u-boot/drivers/net/pfe_eth/pfe_mdio.c

292 lines
6.9 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0+
/*
* Copyright 2015-2016 Freescale Semiconductor, Inc.
* Copyright 2017 NXP
*/
#include <common.h>
#include <dm.h>
#include <log.h>
#include <malloc.h>
#include <dm/platform_data/pfe_dm_eth.h>
#include <net.h>
#include <linux/delay.h>
#include <net/pfe_eth/pfe_eth.h>
extern struct gemac_s gem_info[];
#if defined(CONFIG_PHYLIB)
#define MDIO_TIMEOUT 5000
static int pfe_write_addr(struct mii_dev *bus, int phy_addr, int dev_addr,
int reg_addr)
{
void *reg_base = bus->priv;
u32 devadr;
u32 phy;
u32 reg_data;
int timeout = MDIO_TIMEOUT;
devadr = ((dev_addr & EMAC_MII_DATA_RA_MASK) << EMAC_MII_DATA_RA_SHIFT);
phy = ((phy_addr & EMAC_MII_DATA_PA_MASK) << EMAC_MII_DATA_PA_SHIFT);
reg_data = (EMAC_MII_DATA_TA | phy | devadr | reg_addr);
writel(reg_data, reg_base + EMAC_MII_DATA_REG);
/*
* wait for the MII interrupt
*/
while (!(readl(reg_base + EMAC_IEVENT_REG) & EMAC_IEVENT_MII)) {
if (timeout-- <= 0) {
printf("Phy MDIO read/write timeout\n");
return -1;
}
}
/*
* clear MII interrupt
*/
writel(EMAC_IEVENT_MII, reg_base + EMAC_IEVENT_REG);
return 0;
}
static int pfe_phy_read(struct mii_dev *bus, int phy_addr, int dev_addr,
int reg_addr)
{
void *reg_base = bus->priv;
u32 reg;
u32 phy;
u32 reg_data;
u16 val;
int timeout = MDIO_TIMEOUT;
if (dev_addr == MDIO_DEVAD_NONE) {
reg = ((reg_addr & EMAC_MII_DATA_RA_MASK) <<
EMAC_MII_DATA_RA_SHIFT);
} else {
pfe_write_addr(bus, phy_addr, dev_addr, reg_addr);
reg = ((dev_addr & EMAC_MII_DATA_RA_MASK) <<
EMAC_MII_DATA_RA_SHIFT);
}
phy = ((phy_addr & EMAC_MII_DATA_PA_MASK) << EMAC_MII_DATA_PA_SHIFT);
if (dev_addr == MDIO_DEVAD_NONE)
reg_data = (EMAC_MII_DATA_ST | EMAC_MII_DATA_OP_RD |
EMAC_MII_DATA_TA | phy | reg);
else
reg_data = (EMAC_MII_DATA_OP_CL45_RD | EMAC_MII_DATA_TA |
phy | reg);
writel(reg_data, reg_base + EMAC_MII_DATA_REG);
/*
* wait for the MII interrupt
*/
while (!(readl(reg_base + EMAC_IEVENT_REG) & EMAC_IEVENT_MII)) {
if (timeout-- <= 0) {
printf("Phy MDIO read/write timeout\n");
return -1;
}
}
/*
* clear MII interrupt
*/
writel(EMAC_IEVENT_MII, reg_base + EMAC_IEVENT_REG);
/*
* it's now safe to read the PHY's register
*/
val = (u16)readl(reg_base + EMAC_MII_DATA_REG);
debug("%s: %p phy: 0x%x reg:0x%08x val:%#x\n", __func__, reg_base,
phy_addr, reg_addr, val);
return val;
}
static int pfe_phy_write(struct mii_dev *bus, int phy_addr, int dev_addr,
int reg_addr, u16 data)
{
void *reg_base = bus->priv;
u32 reg;
u32 phy;
u32 reg_data;
int timeout = MDIO_TIMEOUT;
if (dev_addr == MDIO_DEVAD_NONE) {
reg = ((reg_addr & EMAC_MII_DATA_RA_MASK) <<
EMAC_MII_DATA_RA_SHIFT);
} else {
pfe_write_addr(bus, phy_addr, dev_addr, reg_addr);
reg = ((dev_addr & EMAC_MII_DATA_RA_MASK) <<
EMAC_MII_DATA_RA_SHIFT);
}
phy = ((phy_addr & EMAC_MII_DATA_PA_MASK) << EMAC_MII_DATA_PA_SHIFT);
if (dev_addr == MDIO_DEVAD_NONE)
reg_data = (EMAC_MII_DATA_ST | EMAC_MII_DATA_OP_WR |
EMAC_MII_DATA_TA | phy | reg | data);
else
reg_data = (EMAC_MII_DATA_OP_CL45_WR | EMAC_MII_DATA_TA |
phy | reg | data);
writel(reg_data, reg_base + EMAC_MII_DATA_REG);
/*
* wait for the MII interrupt
*/
while (!(readl(reg_base + EMAC_IEVENT_REG) & EMAC_IEVENT_MII)) {
if (timeout-- <= 0) {
printf("Phy MDIO read/write timeout\n");
return -1;
}
}
/*
* clear MII interrupt
*/
writel(EMAC_IEVENT_MII, reg_base + EMAC_IEVENT_REG);
debug("%s: phy: %02x reg:%02x val:%#x\n", __func__, phy_addr,
reg_addr, data);
return 0;
}
static void pfe_configure_serdes(struct pfe_eth_dev *priv)
{
struct mii_dev bus;
int value, sgmii_2500 = 0;
struct gemac_s *gem = priv->gem;
net: freescale: replace usage of phy-mode = "sgmii-2500" with "2500base-x" After the discussion here: https://lore.kernel.org/netdev/20210603143453.if7hgifupx5k433b@pali/ which resulted in this patch: https://patchwork.kernel.org/project/netdevbpf/patch/20210704134325.24842-1-pali@kernel.org/ and many other discussions before it, notably: https://patchwork.kernel.org/project/linux-arm-kernel/patch/1512016235-15909-1-git-send-email-Bhaskar.Upadhaya@nxp.com/ it became apparent that nobody really knows what "SGMII 2500" is. Certainly, Freescale/NXP hardware engineers name this protocol "SGMII 2500" in the reference manuals, but the PCS devices do not support any "SGMII" specific features when operating at the speed of 2500 Mbps, no in-band autoneg and no speed change via symbol replication . So that leaves a fixed speed of 2500 Mbps using a coding of 8b/10b with a SERDES lane frequency of 3.125 GHz. In fact, "SGMII 2500 without in-band autoneg and at a fixed speed" is indistinguishable from "2500base-x without in-band autoneg", which is precisely what these NXP devices support. So it just appears that "SGMII 2500" is an unclear name with no clear definition that stuck. As such, in the Linux kernel, the drivers which use this SERDES protocol use the 2500base-x phy-mode. This patch converts U-Boot to use 2500base-x too, or at least, as much as it can. Note that I would have really liked to delete PHY_INTERFACE_MODE_SGMII_2500 completely, but the mvpp2 driver seems to even distinguish between SGMII 2500 and 2500base-X. Namely, it enables in-band autoneg for one but not the other, and forces flow control for one but not the other. This goes back to the idea that maybe 2500base-X is a fiber protocol and SGMII-2500 is an MII protocol (connects a MAC to a PHY such as Aquantia), but the two are practically indistinguishable through everything except use case. NXP devices can support both use cases through an identical configuration, for example RX flow control can be unconditionally enabled in order to support rate adaptation performed by an Aquantia PHY. At least I can find no indication in online documents published by Cisco which would point towards "SGMII-2500" being an actual standard with an actual definition, so I cannot say "yes, NXP devices support it". Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
2021-09-18 12:32:35 +00:00
if (gem->phy_mode == PHY_INTERFACE_MODE_2500BASEX)
sgmii_2500 = 1;
/* PCS configuration done with corresponding GEMAC */
bus.priv = gem_info[priv->gemac_port].gemac_base;
pfe_phy_read(&bus, 0, MDIO_DEVAD_NONE, 0x0);
pfe_phy_read(&bus, 0, MDIO_DEVAD_NONE, 0x1);
pfe_phy_read(&bus, 0, MDIO_DEVAD_NONE, 0x2);
pfe_phy_read(&bus, 0, MDIO_DEVAD_NONE, 0x3);
/* Reset serdes */
pfe_phy_write(&bus, 0, MDIO_DEVAD_NONE, 0x0, 0x8000);
/* SGMII IF mode + AN enable only for 1G SGMII, not for 2.5G */
value = PHY_SGMII_IF_MODE_SGMII;
if (!sgmii_2500)
value |= PHY_SGMII_IF_MODE_AN;
else
value |= PHY_SGMII_IF_MODE_SGMII_GBT;
pfe_phy_write(&bus, 0, MDIO_DEVAD_NONE, 0x14, value);
/* Dev ability according to SGMII specification */
value = PHY_SGMII_DEV_ABILITY_SGMII;
pfe_phy_write(&bus, 0, MDIO_DEVAD_NONE, 0x4, value);
/* These values taken from validation team */
if (!sgmii_2500) {
pfe_phy_write(&bus, 0, MDIO_DEVAD_NONE, 0x13, 0x0);
pfe_phy_write(&bus, 0, MDIO_DEVAD_NONE, 0x12, 0x400);
} else {
pfe_phy_write(&bus, 0, MDIO_DEVAD_NONE, 0x13, 0x7);
pfe_phy_write(&bus, 0, MDIO_DEVAD_NONE, 0x12, 0xa120);
}
/* Restart AN */
value = PHY_SGMII_CR_DEF_VAL;
if (!sgmii_2500)
value |= PHY_SGMII_CR_RESET_AN;
/* Disable Auto neg for 2.5G SGMII as it doesn't support auto neg*/
if (sgmii_2500)
value &= ~PHY_SGMII_ENABLE_AN;
pfe_phy_write(&bus, 0, MDIO_DEVAD_NONE, 0, value);
}
int pfe_phy_configure(struct pfe_eth_dev *priv, int dev_id, int phy_id)
{
struct phy_device *phydev = NULL;
struct udevice *dev = priv->dev;
struct gemac_s *gem = priv->gem;
struct ccsr_scfg *scfg = (struct ccsr_scfg *)CONFIG_SYS_FSL_SCFG_ADDR;
if (!gem->bus)
return -1;
/* Configure SGMII PCS */
if (gem->phy_mode == PHY_INTERFACE_MODE_SGMII ||
net: freescale: replace usage of phy-mode = "sgmii-2500" with "2500base-x" After the discussion here: https://lore.kernel.org/netdev/20210603143453.if7hgifupx5k433b@pali/ which resulted in this patch: https://patchwork.kernel.org/project/netdevbpf/patch/20210704134325.24842-1-pali@kernel.org/ and many other discussions before it, notably: https://patchwork.kernel.org/project/linux-arm-kernel/patch/1512016235-15909-1-git-send-email-Bhaskar.Upadhaya@nxp.com/ it became apparent that nobody really knows what "SGMII 2500" is. Certainly, Freescale/NXP hardware engineers name this protocol "SGMII 2500" in the reference manuals, but the PCS devices do not support any "SGMII" specific features when operating at the speed of 2500 Mbps, no in-band autoneg and no speed change via symbol replication . So that leaves a fixed speed of 2500 Mbps using a coding of 8b/10b with a SERDES lane frequency of 3.125 GHz. In fact, "SGMII 2500 without in-band autoneg and at a fixed speed" is indistinguishable from "2500base-x without in-band autoneg", which is precisely what these NXP devices support. So it just appears that "SGMII 2500" is an unclear name with no clear definition that stuck. As such, in the Linux kernel, the drivers which use this SERDES protocol use the 2500base-x phy-mode. This patch converts U-Boot to use 2500base-x too, or at least, as much as it can. Note that I would have really liked to delete PHY_INTERFACE_MODE_SGMII_2500 completely, but the mvpp2 driver seems to even distinguish between SGMII 2500 and 2500base-X. Namely, it enables in-band autoneg for one but not the other, and forces flow control for one but not the other. This goes back to the idea that maybe 2500base-X is a fiber protocol and SGMII-2500 is an MII protocol (connects a MAC to a PHY such as Aquantia), but the two are practically indistinguishable through everything except use case. NXP devices can support both use cases through an identical configuration, for example RX flow control can be unconditionally enabled in order to support rate adaptation performed by an Aquantia PHY. At least I can find no indication in online documents published by Cisco which would point towards "SGMII-2500" being an actual standard with an actual definition, so I cannot say "yes, NXP devices support it". Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by: Ramon Fried <rfried.dev@gmail.com>
2021-09-18 12:32:35 +00:00
gem->phy_mode == PHY_INTERFACE_MODE_2500BASEX) {
out_be32(&scfg->mdioselcr, 0x00000000);
pfe_configure_serdes(priv);
}
mdelay(100);
/* By this time on-chip SGMII initialization is done
* we can switch mdio interface to external PHYs
*/
out_be32(&scfg->mdioselcr, 0x80000000);
phydev = phy_connect(gem->bus, phy_id, dev, gem->phy_mode);
if (!phydev) {
printf("phy_connect failed\n");
return -ENODEV;
}
phy_config(phydev);
priv->phydev = phydev;
return 0;
}
#endif
struct mii_dev *pfe_mdio_init(struct pfe_mdio_info *mdio_info)
{
struct mii_dev *bus;
int ret;
u32 mdio_speed;
u32 pclk = 250000000;
bus = mdio_alloc();
if (!bus) {
printf("mdio_alloc failed\n");
return NULL;
}
bus->read = pfe_phy_read;
bus->write = pfe_phy_write;
/* MAC1 MDIO used to communicate with external PHYS */
bus->priv = mdio_info->reg_base;
sprintf(bus->name, mdio_info->name);
/* configure mdio speed */
mdio_speed = (DIV_ROUND_UP(pclk, 4000000) << EMAC_MII_SPEED_SHIFT);
mdio_speed |= EMAC_HOLDTIME(0x5);
writel(mdio_speed, mdio_info->reg_base + EMAC_MII_CTRL_REG);
ret = mdio_register(bus);
if (ret) {
printf("mdio_register failed\n");
free(bus);
return NULL;
}
return bus;
}
void pfe_set_mdio(int dev_id, struct mii_dev *bus)
{
gem_info[dev_id].bus = bus;
}
void pfe_set_phy_address_mode(int dev_id, int phy_id, int phy_mode)
{
gem_info[dev_id].phy_address = phy_id;
gem_info[dev_id].phy_mode = phy_mode;
}