From nobody Sun Dec 14 19:55:19 2025 Received: from mail-pj1-f74.google.com (mail-pj1-f74.google.com [209.85.216.74]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id A7C211E2834 for ; Thu, 8 May 2025 00:48:41 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.216.74 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1746665325; cv=none; b=pnKC7WT9UD06O+BMVthfMOHltdVXUGJoC6O1Qt99AECLpA8sw/E4xLQhCGBEFNzIg3BC9h13JNynoTo89SVDBCkilP4hT+y5RxNbnXySebK4Vjqi3BDuM/3swOZNURZRy/fnNeAEITk8Ol/c7qCe7p7Fanf3TETHBtR6Es7qlgo= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1746665325; c=relaxed/simple; bh=3faivae3XYstrRHw5QOydPRDb8eLp/UXyJazTs99n5U=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=Ik+6BXDlKL5LLPbLHlEyTDa47KfIQorEW9Sk2gO5+xGGbGGUKy2Tby1DS5EJc8kINBxl8JyECVgFCjC/bWNM+/hELzZ+KRhZqLt3n7Bh9QkWBTcTcATHkbfH8vJYce4J4FiOPnamtU+ngTzutyDOdwGXiwmCuPnooMu89RVn/G0= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=gRmv4xB/; arc=none smtp.client-ip=209.85.216.74 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--almasrymina.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="gRmv4xB/" Received: by mail-pj1-f74.google.com with SMTP id 98e67ed59e1d1-30a39fa0765so713332a91.3 for ; Wed, 07 May 2025 17:48:41 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1746665321; x=1747270121; darn=vger.kernel.org; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:from:to:cc:subject:date:message-id:reply-to; bh=jdymNESOtoqlSKfOld7O62q+ATDbJOYzEoCPoXvdsNo=; b=gRmv4xB/qXHvXnDcPtV+aPIN7jGW1O2R8vgXKqqALbjJXUR9g5wpokqIU/pIMZcmxm SsUkfldx2nLtkU1HXNPxC0hgJkaZIIkpmqx8nppSWP6AJ9OGVcoqICLl/IVHx+zTs00g 2tT8tYb9hUJbjKbXBh1Mg5xrqpHQIhlBF6zC7wX5E435G2QHGF8m0XEt9CtJsvqyFQ2v CIkWSkIs4P6wbyVhrJYUciNgLhYJpcxbAJJUejd9WGs0Y67bxE/XAevU/1U7V+2+i7gT LajCJX3/ZewZNs04IyM0mpE051QBhreGjPc43pFvHI6Csg64gXut93FSsrquuudeZmJQ 5FaA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1746665321; x=1747270121; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=jdymNESOtoqlSKfOld7O62q+ATDbJOYzEoCPoXvdsNo=; b=YR+mIP+7T9cLmneSscwKv/61QDdkSaVtW0F/X5UH3DwO3/n+7p4LK/w8zLIucc8vK/ N/bBYUy8pezXCSAiLnacj7LCn8TZzPFo0xFa9QRSezZ12dIHiPXMnBywt/XGlhKdTy3e JzxdoHC/MLUe31Rp8ezEnImSmifxo5axSQ19gZSX/QxwRuEzSNM3SBIwmUDiffXysMoH p9lvsqwHnF3t9dVpRtRfI6kZeTt/+zmEXdvTo6zF5a3b9MQg2RXvtfj1/F3PpVfBb2VU 31wjg5NhsMJWX97UH0hTr/GKtLB6Q9HZb5W9OlfK9b703hFISNS1gyubU5h1A3wPXFcm Bprg== X-Forwarded-Encrypted: i=1; AJvYcCVc60qkzXq2IA2b6ZzFib160trzxfb7WK75k6v8sAp+A/3sMdw5azyziy+FwJoCSOBA05795PfuYL0guY8=@vger.kernel.org X-Gm-Message-State: AOJu0YzGheiw2qeSggnxMocxstg/6N0VQIB4hN5MQe/GxqHmr/uS0161 dwB8R67Px+N3H9wmjWde7fDp7/S9K/F/nsd/wiaY8/Q1P40YlXUS36qZchmYhTihbRwBGik63Km fo58SZup9klcAjyPa0paUGA== X-Google-Smtp-Source: AGHT+IGrRc+EZyVOfKps7I5sxHBKTIU92vkDXCFafwp1D3P2RsQ5r5ep19XSR3PvoNGHNnRfQZQE+Iw3q1aaEIC8hA== X-Received: from pjbsp16.prod.google.com ([2002:a17:90b:52d0:b0:2ea:29de:af10]) (user=almasrymina job=prod-delivery.src-stubby-dispatcher) by 2002:a17:90b:38c6:b0:2fe:b174:31fe with SMTP id 98e67ed59e1d1-30aac18622dmr7934268a91.2.1746665321013; Wed, 07 May 2025 17:48:41 -0700 (PDT) Date: Thu, 8 May 2025 00:48:25 +0000 In-Reply-To: <20250508004830.4100853-1-almasrymina@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20250508004830.4100853-1-almasrymina@google.com> X-Mailer: git-send-email 2.49.0.987.g0cc8ee98dc-goog Message-ID: <20250508004830.4100853-6-almasrymina@google.com> Subject: [PATCH net-next v14 5/9] net: add devmem TCP TX documentation From: Mina Almasry To: netdev@vger.kernel.org, linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org, io-uring@vger.kernel.org, virtualization@lists.linux.dev, kvm@vger.kernel.org, linux-kselftest@vger.kernel.org Cc: Mina Almasry , Donald Hunter , Jakub Kicinski , "David S. Miller" , Eric Dumazet , Paolo Abeni , Simon Horman , Jonathan Corbet , Andrew Lunn , Jeroen de Borst , Harshitha Ramamurthy , Kuniyuki Iwashima , Willem de Bruijn , Jens Axboe , Pavel Begunkov , David Ahern , Neal Cardwell , "Michael S. Tsirkin" , Jason Wang , Xuan Zhuo , "=?UTF-8?q?Eugenio=20P=C3=A9rez?=" , Stefan Hajnoczi , Stefano Garzarella , Shuah Khan , sdf@fomichev.me, dw@davidwei.uk, Jamal Hadi Salim , Victor Nogueira , Pedro Tammela , Samiullah Khawaja Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Add documentation outlining the usage and details of the devmem TCP TX API. Signed-off-by: Mina Almasry Acked-by: Stanislav Fomichev --- v5: - Address comments from Stan and Bagas v4: - Mention SO_BINDTODEVICE is recommended (me/Pavel). v2: - Update documentation for iov_base is the dmabuf offset (Stan) --- Documentation/networking/devmem.rst | 150 +++++++++++++++++++++++++++- 1 file changed, 146 insertions(+), 4 deletions(-) diff --git a/Documentation/networking/devmem.rst b/Documentation/networking= /devmem.rst index eb678ca454968..a6cd7236bfbd2 100644 --- a/Documentation/networking/devmem.rst +++ b/Documentation/networking/devmem.rst @@ -62,15 +62,15 @@ More Info https://lore.kernel.org/netdev/20240831004313.3713467-1-almasrymina@go= ogle.com/ =20 =20 -Interface -=3D=3D=3D=3D=3D=3D=3D=3D=3D +RX Interface +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D =20 =20 Example ------- =20 -tools/testing/selftests/net/ncdevmem.c:do_server shows an example of setti= ng up -the RX path of this API. +./tools/testing/selftests/drivers/net/hw/ncdevmem:do_server shows an examp= le of +setting up the RX path of this API. =20 =20 NIC Setup @@ -235,6 +235,148 @@ can be less than the tokens provided by the user in c= ase of: (a) an internal kernel leak bug. (b) the user passed more than 1024 frags. =20 +TX Interface +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D + + +Example +------- + +./tools/testing/selftests/drivers/net/hw/ncdevmem:do_client shows an examp= le of +setting up the TX path of this API. + + +NIC Setup +--------- + +The user must bind a TX dmabuf to a given NIC using the netlink API:: + + struct netdev_bind_tx_req *req =3D NULL; + struct netdev_bind_tx_rsp *rsp =3D NULL; + struct ynl_error yerr; + + *ys =3D ynl_sock_create(&ynl_netdev_family, &yerr); + + req =3D netdev_bind_tx_req_alloc(); + netdev_bind_tx_req_set_ifindex(req, ifindex); + netdev_bind_tx_req_set_fd(req, dmabuf_fd); + + rsp =3D netdev_bind_tx(*ys, req); + + tx_dmabuf_id =3D rsp->id; + + +The netlink API returns a dmabuf_id: a unique ID that refers to this dmabuf +that has been bound. + +The user can unbind the dmabuf from the netdevice by closing the netlink s= ocket +that established the binding. We do this so that the binding is automatica= lly +unbound even if the userspace process crashes. + +Note that any reasonably well-behaved dmabuf from any exporter should work= with +devmem TCP, even if the dmabuf is not actually backed by devmem. An exampl= e of +this is udmabuf, which wraps user memory (non-devmem) in a dmabuf. + +Socket Setup +------------ + +The user application must use MSG_ZEROCOPY flag when sending devmem TCP. D= evmem +cannot be copied by the kernel, so the semantics of the devmem TX are simi= lar +to the semantics of MSG_ZEROCOPY:: + + setsockopt(socket_fd, SOL_SOCKET, SO_ZEROCOPY, &opt, sizeof(opt)); + +It is also recommended that the user binds the TX socket to the same inter= face +the dma-buf has been bound to via SO_BINDTODEVICE:: + + setsockopt(socket_fd, SOL_SOCKET, SO_BINDTODEVICE, ifname, strlen(ifname)= + 1); + + +Sending data +------------ + +Devmem data is sent using the SCM_DEVMEM_DMABUF cmsg. + +The user should create a msghdr where, + +* iov_base is set to the offset into the dmabuf to start sending from +* iov_len is set to the number of bytes to be sent from the dmabuf + +The user passes the dma-buf id to send from via the dmabuf_tx_cmsg.dmabuf_= id. + +The example below sends 1024 bytes from offset 100 into the dmabuf, and 20= 48 +from offset 2000 into the dmabuf. The dmabuf to send from is tx_dmabuf_id:: + + char ctrl_data[CMSG_SPACE(sizeof(struct dmabuf_tx_cmsg))]; + struct dmabuf_tx_cmsg ddmabuf; + struct msghdr msg =3D {}; + struct cmsghdr *cmsg; + struct iovec iov[2]; + + iov[0].iov_base =3D (void*)100; + iov[0].iov_len =3D 1024; + iov[1].iov_base =3D (void*)2000; + iov[1].iov_len =3D 2048; + + msg.msg_iov =3D iov; + msg.msg_iovlen =3D 2; + + msg.msg_control =3D ctrl_data; + msg.msg_controllen =3D sizeof(ctrl_data); + + cmsg =3D CMSG_FIRSTHDR(&msg); + cmsg->cmsg_level =3D SOL_SOCKET; + cmsg->cmsg_type =3D SCM_DEVMEM_DMABUF; + cmsg->cmsg_len =3D CMSG_LEN(sizeof(struct dmabuf_tx_cmsg)); + + ddmabuf.dmabuf_id =3D tx_dmabuf_id; + + *((struct dmabuf_tx_cmsg *)CMSG_DATA(cmsg)) =3D ddmabuf; + + sendmsg(socket_fd, &msg, MSG_ZEROCOPY); + + +Reusing TX dmabufs +------------------ + +Similar to MSG_ZEROCOPY with regular memory, the user should not modify the +contents of the dma-buf while a send operation is in progress. This is bec= ause +the kernel does not keep a copy of the dmabuf contents. Instead, the kernel +will pin and send data from the buffer available to the userspace. + +Just as in MSG_ZEROCOPY, the kernel notifies the userspace of send complet= ions +using MSG_ERRQUEUE:: + + int64_t tstop =3D gettimeofday_ms() + waittime_ms; + char control[CMSG_SPACE(100)] =3D {}; + struct sock_extended_err *serr; + struct msghdr msg =3D {}; + struct cmsghdr *cm; + int retries =3D 10; + __u32 hi, lo; + + msg.msg_control =3D control; + msg.msg_controllen =3D sizeof(control); + + while (gettimeofday_ms() < tstop) { + if (!do_poll(fd)) continue; + + ret =3D recvmsg(fd, &msg, MSG_ERRQUEUE); + + for (cm =3D CMSG_FIRSTHDR(&msg); cm; cm =3D CMSG_NXTHDR(&m= sg, cm)) { + serr =3D (void *)CMSG_DATA(cm); + + hi =3D serr->ee_data; + lo =3D serr->ee_info; + + fprintf(stdout, "tx complete [%d,%d]\n", lo, hi); + } + } + +After the associated sendmsg has been completed, the dmabuf can be reused = by +the userspace. + + Implementation & Caveats =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D =20 --=20 2.49.0.987.g0cc8ee98dc-goog