From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from out30-101.freemail.mail.aliyun.com (out30-101.freemail.mail.aliyun.com [115.124.30.101]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id C747D1A8C09 for ; Thu, 20 Jun 2024 10:40:05 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=115.124.30.101 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1718880008; cv=none; b=ZvxFkYPIttcmODNQ4VRD36XappiAZ8dOwzl/nt/YAS8wgaQPmgC8NKzXlvjaB+VoCI+i07xlouP5KXc+Lr9xL1Q52r6YspneOEL/YdFikLI2iKMUDATQvfRLte83mcZOqAoxWifxNtZ4duxuQRrAwyVgR2e83Y338jgAc6Z0fVI= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1718880008; c=relaxed/simple; bh=Zw9qpO+r1nzf+ORBN6hO/Oqm1MuBeBSQsnFcVxfvzmk=; h=Message-ID:Subject:Date:From:To:Cc:References:In-Reply-To: Content-Type; b=r1UYCE8w0D5dbq6KXbA09topejfILHyH+JYbwrgUnAJf2hBcyql8CPmBpvW1BLdS5bSNqzadv1PFCOru31+LDyxxYTo4JivrG6MQ9leP1yjyEnZjPqOE6N7V+G2r75VO0MxkZwepGB2cV0UDwwbzs5hDZ8nqoD8vJgMjJxx9ox4= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.alibaba.com; spf=pass smtp.mailfrom=linux.alibaba.com; dkim=pass (1024-bit key) header.d=linux.alibaba.com header.i=@linux.alibaba.com header.b=qZH43+Yy; arc=none smtp.client-ip=115.124.30.101 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.alibaba.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=linux.alibaba.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=linux.alibaba.com header.i=@linux.alibaba.com header.b="qZH43+Yy" DKIM-Signature:v=1; a=rsa-sha256; c=relaxed/relaxed; d=linux.alibaba.com; s=default; t=1718880003; h=Message-ID:Subject:Date:From:To:Content-Type; bh=d9K68yXD0uUIPSuPzoHQIO6969QfhsA4Fem+RZyg2dM=; b=qZH43+YynphvetbbUwKMjhnz3TrC1livzteqAgYuloHW1vT6oj8BiOI30lqiDEVLiMFcB3ERw8TtPU4aThAOE95JUIh8SGTzh5FquqlTPEsQCXINSSL9HFh8Ryx8h/YhQzgdYFvbN/wnvBM0U9h8JyrukgKVowc3pRP/smGvWFI= X-Alimail-AntiSpam:AC=PASS;BC=-1|-1;BR=01201311R151e4;CH=green;DM=||false|;DS=||;FP=0|-1|-1|-1|0|-1|-1|-1;HT=maildocker-contentspam033045075189;MF=hengqi@linux.alibaba.com;NM=1;PH=DS;RN=16;SR=0;TI=SMTPD_---0W8s9bm8_1718880002; Received: from localhost(mailfrom:hengqi@linux.alibaba.com fp:SMTPD_---0W8s9bm8_1718880002) by smtp.aliyun-inc.com; Thu, 20 Jun 2024 18:40:03 +0800 Message-ID: <1718879929.7095907-12-hengqi@linux.alibaba.com> Subject: Re: [PATCH 2/2] virtio_net: fixing XDP for fully checksummed packets handling Date: Thu, 20 Jun 2024 18:38:49 +0800 From: Heng Qi To: Heng Qi Cc: Jason Wang , netdev@vger.kernel.org, virtualization@lists.linux.dev, Thomas Huth , Xuan Zhuo , =?utf-8?q?Eugenio_P=C3=A9rez?= , Eric Dumazet , "David S. Miller" , Jakub Kicinski , Paolo Abeni , Alexei Starovoitov , Daniel Borkmann , Jesper Dangaard Brouer , John Fastabend , "Michael S. Tsirkin" References: <20240617131524.63662-1-hengqi@linux.alibaba.com> <20240617131524.63662-3-hengqi@linux.alibaba.com> <1718680517.8370645-12-hengqi@linux.alibaba.com> <1718875728.9338605-7-hengqi@linux.alibaba.com> <20240620061710-mutt-send-email-mst@kernel.org> <1718879236.1007793-10-hengqi@linux.alibaba.com> In-Reply-To: <1718879236.1007793-10-hengqi@linux.alibaba.com> Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Precedence: bulk X-Mailing-List: netdev@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: On Thu, 20 Jun 2024 18:27:16 +0800, Heng Qi wrot= e: > On Thu, 20 Jun 2024 06:19:01 -0400, "Michael S. Tsirkin" = wrote: > > On Thu, Jun 20, 2024 at 05:28:48PM +0800, Heng Qi wrote: > > > On Thu, 20 Jun 2024 16:33:35 +0800, Jason Wang = wrote: > > > > On Tue, Jun 18, 2024 at 11:17=E2=80=AFAM Heng Qi wrote: > > > > > > > > > > On Tue, 18 Jun 2024 11:10:26 +0800, Jason Wang wrote: > > > > > > On Mon, Jun 17, 2024 at 9:15=E2=80=AFPM Heng Qi wrote: > > > > > > > > > > > > > > The XDP program can't correctly handle partially checksummed > > > > > > > packets, but works fine with fully checksummed packets. > > > > > > > > > > > > Not sure this is ture, if I was not wrong, XDP can try to calcu= late checksum. > > > > > > > > > > XDP's interface serves a full checksum, > > > >=20 > > > > What do you mean by "serve" here? I mean, XDP can calculate the > > > > checksum and fill it in the packet by itself. > > > >=20 > > >=20 > > > Yes, XDP can parse and calculate checksums for all packets. > > > However, the bpf_csum_diff and bpf_l4_csum_replace APIs provided by X= DP assume > > > that the packets being processed are fully checksumed packets. That i= s, > > > after the XDP program modified the packets, the incremental checksum = can be > > > calculated (for example, samples/bpf/tcbpf1_kern.c, samples/bpf/test_= lwt_bpf.c). > > >=20 > > > Therefore, partially checksummed packets cannot be processed normally= in these > > > examples and need to be discarded. > > >=20 > > > > > and this is why we disabled the > > > > > offloading of VIRTIO_NET_F_GUEST_CSUM when loading XDP. > > > >=20 > > > > If we trust the device to disable VIRTIO_NET_F_GUEST_CSUM, any reas= on > > > > to check VIRTIO_NET_HDR_F_NEEDS_CSUM again in the receive path? > > >=20 > > > There doesn't seem to be a mandatory constraint in the spec that devi= ces that > > > haven't negotiated VIRTIO_NET_F_GUEST_CSUM cannot set NEEDS_CSUM bit,= so I check this. > > >=20 > > > Thanks. > >=20 > > The spec says: > >=20 > > \item If the VIRTIO_NET_F_GUEST_CSUM feature was negotiated, the > > VIRTIO_NET_HDR_F_NEEDS_CSUM bit in \field{flags} can be > > set: if so, the packet checksum at offset \field{csum_offset}=20 > > from \field{csum_start} and any preceding checksums > > have been validated. The checksum on the packet is incomplete and > > if bit VIRTIO_NET_HDR_F_RSC_INFO is not set in \field{flags}, > > then \field{csum_start} and \field{csum_offset} indicate how to calcu= late it > > (see Packet Transmission point 1). > >=20 > >=20 > > So yes, NEEDS_CSUM without VIRTIO_NET_F_GUEST_CSUM is at best undefined. > > Please do not try to use it unless VIRTIO_NET_F_GUEST_CSUM is set. >=20 > I've seen it before, but thought something like > "The device MUST NOT set the NEEDS_CSUM bit if GUEST_CSUM has not been n= egotiated" > would be clearer. >=20 > Furthermore, it is still possible for a malicious device to set the bit. Hint: We previously checked and used DATA_VALID and NEEDS_CSUM bits, but never ch= ecked to see if GUEST_CSUM was negotiated. >=20 > Thanks. >=20 > >=20 > > And if you want to be flexible, ignore it unless VIRTIO_NET_F_GUEST_CSUM > > has been negotiated. > >=20 > >=20 > >=20 > >=20 > >=20 > > > >=20 > > > > > > > > > > Thanks. > > > >=20 > > > > Thanks > > > >=20 > > > > > > > > > > > > > > > > > Thanks > > > > > > > > > > > > > If the > > > > > > > device has already validated fully checksummed packets, then > > > > > > > the driver doesn't need to re-validate them, saving CPU resou= rces. > > > > > > > > > > > > > > Additionally, the driver does not drop all partially checksum= med > > > > > > > packets when VIRTIO_NET_F_GUEST_CSUM is not negotiated. This = is > > > > > > > not a bug, as the driver has always done this. > > > > > > > > > > > > > > Fixes: 436c9453a1ac ("virtio-net: keep vnet header zeroed aft= er processing XDP") > > > > > > > Signed-off-by: Heng Qi > > > > > > > --- > > > > > > > drivers/net/virtio_net.c | 20 +++++++++++++++++++- > > > > > > > 1 file changed, 19 insertions(+), 1 deletion(-) > > > > > > > > > > > > > > diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_ne= t.c > > > > > > > index aa70a7ed8072..ea10db9a09fa 100644 > > > > > > > --- a/drivers/net/virtio_net.c > > > > > > > +++ b/drivers/net/virtio_net.c > > > > > > > @@ -1360,6 +1360,10 @@ static struct sk_buff *receive_small_x= dp(struct net_device *dev, > > > > > > > if (unlikely(hdr->hdr.gso_type)) > > > > > > > goto err_xdp; > > > > > > > > > > > > > > + /* Partially checksummed packets must be dropped. */ > > > > > > > + if (unlikely(hdr->hdr.flags & VIRTIO_NET_HDR_F_NEEDS_= CSUM)) > > > > > > > + goto err_xdp; > > > > > > > + > > > > > > > buflen =3D SKB_DATA_ALIGN(GOOD_PACKET_LEN + headroom)= + > > > > > > > SKB_DATA_ALIGN(sizeof(struct skb_shared_info)= ); > > > > > > > > > > > > > > @@ -1677,6 +1681,10 @@ static void *mergeable_xdp_get_buf(str= uct virtnet_info *vi, > > > > > > > if (unlikely(hdr->hdr.gso_type)) > > > > > > > return NULL; > > > > > > > > > > > > > > + /* Partially checksummed packets must be dropped. */ > > > > > > > + if (unlikely(hdr->hdr.flags & VIRTIO_NET_HDR_F_NEEDS_= CSUM)) > > > > > > > + return NULL; > > > > > > > + > > > > > > > /* Now XDP core assumes frag size is PAGE_SIZE, but b= uffers > > > > > > > * with headroom may add hole in truesize, which > > > > > > > * make their length exceed PAGE_SIZE. So we disabled= the > > > > > > > @@ -1943,6 +1951,7 @@ static void receive_buf(struct virtnet_= info *vi, struct receive_queue *rq, > > > > > > > struct net_device *dev =3D vi->dev; > > > > > > > struct sk_buff *skb; > > > > > > > struct virtio_net_common_hdr *hdr; > > > > > > > + u8 flags; > > > > > > > > > > > > > > if (unlikely(len < vi->hdr_len + ETH_HLEN)) { > > > > > > > pr_debug("%s: short packet %i\n", dev->name, = len); > > > > > > > @@ -1951,6 +1960,15 @@ static void receive_buf(struct virtnet= _info *vi, struct receive_queue *rq, > > > > > > > return; > > > > > > > } > > > > > > > > > > > > > > + /* 1. Save the flags early, as the XDP program might = overwrite them. > > > > > > > + * These flags ensure packets marked as VIRTIO_NET_HD= R_F_DATA_VALID > > > > > > > + * stay valid after XDP processing. > > > > > > > + * 2. XDP doesn't work with partially checksummed pac= kets (refer to > > > > > > > + * virtnet_xdp_set()), so packets marked as > > > > > > > + * VIRTIO_NET_HDR_F_NEEDS_CSUM get dropped during XDP= processing. > > > > > > > + */ > > > > > > > + flags =3D ((struct virtio_net_common_hdr *)buf)->hdr.= flags; > > > > > > > + > > > > > > > if (vi->mergeable_rx_bufs) > > > > > > > skb =3D receive_mergeable(dev, vi, rq, buf, c= tx, len, xdp_xmit, > > > > > > > stats); > > > > > > > @@ -1966,7 +1984,7 @@ static void receive_buf(struct virtnet_= info *vi, struct receive_queue *rq, > > > > > > > if (dev->features & NETIF_F_RXHASH && vi->has_rss_has= h_report) > > > > > > > virtio_skb_set_hash(&hdr->hash_v1_hdr, skb); > > > > > > > > > > > > > > - if (hdr->hdr.flags & VIRTIO_NET_HDR_F_DATA_VALID) > > > > > > > + if (flags & VIRTIO_NET_HDR_F_DATA_VALID) > > > > > > > skb->ip_summed =3D CHECKSUM_UNNECESSARY; > > > > > > > > > > > > > > if (virtio_net_hdr_to_skb(skb, &hdr->hdr, > > > > > > > -- > > > > > > > 2.32.0.3.g01195cf9f > > > > > > > > > > > > > > > > > > > > > >=20 > >=20 >=20