From patchwork Sun Dec 12 14:50:25 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Stefan Wahren X-Patchwork-Id: 12672283 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 Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id CA829C433F5 for ; Sun, 12 Dec 2021 14:51:19 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S231496AbhLLOvL (ORCPT ); Sun, 12 Dec 2021 09:51:11 -0500 Received: from mout.kundenserver.de ([212.227.126.133]:51491 "EHLO mout.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230192AbhLLOvK (ORCPT ); Sun, 12 Dec 2021 09:51:10 -0500 Received: from localhost.localdomain ([37.4.249.122]) by mrelayeu.kundenserver.de (mreue011 [212.227.15.167]) with ESMTPSA (Nemesis) id 1N8GhM-1mRsiq1NVQ-014Ejq; Sun, 12 Dec 2021 15:50:52 +0100 From: Stefan Wahren To: "David S . Miller" , Jakub Kicinski Cc: Rob Herring , Andrew Lunn , Michael Heimpold , jimmy.shen@vertexcom.com, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, devicetree@vger.kernel.org, Stefan Wahren Subject: [PATCH V3 net-next 1/3] dt-bindings: add vendor Vertexcom Date: Sun, 12 Dec 2021 15:50:25 +0100 Message-Id: <1639320627-8827-2-git-send-email-stefan.wahren@i2se.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1639320627-8827-1-git-send-email-stefan.wahren@i2se.com> References: <1639320627-8827-1-git-send-email-stefan.wahren@i2se.com> X-Provags-ID: V03:K1:bGM+akImUbUuH0wK7AjcJi62/klv4j3a5IXQv+vN0ZOMabhrLKk TO4CBtdWL3A2Rs6cgZm7KpxYykj+tH7b94G97Qdfb/jaSUFS9tVb4AS/HI0V087kmT22JSz vhCUBUUgyUf5Vp1/RDEY6D58nCpsqNitJXE2DKyC0mLv6rSAiPH+BajlEAzgqKw3u4KSBm4 WieVnTzqL011rPKGiBERQ== X-UI-Out-Filterresults: notjunk:1;V03:K0:CeUaUEyX+ic=:KLO/OFRfFJ+pzKzovht1C4 eTAxGpNXKoKHFVOyGS2hPZKC+AzYmyTsOh7PhGWiNJ4zkNEfngV7d+C4H/oxiP0fOeEEbKr/Q IFTJpmV1doKZprXT4KBIWd/JrF2HpgWaqw8GWjRxnwwBR9zhGBU1uOd8aNZbbeHZSLk+ibyJN ZQ8NxhgpwGe4YbHEfswRbicSG5Mmbg/YY82qYB8/F+BnG3vZwnjd/762U+IgTx9aZ/GHDNnWE k2ff64iUY4gemLKL5mn4kUfYSo/rHvS6GC0htvLSGCFib6vO6aTgsHViTVcBw5QPt0hJhIl3x H3cubtr09CHch3Lyxa/qEJjdC6AdUTZgoq/jvcOJt9Jm1t1mjg0dysYKAkD9mV9UerhR3hpE2 eacFqGk67eIq9bDVJ/7i8Rp14tfD8yphMoyRwtibr9fKiiAJubQbzCPwb+pUQuj8yhoMs305A g1z7EMCrwfFqESXM1XurGg+nN4twGwu/G+jQan5oN83TNKfSgCfSlHN9O9O3Wj/Px7PJWbxmG VgXcbmXy8SVjQ6wpH4Q81JUrMROsVftySjrHkoeA9/h5HYxOQX/rrvcCNp6y4ZlkrqWeABhQX lfIGGCZh6SUUKd7vn0cBk0zyJJPxKb7esaSmpVlqRQECfYYhCZ/O9WrA7mkPYqLFJVPyUtVzH qHiulYRWvuBz5kronsPqv8+VpcWI1nS/NTQBQRtDlJ7ggXjRnJ8IPa/Swuumyqsatf/sjNkya 60gFHAi7OxWMD+cb Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org Add vendor prefix for Vertexcom Technologies, Inc [1]. [1] - http://www.vertexcom.com/ Signed-off-by: Stefan Wahren Acked-by: Rob Herring --- Documentation/devicetree/bindings/vendor-prefixes.yaml | 2 ++ 1 file changed, 2 insertions(+) diff --git a/Documentation/devicetree/bindings/vendor-prefixes.yaml b/Documentation/devicetree/bindings/vendor-prefixes.yaml index ed76d77..932c5de 100644 --- a/Documentation/devicetree/bindings/vendor-prefixes.yaml +++ b/Documentation/devicetree/bindings/vendor-prefixes.yaml @@ -1278,6 +1278,8 @@ patternProperties: description: Variscite Ltd. "^vdl,.*": description: Van der Laan b.v. + "^vertexcom,.*": + description: Vertexcom Technologies, Inc. "^via,.*": description: VIA Technologies, Inc. "^videostrong,.*": From patchwork Sun Dec 12 14:50:26 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Stefan Wahren X-Patchwork-Id: 12672281 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 Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 5911FC433EF for ; Sun, 12 Dec 2021 14:51:14 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S231503AbhLLOvM (ORCPT ); Sun, 12 Dec 2021 09:51:12 -0500 Received: from mout.kundenserver.de ([212.227.126.130]:35105 "EHLO mout.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S231484AbhLLOvL (ORCPT ); Sun, 12 Dec 2021 09:51:11 -0500 Received: from localhost.localdomain ([37.4.249.122]) by mrelayeu.kundenserver.de (mreue011 [212.227.15.167]) with ESMTPSA (Nemesis) id 1MbRwP-1mPpSP3QDJ-00bppz; Sun, 12 Dec 2021 15:50:52 +0100 From: Stefan Wahren To: "David S . Miller" , Jakub Kicinski Cc: Rob Herring , Andrew Lunn , Michael Heimpold , jimmy.shen@vertexcom.com, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, devicetree@vger.kernel.org, Stefan Wahren Subject: [PATCH V3 net-next 2/3] dt-bindings: net: add Vertexcom MSE102x support Date: Sun, 12 Dec 2021 15:50:26 +0100 Message-Id: <1639320627-8827-3-git-send-email-stefan.wahren@i2se.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1639320627-8827-1-git-send-email-stefan.wahren@i2se.com> References: <1639320627-8827-1-git-send-email-stefan.wahren@i2se.com> X-Provags-ID: V03:K1:k1exPWyfeyTQ/GsfXXPCmiGHaAWc5aA/3zMi644sT3GsBEt8EX9 sL6szsBpPKf7Ly6Z+0WAi02Xt1oKsMhe46MqwEeapBpCeXqQXybW4aX1tFW5V1sCk40C/hW JZzoJklLjEY1oGlCsKfe6Lf5Cw0Ec4h1kIh9bvlLFs7ZbDh37Qc3Vk0z2imB5M3fbSA0ni7 fTAQKE977ukl/Fw3zDx6g== X-UI-Out-Filterresults: notjunk:1;V03:K0:W9dumw6KuJE=:GRsGTtG6HvkbgURhtuCCUu mgGmIQXVk0hGzQAXV54t5J+joySlkdKfvKZpskLofvcpkbQGfmp2SQj/VLJ9WyeY3DR7PITlC JlAVMLw7i92u+yeccn7eoZIDWvsIpHnu3sIe39irls/ztIm+BWSW9gn6wV6gu2NHoK0M+WXdC ML63YEE3se29CXxKgJJu8RcYpZBToQRQZMOgd2nUpN5RD7uD7bwdtbBJINoUPZnLvX8O9ZVCg xNLfNygSfotOuD5pA+soixnS4PmT9sW6JJYmSlZVw4KqLMmmnwxCz1rPwXmkSxmjlHRf9+A2U 5UNBlKT6FYmlFosH/PoxjvoIOFI/3/L7D8eRf7G+Zj7D6kk/deF0eohWxBXpIqPeCLaHAizlN 8m9bsfjrapbxF3Qc4tjP+zh1X+ixJkE+K5bDE0ZIjIvhvFbBfXahXd9i6E0EuolWAjNepAi78 I4/HKzrZ4JYLtIiEQ7XUVTG1GF6sUzED9Q+Li/ZSHojhdqNYyo1UXlSk9nSW6rSpUhkpwFFbE 0xGpjNqU7A01nulFCLrVtNCB7MKcb5wPq6cVPy9XyKE8mX+wTXdBDGntBUMLLpak/f826BKuU si/13aqcJtJfSvaSUdu7ouchSdE5iIlAbX/QL3pC6UiaaX/c8sFpOqOnMB8igqdrAHpuixufs Y32drLz3g6B0mwo5gcBhypLl7MLNkI/67u8rGwMq+HBqDFRLwzC9zyFhio4dTxTMOAV3MiaKH Ib9tbPymTFdvfGXx Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org Add devicetree binding for the Vertexcom MSE102x Homeplug GreenPHY chip as SPI device. Signed-off-by: Stefan Wahren Reviewed-by: Rob Herring --- .../devicetree/bindings/net/vertexcom-mse102x.yaml | 71 ++++++++++++++++++++++ 1 file changed, 71 insertions(+) create mode 100644 Documentation/devicetree/bindings/net/vertexcom-mse102x.yaml diff --git a/Documentation/devicetree/bindings/net/vertexcom-mse102x.yaml b/Documentation/devicetree/bindings/net/vertexcom-mse102x.yaml new file mode 100644 index 0000000..8156a9a --- /dev/null +++ b/Documentation/devicetree/bindings/net/vertexcom-mse102x.yaml @@ -0,0 +1,71 @@ +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) +%YAML 1.2 +--- +$id: "http://devicetree.org/schemas/net/vertexcom-mse102x.yaml#" +$schema: "http://devicetree.org/meta-schemas/core.yaml#" + +title: The Vertexcom MSE102x (SPI) Device Tree Bindings + +maintainers: + - Stefan Wahren + +description: + Vertexcom's MSE102x are a family of HomePlug GreenPHY chips. + They can be connected either via RGMII, RMII or SPI to a host CPU. + + In order to use a MSE102x chip as SPI device, it must be defined as + a child of an SPI master device in the device tree. + + More information can be found at + http://www.vertexcom.com/doc/MSE1022%20Product%20Brief.pdf + +allOf: + - $ref: ethernet-controller.yaml# + +properties: + compatible: + enum: + - vertexcom,mse1021 + - vertexcom,mse1022 + + reg: + maxItems: 1 + + interrupts: + maxItems: 1 + + spi-cpha: true + + spi-cpol: true + + spi-max-frequency: + minimum: 6000000 + maximum: 7142857 + +required: + - compatible + - reg + - interrupts + - spi-cpha + - spi-cpol + - spi-max-frequency + +additionalProperties: false + +examples: + - | + #include + spi0 { + #address-cells = <1>; + #size-cells = <0>; + + ethernet@0 { + compatible = "vertexcom,mse1021"; + reg = <0>; + interrupt-parent = <&gpio>; + interrupts = <23 IRQ_TYPE_EDGE_RISING>; + spi-cpha; + spi-cpol; + spi-max-frequency = <7142857>; + }; + }; From patchwork Sun Dec 12 14:50:27 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Stefan Wahren X-Patchwork-Id: 12672285 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 Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id CE360C4332F for ; Sun, 12 Dec 2021 14:51:21 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S231544AbhLLOvU (ORCPT ); Sun, 12 Dec 2021 09:51:20 -0500 Received: from mout.kundenserver.de ([212.227.126.130]:34131 "EHLO mout.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230200AbhLLOvL (ORCPT ); Sun, 12 Dec 2021 09:51:11 -0500 Received: from localhost.localdomain ([37.4.249.122]) by mrelayeu.kundenserver.de (mreue011 [212.227.15.167]) with ESMTPSA (Nemesis) id 1MgNMV-1mJnJi0sFb-00hsFB; Sun, 12 Dec 2021 15:50:53 +0100 From: Stefan Wahren To: "David S . Miller" , Jakub Kicinski Cc: Rob Herring , Andrew Lunn , Michael Heimpold , jimmy.shen@vertexcom.com, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, devicetree@vger.kernel.org, Stefan Wahren Subject: [PATCH V3 net-next 3/3] net: vertexcom: Add MSE102x SPI support Date: Sun, 12 Dec 2021 15:50:27 +0100 Message-Id: <1639320627-8827-4-git-send-email-stefan.wahren@i2se.com> X-Mailer: git-send-email 2.7.4 In-Reply-To: <1639320627-8827-1-git-send-email-stefan.wahren@i2se.com> References: <1639320627-8827-1-git-send-email-stefan.wahren@i2se.com> X-Provags-ID: V03:K1:ZrgPtGvXG4nDZsxwHCCbDcYTCMxn+OHARsfk38rkmsUODif8Qb7 9Z/Exu+Zi7U4P7fm0oITUHVXr2s+KIB9NxUngUatRsah5RGNEBHDqf++iVOqy2gWrKluo7r lk3QoCDOpZtcFpC3s011JwviTKV00VayMsA9KNxjsm+zDbShmvsntX0Q/fNE3g9ZEtahBlV D0SITrFOHdGJrr3+72JLg== X-UI-Out-Filterresults: notjunk:1;V03:K0:Rn7fH91TfW0=:+qRQSpStUH6ffvY143c++q MHPxNAXFZ6N3yMyKeYZgry0qxogsF48StqIyDfCbjSAwblavsa/TMew3Xx9P9mrpxGRbBjEkJ kLhwI5t4Mo00+XZbe2hLeKI+6DWtEYQSouenI7Tu4NCh5lT/1e8s1GxBC1oSqxaRMfUiN2oQD b3wdmc+VRtnmm8ym3jN4CFlgXfeCcebDnu6bx/VCSZKQggOjWYlyspN7PhRLDobjuBIEvBfkD VKiPIg9PF1MW9TUgpfskwtN976X2ZMmtrPhQsfwNoczuxGVhp3D9zeapXMYYEsXnP7WsrOYTN 2pytJf9d+qYL/YEu0MGpfruRAEnWv3FoQHFy2L+ts0hVFzuNJGQak8gn8tSOHaPXTdk0R7n/r SnXYXJEaah93HtH0+FcbwMnfOEykPt1djg+08A6YAHPz6OB1XZYPOVIGw5crDwnpDfelya0Mh Dk2pNfnkyFrd85IqFK4XSRThcj4Fy9EqszOA/NBjvJWHVMWJaYBJEOp/26ExUMXNxfeDA9vEJ ZEHRPSkNtnIBvMVltlRg8OLPtNzyYFkaXUS3fsZ8uF5M8SLJT3EQmTMQSV3U/mOuTp/dtoZKW UjJLzQ/EIlfUphbv3x0lKITdggsIMYgG1VFxnXuNuTCk1pY5NM9Zyp17UwyG0SJzfYNFvU7pM gZs7ODa7UzBt2Hejs5XZx3r4pEiEl0TegeKtIk+IZ14WL65cjL1qtRpwu4OhsfKJwBpLjNICr /pDk4kiCaNaUovyI Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org This implements an SPI protocol driver for Vertexcom MSE102x Homeplug GreenPHY chip. Signed-off-by: Stefan Wahren --- drivers/net/ethernet/Kconfig | 1 + drivers/net/ethernet/Makefile | 1 + drivers/net/ethernet/vertexcom/Kconfig | 25 + drivers/net/ethernet/vertexcom/Makefile | 6 + drivers/net/ethernet/vertexcom/mse102x.c | 769 +++++++++++++++++++++++++++++++ 5 files changed, 802 insertions(+) create mode 100644 drivers/net/ethernet/vertexcom/Kconfig create mode 100644 drivers/net/ethernet/vertexcom/Makefile create mode 100644 drivers/net/ethernet/vertexcom/mse102x.c diff --git a/drivers/net/ethernet/Kconfig b/drivers/net/ethernet/Kconfig index 027cbac..db3ec47 100644 --- a/drivers/net/ethernet/Kconfig +++ b/drivers/net/ethernet/Kconfig @@ -183,6 +183,7 @@ source "drivers/net/ethernet/tehuti/Kconfig" source "drivers/net/ethernet/ti/Kconfig" source "drivers/net/ethernet/toshiba/Kconfig" source "drivers/net/ethernet/tundra/Kconfig" +source "drivers/net/ethernet/vertexcom/Kconfig" source "drivers/net/ethernet/via/Kconfig" source "drivers/net/ethernet/wiznet/Kconfig" source "drivers/net/ethernet/xilinx/Kconfig" diff --git a/drivers/net/ethernet/Makefile b/drivers/net/ethernet/Makefile index 33d30b6..8a87c10 100644 --- a/drivers/net/ethernet/Makefile +++ b/drivers/net/ethernet/Makefile @@ -93,6 +93,7 @@ obj-$(CONFIG_NET_VENDOR_TEHUTI) += tehuti/ obj-$(CONFIG_NET_VENDOR_TI) += ti/ obj-$(CONFIG_NET_VENDOR_TOSHIBA) += toshiba/ obj-$(CONFIG_NET_VENDOR_TUNDRA) += tundra/ +obj-$(CONFIG_NET_VENDOR_VERTEXCOM) += vertexcom/ obj-$(CONFIG_NET_VENDOR_VIA) += via/ obj-$(CONFIG_NET_VENDOR_WIZNET) += wiznet/ obj-$(CONFIG_NET_VENDOR_XILINX) += xilinx/ diff --git a/drivers/net/ethernet/vertexcom/Kconfig b/drivers/net/ethernet/vertexcom/Kconfig new file mode 100644 index 0000000..4184a63 --- /dev/null +++ b/drivers/net/ethernet/vertexcom/Kconfig @@ -0,0 +1,25 @@ +# SPDX-License-Identifier: GPL-2.0-only +# +# Vertexcom network device configuration +# + +config NET_VENDOR_VERTEXCOM + bool "Vertexcom devices" + default y + help + If you have a network (Ethernet) card belonging to this class, say Y. + + Note that the answer to this question doesn't directly affect the + kernel: saying N will just cause the configurator to skip all + the questions about Vertexcom cards. If you say Y, you will be asked + for your specific card in the following questions. + +if NET_VENDOR_VERTEXCOM + +config MSE102X + tristate "Vertexcom MSE102x SPI" + depends on SPI + help + SPI driver for Vertexcom MSE102x SPI attached network chip. + +endif # NET_VENDOR_VERTEXCOM diff --git a/drivers/net/ethernet/vertexcom/Makefile b/drivers/net/ethernet/vertexcom/Makefile new file mode 100644 index 0000000..f8b12e3 --- /dev/null +++ b/drivers/net/ethernet/vertexcom/Makefile @@ -0,0 +1,6 @@ +# SPDX-License-Identifier: GPL-2.0 +# +# Makefile for the Vertexcom network device drivers. +# + +obj-$(CONFIG_MSE102X) += mse102x.o diff --git a/drivers/net/ethernet/vertexcom/mse102x.c b/drivers/net/ethernet/vertexcom/mse102x.c new file mode 100644 index 0000000..a3c2426 --- /dev/null +++ b/drivers/net/ethernet/vertexcom/mse102x.c @@ -0,0 +1,769 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* Copyright (C) 2021 in-tech smart charging GmbH + * + * driver is based on micrel/ks8851_spi.c + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include + +#define MSG_DEFAULT (NETIF_MSG_DRV | NETIF_MSG_PROBE | NETIF_MSG_LINK | \ + NETIF_MSG_TIMER) + +#define DRV_NAME "mse102x" + +#define DET_CMD 0x0001 +#define DET_SOF 0x0002 +#define DET_DFT 0x55AA + +#define CMD_SHIFT 12 +#define CMD_RTS (0x1 << CMD_SHIFT) +#define CMD_CTR (0x2 << CMD_SHIFT) + +#define CMD_MASK GENMASK(15, CMD_SHIFT) +#define LEN_MASK GENMASK(CMD_SHIFT - 1, 0) + +#define DET_CMD_LEN 4 +#define DET_SOF_LEN 2 +#define DET_DFT_LEN 2 + +#define MIN_FREQ_HZ 6000000 +#define MAX_FREQ_HZ 7142857 + +struct mse102x_stats { + u64 xfer_err; + u64 invalid_cmd; + u64 invalid_ctr; + u64 invalid_dft; + u64 invalid_len; + u64 invalid_rts; + u64 invalid_sof; + u64 tx_timeout; +}; + +static const char mse102x_gstrings_stats[][ETH_GSTRING_LEN] = { + "SPI transfer errors", + "Invalid command", + "Invalid CTR", + "Invalid DFT", + "Invalid frame length", + "Invalid RTS", + "Invalid SOF", + "TX timeout", +}; + +struct mse102x_net { + struct net_device *ndev; + + u8 rxd[8]; + u8 txd[8]; + + u32 msg_enable ____cacheline_aligned; + + struct sk_buff_head txq; + struct mse102x_stats stats; +}; + +struct mse102x_net_spi { + struct mse102x_net mse102x; + struct mutex lock; /* Protect SPI frame transfer */ + struct work_struct tx_work; + struct spi_device *spidev; + struct spi_message spi_msg; + struct spi_transfer spi_xfer; + +#ifdef CONFIG_DEBUG_FS + struct dentry *device_root; +#endif +}; + +#define to_mse102x_spi(mse) container_of((mse), struct mse102x_net_spi, mse102x) + +#ifdef CONFIG_DEBUG_FS + +static int mse102x_info_show(struct seq_file *s, void *what) +{ + struct mse102x_net_spi *mses = s->private; + + seq_printf(s, "TX ring size : %u\n", + skb_queue_len(&mses->mse102x.txq)); + + seq_printf(s, "IRQ : %d\n", + mses->spidev->irq); + + seq_printf(s, "SPI effective speed : %lu\n", + (unsigned long)mses->spi_xfer.effective_speed_hz); + seq_printf(s, "SPI mode : %x\n", + mses->spidev->mode); + + return 0; +} +DEFINE_SHOW_ATTRIBUTE(mse102x_info); + +static void mse102x_init_device_debugfs(struct mse102x_net_spi *mses) +{ + mses->device_root = debugfs_create_dir(dev_name(&mses->mse102x.ndev->dev), + NULL); + + debugfs_create_file("info", S_IFREG | 0444, mses->device_root, mses, + &mse102x_info_fops); +} + +static void mse102x_remove_device_debugfs(struct mse102x_net_spi *mses) +{ + debugfs_remove_recursive(mses->device_root); +} + +#else /* CONFIG_DEBUG_FS */ + +static void mse102x_init_device_debugfs(struct mse102x_net_spi *mses) +{ +} + +static void mse102x_remove_device_debugfs(struct mse102x_net_spi *mses) +{ +} + +#endif + +/* SPI register read/write calls. + * + * All these calls issue SPI transactions to access the chip's registers. They + * all require that the necessary lock is held to prevent accesses when the + * chip is busy transferring packet data. + */ + +static void mse102x_tx_cmd_spi(struct mse102x_net *mse, u16 cmd) +{ + struct mse102x_net_spi *mses = to_mse102x_spi(mse); + struct spi_transfer *xfer = &mses->spi_xfer; + struct spi_message *msg = &mses->spi_msg; + __be16 txb[2]; + int ret; + + txb[0] = cpu_to_be16(DET_CMD); + txb[1] = cpu_to_be16(cmd); + + xfer->tx_buf = txb; + xfer->rx_buf = NULL; + xfer->len = DET_CMD_LEN; + + ret = spi_sync(mses->spidev, msg); + if (ret < 0) { + netdev_err(mse->ndev, "%s: spi_sync() failed: %d\n", + __func__, ret); + mse->stats.xfer_err++; + } +} + +static int mse102x_rx_cmd_spi(struct mse102x_net *mse, u8 *rxb) +{ + struct mse102x_net_spi *mses = to_mse102x_spi(mse); + struct spi_transfer *xfer = &mses->spi_xfer; + struct spi_message *msg = &mses->spi_msg; + __be16 *txb = (__be16 *)mse->txd; + __be16 *cmd = (__be16 *)mse->rxd; + u8 *trx = mse->rxd; + int ret; + + txb[0] = 0; + txb[1] = 0; + + xfer->tx_buf = txb; + xfer->rx_buf = trx; + xfer->len = DET_CMD_LEN; + + ret = spi_sync(mses->spidev, msg); + if (ret < 0) { + netdev_err(mse->ndev, "%s: spi_sync() failed: %d\n", + __func__, ret); + mse->stats.xfer_err++; + } else if (*cmd != cpu_to_be16(DET_CMD)) { + net_dbg_ratelimited("%s: Unexpected response (0x%04x)\n", + __func__, *cmd); + mse->stats.invalid_cmd++; + ret = -EIO; + } else { + memcpy(rxb, trx + 2, 2); + } + + return ret; +} + +static inline void mse102x_push_header(struct sk_buff *skb) +{ + __be16 *header = skb_push(skb, DET_SOF_LEN); + + *header = cpu_to_be16(DET_SOF); +} + +static inline void mse102x_put_footer(struct sk_buff *skb) +{ + __be16 *footer = skb_put(skb, DET_DFT_LEN); + + *footer = cpu_to_be16(DET_DFT); +} + +static int mse102x_tx_frame_spi(struct mse102x_net *mse, struct sk_buff *txp, + unsigned int pad) +{ + struct mse102x_net_spi *mses = to_mse102x_spi(mse); + struct spi_transfer *xfer = &mses->spi_xfer; + struct spi_message *msg = &mses->spi_msg; + struct sk_buff *tskb; + int ret; + + netif_dbg(mse, tx_queued, mse->ndev, "%s: skb %p, %d@%p\n", + __func__, txp, txp->len, txp->data); + + if ((skb_headroom(txp) < DET_SOF_LEN) || + (skb_tailroom(txp) < DET_DFT_LEN + pad)) { + tskb = skb_copy_expand(txp, DET_SOF_LEN, DET_DFT_LEN + pad, + GFP_KERNEL); + if (!tskb) + return -ENOMEM; + + dev_kfree_skb(txp); + txp = tskb; + } + + mse102x_push_header(txp); + + if (pad) + skb_put_zero(txp, pad); + + mse102x_put_footer(txp); + + xfer->tx_buf = txp->data; + xfer->rx_buf = NULL; + xfer->len = txp->len; + + ret = spi_sync(mses->spidev, msg); + if (ret < 0) { + netdev_err(mse->ndev, "%s: spi_sync() failed: %d\n", + __func__, ret); + mse->stats.xfer_err++; + } + + return ret; +} + +static int mse102x_rx_frame_spi(struct mse102x_net *mse, u8 *buff, + unsigned int frame_len) +{ + struct mse102x_net_spi *mses = to_mse102x_spi(mse); + struct spi_transfer *xfer = &mses->spi_xfer; + struct spi_message *msg = &mses->spi_msg; + __be16 *sof = (__be16 *)buff; + __be16 *dft = (__be16 *)(buff + DET_SOF_LEN + frame_len); + int ret; + + xfer->rx_buf = buff; + xfer->tx_buf = NULL; + xfer->len = DET_SOF_LEN + frame_len + DET_DFT_LEN; + + ret = spi_sync(mses->spidev, msg); + if (ret < 0) { + netdev_err(mse->ndev, "%s: spi_sync() failed: %d\n", + __func__, ret); + mse->stats.xfer_err++; + } else if (*sof != cpu_to_be16(DET_SOF)) { + netdev_dbg(mse->ndev, "%s: SPI start of frame is invalid (0x%04x)\n", + __func__, *sof); + mse->stats.invalid_sof++; + ret = -EIO; + } else if (*dft != cpu_to_be16(DET_DFT)) { + netdev_dbg(mse->ndev, "%s: SPI frame tail is invalid (0x%04x)\n", + __func__, *dft); + mse->stats.invalid_dft++; + ret = -EIO; + } + + return ret; +} + +static void mse102x_dump_packet(const char *msg, int len, const char *data) +{ + printk(KERN_DEBUG ": %s - packet len:%d\n", msg, len); + print_hex_dump(KERN_DEBUG, "pk data: ", DUMP_PREFIX_OFFSET, 16, 1, + data, len, true); +} + +static void mse102x_rx_pkt_spi(struct mse102x_net *mse) +{ + struct sk_buff *skb; + unsigned int rxalign; + unsigned int rxlen; + __be16 rx = 0; + u16 cmd_resp; + u8 *rxpkt; + int ret; + + mse102x_tx_cmd_spi(mse, CMD_CTR); + ret = mse102x_rx_cmd_spi(mse, (u8 *)&rx); + cmd_resp = be16_to_cpu(rx); + + if (ret || ((cmd_resp & CMD_MASK) != CMD_RTS)) { + usleep_range(50, 100); + + mse102x_tx_cmd_spi(mse, CMD_CTR); + ret = mse102x_rx_cmd_spi(mse, (u8 *)&rx); + if (ret) + return; + + cmd_resp = be16_to_cpu(rx); + if ((cmd_resp & CMD_MASK) != CMD_RTS) { + net_dbg_ratelimited("%s: Unexpected response (0x%04x)\n", + __func__, cmd_resp); + mse->stats.invalid_rts++; + return; + } + + net_dbg_ratelimited("%s: Unexpected response to first CMD\n", + __func__); + } + + rxlen = cmd_resp & LEN_MASK; + if (!rxlen) { + net_dbg_ratelimited("%s: No frame length defined\n", __func__); + mse->stats.invalid_len++; + return; + } + + rxalign = ALIGN(rxlen + DET_SOF_LEN + DET_DFT_LEN, 4); + skb = netdev_alloc_skb_ip_align(mse->ndev, rxalign); + if (!skb) + return; + + /* 2 bytes Start of frame (before ethernet header) + * 2 bytes Data frame tail (after ethernet frame) + * They are copied, but ignored. + */ + rxpkt = skb_put(skb, rxlen) - DET_SOF_LEN; + if (mse102x_rx_frame_spi(mse, rxpkt, rxlen)) { + mse->ndev->stats.rx_errors++; + dev_kfree_skb(skb); + return; + } + + if (netif_msg_pktdata(mse)) + mse102x_dump_packet(__func__, skb->len, skb->data); + + skb->protocol = eth_type_trans(skb, mse->ndev); + netif_rx_ni(skb); + + mse->ndev->stats.rx_packets++; + mse->ndev->stats.rx_bytes += rxlen; +} + +static int mse102x_tx_pkt_spi(struct mse102x_net *mse, struct sk_buff *txb, + unsigned long work_timeout) +{ + unsigned int pad = 0; + __be16 rx = 0; + u16 cmd_resp; + int ret; + bool first = true; + + if (txb->len < 60) + pad = 60 - txb->len; + + while (1) { + mse102x_tx_cmd_spi(mse, CMD_RTS | (txb->len + pad)); + ret = mse102x_rx_cmd_spi(mse, (u8 *)&rx); + cmd_resp = be16_to_cpu(rx); + + if (!ret) { + /* ready to send frame ? */ + if (cmd_resp == CMD_CTR) + break; + + net_dbg_ratelimited("%s: Unexpected response (0x%04x)\n", + __func__, cmd_resp); + mse->stats.invalid_ctr++; + } + + /* It's not predictable how long / many retries it takes to + * send at least one packet, so TX timeouts are possible. + * That's the reason why the netdev watchdog is not used here. + */ + if (time_after(jiffies, work_timeout)) + return -ETIMEDOUT; + + if (first) { + /* throttle at first issue */ + netif_stop_queue(mse->ndev); + /* fast retry */ + usleep_range(50, 100); + first = false; + } else { + msleep(20); + } + }; + + ret = mse102x_tx_frame_spi(mse, txb, pad); + if (ret) + net_dbg_ratelimited("%s: Failed to send (%d), drop frame\n", + __func__, ret); + + return ret; +} + +#define TX_QUEUE_MAX 10 + +static void mse102x_tx_work(struct work_struct *work) +{ + /* Make sure timeout is sufficient to transfer TX_QUEUE_MAX frames */ + unsigned long work_timeout = jiffies + msecs_to_jiffies(1000); + struct mse102x_net_spi *mses; + struct mse102x_net *mse; + struct sk_buff *txb; + int ret = 0; + + mses = container_of(work, struct mse102x_net_spi, tx_work); + mse = &mses->mse102x; + + while ((txb = skb_dequeue(&mse->txq))) { + mutex_lock(&mses->lock); + ret = mse102x_tx_pkt_spi(mse, txb, work_timeout); + mutex_unlock(&mses->lock); + if (ret) { + mse->ndev->stats.tx_dropped++; + } else { + mse->ndev->stats.tx_bytes += txb->len; + mse->ndev->stats.tx_packets++; + } + + dev_kfree_skb(txb); + } + + if (ret == -ETIMEDOUT) { + if (netif_msg_timer(mse)) + netdev_err(mse->ndev, "tx work timeout\n"); + + mse->stats.tx_timeout++; + } + + netif_wake_queue(mse->ndev); +} + +static netdev_tx_t mse102x_start_xmit_spi(struct sk_buff *skb, + struct net_device *ndev) +{ + struct mse102x_net *mse = netdev_priv(ndev); + struct mse102x_net_spi *mses = to_mse102x_spi(mse); + + netif_dbg(mse, tx_queued, ndev, + "%s: skb %p, %d@%p\n", __func__, skb, skb->len, skb->data); + + skb_queue_tail(&mse->txq, skb); + + if (skb_queue_len(&mse->txq) >= TX_QUEUE_MAX) + netif_stop_queue(ndev); + + schedule_work(&mses->tx_work); + + return NETDEV_TX_OK; +} + +static void mse102x_init_mac(struct mse102x_net *mse, struct device_node *np) +{ + struct net_device *ndev = mse->ndev; + int ret = of_get_ethdev_address(np, ndev); + + if (ret) { + eth_hw_addr_random(ndev); + netdev_err(ndev, "Using random MAC address: %pM\n", + ndev->dev_addr); + } +} + +/* Assumption: this is called for every incoming packet */ +static irqreturn_t mse102x_irq(int irq, void *_mse) +{ + struct mse102x_net *mse = _mse; + struct mse102x_net_spi *mses = to_mse102x_spi(mse); + + mutex_lock(&mses->lock); + mse102x_rx_pkt_spi(mse); + mutex_unlock(&mses->lock); + + return IRQ_HANDLED; +} + +static int mse102x_net_open(struct net_device *ndev) +{ + struct mse102x_net *mse = netdev_priv(ndev); + int ret; + + ret = request_threaded_irq(ndev->irq, NULL, mse102x_irq, IRQF_ONESHOT, + ndev->name, mse); + if (ret < 0) { + netdev_err(ndev, "Failed to get irq: %d\n", ret); + return ret; + } + + netif_dbg(mse, ifup, ndev, "opening\n"); + + netif_start_queue(ndev); + + netif_carrier_on(ndev); + + netif_dbg(mse, ifup, ndev, "network device up\n"); + + return 0; +} + +static int mse102x_net_stop(struct net_device *ndev) +{ + struct mse102x_net *mse = netdev_priv(ndev); + struct mse102x_net_spi *mses = to_mse102x_spi(mse); + + netif_info(mse, ifdown, ndev, "shutting down\n"); + + netif_carrier_off(mse->ndev); + + /* stop any outstanding work */ + flush_work(&mses->tx_work); + + netif_stop_queue(ndev); + + skb_queue_purge(&mse->txq); + + free_irq(ndev->irq, mse); + + return 0; +} + +static const struct net_device_ops mse102x_netdev_ops = { + .ndo_open = mse102x_net_open, + .ndo_stop = mse102x_net_stop, + .ndo_start_xmit = mse102x_start_xmit_spi, + .ndo_set_mac_address = eth_mac_addr, + .ndo_validate_addr = eth_validate_addr, +}; + +/* ethtool support */ + +static void mse102x_get_drvinfo(struct net_device *ndev, + struct ethtool_drvinfo *di) +{ + strscpy(di->driver, DRV_NAME, sizeof(di->driver)); + strscpy(di->bus_info, dev_name(ndev->dev.parent), sizeof(di->bus_info)); +} + +static u32 mse102x_get_msglevel(struct net_device *ndev) +{ + struct mse102x_net *mse = netdev_priv(ndev); + + return mse->msg_enable; +} + +static void mse102x_set_msglevel(struct net_device *ndev, u32 to) +{ + struct mse102x_net *mse = netdev_priv(ndev); + + mse->msg_enable = to; +} + +static void mse102x_get_ethtool_stats(struct net_device *ndev, + struct ethtool_stats *estats, u64 *data) +{ + struct mse102x_net *mse = netdev_priv(ndev); + struct mse102x_stats *st = &mse->stats; + + memcpy(data, st, ARRAY_SIZE(mse102x_gstrings_stats) * sizeof(u64)); +} + +static void mse102x_get_strings(struct net_device *ndev, u32 stringset, u8 *buf) +{ + switch (stringset) { + case ETH_SS_STATS: + memcpy(buf, &mse102x_gstrings_stats, + sizeof(mse102x_gstrings_stats)); + break; + default: + WARN_ON(1); + break; + } +} + +static int mse102x_get_sset_count(struct net_device *ndev, int sset) +{ + switch (sset) { + case ETH_SS_STATS: + return ARRAY_SIZE(mse102x_gstrings_stats); + default: + return -EINVAL; + } +} + +static const struct ethtool_ops mse102x_ethtool_ops = { + .get_drvinfo = mse102x_get_drvinfo, + .get_link = ethtool_op_get_link, + .get_msglevel = mse102x_get_msglevel, + .set_msglevel = mse102x_set_msglevel, + .get_ethtool_stats = mse102x_get_ethtool_stats, + .get_strings = mse102x_get_strings, + .get_sset_count = mse102x_get_sset_count, +}; + +/* driver bus management functions */ + +#ifdef CONFIG_PM_SLEEP + +static int mse102x_suspend(struct device *dev) +{ + struct mse102x_net *mse = dev_get_drvdata(dev); + struct net_device *ndev = mse->ndev; + + if (netif_running(ndev)) { + netif_device_detach(ndev); + mse102x_net_stop(ndev); + } + + return 0; +} + +static int mse102x_resume(struct device *dev) +{ + struct mse102x_net *mse = dev_get_drvdata(dev); + struct net_device *ndev = mse->ndev; + + if (netif_running(ndev)) { + mse102x_net_open(ndev); + netif_device_attach(ndev); + } + + return 0; +} +#endif + +static SIMPLE_DEV_PM_OPS(mse102x_pm_ops, mse102x_suspend, mse102x_resume); + +static int mse102x_probe_spi(struct spi_device *spi) +{ + struct device *dev = &spi->dev; + struct mse102x_net_spi *mses; + struct net_device *ndev; + struct mse102x_net *mse; + int ret; + + spi->bits_per_word = 8; + spi->mode |= SPI_MODE_3; + /* enforce minimum speed to ensure device functionality */ + spi->master->min_speed_hz = MIN_FREQ_HZ; + + if (!spi->max_speed_hz) + spi->max_speed_hz = MAX_FREQ_HZ; + + if (spi->max_speed_hz < MIN_FREQ_HZ || + spi->max_speed_hz > MAX_FREQ_HZ) { + dev_err(&spi->dev, "SPI max frequency out of range (min: %u, max: %u)\n", + MIN_FREQ_HZ, MAX_FREQ_HZ); + return -EINVAL; + } + + ret = spi_setup(spi); + if (ret < 0) { + dev_err(&spi->dev, "Unable to setup SPI device: %d\n", ret); + return ret; + } + + ndev = devm_alloc_etherdev(dev, sizeof(struct mse102x_net_spi)); + if (!ndev) + return -ENOMEM; + + ndev->needed_tailroom += ALIGN(DET_DFT_LEN, 4); + ndev->needed_headroom += ALIGN(DET_SOF_LEN, 4); + ndev->priv_flags &= ~IFF_TX_SKB_SHARING; + ndev->tx_queue_len = 100; + + mse = netdev_priv(ndev); + mses = to_mse102x_spi(mse); + + mses->spidev = spi; + mutex_init(&mses->lock); + INIT_WORK(&mses->tx_work, mse102x_tx_work); + + /* initialise pre-made spi transfer messages */ + spi_message_init(&mses->spi_msg); + spi_message_add_tail(&mses->spi_xfer, &mses->spi_msg); + + ndev->irq = spi->irq; + mse->ndev = ndev; + + /* set the default message enable */ + mse->msg_enable = netif_msg_init(-1, MSG_DEFAULT); + + skb_queue_head_init(&mse->txq); + + SET_NETDEV_DEV(ndev, dev); + + dev_set_drvdata(dev, mse); + + netif_carrier_off(mse->ndev); + ndev->netdev_ops = &mse102x_netdev_ops; + ndev->ethtool_ops = &mse102x_ethtool_ops; + + mse102x_init_mac(mse, dev->of_node); + + ret = register_netdev(ndev); + if (ret) { + dev_err(dev, "failed to register network device: %d\n", ret); + return ret; + } + + mse102x_init_device_debugfs(mses); + + return 0; +} + +static int mse102x_remove_spi(struct spi_device *spi) +{ + struct mse102x_net *mse = dev_get_drvdata(&spi->dev); + struct mse102x_net_spi *mses = to_mse102x_spi(mse); + + if (netif_msg_drv(mse)) + dev_info(&spi->dev, "remove\n"); + + mse102x_remove_device_debugfs(mses); + unregister_netdev(mse->ndev); + + return 0; +} + +static const struct of_device_id mse102x_match_table[] = { + { .compatible = "vertexcom,mse1021" }, + { .compatible = "vertexcom,mse1022" }, + { } +}; +MODULE_DEVICE_TABLE(of, mse102x_match_table); + +static struct spi_driver mse102x_driver = { + .driver = { + .name = DRV_NAME, + .of_match_table = mse102x_match_table, + .pm = &mse102x_pm_ops, + }, + .probe = mse102x_probe_spi, + .remove = mse102x_remove_spi, +}; +module_spi_driver(mse102x_driver); + +MODULE_DESCRIPTION("MSE102x Network driver"); +MODULE_AUTHOR("Stefan Wahren "); +MODULE_LICENSE("GPL"); +MODULE_ALIAS("spi:" DRV_NAME);