All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH v3 0/2] TWL6030, TWL6032 GPADC driver
@ 2013-07-12  7:18 ` Oleksandr Kozaruk
  0 siblings, 0 replies; 45+ messages in thread
From: Oleksandr Kozaruk @ 2013-07-12  7:18 UTC (permalink / raw)
  To: tony, benoit.cousson, rnayak, peter.ujfalusi, kishon, jic23,
	grant.likely, rob.herring, lars, sameo, ch.naveen, poeschel,
	milo.kim, balajitk, gg
  Cc: linux-arm-kernel, linux-kernel, linux-iio, devicetree-discuss,
	linux-omap

Hello,

v3 - fixed compiler warning
v2 - the driver put in drivers/iio, and
converted using iio facilities as suggested by Graeme.

TWL603[02] GPADC is used to measure battery voltage,
battery temperature, battery presence ID, and could
be used to measure twl603[02] die temperature.
This is used on TI blaze, blaze tablet platforms.

The TWL6030/TWL6032 is a PMIC that has a GPADC with 17/19
channels respectively. Some channels have current
source and are used for measuring voltage drop
on resistive load for detecting battery ID resistance,
or measuring voltage drop on NTC resistors for external
temperature measurements, other channels measure voltage,
(i.e. battery voltage), and have inbuilt voltage dividers,
thus, capable to scale voltage. Some channels are dedicated
for measuring die temperature.

Some channels could be calibrated in 2 points, having
offsets from ideal values in trim registers.

The difference between GPADC in TWL6030 and TWL6032:
- 10 bit vs 12 bit ADC;
- 17 vs 19 channels;
- channels have different purpose(i. e. battery voltage
  channel 8 vs channel 18);
- trim values are interpreted differently.

The driver exports function returning converted value for
requested channels.

Sysfs entries are added to start and read conversion result
in millivolts for chosen channel.

The driver is derived from git://git.omapzoom.org/kernel/omap.git
The original driver's authors and contributors are Balaji T K,
Graeme Gregory, Ambresh K, Girish S Ghongdemath.

The changes to the original driver:
- device tree adaptation;
- drop ioctl support - never been used;
- unified measurement method for both devices;
- get rid of "if (device == X)" code style to data driven;
- drop polling end of conversion and use interrupt instead;
- iio framework is used

Tested with on blaze tablet 2 with OMAP4430(twl6030), and
OMAP4470(twl6032) SOMs.

The patches were tested against 3.10-rc7

Oleksandr Kozaruk (2):
  ARM: dts: twl: Add GPADC data to device tree
  iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver

 arch/arm/boot/dts/twl6030.dtsi  |    6 +
 drivers/iio/adc/Kconfig         |    8 +
 drivers/iio/adc/Makefile        |    1 +
 drivers/iio/adc/twl6030-gpadc.c | 1019 +++++++++++++++++++++++++++++++++++++++
 4 files changed, 1034 insertions(+)
 create mode 100644 drivers/iio/adc/twl6030-gpadc.c

-- 
1.7.9.5


^ permalink raw reply	[flat|nested] 45+ messages in thread

* [PATCH v3 0/2] TWL6030, TWL6032 GPADC driver
@ 2013-07-12  7:18 ` Oleksandr Kozaruk
  0 siblings, 0 replies; 45+ messages in thread
From: Oleksandr Kozaruk @ 2013-07-12  7:18 UTC (permalink / raw)
  To: tony-4v6yS6AI5VpBDgjK7y7TUQ,
	benoit.cousson-QSEj5FYQhm4dnm+yROfE0A, rnayak-l0cyMroinI0,
	peter.ujfalusi-l0cyMroinI0, kishon-l0cyMroinI0,
	jic23-KWPb1pKIrIJaa/9Udqfwiw,
	grant.likely-QSEj5FYQhm4dnm+yROfE0A,
	rob.herring-bsGFqQB8/DxBDgjK7y7TUQ, lars-Qo5EllUWu/uELgA04lAiVw,
	sameo-VuQAYsv1563Yd54FQh9/CA, ch.naveen-Sze3O3UU22JBDgjK7y7TUQ,
	poeschel-Xtl8qvBWbHwb1SvskN2V4Q, milo.kim-l0cyMroinI0,
	balajitk-l0cyMroinI0, gg-kDsPt+C1G03kYMGBc/C6ZA
  Cc: linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA,
	linux-iio-u79uwXL29TY76Z2rM5mHXA,
	devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ,
	linux-omap-u79uwXL29TY76Z2rM5mHXA

Hello,

v3 - fixed compiler warning
v2 - the driver put in drivers/iio, and
converted using iio facilities as suggested by Graeme.

TWL603[02] GPADC is used to measure battery voltage,
battery temperature, battery presence ID, and could
be used to measure twl603[02] die temperature.
This is used on TI blaze, blaze tablet platforms.

The TWL6030/TWL6032 is a PMIC that has a GPADC with 17/19
channels respectively. Some channels have current
source and are used for measuring voltage drop
on resistive load for detecting battery ID resistance,
or measuring voltage drop on NTC resistors for external
temperature measurements, other channels measure voltage,
(i.e. battery voltage), and have inbuilt voltage dividers,
thus, capable to scale voltage. Some channels are dedicated
for measuring die temperature.

Some channels could be calibrated in 2 points, having
offsets from ideal values in trim registers.

The difference between GPADC in TWL6030 and TWL6032:
- 10 bit vs 12 bit ADC;
- 17 vs 19 channels;
- channels have different purpose(i. e. battery voltage
  channel 8 vs channel 18);
- trim values are interpreted differently.

The driver exports function returning converted value for
requested channels.

Sysfs entries are added to start and read conversion result
in millivolts for chosen channel.

The driver is derived from git://git.omapzoom.org/kernel/omap.git
The original driver's authors and contributors are Balaji T K,
Graeme Gregory, Ambresh K, Girish S Ghongdemath.

The changes to the original driver:
- device tree adaptation;
- drop ioctl support - never been used;
- unified measurement method for both devices;
- get rid of "if (device == X)" code style to data driven;
- drop polling end of conversion and use interrupt instead;
- iio framework is used

Tested with on blaze tablet 2 with OMAP4430(twl6030), and
OMAP4470(twl6032) SOMs.

The patches were tested against 3.10-rc7

Oleksandr Kozaruk (2):
  ARM: dts: twl: Add GPADC data to device tree
  iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver

 arch/arm/boot/dts/twl6030.dtsi  |    6 +
 drivers/iio/adc/Kconfig         |    8 +
 drivers/iio/adc/Makefile        |    1 +
 drivers/iio/adc/twl6030-gpadc.c | 1019 +++++++++++++++++++++++++++++++++++++++
 4 files changed, 1034 insertions(+)
 create mode 100644 drivers/iio/adc/twl6030-gpadc.c

-- 
1.7.9.5

^ permalink raw reply	[flat|nested] 45+ messages in thread

* [PATCH v3 0/2] TWL6030, TWL6032 GPADC driver
@ 2013-07-12  7:18 ` Oleksandr Kozaruk
  0 siblings, 0 replies; 45+ messages in thread
From: Oleksandr Kozaruk @ 2013-07-12  7:18 UTC (permalink / raw)
  To: linux-arm-kernel

Hello,

v3 - fixed compiler warning
v2 - the driver put in drivers/iio, and
converted using iio facilities as suggested by Graeme.

TWL603[02] GPADC is used to measure battery voltage,
battery temperature, battery presence ID, and could
be used to measure twl603[02] die temperature.
This is used on TI blaze, blaze tablet platforms.

The TWL6030/TWL6032 is a PMIC that has a GPADC with 17/19
channels respectively. Some channels have current
source and are used for measuring voltage drop
on resistive load for detecting battery ID resistance,
or measuring voltage drop on NTC resistors for external
temperature measurements, other channels measure voltage,
(i.e. battery voltage), and have inbuilt voltage dividers,
thus, capable to scale voltage. Some channels are dedicated
for measuring die temperature.

Some channels could be calibrated in 2 points, having
offsets from ideal values in trim registers.

The difference between GPADC in TWL6030 and TWL6032:
- 10 bit vs 12 bit ADC;
- 17 vs 19 channels;
- channels have different purpose(i. e. battery voltage
  channel 8 vs channel 18);
- trim values are interpreted differently.

The driver exports function returning converted value for
requested channels.

Sysfs entries are added to start and read conversion result
in millivolts for chosen channel.

The driver is derived from git://git.omapzoom.org/kernel/omap.git
The original driver's authors and contributors are Balaji T K,
Graeme Gregory, Ambresh K, Girish S Ghongdemath.

The changes to the original driver:
- device tree adaptation;
- drop ioctl support - never been used;
- unified measurement method for both devices;
- get rid of "if (device == X)" code style to data driven;
- drop polling end of conversion and use interrupt instead;
- iio framework is used

Tested with on blaze tablet 2 with OMAP4430(twl6030), and
OMAP4470(twl6032) SOMs.

The patches were tested against 3.10-rc7

Oleksandr Kozaruk (2):
  ARM: dts: twl: Add GPADC data to device tree
  iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver

 arch/arm/boot/dts/twl6030.dtsi  |    6 +
 drivers/iio/adc/Kconfig         |    8 +
 drivers/iio/adc/Makefile        |    1 +
 drivers/iio/adc/twl6030-gpadc.c | 1019 +++++++++++++++++++++++++++++++++++++++
 4 files changed, 1034 insertions(+)
 create mode 100644 drivers/iio/adc/twl6030-gpadc.c

-- 
1.7.9.5

^ permalink raw reply	[flat|nested] 45+ messages in thread

* [PATCH v3 1/2] ARM: dts: twl: Add GPADC data to device tree
  2013-07-12  7:18 ` Oleksandr Kozaruk
  (?)
@ 2013-07-12  7:18   ` Oleksandr Kozaruk
  -1 siblings, 0 replies; 45+ messages in thread
From: Oleksandr Kozaruk @ 2013-07-12  7:18 UTC (permalink / raw)
  To: tony, benoit.cousson, rnayak, peter.ujfalusi, kishon, jic23,
	grant.likely, rob.herring, lars, sameo, ch.naveen, poeschel,
	milo.kim, balajitk, gg
  Cc: linux-arm-kernel, linux-kernel, linux-iio, devicetree-discuss,
	linux-omap

GPADC is the general purpose ADC present on twl6030.
The dt data is interrupt used to trigger end of ADC
conversion.

Signed-off-by: Oleksandr Kozaruk <oleksandr.kozaruk@ti.com>
---
 arch/arm/boot/dts/twl6030.dtsi |    6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/arch/arm/boot/dts/twl6030.dtsi b/arch/arm/boot/dts/twl6030.dtsi
index 2e3bd31..434842c 100644
--- a/arch/arm/boot/dts/twl6030.dtsi
+++ b/arch/arm/boot/dts/twl6030.dtsi
@@ -103,4 +103,10 @@
 		compatible = "ti,twl6030-pwmled";
 		#pwm-cells = <2>;
 	};
+
+	gpadc: twl6030_gpadc {
+		compatible = "ti,twl6030_gpadc";
+		interrupts = <3>;
+		#io-channel-cells = <1>;
+	};
 };
-- 
1.7.9.5


^ permalink raw reply related	[flat|nested] 45+ messages in thread

* [PATCH v3 1/2] ARM: dts: twl: Add GPADC data to device tree
@ 2013-07-12  7:18   ` Oleksandr Kozaruk
  0 siblings, 0 replies; 45+ messages in thread
From: Oleksandr Kozaruk @ 2013-07-12  7:18 UTC (permalink / raw)
  To: tony, benoit.cousson, rnayak, peter.ujfalusi, kishon, jic23,
	grant.likely, rob.herring, lars, sameo, ch.naveen, poeschel,
	milo.kim, balajitk, gg
  Cc: linux-arm-kernel, linux-kernel, linux-iio, devicetree-discuss,
	linux-omap

GPADC is the general purpose ADC present on twl6030.
The dt data is interrupt used to trigger end of ADC
conversion.

Signed-off-by: Oleksandr Kozaruk <oleksandr.kozaruk@ti.com>
---
 arch/arm/boot/dts/twl6030.dtsi |    6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/arch/arm/boot/dts/twl6030.dtsi b/arch/arm/boot/dts/twl6030.dtsi
index 2e3bd31..434842c 100644
--- a/arch/arm/boot/dts/twl6030.dtsi
+++ b/arch/arm/boot/dts/twl6030.dtsi
@@ -103,4 +103,10 @@
 		compatible = "ti,twl6030-pwmled";
 		#pwm-cells = <2>;
 	};
+
+	gpadc: twl6030_gpadc {
+		compatible = "ti,twl6030_gpadc";
+		interrupts = <3>;
+		#io-channel-cells = <1>;
+	};
 };
-- 
1.7.9.5

^ permalink raw reply related	[flat|nested] 45+ messages in thread

* [PATCH v3 1/2] ARM: dts: twl: Add GPADC data to device tree
@ 2013-07-12  7:18   ` Oleksandr Kozaruk
  0 siblings, 0 replies; 45+ messages in thread
From: Oleksandr Kozaruk @ 2013-07-12  7:18 UTC (permalink / raw)
  To: linux-arm-kernel

GPADC is the general purpose ADC present on twl6030.
The dt data is interrupt used to trigger end of ADC
conversion.

Signed-off-by: Oleksandr Kozaruk <oleksandr.kozaruk@ti.com>
---
 arch/arm/boot/dts/twl6030.dtsi |    6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/arch/arm/boot/dts/twl6030.dtsi b/arch/arm/boot/dts/twl6030.dtsi
index 2e3bd31..434842c 100644
--- a/arch/arm/boot/dts/twl6030.dtsi
+++ b/arch/arm/boot/dts/twl6030.dtsi
@@ -103,4 +103,10 @@
 		compatible = "ti,twl6030-pwmled";
 		#pwm-cells = <2>;
 	};
+
+	gpadc: twl6030_gpadc {
+		compatible = "ti,twl6030_gpadc";
+		interrupts = <3>;
+		#io-channel-cells = <1>;
+	};
 };
-- 
1.7.9.5

^ permalink raw reply related	[flat|nested] 45+ messages in thread

* [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
  2013-07-12  7:18 ` Oleksandr Kozaruk
  (?)
@ 2013-07-12  7:18   ` Oleksandr Kozaruk
  -1 siblings, 0 replies; 45+ messages in thread
From: Oleksandr Kozaruk @ 2013-07-12  7:18 UTC (permalink / raw)
  To: tony, benoit.cousson, rnayak, peter.ujfalusi, kishon, jic23,
	grant.likely, rob.herring, lars, sameo, ch.naveen, poeschel,
	milo.kim, balajitk, gg
  Cc: linux-arm-kernel, linux-kernel, linux-iio, devicetree-discuss,
	linux-omap

The GPADC is general purpose ADC found on TWL6030,
and TWL6032 PMIC, known also as Phoenix and PhoenixLite.

The TWL6030 and TWL6032 have GPADC with 17 and 19
channels respectively. Some channels have current
source and are used for measuring voltage drop
on resistive load for detecting battery ID resistance,
or measuring voltage drop on NTC resistors for external
temperature measurements. Some channels measure voltage,
(i.e. battery voltage), and have voltage dividers,
thus, capable to scale voltage. Some channels are dedicated
for measuring die temperature.

Some channels are calibrated in 2 points, having
offsets from ideal values kept in trim registers. This
is used to correct measurements.

The differences between GPADC in TWL6030 and TWL6032:
- 10 bit vs 12 bit ADC;
- 17 vs 19 channels;
- channels have different purpose(i. e. battery voltage
  channel 8 vs channel 18);
- trim values are interpreted differently.

The driver exports function returning converted value for
requested channels.

Based on the driver patched from Balaji TK, Graeme Gregory, Ambresh K,
Girish S Ghongdemath.

Signed-off-by: Balaji T K <balajitk@ti.com>
Graeme Gregory <gg@slimlogic.co.uk>
Signed-off-by: Oleksandr Kozaruk <oleksandr.kozaruk@ti.com>
---
 drivers/iio/adc/Kconfig         |    8 +
 drivers/iio/adc/Makefile        |    1 +
 drivers/iio/adc/twl6030-gpadc.c | 1019 +++++++++++++++++++++++++++++++++++++++
 3 files changed, 1028 insertions(+)
 create mode 100644 drivers/iio/adc/twl6030-gpadc.c

diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig
index ab0767e6..87d699e 100644
--- a/drivers/iio/adc/Kconfig
+++ b/drivers/iio/adc/Kconfig
@@ -157,4 +157,12 @@ config VIPERBOARD_ADC
 	  Say yes here to access the ADC part of the Nano River
 	  Technologies Viperboard.
 
+config TWL6030_GPADC
+	tristate "TWL6030 GPADC (General Purpose A/D Convertor) Support"
+	depends on TWL4030_CORE
+	default n
+	help
+	  Say yes here if you want support for the TWL6030 General Purpose
+	  A/D Convertor.
+
 endmenu
diff --git a/drivers/iio/adc/Makefile b/drivers/iio/adc/Makefile
index 0a825be..8b05633 100644
--- a/drivers/iio/adc/Makefile
+++ b/drivers/iio/adc/Makefile
@@ -17,3 +17,4 @@ obj-$(CONFIG_MAX1363) += max1363.o
 obj-$(CONFIG_TI_ADC081C) += ti-adc081c.o
 obj-$(CONFIG_TI_AM335X_ADC) += ti_am335x_adc.o
 obj-$(CONFIG_VIPERBOARD_ADC) += viperboard_adc.o
+obj-$(CONFIG_TWL6030_GPADC) += twl6030-gpadc.o
diff --git a/drivers/iio/adc/twl6030-gpadc.c b/drivers/iio/adc/twl6030-gpadc.c
new file mode 100644
index 0000000..6ceb789
--- /dev/null
+++ b/drivers/iio/adc/twl6030-gpadc.c
@@ -0,0 +1,1019 @@
+/*
+ * TWL6030 GPADC module driver
+ *
+ * Copyright (C) 2009-2013 Texas Instruments Inc.
+ * Nishant Kamat <nskamat@ti.com>
+ * Balaji T K <balajitk@ti.com>
+ * Graeme Gregory <gg@slimlogic.co.uk>
+ * Girish S Ghongdemath <girishsg@ti.com>
+ * Ambresh K <ambresh@ti.com>
+ * Oleksandr Kozaruk <oleksandr.kozaruk@ti.com
+ *
+ * Based on twl4030-madc.c
+ * Copyright (C) 2008 Nokia Corporation
+ * Mikko Ylinen <mikko.k.ylinen@nokia.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA
+ * 02110-1301 USA
+ *
+ */
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/of_platform.h>
+#include <linux/i2c/twl.h>
+#include <linux/iio/iio.h>
+#include <linux/iio/sysfs.h>
+
+#define DRIVER_NAME		"twl6030_gpadc"
+
+#define TWL6030_GPADC_MAX_CHANNELS 17
+#define TWL6032_GPADC_MAX_CHANNELS 19
+/* Define this as the biggest of all chips using this driver */
+#define GPADC_MAX_CHANNELS TWL6032_GPADC_MAX_CHANNELS
+
+#define TWL6030_GPADC_CTRL			0x00    /* 0x2e */
+#define TWL6030_GPADC_CTRL2			0x01	/* 0x2f */
+
+#define TWL6030_GPADC_CTRL_P1			0x05
+#define TWL6030_GPADC_CTRL_P2			0x06
+
+#define TWL6032_GPADC_GPSELECT_ISB		0x07
+#define TWL6032_GPADC_CTRL_P1			0x08
+
+#define TWL6032_GPADC_RTCH0_LSB			0x09
+#define TWL6032_GPADC_RTCH0_MSB			0x0a
+#define TWL6032_GPADC_RTCH1_LSB			0x0b
+#define TWL6032_GPADC_RTCH1_MSB			0x0c
+#define TWL6032_GPADC_GPCH0_LSB			0x0d
+#define TWL6032_GPADC_GPCH0_MSB			0x0e
+
+#define TWL6030_GPADC_CTRL_P1_SP1		BIT(3)
+#define TWL6030_GPADC_CTRL_P1_EOCRT		BIT(2)
+#define TWL6030_GPADC_CTRL_P1_EOCP1		BIT(1)
+#define TWL6030_GPADC_CTRL_P1_BUSY		BIT(0)
+
+#define TWL6030_GPADC_CTRL_P2_SP2		BIT(2)
+#define TWL6030_GPADC_CTRL_P2_EOCP2		BIT(1)
+#define TWL6030_GPADC_CTRL_P1_BUSY		BIT(0)
+
+#define TWL6030_GPADC_EOC_SW			BIT(1)
+#define TWL6030_GPADC_BUSY			BIT(0)
+
+#define TWL6030_GPADC_RTCH0_LSB			(0x07)
+#define TWL6030_GPADC_GPCH0_LSB			(0x29)
+
+/* Fixed channels */
+#define TWL6030_GPADC_CTRL_TEMP1_EN		BIT(0)    /* input ch 1 */
+#define TWL6030_GPADC_CTRL_TEMP2_EN		BIT(1)    /* input ch 4 */
+#define TWL6030_GPADC_CTRL_SCALER_EN		BIT(2)    /* input ch 2 */
+#define TWL6030_GPADC_CTRL_SCALER_DIV4		BIT(3)
+#define TWL6030_GPADC_CTRL_SCALER_EN_CH11	BIT(4)    /* input ch 11 */
+#define TWL6030_GPADC_CTRL_TEMP1_EN_MONITOR	BIT(5)
+#define TWL6030_GPADC_CTRL_TEMP2_EN_MONITOR	BIT(6)
+#define TWL6030_GPADC_CTRL_ISOURCE_EN		BIT(7)
+
+#define TWL6030_GPADC_CTRL2_REMSENSE_0		BIT(0)
+#define TWL6030_GPADC_CTRL2_REMSENSE_1		BIT(1)
+#define TWL6030_GPADC_CTRL2_SCALER_EN_CH18	BIT(2)
+#define TWL6030_GPADC_CTRL2_VBAT_SCALER_DIV4	BIT(3)
+
+#define TWL6030_GPADC_RT_SW1_EOC_MASK		BIT(5)
+#define TWL6030_GPADC_SW2_EOC_MASK		BIT(6)
+
+#define TWL6032_GPADC_RT_EOC_MASK		BIT(4)
+#define TWL6032_GPADC_SW_EOC_MASK		BIT(5)
+
+#define TWL6030_GPADC_TRIM1			0xCD
+
+#define TWL6030_REG_TOGGLE1			0x90
+#define TWL6030_GPADCS				BIT(1)
+#define TWL6030_GPADCR				BIT(0)
+
+/**
+ * struct twl6030_chnl_calib - channel calibration
+ * @gain:		slope coefficient for ideal curve
+ * @gain_error:		gain error
+ * @offset_error:	offset of the real curve
+ */
+struct twl6030_chnl_calib {
+	s32 gain;
+	s32 gain_error;
+	s32 offset_error;
+};
+
+/**
+ * struct twl6030_ideal_code - GPADC calibration parameters
+ * GPADC is calibrated in two points: close to the beginning and
+ * to the and of the measurable input range
+ *
+ * @code1:	ideal code for the input at the beginning
+ * @code2:	ideal code for at the end of the range
+ * @v1:		voltage input at the beginning(low voltage)
+ * @v2:		voltage input at the end(high voltage)
+ */
+struct twl6030_ideal_code {
+	u16 code1;
+	u16 code2;
+	u16 v1;
+	u16 v2;
+};
+
+struct twl6030_gpadc_data;
+
+/**
+ * struct twl6030_gpadc_platform_data - platform specific data
+ * @nchannels:		number of GPADC channels
+ * @iio_channels:	iio channels
+ * @twl6030_ideal:	pointer to calibration parameters
+ * @start_conversion:	pointer to ADC start conversion function
+ * @channel_to_reg	pointer to ADC function to convert channel to
+ * 			register address for reading conversion result
+ * @calibrate:		pointer to calibration function
+ */
+struct twl6030_gpadc_platform_data {
+	const int nchannels;
+	const struct iio_chan_spec *iio_channels;
+	const struct twl6030_ideal_code *ideal;
+	int (*start_conversion)(int channel);
+	u8 (*channel_to_reg)(int channel);
+	int (*calibrate)(struct twl6030_gpadc_data *gpadc);
+};
+
+/**
+ * struct twl6030_gpadc_data - GPADC data
+ * @dev:		device pointer
+ * @lock:		mutual exclusion lock for the structure
+ * @irq_complete:	completion to signal end of conversion
+ * @twl6030_cal_tbl:	pointer to calibration data for each
+ * 			channel with gain error and offset
+ * @pdata:		pointer to device specific data
+ */
+struct twl6030_gpadc_data {
+	struct device	*dev;
+	struct mutex	lock;
+	struct completion	irq_complete;
+	struct twl6030_chnl_calib	*twl6030_cal_tbl;
+	const struct twl6030_gpadc_platform_data *pdata;
+};
+
+/*
+ * channels 11, 12, 13, 15 and 16 have no calibration data
+ * calibration offset is same for channels 1, 3, 4, 5
+ *
+ * The data is taken from GPADC_TRIM registers description.
+ * GPADC_TRIM registers keeps difference between the code measured
+ * at v1 and v2 input voltages and corresponding code1 and code2
+ */
+static const struct twl6030_ideal_code
+	twl6030_ideal[TWL6030_GPADC_MAX_CHANNELS] = {
+	{ /* ch 0, external, battery type, resistor value */
+		.code1 = 116,
+		.code2 = 745,
+		.v1 = 141,
+		.v2 = 910,
+	},
+	{ /* ch 1, external, battery temperature, NTC resistor value */
+		.code1 = 82,
+		.code2 = 900,
+		.v1 = 100,
+		.v2 = 1100,
+	},
+	{ /* ch 2, external, audio accessory/general purpose */
+		.code1 = 55,
+		.code2 = 818,
+		.v1 = 101,
+		.v2 = 1499,
+	},
+	{ /* ch 3, external, general purpose */
+		.code1 = 82,
+		.code2 = 900,
+		.v1 = 100,
+		.v2 = 1100,
+	},
+	{ /* ch 4, external, temperature measurement/general purpose */
+		.code1 = 82,
+		.code2 = 900,
+		.v1 = 100,
+		.v2 = 1100,
+	},
+	{ /* ch 5, external, general purpose */
+		.code1 = 82,
+		.code2 = 900,
+		.v1 = 100,
+		.v2 = 1100,
+	},
+	{ /* ch 6, external, general purpose */
+		.code1 = 82,
+		.code2 = 900,
+		.v1 = 100,
+		.v2 = 1100,
+	},
+	{ /* ch 7, internal, main battery */
+		.code1 = 614,
+		.code2 = 941,
+		.v1 = 3001,
+		.v2 = 4599,
+	},
+	{ /* ch 8, internal, backup battery */
+		.code1 = 82,
+		.code2 = 688,
+		.v1 = 501,
+		.v2 = 4203,
+	},
+	{ /* ch 9, internal, external charger input */
+		.code1 = 182,
+		.code2 = 818,
+		.v1 = 2001,
+		.v2 = 8996,
+	},
+	{ /* ch 10, internal, VBUS */
+		.code1 = 149,
+		.code2 = 818,
+		.v1 = 1001,
+		.v2 = 5497,
+	},
+	{},	/* ch 11, internal, VBUS charging current */
+	{},	/* ch 12, internal, Die temperature */
+	{},	/* ch 13, internal, Die temperature */
+	{ /* ch 14, internal, USB ID line */
+		.code1 = 48,
+		.code2 = 714,
+		.v1 = 323,
+		.v2 = 4800,
+	},
+	{},	/* ch 15, internal, test network */
+	{},	/* ch 16, internal, test network */
+};
+
+static const struct twl6030_ideal_code
+			twl6032_ideal[TWL6032_GPADC_MAX_CHANNELS] = {
+	{ /* ch 0, external, battery type, resistor value */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 1, external, battery temperature, NTC resistor value */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 2, external, audio accessory/general purpose */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 660,
+		.v2 = 1500,
+	},
+	{ /* ch 3, external, temperature with external diode/general purpose */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 4, external, temperature measurement/general purpose */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 5, external, general purpose */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 6, external, general purpose */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch7, internal, system supply */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 2200,
+		.v2 = 5000,
+	},
+	{ /* ch8, internal, backup battery */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 2200,
+		.v2 = 5000,
+	},
+	{ /* ch 9, internal, external charger input */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 3960,
+		.v2 = 9000,
+	},
+	{ /* ch10, internal, VBUS */
+		.code1 = 150,
+		.code2 = 751,
+		.v1 = 1000,
+		.v2 = 5000,
+	},
+	{ /* ch 11, internal, VBUS DC-DC output current */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 660,
+		.v2 = 1500,
+	},
+	{ /* ch 12, internal, Die temperature */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 13, internal, Die temperature */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 14, internal, USB ID line */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 2420,
+		.v2 = 5500,
+	},
+	{},	/* ch 15, internal, test network */
+	{},	/* ch 16, internal, test network */
+	{},	/* ch 17, internal, battery charging current */
+	{ /* ch 18, internal, battery voltage */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 2200,
+		.v2 = 5000,
+	},
+};
+
+static inline int twl6030_gpadc_write(u8 reg, u8 val)
+{
+	return twl_i2c_write_u8(TWL6030_MODULE_GPADC, val, reg);
+}
+
+static inline int twl6030_gpadc_read(u8 *val, u8 reg)
+{
+
+	return twl_i2c_read(TWL6030_MODULE_GPADC, val, reg, 2);
+}
+
+static int twl6030_gpadc_enable_irq(u8 mask)
+{
+	int ret;
+
+	ret = twl6030_interrupt_unmask(mask, REG_INT_MSK_LINE_B);
+	ret |= twl6030_interrupt_unmask(mask, REG_INT_MSK_STS_B);
+
+	return ret;
+}
+
+static void twl6030_gpadc_disable_irq(u8 mask)
+{
+	twl6030_interrupt_mask(mask, REG_INT_MSK_LINE_B);
+	twl6030_interrupt_mask(mask, REG_INT_MSK_STS_B);
+}
+
+static irqreturn_t twl6030_gpadc_irq_handler(int irq, void *_gpadc)
+{
+	struct twl6030_gpadc_data *gpadc = _gpadc;
+
+	complete(&gpadc->irq_complete);
+
+	return IRQ_HANDLED;
+}
+
+static int twl6030_start_conversion(int channel)
+{
+	return twl6030_gpadc_write(TWL6030_GPADC_CTRL_P1,
+					TWL6030_GPADC_CTRL_P1_SP1);
+}
+
+static int twl6032_start_conversion(int channel)
+{
+	int ret;
+
+	ret = twl6030_gpadc_write(TWL6032_GPADC_GPSELECT_ISB, channel);
+	if (ret)
+		return ret;
+
+	return twl6030_gpadc_write(TWL6032_GPADC_CTRL_P1,
+						TWL6030_GPADC_CTRL_P1_SP1);
+}
+
+static u8 twl6030_channel_to_reg(int channel)
+{
+	return TWL6030_GPADC_GPCH0_LSB + 2 * channel;
+}
+
+static u8 twl6032_channel_to_reg(int channel)
+{
+	return TWL6032_GPADC_GPCH0_LSB;
+}
+
+static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,
+		int channel, int *res)
+{
+	u8 reg = gpadc->pdata->channel_to_reg(channel);
+	u8 val[2];
+	int ret;
+
+	ret = twl6030_gpadc_read(val, reg);
+	if (ret) {
+		dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);
+		return ret;
+	}
+
+	*res = (val[1] << 8) | val[0];
+
+	return ret;
+}
+
+static int twl6030_gpadc_get_processed(struct twl6030_gpadc_data *gpadc,
+		int channel, int *val)
+{
+	int raw_code;
+	int corrected_code;
+	int channel_value;
+	int ret;
+
+	ret = twl6030_gpadc_get_raw(gpadc, channel, &raw_code);
+	if (ret)
+		return ret;
+
+	corrected_code = ((raw_code * 1000) -
+		gpadc->twl6030_cal_tbl[channel].offset_error) /
+		gpadc->twl6030_cal_tbl[channel].gain_error;
+
+	channel_value = corrected_code *
+			gpadc->twl6030_cal_tbl[channel].gain;
+
+	/* Shift back into mV range */
+	channel_value /= 1000;
+
+	dev_dbg(gpadc->dev, "GPADC raw: %d", raw_code);
+	dev_dbg(gpadc->dev, "GPADC cor: %d", corrected_code);
+	dev_dbg(gpadc->dev, "GPADC val: %d", channel_value);
+
+	*val = channel_value;
+
+	return ret;
+}
+
+static int twl6030_gpadc_read_raw(struct iio_dev *indio_dev,
+			     const struct iio_chan_spec *chan,
+			     int *val, int *val2, long mask)
+{
+	struct twl6030_gpadc_data *gpadc = iio_priv(indio_dev);
+	int ret = -EINVAL;
+
+	mutex_lock(&gpadc->lock);
+
+	ret = gpadc->pdata->start_conversion(chan->channel);
+	if (ret) {
+		dev_err(gpadc->dev, "failed to start conversion\n");
+		goto err;
+	}
+	/* wait for conversion to complete */
+	wait_for_completion_interruptible_timeout(&gpadc->irq_complete,
+						msecs_to_jiffies(5000));
+
+	switch (mask) {
+	case IIO_CHAN_INFO_RAW:
+		ret = twl6030_gpadc_get_raw(gpadc, chan->channel, val);
+		ret = ret ? -EIO : IIO_VAL_INT;
+		break;
+
+	case IIO_CHAN_INFO_PROCESSED:
+		ret = twl6030_gpadc_get_processed(gpadc, chan->channel, val);
+		ret = ret ? -EIO : IIO_VAL_INT;
+		break;
+
+	default:
+		break;
+	}
+err:
+	mutex_unlock(&gpadc->lock);
+
+	return ret;
+}
+
+static void twl6030_calibrate_channel(struct twl6030_gpadc_data *gpadc,
+					int channel, int d1, int d2)
+{
+	int b, k, gain, x1, x2;
+	const struct twl6030_ideal_code *ideal = gpadc->pdata->ideal;
+
+	/* Gain */
+	gain = ((ideal[channel].v2 - ideal[channel].v1) * 1000) /
+		(ideal[channel].code2 - ideal[channel].code1);
+
+	x1 = ideal[channel].code1;
+	x2 = ideal[channel].code2;
+
+	/* k */
+	k = 1000 + (((d2 - d1) * 1000) / (x2 - x1));
+
+	/* b */
+	b = (d1 * 1000) - (k - 1000) * x1;
+
+	gpadc->twl6030_cal_tbl[channel].gain = gain;
+	gpadc->twl6030_cal_tbl[channel].gain_error = k;
+	gpadc->twl6030_cal_tbl[channel].offset_error = b;
+
+	dev_dbg(gpadc->dev, "GPADC d1   for Chn: %d = %d\n", channel, d1);
+	dev_dbg(gpadc->dev, "GPADC d2   for Chn: %d = %d\n", channel, d2);
+	dev_dbg(gpadc->dev, "GPADC x1   for Chn: %d = %d\n", channel, x1);
+	dev_dbg(gpadc->dev, "GPADC x2   for Chn: %d = %d\n", channel, x2);
+	dev_dbg(gpadc->dev, "GPADC Gain for Chn: %d = %d\n", channel, gain);
+	dev_dbg(gpadc->dev, "GPADC k    for Chn: %d = %d\n", channel, k);
+	dev_dbg(gpadc->dev, "GPADC b    for Chn: %d = %d\n", channel, b);
+}
+
+static inline int twl6030_gpadc_get_trim_offset(s8 d)
+{
+	int sign;
+
+	/*
+	 * XXX NOTE!
+	 * bit 0 - sign, bit 7 - reserved, 6..1 - trim value
+	 * though, the documentation states that trim value
+	 * is absolute value, the correct conversion results are
+	 * obtained if the value is interpreted as 2's complement.
+	 */
+	sign = d & 1;
+	d = (d & 0x7f) >> 1;
+
+	return sign ? (d | 0xc0) : d;
+}
+
+static int twl6030_calibration(struct twl6030_gpadc_data *gpadc)
+{
+	int ret;
+	int chn;
+	u8 trim_regs[TWL6030_GPADC_MAX_CHANNELS];
+	s8 d1, d2;
+
+	/*
+	 * for calibration two measurements have been performed at
+	 * factory, for some channels, during the production test and
+	 * have been stored in registers. This two stored values are
+	 * used to correct the measurements. The values represent
+	 * offsets for the given input from the output on ideal curve.
+	 */
+	ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs,
+			TWL6030_GPADC_TRIM1, 16);
+	if (ret < 0) {
+		dev_err(gpadc->dev, "calibration failed\n");
+		return ret;
+	}
+
+	for (chn = 0; chn < TWL6030_GPADC_MAX_CHANNELS; chn++) {
+
+		switch (chn) {
+		case 0:
+			d1 = trim_regs[0];
+			d2 = trim_regs[1];
+			break;
+		case 1:
+			d1 = trim_regs[4];
+			d2 = trim_regs[5];
+			break;
+		case 2:
+			d1 = trim_regs[12];
+			d2 = trim_regs[13];
+			break;
+		case 3:
+		case 4:
+		case 5:
+		case 6:
+			d1 = trim_regs[4];
+			d2 = trim_regs[5];
+			break;
+		case 7:
+			d1 = trim_regs[6];
+			d2 = trim_regs[7];
+			break;
+		case 8:
+			d1 = trim_regs[2];
+			d2 = trim_regs[3];
+			break;
+		case 9:
+			d1 = trim_regs[8];
+			d2 = trim_regs[9];
+			break;
+		case 10:
+			d1 = trim_regs[10];
+			d2 = trim_regs[11];
+			break;
+		case 14:
+			d1 = trim_regs[14];
+			d2 = trim_regs[15];
+			break;
+		default:
+			continue;
+		}
+
+		d1 = twl6030_gpadc_get_trim_offset(d1);
+		d2 = twl6030_gpadc_get_trim_offset(d2);
+
+		twl6030_calibrate_channel(gpadc, chn, d1, d2);
+	}
+
+	return 0;
+}
+
+static int twl6032_calibration(struct twl6030_gpadc_data *gpadc)
+{
+	int chn, d1 = 0, d2 = 0, temp;
+	u8 trim_regs[17];
+	int ret;
+
+	ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs + 1,
+			TWL6030_GPADC_TRIM1, 16);
+	if (ret < 0) {
+		dev_err(gpadc->dev, "calibration failed\n");
+		return ret;
+	}
+
+	/* Loop to calculate the value needed for returning voltages from
+	 * GPADC not values.
+	 *
+	 * gain is calculated to 3 decimal places fixed point.
+	 */
+	for (chn = 0; chn < TWL6032_GPADC_MAX_CHANNELS; chn++) {
+
+		switch (chn) {
+		case 0:
+		case 1:
+		case 2:
+		case 3:
+		case 4:
+		case 5:
+		case 6:
+		case 11:
+		case 12:
+		case 13:
+		case 14:
+			/* D1 */
+			d1 = (trim_regs[3] & 0x1F) << 2;
+			d1 |= (trim_regs[1] & 0x06) >> 1;
+			if (trim_regs[1] & 0x01)
+				d1 = -d1;
+
+			/* D2 */
+			d2 = (trim_regs[4] & 0x3F) << 2;
+			d2 |= (trim_regs[2] & 0x06) >> 1;
+			if (trim_regs[2] & 0x01)
+				d2 = -d2;
+			break;
+		case 8:
+			/* D1 */
+			temp = (trim_regs[3] & 0x1F) << 2;
+			temp |= (trim_regs[1] & 0x06) >> 1;
+			if (trim_regs[1] & 0x01)
+				temp = -temp;
+
+			d1 = (trim_regs[8] & 0x18) << 1;
+			d1 |= (trim_regs[7] & 0x1E) >> 1;
+			if (trim_regs[7] & 0x01)
+				d1 = -d1;
+
+			d1 += temp;
+
+			/* D2 */
+			temp = (trim_regs[4] & 0x3F) << 2;
+			temp |= (trim_regs[2] & 0x06) >> 1;
+			if (trim_regs[2] & 0x01)
+				temp = -temp;
+
+			d2 = (trim_regs[10] & 0x1F) << 2;
+			d2 |= (trim_regs[8] & 0x06) >> 1;
+			if (trim_regs[8] & 0x01)
+				d2 = -d2;
+
+			d2 += temp;
+			break;
+		case 9:
+			/* D1 */
+			temp = (trim_regs[3] & 0x1F) << 2;
+			temp |= (trim_regs[1] & 0x06) >> 1;
+			if (trim_regs[1] & 0x01)
+				temp = -temp;
+
+			d1 = (trim_regs[14] & 0x18) << 1;
+			d1 |= (trim_regs[12] & 0x1E) >> 1;
+			if (trim_regs[12] & 0x01)
+				d1 = -d1;
+
+			d1 += temp;
+
+			/* D2 */
+			temp = (trim_regs[4] & 0x3F) << 2;
+			temp |= (trim_regs[2] & 0x06) >> 1;
+			if (trim_regs[2] & 0x01)
+				temp = -temp;
+
+			d2 = (trim_regs[16] & 0x1F) << 2;
+			d2 |= (trim_regs[14] & 0x06) >> 1;
+			if (trim_regs[14] & 0x01)
+				d2 = -d2;
+
+			d2 += temp;
+		case 10:
+			/* D1 */
+			d1 = (trim_regs[11] & 0x0F) << 3;
+			d1 |= (trim_regs[9] & 0x0E) >> 1;
+			if (trim_regs[9] & 0x01)
+				d1 = -d1;
+
+			/* D2 */
+			d2 = (trim_regs[15] & 0x0F) << 3;
+			d2 |= (trim_regs[13] & 0x0E) >> 1;
+			if (trim_regs[13] & 0x01)
+				d2 = -d2;
+			break;
+		case 7:
+		case 18:
+			/* D1 */
+			temp = (trim_regs[3] & 0x1F) << 2;
+			temp |= (trim_regs[1] & 0x06) >> 1;
+			if (trim_regs[1] & 0x01)
+				temp = -temp;
+
+			d1 = (trim_regs[5] & 0x7E) >> 1;
+			if (trim_regs[5] & 0x01)
+				d1 = -d1;
+
+			d1 += temp;
+
+			/* D2 */
+			temp = (trim_regs[4] & 0x3F) << 2;
+			temp |= (trim_regs[2] & 0x06) >> 1;
+			if (trim_regs[2] & 0x01)
+				temp = -temp;
+
+			d2 = (trim_regs[6] & 0xFE) >> 1;
+			if (trim_regs[6] & 0x01)
+				d2 = -d2;
+
+			d2 += temp;
+			break;
+		default:
+			/* No data for other channels */
+			continue;
+		}
+
+		twl6030_calibrate_channel(gpadc, chn, d1, d2);
+	}
+
+	return 0;
+}
+
+#define TWL6030_GPADC_CHAN(chn, chan_info) {		\
+	.type = IIO_VOLTAGE,				\
+	.channel = chn,					\
+	.scan_type = {					\
+		.sign = 'u',				\
+		.storagebits = 16,			\
+	},						\
+	.info_mask = BIT(chan_info),			\
+	.info_mask_separate = BIT(chan_info), 		\
+	.indexed = 1,					\
+}
+
+static const struct iio_chan_spec twl6030_gpadc_iio_channels[] = {
+	TWL6030_GPADC_CHAN(0, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(1, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(2, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(3, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(4, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(5, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(6, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(7, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(8, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(9, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(10, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(11, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(12, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(13, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(14, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(15, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(16, IIO_CHAN_INFO_RAW),
+};
+
+static const struct iio_chan_spec twl6032_gpadc_iio_channels[] = {
+	TWL6030_GPADC_CHAN(0, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(1, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(2, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(3, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(4, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(5, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(6, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(7, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(8, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(9, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(10, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(11, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(12, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(13, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(14, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(15, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(16, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(17, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(18, IIO_CHAN_INFO_PROCESSED),
+};
+
+static const struct iio_info twl6030_gpadc_iio_info = {
+	.read_raw = &twl6030_gpadc_read_raw,
+	.driver_module = THIS_MODULE,
+};
+
+static const struct twl6030_gpadc_platform_data twl6030_pdata = {
+	.iio_channels = twl6030_gpadc_iio_channels,
+	.nchannels = TWL6030_GPADC_MAX_CHANNELS,
+	.ideal = twl6030_ideal,
+	.start_conversion = twl6030_start_conversion,
+	.channel_to_reg = twl6030_channel_to_reg,
+	.calibrate = twl6030_calibration,
+};
+
+static const struct twl6030_gpadc_platform_data twl6032_pdata = {
+	.iio_channels = twl6032_gpadc_iio_channels,
+	.nchannels = TWL6032_GPADC_MAX_CHANNELS,
+	.ideal = twl6032_ideal,
+	.start_conversion = twl6032_start_conversion,
+	.channel_to_reg = twl6032_channel_to_reg,
+	.calibrate = twl6032_calibration,
+};
+
+static const struct of_device_id of_twl6030_match_tbl[] = {
+	{
+		.compatible = "ti,twl6030_gpadc",
+		.data = &twl6030_pdata,
+	},
+	{
+		.compatible = "ti,twl6032_gpadc",
+		.data = &twl6032_pdata,
+	},
+	{ /* end */ }
+};
+
+static int twl6030_gpadc_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct twl6030_gpadc_data *gpadc;
+	const struct twl6030_gpadc_platform_data *pdata;
+	const struct of_device_id *match;
+	struct iio_dev *indio_dev;
+	int irq;
+	int ret = 0;
+
+	match = of_match_device(of_match_ptr(of_twl6030_match_tbl), dev);
+	pdata = match ? match->data : dev->platform_data;
+
+	if (!pdata)
+		return -EINVAL;
+
+	indio_dev = iio_device_alloc(sizeof(struct twl6030_gpadc_data));
+	if (!indio_dev) {
+		dev_err(dev, "failed allocating iio device\n");
+		ret = -ENOMEM;
+	}
+
+	gpadc = iio_priv(indio_dev);
+
+	gpadc->twl6030_cal_tbl = devm_kzalloc(dev,
+					sizeof(struct twl6030_chnl_calib) *
+					pdata->nchannels, GFP_KERNEL);
+	if (!gpadc->twl6030_cal_tbl)
+		goto err;
+
+	gpadc->dev = dev;
+	gpadc->pdata = pdata;
+
+	platform_set_drvdata(pdev, indio_dev);
+	mutex_init(&gpadc->lock);
+	init_completion(&gpadc->irq_complete);
+
+	ret = pdata->calibrate(gpadc);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to read calibration registers\n");
+		goto err;
+	}
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(&pdev->dev, "failed to get irq\n");
+		goto err;
+	}
+
+	ret = devm_request_threaded_irq(dev, irq, NULL,
+					twl6030_gpadc_irq_handler,
+					IRQF_ONESHOT, "twl6030_gpadc", gpadc);
+	if (ret) {
+		dev_dbg(&pdev->dev, "could not request irq\n");
+		goto err;
+	}
+
+	ret = twl6030_gpadc_enable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to enable GPADC interrupt\n");
+		goto err;
+	}
+	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
+					TWL6030_REG_TOGGLE1);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to enable GPADC module\n");
+		goto err;
+	}
+
+	indio_dev->name = DRIVER_NAME;
+	indio_dev->dev.parent = dev;
+	indio_dev->info = &twl6030_gpadc_iio_info;
+	indio_dev->modes = INDIO_DIRECT_MODE;
+	indio_dev->channels = pdata->iio_channels;
+	indio_dev->num_channels = pdata->nchannels;
+
+	ret = iio_device_register(indio_dev);
+	if (ret)
+		goto err;
+
+	return ret;
+err:
+	iio_device_free(indio_dev);
+	return ret;
+}
+
+static int twl6030_gpadc_remove(struct platform_device *pdev)
+{
+	struct iio_dev *indio_dev = platform_get_drvdata(pdev);
+
+	twl6030_gpadc_disable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
+	iio_device_unregister(indio_dev);
+	iio_device_free(indio_dev);
+
+	return 0;
+}
+
+static int twl6030_gpadc_suspend(struct device *pdev)
+{
+	int ret;
+
+	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCR,
+				TWL6030_REG_TOGGLE1);
+	if (ret)
+		dev_err(pdev, "error reseting GPADC (%d)!\n", ret);
+
+	return 0;
+};
+
+static int twl6030_gpadc_resume(struct device *pdev)
+{
+	int ret;
+
+	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
+				TWL6030_REG_TOGGLE1);
+	if (ret)
+		dev_err(pdev, "error setting GPADC (%d)!\n", ret);
+
+	return 0;
+};
+
+static const struct dev_pm_ops twl6030_gpadc_pm_ops = {
+	.suspend = twl6030_gpadc_suspend,
+	.resume = twl6030_gpadc_resume,
+};
+
+static struct platform_driver twl6030_gpadc_driver = {
+	.probe		= twl6030_gpadc_probe,
+	.remove		= twl6030_gpadc_remove,
+	.driver		= {
+		.name	= DRIVER_NAME,
+		.owner	= THIS_MODULE,
+		.pm	= &twl6030_gpadc_pm_ops,
+		.of_match_table = of_twl6030_match_tbl,
+	},
+};
+
+module_platform_driver(twl6030_gpadc_driver);
+
+MODULE_ALIAS("platform: " DRIVER_NAME);
+MODULE_AUTHOR("Texas Instruments Inc.");
+MODULE_DESCRIPTION("twl6030 ADC driver");
+MODULE_LICENSE("GPL");
-- 
1.7.9.5


^ permalink raw reply related	[flat|nested] 45+ messages in thread

* [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-12  7:18   ` Oleksandr Kozaruk
  0 siblings, 0 replies; 45+ messages in thread
From: Oleksandr Kozaruk @ 2013-07-12  7:18 UTC (permalink / raw)
  To: tony, benoit.cousson, rnayak, peter.ujfalusi, kishon, jic23,
	grant.likely, rob.herring, lars, sameo, ch.naveen, poeschel,
	milo.kim, balajitk, gg
  Cc: linux-iio, devicetree-discuss, linux-omap, linux-kernel,
	linux-arm-kernel

The GPADC is general purpose ADC found on TWL6030,
and TWL6032 PMIC, known also as Phoenix and PhoenixLite.

The TWL6030 and TWL6032 have GPADC with 17 and 19
channels respectively. Some channels have current
source and are used for measuring voltage drop
on resistive load for detecting battery ID resistance,
or measuring voltage drop on NTC resistors for external
temperature measurements. Some channels measure voltage,
(i.e. battery voltage), and have voltage dividers,
thus, capable to scale voltage. Some channels are dedicated
for measuring die temperature.

Some channels are calibrated in 2 points, having
offsets from ideal values kept in trim registers. This
is used to correct measurements.

The differences between GPADC in TWL6030 and TWL6032:
- 10 bit vs 12 bit ADC;
- 17 vs 19 channels;
- channels have different purpose(i. e. battery voltage
  channel 8 vs channel 18);
- trim values are interpreted differently.

The driver exports function returning converted value for
requested channels.

Based on the driver patched from Balaji TK, Graeme Gregory, Ambresh K,
Girish S Ghongdemath.

Signed-off-by: Balaji T K <balajitk@ti.com>
Graeme Gregory <gg@slimlogic.co.uk>
Signed-off-by: Oleksandr Kozaruk <oleksandr.kozaruk@ti.com>
---
 drivers/iio/adc/Kconfig         |    8 +
 drivers/iio/adc/Makefile        |    1 +
 drivers/iio/adc/twl6030-gpadc.c | 1019 +++++++++++++++++++++++++++++++++++++++
 3 files changed, 1028 insertions(+)
 create mode 100644 drivers/iio/adc/twl6030-gpadc.c

diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig
index ab0767e6..87d699e 100644
--- a/drivers/iio/adc/Kconfig
+++ b/drivers/iio/adc/Kconfig
@@ -157,4 +157,12 @@ config VIPERBOARD_ADC
 	  Say yes here to access the ADC part of the Nano River
 	  Technologies Viperboard.
 
+config TWL6030_GPADC
+	tristate "TWL6030 GPADC (General Purpose A/D Convertor) Support"
+	depends on TWL4030_CORE
+	default n
+	help
+	  Say yes here if you want support for the TWL6030 General Purpose
+	  A/D Convertor.
+
 endmenu
diff --git a/drivers/iio/adc/Makefile b/drivers/iio/adc/Makefile
index 0a825be..8b05633 100644
--- a/drivers/iio/adc/Makefile
+++ b/drivers/iio/adc/Makefile
@@ -17,3 +17,4 @@ obj-$(CONFIG_MAX1363) += max1363.o
 obj-$(CONFIG_TI_ADC081C) += ti-adc081c.o
 obj-$(CONFIG_TI_AM335X_ADC) += ti_am335x_adc.o
 obj-$(CONFIG_VIPERBOARD_ADC) += viperboard_adc.o
+obj-$(CONFIG_TWL6030_GPADC) += twl6030-gpadc.o
diff --git a/drivers/iio/adc/twl6030-gpadc.c b/drivers/iio/adc/twl6030-gpadc.c
new file mode 100644
index 0000000..6ceb789
--- /dev/null
+++ b/drivers/iio/adc/twl6030-gpadc.c
@@ -0,0 +1,1019 @@
+/*
+ * TWL6030 GPADC module driver
+ *
+ * Copyright (C) 2009-2013 Texas Instruments Inc.
+ * Nishant Kamat <nskamat@ti.com>
+ * Balaji T K <balajitk@ti.com>
+ * Graeme Gregory <gg@slimlogic.co.uk>
+ * Girish S Ghongdemath <girishsg@ti.com>
+ * Ambresh K <ambresh@ti.com>
+ * Oleksandr Kozaruk <oleksandr.kozaruk@ti.com
+ *
+ * Based on twl4030-madc.c
+ * Copyright (C) 2008 Nokia Corporation
+ * Mikko Ylinen <mikko.k.ylinen@nokia.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA
+ * 02110-1301 USA
+ *
+ */
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/of_platform.h>
+#include <linux/i2c/twl.h>
+#include <linux/iio/iio.h>
+#include <linux/iio/sysfs.h>
+
+#define DRIVER_NAME		"twl6030_gpadc"
+
+#define TWL6030_GPADC_MAX_CHANNELS 17
+#define TWL6032_GPADC_MAX_CHANNELS 19
+/* Define this as the biggest of all chips using this driver */
+#define GPADC_MAX_CHANNELS TWL6032_GPADC_MAX_CHANNELS
+
+#define TWL6030_GPADC_CTRL			0x00    /* 0x2e */
+#define TWL6030_GPADC_CTRL2			0x01	/* 0x2f */
+
+#define TWL6030_GPADC_CTRL_P1			0x05
+#define TWL6030_GPADC_CTRL_P2			0x06
+
+#define TWL6032_GPADC_GPSELECT_ISB		0x07
+#define TWL6032_GPADC_CTRL_P1			0x08
+
+#define TWL6032_GPADC_RTCH0_LSB			0x09
+#define TWL6032_GPADC_RTCH0_MSB			0x0a
+#define TWL6032_GPADC_RTCH1_LSB			0x0b
+#define TWL6032_GPADC_RTCH1_MSB			0x0c
+#define TWL6032_GPADC_GPCH0_LSB			0x0d
+#define TWL6032_GPADC_GPCH0_MSB			0x0e
+
+#define TWL6030_GPADC_CTRL_P1_SP1		BIT(3)
+#define TWL6030_GPADC_CTRL_P1_EOCRT		BIT(2)
+#define TWL6030_GPADC_CTRL_P1_EOCP1		BIT(1)
+#define TWL6030_GPADC_CTRL_P1_BUSY		BIT(0)
+
+#define TWL6030_GPADC_CTRL_P2_SP2		BIT(2)
+#define TWL6030_GPADC_CTRL_P2_EOCP2		BIT(1)
+#define TWL6030_GPADC_CTRL_P1_BUSY		BIT(0)
+
+#define TWL6030_GPADC_EOC_SW			BIT(1)
+#define TWL6030_GPADC_BUSY			BIT(0)
+
+#define TWL6030_GPADC_RTCH0_LSB			(0x07)
+#define TWL6030_GPADC_GPCH0_LSB			(0x29)
+
+/* Fixed channels */
+#define TWL6030_GPADC_CTRL_TEMP1_EN		BIT(0)    /* input ch 1 */
+#define TWL6030_GPADC_CTRL_TEMP2_EN		BIT(1)    /* input ch 4 */
+#define TWL6030_GPADC_CTRL_SCALER_EN		BIT(2)    /* input ch 2 */
+#define TWL6030_GPADC_CTRL_SCALER_DIV4		BIT(3)
+#define TWL6030_GPADC_CTRL_SCALER_EN_CH11	BIT(4)    /* input ch 11 */
+#define TWL6030_GPADC_CTRL_TEMP1_EN_MONITOR	BIT(5)
+#define TWL6030_GPADC_CTRL_TEMP2_EN_MONITOR	BIT(6)
+#define TWL6030_GPADC_CTRL_ISOURCE_EN		BIT(7)
+
+#define TWL6030_GPADC_CTRL2_REMSENSE_0		BIT(0)
+#define TWL6030_GPADC_CTRL2_REMSENSE_1		BIT(1)
+#define TWL6030_GPADC_CTRL2_SCALER_EN_CH18	BIT(2)
+#define TWL6030_GPADC_CTRL2_VBAT_SCALER_DIV4	BIT(3)
+
+#define TWL6030_GPADC_RT_SW1_EOC_MASK		BIT(5)
+#define TWL6030_GPADC_SW2_EOC_MASK		BIT(6)
+
+#define TWL6032_GPADC_RT_EOC_MASK		BIT(4)
+#define TWL6032_GPADC_SW_EOC_MASK		BIT(5)
+
+#define TWL6030_GPADC_TRIM1			0xCD
+
+#define TWL6030_REG_TOGGLE1			0x90
+#define TWL6030_GPADCS				BIT(1)
+#define TWL6030_GPADCR				BIT(0)
+
+/**
+ * struct twl6030_chnl_calib - channel calibration
+ * @gain:		slope coefficient for ideal curve
+ * @gain_error:		gain error
+ * @offset_error:	offset of the real curve
+ */
+struct twl6030_chnl_calib {
+	s32 gain;
+	s32 gain_error;
+	s32 offset_error;
+};
+
+/**
+ * struct twl6030_ideal_code - GPADC calibration parameters
+ * GPADC is calibrated in two points: close to the beginning and
+ * to the and of the measurable input range
+ *
+ * @code1:	ideal code for the input at the beginning
+ * @code2:	ideal code for at the end of the range
+ * @v1:		voltage input at the beginning(low voltage)
+ * @v2:		voltage input at the end(high voltage)
+ */
+struct twl6030_ideal_code {
+	u16 code1;
+	u16 code2;
+	u16 v1;
+	u16 v2;
+};
+
+struct twl6030_gpadc_data;
+
+/**
+ * struct twl6030_gpadc_platform_data - platform specific data
+ * @nchannels:		number of GPADC channels
+ * @iio_channels:	iio channels
+ * @twl6030_ideal:	pointer to calibration parameters
+ * @start_conversion:	pointer to ADC start conversion function
+ * @channel_to_reg	pointer to ADC function to convert channel to
+ * 			register address for reading conversion result
+ * @calibrate:		pointer to calibration function
+ */
+struct twl6030_gpadc_platform_data {
+	const int nchannels;
+	const struct iio_chan_spec *iio_channels;
+	const struct twl6030_ideal_code *ideal;
+	int (*start_conversion)(int channel);
+	u8 (*channel_to_reg)(int channel);
+	int (*calibrate)(struct twl6030_gpadc_data *gpadc);
+};
+
+/**
+ * struct twl6030_gpadc_data - GPADC data
+ * @dev:		device pointer
+ * @lock:		mutual exclusion lock for the structure
+ * @irq_complete:	completion to signal end of conversion
+ * @twl6030_cal_tbl:	pointer to calibration data for each
+ * 			channel with gain error and offset
+ * @pdata:		pointer to device specific data
+ */
+struct twl6030_gpadc_data {
+	struct device	*dev;
+	struct mutex	lock;
+	struct completion	irq_complete;
+	struct twl6030_chnl_calib	*twl6030_cal_tbl;
+	const struct twl6030_gpadc_platform_data *pdata;
+};
+
+/*
+ * channels 11, 12, 13, 15 and 16 have no calibration data
+ * calibration offset is same for channels 1, 3, 4, 5
+ *
+ * The data is taken from GPADC_TRIM registers description.
+ * GPADC_TRIM registers keeps difference between the code measured
+ * at v1 and v2 input voltages and corresponding code1 and code2
+ */
+static const struct twl6030_ideal_code
+	twl6030_ideal[TWL6030_GPADC_MAX_CHANNELS] = {
+	{ /* ch 0, external, battery type, resistor value */
+		.code1 = 116,
+		.code2 = 745,
+		.v1 = 141,
+		.v2 = 910,
+	},
+	{ /* ch 1, external, battery temperature, NTC resistor value */
+		.code1 = 82,
+		.code2 = 900,
+		.v1 = 100,
+		.v2 = 1100,
+	},
+	{ /* ch 2, external, audio accessory/general purpose */
+		.code1 = 55,
+		.code2 = 818,
+		.v1 = 101,
+		.v2 = 1499,
+	},
+	{ /* ch 3, external, general purpose */
+		.code1 = 82,
+		.code2 = 900,
+		.v1 = 100,
+		.v2 = 1100,
+	},
+	{ /* ch 4, external, temperature measurement/general purpose */
+		.code1 = 82,
+		.code2 = 900,
+		.v1 = 100,
+		.v2 = 1100,
+	},
+	{ /* ch 5, external, general purpose */
+		.code1 = 82,
+		.code2 = 900,
+		.v1 = 100,
+		.v2 = 1100,
+	},
+	{ /* ch 6, external, general purpose */
+		.code1 = 82,
+		.code2 = 900,
+		.v1 = 100,
+		.v2 = 1100,
+	},
+	{ /* ch 7, internal, main battery */
+		.code1 = 614,
+		.code2 = 941,
+		.v1 = 3001,
+		.v2 = 4599,
+	},
+	{ /* ch 8, internal, backup battery */
+		.code1 = 82,
+		.code2 = 688,
+		.v1 = 501,
+		.v2 = 4203,
+	},
+	{ /* ch 9, internal, external charger input */
+		.code1 = 182,
+		.code2 = 818,
+		.v1 = 2001,
+		.v2 = 8996,
+	},
+	{ /* ch 10, internal, VBUS */
+		.code1 = 149,
+		.code2 = 818,
+		.v1 = 1001,
+		.v2 = 5497,
+	},
+	{},	/* ch 11, internal, VBUS charging current */
+	{},	/* ch 12, internal, Die temperature */
+	{},	/* ch 13, internal, Die temperature */
+	{ /* ch 14, internal, USB ID line */
+		.code1 = 48,
+		.code2 = 714,
+		.v1 = 323,
+		.v2 = 4800,
+	},
+	{},	/* ch 15, internal, test network */
+	{},	/* ch 16, internal, test network */
+};
+
+static const struct twl6030_ideal_code
+			twl6032_ideal[TWL6032_GPADC_MAX_CHANNELS] = {
+	{ /* ch 0, external, battery type, resistor value */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 1, external, battery temperature, NTC resistor value */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 2, external, audio accessory/general purpose */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 660,
+		.v2 = 1500,
+	},
+	{ /* ch 3, external, temperature with external diode/general purpose */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 4, external, temperature measurement/general purpose */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 5, external, general purpose */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 6, external, general purpose */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch7, internal, system supply */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 2200,
+		.v2 = 5000,
+	},
+	{ /* ch8, internal, backup battery */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 2200,
+		.v2 = 5000,
+	},
+	{ /* ch 9, internal, external charger input */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 3960,
+		.v2 = 9000,
+	},
+	{ /* ch10, internal, VBUS */
+		.code1 = 150,
+		.code2 = 751,
+		.v1 = 1000,
+		.v2 = 5000,
+	},
+	{ /* ch 11, internal, VBUS DC-DC output current */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 660,
+		.v2 = 1500,
+	},
+	{ /* ch 12, internal, Die temperature */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 13, internal, Die temperature */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 14, internal, USB ID line */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 2420,
+		.v2 = 5500,
+	},
+	{},	/* ch 15, internal, test network */
+	{},	/* ch 16, internal, test network */
+	{},	/* ch 17, internal, battery charging current */
+	{ /* ch 18, internal, battery voltage */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 2200,
+		.v2 = 5000,
+	},
+};
+
+static inline int twl6030_gpadc_write(u8 reg, u8 val)
+{
+	return twl_i2c_write_u8(TWL6030_MODULE_GPADC, val, reg);
+}
+
+static inline int twl6030_gpadc_read(u8 *val, u8 reg)
+{
+
+	return twl_i2c_read(TWL6030_MODULE_GPADC, val, reg, 2);
+}
+
+static int twl6030_gpadc_enable_irq(u8 mask)
+{
+	int ret;
+
+	ret = twl6030_interrupt_unmask(mask, REG_INT_MSK_LINE_B);
+	ret |= twl6030_interrupt_unmask(mask, REG_INT_MSK_STS_B);
+
+	return ret;
+}
+
+static void twl6030_gpadc_disable_irq(u8 mask)
+{
+	twl6030_interrupt_mask(mask, REG_INT_MSK_LINE_B);
+	twl6030_interrupt_mask(mask, REG_INT_MSK_STS_B);
+}
+
+static irqreturn_t twl6030_gpadc_irq_handler(int irq, void *_gpadc)
+{
+	struct twl6030_gpadc_data *gpadc = _gpadc;
+
+	complete(&gpadc->irq_complete);
+
+	return IRQ_HANDLED;
+}
+
+static int twl6030_start_conversion(int channel)
+{
+	return twl6030_gpadc_write(TWL6030_GPADC_CTRL_P1,
+					TWL6030_GPADC_CTRL_P1_SP1);
+}
+
+static int twl6032_start_conversion(int channel)
+{
+	int ret;
+
+	ret = twl6030_gpadc_write(TWL6032_GPADC_GPSELECT_ISB, channel);
+	if (ret)
+		return ret;
+
+	return twl6030_gpadc_write(TWL6032_GPADC_CTRL_P1,
+						TWL6030_GPADC_CTRL_P1_SP1);
+}
+
+static u8 twl6030_channel_to_reg(int channel)
+{
+	return TWL6030_GPADC_GPCH0_LSB + 2 * channel;
+}
+
+static u8 twl6032_channel_to_reg(int channel)
+{
+	return TWL6032_GPADC_GPCH0_LSB;
+}
+
+static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,
+		int channel, int *res)
+{
+	u8 reg = gpadc->pdata->channel_to_reg(channel);
+	u8 val[2];
+	int ret;
+
+	ret = twl6030_gpadc_read(val, reg);
+	if (ret) {
+		dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);
+		return ret;
+	}
+
+	*res = (val[1] << 8) | val[0];
+
+	return ret;
+}
+
+static int twl6030_gpadc_get_processed(struct twl6030_gpadc_data *gpadc,
+		int channel, int *val)
+{
+	int raw_code;
+	int corrected_code;
+	int channel_value;
+	int ret;
+
+	ret = twl6030_gpadc_get_raw(gpadc, channel, &raw_code);
+	if (ret)
+		return ret;
+
+	corrected_code = ((raw_code * 1000) -
+		gpadc->twl6030_cal_tbl[channel].offset_error) /
+		gpadc->twl6030_cal_tbl[channel].gain_error;
+
+	channel_value = corrected_code *
+			gpadc->twl6030_cal_tbl[channel].gain;
+
+	/* Shift back into mV range */
+	channel_value /= 1000;
+
+	dev_dbg(gpadc->dev, "GPADC raw: %d", raw_code);
+	dev_dbg(gpadc->dev, "GPADC cor: %d", corrected_code);
+	dev_dbg(gpadc->dev, "GPADC val: %d", channel_value);
+
+	*val = channel_value;
+
+	return ret;
+}
+
+static int twl6030_gpadc_read_raw(struct iio_dev *indio_dev,
+			     const struct iio_chan_spec *chan,
+			     int *val, int *val2, long mask)
+{
+	struct twl6030_gpadc_data *gpadc = iio_priv(indio_dev);
+	int ret = -EINVAL;
+
+	mutex_lock(&gpadc->lock);
+
+	ret = gpadc->pdata->start_conversion(chan->channel);
+	if (ret) {
+		dev_err(gpadc->dev, "failed to start conversion\n");
+		goto err;
+	}
+	/* wait for conversion to complete */
+	wait_for_completion_interruptible_timeout(&gpadc->irq_complete,
+						msecs_to_jiffies(5000));
+
+	switch (mask) {
+	case IIO_CHAN_INFO_RAW:
+		ret = twl6030_gpadc_get_raw(gpadc, chan->channel, val);
+		ret = ret ? -EIO : IIO_VAL_INT;
+		break;
+
+	case IIO_CHAN_INFO_PROCESSED:
+		ret = twl6030_gpadc_get_processed(gpadc, chan->channel, val);
+		ret = ret ? -EIO : IIO_VAL_INT;
+		break;
+
+	default:
+		break;
+	}
+err:
+	mutex_unlock(&gpadc->lock);
+
+	return ret;
+}
+
+static void twl6030_calibrate_channel(struct twl6030_gpadc_data *gpadc,
+					int channel, int d1, int d2)
+{
+	int b, k, gain, x1, x2;
+	const struct twl6030_ideal_code *ideal = gpadc->pdata->ideal;
+
+	/* Gain */
+	gain = ((ideal[channel].v2 - ideal[channel].v1) * 1000) /
+		(ideal[channel].code2 - ideal[channel].code1);
+
+	x1 = ideal[channel].code1;
+	x2 = ideal[channel].code2;
+
+	/* k */
+	k = 1000 + (((d2 - d1) * 1000) / (x2 - x1));
+
+	/* b */
+	b = (d1 * 1000) - (k - 1000) * x1;
+
+	gpadc->twl6030_cal_tbl[channel].gain = gain;
+	gpadc->twl6030_cal_tbl[channel].gain_error = k;
+	gpadc->twl6030_cal_tbl[channel].offset_error = b;
+
+	dev_dbg(gpadc->dev, "GPADC d1   for Chn: %d = %d\n", channel, d1);
+	dev_dbg(gpadc->dev, "GPADC d2   for Chn: %d = %d\n", channel, d2);
+	dev_dbg(gpadc->dev, "GPADC x1   for Chn: %d = %d\n", channel, x1);
+	dev_dbg(gpadc->dev, "GPADC x2   for Chn: %d = %d\n", channel, x2);
+	dev_dbg(gpadc->dev, "GPADC Gain for Chn: %d = %d\n", channel, gain);
+	dev_dbg(gpadc->dev, "GPADC k    for Chn: %d = %d\n", channel, k);
+	dev_dbg(gpadc->dev, "GPADC b    for Chn: %d = %d\n", channel, b);
+}
+
+static inline int twl6030_gpadc_get_trim_offset(s8 d)
+{
+	int sign;
+
+	/*
+	 * XXX NOTE!
+	 * bit 0 - sign, bit 7 - reserved, 6..1 - trim value
+	 * though, the documentation states that trim value
+	 * is absolute value, the correct conversion results are
+	 * obtained if the value is interpreted as 2's complement.
+	 */
+	sign = d & 1;
+	d = (d & 0x7f) >> 1;
+
+	return sign ? (d | 0xc0) : d;
+}
+
+static int twl6030_calibration(struct twl6030_gpadc_data *gpadc)
+{
+	int ret;
+	int chn;
+	u8 trim_regs[TWL6030_GPADC_MAX_CHANNELS];
+	s8 d1, d2;
+
+	/*
+	 * for calibration two measurements have been performed at
+	 * factory, for some channels, during the production test and
+	 * have been stored in registers. This two stored values are
+	 * used to correct the measurements. The values represent
+	 * offsets for the given input from the output on ideal curve.
+	 */
+	ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs,
+			TWL6030_GPADC_TRIM1, 16);
+	if (ret < 0) {
+		dev_err(gpadc->dev, "calibration failed\n");
+		return ret;
+	}
+
+	for (chn = 0; chn < TWL6030_GPADC_MAX_CHANNELS; chn++) {
+
+		switch (chn) {
+		case 0:
+			d1 = trim_regs[0];
+			d2 = trim_regs[1];
+			break;
+		case 1:
+			d1 = trim_regs[4];
+			d2 = trim_regs[5];
+			break;
+		case 2:
+			d1 = trim_regs[12];
+			d2 = trim_regs[13];
+			break;
+		case 3:
+		case 4:
+		case 5:
+		case 6:
+			d1 = trim_regs[4];
+			d2 = trim_regs[5];
+			break;
+		case 7:
+			d1 = trim_regs[6];
+			d2 = trim_regs[7];
+			break;
+		case 8:
+			d1 = trim_regs[2];
+			d2 = trim_regs[3];
+			break;
+		case 9:
+			d1 = trim_regs[8];
+			d2 = trim_regs[9];
+			break;
+		case 10:
+			d1 = trim_regs[10];
+			d2 = trim_regs[11];
+			break;
+		case 14:
+			d1 = trim_regs[14];
+			d2 = trim_regs[15];
+			break;
+		default:
+			continue;
+		}
+
+		d1 = twl6030_gpadc_get_trim_offset(d1);
+		d2 = twl6030_gpadc_get_trim_offset(d2);
+
+		twl6030_calibrate_channel(gpadc, chn, d1, d2);
+	}
+
+	return 0;
+}
+
+static int twl6032_calibration(struct twl6030_gpadc_data *gpadc)
+{
+	int chn, d1 = 0, d2 = 0, temp;
+	u8 trim_regs[17];
+	int ret;
+
+	ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs + 1,
+			TWL6030_GPADC_TRIM1, 16);
+	if (ret < 0) {
+		dev_err(gpadc->dev, "calibration failed\n");
+		return ret;
+	}
+
+	/* Loop to calculate the value needed for returning voltages from
+	 * GPADC not values.
+	 *
+	 * gain is calculated to 3 decimal places fixed point.
+	 */
+	for (chn = 0; chn < TWL6032_GPADC_MAX_CHANNELS; chn++) {
+
+		switch (chn) {
+		case 0:
+		case 1:
+		case 2:
+		case 3:
+		case 4:
+		case 5:
+		case 6:
+		case 11:
+		case 12:
+		case 13:
+		case 14:
+			/* D1 */
+			d1 = (trim_regs[3] & 0x1F) << 2;
+			d1 |= (trim_regs[1] & 0x06) >> 1;
+			if (trim_regs[1] & 0x01)
+				d1 = -d1;
+
+			/* D2 */
+			d2 = (trim_regs[4] & 0x3F) << 2;
+			d2 |= (trim_regs[2] & 0x06) >> 1;
+			if (trim_regs[2] & 0x01)
+				d2 = -d2;
+			break;
+		case 8:
+			/* D1 */
+			temp = (trim_regs[3] & 0x1F) << 2;
+			temp |= (trim_regs[1] & 0x06) >> 1;
+			if (trim_regs[1] & 0x01)
+				temp = -temp;
+
+			d1 = (trim_regs[8] & 0x18) << 1;
+			d1 |= (trim_regs[7] & 0x1E) >> 1;
+			if (trim_regs[7] & 0x01)
+				d1 = -d1;
+
+			d1 += temp;
+
+			/* D2 */
+			temp = (trim_regs[4] & 0x3F) << 2;
+			temp |= (trim_regs[2] & 0x06) >> 1;
+			if (trim_regs[2] & 0x01)
+				temp = -temp;
+
+			d2 = (trim_regs[10] & 0x1F) << 2;
+			d2 |= (trim_regs[8] & 0x06) >> 1;
+			if (trim_regs[8] & 0x01)
+				d2 = -d2;
+
+			d2 += temp;
+			break;
+		case 9:
+			/* D1 */
+			temp = (trim_regs[3] & 0x1F) << 2;
+			temp |= (trim_regs[1] & 0x06) >> 1;
+			if (trim_regs[1] & 0x01)
+				temp = -temp;
+
+			d1 = (trim_regs[14] & 0x18) << 1;
+			d1 |= (trim_regs[12] & 0x1E) >> 1;
+			if (trim_regs[12] & 0x01)
+				d1 = -d1;
+
+			d1 += temp;
+
+			/* D2 */
+			temp = (trim_regs[4] & 0x3F) << 2;
+			temp |= (trim_regs[2] & 0x06) >> 1;
+			if (trim_regs[2] & 0x01)
+				temp = -temp;
+
+			d2 = (trim_regs[16] & 0x1F) << 2;
+			d2 |= (trim_regs[14] & 0x06) >> 1;
+			if (trim_regs[14] & 0x01)
+				d2 = -d2;
+
+			d2 += temp;
+		case 10:
+			/* D1 */
+			d1 = (trim_regs[11] & 0x0F) << 3;
+			d1 |= (trim_regs[9] & 0x0E) >> 1;
+			if (trim_regs[9] & 0x01)
+				d1 = -d1;
+
+			/* D2 */
+			d2 = (trim_regs[15] & 0x0F) << 3;
+			d2 |= (trim_regs[13] & 0x0E) >> 1;
+			if (trim_regs[13] & 0x01)
+				d2 = -d2;
+			break;
+		case 7:
+		case 18:
+			/* D1 */
+			temp = (trim_regs[3] & 0x1F) << 2;
+			temp |= (trim_regs[1] & 0x06) >> 1;
+			if (trim_regs[1] & 0x01)
+				temp = -temp;
+
+			d1 = (trim_regs[5] & 0x7E) >> 1;
+			if (trim_regs[5] & 0x01)
+				d1 = -d1;
+
+			d1 += temp;
+
+			/* D2 */
+			temp = (trim_regs[4] & 0x3F) << 2;
+			temp |= (trim_regs[2] & 0x06) >> 1;
+			if (trim_regs[2] & 0x01)
+				temp = -temp;
+
+			d2 = (trim_regs[6] & 0xFE) >> 1;
+			if (trim_regs[6] & 0x01)
+				d2 = -d2;
+
+			d2 += temp;
+			break;
+		default:
+			/* No data for other channels */
+			continue;
+		}
+
+		twl6030_calibrate_channel(gpadc, chn, d1, d2);
+	}
+
+	return 0;
+}
+
+#define TWL6030_GPADC_CHAN(chn, chan_info) {		\
+	.type = IIO_VOLTAGE,				\
+	.channel = chn,					\
+	.scan_type = {					\
+		.sign = 'u',				\
+		.storagebits = 16,			\
+	},						\
+	.info_mask = BIT(chan_info),			\
+	.info_mask_separate = BIT(chan_info), 		\
+	.indexed = 1,					\
+}
+
+static const struct iio_chan_spec twl6030_gpadc_iio_channels[] = {
+	TWL6030_GPADC_CHAN(0, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(1, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(2, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(3, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(4, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(5, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(6, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(7, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(8, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(9, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(10, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(11, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(12, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(13, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(14, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(15, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(16, IIO_CHAN_INFO_RAW),
+};
+
+static const struct iio_chan_spec twl6032_gpadc_iio_channels[] = {
+	TWL6030_GPADC_CHAN(0, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(1, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(2, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(3, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(4, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(5, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(6, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(7, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(8, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(9, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(10, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(11, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(12, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(13, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(14, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(15, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(16, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(17, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(18, IIO_CHAN_INFO_PROCESSED),
+};
+
+static const struct iio_info twl6030_gpadc_iio_info = {
+	.read_raw = &twl6030_gpadc_read_raw,
+	.driver_module = THIS_MODULE,
+};
+
+static const struct twl6030_gpadc_platform_data twl6030_pdata = {
+	.iio_channels = twl6030_gpadc_iio_channels,
+	.nchannels = TWL6030_GPADC_MAX_CHANNELS,
+	.ideal = twl6030_ideal,
+	.start_conversion = twl6030_start_conversion,
+	.channel_to_reg = twl6030_channel_to_reg,
+	.calibrate = twl6030_calibration,
+};
+
+static const struct twl6030_gpadc_platform_data twl6032_pdata = {
+	.iio_channels = twl6032_gpadc_iio_channels,
+	.nchannels = TWL6032_GPADC_MAX_CHANNELS,
+	.ideal = twl6032_ideal,
+	.start_conversion = twl6032_start_conversion,
+	.channel_to_reg = twl6032_channel_to_reg,
+	.calibrate = twl6032_calibration,
+};
+
+static const struct of_device_id of_twl6030_match_tbl[] = {
+	{
+		.compatible = "ti,twl6030_gpadc",
+		.data = &twl6030_pdata,
+	},
+	{
+		.compatible = "ti,twl6032_gpadc",
+		.data = &twl6032_pdata,
+	},
+	{ /* end */ }
+};
+
+static int twl6030_gpadc_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct twl6030_gpadc_data *gpadc;
+	const struct twl6030_gpadc_platform_data *pdata;
+	const struct of_device_id *match;
+	struct iio_dev *indio_dev;
+	int irq;
+	int ret = 0;
+
+	match = of_match_device(of_match_ptr(of_twl6030_match_tbl), dev);
+	pdata = match ? match->data : dev->platform_data;
+
+	if (!pdata)
+		return -EINVAL;
+
+	indio_dev = iio_device_alloc(sizeof(struct twl6030_gpadc_data));
+	if (!indio_dev) {
+		dev_err(dev, "failed allocating iio device\n");
+		ret = -ENOMEM;
+	}
+
+	gpadc = iio_priv(indio_dev);
+
+	gpadc->twl6030_cal_tbl = devm_kzalloc(dev,
+					sizeof(struct twl6030_chnl_calib) *
+					pdata->nchannels, GFP_KERNEL);
+	if (!gpadc->twl6030_cal_tbl)
+		goto err;
+
+	gpadc->dev = dev;
+	gpadc->pdata = pdata;
+
+	platform_set_drvdata(pdev, indio_dev);
+	mutex_init(&gpadc->lock);
+	init_completion(&gpadc->irq_complete);
+
+	ret = pdata->calibrate(gpadc);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to read calibration registers\n");
+		goto err;
+	}
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(&pdev->dev, "failed to get irq\n");
+		goto err;
+	}
+
+	ret = devm_request_threaded_irq(dev, irq, NULL,
+					twl6030_gpadc_irq_handler,
+					IRQF_ONESHOT, "twl6030_gpadc", gpadc);
+	if (ret) {
+		dev_dbg(&pdev->dev, "could not request irq\n");
+		goto err;
+	}
+
+	ret = twl6030_gpadc_enable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to enable GPADC interrupt\n");
+		goto err;
+	}
+	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
+					TWL6030_REG_TOGGLE1);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to enable GPADC module\n");
+		goto err;
+	}
+
+	indio_dev->name = DRIVER_NAME;
+	indio_dev->dev.parent = dev;
+	indio_dev->info = &twl6030_gpadc_iio_info;
+	indio_dev->modes = INDIO_DIRECT_MODE;
+	indio_dev->channels = pdata->iio_channels;
+	indio_dev->num_channels = pdata->nchannels;
+
+	ret = iio_device_register(indio_dev);
+	if (ret)
+		goto err;
+
+	return ret;
+err:
+	iio_device_free(indio_dev);
+	return ret;
+}
+
+static int twl6030_gpadc_remove(struct platform_device *pdev)
+{
+	struct iio_dev *indio_dev = platform_get_drvdata(pdev);
+
+	twl6030_gpadc_disable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
+	iio_device_unregister(indio_dev);
+	iio_device_free(indio_dev);
+
+	return 0;
+}
+
+static int twl6030_gpadc_suspend(struct device *pdev)
+{
+	int ret;
+
+	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCR,
+				TWL6030_REG_TOGGLE1);
+	if (ret)
+		dev_err(pdev, "error reseting GPADC (%d)!\n", ret);
+
+	return 0;
+};
+
+static int twl6030_gpadc_resume(struct device *pdev)
+{
+	int ret;
+
+	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
+				TWL6030_REG_TOGGLE1);
+	if (ret)
+		dev_err(pdev, "error setting GPADC (%d)!\n", ret);
+
+	return 0;
+};
+
+static const struct dev_pm_ops twl6030_gpadc_pm_ops = {
+	.suspend = twl6030_gpadc_suspend,
+	.resume = twl6030_gpadc_resume,
+};
+
+static struct platform_driver twl6030_gpadc_driver = {
+	.probe		= twl6030_gpadc_probe,
+	.remove		= twl6030_gpadc_remove,
+	.driver		= {
+		.name	= DRIVER_NAME,
+		.owner	= THIS_MODULE,
+		.pm	= &twl6030_gpadc_pm_ops,
+		.of_match_table = of_twl6030_match_tbl,
+	},
+};
+
+module_platform_driver(twl6030_gpadc_driver);
+
+MODULE_ALIAS("platform: " DRIVER_NAME);
+MODULE_AUTHOR("Texas Instruments Inc.");
+MODULE_DESCRIPTION("twl6030 ADC driver");
+MODULE_LICENSE("GPL");
-- 
1.7.9.5

^ permalink raw reply related	[flat|nested] 45+ messages in thread

* [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-12  7:18   ` Oleksandr Kozaruk
  0 siblings, 0 replies; 45+ messages in thread
From: Oleksandr Kozaruk @ 2013-07-12  7:18 UTC (permalink / raw)
  To: linux-arm-kernel

The GPADC is general purpose ADC found on TWL6030,
and TWL6032 PMIC, known also as Phoenix and PhoenixLite.

The TWL6030 and TWL6032 have GPADC with 17 and 19
channels respectively. Some channels have current
source and are used for measuring voltage drop
on resistive load for detecting battery ID resistance,
or measuring voltage drop on NTC resistors for external
temperature measurements. Some channels measure voltage,
(i.e. battery voltage), and have voltage dividers,
thus, capable to scale voltage. Some channels are dedicated
for measuring die temperature.

Some channels are calibrated in 2 points, having
offsets from ideal values kept in trim registers. This
is used to correct measurements.

The differences between GPADC in TWL6030 and TWL6032:
- 10 bit vs 12 bit ADC;
- 17 vs 19 channels;
- channels have different purpose(i. e. battery voltage
  channel 8 vs channel 18);
- trim values are interpreted differently.

The driver exports function returning converted value for
requested channels.

Based on the driver patched from Balaji TK, Graeme Gregory, Ambresh K,
Girish S Ghongdemath.

Signed-off-by: Balaji T K <balajitk@ti.com>
Graeme Gregory <gg@slimlogic.co.uk>
Signed-off-by: Oleksandr Kozaruk <oleksandr.kozaruk@ti.com>
---
 drivers/iio/adc/Kconfig         |    8 +
 drivers/iio/adc/Makefile        |    1 +
 drivers/iio/adc/twl6030-gpadc.c | 1019 +++++++++++++++++++++++++++++++++++++++
 3 files changed, 1028 insertions(+)
 create mode 100644 drivers/iio/adc/twl6030-gpadc.c

diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig
index ab0767e6..87d699e 100644
--- a/drivers/iio/adc/Kconfig
+++ b/drivers/iio/adc/Kconfig
@@ -157,4 +157,12 @@ config VIPERBOARD_ADC
 	  Say yes here to access the ADC part of the Nano River
 	  Technologies Viperboard.
 
+config TWL6030_GPADC
+	tristate "TWL6030 GPADC (General Purpose A/D Convertor) Support"
+	depends on TWL4030_CORE
+	default n
+	help
+	  Say yes here if you want support for the TWL6030 General Purpose
+	  A/D Convertor.
+
 endmenu
diff --git a/drivers/iio/adc/Makefile b/drivers/iio/adc/Makefile
index 0a825be..8b05633 100644
--- a/drivers/iio/adc/Makefile
+++ b/drivers/iio/adc/Makefile
@@ -17,3 +17,4 @@ obj-$(CONFIG_MAX1363) += max1363.o
 obj-$(CONFIG_TI_ADC081C) += ti-adc081c.o
 obj-$(CONFIG_TI_AM335X_ADC) += ti_am335x_adc.o
 obj-$(CONFIG_VIPERBOARD_ADC) += viperboard_adc.o
+obj-$(CONFIG_TWL6030_GPADC) += twl6030-gpadc.o
diff --git a/drivers/iio/adc/twl6030-gpadc.c b/drivers/iio/adc/twl6030-gpadc.c
new file mode 100644
index 0000000..6ceb789
--- /dev/null
+++ b/drivers/iio/adc/twl6030-gpadc.c
@@ -0,0 +1,1019 @@
+/*
+ * TWL6030 GPADC module driver
+ *
+ * Copyright (C) 2009-2013 Texas Instruments Inc.
+ * Nishant Kamat <nskamat@ti.com>
+ * Balaji T K <balajitk@ti.com>
+ * Graeme Gregory <gg@slimlogic.co.uk>
+ * Girish S Ghongdemath <girishsg@ti.com>
+ * Ambresh K <ambresh@ti.com>
+ * Oleksandr Kozaruk <oleksandr.kozaruk at ti.com
+ *
+ * Based on twl4030-madc.c
+ * Copyright (C) 2008 Nokia Corporation
+ * Mikko Ylinen <mikko.k.ylinen@nokia.com>
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * version 2 as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA
+ * 02110-1301 USA
+ *
+ */
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/of_platform.h>
+#include <linux/i2c/twl.h>
+#include <linux/iio/iio.h>
+#include <linux/iio/sysfs.h>
+
+#define DRIVER_NAME		"twl6030_gpadc"
+
+#define TWL6030_GPADC_MAX_CHANNELS 17
+#define TWL6032_GPADC_MAX_CHANNELS 19
+/* Define this as the biggest of all chips using this driver */
+#define GPADC_MAX_CHANNELS TWL6032_GPADC_MAX_CHANNELS
+
+#define TWL6030_GPADC_CTRL			0x00    /* 0x2e */
+#define TWL6030_GPADC_CTRL2			0x01	/* 0x2f */
+
+#define TWL6030_GPADC_CTRL_P1			0x05
+#define TWL6030_GPADC_CTRL_P2			0x06
+
+#define TWL6032_GPADC_GPSELECT_ISB		0x07
+#define TWL6032_GPADC_CTRL_P1			0x08
+
+#define TWL6032_GPADC_RTCH0_LSB			0x09
+#define TWL6032_GPADC_RTCH0_MSB			0x0a
+#define TWL6032_GPADC_RTCH1_LSB			0x0b
+#define TWL6032_GPADC_RTCH1_MSB			0x0c
+#define TWL6032_GPADC_GPCH0_LSB			0x0d
+#define TWL6032_GPADC_GPCH0_MSB			0x0e
+
+#define TWL6030_GPADC_CTRL_P1_SP1		BIT(3)
+#define TWL6030_GPADC_CTRL_P1_EOCRT		BIT(2)
+#define TWL6030_GPADC_CTRL_P1_EOCP1		BIT(1)
+#define TWL6030_GPADC_CTRL_P1_BUSY		BIT(0)
+
+#define TWL6030_GPADC_CTRL_P2_SP2		BIT(2)
+#define TWL6030_GPADC_CTRL_P2_EOCP2		BIT(1)
+#define TWL6030_GPADC_CTRL_P1_BUSY		BIT(0)
+
+#define TWL6030_GPADC_EOC_SW			BIT(1)
+#define TWL6030_GPADC_BUSY			BIT(0)
+
+#define TWL6030_GPADC_RTCH0_LSB			(0x07)
+#define TWL6030_GPADC_GPCH0_LSB			(0x29)
+
+/* Fixed channels */
+#define TWL6030_GPADC_CTRL_TEMP1_EN		BIT(0)    /* input ch 1 */
+#define TWL6030_GPADC_CTRL_TEMP2_EN		BIT(1)    /* input ch 4 */
+#define TWL6030_GPADC_CTRL_SCALER_EN		BIT(2)    /* input ch 2 */
+#define TWL6030_GPADC_CTRL_SCALER_DIV4		BIT(3)
+#define TWL6030_GPADC_CTRL_SCALER_EN_CH11	BIT(4)    /* input ch 11 */
+#define TWL6030_GPADC_CTRL_TEMP1_EN_MONITOR	BIT(5)
+#define TWL6030_GPADC_CTRL_TEMP2_EN_MONITOR	BIT(6)
+#define TWL6030_GPADC_CTRL_ISOURCE_EN		BIT(7)
+
+#define TWL6030_GPADC_CTRL2_REMSENSE_0		BIT(0)
+#define TWL6030_GPADC_CTRL2_REMSENSE_1		BIT(1)
+#define TWL6030_GPADC_CTRL2_SCALER_EN_CH18	BIT(2)
+#define TWL6030_GPADC_CTRL2_VBAT_SCALER_DIV4	BIT(3)
+
+#define TWL6030_GPADC_RT_SW1_EOC_MASK		BIT(5)
+#define TWL6030_GPADC_SW2_EOC_MASK		BIT(6)
+
+#define TWL6032_GPADC_RT_EOC_MASK		BIT(4)
+#define TWL6032_GPADC_SW_EOC_MASK		BIT(5)
+
+#define TWL6030_GPADC_TRIM1			0xCD
+
+#define TWL6030_REG_TOGGLE1			0x90
+#define TWL6030_GPADCS				BIT(1)
+#define TWL6030_GPADCR				BIT(0)
+
+/**
+ * struct twl6030_chnl_calib - channel calibration
+ * @gain:		slope coefficient for ideal curve
+ * @gain_error:		gain error
+ * @offset_error:	offset of the real curve
+ */
+struct twl6030_chnl_calib {
+	s32 gain;
+	s32 gain_error;
+	s32 offset_error;
+};
+
+/**
+ * struct twl6030_ideal_code - GPADC calibration parameters
+ * GPADC is calibrated in two points: close to the beginning and
+ * to the and of the measurable input range
+ *
+ * @code1:	ideal code for the input at the beginning
+ * @code2:	ideal code for at the end of the range
+ * @v1:		voltage input at the beginning(low voltage)
+ * @v2:		voltage input at the end(high voltage)
+ */
+struct twl6030_ideal_code {
+	u16 code1;
+	u16 code2;
+	u16 v1;
+	u16 v2;
+};
+
+struct twl6030_gpadc_data;
+
+/**
+ * struct twl6030_gpadc_platform_data - platform specific data
+ * @nchannels:		number of GPADC channels
+ * @iio_channels:	iio channels
+ * @twl6030_ideal:	pointer to calibration parameters
+ * @start_conversion:	pointer to ADC start conversion function
+ * @channel_to_reg	pointer to ADC function to convert channel to
+ * 			register address for reading conversion result
+ * @calibrate:		pointer to calibration function
+ */
+struct twl6030_gpadc_platform_data {
+	const int nchannels;
+	const struct iio_chan_spec *iio_channels;
+	const struct twl6030_ideal_code *ideal;
+	int (*start_conversion)(int channel);
+	u8 (*channel_to_reg)(int channel);
+	int (*calibrate)(struct twl6030_gpadc_data *gpadc);
+};
+
+/**
+ * struct twl6030_gpadc_data - GPADC data
+ * @dev:		device pointer
+ * @lock:		mutual exclusion lock for the structure
+ * @irq_complete:	completion to signal end of conversion
+ * @twl6030_cal_tbl:	pointer to calibration data for each
+ * 			channel with gain error and offset
+ * @pdata:		pointer to device specific data
+ */
+struct twl6030_gpadc_data {
+	struct device	*dev;
+	struct mutex	lock;
+	struct completion	irq_complete;
+	struct twl6030_chnl_calib	*twl6030_cal_tbl;
+	const struct twl6030_gpadc_platform_data *pdata;
+};
+
+/*
+ * channels 11, 12, 13, 15 and 16 have no calibration data
+ * calibration offset is same for channels 1, 3, 4, 5
+ *
+ * The data is taken from GPADC_TRIM registers description.
+ * GPADC_TRIM registers keeps difference between the code measured
+ * at v1 and v2 input voltages and corresponding code1 and code2
+ */
+static const struct twl6030_ideal_code
+	twl6030_ideal[TWL6030_GPADC_MAX_CHANNELS] = {
+	{ /* ch 0, external, battery type, resistor value */
+		.code1 = 116,
+		.code2 = 745,
+		.v1 = 141,
+		.v2 = 910,
+	},
+	{ /* ch 1, external, battery temperature, NTC resistor value */
+		.code1 = 82,
+		.code2 = 900,
+		.v1 = 100,
+		.v2 = 1100,
+	},
+	{ /* ch 2, external, audio accessory/general purpose */
+		.code1 = 55,
+		.code2 = 818,
+		.v1 = 101,
+		.v2 = 1499,
+	},
+	{ /* ch 3, external, general purpose */
+		.code1 = 82,
+		.code2 = 900,
+		.v1 = 100,
+		.v2 = 1100,
+	},
+	{ /* ch 4, external, temperature measurement/general purpose */
+		.code1 = 82,
+		.code2 = 900,
+		.v1 = 100,
+		.v2 = 1100,
+	},
+	{ /* ch 5, external, general purpose */
+		.code1 = 82,
+		.code2 = 900,
+		.v1 = 100,
+		.v2 = 1100,
+	},
+	{ /* ch 6, external, general purpose */
+		.code1 = 82,
+		.code2 = 900,
+		.v1 = 100,
+		.v2 = 1100,
+	},
+	{ /* ch 7, internal, main battery */
+		.code1 = 614,
+		.code2 = 941,
+		.v1 = 3001,
+		.v2 = 4599,
+	},
+	{ /* ch 8, internal, backup battery */
+		.code1 = 82,
+		.code2 = 688,
+		.v1 = 501,
+		.v2 = 4203,
+	},
+	{ /* ch 9, internal, external charger input */
+		.code1 = 182,
+		.code2 = 818,
+		.v1 = 2001,
+		.v2 = 8996,
+	},
+	{ /* ch 10, internal, VBUS */
+		.code1 = 149,
+		.code2 = 818,
+		.v1 = 1001,
+		.v2 = 5497,
+	},
+	{},	/* ch 11, internal, VBUS charging current */
+	{},	/* ch 12, internal, Die temperature */
+	{},	/* ch 13, internal, Die temperature */
+	{ /* ch 14, internal, USB ID line */
+		.code1 = 48,
+		.code2 = 714,
+		.v1 = 323,
+		.v2 = 4800,
+	},
+	{},	/* ch 15, internal, test network */
+	{},	/* ch 16, internal, test network */
+};
+
+static const struct twl6030_ideal_code
+			twl6032_ideal[TWL6032_GPADC_MAX_CHANNELS] = {
+	{ /* ch 0, external, battery type, resistor value */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 1, external, battery temperature, NTC resistor value */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 2, external, audio accessory/general purpose */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 660,
+		.v2 = 1500,
+	},
+	{ /* ch 3, external, temperature with external diode/general purpose */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 4, external, temperature measurement/general purpose */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 5, external, general purpose */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 6, external, general purpose */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch7, internal, system supply */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 2200,
+		.v2 = 5000,
+	},
+	{ /* ch8, internal, backup battery */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 2200,
+		.v2 = 5000,
+	},
+	{ /* ch 9, internal, external charger input */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 3960,
+		.v2 = 9000,
+	},
+	{ /* ch10, internal, VBUS */
+		.code1 = 150,
+		.code2 = 751,
+		.v1 = 1000,
+		.v2 = 5000,
+	},
+	{ /* ch 11, internal, VBUS DC-DC output current */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 660,
+		.v2 = 1500,
+	},
+	{ /* ch 12, internal, Die temperature */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 13, internal, Die temperature */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 440,
+		.v2 = 1000,
+	},
+	{ /* ch 14, internal, USB ID line */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 2420,
+		.v2 = 5500,
+	},
+	{},	/* ch 15, internal, test network */
+	{},	/* ch 16, internal, test network */
+	{},	/* ch 17, internal, battery charging current */
+	{ /* ch 18, internal, battery voltage */
+		.code1 = 1441,
+		.code2 = 3276,
+		.v1 = 2200,
+		.v2 = 5000,
+	},
+};
+
+static inline int twl6030_gpadc_write(u8 reg, u8 val)
+{
+	return twl_i2c_write_u8(TWL6030_MODULE_GPADC, val, reg);
+}
+
+static inline int twl6030_gpadc_read(u8 *val, u8 reg)
+{
+
+	return twl_i2c_read(TWL6030_MODULE_GPADC, val, reg, 2);
+}
+
+static int twl6030_gpadc_enable_irq(u8 mask)
+{
+	int ret;
+
+	ret = twl6030_interrupt_unmask(mask, REG_INT_MSK_LINE_B);
+	ret |= twl6030_interrupt_unmask(mask, REG_INT_MSK_STS_B);
+
+	return ret;
+}
+
+static void twl6030_gpadc_disable_irq(u8 mask)
+{
+	twl6030_interrupt_mask(mask, REG_INT_MSK_LINE_B);
+	twl6030_interrupt_mask(mask, REG_INT_MSK_STS_B);
+}
+
+static irqreturn_t twl6030_gpadc_irq_handler(int irq, void *_gpadc)
+{
+	struct twl6030_gpadc_data *gpadc = _gpadc;
+
+	complete(&gpadc->irq_complete);
+
+	return IRQ_HANDLED;
+}
+
+static int twl6030_start_conversion(int channel)
+{
+	return twl6030_gpadc_write(TWL6030_GPADC_CTRL_P1,
+					TWL6030_GPADC_CTRL_P1_SP1);
+}
+
+static int twl6032_start_conversion(int channel)
+{
+	int ret;
+
+	ret = twl6030_gpadc_write(TWL6032_GPADC_GPSELECT_ISB, channel);
+	if (ret)
+		return ret;
+
+	return twl6030_gpadc_write(TWL6032_GPADC_CTRL_P1,
+						TWL6030_GPADC_CTRL_P1_SP1);
+}
+
+static u8 twl6030_channel_to_reg(int channel)
+{
+	return TWL6030_GPADC_GPCH0_LSB + 2 * channel;
+}
+
+static u8 twl6032_channel_to_reg(int channel)
+{
+	return TWL6032_GPADC_GPCH0_LSB;
+}
+
+static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,
+		int channel, int *res)
+{
+	u8 reg = gpadc->pdata->channel_to_reg(channel);
+	u8 val[2];
+	int ret;
+
+	ret = twl6030_gpadc_read(val, reg);
+	if (ret) {
+		dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);
+		return ret;
+	}
+
+	*res = (val[1] << 8) | val[0];
+
+	return ret;
+}
+
+static int twl6030_gpadc_get_processed(struct twl6030_gpadc_data *gpadc,
+		int channel, int *val)
+{
+	int raw_code;
+	int corrected_code;
+	int channel_value;
+	int ret;
+
+	ret = twl6030_gpadc_get_raw(gpadc, channel, &raw_code);
+	if (ret)
+		return ret;
+
+	corrected_code = ((raw_code * 1000) -
+		gpadc->twl6030_cal_tbl[channel].offset_error) /
+		gpadc->twl6030_cal_tbl[channel].gain_error;
+
+	channel_value = corrected_code *
+			gpadc->twl6030_cal_tbl[channel].gain;
+
+	/* Shift back into mV range */
+	channel_value /= 1000;
+
+	dev_dbg(gpadc->dev, "GPADC raw: %d", raw_code);
+	dev_dbg(gpadc->dev, "GPADC cor: %d", corrected_code);
+	dev_dbg(gpadc->dev, "GPADC val: %d", channel_value);
+
+	*val = channel_value;
+
+	return ret;
+}
+
+static int twl6030_gpadc_read_raw(struct iio_dev *indio_dev,
+			     const struct iio_chan_spec *chan,
+			     int *val, int *val2, long mask)
+{
+	struct twl6030_gpadc_data *gpadc = iio_priv(indio_dev);
+	int ret = -EINVAL;
+
+	mutex_lock(&gpadc->lock);
+
+	ret = gpadc->pdata->start_conversion(chan->channel);
+	if (ret) {
+		dev_err(gpadc->dev, "failed to start conversion\n");
+		goto err;
+	}
+	/* wait for conversion to complete */
+	wait_for_completion_interruptible_timeout(&gpadc->irq_complete,
+						msecs_to_jiffies(5000));
+
+	switch (mask) {
+	case IIO_CHAN_INFO_RAW:
+		ret = twl6030_gpadc_get_raw(gpadc, chan->channel, val);
+		ret = ret ? -EIO : IIO_VAL_INT;
+		break;
+
+	case IIO_CHAN_INFO_PROCESSED:
+		ret = twl6030_gpadc_get_processed(gpadc, chan->channel, val);
+		ret = ret ? -EIO : IIO_VAL_INT;
+		break;
+
+	default:
+		break;
+	}
+err:
+	mutex_unlock(&gpadc->lock);
+
+	return ret;
+}
+
+static void twl6030_calibrate_channel(struct twl6030_gpadc_data *gpadc,
+					int channel, int d1, int d2)
+{
+	int b, k, gain, x1, x2;
+	const struct twl6030_ideal_code *ideal = gpadc->pdata->ideal;
+
+	/* Gain */
+	gain = ((ideal[channel].v2 - ideal[channel].v1) * 1000) /
+		(ideal[channel].code2 - ideal[channel].code1);
+
+	x1 = ideal[channel].code1;
+	x2 = ideal[channel].code2;
+
+	/* k */
+	k = 1000 + (((d2 - d1) * 1000) / (x2 - x1));
+
+	/* b */
+	b = (d1 * 1000) - (k - 1000) * x1;
+
+	gpadc->twl6030_cal_tbl[channel].gain = gain;
+	gpadc->twl6030_cal_tbl[channel].gain_error = k;
+	gpadc->twl6030_cal_tbl[channel].offset_error = b;
+
+	dev_dbg(gpadc->dev, "GPADC d1   for Chn: %d = %d\n", channel, d1);
+	dev_dbg(gpadc->dev, "GPADC d2   for Chn: %d = %d\n", channel, d2);
+	dev_dbg(gpadc->dev, "GPADC x1   for Chn: %d = %d\n", channel, x1);
+	dev_dbg(gpadc->dev, "GPADC x2   for Chn: %d = %d\n", channel, x2);
+	dev_dbg(gpadc->dev, "GPADC Gain for Chn: %d = %d\n", channel, gain);
+	dev_dbg(gpadc->dev, "GPADC k    for Chn: %d = %d\n", channel, k);
+	dev_dbg(gpadc->dev, "GPADC b    for Chn: %d = %d\n", channel, b);
+}
+
+static inline int twl6030_gpadc_get_trim_offset(s8 d)
+{
+	int sign;
+
+	/*
+	 * XXX NOTE!
+	 * bit 0 - sign, bit 7 - reserved, 6..1 - trim value
+	 * though, the documentation states that trim value
+	 * is absolute value, the correct conversion results are
+	 * obtained if the value is interpreted as 2's complement.
+	 */
+	sign = d & 1;
+	d = (d & 0x7f) >> 1;
+
+	return sign ? (d | 0xc0) : d;
+}
+
+static int twl6030_calibration(struct twl6030_gpadc_data *gpadc)
+{
+	int ret;
+	int chn;
+	u8 trim_regs[TWL6030_GPADC_MAX_CHANNELS];
+	s8 d1, d2;
+
+	/*
+	 * for calibration two measurements have been performed at
+	 * factory, for some channels, during the production test and
+	 * have been stored in registers. This two stored values are
+	 * used to correct the measurements. The values represent
+	 * offsets for the given input from the output on ideal curve.
+	 */
+	ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs,
+			TWL6030_GPADC_TRIM1, 16);
+	if (ret < 0) {
+		dev_err(gpadc->dev, "calibration failed\n");
+		return ret;
+	}
+
+	for (chn = 0; chn < TWL6030_GPADC_MAX_CHANNELS; chn++) {
+
+		switch (chn) {
+		case 0:
+			d1 = trim_regs[0];
+			d2 = trim_regs[1];
+			break;
+		case 1:
+			d1 = trim_regs[4];
+			d2 = trim_regs[5];
+			break;
+		case 2:
+			d1 = trim_regs[12];
+			d2 = trim_regs[13];
+			break;
+		case 3:
+		case 4:
+		case 5:
+		case 6:
+			d1 = trim_regs[4];
+			d2 = trim_regs[5];
+			break;
+		case 7:
+			d1 = trim_regs[6];
+			d2 = trim_regs[7];
+			break;
+		case 8:
+			d1 = trim_regs[2];
+			d2 = trim_regs[3];
+			break;
+		case 9:
+			d1 = trim_regs[8];
+			d2 = trim_regs[9];
+			break;
+		case 10:
+			d1 = trim_regs[10];
+			d2 = trim_regs[11];
+			break;
+		case 14:
+			d1 = trim_regs[14];
+			d2 = trim_regs[15];
+			break;
+		default:
+			continue;
+		}
+
+		d1 = twl6030_gpadc_get_trim_offset(d1);
+		d2 = twl6030_gpadc_get_trim_offset(d2);
+
+		twl6030_calibrate_channel(gpadc, chn, d1, d2);
+	}
+
+	return 0;
+}
+
+static int twl6032_calibration(struct twl6030_gpadc_data *gpadc)
+{
+	int chn, d1 = 0, d2 = 0, temp;
+	u8 trim_regs[17];
+	int ret;
+
+	ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs + 1,
+			TWL6030_GPADC_TRIM1, 16);
+	if (ret < 0) {
+		dev_err(gpadc->dev, "calibration failed\n");
+		return ret;
+	}
+
+	/* Loop to calculate the value needed for returning voltages from
+	 * GPADC not values.
+	 *
+	 * gain is calculated to 3 decimal places fixed point.
+	 */
+	for (chn = 0; chn < TWL6032_GPADC_MAX_CHANNELS; chn++) {
+
+		switch (chn) {
+		case 0:
+		case 1:
+		case 2:
+		case 3:
+		case 4:
+		case 5:
+		case 6:
+		case 11:
+		case 12:
+		case 13:
+		case 14:
+			/* D1 */
+			d1 = (trim_regs[3] & 0x1F) << 2;
+			d1 |= (trim_regs[1] & 0x06) >> 1;
+			if (trim_regs[1] & 0x01)
+				d1 = -d1;
+
+			/* D2 */
+			d2 = (trim_regs[4] & 0x3F) << 2;
+			d2 |= (trim_regs[2] & 0x06) >> 1;
+			if (trim_regs[2] & 0x01)
+				d2 = -d2;
+			break;
+		case 8:
+			/* D1 */
+			temp = (trim_regs[3] & 0x1F) << 2;
+			temp |= (trim_regs[1] & 0x06) >> 1;
+			if (trim_regs[1] & 0x01)
+				temp = -temp;
+
+			d1 = (trim_regs[8] & 0x18) << 1;
+			d1 |= (trim_regs[7] & 0x1E) >> 1;
+			if (trim_regs[7] & 0x01)
+				d1 = -d1;
+
+			d1 += temp;
+
+			/* D2 */
+			temp = (trim_regs[4] & 0x3F) << 2;
+			temp |= (trim_regs[2] & 0x06) >> 1;
+			if (trim_regs[2] & 0x01)
+				temp = -temp;
+
+			d2 = (trim_regs[10] & 0x1F) << 2;
+			d2 |= (trim_regs[8] & 0x06) >> 1;
+			if (trim_regs[8] & 0x01)
+				d2 = -d2;
+
+			d2 += temp;
+			break;
+		case 9:
+			/* D1 */
+			temp = (trim_regs[3] & 0x1F) << 2;
+			temp |= (trim_regs[1] & 0x06) >> 1;
+			if (trim_regs[1] & 0x01)
+				temp = -temp;
+
+			d1 = (trim_regs[14] & 0x18) << 1;
+			d1 |= (trim_regs[12] & 0x1E) >> 1;
+			if (trim_regs[12] & 0x01)
+				d1 = -d1;
+
+			d1 += temp;
+
+			/* D2 */
+			temp = (trim_regs[4] & 0x3F) << 2;
+			temp |= (trim_regs[2] & 0x06) >> 1;
+			if (trim_regs[2] & 0x01)
+				temp = -temp;
+
+			d2 = (trim_regs[16] & 0x1F) << 2;
+			d2 |= (trim_regs[14] & 0x06) >> 1;
+			if (trim_regs[14] & 0x01)
+				d2 = -d2;
+
+			d2 += temp;
+		case 10:
+			/* D1 */
+			d1 = (trim_regs[11] & 0x0F) << 3;
+			d1 |= (trim_regs[9] & 0x0E) >> 1;
+			if (trim_regs[9] & 0x01)
+				d1 = -d1;
+
+			/* D2 */
+			d2 = (trim_regs[15] & 0x0F) << 3;
+			d2 |= (trim_regs[13] & 0x0E) >> 1;
+			if (trim_regs[13] & 0x01)
+				d2 = -d2;
+			break;
+		case 7:
+		case 18:
+			/* D1 */
+			temp = (trim_regs[3] & 0x1F) << 2;
+			temp |= (trim_regs[1] & 0x06) >> 1;
+			if (trim_regs[1] & 0x01)
+				temp = -temp;
+
+			d1 = (trim_regs[5] & 0x7E) >> 1;
+			if (trim_regs[5] & 0x01)
+				d1 = -d1;
+
+			d1 += temp;
+
+			/* D2 */
+			temp = (trim_regs[4] & 0x3F) << 2;
+			temp |= (trim_regs[2] & 0x06) >> 1;
+			if (trim_regs[2] & 0x01)
+				temp = -temp;
+
+			d2 = (trim_regs[6] & 0xFE) >> 1;
+			if (trim_regs[6] & 0x01)
+				d2 = -d2;
+
+			d2 += temp;
+			break;
+		default:
+			/* No data for other channels */
+			continue;
+		}
+
+		twl6030_calibrate_channel(gpadc, chn, d1, d2);
+	}
+
+	return 0;
+}
+
+#define TWL6030_GPADC_CHAN(chn, chan_info) {		\
+	.type = IIO_VOLTAGE,				\
+	.channel = chn,					\
+	.scan_type = {					\
+		.sign = 'u',				\
+		.storagebits = 16,			\
+	},						\
+	.info_mask = BIT(chan_info),			\
+	.info_mask_separate = BIT(chan_info), 		\
+	.indexed = 1,					\
+}
+
+static const struct iio_chan_spec twl6030_gpadc_iio_channels[] = {
+	TWL6030_GPADC_CHAN(0, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(1, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(2, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(3, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(4, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(5, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(6, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(7, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(8, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(9, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(10, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(11, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(12, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(13, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(14, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(15, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(16, IIO_CHAN_INFO_RAW),
+};
+
+static const struct iio_chan_spec twl6032_gpadc_iio_channels[] = {
+	TWL6030_GPADC_CHAN(0, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(1, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(2, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(3, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(4, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(5, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(6, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(7, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(8, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(9, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(10, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(11, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(12, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(13, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(14, IIO_CHAN_INFO_PROCESSED),
+	TWL6030_GPADC_CHAN(15, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(16, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(17, IIO_CHAN_INFO_RAW),
+	TWL6030_GPADC_CHAN(18, IIO_CHAN_INFO_PROCESSED),
+};
+
+static const struct iio_info twl6030_gpadc_iio_info = {
+	.read_raw = &twl6030_gpadc_read_raw,
+	.driver_module = THIS_MODULE,
+};
+
+static const struct twl6030_gpadc_platform_data twl6030_pdata = {
+	.iio_channels = twl6030_gpadc_iio_channels,
+	.nchannels = TWL6030_GPADC_MAX_CHANNELS,
+	.ideal = twl6030_ideal,
+	.start_conversion = twl6030_start_conversion,
+	.channel_to_reg = twl6030_channel_to_reg,
+	.calibrate = twl6030_calibration,
+};
+
+static const struct twl6030_gpadc_platform_data twl6032_pdata = {
+	.iio_channels = twl6032_gpadc_iio_channels,
+	.nchannels = TWL6032_GPADC_MAX_CHANNELS,
+	.ideal = twl6032_ideal,
+	.start_conversion = twl6032_start_conversion,
+	.channel_to_reg = twl6032_channel_to_reg,
+	.calibrate = twl6032_calibration,
+};
+
+static const struct of_device_id of_twl6030_match_tbl[] = {
+	{
+		.compatible = "ti,twl6030_gpadc",
+		.data = &twl6030_pdata,
+	},
+	{
+		.compatible = "ti,twl6032_gpadc",
+		.data = &twl6032_pdata,
+	},
+	{ /* end */ }
+};
+
+static int twl6030_gpadc_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct twl6030_gpadc_data *gpadc;
+	const struct twl6030_gpadc_platform_data *pdata;
+	const struct of_device_id *match;
+	struct iio_dev *indio_dev;
+	int irq;
+	int ret = 0;
+
+	match = of_match_device(of_match_ptr(of_twl6030_match_tbl), dev);
+	pdata = match ? match->data : dev->platform_data;
+
+	if (!pdata)
+		return -EINVAL;
+
+	indio_dev = iio_device_alloc(sizeof(struct twl6030_gpadc_data));
+	if (!indio_dev) {
+		dev_err(dev, "failed allocating iio device\n");
+		ret = -ENOMEM;
+	}
+
+	gpadc = iio_priv(indio_dev);
+
+	gpadc->twl6030_cal_tbl = devm_kzalloc(dev,
+					sizeof(struct twl6030_chnl_calib) *
+					pdata->nchannels, GFP_KERNEL);
+	if (!gpadc->twl6030_cal_tbl)
+		goto err;
+
+	gpadc->dev = dev;
+	gpadc->pdata = pdata;
+
+	platform_set_drvdata(pdev, indio_dev);
+	mutex_init(&gpadc->lock);
+	init_completion(&gpadc->irq_complete);
+
+	ret = pdata->calibrate(gpadc);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to read calibration registers\n");
+		goto err;
+	}
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(&pdev->dev, "failed to get irq\n");
+		goto err;
+	}
+
+	ret = devm_request_threaded_irq(dev, irq, NULL,
+					twl6030_gpadc_irq_handler,
+					IRQF_ONESHOT, "twl6030_gpadc", gpadc);
+	if (ret) {
+		dev_dbg(&pdev->dev, "could not request irq\n");
+		goto err;
+	}
+
+	ret = twl6030_gpadc_enable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to enable GPADC interrupt\n");
+		goto err;
+	}
+	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
+					TWL6030_REG_TOGGLE1);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "failed to enable GPADC module\n");
+		goto err;
+	}
+
+	indio_dev->name = DRIVER_NAME;
+	indio_dev->dev.parent = dev;
+	indio_dev->info = &twl6030_gpadc_iio_info;
+	indio_dev->modes = INDIO_DIRECT_MODE;
+	indio_dev->channels = pdata->iio_channels;
+	indio_dev->num_channels = pdata->nchannels;
+
+	ret = iio_device_register(indio_dev);
+	if (ret)
+		goto err;
+
+	return ret;
+err:
+	iio_device_free(indio_dev);
+	return ret;
+}
+
+static int twl6030_gpadc_remove(struct platform_device *pdev)
+{
+	struct iio_dev *indio_dev = platform_get_drvdata(pdev);
+
+	twl6030_gpadc_disable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
+	iio_device_unregister(indio_dev);
+	iio_device_free(indio_dev);
+
+	return 0;
+}
+
+static int twl6030_gpadc_suspend(struct device *pdev)
+{
+	int ret;
+
+	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCR,
+				TWL6030_REG_TOGGLE1);
+	if (ret)
+		dev_err(pdev, "error reseting GPADC (%d)!\n", ret);
+
+	return 0;
+};
+
+static int twl6030_gpadc_resume(struct device *pdev)
+{
+	int ret;
+
+	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
+				TWL6030_REG_TOGGLE1);
+	if (ret)
+		dev_err(pdev, "error setting GPADC (%d)!\n", ret);
+
+	return 0;
+};
+
+static const struct dev_pm_ops twl6030_gpadc_pm_ops = {
+	.suspend = twl6030_gpadc_suspend,
+	.resume = twl6030_gpadc_resume,
+};
+
+static struct platform_driver twl6030_gpadc_driver = {
+	.probe		= twl6030_gpadc_probe,
+	.remove		= twl6030_gpadc_remove,
+	.driver		= {
+		.name	= DRIVER_NAME,
+		.owner	= THIS_MODULE,
+		.pm	= &twl6030_gpadc_pm_ops,
+		.of_match_table = of_twl6030_match_tbl,
+	},
+};
+
+module_platform_driver(twl6030_gpadc_driver);
+
+MODULE_ALIAS("platform: " DRIVER_NAME);
+MODULE_AUTHOR("Texas Instruments Inc.");
+MODULE_DESCRIPTION("twl6030 ADC driver");
+MODULE_LICENSE("GPL");
-- 
1.7.9.5

^ permalink raw reply related	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-12 19:42     ` Jonathan Cameron
  0 siblings, 0 replies; 45+ messages in thread
From: Jonathan Cameron @ 2013-07-12 19:42 UTC (permalink / raw)
  To: Oleksandr Kozaruk
  Cc: tony, benoit.cousson, rnayak, peter.ujfalusi, kishon, jic23,
	grant.likely, rob.herring, lars, sameo, ch.naveen, poeschel,
	milo.kim, balajitk, gg, linux-arm-kernel, linux-kernel,
	linux-iio, devicetree-discuss, linux-omap

On 07/12/2013 08:18 AM, Oleksandr Kozaruk wrote:
> The GPADC is general purpose ADC found on TWL6030,
> and TWL6032 PMIC, known also as Phoenix and PhoenixLite.
>
> The TWL6030 and TWL6032 have GPADC with 17 and 19
> channels respectively. Some channels have current
> source and are used for measuring voltage drop
> on resistive load for detecting battery ID resistance,
> or measuring voltage drop on NTC resistors for external
> temperature measurements. Some channels measure voltage,
> (i.e. battery voltage), and have voltage dividers,
> thus, capable to scale voltage. Some channels are dedicated
> for measuring die temperature.
>
> Some channels are calibrated in 2 points, having
> offsets from ideal values kept in trim registers. This
> is used to correct measurements.
>
> The differences between GPADC in TWL6030 and TWL6032:
> - 10 bit vs 12 bit ADC;
> - 17 vs 19 channels;
> - channels have different purpose(i. e. battery voltage
>   channel 8 vs channel 18);
> - trim values are interpreted differently.
>
> The driver exports function returning converted value for
> requested channels.
>
> Based on the driver patched from Balaji TK, Graeme Gregory, Ambresh K,
> Girish S Ghongdemath.

Couple of things:

1) It looks from the driver that a lot of the channels are not measuring
   voltages but rather temperature or currents etc.  If so then their
   types in the channel mask should be appropriately set.  Also if some
   of the channels are entirely internal test networks, what is the benefit
   of exposing them to userspace as readable channels?
   If channels are merely 'suggested' as being used for temperatures etc
   then it is fine to keep them as voltages.

2) You have my sympathy when it comes to the way those trim values are packed
   into the registers.  That is truely hideous ;)

Various trivial little bits but all in all a nice clean driver. Thanks,

Jonathan
>
> Signed-off-by: Balaji T K <balajitk@ti.com>
> Graeme Gregory <gg@slimlogic.co.uk>
> Signed-off-by: Oleksandr Kozaruk <oleksandr.kozaruk@ti.com>
> ---
>  drivers/iio/adc/Kconfig         |    8 +
>  drivers/iio/adc/Makefile        |    1 +
>  drivers/iio/adc/twl6030-gpadc.c | 1019 +++++++++++++++++++++++++++++++++++++++
>  3 files changed, 1028 insertions(+)
>  create mode 100644 drivers/iio/adc/twl6030-gpadc.c
>
> diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig
> index ab0767e6..87d699e 100644
> --- a/drivers/iio/adc/Kconfig
> +++ b/drivers/iio/adc/Kconfig
> @@ -157,4 +157,12 @@ config VIPERBOARD_ADC
>  	  Say yes here to access the ADC part of the Nano River
>  	  Technologies Viperboard.
>
> +config TWL6030_GPADC
> +	tristate "TWL6030 GPADC (General Purpose A/D Convertor) Support"
> +	depends on TWL4030_CORE
> +	default n
> +	help
> +	  Say yes here if you want support for the TWL6030 General Purpose
> +	  A/D Convertor.
Perhaps a little more detail for such a complex device?
> +
>  endmenu
> diff --git a/drivers/iio/adc/Makefile b/drivers/iio/adc/Makefile
> index 0a825be..8b05633 100644
> --- a/drivers/iio/adc/Makefile
> +++ b/drivers/iio/adc/Makefile
> @@ -17,3 +17,4 @@ obj-$(CONFIG_MAX1363) += max1363.o
>  obj-$(CONFIG_TI_ADC081C) += ti-adc081c.o
>  obj-$(CONFIG_TI_AM335X_ADC) += ti_am335x_adc.o
>  obj-$(CONFIG_VIPERBOARD_ADC) += viperboard_adc.o
> +obj-$(CONFIG_TWL6030_GPADC) += twl6030-gpadc.o
> diff --git a/drivers/iio/adc/twl6030-gpadc.c b/drivers/iio/adc/twl6030-gpadc.c
> new file mode 100644
> index 0000000..6ceb789
> --- /dev/null
> +++ b/drivers/iio/adc/twl6030-gpadc.c
> @@ -0,0 +1,1019 @@
> +/*
> + * TWL6030 GPADC module driver
> + *
> + * Copyright (C) 2009-2013 Texas Instruments Inc.
> + * Nishant Kamat <nskamat@ti.com>
> + * Balaji T K <balajitk@ti.com>
> + * Graeme Gregory <gg@slimlogic.co.uk>
> + * Girish S Ghongdemath <girishsg@ti.com>
> + * Ambresh K <ambresh@ti.com>
> + * Oleksandr Kozaruk <oleksandr.kozaruk@ti.com
> + *
> + * Based on twl4030-madc.c
> + * Copyright (C) 2008 Nokia Corporation
> + * Mikko Ylinen <mikko.k.ylinen@nokia.com>
> + *
> + * This program is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU General Public License
> + * version 2 as published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful, but
> + * WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
> + * General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; if not, write to the Free Software
> + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA
> + * 02110-1301 USA
> + *
> + */
> +#include <linux/init.h>
> +#include <linux/interrupt.h>
> +#include <linux/kernel.h>
> +#include <linux/module.h>
> +#include <linux/platform_device.h>
> +#include <linux/of_platform.h>
> +#include <linux/i2c/twl.h>
> +#include <linux/iio/iio.h>
> +#include <linux/iio/sysfs.h>
> +
> +#define DRIVER_NAME		"twl6030_gpadc"
> +
> +#define TWL6030_GPADC_MAX_CHANNELS 17
> +#define TWL6032_GPADC_MAX_CHANNELS 19
> +/* Define this as the biggest of all chips using this driver */
> +#define GPADC_MAX_CHANNELS TWL6032_GPADC_MAX_CHANNELS
> +
> +#define TWL6030_GPADC_CTRL			0x00    /* 0x2e */
> +#define TWL6030_GPADC_CTRL2			0x01	/* 0x2f */
> +
> +#define TWL6030_GPADC_CTRL_P1			0x05
> +#define TWL6030_GPADC_CTRL_P2			0x06
> +
> +#define TWL6032_GPADC_GPSELECT_ISB		0x07
> +#define TWL6032_GPADC_CTRL_P1			0x08
> +
> +#define TWL6032_GPADC_RTCH0_LSB			0x09
> +#define TWL6032_GPADC_RTCH0_MSB			0x0a
> +#define TWL6032_GPADC_RTCH1_LSB			0x0b
> +#define TWL6032_GPADC_RTCH1_MSB			0x0c
> +#define TWL6032_GPADC_GPCH0_LSB			0x0d
> +#define TWL6032_GPADC_GPCH0_MSB			0x0e
> +
> +#define TWL6030_GPADC_CTRL_P1_SP1		BIT(3)
> +#define TWL6030_GPADC_CTRL_P1_EOCRT		BIT(2)
> +#define TWL6030_GPADC_CTRL_P1_EOCP1		BIT(1)
> +#define TWL6030_GPADC_CTRL_P1_BUSY		BIT(0)
> +
> +#define TWL6030_GPADC_CTRL_P2_SP2		BIT(2)
> +#define TWL6030_GPADC_CTRL_P2_EOCP2		BIT(1)
> +#define TWL6030_GPADC_CTRL_P1_BUSY		BIT(0)
> +
> +#define TWL6030_GPADC_EOC_SW			BIT(1)
> +#define TWL6030_GPADC_BUSY			BIT(0)
> +
> +#define TWL6030_GPADC_RTCH0_LSB			(0x07)
> +#define TWL6030_GPADC_GPCH0_LSB			(0x29)
> +
> +/* Fixed channels */
> +#define TWL6030_GPADC_CTRL_TEMP1_EN		BIT(0)    /* input ch 1 */
> +#define TWL6030_GPADC_CTRL_TEMP2_EN		BIT(1)    /* input ch 4 */
> +#define TWL6030_GPADC_CTRL_SCALER_EN		BIT(2)    /* input ch 2 */
> +#define TWL6030_GPADC_CTRL_SCALER_DIV4		BIT(3)
> +#define TWL6030_GPADC_CTRL_SCALER_EN_CH11	BIT(4)    /* input ch 11 */
> +#define TWL6030_GPADC_CTRL_TEMP1_EN_MONITOR	BIT(5)
> +#define TWL6030_GPADC_CTRL_TEMP2_EN_MONITOR	BIT(6)
> +#define TWL6030_GPADC_CTRL_ISOURCE_EN		BIT(7)
> +
> +#define TWL6030_GPADC_CTRL2_REMSENSE_0		BIT(0)
> +#define TWL6030_GPADC_CTRL2_REMSENSE_1		BIT(1)
> +#define TWL6030_GPADC_CTRL2_SCALER_EN_CH18	BIT(2)
> +#define TWL6030_GPADC_CTRL2_VBAT_SCALER_DIV4	BIT(3)
> +
> +#define TWL6030_GPADC_RT_SW1_EOC_MASK		BIT(5)
> +#define TWL6030_GPADC_SW2_EOC_MASK		BIT(6)
> +
> +#define TWL6032_GPADC_RT_EOC_MASK		BIT(4)
> +#define TWL6032_GPADC_SW_EOC_MASK		BIT(5)
> +
> +#define TWL6030_GPADC_TRIM1			0xCD
> +
> +#define TWL6030_REG_TOGGLE1			0x90
> +#define TWL6030_GPADCS				BIT(1)
> +#define TWL6030_GPADCR				BIT(0)
> +
> +/**
> + * struct twl6030_chnl_calib - channel calibration
> + * @gain:		slope coefficient for ideal curve
> + * @gain_error:		gain error
> + * @offset_error:	offset of the real curve
> + */
> +struct twl6030_chnl_calib {
> +	s32 gain;
> +	s32 gain_error;
> +	s32 offset_error;
> +};
> +
> +/**
> + * struct twl6030_ideal_code - GPADC calibration parameters
> + * GPADC is calibrated in two points: close to the beginning and
> + * to the and of the measurable input range
> + *
> + * @code1:	ideal code for the input at the beginning
> + * @code2:	ideal code for at the end of the range
> + * @v1:		voltage input at the beginning(low voltage)
> + * @v2:		voltage input at the end(high voltage)
> + */
> +struct twl6030_ideal_code {
> +	u16 code1;
> +	u16 code2;
> +	u16 v1;
> +	u16 v2;
> +};
> +
> +struct twl6030_gpadc_data;
> +
> +/**
> + * struct twl6030_gpadc_platform_data - platform specific data
> + * @nchannels:		number of GPADC channels
> + * @iio_channels:	iio channels
> + * @twl6030_ideal:	pointer to calibration parameters
> + * @start_conversion:	pointer to ADC start conversion function
> + * @channel_to_reg	pointer to ADC function to convert channel to
> + * 			register address for reading conversion result
> + * @calibrate:		pointer to calibration function
> + */
> +struct twl6030_gpadc_platform_data {
> +	const int nchannels;
> +	const struct iio_chan_spec *iio_channels;
> +	const struct twl6030_ideal_code *ideal;
> +	int (*start_conversion)(int channel);
> +	u8 (*channel_to_reg)(int channel);
> +	int (*calibrate)(struct twl6030_gpadc_data *gpadc);
> +};
> +
> +/**
> + * struct twl6030_gpadc_data - GPADC data
> + * @dev:		device pointer
> + * @lock:		mutual exclusion lock for the structure
> + * @irq_complete:	completion to signal end of conversion
> + * @twl6030_cal_tbl:	pointer to calibration data for each
> + * 			channel with gain error and offset
> + * @pdata:		pointer to device specific data
> + */
> +struct twl6030_gpadc_data {
> +	struct device	*dev;
> +	struct mutex	lock;
> +	struct completion	irq_complete;
> +	struct twl6030_chnl_calib	*twl6030_cal_tbl;
> +	const struct twl6030_gpadc_platform_data *pdata;
> +};
> +
> +/*
> + * channels 11, 12, 13, 15 and 16 have no calibration data
> + * calibration offset is same for channels 1, 3, 4, 5
> + *
> + * The data is taken from GPADC_TRIM registers description.
> + * GPADC_TRIM registers keeps difference between the code measured
> + * at v1 and v2 input voltages and corresponding code1 and code2
> + */
> +static const struct twl6030_ideal_code
> +	twl6030_ideal[TWL6030_GPADC_MAX_CHANNELS] = {
> +	{ /* ch 0, external, battery type, resistor value */
> +		.code1 = 116,
> +		.code2 = 745,
> +		.v1 = 141,
> +		.v2 = 910,
> +	},
> +	{ /* ch 1, external, battery temperature, NTC resistor value */
> +		.code1 = 82,
> +		.code2 = 900,
> +		.v1 = 100,
> +		.v2 = 1100,
> +	},
> +	{ /* ch 2, external, audio accessory/general purpose */
> +		.code1 = 55,
> +		.code2 = 818,
> +		.v1 = 101,
> +		.v2 = 1499,
> +	},
> +	{ /* ch 3, external, general purpose */
> +		.code1 = 82,
> +		.code2 = 900,
> +		.v1 = 100,
> +		.v2 = 1100,
> +	},
> +	{ /* ch 4, external, temperature measurement/general purpose */
> +		.code1 = 82,
> +		.code2 = 900,
> +		.v1 = 100,
> +		.v2 = 1100,
> +	},
> +	{ /* ch 5, external, general purpose */
> +		.code1 = 82,
> +		.code2 = 900,
> +		.v1 = 100,
> +		.v2 = 1100,
> +	},
> +	{ /* ch 6, external, general purpose */
> +		.code1 = 82,
> +		.code2 = 900,
> +		.v1 = 100,
> +		.v2 = 1100,
> +	},
> +	{ /* ch 7, internal, main battery */
> +		.code1 = 614,
> +		.code2 = 941,
> +		.v1 = 3001,
> +		.v2 = 4599,
> +	},
> +	{ /* ch 8, internal, backup battery */
> +		.code1 = 82,
> +		.code2 = 688,
> +		.v1 = 501,
> +		.v2 = 4203,
> +	},
> +	{ /* ch 9, internal, external charger input */
> +		.code1 = 182,
> +		.code2 = 818,
> +		.v1 = 2001,
> +		.v2 = 8996,
> +	},
> +	{ /* ch 10, internal, VBUS */
> +		.code1 = 149,
> +		.code2 = 818,
> +		.v1 = 1001,
> +		.v2 = 5497,
> +	},
> +	{},	/* ch 11, internal, VBUS charging current */
> +	{},	/* ch 12, internal, Die temperature */
> +	{},	/* ch 13, internal, Die temperature */
> +	{ /* ch 14, internal, USB ID line */
> +		.code1 = 48,
> +		.code2 = 714,
> +		.v1 = 323,
> +		.v2 = 4800,
> +	},
> +	{},	/* ch 15, internal, test network */
> +	{},	/* ch 16, internal, test network */
> +};
> +
> +static const struct twl6030_ideal_code
> +			twl6032_ideal[TWL6032_GPADC_MAX_CHANNELS] = {

[0] = {
will make this clearer to read.
> +	{ /* ch 0, external, battery type, resistor value */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 1, external, battery temperature, NTC resistor value */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 2, external, audio accessory/general purpose */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 660,
> +		.v2 = 1500,
> +	},
> +	{ /* ch 3, external, temperature with external diode/general purpose */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 4, external, temperature measurement/general purpose */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 5, external, general purpose */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 6, external, general purpose */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch7, internal, system supply */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 2200,
> +		.v2 = 5000,
> +	},
> +	{ /* ch8, internal, backup battery */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 2200,
> +		.v2 = 5000,
> +	},
> +	{ /* ch 9, internal, external charger input */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 3960,
> +		.v2 = 9000,
> +	},
> +	{ /* ch10, internal, VBUS */
> +		.code1 = 150,
> +		.code2 = 751,
> +		.v1 = 1000,
> +		.v2 = 5000,
> +	},
> +	{ /* ch 11, internal, VBUS DC-DC output current */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 660,
> +		.v2 = 1500,
> +	},
> +	{ /* ch 12, internal, Die temperature */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 13, internal, Die temperature */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 14, internal, USB ID line */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 2420,
> +		.v2 = 5500,
> +	},
> +	{},	/* ch 15, internal, test network */
> +	{},	/* ch 16, internal, test network */
> +	{},	/* ch 17, internal, battery charging current */
> +	{ /* ch 18, internal, battery voltage */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 2200,
> +		.v2 = 5000,
> +	},
> +};
> +
> +static inline int twl6030_gpadc_write(u8 reg, u8 val)
> +{
> +	return twl_i2c_write_u8(TWL6030_MODULE_GPADC, val, reg);
> +}
> +
> +static inline int twl6030_gpadc_read(u8 *val, u8 reg)
> +{
> +
> +	return twl_i2c_read(TWL6030_MODULE_GPADC, val, reg, 2);
> +}
> +
> +static int twl6030_gpadc_enable_irq(u8 mask)
> +{
> +	int ret;
> +
> +	ret = twl6030_interrupt_unmask(mask, REG_INT_MSK_LINE_B);
> +	ret |= twl6030_interrupt_unmask(mask, REG_INT_MSK_STS_B);
I'd just roll these into 1 if it makes sense not to check the value
of ret in between.
> +
> +	return ret;
> +}
> +
> +static void twl6030_gpadc_disable_irq(u8 mask)
> +{
> +	twl6030_interrupt_mask(mask, REG_INT_MSK_LINE_B);
> +	twl6030_interrupt_mask(mask, REG_INT_MSK_STS_B);
> +}
> +
> +static irqreturn_t twl6030_gpadc_irq_handler(int irq, void *_gpadc)
> +{
> +	struct twl6030_gpadc_data *gpadc = _gpadc;
> +
> +	complete(&gpadc->irq_complete);
> +
> +	return IRQ_HANDLED;
> +}
> +
> +static int twl6030_start_conversion(int channel)
> +{
> +	return twl6030_gpadc_write(TWL6030_GPADC_CTRL_P1,
> +					TWL6030_GPADC_CTRL_P1_SP1);
> +}
> +
> +static int twl6032_start_conversion(int channel)
> +{
> +	int ret;
> +
> +	ret = twl6030_gpadc_write(TWL6032_GPADC_GPSELECT_ISB, channel);
> +	if (ret)
> +		return ret;
> +
> +	return twl6030_gpadc_write(TWL6032_GPADC_CTRL_P1,
> +						TWL6030_GPADC_CTRL_P1_SP1);
> +}
> +
> +static u8 twl6030_channel_to_reg(int channel)
> +{
> +	return TWL6030_GPADC_GPCH0_LSB + 2 * channel;
> +}
> +
> +static u8 twl6032_channel_to_reg(int channel)
> +{
> +	return TWL6032_GPADC_GPCH0_LSB;
> +}
> +
> +static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,
> +		int channel, int *res)
> +{
> +	u8 reg = gpadc->pdata->channel_to_reg(channel);
> +	u8 val[2];

le16 val;
> +	int ret;
> +
ret = twl6030_gpadc_read(val, reg);
(note that (reg, val) ordering of parameters would be a more common choice)

> +	ret = twl6030_gpadc_read(val, reg);
> +	if (ret) {
> +		dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);
> +		return ret;
> +	}
> +
res = le16_to_cpup(val);

> +	*res = (val[1] << 8) | val[0];
> +
> +	return ret;
> +}
> +
> +static int twl6030_gpadc_get_processed(struct twl6030_gpadc_data *gpadc,
> +		int channel, int *val)
> +{
> +	int raw_code;
> +	int corrected_code;
> +	int channel_value;
> +	int ret;
> +
> +	ret = twl6030_gpadc_get_raw(gpadc, channel, &raw_code);
> +	if (ret)
> +		return ret;
> +

Might first thought on seeing this is that it would be better to return
raw for all channels and provide the scale and offset info_mask elements
where possible rather than doing the conversion in the kernel.  Note we
allow really quite a bit of precision on those values so I doubt it
will be a problem.

If nothing else it would make everything rather more consistent.


> +	corrected_code = ((raw_code * 1000) -
> +		gpadc->twl6030_cal_tbl[channel].offset_error) /
> +		gpadc->twl6030_cal_tbl[channel].gain_error;
> +
> +	channel_value = corrected_code *
> +			gpadc->twl6030_cal_tbl[channel].gain;
> +
> +	/* Shift back into mV range */
> +	channel_value /= 1000;
> +
> +	dev_dbg(gpadc->dev, "GPADC raw: %d", raw_code);
> +	dev_dbg(gpadc->dev, "GPADC cor: %d", corrected_code);
> +	dev_dbg(gpadc->dev, "GPADC val: %d", channel_value);
> +
> +	*val = channel_value;
> +
> +	return ret;
> +}
> +
> +static int twl6030_gpadc_read_raw(struct iio_dev *indio_dev,
> +			     const struct iio_chan_spec *chan,
> +			     int *val, int *val2, long mask)
> +{
> +	struct twl6030_gpadc_data *gpadc = iio_priv(indio_dev);
> +	int ret = -EINVAL;
> +
> +	mutex_lock(&gpadc->lock);
> +
> +	ret = gpadc->pdata->start_conversion(chan->channel);
> +	if (ret) {
> +		dev_err(gpadc->dev, "failed to start conversion\n");
> +		goto err;
> +	}
> +	/* wait for conversion to complete */
> +	wait_for_completion_interruptible_timeout(&gpadc->irq_complete,
> +						msecs_to_jiffies(5000));
> +
> +	switch (mask) {
> +	case IIO_CHAN_INFO_RAW:
> +		ret = twl6030_gpadc_get_raw(gpadc, chan->channel, val);
> +		ret = ret ? -EIO : IIO_VAL_INT;
> +		break;
> +
> +	case IIO_CHAN_INFO_PROCESSED:
> +		ret = twl6030_gpadc_get_processed(gpadc, chan->channel, val);
> +		ret = ret ? -EIO : IIO_VAL_INT;
> +		break;
> +
> +	default:
> +		break;
> +	}
> +err:
> +	mutex_unlock(&gpadc->lock);
> +
> +	return ret;
> +}
> +
> +static void twl6030_calibrate_channel(struct twl6030_gpadc_data *gpadc,
> +					int channel, int d1, int d2)
> +{
> +	int b, k, gain, x1, x2;
> +	const struct twl6030_ideal_code *ideal = gpadc->pdata->ideal;
> +
> +	/* Gain */
> +	gain = ((ideal[channel].v2 - ideal[channel].v1) * 1000) /
> +		(ideal[channel].code2 - ideal[channel].code1);
> +
> +	x1 = ideal[channel].code1;
> +	x2 = ideal[channel].code2;
> +
> +	/* k */
> +	k = 1000 + (((d2 - d1) * 1000) / (x2 - x1));
> +
> +	/* b */
> +	b = (d1 * 1000) - (k - 1000) * x1;
> +
> +	gpadc->twl6030_cal_tbl[channel].gain = gain;
> +	gpadc->twl6030_cal_tbl[channel].gain_error = k;
> +	gpadc->twl6030_cal_tbl[channel].offset_error = b;
> +
> +	dev_dbg(gpadc->dev, "GPADC d1   for Chn: %d = %d\n", channel, d1);
> +	dev_dbg(gpadc->dev, "GPADC d2   for Chn: %d = %d\n", channel, d2);
> +	dev_dbg(gpadc->dev, "GPADC x1   for Chn: %d = %d\n", channel, x1);
> +	dev_dbg(gpadc->dev, "GPADC x2   for Chn: %d = %d\n", channel, x2);
> +	dev_dbg(gpadc->dev, "GPADC Gain for Chn: %d = %d\n", channel, gain);
> +	dev_dbg(gpadc->dev, "GPADC k    for Chn: %d = %d\n", channel, k);
> +	dev_dbg(gpadc->dev, "GPADC b    for Chn: %d = %d\n", channel, b);
> +}
> +
> +static inline int twl6030_gpadc_get_trim_offset(s8 d)
> +{
> +	int sign;
> +
> +	/*
> +	 * XXX NOTE!
> +	 * bit 0 - sign, bit 7 - reserved, 6..1 - trim value
> +	 * though, the documentation states that trim value
> +	 * is absolute value, the correct conversion results are
> +	 * obtained if the value is interpreted as 2's complement.
> +	 */
I'd be tempted to reorder this and then use the sign extend standard
functions.
	temp = ((d & 0x7f) >> 1) | ((d & 1) << 7);
	return sign_extend32(temp, 7);

(or something like that)

> +	sign = d & 1;
> +	d = (d & 0x7f) >> 1;
> +
> +	return sign ? (d | 0xc0) : d;
> +}
> +
> +static int twl6030_calibration(struct twl6030_gpadc_data *gpadc)
> +{
> +	int ret;
> +	int chn;
> +	u8 trim_regs[TWL6030_GPADC_MAX_CHANNELS];
> +	s8 d1, d2;
> +
> +	/*
> +	 * for calibration two measurements have been performed at
> +	 * factory, for some channels, during the production test and
> +	 * have been stored in registers. This two stored values are
> +	 * used to correct the measurements. The values represent
> +	 * offsets for the given input from the output on ideal curve.
> +	 */
> +	ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs,
> +			TWL6030_GPADC_TRIM1, 16);
> +	if (ret < 0) {
> +		dev_err(gpadc->dev, "calibration failed\n");
> +		return ret;
> +	}
> +
> +	for (chn = 0; chn < TWL6030_GPADC_MAX_CHANNELS; chn++) {
> +
> +		switch (chn) {
> +		case 0:
> +			d1 = trim_regs[0];
> +			d2 = trim_regs[1];
> +			break;
> +		case 1:
> +			d1 = trim_regs[4];
> +			d2 = trim_regs[5];
> +			break;
> +		case 2:
> +			d1 = trim_regs[12];
> +			d2 = trim_regs[13];
> +			break;
Given you give up entirely on ordering in the next calibration function
I'd be tempted to do the same here and have all the d1 = ...[4] etc in one
case statement.
> +		case 3:
> +		case 4:
> +		case 5:
> +		case 6:
> +			d1 = trim_regs[4];
> +			d2 = trim_regs[5];
> +			break;
> +		case 7:
> +			d1 = trim_regs[6];
> +			d2 = trim_regs[7];
> +			break;
> +		case 8:
> +			d1 = trim_regs[2];
> +			d2 = trim_regs[3];
> +			break;
> +		case 9:
> +			d1 = trim_regs[8];
> +			d2 = trim_regs[9];
> +			break;
> +		case 10:
> +			d1 = trim_regs[10];
> +			d2 = trim_regs[11];
> +			break;
> +		case 14:
> +			d1 = trim_regs[14];
> +			d2 = trim_regs[15];
> +			break;
> +		default:
> +			continue;
> +		}
> +
> +		d1 = twl6030_gpadc_get_trim_offset(d1);
> +		d2 = twl6030_gpadc_get_trim_offset(d2);
> +
> +		twl6030_calibrate_channel(gpadc, chn, d1, d2);
> +	}
> +
> +	return 0;
> +}
> +
> +static int twl6032_calibration(struct twl6030_gpadc_data *gpadc)
> +{
> +	int chn, d1 = 0, d2 = 0, temp;
> +	u8 trim_regs[17];
> +	int ret;
> +
> +	ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs + 1,
> +			TWL6030_GPADC_TRIM1, 16);
> +	if (ret < 0) {
> +		dev_err(gpadc->dev, "calibration failed\n");
> +		return ret;
> +	}
> +
/*
 * Loop
please for kernel code.
> +	/* Loop to calculate the value needed for returning voltages from
> +	 * GPADC not values.
> +	 *
> +	 * gain is calculated to 3 decimal places fixed point.
> +	 */
> +	for (chn = 0; chn < TWL6032_GPADC_MAX_CHANNELS; chn++) {
> +
> +		switch (chn) {
> +		case 0:
> +		case 1:
> +		case 2:
> +		case 3:
> +		case 4:
> +		case 5:
> +		case 6:
> +		case 11:
> +		case 12:
> +		case 13:
> +		case 14:
> +			/* D1 */
> +			d1 = (trim_regs[3] & 0x1F) << 2;
> +			d1 |= (trim_regs[1] & 0x06) >> 1;
> +			if (trim_regs[1] & 0x01)
> +				d1 = -d1;
> +
> +			/* D2 */
> +			d2 = (trim_regs[4] & 0x3F) << 2;
> +			d2 |= (trim_regs[2] & 0x06) >> 1;
> +			if (trim_regs[2] & 0x01)
> +				d2 = -d2;
> +			break;
> +		case 8:
No coment on your code, but this is an 'interesting' bit
of bit packing...
I did vaguely wonder if this could be mapped to a big lookup table,
but having tried it I think I end up with something almost as tricky
to read.. Oh well that was a fun 10 minute diversion ;)

> +			/* D1 */
> +			temp = (trim_regs[3] & 0x1F) << 2;
> +			temp |= (trim_regs[1] & 0x06) >> 1;
> +			if (trim_regs[1] & 0x01)
> +				temp = -temp;
> +
> +			d1 = (trim_regs[8] & 0x18) << 1;
> +			d1 |= (trim_regs[7] & 0x1E) >> 1;
> +			if (trim_regs[7] & 0x01)
> +				d1 = -d1;
> +
> +			d1 += temp;
> +
> +			/* D2 */
> +			temp = (trim_regs[4] & 0x3F) << 2;
> +			temp |= (trim_regs[2] & 0x06) >> 1;
> +			if (trim_regs[2] & 0x01)
> +				temp = -temp;
> +
> +			d2 = (trim_regs[10] & 0x1F) << 2;
> +			d2 |= (trim_regs[8] & 0x06) >> 1;
> +			if (trim_regs[8] & 0x01)
> +				d2 = -d2;
> +
> +			d2 += temp;
> +			break;
> +		case 9:
> +			/* D1 */
> +			temp = (trim_regs[3] & 0x1F) << 2;
> +			temp |= (trim_regs[1] & 0x06) >> 1;
> +			if (trim_regs[1] & 0x01)
> +				temp = -temp;
> +
> +			d1 = (trim_regs[14] & 0x18) << 1;
> +			d1 |= (trim_regs[12] & 0x1E) >> 1;
> +			if (trim_regs[12] & 0x01)
> +				d1 = -d1;
> +
> +			d1 += temp;
> +
> +			/* D2 */
> +			temp = (trim_regs[4] & 0x3F) << 2;
> +			temp |= (trim_regs[2] & 0x06) >> 1;
> +			if (trim_regs[2] & 0x01)
> +				temp = -temp;
> +
> +			d2 = (trim_regs[16] & 0x1F) << 2;
> +			d2 |= (trim_regs[14] & 0x06) >> 1;
> +			if (trim_regs[14] & 0x01)
> +				d2 = -d2;
> +
> +			d2 += temp;
> +		case 10:
> +			/* D1 */
> +			d1 = (trim_regs[11] & 0x0F) << 3;
> +			d1 |= (trim_regs[9] & 0x0E) >> 1;
> +			if (trim_regs[9] & 0x01)
> +				d1 = -d1;
> +
> +			/* D2 */
> +			d2 = (trim_regs[15] & 0x0F) << 3;
> +			d2 |= (trim_regs[13] & 0x0E) >> 1;
> +			if (trim_regs[13] & 0x01)
> +				d2 = -d2;
> +			break;
> +		case 7:
> +		case 18:
> +			/* D1 */
> +			temp = (trim_regs[3] & 0x1F) << 2;
> +			temp |= (trim_regs[1] & 0x06) >> 1;
> +			if (trim_regs[1] & 0x01)
> +				temp = -temp;
> +
> +			d1 = (trim_regs[5] & 0x7E) >> 1;
> +			if (trim_regs[5] & 0x01)
> +				d1 = -d1;
> +
> +			d1 += temp;
> +
> +			/* D2 */
> +			temp = (trim_regs[4] & 0x3F) << 2;
> +			temp |= (trim_regs[2] & 0x06) >> 1;
> +			if (trim_regs[2] & 0x01)
> +				temp = -temp;
> +
> +			d2 = (trim_regs[6] & 0xFE) >> 1;
> +			if (trim_regs[6] & 0x01)
> +				d2 = -d2;
> +
> +			d2 += temp;
> +			break;
> +		default:
> +			/* No data for other channels */
> +			continue;
> +		}
> +
> +		twl6030_calibrate_channel(gpadc, chn, d1, d2);
> +	}
> +
> +	return 0;
> +}
> +
> +#define TWL6030_GPADC_CHAN(chn, chan_info) {		\
> +	.type = IIO_VOLTAGE,				\
A lot of the channels appear not to be measuring voltages.
So they should all be appropriately specified.

> +	.channel = chn,					\

Unless you are doing buffered output this stuff is of no interest to
anyone so drop .scan_type from here.

> +	.scan_type = {					\
> +		.sign = 'u',				\
> +		.storagebits = 16,			\
> +	},						\
> +	.info_mask = BIT(chan_info),			\
> +	.info_mask_separate = BIT(chan_info), 		\
> +	.indexed = 1,					\
> +}
> +
> +static const struct iio_chan_spec twl6030_gpadc_iio_channels[] = {
> +	TWL6030_GPADC_CHAN(0, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(1, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(2, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(3, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(4, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(5, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(6, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(7, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(8, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(9, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(10, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(11, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(12, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(13, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(14, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(15, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(16, IIO_CHAN_INFO_RAW),
> +};
> +
> +static const struct iio_chan_spec twl6032_gpadc_iio_channels[] = {
> +	TWL6030_GPADC_CHAN(0, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(1, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(2, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(3, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(4, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(5, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(6, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(7, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(8, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(9, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(10, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(11, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(12, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(13, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(14, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(15, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(16, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(17, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(18, IIO_CHAN_INFO_PROCESSED),
> +};
> +
> +static const struct iio_info twl6030_gpadc_iio_info = {
> +	.read_raw = &twl6030_gpadc_read_raw,
> +	.driver_module = THIS_MODULE,
> +};
> +
> +static const struct twl6030_gpadc_platform_data twl6030_pdata = {
> +	.iio_channels = twl6030_gpadc_iio_channels,
> +	.nchannels = TWL6030_GPADC_MAX_CHANNELS,
> +	.ideal = twl6030_ideal,
> +	.start_conversion = twl6030_start_conversion,
> +	.channel_to_reg = twl6030_channel_to_reg,
> +	.calibrate = twl6030_calibration,
> +};
> +
> +static const struct twl6030_gpadc_platform_data twl6032_pdata = {
> +	.iio_channels = twl6032_gpadc_iio_channels,
> +	.nchannels = TWL6032_GPADC_MAX_CHANNELS,
> +	.ideal = twl6032_ideal,
> +	.start_conversion = twl6032_start_conversion,
> +	.channel_to_reg = twl6032_channel_to_reg,
> +	.calibrate = twl6032_calibration,
> +};
> +
> +static const struct of_device_id of_twl6030_match_tbl[] = {
> +	{
> +		.compatible = "ti,twl6030_gpadc",
> +		.data = &twl6030_pdata,
> +	},
> +	{
> +		.compatible = "ti,twl6032_gpadc",
> +		.data = &twl6032_pdata,
> +	},
> +	{ /* end */ }
> +};
> +
> +static int twl6030_gpadc_probe(struct platform_device *pdev)
> +{
> +	struct device *dev = &pdev->dev;
> +	struct twl6030_gpadc_data *gpadc;
> +	const struct twl6030_gpadc_platform_data *pdata;
> +	const struct of_device_id *match;
> +	struct iio_dev *indio_dev;
> +	int irq;
> +	int ret = 0;
> +
> +	match = of_match_device(of_match_ptr(of_twl6030_match_tbl), dev);
> +	pdata = match ? match->data : dev->platform_data;
> +
> +	if (!pdata)
> +		return -EINVAL;
> +
> +	indio_dev = iio_device_alloc(sizeof(struct twl6030_gpadc_data));
> +	if (!indio_dev) {
> +		dev_err(dev, "failed allocating iio device\n");
> +		ret = -ENOMEM;
> +	}
> +
> +	gpadc = iio_priv(indio_dev);
> +
> +	gpadc->twl6030_cal_tbl = devm_kzalloc(dev,
> +					sizeof(struct twl6030_chnl_calib) *
> +					pdata->nchannels, GFP_KERNEL);
> +	if (!gpadc->twl6030_cal_tbl)
> +		goto err;
> +
> +	gpadc->dev = dev;
> +	gpadc->pdata = pdata;
> +
> +	platform_set_drvdata(pdev, indio_dev);
> +	mutex_init(&gpadc->lock);
> +	init_completion(&gpadc->irq_complete);
> +
> +	ret = pdata->calibrate(gpadc);
> +	if (ret < 0) {
> +		dev_err(&pdev->dev, "failed to read calibration registers\n");
> +		goto err;
> +	}
> +
> +	irq = platform_get_irq(pdev, 0);
> +	if (irq < 0) {
> +		dev_err(&pdev->dev, "failed to get irq\n");
> +		goto err;
> +	}
> +
> +	ret = devm_request_threaded_irq(dev, irq, NULL,
> +					twl6030_gpadc_irq_handler,
> +					IRQF_ONESHOT, "twl6030_gpadc", gpadc);
> +	if (ret) {
> +		dev_dbg(&pdev->dev, "could not request irq\n");
> +		goto err;
> +	}
> +
> +	ret = twl6030_gpadc_enable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
> +	if (ret < 0) {
> +		dev_err(&pdev->dev, "failed to enable GPADC interrupt\n");
> +		goto err;
> +	}
> +	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
> +					TWL6030_REG_TOGGLE1);
> +	if (ret < 0) {
> +		dev_err(&pdev->dev, "failed to enable GPADC module\n");
> +		goto err;
> +	}
> +
> +	indio_dev->name = DRIVER_NAME;
> +	indio_dev->dev.parent = dev;
> +	indio_dev->info = &twl6030_gpadc_iio_info;
> +	indio_dev->modes = INDIO_DIRECT_MODE;
> +	indio_dev->channels = pdata->iio_channels;
> +	indio_dev->num_channels = pdata->nchannels;
> +
> +	ret = iio_device_register(indio_dev);
> +	if (ret)
> +		goto err;
> +
> +	return ret;
> +err:
> +	iio_device_free(indio_dev);
> +	return ret;
> +}
> +
> +static int twl6030_gpadc_remove(struct platform_device *pdev)
> +{
> +	struct iio_dev *indio_dev = platform_get_drvdata(pdev);
> +
> +	twl6030_gpadc_disable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
> +	iio_device_unregister(indio_dev);
> +	iio_device_free(indio_dev);
> +
> +	return 0;
> +}
> +
> +static int twl6030_gpadc_suspend(struct device *pdev)
> +{
> +	int ret;
> +
> +	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCR,
> +				TWL6030_REG_TOGGLE1);
> +	if (ret)
> +		dev_err(pdev, "error reseting GPADC (%d)!\n", ret);
> +
> +	return 0;
> +};
> +
> +static int twl6030_gpadc_resume(struct device *pdev)
> +{
> +	int ret;
> +
> +	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
> +				TWL6030_REG_TOGGLE1);
> +	if (ret)
> +		dev_err(pdev, "error setting GPADC (%d)!\n", ret);
> +
> +	return 0;
> +};
> +
> +static const struct dev_pm_ops twl6030_gpadc_pm_ops = {
> +	.suspend = twl6030_gpadc_suspend,
> +	.resume = twl6030_gpadc_resume,
> +};
> +
> +static struct platform_driver twl6030_gpadc_driver = {
> +	.probe		= twl6030_gpadc_probe,
> +	.remove		= twl6030_gpadc_remove,
> +	.driver		= {
> +		.name	= DRIVER_NAME,
> +		.owner	= THIS_MODULE,
> +		.pm	= &twl6030_gpadc_pm_ops,
> +		.of_match_table = of_twl6030_match_tbl,
> +	},
> +};
> +
> +module_platform_driver(twl6030_gpadc_driver);
> +
> +MODULE_ALIAS("platform: " DRIVER_NAME);
> +MODULE_AUTHOR("Texas Instruments Inc.");
> +MODULE_DESCRIPTION("twl6030 ADC driver");
> +MODULE_LICENSE("GPL");
>

^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-12 19:42     ` Jonathan Cameron
  0 siblings, 0 replies; 45+ messages in thread
From: Jonathan Cameron @ 2013-07-12 19:42 UTC (permalink / raw)
  To: Oleksandr Kozaruk
  Cc: tony-4v6yS6AI5VpBDgjK7y7TUQ,
	benoit.cousson-QSEj5FYQhm4dnm+yROfE0A, rnayak-l0cyMroinI0,
	peter.ujfalusi-l0cyMroinI0, kishon-l0cyMroinI0,
	jic23-KWPb1pKIrIJaa/9Udqfwiw,
	grant.likely-QSEj5FYQhm4dnm+yROfE0A,
	rob.herring-bsGFqQB8/DxBDgjK7y7TUQ, lars-Qo5EllUWu/uELgA04lAiVw,
	sameo-VuQAYsv1563Yd54FQh9/CA, ch.naveen-Sze3O3UU22JBDgjK7y7TUQ,
	poeschel-Xtl8qvBWbHwb1SvskN2V4Q, milo.kim-l0cyMroinI0,
	balajitk-l0cyMroinI0, gg-kDsPt+C1G03kYMGBc/C6ZA,
	linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA,
	linux-iio-u79uwXL29TY76Z2rM5mHXA,
	devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ,
	linux-omap-u79uwXL29TY76Z2rM5mHXA

On 07/12/2013 08:18 AM, Oleksandr Kozaruk wrote:
> The GPADC is general purpose ADC found on TWL6030,
> and TWL6032 PMIC, known also as Phoenix and PhoenixLite.
>
> The TWL6030 and TWL6032 have GPADC with 17 and 19
> channels respectively. Some channels have current
> source and are used for measuring voltage drop
> on resistive load for detecting battery ID resistance,
> or measuring voltage drop on NTC resistors for external
> temperature measurements. Some channels measure voltage,
> (i.e. battery voltage), and have voltage dividers,
> thus, capable to scale voltage. Some channels are dedicated
> for measuring die temperature.
>
> Some channels are calibrated in 2 points, having
> offsets from ideal values kept in trim registers. This
> is used to correct measurements.
>
> The differences between GPADC in TWL6030 and TWL6032:
> - 10 bit vs 12 bit ADC;
> - 17 vs 19 channels;
> - channels have different purpose(i. e. battery voltage
>   channel 8 vs channel 18);
> - trim values are interpreted differently.
>
> The driver exports function returning converted value for
> requested channels.
>
> Based on the driver patched from Balaji TK, Graeme Gregory, Ambresh K,
> Girish S Ghongdemath.

Couple of things:

1) It looks from the driver that a lot of the channels are not measuring
   voltages but rather temperature or currents etc.  If so then their
   types in the channel mask should be appropriately set.  Also if some
   of the channels are entirely internal test networks, what is the benefit
   of exposing them to userspace as readable channels?
   If channels are merely 'suggested' as being used for temperatures etc
   then it is fine to keep them as voltages.

2) You have my sympathy when it comes to the way those trim values are packed
   into the registers.  That is truely hideous ;)

Various trivial little bits but all in all a nice clean driver. Thanks,

Jonathan
>
> Signed-off-by: Balaji T K <balajitk-l0cyMroinI0@public.gmane.org>
> Graeme Gregory <gg-kDsPt+C1G03kYMGBc/C6ZA@public.gmane.org>
> Signed-off-by: Oleksandr Kozaruk <oleksandr.kozaruk-l0cyMroinI0@public.gmane.org>
> ---
>  drivers/iio/adc/Kconfig         |    8 +
>  drivers/iio/adc/Makefile        |    1 +
>  drivers/iio/adc/twl6030-gpadc.c | 1019 +++++++++++++++++++++++++++++++++++++++
>  3 files changed, 1028 insertions(+)
>  create mode 100644 drivers/iio/adc/twl6030-gpadc.c
>
> diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig
> index ab0767e6..87d699e 100644
> --- a/drivers/iio/adc/Kconfig
> +++ b/drivers/iio/adc/Kconfig
> @@ -157,4 +157,12 @@ config VIPERBOARD_ADC
>  	  Say yes here to access the ADC part of the Nano River
>  	  Technologies Viperboard.
>
> +config TWL6030_GPADC
> +	tristate "TWL6030 GPADC (General Purpose A/D Convertor) Support"
> +	depends on TWL4030_CORE
> +	default n
> +	help
> +	  Say yes here if you want support for the TWL6030 General Purpose
> +	  A/D Convertor.
Perhaps a little more detail for such a complex device?
> +
>  endmenu
> diff --git a/drivers/iio/adc/Makefile b/drivers/iio/adc/Makefile
> index 0a825be..8b05633 100644
> --- a/drivers/iio/adc/Makefile
> +++ b/drivers/iio/adc/Makefile
> @@ -17,3 +17,4 @@ obj-$(CONFIG_MAX1363) += max1363.o
>  obj-$(CONFIG_TI_ADC081C) += ti-adc081c.o
>  obj-$(CONFIG_TI_AM335X_ADC) += ti_am335x_adc.o
>  obj-$(CONFIG_VIPERBOARD_ADC) += viperboard_adc.o
> +obj-$(CONFIG_TWL6030_GPADC) += twl6030-gpadc.o
> diff --git a/drivers/iio/adc/twl6030-gpadc.c b/drivers/iio/adc/twl6030-gpadc.c
> new file mode 100644
> index 0000000..6ceb789
> --- /dev/null
> +++ b/drivers/iio/adc/twl6030-gpadc.c
> @@ -0,0 +1,1019 @@
> +/*
> + * TWL6030 GPADC module driver
> + *
> + * Copyright (C) 2009-2013 Texas Instruments Inc.
> + * Nishant Kamat <nskamat-l0cyMroinI0@public.gmane.org>
> + * Balaji T K <balajitk-l0cyMroinI0@public.gmane.org>
> + * Graeme Gregory <gg-kDsPt+C1G03kYMGBc/C6ZA@public.gmane.org>
> + * Girish S Ghongdemath <girishsg-l0cyMroinI0@public.gmane.org>
> + * Ambresh K <ambresh-l0cyMroinI0@public.gmane.org>
> + * Oleksandr Kozaruk <oleksandr.kozaruk-l0cyMroinI0@public.gmane.org
> + *
> + * Based on twl4030-madc.c
> + * Copyright (C) 2008 Nokia Corporation
> + * Mikko Ylinen <mikko.k.ylinen-xNZwKgViW5gAvxtiuMwx3w@public.gmane.org>
> + *
> + * This program is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU General Public License
> + * version 2 as published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful, but
> + * WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
> + * General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; if not, write to the Free Software
> + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA
> + * 02110-1301 USA
> + *
> + */
> +#include <linux/init.h>
> +#include <linux/interrupt.h>
> +#include <linux/kernel.h>
> +#include <linux/module.h>
> +#include <linux/platform_device.h>
> +#include <linux/of_platform.h>
> +#include <linux/i2c/twl.h>
> +#include <linux/iio/iio.h>
> +#include <linux/iio/sysfs.h>
> +
> +#define DRIVER_NAME		"twl6030_gpadc"
> +
> +#define TWL6030_GPADC_MAX_CHANNELS 17
> +#define TWL6032_GPADC_MAX_CHANNELS 19
> +/* Define this as the biggest of all chips using this driver */
> +#define GPADC_MAX_CHANNELS TWL6032_GPADC_MAX_CHANNELS
> +
> +#define TWL6030_GPADC_CTRL			0x00    /* 0x2e */
> +#define TWL6030_GPADC_CTRL2			0x01	/* 0x2f */
> +
> +#define TWL6030_GPADC_CTRL_P1			0x05
> +#define TWL6030_GPADC_CTRL_P2			0x06
> +
> +#define TWL6032_GPADC_GPSELECT_ISB		0x07
> +#define TWL6032_GPADC_CTRL_P1			0x08
> +
> +#define TWL6032_GPADC_RTCH0_LSB			0x09
> +#define TWL6032_GPADC_RTCH0_MSB			0x0a
> +#define TWL6032_GPADC_RTCH1_LSB			0x0b
> +#define TWL6032_GPADC_RTCH1_MSB			0x0c
> +#define TWL6032_GPADC_GPCH0_LSB			0x0d
> +#define TWL6032_GPADC_GPCH0_MSB			0x0e
> +
> +#define TWL6030_GPADC_CTRL_P1_SP1		BIT(3)
> +#define TWL6030_GPADC_CTRL_P1_EOCRT		BIT(2)
> +#define TWL6030_GPADC_CTRL_P1_EOCP1		BIT(1)
> +#define TWL6030_GPADC_CTRL_P1_BUSY		BIT(0)
> +
> +#define TWL6030_GPADC_CTRL_P2_SP2		BIT(2)
> +#define TWL6030_GPADC_CTRL_P2_EOCP2		BIT(1)
> +#define TWL6030_GPADC_CTRL_P1_BUSY		BIT(0)
> +
> +#define TWL6030_GPADC_EOC_SW			BIT(1)
> +#define TWL6030_GPADC_BUSY			BIT(0)
> +
> +#define TWL6030_GPADC_RTCH0_LSB			(0x07)
> +#define TWL6030_GPADC_GPCH0_LSB			(0x29)
> +
> +/* Fixed channels */
> +#define TWL6030_GPADC_CTRL_TEMP1_EN		BIT(0)    /* input ch 1 */
> +#define TWL6030_GPADC_CTRL_TEMP2_EN		BIT(1)    /* input ch 4 */
> +#define TWL6030_GPADC_CTRL_SCALER_EN		BIT(2)    /* input ch 2 */
> +#define TWL6030_GPADC_CTRL_SCALER_DIV4		BIT(3)
> +#define TWL6030_GPADC_CTRL_SCALER_EN_CH11	BIT(4)    /* input ch 11 */
> +#define TWL6030_GPADC_CTRL_TEMP1_EN_MONITOR	BIT(5)
> +#define TWL6030_GPADC_CTRL_TEMP2_EN_MONITOR	BIT(6)
> +#define TWL6030_GPADC_CTRL_ISOURCE_EN		BIT(7)
> +
> +#define TWL6030_GPADC_CTRL2_REMSENSE_0		BIT(0)
> +#define TWL6030_GPADC_CTRL2_REMSENSE_1		BIT(1)
> +#define TWL6030_GPADC_CTRL2_SCALER_EN_CH18	BIT(2)
> +#define TWL6030_GPADC_CTRL2_VBAT_SCALER_DIV4	BIT(3)
> +
> +#define TWL6030_GPADC_RT_SW1_EOC_MASK		BIT(5)
> +#define TWL6030_GPADC_SW2_EOC_MASK		BIT(6)
> +
> +#define TWL6032_GPADC_RT_EOC_MASK		BIT(4)
> +#define TWL6032_GPADC_SW_EOC_MASK		BIT(5)
> +
> +#define TWL6030_GPADC_TRIM1			0xCD
> +
> +#define TWL6030_REG_TOGGLE1			0x90
> +#define TWL6030_GPADCS				BIT(1)
> +#define TWL6030_GPADCR				BIT(0)
> +
> +/**
> + * struct twl6030_chnl_calib - channel calibration
> + * @gain:		slope coefficient for ideal curve
> + * @gain_error:		gain error
> + * @offset_error:	offset of the real curve
> + */
> +struct twl6030_chnl_calib {
> +	s32 gain;
> +	s32 gain_error;
> +	s32 offset_error;
> +};
> +
> +/**
> + * struct twl6030_ideal_code - GPADC calibration parameters
> + * GPADC is calibrated in two points: close to the beginning and
> + * to the and of the measurable input range
> + *
> + * @code1:	ideal code for the input at the beginning
> + * @code2:	ideal code for at the end of the range
> + * @v1:		voltage input at the beginning(low voltage)
> + * @v2:		voltage input at the end(high voltage)
> + */
> +struct twl6030_ideal_code {
> +	u16 code1;
> +	u16 code2;
> +	u16 v1;
> +	u16 v2;
> +};
> +
> +struct twl6030_gpadc_data;
> +
> +/**
> + * struct twl6030_gpadc_platform_data - platform specific data
> + * @nchannels:		number of GPADC channels
> + * @iio_channels:	iio channels
> + * @twl6030_ideal:	pointer to calibration parameters
> + * @start_conversion:	pointer to ADC start conversion function
> + * @channel_to_reg	pointer to ADC function to convert channel to
> + * 			register address for reading conversion result
> + * @calibrate:		pointer to calibration function
> + */
> +struct twl6030_gpadc_platform_data {
> +	const int nchannels;
> +	const struct iio_chan_spec *iio_channels;
> +	const struct twl6030_ideal_code *ideal;
> +	int (*start_conversion)(int channel);
> +	u8 (*channel_to_reg)(int channel);
> +	int (*calibrate)(struct twl6030_gpadc_data *gpadc);
> +};
> +
> +/**
> + * struct twl6030_gpadc_data - GPADC data
> + * @dev:		device pointer
> + * @lock:		mutual exclusion lock for the structure
> + * @irq_complete:	completion to signal end of conversion
> + * @twl6030_cal_tbl:	pointer to calibration data for each
> + * 			channel with gain error and offset
> + * @pdata:		pointer to device specific data
> + */
> +struct twl6030_gpadc_data {
> +	struct device	*dev;
> +	struct mutex	lock;
> +	struct completion	irq_complete;
> +	struct twl6030_chnl_calib	*twl6030_cal_tbl;
> +	const struct twl6030_gpadc_platform_data *pdata;
> +};
> +
> +/*
> + * channels 11, 12, 13, 15 and 16 have no calibration data
> + * calibration offset is same for channels 1, 3, 4, 5
> + *
> + * The data is taken from GPADC_TRIM registers description.
> + * GPADC_TRIM registers keeps difference between the code measured
> + * at v1 and v2 input voltages and corresponding code1 and code2
> + */
> +static const struct twl6030_ideal_code
> +	twl6030_ideal[TWL6030_GPADC_MAX_CHANNELS] = {
> +	{ /* ch 0, external, battery type, resistor value */
> +		.code1 = 116,
> +		.code2 = 745,
> +		.v1 = 141,
> +		.v2 = 910,
> +	},
> +	{ /* ch 1, external, battery temperature, NTC resistor value */
> +		.code1 = 82,
> +		.code2 = 900,
> +		.v1 = 100,
> +		.v2 = 1100,
> +	},
> +	{ /* ch 2, external, audio accessory/general purpose */
> +		.code1 = 55,
> +		.code2 = 818,
> +		.v1 = 101,
> +		.v2 = 1499,
> +	},
> +	{ /* ch 3, external, general purpose */
> +		.code1 = 82,
> +		.code2 = 900,
> +		.v1 = 100,
> +		.v2 = 1100,
> +	},
> +	{ /* ch 4, external, temperature measurement/general purpose */
> +		.code1 = 82,
> +		.code2 = 900,
> +		.v1 = 100,
> +		.v2 = 1100,
> +	},
> +	{ /* ch 5, external, general purpose */
> +		.code1 = 82,
> +		.code2 = 900,
> +		.v1 = 100,
> +		.v2 = 1100,
> +	},
> +	{ /* ch 6, external, general purpose */
> +		.code1 = 82,
> +		.code2 = 900,
> +		.v1 = 100,
> +		.v2 = 1100,
> +	},
> +	{ /* ch 7, internal, main battery */
> +		.code1 = 614,
> +		.code2 = 941,
> +		.v1 = 3001,
> +		.v2 = 4599,
> +	},
> +	{ /* ch 8, internal, backup battery */
> +		.code1 = 82,
> +		.code2 = 688,
> +		.v1 = 501,
> +		.v2 = 4203,
> +	},
> +	{ /* ch 9, internal, external charger input */
> +		.code1 = 182,
> +		.code2 = 818,
> +		.v1 = 2001,
> +		.v2 = 8996,
> +	},
> +	{ /* ch 10, internal, VBUS */
> +		.code1 = 149,
> +		.code2 = 818,
> +		.v1 = 1001,
> +		.v2 = 5497,
> +	},
> +	{},	/* ch 11, internal, VBUS charging current */
> +	{},	/* ch 12, internal, Die temperature */
> +	{},	/* ch 13, internal, Die temperature */
> +	{ /* ch 14, internal, USB ID line */
> +		.code1 = 48,
> +		.code2 = 714,
> +		.v1 = 323,
> +		.v2 = 4800,
> +	},
> +	{},	/* ch 15, internal, test network */
> +	{},	/* ch 16, internal, test network */
> +};
> +
> +static const struct twl6030_ideal_code
> +			twl6032_ideal[TWL6032_GPADC_MAX_CHANNELS] = {

[0] = {
will make this clearer to read.
> +	{ /* ch 0, external, battery type, resistor value */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 1, external, battery temperature, NTC resistor value */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 2, external, audio accessory/general purpose */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 660,
> +		.v2 = 1500,
> +	},
> +	{ /* ch 3, external, temperature with external diode/general purpose */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 4, external, temperature measurement/general purpose */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 5, external, general purpose */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 6, external, general purpose */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch7, internal, system supply */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 2200,
> +		.v2 = 5000,
> +	},
> +	{ /* ch8, internal, backup battery */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 2200,
> +		.v2 = 5000,
> +	},
> +	{ /* ch 9, internal, external charger input */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 3960,
> +		.v2 = 9000,
> +	},
> +	{ /* ch10, internal, VBUS */
> +		.code1 = 150,
> +		.code2 = 751,
> +		.v1 = 1000,
> +		.v2 = 5000,
> +	},
> +	{ /* ch 11, internal, VBUS DC-DC output current */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 660,
> +		.v2 = 1500,
> +	},
> +	{ /* ch 12, internal, Die temperature */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 13, internal, Die temperature */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 14, internal, USB ID line */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 2420,
> +		.v2 = 5500,
> +	},
> +	{},	/* ch 15, internal, test network */
> +	{},	/* ch 16, internal, test network */
> +	{},	/* ch 17, internal, battery charging current */
> +	{ /* ch 18, internal, battery voltage */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 2200,
> +		.v2 = 5000,
> +	},
> +};
> +
> +static inline int twl6030_gpadc_write(u8 reg, u8 val)
> +{
> +	return twl_i2c_write_u8(TWL6030_MODULE_GPADC, val, reg);
> +}
> +
> +static inline int twl6030_gpadc_read(u8 *val, u8 reg)
> +{
> +
> +	return twl_i2c_read(TWL6030_MODULE_GPADC, val, reg, 2);
> +}
> +
> +static int twl6030_gpadc_enable_irq(u8 mask)
> +{
> +	int ret;
> +
> +	ret = twl6030_interrupt_unmask(mask, REG_INT_MSK_LINE_B);
> +	ret |= twl6030_interrupt_unmask(mask, REG_INT_MSK_STS_B);
I'd just roll these into 1 if it makes sense not to check the value
of ret in between.
> +
> +	return ret;
> +}
> +
> +static void twl6030_gpadc_disable_irq(u8 mask)
> +{
> +	twl6030_interrupt_mask(mask, REG_INT_MSK_LINE_B);
> +	twl6030_interrupt_mask(mask, REG_INT_MSK_STS_B);
> +}
> +
> +static irqreturn_t twl6030_gpadc_irq_handler(int irq, void *_gpadc)
> +{
> +	struct twl6030_gpadc_data *gpadc = _gpadc;
> +
> +	complete(&gpadc->irq_complete);
> +
> +	return IRQ_HANDLED;
> +}
> +
> +static int twl6030_start_conversion(int channel)
> +{
> +	return twl6030_gpadc_write(TWL6030_GPADC_CTRL_P1,
> +					TWL6030_GPADC_CTRL_P1_SP1);
> +}
> +
> +static int twl6032_start_conversion(int channel)
> +{
> +	int ret;
> +
> +	ret = twl6030_gpadc_write(TWL6032_GPADC_GPSELECT_ISB, channel);
> +	if (ret)
> +		return ret;
> +
> +	return twl6030_gpadc_write(TWL6032_GPADC_CTRL_P1,
> +						TWL6030_GPADC_CTRL_P1_SP1);
> +}
> +
> +static u8 twl6030_channel_to_reg(int channel)
> +{
> +	return TWL6030_GPADC_GPCH0_LSB + 2 * channel;
> +}
> +
> +static u8 twl6032_channel_to_reg(int channel)
> +{
> +	return TWL6032_GPADC_GPCH0_LSB;
> +}
> +
> +static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,
> +		int channel, int *res)
> +{
> +	u8 reg = gpadc->pdata->channel_to_reg(channel);
> +	u8 val[2];

le16 val;
> +	int ret;
> +
ret = twl6030_gpadc_read(val, reg);
(note that (reg, val) ordering of parameters would be a more common choice)

> +	ret = twl6030_gpadc_read(val, reg);
> +	if (ret) {
> +		dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);
> +		return ret;
> +	}
> +
res = le16_to_cpup(val);

> +	*res = (val[1] << 8) | val[0];
> +
> +	return ret;
> +}
> +
> +static int twl6030_gpadc_get_processed(struct twl6030_gpadc_data *gpadc,
> +		int channel, int *val)
> +{
> +	int raw_code;
> +	int corrected_code;
> +	int channel_value;
> +	int ret;
> +
> +	ret = twl6030_gpadc_get_raw(gpadc, channel, &raw_code);
> +	if (ret)
> +		return ret;
> +

Might first thought on seeing this is that it would be better to return
raw for all channels and provide the scale and offset info_mask elements
where possible rather than doing the conversion in the kernel.  Note we
allow really quite a bit of precision on those values so I doubt it
will be a problem.

If nothing else it would make everything rather more consistent.


> +	corrected_code = ((raw_code * 1000) -
> +		gpadc->twl6030_cal_tbl[channel].offset_error) /
> +		gpadc->twl6030_cal_tbl[channel].gain_error;
> +
> +	channel_value = corrected_code *
> +			gpadc->twl6030_cal_tbl[channel].gain;
> +
> +	/* Shift back into mV range */
> +	channel_value /= 1000;
> +
> +	dev_dbg(gpadc->dev, "GPADC raw: %d", raw_code);
> +	dev_dbg(gpadc->dev, "GPADC cor: %d", corrected_code);
> +	dev_dbg(gpadc->dev, "GPADC val: %d", channel_value);
> +
> +	*val = channel_value;
> +
> +	return ret;
> +}
> +
> +static int twl6030_gpadc_read_raw(struct iio_dev *indio_dev,
> +			     const struct iio_chan_spec *chan,
> +			     int *val, int *val2, long mask)
> +{
> +	struct twl6030_gpadc_data *gpadc = iio_priv(indio_dev);
> +	int ret = -EINVAL;
> +
> +	mutex_lock(&gpadc->lock);
> +
> +	ret = gpadc->pdata->start_conversion(chan->channel);
> +	if (ret) {
> +		dev_err(gpadc->dev, "failed to start conversion\n");
> +		goto err;
> +	}
> +	/* wait for conversion to complete */
> +	wait_for_completion_interruptible_timeout(&gpadc->irq_complete,
> +						msecs_to_jiffies(5000));
> +
> +	switch (mask) {
> +	case IIO_CHAN_INFO_RAW:
> +		ret = twl6030_gpadc_get_raw(gpadc, chan->channel, val);
> +		ret = ret ? -EIO : IIO_VAL_INT;
> +		break;
> +
> +	case IIO_CHAN_INFO_PROCESSED:
> +		ret = twl6030_gpadc_get_processed(gpadc, chan->channel, val);
> +		ret = ret ? -EIO : IIO_VAL_INT;
> +		break;
> +
> +	default:
> +		break;
> +	}
> +err:
> +	mutex_unlock(&gpadc->lock);
> +
> +	return ret;
> +}
> +
> +static void twl6030_calibrate_channel(struct twl6030_gpadc_data *gpadc,
> +					int channel, int d1, int d2)
> +{
> +	int b, k, gain, x1, x2;
> +	const struct twl6030_ideal_code *ideal = gpadc->pdata->ideal;
> +
> +	/* Gain */
> +	gain = ((ideal[channel].v2 - ideal[channel].v1) * 1000) /
> +		(ideal[channel].code2 - ideal[channel].code1);
> +
> +	x1 = ideal[channel].code1;
> +	x2 = ideal[channel].code2;
> +
> +	/* k */
> +	k = 1000 + (((d2 - d1) * 1000) / (x2 - x1));
> +
> +	/* b */
> +	b = (d1 * 1000) - (k - 1000) * x1;
> +
> +	gpadc->twl6030_cal_tbl[channel].gain = gain;
> +	gpadc->twl6030_cal_tbl[channel].gain_error = k;
> +	gpadc->twl6030_cal_tbl[channel].offset_error = b;
> +
> +	dev_dbg(gpadc->dev, "GPADC d1   for Chn: %d = %d\n", channel, d1);
> +	dev_dbg(gpadc->dev, "GPADC d2   for Chn: %d = %d\n", channel, d2);
> +	dev_dbg(gpadc->dev, "GPADC x1   for Chn: %d = %d\n", channel, x1);
> +	dev_dbg(gpadc->dev, "GPADC x2   for Chn: %d = %d\n", channel, x2);
> +	dev_dbg(gpadc->dev, "GPADC Gain for Chn: %d = %d\n", channel, gain);
> +	dev_dbg(gpadc->dev, "GPADC k    for Chn: %d = %d\n", channel, k);
> +	dev_dbg(gpadc->dev, "GPADC b    for Chn: %d = %d\n", channel, b);
> +}
> +
> +static inline int twl6030_gpadc_get_trim_offset(s8 d)
> +{
> +	int sign;
> +
> +	/*
> +	 * XXX NOTE!
> +	 * bit 0 - sign, bit 7 - reserved, 6..1 - trim value
> +	 * though, the documentation states that trim value
> +	 * is absolute value, the correct conversion results are
> +	 * obtained if the value is interpreted as 2's complement.
> +	 */
I'd be tempted to reorder this and then use the sign extend standard
functions.
	temp = ((d & 0x7f) >> 1) | ((d & 1) << 7);
	return sign_extend32(temp, 7);

(or something like that)

> +	sign = d & 1;
> +	d = (d & 0x7f) >> 1;
> +
> +	return sign ? (d | 0xc0) : d;
> +}
> +
> +static int twl6030_calibration(struct twl6030_gpadc_data *gpadc)
> +{
> +	int ret;
> +	int chn;
> +	u8 trim_regs[TWL6030_GPADC_MAX_CHANNELS];
> +	s8 d1, d2;
> +
> +	/*
> +	 * for calibration two measurements have been performed at
> +	 * factory, for some channels, during the production test and
> +	 * have been stored in registers. This two stored values are
> +	 * used to correct the measurements. The values represent
> +	 * offsets for the given input from the output on ideal curve.
> +	 */
> +	ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs,
> +			TWL6030_GPADC_TRIM1, 16);
> +	if (ret < 0) {
> +		dev_err(gpadc->dev, "calibration failed\n");
> +		return ret;
> +	}
> +
> +	for (chn = 0; chn < TWL6030_GPADC_MAX_CHANNELS; chn++) {
> +
> +		switch (chn) {
> +		case 0:
> +			d1 = trim_regs[0];
> +			d2 = trim_regs[1];
> +			break;
> +		case 1:
> +			d1 = trim_regs[4];
> +			d2 = trim_regs[5];
> +			break;
> +		case 2:
> +			d1 = trim_regs[12];
> +			d2 = trim_regs[13];
> +			break;
Given you give up entirely on ordering in the next calibration function
I'd be tempted to do the same here and have all the d1 = ...[4] etc in one
case statement.
> +		case 3:
> +		case 4:
> +		case 5:
> +		case 6:
> +			d1 = trim_regs[4];
> +			d2 = trim_regs[5];
> +			break;
> +		case 7:
> +			d1 = trim_regs[6];
> +			d2 = trim_regs[7];
> +			break;
> +		case 8:
> +			d1 = trim_regs[2];
> +			d2 = trim_regs[3];
> +			break;
> +		case 9:
> +			d1 = trim_regs[8];
> +			d2 = trim_regs[9];
> +			break;
> +		case 10:
> +			d1 = trim_regs[10];
> +			d2 = trim_regs[11];
> +			break;
> +		case 14:
> +			d1 = trim_regs[14];
> +			d2 = trim_regs[15];
> +			break;
> +		default:
> +			continue;
> +		}
> +
> +		d1 = twl6030_gpadc_get_trim_offset(d1);
> +		d2 = twl6030_gpadc_get_trim_offset(d2);
> +
> +		twl6030_calibrate_channel(gpadc, chn, d1, d2);
> +	}
> +
> +	return 0;
> +}
> +
> +static int twl6032_calibration(struct twl6030_gpadc_data *gpadc)
> +{
> +	int chn, d1 = 0, d2 = 0, temp;
> +	u8 trim_regs[17];
> +	int ret;
> +
> +	ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs + 1,
> +			TWL6030_GPADC_TRIM1, 16);
> +	if (ret < 0) {
> +		dev_err(gpadc->dev, "calibration failed\n");
> +		return ret;
> +	}
> +
/*
 * Loop
please for kernel code.
> +	/* Loop to calculate the value needed for returning voltages from
> +	 * GPADC not values.
> +	 *
> +	 * gain is calculated to 3 decimal places fixed point.
> +	 */
> +	for (chn = 0; chn < TWL6032_GPADC_MAX_CHANNELS; chn++) {
> +
> +		switch (chn) {
> +		case 0:
> +		case 1:
> +		case 2:
> +		case 3:
> +		case 4:
> +		case 5:
> +		case 6:
> +		case 11:
> +		case 12:
> +		case 13:
> +		case 14:
> +			/* D1 */
> +			d1 = (trim_regs[3] & 0x1F) << 2;
> +			d1 |= (trim_regs[1] & 0x06) >> 1;
> +			if (trim_regs[1] & 0x01)
> +				d1 = -d1;
> +
> +			/* D2 */
> +			d2 = (trim_regs[4] & 0x3F) << 2;
> +			d2 |= (trim_regs[2] & 0x06) >> 1;
> +			if (trim_regs[2] & 0x01)
> +				d2 = -d2;
> +			break;
> +		case 8:
No coment on your code, but this is an 'interesting' bit
of bit packing...
I did vaguely wonder if this could be mapped to a big lookup table,
but having tried it I think I end up with something almost as tricky
to read.. Oh well that was a fun 10 minute diversion ;)

> +			/* D1 */
> +			temp = (trim_regs[3] & 0x1F) << 2;
> +			temp |= (trim_regs[1] & 0x06) >> 1;
> +			if (trim_regs[1] & 0x01)
> +				temp = -temp;
> +
> +			d1 = (trim_regs[8] & 0x18) << 1;
> +			d1 |= (trim_regs[7] & 0x1E) >> 1;
> +			if (trim_regs[7] & 0x01)
> +				d1 = -d1;
> +
> +			d1 += temp;
> +
> +			/* D2 */
> +			temp = (trim_regs[4] & 0x3F) << 2;
> +			temp |= (trim_regs[2] & 0x06) >> 1;
> +			if (trim_regs[2] & 0x01)
> +				temp = -temp;
> +
> +			d2 = (trim_regs[10] & 0x1F) << 2;
> +			d2 |= (trim_regs[8] & 0x06) >> 1;
> +			if (trim_regs[8] & 0x01)
> +				d2 = -d2;
> +
> +			d2 += temp;
> +			break;
> +		case 9:
> +			/* D1 */
> +			temp = (trim_regs[3] & 0x1F) << 2;
> +			temp |= (trim_regs[1] & 0x06) >> 1;
> +			if (trim_regs[1] & 0x01)
> +				temp = -temp;
> +
> +			d1 = (trim_regs[14] & 0x18) << 1;
> +			d1 |= (trim_regs[12] & 0x1E) >> 1;
> +			if (trim_regs[12] & 0x01)
> +				d1 = -d1;
> +
> +			d1 += temp;
> +
> +			/* D2 */
> +			temp = (trim_regs[4] & 0x3F) << 2;
> +			temp |= (trim_regs[2] & 0x06) >> 1;
> +			if (trim_regs[2] & 0x01)
> +				temp = -temp;
> +
> +			d2 = (trim_regs[16] & 0x1F) << 2;
> +			d2 |= (trim_regs[14] & 0x06) >> 1;
> +			if (trim_regs[14] & 0x01)
> +				d2 = -d2;
> +
> +			d2 += temp;
> +		case 10:
> +			/* D1 */
> +			d1 = (trim_regs[11] & 0x0F) << 3;
> +			d1 |= (trim_regs[9] & 0x0E) >> 1;
> +			if (trim_regs[9] & 0x01)
> +				d1 = -d1;
> +
> +			/* D2 */
> +			d2 = (trim_regs[15] & 0x0F) << 3;
> +			d2 |= (trim_regs[13] & 0x0E) >> 1;
> +			if (trim_regs[13] & 0x01)
> +				d2 = -d2;
> +			break;
> +		case 7:
> +		case 18:
> +			/* D1 */
> +			temp = (trim_regs[3] & 0x1F) << 2;
> +			temp |= (trim_regs[1] & 0x06) >> 1;
> +			if (trim_regs[1] & 0x01)
> +				temp = -temp;
> +
> +			d1 = (trim_regs[5] & 0x7E) >> 1;
> +			if (trim_regs[5] & 0x01)
> +				d1 = -d1;
> +
> +			d1 += temp;
> +
> +			/* D2 */
> +			temp = (trim_regs[4] & 0x3F) << 2;
> +			temp |= (trim_regs[2] & 0x06) >> 1;
> +			if (trim_regs[2] & 0x01)
> +				temp = -temp;
> +
> +			d2 = (trim_regs[6] & 0xFE) >> 1;
> +			if (trim_regs[6] & 0x01)
> +				d2 = -d2;
> +
> +			d2 += temp;
> +			break;
> +		default:
> +			/* No data for other channels */
> +			continue;
> +		}
> +
> +		twl6030_calibrate_channel(gpadc, chn, d1, d2);
> +	}
> +
> +	return 0;
> +}
> +
> +#define TWL6030_GPADC_CHAN(chn, chan_info) {		\
> +	.type = IIO_VOLTAGE,				\
A lot of the channels appear not to be measuring voltages.
So they should all be appropriately specified.

> +	.channel = chn,					\

Unless you are doing buffered output this stuff is of no interest to
anyone so drop .scan_type from here.

> +	.scan_type = {					\
> +		.sign = 'u',				\
> +		.storagebits = 16,			\
> +	},						\
> +	.info_mask = BIT(chan_info),			\
> +	.info_mask_separate = BIT(chan_info), 		\
> +	.indexed = 1,					\
> +}
> +
> +static const struct iio_chan_spec twl6030_gpadc_iio_channels[] = {
> +	TWL6030_GPADC_CHAN(0, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(1, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(2, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(3, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(4, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(5, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(6, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(7, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(8, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(9, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(10, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(11, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(12, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(13, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(14, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(15, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(16, IIO_CHAN_INFO_RAW),
> +};
> +
> +static const struct iio_chan_spec twl6032_gpadc_iio_channels[] = {
> +	TWL6030_GPADC_CHAN(0, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(1, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(2, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(3, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(4, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(5, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(6, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(7, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(8, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(9, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(10, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(11, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(12, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(13, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(14, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(15, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(16, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(17, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(18, IIO_CHAN_INFO_PROCESSED),
> +};
> +
> +static const struct iio_info twl6030_gpadc_iio_info = {
> +	.read_raw = &twl6030_gpadc_read_raw,
> +	.driver_module = THIS_MODULE,
> +};
> +
> +static const struct twl6030_gpadc_platform_data twl6030_pdata = {
> +	.iio_channels = twl6030_gpadc_iio_channels,
> +	.nchannels = TWL6030_GPADC_MAX_CHANNELS,
> +	.ideal = twl6030_ideal,
> +	.start_conversion = twl6030_start_conversion,
> +	.channel_to_reg = twl6030_channel_to_reg,
> +	.calibrate = twl6030_calibration,
> +};
> +
> +static const struct twl6030_gpadc_platform_data twl6032_pdata = {
> +	.iio_channels = twl6032_gpadc_iio_channels,
> +	.nchannels = TWL6032_GPADC_MAX_CHANNELS,
> +	.ideal = twl6032_ideal,
> +	.start_conversion = twl6032_start_conversion,
> +	.channel_to_reg = twl6032_channel_to_reg,
> +	.calibrate = twl6032_calibration,
> +};
> +
> +static const struct of_device_id of_twl6030_match_tbl[] = {
> +	{
> +		.compatible = "ti,twl6030_gpadc",
> +		.data = &twl6030_pdata,
> +	},
> +	{
> +		.compatible = "ti,twl6032_gpadc",
> +		.data = &twl6032_pdata,
> +	},
> +	{ /* end */ }
> +};
> +
> +static int twl6030_gpadc_probe(struct platform_device *pdev)
> +{
> +	struct device *dev = &pdev->dev;
> +	struct twl6030_gpadc_data *gpadc;
> +	const struct twl6030_gpadc_platform_data *pdata;
> +	const struct of_device_id *match;
> +	struct iio_dev *indio_dev;
> +	int irq;
> +	int ret = 0;
> +
> +	match = of_match_device(of_match_ptr(of_twl6030_match_tbl), dev);
> +	pdata = match ? match->data : dev->platform_data;
> +
> +	if (!pdata)
> +		return -EINVAL;
> +
> +	indio_dev = iio_device_alloc(sizeof(struct twl6030_gpadc_data));
> +	if (!indio_dev) {
> +		dev_err(dev, "failed allocating iio device\n");
> +		ret = -ENOMEM;
> +	}
> +
> +	gpadc = iio_priv(indio_dev);
> +
> +	gpadc->twl6030_cal_tbl = devm_kzalloc(dev,
> +					sizeof(struct twl6030_chnl_calib) *
> +					pdata->nchannels, GFP_KERNEL);
> +	if (!gpadc->twl6030_cal_tbl)
> +		goto err;
> +
> +	gpadc->dev = dev;
> +	gpadc->pdata = pdata;
> +
> +	platform_set_drvdata(pdev, indio_dev);
> +	mutex_init(&gpadc->lock);
> +	init_completion(&gpadc->irq_complete);
> +
> +	ret = pdata->calibrate(gpadc);
> +	if (ret < 0) {
> +		dev_err(&pdev->dev, "failed to read calibration registers\n");
> +		goto err;
> +	}
> +
> +	irq = platform_get_irq(pdev, 0);
> +	if (irq < 0) {
> +		dev_err(&pdev->dev, "failed to get irq\n");
> +		goto err;
> +	}
> +
> +	ret = devm_request_threaded_irq(dev, irq, NULL,
> +					twl6030_gpadc_irq_handler,
> +					IRQF_ONESHOT, "twl6030_gpadc", gpadc);
> +	if (ret) {
> +		dev_dbg(&pdev->dev, "could not request irq\n");
> +		goto err;
> +	}
> +
> +	ret = twl6030_gpadc_enable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
> +	if (ret < 0) {
> +		dev_err(&pdev->dev, "failed to enable GPADC interrupt\n");
> +		goto err;
> +	}
> +	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
> +					TWL6030_REG_TOGGLE1);
> +	if (ret < 0) {
> +		dev_err(&pdev->dev, "failed to enable GPADC module\n");
> +		goto err;
> +	}
> +
> +	indio_dev->name = DRIVER_NAME;
> +	indio_dev->dev.parent = dev;
> +	indio_dev->info = &twl6030_gpadc_iio_info;
> +	indio_dev->modes = INDIO_DIRECT_MODE;
> +	indio_dev->channels = pdata->iio_channels;
> +	indio_dev->num_channels = pdata->nchannels;
> +
> +	ret = iio_device_register(indio_dev);
> +	if (ret)
> +		goto err;
> +
> +	return ret;
> +err:
> +	iio_device_free(indio_dev);
> +	return ret;
> +}
> +
> +static int twl6030_gpadc_remove(struct platform_device *pdev)
> +{
> +	struct iio_dev *indio_dev = platform_get_drvdata(pdev);
> +
> +	twl6030_gpadc_disable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
> +	iio_device_unregister(indio_dev);
> +	iio_device_free(indio_dev);
> +
> +	return 0;
> +}
> +
> +static int twl6030_gpadc_suspend(struct device *pdev)
> +{
> +	int ret;
> +
> +	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCR,
> +				TWL6030_REG_TOGGLE1);
> +	if (ret)
> +		dev_err(pdev, "error reseting GPADC (%d)!\n", ret);
> +
> +	return 0;
> +};
> +
> +static int twl6030_gpadc_resume(struct device *pdev)
> +{
> +	int ret;
> +
> +	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
> +				TWL6030_REG_TOGGLE1);
> +	if (ret)
> +		dev_err(pdev, "error setting GPADC (%d)!\n", ret);
> +
> +	return 0;
> +};
> +
> +static const struct dev_pm_ops twl6030_gpadc_pm_ops = {
> +	.suspend = twl6030_gpadc_suspend,
> +	.resume = twl6030_gpadc_resume,
> +};
> +
> +static struct platform_driver twl6030_gpadc_driver = {
> +	.probe		= twl6030_gpadc_probe,
> +	.remove		= twl6030_gpadc_remove,
> +	.driver		= {
> +		.name	= DRIVER_NAME,
> +		.owner	= THIS_MODULE,
> +		.pm	= &twl6030_gpadc_pm_ops,
> +		.of_match_table = of_twl6030_match_tbl,
> +	},
> +};
> +
> +module_platform_driver(twl6030_gpadc_driver);
> +
> +MODULE_ALIAS("platform: " DRIVER_NAME);
> +MODULE_AUTHOR("Texas Instruments Inc.");
> +MODULE_DESCRIPTION("twl6030 ADC driver");
> +MODULE_LICENSE("GPL");
>

^ permalink raw reply	[flat|nested] 45+ messages in thread

* [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-12 19:42     ` Jonathan Cameron
  0 siblings, 0 replies; 45+ messages in thread
From: Jonathan Cameron @ 2013-07-12 19:42 UTC (permalink / raw)
  To: linux-arm-kernel

On 07/12/2013 08:18 AM, Oleksandr Kozaruk wrote:
> The GPADC is general purpose ADC found on TWL6030,
> and TWL6032 PMIC, known also as Phoenix and PhoenixLite.
>
> The TWL6030 and TWL6032 have GPADC with 17 and 19
> channels respectively. Some channels have current
> source and are used for measuring voltage drop
> on resistive load for detecting battery ID resistance,
> or measuring voltage drop on NTC resistors for external
> temperature measurements. Some channels measure voltage,
> (i.e. battery voltage), and have voltage dividers,
> thus, capable to scale voltage. Some channels are dedicated
> for measuring die temperature.
>
> Some channels are calibrated in 2 points, having
> offsets from ideal values kept in trim registers. This
> is used to correct measurements.
>
> The differences between GPADC in TWL6030 and TWL6032:
> - 10 bit vs 12 bit ADC;
> - 17 vs 19 channels;
> - channels have different purpose(i. e. battery voltage
>   channel 8 vs channel 18);
> - trim values are interpreted differently.
>
> The driver exports function returning converted value for
> requested channels.
>
> Based on the driver patched from Balaji TK, Graeme Gregory, Ambresh K,
> Girish S Ghongdemath.

Couple of things:

1) It looks from the driver that a lot of the channels are not measuring
   voltages but rather temperature or currents etc.  If so then their
   types in the channel mask should be appropriately set.  Also if some
   of the channels are entirely internal test networks, what is the benefit
   of exposing them to userspace as readable channels?
   If channels are merely 'suggested' as being used for temperatures etc
   then it is fine to keep them as voltages.

2) You have my sympathy when it comes to the way those trim values are packed
   into the registers.  That is truely hideous ;)

Various trivial little bits but all in all a nice clean driver. Thanks,

Jonathan
>
> Signed-off-by: Balaji T K <balajitk@ti.com>
> Graeme Gregory <gg@slimlogic.co.uk>
> Signed-off-by: Oleksandr Kozaruk <oleksandr.kozaruk@ti.com>
> ---
>  drivers/iio/adc/Kconfig         |    8 +
>  drivers/iio/adc/Makefile        |    1 +
>  drivers/iio/adc/twl6030-gpadc.c | 1019 +++++++++++++++++++++++++++++++++++++++
>  3 files changed, 1028 insertions(+)
>  create mode 100644 drivers/iio/adc/twl6030-gpadc.c
>
> diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig
> index ab0767e6..87d699e 100644
> --- a/drivers/iio/adc/Kconfig
> +++ b/drivers/iio/adc/Kconfig
> @@ -157,4 +157,12 @@ config VIPERBOARD_ADC
>  	  Say yes here to access the ADC part of the Nano River
>  	  Technologies Viperboard.
>
> +config TWL6030_GPADC
> +	tristate "TWL6030 GPADC (General Purpose A/D Convertor) Support"
> +	depends on TWL4030_CORE
> +	default n
> +	help
> +	  Say yes here if you want support for the TWL6030 General Purpose
> +	  A/D Convertor.
Perhaps a little more detail for such a complex device?
> +
>  endmenu
> diff --git a/drivers/iio/adc/Makefile b/drivers/iio/adc/Makefile
> index 0a825be..8b05633 100644
> --- a/drivers/iio/adc/Makefile
> +++ b/drivers/iio/adc/Makefile
> @@ -17,3 +17,4 @@ obj-$(CONFIG_MAX1363) += max1363.o
>  obj-$(CONFIG_TI_ADC081C) += ti-adc081c.o
>  obj-$(CONFIG_TI_AM335X_ADC) += ti_am335x_adc.o
>  obj-$(CONFIG_VIPERBOARD_ADC) += viperboard_adc.o
> +obj-$(CONFIG_TWL6030_GPADC) += twl6030-gpadc.o
> diff --git a/drivers/iio/adc/twl6030-gpadc.c b/drivers/iio/adc/twl6030-gpadc.c
> new file mode 100644
> index 0000000..6ceb789
> --- /dev/null
> +++ b/drivers/iio/adc/twl6030-gpadc.c
> @@ -0,0 +1,1019 @@
> +/*
> + * TWL6030 GPADC module driver
> + *
> + * Copyright (C) 2009-2013 Texas Instruments Inc.
> + * Nishant Kamat <nskamat@ti.com>
> + * Balaji T K <balajitk@ti.com>
> + * Graeme Gregory <gg@slimlogic.co.uk>
> + * Girish S Ghongdemath <girishsg@ti.com>
> + * Ambresh K <ambresh@ti.com>
> + * Oleksandr Kozaruk <oleksandr.kozaruk@ti.com
> + *
> + * Based on twl4030-madc.c
> + * Copyright (C) 2008 Nokia Corporation
> + * Mikko Ylinen <mikko.k.ylinen@nokia.com>
> + *
> + * This program is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU General Public License
> + * version 2 as published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful, but
> + * WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
> + * General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; if not, write to the Free Software
> + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA
> + * 02110-1301 USA
> + *
> + */
> +#include <linux/init.h>
> +#include <linux/interrupt.h>
> +#include <linux/kernel.h>
> +#include <linux/module.h>
> +#include <linux/platform_device.h>
> +#include <linux/of_platform.h>
> +#include <linux/i2c/twl.h>
> +#include <linux/iio/iio.h>
> +#include <linux/iio/sysfs.h>
> +
> +#define DRIVER_NAME		"twl6030_gpadc"
> +
> +#define TWL6030_GPADC_MAX_CHANNELS 17
> +#define TWL6032_GPADC_MAX_CHANNELS 19
> +/* Define this as the biggest of all chips using this driver */
> +#define GPADC_MAX_CHANNELS TWL6032_GPADC_MAX_CHANNELS
> +
> +#define TWL6030_GPADC_CTRL			0x00    /* 0x2e */
> +#define TWL6030_GPADC_CTRL2			0x01	/* 0x2f */
> +
> +#define TWL6030_GPADC_CTRL_P1			0x05
> +#define TWL6030_GPADC_CTRL_P2			0x06
> +
> +#define TWL6032_GPADC_GPSELECT_ISB		0x07
> +#define TWL6032_GPADC_CTRL_P1			0x08
> +
> +#define TWL6032_GPADC_RTCH0_LSB			0x09
> +#define TWL6032_GPADC_RTCH0_MSB			0x0a
> +#define TWL6032_GPADC_RTCH1_LSB			0x0b
> +#define TWL6032_GPADC_RTCH1_MSB			0x0c
> +#define TWL6032_GPADC_GPCH0_LSB			0x0d
> +#define TWL6032_GPADC_GPCH0_MSB			0x0e
> +
> +#define TWL6030_GPADC_CTRL_P1_SP1		BIT(3)
> +#define TWL6030_GPADC_CTRL_P1_EOCRT		BIT(2)
> +#define TWL6030_GPADC_CTRL_P1_EOCP1		BIT(1)
> +#define TWL6030_GPADC_CTRL_P1_BUSY		BIT(0)
> +
> +#define TWL6030_GPADC_CTRL_P2_SP2		BIT(2)
> +#define TWL6030_GPADC_CTRL_P2_EOCP2		BIT(1)
> +#define TWL6030_GPADC_CTRL_P1_BUSY		BIT(0)
> +
> +#define TWL6030_GPADC_EOC_SW			BIT(1)
> +#define TWL6030_GPADC_BUSY			BIT(0)
> +
> +#define TWL6030_GPADC_RTCH0_LSB			(0x07)
> +#define TWL6030_GPADC_GPCH0_LSB			(0x29)
> +
> +/* Fixed channels */
> +#define TWL6030_GPADC_CTRL_TEMP1_EN		BIT(0)    /* input ch 1 */
> +#define TWL6030_GPADC_CTRL_TEMP2_EN		BIT(1)    /* input ch 4 */
> +#define TWL6030_GPADC_CTRL_SCALER_EN		BIT(2)    /* input ch 2 */
> +#define TWL6030_GPADC_CTRL_SCALER_DIV4		BIT(3)
> +#define TWL6030_GPADC_CTRL_SCALER_EN_CH11	BIT(4)    /* input ch 11 */
> +#define TWL6030_GPADC_CTRL_TEMP1_EN_MONITOR	BIT(5)
> +#define TWL6030_GPADC_CTRL_TEMP2_EN_MONITOR	BIT(6)
> +#define TWL6030_GPADC_CTRL_ISOURCE_EN		BIT(7)
> +
> +#define TWL6030_GPADC_CTRL2_REMSENSE_0		BIT(0)
> +#define TWL6030_GPADC_CTRL2_REMSENSE_1		BIT(1)
> +#define TWL6030_GPADC_CTRL2_SCALER_EN_CH18	BIT(2)
> +#define TWL6030_GPADC_CTRL2_VBAT_SCALER_DIV4	BIT(3)
> +
> +#define TWL6030_GPADC_RT_SW1_EOC_MASK		BIT(5)
> +#define TWL6030_GPADC_SW2_EOC_MASK		BIT(6)
> +
> +#define TWL6032_GPADC_RT_EOC_MASK		BIT(4)
> +#define TWL6032_GPADC_SW_EOC_MASK		BIT(5)
> +
> +#define TWL6030_GPADC_TRIM1			0xCD
> +
> +#define TWL6030_REG_TOGGLE1			0x90
> +#define TWL6030_GPADCS				BIT(1)
> +#define TWL6030_GPADCR				BIT(0)
> +
> +/**
> + * struct twl6030_chnl_calib - channel calibration
> + * @gain:		slope coefficient for ideal curve
> + * @gain_error:		gain error
> + * @offset_error:	offset of the real curve
> + */
> +struct twl6030_chnl_calib {
> +	s32 gain;
> +	s32 gain_error;
> +	s32 offset_error;
> +};
> +
> +/**
> + * struct twl6030_ideal_code - GPADC calibration parameters
> + * GPADC is calibrated in two points: close to the beginning and
> + * to the and of the measurable input range
> + *
> + * @code1:	ideal code for the input at the beginning
> + * @code2:	ideal code for at the end of the range
> + * @v1:		voltage input at the beginning(low voltage)
> + * @v2:		voltage input at the end(high voltage)
> + */
> +struct twl6030_ideal_code {
> +	u16 code1;
> +	u16 code2;
> +	u16 v1;
> +	u16 v2;
> +};
> +
> +struct twl6030_gpadc_data;
> +
> +/**
> + * struct twl6030_gpadc_platform_data - platform specific data
> + * @nchannels:		number of GPADC channels
> + * @iio_channels:	iio channels
> + * @twl6030_ideal:	pointer to calibration parameters
> + * @start_conversion:	pointer to ADC start conversion function
> + * @channel_to_reg	pointer to ADC function to convert channel to
> + * 			register address for reading conversion result
> + * @calibrate:		pointer to calibration function
> + */
> +struct twl6030_gpadc_platform_data {
> +	const int nchannels;
> +	const struct iio_chan_spec *iio_channels;
> +	const struct twl6030_ideal_code *ideal;
> +	int (*start_conversion)(int channel);
> +	u8 (*channel_to_reg)(int channel);
> +	int (*calibrate)(struct twl6030_gpadc_data *gpadc);
> +};
> +
> +/**
> + * struct twl6030_gpadc_data - GPADC data
> + * @dev:		device pointer
> + * @lock:		mutual exclusion lock for the structure
> + * @irq_complete:	completion to signal end of conversion
> + * @twl6030_cal_tbl:	pointer to calibration data for each
> + * 			channel with gain error and offset
> + * @pdata:		pointer to device specific data
> + */
> +struct twl6030_gpadc_data {
> +	struct device	*dev;
> +	struct mutex	lock;
> +	struct completion	irq_complete;
> +	struct twl6030_chnl_calib	*twl6030_cal_tbl;
> +	const struct twl6030_gpadc_platform_data *pdata;
> +};
> +
> +/*
> + * channels 11, 12, 13, 15 and 16 have no calibration data
> + * calibration offset is same for channels 1, 3, 4, 5
> + *
> + * The data is taken from GPADC_TRIM registers description.
> + * GPADC_TRIM registers keeps difference between the code measured
> + * at v1 and v2 input voltages and corresponding code1 and code2
> + */
> +static const struct twl6030_ideal_code
> +	twl6030_ideal[TWL6030_GPADC_MAX_CHANNELS] = {
> +	{ /* ch 0, external, battery type, resistor value */
> +		.code1 = 116,
> +		.code2 = 745,
> +		.v1 = 141,
> +		.v2 = 910,
> +	},
> +	{ /* ch 1, external, battery temperature, NTC resistor value */
> +		.code1 = 82,
> +		.code2 = 900,
> +		.v1 = 100,
> +		.v2 = 1100,
> +	},
> +	{ /* ch 2, external, audio accessory/general purpose */
> +		.code1 = 55,
> +		.code2 = 818,
> +		.v1 = 101,
> +		.v2 = 1499,
> +	},
> +	{ /* ch 3, external, general purpose */
> +		.code1 = 82,
> +		.code2 = 900,
> +		.v1 = 100,
> +		.v2 = 1100,
> +	},
> +	{ /* ch 4, external, temperature measurement/general purpose */
> +		.code1 = 82,
> +		.code2 = 900,
> +		.v1 = 100,
> +		.v2 = 1100,
> +	},
> +	{ /* ch 5, external, general purpose */
> +		.code1 = 82,
> +		.code2 = 900,
> +		.v1 = 100,
> +		.v2 = 1100,
> +	},
> +	{ /* ch 6, external, general purpose */
> +		.code1 = 82,
> +		.code2 = 900,
> +		.v1 = 100,
> +		.v2 = 1100,
> +	},
> +	{ /* ch 7, internal, main battery */
> +		.code1 = 614,
> +		.code2 = 941,
> +		.v1 = 3001,
> +		.v2 = 4599,
> +	},
> +	{ /* ch 8, internal, backup battery */
> +		.code1 = 82,
> +		.code2 = 688,
> +		.v1 = 501,
> +		.v2 = 4203,
> +	},
> +	{ /* ch 9, internal, external charger input */
> +		.code1 = 182,
> +		.code2 = 818,
> +		.v1 = 2001,
> +		.v2 = 8996,
> +	},
> +	{ /* ch 10, internal, VBUS */
> +		.code1 = 149,
> +		.code2 = 818,
> +		.v1 = 1001,
> +		.v2 = 5497,
> +	},
> +	{},	/* ch 11, internal, VBUS charging current */
> +	{},	/* ch 12, internal, Die temperature */
> +	{},	/* ch 13, internal, Die temperature */
> +	{ /* ch 14, internal, USB ID line */
> +		.code1 = 48,
> +		.code2 = 714,
> +		.v1 = 323,
> +		.v2 = 4800,
> +	},
> +	{},	/* ch 15, internal, test network */
> +	{},	/* ch 16, internal, test network */
> +};
> +
> +static const struct twl6030_ideal_code
> +			twl6032_ideal[TWL6032_GPADC_MAX_CHANNELS] = {

[0] = {
will make this clearer to read.
> +	{ /* ch 0, external, battery type, resistor value */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 1, external, battery temperature, NTC resistor value */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 2, external, audio accessory/general purpose */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 660,
> +		.v2 = 1500,
> +	},
> +	{ /* ch 3, external, temperature with external diode/general purpose */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 4, external, temperature measurement/general purpose */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 5, external, general purpose */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 6, external, general purpose */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch7, internal, system supply */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 2200,
> +		.v2 = 5000,
> +	},
> +	{ /* ch8, internal, backup battery */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 2200,
> +		.v2 = 5000,
> +	},
> +	{ /* ch 9, internal, external charger input */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 3960,
> +		.v2 = 9000,
> +	},
> +	{ /* ch10, internal, VBUS */
> +		.code1 = 150,
> +		.code2 = 751,
> +		.v1 = 1000,
> +		.v2 = 5000,
> +	},
> +	{ /* ch 11, internal, VBUS DC-DC output current */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 660,
> +		.v2 = 1500,
> +	},
> +	{ /* ch 12, internal, Die temperature */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 13, internal, Die temperature */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 440,
> +		.v2 = 1000,
> +	},
> +	{ /* ch 14, internal, USB ID line */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 2420,
> +		.v2 = 5500,
> +	},
> +	{},	/* ch 15, internal, test network */
> +	{},	/* ch 16, internal, test network */
> +	{},	/* ch 17, internal, battery charging current */
> +	{ /* ch 18, internal, battery voltage */
> +		.code1 = 1441,
> +		.code2 = 3276,
> +		.v1 = 2200,
> +		.v2 = 5000,
> +	},
> +};
> +
> +static inline int twl6030_gpadc_write(u8 reg, u8 val)
> +{
> +	return twl_i2c_write_u8(TWL6030_MODULE_GPADC, val, reg);
> +}
> +
> +static inline int twl6030_gpadc_read(u8 *val, u8 reg)
> +{
> +
> +	return twl_i2c_read(TWL6030_MODULE_GPADC, val, reg, 2);
> +}
> +
> +static int twl6030_gpadc_enable_irq(u8 mask)
> +{
> +	int ret;
> +
> +	ret = twl6030_interrupt_unmask(mask, REG_INT_MSK_LINE_B);
> +	ret |= twl6030_interrupt_unmask(mask, REG_INT_MSK_STS_B);
I'd just roll these into 1 if it makes sense not to check the value
of ret in between.
> +
> +	return ret;
> +}
> +
> +static void twl6030_gpadc_disable_irq(u8 mask)
> +{
> +	twl6030_interrupt_mask(mask, REG_INT_MSK_LINE_B);
> +	twl6030_interrupt_mask(mask, REG_INT_MSK_STS_B);
> +}
> +
> +static irqreturn_t twl6030_gpadc_irq_handler(int irq, void *_gpadc)
> +{
> +	struct twl6030_gpadc_data *gpadc = _gpadc;
> +
> +	complete(&gpadc->irq_complete);
> +
> +	return IRQ_HANDLED;
> +}
> +
> +static int twl6030_start_conversion(int channel)
> +{
> +	return twl6030_gpadc_write(TWL6030_GPADC_CTRL_P1,
> +					TWL6030_GPADC_CTRL_P1_SP1);
> +}
> +
> +static int twl6032_start_conversion(int channel)
> +{
> +	int ret;
> +
> +	ret = twl6030_gpadc_write(TWL6032_GPADC_GPSELECT_ISB, channel);
> +	if (ret)
> +		return ret;
> +
> +	return twl6030_gpadc_write(TWL6032_GPADC_CTRL_P1,
> +						TWL6030_GPADC_CTRL_P1_SP1);
> +}
> +
> +static u8 twl6030_channel_to_reg(int channel)
> +{
> +	return TWL6030_GPADC_GPCH0_LSB + 2 * channel;
> +}
> +
> +static u8 twl6032_channel_to_reg(int channel)
> +{
> +	return TWL6032_GPADC_GPCH0_LSB;
> +}
> +
> +static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,
> +		int channel, int *res)
> +{
> +	u8 reg = gpadc->pdata->channel_to_reg(channel);
> +	u8 val[2];

le16 val;
> +	int ret;
> +
ret = twl6030_gpadc_read(val, reg);
(note that (reg, val) ordering of parameters would be a more common choice)

> +	ret = twl6030_gpadc_read(val, reg);
> +	if (ret) {
> +		dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);
> +		return ret;
> +	}
> +
res = le16_to_cpup(val);

> +	*res = (val[1] << 8) | val[0];
> +
> +	return ret;
> +}
> +
> +static int twl6030_gpadc_get_processed(struct twl6030_gpadc_data *gpadc,
> +		int channel, int *val)
> +{
> +	int raw_code;
> +	int corrected_code;
> +	int channel_value;
> +	int ret;
> +
> +	ret = twl6030_gpadc_get_raw(gpadc, channel, &raw_code);
> +	if (ret)
> +		return ret;
> +

Might first thought on seeing this is that it would be better to return
raw for all channels and provide the scale and offset info_mask elements
where possible rather than doing the conversion in the kernel.  Note we
allow really quite a bit of precision on those values so I doubt it
will be a problem.

If nothing else it would make everything rather more consistent.


> +	corrected_code = ((raw_code * 1000) -
> +		gpadc->twl6030_cal_tbl[channel].offset_error) /
> +		gpadc->twl6030_cal_tbl[channel].gain_error;
> +
> +	channel_value = corrected_code *
> +			gpadc->twl6030_cal_tbl[channel].gain;
> +
> +	/* Shift back into mV range */
> +	channel_value /= 1000;
> +
> +	dev_dbg(gpadc->dev, "GPADC raw: %d", raw_code);
> +	dev_dbg(gpadc->dev, "GPADC cor: %d", corrected_code);
> +	dev_dbg(gpadc->dev, "GPADC val: %d", channel_value);
> +
> +	*val = channel_value;
> +
> +	return ret;
> +}
> +
> +static int twl6030_gpadc_read_raw(struct iio_dev *indio_dev,
> +			     const struct iio_chan_spec *chan,
> +			     int *val, int *val2, long mask)
> +{
> +	struct twl6030_gpadc_data *gpadc = iio_priv(indio_dev);
> +	int ret = -EINVAL;
> +
> +	mutex_lock(&gpadc->lock);
> +
> +	ret = gpadc->pdata->start_conversion(chan->channel);
> +	if (ret) {
> +		dev_err(gpadc->dev, "failed to start conversion\n");
> +		goto err;
> +	}
> +	/* wait for conversion to complete */
> +	wait_for_completion_interruptible_timeout(&gpadc->irq_complete,
> +						msecs_to_jiffies(5000));
> +
> +	switch (mask) {
> +	case IIO_CHAN_INFO_RAW:
> +		ret = twl6030_gpadc_get_raw(gpadc, chan->channel, val);
> +		ret = ret ? -EIO : IIO_VAL_INT;
> +		break;
> +
> +	case IIO_CHAN_INFO_PROCESSED:
> +		ret = twl6030_gpadc_get_processed(gpadc, chan->channel, val);
> +		ret = ret ? -EIO : IIO_VAL_INT;
> +		break;
> +
> +	default:
> +		break;
> +	}
> +err:
> +	mutex_unlock(&gpadc->lock);
> +
> +	return ret;
> +}
> +
> +static void twl6030_calibrate_channel(struct twl6030_gpadc_data *gpadc,
> +					int channel, int d1, int d2)
> +{
> +	int b, k, gain, x1, x2;
> +	const struct twl6030_ideal_code *ideal = gpadc->pdata->ideal;
> +
> +	/* Gain */
> +	gain = ((ideal[channel].v2 - ideal[channel].v1) * 1000) /
> +		(ideal[channel].code2 - ideal[channel].code1);
> +
> +	x1 = ideal[channel].code1;
> +	x2 = ideal[channel].code2;
> +
> +	/* k */
> +	k = 1000 + (((d2 - d1) * 1000) / (x2 - x1));
> +
> +	/* b */
> +	b = (d1 * 1000) - (k - 1000) * x1;
> +
> +	gpadc->twl6030_cal_tbl[channel].gain = gain;
> +	gpadc->twl6030_cal_tbl[channel].gain_error = k;
> +	gpadc->twl6030_cal_tbl[channel].offset_error = b;
> +
> +	dev_dbg(gpadc->dev, "GPADC d1   for Chn: %d = %d\n", channel, d1);
> +	dev_dbg(gpadc->dev, "GPADC d2   for Chn: %d = %d\n", channel, d2);
> +	dev_dbg(gpadc->dev, "GPADC x1   for Chn: %d = %d\n", channel, x1);
> +	dev_dbg(gpadc->dev, "GPADC x2   for Chn: %d = %d\n", channel, x2);
> +	dev_dbg(gpadc->dev, "GPADC Gain for Chn: %d = %d\n", channel, gain);
> +	dev_dbg(gpadc->dev, "GPADC k    for Chn: %d = %d\n", channel, k);
> +	dev_dbg(gpadc->dev, "GPADC b    for Chn: %d = %d\n", channel, b);
> +}
> +
> +static inline int twl6030_gpadc_get_trim_offset(s8 d)
> +{
> +	int sign;
> +
> +	/*
> +	 * XXX NOTE!
> +	 * bit 0 - sign, bit 7 - reserved, 6..1 - trim value
> +	 * though, the documentation states that trim value
> +	 * is absolute value, the correct conversion results are
> +	 * obtained if the value is interpreted as 2's complement.
> +	 */
I'd be tempted to reorder this and then use the sign extend standard
functions.
	temp = ((d & 0x7f) >> 1) | ((d & 1) << 7);
	return sign_extend32(temp, 7);

(or something like that)

> +	sign = d & 1;
> +	d = (d & 0x7f) >> 1;
> +
> +	return sign ? (d | 0xc0) : d;
> +}
> +
> +static int twl6030_calibration(struct twl6030_gpadc_data *gpadc)
> +{
> +	int ret;
> +	int chn;
> +	u8 trim_regs[TWL6030_GPADC_MAX_CHANNELS];
> +	s8 d1, d2;
> +
> +	/*
> +	 * for calibration two measurements have been performed at
> +	 * factory, for some channels, during the production test and
> +	 * have been stored in registers. This two stored values are
> +	 * used to correct the measurements. The values represent
> +	 * offsets for the given input from the output on ideal curve.
> +	 */
> +	ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs,
> +			TWL6030_GPADC_TRIM1, 16);
> +	if (ret < 0) {
> +		dev_err(gpadc->dev, "calibration failed\n");
> +		return ret;
> +	}
> +
> +	for (chn = 0; chn < TWL6030_GPADC_MAX_CHANNELS; chn++) {
> +
> +		switch (chn) {
> +		case 0:
> +			d1 = trim_regs[0];
> +			d2 = trim_regs[1];
> +			break;
> +		case 1:
> +			d1 = trim_regs[4];
> +			d2 = trim_regs[5];
> +			break;
> +		case 2:
> +			d1 = trim_regs[12];
> +			d2 = trim_regs[13];
> +			break;
Given you give up entirely on ordering in the next calibration function
I'd be tempted to do the same here and have all the d1 = ...[4] etc in one
case statement.
> +		case 3:
> +		case 4:
> +		case 5:
> +		case 6:
> +			d1 = trim_regs[4];
> +			d2 = trim_regs[5];
> +			break;
> +		case 7:
> +			d1 = trim_regs[6];
> +			d2 = trim_regs[7];
> +			break;
> +		case 8:
> +			d1 = trim_regs[2];
> +			d2 = trim_regs[3];
> +			break;
> +		case 9:
> +			d1 = trim_regs[8];
> +			d2 = trim_regs[9];
> +			break;
> +		case 10:
> +			d1 = trim_regs[10];
> +			d2 = trim_regs[11];
> +			break;
> +		case 14:
> +			d1 = trim_regs[14];
> +			d2 = trim_regs[15];
> +			break;
> +		default:
> +			continue;
> +		}
> +
> +		d1 = twl6030_gpadc_get_trim_offset(d1);
> +		d2 = twl6030_gpadc_get_trim_offset(d2);
> +
> +		twl6030_calibrate_channel(gpadc, chn, d1, d2);
> +	}
> +
> +	return 0;
> +}
> +
> +static int twl6032_calibration(struct twl6030_gpadc_data *gpadc)
> +{
> +	int chn, d1 = 0, d2 = 0, temp;
> +	u8 trim_regs[17];
> +	int ret;
> +
> +	ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs + 1,
> +			TWL6030_GPADC_TRIM1, 16);
> +	if (ret < 0) {
> +		dev_err(gpadc->dev, "calibration failed\n");
> +		return ret;
> +	}
> +
/*
 * Loop
please for kernel code.
> +	/* Loop to calculate the value needed for returning voltages from
> +	 * GPADC not values.
> +	 *
> +	 * gain is calculated to 3 decimal places fixed point.
> +	 */
> +	for (chn = 0; chn < TWL6032_GPADC_MAX_CHANNELS; chn++) {
> +
> +		switch (chn) {
> +		case 0:
> +		case 1:
> +		case 2:
> +		case 3:
> +		case 4:
> +		case 5:
> +		case 6:
> +		case 11:
> +		case 12:
> +		case 13:
> +		case 14:
> +			/* D1 */
> +			d1 = (trim_regs[3] & 0x1F) << 2;
> +			d1 |= (trim_regs[1] & 0x06) >> 1;
> +			if (trim_regs[1] & 0x01)
> +				d1 = -d1;
> +
> +			/* D2 */
> +			d2 = (trim_regs[4] & 0x3F) << 2;
> +			d2 |= (trim_regs[2] & 0x06) >> 1;
> +			if (trim_regs[2] & 0x01)
> +				d2 = -d2;
> +			break;
> +		case 8:
No coment on your code, but this is an 'interesting' bit
of bit packing...
I did vaguely wonder if this could be mapped to a big lookup table,
but having tried it I think I end up with something almost as tricky
to read.. Oh well that was a fun 10 minute diversion ;)

> +			/* D1 */
> +			temp = (trim_regs[3] & 0x1F) << 2;
> +			temp |= (trim_regs[1] & 0x06) >> 1;
> +			if (trim_regs[1] & 0x01)
> +				temp = -temp;
> +
> +			d1 = (trim_regs[8] & 0x18) << 1;
> +			d1 |= (trim_regs[7] & 0x1E) >> 1;
> +			if (trim_regs[7] & 0x01)
> +				d1 = -d1;
> +
> +			d1 += temp;
> +
> +			/* D2 */
> +			temp = (trim_regs[4] & 0x3F) << 2;
> +			temp |= (trim_regs[2] & 0x06) >> 1;
> +			if (trim_regs[2] & 0x01)
> +				temp = -temp;
> +
> +			d2 = (trim_regs[10] & 0x1F) << 2;
> +			d2 |= (trim_regs[8] & 0x06) >> 1;
> +			if (trim_regs[8] & 0x01)
> +				d2 = -d2;
> +
> +			d2 += temp;
> +			break;
> +		case 9:
> +			/* D1 */
> +			temp = (trim_regs[3] & 0x1F) << 2;
> +			temp |= (trim_regs[1] & 0x06) >> 1;
> +			if (trim_regs[1] & 0x01)
> +				temp = -temp;
> +
> +			d1 = (trim_regs[14] & 0x18) << 1;
> +			d1 |= (trim_regs[12] & 0x1E) >> 1;
> +			if (trim_regs[12] & 0x01)
> +				d1 = -d1;
> +
> +			d1 += temp;
> +
> +			/* D2 */
> +			temp = (trim_regs[4] & 0x3F) << 2;
> +			temp |= (trim_regs[2] & 0x06) >> 1;
> +			if (trim_regs[2] & 0x01)
> +				temp = -temp;
> +
> +			d2 = (trim_regs[16] & 0x1F) << 2;
> +			d2 |= (trim_regs[14] & 0x06) >> 1;
> +			if (trim_regs[14] & 0x01)
> +				d2 = -d2;
> +
> +			d2 += temp;
> +		case 10:
> +			/* D1 */
> +			d1 = (trim_regs[11] & 0x0F) << 3;
> +			d1 |= (trim_regs[9] & 0x0E) >> 1;
> +			if (trim_regs[9] & 0x01)
> +				d1 = -d1;
> +
> +			/* D2 */
> +			d2 = (trim_regs[15] & 0x0F) << 3;
> +			d2 |= (trim_regs[13] & 0x0E) >> 1;
> +			if (trim_regs[13] & 0x01)
> +				d2 = -d2;
> +			break;
> +		case 7:
> +		case 18:
> +			/* D1 */
> +			temp = (trim_regs[3] & 0x1F) << 2;
> +			temp |= (trim_regs[1] & 0x06) >> 1;
> +			if (trim_regs[1] & 0x01)
> +				temp = -temp;
> +
> +			d1 = (trim_regs[5] & 0x7E) >> 1;
> +			if (trim_regs[5] & 0x01)
> +				d1 = -d1;
> +
> +			d1 += temp;
> +
> +			/* D2 */
> +			temp = (trim_regs[4] & 0x3F) << 2;
> +			temp |= (trim_regs[2] & 0x06) >> 1;
> +			if (trim_regs[2] & 0x01)
> +				temp = -temp;
> +
> +			d2 = (trim_regs[6] & 0xFE) >> 1;
> +			if (trim_regs[6] & 0x01)
> +				d2 = -d2;
> +
> +			d2 += temp;
> +			break;
> +		default:
> +			/* No data for other channels */
> +			continue;
> +		}
> +
> +		twl6030_calibrate_channel(gpadc, chn, d1, d2);
> +	}
> +
> +	return 0;
> +}
> +
> +#define TWL6030_GPADC_CHAN(chn, chan_info) {		\
> +	.type = IIO_VOLTAGE,				\
A lot of the channels appear not to be measuring voltages.
So they should all be appropriately specified.

> +	.channel = chn,					\

Unless you are doing buffered output this stuff is of no interest to
anyone so drop .scan_type from here.

> +	.scan_type = {					\
> +		.sign = 'u',				\
> +		.storagebits = 16,			\
> +	},						\
> +	.info_mask = BIT(chan_info),			\
> +	.info_mask_separate = BIT(chan_info), 		\
> +	.indexed = 1,					\
> +}
> +
> +static const struct iio_chan_spec twl6030_gpadc_iio_channels[] = {
> +	TWL6030_GPADC_CHAN(0, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(1, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(2, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(3, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(4, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(5, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(6, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(7, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(8, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(9, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(10, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(11, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(12, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(13, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(14, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(15, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(16, IIO_CHAN_INFO_RAW),
> +};
> +
> +static const struct iio_chan_spec twl6032_gpadc_iio_channels[] = {
> +	TWL6030_GPADC_CHAN(0, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(1, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(2, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(3, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(4, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(5, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(6, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(7, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(8, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(9, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(10, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(11, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(12, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(13, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(14, IIO_CHAN_INFO_PROCESSED),
> +	TWL6030_GPADC_CHAN(15, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(16, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(17, IIO_CHAN_INFO_RAW),
> +	TWL6030_GPADC_CHAN(18, IIO_CHAN_INFO_PROCESSED),
> +};
> +
> +static const struct iio_info twl6030_gpadc_iio_info = {
> +	.read_raw = &twl6030_gpadc_read_raw,
> +	.driver_module = THIS_MODULE,
> +};
> +
> +static const struct twl6030_gpadc_platform_data twl6030_pdata = {
> +	.iio_channels = twl6030_gpadc_iio_channels,
> +	.nchannels = TWL6030_GPADC_MAX_CHANNELS,
> +	.ideal = twl6030_ideal,
> +	.start_conversion = twl6030_start_conversion,
> +	.channel_to_reg = twl6030_channel_to_reg,
> +	.calibrate = twl6030_calibration,
> +};
> +
> +static const struct twl6030_gpadc_platform_data twl6032_pdata = {
> +	.iio_channels = twl6032_gpadc_iio_channels,
> +	.nchannels = TWL6032_GPADC_MAX_CHANNELS,
> +	.ideal = twl6032_ideal,
> +	.start_conversion = twl6032_start_conversion,
> +	.channel_to_reg = twl6032_channel_to_reg,
> +	.calibrate = twl6032_calibration,
> +};
> +
> +static const struct of_device_id of_twl6030_match_tbl[] = {
> +	{
> +		.compatible = "ti,twl6030_gpadc",
> +		.data = &twl6030_pdata,
> +	},
> +	{
> +		.compatible = "ti,twl6032_gpadc",
> +		.data = &twl6032_pdata,
> +	},
> +	{ /* end */ }
> +};
> +
> +static int twl6030_gpadc_probe(struct platform_device *pdev)
> +{
> +	struct device *dev = &pdev->dev;
> +	struct twl6030_gpadc_data *gpadc;
> +	const struct twl6030_gpadc_platform_data *pdata;
> +	const struct of_device_id *match;
> +	struct iio_dev *indio_dev;
> +	int irq;
> +	int ret = 0;
> +
> +	match = of_match_device(of_match_ptr(of_twl6030_match_tbl), dev);
> +	pdata = match ? match->data : dev->platform_data;
> +
> +	if (!pdata)
> +		return -EINVAL;
> +
> +	indio_dev = iio_device_alloc(sizeof(struct twl6030_gpadc_data));
> +	if (!indio_dev) {
> +		dev_err(dev, "failed allocating iio device\n");
> +		ret = -ENOMEM;
> +	}
> +
> +	gpadc = iio_priv(indio_dev);
> +
> +	gpadc->twl6030_cal_tbl = devm_kzalloc(dev,
> +					sizeof(struct twl6030_chnl_calib) *
> +					pdata->nchannels, GFP_KERNEL);
> +	if (!gpadc->twl6030_cal_tbl)
> +		goto err;
> +
> +	gpadc->dev = dev;
> +	gpadc->pdata = pdata;
> +
> +	platform_set_drvdata(pdev, indio_dev);
> +	mutex_init(&gpadc->lock);
> +	init_completion(&gpadc->irq_complete);
> +
> +	ret = pdata->calibrate(gpadc);
> +	if (ret < 0) {
> +		dev_err(&pdev->dev, "failed to read calibration registers\n");
> +		goto err;
> +	}
> +
> +	irq = platform_get_irq(pdev, 0);
> +	if (irq < 0) {
> +		dev_err(&pdev->dev, "failed to get irq\n");
> +		goto err;
> +	}
> +
> +	ret = devm_request_threaded_irq(dev, irq, NULL,
> +					twl6030_gpadc_irq_handler,
> +					IRQF_ONESHOT, "twl6030_gpadc", gpadc);
> +	if (ret) {
> +		dev_dbg(&pdev->dev, "could not request irq\n");
> +		goto err;
> +	}
> +
> +	ret = twl6030_gpadc_enable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
> +	if (ret < 0) {
> +		dev_err(&pdev->dev, "failed to enable GPADC interrupt\n");
> +		goto err;
> +	}
> +	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
> +					TWL6030_REG_TOGGLE1);
> +	if (ret < 0) {
> +		dev_err(&pdev->dev, "failed to enable GPADC module\n");
> +		goto err;
> +	}
> +
> +	indio_dev->name = DRIVER_NAME;
> +	indio_dev->dev.parent = dev;
> +	indio_dev->info = &twl6030_gpadc_iio_info;
> +	indio_dev->modes = INDIO_DIRECT_MODE;
> +	indio_dev->channels = pdata->iio_channels;
> +	indio_dev->num_channels = pdata->nchannels;
> +
> +	ret = iio_device_register(indio_dev);
> +	if (ret)
> +		goto err;
> +
> +	return ret;
> +err:
> +	iio_device_free(indio_dev);
> +	return ret;
> +}
> +
> +static int twl6030_gpadc_remove(struct platform_device *pdev)
> +{
> +	struct iio_dev *indio_dev = platform_get_drvdata(pdev);
> +
> +	twl6030_gpadc_disable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
> +	iio_device_unregister(indio_dev);
> +	iio_device_free(indio_dev);
> +
> +	return 0;
> +}
> +
> +static int twl6030_gpadc_suspend(struct device *pdev)
> +{
> +	int ret;
> +
> +	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCR,
> +				TWL6030_REG_TOGGLE1);
> +	if (ret)
> +		dev_err(pdev, "error reseting GPADC (%d)!\n", ret);
> +
> +	return 0;
> +};
> +
> +static int twl6030_gpadc_resume(struct device *pdev)
> +{
> +	int ret;
> +
> +	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
> +				TWL6030_REG_TOGGLE1);
> +	if (ret)
> +		dev_err(pdev, "error setting GPADC (%d)!\n", ret);
> +
> +	return 0;
> +};
> +
> +static const struct dev_pm_ops twl6030_gpadc_pm_ops = {
> +	.suspend = twl6030_gpadc_suspend,
> +	.resume = twl6030_gpadc_resume,
> +};
> +
> +static struct platform_driver twl6030_gpadc_driver = {
> +	.probe		= twl6030_gpadc_probe,
> +	.remove		= twl6030_gpadc_remove,
> +	.driver		= {
> +		.name	= DRIVER_NAME,
> +		.owner	= THIS_MODULE,
> +		.pm	= &twl6030_gpadc_pm_ops,
> +		.of_match_table = of_twl6030_match_tbl,
> +	},
> +};
> +
> +module_platform_driver(twl6030_gpadc_driver);
> +
> +MODULE_ALIAS("platform: " DRIVER_NAME);
> +MODULE_AUTHOR("Texas Instruments Inc.");
> +MODULE_DESCRIPTION("twl6030 ADC driver");
> +MODULE_LICENSE("GPL");
>

^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-12 19:56     ` Lars-Peter Clausen
  0 siblings, 0 replies; 45+ messages in thread
From: Lars-Peter Clausen @ 2013-07-12 19:56 UTC (permalink / raw)
  To: Oleksandr Kozaruk
  Cc: tony, benoit.cousson, rnayak, peter.ujfalusi, kishon, jic23,
	grant.likely, rob.herring, sameo, ch.naveen, poeschel, milo.kim,
	balajitk, gg, linux-arm-kernel, linux-kernel, linux-iio,
	devicetree-discuss, linux-omap


A couple of comments inline.

On 07/12/2013 09:18 AM, Oleksandr Kozaruk wrote:
> diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig
> index ab0767e6..87d699e 100644
> --- a/drivers/iio/adc/Kconfig
> +++ b/drivers/iio/adc/Kconfig
> @@ -157,4 +157,12 @@ config VIPERBOARD_ADC
>   	  Say yes here to access the ADC part of the Nano River
>   	  Technologies Viperboard.
>
> +config TWL6030_GPADC
> +	tristate "TWL6030 GPADC (General Purpose A/D Convertor) Support"
> +	depends on TWL4030_CORE
> +	default n
> +	help
> +	  Say yes here if you want support for the TWL6030 General Purpose
> +	  A/D Convertor.
> +

Keep it in alphabetical order

>   endmenu
> diff --git a/drivers/iio/adc/Makefile b/drivers/iio/adc/Makefile
> index 0a825be..8b05633 100644
> --- a/drivers/iio/adc/Makefile
> +++ b/drivers/iio/adc/Makefile
> @@ -17,3 +17,4 @@ obj-$(CONFIG_MAX1363) += max1363.o
>   obj-$(CONFIG_TI_ADC081C) += ti-adc081c.o
>   obj-$(CONFIG_TI_AM335X_ADC) += ti_am335x_adc.o
>   obj-$(CONFIG_VIPERBOARD_ADC) += viperboard_adc.o
> +obj-$(CONFIG_TWL6030_GPADC) += twl6030-gpadc.o

Same here.

> diff --git a/drivers/iio/adc/twl6030-gpadc.c b/drivers/iio/adc/twl6030-gpadc.c
> new file mode 100644
> index 0000000..6ceb789
> --- /dev/null
> +++ b/drivers/iio/adc/twl6030-gpadc.c
> @@ -0,0 +1,1019 @@
[...]
> +static u8 twl6032_channel_to_reg(int channel)
> +{
> +	return TWL6032_GPADC_GPCH0_LSB;

There is more than one channel, isn't there?

[...]
 > +static int twl6030_gpadc_read_raw(struct iio_dev *indio_dev,
 > +			     const struct iio_chan_spec *chan,
 > +			     int *val, int *val2, long mask)
 > +{
 > +	struct twl6030_gpadc_data *gpadc = iio_priv(indio_dev);
 > +	int ret = -EINVAL;
 > +
 > +	mutex_lock(&gpadc->lock);
 > +
 > +	ret = gpadc->pdata->start_conversion(chan->channel);
 > +	if (ret) {
 > +		dev_err(gpadc->dev, "failed to start conversion\n");
 > +		goto err;
 > +	}
 > +	/* wait for conversion to complete */
 > +	wait_for_completion_interruptible_timeout(&gpadc->irq_complete,
 > +						msecs_to_jiffies(5000));

wait_for_completion_interruptible_timeout() can return either a negative error 
code if it was interrupted, 0 if it timed out, or a positive value if it was 
successfully completed. You need to handle all three cases. Have a look at 
other existing drivers to see how to do this properly.

 > +
 > +	switch (mask) {
 > +	case IIO_CHAN_INFO_RAW:
 > +		ret = twl6030_gpadc_get_raw(gpadc, chan->channel, val);
 > +		ret = ret ? -EIO : IIO_VAL_INT;
 > +		break;
 > +
 > +	case IIO_CHAN_INFO_PROCESSED:
 > +		ret = twl6030_gpadc_get_processed(gpadc, chan->channel, val);
 > +		ret = ret ? -EIO : IIO_VAL_INT;
 > +		break;
 > +
 > +	default:
 > +		break;
 > +	}
 > +err:
 > +	mutex_unlock(&gpadc->lock);
 > +
 > +	return ret;
 > +}

> +}
> +
[...]
> +static int twl6030_gpadc_probe(struct platform_device *pdev)
> +{
> +	struct device *dev = &pdev->dev;
> +	struct twl6030_gpadc_data *gpadc;
> +	const struct twl6030_gpadc_platform_data *pdata;
> +	const struct of_device_id *match;
> +	struct iio_dev *indio_dev;
> +	int irq;
> +	int ret = 0;
> +
> +	match = of_match_device(of_match_ptr(of_twl6030_match_tbl), dev);
> +	pdata = match ? match->data : dev->platform_data;
> +
> +	if (!pdata)
> +		return -EINVAL;
> +
> +	indio_dev = iio_device_alloc(sizeof(struct twl6030_gpadc_data));

sizeof(*gpadc)

> +	if (!indio_dev) {
> +		dev_err(dev, "failed allocating iio device\n");
> +		ret = -ENOMEM;
> +	}
> +
> +	gpadc = iio_priv(indio_dev);
> +
> +	gpadc->twl6030_cal_tbl = devm_kzalloc(dev,
> +					sizeof(struct twl6030_chnl_calib) *
> +					pdata->nchannels, GFP_KERNEL);
> +	if (!gpadc->twl6030_cal_tbl)
> +		goto err;
> +
> +	gpadc->dev = dev;
> +	gpadc->pdata = pdata;
> +
> +	platform_set_drvdata(pdev, indio_dev);
> +	mutex_init(&gpadc->lock);
> +	init_completion(&gpadc->irq_complete);
> +
> +	ret = pdata->calibrate(gpadc);
> +	if (ret < 0) {
> +		dev_err(&pdev->dev, "failed to read calibration registers\n");
> +		goto err;
> +	}
> +
> +	irq = platform_get_irq(pdev, 0);
> +	if (irq < 0) {
> +		dev_err(&pdev->dev, "failed to get irq\n");
> +		goto err;
> +	}
> +
> +	ret = devm_request_threaded_irq(dev, irq, NULL,
> +					twl6030_gpadc_irq_handler,
> +					IRQF_ONESHOT, "twl6030_gpadc", gpadc);

You access memory in the interrupt handler which is freed before the interrupt 
handler is freed.

> +	if (ret) {
> +		dev_dbg(&pdev->dev, "could not request irq\n");
> +		goto err;
> +	}
> +
> +	ret = twl6030_gpadc_enable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
> +	if (ret < 0) {
> +		dev_err(&pdev->dev, "failed to enable GPADC interrupt\n");
> +		goto err;
> +	}
> +	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
> +					TWL6030_REG_TOGGLE1);
> +	if (ret < 0) {
> +		dev_err(&pdev->dev, "failed to enable GPADC module\n");
> +		goto err;
> +	}
> +
> +	indio_dev->name = DRIVER_NAME;
> +	indio_dev->dev.parent = dev;
> +	indio_dev->info = &twl6030_gpadc_iio_info;
> +	indio_dev->modes = INDIO_DIRECT_MODE;
> +	indio_dev->channels = pdata->iio_channels;
> +	indio_dev->num_channels = pdata->nchannels;
> +
> +	ret = iio_device_register(indio_dev);
> +	if (ret)
> +		goto err;
> +
> +	return ret;
> +err:
> +	iio_device_free(indio_dev);
> +	return ret;
> +}
> +
[...]
> +static int twl6030_gpadc_suspend(struct device *pdev)
> +{
> +	int ret;
> +
> +	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCR,
> +				TWL6030_REG_TOGGLE1);
> +	if (ret)
> +		dev_err(pdev, "error reseting GPADC (%d)!\n", ret);
> +
> +	return 0;
> +};
> +
> +static int twl6030_gpadc_resume(struct device *pdev)
> +{
> +	int ret;
> +
> +	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
> +				TWL6030_REG_TOGGLE1);
> +	if (ret)
> +		dev_err(pdev, "error setting GPADC (%d)!\n", ret);
> +
> +	return 0;
> +};
> +
> +static const struct dev_pm_ops twl6030_gpadc_pm_ops = {
> +	.suspend = twl6030_gpadc_suspend,
> +	.resume = twl6030_gpadc_resume,
> +};

SIMPLE_DEV_PM_OPS?

[...]

^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-12 19:56     ` Lars-Peter Clausen
  0 siblings, 0 replies; 45+ messages in thread
From: Lars-Peter Clausen @ 2013-07-12 19:56 UTC (permalink / raw)
  To: Oleksandr Kozaruk
  Cc: tony-4v6yS6AI5VpBDgjK7y7TUQ,
	benoit.cousson-QSEj5FYQhm4dnm+yROfE0A, rnayak-l0cyMroinI0,
	peter.ujfalusi-l0cyMroinI0, kishon-l0cyMroinI0,
	jic23-KWPb1pKIrIJaa/9Udqfwiw,
	grant.likely-QSEj5FYQhm4dnm+yROfE0A,
	rob.herring-bsGFqQB8/DxBDgjK7y7TUQ, sameo-VuQAYsv1563Yd54FQh9/CA,
	ch.naveen-Sze3O3UU22JBDgjK7y7TUQ,
	poeschel-Xtl8qvBWbHwb1SvskN2V4Q, milo.kim-l0cyMroinI0,
	balajitk-l0cyMroinI0, gg-kDsPt+C1G03kYMGBc/C6ZA,
	linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA,
	linux-iio-u79uwXL29TY76Z2rM5mHXA,
	devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ,
	linux-omap-u79uwXL29TY76Z2rM5mHXA


A couple of comments inline.

On 07/12/2013 09:18 AM, Oleksandr Kozaruk wrote:
> diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig
> index ab0767e6..87d699e 100644
> --- a/drivers/iio/adc/Kconfig
> +++ b/drivers/iio/adc/Kconfig
> @@ -157,4 +157,12 @@ config VIPERBOARD_ADC
>   	  Say yes here to access the ADC part of the Nano River
>   	  Technologies Viperboard.
>
> +config TWL6030_GPADC
> +	tristate "TWL6030 GPADC (General Purpose A/D Convertor) Support"
> +	depends on TWL4030_CORE
> +	default n
> +	help
> +	  Say yes here if you want support for the TWL6030 General Purpose
> +	  A/D Convertor.
> +

Keep it in alphabetical order

>   endmenu
> diff --git a/drivers/iio/adc/Makefile b/drivers/iio/adc/Makefile
> index 0a825be..8b05633 100644
> --- a/drivers/iio/adc/Makefile
> +++ b/drivers/iio/adc/Makefile
> @@ -17,3 +17,4 @@ obj-$(CONFIG_MAX1363) += max1363.o
>   obj-$(CONFIG_TI_ADC081C) += ti-adc081c.o
>   obj-$(CONFIG_TI_AM335X_ADC) += ti_am335x_adc.o
>   obj-$(CONFIG_VIPERBOARD_ADC) += viperboard_adc.o
> +obj-$(CONFIG_TWL6030_GPADC) += twl6030-gpadc.o

Same here.

> diff --git a/drivers/iio/adc/twl6030-gpadc.c b/drivers/iio/adc/twl6030-gpadc.c
> new file mode 100644
> index 0000000..6ceb789
> --- /dev/null
> +++ b/drivers/iio/adc/twl6030-gpadc.c
> @@ -0,0 +1,1019 @@
[...]
> +static u8 twl6032_channel_to_reg(int channel)
> +{
> +	return TWL6032_GPADC_GPCH0_LSB;

There is more than one channel, isn't there?

[...]
 > +static int twl6030_gpadc_read_raw(struct iio_dev *indio_dev,
 > +			     const struct iio_chan_spec *chan,
 > +			     int *val, int *val2, long mask)
 > +{
 > +	struct twl6030_gpadc_data *gpadc = iio_priv(indio_dev);
 > +	int ret = -EINVAL;
 > +
 > +	mutex_lock(&gpadc->lock);
 > +
 > +	ret = gpadc->pdata->start_conversion(chan->channel);
 > +	if (ret) {
 > +		dev_err(gpadc->dev, "failed to start conversion\n");
 > +		goto err;
 > +	}
 > +	/* wait for conversion to complete */
 > +	wait_for_completion_interruptible_timeout(&gpadc->irq_complete,
 > +						msecs_to_jiffies(5000));

wait_for_completion_interruptible_timeout() can return either a negative error 
code if it was interrupted, 0 if it timed out, or a positive value if it was 
successfully completed. You need to handle all three cases. Have a look at 
other existing drivers to see how to do this properly.

 > +
 > +	switch (mask) {
 > +	case IIO_CHAN_INFO_RAW:
 > +		ret = twl6030_gpadc_get_raw(gpadc, chan->channel, val);
 > +		ret = ret ? -EIO : IIO_VAL_INT;
 > +		break;
 > +
 > +	case IIO_CHAN_INFO_PROCESSED:
 > +		ret = twl6030_gpadc_get_processed(gpadc, chan->channel, val);
 > +		ret = ret ? -EIO : IIO_VAL_INT;
 > +		break;
 > +
 > +	default:
 > +		break;
 > +	}
 > +err:
 > +	mutex_unlock(&gpadc->lock);
 > +
 > +	return ret;
 > +}

> +}
> +
[...]
> +static int twl6030_gpadc_probe(struct platform_device *pdev)
> +{
> +	struct device *dev = &pdev->dev;
> +	struct twl6030_gpadc_data *gpadc;
> +	const struct twl6030_gpadc_platform_data *pdata;
> +	const struct of_device_id *match;
> +	struct iio_dev *indio_dev;
> +	int irq;
> +	int ret = 0;
> +
> +	match = of_match_device(of_match_ptr(of_twl6030_match_tbl), dev);
> +	pdata = match ? match->data : dev->platform_data;
> +
> +	if (!pdata)
> +		return -EINVAL;
> +
> +	indio_dev = iio_device_alloc(sizeof(struct twl6030_gpadc_data));

sizeof(*gpadc)

> +	if (!indio_dev) {
> +		dev_err(dev, "failed allocating iio device\n");
> +		ret = -ENOMEM;
> +	}
> +
> +	gpadc = iio_priv(indio_dev);
> +
> +	gpadc->twl6030_cal_tbl = devm_kzalloc(dev,
> +					sizeof(struct twl6030_chnl_calib) *
> +					pdata->nchannels, GFP_KERNEL);
> +	if (!gpadc->twl6030_cal_tbl)
> +		goto err;
> +
> +	gpadc->dev = dev;
> +	gpadc->pdata = pdata;
> +
> +	platform_set_drvdata(pdev, indio_dev);
> +	mutex_init(&gpadc->lock);
> +	init_completion(&gpadc->irq_complete);
> +
> +	ret = pdata->calibrate(gpadc);
> +	if (ret < 0) {
> +		dev_err(&pdev->dev, "failed to read calibration registers\n");
> +		goto err;
> +	}
> +
> +	irq = platform_get_irq(pdev, 0);
> +	if (irq < 0) {
> +		dev_err(&pdev->dev, "failed to get irq\n");
> +		goto err;
> +	}
> +
> +	ret = devm_request_threaded_irq(dev, irq, NULL,
> +					twl6030_gpadc_irq_handler,
> +					IRQF_ONESHOT, "twl6030_gpadc", gpadc);

You access memory in the interrupt handler which is freed before the interrupt 
handler is freed.

> +	if (ret) {
> +		dev_dbg(&pdev->dev, "could not request irq\n");
> +		goto err;
> +	}
> +
> +	ret = twl6030_gpadc_enable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
> +	if (ret < 0) {
> +		dev_err(&pdev->dev, "failed to enable GPADC interrupt\n");
> +		goto err;
> +	}
> +	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
> +					TWL6030_REG_TOGGLE1);
> +	if (ret < 0) {
> +		dev_err(&pdev->dev, "failed to enable GPADC module\n");
> +		goto err;
> +	}
> +
> +	indio_dev->name = DRIVER_NAME;
> +	indio_dev->dev.parent = dev;
> +	indio_dev->info = &twl6030_gpadc_iio_info;
> +	indio_dev->modes = INDIO_DIRECT_MODE;
> +	indio_dev->channels = pdata->iio_channels;
> +	indio_dev->num_channels = pdata->nchannels;
> +
> +	ret = iio_device_register(indio_dev);
> +	if (ret)
> +		goto err;
> +
> +	return ret;
> +err:
> +	iio_device_free(indio_dev);
> +	return ret;
> +}
> +
[...]
> +static int twl6030_gpadc_suspend(struct device *pdev)
> +{
> +	int ret;
> +
> +	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCR,
> +				TWL6030_REG_TOGGLE1);
> +	if (ret)
> +		dev_err(pdev, "error reseting GPADC (%d)!\n", ret);
> +
> +	return 0;
> +};
> +
> +static int twl6030_gpadc_resume(struct device *pdev)
> +{
> +	int ret;
> +
> +	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
> +				TWL6030_REG_TOGGLE1);
> +	if (ret)
> +		dev_err(pdev, "error setting GPADC (%d)!\n", ret);
> +
> +	return 0;
> +};
> +
> +static const struct dev_pm_ops twl6030_gpadc_pm_ops = {
> +	.suspend = twl6030_gpadc_suspend,
> +	.resume = twl6030_gpadc_resume,
> +};

SIMPLE_DEV_PM_OPS?

[...]

^ permalink raw reply	[flat|nested] 45+ messages in thread

* [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-12 19:56     ` Lars-Peter Clausen
  0 siblings, 0 replies; 45+ messages in thread
From: Lars-Peter Clausen @ 2013-07-12 19:56 UTC (permalink / raw)
  To: linux-arm-kernel


A couple of comments inline.

On 07/12/2013 09:18 AM, Oleksandr Kozaruk wrote:
> diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig
> index ab0767e6..87d699e 100644
> --- a/drivers/iio/adc/Kconfig
> +++ b/drivers/iio/adc/Kconfig
> @@ -157,4 +157,12 @@ config VIPERBOARD_ADC
>   	  Say yes here to access the ADC part of the Nano River
>   	  Technologies Viperboard.
>
> +config TWL6030_GPADC
> +	tristate "TWL6030 GPADC (General Purpose A/D Convertor) Support"
> +	depends on TWL4030_CORE
> +	default n
> +	help
> +	  Say yes here if you want support for the TWL6030 General Purpose
> +	  A/D Convertor.
> +

Keep it in alphabetical order

>   endmenu
> diff --git a/drivers/iio/adc/Makefile b/drivers/iio/adc/Makefile
> index 0a825be..8b05633 100644
> --- a/drivers/iio/adc/Makefile
> +++ b/drivers/iio/adc/Makefile
> @@ -17,3 +17,4 @@ obj-$(CONFIG_MAX1363) += max1363.o
>   obj-$(CONFIG_TI_ADC081C) += ti-adc081c.o
>   obj-$(CONFIG_TI_AM335X_ADC) += ti_am335x_adc.o
>   obj-$(CONFIG_VIPERBOARD_ADC) += viperboard_adc.o
> +obj-$(CONFIG_TWL6030_GPADC) += twl6030-gpadc.o

Same here.

> diff --git a/drivers/iio/adc/twl6030-gpadc.c b/drivers/iio/adc/twl6030-gpadc.c
> new file mode 100644
> index 0000000..6ceb789
> --- /dev/null
> +++ b/drivers/iio/adc/twl6030-gpadc.c
> @@ -0,0 +1,1019 @@
[...]
> +static u8 twl6032_channel_to_reg(int channel)
> +{
> +	return TWL6032_GPADC_GPCH0_LSB;

There is more than one channel, isn't there?

[...]
 > +static int twl6030_gpadc_read_raw(struct iio_dev *indio_dev,
 > +			     const struct iio_chan_spec *chan,
 > +			     int *val, int *val2, long mask)
 > +{
 > +	struct twl6030_gpadc_data *gpadc = iio_priv(indio_dev);
 > +	int ret = -EINVAL;
 > +
 > +	mutex_lock(&gpadc->lock);
 > +
 > +	ret = gpadc->pdata->start_conversion(chan->channel);
 > +	if (ret) {
 > +		dev_err(gpadc->dev, "failed to start conversion\n");
 > +		goto err;
 > +	}
 > +	/* wait for conversion to complete */
 > +	wait_for_completion_interruptible_timeout(&gpadc->irq_complete,
 > +						msecs_to_jiffies(5000));

wait_for_completion_interruptible_timeout() can return either a negative error 
code if it was interrupted, 0 if it timed out, or a positive value if it was 
successfully completed. You need to handle all three cases. Have a look at 
other existing drivers to see how to do this properly.

 > +
 > +	switch (mask) {
 > +	case IIO_CHAN_INFO_RAW:
 > +		ret = twl6030_gpadc_get_raw(gpadc, chan->channel, val);
 > +		ret = ret ? -EIO : IIO_VAL_INT;
 > +		break;
 > +
 > +	case IIO_CHAN_INFO_PROCESSED:
 > +		ret = twl6030_gpadc_get_processed(gpadc, chan->channel, val);
 > +		ret = ret ? -EIO : IIO_VAL_INT;
 > +		break;
 > +
 > +	default:
 > +		break;
 > +	}
 > +err:
 > +	mutex_unlock(&gpadc->lock);
 > +
 > +	return ret;
 > +}

> +}
> +
[...]
> +static int twl6030_gpadc_probe(struct platform_device *pdev)
> +{
> +	struct device *dev = &pdev->dev;
> +	struct twl6030_gpadc_data *gpadc;
> +	const struct twl6030_gpadc_platform_data *pdata;
> +	const struct of_device_id *match;
> +	struct iio_dev *indio_dev;
> +	int irq;
> +	int ret = 0;
> +
> +	match = of_match_device(of_match_ptr(of_twl6030_match_tbl), dev);
> +	pdata = match ? match->data : dev->platform_data;
> +
> +	if (!pdata)
> +		return -EINVAL;
> +
> +	indio_dev = iio_device_alloc(sizeof(struct twl6030_gpadc_data));

sizeof(*gpadc)

> +	if (!indio_dev) {
> +		dev_err(dev, "failed allocating iio device\n");
> +		ret = -ENOMEM;
> +	}
> +
> +	gpadc = iio_priv(indio_dev);
> +
> +	gpadc->twl6030_cal_tbl = devm_kzalloc(dev,
> +					sizeof(struct twl6030_chnl_calib) *
> +					pdata->nchannels, GFP_KERNEL);
> +	if (!gpadc->twl6030_cal_tbl)
> +		goto err;
> +
> +	gpadc->dev = dev;
> +	gpadc->pdata = pdata;
> +
> +	platform_set_drvdata(pdev, indio_dev);
> +	mutex_init(&gpadc->lock);
> +	init_completion(&gpadc->irq_complete);
> +
> +	ret = pdata->calibrate(gpadc);
> +	if (ret < 0) {
> +		dev_err(&pdev->dev, "failed to read calibration registers\n");
> +		goto err;
> +	}
> +
> +	irq = platform_get_irq(pdev, 0);
> +	if (irq < 0) {
> +		dev_err(&pdev->dev, "failed to get irq\n");
> +		goto err;
> +	}
> +
> +	ret = devm_request_threaded_irq(dev, irq, NULL,
> +					twl6030_gpadc_irq_handler,
> +					IRQF_ONESHOT, "twl6030_gpadc", gpadc);

You access memory in the interrupt handler which is freed before the interrupt 
handler is freed.

> +	if (ret) {
> +		dev_dbg(&pdev->dev, "could not request irq\n");
> +		goto err;
> +	}
> +
> +	ret = twl6030_gpadc_enable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
> +	if (ret < 0) {
> +		dev_err(&pdev->dev, "failed to enable GPADC interrupt\n");
> +		goto err;
> +	}
> +	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
> +					TWL6030_REG_TOGGLE1);
> +	if (ret < 0) {
> +		dev_err(&pdev->dev, "failed to enable GPADC module\n");
> +		goto err;
> +	}
> +
> +	indio_dev->name = DRIVER_NAME;
> +	indio_dev->dev.parent = dev;
> +	indio_dev->info = &twl6030_gpadc_iio_info;
> +	indio_dev->modes = INDIO_DIRECT_MODE;
> +	indio_dev->channels = pdata->iio_channels;
> +	indio_dev->num_channels = pdata->nchannels;
> +
> +	ret = iio_device_register(indio_dev);
> +	if (ret)
> +		goto err;
> +
> +	return ret;
> +err:
> +	iio_device_free(indio_dev);
> +	return ret;
> +}
> +
[...]
> +static int twl6030_gpadc_suspend(struct device *pdev)
> +{
> +	int ret;
> +
> +	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCR,
> +				TWL6030_REG_TOGGLE1);
> +	if (ret)
> +		dev_err(pdev, "error reseting GPADC (%d)!\n", ret);
> +
> +	return 0;
> +};
> +
> +static int twl6030_gpadc_resume(struct device *pdev)
> +{
> +	int ret;
> +
> +	ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
> +				TWL6030_REG_TOGGLE1);
> +	if (ret)
> +		dev_err(pdev, "error setting GPADC (%d)!\n", ret);
> +
> +	return 0;
> +};
> +
> +static const struct dev_pm_ops twl6030_gpadc_pm_ops = {
> +	.suspend = twl6030_gpadc_suspend,
> +	.resume = twl6030_gpadc_resume,
> +};

SIMPLE_DEV_PM_OPS?

[...]

^ permalink raw reply	[flat|nested] 45+ messages in thread

* RE: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
  2013-07-12 19:56     ` Lars-Peter Clausen
  (?)
  (?)
@ 2013-07-15 11:09       ` Kozaruk, Oleksandr
  -1 siblings, 0 replies; 45+ messages in thread
From: Kozaruk, Oleksandr @ 2013-07-15 11:09 UTC (permalink / raw)
  To: Lars-Peter Clausen
  Cc: tony, benoit.cousson, Nayak, Rajendra, Ujfalusi, Peter, ABRAHAM,
	KISHON VIJAY, jic23, grant.likely, rob.herring, sameo, ch.naveen,
	poeschel, Kim, Milo, Krishnamoorthy, Balaji T, gg,
	linux-arm-kernel, linux-kernel, linux-iio, devicetree-discuss,
	linux-omap

Hello Lars-Peter,
Thank you for the review.

>> diff --git a/drivers/iio/adc/twl6030-gpadc.c b/drivers/iio/adc/twl6030-gpadc.c
>> new file mode 100644
>> index 0000000..6ceb789
>> --- /dev/null
>> +++ b/drivers/iio/adc/twl6030-gpadc.c
>> @@ -0,0 +1,1019 @@
>[...]
>> +static u8 twl6032_channel_to_reg(int channel)
>> +{
>> +     return TWL6032_GPADC_GPCH0_LSB;
>
>There is more than one channel, isn't there?
Yes. But for twl6032 channel of interest is chosen first. When the conversion
is ready tre result is available in GPCH0_LSB/GPCH1_MSB for any cosen
channel. For twl6030 there are as many result register pairs as many of
channels.

>> +     ret = devm_request_threaded_irq(dev, irq, NULL,
>> +                                     twl6030_gpadc_irq_handler,
>> +                                     IRQF_ONESHOT, "twl6030_gpadc", gpadc);
>
>You access memory in the interrupt handler which is freed before the interrupt
>handler is freed.
Thanks for pointing this. devm_* will free memory for irq after the driver
is removed and memory for the device is freed. I took me awhile to understand
this. Is there going to be something like devm_iio_device_alloc? whould it be helpfull?

^ permalink raw reply	[flat|nested] 45+ messages in thread

* RE: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 11:09       ` Kozaruk, Oleksandr
  0 siblings, 0 replies; 45+ messages in thread
From: Kozaruk, Oleksandr @ 2013-07-15 11:09 UTC (permalink / raw)
  To: Lars-Peter Clausen
  Cc: Kim, Milo, linux-omap, sameo, tony, devicetree-discuss, Nayak,
	Rajendra, poeschel, linux-kernel, rob.herring, ABRAHAM,
	KISHON VIJAY, Ujfalusi, Peter, linux-iio, jic23, grant.likely,
	benoit.cousson, gg, ch.naveen,
	linux-arm-kernel@lists.infradead.org

Hello Lars-Peter,
Thank you for the review.

>> diff --git a/drivers/iio/adc/twl6030-gpadc.c b/drivers/iio/adc/twl6030-gpadc.c
>> new file mode 100644
>> index 0000000..6ceb789
>> --- /dev/null
>> +++ b/drivers/iio/adc/twl6030-gpadc.c
>> @@ -0,0 +1,1019 @@
>[...]
>> +static u8 twl6032_channel_to_reg(int channel)
>> +{
>> +     return TWL6032_GPADC_GPCH0_LSB;
>
>There is more than one channel, isn't there?
Yes. But for twl6032 channel of interest is chosen first. When the conversion
is ready tre result is available in GPCH0_LSB/GPCH1_MSB for any cosen
channel. For twl6030 there are as many result register pairs as many of
channels.

>> +     ret = devm_request_threaded_irq(dev, irq, NULL,
>> +                                     twl6030_gpadc_irq_handler,
>> +                                     IRQF_ONESHOT, "twl6030_gpadc", gpadc);
>
>You access memory in the interrupt handler which is freed before the interrupt
>handler is freed.
Thanks for pointing this. devm_* will free memory for irq after the driver
is removed and memory for the device is freed. I took me awhile to understand
this. Is there going to be something like devm_iio_device_alloc? whould it be helpfull?

^ permalink raw reply	[flat|nested] 45+ messages in thread

* RE: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 11:09       ` Kozaruk, Oleksandr
  0 siblings, 0 replies; 45+ messages in thread
From: Kozaruk, Oleksandr @ 2013-07-15 11:09 UTC (permalink / raw)
  To: Lars-Peter Clausen
  Cc: tony, benoit.cousson, Nayak, Rajendra, Ujfalusi, Peter, ABRAHAM,
	KISHON VIJAY, jic23, grant.likely, rob.herring, sameo, ch.naveen,
	poeschel, Kim, Milo, Krishnamoorthy, Balaji T, gg,
	linux-arm-kernel, linux-kernel, linux-iio, devicetree-discuss,
	linux-omap

Hello Lars-Peter,
Thank you for the review.

>> diff --git a/drivers/iio/adc/twl6030-gpadc.c b/drivers/iio/adc/twl6030-gpadc.c
>> new file mode 100644
>> index 0000000..6ceb789
>> --- /dev/null
>> +++ b/drivers/iio/adc/twl6030-gpadc.c
>> @@ -0,0 +1,1019 @@
>[...]
>> +static u8 twl6032_channel_to_reg(int channel)
>> +{
>> +     return TWL6032_GPADC_GPCH0_LSB;
>
>There is more than one channel, isn't there?
Yes. But for twl6032 channel of interest is chosen first. When the conversion
is ready tre result is available in GPCH0_LSB/GPCH1_MSB for any cosen
channel. For twl6030 there are as many result register pairs as many of
channels.

>> +     ret = devm_request_threaded_irq(dev, irq, NULL,
>> +                                     twl6030_gpadc_irq_handler,
>> +                                     IRQF_ONESHOT, "twl6030_gpadc", gpadc);
>
>You access memory in the interrupt handler which is freed before the interrupt
>handler is freed.
Thanks for pointing this. devm_* will free memory for irq after the driver
is removed and memory for the device is freed. I took me awhile to understand
this. Is there going to be something like devm_iio_device_alloc? whould it be helpfull?

^ permalink raw reply	[flat|nested] 45+ messages in thread

* [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 11:09       ` Kozaruk, Oleksandr
  0 siblings, 0 replies; 45+ messages in thread
From: Kozaruk, Oleksandr @ 2013-07-15 11:09 UTC (permalink / raw)
  To: linux-arm-kernel

Hello Lars-Peter,
Thank you for the review.

>> diff --git a/drivers/iio/adc/twl6030-gpadc.c b/drivers/iio/adc/twl6030-gpadc.c
>> new file mode 100644
>> index 0000000..6ceb789
>> --- /dev/null
>> +++ b/drivers/iio/adc/twl6030-gpadc.c
>> @@ -0,0 +1,1019 @@
>[...]
>> +static u8 twl6032_channel_to_reg(int channel)
>> +{
>> +     return TWL6032_GPADC_GPCH0_LSB;
>
>There is more than one channel, isn't there?
Yes. But for twl6032 channel of interest is chosen first. When the conversion
is ready tre result is available in GPCH0_LSB/GPCH1_MSB for any cosen
channel. For twl6030 there are as many result register pairs as many of
channels.

>> +     ret = devm_request_threaded_irq(dev, irq, NULL,
>> +                                     twl6030_gpadc_irq_handler,
>> +                                     IRQF_ONESHOT, "twl6030_gpadc", gpadc);
>
>You access memory in the interrupt handler which is freed before the interrupt
>handler is freed.
Thanks for pointing this. devm_* will free memory for irq after the driver
is removed and memory for the device is freed. I took me awhile to understand
this. Is there going to be something like devm_iio_device_alloc? whould it be helpfull?

^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 11:33         ` Lars-Peter Clausen
  0 siblings, 0 replies; 45+ messages in thread
From: Lars-Peter Clausen @ 2013-07-15 11:33 UTC (permalink / raw)
  To: Kozaruk, Oleksandr
  Cc: tony, benoit.cousson, Nayak, Rajendra, Ujfalusi, Peter, ABRAHAM,
	KISHON VIJAY, jic23, grant.likely, rob.herring, sameo, ch.naveen,
	poeschel, Kim, Milo, Krishnamoorthy, Balaji T, gg,
	linux-arm-kernel, linux-kernel, linux-iio, devicetree-discuss,
	linux-omap

On 07/15/2013 01:09 PM, Kozaruk, Oleksandr wrote:
[...]
> 
>>> +     ret = devm_request_threaded_irq(dev, irq, NULL,
>>> +                                     twl6030_gpadc_irq_handler,
>>> +                                     IRQF_ONESHOT, "twl6030_gpadc", gpadc);
>>
>> You access memory in the interrupt handler which is freed before the interrupt
>> handler is freed.
> Thanks for pointing this. devm_* will free memory for irq after the driver
> is removed and memory for the device is freed. I took me awhile to understand
> this. Is there going to be something like devm_iio_device_alloc? whould it be helpfull?
> 

Yes, I think it certainly makes sense to add a devm_iio_device_alloc(), care
to send a patch?

^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 11:33         ` Lars-Peter Clausen
  0 siblings, 0 replies; 45+ messages in thread
From: Lars-Peter Clausen @ 2013-07-15 11:33 UTC (permalink / raw)
  To: Kozaruk, Oleksandr
  Cc: tony-4v6yS6AI5VpBDgjK7y7TUQ,
	benoit.cousson-QSEj5FYQhm4dnm+yROfE0A, Nayak, Rajendra, Ujfalusi,
	Peter, ABRAHAM, KISHON VIJAY, jic23-KWPb1pKIrIJaa/9Udqfwiw,
	grant.likely-QSEj5FYQhm4dnm+yROfE0A,
	rob.herring-bsGFqQB8/DxBDgjK7y7TUQ, sameo-VuQAYsv1563Yd54FQh9/CA,
	ch.naveen-Sze3O3UU22JBDgjK7y7TUQ,
	poeschel-Xtl8qvBWbHwb1SvskN2V4Q, Kim, Milo, Krishnamoorthy,
	Balaji T, gg-kDsPt+C1G03kYMGBc/C6ZA,
	linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA,
	linux-iio-u79uwXL29TY76Z2rM5mHXA,
	devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ@public.gmane.org

On 07/15/2013 01:09 PM, Kozaruk, Oleksandr wrote:
[...]
> 
>>> +     ret = devm_request_threaded_irq(dev, irq, NULL,
>>> +                                     twl6030_gpadc_irq_handler,
>>> +                                     IRQF_ONESHOT, "twl6030_gpadc", gpadc);
>>
>> You access memory in the interrupt handler which is freed before the interrupt
>> handler is freed.
> Thanks for pointing this. devm_* will free memory for irq after the driver
> is removed and memory for the device is freed. I took me awhile to understand
> this. Is there going to be something like devm_iio_device_alloc? whould it be helpfull?
> 

Yes, I think it certainly makes sense to add a devm_iio_device_alloc(), care
to send a patch?

^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 11:33         ` Lars-Peter Clausen
  0 siblings, 0 replies; 45+ messages in thread
From: Lars-Peter Clausen @ 2013-07-15 11:33 UTC (permalink / raw)
  To: Kozaruk, Oleksandr
  Cc: tony, benoit.cousson, Nayak, Rajendra, Ujfalusi, Peter, ABRAHAM,
	KISHON VIJAY, jic23, grant.likely, rob.herring, sameo, ch.naveen,
	poeschel, Kim, Milo, Krishnamoorthy, Balaji T, gg,
	linux-arm-kernel, linux-kernel, linux-iio, devicetree-discuss,
	linux-omap

On 07/15/2013 01:09 PM, Kozaruk, Oleksandr wrote:
[...]
> 
>>> +     ret = devm_request_threaded_irq(dev, irq, NULL,
>>> +                                     twl6030_gpadc_irq_handler,
>>> +                                     IRQF_ONESHOT, "twl6030_gpadc", gpadc);
>>
>> You access memory in the interrupt handler which is freed before the interrupt
>> handler is freed.
> Thanks for pointing this. devm_* will free memory for irq after the driver
> is removed and memory for the device is freed. I took me awhile to understand
> this. Is there going to be something like devm_iio_device_alloc? whould it be helpfull?
> 

Yes, I think it certainly makes sense to add a devm_iio_device_alloc(), care
to send a patch?

^ permalink raw reply	[flat|nested] 45+ messages in thread

* [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 11:33         ` Lars-Peter Clausen
  0 siblings, 0 replies; 45+ messages in thread
From: Lars-Peter Clausen @ 2013-07-15 11:33 UTC (permalink / raw)
  To: linux-arm-kernel

On 07/15/2013 01:09 PM, Kozaruk, Oleksandr wrote:
[...]
> 
>>> +     ret = devm_request_threaded_irq(dev, irq, NULL,
>>> +                                     twl6030_gpadc_irq_handler,
>>> +                                     IRQF_ONESHOT, "twl6030_gpadc", gpadc);
>>
>> You access memory in the interrupt handler which is freed before the interrupt
>> handler is freed.
> Thanks for pointing this. devm_* will free memory for irq after the driver
> is removed and memory for the device is freed. I took me awhile to understand
> this. Is there going to be something like devm_iio_device_alloc? whould it be helpfull?
> 

Yes, I think it certainly makes sense to add a devm_iio_device_alloc(), care
to send a patch?

^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 11:56       ` Grygorii Strashko
  0 siblings, 0 replies; 45+ messages in thread
From: Grygorii Strashko @ 2013-07-15 11:56 UTC (permalink / raw)
  To: Lars-Peter Clausen
  Cc: Oleksandr Kozaruk, tony, benoit.cousson, rnayak, peter.ujfalusi,
	kishon, jic23, grant.likely, rob.herring, sameo, ch.naveen,
	poeschel, milo.kim, balajitk, gg, linux-arm-kernel, linux-kernel,
	linux-iio, devicetree-discuss, linux-omap

Hi All,

I have a question regarding this patch and IIO in general
- Does IIO provide sync mechanism with system wide suspend/resume or 
this should be handled by each driver itself?

What if during system suspend iio_read_channel_raw() (or any other 
consumer API) will be called after gpadc driver have been suspended 
already? (I did some investigation and seems it's possible - Am I right?)

If no, could "info_exist_lock" be reused for such purposes?

Regards,
-grygorii


On 07/12/2013 10:56 PM, Lars-Peter Clausen wrote:
>
> A couple of comments inline.
>
> On 07/12/2013 09:18 AM, Oleksandr Kozaruk wrote:
>> diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig
>> index ab0767e6..87d699e 100644
>> --- a/drivers/iio/adc/Kconfig
>> +++ b/drivers/iio/adc/Kconfig
>> @@ -157,4 +157,12 @@ config VIPERBOARD_ADC
>>         Say yes here to access the ADC part of the Nano River
>>         Technologies Viperboard.
>>
>> +config TWL6030_GPADC
>> +    tristate "TWL6030 GPADC (General Purpose A/D Convertor) Support"
>> +    depends on TWL4030_CORE
>> +    default n
>> +    help
>> +      Say yes here if you want support for the TWL6030 General Purpose
>> +      A/D Convertor.
>> +
>
> Keep it in alphabetical order
>
>>   endmenu
>> diff --git a/drivers/iio/adc/Makefile b/drivers/iio/adc/Makefile
>> index 0a825be..8b05633 100644
>> --- a/drivers/iio/adc/Makefile
>> +++ b/drivers/iio/adc/Makefile
>> @@ -17,3 +17,4 @@ obj-$(CONFIG_MAX1363) += max1363.o
>>   obj-$(CONFIG_TI_ADC081C) += ti-adc081c.o
>>   obj-$(CONFIG_TI_AM335X_ADC) += ti_am335x_adc.o
>>   obj-$(CONFIG_VIPERBOARD_ADC) += viperboard_adc.o
>> +obj-$(CONFIG_TWL6030_GPADC) += twl6030-gpadc.o
>
> Same here.
>
>> diff --git a/drivers/iio/adc/twl6030-gpadc.c
>> b/drivers/iio/adc/twl6030-gpadc.c
>> new file mode 100644
>> index 0000000..6ceb789
>> --- /dev/null
>> +++ b/drivers/iio/adc/twl6030-gpadc.c
>> @@ -0,0 +1,1019 @@
> [...]
>> +static u8 twl6032_channel_to_reg(int channel)
>> +{
>> +    return TWL6032_GPADC_GPCH0_LSB;
>
> There is more than one channel, isn't there?
>
> [...]
>  > +static int twl6030_gpadc_read_raw(struct iio_dev *indio_dev,
>  > +                 const struct iio_chan_spec *chan,
>  > +                 int *val, int *val2, long mask)
>  > +{
>  > +    struct twl6030_gpadc_data *gpadc = iio_priv(indio_dev);
>  > +    int ret = -EINVAL;
>  > +
>  > +    mutex_lock(&gpadc->lock);
>  > +
>  > +    ret = gpadc->pdata->start_conversion(chan->channel);
>  > +    if (ret) {
>  > +        dev_err(gpadc->dev, "failed to start conversion\n");
>  > +        goto err;
>  > +    }
>  > +    /* wait for conversion to complete */
>  > +    wait_for_completion_interruptible_timeout(&gpadc->irq_complete,
>  > +                        msecs_to_jiffies(5000));
>
> wait_for_completion_interruptible_timeout() can return either a negative
> error code if it was interrupted, 0 if it timed out, or a positive value
> if it was successfully completed. You need to handle all three cases.
> Have a look at other existing drivers to see how to do this properly.
>
>  > +
>  > +    switch (mask) {
>  > +    case IIO_CHAN_INFO_RAW:
>  > +        ret = twl6030_gpadc_get_raw(gpadc, chan->channel, val);
>  > +        ret = ret ? -EIO : IIO_VAL_INT;
>  > +        break;
>  > +
>  > +    case IIO_CHAN_INFO_PROCESSED:
>  > +        ret = twl6030_gpadc_get_processed(gpadc, chan->channel, val);
>  > +        ret = ret ? -EIO : IIO_VAL_INT;
>  > +        break;
>  > +
>  > +    default:
>  > +        break;
>  > +    }
>  > +err:
>  > +    mutex_unlock(&gpadc->lock);
>  > +
>  > +    return ret;
>  > +}
>
>> +}
>> +
> [...]
>> +static int twl6030_gpadc_probe(struct platform_device *pdev)
>> +{
>> +    struct device *dev = &pdev->dev;
>> +    struct twl6030_gpadc_data *gpadc;
>> +    const struct twl6030_gpadc_platform_data *pdata;
>> +    const struct of_device_id *match;
>> +    struct iio_dev *indio_dev;
>> +    int irq;
>> +    int ret = 0;
>> +
>> +    match = of_match_device(of_match_ptr(of_twl6030_match_tbl), dev);
>> +    pdata = match ? match->data : dev->platform_data;
>> +
>> +    if (!pdata)
>> +        return -EINVAL;
>> +
>> +    indio_dev = iio_device_alloc(sizeof(struct twl6030_gpadc_data));
>
> sizeof(*gpadc)
>
>> +    if (!indio_dev) {
>> +        dev_err(dev, "failed allocating iio device\n");
>> +        ret = -ENOMEM;
>> +    }
>> +
>> +    gpadc = iio_priv(indio_dev);
>> +
>> +    gpadc->twl6030_cal_tbl = devm_kzalloc(dev,
>> +                    sizeof(struct twl6030_chnl_calib) *
>> +                    pdata->nchannels, GFP_KERNEL);
>> +    if (!gpadc->twl6030_cal_tbl)
>> +        goto err;
>> +
>> +    gpadc->dev = dev;
>> +    gpadc->pdata = pdata;
>> +
>> +    platform_set_drvdata(pdev, indio_dev);
>> +    mutex_init(&gpadc->lock);
>> +    init_completion(&gpadc->irq_complete);
>> +
>> +    ret = pdata->calibrate(gpadc);
>> +    if (ret < 0) {
>> +        dev_err(&pdev->dev, "failed to read calibration registers\n");
>> +        goto err;
>> +    }
>> +
>> +    irq = platform_get_irq(pdev, 0);
>> +    if (irq < 0) {
>> +        dev_err(&pdev->dev, "failed to get irq\n");
>> +        goto err;
>> +    }
>> +
>> +    ret = devm_request_threaded_irq(dev, irq, NULL,
>> +                    twl6030_gpadc_irq_handler,
>> +                    IRQF_ONESHOT, "twl6030_gpadc", gpadc);
>
> You access memory in the interrupt handler which is freed before the
> interrupt handler is freed.
>
>> +    if (ret) {
>> +        dev_dbg(&pdev->dev, "could not request irq\n");
>> +        goto err;
>> +    }
>> +
>> +    ret = twl6030_gpadc_enable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
>> +    if (ret < 0) {
>> +        dev_err(&pdev->dev, "failed to enable GPADC interrupt\n");
>> +        goto err;
>> +    }
>> +    ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
>> +                    TWL6030_REG_TOGGLE1);
>> +    if (ret < 0) {
>> +        dev_err(&pdev->dev, "failed to enable GPADC module\n");
>> +        goto err;
>> +    }
>> +
>> +    indio_dev->name = DRIVER_NAME;
>> +    indio_dev->dev.parent = dev;
>> +    indio_dev->info = &twl6030_gpadc_iio_info;
>> +    indio_dev->modes = INDIO_DIRECT_MODE;
>> +    indio_dev->channels = pdata->iio_channels;
>> +    indio_dev->num_channels = pdata->nchannels;
>> +
>> +    ret = iio_device_register(indio_dev);
>> +    if (ret)
>> +        goto err;
>> +
>> +    return ret;
>> +err:
>> +    iio_device_free(indio_dev);
>> +    return ret;
>> +}
>> +
> [...]
>> +static int twl6030_gpadc_suspend(struct device *pdev)
>> +{
>> +    int ret;
>> +
>> +    ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCR,
>> +                TWL6030_REG_TOGGLE1);
>> +    if (ret)
>> +        dev_err(pdev, "error reseting GPADC (%d)!\n", ret);
>> +
>> +    return 0;
>> +};
>> +
>> +static int twl6030_gpadc_resume(struct device *pdev)
>> +{
>> +    int ret;
>> +
>> +    ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
>> +                TWL6030_REG_TOGGLE1);
>> +    if (ret)
>> +        dev_err(pdev, "error setting GPADC (%d)!\n", ret);
>> +
>> +    return 0;
>> +};
>> +
>> +static const struct dev_pm_ops twl6030_gpadc_pm_ops = {
>> +    .suspend = twl6030_gpadc_suspend,
>> +    .resume = twl6030_gpadc_resume,
>> +};
>
> SIMPLE_DEV_PM_OPS?
>
> [...]
> --
> To unsubscribe from this list: send the line "unsubscribe linux-omap" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html


^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 11:56       ` Grygorii Strashko
  0 siblings, 0 replies; 45+ messages in thread
From: Grygorii Strashko @ 2013-07-15 11:56 UTC (permalink / raw)
  To: Lars-Peter Clausen
  Cc: Oleksandr Kozaruk, tony-4v6yS6AI5VpBDgjK7y7TUQ,
	benoit.cousson-QSEj5FYQhm4dnm+yROfE0A, rnayak-l0cyMroinI0,
	peter.ujfalusi-l0cyMroinI0, kishon-l0cyMroinI0,
	jic23-KWPb1pKIrIJaa/9Udqfwiw,
	grant.likely-QSEj5FYQhm4dnm+yROfE0A,
	rob.herring-bsGFqQB8/DxBDgjK7y7TUQ, sameo-VuQAYsv1563Yd54FQh9/CA,
	ch.naveen-Sze3O3UU22JBDgjK7y7TUQ,
	poeschel-Xtl8qvBWbHwb1SvskN2V4Q, milo.kim-l0cyMroinI0,
	balajitk-l0cyMroinI0, gg-kDsPt+C1G03kYMGBc/C6ZA,
	linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA,
	linux-iio-u79uwXL29TY76Z2rM5mHXA,
	devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ,
	linux-omap-u79uwXL29TY76Z2rM5mHXA

Hi All,

I have a question regarding this patch and IIO in general
- Does IIO provide sync mechanism with system wide suspend/resume or 
this should be handled by each driver itself?

What if during system suspend iio_read_channel_raw() (or any other 
consumer API) will be called after gpadc driver have been suspended 
already? (I did some investigation and seems it's possible - Am I right?)

If no, could "info_exist_lock" be reused for such purposes?

Regards,
-grygorii


On 07/12/2013 10:56 PM, Lars-Peter Clausen wrote:
>
> A couple of comments inline.
>
> On 07/12/2013 09:18 AM, Oleksandr Kozaruk wrote:
>> diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig
>> index ab0767e6..87d699e 100644
>> --- a/drivers/iio/adc/Kconfig
>> +++ b/drivers/iio/adc/Kconfig
>> @@ -157,4 +157,12 @@ config VIPERBOARD_ADC
>>         Say yes here to access the ADC part of the Nano River
>>         Technologies Viperboard.
>>
>> +config TWL6030_GPADC
>> +    tristate "TWL6030 GPADC (General Purpose A/D Convertor) Support"
>> +    depends on TWL4030_CORE
>> +    default n
>> +    help
>> +      Say yes here if you want support for the TWL6030 General Purpose
>> +      A/D Convertor.
>> +
>
> Keep it in alphabetical order
>
>>   endmenu
>> diff --git a/drivers/iio/adc/Makefile b/drivers/iio/adc/Makefile
>> index 0a825be..8b05633 100644
>> --- a/drivers/iio/adc/Makefile
>> +++ b/drivers/iio/adc/Makefile
>> @@ -17,3 +17,4 @@ obj-$(CONFIG_MAX1363) += max1363.o
>>   obj-$(CONFIG_TI_ADC081C) += ti-adc081c.o
>>   obj-$(CONFIG_TI_AM335X_ADC) += ti_am335x_adc.o
>>   obj-$(CONFIG_VIPERBOARD_ADC) += viperboard_adc.o
>> +obj-$(CONFIG_TWL6030_GPADC) += twl6030-gpadc.o
>
> Same here.
>
>> diff --git a/drivers/iio/adc/twl6030-gpadc.c
>> b/drivers/iio/adc/twl6030-gpadc.c
>> new file mode 100644
>> index 0000000..6ceb789
>> --- /dev/null
>> +++ b/drivers/iio/adc/twl6030-gpadc.c
>> @@ -0,0 +1,1019 @@
> [...]
>> +static u8 twl6032_channel_to_reg(int channel)
>> +{
>> +    return TWL6032_GPADC_GPCH0_LSB;
>
> There is more than one channel, isn't there?
>
> [...]
>  > +static int twl6030_gpadc_read_raw(struct iio_dev *indio_dev,
>  > +                 const struct iio_chan_spec *chan,
>  > +                 int *val, int *val2, long mask)
>  > +{
>  > +    struct twl6030_gpadc_data *gpadc = iio_priv(indio_dev);
>  > +    int ret = -EINVAL;
>  > +
>  > +    mutex_lock(&gpadc->lock);
>  > +
>  > +    ret = gpadc->pdata->start_conversion(chan->channel);
>  > +    if (ret) {
>  > +        dev_err(gpadc->dev, "failed to start conversion\n");
>  > +        goto err;
>  > +    }
>  > +    /* wait for conversion to complete */
>  > +    wait_for_completion_interruptible_timeout(&gpadc->irq_complete,
>  > +                        msecs_to_jiffies(5000));
>
> wait_for_completion_interruptible_timeout() can return either a negative
> error code if it was interrupted, 0 if it timed out, or a positive value
> if it was successfully completed. You need to handle all three cases.
> Have a look at other existing drivers to see how to do this properly.
>
>  > +
>  > +    switch (mask) {
>  > +    case IIO_CHAN_INFO_RAW:
>  > +        ret = twl6030_gpadc_get_raw(gpadc, chan->channel, val);
>  > +        ret = ret ? -EIO : IIO_VAL_INT;
>  > +        break;
>  > +
>  > +    case IIO_CHAN_INFO_PROCESSED:
>  > +        ret = twl6030_gpadc_get_processed(gpadc, chan->channel, val);
>  > +        ret = ret ? -EIO : IIO_VAL_INT;
>  > +        break;
>  > +
>  > +    default:
>  > +        break;
>  > +    }
>  > +err:
>  > +    mutex_unlock(&gpadc->lock);
>  > +
>  > +    return ret;
>  > +}
>
>> +}
>> +
> [...]
>> +static int twl6030_gpadc_probe(struct platform_device *pdev)
>> +{
>> +    struct device *dev = &pdev->dev;
>> +    struct twl6030_gpadc_data *gpadc;
>> +    const struct twl6030_gpadc_platform_data *pdata;
>> +    const struct of_device_id *match;
>> +    struct iio_dev *indio_dev;
>> +    int irq;
>> +    int ret = 0;
>> +
>> +    match = of_match_device(of_match_ptr(of_twl6030_match_tbl), dev);
>> +    pdata = match ? match->data : dev->platform_data;
>> +
>> +    if (!pdata)
>> +        return -EINVAL;
>> +
>> +    indio_dev = iio_device_alloc(sizeof(struct twl6030_gpadc_data));
>
> sizeof(*gpadc)
>
>> +    if (!indio_dev) {
>> +        dev_err(dev, "failed allocating iio device\n");
>> +        ret = -ENOMEM;
>> +    }
>> +
>> +    gpadc = iio_priv(indio_dev);
>> +
>> +    gpadc->twl6030_cal_tbl = devm_kzalloc(dev,
>> +                    sizeof(struct twl6030_chnl_calib) *
>> +                    pdata->nchannels, GFP_KERNEL);
>> +    if (!gpadc->twl6030_cal_tbl)
>> +        goto err;
>> +
>> +    gpadc->dev = dev;
>> +    gpadc->pdata = pdata;
>> +
>> +    platform_set_drvdata(pdev, indio_dev);
>> +    mutex_init(&gpadc->lock);
>> +    init_completion(&gpadc->irq_complete);
>> +
>> +    ret = pdata->calibrate(gpadc);
>> +    if (ret < 0) {
>> +        dev_err(&pdev->dev, "failed to read calibration registers\n");
>> +        goto err;
>> +    }
>> +
>> +    irq = platform_get_irq(pdev, 0);
>> +    if (irq < 0) {
>> +        dev_err(&pdev->dev, "failed to get irq\n");
>> +        goto err;
>> +    }
>> +
>> +    ret = devm_request_threaded_irq(dev, irq, NULL,
>> +                    twl6030_gpadc_irq_handler,
>> +                    IRQF_ONESHOT, "twl6030_gpadc", gpadc);
>
> You access memory in the interrupt handler which is freed before the
> interrupt handler is freed.
>
>> +    if (ret) {
>> +        dev_dbg(&pdev->dev, "could not request irq\n");
>> +        goto err;
>> +    }
>> +
>> +    ret = twl6030_gpadc_enable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
>> +    if (ret < 0) {
>> +        dev_err(&pdev->dev, "failed to enable GPADC interrupt\n");
>> +        goto err;
>> +    }
>> +    ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
>> +                    TWL6030_REG_TOGGLE1);
>> +    if (ret < 0) {
>> +        dev_err(&pdev->dev, "failed to enable GPADC module\n");
>> +        goto err;
>> +    }
>> +
>> +    indio_dev->name = DRIVER_NAME;
>> +    indio_dev->dev.parent = dev;
>> +    indio_dev->info = &twl6030_gpadc_iio_info;
>> +    indio_dev->modes = INDIO_DIRECT_MODE;
>> +    indio_dev->channels = pdata->iio_channels;
>> +    indio_dev->num_channels = pdata->nchannels;
>> +
>> +    ret = iio_device_register(indio_dev);
>> +    if (ret)
>> +        goto err;
>> +
>> +    return ret;
>> +err:
>> +    iio_device_free(indio_dev);
>> +    return ret;
>> +}
>> +
> [...]
>> +static int twl6030_gpadc_suspend(struct device *pdev)
>> +{
>> +    int ret;
>> +
>> +    ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCR,
>> +                TWL6030_REG_TOGGLE1);
>> +    if (ret)
>> +        dev_err(pdev, "error reseting GPADC (%d)!\n", ret);
>> +
>> +    return 0;
>> +};
>> +
>> +static int twl6030_gpadc_resume(struct device *pdev)
>> +{
>> +    int ret;
>> +
>> +    ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
>> +                TWL6030_REG_TOGGLE1);
>> +    if (ret)
>> +        dev_err(pdev, "error setting GPADC (%d)!\n", ret);
>> +
>> +    return 0;
>> +};
>> +
>> +static const struct dev_pm_ops twl6030_gpadc_pm_ops = {
>> +    .suspend = twl6030_gpadc_suspend,
>> +    .resume = twl6030_gpadc_resume,
>> +};
>
> SIMPLE_DEV_PM_OPS?
>
> [...]
> --
> To unsubscribe from this list: send the line "unsubscribe linux-omap" in
> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

^ permalink raw reply	[flat|nested] 45+ messages in thread

* [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 11:56       ` Grygorii Strashko
  0 siblings, 0 replies; 45+ messages in thread
From: Grygorii Strashko @ 2013-07-15 11:56 UTC (permalink / raw)
  To: linux-arm-kernel

Hi All,

I have a question regarding this patch and IIO in general
- Does IIO provide sync mechanism with system wide suspend/resume or 
this should be handled by each driver itself?

What if during system suspend iio_read_channel_raw() (or any other 
consumer API) will be called after gpadc driver have been suspended 
already? (I did some investigation and seems it's possible - Am I right?)

If no, could "info_exist_lock" be reused for such purposes?

Regards,
-grygorii


On 07/12/2013 10:56 PM, Lars-Peter Clausen wrote:
>
> A couple of comments inline.
>
> On 07/12/2013 09:18 AM, Oleksandr Kozaruk wrote:
>> diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig
>> index ab0767e6..87d699e 100644
>> --- a/drivers/iio/adc/Kconfig
>> +++ b/drivers/iio/adc/Kconfig
>> @@ -157,4 +157,12 @@ config VIPERBOARD_ADC
>>         Say yes here to access the ADC part of the Nano River
>>         Technologies Viperboard.
>>
>> +config TWL6030_GPADC
>> +    tristate "TWL6030 GPADC (General Purpose A/D Convertor) Support"
>> +    depends on TWL4030_CORE
>> +    default n
>> +    help
>> +      Say yes here if you want support for the TWL6030 General Purpose
>> +      A/D Convertor.
>> +
>
> Keep it in alphabetical order
>
>>   endmenu
>> diff --git a/drivers/iio/adc/Makefile b/drivers/iio/adc/Makefile
>> index 0a825be..8b05633 100644
>> --- a/drivers/iio/adc/Makefile
>> +++ b/drivers/iio/adc/Makefile
>> @@ -17,3 +17,4 @@ obj-$(CONFIG_MAX1363) += max1363.o
>>   obj-$(CONFIG_TI_ADC081C) += ti-adc081c.o
>>   obj-$(CONFIG_TI_AM335X_ADC) += ti_am335x_adc.o
>>   obj-$(CONFIG_VIPERBOARD_ADC) += viperboard_adc.o
>> +obj-$(CONFIG_TWL6030_GPADC) += twl6030-gpadc.o
>
> Same here.
>
>> diff --git a/drivers/iio/adc/twl6030-gpadc.c
>> b/drivers/iio/adc/twl6030-gpadc.c
>> new file mode 100644
>> index 0000000..6ceb789
>> --- /dev/null
>> +++ b/drivers/iio/adc/twl6030-gpadc.c
>> @@ -0,0 +1,1019 @@
> [...]
>> +static u8 twl6032_channel_to_reg(int channel)
>> +{
>> +    return TWL6032_GPADC_GPCH0_LSB;
>
> There is more than one channel, isn't there?
>
> [...]
>  > +static int twl6030_gpadc_read_raw(struct iio_dev *indio_dev,
>  > +                 const struct iio_chan_spec *chan,
>  > +                 int *val, int *val2, long mask)
>  > +{
>  > +    struct twl6030_gpadc_data *gpadc = iio_priv(indio_dev);
>  > +    int ret = -EINVAL;
>  > +
>  > +    mutex_lock(&gpadc->lock);
>  > +
>  > +    ret = gpadc->pdata->start_conversion(chan->channel);
>  > +    if (ret) {
>  > +        dev_err(gpadc->dev, "failed to start conversion\n");
>  > +        goto err;
>  > +    }
>  > +    /* wait for conversion to complete */
>  > +    wait_for_completion_interruptible_timeout(&gpadc->irq_complete,
>  > +                        msecs_to_jiffies(5000));
>
> wait_for_completion_interruptible_timeout() can return either a negative
> error code if it was interrupted, 0 if it timed out, or a positive value
> if it was successfully completed. You need to handle all three cases.
> Have a look at other existing drivers to see how to do this properly.
>
>  > +
>  > +    switch (mask) {
>  > +    case IIO_CHAN_INFO_RAW:
>  > +        ret = twl6030_gpadc_get_raw(gpadc, chan->channel, val);
>  > +        ret = ret ? -EIO : IIO_VAL_INT;
>  > +        break;
>  > +
>  > +    case IIO_CHAN_INFO_PROCESSED:
>  > +        ret = twl6030_gpadc_get_processed(gpadc, chan->channel, val);
>  > +        ret = ret ? -EIO : IIO_VAL_INT;
>  > +        break;
>  > +
>  > +    default:
>  > +        break;
>  > +    }
>  > +err:
>  > +    mutex_unlock(&gpadc->lock);
>  > +
>  > +    return ret;
>  > +}
>
>> +}
>> +
> [...]
>> +static int twl6030_gpadc_probe(struct platform_device *pdev)
>> +{
>> +    struct device *dev = &pdev->dev;
>> +    struct twl6030_gpadc_data *gpadc;
>> +    const struct twl6030_gpadc_platform_data *pdata;
>> +    const struct of_device_id *match;
>> +    struct iio_dev *indio_dev;
>> +    int irq;
>> +    int ret = 0;
>> +
>> +    match = of_match_device(of_match_ptr(of_twl6030_match_tbl), dev);
>> +    pdata = match ? match->data : dev->platform_data;
>> +
>> +    if (!pdata)
>> +        return -EINVAL;
>> +
>> +    indio_dev = iio_device_alloc(sizeof(struct twl6030_gpadc_data));
>
> sizeof(*gpadc)
>
>> +    if (!indio_dev) {
>> +        dev_err(dev, "failed allocating iio device\n");
>> +        ret = -ENOMEM;
>> +    }
>> +
>> +    gpadc = iio_priv(indio_dev);
>> +
>> +    gpadc->twl6030_cal_tbl = devm_kzalloc(dev,
>> +                    sizeof(struct twl6030_chnl_calib) *
>> +                    pdata->nchannels, GFP_KERNEL);
>> +    if (!gpadc->twl6030_cal_tbl)
>> +        goto err;
>> +
>> +    gpadc->dev = dev;
>> +    gpadc->pdata = pdata;
>> +
>> +    platform_set_drvdata(pdev, indio_dev);
>> +    mutex_init(&gpadc->lock);
>> +    init_completion(&gpadc->irq_complete);
>> +
>> +    ret = pdata->calibrate(gpadc);
>> +    if (ret < 0) {
>> +        dev_err(&pdev->dev, "failed to read calibration registers\n");
>> +        goto err;
>> +    }
>> +
>> +    irq = platform_get_irq(pdev, 0);
>> +    if (irq < 0) {
>> +        dev_err(&pdev->dev, "failed to get irq\n");
>> +        goto err;
>> +    }
>> +
>> +    ret = devm_request_threaded_irq(dev, irq, NULL,
>> +                    twl6030_gpadc_irq_handler,
>> +                    IRQF_ONESHOT, "twl6030_gpadc", gpadc);
>
> You access memory in the interrupt handler which is freed before the
> interrupt handler is freed.
>
>> +    if (ret) {
>> +        dev_dbg(&pdev->dev, "could not request irq\n");
>> +        goto err;
>> +    }
>> +
>> +    ret = twl6030_gpadc_enable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
>> +    if (ret < 0) {
>> +        dev_err(&pdev->dev, "failed to enable GPADC interrupt\n");
>> +        goto err;
>> +    }
>> +    ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
>> +                    TWL6030_REG_TOGGLE1);
>> +    if (ret < 0) {
>> +        dev_err(&pdev->dev, "failed to enable GPADC module\n");
>> +        goto err;
>> +    }
>> +
>> +    indio_dev->name = DRIVER_NAME;
>> +    indio_dev->dev.parent = dev;
>> +    indio_dev->info = &twl6030_gpadc_iio_info;
>> +    indio_dev->modes = INDIO_DIRECT_MODE;
>> +    indio_dev->channels = pdata->iio_channels;
>> +    indio_dev->num_channels = pdata->nchannels;
>> +
>> +    ret = iio_device_register(indio_dev);
>> +    if (ret)
>> +        goto err;
>> +
>> +    return ret;
>> +err:
>> +    iio_device_free(indio_dev);
>> +    return ret;
>> +}
>> +
> [...]
>> +static int twl6030_gpadc_suspend(struct device *pdev)
>> +{
>> +    int ret;
>> +
>> +    ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCR,
>> +                TWL6030_REG_TOGGLE1);
>> +    if (ret)
>> +        dev_err(pdev, "error reseting GPADC (%d)!\n", ret);
>> +
>> +    return 0;
>> +};
>> +
>> +static int twl6030_gpadc_resume(struct device *pdev)
>> +{
>> +    int ret;
>> +
>> +    ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
>> +                TWL6030_REG_TOGGLE1);
>> +    if (ret)
>> +        dev_err(pdev, "error setting GPADC (%d)!\n", ret);
>> +
>> +    return 0;
>> +};
>> +
>> +static const struct dev_pm_ops twl6030_gpadc_pm_ops = {
>> +    .suspend = twl6030_gpadc_suspend,
>> +    .resume = twl6030_gpadc_resume,
>> +};
>
> SIMPLE_DEV_PM_OPS?
>
> [...]
> --
> To unsubscribe from this list: send the line "unsubscribe linux-omap" in
> the body of a message to majordomo at vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 12:33         ` Lars-Peter Clausen
  0 siblings, 0 replies; 45+ messages in thread
From: Lars-Peter Clausen @ 2013-07-15 12:33 UTC (permalink / raw)
  To: Grygorii Strashko
  Cc: Oleksandr Kozaruk, tony, benoit.cousson, rnayak, peter.ujfalusi,
	kishon, jic23, grant.likely, rob.herring, sameo, ch.naveen,
	poeschel, milo.kim, balajitk, gg, linux-arm-kernel, linux-kernel,
	linux-iio, devicetree-discuss, linux-omap

On 07/15/2013 01:56 PM, Grygorii Strashko wrote:
> Hi All,
> 
> I have a question regarding this patch and IIO in general
> - Does IIO provide sync mechanism with system wide suspend/resume or this
> should be handled by each driver itself?
> 
> What if during system suspend iio_read_channel_raw() (or any other consumer
> API) will be called after gpadc driver have been suspended already? (I did
> some investigation and seems it's possible - Am I right?)
> 
> If no, could "info_exist_lock" be reused for such purposes?

I think you can use the mlock for this purpose. Usually you'd have a flag in
your driver struct which you'd set to true in suspend and to false in
resume. And in the read_raw callback you'd check that flag before accessing
the hardware. If it turns out that this is a common pattern it probably
makes sense to add infrastructure for this to the IIO core. Something along
the lines of:

static int drv_suspend(...)
{
	...
	iio_device_set_suspended(iio_dev, true);
	...
}

static int drv_suspend(...)
{
	...
	iio_device_set_suspended(iio_dev, false);
	...
}

and then have the IIO core check that flag before calling the read_raw callback.

- Lars

> 
> Regards,
> -grygorii
> 
> 
> On 07/12/2013 10:56 PM, Lars-Peter Clausen wrote:
>>
>> A couple of comments inline.
>>
>> On 07/12/2013 09:18 AM, Oleksandr Kozaruk wrote:
>>> diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig
>>> index ab0767e6..87d699e 100644
>>> --- a/drivers/iio/adc/Kconfig
>>> +++ b/drivers/iio/adc/Kconfig
>>> @@ -157,4 +157,12 @@ config VIPERBOARD_ADC
>>>         Say yes here to access the ADC part of the Nano River
>>>         Technologies Viperboard.
>>>
>>> +config TWL6030_GPADC
>>> +    tristate "TWL6030 GPADC (General Purpose A/D Convertor) Support"
>>> +    depends on TWL4030_CORE
>>> +    default n
>>> +    help
>>> +      Say yes here if you want support for the TWL6030 General Purpose
>>> +      A/D Convertor.
>>> +
>>
>> Keep it in alphabetical order
>>
>>>   endmenu
>>> diff --git a/drivers/iio/adc/Makefile b/drivers/iio/adc/Makefile
>>> index 0a825be..8b05633 100644
>>> --- a/drivers/iio/adc/Makefile
>>> +++ b/drivers/iio/adc/Makefile
>>> @@ -17,3 +17,4 @@ obj-$(CONFIG_MAX1363) += max1363.o
>>>   obj-$(CONFIG_TI_ADC081C) += ti-adc081c.o
>>>   obj-$(CONFIG_TI_AM335X_ADC) += ti_am335x_adc.o
>>>   obj-$(CONFIG_VIPERBOARD_ADC) += viperboard_adc.o
>>> +obj-$(CONFIG_TWL6030_GPADC) += twl6030-gpadc.o
>>
>> Same here.
>>
>>> diff --git a/drivers/iio/adc/twl6030-gpadc.c
>>> b/drivers/iio/adc/twl6030-gpadc.c
>>> new file mode 100644
>>> index 0000000..6ceb789
>>> --- /dev/null
>>> +++ b/drivers/iio/adc/twl6030-gpadc.c
>>> @@ -0,0 +1,1019 @@
>> [...]
>>> +static u8 twl6032_channel_to_reg(int channel)
>>> +{
>>> +    return TWL6032_GPADC_GPCH0_LSB;
>>
>> There is more than one channel, isn't there?
>>
>> [...]
>>  > +static int twl6030_gpadc_read_raw(struct iio_dev *indio_dev,
>>  > +                 const struct iio_chan_spec *chan,
>>  > +                 int *val, int *val2, long mask)
>>  > +{
>>  > +    struct twl6030_gpadc_data *gpadc = iio_priv(indio_dev);
>>  > +    int ret = -EINVAL;
>>  > +
>>  > +    mutex_lock(&gpadc->lock);
>>  > +
>>  > +    ret = gpadc->pdata->start_conversion(chan->channel);
>>  > +    if (ret) {
>>  > +        dev_err(gpadc->dev, "failed to start conversion\n");
>>  > +        goto err;
>>  > +    }
>>  > +    /* wait for conversion to complete */
>>  > +    wait_for_completion_interruptible_timeout(&gpadc->irq_complete,
>>  > +                        msecs_to_jiffies(5000));
>>
>> wait_for_completion_interruptible_timeout() can return either a negative
>> error code if it was interrupted, 0 if it timed out, or a positive value
>> if it was successfully completed. You need to handle all three cases.
>> Have a look at other existing drivers to see how to do this properly.
>>
>>  > +
>>  > +    switch (mask) {
>>  > +    case IIO_CHAN_INFO_RAW:
>>  > +        ret = twl6030_gpadc_get_raw(gpadc, chan->channel, val);
>>  > +        ret = ret ? -EIO : IIO_VAL_INT;
>>  > +        break;
>>  > +
>>  > +    case IIO_CHAN_INFO_PROCESSED:
>>  > +        ret = twl6030_gpadc_get_processed(gpadc, chan->channel, val);
>>  > +        ret = ret ? -EIO : IIO_VAL_INT;
>>  > +        break;
>>  > +
>>  > +    default:
>>  > +        break;
>>  > +    }
>>  > +err:
>>  > +    mutex_unlock(&gpadc->lock);
>>  > +
>>  > +    return ret;
>>  > +}
>>
>>> +}
>>> +
>> [...]
>>> +static int twl6030_gpadc_probe(struct platform_device *pdev)
>>> +{
>>> +    struct device *dev = &pdev->dev;
>>> +    struct twl6030_gpadc_data *gpadc;
>>> +    const struct twl6030_gpadc_platform_data *pdata;
>>> +    const struct of_device_id *match;
>>> +    struct iio_dev *indio_dev;
>>> +    int irq;
>>> +    int ret = 0;
>>> +
>>> +    match = of_match_device(of_match_ptr(of_twl6030_match_tbl), dev);
>>> +    pdata = match ? match->data : dev->platform_data;
>>> +
>>> +    if (!pdata)
>>> +        return -EINVAL;
>>> +
>>> +    indio_dev = iio_device_alloc(sizeof(struct twl6030_gpadc_data));
>>
>> sizeof(*gpadc)
>>
>>> +    if (!indio_dev) {
>>> +        dev_err(dev, "failed allocating iio device\n");
>>> +        ret = -ENOMEM;
>>> +    }
>>> +
>>> +    gpadc = iio_priv(indio_dev);
>>> +
>>> +    gpadc->twl6030_cal_tbl = devm_kzalloc(dev,
>>> +                    sizeof(struct twl6030_chnl_calib) *
>>> +                    pdata->nchannels, GFP_KERNEL);
>>> +    if (!gpadc->twl6030_cal_tbl)
>>> +        goto err;
>>> +
>>> +    gpadc->dev = dev;
>>> +    gpadc->pdata = pdata;
>>> +
>>> +    platform_set_drvdata(pdev, indio_dev);
>>> +    mutex_init(&gpadc->lock);
>>> +    init_completion(&gpadc->irq_complete);
>>> +
>>> +    ret = pdata->calibrate(gpadc);
>>> +    if (ret < 0) {
>>> +        dev_err(&pdev->dev, "failed to read calibration registers\n");
>>> +        goto err;
>>> +    }
>>> +
>>> +    irq = platform_get_irq(pdev, 0);
>>> +    if (irq < 0) {
>>> +        dev_err(&pdev->dev, "failed to get irq\n");
>>> +        goto err;
>>> +    }
>>> +
>>> +    ret = devm_request_threaded_irq(dev, irq, NULL,
>>> +                    twl6030_gpadc_irq_handler,
>>> +                    IRQF_ONESHOT, "twl6030_gpadc", gpadc);
>>
>> You access memory in the interrupt handler which is freed before the
>> interrupt handler is freed.
>>
>>> +    if (ret) {
>>> +        dev_dbg(&pdev->dev, "could not request irq\n");
>>> +        goto err;
>>> +    }
>>> +
>>> +    ret = twl6030_gpadc_enable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
>>> +    if (ret < 0) {
>>> +        dev_err(&pdev->dev, "failed to enable GPADC interrupt\n");
>>> +        goto err;
>>> +    }
>>> +    ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
>>> +                    TWL6030_REG_TOGGLE1);
>>> +    if (ret < 0) {
>>> +        dev_err(&pdev->dev, "failed to enable GPADC module\n");
>>> +        goto err;
>>> +    }
>>> +
>>> +    indio_dev->name = DRIVER_NAME;
>>> +    indio_dev->dev.parent = dev;
>>> +    indio_dev->info = &twl6030_gpadc_iio_info;
>>> +    indio_dev->modes = INDIO_DIRECT_MODE;
>>> +    indio_dev->channels = pdata->iio_channels;
>>> +    indio_dev->num_channels = pdata->nchannels;
>>> +
>>> +    ret = iio_device_register(indio_dev);
>>> +    if (ret)
>>> +        goto err;
>>> +
>>> +    return ret;
>>> +err:
>>> +    iio_device_free(indio_dev);
>>> +    return ret;
>>> +}
>>> +
>> [...]
>>> +static int twl6030_gpadc_suspend(struct device *pdev)
>>> +{
>>> +    int ret;
>>> +
>>> +    ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCR,
>>> +                TWL6030_REG_TOGGLE1);
>>> +    if (ret)
>>> +        dev_err(pdev, "error reseting GPADC (%d)!\n", ret);
>>> +
>>> +    return 0;
>>> +};
>>> +
>>> +static int twl6030_gpadc_resume(struct device *pdev)
>>> +{
>>> +    int ret;
>>> +
>>> +    ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
>>> +                TWL6030_REG_TOGGLE1);
>>> +    if (ret)
>>> +        dev_err(pdev, "error setting GPADC (%d)!\n", ret);
>>> +
>>> +    return 0;
>>> +};
>>> +
>>> +static const struct dev_pm_ops twl6030_gpadc_pm_ops = {
>>> +    .suspend = twl6030_gpadc_suspend,
>>> +    .resume = twl6030_gpadc_resume,
>>> +};
>>
>> SIMPLE_DEV_PM_OPS?
>>
>> [...]
>> -- 
>> To unsubscribe from this list: send the line "unsubscribe linux-omap" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 


^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 12:33         ` Lars-Peter Clausen
  0 siblings, 0 replies; 45+ messages in thread
From: Lars-Peter Clausen @ 2013-07-15 12:33 UTC (permalink / raw)
  To: Grygorii Strashko
  Cc: Oleksandr Kozaruk, tony-4v6yS6AI5VpBDgjK7y7TUQ,
	benoit.cousson-QSEj5FYQhm4dnm+yROfE0A, rnayak-l0cyMroinI0,
	peter.ujfalusi-l0cyMroinI0, kishon-l0cyMroinI0,
	jic23-KWPb1pKIrIJaa/9Udqfwiw,
	grant.likely-QSEj5FYQhm4dnm+yROfE0A,
	rob.herring-bsGFqQB8/DxBDgjK7y7TUQ, sameo-VuQAYsv1563Yd54FQh9/CA,
	ch.naveen-Sze3O3UU22JBDgjK7y7TUQ,
	poeschel-Xtl8qvBWbHwb1SvskN2V4Q, milo.kim-l0cyMroinI0,
	balajitk-l0cyMroinI0, gg-kDsPt+C1G03kYMGBc/C6ZA,
	linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA,
	linux-iio-u79uwXL29TY76Z2rM5mHXA,
	devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ,
	linux-omap-u79uwXL29TY76Z2rM5mHXA

On 07/15/2013 01:56 PM, Grygorii Strashko wrote:
> Hi All,
> 
> I have a question regarding this patch and IIO in general
> - Does IIO provide sync mechanism with system wide suspend/resume or this
> should be handled by each driver itself?
> 
> What if during system suspend iio_read_channel_raw() (or any other consumer
> API) will be called after gpadc driver have been suspended already? (I did
> some investigation and seems it's possible - Am I right?)
> 
> If no, could "info_exist_lock" be reused for such purposes?

I think you can use the mlock for this purpose. Usually you'd have a flag in
your driver struct which you'd set to true in suspend and to false in
resume. And in the read_raw callback you'd check that flag before accessing
the hardware. If it turns out that this is a common pattern it probably
makes sense to add infrastructure for this to the IIO core. Something along
the lines of:

static int drv_suspend(...)
{
	...
	iio_device_set_suspended(iio_dev, true);
	...
}

static int drv_suspend(...)
{
	...
	iio_device_set_suspended(iio_dev, false);
	...
}

and then have the IIO core check that flag before calling the read_raw callback.

- Lars

> 
> Regards,
> -grygorii
> 
> 
> On 07/12/2013 10:56 PM, Lars-Peter Clausen wrote:
>>
>> A couple of comments inline.
>>
>> On 07/12/2013 09:18 AM, Oleksandr Kozaruk wrote:
>>> diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig
>>> index ab0767e6..87d699e 100644
>>> --- a/drivers/iio/adc/Kconfig
>>> +++ b/drivers/iio/adc/Kconfig
>>> @@ -157,4 +157,12 @@ config VIPERBOARD_ADC
>>>         Say yes here to access the ADC part of the Nano River
>>>         Technologies Viperboard.
>>>
>>> +config TWL6030_GPADC
>>> +    tristate "TWL6030 GPADC (General Purpose A/D Convertor) Support"
>>> +    depends on TWL4030_CORE
>>> +    default n
>>> +    help
>>> +      Say yes here if you want support for the TWL6030 General Purpose
>>> +      A/D Convertor.
>>> +
>>
>> Keep it in alphabetical order
>>
>>>   endmenu
>>> diff --git a/drivers/iio/adc/Makefile b/drivers/iio/adc/Makefile
>>> index 0a825be..8b05633 100644
>>> --- a/drivers/iio/adc/Makefile
>>> +++ b/drivers/iio/adc/Makefile
>>> @@ -17,3 +17,4 @@ obj-$(CONFIG_MAX1363) += max1363.o
>>>   obj-$(CONFIG_TI_ADC081C) += ti-adc081c.o
>>>   obj-$(CONFIG_TI_AM335X_ADC) += ti_am335x_adc.o
>>>   obj-$(CONFIG_VIPERBOARD_ADC) += viperboard_adc.o
>>> +obj-$(CONFIG_TWL6030_GPADC) += twl6030-gpadc.o
>>
>> Same here.
>>
>>> diff --git a/drivers/iio/adc/twl6030-gpadc.c
>>> b/drivers/iio/adc/twl6030-gpadc.c
>>> new file mode 100644
>>> index 0000000..6ceb789
>>> --- /dev/null
>>> +++ b/drivers/iio/adc/twl6030-gpadc.c
>>> @@ -0,0 +1,1019 @@
>> [...]
>>> +static u8 twl6032_channel_to_reg(int channel)
>>> +{
>>> +    return TWL6032_GPADC_GPCH0_LSB;
>>
>> There is more than one channel, isn't there?
>>
>> [...]
>>  > +static int twl6030_gpadc_read_raw(struct iio_dev *indio_dev,
>>  > +                 const struct iio_chan_spec *chan,
>>  > +                 int *val, int *val2, long mask)
>>  > +{
>>  > +    struct twl6030_gpadc_data *gpadc = iio_priv(indio_dev);
>>  > +    int ret = -EINVAL;
>>  > +
>>  > +    mutex_lock(&gpadc->lock);
>>  > +
>>  > +    ret = gpadc->pdata->start_conversion(chan->channel);
>>  > +    if (ret) {
>>  > +        dev_err(gpadc->dev, "failed to start conversion\n");
>>  > +        goto err;
>>  > +    }
>>  > +    /* wait for conversion to complete */
>>  > +    wait_for_completion_interruptible_timeout(&gpadc->irq_complete,
>>  > +                        msecs_to_jiffies(5000));
>>
>> wait_for_completion_interruptible_timeout() can return either a negative
>> error code if it was interrupted, 0 if it timed out, or a positive value
>> if it was successfully completed. You need to handle all three cases.
>> Have a look at other existing drivers to see how to do this properly.
>>
>>  > +
>>  > +    switch (mask) {
>>  > +    case IIO_CHAN_INFO_RAW:
>>  > +        ret = twl6030_gpadc_get_raw(gpadc, chan->channel, val);
>>  > +        ret = ret ? -EIO : IIO_VAL_INT;
>>  > +        break;
>>  > +
>>  > +    case IIO_CHAN_INFO_PROCESSED:
>>  > +        ret = twl6030_gpadc_get_processed(gpadc, chan->channel, val);
>>  > +        ret = ret ? -EIO : IIO_VAL_INT;
>>  > +        break;
>>  > +
>>  > +    default:
>>  > +        break;
>>  > +    }
>>  > +err:
>>  > +    mutex_unlock(&gpadc->lock);
>>  > +
>>  > +    return ret;
>>  > +}
>>
>>> +}
>>> +
>> [...]
>>> +static int twl6030_gpadc_probe(struct platform_device *pdev)
>>> +{
>>> +    struct device *dev = &pdev->dev;
>>> +    struct twl6030_gpadc_data *gpadc;
>>> +    const struct twl6030_gpadc_platform_data *pdata;
>>> +    const struct of_device_id *match;
>>> +    struct iio_dev *indio_dev;
>>> +    int irq;
>>> +    int ret = 0;
>>> +
>>> +    match = of_match_device(of_match_ptr(of_twl6030_match_tbl), dev);
>>> +    pdata = match ? match->data : dev->platform_data;
>>> +
>>> +    if (!pdata)
>>> +        return -EINVAL;
>>> +
>>> +    indio_dev = iio_device_alloc(sizeof(struct twl6030_gpadc_data));
>>
>> sizeof(*gpadc)
>>
>>> +    if (!indio_dev) {
>>> +        dev_err(dev, "failed allocating iio device\n");
>>> +        ret = -ENOMEM;
>>> +    }
>>> +
>>> +    gpadc = iio_priv(indio_dev);
>>> +
>>> +    gpadc->twl6030_cal_tbl = devm_kzalloc(dev,
>>> +                    sizeof(struct twl6030_chnl_calib) *
>>> +                    pdata->nchannels, GFP_KERNEL);
>>> +    if (!gpadc->twl6030_cal_tbl)
>>> +        goto err;
>>> +
>>> +    gpadc->dev = dev;
>>> +    gpadc->pdata = pdata;
>>> +
>>> +    platform_set_drvdata(pdev, indio_dev);
>>> +    mutex_init(&gpadc->lock);
>>> +    init_completion(&gpadc->irq_complete);
>>> +
>>> +    ret = pdata->calibrate(gpadc);
>>> +    if (ret < 0) {
>>> +        dev_err(&pdev->dev, "failed to read calibration registers\n");
>>> +        goto err;
>>> +    }
>>> +
>>> +    irq = platform_get_irq(pdev, 0);
>>> +    if (irq < 0) {
>>> +        dev_err(&pdev->dev, "failed to get irq\n");
>>> +        goto err;
>>> +    }
>>> +
>>> +    ret = devm_request_threaded_irq(dev, irq, NULL,
>>> +                    twl6030_gpadc_irq_handler,
>>> +                    IRQF_ONESHOT, "twl6030_gpadc", gpadc);
>>
>> You access memory in the interrupt handler which is freed before the
>> interrupt handler is freed.
>>
>>> +    if (ret) {
>>> +        dev_dbg(&pdev->dev, "could not request irq\n");
>>> +        goto err;
>>> +    }
>>> +
>>> +    ret = twl6030_gpadc_enable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
>>> +    if (ret < 0) {
>>> +        dev_err(&pdev->dev, "failed to enable GPADC interrupt\n");
>>> +        goto err;
>>> +    }
>>> +    ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
>>> +                    TWL6030_REG_TOGGLE1);
>>> +    if (ret < 0) {
>>> +        dev_err(&pdev->dev, "failed to enable GPADC module\n");
>>> +        goto err;
>>> +    }
>>> +
>>> +    indio_dev->name = DRIVER_NAME;
>>> +    indio_dev->dev.parent = dev;
>>> +    indio_dev->info = &twl6030_gpadc_iio_info;
>>> +    indio_dev->modes = INDIO_DIRECT_MODE;
>>> +    indio_dev->channels = pdata->iio_channels;
>>> +    indio_dev->num_channels = pdata->nchannels;
>>> +
>>> +    ret = iio_device_register(indio_dev);
>>> +    if (ret)
>>> +        goto err;
>>> +
>>> +    return ret;
>>> +err:
>>> +    iio_device_free(indio_dev);
>>> +    return ret;
>>> +}
>>> +
>> [...]
>>> +static int twl6030_gpadc_suspend(struct device *pdev)
>>> +{
>>> +    int ret;
>>> +
>>> +    ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCR,
>>> +                TWL6030_REG_TOGGLE1);
>>> +    if (ret)
>>> +        dev_err(pdev, "error reseting GPADC (%d)!\n", ret);
>>> +
>>> +    return 0;
>>> +};
>>> +
>>> +static int twl6030_gpadc_resume(struct device *pdev)
>>> +{
>>> +    int ret;
>>> +
>>> +    ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
>>> +                TWL6030_REG_TOGGLE1);
>>> +    if (ret)
>>> +        dev_err(pdev, "error setting GPADC (%d)!\n", ret);
>>> +
>>> +    return 0;
>>> +};
>>> +
>>> +static const struct dev_pm_ops twl6030_gpadc_pm_ops = {
>>> +    .suspend = twl6030_gpadc_suspend,
>>> +    .resume = twl6030_gpadc_resume,
>>> +};
>>
>> SIMPLE_DEV_PM_OPS?
>>
>> [...]
>> -- 
>> To unsubscribe from this list: send the line "unsubscribe linux-omap" in
>> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 

^ permalink raw reply	[flat|nested] 45+ messages in thread

* [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 12:33         ` Lars-Peter Clausen
  0 siblings, 0 replies; 45+ messages in thread
From: Lars-Peter Clausen @ 2013-07-15 12:33 UTC (permalink / raw)
  To: linux-arm-kernel

On 07/15/2013 01:56 PM, Grygorii Strashko wrote:
> Hi All,
> 
> I have a question regarding this patch and IIO in general
> - Does IIO provide sync mechanism with system wide suspend/resume or this
> should be handled by each driver itself?
> 
> What if during system suspend iio_read_channel_raw() (or any other consumer
> API) will be called after gpadc driver have been suspended already? (I did
> some investigation and seems it's possible - Am I right?)
> 
> If no, could "info_exist_lock" be reused for such purposes?

I think you can use the mlock for this purpose. Usually you'd have a flag in
your driver struct which you'd set to true in suspend and to false in
resume. And in the read_raw callback you'd check that flag before accessing
the hardware. If it turns out that this is a common pattern it probably
makes sense to add infrastructure for this to the IIO core. Something along
the lines of:

static int drv_suspend(...)
{
	...
	iio_device_set_suspended(iio_dev, true);
	...
}

static int drv_suspend(...)
{
	...
	iio_device_set_suspended(iio_dev, false);
	...
}

and then have the IIO core check that flag before calling the read_raw callback.

- Lars

> 
> Regards,
> -grygorii
> 
> 
> On 07/12/2013 10:56 PM, Lars-Peter Clausen wrote:
>>
>> A couple of comments inline.
>>
>> On 07/12/2013 09:18 AM, Oleksandr Kozaruk wrote:
>>> diff --git a/drivers/iio/adc/Kconfig b/drivers/iio/adc/Kconfig
>>> index ab0767e6..87d699e 100644
>>> --- a/drivers/iio/adc/Kconfig
>>> +++ b/drivers/iio/adc/Kconfig
>>> @@ -157,4 +157,12 @@ config VIPERBOARD_ADC
>>>         Say yes here to access the ADC part of the Nano River
>>>         Technologies Viperboard.
>>>
>>> +config TWL6030_GPADC
>>> +    tristate "TWL6030 GPADC (General Purpose A/D Convertor) Support"
>>> +    depends on TWL4030_CORE
>>> +    default n
>>> +    help
>>> +      Say yes here if you want support for the TWL6030 General Purpose
>>> +      A/D Convertor.
>>> +
>>
>> Keep it in alphabetical order
>>
>>>   endmenu
>>> diff --git a/drivers/iio/adc/Makefile b/drivers/iio/adc/Makefile
>>> index 0a825be..8b05633 100644
>>> --- a/drivers/iio/adc/Makefile
>>> +++ b/drivers/iio/adc/Makefile
>>> @@ -17,3 +17,4 @@ obj-$(CONFIG_MAX1363) += max1363.o
>>>   obj-$(CONFIG_TI_ADC081C) += ti-adc081c.o
>>>   obj-$(CONFIG_TI_AM335X_ADC) += ti_am335x_adc.o
>>>   obj-$(CONFIG_VIPERBOARD_ADC) += viperboard_adc.o
>>> +obj-$(CONFIG_TWL6030_GPADC) += twl6030-gpadc.o
>>
>> Same here.
>>
>>> diff --git a/drivers/iio/adc/twl6030-gpadc.c
>>> b/drivers/iio/adc/twl6030-gpadc.c
>>> new file mode 100644
>>> index 0000000..6ceb789
>>> --- /dev/null
>>> +++ b/drivers/iio/adc/twl6030-gpadc.c
>>> @@ -0,0 +1,1019 @@
>> [...]
>>> +static u8 twl6032_channel_to_reg(int channel)
>>> +{
>>> +    return TWL6032_GPADC_GPCH0_LSB;
>>
>> There is more than one channel, isn't there?
>>
>> [...]
>>  > +static int twl6030_gpadc_read_raw(struct iio_dev *indio_dev,
>>  > +                 const struct iio_chan_spec *chan,
>>  > +                 int *val, int *val2, long mask)
>>  > +{
>>  > +    struct twl6030_gpadc_data *gpadc = iio_priv(indio_dev);
>>  > +    int ret = -EINVAL;
>>  > +
>>  > +    mutex_lock(&gpadc->lock);
>>  > +
>>  > +    ret = gpadc->pdata->start_conversion(chan->channel);
>>  > +    if (ret) {
>>  > +        dev_err(gpadc->dev, "failed to start conversion\n");
>>  > +        goto err;
>>  > +    }
>>  > +    /* wait for conversion to complete */
>>  > +    wait_for_completion_interruptible_timeout(&gpadc->irq_complete,
>>  > +                        msecs_to_jiffies(5000));
>>
>> wait_for_completion_interruptible_timeout() can return either a negative
>> error code if it was interrupted, 0 if it timed out, or a positive value
>> if it was successfully completed. You need to handle all three cases.
>> Have a look at other existing drivers to see how to do this properly.
>>
>>  > +
>>  > +    switch (mask) {
>>  > +    case IIO_CHAN_INFO_RAW:
>>  > +        ret = twl6030_gpadc_get_raw(gpadc, chan->channel, val);
>>  > +        ret = ret ? -EIO : IIO_VAL_INT;
>>  > +        break;
>>  > +
>>  > +    case IIO_CHAN_INFO_PROCESSED:
>>  > +        ret = twl6030_gpadc_get_processed(gpadc, chan->channel, val);
>>  > +        ret = ret ? -EIO : IIO_VAL_INT;
>>  > +        break;
>>  > +
>>  > +    default:
>>  > +        break;
>>  > +    }
>>  > +err:
>>  > +    mutex_unlock(&gpadc->lock);
>>  > +
>>  > +    return ret;
>>  > +}
>>
>>> +}
>>> +
>> [...]
>>> +static int twl6030_gpadc_probe(struct platform_device *pdev)
>>> +{
>>> +    struct device *dev = &pdev->dev;
>>> +    struct twl6030_gpadc_data *gpadc;
>>> +    const struct twl6030_gpadc_platform_data *pdata;
>>> +    const struct of_device_id *match;
>>> +    struct iio_dev *indio_dev;
>>> +    int irq;
>>> +    int ret = 0;
>>> +
>>> +    match = of_match_device(of_match_ptr(of_twl6030_match_tbl), dev);
>>> +    pdata = match ? match->data : dev->platform_data;
>>> +
>>> +    if (!pdata)
>>> +        return -EINVAL;
>>> +
>>> +    indio_dev = iio_device_alloc(sizeof(struct twl6030_gpadc_data));
>>
>> sizeof(*gpadc)
>>
>>> +    if (!indio_dev) {
>>> +        dev_err(dev, "failed allocating iio device\n");
>>> +        ret = -ENOMEM;
>>> +    }
>>> +
>>> +    gpadc = iio_priv(indio_dev);
>>> +
>>> +    gpadc->twl6030_cal_tbl = devm_kzalloc(dev,
>>> +                    sizeof(struct twl6030_chnl_calib) *
>>> +                    pdata->nchannels, GFP_KERNEL);
>>> +    if (!gpadc->twl6030_cal_tbl)
>>> +        goto err;
>>> +
>>> +    gpadc->dev = dev;
>>> +    gpadc->pdata = pdata;
>>> +
>>> +    platform_set_drvdata(pdev, indio_dev);
>>> +    mutex_init(&gpadc->lock);
>>> +    init_completion(&gpadc->irq_complete);
>>> +
>>> +    ret = pdata->calibrate(gpadc);
>>> +    if (ret < 0) {
>>> +        dev_err(&pdev->dev, "failed to read calibration registers\n");
>>> +        goto err;
>>> +    }
>>> +
>>> +    irq = platform_get_irq(pdev, 0);
>>> +    if (irq < 0) {
>>> +        dev_err(&pdev->dev, "failed to get irq\n");
>>> +        goto err;
>>> +    }
>>> +
>>> +    ret = devm_request_threaded_irq(dev, irq, NULL,
>>> +                    twl6030_gpadc_irq_handler,
>>> +                    IRQF_ONESHOT, "twl6030_gpadc", gpadc);
>>
>> You access memory in the interrupt handler which is freed before the
>> interrupt handler is freed.
>>
>>> +    if (ret) {
>>> +        dev_dbg(&pdev->dev, "could not request irq\n");
>>> +        goto err;
>>> +    }
>>> +
>>> +    ret = twl6030_gpadc_enable_irq(TWL6030_GPADC_RT_SW1_EOC_MASK);
>>> +    if (ret < 0) {
>>> +        dev_err(&pdev->dev, "failed to enable GPADC interrupt\n");
>>> +        goto err;
>>> +    }
>>> +    ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
>>> +                    TWL6030_REG_TOGGLE1);
>>> +    if (ret < 0) {
>>> +        dev_err(&pdev->dev, "failed to enable GPADC module\n");
>>> +        goto err;
>>> +    }
>>> +
>>> +    indio_dev->name = DRIVER_NAME;
>>> +    indio_dev->dev.parent = dev;
>>> +    indio_dev->info = &twl6030_gpadc_iio_info;
>>> +    indio_dev->modes = INDIO_DIRECT_MODE;
>>> +    indio_dev->channels = pdata->iio_channels;
>>> +    indio_dev->num_channels = pdata->nchannels;
>>> +
>>> +    ret = iio_device_register(indio_dev);
>>> +    if (ret)
>>> +        goto err;
>>> +
>>> +    return ret;
>>> +err:
>>> +    iio_device_free(indio_dev);
>>> +    return ret;
>>> +}
>>> +
>> [...]
>>> +static int twl6030_gpadc_suspend(struct device *pdev)
>>> +{
>>> +    int ret;
>>> +
>>> +    ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCR,
>>> +                TWL6030_REG_TOGGLE1);
>>> +    if (ret)
>>> +        dev_err(pdev, "error reseting GPADC (%d)!\n", ret);
>>> +
>>> +    return 0;
>>> +};
>>> +
>>> +static int twl6030_gpadc_resume(struct device *pdev)
>>> +{
>>> +    int ret;
>>> +
>>> +    ret = twl_i2c_write_u8(TWL6030_MODULE_ID1, TWL6030_GPADCS,
>>> +                TWL6030_REG_TOGGLE1);
>>> +    if (ret)
>>> +        dev_err(pdev, "error setting GPADC (%d)!\n", ret);
>>> +
>>> +    return 0;
>>> +};
>>> +
>>> +static const struct dev_pm_ops twl6030_gpadc_pm_ops = {
>>> +    .suspend = twl6030_gpadc_suspend,
>>> +    .resume = twl6030_gpadc_resume,
>>> +};
>>
>> SIMPLE_DEV_PM_OPS?
>>
>> [...]
>> -- 
>> To unsubscribe from this list: send the line "unsubscribe linux-omap" in
>> the body of a message to majordomo at vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 

^ permalink raw reply	[flat|nested] 45+ messages in thread

* RE: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 13:30       ` Kozaruk, Oleksandr
  0 siblings, 0 replies; 45+ messages in thread
From: Kozaruk, Oleksandr @ 2013-07-15 13:30 UTC (permalink / raw)
  To: Jonathan Cameron
  Cc: tony, benoit.cousson, Nayak, Rajendra, Ujfalusi, Peter, ABRAHAM,
	KISHON VIJAY, jic23, grant.likely, rob.herring, lars, sameo,
	ch.naveen, poeschel, Kim, Milo, Krishnamoorthy, Balaji T, gg,
	linux-arm-kernel, linux-kernel, linux-iio, devicetree-discuss,
	linux-omap

Hello Jonathan,
Thanks for the review.

>Couple of things:
>
>1) It looks from the driver that a lot of the channels are not measuring
>voltages but rather temperature or currents etc.  If so then their
>types in the channel mask should be appropriately set.  Also if some
>of the channels are entirely internal test networks, what is the benefit
>of exposing them to userspace as readable channels?
>If channels are merely 'suggested' as being used for temperatures etc
>then it is fine to keep them as voltages.
There are two kinds of channel for measuring temperature: die temperature
and those that measure temperature indirectly measuring voltage on resistive
load(NTC sensor).
die temperature is calculated by some formulas which convert ADC code to
temperature. This is not implemented in the driver.
Channels intended to measure temperature with NTC sensor have inbuilt current
sources. Voltage measured by this channels and specific NTC could be converted
to temperature.
This is the reason they chosen to be voltage channels.
As for test network I'll remove them from exposing to userspace.
[...]

>> +static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,
>> +             int channel, int *res)
>> +{
>> +     u8 reg = gpadc->pdata->channel_to_reg(channel);
>> +     u8 val[2];
>
>le16 val;
>> +     int ret;
>> +
>ret = twl6030_gpadc_read(val, reg);
>(note that (reg, val) ordering of parameters would be a more common choice)
>
>> +     ret = twl6030_gpadc_read(val, reg);
>> +     if (ret) {
>> +             dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);
>> +             return ret;
>> +     }
>> +
>res = le16_to_cpup(val);
>
>> +     *res = (val[1] << 8) | val[0];
>> +
>> +     return ret;
>> +}
>> +
You mean something like this:
static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,                                                                                                                                                  
						int channel, int *res)                                                                                                                                                                              
{                                                                                                                                                                                                                   
	u8 reg = gpadc->pdata->channel_to_reg(channel);                                                                                                                                                             
	__le16 val;                                                                                                                                                                                                 
	int ret;                                                                                                                                                                                                    

	ret = twl6030_gpadc_read(reg, (u8 *)&val);                                                                                                                                                                  
	if (ret) {                                                                                                                                                                                                  
		dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);                                                                                                                                         
		return ret;                                                                                                                                                                                         
	}                                                                                                                                                                                                           

	*res = le16_to_cpup(&val);                                                                                                                                                                                  

	return ret;                                                                                                                                                                                                 
}
Note, that twl6030_gpadc_read() is just wrapper for twl_i2c_read() which takes
"an array of num_bytes containing data to be read"
I like the original implementation better then this(if I did it correctly)
Do you insist on this change?
[...]

>> +static int twl6030_gpadc_get_processed(struct twl6030_gpadc_data *gpadc,
>> +             int channel, int *val)
>> +{
>> +     int raw_code;
>> +     int corrected_code;
>> +     int channel_value;
>> +     int ret;
>> +
>> +     ret = twl6030_gpadc_get_raw(gpadc, channel, &raw_code);
>> +     if (ret)
>> +             return ret;
>> +
>
>Might first thought on seeing this is that it would be better to return
>raw for all channels and provide the scale and offset info_mask elements
>where possible rather than doing the conversion in the kernel.  Note we

In our custom kernel branch battery driver needs battery voltage and
temperature.  Thus the conversion anyway is done in the kernel, either
in ADC driver or battery driver.

>allow really quite a bit of precision on those values so I doubt it
>will be a problem.
>
>If nothing else it would make everything rather more consistent.
>
[...]

>> +static int twl6032_calibration(struct twl6030_gpadc_data *gpadc)
>> +{
>> +     int chn, d1 = 0, d2 = 0, temp;
>> +     u8 trim_regs[17];
>> +     int ret;
>> +
>> +     ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs + 1,
>> +                     TWL6030_GPADC_TRIM1, 16);
>> +     if (ret < 0) {
>> +             dev_err(gpadc->dev, "calibration failed\n");
>> +             return ret;
>> +     }
>> +
>/*
>* Loop
>please for kernel code.
>> +     /* Loop to calculate the value needed for returning voltages from
>> +      * GPADC not values.
>> +      *
>> +      * gain is calculated to 3 decimal places fixed point.
>> +      */
>> +     for (chn = 0; chn < TWL6032_GPADC_MAX_CHANNELS; chn++) {
>> +
>> +             switch (chn) {
>> +             case 0:
>> +             case 1:
>> +             case 2:
>> +             case 3:
>> +             case 4:
>> +             case 5:
>> +             case 6:
>> +             case 11:
>> +             case 12:
>> +             case 13:
>> +             case 14:
>> +                     /* D1 */
>> +                     d1 = (trim_regs[3] & 0x1F) << 2;
>> +                     d1 |= (trim_regs[1] & 0x06) >> 1;
>> +                     if (trim_regs[1] & 0x01)
>> +                             d1 = -d1;
>> +
>> +                     /* D2 */
>> +                     d2 = (trim_regs[4] & 0x3F) << 2;
>> +                     d2 |= (trim_regs[2] & 0x06) >> 1;
>> +                     if (trim_regs[2] & 0x01)
>> +                             d2 = -d2;
>> +                     break;
>> +             case 8:
>No coment on your code, but this is an 'interesting' bit
>of bit packing...
>I did vaguely wonder if this could be mapped to a big lookup table,
>but having tried it I think I end up with something almost as tricky
>to read.. Oh well that was a fun 10 minute diversion ;)
This is inherited code from Graeme - original author of implementation
for twl6032.
[...]

Regards,
OK.

^ permalink raw reply	[flat|nested] 45+ messages in thread

* RE: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 13:30       ` Kozaruk, Oleksandr
  0 siblings, 0 replies; 45+ messages in thread
From: Kozaruk, Oleksandr @ 2013-07-15 13:30 UTC (permalink / raw)
  To: Jonathan Cameron
  Cc: tony-4v6yS6AI5VpBDgjK7y7TUQ,
	benoit.cousson-QSEj5FYQhm4dnm+yROfE0A, Nayak, Rajendra, Ujfalusi,
	Peter, ABRAHAM, KISHON VIJAY, jic23-KWPb1pKIrIJaa/9Udqfwiw,
	grant.likely-QSEj5FYQhm4dnm+yROfE0A,
	rob.herring-bsGFqQB8/DxBDgjK7y7TUQ, lars-Qo5EllUWu/uELgA04lAiVw,
	sameo-VuQAYsv1563Yd54FQh9/CA, ch.naveen-Sze3O3UU22JBDgjK7y7TUQ,
	poeschel-Xtl8qvBWbHwb1SvskN2V4Q, Kim, Milo, Krishnamoorthy,
	Balaji T, gg-kDsPt+C1G03kYMGBc/C6ZA,
	linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA,
	linux-iio-u79uwXL29TY76Z2rM5mHXA, devicetree-discu

Hello Jonathan,
Thanks for the review.

>Couple of things:
>
>1) It looks from the driver that a lot of the channels are not measuring
>voltages but rather temperature or currents etc.  If so then their
>types in the channel mask should be appropriately set.  Also if some
>of the channels are entirely internal test networks, what is the benefit
>of exposing them to userspace as readable channels?
>If channels are merely 'suggested' as being used for temperatures etc
>then it is fine to keep them as voltages.
There are two kinds of channel for measuring temperature: die temperature
and those that measure temperature indirectly measuring voltage on resistive
load(NTC sensor).
die temperature is calculated by some formulas which convert ADC code to
temperature. This is not implemented in the driver.
Channels intended to measure temperature with NTC sensor have inbuilt current
sources. Voltage measured by this channels and specific NTC could be converted
to temperature.
This is the reason they chosen to be voltage channels.
As for test network I'll remove them from exposing to userspace.
[...]

>> +static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,
>> +             int channel, int *res)
>> +{
>> +     u8 reg = gpadc->pdata->channel_to_reg(channel);
>> +     u8 val[2];
>
>le16 val;
>> +     int ret;
>> +
>ret = twl6030_gpadc_read(val, reg);
>(note that (reg, val) ordering of parameters would be a more common choice)
>
>> +     ret = twl6030_gpadc_read(val, reg);
>> +     if (ret) {
>> +             dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);
>> +             return ret;
>> +     }
>> +
>res = le16_to_cpup(val);
>
>> +     *res = (val[1] << 8) | val[0];
>> +
>> +     return ret;
>> +}
>> +
You mean something like this:
static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,                                                                                                                                                  
						int channel, int *res)                                                                                                                                                                              
{                                                                                                                                                                                                                   
	u8 reg = gpadc->pdata->channel_to_reg(channel);                                                                                                                                                             
	__le16 val;                                                                                                                                                                                                 
	int ret;                                                                                                                                                                                                    

	ret = twl6030_gpadc_read(reg, (u8 *)&val);                                                                                                                                                                  
	if (ret) {                                                                                                                                                                                                  
		dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);                                                                                                                                         
		return ret;                                                                                                                                                                                         
	}                                                                                                                                                                                                           

	*res = le16_to_cpup(&val);                                                                                                                                                                                  

	return ret;                                                                                                                                                                                                 
}
Note, that twl6030_gpadc_read() is just wrapper for twl_i2c_read() which takes
"an array of num_bytes containing data to be read"
I like the original implementation better then this(if I did it correctly)
Do you insist on this change?
[...]

>> +static int twl6030_gpadc_get_processed(struct twl6030_gpadc_data *gpadc,
>> +             int channel, int *val)
>> +{
>> +     int raw_code;
>> +     int corrected_code;
>> +     int channel_value;
>> +     int ret;
>> +
>> +     ret = twl6030_gpadc_get_raw(gpadc, channel, &raw_code);
>> +     if (ret)
>> +             return ret;
>> +
>
>Might first thought on seeing this is that it would be better to return
>raw for all channels and provide the scale and offset info_mask elements
>where possible rather than doing the conversion in the kernel.  Note we

In our custom kernel branch battery driver needs battery voltage and
temperature.  Thus the conversion anyway is done in the kernel, either
in ADC driver or battery driver.

>allow really quite a bit of precision on those values so I doubt it
>will be a problem.
>
>If nothing else it would make everything rather more consistent.
>
[...]

>> +static int twl6032_calibration(struct twl6030_gpadc_data *gpadc)
>> +{
>> +     int chn, d1 = 0, d2 = 0, temp;
>> +     u8 trim_regs[17];
>> +     int ret;
>> +
>> +     ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs + 1,
>> +                     TWL6030_GPADC_TRIM1, 16);
>> +     if (ret < 0) {
>> +             dev_err(gpadc->dev, "calibration failed\n");
>> +             return ret;
>> +     }
>> +
>/*
>* Loop
>please for kernel code.
>> +     /* Loop to calculate the value needed for returning voltages from
>> +      * GPADC not values.
>> +      *
>> +      * gain is calculated to 3 decimal places fixed point.
>> +      */
>> +     for (chn = 0; chn < TWL6032_GPADC_MAX_CHANNELS; chn++) {
>> +
>> +             switch (chn) {
>> +             case 0:
>> +             case 1:
>> +             case 2:
>> +             case 3:
>> +             case 4:
>> +             case 5:
>> +             case 6:
>> +             case 11:
>> +             case 12:
>> +             case 13:
>> +             case 14:
>> +                     /* D1 */
>> +                     d1 = (trim_regs[3] & 0x1F) << 2;
>> +                     d1 |= (trim_regs[1] & 0x06) >> 1;
>> +                     if (trim_regs[1] & 0x01)
>> +                             d1 = -d1;
>> +
>> +                     /* D2 */
>> +                     d2 = (trim_regs[4] & 0x3F) << 2;
>> +                     d2 |= (trim_regs[2] & 0x06) >> 1;
>> +                     if (trim_regs[2] & 0x01)
>> +                             d2 = -d2;
>> +                     break;
>> +             case 8:
>No coment on your code, but this is an 'interesting' bit
>of bit packing...
>I did vaguely wonder if this could be mapped to a big lookup table,
>but having tried it I think I end up with something almost as tricky
>to read.. Oh well that was a fun 10 minute diversion ;)
This is inherited code from Graeme - original author of implementation
for twl6032.
[...]

Regards,
OK.

^ permalink raw reply	[flat|nested] 45+ messages in thread

* RE: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 13:30       ` Kozaruk, Oleksandr
  0 siblings, 0 replies; 45+ messages in thread
From: Kozaruk, Oleksandr @ 2013-07-15 13:30 UTC (permalink / raw)
  To: Jonathan Cameron
  Cc: tony, benoit.cousson, Nayak, Rajendra, Ujfalusi, Peter, ABRAHAM,
	KISHON VIJAY, jic23, grant.likely, rob.herring, lars, sameo,
	ch.naveen, poeschel, Kim, Milo, Krishnamoorthy, Balaji T, gg,
	linux-arm-kernel, linux-kernel, linux-iio, devicetree-discuss,
	linux-omap

Hello Jonathan,=0A=
Thanks for the review.=0A=
=0A=
>Couple of things:=0A=
>=0A=
>1) It looks from the driver that a lot of the channels are not measuring=
=0A=
>voltages but rather temperature or currents etc.  If so then their=0A=
>types in the channel mask should be appropriately set.  Also if some=0A=
>of the channels are entirely internal test networks, what is the benefit=
=0A=
>of exposing them to userspace as readable channels?=0A=
>If channels are merely 'suggested' as being used for temperatures etc=0A=
>then it is fine to keep them as voltages.=0A=
There are two kinds of channel for measuring temperature: die temperature=
=0A=
and those that measure temperature indirectly measuring voltage on resistiv=
e=0A=
load(NTC sensor).=0A=
die temperature is calculated by some formulas which convert ADC code to=0A=
temperature. This is not implemented in the driver.=0A=
Channels intended to measure temperature with NTC sensor have inbuilt curre=
nt=0A=
sources. Voltage measured by this channels and specific NTC could be conver=
ted=0A=
to temperature.=0A=
This is the reason they chosen to be voltage channels.=0A=
As for test network I'll remove them from exposing to userspace.=0A=
[...]=0A=
=0A=
>> +static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,=0A=
>> +             int channel, int *res)=0A=
>> +{=0A=
>> +     u8 reg =3D gpadc->pdata->channel_to_reg(channel);=0A=
>> +     u8 val[2];=0A=
>=0A=
>le16 val;=0A=
>> +     int ret;=0A=
>> +=0A=
>ret =3D twl6030_gpadc_read(val, reg);=0A=
>(note that (reg, val) ordering of parameters would be a more common choice=
)=0A=
>=0A=
>> +     ret =3D twl6030_gpadc_read(val, reg);=0A=
>> +     if (ret) {=0A=
>> +             dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg)=
;=0A=
>> +             return ret;=0A=
>> +     }=0A=
>> +=0A=
>res =3D le16_to_cpup(val);=0A=
>=0A=
>> +     *res =3D (val[1] << 8) | val[0];=0A=
>> +=0A=
>> +     return ret;=0A=
>> +}=0A=
>> +=0A=
You mean something like this:=0A=
static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,         =
                                                                           =
                                                              =0A=
						int channel, int *res)                                               =
                                                                           =
                                                    =0A=
{                                                                          =
                                                                           =
                                                              =0A=
	u8 reg =3D gpadc->pdata->channel_to_reg(channel);                         =
                                                                           =
                                                         =0A=
	__le16 val;                                                               =
                                                                           =
                                                       =0A=
	int ret;                                                                  =
                                                                           =
                                                       =0A=
=0A=
	ret =3D twl6030_gpadc_read(reg, (u8 *)&val);                              =
                                                                           =
                                                         =0A=
	if (ret) {                                                                =
                                                                           =
                                                       =0A=
		dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);              =
                                                                           =
                                                =0A=
		return ret;                                                              =
                                                                           =
                                                =0A=
	}                                                                         =
                                                                           =
                                                       =0A=
=0A=
	*res =3D le16_to_cpup(&val);                                              =
                                                                           =
                                                         =0A=
=0A=
	return ret;                                                               =
                                                                           =
                                                       =0A=
}=0A=
Note, that twl6030_gpadc_read() is just wrapper for twl_i2c_read() which ta=
kes=0A=
"an array of num_bytes containing data to be read"=0A=
I like the original implementation better then this(if I did it correctly)=
=0A=
Do you insist on this change?=0A=
[...]=0A=
=0A=
>> +static int twl6030_gpadc_get_processed(struct twl6030_gpadc_data *gpadc=
,=0A=
>> +             int channel, int *val)=0A=
>> +{=0A=
>> +     int raw_code;=0A=
>> +     int corrected_code;=0A=
>> +     int channel_value;=0A=
>> +     int ret;=0A=
>> +=0A=
>> +     ret =3D twl6030_gpadc_get_raw(gpadc, channel, &raw_code);=0A=
>> +     if (ret)=0A=
>> +             return ret;=0A=
>> +=0A=
>=0A=
>Might first thought on seeing this is that it would be better to return=0A=
>raw for all channels and provide the scale and offset info_mask elements=
=0A=
>where possible rather than doing the conversion in the kernel.  Note we=0A=
=0A=
In our custom kernel branch battery driver needs battery voltage and=0A=
temperature.  Thus the conversion anyway is done in the kernel, either=0A=
in ADC driver or battery driver.=0A=
=0A=
>allow really quite a bit of precision on those values so I doubt it=0A=
>will be a problem.=0A=
>=0A=
>If nothing else it would make everything rather more consistent.=0A=
>=0A=
[...]=0A=
=0A=
>> +static int twl6032_calibration(struct twl6030_gpadc_data *gpadc)=0A=
>> +{=0A=
>> +     int chn, d1 =3D 0, d2 =3D 0, temp;=0A=
>> +     u8 trim_regs[17];=0A=
>> +     int ret;=0A=
>> +=0A=
>> +     ret =3D twl_i2c_read(TWL6030_MODULE_ID2, trim_regs + 1,=0A=
>> +                     TWL6030_GPADC_TRIM1, 16);=0A=
>> +     if (ret < 0) {=0A=
>> +             dev_err(gpadc->dev, "calibration failed\n");=0A=
>> +             return ret;=0A=
>> +     }=0A=
>> +=0A=
>/*=0A=
>* Loop=0A=
>please for kernel code.=0A=
>> +     /* Loop to calculate the value needed for returning voltages from=
=0A=
>> +      * GPADC not values.=0A=
>> +      *=0A=
>> +      * gain is calculated to 3 decimal places fixed point.=0A=
>> +      */=0A=
>> +     for (chn =3D 0; chn < TWL6032_GPADC_MAX_CHANNELS; chn++) {=0A=
>> +=0A=
>> +             switch (chn) {=0A=
>> +             case 0:=0A=
>> +             case 1:=0A=
>> +             case 2:=0A=
>> +             case 3:=0A=
>> +             case 4:=0A=
>> +             case 5:=0A=
>> +             case 6:=0A=
>> +             case 11:=0A=
>> +             case 12:=0A=
>> +             case 13:=0A=
>> +             case 14:=0A=
>> +                     /* D1 */=0A=
>> +                     d1 =3D (trim_regs[3] & 0x1F) << 2;=0A=
>> +                     d1 |=3D (trim_regs[1] & 0x06) >> 1;=0A=
>> +                     if (trim_regs[1] & 0x01)=0A=
>> +                             d1 =3D -d1;=0A=
>> +=0A=
>> +                     /* D2 */=0A=
>> +                     d2 =3D (trim_regs[4] & 0x3F) << 2;=0A=
>> +                     d2 |=3D (trim_regs[2] & 0x06) >> 1;=0A=
>> +                     if (trim_regs[2] & 0x01)=0A=
>> +                             d2 =3D -d2;=0A=
>> +                     break;=0A=
>> +             case 8:=0A=
>No coment on your code, but this is an 'interesting' bit=0A=
>of bit packing...=0A=
>I did vaguely wonder if this could be mapped to a big lookup table,=0A=
>but having tried it I think I end up with something almost as tricky=0A=
>to read.. Oh well that was a fun 10 minute diversion ;)=0A=
This is inherited code from Graeme - original author of implementation=0A=
for twl6032.=0A=
[...]=0A=
=0A=
Regards,=0A=
OK.=0A=

^ permalink raw reply	[flat|nested] 45+ messages in thread

* [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 13:30       ` Kozaruk, Oleksandr
  0 siblings, 0 replies; 45+ messages in thread
From: Kozaruk, Oleksandr @ 2013-07-15 13:30 UTC (permalink / raw)
  To: linux-arm-kernel

Hello Jonathan,
Thanks for the review.

>Couple of things:
>
>1) It looks from the driver that a lot of the channels are not measuring
>voltages but rather temperature or currents etc.  If so then their
>types in the channel mask should be appropriately set.  Also if some
>of the channels are entirely internal test networks, what is the benefit
>of exposing them to userspace as readable channels?
>If channels are merely 'suggested' as being used for temperatures etc
>then it is fine to keep them as voltages.
There are two kinds of channel for measuring temperature: die temperature
and those that measure temperature indirectly measuring voltage on resistive
load(NTC sensor).
die temperature is calculated by some formulas which convert ADC code to
temperature. This is not implemented in the driver.
Channels intended to measure temperature with NTC sensor have inbuilt current
sources. Voltage measured by this channels and specific NTC could be converted
to temperature.
This is the reason they chosen to be voltage channels.
As for test network I'll remove them from exposing to userspace.
[...]

>> +static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,
>> +             int channel, int *res)
>> +{
>> +     u8 reg = gpadc->pdata->channel_to_reg(channel);
>> +     u8 val[2];
>
>le16 val;
>> +     int ret;
>> +
>ret = twl6030_gpadc_read(val, reg);
>(note that (reg, val) ordering of parameters would be a more common choice)
>
>> +     ret = twl6030_gpadc_read(val, reg);
>> +     if (ret) {
>> +             dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);
>> +             return ret;
>> +     }
>> +
>res = le16_to_cpup(val);
>
>> +     *res = (val[1] << 8) | val[0];
>> +
>> +     return ret;
>> +}
>> +
You mean something like this:
static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,                                                                                                                                                  
						int channel, int *res)                                                                                                                                                                              
{                                                                                                                                                                                                                   
	u8 reg = gpadc->pdata->channel_to_reg(channel);                                                                                                                                                             
	__le16 val;                                                                                                                                                                                                 
	int ret;                                                                                                                                                                                                    

	ret = twl6030_gpadc_read(reg, (u8 *)&val);                                                                                                                                                                  
	if (ret) {                                                                                                                                                                                                  
		dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);                                                                                                                                         
		return ret;                                                                                                                                                                                         
	}                                                                                                                                                                                                           

	*res = le16_to_cpup(&val);                                                                                                                                                                                  

	return ret;                                                                                                                                                                                                 
}
Note, that twl6030_gpadc_read() is just wrapper for twl_i2c_read() which takes
"an array of num_bytes containing data to be read"
I like the original implementation better then this(if I did it correctly)
Do you insist on this change?
[...]

>> +static int twl6030_gpadc_get_processed(struct twl6030_gpadc_data *gpadc,
>> +             int channel, int *val)
>> +{
>> +     int raw_code;
>> +     int corrected_code;
>> +     int channel_value;
>> +     int ret;
>> +
>> +     ret = twl6030_gpadc_get_raw(gpadc, channel, &raw_code);
>> +     if (ret)
>> +             return ret;
>> +
>
>Might first thought on seeing this is that it would be better to return
>raw for all channels and provide the scale and offset info_mask elements
>where possible rather than doing the conversion in the kernel.  Note we

In our custom kernel branch battery driver needs battery voltage and
temperature.  Thus the conversion anyway is done in the kernel, either
in ADC driver or battery driver.

>allow really quite a bit of precision on those values so I doubt it
>will be a problem.
>
>If nothing else it would make everything rather more consistent.
>
[...]

>> +static int twl6032_calibration(struct twl6030_gpadc_data *gpadc)
>> +{
>> +     int chn, d1 = 0, d2 = 0, temp;
>> +     u8 trim_regs[17];
>> +     int ret;
>> +
>> +     ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs + 1,
>> +                     TWL6030_GPADC_TRIM1, 16);
>> +     if (ret < 0) {
>> +             dev_err(gpadc->dev, "calibration failed\n");
>> +             return ret;
>> +     }
>> +
>/*
>* Loop
>please for kernel code.
>> +     /* Loop to calculate the value needed for returning voltages from
>> +      * GPADC not values.
>> +      *
>> +      * gain is calculated to 3 decimal places fixed point.
>> +      */
>> +     for (chn = 0; chn < TWL6032_GPADC_MAX_CHANNELS; chn++) {
>> +
>> +             switch (chn) {
>> +             case 0:
>> +             case 1:
>> +             case 2:
>> +             case 3:
>> +             case 4:
>> +             case 5:
>> +             case 6:
>> +             case 11:
>> +             case 12:
>> +             case 13:
>> +             case 14:
>> +                     /* D1 */
>> +                     d1 = (trim_regs[3] & 0x1F) << 2;
>> +                     d1 |= (trim_regs[1] & 0x06) >> 1;
>> +                     if (trim_regs[1] & 0x01)
>> +                             d1 = -d1;
>> +
>> +                     /* D2 */
>> +                     d2 = (trim_regs[4] & 0x3F) << 2;
>> +                     d2 |= (trim_regs[2] & 0x06) >> 1;
>> +                     if (trim_regs[2] & 0x01)
>> +                             d2 = -d2;
>> +                     break;
>> +             case 8:
>No coment on your code, but this is an 'interesting' bit
>of bit packing...
>I did vaguely wonder if this could be mapped to a big lookup table,
>but having tried it I think I end up with something almost as tricky
>to read.. Oh well that was a fun 10 minute diversion ;)
This is inherited code from Graeme - original author of implementation
for twl6032.
[...]

Regards,
OK.

^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
  2013-07-15 13:30       ` Kozaruk, Oleksandr
  (?)
  (?)
@ 2013-07-15 14:05         ` Graeme Gregory
  -1 siblings, 0 replies; 45+ messages in thread
From: Graeme Gregory @ 2013-07-15 14:05 UTC (permalink / raw)
  To: Kozaruk, Oleksandr
  Cc: Jonathan Cameron, tony, benoit.cousson, Nayak, Rajendra,
	Ujfalusi, Peter, ABRAHAM, KISHON VIJAY, jic23, grant.likely,
	rob.herring, lars, sameo, ch.naveen, poeschel, Kim, Milo,
	Krishnamoorthy, Balaji T, linux-arm-kernel, linux-kernel,
	linux-iio, devicetree-discuss, linux-omap

On 15/07/13 14:30, Kozaruk, Oleksandr wrote:
> Hello Jonathan,
> Thanks for the review.
>
>> Couple of things:
>>
>> 1) It looks from the driver that a lot of the channels are not measuring
>> voltages but rather temperature or currents etc.  If so then their
>> types in the channel mask should be appropriately set.  Also if some
>> of the channels are entirely internal test networks, what is the benefit
>> of exposing them to userspace as readable channels?
>> If channels are merely 'suggested' as being used for temperatures etc
>> then it is fine to keep them as voltages.
> There are two kinds of channel for measuring temperature: die temperature
> and those that measure temperature indirectly measuring voltage on resistive
> load(NTC sensor).
> die temperature is calculated by some formulas which convert ADC code to
> temperature. This is not implemented in the driver.
> Channels intended to measure temperature with NTC sensor have inbuilt current
> sources. Voltage measured by this channels and specific NTC could be converted
> to temperature.
> This is the reason they chosen to be voltage channels.
> As for test network I'll remove them from exposing to userspace.
> [...]
>
>>> +static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,
>>> +             int channel, int *res)
>>> +{
>>> +     u8 reg = gpadc->pdata->channel_to_reg(channel);
>>> +     u8 val[2];
>> le16 val;
>>> +     int ret;
>>> +
>> ret = twl6030_gpadc_read(val, reg);
>> (note that (reg, val) ordering of parameters would be a more common choice)
>>
>>> +     ret = twl6030_gpadc_read(val, reg);
>>> +     if (ret) {
>>> +             dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);
>>> +             return ret;
>>> +     }
>>> +
>> res = le16_to_cpup(val);
>>
>>> +     *res = (val[1] << 8) | val[0];
>>> +
>>> +     return ret;
>>> +}
>>> +
> You mean something like this:
> static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,                                                                                                                                                  
> 						int channel, int *res)                                                                                                                                                                              
> {                                                                                                                                                                                                                   
> 	u8 reg = gpadc->pdata->channel_to_reg(channel);                                                                                                                                                             
> 	__le16 val;                                                                                                                                                                                                 
> 	int ret;                                                                                                                                                                                                    
>
> 	ret = twl6030_gpadc_read(reg, (u8 *)&val);                                                                                                                                                                  
> 	if (ret) {                                                                                                                                                                                                  
> 		dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);                                                                                                                                         
> 		return ret;                                                                                                                                                                                         
> 	}                                                                                                                                                                                                           
>
> 	*res = le16_to_cpup(&val);                                                                                                                                                                                  
>
> 	return ret;                                                                                                                                                                                                 
> }
> Note, that twl6030_gpadc_read() is just wrapper for twl_i2c_read() which takes
> "an array of num_bytes containing data to be read"
> I like the original implementation better then this(if I did it correctly)
> Do you insist on this change?
> [...]
>
>>> +static int twl6030_gpadc_get_processed(struct twl6030_gpadc_data *gpadc,
>>> +             int channel, int *val)
>>> +{
>>> +     int raw_code;
>>> +     int corrected_code;
>>> +     int channel_value;
>>> +     int ret;
>>> +
>>> +     ret = twl6030_gpadc_get_raw(gpadc, channel, &raw_code);
>>> +     if (ret)
>>> +             return ret;
>>> +
>> Might first thought on seeing this is that it would be better to return
>> raw for all channels and provide the scale and offset info_mask elements
>> where possible rather than doing the conversion in the kernel.  Note we
> In our custom kernel branch battery driver needs battery voltage and
> temperature.  Thus the conversion anyway is done in the kernel, either
> in ADC driver or battery driver.
>
>> allow really quite a bit of precision on those values so I doubt it
>> will be a problem.
>>
>> If nothing else it would make everything rather more consistent.
>>
> [...]
>
>>> +static int twl6032_calibration(struct twl6030_gpadc_data *gpadc)
>>> +{
>>> +     int chn, d1 = 0, d2 = 0, temp;
>>> +     u8 trim_regs[17];
>>> +     int ret;
>>> +
>>> +     ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs + 1,
>>> +                     TWL6030_GPADC_TRIM1, 16);
>>> +     if (ret < 0) {
>>> +             dev_err(gpadc->dev, "calibration failed\n");
>>> +             return ret;
>>> +     }
>>> +
>> /*
>> * Loop
>> please for kernel code.
>>> +     /* Loop to calculate the value needed for returning voltages from
>>> +      * GPADC not values.
>>> +      *
>>> +      * gain is calculated to 3 decimal places fixed point.
>>> +      */
>>> +     for (chn = 0; chn < TWL6032_GPADC_MAX_CHANNELS; chn++) {
>>> +
>>> +             switch (chn) {
>>> +             case 0:
>>> +             case 1:
>>> +             case 2:
>>> +             case 3:
>>> +             case 4:
>>> +             case 5:
>>> +             case 6:
>>> +             case 11:
>>> +             case 12:
>>> +             case 13:
>>> +             case 14:
>>> +                     /* D1 */
>>> +                     d1 = (trim_regs[3] & 0x1F) << 2;
>>> +                     d1 |= (trim_regs[1] & 0x06) >> 1;
>>> +                     if (trim_regs[1] & 0x01)
>>> +                             d1 = -d1;
>>> +
>>> +                     /* D2 */
>>> +                     d2 = (trim_regs[4] & 0x3F) << 2;
>>> +                     d2 |= (trim_regs[2] & 0x06) >> 1;
>>> +                     if (trim_regs[2] & 0x01)
>>> +                             d2 = -d2;
>>> +                     break;
>>> +             case 8:
>> No coment on your code, but this is an 'interesting' bit
>> of bit packing...
>> I did vaguely wonder if this could be mapped to a big lookup table,
>> but having tried it I think I end up with something almost as tricky
>> to read.. Oh well that was a fun 10 minute diversion ;)
> This is inherited code from Graeme - original author of implementation
> for twl6032.
> [...]
>
That bit is the bane of my life :-( I think the data sheet used to
changed weekly as the hardware guys did not understand the difference
between ones and twos complement!

The reason for the crazy packing is there was not enough space allocated
in hardware for the change from 10 to 12 bits for the error values
between 6030 and 6032.

Graeme


^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 14:05         ` Graeme Gregory
  0 siblings, 0 replies; 45+ messages in thread
From: Graeme Gregory @ 2013-07-15 14:05 UTC (permalink / raw)
  To: Kozaruk, Oleksandr
  Cc: Jonathan Cameron, tony, benoit.cousson, Nayak, Rajendra,
	Ujfalusi, Peter, ABRAHAM, KISHON VIJAY, jic23, grant.likely,
	rob.herring, lars, sameo, ch.naveen, poeschel, Kim, Milo,
	Krishnamoorthy, Balaji T, linux-arm-kernel, linux-kernel,
	linux-iio, devicetree-discuss

On 15/07/13 14:30, Kozaruk, Oleksandr wrote:
> Hello Jonathan,
> Thanks for the review.
>
>> Couple of things:
>>
>> 1) It looks from the driver that a lot of the channels are not measuring
>> voltages but rather temperature or currents etc.  If so then their
>> types in the channel mask should be appropriately set.  Also if some
>> of the channels are entirely internal test networks, what is the benefit
>> of exposing them to userspace as readable channels?
>> If channels are merely 'suggested' as being used for temperatures etc
>> then it is fine to keep them as voltages.
> There are two kinds of channel for measuring temperature: die temperature
> and those that measure temperature indirectly measuring voltage on resistive
> load(NTC sensor).
> die temperature is calculated by some formulas which convert ADC code to
> temperature. This is not implemented in the driver.
> Channels intended to measure temperature with NTC sensor have inbuilt current
> sources. Voltage measured by this channels and specific NTC could be converted
> to temperature.
> This is the reason they chosen to be voltage channels.
> As for test network I'll remove them from exposing to userspace.
> [...]
>
>>> +static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,
>>> +             int channel, int *res)
>>> +{
>>> +     u8 reg = gpadc->pdata->channel_to_reg(channel);
>>> +     u8 val[2];
>> le16 val;
>>> +     int ret;
>>> +
>> ret = twl6030_gpadc_read(val, reg);
>> (note that (reg, val) ordering of parameters would be a more common choice)
>>
>>> +     ret = twl6030_gpadc_read(val, reg);
>>> +     if (ret) {
>>> +             dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);
>>> +             return ret;
>>> +     }
>>> +
>> res = le16_to_cpup(val);
>>
>>> +     *res = (val[1] << 8) | val[0];
>>> +
>>> +     return ret;
>>> +}
>>> +
> You mean something like this:
> static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,                                                                                                                                                  
> 						int channel, int *res)                                                                                                                                                                              
> {                                                                                                                                                                                                                   
> 	u8 reg = gpadc->pdata->channel_to_reg(channel);                                                                                                                                                             
> 	__le16 val;                                                                                                                                                                                                 
> 	int ret;                                                                                                                                                                                                    
>
> 	ret = twl6030_gpadc_read(reg, (u8 *)&val);                                                                                                                                                                  
> 	if (ret) {                                                                                                                                                                                                  
> 		dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);                                                                                                                                         
> 		return ret;                                                                                                                                                                                         
> 	}                                                                                                                                                                                                           
>
> 	*res = le16_to_cpup(&val);                                                                                                                                                                                  
>
> 	return ret;                                                                                                                                                                                                 
> }
> Note, that twl6030_gpadc_read() is just wrapper for twl_i2c_read() which takes
> "an array of num_bytes containing data to be read"
> I like the original implementation better then this(if I did it correctly)
> Do you insist on this change?
> [...]
>
>>> +static int twl6030_gpadc_get_processed(struct twl6030_gpadc_data *gpadc,
>>> +             int channel, int *val)
>>> +{
>>> +     int raw_code;
>>> +     int corrected_code;
>>> +     int channel_value;
>>> +     int ret;
>>> +
>>> +     ret = twl6030_gpadc_get_raw(gpadc, channel, &raw_code);
>>> +     if (ret)
>>> +             return ret;
>>> +
>> Might first thought on seeing this is that it would be better to return
>> raw for all channels and provide the scale and offset info_mask elements
>> where possible rather than doing the conversion in the kernel.  Note we
> In our custom kernel branch battery driver needs battery voltage and
> temperature.  Thus the conversion anyway is done in the kernel, either
> in ADC driver or battery driver.
>
>> allow really quite a bit of precision on those values so I doubt it
>> will be a problem.
>>
>> If nothing else it would make everything rather more consistent.
>>
> [...]
>
>>> +static int twl6032_calibration(struct twl6030_gpadc_data *gpadc)
>>> +{
>>> +     int chn, d1 = 0, d2 = 0, temp;
>>> +     u8 trim_regs[17];
>>> +     int ret;
>>> +
>>> +     ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs + 1,
>>> +                     TWL6030_GPADC_TRIM1, 16);
>>> +     if (ret < 0) {
>>> +             dev_err(gpadc->dev, "calibration failed\n");
>>> +             return ret;
>>> +     }
>>> +
>> /*
>> * Loop
>> please for kernel code.
>>> +     /* Loop to calculate the value needed for returning voltages from
>>> +      * GPADC not values.
>>> +      *
>>> +      * gain is calculated to 3 decimal places fixed point.
>>> +      */
>>> +     for (chn = 0; chn < TWL6032_GPADC_MAX_CHANNELS; chn++) {
>>> +
>>> +             switch (chn) {
>>> +             case 0:
>>> +             case 1:
>>> +             case 2:
>>> +             case 3:
>>> +             case 4:
>>> +             case 5:
>>> +             case 6:
>>> +             case 11:
>>> +             case 12:
>>> +             case 13:
>>> +             case 14:
>>> +                     /* D1 */
>>> +                     d1 = (trim_regs[3] & 0x1F) << 2;
>>> +                     d1 |= (trim_regs[1] & 0x06) >> 1;
>>> +                     if (trim_regs[1] & 0x01)
>>> +                             d1 = -d1;
>>> +
>>> +                     /* D2 */
>>> +                     d2 = (trim_regs[4] & 0x3F) << 2;
>>> +                     d2 |= (trim_regs[2] & 0x06) >> 1;
>>> +                     if (trim_regs[2] & 0x01)
>>> +                             d2 = -d2;
>>> +                     break;
>>> +             case 8:
>> No coment on your code, but this is an 'interesting' bit
>> of bit packing...
>> I did vaguely wonder if this could be mapped to a big lookup table,
>> but having tried it I think I end up with something almost as tricky
>> to read.. Oh well that was a fun 10 minute diversion ;)
> This is inherited code from Graeme - original author of implementation
> for twl6032.
> [...]
>
That bit is the bane of my life :-( I think the data sheet used to
changed weekly as the hardware guys did not understand the difference
between ones and twos complement!

The reason for the crazy packing is there was not enough space allocated
in hardware for the change from 10 to 12 bits for the error values
between 6030 and 6032.

Graeme


^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 14:05         ` Graeme Gregory
  0 siblings, 0 replies; 45+ messages in thread
From: Graeme Gregory @ 2013-07-15 14:05 UTC (permalink / raw)
  To: Kozaruk, Oleksandr
  Cc: Jonathan Cameron, tony, benoit.cousson, Nayak, Rajendra,
	Ujfalusi, Peter, ABRAHAM, KISHON VIJAY, jic23, grant.likely,
	rob.herring, lars, sameo, ch.naveen, poeschel, Kim, Milo,
	Krishnamoorthy, Balaji T, linux-arm-kernel, linux-kernel,
	linux-iio, devicetree-discuss, linux-omap

On 15/07/13 14:30, Kozaruk, Oleksandr wrote:
> Hello Jonathan,
> Thanks for the review.
>
>> Couple of things:
>>
>> 1) It looks from the driver that a lot of the channels are not measuring
>> voltages but rather temperature or currents etc.  If so then their
>> types in the channel mask should be appropriately set.  Also if some
>> of the channels are entirely internal test networks, what is the benefit
>> of exposing them to userspace as readable channels?
>> If channels are merely 'suggested' as being used for temperatures etc
>> then it is fine to keep them as voltages.
> There are two kinds of channel for measuring temperature: die temperature
> and those that measure temperature indirectly measuring voltage on resistive
> load(NTC sensor).
> die temperature is calculated by some formulas which convert ADC code to
> temperature. This is not implemented in the driver.
> Channels intended to measure temperature with NTC sensor have inbuilt current
> sources. Voltage measured by this channels and specific NTC could be converted
> to temperature.
> This is the reason they chosen to be voltage channels.
> As for test network I'll remove them from exposing to userspace.
> [...]
>
>>> +static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,
>>> +             int channel, int *res)
>>> +{
>>> +     u8 reg = gpadc->pdata->channel_to_reg(channel);
>>> +     u8 val[2];
>> le16 val;
>>> +     int ret;
>>> +
>> ret = twl6030_gpadc_read(val, reg);
>> (note that (reg, val) ordering of parameters would be a more common choice)
>>
>>> +     ret = twl6030_gpadc_read(val, reg);
>>> +     if (ret) {
>>> +             dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);
>>> +             return ret;
>>> +     }
>>> +
>> res = le16_to_cpup(val);
>>
>>> +     *res = (val[1] << 8) | val[0];
>>> +
>>> +     return ret;
>>> +}
>>> +
> You mean something like this:
> static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,                                                                                                                                                  
> 						int channel, int *res)                                                                                                                                                                              
> {                                                                                                                                                                                                                   
> 	u8 reg = gpadc->pdata->channel_to_reg(channel);                                                                                                                                                             
> 	__le16 val;                                                                                                                                                                                                 
> 	int ret;                                                                                                                                                                                                    
>
> 	ret = twl6030_gpadc_read(reg, (u8 *)&val);                                                                                                                                                                  
> 	if (ret) {                                                                                                                                                                                                  
> 		dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);                                                                                                                                         
> 		return ret;                                                                                                                                                                                         
> 	}                                                                                                                                                                                                           
>
> 	*res = le16_to_cpup(&val);                                                                                                                                                                                  
>
> 	return ret;                                                                                                                                                                                                 
> }
> Note, that twl6030_gpadc_read() is just wrapper for twl_i2c_read() which takes
> "an array of num_bytes containing data to be read"
> I like the original implementation better then this(if I did it correctly)
> Do you insist on this change?
> [...]
>
>>> +static int twl6030_gpadc_get_processed(struct twl6030_gpadc_data *gpadc,
>>> +             int channel, int *val)
>>> +{
>>> +     int raw_code;
>>> +     int corrected_code;
>>> +     int channel_value;
>>> +     int ret;
>>> +
>>> +     ret = twl6030_gpadc_get_raw(gpadc, channel, &raw_code);
>>> +     if (ret)
>>> +             return ret;
>>> +
>> Might first thought on seeing this is that it would be better to return
>> raw for all channels and provide the scale and offset info_mask elements
>> where possible rather than doing the conversion in the kernel.  Note we
> In our custom kernel branch battery driver needs battery voltage and
> temperature.  Thus the conversion anyway is done in the kernel, either
> in ADC driver or battery driver.
>
>> allow really quite a bit of precision on those values so I doubt it
>> will be a problem.
>>
>> If nothing else it would make everything rather more consistent.
>>
> [...]
>
>>> +static int twl6032_calibration(struct twl6030_gpadc_data *gpadc)
>>> +{
>>> +     int chn, d1 = 0, d2 = 0, temp;
>>> +     u8 trim_regs[17];
>>> +     int ret;
>>> +
>>> +     ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs + 1,
>>> +                     TWL6030_GPADC_TRIM1, 16);
>>> +     if (ret < 0) {
>>> +             dev_err(gpadc->dev, "calibration failed\n");
>>> +             return ret;
>>> +     }
>>> +
>> /*
>> * Loop
>> please for kernel code.
>>> +     /* Loop to calculate the value needed for returning voltages from
>>> +      * GPADC not values.
>>> +      *
>>> +      * gain is calculated to 3 decimal places fixed point.
>>> +      */
>>> +     for (chn = 0; chn < TWL6032_GPADC_MAX_CHANNELS; chn++) {
>>> +
>>> +             switch (chn) {
>>> +             case 0:
>>> +             case 1:
>>> +             case 2:
>>> +             case 3:
>>> +             case 4:
>>> +             case 5:
>>> +             case 6:
>>> +             case 11:
>>> +             case 12:
>>> +             case 13:
>>> +             case 14:
>>> +                     /* D1 */
>>> +                     d1 = (trim_regs[3] & 0x1F) << 2;
>>> +                     d1 |= (trim_regs[1] & 0x06) >> 1;
>>> +                     if (trim_regs[1] & 0x01)
>>> +                             d1 = -d1;
>>> +
>>> +                     /* D2 */
>>> +                     d2 = (trim_regs[4] & 0x3F) << 2;
>>> +                     d2 |= (trim_regs[2] & 0x06) >> 1;
>>> +                     if (trim_regs[2] & 0x01)
>>> +                             d2 = -d2;
>>> +                     break;
>>> +             case 8:
>> No coment on your code, but this is an 'interesting' bit
>> of bit packing...
>> I did vaguely wonder if this could be mapped to a big lookup table,
>> but having tried it I think I end up with something almost as tricky
>> to read.. Oh well that was a fun 10 minute diversion ;)
> This is inherited code from Graeme - original author of implementation
> for twl6032.
> [...]
>
That bit is the bane of my life :-( I think the data sheet used to
changed weekly as the hardware guys did not understand the difference
between ones and twos complement!

The reason for the crazy packing is there was not enough space allocated
in hardware for the change from 10 to 12 bits for the error values
between 6030 and 6032.

Graeme


^ permalink raw reply	[flat|nested] 45+ messages in thread

* [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-15 14:05         ` Graeme Gregory
  0 siblings, 0 replies; 45+ messages in thread
From: Graeme Gregory @ 2013-07-15 14:05 UTC (permalink / raw)
  To: linux-arm-kernel

On 15/07/13 14:30, Kozaruk, Oleksandr wrote:
> Hello Jonathan,
> Thanks for the review.
>
>> Couple of things:
>>
>> 1) It looks from the driver that a lot of the channels are not measuring
>> voltages but rather temperature or currents etc.  If so then their
>> types in the channel mask should be appropriately set.  Also if some
>> of the channels are entirely internal test networks, what is the benefit
>> of exposing them to userspace as readable channels?
>> If channels are merely 'suggested' as being used for temperatures etc
>> then it is fine to keep them as voltages.
> There are two kinds of channel for measuring temperature: die temperature
> and those that measure temperature indirectly measuring voltage on resistive
> load(NTC sensor).
> die temperature is calculated by some formulas which convert ADC code to
> temperature. This is not implemented in the driver.
> Channels intended to measure temperature with NTC sensor have inbuilt current
> sources. Voltage measured by this channels and specific NTC could be converted
> to temperature.
> This is the reason they chosen to be voltage channels.
> As for test network I'll remove them from exposing to userspace.
> [...]
>
>>> +static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,
>>> +             int channel, int *res)
>>> +{
>>> +     u8 reg = gpadc->pdata->channel_to_reg(channel);
>>> +     u8 val[2];
>> le16 val;
>>> +     int ret;
>>> +
>> ret = twl6030_gpadc_read(val, reg);
>> (note that (reg, val) ordering of parameters would be a more common choice)
>>
>>> +     ret = twl6030_gpadc_read(val, reg);
>>> +     if (ret) {
>>> +             dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);
>>> +             return ret;
>>> +     }
>>> +
>> res = le16_to_cpup(val);
>>
>>> +     *res = (val[1] << 8) | val[0];
>>> +
>>> +     return ret;
>>> +}
>>> +
> You mean something like this:
> static int twl6030_gpadc_get_raw(struct twl6030_gpadc_data *gpadc,                                                                                                                                                  
> 						int channel, int *res)                                                                                                                                                                              
> {                                                                                                                                                                                                                   
> 	u8 reg = gpadc->pdata->channel_to_reg(channel);                                                                                                                                                             
> 	__le16 val;                                                                                                                                                                                                 
> 	int ret;                                                                                                                                                                                                    
>
> 	ret = twl6030_gpadc_read(reg, (u8 *)&val);                                                                                                                                                                  
> 	if (ret) {                                                                                                                                                                                                  
> 		dev_dbg(gpadc->dev, "unable to read register 0x%X\n", reg);                                                                                                                                         
> 		return ret;                                                                                                                                                                                         
> 	}                                                                                                                                                                                                           
>
> 	*res = le16_to_cpup(&val);                                                                                                                                                                                  
>
> 	return ret;                                                                                                                                                                                                 
> }
> Note, that twl6030_gpadc_read() is just wrapper for twl_i2c_read() which takes
> "an array of num_bytes containing data to be read"
> I like the original implementation better then this(if I did it correctly)
> Do you insist on this change?
> [...]
>
>>> +static int twl6030_gpadc_get_processed(struct twl6030_gpadc_data *gpadc,
>>> +             int channel, int *val)
>>> +{
>>> +     int raw_code;
>>> +     int corrected_code;
>>> +     int channel_value;
>>> +     int ret;
>>> +
>>> +     ret = twl6030_gpadc_get_raw(gpadc, channel, &raw_code);
>>> +     if (ret)
>>> +             return ret;
>>> +
>> Might first thought on seeing this is that it would be better to return
>> raw for all channels and provide the scale and offset info_mask elements
>> where possible rather than doing the conversion in the kernel.  Note we
> In our custom kernel branch battery driver needs battery voltage and
> temperature.  Thus the conversion anyway is done in the kernel, either
> in ADC driver or battery driver.
>
>> allow really quite a bit of precision on those values so I doubt it
>> will be a problem.
>>
>> If nothing else it would make everything rather more consistent.
>>
> [...]
>
>>> +static int twl6032_calibration(struct twl6030_gpadc_data *gpadc)
>>> +{
>>> +     int chn, d1 = 0, d2 = 0, temp;
>>> +     u8 trim_regs[17];
>>> +     int ret;
>>> +
>>> +     ret = twl_i2c_read(TWL6030_MODULE_ID2, trim_regs + 1,
>>> +                     TWL6030_GPADC_TRIM1, 16);
>>> +     if (ret < 0) {
>>> +             dev_err(gpadc->dev, "calibration failed\n");
>>> +             return ret;
>>> +     }
>>> +
>> /*
>> * Loop
>> please for kernel code.
>>> +     /* Loop to calculate the value needed for returning voltages from
>>> +      * GPADC not values.
>>> +      *
>>> +      * gain is calculated to 3 decimal places fixed point.
>>> +      */
>>> +     for (chn = 0; chn < TWL6032_GPADC_MAX_CHANNELS; chn++) {
>>> +
>>> +             switch (chn) {
>>> +             case 0:
>>> +             case 1:
>>> +             case 2:
>>> +             case 3:
>>> +             case 4:
>>> +             case 5:
>>> +             case 6:
>>> +             case 11:
>>> +             case 12:
>>> +             case 13:
>>> +             case 14:
>>> +                     /* D1 */
>>> +                     d1 = (trim_regs[3] & 0x1F) << 2;
>>> +                     d1 |= (trim_regs[1] & 0x06) >> 1;
>>> +                     if (trim_regs[1] & 0x01)
>>> +                             d1 = -d1;
>>> +
>>> +                     /* D2 */
>>> +                     d2 = (trim_regs[4] & 0x3F) << 2;
>>> +                     d2 |= (trim_regs[2] & 0x06) >> 1;
>>> +                     if (trim_regs[2] & 0x01)
>>> +                             d2 = -d2;
>>> +                     break;
>>> +             case 8:
>> No coment on your code, but this is an 'interesting' bit
>> of bit packing...
>> I did vaguely wonder if this could be mapped to a big lookup table,
>> but having tried it I think I end up with something almost as tricky
>> to read.. Oh well that was a fun 10 minute diversion ;)
> This is inherited code from Graeme - original author of implementation
> for twl6032.
> [...]
>
That bit is the bane of my life :-( I think the data sheet used to
changed weekly as the hardware guys did not understand the difference
between ones and twos complement!

The reason for the crazy packing is there was not enough space allocated
in hardware for the change from 10 to 12 bits for the error values
between 6030 and 6032.

Graeme

^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-17 13:45           ` Oleksandr Kozaruk
  0 siblings, 0 replies; 45+ messages in thread
From: Oleksandr Kozaruk @ 2013-07-17 13:45 UTC (permalink / raw)
  To: Lars-Peter Clausen
  Cc: tony, benoit.cousson, Nayak, Rajendra, Ujfalusi, Peter, ABRAHAM,
	KISHON VIJAY, jic23, grant.likely, rob.herring, sameo, ch.naveen,
	poeschel, Kim, Milo, Krishnamoorthy, Balaji T, gg,
	linux-arm-kernel, linux-kernel, linux-iio, devicetree-discuss,
	linux-omap

On Mon, Jul 15, 2013 at 01:33:53PM +0200, Lars-Peter Clausen wrote:
 > On 07/15/2013 01:09 PM, Kozaruk, Oleksandr wrote:
 > [...]
 > >
 > >>> +     ret = devm_request_threaded_irq(dev, irq, NULL,
 > >>> + twl6030_gpadc_irq_handler,
 > >>> + IRQF_ONESHOT, "twl6030_gpadc", gpadc);
 > >>
 > >> You access memory in the interrupt handler which is freed before 
the interrupt
 > >> handler is freed.
 > > Thanks for pointing this. devm_* will free memory for irq after the 
driver
 > > is removed and memory for the device is freed. I took me awhile to 
understand
 > > this. Is there going to be something like devm_iio_device_alloc? 
whould it be helpfull?
 > >
 >
 > Yes, I think it certainly makes sense to add a 
devm_iio_device_alloc(), care
 > to send a patch?

Anything like this? (of course it's not a patch)

struct iio_dev *devm_iio_device_alloc(struct device *dev, int sizeof_priv)
{
     struct iio_dev *indio_dev;
     size_t alloc_size;

     alloc_size = sizeof(struct iio_dev);
     if (sizeof_priv) {
         alloc_size = ALIGN(alloc_size, IIO_ALIGN);
         alloc_size += sizeof_priv;
     }
     /* ensure 32-byte alignment of whole construct ? */
     alloc_size += IIO_ALIGN - 1;

     indio_dev = devm_kzalloc(dev, alloc_size, GFP_KERNEL);
     if (indio_dev) {
         indio_dev->dev.groups = indio_dev->groups;
         indio_dev->dev.type = &iio_device_type;
         indio_dev->dev.bus = &iio_bus_type;
         device_initialize(&indio_dev->dev);
         dev_set_drvdata(&indio_dev->dev, (void *)indio_dev);
         mutex_init(&indio_dev->mlock);
         mutex_init(&indio_dev->info_exist_lock);
         INIT_LIST_HEAD(&indio_dev->channel_attr_list);

         indio_dev->id = ida_simple_get(&iio_ida, 0, 0, GFP_KERNEL);
         if (indio_dev->id < 0) {
             /* cannot use a dev_err as the name isn't available */
             printk(KERN_ERR "Failed to get id\n");
             kfree(dev);
             return NULL;
         }
         dev_set_name(&indio_dev->dev, "iio:device%d", indio_dev->id);
         INIT_LIST_HEAD(&indio_dev->buffer_list);
     }

     return indio_dev;
}
EXPORT_SYMBOL(devm_iio_device_alloc);

Regards,
OK

^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-17 13:45           ` Oleksandr Kozaruk
  0 siblings, 0 replies; 45+ messages in thread
From: Oleksandr Kozaruk @ 2013-07-17 13:45 UTC (permalink / raw)
  To: Lars-Peter Clausen
  Cc: tony-4v6yS6AI5VpBDgjK7y7TUQ,
	benoit.cousson-QSEj5FYQhm4dnm+yROfE0A, Nayak, Rajendra, Ujfalusi,
	Peter, ABRAHAM, KISHON VIJAY, jic23-KWPb1pKIrIJaa/9Udqfwiw,
	grant.likely-QSEj5FYQhm4dnm+yROfE0A,
	rob.herring-bsGFqQB8/DxBDgjK7y7TUQ, sameo-VuQAYsv1563Yd54FQh9/CA,
	ch.naveen-Sze3O3UU22JBDgjK7y7TUQ,
	poeschel-Xtl8qvBWbHwb1SvskN2V4Q, Kim, Milo, Krishnamoorthy,
	Balaji T, gg-kDsPt+C1G03kYMGBc/C6ZA,
	linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA,
	linux-iio-u79uwXL29TY76Z2rM5mHXA,
	devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ@public.gmane.org

On Mon, Jul 15, 2013 at 01:33:53PM +0200, Lars-Peter Clausen wrote:
 > On 07/15/2013 01:09 PM, Kozaruk, Oleksandr wrote:
 > [...]
 > >
 > >>> +     ret = devm_request_threaded_irq(dev, irq, NULL,
 > >>> + twl6030_gpadc_irq_handler,
 > >>> + IRQF_ONESHOT, "twl6030_gpadc", gpadc);
 > >>
 > >> You access memory in the interrupt handler which is freed before 
the interrupt
 > >> handler is freed.
 > > Thanks for pointing this. devm_* will free memory for irq after the 
driver
 > > is removed and memory for the device is freed. I took me awhile to 
understand
 > > this. Is there going to be something like devm_iio_device_alloc? 
whould it be helpfull?
 > >
 >
 > Yes, I think it certainly makes sense to add a 
devm_iio_device_alloc(), care
 > to send a patch?

Anything like this? (of course it's not a patch)

struct iio_dev *devm_iio_device_alloc(struct device *dev, int sizeof_priv)
{
     struct iio_dev *indio_dev;
     size_t alloc_size;

     alloc_size = sizeof(struct iio_dev);
     if (sizeof_priv) {
         alloc_size = ALIGN(alloc_size, IIO_ALIGN);
         alloc_size += sizeof_priv;
     }
     /* ensure 32-byte alignment of whole construct ? */
     alloc_size += IIO_ALIGN - 1;

     indio_dev = devm_kzalloc(dev, alloc_size, GFP_KERNEL);
     if (indio_dev) {
         indio_dev->dev.groups = indio_dev->groups;
         indio_dev->dev.type = &iio_device_type;
         indio_dev->dev.bus = &iio_bus_type;
         device_initialize(&indio_dev->dev);
         dev_set_drvdata(&indio_dev->dev, (void *)indio_dev);
         mutex_init(&indio_dev->mlock);
         mutex_init(&indio_dev->info_exist_lock);
         INIT_LIST_HEAD(&indio_dev->channel_attr_list);

         indio_dev->id = ida_simple_get(&iio_ida, 0, 0, GFP_KERNEL);
         if (indio_dev->id < 0) {
             /* cannot use a dev_err as the name isn't available */
             printk(KERN_ERR "Failed to get id\n");
             kfree(dev);
             return NULL;
         }
         dev_set_name(&indio_dev->dev, "iio:device%d", indio_dev->id);
         INIT_LIST_HEAD(&indio_dev->buffer_list);
     }

     return indio_dev;
}
EXPORT_SYMBOL(devm_iio_device_alloc);

Regards,
OK

^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-17 13:45           ` Oleksandr Kozaruk
  0 siblings, 0 replies; 45+ messages in thread
From: Oleksandr Kozaruk @ 2013-07-17 13:45 UTC (permalink / raw)
  To: Lars-Peter Clausen
  Cc: tony, benoit.cousson, Nayak, Rajendra, Ujfalusi, Peter, ABRAHAM,
	KISHON VIJAY, jic23, grant.likely, rob.herring, sameo, ch.naveen,
	poeschel, Kim, Milo, Krishnamoorthy, Balaji T, gg,
	linux-arm-kernel, linux-kernel, linux-iio, devicetree-discuss,
	linux-omap

On Mon, Jul 15, 2013 at 01:33:53PM +0200, Lars-Peter Clausen wrote:
 > On 07/15/2013 01:09 PM, Kozaruk, Oleksandr wrote:
 > [...]
 > >
 > >>> +     ret = devm_request_threaded_irq(dev, irq, NULL,
 > >>> + twl6030_gpadc_irq_handler,
 > >>> + IRQF_ONESHOT, "twl6030_gpadc", gpadc);
 > >>
 > >> You access memory in the interrupt handler which is freed before 
the interrupt
 > >> handler is freed.
 > > Thanks for pointing this. devm_* will free memory for irq after the 
driver
 > > is removed and memory for the device is freed. I took me awhile to 
understand
 > > this. Is there going to be something like devm_iio_device_alloc? 
whould it be helpfull?
 > >
 >
 > Yes, I think it certainly makes sense to add a 
devm_iio_device_alloc(), care
 > to send a patch?

Anything like this? (of course it's not a patch)

struct iio_dev *devm_iio_device_alloc(struct device *dev, int sizeof_priv)
{
     struct iio_dev *indio_dev;
     size_t alloc_size;

     alloc_size = sizeof(struct iio_dev);
     if (sizeof_priv) {
         alloc_size = ALIGN(alloc_size, IIO_ALIGN);
         alloc_size += sizeof_priv;
     }
     /* ensure 32-byte alignment of whole construct ? */
     alloc_size += IIO_ALIGN - 1;

     indio_dev = devm_kzalloc(dev, alloc_size, GFP_KERNEL);
     if (indio_dev) {
         indio_dev->dev.groups = indio_dev->groups;
         indio_dev->dev.type = &iio_device_type;
         indio_dev->dev.bus = &iio_bus_type;
         device_initialize(&indio_dev->dev);
         dev_set_drvdata(&indio_dev->dev, (void *)indio_dev);
         mutex_init(&indio_dev->mlock);
         mutex_init(&indio_dev->info_exist_lock);
         INIT_LIST_HEAD(&indio_dev->channel_attr_list);

         indio_dev->id = ida_simple_get(&iio_ida, 0, 0, GFP_KERNEL);
         if (indio_dev->id < 0) {
             /* cannot use a dev_err as the name isn't available */
             printk(KERN_ERR "Failed to get id\n");
             kfree(dev);
             return NULL;
         }
         dev_set_name(&indio_dev->dev, "iio:device%d", indio_dev->id);
         INIT_LIST_HEAD(&indio_dev->buffer_list);
     }

     return indio_dev;
}
EXPORT_SYMBOL(devm_iio_device_alloc);

Regards,
OK

^ permalink raw reply	[flat|nested] 45+ messages in thread

* [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-17 13:45           ` Oleksandr Kozaruk
  0 siblings, 0 replies; 45+ messages in thread
From: Oleksandr Kozaruk @ 2013-07-17 13:45 UTC (permalink / raw)
  To: linux-arm-kernel

On Mon, Jul 15, 2013 at 01:33:53PM +0200, Lars-Peter Clausen wrote:
 > On 07/15/2013 01:09 PM, Kozaruk, Oleksandr wrote:
 > [...]
 > >
 > >>> +     ret = devm_request_threaded_irq(dev, irq, NULL,
 > >>> + twl6030_gpadc_irq_handler,
 > >>> + IRQF_ONESHOT, "twl6030_gpadc", gpadc);
 > >>
 > >> You access memory in the interrupt handler which is freed before 
the interrupt
 > >> handler is freed.
 > > Thanks for pointing this. devm_* will free memory for irq after the 
driver
 > > is removed and memory for the device is freed. I took me awhile to 
understand
 > > this. Is there going to be something like devm_iio_device_alloc? 
whould it be helpfull?
 > >
 >
 > Yes, I think it certainly makes sense to add a 
devm_iio_device_alloc(), care
 > to send a patch?

Anything like this? (of course it's not a patch)

struct iio_dev *devm_iio_device_alloc(struct device *dev, int sizeof_priv)
{
     struct iio_dev *indio_dev;
     size_t alloc_size;

     alloc_size = sizeof(struct iio_dev);
     if (sizeof_priv) {
         alloc_size = ALIGN(alloc_size, IIO_ALIGN);
         alloc_size += sizeof_priv;
     }
     /* ensure 32-byte alignment of whole construct ? */
     alloc_size += IIO_ALIGN - 1;

     indio_dev = devm_kzalloc(dev, alloc_size, GFP_KERNEL);
     if (indio_dev) {
         indio_dev->dev.groups = indio_dev->groups;
         indio_dev->dev.type = &iio_device_type;
         indio_dev->dev.bus = &iio_bus_type;
         device_initialize(&indio_dev->dev);
         dev_set_drvdata(&indio_dev->dev, (void *)indio_dev);
         mutex_init(&indio_dev->mlock);
         mutex_init(&indio_dev->info_exist_lock);
         INIT_LIST_HEAD(&indio_dev->channel_attr_list);

         indio_dev->id = ida_simple_get(&iio_ida, 0, 0, GFP_KERNEL);
         if (indio_dev->id < 0) {
             /* cannot use a dev_err as the name isn't available */
             printk(KERN_ERR "Failed to get id\n");
             kfree(dev);
             return NULL;
         }
         dev_set_name(&indio_dev->dev, "iio:device%d", indio_dev->id);
         INIT_LIST_HEAD(&indio_dev->buffer_list);
     }

     return indio_dev;
}
EXPORT_SYMBOL(devm_iio_device_alloc);

Regards,
OK

^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
  2013-07-17 13:45           ` Oleksandr Kozaruk
  (?)
  (?)
@ 2013-07-17 14:47             ` Lars-Peter Clausen
  -1 siblings, 0 replies; 45+ messages in thread
From: Lars-Peter Clausen @ 2013-07-17 14:47 UTC (permalink / raw)
  To: Oleksandr Kozaruk
  Cc: tony, benoit.cousson, Nayak, Rajendra, Ujfalusi, Peter, ABRAHAM,
	KISHON VIJAY, jic23, grant.likely, rob.herring, sameo, ch.naveen,
	poeschel, Kim, Milo, Krishnamoorthy, Balaji T, gg,
	linux-arm-kernel, linux-kernel, linux-iio, devicetree-discuss,
	linux-omap

On 07/17/2013 03:45 PM, Oleksandr Kozaruk wrote:
> On Mon, Jul 15, 2013 at 01:33:53PM +0200, Lars-Peter Clausen wrote:
>> On 07/15/2013 01:09 PM, Kozaruk, Oleksandr wrote:
>> [...]
>> >
>> >>> +     ret = devm_request_threaded_irq(dev, irq, NULL,
>> >>> + twl6030_gpadc_irq_handler,
>> >>> + IRQF_ONESHOT, "twl6030_gpadc", gpadc);
>> >>
>> >> You access memory in the interrupt handler which is freed before the
> interrupt
>> >> handler is freed.
>> > Thanks for pointing this. devm_* will free memory for irq after the driver
>> > is removed and memory for the device is freed. I took me awhile to
> understand
>> > this. Is there going to be something like devm_iio_device_alloc? whould
> it be helpfull?
>> >
>>
>> Yes, I think it certainly makes sense to add a devm_iio_device_alloc(), care
>> to send a patch?
> 
> Anything like this? (of course it's not a patch)
> 

No. I think you can for example use devm_regulator_get() as a template. But
instead of regulator_get() and regulator_put() use iio_device_alloc() and
iio_device_free().

> struct iio_dev *devm_iio_device_alloc(struct device *dev, int sizeof_priv)
> {
>     struct iio_dev *indio_dev;
>     size_t alloc_size;
> 
>     alloc_size = sizeof(struct iio_dev);
>     if (sizeof_priv) {
>         alloc_size = ALIGN(alloc_size, IIO_ALIGN);
>         alloc_size += sizeof_priv;
>     }
>     /* ensure 32-byte alignment of whole construct ? */
>     alloc_size += IIO_ALIGN - 1;
> 
>     indio_dev = devm_kzalloc(dev, alloc_size, GFP_KERNEL);
>     if (indio_dev) {
>         indio_dev->dev.groups = indio_dev->groups;
>         indio_dev->dev.type = &iio_device_type;
>         indio_dev->dev.bus = &iio_bus_type;
>         device_initialize(&indio_dev->dev);
>         dev_set_drvdata(&indio_dev->dev, (void *)indio_dev);
>         mutex_init(&indio_dev->mlock);
>         mutex_init(&indio_dev->info_exist_lock);
>         INIT_LIST_HEAD(&indio_dev->channel_attr_list);
> 
>         indio_dev->id = ida_simple_get(&iio_ida, 0, 0, GFP_KERNEL);
>         if (indio_dev->id < 0) {
>             /* cannot use a dev_err as the name isn't available */
>             printk(KERN_ERR "Failed to get id\n");
>             kfree(dev);
>             return NULL;
>         }
>         dev_set_name(&indio_dev->dev, "iio:device%d", indio_dev->id);
>         INIT_LIST_HEAD(&indio_dev->buffer_list);
>     }
> 
>     return indio_dev;
> }
> EXPORT_SYMBOL(devm_iio_device_alloc);
> 
> Regards,
> OK


^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-17 14:47             ` Lars-Peter Clausen
  0 siblings, 0 replies; 45+ messages in thread
From: Lars-Peter Clausen @ 2013-07-17 14:47 UTC (permalink / raw)
  To: Oleksandr Kozaruk
  Cc: tony, benoit.cousson, Nayak, Rajendra, Ujfalusi, Peter, ABRAHAM,
	KISHON VIJAY, jic23, grant.likely, rob.herring, sameo, ch.naveen,
	poeschel, Kim, Milo, Krishnamoorthy, Balaji T, gg,
	linux-arm-kernel, linux-kernel, linux-iio,
	devicetree-discuss@lists.ozlabs.org

On 07/17/2013 03:45 PM, Oleksandr Kozaruk wrote:
> On Mon, Jul 15, 2013 at 01:33:53PM +0200, Lars-Peter Clausen wrote:
>> On 07/15/2013 01:09 PM, Kozaruk, Oleksandr wrote:
>> [...]
>> >
>> >>> +     ret = devm_request_threaded_irq(dev, irq, NULL,
>> >>> + twl6030_gpadc_irq_handler,
>> >>> + IRQF_ONESHOT, "twl6030_gpadc", gpadc);
>> >>
>> >> You access memory in the interrupt handler which is freed before the
> interrupt
>> >> handler is freed.
>> > Thanks for pointing this. devm_* will free memory for irq after the driver
>> > is removed and memory for the device is freed. I took me awhile to
> understand
>> > this. Is there going to be something like devm_iio_device_alloc? whould
> it be helpfull?
>> >
>>
>> Yes, I think it certainly makes sense to add a devm_iio_device_alloc(), care
>> to send a patch?
> 
> Anything like this? (of course it's not a patch)
> 

No. I think you can for example use devm_regulator_get() as a template. But
instead of regulator_get() and regulator_put() use iio_device_alloc() and
iio_device_free().

> struct iio_dev *devm_iio_device_alloc(struct device *dev, int sizeof_priv)
> {
>     struct iio_dev *indio_dev;
>     size_t alloc_size;
> 
>     alloc_size = sizeof(struct iio_dev);
>     if (sizeof_priv) {
>         alloc_size = ALIGN(alloc_size, IIO_ALIGN);
>         alloc_size += sizeof_priv;
>     }
>     /* ensure 32-byte alignment of whole construct ? */
>     alloc_size += IIO_ALIGN - 1;
> 
>     indio_dev = devm_kzalloc(dev, alloc_size, GFP_KERNEL);
>     if (indio_dev) {
>         indio_dev->dev.groups = indio_dev->groups;
>         indio_dev->dev.type = &iio_device_type;
>         indio_dev->dev.bus = &iio_bus_type;
>         device_initialize(&indio_dev->dev);
>         dev_set_drvdata(&indio_dev->dev, (void *)indio_dev);
>         mutex_init(&indio_dev->mlock);
>         mutex_init(&indio_dev->info_exist_lock);
>         INIT_LIST_HEAD(&indio_dev->channel_attr_list);
> 
>         indio_dev->id = ida_simple_get(&iio_ida, 0, 0, GFP_KERNEL);
>         if (indio_dev->id < 0) {
>             /* cannot use a dev_err as the name isn't available */
>             printk(KERN_ERR "Failed to get id\n");
>             kfree(dev);
>             return NULL;
>         }
>         dev_set_name(&indio_dev->dev, "iio:device%d", indio_dev->id);
>         INIT_LIST_HEAD(&indio_dev->buffer_list);
>     }
> 
>     return indio_dev;
> }
> EXPORT_SYMBOL(devm_iio_device_alloc);
> 
> Regards,
> OK

^ permalink raw reply	[flat|nested] 45+ messages in thread

* Re: [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-17 14:47             ` Lars-Peter Clausen
  0 siblings, 0 replies; 45+ messages in thread
From: Lars-Peter Clausen @ 2013-07-17 14:47 UTC (permalink / raw)
  To: Oleksandr Kozaruk
  Cc: tony, benoit.cousson, Nayak, Rajendra, Ujfalusi, Peter, ABRAHAM,
	KISHON VIJAY, jic23, grant.likely, rob.herring, sameo, ch.naveen,
	poeschel, Kim, Milo, Krishnamoorthy, Balaji T, gg,
	linux-arm-kernel, linux-kernel, linux-iio, devicetree-discuss,
	linux-omap

On 07/17/2013 03:45 PM, Oleksandr Kozaruk wrote:
> On Mon, Jul 15, 2013 at 01:33:53PM +0200, Lars-Peter Clausen wrote:
>> On 07/15/2013 01:09 PM, Kozaruk, Oleksandr wrote:
>> [...]
>> >
>> >>> +     ret = devm_request_threaded_irq(dev, irq, NULL,
>> >>> + twl6030_gpadc_irq_handler,
>> >>> + IRQF_ONESHOT, "twl6030_gpadc", gpadc);
>> >>
>> >> You access memory in the interrupt handler which is freed before the
> interrupt
>> >> handler is freed.
>> > Thanks for pointing this. devm_* will free memory for irq after the driver
>> > is removed and memory for the device is freed. I took me awhile to
> understand
>> > this. Is there going to be something like devm_iio_device_alloc? whould
> it be helpfull?
>> >
>>
>> Yes, I think it certainly makes sense to add a devm_iio_device_alloc(), care
>> to send a patch?
> 
> Anything like this? (of course it's not a patch)
> 

No. I think you can for example use devm_regulator_get() as a template. But
instead of regulator_get() and regulator_put() use iio_device_alloc() and
iio_device_free().

> struct iio_dev *devm_iio_device_alloc(struct device *dev, int sizeof_priv)
> {
>     struct iio_dev *indio_dev;
>     size_t alloc_size;
> 
>     alloc_size = sizeof(struct iio_dev);
>     if (sizeof_priv) {
>         alloc_size = ALIGN(alloc_size, IIO_ALIGN);
>         alloc_size += sizeof_priv;
>     }
>     /* ensure 32-byte alignment of whole construct ? */
>     alloc_size += IIO_ALIGN - 1;
> 
>     indio_dev = devm_kzalloc(dev, alloc_size, GFP_KERNEL);
>     if (indio_dev) {
>         indio_dev->dev.groups = indio_dev->groups;
>         indio_dev->dev.type = &iio_device_type;
>         indio_dev->dev.bus = &iio_bus_type;
>         device_initialize(&indio_dev->dev);
>         dev_set_drvdata(&indio_dev->dev, (void *)indio_dev);
>         mutex_init(&indio_dev->mlock);
>         mutex_init(&indio_dev->info_exist_lock);
>         INIT_LIST_HEAD(&indio_dev->channel_attr_list);
> 
>         indio_dev->id = ida_simple_get(&iio_ida, 0, 0, GFP_KERNEL);
>         if (indio_dev->id < 0) {
>             /* cannot use a dev_err as the name isn't available */
>             printk(KERN_ERR "Failed to get id\n");
>             kfree(dev);
>             return NULL;
>         }
>         dev_set_name(&indio_dev->dev, "iio:device%d", indio_dev->id);
>         INIT_LIST_HEAD(&indio_dev->buffer_list);
>     }
> 
>     return indio_dev;
> }
> EXPORT_SYMBOL(devm_iio_device_alloc);
> 
> Regards,
> OK


^ permalink raw reply	[flat|nested] 45+ messages in thread

* [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver
@ 2013-07-17 14:47             ` Lars-Peter Clausen
  0 siblings, 0 replies; 45+ messages in thread
From: Lars-Peter Clausen @ 2013-07-17 14:47 UTC (permalink / raw)
  To: linux-arm-kernel

On 07/17/2013 03:45 PM, Oleksandr Kozaruk wrote:
> On Mon, Jul 15, 2013 at 01:33:53PM +0200, Lars-Peter Clausen wrote:
>> On 07/15/2013 01:09 PM, Kozaruk, Oleksandr wrote:
>> [...]
>> >
>> >>> +     ret = devm_request_threaded_irq(dev, irq, NULL,
>> >>> + twl6030_gpadc_irq_handler,
>> >>> + IRQF_ONESHOT, "twl6030_gpadc", gpadc);
>> >>
>> >> You access memory in the interrupt handler which is freed before the
> interrupt
>> >> handler is freed.
>> > Thanks for pointing this. devm_* will free memory for irq after the driver
>> > is removed and memory for the device is freed. I took me awhile to
> understand
>> > this. Is there going to be something like devm_iio_device_alloc? whould
> it be helpfull?
>> >
>>
>> Yes, I think it certainly makes sense to add a devm_iio_device_alloc(), care
>> to send a patch?
> 
> Anything like this? (of course it's not a patch)
> 

No. I think you can for example use devm_regulator_get() as a template. But
instead of regulator_get() and regulator_put() use iio_device_alloc() and
iio_device_free().

> struct iio_dev *devm_iio_device_alloc(struct device *dev, int sizeof_priv)
> {
>     struct iio_dev *indio_dev;
>     size_t alloc_size;
> 
>     alloc_size = sizeof(struct iio_dev);
>     if (sizeof_priv) {
>         alloc_size = ALIGN(alloc_size, IIO_ALIGN);
>         alloc_size += sizeof_priv;
>     }
>     /* ensure 32-byte alignment of whole construct ? */
>     alloc_size += IIO_ALIGN - 1;
> 
>     indio_dev = devm_kzalloc(dev, alloc_size, GFP_KERNEL);
>     if (indio_dev) {
>         indio_dev->dev.groups = indio_dev->groups;
>         indio_dev->dev.type = &iio_device_type;
>         indio_dev->dev.bus = &iio_bus_type;
>         device_initialize(&indio_dev->dev);
>         dev_set_drvdata(&indio_dev->dev, (void *)indio_dev);
>         mutex_init(&indio_dev->mlock);
>         mutex_init(&indio_dev->info_exist_lock);
>         INIT_LIST_HEAD(&indio_dev->channel_attr_list);
> 
>         indio_dev->id = ida_simple_get(&iio_ida, 0, 0, GFP_KERNEL);
>         if (indio_dev->id < 0) {
>             /* cannot use a dev_err as the name isn't available */
>             printk(KERN_ERR "Failed to get id\n");
>             kfree(dev);
>             return NULL;
>         }
>         dev_set_name(&indio_dev->dev, "iio:device%d", indio_dev->id);
>         INIT_LIST_HEAD(&indio_dev->buffer_list);
>     }
> 
>     return indio_dev;
> }
> EXPORT_SYMBOL(devm_iio_device_alloc);
> 
> Regards,
> OK

^ permalink raw reply	[flat|nested] 45+ messages in thread

end of thread, other threads:[~2013-07-17 14:47 UTC | newest]

Thread overview: 45+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-07-12  7:18 [PATCH v3 0/2] TWL6030, TWL6032 GPADC driver Oleksandr Kozaruk
2013-07-12  7:18 ` Oleksandr Kozaruk
2013-07-12  7:18 ` Oleksandr Kozaruk
2013-07-12  7:18 ` [PATCH v3 1/2] ARM: dts: twl: Add GPADC data to device tree Oleksandr Kozaruk
2013-07-12  7:18   ` Oleksandr Kozaruk
2013-07-12  7:18   ` Oleksandr Kozaruk
2013-07-12  7:18 ` [PATCH v3 2/2] iio: twl6030-gpadc: TWL6030, TWL6032 GPADC driver Oleksandr Kozaruk
2013-07-12  7:18   ` Oleksandr Kozaruk
2013-07-12  7:18   ` Oleksandr Kozaruk
2013-07-12 19:42   ` Jonathan Cameron
2013-07-12 19:42     ` Jonathan Cameron
2013-07-12 19:42     ` Jonathan Cameron
2013-07-15 13:30     ` Kozaruk, Oleksandr
2013-07-15 13:30       ` Kozaruk, Oleksandr
2013-07-15 13:30       ` Kozaruk, Oleksandr
2013-07-15 13:30       ` Kozaruk, Oleksandr
2013-07-15 14:05       ` Graeme Gregory
2013-07-15 14:05         ` Graeme Gregory
2013-07-15 14:05         ` Graeme Gregory
2013-07-15 14:05         ` Graeme Gregory
2013-07-12 19:56   ` Lars-Peter Clausen
2013-07-12 19:56     ` Lars-Peter Clausen
2013-07-12 19:56     ` Lars-Peter Clausen
2013-07-15 11:09     ` Kozaruk, Oleksandr
2013-07-15 11:09       ` Kozaruk, Oleksandr
2013-07-15 11:09       ` Kozaruk, Oleksandr
2013-07-15 11:09       ` Kozaruk, Oleksandr
2013-07-15 11:33       ` Lars-Peter Clausen
2013-07-15 11:33         ` Lars-Peter Clausen
2013-07-15 11:33         ` Lars-Peter Clausen
2013-07-15 11:33         ` Lars-Peter Clausen
2013-07-17 13:45         ` Oleksandr Kozaruk
2013-07-17 13:45           ` Oleksandr Kozaruk
2013-07-17 13:45           ` Oleksandr Kozaruk
2013-07-17 13:45           ` Oleksandr Kozaruk
2013-07-17 14:47           ` Lars-Peter Clausen
2013-07-17 14:47             ` Lars-Peter Clausen
2013-07-17 14:47             ` Lars-Peter Clausen
2013-07-17 14:47             ` Lars-Peter Clausen
2013-07-15 11:56     ` Grygorii Strashko
2013-07-15 11:56       ` Grygorii Strashko
2013-07-15 11:56       ` Grygorii Strashko
2013-07-15 12:33       ` Lars-Peter Clausen
2013-07-15 12:33         ` Lars-Peter Clausen
2013-07-15 12:33         ` Lars-Peter Clausen

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.