From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [PATCH 1/3] skb: Drop bad code from pskb_expand_head Date: Sat, 05 May 2012 07:35:30 +0200 Message-ID: <1336196130.3752.483.camel@edumazet-glaptop> References: <20120505001059.21292.31647.stgit@gitlad.jf.intel.com> <20120505002645.21292.38368.stgit@gitlad.jf.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org, davem@davemloft.net, jeffrey.t.kirsher@intel.com To: Alexander Duyck Return-path: Received: from mail-wi0-f172.google.com ([209.85.212.172]:57316 "EHLO mail-wi0-f172.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752025Ab2EEFff (ORCPT ); Sat, 5 May 2012 01:35:35 -0400 Received: by wibhj6 with SMTP id hj6so1911212wib.1 for ; Fri, 04 May 2012 22:35:34 -0700 (PDT) In-Reply-To: <20120505002645.21292.38368.stgit@gitlad.jf.intel.com> Sender: netdev-owner@vger.kernel.org List-ID: On Fri, 2012-05-04 at 17:26 -0700, Alexander Duyck wrote: > The fast-path for pskb_expand_head contains a check where the size plus the > unaligned size of skb_shared_info is compared against the size of the data > buffer. This code path has two issues. First is the fact that after the > recent changes by Eric Dumazet to __alloc_skb and build_skb the shared info > is always placed in the optimal spot for a buffer size making this check > unnecessary. The second issue is the fact that the check doesn't take into > account the aligned size of shared info. As a result the code burns cycles > doing a memcpy with nothing actually being shifted. > > Signed-off-by: Alexander Duyck > --- > > net/core/skbuff.c | 12 ------------ > 1 files changed, 0 insertions(+), 12 deletions(-) > > diff --git a/net/core/skbuff.c b/net/core/skbuff.c > index c199aa4..4d085d4 100644 > --- a/net/core/skbuff.c > +++ b/net/core/skbuff.c > @@ -951,17 +951,6 @@ int pskb_expand_head(struct sk_buff *skb, int nhead, int ntail, > fastpath = atomic_read(&skb_shinfo(skb)->dataref) == delta; > } > > - if (fastpath && !skb->head_frag && > - size + sizeof(struct skb_shared_info) <= ksize(skb->head)) { > - memmove(skb->head + size, skb_shinfo(skb), > - offsetof(struct skb_shared_info, > - frags[skb_shinfo(skb)->nr_frags])); > - memmove(skb->head + nhead, skb->head, > - skb_tail_pointer(skb) - skb->head); > - off = nhead; > - goto adjust_others; > - } > - > data = kmalloc(size + SKB_DATA_ALIGN(sizeof(struct skb_shared_info)), > gfp_mask); > if (!data) > @@ -997,7 +986,6 @@ int pskb_expand_head(struct sk_buff *skb, int nhead, int ntail, > > skb->head = data; > skb->head_frag = 0; > -adjust_others: > skb->data += off; > #ifdef NET_SKBUFF_DATA_USES_OFFSET > skb->end = size; > I totally agree this code is no longer needed, we already have the skb_shared_info at the end of the buffer. Acked-by: Eric Dumazet