Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1161067AbWHDF76 (ORCPT ); Fri, 4 Aug 2006 01:59:58 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1161065AbWHDF75 (ORCPT ); Fri, 4 Aug 2006 01:59:57 -0400 Received: from rhun.apana.org.au ([64.62.148.172]:35598 "EHLO arnor.apana.org.au") by vger.kernel.org with ESMTP id S1161061AbWHDF7z (ORCPT ); Fri, 4 Aug 2006 01:59:55 -0400 From: Herbert Xu To: chris.leech@gmail.com (Chris Leech) Subject: Re: problems with e1000 and jumboframes Cc: arnd@arndnet.de, johnpol@2ka.mipt.ru, olel@ans.pl, linux-kernel@vger.kernel.org, netdev@vger.kernel.org Organization: Core In-Reply-To: <41b516cb0608031334s6e159e99tb749240f44ae608d@mail.gmail.com> X-Newsgroups: apana.lists.os.linux.kernel,apana.lists.os.linux.netdev User-Agent: tin/1.7.4-20040225 ("Benbecula") (UNIX) (Linux/2.6.17-rc4 (i686)) Message-Id: Date: Fri, 04 Aug 2006 15:59:37 +1000 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1060 Lines: 22 Chris Leech wrote: > > We could try and only use page allocations for older e1000 devices, > putting headers and payload into skb->frags and copying the headers > out into the skb->data area as needed for processing. That would do > away with large allocations, but in Jesse's experiments calling > alloc_page() is slower than kmalloc(), so there can actually be a > performance hit from trying to use page allocations all the time. Interesting. Could you guys post figures on alloc_page speed vs. kmalloc? Also, getting memory slower is better than not getting them at all :) -- Visit Openswan at http://www.openswan.org/ Email: Herbert Xu ~{PmV>HI~} Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/