u-boot/board/CZ.NIC/turris_omnia/turris_omnia.c

1078 lines
27 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0+
/*
* Copyright (C) 2017 Marek Behún <kabel@kernel.org>
* Copyright (C) 2016 Tomas Hlavacek <tomas.hlavacek@nic.cz>
*
* Derived from the code for
* Marvell/db-88f6820-gp by Stefan Roese <sr@denx.de>
*/
#include <common.h>
#include <env.h>
#include <i2c.h>
#include <init.h>
#include <log.h>
#include <miiphy.h>
#include <mtd.h>
#include <asm/global_data.h>
#include <asm/io.h>
#include <asm/arch/cpu.h>
#include <asm/arch/soc.h>
#include <dm/uclass.h>
arm: mvebu: turris_omnia: Add support for design with SW reset signals New Turris Omnia HW board revision requires that software controls peripheral reset signals, namely PERST# signals on mPCIe slots, ethernet phy reset and lan switch reset. Those pins are connected to MCU controlled by MCU i2c API as GPIOs. On new HW board revision those pins stay in reset after board reset and software has to release these peripherals from reset manually. MCU announce this requirement by FEAT_PERIPH_MCU bit in CMD_GET_FEATURES command. On older HW board revisions when FEAT_PERIPH_MCU is not announced, all those reset signals are automatically released after board finish reset. Detect FEAT_PERIPH_MCU bit in board_fix_fdt() and ft_board_setup() functions and insert into device tree blob pcie "reset-gpios" and eth phy "phy-reset-gpios" properties with corresponding MCU gpio definitions. PCIe and eth PHY drivers then automatically release resets during device initialization. Both U-Boot and Linux kernel drivers support those device tree reset properties. Initialization of lan switch on new HW board revision is more complicated. Switch strapping pins are shared with switch RGMII pins. And strapping pins must be in specific configuration after releasing switch reset. Due to pin sharing, it is first required to switch A385 side of switch pins into GPIO mode, set strapping configuration, release switch from reset and after that switch A385 pins back to RGMII mode. Because this complicated setup is not supported by switch DSA drivers and cannot be expressed easily in device tree, implement it manually in SPL function spl_board_init(). So in proper U-Boot and OS/kernel would be lan switch initialized and be in same configuration like it was on old HW board revisions (where reset sequence did those steps at hardware level). Signed-off-by: Pali Rohár <pali@kernel.org> Reviewed-by: Stefan Roese <sr@denx.de> Reviewed-by: Marek Behún <kabel@kernel.org>
2022-07-29 11:29:07 +00:00
#include <dt-bindings/gpio/gpio.h>
#include <fdt_support.h>
#include <time.h>
#include <linux/bitops.h>
arm: mvebu: turris_omnia: Add support for design with SW reset signals New Turris Omnia HW board revision requires that software controls peripheral reset signals, namely PERST# signals on mPCIe slots, ethernet phy reset and lan switch reset. Those pins are connected to MCU controlled by MCU i2c API as GPIOs. On new HW board revision those pins stay in reset after board reset and software has to release these peripherals from reset manually. MCU announce this requirement by FEAT_PERIPH_MCU bit in CMD_GET_FEATURES command. On older HW board revisions when FEAT_PERIPH_MCU is not announced, all those reset signals are automatically released after board finish reset. Detect FEAT_PERIPH_MCU bit in board_fix_fdt() and ft_board_setup() functions and insert into device tree blob pcie "reset-gpios" and eth phy "phy-reset-gpios" properties with corresponding MCU gpio definitions. PCIe and eth PHY drivers then automatically release resets during device initialization. Both U-Boot and Linux kernel drivers support those device tree reset properties. Initialization of lan switch on new HW board revision is more complicated. Switch strapping pins are shared with switch RGMII pins. And strapping pins must be in specific configuration after releasing switch reset. Due to pin sharing, it is first required to switch A385 side of switch pins into GPIO mode, set strapping configuration, release switch from reset and after that switch A385 pins back to RGMII mode. Because this complicated setup is not supported by switch DSA drivers and cannot be expressed easily in device tree, implement it manually in SPL function spl_board_init(). So in proper U-Boot and OS/kernel would be lan switch initialized and be in same configuration like it was on old HW board revisions (where reset sequence did those steps at hardware level). Signed-off-by: Pali Rohár <pali@kernel.org> Reviewed-by: Stefan Roese <sr@denx.de> Reviewed-by: Marek Behún <kabel@kernel.org>
2022-07-29 11:29:07 +00:00
#include <linux/delay.h>
#include <u-boot/crc.h>
#include "../drivers/ddr/marvell/a38x/ddr3_init.h"
#include <../serdes/a38x/high_speed_env_spec.h>
#include "../turris_atsha_otp.h"
DECLARE_GLOBAL_DATA_PTR;
#define OMNIA_SPI_NOR_PATH "/soc/spi@10600/spi-nor@0"
#define OMNIA_I2C_BUS_NAME "i2c@11000->i2cmux@70->i2c@0"
#define OMNIA_I2C_MCU_CHIP_ADDR 0x2a
#define OMNIA_I2C_MCU_CHIP_LEN 1
#define OMNIA_I2C_EEPROM_CHIP_ADDR 0x54
#define OMNIA_I2C_EEPROM_CHIP_LEN 2
#define OMNIA_I2C_EEPROM_MAGIC 0x0341a034
#define A385_SYS_RSTOUT_MASK MVEBU_REGISTER(0x18260)
#define A385_SYS_RSTOUT_MASK_WD BIT(10)
arm: mvebu: turris_omnia: enable A385 watchdog before disabling MCU watchdog Commit aeb0ca64dbb5 ("arm: mvebu: turris_omnia: disable MCU watchdog in SPL when booting over UART") disabled MCU watchdog when booting over UART to ensure that watchdog does not reboot the board before UART transfer finishes. But if UART transfer fails for some reason, or if U-Boot binary crashes, then board hangs forever as there is no watchdog running which could reset it. To fix this issue, enable A385 watchdog with very high timeout before disabling MCU watchdog to ensure that even slow transfer can finish successfully before watchdog timer expires and also to ensure that if board hangs for some reason, watchdog will reset it. Omnia's MCU watchdog has fixed 120 seconds timer and it cannot be changed (without updating MCU firmware). A385 watchdog by default uses 25 MHz input clock and so the largest timeout value (2^32-1) can be just 171 seconds. But A385 watchdog can be switched to use NBCLK (L2) as input clock (on Turris Omnia it is 800 MHz clock) and in this case final watchdog clock frequency is calculated as: freq = NBCLK / 2 / (2 ^ R) So A385 watchdog on Turris Omnia can be configured to at most 1374 seconds (about 22 minutes). We set it to 10 minutes, which should be enough even for bigger U-Boot binaries or slower UART transfers. Both U-Boot and Linux kernel, when initializing A385 watchdog, switch watchdog timer to 25 MHz input clock, so usage of NBCLK input clock in U-Boot SPL does not cause any issues. Fixes: aeb0ca64dbb5 ("arm: mvebu: turris_omnia: disable MCU watchdog in SPL when booting over UART") Signed-off-by: Pali Rohár <pali@kernel.org> Signed-off-by: Marek Behún <marek.behun@nic.cz>
2021-11-09 16:14:02 +00:00
#define A385_WDT_GLOBAL_CTRL MVEBU_REGISTER(0x20300)
#define A385_WDT_GLOBAL_RATIO_MASK GENMASK(18, 16)
#define A385_WDT_GLOBAL_RATIO_SHIFT 16
#define A385_WDT_GLOBAL_25MHZ BIT(10)
#define A385_WDT_GLOBAL_ENABLE BIT(8)
#define A385_WDT_GLOBAL_STATUS MVEBU_REGISTER(0x20304)
#define A385_WDT_GLOBAL_EXPIRED BIT(31)
#define A385_WDT_DURATION MVEBU_REGISTER(0x20334)
#define A385_WD_RSTOUT_UNMASK MVEBU_REGISTER(0x20704)
#define A385_WD_RSTOUT_UNMASK_GLOBAL BIT(8)
enum mcu_commands {
CMD_GET_STATUS_WORD = 0x01,
CMD_GET_RESET = 0x09,
CMD_WATCHDOG_STATE = 0x0b,
/* available if STS_FEATURES_SUPPORTED bit set in status word */
CMD_GET_FEATURES = 0x10,
arm: mvebu: turris_omnia: Add support for design with SW reset signals New Turris Omnia HW board revision requires that software controls peripheral reset signals, namely PERST# signals on mPCIe slots, ethernet phy reset and lan switch reset. Those pins are connected to MCU controlled by MCU i2c API as GPIOs. On new HW board revision those pins stay in reset after board reset and software has to release these peripherals from reset manually. MCU announce this requirement by FEAT_PERIPH_MCU bit in CMD_GET_FEATURES command. On older HW board revisions when FEAT_PERIPH_MCU is not announced, all those reset signals are automatically released after board finish reset. Detect FEAT_PERIPH_MCU bit in board_fix_fdt() and ft_board_setup() functions and insert into device tree blob pcie "reset-gpios" and eth phy "phy-reset-gpios" properties with corresponding MCU gpio definitions. PCIe and eth PHY drivers then automatically release resets during device initialization. Both U-Boot and Linux kernel drivers support those device tree reset properties. Initialization of lan switch on new HW board revision is more complicated. Switch strapping pins are shared with switch RGMII pins. And strapping pins must be in specific configuration after releasing switch reset. Due to pin sharing, it is first required to switch A385 side of switch pins into GPIO mode, set strapping configuration, release switch from reset and after that switch A385 pins back to RGMII mode. Because this complicated setup is not supported by switch DSA drivers and cannot be expressed easily in device tree, implement it manually in SPL function spl_board_init(). So in proper U-Boot and OS/kernel would be lan switch initialized and be in same configuration like it was on old HW board revisions (where reset sequence did those steps at hardware level). Signed-off-by: Pali Rohár <pali@kernel.org> Reviewed-by: Stefan Roese <sr@denx.de> Reviewed-by: Marek Behún <kabel@kernel.org>
2022-07-29 11:29:07 +00:00
/* available if EXT_CMD bit set in features */
CMD_EXT_CONTROL = 0x12,
};
enum status_word_bits {
STS_MCU_TYPE_MASK = GENMASK(1, 0),
STS_MCU_TYPE_STM32 = 0,
STS_MCU_TYPE_GD32 = 1,
STS_MCU_TYPE_MKL = 2,
STS_MCU_TYPE_UNKN = 3,
STS_FEATURES_SUPPORTED = BIT(2),
CARD_DET_STSBIT = 0x0010,
MSATA_IND_STSBIT = 0x0020,
};
/* CMD_GET_FEATURES */
enum features_e {
FEAT_PERIPH_MCU = BIT(0),
arm: mvebu: turris_omnia: Add support for design with SW reset signals New Turris Omnia HW board revision requires that software controls peripheral reset signals, namely PERST# signals on mPCIe slots, ethernet phy reset and lan switch reset. Those pins are connected to MCU controlled by MCU i2c API as GPIOs. On new HW board revision those pins stay in reset after board reset and software has to release these peripherals from reset manually. MCU announce this requirement by FEAT_PERIPH_MCU bit in CMD_GET_FEATURES command. On older HW board revisions when FEAT_PERIPH_MCU is not announced, all those reset signals are automatically released after board finish reset. Detect FEAT_PERIPH_MCU bit in board_fix_fdt() and ft_board_setup() functions and insert into device tree blob pcie "reset-gpios" and eth phy "phy-reset-gpios" properties with corresponding MCU gpio definitions. PCIe and eth PHY drivers then automatically release resets during device initialization. Both U-Boot and Linux kernel drivers support those device tree reset properties. Initialization of lan switch on new HW board revision is more complicated. Switch strapping pins are shared with switch RGMII pins. And strapping pins must be in specific configuration after releasing switch reset. Due to pin sharing, it is first required to switch A385 side of switch pins into GPIO mode, set strapping configuration, release switch from reset and after that switch A385 pins back to RGMII mode. Because this complicated setup is not supported by switch DSA drivers and cannot be expressed easily in device tree, implement it manually in SPL function spl_board_init(). So in proper U-Boot and OS/kernel would be lan switch initialized and be in same configuration like it was on old HW board revisions (where reset sequence did those steps at hardware level). Signed-off-by: Pali Rohár <pali@kernel.org> Reviewed-by: Stefan Roese <sr@denx.de> Reviewed-by: Marek Behún <kabel@kernel.org>
2022-07-29 11:29:07 +00:00
FEAT_EXT_CMDS = BIT(1),
};
/* CMD_EXT_CONTROL */
enum ext_ctl_e {
EXT_CTL_nRES_LAN = BIT(1),
EXT_CTL_nRES_PHY = BIT(2),
EXT_CTL_nPERST0 = BIT(3),
EXT_CTL_nPERST1 = BIT(4),
EXT_CTL_nPERST2 = BIT(5),
};
/*
* Those values and defines are taken from the Marvell U-Boot version
* "u-boot-2013.01-2014_T3.0"
*/
#define OMNIA_GPP_OUT_ENA_LOW \
(~(BIT(1) | BIT(4) | BIT(6) | BIT(7) | BIT(8) | BIT(9) | \
BIT(10) | BIT(11) | BIT(19) | BIT(22) | BIT(23) | BIT(25) | \
BIT(26) | BIT(27) | BIT(29) | BIT(30) | BIT(31)))
#define OMNIA_GPP_OUT_ENA_MID \
(~(BIT(0) | BIT(1) | BIT(2) | BIT(3) | BIT(4) | BIT(15) | \
BIT(16) | BIT(17) | BIT(18)))
#define OMNIA_GPP_OUT_VAL_LOW 0x0
#define OMNIA_GPP_OUT_VAL_MID 0x0
#define OMNIA_GPP_POL_LOW 0x0
#define OMNIA_GPP_POL_MID 0x0
static struct serdes_map board_serdes_map[] = {
{PEX0, SERDES_SPEED_5_GBPS, PEX_ROOT_COMPLEX_X1, 0, 0},
{USB3_HOST0, SERDES_SPEED_5_GBPS, SERDES_DEFAULT_MODE, 0, 0},
{PEX1, SERDES_SPEED_5_GBPS, PEX_ROOT_COMPLEX_X1, 0, 0},
{USB3_HOST1, SERDES_SPEED_5_GBPS, SERDES_DEFAULT_MODE, 0, 0},
{PEX2, SERDES_SPEED_5_GBPS, PEX_ROOT_COMPLEX_X1, 0, 0},
{SGMII2, SERDES_SPEED_1_25_GBPS, SERDES_DEFAULT_MODE, 0, 0}
};
static struct udevice *omnia_get_i2c_chip(const char *name, uint addr,
uint offset_len)
{
struct udevice *bus, *dev;
int ret;
ret = uclass_get_device_by_name(UCLASS_I2C, OMNIA_I2C_BUS_NAME, &bus);
if (ret) {
printf("Cannot get I2C bus %s: uclass_get_device_by_name failed: %i\n",
OMNIA_I2C_BUS_NAME, ret);
return NULL;
}
ret = i2c_get_chip(bus, addr, offset_len, &dev);
if (ret) {
printf("Cannot get %s I2C chip: i2c_get_chip failed: %i\n",
name, ret);
return NULL;
}
return dev;
}
static int omnia_mcu_read(u8 cmd, void *buf, int len)
{
struct udevice *chip;
chip = omnia_get_i2c_chip("MCU", OMNIA_I2C_MCU_CHIP_ADDR,
OMNIA_I2C_MCU_CHIP_LEN);
if (!chip)
return -ENODEV;
return dm_i2c_read(chip, cmd, buf, len);
}
static int omnia_mcu_write(u8 cmd, const void *buf, int len)
{
struct udevice *chip;
chip = omnia_get_i2c_chip("MCU", OMNIA_I2C_MCU_CHIP_ADDR,
OMNIA_I2C_MCU_CHIP_LEN);
if (!chip)
return -ENODEV;
return dm_i2c_write(chip, cmd, buf, len);
}
arm: mvebu: turris_omnia: enable A385 watchdog before disabling MCU watchdog Commit aeb0ca64dbb5 ("arm: mvebu: turris_omnia: disable MCU watchdog in SPL when booting over UART") disabled MCU watchdog when booting over UART to ensure that watchdog does not reboot the board before UART transfer finishes. But if UART transfer fails for some reason, or if U-Boot binary crashes, then board hangs forever as there is no watchdog running which could reset it. To fix this issue, enable A385 watchdog with very high timeout before disabling MCU watchdog to ensure that even slow transfer can finish successfully before watchdog timer expires and also to ensure that if board hangs for some reason, watchdog will reset it. Omnia's MCU watchdog has fixed 120 seconds timer and it cannot be changed (without updating MCU firmware). A385 watchdog by default uses 25 MHz input clock and so the largest timeout value (2^32-1) can be just 171 seconds. But A385 watchdog can be switched to use NBCLK (L2) as input clock (on Turris Omnia it is 800 MHz clock) and in this case final watchdog clock frequency is calculated as: freq = NBCLK / 2 / (2 ^ R) So A385 watchdog on Turris Omnia can be configured to at most 1374 seconds (about 22 minutes). We set it to 10 minutes, which should be enough even for bigger U-Boot binaries or slower UART transfers. Both U-Boot and Linux kernel, when initializing A385 watchdog, switch watchdog timer to 25 MHz input clock, so usage of NBCLK input clock in U-Boot SPL does not cause any issues. Fixes: aeb0ca64dbb5 ("arm: mvebu: turris_omnia: disable MCU watchdog in SPL when booting over UART") Signed-off-by: Pali Rohár <pali@kernel.org> Signed-off-by: Marek Behún <marek.behun@nic.cz>
2021-11-09 16:14:02 +00:00
static void enable_a385_watchdog(unsigned int timeout_minutes)
{
struct sar_freq_modes sar_freq;
u32 watchdog_freq;
printf("Enabling A385 watchdog with %u minutes timeout...\n",
timeout_minutes);
/*
* Use NBCLK clock (a.k.a. L2 clock) as watchdog input clock with
* its maximal ratio 7 instead of default fixed 25 MHz clock.
* It allows to set watchdog duration up to the 22 minutes.
*/
clrsetbits_32(A385_WDT_GLOBAL_CTRL,
A385_WDT_GLOBAL_25MHZ | A385_WDT_GLOBAL_RATIO_MASK,
7 << A385_WDT_GLOBAL_RATIO_SHIFT);
/*
* Calculate watchdog clock frequency. It is defined by formula:
* freq = NBCLK / 2 / (2 ^ ratio)
* We set ratio to the maximal possible value 7.
*/
get_sar_freq(&sar_freq);
watchdog_freq = sar_freq.nb_clk * 1000000 / 2 / (1 << 7);
/* Set watchdog duration */
writel(timeout_minutes * 60 * watchdog_freq, A385_WDT_DURATION);
/* Clear the watchdog expiration bit */
clrbits_32(A385_WDT_GLOBAL_STATUS, A385_WDT_GLOBAL_EXPIRED);
/* Enable watchdog timer */
setbits_32(A385_WDT_GLOBAL_CTRL, A385_WDT_GLOBAL_ENABLE);
/* Enable reset on watchdog */
setbits_32(A385_WD_RSTOUT_UNMASK, A385_WD_RSTOUT_UNMASK_GLOBAL);
/* Unmask reset for watchdog */
clrbits_32(A385_SYS_RSTOUT_MASK, A385_SYS_RSTOUT_MASK_WD);
arm: mvebu: turris_omnia: enable A385 watchdog before disabling MCU watchdog Commit aeb0ca64dbb5 ("arm: mvebu: turris_omnia: disable MCU watchdog in SPL when booting over UART") disabled MCU watchdog when booting over UART to ensure that watchdog does not reboot the board before UART transfer finishes. But if UART transfer fails for some reason, or if U-Boot binary crashes, then board hangs forever as there is no watchdog running which could reset it. To fix this issue, enable A385 watchdog with very high timeout before disabling MCU watchdog to ensure that even slow transfer can finish successfully before watchdog timer expires and also to ensure that if board hangs for some reason, watchdog will reset it. Omnia's MCU watchdog has fixed 120 seconds timer and it cannot be changed (without updating MCU firmware). A385 watchdog by default uses 25 MHz input clock and so the largest timeout value (2^32-1) can be just 171 seconds. But A385 watchdog can be switched to use NBCLK (L2) as input clock (on Turris Omnia it is 800 MHz clock) and in this case final watchdog clock frequency is calculated as: freq = NBCLK / 2 / (2 ^ R) So A385 watchdog on Turris Omnia can be configured to at most 1374 seconds (about 22 minutes). We set it to 10 minutes, which should be enough even for bigger U-Boot binaries or slower UART transfers. Both U-Boot and Linux kernel, when initializing A385 watchdog, switch watchdog timer to 25 MHz input clock, so usage of NBCLK input clock in U-Boot SPL does not cause any issues. Fixes: aeb0ca64dbb5 ("arm: mvebu: turris_omnia: disable MCU watchdog in SPL when booting over UART") Signed-off-by: Pali Rohár <pali@kernel.org> Signed-off-by: Marek Behún <marek.behun@nic.cz>
2021-11-09 16:14:02 +00:00
}
static bool disable_mcu_watchdog(void)
{
int ret;
puts("Disabling MCU watchdog... ");
ret = omnia_mcu_write(CMD_WATCHDOG_STATE, "\x00", 1);
if (ret) {
printf("omnia_mcu_write failed: %i\n", ret);
return false;
}
puts("disabled\n");
return true;
}
static bool omnia_detect_sata(const char *msata_slot)
{
int ret;
u16 stsword;
puts("MiniPCIe/mSATA card detection... ");
if (msata_slot) {
if (strcmp(msata_slot, "pcie") == 0) {
puts("forced to MiniPCIe via env\n");
return false;
} else if (strcmp(msata_slot, "sata") == 0) {
puts("forced to mSATA via env\n");
return true;
} else if (strcmp(msata_slot, "auto") != 0) {
printf("unsupported env value '%s', fallback to... ", msata_slot);
}
}
ret = omnia_mcu_read(CMD_GET_STATUS_WORD, &stsword, sizeof(stsword));
if (ret) {
printf("omnia_mcu_read failed: %i, defaulting to MiniPCIe card\n",
ret);
return false;
}
if (!(stsword & CARD_DET_STSBIT)) {
puts("none\n");
return false;
}
if (stsword & MSATA_IND_STSBIT)
puts("mSATA\n");
else
puts("MiniPCIe\n");
return stsword & MSATA_IND_STSBIT ? true : false;
}
arm: mvebu: turris_omnia: Add support for USB3.0 mode in WWAN MiniPCIe slot PCIe Mini CEM 2.1 spec added support for USB3.0 mode on MiniPCIe cards. USB3.0 and PCIe share same pins and only one function can be active at the same time. PCIe Mini CEM 2.1 spec says that determining function is platform specific and spec does not define any dedicated pin which could say if card is USB3.0-based or PCIe-based. Implement this platform specific decision (USB3.0 vs PCIe) for WWAN MiniPCIe slot on Turris Omnia via U-Boot env variable "omnia_wwan_slot", similarly like is implemented forced mode for MiniPCIe/mSATA slot via "omnia_msata_slot" env variable. Value "usb3" for "omnia_wwan_slot" would mean to set USB3.0 mode and value "pcie" original PCIe mode. A385 SoC on Turris Omnia has configurable fifth SerDes line (exported to MiniPCIe WWAN slot with SIM card) either to USB3.0 or PCIe functionality, so implementation of this new PCIe Mini CEM 2.1 feature is simple, by just configuring SerDes to USB 3.0 mode. Other twos MiniPCIe slots on Turris Omnia do not have this new functionality as their SerDes lines cannot be switched to USB3.0 functionality. Note that A385 SoC does not have too many USB3.0 blocks, so activating USB3.0 in MiniPCIe cause that one external USB3.0 USB-A port would loose USB3.0 functionality and would be downgraded just to USB2.0. By default this MiniPCIe WWAN slot is in PCIe mode, like before. To set this MiniPCIe WWAN slot to USB3.0 mode, call U-Boot commands: => setenv omnia_wwan_slot usb3 => saveenv => reset Signed-off-by: Pali Rohár <pali@kernel.org>
2022-03-02 11:47:58 +00:00
static bool omnia_detect_wwan_usb3(const char *wwan_slot)
{
puts("WWAN slot configuration... ");
if (wwan_slot && strcmp(wwan_slot, "usb3") == 0) {
puts("USB3.0\n");
return true;
}
if (wwan_slot && strcmp(wwan_slot, "pcie") != 0)
printf("unsupported env value '%s', fallback to... ", wwan_slot);
puts("PCIe+USB2.0\n");
return false;
}
void *env_sf_get_env_addr(void)
{
/* SPI Flash is mapped to address 0xD4000000 only in SPL */
#ifdef CONFIG_SPL_BUILD
return (void *)0xD4000000 + CONFIG_ENV_OFFSET;
#else
return NULL;
#endif
}
int hws_board_topology_load(struct serdes_map **serdes_map_array, u8 *count)
{
#ifdef CONFIG_SPL_ENV_SUPPORT
/* Do not use env_load() as malloc() pool is too small at this stage */
bool has_env = (env_init() == 0);
#endif
const char *env_value = NULL;
#ifdef CONFIG_SPL_ENV_SUPPORT
/* beware that env_get() returns static allocated memory */
env_value = has_env ? env_get("omnia_msata_slot") : NULL;
#endif
if (omnia_detect_sata(env_value)) {
/* Change SerDes for first mPCIe port (mSATA) from PCIe to SATA */
board_serdes_map[0].serdes_type = SATA0;
board_serdes_map[0].serdes_speed = SERDES_SPEED_6_GBPS;
board_serdes_map[0].serdes_mode = SERDES_DEFAULT_MODE;
}
arm: mvebu: turris_omnia: Add support for USB3.0 mode in WWAN MiniPCIe slot PCIe Mini CEM 2.1 spec added support for USB3.0 mode on MiniPCIe cards. USB3.0 and PCIe share same pins and only one function can be active at the same time. PCIe Mini CEM 2.1 spec says that determining function is platform specific and spec does not define any dedicated pin which could say if card is USB3.0-based or PCIe-based. Implement this platform specific decision (USB3.0 vs PCIe) for WWAN MiniPCIe slot on Turris Omnia via U-Boot env variable "omnia_wwan_slot", similarly like is implemented forced mode for MiniPCIe/mSATA slot via "omnia_msata_slot" env variable. Value "usb3" for "omnia_wwan_slot" would mean to set USB3.0 mode and value "pcie" original PCIe mode. A385 SoC on Turris Omnia has configurable fifth SerDes line (exported to MiniPCIe WWAN slot with SIM card) either to USB3.0 or PCIe functionality, so implementation of this new PCIe Mini CEM 2.1 feature is simple, by just configuring SerDes to USB 3.0 mode. Other twos MiniPCIe slots on Turris Omnia do not have this new functionality as their SerDes lines cannot be switched to USB3.0 functionality. Note that A385 SoC does not have too many USB3.0 blocks, so activating USB3.0 in MiniPCIe cause that one external USB3.0 USB-A port would loose USB3.0 functionality and would be downgraded just to USB2.0. By default this MiniPCIe WWAN slot is in PCIe mode, like before. To set this MiniPCIe WWAN slot to USB3.0 mode, call U-Boot commands: => setenv omnia_wwan_slot usb3 => saveenv => reset Signed-off-by: Pali Rohár <pali@kernel.org>
2022-03-02 11:47:58 +00:00
#ifdef CONFIG_SPL_ENV_SUPPORT
/* beware that env_get() returns static allocated memory */
env_value = has_env ? env_get("omnia_wwan_slot") : NULL;
#endif
if (omnia_detect_wwan_usb3(env_value)) {
/* Disable SerDes for USB 3.0 pins on the front USB-A port */
board_serdes_map[1].serdes_type = DEFAULT_SERDES;
/* Change SerDes for third mPCIe port (WWAN) from PCIe to USB 3.0 */
board_serdes_map[4].serdes_type = USB3_HOST0;
board_serdes_map[4].serdes_speed = SERDES_SPEED_5_GBPS;
board_serdes_map[4].serdes_mode = SERDES_DEFAULT_MODE;
}
*serdes_map_array = board_serdes_map;
*count = ARRAY_SIZE(board_serdes_map);
return 0;
}
struct omnia_eeprom {
u32 magic;
u32 ramsize;
char region[4];
u32 crc;
};
static bool omnia_read_eeprom(struct omnia_eeprom *oep)
{
struct udevice *chip;
u32 crc;
int ret;
chip = omnia_get_i2c_chip("EEPROM", OMNIA_I2C_EEPROM_CHIP_ADDR,
OMNIA_I2C_EEPROM_CHIP_LEN);
if (!chip)
return false;
ret = dm_i2c_read(chip, 0, (void *)oep, sizeof(*oep));
if (ret) {
printf("dm_i2c_read failed: %i, cannot read EEPROM\n", ret);
return false;
}
if (oep->magic != OMNIA_I2C_EEPROM_MAGIC) {
printf("bad EEPROM magic number (%08x, should be %08x)\n",
oep->magic, OMNIA_I2C_EEPROM_MAGIC);
return false;
}
crc = crc32(0, (void *)oep, sizeof(*oep) - 4);
if (crc != oep->crc) {
printf("bad EEPROM CRC (stored %08x, computed %08x)\n",
oep->crc, crc);
return false;
}
return true;
}
static int omnia_get_ram_size_gb(void)
{
static int ram_size;
struct omnia_eeprom oep;
if (!ram_size) {
/* Get the board config from EEPROM */
if (omnia_read_eeprom(&oep)) {
debug("Memory config in EEPROM: 0x%02x\n", oep.ramsize);
if (oep.ramsize == 0x2)
ram_size = 2;
else
ram_size = 1;
} else {
/* Hardcoded fallback */
puts("Memory config from EEPROM read failed!\n");
puts("Falling back to default 1 GiB!\n");
ram_size = 1;
}
}
return ram_size;
}
static const char * const omnia_get_mcu_type(void)
{
static const char * const mcu_types[] = {
[STS_MCU_TYPE_STM32] = "STM32",
[STS_MCU_TYPE_GD32] = "GD32",
[STS_MCU_TYPE_MKL] = "MKL",
[STS_MCU_TYPE_UNKN] = "unknown",
};
static const char * const mcu_types_with_perip_resets[] = {
[STS_MCU_TYPE_STM32] = "STM32 (with peripheral resets)",
[STS_MCU_TYPE_GD32] = "GD32 (with peripheral resets)",
[STS_MCU_TYPE_MKL] = "MKL (with peripheral resets)",
[STS_MCU_TYPE_UNKN] = "unknown (with peripheral resets)",
};
u16 stsword, features;
int ret;
ret = omnia_mcu_read(CMD_GET_STATUS_WORD, &stsword, sizeof(stsword));
if (ret)
return "unknown";
if (stsword & STS_FEATURES_SUPPORTED) {
ret = omnia_mcu_read(CMD_GET_FEATURES, &features, sizeof(features));
if (ret == 0 && (features & FEAT_PERIPH_MCU))
return mcu_types_with_perip_resets[stsword & STS_MCU_TYPE_MASK];
}
return mcu_types[stsword & STS_MCU_TYPE_MASK];
}
/*
* Define the DDR layout / topology here in the board file. This will
* be used by the DDR3 init code in the SPL U-Boot version to configure
* the DDR3 controller.
*/
static struct mv_ddr_topology_map board_topology_map_1g = {
DEBUG_LEVEL_ERROR,
0x1, /* active interfaces */
/* cs_mask, mirror, dqs_swap, ck_swap X PUPs */
{ { { {0x1, 0, 0, 0},
{0x1, 0, 0, 0},
{0x1, 0, 0, 0},
{0x1, 0, 0, 0},
{0x1, 0, 0, 0} },
SPEED_BIN_DDR_1600K, /* speed_bin */
MV_DDR_DEV_WIDTH_16BIT, /* memory_width */
MV_DDR_DIE_CAP_4GBIT, /* mem_size */
MV_DDR_FREQ_800, /* frequency */
0, 0, /* cas_wl cas_l */
MV_DDR_TEMP_NORMAL, /* temperature */
MV_DDR_TIM_2T} }, /* timing */
BUS_MASK_32BIT, /* Busses mask */
MV_DDR_CFG_DEFAULT, /* ddr configuration data source */
NOT_COMBINED, /* ddr twin-die combined */
{ {0} }, /* raw spd data */
{0} /* timing parameters */
};
static struct mv_ddr_topology_map board_topology_map_2g = {
DEBUG_LEVEL_ERROR,
0x1, /* active interfaces */
/* cs_mask, mirror, dqs_swap, ck_swap X PUPs */
{ { { {0x1, 0, 0, 0},
{0x1, 0, 0, 0},
{0x1, 0, 0, 0},
{0x1, 0, 0, 0},
{0x1, 0, 0, 0} },
SPEED_BIN_DDR_1600K, /* speed_bin */
MV_DDR_DEV_WIDTH_16BIT, /* memory_width */
MV_DDR_DIE_CAP_8GBIT, /* mem_size */
MV_DDR_FREQ_800, /* frequency */
0, 0, /* cas_wl cas_l */
MV_DDR_TEMP_NORMAL, /* temperature */
MV_DDR_TIM_2T} }, /* timing */
BUS_MASK_32BIT, /* Busses mask */
MV_DDR_CFG_DEFAULT, /* ddr configuration data source */
NOT_COMBINED, /* ddr twin-die combined */
{ {0} }, /* raw spd data */
{0} /* timing parameters */
};
struct mv_ddr_topology_map *mv_ddr_topology_map_get(void)
{
if (omnia_get_ram_size_gb() == 2)
return &board_topology_map_2g;
else
return &board_topology_map_1g;
}
static int set_regdomain(void)
{
struct omnia_eeprom oep;
char rd[3] = {' ', ' ', 0};
if (omnia_read_eeprom(&oep))
memcpy(rd, &oep.region, 2);
else
puts("EEPROM regdomain read failed.\n");
printf("Regdomain set to %s\n", rd);
return env_set("regdomain", rd);
}
static void handle_reset_button(void)
{
const char * const vars[1] = { "bootcmd_rescue", };
int ret;
u8 reset_status;
/*
* Ensure that bootcmd_rescue has always stock value, so that running
* run bootcmd_rescue
* always works correctly.
*/
env_set_default_vars(1, (char * const *)vars, 0);
ret = omnia_mcu_read(CMD_GET_RESET, &reset_status, 1);
if (ret) {
printf("omnia_mcu_read failed: %i, reset status unknown!\n",
ret);
return;
}
env_set_ulong("omnia_reset", reset_status);
if (reset_status) {
const char * const vars[2] = {
"bootcmd",
"distro_bootcmd",
};
/*
* Set the above envs to their default values, in case the user
* managed to break them.
*/
env_set_default_vars(2, (char * const *)vars, 0);
/* Ensure bootcmd_rescue is used by distroboot */
env_set("boot_targets", "rescue");
printf("RESET button was pressed, overwriting boot_targets!\n");
} else {
/*
* In case the user somehow managed to save environment with
* boot_targets=rescue, reset boot_targets to default value.
* This could happen in subsequent commands if bootcmd_rescue
* failed.
*/
if (!strcmp(env_get("boot_targets"), "rescue")) {
const char * const vars[1] = {
"boot_targets",
};
env_set_default_vars(1, (char * const *)vars, 0);
}
}
}
arm: mvebu: turris_omnia: Add support for design with SW reset signals New Turris Omnia HW board revision requires that software controls peripheral reset signals, namely PERST# signals on mPCIe slots, ethernet phy reset and lan switch reset. Those pins are connected to MCU controlled by MCU i2c API as GPIOs. On new HW board revision those pins stay in reset after board reset and software has to release these peripherals from reset manually. MCU announce this requirement by FEAT_PERIPH_MCU bit in CMD_GET_FEATURES command. On older HW board revisions when FEAT_PERIPH_MCU is not announced, all those reset signals are automatically released after board finish reset. Detect FEAT_PERIPH_MCU bit in board_fix_fdt() and ft_board_setup() functions and insert into device tree blob pcie "reset-gpios" and eth phy "phy-reset-gpios" properties with corresponding MCU gpio definitions. PCIe and eth PHY drivers then automatically release resets during device initialization. Both U-Boot and Linux kernel drivers support those device tree reset properties. Initialization of lan switch on new HW board revision is more complicated. Switch strapping pins are shared with switch RGMII pins. And strapping pins must be in specific configuration after releasing switch reset. Due to pin sharing, it is first required to switch A385 side of switch pins into GPIO mode, set strapping configuration, release switch from reset and after that switch A385 pins back to RGMII mode. Because this complicated setup is not supported by switch DSA drivers and cannot be expressed easily in device tree, implement it manually in SPL function spl_board_init(). So in proper U-Boot and OS/kernel would be lan switch initialized and be in same configuration like it was on old HW board revisions (where reset sequence did those steps at hardware level). Signed-off-by: Pali Rohár <pali@kernel.org> Reviewed-by: Stefan Roese <sr@denx.de> Reviewed-by: Marek Behún <kabel@kernel.org>
2022-07-29 11:29:07 +00:00
static void initialize_switch(void)
{
u32 val, val04, val08, val10, val14;
u16 ctrl[2];
int err;
printf("Initializing LAN eth switch... ");
/* Change RGMII pins to GPIO mode */
val = val04 = readl(MVEBU_MPP_BASE + 0x04);
val &= ~GENMASK(19, 16); /* MPP[12] := GPIO */
val &= ~GENMASK(23, 20); /* MPP[13] := GPIO */
val &= ~GENMASK(27, 24); /* MPP[14] := GPIO */
val &= ~GENMASK(31, 28); /* MPP[15] := GPIO */
writel(val, MVEBU_MPP_BASE + 0x04);
val = val08 = readl(MVEBU_MPP_BASE + 0x08);
val &= ~GENMASK(3, 0); /* MPP[16] := GPIO */
val &= ~GENMASK(23, 20); /* MPP[21] := GPIO */
writel(val, MVEBU_MPP_BASE + 0x08);
val = val10 = readl(MVEBU_MPP_BASE + 0x10);
val &= ~GENMASK(27, 24); /* MPP[38] := GPIO */
val &= ~GENMASK(31, 28); /* MPP[39] := GPIO */
writel(val, MVEBU_MPP_BASE + 0x10);
val = val14 = readl(MVEBU_MPP_BASE + 0x14);
val &= ~GENMASK(3, 0); /* MPP[40] := GPIO */
val &= ~GENMASK(7, 4); /* MPP[41] := GPIO */
writel(val, MVEBU_MPP_BASE + 0x14);
/* Set initial values for switch reset strapping pins */
val = readl(MVEBU_GPIO0_BASE + 0x00);
val |= BIT(12); /* GPIO[12] := 1 */
val |= BIT(13); /* GPIO[13] := 1 */
val |= BIT(14); /* GPIO[14] := 1 */
val |= BIT(15); /* GPIO[15] := 1 */
val &= ~BIT(16); /* GPIO[16] := 0 */
val |= BIT(21); /* GPIO[21] := 1 */
writel(val, MVEBU_GPIO0_BASE + 0x00);
val = readl(MVEBU_GPIO1_BASE + 0x00);
val |= BIT(6); /* GPIO[38] := 1 */
val |= BIT(7); /* GPIO[39] := 1 */
val |= BIT(8); /* GPIO[40] := 1 */
val &= ~BIT(9); /* GPIO[41] := 0 */
writel(val, MVEBU_GPIO1_BASE + 0x00);
val = readl(MVEBU_GPIO0_BASE + 0x04);
val &= ~BIT(12); /* GPIO[12] := Out Enable */
val &= ~BIT(13); /* GPIO[13] := Out Enable */
val &= ~BIT(14); /* GPIO[14] := Out Enable */
val &= ~BIT(15); /* GPIO[15] := Out Enable */
val &= ~BIT(16); /* GPIO[16] := Out Enable */
val &= ~BIT(21); /* GPIO[21] := Out Enable */
writel(val, MVEBU_GPIO0_BASE + 0x04);
val = readl(MVEBU_GPIO1_BASE + 0x04);
val &= ~BIT(6); /* GPIO[38] := Out Enable */
val &= ~BIT(7); /* GPIO[39] := Out Enable */
val &= ~BIT(8); /* GPIO[40] := Out Enable */
val &= ~BIT(9); /* GPIO[41] := Out Enable */
writel(val, MVEBU_GPIO1_BASE + 0x04);
/* Release switch reset */
ctrl[0] = EXT_CTL_nRES_LAN;
ctrl[1] = EXT_CTL_nRES_LAN;
err = omnia_mcu_write(CMD_EXT_CONTROL, ctrl, sizeof(ctrl));
mdelay(10);
/* Change RGMII pins back to RGMII mode */
writel(val04, MVEBU_MPP_BASE + 0x04);
writel(val08, MVEBU_MPP_BASE + 0x08);
writel(val10, MVEBU_MPP_BASE + 0x10);
writel(val14, MVEBU_MPP_BASE + 0x14);
puts(err ? "failed\n" : "done\n");
}
int board_early_init_f(void)
{
/* Configure MPP */
writel(0x11111111, MVEBU_MPP_BASE + 0x00);
writel(0x11111111, MVEBU_MPP_BASE + 0x04);
writel(0x11244011, MVEBU_MPP_BASE + 0x08);
writel(0x22222111, MVEBU_MPP_BASE + 0x0c);
writel(0x22200002, MVEBU_MPP_BASE + 0x10);
writel(0x30042022, MVEBU_MPP_BASE + 0x14);
writel(0x55550555, MVEBU_MPP_BASE + 0x18);
writel(0x00005550, MVEBU_MPP_BASE + 0x1c);
/* Set GPP Out value */
writel(OMNIA_GPP_OUT_VAL_LOW, MVEBU_GPIO0_BASE + 0x00);
writel(OMNIA_GPP_OUT_VAL_MID, MVEBU_GPIO1_BASE + 0x00);
/* Set GPP Polarity */
writel(OMNIA_GPP_POL_LOW, MVEBU_GPIO0_BASE + 0x0c);
writel(OMNIA_GPP_POL_MID, MVEBU_GPIO1_BASE + 0x0c);
/* Set GPP Out Enable */
writel(OMNIA_GPP_OUT_ENA_LOW, MVEBU_GPIO0_BASE + 0x04);
writel(OMNIA_GPP_OUT_ENA_MID, MVEBU_GPIO1_BASE + 0x04);
return 0;
}
void spl_board_init(void)
{
arm: mvebu: turris_omnia: Add support for design with SW reset signals New Turris Omnia HW board revision requires that software controls peripheral reset signals, namely PERST# signals on mPCIe slots, ethernet phy reset and lan switch reset. Those pins are connected to MCU controlled by MCU i2c API as GPIOs. On new HW board revision those pins stay in reset after board reset and software has to release these peripherals from reset manually. MCU announce this requirement by FEAT_PERIPH_MCU bit in CMD_GET_FEATURES command. On older HW board revisions when FEAT_PERIPH_MCU is not announced, all those reset signals are automatically released after board finish reset. Detect FEAT_PERIPH_MCU bit in board_fix_fdt() and ft_board_setup() functions and insert into device tree blob pcie "reset-gpios" and eth phy "phy-reset-gpios" properties with corresponding MCU gpio definitions. PCIe and eth PHY drivers then automatically release resets during device initialization. Both U-Boot and Linux kernel drivers support those device tree reset properties. Initialization of lan switch on new HW board revision is more complicated. Switch strapping pins are shared with switch RGMII pins. And strapping pins must be in specific configuration after releasing switch reset. Due to pin sharing, it is first required to switch A385 side of switch pins into GPIO mode, set strapping configuration, release switch from reset and after that switch A385 pins back to RGMII mode. Because this complicated setup is not supported by switch DSA drivers and cannot be expressed easily in device tree, implement it manually in SPL function spl_board_init(). So in proper U-Boot and OS/kernel would be lan switch initialized and be in same configuration like it was on old HW board revisions (where reset sequence did those steps at hardware level). Signed-off-by: Pali Rohár <pali@kernel.org> Reviewed-by: Stefan Roese <sr@denx.de> Reviewed-by: Marek Behún <kabel@kernel.org>
2022-07-29 11:29:07 +00:00
u16 val;
int ret;
/*
* If booting from UART, disable MCU watchdog in SPL, since uploading
arm: mvebu: turris_omnia: enable A385 watchdog before disabling MCU watchdog Commit aeb0ca64dbb5 ("arm: mvebu: turris_omnia: disable MCU watchdog in SPL when booting over UART") disabled MCU watchdog when booting over UART to ensure that watchdog does not reboot the board before UART transfer finishes. But if UART transfer fails for some reason, or if U-Boot binary crashes, then board hangs forever as there is no watchdog running which could reset it. To fix this issue, enable A385 watchdog with very high timeout before disabling MCU watchdog to ensure that even slow transfer can finish successfully before watchdog timer expires and also to ensure that if board hangs for some reason, watchdog will reset it. Omnia's MCU watchdog has fixed 120 seconds timer and it cannot be changed (without updating MCU firmware). A385 watchdog by default uses 25 MHz input clock and so the largest timeout value (2^32-1) can be just 171 seconds. But A385 watchdog can be switched to use NBCLK (L2) as input clock (on Turris Omnia it is 800 MHz clock) and in this case final watchdog clock frequency is calculated as: freq = NBCLK / 2 / (2 ^ R) So A385 watchdog on Turris Omnia can be configured to at most 1374 seconds (about 22 minutes). We set it to 10 minutes, which should be enough even for bigger U-Boot binaries or slower UART transfers. Both U-Boot and Linux kernel, when initializing A385 watchdog, switch watchdog timer to 25 MHz input clock, so usage of NBCLK input clock in U-Boot SPL does not cause any issues. Fixes: aeb0ca64dbb5 ("arm: mvebu: turris_omnia: disable MCU watchdog in SPL when booting over UART") Signed-off-by: Pali Rohár <pali@kernel.org> Signed-off-by: Marek Behún <marek.behun@nic.cz>
2021-11-09 16:14:02 +00:00
* U-Boot proper can take too much time and trigger it. Instead enable
* A385 watchdog with very high timeout (10 minutes) to prevent hangup.
*/
arm: mvebu: turris_omnia: enable A385 watchdog before disabling MCU watchdog Commit aeb0ca64dbb5 ("arm: mvebu: turris_omnia: disable MCU watchdog in SPL when booting over UART") disabled MCU watchdog when booting over UART to ensure that watchdog does not reboot the board before UART transfer finishes. But if UART transfer fails for some reason, or if U-Boot binary crashes, then board hangs forever as there is no watchdog running which could reset it. To fix this issue, enable A385 watchdog with very high timeout before disabling MCU watchdog to ensure that even slow transfer can finish successfully before watchdog timer expires and also to ensure that if board hangs for some reason, watchdog will reset it. Omnia's MCU watchdog has fixed 120 seconds timer and it cannot be changed (without updating MCU firmware). A385 watchdog by default uses 25 MHz input clock and so the largest timeout value (2^32-1) can be just 171 seconds. But A385 watchdog can be switched to use NBCLK (L2) as input clock (on Turris Omnia it is 800 MHz clock) and in this case final watchdog clock frequency is calculated as: freq = NBCLK / 2 / (2 ^ R) So A385 watchdog on Turris Omnia can be configured to at most 1374 seconds (about 22 minutes). We set it to 10 minutes, which should be enough even for bigger U-Boot binaries or slower UART transfers. Both U-Boot and Linux kernel, when initializing A385 watchdog, switch watchdog timer to 25 MHz input clock, so usage of NBCLK input clock in U-Boot SPL does not cause any issues. Fixes: aeb0ca64dbb5 ("arm: mvebu: turris_omnia: disable MCU watchdog in SPL when booting over UART") Signed-off-by: Pali Rohár <pali@kernel.org> Signed-off-by: Marek Behún <marek.behun@nic.cz>
2021-11-09 16:14:02 +00:00
if (get_boot_device() == BOOT_DEVICE_UART) {
enable_a385_watchdog(10);
disable_mcu_watchdog();
arm: mvebu: turris_omnia: enable A385 watchdog before disabling MCU watchdog Commit aeb0ca64dbb5 ("arm: mvebu: turris_omnia: disable MCU watchdog in SPL when booting over UART") disabled MCU watchdog when booting over UART to ensure that watchdog does not reboot the board before UART transfer finishes. But if UART transfer fails for some reason, or if U-Boot binary crashes, then board hangs forever as there is no watchdog running which could reset it. To fix this issue, enable A385 watchdog with very high timeout before disabling MCU watchdog to ensure that even slow transfer can finish successfully before watchdog timer expires and also to ensure that if board hangs for some reason, watchdog will reset it. Omnia's MCU watchdog has fixed 120 seconds timer and it cannot be changed (without updating MCU firmware). A385 watchdog by default uses 25 MHz input clock and so the largest timeout value (2^32-1) can be just 171 seconds. But A385 watchdog can be switched to use NBCLK (L2) as input clock (on Turris Omnia it is 800 MHz clock) and in this case final watchdog clock frequency is calculated as: freq = NBCLK / 2 / (2 ^ R) So A385 watchdog on Turris Omnia can be configured to at most 1374 seconds (about 22 minutes). We set it to 10 minutes, which should be enough even for bigger U-Boot binaries or slower UART transfers. Both U-Boot and Linux kernel, when initializing A385 watchdog, switch watchdog timer to 25 MHz input clock, so usage of NBCLK input clock in U-Boot SPL does not cause any issues. Fixes: aeb0ca64dbb5 ("arm: mvebu: turris_omnia: disable MCU watchdog in SPL when booting over UART") Signed-off-by: Pali Rohár <pali@kernel.org> Signed-off-by: Marek Behún <marek.behun@nic.cz>
2021-11-09 16:14:02 +00:00
}
arm: mvebu: turris_omnia: Add support for design with SW reset signals New Turris Omnia HW board revision requires that software controls peripheral reset signals, namely PERST# signals on mPCIe slots, ethernet phy reset and lan switch reset. Those pins are connected to MCU controlled by MCU i2c API as GPIOs. On new HW board revision those pins stay in reset after board reset and software has to release these peripherals from reset manually. MCU announce this requirement by FEAT_PERIPH_MCU bit in CMD_GET_FEATURES command. On older HW board revisions when FEAT_PERIPH_MCU is not announced, all those reset signals are automatically released after board finish reset. Detect FEAT_PERIPH_MCU bit in board_fix_fdt() and ft_board_setup() functions and insert into device tree blob pcie "reset-gpios" and eth phy "phy-reset-gpios" properties with corresponding MCU gpio definitions. PCIe and eth PHY drivers then automatically release resets during device initialization. Both U-Boot and Linux kernel drivers support those device tree reset properties. Initialization of lan switch on new HW board revision is more complicated. Switch strapping pins are shared with switch RGMII pins. And strapping pins must be in specific configuration after releasing switch reset. Due to pin sharing, it is first required to switch A385 side of switch pins into GPIO mode, set strapping configuration, release switch from reset and after that switch A385 pins back to RGMII mode. Because this complicated setup is not supported by switch DSA drivers and cannot be expressed easily in device tree, implement it manually in SPL function spl_board_init(). So in proper U-Boot and OS/kernel would be lan switch initialized and be in same configuration like it was on old HW board revisions (where reset sequence did those steps at hardware level). Signed-off-by: Pali Rohár <pali@kernel.org> Reviewed-by: Stefan Roese <sr@denx.de> Reviewed-by: Marek Behún <kabel@kernel.org>
2022-07-29 11:29:07 +00:00
/*
* When MCU controls peripheral resets then release LAN eth switch from
* the reset and initialize it. When MCU does not control peripheral
* resets then LAN eth switch is initialized automatically by bootstrap
* pins when A385 is released from the reset.
*/
ret = omnia_mcu_read(CMD_GET_STATUS_WORD, &val, sizeof(val));
if (ret == 0 && (val & STS_FEATURES_SUPPORTED)) {
ret = omnia_mcu_read(CMD_GET_FEATURES, &val, sizeof(val));
if (ret == 0 && (val & FEAT_PERIPH_MCU))
initialize_switch();
}
}
#if IS_ENABLED(CONFIG_OF_BOARD_FIXUP) || IS_ENABLED(CONFIG_OF_BOARD_SETUP)
static void disable_sata_node(void *blob)
{
int node;
fdt_for_each_node_by_compatible(node, blob, -1, "marvell,armada-380-ahci") {
if (!fdtdec_get_is_enabled(blob, node))
continue;
if (fdt_status_disabled(blob, node) < 0)
printf("Cannot disable SATA DT node!\n");
else
debug("Disabled SATA DT node\n");
return;
}
printf("Cannot find SATA DT node!\n");
}
static void disable_pcie_node(void *blob, int port)
{
int node;
fdt_for_each_node_by_compatible(node, blob, -1, "marvell,armada-370-pcie") {
int port_node;
if (!fdtdec_get_is_enabled(blob, node))
continue;
fdt_for_each_subnode (port_node, blob, node) {
if (!fdtdec_get_is_enabled(blob, port_node))
continue;
if (fdtdec_get_int(blob, port_node, "marvell,pcie-port", -1) != port)
continue;
if (fdt_status_disabled(blob, port_node) < 0)
printf("Cannot disable PCIe port %d DT node!\n", port);
else
debug("Disabled PCIe port %d DT node\n", port);
return;
}
}
printf("Cannot find PCIe port %d DT node!\n", port);
}
static void fixup_msata_port_nodes(void *blob)
{
bool mode_sata;
/*
* Determine if SerDes 0 is configured to SATA mode.
* We do this instead of calling omnia_detect_sata() to avoid another
* call to the MCU. By this time the common PHYs are initialized (it is
* done in SPL), so we can read this common PHY register.
*/
mode_sata = (readl(MVEBU_REGISTER(0x183fc)) & GENMASK(3, 0)) == 2;
/*
* We're either adding status = "disabled" property, or changing
* status = "okay" to status = "disabled". In both cases we'll need more
* space. Increase the size a little.
*/
if (fdt_increase_size(blob, 32) < 0) {
printf("Cannot increase FDT size!\n");
return;
}
if (!mode_sata) {
/* If mSATA card is not present, disable SATA DT node */
disable_sata_node(blob);
} else {
/* Otherwise disable PCIe port 0 DT node (MiniPCIe / mSATA port) */
disable_pcie_node(blob, 0);
}
}
arm: mvebu: turris_omnia: Add support for USB3.0 mode in WWAN MiniPCIe slot PCIe Mini CEM 2.1 spec added support for USB3.0 mode on MiniPCIe cards. USB3.0 and PCIe share same pins and only one function can be active at the same time. PCIe Mini CEM 2.1 spec says that determining function is platform specific and spec does not define any dedicated pin which could say if card is USB3.0-based or PCIe-based. Implement this platform specific decision (USB3.0 vs PCIe) for WWAN MiniPCIe slot on Turris Omnia via U-Boot env variable "omnia_wwan_slot", similarly like is implemented forced mode for MiniPCIe/mSATA slot via "omnia_msata_slot" env variable. Value "usb3" for "omnia_wwan_slot" would mean to set USB3.0 mode and value "pcie" original PCIe mode. A385 SoC on Turris Omnia has configurable fifth SerDes line (exported to MiniPCIe WWAN slot with SIM card) either to USB3.0 or PCIe functionality, so implementation of this new PCIe Mini CEM 2.1 feature is simple, by just configuring SerDes to USB 3.0 mode. Other twos MiniPCIe slots on Turris Omnia do not have this new functionality as their SerDes lines cannot be switched to USB3.0 functionality. Note that A385 SoC does not have too many USB3.0 blocks, so activating USB3.0 in MiniPCIe cause that one external USB3.0 USB-A port would loose USB3.0 functionality and would be downgraded just to USB2.0. By default this MiniPCIe WWAN slot is in PCIe mode, like before. To set this MiniPCIe WWAN slot to USB3.0 mode, call U-Boot commands: => setenv omnia_wwan_slot usb3 => saveenv => reset Signed-off-by: Pali Rohár <pali@kernel.org>
2022-03-02 11:47:58 +00:00
static void fixup_wwan_port_nodes(void *blob)
{
bool mode_usb3;
/* Determine if SerDes 4 is configured to USB3 mode */
mode_usb3 = ((readl(MVEBU_REGISTER(0x183fc)) & GENMASK(19, 16)) >> 16) == 4;
/* If SerDes 4 is not configured to USB3 mode then nothing is needed to fixup */
if (!mode_usb3)
return;
/*
* We're either adding status = "disabled" property, or changing
* status = "okay" to status = "disabled". In both cases we'll need more
* space. Increase the size a little.
*/
if (fdt_increase_size(blob, 32) < 0) {
printf("Cannot increase FDT size!\n");
return;
}
/* Disable PCIe port 2 DT node (WWAN) */
disable_pcie_node(blob, 2);
}
arm: mvebu: turris_omnia: Add support for design with SW reset signals New Turris Omnia HW board revision requires that software controls peripheral reset signals, namely PERST# signals on mPCIe slots, ethernet phy reset and lan switch reset. Those pins are connected to MCU controlled by MCU i2c API as GPIOs. On new HW board revision those pins stay in reset after board reset and software has to release these peripherals from reset manually. MCU announce this requirement by FEAT_PERIPH_MCU bit in CMD_GET_FEATURES command. On older HW board revisions when FEAT_PERIPH_MCU is not announced, all those reset signals are automatically released after board finish reset. Detect FEAT_PERIPH_MCU bit in board_fix_fdt() and ft_board_setup() functions and insert into device tree blob pcie "reset-gpios" and eth phy "phy-reset-gpios" properties with corresponding MCU gpio definitions. PCIe and eth PHY drivers then automatically release resets during device initialization. Both U-Boot and Linux kernel drivers support those device tree reset properties. Initialization of lan switch on new HW board revision is more complicated. Switch strapping pins are shared with switch RGMII pins. And strapping pins must be in specific configuration after releasing switch reset. Due to pin sharing, it is first required to switch A385 side of switch pins into GPIO mode, set strapping configuration, release switch from reset and after that switch A385 pins back to RGMII mode. Because this complicated setup is not supported by switch DSA drivers and cannot be expressed easily in device tree, implement it manually in SPL function spl_board_init(). So in proper U-Boot and OS/kernel would be lan switch initialized and be in same configuration like it was on old HW board revisions (where reset sequence did those steps at hardware level). Signed-off-by: Pali Rohár <pali@kernel.org> Reviewed-by: Stefan Roese <sr@denx.de> Reviewed-by: Marek Behún <kabel@kernel.org>
2022-07-29 11:29:07 +00:00
static int insert_mcu_gpio_prop(void *blob, int node, const char *prop,
unsigned int phandle, u32 bank, u32 gpio,
u32 flags)
{
fdt32_t val[4] = { cpu_to_fdt32(phandle), cpu_to_fdt32(bank),
cpu_to_fdt32(gpio), cpu_to_fdt32(flags) };
return fdt_setprop(blob, node, prop, &val, sizeof(val));
}
static int fixup_mcu_gpio_in_pcie_nodes(void *blob)
{
unsigned int mcu_phandle;
int port, gpio;
int pcie_node;
int port_node;
int ret;
ret = fdt_increase_size(blob, 128);
if (ret < 0) {
printf("Cannot increase FDT size!\n");
return ret;
}
mcu_phandle = fdt_create_phandle_by_compatible(blob, "cznic,turris-omnia-mcu");
if (!mcu_phandle)
return -FDT_ERR_NOPHANDLES;
fdt_for_each_node_by_compatible(pcie_node, blob, -1, "marvell,armada-370-pcie") {
if (!fdtdec_get_is_enabled(blob, pcie_node))
continue;
fdt_for_each_subnode(port_node, blob, pcie_node) {
if (!fdtdec_get_is_enabled(blob, port_node))
continue;
port = fdtdec_get_int(blob, port_node, "marvell,pcie-port", -1);
if (port == 0)
gpio = ilog2(EXT_CTL_nPERST0);
else if (port == 1)
gpio = ilog2(EXT_CTL_nPERST1);
else if (port == 2)
gpio = ilog2(EXT_CTL_nPERST2);
else
continue;
/* insert: reset-gpios = <&mcu 2 gpio GPIO_ACTIVE_LOW>; */
ret = insert_mcu_gpio_prop(blob, port_node, "reset-gpios",
mcu_phandle, 2, gpio, GPIO_ACTIVE_LOW);
if (ret < 0)
return ret;
}
}
return 0;
}
static int fixup_mcu_gpio_in_eth_wan_node(void *blob)
{
unsigned int mcu_phandle;
int eth_wan_node;
int ret;
ret = fdt_increase_size(blob, 64);
if (ret < 0) {
printf("Cannot increase FDT size!\n");
return ret;
}
eth_wan_node = fdt_path_offset(blob, "ethernet2");
if (eth_wan_node < 0)
return eth_wan_node;
mcu_phandle = fdt_create_phandle_by_compatible(blob, "cznic,turris-omnia-mcu");
if (!mcu_phandle)
return -FDT_ERR_NOPHANDLES;
/* insert: phy-reset-gpios = <&mcu 2 gpio GPIO_ACTIVE_LOW>; */
ret = insert_mcu_gpio_prop(blob, eth_wan_node, "phy-reset-gpios",
mcu_phandle, 2, ilog2(EXT_CTL_nRES_PHY), GPIO_ACTIVE_LOW);
if (ret < 0)
return ret;
return 0;
}
#endif
#if IS_ENABLED(CONFIG_OF_BOARD_FIXUP)
int board_fix_fdt(void *blob)
{
arm: mvebu: turris_omnia: Add support for design with SW reset signals New Turris Omnia HW board revision requires that software controls peripheral reset signals, namely PERST# signals on mPCIe slots, ethernet phy reset and lan switch reset. Those pins are connected to MCU controlled by MCU i2c API as GPIOs. On new HW board revision those pins stay in reset after board reset and software has to release these peripherals from reset manually. MCU announce this requirement by FEAT_PERIPH_MCU bit in CMD_GET_FEATURES command. On older HW board revisions when FEAT_PERIPH_MCU is not announced, all those reset signals are automatically released after board finish reset. Detect FEAT_PERIPH_MCU bit in board_fix_fdt() and ft_board_setup() functions and insert into device tree blob pcie "reset-gpios" and eth phy "phy-reset-gpios" properties with corresponding MCU gpio definitions. PCIe and eth PHY drivers then automatically release resets during device initialization. Both U-Boot and Linux kernel drivers support those device tree reset properties. Initialization of lan switch on new HW board revision is more complicated. Switch strapping pins are shared with switch RGMII pins. And strapping pins must be in specific configuration after releasing switch reset. Due to pin sharing, it is first required to switch A385 side of switch pins into GPIO mode, set strapping configuration, release switch from reset and after that switch A385 pins back to RGMII mode. Because this complicated setup is not supported by switch DSA drivers and cannot be expressed easily in device tree, implement it manually in SPL function spl_board_init(). So in proper U-Boot and OS/kernel would be lan switch initialized and be in same configuration like it was on old HW board revisions (where reset sequence did those steps at hardware level). Signed-off-by: Pali Rohár <pali@kernel.org> Reviewed-by: Stefan Roese <sr@denx.de> Reviewed-by: Marek Behún <kabel@kernel.org>
2022-07-29 11:29:07 +00:00
u16 val;
int ret;
ret = omnia_mcu_read(CMD_GET_STATUS_WORD, &val, sizeof(val));
if (ret == 0 && (val & STS_FEATURES_SUPPORTED)) {
ret = omnia_mcu_read(CMD_GET_FEATURES, &val, sizeof(val));
if (ret == 0 && (val & FEAT_PERIPH_MCU)) {
fixup_mcu_gpio_in_pcie_nodes(blob);
fixup_mcu_gpio_in_eth_wan_node(blob);
}
}
fixup_msata_port_nodes(blob);
arm: mvebu: turris_omnia: Add support for USB3.0 mode in WWAN MiniPCIe slot PCIe Mini CEM 2.1 spec added support for USB3.0 mode on MiniPCIe cards. USB3.0 and PCIe share same pins and only one function can be active at the same time. PCIe Mini CEM 2.1 spec says that determining function is platform specific and spec does not define any dedicated pin which could say if card is USB3.0-based or PCIe-based. Implement this platform specific decision (USB3.0 vs PCIe) for WWAN MiniPCIe slot on Turris Omnia via U-Boot env variable "omnia_wwan_slot", similarly like is implemented forced mode for MiniPCIe/mSATA slot via "omnia_msata_slot" env variable. Value "usb3" for "omnia_wwan_slot" would mean to set USB3.0 mode and value "pcie" original PCIe mode. A385 SoC on Turris Omnia has configurable fifth SerDes line (exported to MiniPCIe WWAN slot with SIM card) either to USB3.0 or PCIe functionality, so implementation of this new PCIe Mini CEM 2.1 feature is simple, by just configuring SerDes to USB 3.0 mode. Other twos MiniPCIe slots on Turris Omnia do not have this new functionality as their SerDes lines cannot be switched to USB3.0 functionality. Note that A385 SoC does not have too many USB3.0 blocks, so activating USB3.0 in MiniPCIe cause that one external USB3.0 USB-A port would loose USB3.0 functionality and would be downgraded just to USB2.0. By default this MiniPCIe WWAN slot is in PCIe mode, like before. To set this MiniPCIe WWAN slot to USB3.0 mode, call U-Boot commands: => setenv omnia_wwan_slot usb3 => saveenv => reset Signed-off-by: Pali Rohár <pali@kernel.org>
2022-03-02 11:47:58 +00:00
fixup_wwan_port_nodes(blob);
return 0;
}
#endif
int board_init(void)
{
/* address of boot parameters */
gd->bd->bi_boot_params = mvebu_sdram_bar(0) + 0x100;
return 0;
}
int board_late_init(void)
{
/*
* If not booting from UART, MCU watchdog was not disabled in SPL,
* disable it now.
*/
if (get_boot_device() != BOOT_DEVICE_UART)
disable_mcu_watchdog();
set_regdomain();
handle_reset_button();
pci_init();
return 0;
}
int show_board_info(void)
{
u32 version_num, serial_num;
int err;
err = turris_atsha_otp_get_serial_number(&version_num, &serial_num);
printf("Model: Turris Omnia\n");
printf(" MCU type: %s\n", omnia_get_mcu_type());
printf(" RAM size: %i MiB\n", omnia_get_ram_size_gb() * 1024);
if (err)
printf(" Serial Number: unknown\n");
else
printf(" Serial Number: %08X%08X\n", be32_to_cpu(version_num),
be32_to_cpu(serial_num));
return 0;
}
int misc_init_r(void)
{
turris_atsha_otp_init_mac_addresses(1);
return 0;
}
#if defined(CONFIG_OF_BOARD_SETUP)
/*
* I plan to generalize this function and move it to common/fdt_support.c.
* This will require some more work on multiple boards, though, so for now leave
* it here.
*/
static bool fixup_mtd_partitions(void *blob, int offset, struct mtd_info *mtd)
{
struct mtd_info *slave;
int parts;
parts = fdt_subnode_offset(blob, offset, "partitions");
if (parts >= 0) {
if (fdt_del_node(blob, parts) < 0)
return false;
}
if (fdt_increase_size(blob, 512) < 0)
return false;
parts = fdt_add_subnode(blob, offset, "partitions");
if (parts < 0)
return false;
if (fdt_setprop_u32(blob, parts, "#address-cells", 1) < 0)
return false;
if (fdt_setprop_u32(blob, parts, "#size-cells", 1) < 0)
return false;
if (fdt_setprop_string(blob, parts, "compatible",
"fixed-partitions") < 0)
return false;
mtd_probe_devices();
list_for_each_entry_reverse(slave, &mtd->partitions, node) {
char name[32];
int part;
snprintf(name, sizeof(name), "partition@%llx", slave->offset);
part = fdt_add_subnode(blob, parts, name);
if (part < 0)
return false;
if (fdt_setprop_u32(blob, part, "reg", slave->offset) < 0)
return false;
if (fdt_appendprop_u32(blob, part, "reg", slave->size) < 0)
return false;
if (fdt_setprop_string(blob, part, "label", slave->name) < 0)
return false;
if (!(slave->flags & MTD_WRITEABLE))
if (fdt_setprop_empty(blob, part, "read-only") < 0)
return false;
if (slave->flags & MTD_POWERUP_LOCK)
if (fdt_setprop_empty(blob, part, "lock") < 0)
return false;
}
return true;
}
static void fixup_spi_nor_partitions(void *blob)
{
struct mtd_info *mtd;
int node;
mtd = get_mtd_device_nm(OMNIA_SPI_NOR_PATH);
if (IS_ERR_OR_NULL(mtd))
goto fail;
node = fdt_path_offset(blob, OMNIA_SPI_NOR_PATH);
if (node < 0)
goto fail;
if (!fixup_mtd_partitions(blob, node, mtd))
goto fail;
put_mtd_device(mtd);
return;
fail:
printf("Failed fixing SPI NOR partitions!\n");
if (!IS_ERR_OR_NULL(mtd))
put_mtd_device(mtd);
}
int ft_board_setup(void *blob, struct bd_info *bd)
{
arm: mvebu: turris_omnia: Add support for design with SW reset signals New Turris Omnia HW board revision requires that software controls peripheral reset signals, namely PERST# signals on mPCIe slots, ethernet phy reset and lan switch reset. Those pins are connected to MCU controlled by MCU i2c API as GPIOs. On new HW board revision those pins stay in reset after board reset and software has to release these peripherals from reset manually. MCU announce this requirement by FEAT_PERIPH_MCU bit in CMD_GET_FEATURES command. On older HW board revisions when FEAT_PERIPH_MCU is not announced, all those reset signals are automatically released after board finish reset. Detect FEAT_PERIPH_MCU bit in board_fix_fdt() and ft_board_setup() functions and insert into device tree blob pcie "reset-gpios" and eth phy "phy-reset-gpios" properties with corresponding MCU gpio definitions. PCIe and eth PHY drivers then automatically release resets during device initialization. Both U-Boot and Linux kernel drivers support those device tree reset properties. Initialization of lan switch on new HW board revision is more complicated. Switch strapping pins are shared with switch RGMII pins. And strapping pins must be in specific configuration after releasing switch reset. Due to pin sharing, it is first required to switch A385 side of switch pins into GPIO mode, set strapping configuration, release switch from reset and after that switch A385 pins back to RGMII mode. Because this complicated setup is not supported by switch DSA drivers and cannot be expressed easily in device tree, implement it manually in SPL function spl_board_init(). So in proper U-Boot and OS/kernel would be lan switch initialized and be in same configuration like it was on old HW board revisions (where reset sequence did those steps at hardware level). Signed-off-by: Pali Rohár <pali@kernel.org> Reviewed-by: Stefan Roese <sr@denx.de> Reviewed-by: Marek Behún <kabel@kernel.org>
2022-07-29 11:29:07 +00:00
int node;
/*
* U-Boot's FDT blob contains phy-reset-gpios in ethernet2
* node when MCU controls all peripherals resets.
* Fixup MCU GPIO nodes in PCIe and eth wan nodes in this case.
*/
node = fdt_path_offset(gd->fdt_blob, "ethernet2");
if (node >= 0 && fdt_getprop(gd->fdt_blob, node, "phy-reset-gpios", NULL)) {
fixup_mcu_gpio_in_pcie_nodes(blob);
fixup_mcu_gpio_in_eth_wan_node(blob);
}
fixup_spi_nor_partitions(blob);
fixup_msata_port_nodes(blob);
arm: mvebu: turris_omnia: Add support for USB3.0 mode in WWAN MiniPCIe slot PCIe Mini CEM 2.1 spec added support for USB3.0 mode on MiniPCIe cards. USB3.0 and PCIe share same pins and only one function can be active at the same time. PCIe Mini CEM 2.1 spec says that determining function is platform specific and spec does not define any dedicated pin which could say if card is USB3.0-based or PCIe-based. Implement this platform specific decision (USB3.0 vs PCIe) for WWAN MiniPCIe slot on Turris Omnia via U-Boot env variable "omnia_wwan_slot", similarly like is implemented forced mode for MiniPCIe/mSATA slot via "omnia_msata_slot" env variable. Value "usb3" for "omnia_wwan_slot" would mean to set USB3.0 mode and value "pcie" original PCIe mode. A385 SoC on Turris Omnia has configurable fifth SerDes line (exported to MiniPCIe WWAN slot with SIM card) either to USB3.0 or PCIe functionality, so implementation of this new PCIe Mini CEM 2.1 feature is simple, by just configuring SerDes to USB 3.0 mode. Other twos MiniPCIe slots on Turris Omnia do not have this new functionality as their SerDes lines cannot be switched to USB3.0 functionality. Note that A385 SoC does not have too many USB3.0 blocks, so activating USB3.0 in MiniPCIe cause that one external USB3.0 USB-A port would loose USB3.0 functionality and would be downgraded just to USB2.0. By default this MiniPCIe WWAN slot is in PCIe mode, like before. To set this MiniPCIe WWAN slot to USB3.0 mode, call U-Boot commands: => setenv omnia_wwan_slot usb3 => saveenv => reset Signed-off-by: Pali Rohár <pali@kernel.org>
2022-03-02 11:47:58 +00:00
fixup_wwan_port_nodes(blob);
return 0;
}
#endif