[PATCH V2 0/2] Support for HTKW mcx

[PATCH V2 0/2] Support for HTKW mcx

[Ilya Yanok]

Split into two parts:
 1. Device tree only to be used with board-generic (only serial works)
 2. Full board support via custom machine file for board-testing

Ilya Yanok (2):
  mcx: very basic support for HTKW mcx board
  mcx: support for HTKW mcx board

 arch/arm/boot/dts/mcx.dts                    |   27 ++
 arch/arm/mach-omap2/Kconfig                  |    6 +
 arch/arm/mach-omap2/Makefile                 |    1 +
 arch/arm/mach-omap2/board-mcx.c              |  495 ++++++++++++++++++++++++++
 arch/arm/plat-omap/include/plat/uncompress.h |    1 +
 5 files changed, 530 insertions(+), 0 deletions(-)
 create mode 100644 arch/arm/boot/dts/mcx.dts
 create mode 100644 arch/arm/mach-omap2/board-mcx.c

-- 
1.7.6.4

[PATCH V2 1/2] mcx: very basic support for HTKW mcx board

[Ilya Yanok]

Very basic support for HTKW mcx board. Able to boot via board-generic
and ramdisk/initramfs, however most of peripherals is unsupported.
Produces tons of twl4030 related errors as this board doesn't have
twl4030 installed.

Signed-off-by: Ilya Yanok <yanok@emcraft.com>

---
Changes from V1:

 - device tree moved to the separate patch
 - iva node is disabled instead of using custom includes
 - removed bootargs entry

 arch/arm/boot/dts/mcx.dts |   27 +++++++++++++++++++++++++++
 1 files changed, 27 insertions(+), 0 deletions(-)
 create mode 100644 arch/arm/boot/dts/mcx.dts

diff --git a/arch/arm/boot/dts/mcx.dts b/arch/arm/boot/dts/mcx.dts
new file mode 100644
index 0000000..66b81bd
--- /dev/null
+++ b/arch/arm/boot/dts/mcx.dts
@@ -0,0 +1,27 @@
+/*
+ * Copyright (C) 2011 Ilya Yanok, EmCraft Systems
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+/dts-v1/;
+
+/include/ "omap3.dtsi"
+
+/ {
+	model = "HTKW mcx";
+	compatible = "htkw,mcx", "ti,omap3";
+
+	memory {
+		device_type = "memory";
+		reg = <0x80000000 0x10000000>; /* 256 MB */
+	};
+
+	/* AM35xx doesn't have IVA */
+	soc {
+		iva {
+			status = "disabled";
+		};
+	};
+};
-- 
1.7.6.4

[PATCH V2 2/2] mcx: support for HTKW mcx board

[Ilya Yanok]

Support for the HTKW mcx board (TI AM3517 based) including serial,
Ethernet, I2C, USB host, HSMMC, DSS and RTC.

Signed-off-by: Ilya Yanok <yanok@emcraft.com>

---
Requires updated mach-types file and previously posted AM35xx-EMAC
patch: http://article.gmane.org/gmane.linux.ports.arm.omap/66861

Changes from V1:
 - Kconfig option name fixed
 - Makefile entry sanitized
 - Unneeded headers removed
 - EMAC initialization moved to separate file/patch
 - Use gpio_{request,free}_{array,one} where possible
 - don't use platform data for touchscreen, we only need to pass
   irq number, do it via client.irq
 - check mcx_ts_init return value
 - Moved DEBUG_LL_OMAP3 entry to be in aplhabetical order
 - check return value of gpio_request for USB pwr pin
 - use pr_err instead of printk for error printing
 - added a fixed regulator for vdds_dsi
 - added SDcard card-detect pin

 arch/arm/mach-omap2/Kconfig                  |    6 +
 arch/arm/mach-omap2/Makefile                 |    1 +
 arch/arm/mach-omap2/board-mcx.c              |  495 ++++++++++++++++++++++++++
 arch/arm/plat-omap/include/plat/uncompress.h |    1 +
 4 files changed, 503 insertions(+), 0 deletions(-)
 create mode 100644 arch/arm/mach-omap2/board-mcx.c

diff --git a/arch/arm/mach-omap2/Kconfig b/arch/arm/mach-omap2/Kconfig
index 09ea525..535f1a1 100644
--- a/arch/arm/mach-omap2/Kconfig
+++ b/arch/arm/mach-omap2/Kconfig
@@ -230,6 +230,12 @@ config MACH_OMAP_3430SDP
 	default y
 	select OMAP_PACKAGE_CBB
 
+config MACH_MCX
+	bool "HTKW mcx (AM3517 based) board"
+	depends on ARCH_OMAP3
+	select OMAP_PACKAGE_CBB
+	select REGULATOR_FIXED_VOLTAGE
+
 config MACH_NOKIA_N800
        bool
 
diff --git a/arch/arm/mach-omap2/Makefile b/arch/arm/mach-omap2/Makefile
index e88fcfa..4204c97 100644
--- a/arch/arm/mach-omap2/Makefile
+++ b/arch/arm/mach-omap2/Makefile
@@ -232,6 +232,7 @@ obj-$(CONFIG_MACH_CRANEBOARD)		+= board-am3517crane.o
 
 obj-$(CONFIG_MACH_SBC3530)		+= board-omap3stalker.o
 obj-$(CONFIG_MACH_TI8168EVM)		+= board-ti8168evm.o
+obj-$(CONFIG_MACH_MCX)			+= board-mcx.o
 
 # Platform specific device init code
 
diff --git a/arch/arm/mach-omap2/board-mcx.c b/arch/arm/mach-omap2/board-mcx.c
new file mode 100644
index 0000000..6375fa1
--- /dev/null
+++ b/arch/arm/mach-omap2/board-mcx.c
@@ -0,0 +1,495 @@
+/*
+ * Copyright (C) 2011 Ilya Yanok, Emcraft Systems
+ *
+ * Modified from mach-omap2/board-omap3beagle.c
+ *
+ * Initial code: Syed Mohammed Khasim
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/gpio.h>
+#include <linux/mtd/partitions.h>
+#include <linux/mtd/nand.h>
+#include <linux/mmc/host.h>
+#include <linux/regulator/fixed.h>
+#include <linux/regulator/machine.h>
+
+#include <asm/mach-types.h>
+#include <asm/mach/arch.h>
+
+#include <plat/common.h>
+#include <video/omapdss.h>
+#include <video/omap-panel-generic-dpi.h>
+#include <plat/usb.h>
+
+#include "am35xx-emac.h"
+#include "mux.h"
+#include "control.h"
+#include "hsmmc.h"
+#include "common-board-devices.h"
+
+#define MCX_MDIO_FREQUENCY	(1000000)
+
+static struct mtd_partition mcx_nand_partitions[] = {
+	/* All the partition sizes are listed in terms of NAND block size */
+	{
+		.name		= "X-Loader",
+		.offset		= 0,
+		.size		= 4 * NAND_BLOCK_SIZE,
+		.mask_flags	= MTD_WRITEABLE,	/* force read-only */
+	},
+	{
+		.name		= "U-Boot",
+		.offset		= MTDPART_OFS_APPEND,	/* Offset = 0x80000 */
+		.size		= 15 * NAND_BLOCK_SIZE,
+		.mask_flags	= MTD_WRITEABLE,	/* force read-only */
+	},
+	{
+		.name		= "U-Boot Env",
+		.offset		= MTDPART_OFS_APPEND,	/* Offset = 0x260000 */
+		.size		= 1 * NAND_BLOCK_SIZE,
+	},
+	{
+		.name		= "Kernel",
+		.offset		= MTDPART_OFS_APPEND,	/* Offset = 0x280000 */
+		.size		= 32 * NAND_BLOCK_SIZE,
+	},
+	{
+		.name		= "File System",
+		.offset		= MTDPART_OFS_APPEND,	/* Offset = 0x680000 */
+		.size		= MTDPART_SIZ_FULL,
+	},
+};
+
+#define LCD_PWR_ENn		131
+#define HDMI_TRCVR_PDn		133
+#define LCD_BKLIGHT_EN		55
+#define LCD_LVL_SFHT_BUF_ENn	43
+
+static int lcd_enabled;
+static int dvi_enabled;
+
+static int mcx_panel_enable_lcd(struct omap_dss_device *dssdev)
+{
+	if (dvi_enabled) {
+		pr_err("cannot enable LCD, DVI is enabled\n");
+		return -EINVAL;
+	}
+
+	gpio_set_value(LCD_BKLIGHT_EN, 1);
+	lcd_enabled = 1;
+
+	return 0;
+}
+
+static void mcx_panel_disable_lcd(struct omap_dss_device *dssdev)
+{
+	gpio_set_value(LCD_BKLIGHT_EN, 0);
+	lcd_enabled = 0;
+}
+
+static struct panel_generic_dpi_data lcd_panel = {
+	.name			= "focaltech_etm070003dh6",
+	.platform_enable	= mcx_panel_enable_lcd,
+	.platform_disable	= mcx_panel_disable_lcd,
+};
+
+static struct omap_dss_device mcx_lcd_device = {
+	.type			= OMAP_DISPLAY_TYPE_DPI,
+	.name			= "lcd",
+	.driver_name		= "generic_dpi_panel",
+	.data			= &lcd_panel,
+	.phy.dpi.data_lines	= 24,
+};
+
+/*
+ * TV Output
+ */
+
+static int mcx_panel_enable_tv(struct omap_dss_device *dssdev)
+{
+	return 0;
+}
+
+static void mcx_panel_disable_tv(struct omap_dss_device *dssdev)
+{
+}
+
+static struct omap_dss_device mcx_tv_device = {
+	.type			= OMAP_DISPLAY_TYPE_VENC,
+	.name			= "tv",
+	.driver_name		= "venc",
+	.phy.venc.type		= OMAP_DSS_VENC_TYPE_SVIDEO,
+	.platform_enable	= mcx_panel_enable_tv,
+	.platform_disable	= mcx_panel_disable_tv,
+};
+
+/*
+ * DVI/HDMI Output
+ */
+
+static int mcx_panel_enable_dvi(struct omap_dss_device *dssdev)
+{
+	if (lcd_enabled) {
+		pr_err("cannot enable DVI, LCD is enabled\n");
+		return -EINVAL;
+	}
+	dvi_enabled = 1;
+	gpio_set_value(HDMI_TRCVR_PDn, 1);
+	return 0;
+}
+
+static void mcx_panel_disable_dvi(struct omap_dss_device *dssdev)
+{
+	dvi_enabled = 0;
+	gpio_set_value(HDMI_TRCVR_PDn, 0);
+}
+
+static struct panel_generic_dpi_data dvi_panel = {
+	.platform_enable	= mcx_panel_enable_dvi,
+	.platform_disable	= mcx_panel_disable_dvi,
+};
+static struct omap_dss_device mcx_dvi_device = {
+	.type			= OMAP_DISPLAY_TYPE_DPI,
+	.name			= "dvi",
+	.driver_name		= "dvi",
+	.data			= &dvi_panel,
+	.phy.dpi.data_lines	= 24,
+};
+
+static struct omap_dss_device *mcx_dss_devices[] = {
+	&mcx_lcd_device,
+	&mcx_tv_device,
+	&mcx_dvi_device,
+};
+
+static struct omap_dss_board_info mcx_dss_data = {
+	.num_devices	= ARRAY_SIZE(mcx_dss_devices),
+	.devices	= mcx_dss_devices,
+	.default_device	= &mcx_lcd_device,
+};
+
+/*
+ * use fake regulator for vdds_dsi as we can't find this pin inside
+ * AM3517 datasheet.
+ */
+static struct regulator_consumer_supply mcx_vdds_dsi_supply[] = {
+	REGULATOR_SUPPLY("vdds_dsi", "omapdss"),
+	REGULATOR_SUPPLY("vdds_dsi", "omapdss_dsi.0"),
+};
+
+static struct regulator_init_data mcx_vdds_dsi = {
+	.constraints		= {
+		.valid_modes_mask	= REGULATOR_MODE_NORMAL,
+		.always_on		= 1,
+	},
+	.num_consumer_supplies	= ARRAY_SIZE(mcx_vdds_dsi_supply),
+	.consumer_supplies	= mcx_vdds_dsi_supply,
+};
+
+static struct fixed_voltage_config mcx_display = {
+	.supply_name		= "display",
+	.microvolts		= 1800000,
+	.gpio			= -EINVAL,
+	.enabled_at_boot	= 1,
+	.init_data		= &mcx_vdds_dsi,
+};
+
+static struct platform_device mcx_display_device = {
+	.name		= "reg-fixed-voltage",
+	.id		= 0,
+	.dev = {
+		.platform_data	= &mcx_display,
+	},
+};
+
+static struct gpio mcx_dss_gpios[] __initdata = {
+	{ LCD_BKLIGHT_EN, GPIOF_OUT_INIT_LOW, "lcd backlight enable"	},
+	{ LCD_LVL_SFHT_BUF_ENn, GPIOF_OUT_INIT_LOW, "lcd lvl shifter"	},
+	{ LCD_PWR_ENn, GPIOF_OUT_INIT_LOW, "lcd power enable"		},
+	{ HDMI_TRCVR_PDn, GPIOF_OUT_INIT_LOW, "HDMI trcvr power"	},
+};
+
+static void __init mcx_display_init(void)
+{
+	int r;
+
+	r = gpio_request_array(mcx_dss_gpios, ARRAY_SIZE(mcx_dss_gpios));
+	if (r) {
+		pr_err("failed to get DSS control GPIOs\n");
+		return;
+	}
+
+	omap_mux_init_gpio(LCD_BKLIGHT_EN, OMAP_PIN_OUTPUT);
+	omap_mux_init_gpio(LCD_LVL_SFHT_BUF_ENn, OMAP_PIN_OUTPUT);
+	omap_mux_init_gpio(LCD_PWR_ENn, OMAP_PIN_OUTPUT);
+	omap_mux_init_gpio(HDMI_TRCVR_PDn, OMAP_PIN_OUTPUT);
+
+	r = omap_display_init(&mcx_dss_data);
+	if (r) {
+		pr_err("Failed to register DSS device\n");
+		gpio_free_array(mcx_dss_gpios, ARRAY_SIZE(mcx_dss_gpios));
+	}
+}
+
+/* TPS65023 specific initialization */
+/* VDCDC1 -> VDD_CORE */
+static struct regulator_consumer_supply am3517_vdcdc1_supplies[] = {
+	{
+		.supply = "vdd_core",
+	},
+};
+
+/* VDCDC2 -> VDDSHV */
+static struct regulator_consumer_supply am3517_vdcdc2_supplies[] = {
+	{
+		.supply = "vddshv",
+	},
+};
+
+/*
+ * VDCDC2 |-> VDDS
+ *	  |-> VDDS_SRAM_CORE_BG
+ *	  |-> VDDS_SRAM_MPU
+ */
+static struct regulator_consumer_supply am3517_vdcdc3_supplies[] = {
+	{
+		.supply = "vdds",
+	},
+	{
+		.supply = "vdds_sram_core_bg",
+	},
+	{
+		.supply = "vdds_sram_mpu",
+	},
+};
+
+/*
+ * LDO1 |-> VDDA1P8V_USBPHY
+ *	|-> VDDA_DAC
+ */
+static struct regulator_consumer_supply am3517_ldo1_supplies[] = {
+	{
+		.supply = "vdda1p8v_usbphy",
+	},
+	{
+		.supply = "vdda_dac",
+	},
+};
+
+/* LDO2 -> VDDA3P3V_USBPHY */
+static struct regulator_consumer_supply am3517_ldo2_supplies[] = {
+	{
+		.supply = "vdda3p3v_usbphy",
+	},
+};
+
+static struct regulator_init_data mcx_regulator_data[] = {
+	/* DCDC1 */
+	{
+		.constraints = {
+			.min_uV = 1200000,
+			.max_uV = 1200000,
+			.valid_modes_mask = REGULATOR_MODE_NORMAL,
+			.valid_ops_mask = REGULATOR_CHANGE_STATUS,
+			.always_on = true,
+			.apply_uV = false,
+		},
+		.num_consumer_supplies = ARRAY_SIZE(am3517_vdcdc1_supplies),
+		.consumer_supplies = am3517_vdcdc1_supplies,
+	},
+	/* DCDC2 */
+	{
+		.constraints = {
+			.min_uV = 3300000,
+			.max_uV = 3300000,
+			.valid_modes_mask = REGULATOR_MODE_NORMAL,
+			.valid_ops_mask = REGULATOR_CHANGE_STATUS,
+			.always_on = true,
+			.apply_uV = false,
+		},
+		.num_consumer_supplies = ARRAY_SIZE(am3517_vdcdc2_supplies),
+		.consumer_supplies = am3517_vdcdc2_supplies,
+	},
+	/* DCDC3 */
+	{
+		.constraints = {
+			.min_uV = 1800000,
+			.max_uV = 1800000,
+			.valid_modes_mask = REGULATOR_MODE_NORMAL,
+			.valid_ops_mask = REGULATOR_CHANGE_STATUS,
+			.always_on = true,
+			.apply_uV = false,
+		},
+		.num_consumer_supplies = ARRAY_SIZE(am3517_vdcdc3_supplies),
+		.consumer_supplies = am3517_vdcdc3_supplies,
+	},
+	/* LDO1 */
+	{
+		.constraints = {
+			.min_uV = 1800000,
+			.max_uV = 1800000,
+			.valid_modes_mask = REGULATOR_MODE_NORMAL,
+			.valid_ops_mask = REGULATOR_CHANGE_STATUS,
+			.always_on = false,
+			.apply_uV = false,
+		},
+		.num_consumer_supplies = ARRAY_SIZE(am3517_ldo1_supplies),
+		.consumer_supplies = am3517_ldo1_supplies,
+	},
+	/* LDO2 */
+	{
+		.constraints = {
+			.min_uV = 3300000,
+			.max_uV = 3300000,
+			.valid_modes_mask = REGULATOR_MODE_NORMAL,
+			.valid_ops_mask = REGULATOR_CHANGE_STATUS,
+			.always_on = false,
+			.apply_uV = false,
+		},
+		.num_consumer_supplies = ARRAY_SIZE(am3517_ldo2_supplies),
+		.consumer_supplies = am3517_ldo2_supplies,
+	},
+};
+
+static struct i2c_board_info __initdata mcx_i2c1_devices[] = {
+	{
+		I2C_BOARD_INFO("ds1337", 0x68),
+	},
+	{
+		I2C_BOARD_INFO("tps65023", 0x48),
+		.flags = I2C_CLIENT_WAKE,
+		.platform_data = &mcx_regulator_data[0],
+	},
+};
+
+#define TOUCH_INT_GPIO	170
+
+static int __init mcx_ts_init(void)
+{
+	struct i2c_board_info mcx_edt_ts[] = {
+		{
+			I2C_BOARD_INFO("edt_ts", 0x38),
+			.irq = gpio_to_irq(TOUCH_INT_GPIO),
+		},
+	};
+	int err;
+
+	err = gpio_request_one(TOUCH_INT_GPIO, GPIOF_IN, "TOUCH_INT");
+	if (err < 0) {
+		pr_err("failed to get TOUCH_INT gpio\n");
+		return -ENODEV;
+	}
+	omap_mux_init_gpio(TOUCH_INT_GPIO, OMAP_PIN_INPUT);
+
+	return i2c_register_board_info(3, mcx_edt_ts, 1);
+}
+
+static void __init mcx_i2c_init(void)
+{
+	omap_register_i2c_bus(1, 400, mcx_i2c1_devices,
+			ARRAY_SIZE(mcx_i2c1_devices));
+	omap_register_i2c_bus(2, 400, NULL, 0);
+	omap_register_i2c_bus(3, 400, NULL, 0);
+	if (mcx_ts_init() < 0)
+		pr_err("failed to register touchscreen device\n");
+}
+
+#define USB_HOST_PWR_EN		132
+#define USB_PHY1_RESET		154
+#define USB_PHY2_RESET		152
+
+static const struct usbhs_omap_board_data usbhs_bdata __initconst = {
+
+	.port_mode[0] = OMAP_EHCI_PORT_MODE_PHY,
+	.port_mode[1] = OMAP_EHCI_PORT_MODE_PHY,
+	.port_mode[2] = OMAP_USBHS_PORT_MODE_UNUSED,
+
+	.phy_reset  = true,
+	.reset_gpio_port[0]  = USB_PHY1_RESET,
+	.reset_gpio_port[1]  = USB_PHY2_RESET,
+	.reset_gpio_port[2]  = -EINVAL
+};
+
+#define SD_CARD_CD		61
+#define SD_CARD_WP		65
+
+static struct omap2_hsmmc_info mmc[] = {
+	{
+		.mmc		= 1,
+		.caps		= MMC_CAP_4_BIT_DATA,
+		.gpio_cd        = SD_CARD_CD,
+		.gpio_wp        = SD_CARD_WP,
+		.ocr_mask	= MMC_VDD_32_33 | MMC_VDD_33_34 |
+					MMC_VDD_165_195,
+	},
+	{}      /* Terminator */
+};
+
+#ifdef CONFIG_OMAP_MUX
+static struct omap_board_mux board_mux[] __initdata = {
+	OMAP3_MUX(CHASSIS_DMAREQ3, OMAP_MUX_MODE0 | OMAP_PIN_INPUT_PULLDOWN),
+	OMAP3_MUX(UART1_TX, OMAP_MUX_MODE0 | OMAP_PIN_OUTPUT |
+			OMAP_PULL_ENA | OMAP_PULL_UP),
+	OMAP3_MUX(UART1_RX, OMAP_MUX_MODE0 | OMAP_PIN_INPUT),
+	OMAP3_MUX(UART1_RTS, OMAP_MUX_MODE0 | OMAP_PIN_INPUT),
+	OMAP3_MUX(UART1_CTS, OMAP_MUX_MODE0 | OMAP_PIN_OUTPUT |
+			OMAP_PULL_ENA | OMAP_PULL_UP),
+	{ .reg_offset = OMAP_MUX_TERMINATOR },
+};
+#endif
+
+static struct platform_device *mcx_devices[] __initdata = {
+	&mcx_display_device,
+};
+
+static void __init mcx_init(void)
+{
+	int err;
+
+	omap3_mux_init(board_mux, OMAP_PACKAGE_CBB);
+	mcx_i2c_init();
+	platform_add_devices(mcx_devices, ARRAY_SIZE(mcx_devices));
+	omap_serial_init();
+
+	mcx_display_init();
+
+	/* Configure EHCI ports */
+	err = gpio_request_one(USB_HOST_PWR_EN, GPIOF_OUT_INIT_HIGH,
+			"USB_HOST_PWR_EN");
+	if (err)
+		pr_warn("Failed to request USB host power enable GPIO\n");
+	omap_mux_init_gpio(USB_HOST_PWR_EN, OMAP_PIN_OUTPUT);
+	omap_mux_init_gpio(USB_PHY1_RESET, OMAP_PIN_OUTPUT);
+	omap_mux_init_gpio(USB_PHY2_RESET, OMAP_PIN_OUTPUT);
+	usbhs_init(&usbhs_bdata);
+	omap_nand_flash_init(NAND_BUSWIDTH_16, mcx_nand_partitions,
+			     ARRAY_SIZE(mcx_nand_partitions));
+	/* Ethernet */
+	am35xx_ethernet_init(MCX_MDIO_FREQUENCY, 1);
+
+	/* MMC init */
+	omap_mux_init_gpio(SD_CARD_WP, OMAP_PIN_INPUT);
+	omap_mux_init_gpio(SD_CARD_CD, OMAP_PIN_INPUT);
+	omap2_hsmmc_init(mmc);
+}
+
+static const char *mcx_dt_match[] __initdata = {
+	"htkw,mcx",
+	NULL
+};
+
+MACHINE_START(MCX, "htkw mcx")
+	/* Maintainer: Ilya Yanok */
+	.atag_offset	= 0x100,
+	.reserve	= omap_reserve,
+	.map_io		= omap3_map_io,
+	.init_early	= am35xx_init_early,
+	.init_irq	= omap3_init_irq,
+	.init_machine	= mcx_init,
+	.timer		= &omap3_timer,
+	.dt_compat	= mcx_dt_match,
+MACHINE_END
diff --git a/arch/arm/plat-omap/include/plat/uncompress.h b/arch/arm/plat-omap/include/plat/uncompress.h
index 2d45ea3..450962d 100644
--- a/arch/arm/plat-omap/include/plat/uncompress.h
+++ b/arch/arm/plat-omap/include/plat/uncompress.h
@@ -160,6 +160,7 @@ static inline void __arch_decomp_setup(unsigned long arch_id)
 		DEBUG_LL_OMAP3(3, encore);
 		DEBUG_LL_OMAP3(3, igep0020);
 		DEBUG_LL_OMAP3(3, igep0030);
+		DEBUG_LL_OMAP3(3, mcx);
 		DEBUG_LL_OMAP3(3, nokia_rm680);
 		DEBUG_LL_OMAP3(3, nokia_rx51);
 		DEBUG_LL_OMAP3(3, omap3517evm);
-- 
1.7.6.4

Re: [PATCH V2 1/2] mcx: very basic support for HTKW mcx board

[Igor Grinberg]

On 12/15/11 02:53, Ilya Yanok wrote:
> Very basic support for HTKW mcx board. Able to boot via board-generic
> and ramdisk/initramfs, however most of peripherals is unsupported.
> Produces tons of twl4030 related errors as this board doesn't have
> twl4030 installed.
> 
> Signed-off-by: Ilya Yanok <yanok@emcraft.com>
> 
> ---
> Changes from V1:
> 
>  - device tree moved to the separate patch
>  - iva node is disabled instead of using custom includes
>  - removed bootargs entry
> 
>  arch/arm/boot/dts/mcx.dts |   27 +++++++++++++++++++++++++++
>  1 files changed, 27 insertions(+), 0 deletions(-)
>  create mode 100644 arch/arm/boot/dts/mcx.dts
> 
> diff --git a/arch/arm/boot/dts/mcx.dts b/arch/arm/boot/dts/mcx.dts
> new file mode 100644
> index 0000000..66b81bd
> --- /dev/null
> +++ b/arch/arm/boot/dts/mcx.dts
> @@ -0,0 +1,27 @@
> +/*
> + * Copyright (C) 2011 Ilya Yanok, EmCraft Systems
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License version 2 as
> + * published by the Free Software Foundation.
> + */
> +/dts-v1/;
> +
> +/include/ "omap3.dtsi"
> +
> +/ {
> +	model = "HTKW mcx";
> +	compatible = "htkw,mcx", "ti,omap3";
> +
> +	memory {
> +		device_type = "memory";
> +		reg = <0x80000000 0x10000000>; /* 256 MB */
> +	};
> +
> +	/* AM35xx doesn't have IVA */
> +	soc {
> +		iva {
> +			status = "disabled";
> +		};
> +	};

I don't get it...
Why SoCs that do not have those IP blocks should poke
their configuration inside the h/w description
(e.g. disable/enable/workaround/hack)?
This way, why don't we also disable the PCIe which this SoC does not have?
Of course, I'm exaggerating, but this just does not scale...
Soon you will have a bunch of boards disabling stuff,
that they *do not have natively*...
Why don't generic OMAP3 DT file disable the EMAC?
If we will go this way, we will find ourself fixing it later
and producing the renaming/moving "churn", won't we?

-- 
Regards,
Igor.

Re: [PATCH V2 1/2] mcx: very basic support for HTKW mcx board

[Cousson, Benoit]

Hi Igor,

On 12/15/2011 10:51 AM, Igor Grinberg wrote:
> On 12/15/11 02:53, Ilya Yanok wrote:
>> Very basic support for HTKW mcx board. Able to boot via board-generic
>> and ramdisk/initramfs, however most of peripherals is unsupported.
>> Produces tons of twl4030 related errors as this board doesn't have
>> twl4030 installed.
>>
>> Signed-off-by: Ilya Yanok<yanok@emcraft.com>
>>
>> ---
>> Changes from V1:
>>
>>   - device tree moved to the separate patch
>>   - iva node is disabled instead of using custom includes
>>   - removed bootargs entry
>>
>>   arch/arm/boot/dts/mcx.dts |   27 +++++++++++++++++++++++++++
>>   1 files changed, 27 insertions(+), 0 deletions(-)
>>   create mode 100644 arch/arm/boot/dts/mcx.dts
>>
>> diff --git a/arch/arm/boot/dts/mcx.dts b/arch/arm/boot/dts/mcx.dts
>> new file mode 100644
>> index 0000000..66b81bd
>> --- /dev/null
>> +++ b/arch/arm/boot/dts/mcx.dts
>> @@ -0,0 +1,27 @@
>> +/*
>> + * Copyright (C) 2011 Ilya Yanok, EmCraft Systems
>> + *
>> + * This program is free software; you can redistribute it and/or modify
>> + * it under the terms of the GNU General Public License version 2 as
>> + * published by the Free Software Foundation.
>> + */
>> +/dts-v1/;
>> +
>> +/include/ "omap3.dtsi"
>> +
>> +/ {
>> +	model = "HTKW mcx";
>> +	compatible = "htkw,mcx", "ti,omap3";
>> +
>> +	memory {
>> +		device_type = "memory";
>> +		reg =<0x80000000 0x10000000>; /* 256 MB */
>> +	};
>> +
>> +	/* AM35xx doesn't have IVA */
>> +	soc {
>> +		iva {
>> +			status = "disabled";
>> +		};
>> +	};
>
> I don't get it...
> Why SoCs that do not have those IP blocks should poke
> their configuration inside the h/w description
> (e.g. disable/enable/workaround/hack)?

This is indeed the proper way assuming the HW does contain this 
information. I do not know for this OMAP variant, but this kind of 
information is not necessarily well exposed inside the HW.

> This way, why don't we also disable the PCIe which this SoC does not have?
> Of course, I'm exaggerating, but this just does not scale...
> Soon you will have a bunch of boards disabling stuff,
> that they *do not have natively*...
> Why don't generic OMAP3 DT file disable the EMAC?
> If we will go this way, we will find ourself fixing it later
> and producing the renaming/moving "churn", won't we?

You are indeed exaggerating :-)

Assuming that device is an OMAP3 variant, it seems OK to me to define it 
like that. am35xx = omap3 + (new IPs) - (IPs not supported)

The good point with DT is that you can add or *remove* things from an 
included file.

Regards,
Benoit

Re: [PATCH V2 2/2] mcx: support for HTKW mcx board

[Igor Grinberg]

Hi Ilya,

On 12/15/11 02:53, Ilya Yanok wrote:
> Support for the HTKW mcx board (TI AM3517 based) including serial,
> Ethernet, I2C, USB host, HSMMC, DSS and RTC.
> 
> Signed-off-by: Ilya Yanok <yanok@emcraft.com>
> 
> ---
> Requires updated mach-types file and previously posted AM35xx-EMAC
> patch: http://article.gmane.org/gmane.linux.ports.arm.omap/66861
> 
> Changes from V1:
>  - Kconfig option name fixed
>  - Makefile entry sanitized
>  - Unneeded headers removed
>  - EMAC initialization moved to separate file/patch
>  - Use gpio_{request,free}_{array,one} where possible
>  - don't use platform data for touchscreen, we only need to pass
>    irq number, do it via client.irq
>  - check mcx_ts_init return value
>  - Moved DEBUG_LL_OMAP3 entry to be in aplhabetical order
>  - check return value of gpio_request for USB pwr pin
>  - use pr_err instead of printk for error printing
>  - added a fixed regulator for vdds_dsi
>  - added SDcard card-detect pin
> 
>  arch/arm/mach-omap2/Kconfig                  |    6 +
>  arch/arm/mach-omap2/Makefile                 |    1 +
>  arch/arm/mach-omap2/board-mcx.c              |  495 ++++++++++++++++++++++++++
>  arch/arm/plat-omap/include/plat/uncompress.h |    1 +
>  4 files changed, 503 insertions(+), 0 deletions(-)
>  create mode 100644 arch/arm/mach-omap2/board-mcx.c

[...]

> diff --git a/arch/arm/mach-omap2/board-mcx.c b/arch/arm/mach-omap2/board-mcx.c
> new file mode 100644
> index 0000000..6375fa1
> --- /dev/null
> +++ b/arch/arm/mach-omap2/board-mcx.c

[...]

> +static void __init mcx_display_init(void)
> +{
> +	int r;
> +
> +	r = gpio_request_array(mcx_dss_gpios, ARRAY_SIZE(mcx_dss_gpios));
> +	if (r) {
> +		pr_err("failed to get DSS control GPIOs\n");
> +		return;
> +	}
> +
> +	omap_mux_init_gpio(LCD_BKLIGHT_EN, OMAP_PIN_OUTPUT);
> +	omap_mux_init_gpio(LCD_LVL_SFHT_BUF_ENn, OMAP_PIN_OUTPUT);
> +	omap_mux_init_gpio(LCD_PWR_ENn, OMAP_PIN_OUTPUT);
> +	omap_mux_init_gpio(HDMI_TRCVR_PDn, OMAP_PIN_OUTPUT);

Shouldn't you mux the pins, before you access the GPIO
(e.g. before the gpio_request_array()).
Are there any safety problems?

> +
> +	r = omap_display_init(&mcx_dss_data);
> +	if (r) {
> +		pr_err("Failed to register DSS device\n");
> +		gpio_free_array(mcx_dss_gpios, ARRAY_SIZE(mcx_dss_gpios));
> +	}
> +}

[...]

> +#define TOUCH_INT_GPIO	170
> +
> +static int __init mcx_ts_init(void)
> +{
> +	struct i2c_board_info mcx_edt_ts[] = {
> +		{
> +			I2C_BOARD_INFO("edt_ts", 0x38),
> +			.irq = gpio_to_irq(TOUCH_INT_GPIO),
> +		},
> +	};
> +	int err;
> +
> +	err = gpio_request_one(TOUCH_INT_GPIO, GPIOF_IN, "TOUCH_INT");
> +	if (err < 0) {
> +		pr_err("failed to get TOUCH_INT gpio\n");
> +		return -ENODEV;
> +	}
> +	omap_mux_init_gpio(TOUCH_INT_GPIO, OMAP_PIN_INPUT);

Same here...

> +
> +	return i2c_register_board_info(3, mcx_edt_ts, 1);
> +}

[...]

> +#define USB_HOST_PWR_EN		132
> +#define USB_PHY1_RESET		154
> +#define USB_PHY2_RESET		152
> +
> +static const struct usbhs_omap_board_data usbhs_bdata __initconst = {
> +
> +	.port_mode[0] = OMAP_EHCI_PORT_MODE_PHY,
> +	.port_mode[1] = OMAP_EHCI_PORT_MODE_PHY,
> +	.port_mode[2] = OMAP_USBHS_PORT_MODE_UNUSED,
> +
> +	.phy_reset  = true,
> +	.reset_gpio_port[0]  = USB_PHY1_RESET,
> +	.reset_gpio_port[1]  = USB_PHY2_RESET,
> +	.reset_gpio_port[2]  = -EINVAL
> +};
> +
> +#define SD_CARD_CD		61
> +#define SD_CARD_WP		65
> +
> +static struct omap2_hsmmc_info mmc[] = {
> +	{
> +		.mmc		= 1,
> +		.caps		= MMC_CAP_4_BIT_DATA,
> +		.gpio_cd        = SD_CARD_CD,
> +		.gpio_wp        = SD_CARD_WP,
> +		.ocr_mask	= MMC_VDD_32_33 | MMC_VDD_33_34 |
> +					MMC_VDD_165_195,

The ocr_mask will be overridden, by the following patch:
-----------------
commit e89715a7e48d505f42813a4e3ee0f0efb49832ba
Author: Abhilash K V <abhilash.kv@ti.com>
Date:   Fri Dec 9 12:27:36 2011 -0800

    ARM: OMAP: hsmmc: Support for AM3517 MMC1 voltages
--------------

in Tony's hsmmc branch.

IMO it should be fixed, by adding a check if the ocr_mask is
already set...
I can't send a patch for this right now...

[...]

> +	},
> +	{}      /* Terminator */
> +};
> +
> +#ifdef CONFIG_OMAP_MUX
> +static struct omap_board_mux board_mux[] __initdata = {
> +	OMAP3_MUX(CHASSIS_DMAREQ3, OMAP_MUX_MODE0 | OMAP_PIN_INPUT_PULLDOWN),
> +	OMAP3_MUX(UART1_TX, OMAP_MUX_MODE0 | OMAP_PIN_OUTPUT |
> +			OMAP_PULL_ENA | OMAP_PULL_UP),
> +	OMAP3_MUX(UART1_RX, OMAP_MUX_MODE0 | OMAP_PIN_INPUT),
> +	OMAP3_MUX(UART1_RTS, OMAP_MUX_MODE0 | OMAP_PIN_INPUT),
> +	OMAP3_MUX(UART1_CTS, OMAP_MUX_MODE0 | OMAP_PIN_OUTPUT |
> +			OMAP_PULL_ENA | OMAP_PULL_UP),

Hmm... pullup for output? Is this needed for kind of safety?

[...]

> +static void __init mcx_init(void)
> +{
> +	int err;
> +
> +	omap3_mux_init(board_mux, OMAP_PACKAGE_CBB);
> +	mcx_i2c_init();
> +	platform_add_devices(mcx_devices, ARRAY_SIZE(mcx_devices));
> +	omap_serial_init();

Shouldn't this one be before the mcx_i2c_init() call?

> +
> +	mcx_display_init();
> +
> +	/* Configure EHCI ports */
> +	err = gpio_request_one(USB_HOST_PWR_EN, GPIOF_OUT_INIT_HIGH,
> +			"USB_HOST_PWR_EN");
> +	if (err)
> +		pr_warn("Failed to request USB host power enable GPIO\n");

empty line here will improve the readability.

> +	omap_mux_init_gpio(USB_HOST_PWR_EN, OMAP_PIN_OUTPUT);
> +	omap_mux_init_gpio(USB_PHY1_RESET, OMAP_PIN_OUTPUT);
> +	omap_mux_init_gpio(USB_PHY2_RESET, OMAP_PIN_OUTPUT);

once again mux after gpio_request?

> +	usbhs_init(&usbhs_bdata);
> +	omap_nand_flash_init(NAND_BUSWIDTH_16, mcx_nand_partitions,
> +			     ARRAY_SIZE(mcx_nand_partitions));
> +	/* Ethernet */
> +	am35xx_ethernet_init(MCX_MDIO_FREQUENCY, 1);
> +
> +	/* MMC init */
> +	omap_mux_init_gpio(SD_CARD_WP, OMAP_PIN_INPUT);
> +	omap_mux_init_gpio(SD_CARD_CD, OMAP_PIN_INPUT);
> +	omap2_hsmmc_init(mmc);
> +}

[...]


-- 
Regards,
Igor.

Re: [PATCH V2 1/2] mcx: very basic support for HTKW mcx board

[Igor Grinberg]

On 12/15/11 12:17, Cousson, Benoit wrote:
> Hi Igor,
> 
> On 12/15/2011 10:51 AM, Igor Grinberg wrote:
>> On 12/15/11 02:53, Ilya Yanok wrote:
>>> Very basic support for HTKW mcx board. Able to boot via board-generic
>>> and ramdisk/initramfs, however most of peripherals is unsupported.
>>> Produces tons of twl4030 related errors as this board doesn't have
>>> twl4030 installed.
>>>
>>> Signed-off-by: Ilya Yanok<yanok@emcraft.com>
>>>
>>> ---
>>> Changes from V1:
>>>
>>>   - device tree moved to the separate patch
>>>   - iva node is disabled instead of using custom includes
>>>   - removed bootargs entry
>>>
>>>   arch/arm/boot/dts/mcx.dts |   27 +++++++++++++++++++++++++++
>>>   1 files changed, 27 insertions(+), 0 deletions(-)
>>>   create mode 100644 arch/arm/boot/dts/mcx.dts
>>>
>>> diff --git a/arch/arm/boot/dts/mcx.dts b/arch/arm/boot/dts/mcx.dts
>>> new file mode 100644
>>> index 0000000..66b81bd
>>> --- /dev/null
>>> +++ b/arch/arm/boot/dts/mcx.dts
>>> @@ -0,0 +1,27 @@
>>> +/*
>>> + * Copyright (C) 2011 Ilya Yanok, EmCraft Systems
>>> + *
>>> + * This program is free software; you can redistribute it and/or modify
>>> + * it under the terms of the GNU General Public License version 2 as
>>> + * published by the Free Software Foundation.
>>> + */
>>> +/dts-v1/;
>>> +
>>> +/include/ "omap3.dtsi"
>>> +
>>> +/ {
>>> +    model = "HTKW mcx";
>>> +    compatible = "htkw,mcx", "ti,omap3";
>>> +
>>> +    memory {
>>> +        device_type = "memory";
>>> +        reg =<0x80000000 0x10000000>; /* 256 MB */
>>> +    };
>>> +
>>> +    /* AM35xx doesn't have IVA */
>>> +    soc {
>>> +        iva {
>>> +            status = "disabled";
>>> +        };
>>> +    };
>>
>> I don't get it...
>> Why SoCs that do not have those IP blocks should poke
>> their configuration inside the h/w description
>> (e.g. disable/enable/workaround/hack)?
> 
> This is indeed the proper way assuming the HW does contain this information. I do not know for this OMAP variant, but this kind of information is not necessarily well exposed inside the HW.
> 
>> This way, why don't we also disable the PCIe which this SoC does not have?
>> Of course, I'm exaggerating, but this just does not scale...
>> Soon you will have a bunch of boards disabling stuff,
>> that they *do not have natively*...
>> Why don't generic OMAP3 DT file disable the EMAC?
>> If we will go this way, we will find ourself fixing it later
>> and producing the renaming/moving "churn", won't we?
> 
> You are indeed exaggerating :-)
> 
> Assuming that device is an OMAP3 variant, it seems OK to me to define it like that. am35xx = omap3 + (new IPs) - (IPs not supported)

I still haven't seen any real cons about my proposed solution:
Have include file for each IP, for example IP1/2/3/4...
omap3 = (include IPs supported: IP1, IP3, IP4)
am35x = (include IPs supported: IP1, IP2, IP4)

This way, we will be able to easily support any mixture of those whenever
TI releases new version of AM/DM3/4/5/6xxx - just include those you have.

Also, OMAP3 is not a variant of OMAP2, but has several same IPs inside
(or am I wrong?). Same probably will stand for OMAP4/5/6...
In my proposed way we can share those IPs without duplicating them.

> 
> The good point with DT is that you can add or *remove* things from an included file.

Yes, but I don't think board DT file should do it...
It contradicts, the DT=="hardware description" statement.

I also think that it is good thing to try and find a common set of IPs,
but it can be done by including those and still stay scalable...


-- 
Regards,
Igor.

Re: [PATCH V2 1/2] mcx: very basic support for HTKW mcx board

[Ilya Yanok]

Hi Igor,

On 15.12.2011 13:51, Igor Grinberg wrote:
>> +	/* AM35xx doesn't have IVA */
>> +	soc {
>> +		iva {
>> +			status = "disabled";
>> +		};
>> +	};
> 
> I don't get it...
> Why SoCs that do not have those IP blocks should poke
> their configuration inside the h/w description
> (e.g. disable/enable/workaround/hack)?
> This way, why don't we also disable the PCIe which this SoC does not have?

Well, I'm not really happy about this solution... But when we discussed
it last time (http://thread.gmane.org/gmane.linux.ports.arm.omap/66601 )
you and Benoit didn't come to agreement...
I really like your solution with dts built from brick corresponding to
IPs but I don't have much time to invest into patches that will likely
be rejected...

Regards, Ilya.

Re: [PATCH V2 1/2] mcx: very basic support for HTKW mcx board

[Tony Lindgren]

* Ilya Yanok <yanok@emcraft.com> [111214 16:21]:
> Very basic support for HTKW mcx board. Able to boot via board-generic
> and ramdisk/initramfs, however most of peripherals is unsupported.
> Produces tons of twl4030 related errors as this board doesn't have
> twl4030 installed.

Please resend with linux-arm-kernel and devicetree-discuss
lists in Cc. We need an ack from the devicetree people for this.

Tony

Re: [PATCH V2 1/2] mcx: very basic support for HTKW mcx board

[Igor Grinberg]

Forgot to Cc devicetree-discuss - done now.

On 12/15/11 12:53, Igor Grinberg wrote:
> On 12/15/11 12:17, Cousson, Benoit wrote:
>> Hi Igor,
>>
>> On 12/15/2011 10:51 AM, Igor Grinberg wrote:
>>> On 12/15/11 02:53, Ilya Yanok wrote:
>>>> Very basic support for HTKW mcx board. Able to boot via board-generic
>>>> and ramdisk/initramfs, however most of peripherals is unsupported.
>>>> Produces tons of twl4030 related errors as this board doesn't have
>>>> twl4030 installed.
>>>>
>>>> Signed-off-by: Ilya Yanok<yanok@emcraft.com>
>>>>
>>>> ---
>>>> Changes from V1:
>>>>
>>>>   - device tree moved to the separate patch
>>>>   - iva node is disabled instead of using custom includes
>>>>   - removed bootargs entry
>>>>
>>>>   arch/arm/boot/dts/mcx.dts |   27 +++++++++++++++++++++++++++
>>>>   1 files changed, 27 insertions(+), 0 deletions(-)
>>>>   create mode 100644 arch/arm/boot/dts/mcx.dts
>>>>
>>>> diff --git a/arch/arm/boot/dts/mcx.dts b/arch/arm/boot/dts/mcx.dts
>>>> new file mode 100644
>>>> index 0000000..66b81bd
>>>> --- /dev/null
>>>> +++ b/arch/arm/boot/dts/mcx.dts
>>>> @@ -0,0 +1,27 @@
>>>> +/*
>>>> + * Copyright (C) 2011 Ilya Yanok, EmCraft Systems
>>>> + *
>>>> + * This program is free software; you can redistribute it and/or modify
>>>> + * it under the terms of the GNU General Public License version 2 as
>>>> + * published by the Free Software Foundation.
>>>> + */
>>>> +/dts-v1/;
>>>> +
>>>> +/include/ "omap3.dtsi"
>>>> +
>>>> +/ {
>>>> +    model = "HTKW mcx";
>>>> +    compatible = "htkw,mcx", "ti,omap3";
>>>> +
>>>> +    memory {
>>>> +        device_type = "memory";
>>>> +        reg =<0x80000000 0x10000000>; /* 256 MB */
>>>> +    };
>>>> +
>>>> +    /* AM35xx doesn't have IVA */
>>>> +    soc {
>>>> +        iva {
>>>> +            status = "disabled";
>>>> +        };
>>>> +    };
>>>
>>> I don't get it...
>>> Why SoCs that do not have those IP blocks should poke
>>> their configuration inside the h/w description
>>> (e.g. disable/enable/workaround/hack)?
>>
>> This is indeed the proper way assuming the HW does contain this information. I do not know for this OMAP variant, but this kind of information is not necessarily well exposed inside the HW.
>>
>>> This way, why don't we also disable the PCIe which this SoC does not have?
>>> Of course, I'm exaggerating, but this just does not scale...
>>> Soon you will have a bunch of boards disabling stuff,
>>> that they *do not have natively*...
>>> Why don't generic OMAP3 DT file disable the EMAC?
>>> If we will go this way, we will find ourself fixing it later
>>> and producing the renaming/moving "churn", won't we?
>>
>> You are indeed exaggerating :-)
>>
>> Assuming that device is an OMAP3 variant, it seems OK to me to define it like that. am35xx = omap3 + (new IPs) - (IPs not supported)
> 
> I still haven't seen any real cons about my proposed solution:
> Have include file for each IP, for example IP1/2/3/4...
> omap3 = (include IPs supported: IP1, IP3, IP4)
> am35x = (include IPs supported: IP1, IP2, IP4)
> 
> This way, we will be able to easily support any mixture of those whenever
> TI releases new version of AM/DM3/4/5/6xxx - just include those you have.
> 
> Also, OMAP3 is not a variant of OMAP2, but has several same IPs inside
> (or am I wrong?). Same probably will stand for OMAP4/5/6...
> In my proposed way we can share those IPs without duplicating them.
> 
>>
>> The good point with DT is that you can add or *remove* things from an included file.
> 
> Yes, but I don't think board DT file should do it...
> It contradicts, the DT=="hardware description" statement.
> 
> I also think that it is good thing to try and find a common set of IPs,
> but it can be done by including those and still stay scalable...


-- 
Regards,
Igor.

Re: [PATCH V2 2/2] mcx: support for HTKW mcx board

[Ilya Yanok]

Hi Igor,

On 15.12.2011 14:40, Igor Grinberg wrote:
>> +	r = gpio_request_array(mcx_dss_gpios, ARRAY_SIZE(mcx_dss_gpios));
>> +	if (r) {
>> +		pr_err("failed to get DSS control GPIOs\n");
>> +		return;
>> +	}
>> +
>> +	omap_mux_init_gpio(LCD_BKLIGHT_EN, OMAP_PIN_OUTPUT);
>> +	omap_mux_init_gpio(LCD_LVL_SFHT_BUF_ENn, OMAP_PIN_OUTPUT);
>> +	omap_mux_init_gpio(LCD_PWR_ENn, OMAP_PIN_OUTPUT);
>> +	omap_mux_init_gpio(HDMI_TRCVR_PDn, OMAP_PIN_OUTPUT);
> 
> Shouldn't you mux the pins, before you access the GPIO
> (e.g. before the gpio_request_array()).
> Are there any safety problems?

No, there are no problems. I will move mux settings above
gpio_request_array() call.

>> +static struct omap2_hsmmc_info mmc[] = {
>> +	{
>> +		.mmc		= 1,
>> +		.caps		= MMC_CAP_4_BIT_DATA,
>> +		.gpio_cd        = SD_CARD_CD,
>> +		.gpio_wp        = SD_CARD_WP,
>> +		.ocr_mask	= MMC_VDD_32_33 | MMC_VDD_33_34 |
>> +					MMC_VDD_165_195,
> 
> The ocr_mask will be overridden, by the following patch:
> -----------------
> commit e89715a7e48d505f42813a4e3ee0f0efb49832ba
> Author: Abhilash K V <abhilash.kv@ti.com>
> Date:   Fri Dec 9 12:27:36 2011 -0800
> 
>     ARM: OMAP: hsmmc: Support for AM3517 MMC1 voltages
> --------------
> 
> in Tony's hsmmc branch.
> 
> IMO it should be fixed, by adding a check if the ocr_mask is
> already set...
> I can't send a patch for this right now...

Well, I think I should just drop the .ocr_mask field then. Everything
works fine for me with the above mentioned patch.

>> +#ifdef CONFIG_OMAP_MUX
>> +static struct omap_board_mux board_mux[] __initdata = {
>> +	OMAP3_MUX(CHASSIS_DMAREQ3, OMAP_MUX_MODE0 | OMAP_PIN_INPUT_PULLDOWN),
>> +	OMAP3_MUX(UART1_TX, OMAP_MUX_MODE0 | OMAP_PIN_OUTPUT |
>> +			OMAP_PULL_ENA | OMAP_PULL_UP),
>> +	OMAP3_MUX(UART1_RX, OMAP_MUX_MODE0 | OMAP_PIN_INPUT),
>> +	OMAP3_MUX(UART1_RTS, OMAP_MUX_MODE0 | OMAP_PIN_INPUT),
>> +	OMAP3_MUX(UART1_CTS, OMAP_MUX_MODE0 | OMAP_PIN_OUTPUT |
>> +			OMAP_PULL_ENA | OMAP_PULL_UP),
> 
> Hmm... pullup for output? Is this needed for kind of safety?

This is a mistake indeed. Will remove the pullups.

>> +static void __init mcx_init(void)
>> +{
>> +	int err;
>> +
>> +	omap3_mux_init(board_mux, OMAP_PACKAGE_CBB);
>> +	mcx_i2c_init();
>> +	platform_add_devices(mcx_devices, ARRAY_SIZE(mcx_devices));
>> +	omap_serial_init();
> 
> Shouldn't this one be before the mcx_i2c_init() call?

Well, I think I've taken this order from some other board init... I
think the idea was to bring up regulator chip earlier. But I can move
serial up with no problem.

>> +	mcx_display_init();
>> +
>> +	/* Configure EHCI ports */
>> +	err = gpio_request_one(USB_HOST_PWR_EN, GPIOF_OUT_INIT_HIGH,
>> +			"USB_HOST_PWR_EN");
>> +	if (err)
>> +		pr_warn("Failed to request USB host power enable GPIO\n");
> 
> empty line here will improve the readability.

Ok.

Regards, Ilya.

Re: [PATCH V2 2/2] mcx: support for HTKW mcx board

[Igor Grinberg]

Hi Ilya,

On 12/20/11 22:07, Ilya Yanok wrote:
> Hi Igor,
> 
>>> +static struct omap2_hsmmc_info mmc[] = {
>>> +	{
>>> +		.mmc		= 1,
>>> +		.caps		= MMC_CAP_4_BIT_DATA,
>>> +		.gpio_cd        = SD_CARD_CD,
>>> +		.gpio_wp        = SD_CARD_WP,
>>> +		.ocr_mask	= MMC_VDD_32_33 | MMC_VDD_33_34 |
>>> +					MMC_VDD_165_195,
>>
>> The ocr_mask will be overridden, by the following patch:
>> -----------------
>> commit e89715a7e48d505f42813a4e3ee0f0efb49832ba
>> Author: Abhilash K V <abhilash.kv@ti.com>
>> Date:   Fri Dec 9 12:27:36 2011 -0800
>>
>>     ARM: OMAP: hsmmc: Support for AM3517 MMC1 voltages
>> --------------
>>
>> in Tony's hsmmc branch.
>>
>> IMO it should be fixed, by adding a check if the ocr_mask is
>> already set...
>> I can't send a patch for this right now...
> 
> Well, I think I should just drop the .ocr_mask field then. Everything
> works fine for me with the above mentioned patch.

Yes, for your patch, it is correct - just remove it.

My concern is for the common code - it will override any
board special setting (e.g. you want smaller range supported).

>>> +static void __init mcx_init(void)
>>> +{
>>> +	int err;
>>> +
>>> +	omap3_mux_init(board_mux, OMAP_PACKAGE_CBB);
>>> +	mcx_i2c_init();
>>> +	platform_add_devices(mcx_devices, ARRAY_SIZE(mcx_devices));
>>> +	omap_serial_init();
>>
>> Shouldn't this one be before the mcx_i2c_init() call?
> 
> Well, I think I've taken this order from some other board init... I
> think the idea was to bring up regulator chip earlier. But I can move
> serial up with no problem.

I see, no problem.
My "concern" was about the pr_err() inside the mcx_i2c_init()
and the mcx_ts_init() functions.

Also, IMO, mcx_ts_init() can be done at a later stage, than along
with I2C buses initialization, but it does not really meter...


-- 
Regards,
Igor.

Re: [PATCH V2 1/2] mcx: very basic support for HTKW mcx board

[Ilya Yanok]

Hi Tony,

as you probably seen we've got an ACK from Grant Likely for the
mt_ventoux DT-only support patch (which is really the same as this one,
only board name/vendor changed). Probably you can now merge support for
both boards?

Regards, Ilya.

On 17.12.2011 05:37, Tony Lindgren wrote:
> * Ilya Yanok <yanok@emcraft.com> [111214 16:21]:
>> Very basic support for HTKW mcx board. Able to boot via board-generic
>> and ramdisk/initramfs, however most of peripherals is unsupported.
>> Produces tons of twl4030 related errors as this board doesn't have
>> twl4030 installed.
> 
> Please resend with linux-arm-kernel and devicetree-discuss
> lists in Cc. We need an ack from the devicetree people for this.
> 
> Tony