From patchwork Wed Jan 27 17:01:15 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Loic Poulain X-Patchwork-Id: 12050581 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-18.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER, INCLUDES_PATCH,MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED, USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 39E82C433DB for ; Wed, 27 Jan 2021 16:55:10 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id DCDB760187 for ; Wed, 27 Jan 2021 16:55:09 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S233438AbhA0Qyw (ORCPT ); Wed, 27 Jan 2021 11:54:52 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:47540 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S233925AbhA0QyT (ORCPT ); Wed, 27 Jan 2021 11:54:19 -0500 Received: from mail-wr1-x433.google.com (mail-wr1-x433.google.com [IPv6:2a00:1450:4864:20::433]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 93D57C061573 for ; Wed, 27 Jan 2021 08:53:38 -0800 (PST) Received: by mail-wr1-x433.google.com with SMTP id p15so2625146wrq.8 for ; Wed, 27 Jan 2021 08:53:38 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=from:to:cc:subject:date:message-id; bh=Bfd8ZrqYLleMnZ9WSg0Zx5WTuCO4cMXNGVPsnFmssQw=; b=HXICIs76q4gkZ7vHbuwRlFDK7GVp8XMiqzJaGBebJ7wQLQKFWEp7MmoVkOL//Wv1KL W/qsHQn/vEvI2jwC36O5pU8jLDdebab+21Ge3buKALH2IDO50ZKEpmss8sDONG0EDveY KdEPGIhWhuD3QJVuN3NGv2Yuv3P6dyR3LlcJWr4bvlV+O48mxi/QG5UrDr3I100QHLuC g7j7CpZu4jLNQbEo84rvDHMEB0Jfkr05f4sSJYR5weA80+CUxPtFeRFPjzH+zg0+GcFJ G0AvZOHnNjjbIbt/798Zrs84ECtty6of7tAwp5EtXioe9OncygaP126Nbh9Y1P7NajHg Ac/A== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id; bh=Bfd8ZrqYLleMnZ9WSg0Zx5WTuCO4cMXNGVPsnFmssQw=; b=WD30Y68/s20o7o54oEvOmUUhRxjtyBFnES/sp+QfPYjQxpLtIXRx+SSRSvDdvhYoC5 sJwkdlP1Q0J3NoVwQeV9H/oTfmAdyun3+7rPrM6wiSpkiOKi0tBIyQYzsvfVqTa7gw+E 6bPk81FrCBsox61ulg0HZbLtAx2EOIAi76dD0ZYg7hs6fzMbF4kwg3fPsCad504CKajj od8hkeGH94JIPqxNcHy+UexUes56cg48LtQNvK6XCCGrfA5xzoT97DtT4ELd5BNhXrLm Sywb0mn2N/CjVQNvoY1uX0rZImmckjVXJpBbiD5CVJkfpUTI+dqG2N0Xq5/umrrXqeGg Sb3Q== X-Gm-Message-State: AOAM532qz/baN/TkVLjVzFKDgtizJ0DKm8/ARoj2kzMb8y3pi6HeQTCN ILFeKW5zcU031ovaYqqTxC98gw== X-Google-Smtp-Source: ABdhPJyBUQfRIC5GgYmK9mKXS3zzuwbUEFbD0VDIFVMKElugPiwQAuBSJ2r7DHJmpU1HWxc8N1HQpg== X-Received: by 2002:a5d:5384:: with SMTP id d4mr11586388wrv.177.1611766416903; Wed, 27 Jan 2021 08:53:36 -0800 (PST) Received: from localhost.localdomain ([88.122.66.28]) by smtp.gmail.com with ESMTPSA id m82sm3077042wmf.29.2021.01.27.08.53.36 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Wed, 27 Jan 2021 08:53:36 -0800 (PST) From: Loic Poulain To: kuba@kernel.org, davem@davemloft.net Cc: netdev@vger.kernel.org, carl.yin@quectel.com, Loic Poulain Subject: [PATCH net-next 1/3] net: mhi: Add RX/TX fixup callbacks Date: Wed, 27 Jan 2021 18:01:15 +0100 Message-Id: <1611766877-16787-1-git-send-email-loic.poulain@linaro.org> X-Mailer: git-send-email 2.7.4 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org MHI can transport different protocols, some are handled at upper level, like IP and QMAP(rmnet/netlink), but others will need to be inside MHI net driver, like mbim. This change adds support for protocol rx/tx fixup callbacks registration, that can be used to encode/decode the targeted protocol. Signed-off-by: Loic Poulain --- drivers/net/mhi_net.c | 70 ++++++++++++++++++++++++++++++++++++++++++--------- 1 file changed, 58 insertions(+), 12 deletions(-) diff --git a/drivers/net/mhi_net.c b/drivers/net/mhi_net.c index a5a214d..aa3a5e0 100644 --- a/drivers/net/mhi_net.c +++ b/drivers/net/mhi_net.c @@ -33,11 +33,24 @@ struct mhi_net_stats { struct mhi_net_dev { struct mhi_device *mdev; struct net_device *ndev; + const struct mhi_net_proto *proto; + void *proto_data; struct delayed_work rx_refill; struct mhi_net_stats stats; u32 rx_queue_sz; }; +struct mhi_net_proto { + int (*init)(struct mhi_net_dev *dev); + struct sk_buff * (*tx_fixup)(struct net_device *ndev, struct sk_buff *skb); + int (*rx_fixup)(struct net_device *ndev, struct sk_buff *skb); +}; + +struct mhi_device_info { + const char *netname; + const struct mhi_net_proto *proto; +}; + static int mhi_ndo_open(struct net_device *ndev) { struct mhi_net_dev *mhi_netdev = netdev_priv(ndev); @@ -67,26 +80,35 @@ static int mhi_ndo_stop(struct net_device *ndev) static int mhi_ndo_xmit(struct sk_buff *skb, struct net_device *ndev) { struct mhi_net_dev *mhi_netdev = netdev_priv(ndev); + const struct mhi_net_proto *proto = mhi_netdev->proto; struct mhi_device *mdev = mhi_netdev->mdev; int err; + if (proto && proto->tx_fixup) { + skb = proto->tx_fixup(mhi_netdev->ndev, skb); + if (unlikely(!skb)) + goto exit_drop; + } + err = mhi_queue_skb(mdev, DMA_TO_DEVICE, skb, skb->len, MHI_EOT); if (unlikely(err)) { net_err_ratelimited("%s: Failed to queue TX buf (%d)\n", ndev->name, err); - - u64_stats_update_begin(&mhi_netdev->stats.tx_syncp); - u64_stats_inc(&mhi_netdev->stats.tx_dropped); - u64_stats_update_end(&mhi_netdev->stats.tx_syncp); - - /* drop the packet */ dev_kfree_skb_any(skb); + goto exit_drop; } if (mhi_queue_is_full(mdev, DMA_TO_DEVICE)) netif_stop_queue(ndev); return NETDEV_TX_OK; + +exit_drop: + u64_stats_update_begin(&mhi_netdev->stats.tx_syncp); + u64_stats_inc(&mhi_netdev->stats.tx_dropped); + u64_stats_update_end(&mhi_netdev->stats.tx_syncp); + + return NETDEV_TX_OK; } static void mhi_ndo_get_stats64(struct net_device *ndev, @@ -137,6 +159,7 @@ static void mhi_net_dl_callback(struct mhi_device *mhi_dev, struct mhi_result *mhi_res) { struct mhi_net_dev *mhi_netdev = dev_get_drvdata(&mhi_dev->dev); + const struct mhi_net_proto *proto = mhi_netdev->proto; struct sk_buff *skb = mhi_res->buf_addr; int remaining; @@ -171,7 +194,11 @@ static void mhi_net_dl_callback(struct mhi_device *mhi_dev, } skb_put(skb, mhi_res->bytes_xferd); - netif_rx(skb); + + if (proto && proto->rx_fixup) + proto->rx_fixup(mhi_netdev->ndev, skb); + else + netif_rx(skb); } /* Refill if RX buffers queue becomes low */ @@ -255,14 +282,14 @@ static struct device_type wwan_type = { static int mhi_net_probe(struct mhi_device *mhi_dev, const struct mhi_device_id *id) { - const char *netname = (char *)id->driver_data; + const struct mhi_device_info *info = (struct mhi_device_info *)id->driver_data; struct device *dev = &mhi_dev->dev; struct mhi_net_dev *mhi_netdev; struct net_device *ndev; int err; - ndev = alloc_netdev(sizeof(*mhi_netdev), netname, NET_NAME_PREDICTABLE, - mhi_net_setup); + ndev = alloc_netdev(sizeof(*mhi_netdev), info->netname, + NET_NAME_PREDICTABLE, mhi_net_setup); if (!ndev) return -ENOMEM; @@ -270,6 +297,7 @@ static int mhi_net_probe(struct mhi_device *mhi_dev, dev_set_drvdata(dev, mhi_netdev); mhi_netdev->ndev = ndev; mhi_netdev->mdev = mhi_dev; + mhi_netdev->proto = info->proto; SET_NETDEV_DEV(ndev, &mhi_dev->dev); SET_NETDEV_DEVTYPE(ndev, &wwan_type); @@ -289,8 +317,16 @@ static int mhi_net_probe(struct mhi_device *mhi_dev, if (err) goto out_err; + if (mhi_netdev->proto) { + err = mhi_netdev->proto->init(mhi_netdev); + if (err) + goto out_err_proto; + } + return 0; +out_err_proto: + unregister_netdev(ndev); out_err: free_netdev(ndev); return err; @@ -307,9 +343,19 @@ static void mhi_net_remove(struct mhi_device *mhi_dev) free_netdev(mhi_netdev->ndev); } +static const struct mhi_device_info mhi_hwip0 = { + .netname = "mhi_hwip%d", +}; + +static const struct mhi_device_info mhi_swip0 = { + .netname = "mhi_swip%d", +}; + static const struct mhi_device_id mhi_net_id_table[] = { - { .chan = "IP_HW0", .driver_data = (kernel_ulong_t)"mhi_hwip%d" }, - { .chan = "IP_SW0", .driver_data = (kernel_ulong_t)"mhi_swip%d" }, + /* Hardware accelerated data PATH (to modem IPA), protocol agnostic */ + { .chan = "IP_HW0", .driver_data = (kernel_ulong_t)&mhi_hwip0 }, + /* Software data PATH (to modem CPU) */ + { .chan = "IP_SW0", .driver_data = (kernel_ulong_t)&mhi_swip0 }, {} }; MODULE_DEVICE_TABLE(mhi, mhi_net_id_table); From patchwork Wed Jan 27 17:01:16 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Loic Poulain X-Patchwork-Id: 12050585 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-18.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER, INCLUDES_PATCH,MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED, USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id F3D2EC433E9 for ; Wed, 27 Jan 2021 16:55:11 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 8D00F64D9E for ; Wed, 27 Jan 2021 16:55:11 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S234081AbhA0QzE (ORCPT ); Wed, 27 Jan 2021 11:55:04 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:47544 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S233825AbhA0QyU (ORCPT ); Wed, 27 Jan 2021 11:54:20 -0500 Received: from mail-wm1-x32f.google.com (mail-wm1-x32f.google.com [IPv6:2a00:1450:4864:20::32f]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 6A531C061574 for ; Wed, 27 Jan 2021 08:53:39 -0800 (PST) Received: by mail-wm1-x32f.google.com with SMTP id e15so2286612wme.0 for ; Wed, 27 Jan 2021 08:53:39 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=5LKEwNd0woDrm/FI4/GE1OIO9In2ycXgwHk8R+XlR18=; b=y/mZU9k/xDVKhuFr7ASa4Kua81/qVwswTjEHpFQDQwxWO5WoN9hQAOMMLgJcSBQq6V pPyexjHsSGy2jmJ70Y43xqoZ+8KStH6NYEMLvju+FGSJA44npkW4m9woiZq8VcbIBkj/ 1PW/GsZG3M+Iw/1sUIQ0AC+pdv2nOJLzUzQ/nNPXOEyzVWfMLOESIHJdpGOopLBaZrDE xQdjvdoWd4KNWsgO1LHcPx7ceEXw4G3t3QUd86nty/2ZpEuX12mv4qv87GroUi8t/3rb DFw9Suw/saL4LZqB5iNs5FzxcMyq4RWC3+LaeOaflozi1HsgfWfOU7iK+J1OldhT8Tmx 1kLA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=5LKEwNd0woDrm/FI4/GE1OIO9In2ycXgwHk8R+XlR18=; b=F4u2Q8a5K+6zO9YlQFJNWjLIg8a/Hqk5DQZDRPG3sTqZwEY/R363puOuZqLOjfP0E/ dqr0R9Rxa6T3KJLKwOxpNeikEzEBmCbpUoqGx4sMQIApgS/Ojqvvj5rVf//ERW1yZO22 fMm14os0CVQvrGzTeHuWW7gv58cEbACdwK0JtlYpa2NC08PoEv1vNxLsjQFG3A8vxt+j hVW3Eutw5jJIZLZBjgRAdgFoADS9xyOERVgUQ6lwJHfnjd22/ok+JB9tnedumnO9zzFu x+d7Gi4eE364kVEKICzzQwtCgHwGJr7TuLxzCf97WTHQkq2YiAXFcD9wTaVoO/NGUkgb pTLQ== X-Gm-Message-State: AOAM5327UXAOSdWH1kmw7zf0DAWxAkEkR0ElFXkep8/DiJldrJ/OoEwG FDOabU0NtNNTcsUSYGrQo69mJg== X-Google-Smtp-Source: ABdhPJyalz3FU9EWSttgBFUd98dJQEaadHPMuh738dHskmN/SKqurJJheFEMbd14dqNPPG5b/tn6eA== X-Received: by 2002:a1c:a90f:: with SMTP id s15mr5081210wme.154.1611766418020; Wed, 27 Jan 2021 08:53:38 -0800 (PST) Received: from localhost.localdomain ([88.122.66.28]) by smtp.gmail.com with ESMTPSA id m82sm3077042wmf.29.2021.01.27.08.53.36 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Wed, 27 Jan 2021 08:53:37 -0800 (PST) From: Loic Poulain To: kuba@kernel.org, davem@davemloft.net Cc: netdev@vger.kernel.org, carl.yin@quectel.com, Loic Poulain Subject: [PATCH net-next 2/3] net: mhi: Add dedicated folder Date: Wed, 27 Jan 2021 18:01:16 +0100 Message-Id: <1611766877-16787-2-git-send-email-loic.poulain@linaro.org> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611766877-16787-1-git-send-email-loic.poulain@linaro.org> References: <1611766877-16787-1-git-send-email-loic.poulain@linaro.org> Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org Create a dedicated mhi directory for mhi-net, mhi-net is going to be split into differente files (for additional protocol support). Signed-off-by: Loic Poulain --- drivers/net/Makefile | 2 +- drivers/net/mhi/Makefile | 3 + drivers/net/mhi/net.c | 379 +++++++++++++++++++++++++++++++++++++++++++++++ drivers/net/mhi_net.c | 379 ----------------------------------------------- 4 files changed, 383 insertions(+), 380 deletions(-) create mode 100644 drivers/net/mhi/Makefile create mode 100644 drivers/net/mhi/net.c delete mode 100644 drivers/net/mhi_net.c diff --git a/drivers/net/Makefile b/drivers/net/Makefile index 36e2e41..f4990ff 100644 --- a/drivers/net/Makefile +++ b/drivers/net/Makefile @@ -36,7 +36,7 @@ obj-$(CONFIG_GTP) += gtp.o obj-$(CONFIG_NLMON) += nlmon.o obj-$(CONFIG_NET_VRF) += vrf.o obj-$(CONFIG_VSOCKMON) += vsockmon.o -obj-$(CONFIG_MHI_NET) += mhi_net.o +obj-$(CONFIG_MHI_NET) += mhi/ # # Networking Drivers diff --git a/drivers/net/mhi/Makefile b/drivers/net/mhi/Makefile new file mode 100644 index 0000000..0acf989 --- /dev/null +++ b/drivers/net/mhi/Makefile @@ -0,0 +1,3 @@ +obj-$(CONFIG_MHI_NET) += mhi_net.o + +mhi_net-y := net.o diff --git a/drivers/net/mhi/net.c b/drivers/net/mhi/net.c new file mode 100644 index 0000000..aa3a5e0 --- /dev/null +++ b/drivers/net/mhi/net.c @@ -0,0 +1,379 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* MHI Network driver - Network over MHI bus + * + * Copyright (C) 2020 Linaro Ltd + */ + +#include +#include +#include +#include +#include +#include +#include + +#define MHI_NET_MIN_MTU ETH_MIN_MTU +#define MHI_NET_MAX_MTU 0xffff +#define MHI_NET_DEFAULT_MTU 0x4000 + +struct mhi_net_stats { + u64_stats_t rx_packets; + u64_stats_t rx_bytes; + u64_stats_t rx_errors; + u64_stats_t rx_dropped; + u64_stats_t tx_packets; + u64_stats_t tx_bytes; + u64_stats_t tx_errors; + u64_stats_t tx_dropped; + atomic_t rx_queued; + struct u64_stats_sync tx_syncp; + struct u64_stats_sync rx_syncp; +}; + +struct mhi_net_dev { + struct mhi_device *mdev; + struct net_device *ndev; + const struct mhi_net_proto *proto; + void *proto_data; + struct delayed_work rx_refill; + struct mhi_net_stats stats; + u32 rx_queue_sz; +}; + +struct mhi_net_proto { + int (*init)(struct mhi_net_dev *dev); + struct sk_buff * (*tx_fixup)(struct net_device *ndev, struct sk_buff *skb); + int (*rx_fixup)(struct net_device *ndev, struct sk_buff *skb); +}; + +struct mhi_device_info { + const char *netname; + const struct mhi_net_proto *proto; +}; + +static int mhi_ndo_open(struct net_device *ndev) +{ + struct mhi_net_dev *mhi_netdev = netdev_priv(ndev); + + /* Feed the rx buffer pool */ + schedule_delayed_work(&mhi_netdev->rx_refill, 0); + + /* Carrier is established via out-of-band channel (e.g. qmi) */ + netif_carrier_on(ndev); + + netif_start_queue(ndev); + + return 0; +} + +static int mhi_ndo_stop(struct net_device *ndev) +{ + struct mhi_net_dev *mhi_netdev = netdev_priv(ndev); + + netif_stop_queue(ndev); + netif_carrier_off(ndev); + cancel_delayed_work_sync(&mhi_netdev->rx_refill); + + return 0; +} + +static int mhi_ndo_xmit(struct sk_buff *skb, struct net_device *ndev) +{ + struct mhi_net_dev *mhi_netdev = netdev_priv(ndev); + const struct mhi_net_proto *proto = mhi_netdev->proto; + struct mhi_device *mdev = mhi_netdev->mdev; + int err; + + if (proto && proto->tx_fixup) { + skb = proto->tx_fixup(mhi_netdev->ndev, skb); + if (unlikely(!skb)) + goto exit_drop; + } + + err = mhi_queue_skb(mdev, DMA_TO_DEVICE, skb, skb->len, MHI_EOT); + if (unlikely(err)) { + net_err_ratelimited("%s: Failed to queue TX buf (%d)\n", + ndev->name, err); + dev_kfree_skb_any(skb); + goto exit_drop; + } + + if (mhi_queue_is_full(mdev, DMA_TO_DEVICE)) + netif_stop_queue(ndev); + + return NETDEV_TX_OK; + +exit_drop: + u64_stats_update_begin(&mhi_netdev->stats.tx_syncp); + u64_stats_inc(&mhi_netdev->stats.tx_dropped); + u64_stats_update_end(&mhi_netdev->stats.tx_syncp); + + return NETDEV_TX_OK; +} + +static void mhi_ndo_get_stats64(struct net_device *ndev, + struct rtnl_link_stats64 *stats) +{ + struct mhi_net_dev *mhi_netdev = netdev_priv(ndev); + unsigned int start; + + do { + start = u64_stats_fetch_begin_irq(&mhi_netdev->stats.rx_syncp); + stats->rx_packets = u64_stats_read(&mhi_netdev->stats.rx_packets); + stats->rx_bytes = u64_stats_read(&mhi_netdev->stats.rx_bytes); + stats->rx_errors = u64_stats_read(&mhi_netdev->stats.rx_errors); + stats->rx_dropped = u64_stats_read(&mhi_netdev->stats.rx_dropped); + } while (u64_stats_fetch_retry_irq(&mhi_netdev->stats.rx_syncp, start)); + + do { + start = u64_stats_fetch_begin_irq(&mhi_netdev->stats.tx_syncp); + stats->tx_packets = u64_stats_read(&mhi_netdev->stats.tx_packets); + stats->tx_bytes = u64_stats_read(&mhi_netdev->stats.tx_bytes); + stats->tx_errors = u64_stats_read(&mhi_netdev->stats.tx_errors); + stats->tx_dropped = u64_stats_read(&mhi_netdev->stats.tx_dropped); + } while (u64_stats_fetch_retry_irq(&mhi_netdev->stats.tx_syncp, start)); +} + +static const struct net_device_ops mhi_netdev_ops = { + .ndo_open = mhi_ndo_open, + .ndo_stop = mhi_ndo_stop, + .ndo_start_xmit = mhi_ndo_xmit, + .ndo_get_stats64 = mhi_ndo_get_stats64, +}; + +static void mhi_net_setup(struct net_device *ndev) +{ + ndev->header_ops = NULL; /* No header */ + ndev->type = ARPHRD_RAWIP; + ndev->hard_header_len = 0; + ndev->addr_len = 0; + ndev->flags = IFF_POINTOPOINT | IFF_NOARP; + ndev->netdev_ops = &mhi_netdev_ops; + ndev->mtu = MHI_NET_DEFAULT_MTU; + ndev->min_mtu = MHI_NET_MIN_MTU; + ndev->max_mtu = MHI_NET_MAX_MTU; + ndev->tx_queue_len = 1000; +} + +static void mhi_net_dl_callback(struct mhi_device *mhi_dev, + struct mhi_result *mhi_res) +{ + struct mhi_net_dev *mhi_netdev = dev_get_drvdata(&mhi_dev->dev); + const struct mhi_net_proto *proto = mhi_netdev->proto; + struct sk_buff *skb = mhi_res->buf_addr; + int remaining; + + remaining = atomic_dec_return(&mhi_netdev->stats.rx_queued); + + if (unlikely(mhi_res->transaction_status)) { + dev_kfree_skb_any(skb); + + /* MHI layer stopping/resetting the DL channel */ + if (mhi_res->transaction_status == -ENOTCONN) + return; + + u64_stats_update_begin(&mhi_netdev->stats.rx_syncp); + u64_stats_inc(&mhi_netdev->stats.rx_errors); + u64_stats_update_end(&mhi_netdev->stats.rx_syncp); + } else { + u64_stats_update_begin(&mhi_netdev->stats.rx_syncp); + u64_stats_inc(&mhi_netdev->stats.rx_packets); + u64_stats_add(&mhi_netdev->stats.rx_bytes, mhi_res->bytes_xferd); + u64_stats_update_end(&mhi_netdev->stats.rx_syncp); + + switch (skb->data[0] & 0xf0) { + case 0x40: + skb->protocol = htons(ETH_P_IP); + break; + case 0x60: + skb->protocol = htons(ETH_P_IPV6); + break; + default: + skb->protocol = htons(ETH_P_MAP); + break; + } + + skb_put(skb, mhi_res->bytes_xferd); + + if (proto && proto->rx_fixup) + proto->rx_fixup(mhi_netdev->ndev, skb); + else + netif_rx(skb); + } + + /* Refill if RX buffers queue becomes low */ + if (remaining <= mhi_netdev->rx_queue_sz / 2) + schedule_delayed_work(&mhi_netdev->rx_refill, 0); +} + +static void mhi_net_ul_callback(struct mhi_device *mhi_dev, + struct mhi_result *mhi_res) +{ + struct mhi_net_dev *mhi_netdev = dev_get_drvdata(&mhi_dev->dev); + struct net_device *ndev = mhi_netdev->ndev; + struct mhi_device *mdev = mhi_netdev->mdev; + struct sk_buff *skb = mhi_res->buf_addr; + + /* Hardware has consumed the buffer, so free the skb (which is not + * freed by the MHI stack) and perform accounting. + */ + dev_consume_skb_any(skb); + + u64_stats_update_begin(&mhi_netdev->stats.tx_syncp); + if (unlikely(mhi_res->transaction_status)) { + + /* MHI layer stopping/resetting the UL channel */ + if (mhi_res->transaction_status == -ENOTCONN) { + u64_stats_update_end(&mhi_netdev->stats.tx_syncp); + return; + } + + u64_stats_inc(&mhi_netdev->stats.tx_errors); + } else { + u64_stats_inc(&mhi_netdev->stats.tx_packets); + u64_stats_add(&mhi_netdev->stats.tx_bytes, mhi_res->bytes_xferd); + } + u64_stats_update_end(&mhi_netdev->stats.tx_syncp); + + if (netif_queue_stopped(ndev) && !mhi_queue_is_full(mdev, DMA_TO_DEVICE)) + netif_wake_queue(ndev); +} + +static void mhi_net_rx_refill_work(struct work_struct *work) +{ + struct mhi_net_dev *mhi_netdev = container_of(work, struct mhi_net_dev, + rx_refill.work); + struct net_device *ndev = mhi_netdev->ndev; + struct mhi_device *mdev = mhi_netdev->mdev; + int size = READ_ONCE(ndev->mtu); + struct sk_buff *skb; + int err; + + while (atomic_read(&mhi_netdev->stats.rx_queued) < mhi_netdev->rx_queue_sz) { + skb = netdev_alloc_skb(ndev, size); + if (unlikely(!skb)) + break; + + err = mhi_queue_skb(mdev, DMA_FROM_DEVICE, skb, size, MHI_EOT); + if (unlikely(err)) { + net_err_ratelimited("%s: Failed to queue RX buf (%d)\n", + ndev->name, err); + kfree_skb(skb); + break; + } + + atomic_inc(&mhi_netdev->stats.rx_queued); + + /* Do not hog the CPU if rx buffers are consumed faster than + * queued (unlikely). + */ + cond_resched(); + } + + /* If we're still starved of rx buffers, reschedule later */ + if (unlikely(!atomic_read(&mhi_netdev->stats.rx_queued))) + schedule_delayed_work(&mhi_netdev->rx_refill, HZ / 2); +} + +static struct device_type wwan_type = { + .name = "wwan", +}; + +static int mhi_net_probe(struct mhi_device *mhi_dev, + const struct mhi_device_id *id) +{ + const struct mhi_device_info *info = (struct mhi_device_info *)id->driver_data; + struct device *dev = &mhi_dev->dev; + struct mhi_net_dev *mhi_netdev; + struct net_device *ndev; + int err; + + ndev = alloc_netdev(sizeof(*mhi_netdev), info->netname, + NET_NAME_PREDICTABLE, mhi_net_setup); + if (!ndev) + return -ENOMEM; + + mhi_netdev = netdev_priv(ndev); + dev_set_drvdata(dev, mhi_netdev); + mhi_netdev->ndev = ndev; + mhi_netdev->mdev = mhi_dev; + mhi_netdev->proto = info->proto; + SET_NETDEV_DEV(ndev, &mhi_dev->dev); + SET_NETDEV_DEVTYPE(ndev, &wwan_type); + + /* All MHI net channels have 128 ring elements (at least for now) */ + mhi_netdev->rx_queue_sz = 128; + + INIT_DELAYED_WORK(&mhi_netdev->rx_refill, mhi_net_rx_refill_work); + u64_stats_init(&mhi_netdev->stats.rx_syncp); + u64_stats_init(&mhi_netdev->stats.tx_syncp); + + /* Start MHI channels */ + err = mhi_prepare_for_transfer(mhi_dev); + if (err) + goto out_err; + + err = register_netdev(ndev); + if (err) + goto out_err; + + if (mhi_netdev->proto) { + err = mhi_netdev->proto->init(mhi_netdev); + if (err) + goto out_err_proto; + } + + return 0; + +out_err_proto: + unregister_netdev(ndev); +out_err: + free_netdev(ndev); + return err; +} + +static void mhi_net_remove(struct mhi_device *mhi_dev) +{ + struct mhi_net_dev *mhi_netdev = dev_get_drvdata(&mhi_dev->dev); + + unregister_netdev(mhi_netdev->ndev); + + mhi_unprepare_from_transfer(mhi_netdev->mdev); + + free_netdev(mhi_netdev->ndev); +} + +static const struct mhi_device_info mhi_hwip0 = { + .netname = "mhi_hwip%d", +}; + +static const struct mhi_device_info mhi_swip0 = { + .netname = "mhi_swip%d", +}; + +static const struct mhi_device_id mhi_net_id_table[] = { + /* Hardware accelerated data PATH (to modem IPA), protocol agnostic */ + { .chan = "IP_HW0", .driver_data = (kernel_ulong_t)&mhi_hwip0 }, + /* Software data PATH (to modem CPU) */ + { .chan = "IP_SW0", .driver_data = (kernel_ulong_t)&mhi_swip0 }, + {} +}; +MODULE_DEVICE_TABLE(mhi, mhi_net_id_table); + +static struct mhi_driver mhi_net_driver = { + .probe = mhi_net_probe, + .remove = mhi_net_remove, + .dl_xfer_cb = mhi_net_dl_callback, + .ul_xfer_cb = mhi_net_ul_callback, + .id_table = mhi_net_id_table, + .driver = { + .name = "mhi_net", + .owner = THIS_MODULE, + }, +}; + +module_mhi_driver(mhi_net_driver); + +MODULE_AUTHOR("Loic Poulain "); +MODULE_DESCRIPTION("Network over MHI"); +MODULE_LICENSE("GPL v2"); diff --git a/drivers/net/mhi_net.c b/drivers/net/mhi_net.c deleted file mode 100644 index aa3a5e0..0000000 --- a/drivers/net/mhi_net.c +++ /dev/null @@ -1,379 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0-or-later -/* MHI Network driver - Network over MHI bus - * - * Copyright (C) 2020 Linaro Ltd - */ - -#include -#include -#include -#include -#include -#include -#include - -#define MHI_NET_MIN_MTU ETH_MIN_MTU -#define MHI_NET_MAX_MTU 0xffff -#define MHI_NET_DEFAULT_MTU 0x4000 - -struct mhi_net_stats { - u64_stats_t rx_packets; - u64_stats_t rx_bytes; - u64_stats_t rx_errors; - u64_stats_t rx_dropped; - u64_stats_t tx_packets; - u64_stats_t tx_bytes; - u64_stats_t tx_errors; - u64_stats_t tx_dropped; - atomic_t rx_queued; - struct u64_stats_sync tx_syncp; - struct u64_stats_sync rx_syncp; -}; - -struct mhi_net_dev { - struct mhi_device *mdev; - struct net_device *ndev; - const struct mhi_net_proto *proto; - void *proto_data; - struct delayed_work rx_refill; - struct mhi_net_stats stats; - u32 rx_queue_sz; -}; - -struct mhi_net_proto { - int (*init)(struct mhi_net_dev *dev); - struct sk_buff * (*tx_fixup)(struct net_device *ndev, struct sk_buff *skb); - int (*rx_fixup)(struct net_device *ndev, struct sk_buff *skb); -}; - -struct mhi_device_info { - const char *netname; - const struct mhi_net_proto *proto; -}; - -static int mhi_ndo_open(struct net_device *ndev) -{ - struct mhi_net_dev *mhi_netdev = netdev_priv(ndev); - - /* Feed the rx buffer pool */ - schedule_delayed_work(&mhi_netdev->rx_refill, 0); - - /* Carrier is established via out-of-band channel (e.g. qmi) */ - netif_carrier_on(ndev); - - netif_start_queue(ndev); - - return 0; -} - -static int mhi_ndo_stop(struct net_device *ndev) -{ - struct mhi_net_dev *mhi_netdev = netdev_priv(ndev); - - netif_stop_queue(ndev); - netif_carrier_off(ndev); - cancel_delayed_work_sync(&mhi_netdev->rx_refill); - - return 0; -} - -static int mhi_ndo_xmit(struct sk_buff *skb, struct net_device *ndev) -{ - struct mhi_net_dev *mhi_netdev = netdev_priv(ndev); - const struct mhi_net_proto *proto = mhi_netdev->proto; - struct mhi_device *mdev = mhi_netdev->mdev; - int err; - - if (proto && proto->tx_fixup) { - skb = proto->tx_fixup(mhi_netdev->ndev, skb); - if (unlikely(!skb)) - goto exit_drop; - } - - err = mhi_queue_skb(mdev, DMA_TO_DEVICE, skb, skb->len, MHI_EOT); - if (unlikely(err)) { - net_err_ratelimited("%s: Failed to queue TX buf (%d)\n", - ndev->name, err); - dev_kfree_skb_any(skb); - goto exit_drop; - } - - if (mhi_queue_is_full(mdev, DMA_TO_DEVICE)) - netif_stop_queue(ndev); - - return NETDEV_TX_OK; - -exit_drop: - u64_stats_update_begin(&mhi_netdev->stats.tx_syncp); - u64_stats_inc(&mhi_netdev->stats.tx_dropped); - u64_stats_update_end(&mhi_netdev->stats.tx_syncp); - - return NETDEV_TX_OK; -} - -static void mhi_ndo_get_stats64(struct net_device *ndev, - struct rtnl_link_stats64 *stats) -{ - struct mhi_net_dev *mhi_netdev = netdev_priv(ndev); - unsigned int start; - - do { - start = u64_stats_fetch_begin_irq(&mhi_netdev->stats.rx_syncp); - stats->rx_packets = u64_stats_read(&mhi_netdev->stats.rx_packets); - stats->rx_bytes = u64_stats_read(&mhi_netdev->stats.rx_bytes); - stats->rx_errors = u64_stats_read(&mhi_netdev->stats.rx_errors); - stats->rx_dropped = u64_stats_read(&mhi_netdev->stats.rx_dropped); - } while (u64_stats_fetch_retry_irq(&mhi_netdev->stats.rx_syncp, start)); - - do { - start = u64_stats_fetch_begin_irq(&mhi_netdev->stats.tx_syncp); - stats->tx_packets = u64_stats_read(&mhi_netdev->stats.tx_packets); - stats->tx_bytes = u64_stats_read(&mhi_netdev->stats.tx_bytes); - stats->tx_errors = u64_stats_read(&mhi_netdev->stats.tx_errors); - stats->tx_dropped = u64_stats_read(&mhi_netdev->stats.tx_dropped); - } while (u64_stats_fetch_retry_irq(&mhi_netdev->stats.tx_syncp, start)); -} - -static const struct net_device_ops mhi_netdev_ops = { - .ndo_open = mhi_ndo_open, - .ndo_stop = mhi_ndo_stop, - .ndo_start_xmit = mhi_ndo_xmit, - .ndo_get_stats64 = mhi_ndo_get_stats64, -}; - -static void mhi_net_setup(struct net_device *ndev) -{ - ndev->header_ops = NULL; /* No header */ - ndev->type = ARPHRD_RAWIP; - ndev->hard_header_len = 0; - ndev->addr_len = 0; - ndev->flags = IFF_POINTOPOINT | IFF_NOARP; - ndev->netdev_ops = &mhi_netdev_ops; - ndev->mtu = MHI_NET_DEFAULT_MTU; - ndev->min_mtu = MHI_NET_MIN_MTU; - ndev->max_mtu = MHI_NET_MAX_MTU; - ndev->tx_queue_len = 1000; -} - -static void mhi_net_dl_callback(struct mhi_device *mhi_dev, - struct mhi_result *mhi_res) -{ - struct mhi_net_dev *mhi_netdev = dev_get_drvdata(&mhi_dev->dev); - const struct mhi_net_proto *proto = mhi_netdev->proto; - struct sk_buff *skb = mhi_res->buf_addr; - int remaining; - - remaining = atomic_dec_return(&mhi_netdev->stats.rx_queued); - - if (unlikely(mhi_res->transaction_status)) { - dev_kfree_skb_any(skb); - - /* MHI layer stopping/resetting the DL channel */ - if (mhi_res->transaction_status == -ENOTCONN) - return; - - u64_stats_update_begin(&mhi_netdev->stats.rx_syncp); - u64_stats_inc(&mhi_netdev->stats.rx_errors); - u64_stats_update_end(&mhi_netdev->stats.rx_syncp); - } else { - u64_stats_update_begin(&mhi_netdev->stats.rx_syncp); - u64_stats_inc(&mhi_netdev->stats.rx_packets); - u64_stats_add(&mhi_netdev->stats.rx_bytes, mhi_res->bytes_xferd); - u64_stats_update_end(&mhi_netdev->stats.rx_syncp); - - switch (skb->data[0] & 0xf0) { - case 0x40: - skb->protocol = htons(ETH_P_IP); - break; - case 0x60: - skb->protocol = htons(ETH_P_IPV6); - break; - default: - skb->protocol = htons(ETH_P_MAP); - break; - } - - skb_put(skb, mhi_res->bytes_xferd); - - if (proto && proto->rx_fixup) - proto->rx_fixup(mhi_netdev->ndev, skb); - else - netif_rx(skb); - } - - /* Refill if RX buffers queue becomes low */ - if (remaining <= mhi_netdev->rx_queue_sz / 2) - schedule_delayed_work(&mhi_netdev->rx_refill, 0); -} - -static void mhi_net_ul_callback(struct mhi_device *mhi_dev, - struct mhi_result *mhi_res) -{ - struct mhi_net_dev *mhi_netdev = dev_get_drvdata(&mhi_dev->dev); - struct net_device *ndev = mhi_netdev->ndev; - struct mhi_device *mdev = mhi_netdev->mdev; - struct sk_buff *skb = mhi_res->buf_addr; - - /* Hardware has consumed the buffer, so free the skb (which is not - * freed by the MHI stack) and perform accounting. - */ - dev_consume_skb_any(skb); - - u64_stats_update_begin(&mhi_netdev->stats.tx_syncp); - if (unlikely(mhi_res->transaction_status)) { - - /* MHI layer stopping/resetting the UL channel */ - if (mhi_res->transaction_status == -ENOTCONN) { - u64_stats_update_end(&mhi_netdev->stats.tx_syncp); - return; - } - - u64_stats_inc(&mhi_netdev->stats.tx_errors); - } else { - u64_stats_inc(&mhi_netdev->stats.tx_packets); - u64_stats_add(&mhi_netdev->stats.tx_bytes, mhi_res->bytes_xferd); - } - u64_stats_update_end(&mhi_netdev->stats.tx_syncp); - - if (netif_queue_stopped(ndev) && !mhi_queue_is_full(mdev, DMA_TO_DEVICE)) - netif_wake_queue(ndev); -} - -static void mhi_net_rx_refill_work(struct work_struct *work) -{ - struct mhi_net_dev *mhi_netdev = container_of(work, struct mhi_net_dev, - rx_refill.work); - struct net_device *ndev = mhi_netdev->ndev; - struct mhi_device *mdev = mhi_netdev->mdev; - int size = READ_ONCE(ndev->mtu); - struct sk_buff *skb; - int err; - - while (atomic_read(&mhi_netdev->stats.rx_queued) < mhi_netdev->rx_queue_sz) { - skb = netdev_alloc_skb(ndev, size); - if (unlikely(!skb)) - break; - - err = mhi_queue_skb(mdev, DMA_FROM_DEVICE, skb, size, MHI_EOT); - if (unlikely(err)) { - net_err_ratelimited("%s: Failed to queue RX buf (%d)\n", - ndev->name, err); - kfree_skb(skb); - break; - } - - atomic_inc(&mhi_netdev->stats.rx_queued); - - /* Do not hog the CPU if rx buffers are consumed faster than - * queued (unlikely). - */ - cond_resched(); - } - - /* If we're still starved of rx buffers, reschedule later */ - if (unlikely(!atomic_read(&mhi_netdev->stats.rx_queued))) - schedule_delayed_work(&mhi_netdev->rx_refill, HZ / 2); -} - -static struct device_type wwan_type = { - .name = "wwan", -}; - -static int mhi_net_probe(struct mhi_device *mhi_dev, - const struct mhi_device_id *id) -{ - const struct mhi_device_info *info = (struct mhi_device_info *)id->driver_data; - struct device *dev = &mhi_dev->dev; - struct mhi_net_dev *mhi_netdev; - struct net_device *ndev; - int err; - - ndev = alloc_netdev(sizeof(*mhi_netdev), info->netname, - NET_NAME_PREDICTABLE, mhi_net_setup); - if (!ndev) - return -ENOMEM; - - mhi_netdev = netdev_priv(ndev); - dev_set_drvdata(dev, mhi_netdev); - mhi_netdev->ndev = ndev; - mhi_netdev->mdev = mhi_dev; - mhi_netdev->proto = info->proto; - SET_NETDEV_DEV(ndev, &mhi_dev->dev); - SET_NETDEV_DEVTYPE(ndev, &wwan_type); - - /* All MHI net channels have 128 ring elements (at least for now) */ - mhi_netdev->rx_queue_sz = 128; - - INIT_DELAYED_WORK(&mhi_netdev->rx_refill, mhi_net_rx_refill_work); - u64_stats_init(&mhi_netdev->stats.rx_syncp); - u64_stats_init(&mhi_netdev->stats.tx_syncp); - - /* Start MHI channels */ - err = mhi_prepare_for_transfer(mhi_dev); - if (err) - goto out_err; - - err = register_netdev(ndev); - if (err) - goto out_err; - - if (mhi_netdev->proto) { - err = mhi_netdev->proto->init(mhi_netdev); - if (err) - goto out_err_proto; - } - - return 0; - -out_err_proto: - unregister_netdev(ndev); -out_err: - free_netdev(ndev); - return err; -} - -static void mhi_net_remove(struct mhi_device *mhi_dev) -{ - struct mhi_net_dev *mhi_netdev = dev_get_drvdata(&mhi_dev->dev); - - unregister_netdev(mhi_netdev->ndev); - - mhi_unprepare_from_transfer(mhi_netdev->mdev); - - free_netdev(mhi_netdev->ndev); -} - -static const struct mhi_device_info mhi_hwip0 = { - .netname = "mhi_hwip%d", -}; - -static const struct mhi_device_info mhi_swip0 = { - .netname = "mhi_swip%d", -}; - -static const struct mhi_device_id mhi_net_id_table[] = { - /* Hardware accelerated data PATH (to modem IPA), protocol agnostic */ - { .chan = "IP_HW0", .driver_data = (kernel_ulong_t)&mhi_hwip0 }, - /* Software data PATH (to modem CPU) */ - { .chan = "IP_SW0", .driver_data = (kernel_ulong_t)&mhi_swip0 }, - {} -}; -MODULE_DEVICE_TABLE(mhi, mhi_net_id_table); - -static struct mhi_driver mhi_net_driver = { - .probe = mhi_net_probe, - .remove = mhi_net_remove, - .dl_xfer_cb = mhi_net_dl_callback, - .ul_xfer_cb = mhi_net_ul_callback, - .id_table = mhi_net_id_table, - .driver = { - .name = "mhi_net", - .owner = THIS_MODULE, - }, -}; - -module_mhi_driver(mhi_net_driver); - -MODULE_AUTHOR("Loic Poulain "); -MODULE_DESCRIPTION("Network over MHI"); -MODULE_LICENSE("GPL v2"); From patchwork Wed Jan 27 17:01:17 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Loic Poulain X-Patchwork-Id: 12050583 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-18.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER, INCLUDES_PATCH,MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED, USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id BC0B9C433E6 for ; Wed, 27 Jan 2021 16:55:10 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 6164F60187 for ; Wed, 27 Jan 2021 16:55:10 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S234080AbhA0Qy4 (ORCPT ); Wed, 27 Jan 2021 11:54:56 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:47550 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S233964AbhA0QyU (ORCPT ); Wed, 27 Jan 2021 11:54:20 -0500 Received: from mail-wr1-x431.google.com (mail-wr1-x431.google.com [IPv6:2a00:1450:4864:20::431]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 3BBE5C06174A for ; Wed, 27 Jan 2021 08:53:40 -0800 (PST) Received: by mail-wr1-x431.google.com with SMTP id 7so2663506wrz.0 for ; Wed, 27 Jan 2021 08:53:40 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=from:to:cc:subject:date:message-id:in-reply-to:references; bh=bfdBxnBK5Jy768CPQ2cgYOreRpXRg/jTDJOtLuGW1LI=; b=FQZ1FpoQouMAagNKFUvGelvSrToB1Nis1IHxJYOI0K6lHIcrgqCtIoOLMFvickoBIk SUWGHersm/xHwzzwFZus0R4Onl6Wsnbam6WY0VPFwF6QwsTnXeOlUkCqZh56xnt2Enhs xNAx4ZwAm1mOdO6sS8GY75sxM7npRO0vPLE4HbN80D84D/YIDM906TGUQ1uX3fbU+9jn GRSEqOoVZ5XW4z2MjNj6h2i6n6RJhtr+/QQkTDzrp0fKLyEq6ZpOiFIG5GiUDxPk9wPl rnIoBjYIVt3X6zNYmkJzJpgWhUDXrgTmT7A+plG5SyuLNvZI3l6txFxQ1wpbbSolajC5 yGdQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references; bh=bfdBxnBK5Jy768CPQ2cgYOreRpXRg/jTDJOtLuGW1LI=; b=irEOtYVPYEj93sVzvVt2DIu2jL5UUm0nVAU368ivQD2/n60WwByQKZ4VC7fSCpExis VQvzZue+to5PtmqjgtyPu7Hxj0ZrYAmgvTTwY4vo+1ff0I6f2jtBB2ElPrijAcvnOaAz C2Z2EwGhjv4eUaTmMnCwRTITyelmUncBetykzq1ZIigyp9FfkguFxfFklXH86LGRbhhi 8i2WdiWi5Y8SenU4OjRXLgeDsuc0qBf2ZmD9jUQiivVmR74IoGzqCuwPVmeNCIPTj6dI VHV++jR7qV4fOanmWkji+lmPd32lpzijYvbx9R2L9rS9G27GDr+HpCc/xnrT0iwslx1G iICQ== X-Gm-Message-State: AOAM530ZDeB7yUmD8HB19PY2FtEMxzNJWuyYwwXaWMQfaPO4UaOvvWIq 2GWvlPwf+O57dSEE6kDmOBkhnA== X-Google-Smtp-Source: ABdhPJxhyA7y1eghNbiOTX4pmUF0pI2IzG3UA9/YNUCc8PhDuaxe43hJ9jX1jboHpYMEYs0b5eLqkg== X-Received: by 2002:a5d:4e92:: with SMTP id e18mr12552625wru.66.1611766418867; Wed, 27 Jan 2021 08:53:38 -0800 (PST) Received: from localhost.localdomain ([88.122.66.28]) by smtp.gmail.com with ESMTPSA id m82sm3077042wmf.29.2021.01.27.08.53.38 (version=TLS1_2 cipher=ECDHE-ECDSA-AES128-GCM-SHA256 bits=128/128); Wed, 27 Jan 2021 08:53:38 -0800 (PST) From: Loic Poulain To: kuba@kernel.org, davem@davemloft.net Cc: netdev@vger.kernel.org, carl.yin@quectel.com, Loic Poulain Subject: [PATCH net-next 3/3] net: mhi: Add mbim proto Date: Wed, 27 Jan 2021 18:01:17 +0100 Message-Id: <1611766877-16787-3-git-send-email-loic.poulain@linaro.org> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1611766877-16787-1-git-send-email-loic.poulain@linaro.org> References: <1611766877-16787-1-git-send-email-loic.poulain@linaro.org> Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org MBIM has initially been specified by USB-IF for transporting data (IP) between a modem and a host over USB. However some modern modems also support MBIM over PCIe (via MHI). In the same way as QMAP(rmnet), it allows to aggregate IP packets and to perform context multiplexing. This change adds minimal MBIM support to MHI, allowing to support MBIM only modems. MBIM being based on USB NCM, it reuses some helpers from the USB stack, but the cdc-mbim driver is too USB coupled to be reused. At some point it would be interesting to move on a factorized solution, having a generic MBIM network lib or dedicated MBIM netlink virtual interface support. This code has been highly inspired from the mhi_mbim downstream driver (Carl Yin ). Signed-off-by: Loic Poulain --- drivers/net/mhi/Makefile | 2 +- drivers/net/mhi/mhi.h | 39 ++++++++ drivers/net/mhi/net.c | 41 ++------ drivers/net/mhi/proto_mbim.c | 220 +++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 270 insertions(+), 32 deletions(-) create mode 100644 drivers/net/mhi/mhi.h create mode 100644 drivers/net/mhi/proto_mbim.c diff --git a/drivers/net/mhi/Makefile b/drivers/net/mhi/Makefile index 0acf989..f71b9f8 100644 --- a/drivers/net/mhi/Makefile +++ b/drivers/net/mhi/Makefile @@ -1,3 +1,3 @@ obj-$(CONFIG_MHI_NET) += mhi_net.o -mhi_net-y := net.o +mhi_net-y := net.o proto_mbim.o diff --git a/drivers/net/mhi/mhi.h b/drivers/net/mhi/mhi.h new file mode 100644 index 0000000..e7f7246 --- /dev/null +++ b/drivers/net/mhi/mhi.h @@ -0,0 +1,39 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ + +/* MHI Network driver - Network over MHI bus + * + * Copyright (C) 2021 Linaro Ltd + */ + +struct mhi_net_stats { + u64_stats_t rx_packets; + u64_stats_t rx_bytes; + u64_stats_t rx_errors; + u64_stats_t rx_dropped; + u64_stats_t tx_packets; + u64_stats_t tx_bytes; + u64_stats_t tx_errors; + u64_stats_t tx_dropped; + atomic_t rx_queued; + struct u64_stats_sync tx_syncp; + struct u64_stats_sync rx_syncp; +}; + +struct mhi_net_dev { + struct mhi_device *mdev; + struct net_device *ndev; + const struct mhi_device_info *info; + const struct mhi_net_proto *proto; + void *proto_data; + struct delayed_work rx_refill; + struct mhi_net_stats stats; + u32 rx_queue_sz; +}; + +struct mhi_net_proto { + int (*init)(struct mhi_net_dev *dev); + struct sk_buff * (*tx_fixup)(struct net_device *ndev, struct sk_buff *skb); + int (*rx_fixup)(struct net_device *ndev, struct sk_buff *skb); +}; + +extern const struct mhi_net_proto proto_mbim; diff --git a/drivers/net/mhi/net.c b/drivers/net/mhi/net.c index aa3a5e0..e8e458f 100644 --- a/drivers/net/mhi/net.c +++ b/drivers/net/mhi/net.c @@ -12,40 +12,12 @@ #include #include +#include "mhi.h" + #define MHI_NET_MIN_MTU ETH_MIN_MTU #define MHI_NET_MAX_MTU 0xffff #define MHI_NET_DEFAULT_MTU 0x4000 -struct mhi_net_stats { - u64_stats_t rx_packets; - u64_stats_t rx_bytes; - u64_stats_t rx_errors; - u64_stats_t rx_dropped; - u64_stats_t tx_packets; - u64_stats_t tx_bytes; - u64_stats_t tx_errors; - u64_stats_t tx_dropped; - atomic_t rx_queued; - struct u64_stats_sync tx_syncp; - struct u64_stats_sync rx_syncp; -}; - -struct mhi_net_dev { - struct mhi_device *mdev; - struct net_device *ndev; - const struct mhi_net_proto *proto; - void *proto_data; - struct delayed_work rx_refill; - struct mhi_net_stats stats; - u32 rx_queue_sz; -}; - -struct mhi_net_proto { - int (*init)(struct mhi_net_dev *dev); - struct sk_buff * (*tx_fixup)(struct net_device *ndev, struct sk_buff *skb); - int (*rx_fixup)(struct net_device *ndev, struct sk_buff *skb); -}; - struct mhi_device_info { const char *netname; const struct mhi_net_proto *proto; @@ -351,12 +323,19 @@ static const struct mhi_device_info mhi_swip0 = { .netname = "mhi_swip%d", }; +static const struct mhi_device_info mhi_hwip0_mbim = { + .netname = "mhi_mbim%d", + .proto = &proto_mbim, +}; + static const struct mhi_device_id mhi_net_id_table[] = { /* Hardware accelerated data PATH (to modem IPA), protocol agnostic */ { .chan = "IP_HW0", .driver_data = (kernel_ulong_t)&mhi_hwip0 }, /* Software data PATH (to modem CPU) */ { .chan = "IP_SW0", .driver_data = (kernel_ulong_t)&mhi_swip0 }, - {} + /* Hardware accelerated data PATH (to modem IPA), MBIM protocol */ + { .chan = "IP_HW0_MBIM", .driver_data = (kernel_ulong_t)&mhi_hwip0_mbim }, + { } }; MODULE_DEVICE_TABLE(mhi, mhi_net_id_table); diff --git a/drivers/net/mhi/proto_mbim.c b/drivers/net/mhi/proto_mbim.c new file mode 100644 index 0000000..b568078c --- /dev/null +++ b/drivers/net/mhi/proto_mbim.c @@ -0,0 +1,220 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* MHI Network driver - Network over MHI bus + * + * Copyright (C) 2021 Linaro Ltd + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include "mhi.h" + +struct mbim_context { + u16 rx_seq; +}; + +static int mbim_rx_verify_nth16(struct sk_buff *skb) +{ + struct usb_cdc_ncm_nth16 *nth16; + int ret = -EINVAL; + + if (skb->len < (sizeof(struct usb_cdc_ncm_nth16) + + sizeof(struct usb_cdc_ncm_ndp16))) { + goto error; + } + + nth16 = (struct usb_cdc_ncm_nth16 *)skb->data; + + if (nth16->dwSignature != cpu_to_le32(USB_CDC_NCM_NTH16_SIGN)) + goto error; + + ret = le16_to_cpu(nth16->wNdpIndex); +error: + return ret; +} + +static int mbim_rx_verify_ndp16(struct sk_buff *skb, int ndpoffset) +{ + struct usb_cdc_ncm_ndp16 *ndp16; + int ret = -EINVAL; + + if ((ndpoffset + sizeof(struct usb_cdc_ncm_ndp16)) > skb->len) + goto error; + + ndp16 = (struct usb_cdc_ncm_ndp16 *)(skb->data + ndpoffset); + + if (le16_to_cpu(ndp16->wLength) < USB_CDC_NCM_NDP16_LENGTH_MIN) + goto error; + + ret = ((le16_to_cpu(ndp16->wLength) - + sizeof(struct usb_cdc_ncm_ndp16)) / + sizeof(struct usb_cdc_ncm_dpe16)); + ret--; /* Last entry is always a NULL terminator */ + + if ((sizeof(struct usb_cdc_ncm_ndp16) + + ret * (sizeof(struct usb_cdc_ncm_dpe16))) > skb->len) { + ret = -EINVAL; + } +error: + return ret; +} + +static int mbim_rx_fixup(struct net_device *ndev, struct sk_buff *skb) +{ + int ndpoffset; + + /* Check NTB header signature and retrieve first NDP offset */ + ndpoffset = mbim_rx_verify_nth16(skb); + if (ndpoffset < 0) { + netdev_err(ndev, "MBIM: Incorrect NTB header\n"); + goto error; + } + + /* Process each NDP */ + while (1) { + struct usb_cdc_ncm_ndp16 *ndp16; + struct usb_cdc_ncm_dpe16 *dpe16; + int nframes, n; + + /* Check NDP header and retrieve number of datagrams */ + nframes = mbim_rx_verify_ndp16(skb, ndpoffset); + if (nframes < 0) { + netdev_err(ndev, "MBIM: Incorrect NDP16\n"); + goto error; + } + + /* Only support the IPS session 0 for now */ + ndp16 = (struct usb_cdc_ncm_ndp16 *)(skb->data + ndpoffset); + switch (ndp16->dwSignature & cpu_to_le32(0x00ffffff)) { + case cpu_to_le32(USB_CDC_MBIM_NDP16_IPS_SIGN): + break; + default: + netdev_err(ndev, "MBIM: Unsupported NDP type\n"); + goto next_ndp; + } + + /* de-aggregate and deliver IP packets */ + dpe16 = ndp16->dpe16; + for (n = 0; n < nframes; n++, dpe16++) { + u16 dgram_offset = le16_to_cpu(dpe16->wDatagramIndex); + u16 dgram_len = le16_to_cpu(dpe16->wDatagramLength); + struct sk_buff *skbn; + + if (!dgram_offset || !dgram_len) + break; /* null terminator */ + + skbn = netdev_alloc_skb(ndev, dgram_len); + if (!skbn) + continue; + + skb_put(skbn, dgram_len); + memcpy(skbn->data, skb->data + dgram_offset, dgram_len); + + switch (skbn->data[0] & 0xf0) { + case 0x40: + skbn->protocol = htons(ETH_P_IP); + break; + case 0x60: + skbn->protocol = htons(ETH_P_IPV6); + break; + default: + netdev_err(ndev, "MBIM: unknown protocol\n"); + continue; + } + + netif_rx(skbn); + } +next_ndp: + /* Other NDP to process? */ + ndpoffset = le16_to_cpu(ndp16->wNextNdpIndex); + if (!ndpoffset) + break; + } + + /* free skb */ + dev_consume_skb_any(skb); + return 0; +error: + dev_kfree_skb_any(skb); + return -EIO; +} + +struct mbim_tx_hdr { + struct usb_cdc_ncm_nth16 nth16; + struct usb_cdc_ncm_ndp16 ndp16; + struct usb_cdc_ncm_dpe16 dpe16[2]; +} __packed; + +static struct sk_buff *mbim_tx_fixup(struct net_device *ndev, struct sk_buff *skb) +{ + struct mbim_tx_hdr *mbim_hdr; + struct usb_cdc_ncm_nth16 *nth16; + struct usb_cdc_ncm_ndp16 *ndp16; + unsigned int dgram_size = skb->len; + static int seq; + + /* For now, this is a partial implementation of CDC MBIM, only one NDP + * is sent, containing the IP packet (no aggregation). + */ + + if (skb_headroom(skb) < sizeof(struct mbim_tx_hdr)) { + dev_kfree_skb_any(skb); + return NULL; + } + + mbim_hdr = skb_push(skb, sizeof(struct mbim_tx_hdr)); + + /* Fill NTB header */ + nth16 = &mbim_hdr->nth16; + nth16->dwSignature = cpu_to_le32(USB_CDC_NCM_NTH16_SIGN); + nth16->wHeaderLength = cpu_to_le16(sizeof(struct usb_cdc_ncm_nth16)); + nth16->wSequence = cpu_to_le16(seq++); + nth16->wBlockLength = cpu_to_le16(skb->len); + nth16->wNdpIndex = cpu_to_le16(sizeof(struct usb_cdc_ncm_nth16)); + + /* Fill the unique NDP */ + ndp16 = &mbim_hdr->ndp16; + ndp16->dwSignature = cpu_to_le32(USB_CDC_MBIM_NDP16_IPS_SIGN); + ndp16->wLength = cpu_to_le16(sizeof(struct usb_cdc_ncm_ndp16) + + sizeof(struct usb_cdc_ncm_dpe16) * 2); + ndp16->wNextNdpIndex = 0; + + /* Datagram follows the mbim header */ + ndp16->dpe16[0].wDatagramIndex = cpu_to_le16(sizeof(struct mbim_tx_hdr)); + ndp16->dpe16[0].wDatagramLength = cpu_to_le16(dgram_size); + + /* null termination */ + ndp16->dpe16[1].wDatagramIndex = 0; + ndp16->dpe16[1].wDatagramLength = 0; + + return skb; +} + +static int mbim_init(struct mhi_net_dev *mhi_netdev) +{ + struct net_device *ndev = mhi_netdev->ndev; + + mhi_netdev->proto_data = devm_kzalloc(&ndev->dev, + sizeof(struct mbim_context), + GFP_KERNEL); + if (!mhi_netdev->proto_data) + return -ENOMEM; + + ndev->needed_headroom = sizeof(struct mbim_tx_hdr); + + return 0; +} + +const struct mhi_net_proto proto_mbim = { + .init = mbim_init, + .rx_fixup = mbim_rx_fixup, + .tx_fixup = mbim_tx_fixup, +};