netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Eric Dumazet <dada1@cosmosbay.com>
To: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Cc: Jeff Garzik <jgarzik@pobox.com>,
	netdev@vger.kernel.org, David Miller <davem@davemloft.net>,
	John Rociak <john.ronciak@intel.com>,
	Jesse Brandeburg <jesse.brandeburg@intel.com>
Subject: Re: [PATCH 11/11] ixgb: Add prefetch
Date: Sat, 22 Apr 2006 07:34:41 +0200	[thread overview]
Message-ID: <4449C071.5000107@cosmosbay.com> (raw)
In-Reply-To: <20060422010055.24255.71946.stgit@jk-desktop.jf.intel.com>

Jeff Kirsher a écrit :
> - This patch is to improve performance by adding prefetch to the ixgb driver
> - Add driver comments
> 
> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
> Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
> Signed-off-by: John Ronciak <john.ronciak@intel.com>
> ---
> 
>  drivers/net/ixgb/ixgb_main.c |   10 +++++++++-
>  1 files changed, 9 insertions(+), 1 deletions(-)
> 
> diff --git a/drivers/net/ixgb/ixgb_main.c b/drivers/net/ixgb/ixgb_main.c
> index 26cb0d5..98303cb 100644
> --- a/drivers/net/ixgb/ixgb_main.c
> +++ b/drivers/net/ixgb/ixgb_main.c
> @@ -29,6 +29,13 @@
>  #include "ixgb.h"
>  
>  /* Change Log
> + * 1.0.104 10-Jan-2006
> + * - fix for copybreak/recycle
> + * 1.0.103 Oct-3
> + * - suck in some e1000 changes, including copybreak and LLTX
> + * - support for CX4 adapters
> + * 1.0.102 June-20-2005
> + * - add a workaround for a hardware issue when using TSO
>   * 1.0.96 04/19/05
>   * - Make needlessly global code static -- bunk@stusta.de
>   * - ethtool cleanup -- shemminger@osdl.org
> @@ -1916,7 +1923,7 @@ ixgb_clean_rx_irq(struct ixgb_adapter *a
>  		skb = buffer_info->skb;
>  		buffer_info->skb = NULL;
>  
> -		prefetch(skb->data);
> +		prefetch(skb->data - NET_IP_ALIGN);

I doubt this change is usefull.

skb->data and dkb->dta - NET_IP_ALIGN are on the same cache line.
So prefetch(skb->data) is cheaper for he compiler and has the same effect on 
the memory prefetch that is eventually done by the cpu.


>  
>  		if(++i == rx_ring->count) i = 0;
>  		next_rxd = IXGB_RX_DESC(*rx_ring, i);
> @@ -1929,6 +1936,7 @@ ixgb_clean_rx_irq(struct ixgb_adapter *a
>  		next_buffer = &rx_ring->buffer_info[i];
>  		next_skb = next_buffer->skb;
>  		prefetch(next_skb);
> +		prefetch(next_skb->data - NET_IP_ALIGN);

I doubt that next->skb_data is available for free just after a 
prefetch(next_skb). This second prefetch has a hidden cost : The memory 
location (&next_skb->data) must be in L1 cache. So basically the 
prefetch(next_skb) is useless...

prefetch are not magic things. They have a cost (they increase the code size), 
and should be used carefully.

Eric

  reply	other threads:[~2006-04-22 19:26 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-04-22  1:00 [PATCH 00/11] ixgb: driver update (upstream) Jeff Kirsher
2006-04-22  1:00 ` [PATCH 01/11] ixgb: Fix compilation errors by initializing variables Jeff Kirsher
2006-04-22  8:49   ` Francois Romieu
2006-04-22  1:00 ` [PATCH 02/11] ixgb: Fix the use of dprintk rather than printk Jeff Kirsher
2006-04-22  9:03   ` Francois Romieu
2006-04-24 17:24     ` Stephen Hemminger
2006-05-19 18:35     ` Auke Kok
2006-04-22  1:00 ` [PATCH 03/11] ixgb: Fix hard coded numbers Jeff Kirsher
2006-04-22  9:17   ` Francois Romieu
2006-04-22  1:00 ` [PATCH 04/11] ixgb: Fix duplicate code Jeff Kirsher
2006-04-22  1:00 ` [PATCH 05/11] ixgb: Fix timeout code Jeff Kirsher
2006-04-22  1:00 ` [PATCH 06/11] ixgb: Fix TSO Jeff Kirsher
2006-04-22  1:00 ` [PATCH 07/11] ixgb: Fixed flow control parameters Jeff Kirsher
2006-04-22  1:00 ` [PATCH 08/11] ixgb: Add support for copper 10GbE Jeff Kirsher
2006-04-22  1:00 ` [PATCH 09/11] ixgb: Add performance enhancements to the buffer_info struct Jeff Kirsher
2006-04-22  1:00 ` [PATCH 10/11] ixgb: clean up whitespace Jeff Kirsher
2006-04-22  1:00 ` [PATCH 11/11] ixgb: Add prefetch Jeff Kirsher
2006-04-22  5:34   ` Eric Dumazet [this message]
2006-04-26 10:22 ` [PATCH 00/11] ixgb: driver update (upstream) Jeff Garzik

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=4449C071.5000107@cosmosbay.com \
    --to=dada1@cosmosbay.com \
    --cc=davem@davemloft.net \
    --cc=jeffrey.t.kirsher@intel.com \
    --cc=jesse.brandeburg@intel.com \
    --cc=jgarzik@pobox.com \
    --cc=john.ronciak@intel.com \
    --cc=netdev@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).