Received: by 2002:a05:6a10:8c0a:0:0:0:0 with SMTP id go10csp992623pxb; Fri, 22 Jan 2021 04:43:06 -0800 (PST) X-Google-Smtp-Source: ABdhPJwt13xNfKrtuZw9s8GFrH0c1DERx6X+ts3y/PP3qCw0PGK27/umJ982bGtsSMMw+mwIv3fR X-Received: by 2002:a17:906:c1d9:: with SMTP id bw25mr2780513ejb.452.1611319386280; Fri, 22 Jan 2021 04:43:06 -0800 (PST) ARC-Seal: i=1; a=rsa-sha256; t=1611319386; cv=none; d=google.com; s=arc-20160816; b=iEHUimcOxzLTZdMeueLwzmyl8irTlRSNlaKMmS+vL3i5G4Z4uCnuABHdrPDJR6aIHB 97qFuMhq2vIMZcd9dfKzNYIhhSOCkaDxpilyf134rC5spMrVaTu3tV2rCxAsGvBYFRJ3 9bd8oD1qADLHB70BcIOhdh3sKEGrJRdeLn3Jnf0vLhoC07sZhwqgCHDTcdPwTmMas05m rRlJStHIRa2aXXn154R+WaZMtpdb1arjUy0/O0gDFpfnOkax7HRMeR6Fdh+Ba6w0mJxz 06YXyjRgR3hsqpKSbwpZveivovr4uXCv7j2JD/I7UXqjRWsJ/Q0DR7h0oaL+e2F4En5/ 1aJA== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=list-id:precedence:content-transfer-encoding:mime-version :references:in-reply-to:message-id:subject:reply-to:cc:from:to :dkim-signature:date; bh=MurfsX/xFc6EJaZnP4aHHBq5FK0oyXI/IchGUo+j9ZM=; b=kuTCRERl6XIToBA4YrBAm3Dl/XCtODPzq2lFIwNKXjxpDyOHYUd2ttPvtjwg2r5wlG VJszGGI4V40wIwPN80brtNSW0MrI5dLsg273nE9RzRZRp81x7FcHQ9GJxHcDg5Up5sl/ 13FBBJHq4kLnBdsw6qMlygvTG+FyuTa6/p5es+dlum0I0cDXnUirfJZDnJpugwZkFgqX cZ6xYfMK3TVSjE8XVBl4N0utQzY9RCR6ai3TcEoDZU15eLEIAdAASjpRPXw0UvtngACO 0O74dOvmkGq03wj4ElWLlo4NOsa9NIHtPzZCNmwtREP9H+f00gXjHw3awiEAP6khPaOl kJKA== ARC-Authentication-Results: i=1; mx.google.com; dkim=pass header.i=@pm.me header.s=protonmail header.b=P5QmSyiH; spf=pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=pass (p=QUARANTINE sp=QUARANTINE dis=NONE) header.from=pm.me Return-Path: Received: from vger.kernel.org (vger.kernel.org. [23.128.96.18]) by mx.google.com with ESMTP id e6si1217554edz.362.2021.01.22.04.42.42; Fri, 22 Jan 2021 04:43:06 -0800 (PST) Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18; Authentication-Results: mx.google.com; dkim=pass header.i=@pm.me header.s=protonmail header.b=P5QmSyiH; spf=pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=pass (p=QUARANTINE sp=QUARANTINE dis=NONE) header.from=pm.me Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727571AbhAVMk3 (ORCPT + 99 others); Fri, 22 Jan 2021 07:40:29 -0500 Received: from mail-40131.protonmail.ch ([185.70.40.131]:28567 "EHLO mail-40131.protonmail.ch" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726515AbhAVMk1 (ORCPT ); Fri, 22 Jan 2021 07:40:27 -0500 Date: Fri, 22 Jan 2021 12:39:31 +0000 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=pm.me; s=protonmail; t=1611319177; bh=MurfsX/xFc6EJaZnP4aHHBq5FK0oyXI/IchGUo+j9ZM=; h=Date:To:From:Cc:Reply-To:Subject:In-Reply-To:References:From; b=P5QmSyiHLWiyp62ojlSpBj+PHE1FilejDgySIqAe9ggMhdv1NVhIof4xs/N55knKb lYhnbkSzpBVSG7NeKVKCilPmCBlUK9jlPakwiabCSJUR/PM6ZS2gzywSP7DHzKgNjb adi6rBSJEghwHIOB5DdIB89uZ648rDUIclfP0wmc/TQMTg1qmJHGUKAt575ne5q5Jg 5ifBiPmbKYmM+hFb1Pw/GudMxMSHsfTL8zpUD+h+qf9iaG4yXQYGf4ay9gFfsVhcuI 5xho5xgcUZZYbrg5vdnKdSdOHxxJm+P8rVEYObVh9g400FOY3PL1hcB9NR3UklPPR8 jYK62pLW0iOhg== To: Magnus Karlsson From: Alexander Lobakin Cc: Alexander Lobakin , Eric Dumazet , Xuan Zhuo , "Michael S. Tsirkin" , Jason Wang , "David S. Miller" , Jakub Kicinski , Bjorn Topel , Magnus Karlsson , Jonathan Lemon , Alexei Starovoitov , Daniel Borkmann , Jesper Dangaard Brouer , John Fastabend , Andrii Nakryiko , Martin KaFai Lau , Song Liu , Yonghong Song , KP Singh , virtualization@lists.linux-foundation.org, bpf , Network Development , open list Reply-To: Alexander Lobakin Subject: Re: [PATCH bpf-next v3 3/3] xsk: build skb by page Message-ID: <20210122123909.3603-1-alobakin@pm.me> In-Reply-To: References: <340f1dfa40416dd966a56e08507daba82d633088.1611236588.git.xuanzhuo@linux.alibaba.com> <20210122114729.1758-1-alobakin@pm.me> <20210122115519.2183-1-alobakin@pm.me> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable X-Spam-Status: No, score=-1.2 required=10.0 tests=ALL_TRUSTED,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,DKIM_VALID_EF shortcircuit=no autolearn=disabled version=3.4.4 X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on mailout.protonmail.ch Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org From: Magnus Karlsson Date: Fri, 22 Jan 2021 13:18:47 +0100 > On Fri, Jan 22, 2021 at 12:57 PM Alexander Lobakin wrote= : > > > > From: Alexander Lobakin > > Date: Fri, 22 Jan 2021 11:47:45 +0000 > > > > > From: Eric Dumazet > > > Date: Thu, 21 Jan 2021 16:41:33 +0100 > > > > > > > On 1/21/21 2:47 PM, Xuan Zhuo wrote: > > > > > This patch is used to construct skb based on page to save memory = copy > > > > > overhead. > > > > > > > > > > This function is implemented based on IFF_TX_SKB_NO_LINEAR. Only = the > > > > > network card priv_flags supports IFF_TX_SKB_NO_LINEAR will use pa= ge to > > > > > directly construct skb. If this feature is not supported, it is s= till > > > > > necessary to copy data to construct skb. > > > > > > > > > > ---------------- Performance Testing ------------ > > > > > > > > > > The test environment is Aliyun ECS server. > > > > > Test cmd: > > > > > ``` > > > > > xdpsock -i eth0 -t -S -s > > > > > ``` > > > > > > > > > > Test result data: > > > > > > > > > > size 64 512 1024 1500 > > > > > copy 1916747 1775988 1600203 1440054 > > > > > page 1974058 1953655 1945463 1904478 > > > > > percent 3.0% 10.0% 21.58% 32.3% > > > > > > > > > > Signed-off-by: Xuan Zhuo > > > > > Reviewed-by: Dust Li > > > > > --- > > > > > net/xdp/xsk.c | 104 ++++++++++++++++++++++++++++++++++++++++++++= ++++---------- > > > > > 1 file changed, 86 insertions(+), 18 deletions(-) > > > > > > > > > > diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c > > > > > index 4a83117..38af7f1 100644 > > > > > --- a/net/xdp/xsk.c > > > > > +++ b/net/xdp/xsk.c > > > > > @@ -430,6 +430,87 @@ static void xsk_destruct_skb(struct sk_buff = *skb) > > > > > sock_wfree(skb); > > > > > } > > > > > > > > > > +static struct sk_buff *xsk_build_skb_zerocopy(struct xdp_sock *x= s, > > > > > + struct xdp_desc *desc) > > > > > +{ > > > > > + u32 len, offset, copy, copied; > > > > > + struct sk_buff *skb; > > > > > + struct page *page; > > > > > + void *buffer; > > > > > + int err, i; > > > > > + u64 addr; > > > > > + > > > > > + skb =3D sock_alloc_send_skb(&xs->sk, 0, 1, &err); > > > > > + if (unlikely(!skb)) > > > > > + return ERR_PTR(err); > > > > > + > > > > > + addr =3D desc->addr; > > > > > + len =3D desc->len; > > > > > + > > > > > + buffer =3D xsk_buff_raw_get_data(xs->pool, addr); > > > > > + offset =3D offset_in_page(buffer); > > > > > + addr =3D buffer - xs->pool->addrs; > > > > > + > > > > > + for (copied =3D 0, i =3D 0; copied < len; i++) { > > > > > + page =3D xs->pool->umem->pgs[addr >> PAGE_SHIFT]; > > > > > + > > > > > + get_page(page); > > > > > + > > > > > + copy =3D min_t(u32, PAGE_SIZE - offset, len - copied); > > > > > + > > > > > + skb_fill_page_desc(skb, i, page, offset, copy); > > > > > + > > > > > + copied +=3D copy; > > > > > + addr +=3D copy; > > > > > + offset =3D 0; > > > > > + } > > > > > + > > > > > + skb->len +=3D len; > > > > > + skb->data_len +=3D len; > > > > > > > > > + skb->truesize +=3D len; > > > > > > > > This is not the truesize, unfortunately. > > > > > > > > We need to account for the number of pages, not number of bytes. > > > > > > The easiest solution is: > > > > > > skb->truesize +=3D PAGE_SIZE * i; > > > > > > i would be equal to skb_shinfo(skb)->nr_frags after exiting the loop. > > > > Oops, pls ignore this. I forgot that XSK buffers are not > > "one per page". > > We need to count the number of pages manually and then do > > > > skb->truesize +=3D PAGE_SIZE * npages; > > > > Right. >=20 > There are two possible packet buffer (chunks) sizes in a umem, 2K and > 4K on a system with a PAGE_SIZE of 4K. If I remember correctly, and > please correct me if wrong, truesize is used for memory accounting. > But in this code, no kernel memory has been allocated (apart from the > skb). The page is just a part of the umem that has been already > allocated beforehand and by user-space in this case. So what should > truesize be in this case? Do we add 0, chunk_size * i, or the > complicated case of counting exactly how many 4K pages that are used > when the chunk_size is 2K, as two chunks could occupy the same page, > or just the upper bound of PAGE_SIZE * i that is likely a good > approximation in most cases? Just note that there might be other uses > of truesize that I am unaware of that could impact this choice. Truesize is "what amount of memory does this skb occupy with all its fragments, linear space and struct sk_buff itself". The closest it will be to the actual value, the better. In this case, I think adding of chunk_size * i would be enough. (PAGE_SIZE * i can be overwhelming when chunk_size is 2K, especially for setups with PAGE_SIZE > SZ_4K) > > > > > + > > > > > + refcount_add(len, &xs->sk.sk_wmem_alloc); > > > > > + > > > > > + return skb; > > > > > +} > > > > > + > > > > > > Al > > > > Thanks, > > Al Al