[PATCH v2 0/7] VIM3: add support for checking 'Function' button state

[PATCH v2 0/7] VIM3: add support for checking 'Function' button state

[Marek Szyprowski]

Hi All,

This patchset adds all building blocks needed for checking the 'Function'
button state in the boot script on Amlogic A311D based VIM3 board. This
button is connected to the ADC lines of the SoC, so it required to enable
meson SARADC, the clocks needed for it and a simple button-adc drivers.

Once applied, one can use following commands in the boot scripts:
-->8---
echo Checking Func button state: \\c
button Function
if test ${button} = on
then
	echo Selected alternative boot
	...
fi
--->8---

Best regards
Marek Szyprowski
Samsung R&D Institute Poland


Changelog:
v2:
- removed Change-Id tags
- split defconfig changes into ADC and button related

v1: https://lists.denx.de/pipermail/u-boot/2020-December/434875.html
- initial submission


Patch summary:

Marek Szyprowski (7):
  clk: meson: add minimal driver for g12a-ao clocks
  adc: meson-saradc: add G12A variant
  adc: meson-saradc: skip hardware init only if ADC is enabled
  configs: khadas-vim3: enable ADC device support
  button: add a simple ADC-based button driver
  cmd: button: store button state in the 'button' env
  configs: khadas-vim3: enable Function button support

 cmd/button.c                  |   4 +-
 configs/khadas-vim3_defconfig |   4 ++
 drivers/adc/meson-saradc.c    |   9 ++-
 drivers/button/Kconfig        |   8 +++
 drivers/button/Makefile       |   1 +
 drivers/button/button-adc.c   | 117 ++++++++++++++++++++++++++++++++++
 drivers/clk/meson/Makefile    |   1 +
 drivers/clk/meson/g12a-ao.c   |  83 ++++++++++++++++++++++++
 8 files changed, 224 insertions(+), 3 deletions(-)
 create mode 100644 drivers/button/button-adc.c
 create mode 100644 drivers/clk/meson/g12a-ao.c

-- 
2.17.1

[PATCH v2 0/7] VIM3: add support for checking 'Function' button state

[Marek Szyprowski]

Hi All,

This patchset adds all building blocks needed for checking the 'Function'
button state in the boot script on Amlogic A311D based VIM3 board. This
button is connected to the ADC lines of the SoC, so it required to enable
meson SARADC, the clocks needed for it and a simple button-adc drivers.

Once applied, one can use following commands in the boot scripts:
-->8---
echo Checking Func button state: \\c
button Function
if test ${button} = on
then
	echo Selected alternative boot
	...
fi
--->8---

Best regards
Marek Szyprowski
Samsung R&D Institute Poland


Changelog:
v2:
- removed Change-Id tags
- split defconfig changes into ADC and button related

v1: https://lists.denx.de/pipermail/u-boot/2020-December/434875.html
- initial submission


Patch summary:

Marek Szyprowski (7):
  clk: meson: add minimal driver for g12a-ao clocks
  adc: meson-saradc: add G12A variant
  adc: meson-saradc: skip hardware init only if ADC is enabled
  configs: khadas-vim3: enable ADC device support
  button: add a simple ADC-based button driver
  cmd: button: store button state in the 'button' env
  configs: khadas-vim3: enable Function button support

 cmd/button.c                  |   4 +-
 configs/khadas-vim3_defconfig |   4 ++
 drivers/adc/meson-saradc.c    |   9 ++-
 drivers/button/Kconfig        |   8 +++
 drivers/button/Makefile       |   1 +
 drivers/button/button-adc.c   | 117 ++++++++++++++++++++++++++++++++++
 drivers/clk/meson/Makefile    |   1 +
 drivers/clk/meson/g12a-ao.c   |  83 ++++++++++++++++++++++++
 8 files changed, 224 insertions(+), 3 deletions(-)
 create mode 100644 drivers/button/button-adc.c
 create mode 100644 drivers/clk/meson/g12a-ao.c

-- 
2.17.1

[PATCH v2 1/7] clk: meson: add minimal driver for g12a-ao clocks

[Marek Szyprowski]

Add minimal driver AO clocks on meson G12A family. Only ADC related clocks
are supported.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Neil Armstrong <narmstrong@baylibre.com>
Tested-by: Jaehoon Chung <jh80.chung@samsung.com>
Reviewed-by: Jaehoon Chung <jh80.chung@samsung.com>
---
 drivers/clk/meson/Makefile  |  1 +
 drivers/clk/meson/g12a-ao.c | 83 +++++++++++++++++++++++++++++++++++++
 2 files changed, 84 insertions(+)
 create mode 100644 drivers/clk/meson/g12a-ao.c

diff --git a/drivers/clk/meson/Makefile b/drivers/clk/meson/Makefile
index c873d6976f..7204383e17 100644
--- a/drivers/clk/meson/Makefile
+++ b/drivers/clk/meson/Makefile
@@ -6,4 +6,5 @@
 obj-$(CONFIG_CLK_MESON_GX) += gxbb.o
 obj-$(CONFIG_CLK_MESON_AXG) += axg.o
 obj-$(CONFIG_CLK_MESON_G12A) += g12a.o
+obj-$(CONFIG_CLK_MESON_G12A) += g12a-ao.o
 
diff --git a/drivers/clk/meson/g12a-ao.c b/drivers/clk/meson/g12a-ao.c
new file mode 100644
index 0000000000..7a0abea77c
--- /dev/null
+++ b/drivers/clk/meson/g12a-ao.c
@@ -0,0 +1,83 @@
+// SPDX-License-Identifier: GPL-2.0+
+
+#include <common.h>
+#include <log.h>
+#include <asm/io.h>
+#include <clk-uclass.h>
+#include <dm.h>
+#include <regmap.h>
+#include <syscon.h>
+#include <dt-bindings/clock/g12a-aoclkc.h>
+
+#include "clk_meson.h"
+
+struct meson_clk {
+	struct regmap *map;
+};
+
+#define AO_CLK_GATE0		0x4c
+#define AO_SAR_CLK		0x90
+
+static struct meson_gate gates[] = {
+	MESON_GATE(CLKID_AO_SAR_ADC, AO_CLK_GATE0, 8),
+	MESON_GATE(CLKID_AO_SAR_ADC_CLK, AO_SAR_CLK, 8),
+};
+
+static int meson_set_gate(struct clk *clk, bool on)
+{
+	struct meson_clk *priv = dev_get_priv(clk->dev);
+	struct meson_gate *gate;
+
+	if (clk->id >= ARRAY_SIZE(gates))
+		return -ENOENT;
+
+	gate = &gates[clk->id];
+
+	if (gate->reg == 0)
+		return 0;
+
+	regmap_update_bits(priv->map, gate->reg,
+			   BIT(gate->bit), on ? BIT(gate->bit) : 0);
+
+	return 0;
+}
+
+static int meson_clk_enable(struct clk *clk)
+{
+	return meson_set_gate(clk, true);
+}
+
+static int meson_clk_disable(struct clk *clk)
+{
+	return meson_set_gate(clk, false);
+}
+
+static int meson_clk_probe(struct udevice *dev)
+{
+	struct meson_clk *priv = dev_get_priv(dev);
+
+	priv->map = syscon_node_to_regmap(dev_get_parent(dev)->node);
+	if (IS_ERR(priv->map))
+		return PTR_ERR(priv->map);
+
+	return 0;
+}
+
+static struct clk_ops meson_clk_ops = {
+	.disable	= meson_clk_disable,
+	.enable		= meson_clk_enable,
+};
+
+static const struct udevice_id meson_clk_ids[] = {
+	{ .compatible = "amlogic,meson-g12a-aoclkc" },
+	{ }
+};
+
+U_BOOT_DRIVER(meson_clk_axg) = {
+	.name		= "meson_clk_g12a_ao",
+	.id		= UCLASS_CLK,
+	.of_match	= meson_clk_ids,
+	.priv_auto_alloc_size = sizeof(struct meson_clk),
+	.ops		= &meson_clk_ops,
+	.probe		= meson_clk_probe,
+};
-- 
2.17.1

[PATCH v2 1/7] clk: meson: add minimal driver for g12a-ao clocks

[Marek Szyprowski]

Add minimal driver AO clocks on meson G12A family. Only ADC related clocks
are supported.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Neil Armstrong <narmstrong@baylibre.com>
Tested-by: Jaehoon Chung <jh80.chung@samsung.com>
Reviewed-by: Jaehoon Chung <jh80.chung@samsung.com>
---
 drivers/clk/meson/Makefile  |  1 +
 drivers/clk/meson/g12a-ao.c | 83 +++++++++++++++++++++++++++++++++++++
 2 files changed, 84 insertions(+)
 create mode 100644 drivers/clk/meson/g12a-ao.c

diff --git a/drivers/clk/meson/Makefile b/drivers/clk/meson/Makefile
index c873d6976f..7204383e17 100644
--- a/drivers/clk/meson/Makefile
+++ b/drivers/clk/meson/Makefile
@@ -6,4 +6,5 @@
 obj-$(CONFIG_CLK_MESON_GX) += gxbb.o
 obj-$(CONFIG_CLK_MESON_AXG) += axg.o
 obj-$(CONFIG_CLK_MESON_G12A) += g12a.o
+obj-$(CONFIG_CLK_MESON_G12A) += g12a-ao.o
 
diff --git a/drivers/clk/meson/g12a-ao.c b/drivers/clk/meson/g12a-ao.c
new file mode 100644
index 0000000000..7a0abea77c
--- /dev/null
+++ b/drivers/clk/meson/g12a-ao.c
@@ -0,0 +1,83 @@
+// SPDX-License-Identifier: GPL-2.0+
+
+#include <common.h>
+#include <log.h>
+#include <asm/io.h>
+#include <clk-uclass.h>
+#include <dm.h>
+#include <regmap.h>
+#include <syscon.h>
+#include <dt-bindings/clock/g12a-aoclkc.h>
+
+#include "clk_meson.h"
+
+struct meson_clk {
+	struct regmap *map;
+};
+
+#define AO_CLK_GATE0		0x4c
+#define AO_SAR_CLK		0x90
+
+static struct meson_gate gates[] = {
+	MESON_GATE(CLKID_AO_SAR_ADC, AO_CLK_GATE0, 8),
+	MESON_GATE(CLKID_AO_SAR_ADC_CLK, AO_SAR_CLK, 8),
+};
+
+static int meson_set_gate(struct clk *clk, bool on)
+{
+	struct meson_clk *priv = dev_get_priv(clk->dev);
+	struct meson_gate *gate;
+
+	if (clk->id >= ARRAY_SIZE(gates))
+		return -ENOENT;
+
+	gate = &gates[clk->id];
+
+	if (gate->reg == 0)
+		return 0;
+
+	regmap_update_bits(priv->map, gate->reg,
+			   BIT(gate->bit), on ? BIT(gate->bit) : 0);
+
+	return 0;
+}
+
+static int meson_clk_enable(struct clk *clk)
+{
+	return meson_set_gate(clk, true);
+}
+
+static int meson_clk_disable(struct clk *clk)
+{
+	return meson_set_gate(clk, false);
+}
+
+static int meson_clk_probe(struct udevice *dev)
+{
+	struct meson_clk *priv = dev_get_priv(dev);
+
+	priv->map = syscon_node_to_regmap(dev_get_parent(dev)->node);
+	if (IS_ERR(priv->map))
+		return PTR_ERR(priv->map);
+
+	return 0;
+}
+
+static struct clk_ops meson_clk_ops = {
+	.disable	= meson_clk_disable,
+	.enable		= meson_clk_enable,
+};
+
+static const struct udevice_id meson_clk_ids[] = {
+	{ .compatible = "amlogic,meson-g12a-aoclkc" },
+	{ }
+};
+
+U_BOOT_DRIVER(meson_clk_axg) = {
+	.name		= "meson_clk_g12a_ao",
+	.id		= UCLASS_CLK,
+	.of_match	= meson_clk_ids,
+	.priv_auto_alloc_size = sizeof(struct meson_clk),
+	.ops		= &meson_clk_ops,
+	.probe		= meson_clk_probe,
+};
-- 
2.17.1

[PATCH v2 2/7] adc: meson-saradc: add G12A variant

[Marek Szyprowski]

Add support for the SARADC variant found on the G12A SoCs family.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Neil Armstrong <narmstrong@baylibre.com>
Tested-by: Jaehoon Chung <jh80.chung@samsung.com>
Reviewed-by: Jaehoon Chung <jh80.chung@samsung.com>
---
 drivers/adc/meson-saradc.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/adc/meson-saradc.c b/drivers/adc/meson-saradc.c
index 72b0cc4e5b..998cef24d8 100644
--- a/drivers/adc/meson-saradc.c
+++ b/drivers/adc/meson-saradc.c
@@ -711,6 +711,8 @@ static const struct udevice_id meson_saradc_ids[] = {
 	  .data = (ulong)&gxl_saradc_data },
 	{ .compatible = "amlogic,meson-gxm-saradc",
 	  .data = (ulong)&gxl_saradc_data },
+	{ .compatible = "amlogic,meson-g12a-saradc",
+	  .data = (ulong)&gxl_saradc_data },
 	{ }
 };
 
-- 
2.17.1

[PATCH v2 2/7] adc: meson-saradc: add G12A variant

[Marek Szyprowski]

Add support for the SARADC variant found on the G12A SoCs family.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Neil Armstrong <narmstrong@baylibre.com>
Tested-by: Jaehoon Chung <jh80.chung@samsung.com>
Reviewed-by: Jaehoon Chung <jh80.chung@samsung.com>
---
 drivers/adc/meson-saradc.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/adc/meson-saradc.c b/drivers/adc/meson-saradc.c
index 72b0cc4e5b..998cef24d8 100644
--- a/drivers/adc/meson-saradc.c
+++ b/drivers/adc/meson-saradc.c
@@ -711,6 +711,8 @@ static const struct udevice_id meson_saradc_ids[] = {
 	  .data = (ulong)&gxl_saradc_data },
 	{ .compatible = "amlogic,meson-gxm-saradc",
 	  .data = (ulong)&gxl_saradc_data },
+	{ .compatible = "amlogic,meson-g12a-saradc",
+	  .data = (ulong)&gxl_saradc_data },
 	{ }
 };
 
-- 
2.17.1

[PATCH v2 3/7] adc: meson-saradc: skip hardware init only if ADC is enabled

[Marek Szyprowski]

The driver skips hardware initialization if it is already configured by
the earlier bootloader stage (BL30). Skip the initialization only if the
hardware is really initialized and enabled.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Neil Armstrong <narmstrong@baylibre.com>
Tested-by: Jaehoon Chung <jh80.chung@samsung.com>
Reviewed-by: Jaehoon Chung <jh80.chung@samsung.com>
---
 drivers/adc/meson-saradc.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/adc/meson-saradc.c b/drivers/adc/meson-saradc.c
index 998cef24d8..ce7ae990ad 100644
--- a/drivers/adc/meson-saradc.c
+++ b/drivers/adc/meson-saradc.c
@@ -512,8 +512,11 @@ static int meson_saradc_init(struct meson_saradc_priv *priv)
 	 * reading the temperature sensor.
 	 */
 	regmap_read(priv->regmap, MESON_SAR_ADC_REG3, &regval);
-	if (regval & MESON_SAR_ADC_REG3_BL30_INITIALIZED)
-		return 0;
+	if (regval & MESON_SAR_ADC_REG3_BL30_INITIALIZED) {
+		regmap_read(priv->regmap, MESON_SAR_ADC_REG3, &regval);
+		if (regval & MESON_SAR_ADC_REG3_ADC_EN)
+			return 0;
+	}
 
 	meson_saradc_stop_sample_engine(priv);
 
-- 
2.17.1

[PATCH v2 3/7] adc: meson-saradc: skip hardware init only if ADC is enabled

[Marek Szyprowski]

The driver skips hardware initialization if it is already configured by
the earlier bootloader stage (BL30). Skip the initialization only if the
hardware is really initialized and enabled.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: Neil Armstrong <narmstrong@baylibre.com>
Tested-by: Jaehoon Chung <jh80.chung@samsung.com>
Reviewed-by: Jaehoon Chung <jh80.chung@samsung.com>
---
 drivers/adc/meson-saradc.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/adc/meson-saradc.c b/drivers/adc/meson-saradc.c
index 998cef24d8..ce7ae990ad 100644
--- a/drivers/adc/meson-saradc.c
+++ b/drivers/adc/meson-saradc.c
@@ -512,8 +512,11 @@ static int meson_saradc_init(struct meson_saradc_priv *priv)
 	 * reading the temperature sensor.
 	 */
 	regmap_read(priv->regmap, MESON_SAR_ADC_REG3, &regval);
-	if (regval & MESON_SAR_ADC_REG3_BL30_INITIALIZED)
-		return 0;
+	if (regval & MESON_SAR_ADC_REG3_BL30_INITIALIZED) {
+		regmap_read(priv->regmap, MESON_SAR_ADC_REG3, &regval);
+		if (regval & MESON_SAR_ADC_REG3_ADC_EN)
+			return 0;
+	}
 
 	meson_saradc_stop_sample_engine(priv);
 
-- 
2.17.1

[PATCH v2 4/7] configs: khadas-vim3: enable ADC device support

[Marek Szyprowski]

ADC device (Meson SARADC) will be used for probing 'Function' button state.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
 configs/khadas-vim3_defconfig | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/configs/khadas-vim3_defconfig b/configs/khadas-vim3_defconfig
index 9d7ba72d44..5d16652fd6 100644
--- a/configs/khadas-vim3_defconfig
+++ b/configs/khadas-vim3_defconfig
@@ -29,6 +29,8 @@ CONFIG_CMD_REGULATOR=y
 CONFIG_OF_CONTROL=y
 CONFIG_SYS_RELOC_GD_ENV_ADDR=y
 CONFIG_NET_RANDOM_ETHADDR=y
+CONFIG_ADC=y
+CONFIG_SARADC_MESON=y
 CONFIG_DM_I2C=y
 CONFIG_SYS_I2C_MESON=y
 CONFIG_DM_MMC=y
-- 
2.17.1

[PATCH v2 4/7] configs: khadas-vim3: enable ADC device support

[Marek Szyprowski]

ADC device (Meson SARADC) will be used for probing 'Function' button state.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
 configs/khadas-vim3_defconfig | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/configs/khadas-vim3_defconfig b/configs/khadas-vim3_defconfig
index 9d7ba72d44..5d16652fd6 100644
--- a/configs/khadas-vim3_defconfig
+++ b/configs/khadas-vim3_defconfig
@@ -29,6 +29,8 @@ CONFIG_CMD_REGULATOR=y
 CONFIG_OF_CONTROL=y
 CONFIG_SYS_RELOC_GD_ENV_ADDR=y
 CONFIG_NET_RANDOM_ETHADDR=y
+CONFIG_ADC=y
+CONFIG_SARADC_MESON=y
 CONFIG_DM_I2C=y
 CONFIG_SYS_I2C_MESON=y
 CONFIG_DM_MMC=y
-- 
2.17.1

[PATCH v2 5/7] button: add a simple ADC-based button driver

[Marek Szyprowski]

Add a simple ADC-based button driver. This driver binds to the 'adc-keys'
device tree node.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
 drivers/button/Kconfig      |   8 +++
 drivers/button/Makefile     |   1 +
 drivers/button/button-adc.c | 117 ++++++++++++++++++++++++++++++++++++
 3 files changed, 126 insertions(+)
 create mode 100644 drivers/button/button-adc.c

diff --git a/drivers/button/Kconfig b/drivers/button/Kconfig
index 6b3ec7e55d..283367f2bd 100644
--- a/drivers/button/Kconfig
+++ b/drivers/button/Kconfig
@@ -9,6 +9,14 @@ config BUTTON
 	  can provide access to board-specific buttons. Use of the device tree
 	  for configuration is encouraged.
 
+config BUTTON_ADC
+	bool "Button adc"
+	depends on BUTTON
+	help
+	  Enable support for buttons which are connected to ADC lines. The ADC
+	  driver must use driver model. Buttons are configured using the device
+	  tree.
+
 config BUTTON_GPIO
 	bool "Button gpio"
 	depends on BUTTON
diff --git a/drivers/button/Makefile b/drivers/button/Makefile
index fcc10ebe8d..bbd18af149 100644
--- a/drivers/button/Makefile
+++ b/drivers/button/Makefile
@@ -3,4 +3,5 @@
 # Copyright (C) 2020 Philippe Reynes <philippe.reynes@softathome.com>
 
 obj-$(CONFIG_BUTTON) += button-uclass.o
+obj-$(CONFIG_BUTTON_ADC) += button-adc.o
 obj-$(CONFIG_BUTTON_GPIO) += button-gpio.o
diff --git a/drivers/button/button-adc.c b/drivers/button/button-adc.c
new file mode 100644
index 0000000000..086c676c02
--- /dev/null
+++ b/drivers/button/button-adc.c
@@ -0,0 +1,117 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2020 Samsung Electronics Co., Ltd.
+ *		http://www.samsung.com
+ * Author: Marek Szyprowski <m.szyprowski@samsung.com>
+ */
+
+#include <common.h>
+#include <button.h>
+#include <dm.h>
+#include <dm/lists.h>
+#include <dm/uclass-internal.h>
+#include <log.h>
+#include <adc.h>
+
+struct button_adc_priv {
+	struct udevice *adc;
+	int channel;
+};
+
+static enum button_state_t button_adc_get_state(struct udevice *dev)
+{
+	struct button_adc_priv *priv = dev_get_priv(dev);
+	unsigned int val, mask;
+	int ret;
+
+	ret = adc_start_channel(priv->adc, priv->channel);
+	if (ret)
+		return ret;
+
+	ret = adc_channel_data(priv->adc, priv->channel, &val);
+	if (ret)
+		return ret;
+
+	ret = adc_data_mask(priv->adc, &mask);
+	if (ret)
+		return ret;
+
+	/* getting state is simplified a bit */
+	if (ret == 0)
+		return (val < mask / 2) ? BUTTON_ON : BUTTON_OFF;
+
+	return ret;
+}
+
+static int button_adc_probe(struct udevice *dev)
+{
+	struct button_uc_plat *uc_plat = dev_get_uclass_platdata(dev);
+	struct button_adc_priv *priv = dev_get_priv(dev);
+	struct ofnode_phandle_args args;
+	int ret;
+
+	/* Ignore the top-level button node */
+	if (!uc_plat->label)
+		return 0;
+
+	ret = dev_read_phandle_with_args(dev->parent, "io-channels",
+					 "#io-channel-cells", 0, 0, &args);
+	if (ret)
+		return ret;
+
+	ret = uclass_get_device_by_name(UCLASS_ADC, ofnode_get_name(args.node),
+					&priv->adc);
+	if (ret)
+		return ret;
+
+	priv->channel = args.args[0];
+
+	return ret;
+}
+
+static int button_adc_bind(struct udevice *parent)
+{
+	struct udevice *dev;
+	ofnode node;
+	int ret;
+
+	dev_for_each_subnode(node, parent) {
+		struct button_uc_plat *uc_plat;
+		const char *label;
+
+		label = ofnode_read_string(node, "label");
+		if (!label) {
+			debug("%s: node %s has no label\n", __func__,
+			      ofnode_get_name(node));
+			return -EINVAL;
+		}
+		ret = device_bind_driver_to_node(parent, "button_adc",
+						 ofnode_get_name(node),
+						 node, &dev);
+		if (ret)
+			return ret;
+		uc_plat = dev_get_uclass_platdata(dev);
+		uc_plat->label = label;
+	}
+
+	return 0;
+}
+
+static const struct button_ops button_adc_ops = {
+	.get_state	= button_adc_get_state,
+};
+
+static const struct udevice_id button_adc_ids[] = {
+	{ .compatible = "adc-keys" },
+	{ }
+};
+
+U_BOOT_DRIVER(button_adc) = {
+	.name		= "button_adc",
+	.id		= UCLASS_BUTTON,
+	.of_match	= button_adc_ids,
+	.ops		= &button_adc_ops,
+	.priv_auto_alloc_size = sizeof(struct button_adc_priv),
+	.bind		= button_adc_bind,
+	.probe		= button_adc_probe,
+};
-- 
2.17.1

[PATCH v2 5/7] button: add a simple ADC-based button driver

[Marek Szyprowski]

Add a simple ADC-based button driver. This driver binds to the 'adc-keys'
device tree node.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
 drivers/button/Kconfig      |   8 +++
 drivers/button/Makefile     |   1 +
 drivers/button/button-adc.c | 117 ++++++++++++++++++++++++++++++++++++
 3 files changed, 126 insertions(+)
 create mode 100644 drivers/button/button-adc.c

diff --git a/drivers/button/Kconfig b/drivers/button/Kconfig
index 6b3ec7e55d..283367f2bd 100644
--- a/drivers/button/Kconfig
+++ b/drivers/button/Kconfig
@@ -9,6 +9,14 @@ config BUTTON
 	  can provide access to board-specific buttons. Use of the device tree
 	  for configuration is encouraged.
 
+config BUTTON_ADC
+	bool "Button adc"
+	depends on BUTTON
+	help
+	  Enable support for buttons which are connected to ADC lines. The ADC
+	  driver must use driver model. Buttons are configured using the device
+	  tree.
+
 config BUTTON_GPIO
 	bool "Button gpio"
 	depends on BUTTON
diff --git a/drivers/button/Makefile b/drivers/button/Makefile
index fcc10ebe8d..bbd18af149 100644
--- a/drivers/button/Makefile
+++ b/drivers/button/Makefile
@@ -3,4 +3,5 @@
 # Copyright (C) 2020 Philippe Reynes <philippe.reynes@softathome.com>
 
 obj-$(CONFIG_BUTTON) += button-uclass.o
+obj-$(CONFIG_BUTTON_ADC) += button-adc.o
 obj-$(CONFIG_BUTTON_GPIO) += button-gpio.o
diff --git a/drivers/button/button-adc.c b/drivers/button/button-adc.c
new file mode 100644
index 0000000000..086c676c02
--- /dev/null
+++ b/drivers/button/button-adc.c
@@ -0,0 +1,117 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2020 Samsung Electronics Co., Ltd.
+ *		http://www.samsung.com
+ * Author: Marek Szyprowski <m.szyprowski@samsung.com>
+ */
+
+#include <common.h>
+#include <button.h>
+#include <dm.h>
+#include <dm/lists.h>
+#include <dm/uclass-internal.h>
+#include <log.h>
+#include <adc.h>
+
+struct button_adc_priv {
+	struct udevice *adc;
+	int channel;
+};
+
+static enum button_state_t button_adc_get_state(struct udevice *dev)
+{
+	struct button_adc_priv *priv = dev_get_priv(dev);
+	unsigned int val, mask;
+	int ret;
+
+	ret = adc_start_channel(priv->adc, priv->channel);
+	if (ret)
+		return ret;
+
+	ret = adc_channel_data(priv->adc, priv->channel, &val);
+	if (ret)
+		return ret;
+
+	ret = adc_data_mask(priv->adc, &mask);
+	if (ret)
+		return ret;
+
+	/* getting state is simplified a bit */
+	if (ret == 0)
+		return (val < mask / 2) ? BUTTON_ON : BUTTON_OFF;
+
+	return ret;
+}
+
+static int button_adc_probe(struct udevice *dev)
+{
+	struct button_uc_plat *uc_plat = dev_get_uclass_platdata(dev);
+	struct button_adc_priv *priv = dev_get_priv(dev);
+	struct ofnode_phandle_args args;
+	int ret;
+
+	/* Ignore the top-level button node */
+	if (!uc_plat->label)
+		return 0;
+
+	ret = dev_read_phandle_with_args(dev->parent, "io-channels",
+					 "#io-channel-cells", 0, 0, &args);
+	if (ret)
+		return ret;
+
+	ret = uclass_get_device_by_name(UCLASS_ADC, ofnode_get_name(args.node),
+					&priv->adc);
+	if (ret)
+		return ret;
+
+	priv->channel = args.args[0];
+
+	return ret;
+}
+
+static int button_adc_bind(struct udevice *parent)
+{
+	struct udevice *dev;
+	ofnode node;
+	int ret;
+
+	dev_for_each_subnode(node, parent) {
+		struct button_uc_plat *uc_plat;
+		const char *label;
+
+		label = ofnode_read_string(node, "label");
+		if (!label) {
+			debug("%s: node %s has no label\n", __func__,
+			      ofnode_get_name(node));
+			return -EINVAL;
+		}
+		ret = device_bind_driver_to_node(parent, "button_adc",
+						 ofnode_get_name(node),
+						 node, &dev);
+		if (ret)
+			return ret;
+		uc_plat = dev_get_uclass_platdata(dev);
+		uc_plat->label = label;
+	}
+
+	return 0;
+}
+
+static const struct button_ops button_adc_ops = {
+	.get_state	= button_adc_get_state,
+};
+
+static const struct udevice_id button_adc_ids[] = {
+	{ .compatible = "adc-keys" },
+	{ }
+};
+
+U_BOOT_DRIVER(button_adc) = {
+	.name		= "button_adc",
+	.id		= UCLASS_BUTTON,
+	.of_match	= button_adc_ids,
+	.ops		= &button_adc_ops,
+	.priv_auto_alloc_size = sizeof(struct button_adc_priv),
+	.bind		= button_adc_bind,
+	.probe		= button_adc_probe,
+};
-- 
2.17.1

[PATCH v2 6/7] cmd: button: store button state in the 'button' env

[Marek Szyprowski]

Save examined button state in 'button' environment variable to enable
checking button state in the scripts.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Change-Id: I78b539e1516573fcfea4401f75469291844daae4
---
 cmd/button.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/cmd/button.c b/cmd/button.c
index 64c5a8fa04..8da911068a 100644
--- a/cmd/button.c
+++ b/cmd/button.c
@@ -23,8 +23,10 @@ static int show_button_state(struct udevice *dev)
 	ret = button_get_state(dev);
 	if (ret >= BUTTON_COUNT)
 		ret = -EINVAL;
-	if (ret >= 0)
+	if (ret >= 0) {
 		printf("%s\n", state_label[ret]);
+		env_set("button", state_label[ret]);
+	}
 
 	return ret;
 }
-- 
2.17.1

[PATCH v2 6/7] cmd: button: store button state in the 'button' env

[Marek Szyprowski]

Save examined button state in 'button' environment variable to enable
checking button state in the scripts.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
Change-Id: I78b539e1516573fcfea4401f75469291844daae4
---
 cmd/button.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/cmd/button.c b/cmd/button.c
index 64c5a8fa04..8da911068a 100644
--- a/cmd/button.c
+++ b/cmd/button.c
@@ -23,8 +23,10 @@ static int show_button_state(struct udevice *dev)
 	ret = button_get_state(dev);
 	if (ret >= BUTTON_COUNT)
 		ret = -EINVAL;
-	if (ret >= 0)
+	if (ret >= 0) {
 		printf("%s\n", state_label[ret]);
+		env_set("button", state_label[ret]);
+	}
 
 	return ret;
 }
-- 
2.17.1

[PATCH v2 7/7] configs: khadas-vim3: enable Function button support

[Marek Szyprowski]

Add options needed to check the 'Function' button state.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
 configs/khadas-vim3_defconfig | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/configs/khadas-vim3_defconfig b/configs/khadas-vim3_defconfig
index 5d16652fd6..bc17430569 100644
--- a/configs/khadas-vim3_defconfig
+++ b/configs/khadas-vim3_defconfig
@@ -31,6 +31,8 @@ CONFIG_SYS_RELOC_GD_ENV_ADDR=y
 CONFIG_NET_RANDOM_ETHADDR=y
 CONFIG_ADC=y
 CONFIG_SARADC_MESON=y
+CONFIG_BUTTON=y
+CONFIG_BUTTON_ADC=y
 CONFIG_DM_I2C=y
 CONFIG_SYS_I2C_MESON=y
 CONFIG_DM_MMC=y
-- 
2.17.1

[PATCH v2 7/7] configs: khadas-vim3: enable Function button support

[Marek Szyprowski]

Add options needed to check the 'Function' button state.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
 configs/khadas-vim3_defconfig | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/configs/khadas-vim3_defconfig b/configs/khadas-vim3_defconfig
index 5d16652fd6..bc17430569 100644
--- a/configs/khadas-vim3_defconfig
+++ b/configs/khadas-vim3_defconfig
@@ -31,6 +31,8 @@ CONFIG_SYS_RELOC_GD_ENV_ADDR=y
 CONFIG_NET_RANDOM_ETHADDR=y
 CONFIG_ADC=y
 CONFIG_SARADC_MESON=y
+CONFIG_BUTTON=y
+CONFIG_BUTTON_ADC=y
 CONFIG_DM_I2C=y
 CONFIG_SYS_I2C_MESON=y
 CONFIG_DM_MMC=y
-- 
2.17.1

[PATCH v2 6/7 RESEND] cmd: button: store button state in the 'button' env

[Marek Szyprowski]

Save examined button state in 'button' environment variable to enable
checking button state in the scripts.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
Resend reason: get rid of the Change-Id tag, that was still in v2.
---
 cmd/button.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/cmd/button.c b/cmd/button.c
index 64c5a8fa04..8da911068a 100644
--- a/cmd/button.c
+++ b/cmd/button.c
@@ -23,8 +23,10 @@ static int show_button_state(struct udevice *dev)
 	ret = button_get_state(dev);
 	if (ret >= BUTTON_COUNT)
 		ret = -EINVAL;
-	if (ret >= 0)
+	if (ret >= 0) {
 		printf("%s\n", state_label[ret]);
+		env_set("button", state_label[ret]);
+	}
 
 	return ret;
 }
-- 
2.17.1

[PATCH v2 6/7 RESEND] cmd: button: store button state in the 'button' env

[Marek Szyprowski]

Save examined button state in 'button' environment variable to enable
checking button state in the scripts.

Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
---
Resend reason: get rid of the Change-Id tag, that was still in v2.
---
 cmd/button.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/cmd/button.c b/cmd/button.c
index 64c5a8fa04..8da911068a 100644
--- a/cmd/button.c
+++ b/cmd/button.c
@@ -23,8 +23,10 @@ static int show_button_state(struct udevice *dev)
 	ret = button_get_state(dev);
 	if (ret >= BUTTON_COUNT)
 		ret = -EINVAL;
-	if (ret >= 0)
+	if (ret >= 0) {
 		printf("%s\n", state_label[ret]);
+		env_set("button", state_label[ret]);
+	}
 
 	return ret;
 }
-- 
2.17.1

[PATCH v2 6/7 RESEND] cmd: button: store button state in the 'button' env

[Heinrich Schuchardt]

On 12/15/20 5:54 PM, Marek Szyprowski wrote:
> Save examined button state in 'button' environment variable to enable
> checking button state in the scripts.
>
> Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
> ---
> Resend reason: get rid of the Change-Id tag, that was still in v2.
> ---
>   cmd/button.c | 4 +++-
>   1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/cmd/button.c b/cmd/button.c
> index 64c5a8fa04..8da911068a 100644
> --- a/cmd/button.c
> +++ b/cmd/button.c
> @@ -23,8 +23,10 @@ static int show_button_state(struct udevice *dev)
>   	ret = button_get_state(dev);
>   	if (ret >= BUTTON_COUNT)
>   		ret = -EINVAL;
> -	if (ret >= 0)
> +	if (ret >= 0) {
>   		printf("%s\n", state_label[ret]);
> +		env_set("button", state_label[ret]);

Using a hard coded environment variable does not make much sense to me.
The button command has a return value. So just use

button mybutton; setenv myvar $?

Best regards

Heinrich

> +	}
>
>   	return ret;
>   }
>

Re: [PATCH v2 6/7 RESEND] cmd: button: store button state in the 'button' env

[Heinrich Schuchardt]

On 12/15/20 5:54 PM, Marek Szyprowski wrote:
> Save examined button state in 'button' environment variable to enable
> checking button state in the scripts.
>
> Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
> ---
> Resend reason: get rid of the Change-Id tag, that was still in v2.
> ---
>   cmd/button.c | 4 +++-
>   1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/cmd/button.c b/cmd/button.c
> index 64c5a8fa04..8da911068a 100644
> --- a/cmd/button.c
> +++ b/cmd/button.c
> @@ -23,8 +23,10 @@ static int show_button_state(struct udevice *dev)
>   	ret = button_get_state(dev);
>   	if (ret >= BUTTON_COUNT)
>   		ret = -EINVAL;
> -	if (ret >= 0)
> +	if (ret >= 0) {
>   		printf("%s\n", state_label[ret]);
> +		env_set("button", state_label[ret]);

Using a hard coded environment variable does not make much sense to me.
The button command has a return value. So just use

button mybutton; setenv myvar $?

Best regards

Heinrich

> +	}
>
>   	return ret;
>   }
>

[PATCH v2 5/7] button: add a simple ADC-based button driver

[Heinrich Schuchardt]

On 12/15/20 3:42 PM, Marek Szyprowski wrote:
> Add a simple ADC-based button driver. This driver binds to the 'adc-keys'
> device tree node.
>
> Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
> ---
>   drivers/button/Kconfig      |   8 +++
>   drivers/button/Makefile     |   1 +
>   drivers/button/button-adc.c | 117 ++++++++++++++++++++++++++++++++++++
>   3 files changed, 126 insertions(+)
>   create mode 100644 drivers/button/button-adc.c
>
> diff --git a/drivers/button/Kconfig b/drivers/button/Kconfig
> index 6b3ec7e55d..283367f2bd 100644
> --- a/drivers/button/Kconfig
> +++ b/drivers/button/Kconfig
> @@ -9,6 +9,14 @@ config BUTTON
>   	  can provide access to board-specific buttons. Use of the device tree
>   	  for configuration is encouraged.
>
> +config BUTTON_ADC
> +	bool "Button adc"
> +	depends on BUTTON
> +	help
> +	  Enable support for buttons which are connected to ADC lines. The ADC

ADC might stand for "Analog to Digital Converter", or "Automated Data
Collection", or "Automatic Distance Control, or something else. Please,
replace the abbreviation by the full expression.

Best regards

Heinrich

> +	  driver must use driver model. Buttons are configured using the device
> +	  tree.
> +
>   config BUTTON_GPIO
>   	bool "Button gpio"
>   	depends on BUTTON
> diff --git a/drivers/button/Makefile b/drivers/button/Makefile
> index fcc10ebe8d..bbd18af149 100644
> --- a/drivers/button/Makefile
> +++ b/drivers/button/Makefile
> @@ -3,4 +3,5 @@
>   # Copyright (C) 2020 Philippe Reynes <philippe.reynes@softathome.com>
>
>   obj-$(CONFIG_BUTTON) += button-uclass.o
> +obj-$(CONFIG_BUTTON_ADC) += button-adc.o
>   obj-$(CONFIG_BUTTON_GPIO) += button-gpio.o
> diff --git a/drivers/button/button-adc.c b/drivers/button/button-adc.c
> new file mode 100644
> index 0000000000..086c676c02
> --- /dev/null
> +++ b/drivers/button/button-adc.c
> @@ -0,0 +1,117 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * Copyright (C) 2020 Samsung Electronics Co., Ltd.
> + *		http://www.samsung.com
> + * Author: Marek Szyprowski <m.szyprowski@samsung.com>
> + */
> +
> +#include <common.h>
> +#include <button.h>
> +#include <dm.h>
> +#include <dm/lists.h>
> +#include <dm/uclass-internal.h>
> +#include <log.h>
> +#include <adc.h>
> +
> +struct button_adc_priv {
> +	struct udevice *adc;
> +	int channel;
> +};
> +
> +static enum button_state_t button_adc_get_state(struct udevice *dev)
> +{
> +	struct button_adc_priv *priv = dev_get_priv(dev);
> +	unsigned int val, mask;
> +	int ret;
> +
> +	ret = adc_start_channel(priv->adc, priv->channel);
> +	if (ret)
> +		return ret;
> +
> +	ret = adc_channel_data(priv->adc, priv->channel, &val);
> +	if (ret)
> +		return ret;
> +
> +	ret = adc_data_mask(priv->adc, &mask);
> +	if (ret)
> +		return ret;
> +
> +	/* getting state is simplified a bit */
> +	if (ret == 0)
> +		return (val < mask / 2) ? BUTTON_ON : BUTTON_OFF;
> +
> +	return ret;
> +}
> +
> +static int button_adc_probe(struct udevice *dev)
> +{
> +	struct button_uc_plat *uc_plat = dev_get_uclass_platdata(dev);
> +	struct button_adc_priv *priv = dev_get_priv(dev);
> +	struct ofnode_phandle_args args;
> +	int ret;
> +
> +	/* Ignore the top-level button node */
> +	if (!uc_plat->label)
> +		return 0;
> +
> +	ret = dev_read_phandle_with_args(dev->parent, "io-channels",
> +					 "#io-channel-cells", 0, 0, &args);
> +	if (ret)
> +		return ret;
> +
> +	ret = uclass_get_device_by_name(UCLASS_ADC, ofnode_get_name(args.node),
> +					&priv->adc);
> +	if (ret)
> +		return ret;
> +
> +	priv->channel = args.args[0];
> +
> +	return ret;
> +}
> +
> +static int button_adc_bind(struct udevice *parent)
> +{
> +	struct udevice *dev;
> +	ofnode node;
> +	int ret;
> +
> +	dev_for_each_subnode(node, parent) {
> +		struct button_uc_plat *uc_plat;
> +		const char *label;
> +
> +		label = ofnode_read_string(node, "label");
> +		if (!label) {
> +			debug("%s: node %s has no label\n", __func__,
> +			      ofnode_get_name(node));
> +			return -EINVAL;
> +		}
> +		ret = device_bind_driver_to_node(parent, "button_adc",
> +						 ofnode_get_name(node),
> +						 node, &dev);
> +		if (ret)
> +			return ret;
> +		uc_plat = dev_get_uclass_platdata(dev);
> +		uc_plat->label = label;
> +	}
> +
> +	return 0;
> +}
> +
> +static const struct button_ops button_adc_ops = {
> +	.get_state	= button_adc_get_state,
> +};
> +
> +static const struct udevice_id button_adc_ids[] = {
> +	{ .compatible = "adc-keys" },
> +	{ }
> +};
> +
> +U_BOOT_DRIVER(button_adc) = {
> +	.name		= "button_adc",
> +	.id		= UCLASS_BUTTON,
> +	.of_match	= button_adc_ids,
> +	.ops		= &button_adc_ops,
> +	.priv_auto_alloc_size = sizeof(struct button_adc_priv),
> +	.bind		= button_adc_bind,
> +	.probe		= button_adc_probe,
> +};
>

Re: [PATCH v2 5/7] button: add a simple ADC-based button driver

[Heinrich Schuchardt]

On 12/15/20 3:42 PM, Marek Szyprowski wrote:
> Add a simple ADC-based button driver. This driver binds to the 'adc-keys'
> device tree node.
>
> Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
> ---
>   drivers/button/Kconfig      |   8 +++
>   drivers/button/Makefile     |   1 +
>   drivers/button/button-adc.c | 117 ++++++++++++++++++++++++++++++++++++
>   3 files changed, 126 insertions(+)
>   create mode 100644 drivers/button/button-adc.c
>
> diff --git a/drivers/button/Kconfig b/drivers/button/Kconfig
> index 6b3ec7e55d..283367f2bd 100644
> --- a/drivers/button/Kconfig
> +++ b/drivers/button/Kconfig
> @@ -9,6 +9,14 @@ config BUTTON
>   	  can provide access to board-specific buttons. Use of the device tree
>   	  for configuration is encouraged.
>
> +config BUTTON_ADC
> +	bool "Button adc"
> +	depends on BUTTON
> +	help
> +	  Enable support for buttons which are connected to ADC lines. The ADC

ADC might stand for "Analog to Digital Converter", or "Automated Data
Collection", or "Automatic Distance Control, or something else. Please,
replace the abbreviation by the full expression.

Best regards

Heinrich

> +	  driver must use driver model. Buttons are configured using the device
> +	  tree.
> +
>   config BUTTON_GPIO
>   	bool "Button gpio"
>   	depends on BUTTON
> diff --git a/drivers/button/Makefile b/drivers/button/Makefile
> index fcc10ebe8d..bbd18af149 100644
> --- a/drivers/button/Makefile
> +++ b/drivers/button/Makefile
> @@ -3,4 +3,5 @@
>   # Copyright (C) 2020 Philippe Reynes <philippe.reynes@softathome.com>
>
>   obj-$(CONFIG_BUTTON) += button-uclass.o
> +obj-$(CONFIG_BUTTON_ADC) += button-adc.o
>   obj-$(CONFIG_BUTTON_GPIO) += button-gpio.o
> diff --git a/drivers/button/button-adc.c b/drivers/button/button-adc.c
> new file mode 100644
> index 0000000000..086c676c02
> --- /dev/null
> +++ b/drivers/button/button-adc.c
> @@ -0,0 +1,117 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * Copyright (C) 2020 Samsung Electronics Co., Ltd.
> + *		http://www.samsung.com
> + * Author: Marek Szyprowski <m.szyprowski@samsung.com>
> + */
> +
> +#include <common.h>
> +#include <button.h>
> +#include <dm.h>
> +#include <dm/lists.h>
> +#include <dm/uclass-internal.h>
> +#include <log.h>
> +#include <adc.h>
> +
> +struct button_adc_priv {
> +	struct udevice *adc;
> +	int channel;
> +};
> +
> +static enum button_state_t button_adc_get_state(struct udevice *dev)
> +{
> +	struct button_adc_priv *priv = dev_get_priv(dev);
> +	unsigned int val, mask;
> +	int ret;
> +
> +	ret = adc_start_channel(priv->adc, priv->channel);
> +	if (ret)
> +		return ret;
> +
> +	ret = adc_channel_data(priv->adc, priv->channel, &val);
> +	if (ret)
> +		return ret;
> +
> +	ret = adc_data_mask(priv->adc, &mask);
> +	if (ret)
> +		return ret;
> +
> +	/* getting state is simplified a bit */
> +	if (ret == 0)
> +		return (val < mask / 2) ? BUTTON_ON : BUTTON_OFF;
> +
> +	return ret;
> +}
> +
> +static int button_adc_probe(struct udevice *dev)
> +{
> +	struct button_uc_plat *uc_plat = dev_get_uclass_platdata(dev);
> +	struct button_adc_priv *priv = dev_get_priv(dev);
> +	struct ofnode_phandle_args args;
> +	int ret;
> +
> +	/* Ignore the top-level button node */
> +	if (!uc_plat->label)
> +		return 0;
> +
> +	ret = dev_read_phandle_with_args(dev->parent, "io-channels",
> +					 "#io-channel-cells", 0, 0, &args);
> +	if (ret)
> +		return ret;
> +
> +	ret = uclass_get_device_by_name(UCLASS_ADC, ofnode_get_name(args.node),
> +					&priv->adc);
> +	if (ret)
> +		return ret;
> +
> +	priv->channel = args.args[0];
> +
> +	return ret;
> +}
> +
> +static int button_adc_bind(struct udevice *parent)
> +{
> +	struct udevice *dev;
> +	ofnode node;
> +	int ret;
> +
> +	dev_for_each_subnode(node, parent) {
> +		struct button_uc_plat *uc_plat;
> +		const char *label;
> +
> +		label = ofnode_read_string(node, "label");
> +		if (!label) {
> +			debug("%s: node %s has no label\n", __func__,
> +			      ofnode_get_name(node));
> +			return -EINVAL;
> +		}
> +		ret = device_bind_driver_to_node(parent, "button_adc",
> +						 ofnode_get_name(node),
> +						 node, &dev);
> +		if (ret)
> +			return ret;
> +		uc_plat = dev_get_uclass_platdata(dev);
> +		uc_plat->label = label;
> +	}
> +
> +	return 0;
> +}
> +
> +static const struct button_ops button_adc_ops = {
> +	.get_state	= button_adc_get_state,
> +};
> +
> +static const struct udevice_id button_adc_ids[] = {
> +	{ .compatible = "adc-keys" },
> +	{ }
> +};
> +
> +U_BOOT_DRIVER(button_adc) = {
> +	.name		= "button_adc",
> +	.id		= UCLASS_BUTTON,
> +	.of_match	= button_adc_ids,
> +	.ops		= &button_adc_ops,
> +	.priv_auto_alloc_size = sizeof(struct button_adc_priv),
> +	.bind		= button_adc_bind,
> +	.probe		= button_adc_probe,
> +};
>

[PATCH v2 6/7 RESEND] cmd: button: store button state in the 'button' env

[Marek Szyprowski]

On 15.12.2020 20:07, Heinrich Schuchardt wrote:
> On 12/15/20 5:54 PM, Marek Szyprowski wrote:
>> Save examined button state in 'button' environment variable to enable
>> checking button state in the scripts.
>>
>> Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
>> ---
>> Resend reason: get rid of the Change-Id tag, that was still in v2.
>> ---
>> ? cmd/button.c | 4 +++-
>> ? 1 file changed, 3 insertions(+), 1 deletion(-)
>>
>> diff --git a/cmd/button.c b/cmd/button.c
>> index 64c5a8fa04..8da911068a 100644
>> --- a/cmd/button.c
>> +++ b/cmd/button.c
>> @@ -23,8 +23,10 @@ static int show_button_state(struct udevice *dev)
>> ????? ret = button_get_state(dev);
>> ????? if (ret >= BUTTON_COUNT)
>> ????????? ret = -EINVAL;
>> -??? if (ret >= 0)
>> +??? if (ret >= 0) {
>> ????????? printf("%s\n", state_label[ret]);
>> +??????? env_set("button", state_label[ret]);
>
> Using a hard coded environment variable does not make much sense to me.
> The button command has a return value. So just use
>
> button mybutton; setenv myvar $?
>
Thanks for the hint, I wasn't aware that uboot supports '$?'. By setting 
the 'button' env variable I've tried to mimic the behavior of the 
various network and file related commands, which sets 'filesize' env 
variable.

I will need to add the return value propagation to the button command 
anyway to make it usable from the scripts.

Best regards
-- 
Marek Szyprowski, PhD
Samsung R&D Institute Poland

Re: [PATCH v2 6/7 RESEND] cmd: button: store button state in the 'button' env

[Marek Szyprowski]

On 15.12.2020 20:07, Heinrich Schuchardt wrote:
> On 12/15/20 5:54 PM, Marek Szyprowski wrote:
>> Save examined button state in 'button' environment variable to enable
>> checking button state in the scripts.
>>
>> Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
>> ---
>> Resend reason: get rid of the Change-Id tag, that was still in v2.
>> ---
>>   cmd/button.c | 4 +++-
>>   1 file changed, 3 insertions(+), 1 deletion(-)
>>
>> diff --git a/cmd/button.c b/cmd/button.c
>> index 64c5a8fa04..8da911068a 100644
>> --- a/cmd/button.c
>> +++ b/cmd/button.c
>> @@ -23,8 +23,10 @@ static int show_button_state(struct udevice *dev)
>>       ret = button_get_state(dev);
>>       if (ret >= BUTTON_COUNT)
>>           ret = -EINVAL;
>> -    if (ret >= 0)
>> +    if (ret >= 0) {
>>           printf("%s\n", state_label[ret]);
>> +        env_set("button", state_label[ret]);
>
> Using a hard coded environment variable does not make much sense to me.
> The button command has a return value. So just use
>
> button mybutton; setenv myvar $?
>
Thanks for the hint, I wasn't aware that uboot supports '$?'. By setting 
the 'button' env variable I've tried to mimic the behavior of the 
various network and file related commands, which sets 'filesize' env 
variable.

I will need to add the return value propagation to the button command 
anyway to make it usable from the scripts.

Best regards
-- 
Marek Szyprowski, PhD
Samsung R&D Institute Poland

[PATCH v2 6/7 RESEND] cmd: button: store button state in the 'button' env

[Heinrich Schuchardt]

On 12/16/20 8:08 AM, Marek Szyprowski wrote:
> On 15.12.2020 20:07, Heinrich Schuchardt wrote:
>> On 12/15/20 5:54 PM, Marek Szyprowski wrote:
>>> Save examined button state in 'button' environment variable to enable
>>> checking button state in the scripts.
>>>
>>> Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
>>> ---
>>> Resend reason: get rid of the Change-Id tag, that was still in v2.
>>> ---
>>>  ? cmd/button.c | 4 +++-
>>>  ? 1 file changed, 3 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/cmd/button.c b/cmd/button.c
>>> index 64c5a8fa04..8da911068a 100644
>>> --- a/cmd/button.c
>>> +++ b/cmd/button.c
>>> @@ -23,8 +23,10 @@ static int show_button_state(struct udevice *dev)
>>>  ????? ret = button_get_state(dev);
>>>  ????? if (ret >= BUTTON_COUNT)
>>>  ????????? ret = -EINVAL;
>>> -??? if (ret >= 0)
>>> +??? if (ret >= 0) {
>>>  ????????? printf("%s\n", state_label[ret]);
>>> +??????? env_set("button", state_label[ret]);
>>
>> Using a hard coded environment variable does not make much sense to me.
>> The button command has a return value. So just use
>>
>> button mybutton; setenv myvar $?
>>
> Thanks for the hint, I wasn't aware that uboot supports '$?'. By setting
> the 'button' env variable I've tried to mimic the behavior of the
> various network and file related commands, which sets 'filesize' env
> variable.
>
> I will need to add the return value propagation to the button command
> anyway to make it usable from the scripts.

Nothing to be done for the button command. Since

     a6bfd71a96201127836d59736abcb54dc2d5e1a5
     cmd/button: return button status

the button command returns

     0 (true) - pressed, on
     1 (false) - not pressed, off

Best regards

Heinrich

Re: [PATCH v2 6/7 RESEND] cmd: button: store button state in the 'button' env

[Heinrich Schuchardt]

On 12/16/20 8:08 AM, Marek Szyprowski wrote:
> On 15.12.2020 20:07, Heinrich Schuchardt wrote:
>> On 12/15/20 5:54 PM, Marek Szyprowski wrote:
>>> Save examined button state in 'button' environment variable to enable
>>> checking button state in the scripts.
>>>
>>> Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
>>> ---
>>> Resend reason: get rid of the Change-Id tag, that was still in v2.
>>> ---
>>>    cmd/button.c | 4 +++-
>>>    1 file changed, 3 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/cmd/button.c b/cmd/button.c
>>> index 64c5a8fa04..8da911068a 100644
>>> --- a/cmd/button.c
>>> +++ b/cmd/button.c
>>> @@ -23,8 +23,10 @@ static int show_button_state(struct udevice *dev)
>>>        ret = button_get_state(dev);
>>>        if (ret >= BUTTON_COUNT)
>>>            ret = -EINVAL;
>>> -    if (ret >= 0)
>>> +    if (ret >= 0) {
>>>            printf("%s\n", state_label[ret]);
>>> +        env_set("button", state_label[ret]);
>>
>> Using a hard coded environment variable does not make much sense to me.
>> The button command has a return value. So just use
>>
>> button mybutton; setenv myvar $?
>>
> Thanks for the hint, I wasn't aware that uboot supports '$?'. By setting
> the 'button' env variable I've tried to mimic the behavior of the
> various network and file related commands, which sets 'filesize' env
> variable.
>
> I will need to add the return value propagation to the button command
> anyway to make it usable from the scripts.

Nothing to be done for the button command. Since

     a6bfd71a96201127836d59736abcb54dc2d5e1a5
     cmd/button: return button status

the button command returns

     0 (true) - pressed, on
     1 (false) - not pressed, off

Best regards

Heinrich

[PATCH v2 6/7 RESEND] cmd: button: store button state in the 'button' env

[Marek Szyprowski]

On 16.12.2020 08:29, Heinrich Schuchardt wrote:
> On 12/16/20 8:08 AM, Marek Szyprowski wrote:
>> On 15.12.2020 20:07, Heinrich Schuchardt wrote:
>>> On 12/15/20 5:54 PM, Marek Szyprowski wrote:
>>>> Save examined button state in 'button' environment variable to enable
>>>> checking button state in the scripts.
>>>>
>>>> Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
>>>> ---
>>>> Resend reason: get rid of the Change-Id tag, that was still in v2.
>>>> ---
>>>> ?? cmd/button.c | 4 +++-
>>>> ?? 1 file changed, 3 insertions(+), 1 deletion(-)
>>>>
>>>> diff --git a/cmd/button.c b/cmd/button.c
>>>> index 64c5a8fa04..8da911068a 100644
>>>> --- a/cmd/button.c
>>>> +++ b/cmd/button.c
>>>> @@ -23,8 +23,10 @@ static int show_button_state(struct udevice *dev)
>>>> ?????? ret = button_get_state(dev);
>>>> ?????? if (ret >= BUTTON_COUNT)
>>>> ?????????? ret = -EINVAL;
>>>> -??? if (ret >= 0)
>>>> +??? if (ret >= 0) {
>>>> ?????????? printf("%s\n", state_label[ret]);
>>>> +??????? env_set("button", state_label[ret]);
>>>
>>> Using a hard coded environment variable does not make much sense to me.
>>> The button command has a return value. So just use
>>>
>>> button mybutton; setenv myvar $?
>>>
>> Thanks for the hint, I wasn't aware that uboot supports '$?'. By setting
>> the 'button' env variable I've tried to mimic the behavior of the
>> various network and file related commands, which sets 'filesize' env
>> variable.
>>
>> I will need to add the return value propagation to the button command
>> anyway to make it usable from the scripts.
>
> Nothing to be done for the button command. Since
>
> ??? a6bfd71a96201127836d59736abcb54dc2d5e1a5
> ??? cmd/button: return button status
>
> the button command returns
>
> ??? 0 (true) - pressed, on
> ??? 1 (false) - not pressed, off
>
Right, I've missed that commit. I've developed my code on top of 
v2020.10 release.

Best regards
-- 
Marek Szyprowski, PhD
Samsung R&D Institute Poland

Re: [PATCH v2 6/7 RESEND] cmd: button: store button state in the 'button' env

[Marek Szyprowski]

On 16.12.2020 08:29, Heinrich Schuchardt wrote:
> On 12/16/20 8:08 AM, Marek Szyprowski wrote:
>> On 15.12.2020 20:07, Heinrich Schuchardt wrote:
>>> On 12/15/20 5:54 PM, Marek Szyprowski wrote:
>>>> Save examined button state in 'button' environment variable to enable
>>>> checking button state in the scripts.
>>>>
>>>> Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
>>>> ---
>>>> Resend reason: get rid of the Change-Id tag, that was still in v2.
>>>> ---
>>>>    cmd/button.c | 4 +++-
>>>>    1 file changed, 3 insertions(+), 1 deletion(-)
>>>>
>>>> diff --git a/cmd/button.c b/cmd/button.c
>>>> index 64c5a8fa04..8da911068a 100644
>>>> --- a/cmd/button.c
>>>> +++ b/cmd/button.c
>>>> @@ -23,8 +23,10 @@ static int show_button_state(struct udevice *dev)
>>>>        ret = button_get_state(dev);
>>>>        if (ret >= BUTTON_COUNT)
>>>>            ret = -EINVAL;
>>>> -    if (ret >= 0)
>>>> +    if (ret >= 0) {
>>>>            printf("%s\n", state_label[ret]);
>>>> +        env_set("button", state_label[ret]);
>>>
>>> Using a hard coded environment variable does not make much sense to me.
>>> The button command has a return value. So just use
>>>
>>> button mybutton; setenv myvar $?
>>>
>> Thanks for the hint, I wasn't aware that uboot supports '$?'. By setting
>> the 'button' env variable I've tried to mimic the behavior of the
>> various network and file related commands, which sets 'filesize' env
>> variable.
>>
>> I will need to add the return value propagation to the button command
>> anyway to make it usable from the scripts.
>
> Nothing to be done for the button command. Since
>
>     a6bfd71a96201127836d59736abcb54dc2d5e1a5
>     cmd/button: return button status
>
> the button command returns
>
>     0 (true) - pressed, on
>     1 (false) - not pressed, off
>
Right, I've missed that commit. I've developed my code on top of 
v2020.10 release.

Best regards
-- 
Marek Szyprowski, PhD
Samsung R&D Institute Poland

[PATCH v2 6/7 RESEND] cmd: button: store button state in the 'button' env

[Simon Glass]

Hi Marek,

On Tue, 15 Dec 2020 at 09:55, Marek Szyprowski <m.szyprowski@samsung.com> wrote:
>
> Save examined button state in 'button' environment variable to enable
> checking button state in the scripts.
>
> Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
> ---
> Resend reason: get rid of the Change-Id tag, that was still in v2.

If you use patman it does that for you.

Regards,
Simon

Re: [PATCH v2 6/7 RESEND] cmd: button: store button state in the 'button' env

[Simon Glass]

Hi Marek,

On Tue, 15 Dec 2020 at 09:55, Marek Szyprowski <m.szyprowski@samsung.com> wrote:
>
> Save examined button state in 'button' environment variable to enable
> checking button state in the scripts.
>
> Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
> ---
> Resend reason: get rid of the Change-Id tag, that was still in v2.

If you use patman it does that for you.

Regards,
Simon