [All Lists]

Re: [timers] net/core/*

To: Andi Kleen <ak@xxxxxx>
Subject: Re: [timers] net/core/*
From: Jeff Garzik <jgarzik@xxxxxxxxxxxxxxxx>
Date: Wed, 31 May 2000 16:19:25 -0400
Cc: kuznet@xxxxxxxxxxxxx, Andrew Morton <andrewm@xxxxxxxxxx>, netdev@xxxxxxxxxxx
Organization: MandrakeSoft
References: <39352AB3.7C609B8C@uow.edu.au> <200005311903.XAA23547@ms2.inr.ac.ru> <20000531212143.64448@colin.muc.de>
Sender: owner-netdev@xxxxxxxxxxx
Andi Kleen wrote:
> 
> On Wed, May 31, 2000 at 09:05:06PM +0200, kuznet@xxxxxxxxxxxxx wrote:
> >
> > It is consequence of too careful reading intel and alpha programmer's
> > guides. 8) All of them require to help static branch prediction
> > and not to jump forward in the most frequent path. It looks silly, when done
> > in C, but gcc has no means to do this in different way.
> 
> gcc 3.0 will fix this, with -freorder-blocks (which knows some common
> heuristics like == NULL is usually not taken already) and __builtin_expect()
> 
> it'll allow exorcise of about a zillion gotos from the kernel tree.

I wish... :)   We'll still have to support gcc 2.7.2...

-- 
Jeff Garzik              | Liberty is always dangerous, but
Building 1024            | it is the safest thing we have.
MandrakeSoft, Inc.       |      -- Harry Emerson Fosdick

<Prev in Thread] Current Thread [Next in Thread>