net: Distributed Switch Architecture protocol support

Distributed Switch Architecture is a protocol for managing hardware
switch chips.  It consists of a set of MII management registers and
commands to configure the switch, and an ethernet header format to
signal which of the ports of the switch a packet was received from
or is intended to be sent to.

The switches that this driver supports are typically embedded in
access points and routers, and a typical setup with a DSA switch
looks something like this:

	+-----------+       +-----------+
	|           | RGMII |           |
	|           +-------+           +------ 1000baseT MDI ("WAN")
	|           |       |  6-port   +------ 1000baseT MDI ("LAN1")
	|    CPU    |       |  ethernet +------ 1000baseT MDI ("LAN2")
	|           |MIImgmt|  switch   +------ 1000baseT MDI ("LAN3")
	|           +-------+  w/5 PHYs +------ 1000baseT MDI ("LAN4")
	|           |       |           |
	+-----------+       +-----------+

The switch driver presents each port on the switch as a separate
network interface to Linux, polls the switch to maintain software
link state of those ports, forwards MII management interface
accesses to those network interfaces (e.g. as done by ethtool) to
the switch, and exposes the switch's hardware statistics counters
via the appropriate Linux kernel interfaces.

This initial patch supports the MII management interface register
layout of the Marvell 88E6123, 88E6161 and 88E6165 switch chips, and
supports the "Ethertype DSA" packet tagging format.

(There is no officially registered ethertype for the Ethertype DSA
packet format, so we just grab a random one.  The ethertype to use
is programmed into the switch, and the switch driver uses the value
of ETH_P_EDSA for this, so this define can be changed at any time in
the future if the one we chose is allocated to another protocol or
if Ethertype DSA gets its own officially registered ethertype, and
everything will continue to work.)

Signed-off-by: Lennert Buytenhek <buytenh@marvell.com>
Tested-by: Nicolas Pitre <nico@marvell.com>
Tested-by: Byron Bradley <byron.bbradley@gmail.com>
Tested-by: Tim Ellis <tim.ellis@mac.com>
Tested-by: Peter van Valderen <linux@ddcrew.com>
Tested-by: Dirk Teurlings <dirk@upexia.nl>
Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 377 insertions, 0 deletions

diff --git a/net/dsa/mv88e6xxx.c b/net/dsa/mv88e6xxx.c
new file mode 100644
index 00000000000..13d2328a240
--- /dev/null
+++ b/net/dsa/mv88e6xxx.c
@@ -0,0 +1,377 @@
+/*
+ * net/dsa/mv88e6xxx.c - Marvell 88e6xxx switch chip support
+ * Copyright (c) 2008 Marvell Semiconductor
+ *
+ * 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.
+ */
+
+#include <linux/list.h>
+#include <linux/netdevice.h>
+#include <linux/phy.h>
+#include "dsa_priv.h"
+#include "mv88e6xxx.h"
+
+/*
+ * If the switch's ADDR[4:0] strap pins are strapped to zero, it will
+ * use all 32 SMI bus addresses on its SMI bus, and all switch registers
+ * will be directly accessible on some {device address,register address}
+ * pair.  If the ADDR[4:0] pins are not strapped to zero, the switch
+ * will only respond to SMI transactions to that specific address, and
+ * an indirect addressing mechanism needs to be used to access its
+ * registers.
+ */
+static int mv88e6xxx_reg_wait_ready(struct mii_bus *bus, int sw_addr)
+{
+	int ret;
+	int i;
+
+	for (i = 0; i < 16; i++) {
+		ret = mdiobus_read(bus, sw_addr, 0);
+		if (ret < 0)
+			return ret;
+
+		if ((ret & 0x8000) == 0)
+			return 0;
+	}
+
+	return -ETIMEDOUT;
+}
+
+int __mv88e6xxx_reg_read(struct mii_bus *bus, int sw_addr, int addr, int reg)
+{
+	int ret;
+
+	if (sw_addr == 0)
+		return mdiobus_read(bus, addr, reg);
+
+	/*
+	 * Wait for the bus to become free.
+	 */
+	ret = mv88e6xxx_reg_wait_ready(bus, sw_addr);
+	if (ret < 0)
+		return ret;
+
+	/*
+	 * Transmit the read command.
+	 */
+	ret = mdiobus_write(bus, sw_addr, 0, 0x9800 | (addr << 5) | reg);
+	if (ret < 0)
+		return ret;
+
+	/*
+	 * Wait for the read command to complete.
+	 */
+	ret = mv88e6xxx_reg_wait_ready(bus, sw_addr);
+	if (ret < 0)
+		return ret;
+
+	/*
+	 * Read the data.
+	 */
+	ret = mdiobus_read(bus, sw_addr, 1);
+	if (ret < 0)
+		return ret;
+
+	return ret & 0xffff;
+}
+
+int mv88e6xxx_reg_read(struct dsa_switch *ds, int addr, int reg)
+{
+	struct mv88e6xxx_priv_state *ps = (void *)(ds + 1);
+	int ret;
+
+	mutex_lock(&ps->smi_mutex);
+	ret = __mv88e6xxx_reg_read(ds->master_mii_bus,
+				   ds->pd->sw_addr, addr, reg);
+	mutex_unlock(&ps->smi_mutex);
+
+	return ret;
+}
+
+int __mv88e6xxx_reg_write(struct mii_bus *bus, int sw_addr, int addr,
+			  int reg, u16 val)
+{
+	int ret;
+
+	if (sw_addr == 0)
+		return mdiobus_write(bus, addr, reg, val);
+
+	/*
+	 * Wait for the bus to become free.
+	 */
+	ret = mv88e6xxx_reg_wait_ready(bus, sw_addr);
+	if (ret < 0)
+		return ret;
+
+	/*
+	 * Transmit the data to write.
+	 */
+	ret = mdiobus_write(bus, sw_addr, 1, val);
+	if (ret < 0)
+		return ret;
+
+	/*
+	 * Transmit the write command.
+	 */
+	ret = mdiobus_write(bus, sw_addr, 0, 0x9400 | (addr << 5) | reg);
+	if (ret < 0)
+		return ret;
+
+	/*
+	 * Wait for the write command to complete.
+	 */
+	ret = mv88e6xxx_reg_wait_ready(bus, sw_addr);
+	if (ret < 0)
+		return ret;
+
+	return 0;
+}
+
+int mv88e6xxx_reg_write(struct dsa_switch *ds, int addr, int reg, u16 val)
+{
+	struct mv88e6xxx_priv_state *ps = (void *)(ds + 1);
+	int ret;
+
+	mutex_lock(&ps->smi_mutex);
+	ret = __mv88e6xxx_reg_write(ds->master_mii_bus,
+				    ds->pd->sw_addr, addr, reg, val);
+	mutex_unlock(&ps->smi_mutex);
+
+	return ret;
+}
+
+int mv88e6xxx_config_prio(struct dsa_switch *ds)
+{
+	/*
+	 * Configure the IP ToS mapping registers.
+	 */
+	REG_WRITE(REG_GLOBAL, 0x10, 0x0000);
+	REG_WRITE(REG_GLOBAL, 0x11, 0x0000);
+	REG_WRITE(REG_GLOBAL, 0x12, 0x5555);
+	REG_WRITE(REG_GLOBAL, 0x13, 0x5555);
+	REG_WRITE(REG_GLOBAL, 0x14, 0xaaaa);
+	REG_WRITE(REG_GLOBAL, 0x15, 0xaaaa);
+	REG_WRITE(REG_GLOBAL, 0x16, 0xffff);
+	REG_WRITE(REG_GLOBAL, 0x17, 0xffff);
+
+	/*
+	 * Configure the IEEE 802.1p priority mapping register.
+	 */
+	REG_WRITE(REG_GLOBAL, 0x18, 0xfa41);
+
+	return 0;
+}
+
+int mv88e6xxx_set_addr_indirect(struct dsa_switch *ds, u8 *addr)
+{
+	int i;
+	int ret;
+
+	for (i = 0; i < 6; i++) {
+		int j;
+
+		/*
+		 * Write the MAC address byte.
+		 */
+		REG_WRITE(REG_GLOBAL2, 0x0d, 0x8000 | (i << 8) | addr[i]);
+
+		/*
+		 * Wait for the write to complete.
+		 */
+		for (j = 0; j < 16; j++) {
+			ret = REG_READ(REG_GLOBAL2, 0x0d);
+			if ((ret & 0x8000) == 0)
+				break;
+		}
+		if (j == 16)
+			return -ETIMEDOUT;
+	}
+
+	return 0;
+}
+
+int mv88e6xxx_phy_read(struct dsa_switch *ds, int addr, int regnum)
+{
+	if (addr >= 0)
+		return mv88e6xxx_reg_read(ds, addr, regnum);
+	return 0xffff;
+}
+
+int mv88e6xxx_phy_write(struct dsa_switch *ds, int addr, int regnum, u16 val)
+{
+	if (addr >= 0)
+		return mv88e6xxx_reg_write(ds, addr, regnum, val);
+	return 0;
+}
+
+void mv88e6xxx_poll_link(struct dsa_switch *ds)
+{
+	int i;
+
+	for (i = 0; i < DSA_MAX_PORTS; i++) {
+		struct net_device *dev;
+		int port_status;
+		int link;
+		int speed;
+		int duplex;
+		int fc;
+
+		dev = ds->ports[i];
+		if (dev == NULL)
+			continue;
+
+		link = 0;
+		if (dev->flags & IFF_UP) {
+			port_status = mv88e6xxx_reg_read(ds, REG_PORT(i), 0x00);
+			if (port_status < 0)
+				continue;
+
+			link = !!(port_status & 0x0800);
+		}
+
+		if (!link) {
+			if (netif_carrier_ok(dev)) {
+				printk(KERN_INFO "%s: link down\n", dev->name);
+				netif_carrier_off(dev);
+			}
+			continue;
+		}
+
+		switch (port_status & 0x0300) {
+		case 0x0000:
+			speed = 10;
+			break;
+		case 0x0100:
+			speed = 100;
+			break;
+		case 0x0200:
+			speed = 1000;
+			break;
+		default:
+			speed = -1;
+			break;
+		}
+		duplex = (port_status & 0x0400) ? 1 : 0;
+		fc = (port_status & 0x8000) ? 1 : 0;
+
+		if (!netif_carrier_ok(dev)) {
+			printk(KERN_INFO "%s: link up, %d Mb/s, %s duplex, "
+					 "flow control %sabled\n", dev->name,
+					 speed, duplex ? "full" : "half",
+					 fc ? "en" : "dis");
+			netif_carrier_on(dev);
+		}
+	}
+}
+
+static int mv88e6xxx_stats_wait(struct dsa_switch *ds)
+{
+	int ret;
+	int i;
+
+	for (i = 0; i < 10; i++) {
+		ret = REG_READ(REG_GLOBAL2, 0x1d);
+		if ((ret & 0x8000) == 0)
+			return 0;
+	}
+
+	return -ETIMEDOUT;
+}
+
+static int mv88e6xxx_stats_snapshot(struct dsa_switch *ds, int port)
+{
+	int ret;
+
+	/*
+	 * Snapshot the hardware statistics counters for this port.
+	 */
+	REG_WRITE(REG_GLOBAL, 0x1d, 0xdc00 | port);
+
+	/*
+	 * Wait for the snapshotting to complete.
+	 */
+	ret = mv88e6xxx_stats_wait(ds);
+	if (ret < 0)
+		return ret;
+
+	return 0;
+}
+
+static void mv88e6xxx_stats_read(struct dsa_switch *ds, int stat, u32 *val)
+{
+	u32 _val;
+	int ret;
+
+	*val = 0;
+
+	ret = mv88e6xxx_reg_write(ds, REG_GLOBAL, 0x1d, 0xcc00 | stat);
+	if (ret < 0)
+		return;
+
+	ret = mv88e6xxx_stats_wait(ds);
+	if (ret < 0)
+		return;
+
+	ret = mv88e6xxx_reg_read(ds, REG_GLOBAL, 0x1e);
+	if (ret < 0)
+		return;
+
+	_val = ret << 16;
+
+	ret = mv88e6xxx_reg_read(ds, REG_GLOBAL, 0x1f);
+	if (ret < 0)
+		return;
+
+	*val = _val | ret;
+}
+
+void mv88e6xxx_get_strings(struct dsa_switch *ds,
+			   int nr_stats, struct mv88e6xxx_hw_stat *stats,
+			   int port, uint8_t *data)
+{
+	int i;
+
+	for (i = 0; i < nr_stats; i++) {
+		memcpy(data + i * ETH_GSTRING_LEN,
+		       stats[i].string, ETH_GSTRING_LEN);
+	}
+}
+
+void mv88e6xxx_get_ethtool_stats(struct dsa_switch *ds,
+				 int nr_stats, struct mv88e6xxx_hw_stat *stats,
+				 int port, uint64_t *data)
+{
+	struct mv88e6xxx_priv_state *ps = (void *)(ds + 1);
+	int ret;
+	int i;
+
+	mutex_lock(&ps->stats_mutex);
+
+	ret = mv88e6xxx_stats_snapshot(ds, port);
+	if (ret < 0) {
+		mutex_unlock(&ps->stats_mutex);
+		return;
+	}
+
+	/*
+	 * Read each of the counters.
+	 */
+	for (i = 0; i < nr_stats; i++) {
+		struct mv88e6xxx_hw_stat *s = stats + i;
+		u32 low;
+		u32 high;
+
+		mv88e6xxx_stats_read(ds, s->reg, &low);
+		if (s->sizeof_stat == 8)
+			mv88e6xxx_stats_read(ds, s->reg + 1, &high);
+		else
+			high = 0;
+
+		data[i] = (((u64)high) << 32) | low;
+	}
+
+	mutex_unlock(&ps->stats_mutex);
+}