From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [PATCH net-next] net: skb_peek()/skb_peek_tail() cleanups Date: Fri, 04 May 2012 15:01:31 +0200 Message-ID: <1336136491.3752.318.camel@edumazet-glaptop> References: <1335839506.11396.17.camel@edumazet-glaptop> <20120501.094127.190023878792795840.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org To: David Miller Return-path: Received: from mail-bk0-f46.google.com ([209.85.214.46]:37730 "EHLO mail-bk0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750880Ab2EDNBg (ORCPT ); Fri, 4 May 2012 09:01:36 -0400 Received: by bkcji2 with SMTP id ji2so2183592bkc.19 for ; Fri, 04 May 2012 06:01:35 -0700 (PDT) In-Reply-To: <20120501.094127.190023878792795840.davem@davemloft.net> Sender: netdev-owner@vger.kernel.org List-ID: On Tue, 2012-05-01 at 09:41 -0400, David Miller wrote: > From: Eric Dumazet > Date: Tue, 01 May 2012 04:31:46 +0200 > > > From: Eric Dumazet > > > > remove useless casts and rename variables for less confusion. > > > > Signed-off-by: Eric Dumazet > > Applied, I really need to get back to completing the list_head > conversion :-/ Actually doubly linked lists everywhere has a performance issue. When we dequeue one skb, we must bring in cpu cache the next skb as well, to perform the unlink. In some places it would be better to not have a double linked list, and only perform a prefetch(skb->next).