netdev
[Top] [All Lists]

Re: [PATCH] use mmiowb in tg3_poll

To: Arthur Kepner <akepner@xxxxxxx>
Subject: Re: [PATCH] use mmiowb in tg3_poll
From: jamal <hadi@xxxxxxxxxx>
Date: Tue, 31 May 2005 08:53:42 -0400
Cc: Lennert Buytenhek <buytenh@xxxxxxxxxxxxxx>, netdev@xxxxxxxxxxx, jesse.barnes@xxxxxxxxx, gnb@xxxxxxx
In-reply-to: <Pine.LNX.4.61.0505300925570.11645@xxxxxxxxxx>
Organization: unknown
References: <200410211628.06906.jbarnes@xxxxxxxxxxxx> <Pine.LNX.4.33.0410221345400.392-100000@xxxxxxxxxxxxxxxxxxxxx> <20050528231209.GA8456@xxxxxxxxxxxxxxxxx> <Pine.LNX.4.61.0505300925570.11645@xxxxxxxxxx>
Reply-to: hadi@xxxxxxxxxx
Sender: netdev-bounce@xxxxxxxxxxx
On Mon, 2005-30-05 at 09:30 -0700, Arthur Kepner wrote:

> I quickly looked through the records that I have and, unfortunately, 
> wasn't able to locate raw data. But I found some scrawlings in my 
> notes which say the most recent measurements for an Altix are: 
> 
>       i)  PIO read latency ~ 2.4 usec
>       ii) mmiowb() latency ~ 1.1 usec
> 


In terms of CPU cycles, how many do the above compute to be?
The absolute time numbers you have are also useful for quantification
purposes.

Did i read correctly that your reads are about 2x more expensive than
the writes?

cheers,
jamal


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