linux/drivers/usb/gadget/ether.c

/*
 * ether.c -- Ethernet gadget driver, with CDC and non-CDC options
 *
 * Copyright (C) 2003-2005,2008 David Brownell
 * Copyright (C) 2003-2004 Robert Schwebel, Benedikt Spranger
 * Copyright (C) 2008 Nokia Corporation
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 */

/* #define VERBOSE_DEBUG */

#include <linux/kernel.h>
#include <linux/utsname.h>

#include "u_ether.h"


/*
 * Ethernet gadget driver -- with CDC and non-CDC options
 * Builds on hardware support for a full duplex link.
 *
 * CDC Ethernet is the standard USB solution for sending Ethernet frames
 * using USB.  Real hardware tends to use the same framing protocol but look
 * different for control features.  This driver strongly prefers to use
 * this USB-IF standard as its open-systems interoperability solution;
 * most host side USB stacks (except from Microsoft) support it.
 *
 * This is sometimes called "CDC ECM" (Ethernet Control Model) to support
 * TLA-soup.  "CDC ACM" (Abstract Control Model) is for modems, and a new
 * "CDC EEM" (Ethernet Emulation Model) is starting to spread.
 *
 * There's some hardware that can't talk CDC ECM.  We make that hardware
 * implement a "minimalist" vendor-agnostic CDC core:  same framing, but
 * link-level setup only requires activating the configuration.  Only the
 * endpoint descriptors, and product/vendor IDs, are relevant; no control
 * operations are available.  Linux supports it, but other host operating
 * systems may not.  (This is a subset of CDC Ethernet.)
 *
 * It turns out that if you add a few descriptors to that "CDC Subset",
 * (Windows) host side drivers from MCCI can treat it as one submode of
 * a proprietary scheme called "SAFE" ... without needing to know about
 * specific product/vendor IDs.  So we do that, making it easier to use
 * those MS-Windows drivers.  Those added descriptors make it resemble a
 * CDC MDLM device, but they don't change device behavior at all.  (See
 * MCCI Engineering report 950198 "SAFE Networking Functions".)
 *
 * A third option is also in use.  Rather than CDC Ethernet, or something
 * simpler, Microsoft pushes their own approach: RNDIS.  The published
 * RNDIS specs are ambiguous and appear to be incomplete, and are also
 * needlessly complex.  They borrow more from CDC ACM than CDC ECM.
 */

#define DRIVER_DESC		"Ethernet Gadget"
#define DRIVER_VERSION		"Memorial Day 2008"

#ifdef CONFIG_USB_ETH_RNDIS
#define PREFIX			"RNDIS/"
#else
#define PREFIX			""
#endif

/*
 * This driver aims for interoperability by using CDC ECM unless
 *
 *		can_support_ecm()
 *
 * returns false, in which case it supports the CDC Subset.  By default,
 * that returns true; most hardware has no problems with CDC ECM, that's
 * a good default.  Previous versions of this driver had no default; this
 * version changes that, removing overhead for new controller support.
 *
 *	IF YOUR HARDWARE CAN'T SUPPORT CDC ECM, UPDATE THAT ROUTINE!
 */

static inline bool has_rndis(void)
{
#ifdef	CONFIG_USB_ETH_RNDIS
	return true;
#else
	return false;
#endif
}

/*-------------------------------------------------------------------------*/

/*
 * Kbuild is not very cooperative with respect to linking separately
 * compiled library objects into one module.  So for now we won't use
 * separate compilation ... ensuring init/exit sections work to shrink
 * the runtime footprint, and giving us at least some parts of what
 * a "gcc --combine ... part1.c part2.c part3.c ... " build would.
 */
#include "composite.c"
#include "usbstring.c"
#include "config.c"
#include "epautoconf.c"

#include "f_ecm.c"
#include "f_subset.c"
#ifdef	CONFIG_USB_ETH_RNDIS
#include "f_rndis.c"
#include "rndis.c"
#endif
#include "u_ether.c"

/*-------------------------------------------------------------------------*/

/* DO NOT REUSE THESE IDs with a protocol-incompatible driver!!  Ever!!
 * Instead:  allocate your own, using normal USB-IF procedures.
 */

/* Thanks to NetChip Technologies for donating this product ID.
 * It's for devices with only CDC Ethernet configurations.
 */
#define CDC_VENDOR_NUM		0x0525	/* NetChip */
#define CDC_PRODUCT_NUM		0xa4a1	/* Linux-USB Ethernet Gadget */

/* For hardware that can't talk CDC, we use the same vendor ID that
 * ARM Linux has used for ethernet-over-usb, both with sa1100 and
 * with pxa250.  We're protocol-compatible, if the host-side drivers
 * use the endpoint descriptors.  bcdDevice (version) is nonzero, so
 * drivers that need to hard-wire endpoint numbers have a hook.
 *
 * The protocol is a minimal subset of CDC Ether, which works on any bulk
 * hardware that's not deeply broken ... even on hardware that can't talk
 * RNDIS (like SA-1100, with no interrupt endpoint, or anything that
 * doesn't handle control-OUT).
 */
#define	SIMPLE_VENDOR_NUM	0x049f
#define	SIMPLE_PRODUCT_NUM	0x505a

/* For hardware that can talk RNDIS and either of the above protocols,
 * use this ID ... the windows INF files will know it.  Unless it's
 * used with CDC Ethernet, Linux 2.4 hosts will need updates to choose
 * the non-RNDIS configuration.
 */
#define RNDIS_VENDOR_NUM	0x0525	/* NetChip */
#define RNDIS_PRODUCT_NUM	0xa4a2	/* Ethernet/RNDIS Gadget */

/*-------------------------------------------------------------------------*/

static struct usb_device_descriptor device_desc = {
	.bLength =		sizeof device_desc,
	.bDescriptorType =	USB_DT_DEVICE,

	.bcdUSB =		cpu_to_le16 (0x0200),

	.bDeviceClass =		USB_CLASS_COMM,
	.bDeviceSubClass =	0,
	.bDeviceProtocol =	0,
	/* .bMaxPacketSize0 = f(hardware) */

	/* Vendor and product id defaults change according to what configs
	 * we support.  (As does bNumConfigurations.)  These values can
	 * also be overridden by module parameters.
	 */
	.idVendor =		cpu_to_le16 (CDC_VENDOR_NUM),
	.idProduct =		cpu_to_le16 (CDC_PRODUCT_NUM),
	/* .bcdDevice = f(hardware) */
	/* .iManufacturer = DYNAMIC */
	/* .iProduct = DYNAMIC */
	/* NO SERIAL NUMBER */
	.bNumConfigurations =	1,
};

static struct usb_otg_descriptor otg_descriptor = {
	.bLength =		sizeof otg_descriptor,
	.bDescriptorType =	USB_DT_OTG,

	/* REVISIT SRP-only hardware is possible, although
	 * it would not be called "OTG" ...
	 */
	.bmAttributes =		USB_OTG_SRP | USB_OTG_HNP,
};

static const struct usb_descriptor_header *otg_desc[] = {
	(struct usb_descriptor_header *) &otg_descriptor,
	NULL,
};


/* string IDs are assigned dynamically */

#define STRING_MANUFACTURER_IDX		0
#define STRING_PRODUCT_IDX		1

static char manufacturer[50];

static struct usb_string strings_dev[] = {
	[STRING_MANUFACTURER_IDX].s = manufacturer,
	[STRING_PRODUCT_IDX].s = PREFIX DRIVER_DESC,
	{  } /* end of list */
};

static struct usb_gadget_strings stringtab_dev = {
	.language	= 0x0409,	/* en-us */
	.strings	= strings_dev,
};

static struct usb_gadget_strings *dev_strings[] = {
	&stringtab_dev,
	NULL,
};

static u8 hostaddr[ETH_ALEN];

/*-------------------------------------------------------------------------*/

/*
 * We may not have an RNDIS configuration, but if we do it needs to be
 * the first one present.  That's to make Microsoft's drivers happy,
 * and to follow DOCSIS 1.0 (cable modem standard).
 */
static int __init rndis_do_config(struct usb_configuration *c)
{
	/* FIXME alloc iConfiguration string, set it in c->strings */

	if (gadget_is_otg(c->cdev->gadget)) {
		c->descriptors = otg_desc;
		c->bmAttributes |= USB_CONFIG_ATT_WAKEUP;
	}

	return rndis_bind_config(c, hostaddr);
}

static struct usb_configuration rndis_config_driver = {
	.label			= "RNDIS",
	.bind			= rndis_do_config,
	.bConfigurationValue	= 2,
	/* .iConfiguration = DYNAMIC */
	.bmAttributes		= USB_CONFIG_ATT_SELFPOWER,
};

/*-------------------------------------------------------------------------*/

/*
 * We _always_ have an ECM or CDC Subset configuration.
 */
static int __init eth_do_config(struct usb_configuration *c)
{
	/* FIXME alloc iConfiguration string, set it in c->strings */

	if (gadget_is_otg(c->cdev->gadget)) {
		c->descriptors = otg_desc;
		c->bmAttributes |= USB_CONFIG_ATT_WAKEUP;
	}

	if (can_support_ecm(c->cdev->gadget))
		return ecm_bind_config(c, hostaddr);
	else
		return geth_bind_config(c, hostaddr);
}

static struct usb_configuration eth_config_driver = {
	/* .label = f(hardware) */
	.bind			= eth_do_config,
	.bConfigurationValue	= 1,
	/* .iConfiguration = DYNAMIC */
	.bmAttributes		= USB_CONFIG_ATT_SELFPOWER,
};

/*-------------------------------------------------------------------------*/

static int __init eth_bind(struct usb_composite_dev *cdev)
{
	int			gcnum;
	struct usb_gadget	*gadget = cdev->gadget;
	int			status;

	/* set up network link layer */
	status = gether_setup(cdev->gadget, hostaddr);
	if (status < 0)
		return status;

	/* set up main config label and device descriptor */
	if (can_support_ecm(cdev->gadget)) {
		/* ECM */
		eth_config_driver.label = "CDC Ethernet (ECM)";
	} else {
		/* CDC Subset */
		eth_config_driver.label = "CDC Subset/SAFE";

		device_desc.idVendor = cpu_to_le16(SIMPLE_VENDOR_NUM),
		device_desc.idProduct = cpu_to_le16(SIMPLE_PRODUCT_NUM),
		device_desc.bDeviceClass = USB_CLASS_VENDOR_SPEC;
	}

	if (has_rndis()) {
		/* RNDIS plus ECM-or-Subset */
		device_desc.idVendor = cpu_to_le16(RNDIS_VENDOR_NUM),
		device_desc.idProduct = cpu_to_le16(RNDIS_PRODUCT_NUM),
		device_desc.bNumConfigurations = 2;
	}

	gcnum = usb_gadget_controller_number(gadget);
	if (gcnum >= 0)
		device_desc.bcdDevice = cpu_to_le16(0x0300 | gcnum);
	else {
		/* We assume that can_support_ecm() tells the truth;
		 * but if the controller isn't recognized at all then
		 * that assumption is a bit more likely to be wrong.
		 */
		dev_warn(&gadget->dev,
				"controller '%s' not recognized; trying %s\n",
				gadget->name,
				eth_config_driver.label);
		device_desc.bcdDevice =
			cpu_to_le16(0x0300 | 0x0099);
	}


	/* Allocate string descriptor numbers ... note that string
	 * contents can be overridden by the composite_dev glue.
	 */

	/* device descriptor strings: manufacturer, product */
	snprintf(manufacturer, sizeof manufacturer, "%s %s with %s",
		init_utsname()->sysname, init_utsname()->release,
		gadget->name);
	status = usb_string_id(cdev);
	if (status < 0)
		goto fail;
	strings_dev[STRING_MANUFACTURER_IDX].id = status;
	device_desc.iManufacturer = status;

	status = usb_string_id(cdev);
	if (status < 0)
		goto fail;
	strings_dev[STRING_PRODUCT_IDX].id = status;
	device_desc.iProduct = status;

	/* register our configuration(s); RNDIS first, if it's used */
	if (has_rndis()) {
		status = usb_add_config(cdev, &rndis_config_driver);
		if (status < 0)
			goto fail;
	}

	status = usb_add_config(cdev, &eth_config_driver);
	if (status < 0)
		goto fail;

	dev_info(&gadget->dev, "%s, version: " DRIVER_VERSION "\n",
			DRIVER_DESC);

	return 0;

fail:
	gether_cleanup();
	return status;
}

static int __exit eth_unbind(struct usb_composite_dev *cdev)
{
	gether_cleanup();
	return 0;
}

static struct usb_composite_driver eth_driver = {
	.name		= "g_ether",
	.dev		= &device_desc,
	.strings	= dev_strings,
	.bind		= eth_bind,
	.unbind		= __exit_p(eth_unbind),
};

MODULE_DESCRIPTION(PREFIX DRIVER_DESC);
MODULE_AUTHOR("David Brownell, Benedikt Spanger");
MODULE_LICENSE("GPL");

static int __init init(void)
{
	return usb_composite_register(&eth_driver);
}
module_init(init);

static void __exit cleanup(void)
{
	usb_composite_unregister(&eth_driver);
}
module_exit(cleanup);