[U-Boot] [PATCH v5 0/9] Add support for DM_USB for TI's DRA7 EVMs and AM57 EVMs platforms

[U-Boot] [PATCH v5 0/9] Add support for DM_USB for TI's DRA7 EVMs and AM57 EVMs platforms

[Jean-Jacques Hiblot]

Supporting DM USB is required to support DM_ETH and USB network adapters
with the same binary.
This series adds support for DM_USB for the DRA7/AM57 families. As a bonus
it adds support for USB3.

limitation:
- only Host mode is supported. The Device mode can be added later.

Tested on:
- dra7 evm
- dra71 evm
- dra72 evm
- dra72 evm rev C
- dra76 evm
- AM572 evm

Changes in v5:
- Rebased on latest u-boot
- Removed patches that have already been merged

Changes in v4:
- Fix word missing in commit log
- Remove dependency on MISC
- Renamed the commit (DM_USB is already enabled, the only missing option
  is for the USB2 phy driver)
- Remove dependency on MISC
- Remove omap specific glue layer and use dwc3-of-simple glue layer

Changes in v3:
- in syscon_regmap_lookup_by_phandle(), use dev_dbg() instead of printf()
- added unit test for syscon_regmap_lookup_by_phandle()

Changes in v2:
- Add USB3 support to ti-pipe3-phy driver
- omap-usb2-phy: Implement power_on and power_off callbacks
- am57xx boards: when DM_USB is used, turn on the required USB clocks
- split dts changes in 2 commits: one for binding the children of
  ocp2scp at 4a080000, and one to disable USB1  on all DRA7 EVMs
- Instead of disabling USB1 port in dts files, use *-u-boot.dtsi files
- Enable DM_USB in am57xx_evm_defconfig. USB3 (super speed) is supported.

Jean-Jacques Hiblot (7):
  syscon: dm: Add a new method to get a regmap from DTS
  phy: Add a new driver for OMAP's USB2 PHYs
  board: ti: dra7xx-evm: turn on USB clocks in late init stage
  dts: dra7x: make ocp2scp at 4a080000 compatible with simple-bus
  dts: dra7x: Disable USB1 on all evms
  configs: enable OMAP_USB2_PHY for all DRA7 platforms
  configs: am57xx_evm: Enable DM_USB and dependencies

Vignesh R (2):
  phy: ti-pip3-phy: Add support for USB3 PHY
  board; ti: am57xx: turn on USB clocks

 arch/arm/dts/dra7-evm-u-boot.dtsi       |   8 ++
 arch/arm/dts/dra71-evm-u-boot.dtsi      |   9 ++
 arch/arm/dts/dra72-evm-revc-u-boot.dtsi |   8 ++
 arch/arm/dts/dra72-evm-u-boot.dtsi      |  22 ++++
 arch/arm/dts/dra76-evm-u-boot.dtsi      |   9 ++
 arch/arm/dts/omap5-u-boot.dtsi          |   4 +
 arch/sandbox/dts/test.dts               |   6 +-
 board/ti/am57xx/board.c                 |  19 ++++
 board/ti/dra7xx/evm.c                   |  19 ++++
 configs/am57xx_evm_defconfig            |   4 +
 configs/dra7xx_evm_defconfig            |   1 +
 configs/dra7xx_hs_evm_defconfig         |   1 +
 drivers/core/syscon-uclass.c            |  23 ++++
 drivers/phy/Kconfig                     |   8 ++
 drivers/phy/Makefile                    |   1 +
 drivers/phy/omap-usb2-phy.c             | 196 ++++++++++++++++++++++++++++++++
 drivers/phy/ti-pipe3-phy.c              |  30 +++--
 include/syscon.h                        |  13 +++
 test/dm/syscon.c                        |  29 +++++
 19 files changed, 401 insertions(+), 9 deletions(-)
 create mode 100644 arch/arm/dts/dra72-evm-u-boot.dtsi
 create mode 100644 drivers/phy/omap-usb2-phy.c

-- 
2.7.4

[U-Boot] [PATCH v5 1/9] syscon: dm: Add a new method to get a regmap from DTS

[Jean-Jacques Hiblot]

syscon_regmap_lookup_by_phandle() can be used to get the regmap of a syscon
device from a reference in the DTS. It operates similarly to the linux
version of the namesake function.

Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>

---

Changes in v5: None
Changes in v4:
- Fix word missing in commit log

Changes in v3:
- in syscon_regmap_lookup_by_phandle(), use dev_dbg() instead of printf()
- added unit test for syscon_regmap_lookup_by_phandle()

Changes in v2: None

 arch/sandbox/dts/test.dts    |  6 ++++--
 drivers/core/syscon-uclass.c | 23 +++++++++++++++++++++++
 include/syscon.h             | 13 +++++++++++++
 test/dm/syscon.c             | 29 +++++++++++++++++++++++++++++
 4 files changed, 69 insertions(+), 2 deletions(-)

diff --git a/arch/sandbox/dts/test.dts b/arch/sandbox/dts/test.dts
index 06d0e8c..369002e 100644
--- a/arch/sandbox/dts/test.dts
+++ b/arch/sandbox/dts/test.dts
@@ -311,6 +311,8 @@
 
 		test4 {
 			compatible = "denx,u-boot-probe-test";
+			first-syscon = <&syscon0>;
+			second-sys-ctrl = <&another_system_controller>;
 		};
 	};
 
@@ -380,12 +382,12 @@
 		};
 	};
 
-	syscon at 0 {
+	syscon0: syscon at 0 {
 		compatible = "sandbox,syscon0";
 		reg = <0x10 4>;
 	};
 
-	syscon at 1 {
+	another_system_controller: syscon at 1 {
 		compatible = "sandbox,syscon1";
 		reg = <0x20 5
 			0x28 6
diff --git a/drivers/core/syscon-uclass.c b/drivers/core/syscon-uclass.c
index a69937e..739fddd 100644
--- a/drivers/core/syscon-uclass.c
+++ b/drivers/core/syscon-uclass.c
@@ -45,6 +45,29 @@ static int syscon_pre_probe(struct udevice *dev)
 #endif
 }
 
+struct regmap *syscon_regmap_lookup_by_phandle(struct udevice *dev,
+					       const char *name)
+{
+	struct udevice *syscon;
+	struct regmap *r;
+	int err;
+
+	err = uclass_get_device_by_phandle(UCLASS_SYSCON, dev,
+					   name, &syscon);
+	if (err) {
+		dev_dbg(dev, "unable to find syscon device\n");
+		return ERR_PTR(err);
+	}
+
+	r = syscon_get_regmap(syscon);
+	if (!r) {
+		dev_dbg(dev, "unable to find regmap\n");
+		return ERR_PTR(-ENODEV);
+	}
+
+	return r;
+}
+
 int syscon_get_by_driver_data(ulong driver_data, struct udevice **devp)
 {
 	struct udevice *dev;
diff --git a/include/syscon.h b/include/syscon.h
index 5d52b1c..23d257a 100644
--- a/include/syscon.h
+++ b/include/syscon.h
@@ -74,6 +74,19 @@ int syscon_get_by_driver_data(ulong driver_data, struct udevice **devp);
 struct regmap *syscon_get_regmap_by_driver_data(ulong driver_data);
 
 /**
+ * syscon_regmap_lookup_by_phandle() - Look up a controller by a phandle
+ *
+ * This operates by looking up the given name in the device (device
+ * tree property) of the device using the system controller.
+ *
+ * @dev:	Device using the system controller
+ * @name:	Name of property referring to the system controller
+ * @return	A pointer to the regmap if found, ERR_PTR(-ve) on error
+ */
+struct regmap *syscon_regmap_lookup_by_phandle(struct udevice *dev,
+					       const char *name);
+
+/**
  * syscon_get_first_range() - get the first memory range from a syscon regmap
  *
  * @driver_data:	Driver data value to look up
diff --git a/test/dm/syscon.c b/test/dm/syscon.c
index 99bff96..6f15baf 100644
--- a/test/dm/syscon.c
+++ b/test/dm/syscon.c
@@ -7,6 +7,7 @@
 #include <common.h>
 #include <dm.h>
 #include <syscon.h>
+#include <regmap.h>
 #include <asm/test.h>
 #include <dm/test.h>
 #include <test/ut.h>
@@ -44,3 +45,31 @@ static int dm_test_syscon_by_driver_data(struct unit_test_state *uts)
 	return 0;
 }
 DM_TEST(dm_test_syscon_by_driver_data, DM_TESTF_SCAN_PDATA | DM_TESTF_SCAN_FDT);
+
+/* Test system controller by phandle */
+static int dm_test_syscon_by_phandle(struct unit_test_state *uts)
+{
+	struct udevice *dev;
+	struct regmap *map;
+
+	ut_assertok(uclass_get_device_by_name(UCLASS_TEST_PROBE, "test4",
+					      &dev));
+	if (!dev || IS_ERR(dev))
+		return -ENODEV;
+
+	ut_assertok_ptr(syscon_regmap_lookup_by_phandle(dev, "first-syscon"));
+	map = syscon_regmap_lookup_by_phandle(dev, "first-syscon");
+	if (map && !IS_ERR(map))
+		ut_asserteq(1, map->range_count);
+
+	ut_assertok_ptr(syscon_regmap_lookup_by_phandle(dev,
+							"second-sys-ctrl"));
+	map = syscon_regmap_lookup_by_phandle(dev, "second-sys-controller");
+	if (map && !IS_ERR(map))
+		ut_asserteq(4, map->range_count);
+
+	ut_assert(IS_ERR(syscon_regmap_lookup_by_phandle(dev, "not-present")));
+
+	return 0;
+}
+DM_TEST(dm_test_syscon_by_phandle, DM_TESTF_SCAN_PDATA | DM_TESTF_SCAN_FDT);
-- 
2.7.4

[U-Boot] [PATCH v5 1/9] syscon: dm: Add a new method to get a regmap from DTS

[Simon Glass]

On 4 May 2018 at 08:34, Jean-Jacques Hiblot <jjhiblot@ti.com> wrote:
> syscon_regmap_lookup_by_phandle() can be used to get the regmap of a
syscon
> device from a reference in the DTS. It operates similarly to the linux
> version of the namesake function.
>
> Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
>
> ---
>
> Changes in v5: None
> Changes in v4:
> - Fix word missing in commit log
>
> Changes in v3:
> - in syscon_regmap_lookup_by_phandle(), use dev_dbg() instead of printf()
> - added unit test for syscon_regmap_lookup_by_phandle()
>
> Changes in v2: None
>
>  arch/sandbox/dts/test.dts    |  6 ++++--
>  drivers/core/syscon-uclass.c | 23 +++++++++++++++++++++++
>  include/syscon.h             | 13 +++++++++++++
>  test/dm/syscon.c             | 29 +++++++++++++++++++++++++++++
>  4 files changed, 69 insertions(+), 2 deletions(-)
>

Reviewed-by: Simon Glass <sjg@chromium.org>

[U-Boot] [PATCH v5 2/9] phy: ti-pip3-phy: Add support for USB3 PHY

[Jean-Jacques Hiblot]

From: Vignesh R <vigneshr@ti.com>

Add support to handle USB3 PHYs present on AM57xx/DRA7xx SoCs. This is
needed to move AM57xx to DM_USB.

Signed-off-by: Vignesh R <vigneshr@ti.com>

Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
---

Changes in v5: None
Changes in v4: None
Changes in v3: None
Changes in v2:
- Add USB3 support to ti-pipe3-phy driver

 drivers/phy/ti-pipe3-phy.c | 30 +++++++++++++++++++++++-------
 1 file changed, 23 insertions(+), 7 deletions(-)

diff --git a/drivers/phy/ti-pipe3-phy.c b/drivers/phy/ti-pipe3-phy.c
index babf2ff..9fddd01 100644
--- a/drivers/phy/ti-pipe3-phy.c
+++ b/drivers/phy/ti-pipe3-phy.c
@@ -266,10 +266,13 @@ static int pipe3_exit(struct phy *phy)
 		return -EBUSY;
 	}
 
-	val = readl(pipe3->pll_reset_reg);
-	writel(val | SATA_PLL_SOFT_RESET, pipe3->pll_reset_reg);
-	mdelay(1);
-	writel(val & ~SATA_PLL_SOFT_RESET, pipe3->pll_reset_reg);
+	if (pipe3->pll_reset_reg) {
+		val = readl(pipe3->pll_reset_reg);
+		writel(val | SATA_PLL_SOFT_RESET, pipe3->pll_reset_reg);
+		mdelay(1);
+		writel(val & ~SATA_PLL_SOFT_RESET, pipe3->pll_reset_reg);
+	}
+
 	return 0;
 }
 
@@ -332,9 +335,11 @@ static int pipe3_phy_probe(struct udevice *dev)
 	if (!pipe3->power_reg)
 		return -EINVAL;
 
-	pipe3->pll_reset_reg = get_reg(dev, "syscon-pllreset");
-	if (!pipe3->pll_reset_reg)
-		return -EINVAL;
+	if (device_is_compatible(dev, "ti,phy-pipe3-sata")) {
+		pipe3->pll_reset_reg = get_reg(dev, "syscon-pllreset");
+		if (!pipe3->pll_reset_reg)
+			return -EINVAL;
+	}
 
 	pipe3->dpll_map = (struct pipe3_dpll_map *)dev_get_driver_data(dev);
 
@@ -351,8 +356,19 @@ static struct pipe3_dpll_map dpll_map_sata[] = {
 	{ },                                    /* Terminator */
 };
 
+static struct pipe3_dpll_map dpll_map_usb[] = {
+	{12000000, {1250, 5, 4, 20, 0} },	/* 12 MHz */
+	{16800000, {3125, 20, 4, 20, 0} },	/* 16.8 MHz */
+	{19200000, {1172, 8, 4, 20, 65537} },	/* 19.2 MHz */
+	{20000000, {1000, 7, 4, 10, 0} },	/* 20 MHz */
+	{26000000, {1250, 12, 4, 20, 0} },	/* 26 MHz */
+	{38400000, {3125, 47, 4, 20, 92843} },	/* 38.4 MHz */
+	{ },					/* Terminator */
+};
+
 static const struct udevice_id pipe3_phy_ids[] = {
 	{ .compatible = "ti,phy-pipe3-sata", .data = (ulong)&dpll_map_sata },
+	{ .compatible = "ti,omap-usb3", .data = (ulong)&dpll_map_usb},
 	{ }
 };
 
-- 
2.7.4

[U-Boot] [PATCH v5 3/9] phy: Add a new driver for OMAP's USB2 PHYs

[Jean-Jacques Hiblot]

This drivers supports the USB2 PHY found on omap5 and dra7 SOCs.

Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>

---

Changes in v5: None
Changes in v4: None
Changes in v3: None
Changes in v2:
- omap-usb2-phy: Implement power_on and power_off callbacks

 drivers/phy/Kconfig         |   8 ++
 drivers/phy/Makefile        |   1 +
 drivers/phy/omap-usb2-phy.c | 196 ++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 205 insertions(+)
 create mode 100644 drivers/phy/omap-usb2-phy.c

diff --git a/drivers/phy/Kconfig b/drivers/phy/Kconfig
index 119edec..ed2b0a8 100644
--- a/drivers/phy/Kconfig
+++ b/drivers/phy/Kconfig
@@ -118,4 +118,12 @@ config MESON_GXL_USB_PHY
 	  This is the generic phy driver for the Amlogic Meson GXL
 	  USB2 and USB3 PHYS.
 
+config OMAP_USB2_PHY
+	bool "Support OMAP's USB2 PHY"
+	depends on PHY
+	depends on SYSCON
+	help
+	  Support for the OMAP's USB2 PHY.
+	  This PHY is found on OMAP devices supporting USB2.
+
 endmenu
diff --git a/drivers/phy/Makefile b/drivers/phy/Makefile
index 72c1492..d440e59 100644
--- a/drivers/phy/Makefile
+++ b/drivers/phy/Makefile
@@ -15,3 +15,4 @@ obj-$(CONFIG_PHY_SANDBOX) += sandbox-phy.o
 obj-$(CONFIG_$(SPL_)PIPE3_PHY) += ti-pipe3-phy.o
 obj-$(CONFIG_STI_USB_PHY) += sti_usb_phy.o
 obj-$(CONFIG_MESON_GXL_USB_PHY) += meson-gxl-usb2.o meson-gxl-usb3.o
+obj-$(CONFIG_OMAP_USB2_PHY) += omap-usb2-phy.o
diff --git a/drivers/phy/omap-usb2-phy.c b/drivers/phy/omap-usb2-phy.c
new file mode 100644
index 0000000..671539b
--- /dev/null
+++ b/drivers/phy/omap-usb2-phy.c
@@ -0,0 +1,196 @@
+/*
+ * OMAP USB2 PHY LAYER
+ * Copyright (C) 2017 Texas Instruments Incorporated - http://www.ti.com/
+ * Written by Jean-Jacques Hiblot  <jjhiblot@ti.com>
+ *
+ * SPDX-License-Identifier:	GPL-2.0+
+ */
+
+#include <common.h>
+#include <asm/io.h>
+#include <dm.h>
+#include <errno.h>
+#include <generic-phy.h>
+#include <regmap.h>
+#include <syscon.h>
+
+#define OMAP_USB2_CALIBRATE_FALSE_DISCONNECT	BIT(0)
+
+#define OMAP_DEV_PHY_PD		BIT(0)
+#define OMAP_USB2_PHY_PD	BIT(28)
+
+#define USB2PHY_DISCON_BYP_LATCH	BIT(31)
+#define USB2PHY_ANA_CONFIG1		(0x4c)
+
+DECLARE_GLOBAL_DATA_PTR;
+
+struct omap_usb2_phy {
+	struct regmap *pwr_regmap;
+	ulong flags;
+	void *phy_base;
+	u32 pwr_reg_offset;
+};
+
+struct usb_phy_data {
+	const char *label;
+	u8 flags;
+	u32 mask;
+	u32 power_on;
+	u32 power_off;
+};
+
+static const struct usb_phy_data omap5_usb2_data = {
+	.label = "omap5_usb2",
+	.flags = 0,
+	.mask = OMAP_DEV_PHY_PD,
+	.power_off = OMAP_DEV_PHY_PD,
+};
+
+static const struct usb_phy_data dra7x_usb2_data = {
+	.label = "dra7x_usb2",
+	.flags = OMAP_USB2_CALIBRATE_FALSE_DISCONNECT,
+	.mask = OMAP_DEV_PHY_PD,
+	.power_off = OMAP_DEV_PHY_PD,
+};
+
+static const struct usb_phy_data dra7x_usb2_phy2_data = {
+	.label = "dra7x_usb2_phy2",
+	.flags = OMAP_USB2_CALIBRATE_FALSE_DISCONNECT,
+	.mask = OMAP_USB2_PHY_PD,
+	.power_off = OMAP_USB2_PHY_PD,
+};
+
+static const struct udevice_id omap_usb2_id_table[] = {
+	{
+		.compatible = "ti,omap5-usb2",
+		.data = (ulong)&omap5_usb2_data,
+	},
+	{
+		.compatible = "ti,dra7x-usb2",
+		.data = (ulong)&dra7x_usb2_data,
+	},
+	{
+		.compatible = "ti,dra7x-usb2-phy2",
+		.data = (ulong)&dra7x_usb2_phy2_data,
+	},
+	{},
+};
+
+static int omap_usb_phy_power(struct phy *usb_phy, bool on)
+{
+	struct udevice *dev = usb_phy->dev;
+	const struct usb_phy_data *data;
+	const struct omap_usb2_phy *phy = dev_get_priv(dev);
+	u32 val;
+	int rc;
+
+	data = (const struct usb_phy_data *)dev_get_driver_data(dev);
+	if (!data)
+		return -EINVAL;
+
+	rc = regmap_read(phy->pwr_regmap, phy->pwr_reg_offset, &val);
+	if (rc)
+		return rc;
+	val &= ~data->mask;
+	if (on)
+		val |= data->power_on;
+	else
+		val |= data->power_off;
+	rc = regmap_write(phy->pwr_regmap, phy->pwr_reg_offset, val);
+	if (rc)
+		return rc;
+
+	return 0;
+}
+
+static int omap_usb2_phy_init(struct phy *usb_phy)
+{
+	struct udevice *dev = usb_phy->dev;
+	struct omap_usb2_phy *priv = dev_get_priv(dev);
+	u32 val;
+
+	if (priv->flags & OMAP_USB2_CALIBRATE_FALSE_DISCONNECT) {
+		/*
+		 *
+		 * Reduce the sensitivity of internal PHY by enabling the
+		 * DISCON_BYP_LATCH of the USB2PHY_ANA_CONFIG1 register. This
+		 * resolves issues with certain devices which can otherwise
+		 * be prone to false disconnects.
+		 *
+		 */
+		val = readl(priv->phy_base + USB2PHY_ANA_CONFIG1);
+		val |= USB2PHY_DISCON_BYP_LATCH;
+		writel(val, priv->phy_base + USB2PHY_ANA_CONFIG1);
+	}
+
+	return 0;
+}
+
+static int omap_usb2_phy_power_on(struct phy *usb_phy)
+{
+	return omap_usb_phy_power(usb_phy, true);
+}
+
+static int omap_usb2_phy_power_off(struct phy *usb_phy)
+{
+	return omap_usb_phy_power(usb_phy, false);
+}
+
+static int omap_usb2_phy_exit(struct phy *usb_phy)
+{
+	return omap_usb_phy_power(usb_phy, false);
+}
+
+struct phy_ops omap_usb2_phy_ops = {
+	.init = omap_usb2_phy_init,
+	.power_on = omap_usb2_phy_power_on,
+	.power_off = omap_usb2_phy_power_off,
+	.exit = omap_usb2_phy_exit,
+};
+
+int omap_usb2_phy_probe(struct udevice *dev)
+{
+	int rc;
+	struct regmap *regmap;
+	struct omap_usb2_phy *priv = dev_get_priv(dev);
+	const struct usb_phy_data *data;
+	u32 tmp[2];
+
+	data = (const struct usb_phy_data *)dev_get_driver_data(dev);
+	if (!data)
+		return -EINVAL;
+
+	if (data->flags & OMAP_USB2_CALIBRATE_FALSE_DISCONNECT) {
+		u32 base = dev_read_addr(dev);
+
+		if (base == FDT_ADDR_T_NONE)
+			return -EINVAL;
+		priv->phy_base = (void *)base;
+		priv->flags |= OMAP_USB2_CALIBRATE_FALSE_DISCONNECT;
+	}
+
+	regmap = syscon_regmap_lookup_by_phandle(dev, "syscon-phy-power");
+	if (IS_ERR(regmap)) {
+		printf("can't get regmap (err %ld)\n", PTR_ERR(regmap));
+		return PTR_ERR(regmap);
+	}
+	priv->pwr_regmap = regmap;
+
+	rc =  dev_read_u32_array(dev, "syscon-phy-power", tmp, 2);
+	if (rc) {
+		printf("couldn't get power reg. offset (err %d)\n", rc);
+		return rc;
+	}
+	priv->pwr_reg_offset = tmp[1];
+
+	return 0;
+}
+
+U_BOOT_DRIVER(omap_usb2_phy) = {
+	.name = "omap_usb2_phy",
+	.id = UCLASS_PHY,
+	.of_match = omap_usb2_id_table,
+	.probe = omap_usb2_phy_probe,
+	.ops = &omap_usb2_phy_ops,
+	.priv_auto_alloc_size = sizeof(struct omap_usb2_phy),
+};
-- 
2.7.4

[U-Boot] [PATCH v5 3/9] phy: Add a new driver for OMAP's USB2 PHYs

[Marek Vasut]

On 05/04/2018 04:34 PM, Jean-Jacques Hiblot wrote:
> This drivers supports the USB2 PHY found on omap5 and dra7 SOCs.
> 
> Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
> 
> ---
> 
> Changes in v5: None
> Changes in v4: None
> Changes in v3: None
> Changes in v2:
> - omap-usb2-phy: Implement power_on and power_off callbacks
> 
>  drivers/phy/Kconfig         |   8 ++
>  drivers/phy/Makefile        |   1 +
>  drivers/phy/omap-usb2-phy.c | 196 ++++++++++++++++++++++++++++++++++++++++++++
>  3 files changed, 205 insertions(+)
>  create mode 100644 drivers/phy/omap-usb2-phy.c
> 
> diff --git a/drivers/phy/Kconfig b/drivers/phy/Kconfig
> index 119edec..ed2b0a8 100644
> --- a/drivers/phy/Kconfig
> +++ b/drivers/phy/Kconfig
> @@ -118,4 +118,12 @@ config MESON_GXL_USB_PHY
>  	  This is the generic phy driver for the Amlogic Meson GXL
>  	  USB2 and USB3 PHYS.
>  
> +config OMAP_USB2_PHY
> +	bool "Support OMAP's USB2 PHY"

s/OMAP's/TI OMAP/ ?
[...]

> +
> +#define OMAP_USB2_CALIBRATE_FALSE_DISCONNECT	BIT(0)
> +
> +#define OMAP_DEV_PHY_PD		BIT(0)
> +#define OMAP_USB2_PHY_PD	BIT(28)
> +
> +#define USB2PHY_DISCON_BYP_LATCH	BIT(31)
> +#define USB2PHY_ANA_CONFIG1		(0x4c)

Parenthesis not needed

> +DECLARE_GLOBAL_DATA_PTR;

Is GD really needed for DM driver ?

[...]

> +static int omap_usb2_phy_init(struct phy *usb_phy)
> +{
> +	struct udevice *dev = usb_phy->dev;
> +	struct omap_usb2_phy *priv = dev_get_priv(dev);
> +	u32 val;
> +
> +	if (priv->flags & OMAP_USB2_CALIBRATE_FALSE_DISCONNECT) {

Invert the condition to reduce level of indent .

> +		/*
> +		 *
> +		 * Reduce the sensitivity of internal PHY by enabling the
> +		 * DISCON_BYP_LATCH of the USB2PHY_ANA_CONFIG1 register. This
> +		 * resolves issues with certain devices which can otherwise
> +		 * be prone to false disconnects.
> +		 *
> +		 */
> +		val = readl(priv->phy_base + USB2PHY_ANA_CONFIG1);
> +		val |= USB2PHY_DISCON_BYP_LATCH;
> +		writel(val, priv->phy_base + USB2PHY_ANA_CONFIG1);
> +	}
> +
> +	return 0;
> +}
[...]


-- 
Best regards,
Marek Vasut

[U-Boot] [PATCH v5 4/9] board: ti: dra7xx-evm: turn on USB clocks in late init stage

[Jean-Jacques Hiblot]

For USB ports that use the Driver Model, turn on the clocks during the
late init stage.

Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
---

Changes in v5: None
Changes in v4: None
Changes in v3: None
Changes in v2: None

 board/ti/dra7xx/evm.c | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)

diff --git a/board/ti/dra7xx/evm.c b/board/ti/dra7xx/evm.c
index 06f061c..c1e1b8e 100644
--- a/board/ti/dra7xx/evm.c
+++ b/board/ti/dra7xx/evm.c
@@ -647,6 +647,19 @@ int dram_init_banksize(void)
 	return 0;
 }
 
+#if CONFIG_IS_ENABLED(DM_USB) && CONFIG_IS_ENABLED(OF_CONTROL)
+static int device_okay(const char *path)
+{
+	int node;
+
+	node = fdt_path_offset(gd->fdt_blob, path);
+	if (node < 0)
+		return 0;
+
+	return fdtdec_get_is_enabled(gd->fdt_blob, node);
+}
+#endif
+
 int board_late_init(void)
 {
 #ifdef CONFIG_ENV_VARS_UBOOT_RUNTIME_CONFIG
@@ -686,6 +699,12 @@ int board_late_init(void)
 	if (board_is_dra71x_evm())
 		palmas_i2c_write_u8(LP873X_I2C_SLAVE_ADDR, 0x9, 0x7);
 #endif
+#if CONFIG_IS_ENABLED(DM_USB) && CONFIG_IS_ENABLED(OF_CONTROL)
+	if (device_okay("/ocp/omap_dwc3_1 at 48880000"))
+		enable_usb_clocks(0);
+	if (device_okay("/ocp/omap_dwc3_2 at 488c0000"))
+		enable_usb_clocks(1);
+#endif
 	return 0;
 }
 
-- 
2.7.4

[U-Boot] [PATCH v5 4/9] board: ti: dra7xx-evm: turn on USB clocks in late init stage

[Marek Vasut]

On 05/04/2018 04:34 PM, Jean-Jacques Hiblot wrote:
> For USB ports that use the Driver Model, turn on the clocks during the
> late init stage.
> 
> Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
> Reviewed-by: Tom Rini <trini@konsulko.com>
> ---
> 
> Changes in v5: None
> Changes in v4: None
> Changes in v3: None
> Changes in v2: None
> 
>  board/ti/dra7xx/evm.c | 19 +++++++++++++++++++
>  1 file changed, 19 insertions(+)
> 
> diff --git a/board/ti/dra7xx/evm.c b/board/ti/dra7xx/evm.c
> index 06f061c..c1e1b8e 100644
> --- a/board/ti/dra7xx/evm.c
> +++ b/board/ti/dra7xx/evm.c
> @@ -647,6 +647,19 @@ int dram_init_banksize(void)
>  	return 0;
>  }
>  
> +#if CONFIG_IS_ENABLED(DM_USB) && CONFIG_IS_ENABLED(OF_CONTROL)
> +static int device_okay(const char *path)
> +{
> +	int node;
> +
> +	node = fdt_path_offset(gd->fdt_blob, path);
> +	if (node < 0)
> +		return 0;
> +
> +	return fdtdec_get_is_enabled(gd->fdt_blob, node);
> +}
> +#endif
> +
>  int board_late_init(void)
>  {
>  #ifdef CONFIG_ENV_VARS_UBOOT_RUNTIME_CONFIG
> @@ -686,6 +699,12 @@ int board_late_init(void)
>  	if (board_is_dra71x_evm())
>  		palmas_i2c_write_u8(LP873X_I2C_SLAVE_ADDR, 0x9, 0x7);
>  #endif
> +#if CONFIG_IS_ENABLED(DM_USB) && CONFIG_IS_ENABLED(OF_CONTROL)
> +	if (device_okay("/ocp/omap_dwc3_1 at 48880000"))
> +		enable_usb_clocks(0);
> +	if (device_okay("/ocp/omap_dwc3_2 at 488c0000"))
> +		enable_usb_clocks(1);
> +#endif

Shouldnt the driver turn this on?

-- 
Best regards,
Marek Vasut

[U-Boot] [PATCH v5 4/9] board: ti: dra7xx-evm: turn on USB clocks in late init stage

[Jean-Jacques Hiblot]

On 04/05/2018 16:38, Marek Vasut wrote:
> On 05/04/2018 04:34 PM, Jean-Jacques Hiblot wrote:
>> For USB ports that use the Driver Model, turn on the clocks during the
>> late init stage.
>>
>> Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
>> Reviewed-by: Tom Rini <trini@konsulko.com>
>> ---
>>
>> Changes in v5: None
>> Changes in v4: None
>> Changes in v3: None
>> Changes in v2: None
>>
>>   board/ti/dra7xx/evm.c | 19 +++++++++++++++++++
>>   1 file changed, 19 insertions(+)
>>
>> diff --git a/board/ti/dra7xx/evm.c b/board/ti/dra7xx/evm.c
>> index 06f061c..c1e1b8e 100644
>> --- a/board/ti/dra7xx/evm.c
>> +++ b/board/ti/dra7xx/evm.c
>> @@ -647,6 +647,19 @@ int dram_init_banksize(void)
>>   	return 0;
>>   }
>>   
>> +#if CONFIG_IS_ENABLED(DM_USB) && CONFIG_IS_ENABLED(OF_CONTROL)
>> +static int device_okay(const char *path)
>> +{
>> +	int node;
>> +
>> +	node = fdt_path_offset(gd->fdt_blob, path);
>> +	if (node < 0)
>> +		return 0;
>> +
>> +	return fdtdec_get_is_enabled(gd->fdt_blob, node);
>> +}
>> +#endif
>> +
>>   int board_late_init(void)
>>   {
>>   #ifdef CONFIG_ENV_VARS_UBOOT_RUNTIME_CONFIG
>> @@ -686,6 +699,12 @@ int board_late_init(void)
>>   	if (board_is_dra71x_evm())
>>   		palmas_i2c_write_u8(LP873X_I2C_SLAVE_ADDR, 0x9, 0x7);
>>   #endif
>> +#if CONFIG_IS_ENABLED(DM_USB) && CONFIG_IS_ENABLED(OF_CONTROL)
>> +	if (device_okay("/ocp/omap_dwc3_1 at 48880000"))
>> +		enable_usb_clocks(0);
>> +	if (device_okay("/ocp/omap_dwc3_2 at 488c0000"))
>> +		enable_usb_clocks(1);
>> +#endif
> Shouldnt the driver turn this on?
AFAIK there is no clock driver for the OMAP. So we have to do it before 
the driver is probed.

JJ
>

[U-Boot] [PATCH v5 4/9] board: ti: dra7xx-evm: turn on USB clocks in late init stage

[Marek Vasut]

On 05/04/2018 04:42 PM, Jean-Jacques Hiblot wrote:
> 
> 
> On 04/05/2018 16:38, Marek Vasut wrote:
>> On 05/04/2018 04:34 PM, Jean-Jacques Hiblot wrote:
>>> For USB ports that use the Driver Model, turn on the clocks during the
>>> late init stage.
>>>
>>> Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
>>> Reviewed-by: Tom Rini <trini@konsulko.com>
>>> ---
>>>
>>> Changes in v5: None
>>> Changes in v4: None
>>> Changes in v3: None
>>> Changes in v2: None
>>>
>>>   board/ti/dra7xx/evm.c | 19 +++++++++++++++++++
>>>   1 file changed, 19 insertions(+)
>>>
>>> diff --git a/board/ti/dra7xx/evm.c b/board/ti/dra7xx/evm.c
>>> index 06f061c..c1e1b8e 100644
>>> --- a/board/ti/dra7xx/evm.c
>>> +++ b/board/ti/dra7xx/evm.c
>>> @@ -647,6 +647,19 @@ int dram_init_banksize(void)
>>>       return 0;
>>>   }
>>>   +#if CONFIG_IS_ENABLED(DM_USB) && CONFIG_IS_ENABLED(OF_CONTROL)
>>> +static int device_okay(const char *path)
>>> +{
>>> +    int node;
>>> +
>>> +    node = fdt_path_offset(gd->fdt_blob, path);
>>> +    if (node < 0)
>>> +        return 0;
>>> +
>>> +    return fdtdec_get_is_enabled(gd->fdt_blob, node);
>>> +}
>>> +#endif
>>> +
>>>   int board_late_init(void)
>>>   {
>>>   #ifdef CONFIG_ENV_VARS_UBOOT_RUNTIME_CONFIG
>>> @@ -686,6 +699,12 @@ int board_late_init(void)
>>>       if (board_is_dra71x_evm())
>>>           palmas_i2c_write_u8(LP873X_I2C_SLAVE_ADDR, 0x9, 0x7);
>>>   #endif
>>> +#if CONFIG_IS_ENABLED(DM_USB) && CONFIG_IS_ENABLED(OF_CONTROL)
>>> +    if (device_okay("/ocp/omap_dwc3_1 at 48880000"))
>>> +        enable_usb_clocks(0);
>>> +    if (device_okay("/ocp/omap_dwc3_2 at 488c0000"))
>>> +        enable_usb_clocks(1);
>>> +#endif
>> Shouldnt the driver turn this on?
> AFAIK there is no clock driver for the OMAP. So we have to do it before
> the driver is probed.

Maybe it's time to implement it instead of piling up those hacks ?

-- 
Best regards,
Marek Vasut

[U-Boot] [PATCH v5 4/9] board: ti: dra7xx-evm: turn on USB clocks in late init stage

[Jean-Jacques Hiblot]

On 04/05/2018 17:43, Marek Vasut wrote:
> On 05/04/2018 04:42 PM, Jean-Jacques Hiblot wrote:
>>
>> On 04/05/2018 16:38, Marek Vasut wrote:
>>> On 05/04/2018 04:34 PM, Jean-Jacques Hiblot wrote:
>>>> For USB ports that use the Driver Model, turn on the clocks during the
>>>> late init stage.
>>>>
>>>> Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
>>>> Reviewed-by: Tom Rini <trini@konsulko.com>
>>>> ---
>>>>
>>>> Changes in v5: None
>>>> Changes in v4: None
>>>> Changes in v3: None
>>>> Changes in v2: None
>>>>
>>>>    board/ti/dra7xx/evm.c | 19 +++++++++++++++++++
>>>>    1 file changed, 19 insertions(+)
>>>>
>>>> diff --git a/board/ti/dra7xx/evm.c b/board/ti/dra7xx/evm.c
>>>> index 06f061c..c1e1b8e 100644
>>>> --- a/board/ti/dra7xx/evm.c
>>>> +++ b/board/ti/dra7xx/evm.c
>>>> @@ -647,6 +647,19 @@ int dram_init_banksize(void)
>>>>        return 0;
>>>>    }
>>>>    +#if CONFIG_IS_ENABLED(DM_USB) && CONFIG_IS_ENABLED(OF_CONTROL)
>>>> +static int device_okay(const char *path)
>>>> +{
>>>> +    int node;
>>>> +
>>>> +    node = fdt_path_offset(gd->fdt_blob, path);
>>>> +    if (node < 0)
>>>> +        return 0;
>>>> +
>>>> +    return fdtdec_get_is_enabled(gd->fdt_blob, node);
>>>> +}
>>>> +#endif
>>>> +
>>>>    int board_late_init(void)
>>>>    {
>>>>    #ifdef CONFIG_ENV_VARS_UBOOT_RUNTIME_CONFIG
>>>> @@ -686,6 +699,12 @@ int board_late_init(void)
>>>>        if (board_is_dra71x_evm())
>>>>            palmas_i2c_write_u8(LP873X_I2C_SLAVE_ADDR, 0x9, 0x7);
>>>>    #endif
>>>> +#if CONFIG_IS_ENABLED(DM_USB) && CONFIG_IS_ENABLED(OF_CONTROL)
>>>> +    if (device_okay("/ocp/omap_dwc3_1 at 48880000"))
>>>> +        enable_usb_clocks(0);
>>>> +    if (device_okay("/ocp/omap_dwc3_2 at 488c0000"))
>>>> +        enable_usb_clocks(1);
>>>> +#endif
>>> Shouldnt the driver turn this on?
>> AFAIK there is no clock driver for the OMAP. So we have to do it before
>> the driver is probed.
> Maybe it's time to implement it instead of piling up those hacks ?
I was thinking along the same lines at some point but then I realized 
that it will make the SPL really big. And we are already tight on some 
platforms like am335x.

JJ
>

[U-Boot] [PATCH v5 4/9] board: ti: dra7xx-evm: turn on USB clocks in late init stage

[Marek Vasut]

On 05/04/2018 07:58 PM, Jean-Jacques Hiblot wrote:
> 
> 
> On 04/05/2018 17:43, Marek Vasut wrote:
>> On 05/04/2018 04:42 PM, Jean-Jacques Hiblot wrote:
>>>
>>> On 04/05/2018 16:38, Marek Vasut wrote:
>>>> On 05/04/2018 04:34 PM, Jean-Jacques Hiblot wrote:
>>>>> For USB ports that use the Driver Model, turn on the clocks during the
>>>>> late init stage.
>>>>>
>>>>> Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
>>>>> Reviewed-by: Tom Rini <trini@konsulko.com>
>>>>> ---
>>>>>
>>>>> Changes in v5: None
>>>>> Changes in v4: None
>>>>> Changes in v3: None
>>>>> Changes in v2: None
>>>>>
>>>>>    board/ti/dra7xx/evm.c | 19 +++++++++++++++++++
>>>>>    1 file changed, 19 insertions(+)
>>>>>
>>>>> diff --git a/board/ti/dra7xx/evm.c b/board/ti/dra7xx/evm.c
>>>>> index 06f061c..c1e1b8e 100644
>>>>> --- a/board/ti/dra7xx/evm.c
>>>>> +++ b/board/ti/dra7xx/evm.c
>>>>> @@ -647,6 +647,19 @@ int dram_init_banksize(void)
>>>>>        return 0;
>>>>>    }
>>>>>    +#if CONFIG_IS_ENABLED(DM_USB) && CONFIG_IS_ENABLED(OF_CONTROL)
>>>>> +static int device_okay(const char *path)
>>>>> +{
>>>>> +    int node;
>>>>> +
>>>>> +    node = fdt_path_offset(gd->fdt_blob, path);
>>>>> +    if (node < 0)
>>>>> +        return 0;
>>>>> +
>>>>> +    return fdtdec_get_is_enabled(gd->fdt_blob, node);
>>>>> +}
>>>>> +#endif
>>>>> +
>>>>>    int board_late_init(void)
>>>>>    {
>>>>>    #ifdef CONFIG_ENV_VARS_UBOOT_RUNTIME_CONFIG
>>>>> @@ -686,6 +699,12 @@ int board_late_init(void)
>>>>>        if (board_is_dra71x_evm())
>>>>>            palmas_i2c_write_u8(LP873X_I2C_SLAVE_ADDR, 0x9, 0x7);
>>>>>    #endif
>>>>> +#if CONFIG_IS_ENABLED(DM_USB) && CONFIG_IS_ENABLED(OF_CONTROL)
>>>>> +    if (device_okay("/ocp/omap_dwc3_1 at 48880000"))
>>>>> +        enable_usb_clocks(0);
>>>>> +    if (device_okay("/ocp/omap_dwc3_2 at 488c0000"))
>>>>> +        enable_usb_clocks(1);
>>>>> +#endif
>>>> Shouldnt the driver turn this on?
>>> AFAIK there is no clock driver for the OMAP. So we have to do it before
>>> the driver is probed.
>> Maybe it's time to implement it instead of piling up those hacks ?
> I was thinking along the same lines at some point but then I realized
> that it will make the SPL really big. And we are already tight on some
> platforms like am335x.

But this is not only SPL , right ? And yes, for SPL you might need to
poke registers or something.

-- 
Best regards,
Marek Vasut

[U-Boot] [PATCH v5 4/9] board: ti: dra7xx-evm: turn on USB clocks in late init stage

[Jean-Jacques Hiblot]

Hi Marek,


On 04/05/2018 21:06, Marek Vasut wrote:
>>>>>
>>>>> Shouldnt the driver turn this on?
>>>> AFAIK there is no clock driver for the OMAP. So we have to do it before
>>>> the driver is probed.
>>> Maybe it's time to implement it instead of piling up those hacks ?
>> I was thinking along the same lines at some point but then I realized
>> that it will make the SPL really big. And we are already tight on some
>> platforms like am335x.
> But this is not only SPL , right ? And yes, for SPL you might need to
> poke registers or something.

IMHO having 2 different ways to enable the clocks in SPL and in u-boot 
is an not improvement over the current situation.
In any case, This is a subject that can be discussed.

Tom, Lokesh,  do you have an opinion on this ?

Meanwhile It would nice to have DM_USB support for the DRA7/AM57 
platforms. It allows using a USB ether adapter with CONFIG_DM_ETHER 
enabled (default on TI's platform). And this series allows using the USB 
3.0 on AM57 evm.

JJ

[U-Boot] [PATCH v5 4/9] board: ti: dra7xx-evm: turn on USB clocks in late init stage

[Jean-Jacques Hiblot]

+Lokesh


On 15/05/2018 18:10, Jean-Jacques Hiblot wrote:
> Hi Marek,
>
>
> On 04/05/2018 21:06, Marek Vasut wrote:
>>>>>>
>>>>>> Shouldnt the driver turn this on?
>>>>> AFAIK there is no clock driver for the OMAP. So we have to do it 
>>>>> before
>>>>> the driver is probed.
>>>> Maybe it's time to implement it instead of piling up those hacks ?
>>> I was thinking along the same lines at some point but then I realized
>>> that it will make the SPL really big. And we are already tight on some
>>> platforms like am335x.
>> But this is not only SPL , right ? And yes, for SPL you might need to
>> poke registers or something.
>
> IMHO having 2 different ways to enable the clocks in SPL and in u-boot 
> is an not improvement over the current situation.
> In any case, This is a subject that can be discussed.
>
> Tom, Lokesh,  do you have an opinion on this ?
>
> Meanwhile It would nice to have DM_USB support for the DRA7/AM57 
> platforms. It allows using a USB ether adapter with CONFIG_DM_ETHER 
> enabled (default on TI's platform). And this series allows using the 
> USB 3.0 on AM57 evm.
>
> JJ
>
>
>
> _______________________________________________
> U-Boot mailing list
> U-Boot at lists.denx.de
> https://lists.denx.de/listinfo/u-boot

[U-Boot] [PATCH v5 4/9] board: ti: dra7xx-evm: turn on USB clocks in late init stage

[Marek Vasut]

On 05/15/2018 06:10 PM, Jean-Jacques Hiblot wrote:
> Hi Marek,
> 
> 
> On 04/05/2018 21:06, Marek Vasut wrote:
>>>>>>
>>>>>> Shouldnt the driver turn this on?
>>>>> AFAIK there is no clock driver for the OMAP. So we have to do it
>>>>> before
>>>>> the driver is probed.
>>>> Maybe it's time to implement it instead of piling up those hacks ?
>>> I was thinking along the same lines at some point but then I realized
>>> that it will make the SPL really big. And we are already tight on some
>>> platforms like am335x.
>> But this is not only SPL , right ? And yes, for SPL you might need to
>> poke registers or something.
> 
> IMHO having 2 different ways to enable the clocks in SPL and in u-boot
> is an not improvement over the current situation.
> In any case, This is a subject that can be discussed.

Do you have a better idea ?

> Tom, Lokesh,  do you have an opinion on this ?
> 
> Meanwhile It would nice to have DM_USB support for the DRA7/AM57
> platforms. It allows using a USB ether adapter with CONFIG_DM_ETHER
> enabled (default on TI's platform). And this series allows using the USB
> 3.0 on AM57 evm.
> 
> JJ
> 
> 
> 


-- 
Best regards,
Marek Vasut

[U-Boot] [PATCH v5 4/9] board: ti: dra7xx-evm: turn on USB clocks in late init stage

[Jean-Jacques Hiblot]

On 16/05/2018 00:13, Marek Vasut wrote:
> On 05/15/2018 06:10 PM, Jean-Jacques Hiblot wrote:
>> Hi Marek,
>>
>>
>> On 04/05/2018 21:06, Marek Vasut wrote:
>>>>>>> Shouldnt the driver turn this on?
>>>>>> AFAIK there is no clock driver for the OMAP. So we have to do it
>>>>>> before
>>>>>> the driver is probed.
>>>>> Maybe it's time to implement it instead of piling up those hacks ?
>>>> I was thinking along the same lines at some point but then I realized
>>>> that it will make the SPL really big. And we are already tight on some
>>>> platforms like am335x.
>>> But this is not only SPL , right ? And yes, for SPL you might need to
>>> poke registers or something.
>> IMHO having 2 different ways to enable the clocks in SPL and in u-boot
>> is an not improvement over the current situation.
>> In any case, This is a subject that can be discussed.
> Do you have a better idea ?
I would keep it as it is today.

After studying the clock management in Linux, I believe that 
implementing the clock drivers in u-boot won't be an easy task and 
moreover won't make things much cleaner.
The way clocks are handled in Linux for the omap platforms is not 
completely standard. There are a lot of clocks that are handled by the 
clock framework (pll, mux, some gates, ...) but most of the clock gating 
for peripherals is handled the HWMOD subsystem which automatically 
enables/disables the clocks of the peripherals. Parameters for this 
subsystem are hard-coded in platform code (in 
arch/arm/mach-omap2/omap-hwmod_xxx_data.c) not in the DTS. It also 
handle clock and power domains.
HWMOD is probably going to go away some day, I'm not sure that we want 
to implement it in u-boot. Until hwmod has completely replaced by more 
generic frameworks, I would stick to platform code to enable clocks in 
u-boot.

Jean-Jacques
>
>> Tom, Lokesh,  do you have an opinion on this ?
>>
>> Meanwhile It would nice to have DM_USB support for the DRA7/AM57
>> platforms. It allows using a USB ether adapter with CONFIG_DM_ETHER
>> enabled (default on TI's platform). And this series allows using the USB
>> 3.0 on AM57 evm.
>>
>> JJ
>>
>>
>>
>

[U-Boot] [PATCH v5 4/9] board: ti: dra7xx-evm: turn on USB clocks in late init stage

[Marek Vasut]

On 05/16/2018 11:00 AM, Jean-Jacques Hiblot wrote:
> 
> 
> On 16/05/2018 00:13, Marek Vasut wrote:
>> On 05/15/2018 06:10 PM, Jean-Jacques Hiblot wrote:
>>> Hi Marek,
>>>
>>>
>>> On 04/05/2018 21:06, Marek Vasut wrote:
>>>>>>>> Shouldnt the driver turn this on?
>>>>>>> AFAIK there is no clock driver for the OMAP. So we have to do it
>>>>>>> before
>>>>>>> the driver is probed.
>>>>>> Maybe it's time to implement it instead of piling up those hacks ?
>>>>> I was thinking along the same lines at some point but then I realized
>>>>> that it will make the SPL really big. And we are already tight on some
>>>>> platforms like am335x.
>>>> But this is not only SPL , right ? And yes, for SPL you might need to
>>>> poke registers or something.
>>> IMHO having 2 different ways to enable the clocks in SPL and in u-boot
>>> is an not improvement over the current situation.
>>> In any case, This is a subject that can be discussed.
>> Do you have a better idea ?
> I would keep it as it is today.
> 
> After studying the clock management in Linux, I believe that
> implementing the clock drivers in u-boot won't be an easy task and
> moreover won't make things much cleaner.
> The way clocks are handled in Linux for the omap platforms is not
> completely standard. There are a lot of clocks that are handled by the
> clock framework (pll, mux, some gates, ...) but most of the clock gating
> for peripherals is handled the HWMOD subsystem which automatically
> enables/disables the clocks of the peripherals. Parameters for this
> subsystem are hard-coded in platform code (in
> arch/arm/mach-omap2/omap-hwmod_xxx_data.c) not in the DTS. It also
> handle clock and power domains.
> HWMOD is probably going to go away some day, I'm not sure that we want
> to implement it in u-boot. Until hwmod has completely replaced by more
> generic frameworks, I would stick to platform code to enable clocks in
> u-boot.

But if this keeps going on, eventually we will regress to where we were
years ago -- having huge platform specific list of registers to poke to
set a platform up and that'd be it ?

-- 
Best regards,
Marek Vasut

[U-Boot] [PATCH v5 4/9] board: ti: dra7xx-evm: turn on USB clocks in late init stage

[Jean-Jacques Hiblot]

On 16/05/2018 11:04, Marek Vasut wrote:
> On 05/16/2018 11:00 AM, Jean-Jacques Hiblot wrote:
>>
>> On 16/05/2018 00:13, Marek Vasut wrote:
>>> On 05/15/2018 06:10 PM, Jean-Jacques Hiblot wrote:
>>>> Hi Marek,
>>>>
>>>>
>>>> On 04/05/2018 21:06, Marek Vasut wrote:
>>>>>>>>> Shouldnt the driver turn this on?
>>>>>>>> AFAIK there is no clock driver for the OMAP. So we have to do it
>>>>>>>> before
>>>>>>>> the driver is probed.
>>>>>>> Maybe it's time to implement it instead of piling up those hacks ?
>>>>>> I was thinking along the same lines at some point but then I realized
>>>>>> that it will make the SPL really big. And we are already tight on some
>>>>>> platforms like am335x.
>>>>> But this is not only SPL , right ? And yes, for SPL you might need to
>>>>> poke registers or something.
>>>> IMHO having 2 different ways to enable the clocks in SPL and in u-boot
>>>> is an not improvement over the current situation.
>>>> In any case, This is a subject that can be discussed.
>>> Do you have a better idea ?
>> I would keep it as it is today.
>>
>> After studying the clock management in Linux, I believe that
>> implementing the clock drivers in u-boot won't be an easy task and
>> moreover won't make things much cleaner.
>> The way clocks are handled in Linux for the omap platforms is not
>> completely standard. There are a lot of clocks that are handled by the
>> clock framework (pll, mux, some gates, ...) but most of the clock gating
>> for peripherals is handled the HWMOD subsystem which automatically
>> enables/disables the clocks of the peripherals. Parameters for this
>> subsystem are hard-coded in platform code (in
>> arch/arm/mach-omap2/omap-hwmod_xxx_data.c) not in the DTS. It also
>> handle clock and power domains.
>> HWMOD is probably going to go away some day, I'm not sure that we want
>> to implement it in u-boot. Until hwmod has completely replaced by more
>> generic frameworks, I would stick to platform code to enable clocks in
>> u-boot.
> But if this keeps going on, eventually we will regress to where we were
> years ago -- having huge platform specific list of registers to poke to
> set a platform up and that'd be it ?
For omap clocks that would be status quo but the rest is moving toward DM

[U-Boot] [PATCH v5 4/9] board: ti: dra7xx-evm: turn on USB clocks in late init stage

[Tom Rini]

On Wed, May 16, 2018 at 11:04:18AM +0200, Marek Vasut wrote:
> On 05/16/2018 11:00 AM, Jean-Jacques Hiblot wrote:
> > 
> > 
> > On 16/05/2018 00:13, Marek Vasut wrote:
> >> On 05/15/2018 06:10 PM, Jean-Jacques Hiblot wrote:
> >>> Hi Marek,
> >>>
> >>>
> >>> On 04/05/2018 21:06, Marek Vasut wrote:
> >>>>>>>> Shouldnt the driver turn this on?
> >>>>>>> AFAIK there is no clock driver for the OMAP. So we have to do it
> >>>>>>> before
> >>>>>>> the driver is probed.
> >>>>>> Maybe it's time to implement it instead of piling up those hacks ?
> >>>>> I was thinking along the same lines at some point but then I realized
> >>>>> that it will make the SPL really big. And we are already tight on some
> >>>>> platforms like am335x.
> >>>> But this is not only SPL , right ? And yes, for SPL you might need to
> >>>> poke registers or something.
> >>> IMHO having 2 different ways to enable the clocks in SPL and in u-boot
> >>> is an not improvement over the current situation.
> >>> In any case, This is a subject that can be discussed.
> >> Do you have a better idea ?
> > I would keep it as it is today.
> > 
> > After studying the clock management in Linux, I believe that
> > implementing the clock drivers in u-boot won't be an easy task and
> > moreover won't make things much cleaner.
> > The way clocks are handled in Linux for the omap platforms is not
> > completely standard. There are a lot of clocks that are handled by the
> > clock framework (pll, mux, some gates, ...) but most of the clock gating
> > for peripherals is handled the HWMOD subsystem which automatically
> > enables/disables the clocks of the peripherals. Parameters for this
> > subsystem are hard-coded in platform code (in
> > arch/arm/mach-omap2/omap-hwmod_xxx_data.c) not in the DTS. It also
> > handle clock and power domains.
> > HWMOD is probably going to go away some day, I'm not sure that we want
> > to implement it in u-boot. Until hwmod has completely replaced by more
> > generic frameworks, I would stick to platform code to enable clocks in
> > u-boot.
> 
> But if this keeps going on, eventually we will regress to where we were
> years ago -- having huge platform specific list of registers to poke to
> set a platform up and that'd be it ?

We should certainly re-evaluate the situation once hwmod finally gets
replaced in the kernel.  That's a long-standing and slowly-but-surely
happening thing.

-- 
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20180516/1bb5801e/attachment.sig>

[U-Boot] [PATCH v5 4/9] board: ti: dra7xx-evm: turn on USB clocks in late init stage

[Jean-Jacques Hiblot]

On 16/05/2018 17:21, Tom Rini wrote:
> On Wed, May 16, 2018 at 11:04:18AM +0200, Marek Vasut wrote:
>> On 05/16/2018 11:00 AM, Jean-Jacques Hiblot wrote:
>>>
>>> On 16/05/2018 00:13, Marek Vasut wrote:
>>>> On 05/15/2018 06:10 PM, Jean-Jacques Hiblot wrote:
>>>>> Hi Marek,
>>>>>
>>>>>
>>>>> On 04/05/2018 21:06, Marek Vasut wrote:
>>>>>>>>>> Shouldnt the driver turn this on?
>>>>>>>>> AFAIK there is no clock driver for the OMAP. So we have to do it
>>>>>>>>> before
>>>>>>>>> the driver is probed.
>>>>>>>> Maybe it's time to implement it instead of piling up those hacks ?
>>>>>>> I was thinking along the same lines at some point but then I realized
>>>>>>> that it will make the SPL really big. And we are already tight on some
>>>>>>> platforms like am335x.
>>>>>> But this is not only SPL , right ? And yes, for SPL you might need to
>>>>>> poke registers or something.
>>>>> IMHO having 2 different ways to enable the clocks in SPL and in u-boot
>>>>> is an not improvement over the current situation.
>>>>> In any case, This is a subject that can be discussed.
>>>> Do you have a better idea ?
>>> I would keep it as it is today.
>>>
>>> After studying the clock management in Linux, I believe that
>>> implementing the clock drivers in u-boot won't be an easy task and
>>> moreover won't make things much cleaner.
>>> The way clocks are handled in Linux for the omap platforms is not
>>> completely standard. There are a lot of clocks that are handled by the
>>> clock framework (pll, mux, some gates, ...) but most of the clock gating
>>> for peripherals is handled the HWMOD subsystem which automatically
>>> enables/disables the clocks of the peripherals. Parameters for this
>>> subsystem are hard-coded in platform code (in
>>> arch/arm/mach-omap2/omap-hwmod_xxx_data.c) not in the DTS. It also
>>> handle clock and power domains.
>>> HWMOD is probably going to go away some day, I'm not sure that we want
>>> to implement it in u-boot. Until hwmod has completely replaced by more
>>> generic frameworks, I would stick to platform code to enable clocks in
>>> u-boot.
>> But if this keeps going on, eventually we will regress to where we were
>> years ago -- having huge platform specific list of registers to poke to
>> set a platform up and that'd be it ?
> We should certainly re-evaluate the situation once hwmod finally gets
> replaced in the kernel.  That's a long-standing and slowly-but-surely
> happening thing.
I understand it is frustrating to merge this kind of platform code when
a generic framework exists, but it is  one step in having DM_USB
support which in the end will remove more platform code.

>

[U-Boot] [PATCH v5 4/9] board: ti: dra7xx-evm: turn on USB clocks in late init stage

[Tom Rini]

On Wed, May 16, 2018 at 11:00:54AM +0200, Jean-Jacques Hiblot wrote:
> 
> 
> On 16/05/2018 00:13, Marek Vasut wrote:
> >On 05/15/2018 06:10 PM, Jean-Jacques Hiblot wrote:
> >>Hi Marek,
> >>
> >>
> >>On 04/05/2018 21:06, Marek Vasut wrote:
> >>>>>>>Shouldnt the driver turn this on?
> >>>>>>AFAIK there is no clock driver for the OMAP. So we have to do it
> >>>>>>before
> >>>>>>the driver is probed.
> >>>>>Maybe it's time to implement it instead of piling up those hacks ?
> >>>>I was thinking along the same lines at some point but then I realized
> >>>>that it will make the SPL really big. And we are already tight on some
> >>>>platforms like am335x.
> >>>But this is not only SPL , right ? And yes, for SPL you might need to
> >>>poke registers or something.
> >>IMHO having 2 different ways to enable the clocks in SPL and in u-boot
> >>is an not improvement over the current situation.
> >>In any case, This is a subject that can be discussed.
> >Do you have a better idea ?
> I would keep it as it is today.
> 
> After studying the clock management in Linux, I believe that implementing
> the clock drivers in u-boot won't be an easy task and moreover won't make
> things much cleaner.
> The way clocks are handled in Linux for the omap platforms is not completely
> standard. There are a lot of clocks that are handled by the clock framework
> (pll, mux, some gates, ...) but most of the clock gating for peripherals is
> handled the HWMOD subsystem which automatically enables/disables the clocks
> of the peripherals. Parameters for this subsystem are hard-coded in platform
> code (in arch/arm/mach-omap2/omap-hwmod_xxx_data.c) not in the DTS. It also
> handle clock and power domains.
> HWMOD is probably going to go away some day, I'm not sure that we want to
> implement it in u-boot. Until hwmod has completely replaced by more generic
> frameworks, I would stick to platform code to enable clocks in u-boot.

Yes, this is an area that doesn't have a small, clean and generic
solution.

-- 
Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20180516/01c9b66d/attachment.sig>

[U-Boot] [PATCH v5 5/9] board; ti: am57xx: turn on USB clocks

[Jean-Jacques Hiblot]

From: Vignesh R <vigneshr@ti.com>

Enable USB clocks in late init stage to support ports under DM_USB.

Signed-off-by: Vignesh R <vigneshr@ti.com>

Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
---

Changes in v5: None
Changes in v4: None
Changes in v3: None
Changes in v2:
- am57xx boards: when DM_USB is used, turn on the required USB clocks

 board/ti/am57xx/board.c | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)

diff --git a/board/ti/am57xx/board.c b/board/ti/am57xx/board.c
index 5bd8778..cd53776 100644
--- a/board/ti/am57xx/board.c
+++ b/board/ti/am57xx/board.c
@@ -676,6 +676,19 @@ out:
 	return;
 }
 
+#if CONFIG_IS_ENABLED(DM_USB) && CONFIG_IS_ENABLED(OF_CONTROL)
+static int device_okay(const char *path)
+{
+	int node;
+
+	node = fdt_path_offset(gd->fdt_blob, path);
+	if (node < 0)
+		return 0;
+
+	return fdtdec_get_is_enabled(gd->fdt_blob, node);
+}
+#endif
+
 int board_late_init(void)
 {
 	setup_board_eeprom_env();
@@ -715,6 +728,12 @@ int board_late_init(void)
 	board_ti_set_ethaddr(2);
 #endif
 
+#if CONFIG_IS_ENABLED(DM_USB) && CONFIG_IS_ENABLED(OF_CONTROL)
+	if (device_okay("/ocp/omap_dwc3_1 at 48880000"))
+		enable_usb_clocks(0);
+	if (device_okay("/ocp/omap_dwc3_2 at 488c0000"))
+		enable_usb_clocks(1);
+#endif
 	return 0;
 }
 
-- 
2.7.4

[U-Boot] [PATCH v5 6/9] dts: dra7x: make ocp2scp@4a080000 compatible with simple-bus

[Jean-Jacques Hiblot]

This is required when DM_USB is used, to bind the USB phys.

Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>

---

Changes in v5: None
Changes in v4: None
Changes in v3: None
Changes in v2:
- split dts changes in 2 commits: one for binding the children of
  ocp2scp at 4a080000, and one to disable USB1  on all DRA7 EVMs

 arch/arm/dts/omap5-u-boot.dtsi | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/arch/arm/dts/omap5-u-boot.dtsi b/arch/arm/dts/omap5-u-boot.dtsi
index bf2684c..a6a7801 100644
--- a/arch/arm/dts/omap5-u-boot.dtsi
+++ b/arch/arm/dts/omap5-u-boot.dtsi
@@ -15,6 +15,10 @@
 	ocp {
 		u-boot,dm-spl;
 
+		ocp2scp at 4a080000 {
+			compatible = "ti,omap-ocp2scp", "simple-bus";
+		};
+
 		ocp2scp at 4a090000 {
 			compatible = "ti,omap-ocp2scp", "simple-bus";
 		};
-- 
2.7.4

[U-Boot] [PATCH v5 7/9] dts: dra7x: Disable USB1 on all evms

[Jean-Jacques Hiblot]

On all the EVMs featuring a SOC of the DRA7 family, the USB1 port is used
as a device for DFU. This port is managed by the platform code and must not
be advertised to the DWC3 DM driver.
This will be changed when/if support for the device mode is added to the
dwc3-omap driver.

Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>

---

Changes in v5: None
Changes in v4: None
Changes in v3: None
Changes in v2:
- Instead of disabling USB1 port in dts files, use *-u-boot.dtsi files

 arch/arm/dts/dra7-evm-u-boot.dtsi       |  8 ++++++++
 arch/arm/dts/dra71-evm-u-boot.dtsi      |  9 +++++++++
 arch/arm/dts/dra72-evm-revc-u-boot.dtsi |  8 ++++++++
 arch/arm/dts/dra72-evm-u-boot.dtsi      | 22 ++++++++++++++++++++++
 arch/arm/dts/dra76-evm-u-boot.dtsi      |  9 +++++++++
 5 files changed, 56 insertions(+)
 create mode 100644 arch/arm/dts/dra72-evm-u-boot.dtsi

diff --git a/arch/arm/dts/dra7-evm-u-boot.dtsi b/arch/arm/dts/dra7-evm-u-boot.dtsi
index 3e7da7c..59b5643 100644
--- a/arch/arm/dts/dra7-evm-u-boot.dtsi
+++ b/arch/arm/dts/dra7-evm-u-boot.dtsi
@@ -33,3 +33,11 @@
 &mmc2_iodelay_hs200_rev20_conf {
 	u-boot,dm-spl;
 };
+
+&usb1 {
+	status = "disabled";
+};
+
+&omap_dwc3_1 {
+	status = "disabled";
+};
diff --git a/arch/arm/dts/dra71-evm-u-boot.dtsi b/arch/arm/dts/dra71-evm-u-boot.dtsi
index e2ab0bb..a61fcfb 100644
--- a/arch/arm/dts/dra71-evm-u-boot.dtsi
+++ b/arch/arm/dts/dra71-evm-u-boot.dtsi
@@ -45,3 +45,12 @@
 &mmc2_iodelay_hs200_rev20_conf {
 	u-boot,dm-spl;
 };
+
+&usb1 {
+	status = "disabled";
+};
+
+&omap_dwc3_1 {
+	status = "disabled";
+};
+
diff --git a/arch/arm/dts/dra72-evm-revc-u-boot.dtsi b/arch/arm/dts/dra72-evm-revc-u-boot.dtsi
index e2ab0bb..aeca9b2 100644
--- a/arch/arm/dts/dra72-evm-revc-u-boot.dtsi
+++ b/arch/arm/dts/dra72-evm-revc-u-boot.dtsi
@@ -45,3 +45,11 @@
 &mmc2_iodelay_hs200_rev20_conf {
 	u-boot,dm-spl;
 };
+
+&usb1 {
+	status = "disabled";
+};
+
+&omap_dwc3_1 {
+	status = "disabled";
+};
diff --git a/arch/arm/dts/dra72-evm-u-boot.dtsi b/arch/arm/dts/dra72-evm-u-boot.dtsi
new file mode 100644
index 0000000..3d1dd6a
--- /dev/null
+++ b/arch/arm/dts/dra72-evm-u-boot.dtsi
@@ -0,0 +1,22 @@
+/*
+ * Copyright (C) 2018 Texas Instruments Incorporated - http://www.ti.com/
+ * SPDX-License-Identifier:	GPL-2.0+
+ */
+
+#include "omap5-u-boot.dtsi"
+
+&pcf_gpio_21{
+	u-boot,i2c-offset-len = <0>;
+};
+
+&pcf_hdmi{
+	u-boot,i2c-offset-len = <0>;
+};
+
+&usb1 {
+	status = "disabled";
+};
+
+&omap_dwc3_1 {
+	status = "disabled";
+};
diff --git a/arch/arm/dts/dra76-evm-u-boot.dtsi b/arch/arm/dts/dra76-evm-u-boot.dtsi
index a5a0694..c719421 100644
--- a/arch/arm/dts/dra76-evm-u-boot.dtsi
+++ b/arch/arm/dts/dra76-evm-u-boot.dtsi
@@ -25,3 +25,12 @@
 &mmc2_iodelay_hs200_conf {
 	u-boot,dm-spl;
 };
+
+&usb1 {
+	status = "disabled";
+};
+
+&omap_dwc3_1 {
+	status = "disabled";
+};
+
-- 
2.7.4

[U-Boot] [PATCH v5 8/9] configs: enable OMAP_USB2_PHY for all DRA7 platforms

[Jean-Jacques Hiblot]

Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>

---

Changes in v5: None
Changes in v4:
- Remove dependency on MISC
- Renamed the commit (DM_USB is already enabled, the only missing option
  is for the USB2 phy driver)

Changes in v3: None
Changes in v2: None

 configs/dra7xx_evm_defconfig    | 1 +
 configs/dra7xx_hs_evm_defconfig | 1 +
 2 files changed, 2 insertions(+)

diff --git a/configs/dra7xx_evm_defconfig b/configs/dra7xx_evm_defconfig
index 9b81b0c..562046f 100644
--- a/configs/dra7xx_evm_defconfig
+++ b/configs/dra7xx_evm_defconfig
@@ -68,6 +68,7 @@ CONFIG_DRIVER_TI_CPSW=y
 CONFIG_PHY_GIGE=y
 CONFIG_SPL_PHY=y
 CONFIG_PIPE3_PHY=y
+CONFIG_OMAP_USB2_PHY=y
 CONFIG_PMIC_PALMAS=y
 CONFIG_PMIC_LP873X=y
 CONFIG_DM_REGULATOR_FIXED=y
diff --git a/configs/dra7xx_hs_evm_defconfig b/configs/dra7xx_hs_evm_defconfig
index 536946e..6375aad 100644
--- a/configs/dra7xx_hs_evm_defconfig
+++ b/configs/dra7xx_hs_evm_defconfig
@@ -67,6 +67,7 @@ CONFIG_DRIVER_TI_CPSW=y
 CONFIG_PHY_GIGE=y
 CONFIG_SPL_PHY=y
 CONFIG_PIPE3_PHY=y
+CONFIG_OMAP_USB2_PHY=y
 CONFIG_PMIC_PALMAS=y
 CONFIG_PMIC_LP873X=y
 CONFIG_DM_REGULATOR_FIXED=y
-- 
2.7.4

[U-Boot] [PATCH v5 9/9] configs: am57xx_evm: Enable DM_USB and dependencies

[Jean-Jacques Hiblot]

Enable DM_USB for AM57xx based boards.

Signed-off-by: Vignesh R <vigneshr@ti.com>

Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
---
Changes in v5: None
Changes in v4:
- Remove dependency on MISC
- Remove omap specific glue layer and use dwc3-of-simple glue layer

Changes in v3: None
Changes in v2:
- Enable DM_USB in am57xx_evm_defconfig. USB3 (super speed) is supported.

 configs/am57xx_evm_defconfig | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/configs/am57xx_evm_defconfig b/configs/am57xx_evm_defconfig
index 0e99268..dc2f29c 100644
--- a/configs/am57xx_evm_defconfig
+++ b/configs/am57xx_evm_defconfig
@@ -79,3 +79,7 @@ CONFIG_USB_GADGET=y
 CONFIG_USB_GADGET_MANUFACTURER="Texas Instruments"
 CONFIG_USB_GADGET_VENDOR_NUM=0x0451
 CONFIG_USB_GADGET_PRODUCT_NUM=0xd022
+CONFIG_DM_USB=y
+CONFIG_OMAP_USB2_PHY=y
+CONFIG_PIPE3_PHY=y
+CONFIG_PHY=y
-- 
2.7.4