[{"id":3674951,"web_url":"http://patchwork.ozlabs.org/comment/3674951/","msgid":"","list_archive_url":null,"date":"2026-04-08T12:07:06","subject":"Re: [RFC v4 4/5] chardev/socket: add AF_PACKET inject path","submitter":{"id":2694,"url":"http://patchwork.ozlabs.org/api/people/2694/","name":"Daniel P. Berrangé","email":"berrange@redhat.com"},"content":"On Tue, Apr 07, 2026 at 01:05:51PM +0800, Cindy Lu wrote:\n> Add the AF_PACKET inject write path for socket chardevs. When a socket\n> backend is opened with af-packet-mode=inject, tcp_chr_write() no longer\n> sends the redirector stream framing through QIOChannel. Instead it\n> parses the existing 4-byte length header, accumulates one complete\n> packet, and frame on the AF_PACKET fd.\n> \n> Signed-off-by: Cindy Lu \n> ---\n> chardev/char-socket.c | 148 ++++++++++++++++++++++++++++++++++++++++++\n> 1 file changed, 148 insertions(+)\n> \n> diff --git a/chardev/char-socket.c b/chardev/char-socket.c\n> index c710fdb497..45d06fda8f 100644\n> --- a/chardev/char-socket.c\n> +++ b/chardev/char-socket.c\n> @@ -108,11 +108,159 @@ static void tcp_chr_accept(QIONetListener *listener,\n> static int tcp_chr_read_poll(void *opaque);\n> static void tcp_chr_disconnect_locked(Chardev *chr);\n> \n> +#define TCP_CHARDEV_AF_PACKET_MAX_FRAME_SIZE 65536\n> +\n> +static bool tcp_chr_uses_af_packet_inject(SocketChardev *s)\n> +{\n> + return s->is_af_packet &&\n> + s->af_packet_mode_set &&\n> + s->af_packet_mode == CHARDEV_SOCKET_AF_PACKET_MODE_INJECT;\n> +}\n> +\n> +static ssize_t tcp_chr_send_af_packet(SocketChardev *s,\n> + const uint8_t *buf,\n> + size_t len)\n> +{\n> +#ifdef CONFIG_LINUX\n> + struct iovec iov = {\n> + .iov_base = (void *)buf,\n> + .iov_len = len,\n> + };\n> + struct msghdr msg = {\n> + .msg_iov = &iov,\n> + .msg_iovlen = 1,\n> + };\n> + ssize_t ret;\n> +\n> + if (!s->sioc || s->sioc->localAddr.ss_family != AF_PACKET) {\n> + errno = ENOTSOCK;\n> + return -1;\n> + }\n> +\n> + do {\n> + ret = sendmsg(s->sioc->fd, &msg, 0);\n> + } while (ret < 0 && errno == EINTR);\n> +\n> + return ret;\n> +#else\n> + errno = EPROTONOSUPPORT;\n> + return -1;\n> +#endif\n> +}\n> +\n> +static bool tcp_chr_af_packet_prepare_send(SocketChardev *s, uint32_t frame_len)\n> +{\n> + if (frame_len > TCP_CHARDEV_AF_PACKET_MAX_FRAME_SIZE) {\n> + errno = EMSGSIZE;\n> + return false;\n> + }\n> +\n> + if (frame_len == 0) {\n> + s->af_packet_send_len = 0;\n> + s->af_packet_send_offset = 0;\n> + s->af_packet_send_len_bytes = 0;\n> + return true;\n> + }\n> +\n> + if (s->af_packet_send_buf_size < frame_len) {\n> + s->af_packet_send_buf = g_realloc(s->af_packet_send_buf, frame_len);\n> + s->af_packet_send_buf_size = frame_len;\n> + }\n> +\n> + s->af_packet_send_len = frame_len;\n> + s->af_packet_send_offset = 0;\n> + s->af_packet_send_len_bytes = sizeof(s->af_packet_send_len_buf);\n> + return true;\n> +}\n> +\n> +static int tcp_chr_inject_af_packet(Chardev *chr,\n> + SocketChardev *s,\n> + const uint8_t *buf,\n> + int len)\n> +{\n> + size_t offset = 0;\n> + uint32_t frame_len_be;\n> +\n> + while (offset < len) {\n> + size_t copy;\n> +\n> + if (s->af_packet_send_len_bytes < sizeof(s->af_packet_send_len_buf)) {\n> + copy = MIN(sizeof(s->af_packet_send_len_buf) -\n> + s->af_packet_send_len_bytes,\n> + (size_t)len - offset);\n> + memcpy(s->af_packet_send_len_buf + s->af_packet_send_len_bytes,\n> + buf + offset, copy);\n> + s->af_packet_send_len_bytes += copy;\n> + offset += copy;\n> +\n> + if (s->af_packet_send_len_bytes <\n> + sizeof(s->af_packet_send_len_buf)) {\n> + continue;\n> + }\n> +\n> + memcpy(&frame_len_be, s->af_packet_send_len_buf,\n> + sizeof(frame_len_be));\n> + if (!tcp_chr_af_packet_prepare_send(s, ntohl(frame_len_be))) {\n> + return -1;\n> + }\n> + if (s->af_packet_send_len == 0) {\n> + continue;\n> + }\n> + }\n> +\n> + copy = MIN(s->af_packet_send_len - s->af_packet_send_offset,\n> + (size_t)len - offset);\n> + memcpy(s->af_packet_send_buf + s->af_packet_send_offset,\n> + buf + offset, copy);\n> + s->af_packet_send_offset += copy;\n> + offset += copy;\n> +\n> + if (s->af_packet_send_offset == s->af_packet_send_len) {\n> + ssize_t ret;\n> +\n> + ret = tcp_chr_send_af_packet(s, s->af_packet_send_buf,\n> + s->af_packet_send_len);\n> +\n> + if (ret < 0) {\n> + if (errno == EAGAIN || errno == EWOULDBLOCK) {\n> + return -1;\n> + }\n> + if (tcp_chr_read_poll(chr) <= 0) {\n> + trace_chr_socket_poll_err(chr, chr->label);\n> + tcp_chr_disconnect_locked(chr);\n> + }\n> + return -1;\n> + }\n> +\n> + if (ret != (ssize_t)s->af_packet_send_len) {\n> + if (ret >= 0) {\n> + errno = EIO;\n> + }\n> + if (tcp_chr_read_poll(chr) <= 0) {\n> + trace_chr_socket_poll_err(chr, chr->label);\n> + tcp_chr_disconnect_locked(chr);\n> + }\n> + return -1;\n> + }\n> +\n> + s->af_packet_send_len = 0;\n> + s->af_packet_send_offset = 0;\n> + s->af_packet_send_len_bytes = 0;\n> + }\n> + }\n> +\n> + return len;\n> +}\n> +\n> /* Called with chr_write_lock held. */\n> static int tcp_chr_write(Chardev *chr, const uint8_t *buf, int len)\n> {\n> SocketChardev *s = SOCKET_CHARDEV(chr);\n> \n> + if (tcp_chr_uses_af_packet_inject(s)) {\n> + return tcp_chr_inject_af_packet(chr, s, buf, len);\n> + }\n> +\n\nThis code is pretty unpleasant, completely bypassing all of the\nnormal I/O path in the chardev, and completely ignoring the\nQIOChannel too, just poking the socket directly. Essentially\nthis shares nothing in common with the socket chardev functionality.\n\nIf we do want to have AF_PACKET support in the socket chardev then\nIMHO all this buffer parsing code needs to be in the netfilter\nlayer instead. The chardev should just accept a single packet\nbuffer at a time, such that it can directly pass it on to the\nnormal qio_channel_write API which will call sendmsg.\n\n> if (s->state == TCP_CHARDEV_STATE_CONNECTED) {\n> int ret = io_channel_send_full(s->ioc, buf, len,\n> s->write_msgfds,\n> -- \n> 2.52.0\n> \n> \n\nWith regards,\nDaniel","headers":{"Return-Path":"","X-Original-To":"incoming@patchwork.ozlabs.org","Delivered-To":"patchwork-incoming@legolas.ozlabs.org","Authentication-Results":["legolas.ozlabs.org;\n\tdkim=pass (1024-bit key;\n unprotected) header.d=redhat.com header.i=@redhat.com header.a=rsa-sha256\n header.s=mimecast20190719 header.b=Grr7xMyW;\n\tdkim-atps=neutral","legolas.ozlabs.org;\n spf=pass (sender SPF authorized) smtp.mailfrom=nongnu.org\n (client-ip=209.51.188.17; helo=lists.gnu.org;\n envelope-from=qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org;\n receiver=patchwork.ozlabs.org)"],"Received":["from lists.gnu.org (lists1p.gnu.org [209.51.188.17])\n\t(using TLSv1.2 with cipher ECDHE-ECDSA-AES256-GCM-SHA384 (256/256 bits))\n\t(No client certificate requested)\n\tby legolas.ozlabs.org (Postfix) with ESMTPS id 4frY9q62W3z1xv0\n\tfor ; Thu, 09 Apr 2026 05:32:31 +1000 (AEST)","from localhost ([::1] helo=lists1p.gnu.org)\n\tby lists.gnu.org with esmtp (Exim 4.90_1)\n\t(envelope-from )\n\tid 1wAYXz-00021h-2n; Wed, 08 Apr 2026 15:26:47 -0400","from eggs.gnu.org ([2001:470:142:3::10])\n by lists1p.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)\n (Exim 4.90_1) (envelope-from )\n id 1wAY7V-0007if-Pa\n for qemu-devel@nongnu.org; Wed, 08 Apr 2026 14:59:21 -0400","from us-smtp-delivery-124.mimecast.com ([170.10.129.124])\n by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256)\n (Exim 4.90_1) (envelope-from )\n id 1wARgi-0001R8-8o\n for qemu-devel@nongnu.org; Wed, 08 Apr 2026 08:07:18 -0400","from mx-prod-mc-03.mail-002.prod.us-west-2.aws.redhat.com\n (ec2-54-186-198-63.us-west-2.compute.amazonaws.com [54.186.198.63]) by\n relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3,\n cipher=TLS_AES_256_GCM_SHA384) id us-mta-81-yZlxBYgrPhianXDoxlMUkA-1; Wed,\n 08 Apr 2026 08:07:13 -0400","from mx-prod-int-05.mail-002.prod.us-west-2.aws.redhat.com\n (mx-prod-int-05.mail-002.prod.us-west-2.aws.redhat.com [10.30.177.17])\n (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits)\n key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest\n SHA256)\n (No client certificate requested)\n by mx-prod-mc-03.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTPS\n id 198F7195606B; Wed, 8 Apr 2026 12:07:11 +0000 (UTC)","from redhat.com (headnet01.pony-001.prod.iad2.dc.redhat.com\n [10.2.32.101])\n by mx-prod-int-05.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with\n ESMTPS\n id 387371953944; Wed, 8 Apr 2026 12:07:09 +0000 (UTC)"],"DKIM-Signature":"v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com;\n s=mimecast20190719; t=1775650034;\n h=from:from:reply-to:reply-to:subject:subject:date:date:\n message-id:message-id:to:to:cc:cc:mime-version:mime-version:\n content-type:content-type:in-reply-to:in-reply-to: references:references;\n bh=0zipxTMFRRf9pxRki22nji35TS0J1pUw/OKrbIBsgQA=;\n b=Grr7xMyW9S41hvzLQdYFRp/K68clQWgS7GyCvfF2S0dymqv2soXITbkuJvRMFRaFU6am+i\n XGvHci+/QD+dZrbaZgXFDFTwS5arxGGn5SuldeKKSovmH0JwiGTBtmZoIPQPXnSWlWSpji\n iipitzUtc7f5slGjJLMyk+vN6bwWjx4=","X-MC-Unique":"yZlxBYgrPhianXDoxlMUkA-1","X-Mimecast-MFC-AGG-ID":"yZlxBYgrPhianXDoxlMUkA_1775650031","Date":"Wed, 8 Apr 2026 13:07:06 +0100","From":"Daniel =?utf-8?b?UC4gQmVycmFuZ8Op?= ","To":"Cindy Lu ","Cc":"mst@redhat.com, jasowang@redhat.com, zhangckid@gmail.com,\n lizhijian@fujitsu.com, jmarcin@redhat.com, qemu-devel@nongnu.org","Subject":"Re: [RFC v4 4/5] chardev/socket: add AF_PACKET inject path","Message-ID":"","References":"<20260407050818.2249570-1-lulu@redhat.com>\n <20260407050818.2249570-5-lulu@redhat.com>","MIME-Version":"1.0","Content-Type":"text/plain; charset=utf-8","Content-Disposition":"inline","In-Reply-To":"<20260407050818.2249570-5-lulu@redhat.com>","User-Agent":"Mutt/2.2.14 (2025-02-20)","X-Scanned-By":"MIMEDefang 3.0 on 10.30.177.17","Received-SPF":"pass client-ip=170.10.129.124;\n envelope-from=berrange@redhat.com;\n helo=us-smtp-delivery-124.mimecast.com","X-Spam_score_int":"7","X-Spam_score":"0.7","X-Spam_bar":"/","X-Spam_report":"(0.7 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.54,\n DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1,\n RCVD_IN_DNSWL_NONE=-0.0001, RCVD_IN_MSPIKE_H4=0.001, RCVD_IN_MSPIKE_WL=0.001,\n RCVD_IN_SBL_CSS=3.335, RCVD_IN_VALIDITY_RPBL_BLOCKED=0.001,\n RCVD_IN_VALIDITY_SAFE_BLOCKED=0.001, SPF_HELO_PASS=-0.001,\n SPF_PASS=-0.001 autolearn=no autolearn_force=no","X-Spam_action":"no action","X-BeenThere":"qemu-devel@nongnu.org","X-Mailman-Version":"2.1.29","Precedence":"list","List-Id":"qemu development ","List-Unsubscribe":",\n ","List-Archive":"","List-Post":"","List-Help":"","List-Subscribe":",\n ","Reply-To":"Daniel =?utf-8?b?UC4gQmVycmFuZ8Op?= ","Errors-To":"qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org","Sender":"qemu-devel-bounces+incoming=patchwork.ozlabs.org@nongnu.org"}}]