Netdev List
 help / color / mirror / Atom feed
From: Eric Dumazet <eric.dumazet@gmail.com>
To: Giuseppe CAVALLARO <peppe.cavallaro@st.com>
Cc: netdev@vger.kernel.org
Subject: Re: [net-next.git 3/7] stmmac: review private structure fields
Date: Wed, 03 Apr 2013 00:21:31 -0700	[thread overview]
Message-ID: <1364973691.13853.7.camel@edumazet-glaptop> (raw)
In-Reply-To: <1364967689-11155-3-git-send-email-peppe.cavallaro@st.com>

On Wed, 2013-04-03 at 07:41 +0200, Giuseppe CAVALLARO wrote:
> recently many new supports have been added in the stmmac driver w/o taking care
> about where each new field had to be placed inside the private structure for
> guaranteeing the best cache usage.
> This is what I wanted in the beginning, so this patch reorganizes all the fields
> in order to keep adjacent fields for cache effect.
> I have also tried to optimize them by using pahole.
> 
> Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
> ---
>  drivers/net/ethernet/stmicro/stmmac/stmmac.h |   70 +++++++++++++-------------
>  1 files changed, 35 insertions(+), 35 deletions(-)
> 
> diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac.h b/drivers/net/ethernet/stmicro/stmmac/stmmac.h
> index 75f997b..8aa28c5 100644
> --- a/drivers/net/ethernet/stmicro/stmmac/stmmac.h
> +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac.h
> @@ -35,36 +35,45 @@
>  
>  struct stmmac_priv {
>  	/* Frequently used values are kept adjacent for cache effect */
> -	struct dma_desc *dma_tx ____cacheline_aligned;	/* Basic TX desc */
> -	struct dma_extended_desc *dma_etx;	/* Extended TX descriptor */
> -	dma_addr_t dma_tx_phy;
> -	struct sk_buff **tx_skbuff;
> -	dma_addr_t *tx_skbuff_dma;
> +	struct dma_extended_desc *dma_etx;
> +	struct dma_desc *dma_tx ____cacheline_aligned_in_smp;
> +	struct sk_buff **tx_skbuff ____cacheline_aligned_in_smp;

dma_tx & tx_skbuff are readonly, why put them in separate cache lines ?

It seems there is an abuse of ____cacheline_aligned_in_smp in this
driver (especially if this driver only runs on UP arch)

  reply	other threads:[~2013-04-03  7:21 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-04-03  5:41 [net-next.git 1/7] stmmac: review napi gro support Giuseppe CAVALLARO
2013-04-03  5:41 ` [net-next.git 2/7] stmmac: review barriers Giuseppe CAVALLARO
2013-04-03  7:18   ` Eric Dumazet
2013-04-03  7:28     ` Giuseppe CAVALLARO
2013-04-03  8:51   ` Shiraz Hashim
2013-04-04  6:06     ` Giuseppe CAVALLARO
2013-04-04 15:08       ` Eric Dumazet
2013-04-03 14:02   ` Sergei Shtylyov
2013-04-03  5:41 ` [net-next.git 3/7] stmmac: review private structure fields Giuseppe CAVALLARO
2013-04-03  7:21   ` Eric Dumazet [this message]
2013-04-03  7:33     ` Giuseppe CAVALLARO
2013-04-03 16:09       ` Ben Hutchings
2013-04-04  6:11         ` Giuseppe CAVALLARO
2013-04-03  5:41 ` [net-next.git 4/7] stmmac: review driver documentation Giuseppe CAVALLARO
2013-04-03  5:41 ` [net-next.git 5/7] stmmac: improve/review and fix kernel-doc Giuseppe CAVALLARO
2013-04-03  5:41 ` [net-next.git 6/7] stmmac: code tidy-up Giuseppe CAVALLARO
2013-04-03  5:41 ` [net-next.git 7/7] stmmac: prefetch all dma_erx when use extend_desc Giuseppe CAVALLARO
2013-04-03  7:05 ` [net-next.git 1/7] stmmac: review napi gro support Eric Dumazet
2013-04-03  7:41   ` Giuseppe CAVALLARO
2013-04-03 13:39     ` Eric Dumazet
2013-04-04  6:16       ` Giuseppe CAVALLARO
2013-04-03 16:01 ` Ben Hutchings
2013-04-04  6:20   ` Giuseppe CAVALLARO

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=1364973691.13853.7.camel@edumazet-glaptop \
    --to=eric.dumazet@gmail.com \
    --cc=netdev@vger.kernel.org \
    --cc=peppe.cavallaro@st.com \
    /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