netdev
[Top] [All Lists]

Re: 3c59x

To: chas@xxxxxxxxxxxxxxxx
Subject: Re: 3c59x
From: "David S. Miller" <davem@xxxxxxxxxx>
Date: Tue, 10 Jun 2003 10:31:11 -0700 (PDT)
Cc: jgarzik@xxxxxxxxx, ak@xxxxxxx, bogdan.costescu@xxxxxxxxxxxxxxxxxxxxx, hadi@xxxxxxxxxxxxxxxx, ralph+d@xxxxxxxxx, xerox@xxxxxxxxxx, sim@xxxxxxxxxxxxx, fw@xxxxxxxxxxxxx, netdev@xxxxxxxxxxx, linux-net@xxxxxxxxxxxxxxx
In-reply-to: <200306101731.h5AHVlsG008694@ginger.cmf.nrl.navy.mil>
References: <20030610171617.GC1959@gtf.org> <200306101731.h5AHVlsG008694@ginger.cmf.nrl.navy.mil>
Sender: netdev-bounce@xxxxxxxxxxx
   From: chas williams <chas@xxxxxxxxxxxxxxxx>
   Date: Tue, 10 Jun 2003 13:29:54 -0400

   In message <20030610171617.GC1959@xxxxxxx>,Jeff Garzik writes:
   >Making every IO a conditional branch?  Ug.
   
   you could just test once during driver init and setup an indirection
   to the appropriate function.  its a little better than test and branch.
   
Function calls are actually more expensive, you eat an entry in
the cpu's return address cache.

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