Subject: Re: device polling system
To: None <thir@thir.org>
From: Hayakawa Koichi <haya@arch.sony.co.jp>
List: tech-kern
Date: 05/30/2003 18:55:15
From: Takahiro Igarashi <thir@thir.org>
Subject: device polling system
Date: Thu, 29 May 2003 00:54:23 +0900 (JST)
Message-ID: <20030529.005423.68559706.tigara@pop02.odn.ne.jp>

 > I am now working to implement the device polling system on
 > NetBSD.  The project page is at:
 > 
 > http://thir.org/thir/NetBSD/devicepolling.en.html
 > 
 > The patch is still based on FreeBSD's one.  But I now start
 > to select the way of implementation of it and so on.
 > Fortunately my patch seems to work well.
 > 
 > Comments?

Interesting!  It is a good start point, I think.

I have some questions and comments.

 * What NIC did you use for your benchmark?

 * How many clock interrupts per second (hz value)?

 * Did you measure TCP/IP performance?

   I suppose this may hit TCP/IP performance, because ACK
   transmission delay may not be negrectible under low clock
   interrupt rate.

I suppose, for the first step, making generic polling
interrupt handler, without NIC driver changes, is better.
Do you consider sporadic server or some kind of similar
handler on NetBSD?