Skip to contentSkip to conversation

tnetv107x ssp driver stack

PATCH Superseded Cyril Chemparathy 3 participants 21 messages Oct 14, 2010 – Oct 18, 2010 8 messages not in the archive

Subsystems (9)

v1 of 8 Superseded Latest is v8 →

lore
Share & tools
lore
#0 lore raw

cover tnetv107x ssp driver stack

TI's sequencer serial port (TI-SSP) is a jack-of-all-trades type of serial port device. It has a built-in programmable execution engine that can be programmed to operate as almost any serial bus (I2C, SPI, EasyScale, and others).

This patch series implements a driver stack that looks like the following:

                             +--------+
                             | eeprom | . . .
                             +--------+
   +-----------+             +--------------+ +---------+
   | regulator | . . .       |   i2c-gpio   | | 1-wire  | . . .
   +-----------+             +--------------+ +---------+
   +----------------------+  +--------------------------------+
   |       ssp-spi        |  |           ssp-gpio             |
   +----------------------+  +--------------------------------+
   +----------------------------------------------------------+
   |                            ssp                           |
   +----------------------------------------------------------+

Cyril Chemparathy (12):
misc: add driver for sequencer serial port
davinci: add tnetv107x ssp platform device
davinci: add ssp config for tnetv107x evm board
spi: add ti-ssp spi master driver
davinci: add spi devices on tnetv107x evm
regulator: add driver for tps6524x regulator
davinci: add tnetv107x evm regulators
gpio: add ti-ssp virtual gpio driver
davinci: add tnetv107x evm ti-ssp gpio device
backlight: add support for tps6116x controller
davinci: add tnetv107x evm backlight device
davinci: add tnetv107x evm i2c eeprom device

 arch/arm/mach-davinci/board-tnetv107x-evm.c    |  211 ++++++
 arch/arm/mach-davinci/devices-tnetv107x.c      |   25 +
 arch/arm/mach-davinci/include/mach/ti_ssp.h    |  103 +++
 arch/arm/mach-davinci/include/mach/tnetv107x.h |    2 +
 arch/arm/mach-davinci/tnetv107x.c              |    2 +-
 drivers/gpio/Kconfig                           |   10 +
 drivers/gpio/Makefile                          |    1 +
 drivers/gpio/ti-ssp-gpio.c                     |  234 +++++++
 drivers/misc/Kconfig                           |   11 +
 drivers/misc/Makefile                          |    1 +
 drivers/misc/ti_ssp.c                          |  443 +++++++++++++
 drivers/regulator/Kconfig                      |   10 +
 drivers/regulator/Makefile                     |    1 +
 drivers/regulator/tps6524x-regulator.c         |  812 ++++++++++++++++++++++++
 drivers/spi/Kconfig                            |    6 +
 drivers/spi/Makefile                           |    1 +
 drivers/spi/spi_ti_ssp.c                       |  408 ++++++++++++
 drivers/video/backlight/Kconfig                |    7 +
 drivers/video/backlight/Makefile               |    2 +-
 drivers/video/backlight/tps6116x.c             |  339 ++++++++++
 20 files changed, 2627 insertions(+), 2 deletions(-)
 create mode 100644 arch/arm/mach-davinci/include/mach/ti_ssp.h
 create mode 100644 drivers/gpio/ti-ssp-gpio.c
 create mode 100644 drivers/misc/ti_ssp.c
 create mode 100644 drivers/regulator/tps6524x-regulator.c
 create mode 100644 drivers/spi/spi_ti_ssp.c
 create mode 100644 drivers/video/backlight/tps6116x.c

A message that is not in the archive (sent off-list or not yet received): 1287081535-2864-1-git-send-email-cyril-l0cyMroinI0@public.gmane.org

#2 lore raw

PATCH 1/12 misc: add driver for sequencer serial port

TI's sequencer serial port (TI-SSP) is a jack-of-all-trades type of serial port device. It has a built-in programmable execution engine that can be programmed to operate as almost any serial bus (I2C, SPI, EasyScale, and others).

This patch adds a driver for this controller device. The driver does not expose a user-land interface. Protocol drivers built on top of this layer are expected to remain in-kernel.

---
 arch/arm/mach-davinci/include/mach/ti_ssp.h |   92 ++++++
 drivers/misc/Kconfig                        |   11 +
 drivers/misc/Makefile                       |    1 +
 drivers/misc/ti_ssp.c                       |  443 +++++++++++++++++++++++++++
 4 files changed, 547 insertions(+), 0 deletions(-)
 create mode 100644 arch/arm/mach-davinci/include/mach/ti_ssp.h
 create mode 100644 drivers/misc/ti_ssp.c
arch/arm/mach-davinci/include/mach/ti_ssp.h new file +92 -0
new file mode 100644
index 0000000..8365101
@@ -0,0 +1,92 @@
+/*
+ * Sequencer Serial Port (SSP) driver for Texas Instruments' SoCs
+ *
+ * Copyright (C) 2010 Texas Instruments Inc
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * 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., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#ifndef __TI_SSP_H__
+#define __TI_SSP_H__
+
+struct ti_ssp_data {
+	unsigned long		out_clock;
+};
+
+struct ti_ssp_port_data {
+	const char	*ssp_dev_name;
+	int		port;
+	unsigned long	iosel; /* see note below */
+	unsigned long	config;
+};
+
+struct ti_ssp_port;
+
+/*
+ * Sequencer port IO pin configuration bits.  These do not correlate 1-1 with
+ * the hardware.  The iosel field in the port data combines iosel1 and iosel2,
+ * and is therefore not a direct map to register space.  It is best to use the
+ * macros below to construct iosel values.
+ *
+ * least significant 16 bits --> iosel1
+ * most significant 16 bits  --> iosel2
+ */
+
+#define SSP_IN			0x0000
+#define SSP_DATA		0x0001
+#define SSP_CLOCK		0x0002
+#define SSP_CHIPSEL		0x0003
+#define SSP_OUT			0x0004
+#define SSP_PIN_SEL(pin, v)	((v) << ((pin) * 3))
+#define SSP_PIN_MASK(pin)	SSP_PIN_SEL(pin, 0x7)
+#define SSP_INPUT_SEL(pin)	((pin) << 16)
+
+/* Sequencer port config bits */
+#define SSP_EARLY_DIN		BIT(8)
+#define SSP_DELAY_DOUT		BIT(9)
+
+/* Sequence map definitions */
+#define SSP_CLK_HIGH		BIT(0)
+#define SSP_CLK_LOW		0
+#define SSP_DATA_HIGH		BIT(1)
+#define SSP_DATA_LOW		0
+#define SSP_CS_HIGH		BIT(2)
+#define SSP_CS_LOW		0
+#define SSP_OUT_MODE		BIT(3)
+#define SSP_IN_MODE		0
+#define SSP_DATA_REG		BIT(4)
+#define SSP_ADDR_REG		0
+
+#define SSP_OPCODE_DIRECT	((0x0) << 5)
+#define SSP_OPCODE_TOGGLE	((0x1) << 5)
+#define SSP_OPCODE_SHIFT	((0x2) << 5)
+#define SSP_OPCODE_BRANCH0	((0x4) << 5)
+#define SSP_OPCODE_BRANCH1	((0x5) << 5)
+#define SSP_OPCODE_BRANCH	((0x6) << 5)
+#define SSP_OPCODE_STOP		((0x7) << 5)
+#define SSP_BRANCH(addr)	((addr) << 8)
+#define SSP_COUNT(cycles)	((cycles) << 8)
+
+struct ti_ssp_port *ti_ssp_open(const struct ti_ssp_port_data *data);
+int ti_ssp_close(struct ti_ssp_port *dev);
+int ti_ssp_dumpregs(struct ti_ssp_port *dev);
+int ti_ssp_raw_read(struct ti_ssp_port *dev);
+int ti_ssp_raw_write(struct ti_ssp_port *dev, u32 val);
+int ti_ssp_load(struct ti_ssp_port *dev, int offs, u32* prog, int len);
+int ti_ssp_run(struct ti_ssp_port *dev, u32 pc, u32 input, u32 *output);
+int ti_ssp_set_mode(struct ti_ssp_port *dev, int mode);
+int ti_ssp_set_iosel(struct ti_ssp_port *dev, u32 iosel);
+
+#endif /* __TI_SSP_H__ */
drivers/misc/Kconfig +11 -0
index b743312..9fb8470 100644
@@ -390,6 +390,17 @@ config BMP085
 	  To compile this driver as a module, choose M here: the
 	  module will be called bmp085.
 
+config TI_SSP
+	depends on ARCH_DAVINCI_TNETV107X
+	tristate "Sequencer Serial Port support"
+	default y
+	---help---
+	  Say Y here if you want support for the Sequencer Serial Port
+	  in a Texas Instruments TNETV107X SoC.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called ti_ssp.
+
 source "drivers/misc/c2port/Kconfig"
 source "drivers/misc/eeprom/Kconfig"
 source "drivers/misc/cb710/Kconfig"
drivers/misc/Makefile +1 -0
index 42eab95..7568100 100644
@@ -28,6 +28,7 @@ obj-$(CONFIG_SENSORS_TSL2550)	+= tsl2550.o
 obj-$(CONFIG_EP93XX_PWM)	+= ep93xx_pwm.o
 obj-$(CONFIG_DS1682)		+= ds1682.o
 obj-$(CONFIG_TI_DAC7512)	+= ti_dac7512.o
+obj-$(CONFIG_TI_SSP)		+= ti_ssp.o
 obj-$(CONFIG_C2PORT)		+= c2port/
 obj-$(CONFIG_IWMC3200TOP)      += iwmc3200top/
 obj-$(CONFIG_HMC6352)		+= hmc6352.o
drivers/misc/ti_ssp.c new file +443 -0

443 lines not on this page · raw message

Signature · 1 line
1.7.0.4

A message that is not in the archive (sent off-list or not yet received): 1287081535-2864-2-git-send-email-cyril-l0cyMroinI0@public.gmane.org

#4 lore raw

Re: [PATCH 01/12] misc: add driver for sequencer serial port

Quoted 9 lines · Cyril Chemparathy

TI's sequencer serial port (TI-SSP) is a jack-of-all-trades type of serial port device. It has a built-in programmable execution engine that can be programmed to operate as almost any serial bus (I2C, SPI, EasyScale, and others).

This patch adds a driver for this controller device. The driver does not expose a user-land interface. Protocol drivers built on top of this layer are expected to remain in-kernel.

Hi Cyril,

Overall the driver looks pretty clean, but I have an issue with the usage model (see below).

Quoted 229 lines
---
 arch/arm/mach-davinci/include/mach/ti_ssp.h |   92 ++++++
 drivers/misc/Kconfig                        |   11 +
 drivers/misc/Makefile                       |    1 +
 drivers/misc/ti_ssp.c                       |  443 +++++++++++++++++++++++++++
 4 files changed, 547 insertions(+), 0 deletions(-)
 create mode 100644 arch/arm/mach-davinci/include/mach/ti_ssp.h
 create mode 100644 drivers/misc/ti_ssp.c
arch/arm/mach-davinci/include/mach/ti_ssp.h new file +92 -0
new file mode 100644
index 0000000..8365101
@@ -0,0 +1,92 @@
+/*
+ * Sequencer Serial Port (SSP) driver for Texas Instruments' SoCs
+ *
+ * Copyright (C) 2010 Texas Instruments Inc
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * 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., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#ifndef __TI_SSP_H__
+#define __TI_SSP_H__
+
+struct ti_ssp_data {
+	unsigned long		out_clock;
+};
+
+struct ti_ssp_port_data {
+	const char	*ssp_dev_name;
+	int		port;
+	unsigned long	iosel; /* see note below */
+	unsigned long	config;
+};
+
+struct ti_ssp_port;
+
+/*
+ * Sequencer port IO pin configuration bits.  These do not correlate 1-1 with
+ * the hardware.  The iosel field in the port data combines iosel1 and iosel2,
+ * and is therefore not a direct map to register space.  It is best to use the
+ * macros below to construct iosel values.
+ *
+ * least significant 16 bits --> iosel1
+ * most significant 16 bits  --> iosel2
+ */
+
+#define SSP_IN			0x0000
+#define SSP_DATA		0x0001
+#define SSP_CLOCK		0x0002
+#define SSP_CHIPSEL		0x0003
+#define SSP_OUT			0x0004
+#define SSP_PIN_SEL(pin, v)	((v) << ((pin) * 3))
+#define SSP_PIN_MASK(pin)	SSP_PIN_SEL(pin, 0x7)
+#define SSP_INPUT_SEL(pin)	((pin) << 16)
+
+/* Sequencer port config bits */
+#define SSP_EARLY_DIN		BIT(8)
+#define SSP_DELAY_DOUT		BIT(9)
+
+/* Sequence map definitions */
+#define SSP_CLK_HIGH		BIT(0)
+#define SSP_CLK_LOW		0
+#define SSP_DATA_HIGH		BIT(1)
+#define SSP_DATA_LOW		0
+#define SSP_CS_HIGH		BIT(2)
+#define SSP_CS_LOW		0
+#define SSP_OUT_MODE		BIT(3)
+#define SSP_IN_MODE		0
+#define SSP_DATA_REG		BIT(4)
+#define SSP_ADDR_REG		0
+
+#define SSP_OPCODE_DIRECT	((0x0) << 5)
+#define SSP_OPCODE_TOGGLE	((0x1) << 5)
+#define SSP_OPCODE_SHIFT	((0x2) << 5)
+#define SSP_OPCODE_BRANCH0	((0x4) << 5)
+#define SSP_OPCODE_BRANCH1	((0x5) << 5)
+#define SSP_OPCODE_BRANCH	((0x6) << 5)
+#define SSP_OPCODE_STOP		((0x7) << 5)
+#define SSP_BRANCH(addr)	((addr) << 8)
+#define SSP_COUNT(cycles)	((cycles) << 8)
+
+struct ti_ssp_port *ti_ssp_open(const struct ti_ssp_port_data *data);
+int ti_ssp_close(struct ti_ssp_port *dev);
+int ti_ssp_dumpregs(struct ti_ssp_port *dev);
+int ti_ssp_raw_read(struct ti_ssp_port *dev);
+int ti_ssp_raw_write(struct ti_ssp_port *dev, u32 val);
+int ti_ssp_load(struct ti_ssp_port *dev, int offs, u32* prog, int len);
+int ti_ssp_run(struct ti_ssp_port *dev, u32 pc, u32 input, u32 *output);
+int ti_ssp_set_mode(struct ti_ssp_port *dev, int mode);
+int ti_ssp_set_iosel(struct ti_ssp_port *dev, u32 iosel);
+
+#endif /* __TI_SSP_H__ */
drivers/misc/Kconfig +11 -0
index b743312..9fb8470 100644
@@ -390,6 +390,17 @@ config BMP085
 	  To compile this driver as a module, choose M here: the
 	  module will be called bmp085.
 
+config TI_SSP
+	depends on ARCH_DAVINCI_TNETV107X
+	tristate "Sequencer Serial Port support"
+	default y
+	---help---
+	  Say Y here if you want support for the Sequencer Serial Port
+	  in a Texas Instruments TNETV107X SoC.
+
+	  To compile this driver as a module, choose M here: the
+	  module will be called ti_ssp.
+
 source "drivers/misc/c2port/Kconfig"
 source "drivers/misc/eeprom/Kconfig"
 source "drivers/misc/cb710/Kconfig"
drivers/misc/Makefile +1 -0
index 42eab95..7568100 100644
@@ -28,6 +28,7 @@ obj-$(CONFIG_SENSORS_TSL2550)	+= tsl2550.o
 obj-$(CONFIG_EP93XX_PWM)	+= ep93xx_pwm.o
 obj-$(CONFIG_DS1682)		+= ds1682.o
 obj-$(CONFIG_TI_DAC7512)	+= ti_dac7512.o
+obj-$(CONFIG_TI_SSP)		+= ti_ssp.o
 obj-$(CONFIG_C2PORT)		+= c2port/
 obj-$(CONFIG_IWMC3200TOP)      += iwmc3200top/
 obj-$(CONFIG_HMC6352)		+= hmc6352.o
drivers/misc/ti_ssp.c new file +82 -0
new file mode 100644
index 0000000..edbc94d
@@ -0,0 +1,443 @@
+/*
+ * Sequencer Serial Port (SSP) driver for Texas Instruments' SoCs
+ *
+ * Copyright (C) 2010 Texas Instruments Inc
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * 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., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <linux/errno.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/err.h>
+#include <linux/init.h>
+#include <linux/wait.h>
+#include <linux/clk.h>
+#include <linux/interrupt.h>
+#include <linux/device.h>
+#include <linux/spinlock.h>
+#include <linux/platform_device.h>
+#include <linux/delay.h>
+#include <linux/io.h>
+
+#include <mach/ti_ssp.h>
+
+/* Register Offsets */
+#define SSP_REG_REV		0x00
+#define SSP_REG_IOSEL_1		0x04
+#define SSP_REG_IOSEL_2		0x08
+#define SSP_REG_PREDIV		0x0c
+#define SSP_REG_INTR_STAT	0x10
+#define SSP_REG_INTR_EN		0x14
+#define SSP_REG_TEST_CTRL	0x18
+
+/* Per port registers */
+#define SSP_PORT_REG_CFG_2	0x00
+#define SSP_PORT_REG_ADDR	0x04
+#define SSP_PORT_REG_DATA	0x08
+#define SSP_PORT_REG_CFG_1	0x0c
+#define SSP_PORT_REG_STATE	0x10
+
+#define SSP_PORT_CONFIG_MASK	(SSP_EARLY_DIN | SSP_DELAY_DOUT)
+#define SSP_PORT_CLKRATE_MASK	0x0f
+
+#define SSP_SEQRAM_WR_EN	BIT(4)
+#define SSP_SEQRAM_RD_EN	BIT(5)
+#define SSP_START		BIT(15)
+#define SSP_BUSY		BIT(10)
+#define SSP_PORT_ASL		BIT(7)
+#define SSP_PORT_CFO1		BIT(6)
+
+#define SSP_PORT_SEQRAM_SIZE	32
+
+static const int ssp_port_base[]   = {0x040, 0x080};
+static const int ssp_port_seqram[] = {0x100, 0x180};
+
+/* Register Access Macros */
+#define ssp_read(ssp, reg)	 __raw_readl((ssp)->regs + SSP_REG_##reg)
+#define ssp_write(ssp, reg, val) __raw_writel(val, (ssp)->regs + SSP_REG_##reg)
+
+#define ssp_rmw(ssp, reg, mask, bits)				\
+	ssp_write(ssp, reg, (ssp_read(ssp, reg) & ~(mask)) | (bits))
+
+#define ssp_port_read(ssp, port, reg)				\
+	__raw_readl((ssp)->regs + ssp_port_base[port] +		\
+		    SSP_PORT_REG_##reg)
+
+#define ssp_port_write(ssp, port, reg, val)			\
+	__raw_writel(val, (ssp)->regs + ssp_port_base[port] +	\
+		     SSP_PORT_REG_##reg)

drivers/misc/ti_ssp.c line 92

Quoted 10 lines
drivers/misc/ti_ssp.c new file +10 -0
+
+#define ssp_port_rmw(ssp, port, reg, mask, bits)		\
+	ssp_port_write(ssp, port, reg,				\
+		       (ssp_port_read(ssp, port, reg) & ~(mask)) | (bits))
+
+#define ssp_port_clr_bits(ssp, port, reg, bits)	\
+	ssp_port_rmw(ssp, port, reg, bits, 0)
+
+#define ssp_port_set_bits(ssp, port, reg, bits)	\
+	ssp_port_rmw(ssp, port, reg, 0, bits)

Consider programming in c rather than preprocessor. You could use static inlines here.

Quoted 66 lines
drivers/misc/ti_ssp.c new file +66 -0
+
+struct ti_ssp {
+	struct resource			*res;
+	void __iomem			*regs;
+	struct timer_list		timer;
+	struct ti_ssp_port		*ports[2];
+	spinlock_t			lock;
+	struct clk			*clk;
+	struct device			*dev;
+};
+
+struct ti_ssp_port {
+	int		num;
+	struct ti_ssp	*ssp;
+	spinlock_t	lock;
+};
+
+struct ti_ssp_port *ti_ssp_open(const struct ti_ssp_port_data *data)
+{
+	struct ti_ssp		*ssp;
+	struct ti_ssp_port	*port;
+	struct device		*dev;
+	int			error = 0;
+
+	dev = bus_find_device_by_name(&platform_bus_type, NULL,
+				      data->ssp_dev_name);
+	if (!dev || !dev->driver)
+		return ERR_PTR(-ENODEV);
+
+	if (!get_device(dev))
+		return ERR_PTR(-ENODEV);
+
+	ssp = platform_get_drvdata(to_platform_device(dev));
+	if (!ssp) {
+		error = -ENODEV;
+		goto error_put;
+	}
+
+	port = kzalloc(sizeof(*port), GFP_KERNEL);
+	if (!port) {
+		error = -ENOMEM;
+		goto error_put;
+	}
+
+	port->num = data->port;
+	port->ssp = ssp;
+	spin_lock_init(&port->lock);
+
+	spin_lock(&ssp->lock);
+
+	if (ssp->ports[port->num]) {
+		error = -EBUSY;
+		goto error_unlock;
+	}
+
+	ssp->ports[port->num] = port;
+
+	spin_unlock(&ssp->lock);
+
+	ti_ssp_set_iosel(port, data->iosel);
+
+	spin_lock(&port->lock);
+	ssp_port_rmw(ssp, port->num, CFG_1, SSP_PORT_CONFIG_MASK,
+		     data->config);
+	ssp_port_rmw(ssp, port->num, CFG_2, SSP_PORT_CLKRATE_MASK, 0);
+	spin_unlock(&port->lock);

drivers/misc/ti_ssp.c line 168

Quoted 10 lines
drivers/misc/ti_ssp.c new file +10 -0
+
+	return port;
+
+error_unlock:
+	spin_unlock(&ssp->lock);
+error_put:
+	put_device(dev);
+	return ERR_PTR(error);
+}
+EXPORT_SYMBOL(ti_ssp_open);

I'm not thrilled with the ti_ssp_open()/ti_ssp_close() usage model. It appears that the usage model is the board code registers an pile of platform_devices, one for the ssp, and one for each of the behaviours on top of it. Then the various driver instances have to figure out how to find each other and whether or not they are related. Am I correct?

Rather than doing an end-run around the Linux driver model, I strongly recommend using the model to solve your problem. Register only the ssp platform_device in the board support code and pass it data about the intended behaviour (via platform data). When it gets probed, it can then register additional platform devices which will get probed by the requested driver. That way the specific ssp device instance data can be passed reliably to the spi/i2c/whatever driver without any ambiguity, without any uncertainty about whether a port is 'busy', and without the need of these open/close routines.

(Hint: The trick is to set the platform_device's pdev->dev.parent pointer to make use of the Linux device model's hierarchy).

Quoted 247 lines
drivers/misc/ti_ssp.c new file +247 -0
+
+int ti_ssp_close(struct ti_ssp_port *port)
+{
+	if (!port || !port->ssp)
+		return -EINVAL;
+	port->ssp->ports[port->num] = NULL;
+	put_device(port->ssp->dev);
+	kfree(port);
+	return 0;
+}
+EXPORT_SYMBOL(ti_ssp_close);
+
+int ti_ssp_set_mode(struct ti_ssp_port *port, int mode)
+{
+	if (!port || !port->ssp)
+		return -EINVAL;
+
+	spin_lock(&port->lock);
+	mode &= SSP_PORT_CONFIG_MASK;
+	ssp_port_rmw(port->ssp, port->num, CFG_1, SSP_PORT_CONFIG_MASK, mode);
+	spin_unlock(&port->lock);
+
+	return 0;
+}
+EXPORT_SYMBOL(ti_ssp_set_mode);
+
+int ti_ssp_set_iosel(struct ti_ssp_port *port, u32 iosel)
+{
+	unsigned int		val;
+
+	if (!port || !port->ssp)
+		return -EINVAL;
+
+	spin_lock(&port->lock);
+
+	/* IOSEL1 gets the least significant 16 bits */
+	val = ssp_read(port->ssp, IOSEL_1);
+	val &= 0xffff << (port->num ? 0 : 16);
+	val |= (iosel & 0xffff) << (port->num ? 16 : 0);
+	ssp_write(port->ssp, IOSEL_1, val);
+
+	/* IOSEL2 gets the most significant 16 bits */
+	val = ssp_read(port->ssp, IOSEL_2);
+	val &= 0x0007 << (port->num ? 0 : 16);
+	val |= (iosel & 0x00070000) >> (port->num ? 0 : 16);
+	ssp_write(port->ssp, IOSEL_2, val);
+
+	spin_unlock(&port->lock);
+
+	return 0;
+}
+EXPORT_SYMBOL(ti_ssp_set_iosel);
+
+int ti_ssp_load(struct ti_ssp_port *port, int offs, u32* prog, int len)
+{
+	int i;
+
+	if (!port || !port->ssp)
+		return -EINVAL;
+
+	if (len > SSP_PORT_SEQRAM_SIZE)
+		return -ENOSPC;
+
+	spin_lock(&port->lock);
+
+	/* Enable SeqRAM access */
+	ssp_port_set_bits(port->ssp, port->num, CFG_2, SSP_SEQRAM_WR_EN);
+
+	/* Copy code */
+	for (i = 0; i < len; i++) {
+		__raw_writel(prog[i], port->ssp->regs + offs + 4*i +
+			     ssp_port_seqram[port->num]);
+	}
+
+	/* Disable SeqRAM access */
+	ssp_port_clr_bits(port->ssp, port->num, CFG_2, SSP_SEQRAM_WR_EN);
+
+	spin_unlock(&port->lock);
+
+	return 0;
+}
+EXPORT_SYMBOL(ti_ssp_load);
+
+int ti_ssp_raw_read(struct ti_ssp_port *port)
+{
+	u32 val;
+
+	if (!port || !port->ssp)
+		return -EINVAL;
+
+	val = ssp_read(port->ssp, IOSEL_2);
+	val >>= (port->num ? 27 : 11);
+
+	return val & 0x0f;
+}
+EXPORT_SYMBOL(ti_ssp_raw_read);
+
+int ti_ssp_raw_write(struct ti_ssp_port *port, u32 val)
+{
+	u32 mask;
+
+	if (!port || !port->ssp)
+		return -EINVAL;
+
+	spin_lock(&port->ssp->lock);
+	val &= 0x0f;
+	val <<= (port->num ? 22 : 6);
+	mask = 0x0f;
+	mask <<= (port->num ? 22 : 6);
+	ssp_rmw(port->ssp, IOSEL_2, mask, val);
+	spin_unlock(&port->ssp->lock);
+
+	return 0;
+}
+EXPORT_SYMBOL(ti_ssp_raw_write);
+
+int ti_ssp_run(struct ti_ssp_port *port, u32 pc, u32 input, u32 *output)
+{
+	struct ti_ssp *ssp;
+	int count;
+
+	if (!port || !port->ssp)
+		return -EINVAL;
+	ssp = port->ssp;
+
+	if (pc & ~(0x3f))
+		return -EINVAL;
+
+	ssp_port_write(ssp, port->num, ADDR, input >> 16);
+	ssp_port_write(ssp, port->num, DATA, input & 0xffff);
+	ssp_port_rmw(ssp, port->num, CFG_1, 0x3f, pc);
+
+	ssp_port_set_bits(ssp, port->num, CFG_1, SSP_START);
+
+	for (count = 10000; count; count--) {
+		if ((ssp_port_read(ssp, port->num, CFG_1) & SSP_BUSY) == 0)
+			break;
+		udelay(1);
+	}
+
+	if (output) {
+		*(output) = (ssp_port_read(ssp, port->num, ADDR) << 16) |
+			    (ssp_port_read(ssp, port->num, DATA) &  0xffff);
+	}
+
+	if (!count) {
+		dev_err(ssp->dev, "timed out waiting for SSP operation\n");
+		return -EIO;
+	}
+
+	/* return stop address */
+	return ssp_port_read(ssp, port->num, STATE) & 0x3f;
+}
+EXPORT_SYMBOL(ti_ssp_run);
+
+static int __devinit ti_ssp_probe(struct platform_device *pdev)
+{
+	static struct ti_ssp *ssp;
+	const struct ti_ssp_data *pdata = pdev->dev.platform_data;
+	int ret = 0, prediv = 0xff;
+	unsigned long sysclk;
+	struct device *dev = &pdev->dev;
+
+	ssp = kzalloc(sizeof(*ssp), GFP_KERNEL);
+	if (!ssp) {
+		dev_err(dev, "cannot allocate device info\n");
+		return -ENOMEM;
+	}
+
+	ssp->dev = dev;
+	platform_set_drvdata(pdev, ssp);
+
+	ret = -ENODEV;
+	ssp->res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!ssp->res) {
+		dev_err(dev, "cannot determine register area\n");
+		goto error_res;
+	}
+
+	ret = -EINVAL;
+	if (!request_mem_region(ssp->res->start, resource_size(ssp->res),
+				pdev->name)) {
+		dev_err(dev, "cannot claim register memory\n");
+		goto error_res;
+	}
+
+	ret = -ENOMEM;
+	ssp->regs = ioremap(ssp->res->start, resource_size(ssp->res));
+	if (!ssp->regs) {
+		dev_err(dev, "cannot map register memory\n");
+		goto error_map;
+	}
+
+	ret = -EINVAL;
+	ssp->clk = clk_get(dev, NULL);
+	if (IS_ERR(ssp->clk)) {
+		dev_err(dev, "cannot claim device clock\n");
+		goto error_clk;
+	}
+
+	spin_lock_init(&ssp->lock);
+
+	platform_set_drvdata(pdev, ssp);
+
+	/* Power on and initialize SSP */
+	ret = clk_enable(ssp->clk);
+	if (ret)
+		goto error_enable;
+
+	/* Reset registers to a sensible known state */
+	ssp_write(ssp, IOSEL_1, 0);
+	ssp_write(ssp, IOSEL_2, 0);
+	ssp_write(ssp, INTR_EN, 0);
+	ssp_write(ssp, TEST_CTRL, 0);
+	ssp_port_write(ssp, 0, CFG_1, SSP_PORT_ASL);
+	ssp_port_write(ssp, 1, CFG_1, SSP_PORT_ASL);
+	ssp_port_write(ssp, 0, CFG_2, SSP_PORT_CFO1);
+	ssp_port_write(ssp, 1, CFG_2, SSP_PORT_CFO1);
+
+	sysclk = clk_get_rate(ssp->clk);
+	if (pdata && pdata->out_clock)
+		prediv = (sysclk / pdata->out_clock) - 1;
+	prediv = clamp(prediv, 0, 0xff);
+	ssp_rmw(ssp, PREDIV, 0xff, prediv);
+
+	return 0;
+
+error_enable:
+	clk_put(ssp->clk);
+error_clk:
+	iounmap(ssp->regs);
+error_map:
+	release_mem_region(ssp->res->start, resource_size(ssp->res));
+error_res:
+	kfree(ssp);
+	return ret;
+}
+
+static int __devexit ti_ssp_remove(struct platform_device *pdev)
+{
+	struct ti_ssp *ssp = platform_get_drvdata(pdev);
+
+	clk_disable(ssp->clk);
+	clk_put(ssp->clk);
+	iounmap(ssp->regs);
+	release_mem_region(ssp->res->start, resource_size(ssp->res));
+	kfree(ssp);

drivers/misc/ti_ssp.c line 425

Quoted 10 lines
drivers/misc/ti_ssp.c new file +10 -0
+	platform_set_drvdata(pdev, NULL);
+	return 0;
+}
+
+static struct platform_driver ti_ssp_driver = {
+	.probe		= ti_ssp_probe,
+	.remove		= __devexit_p(ti_ssp_remove),
+	.driver.name	= "ti-ssp",
+	.driver.owner	= THIS_MODULE,
+};

Personally, I prefer the following form (comment goes for the whole series):

static struct platform_driver ti_ssp_driver = {
	.probe		= ti_ssp_probe,
	.remove		= __devexit_p(ti_ssp_remove),
	.driver = {
		.name	= "ti-ssp",
		.owner	= THIS_MODULE,
	},
};

drivers/misc/ti_ssp.c line 437

Quoted 12 lines
drivers/misc/ti_ssp.c new file +12 -0
+
+static int __init ti_ssp_init(void)
+{
+	return platform_driver_register(&ti_ssp_driver);
+}
+
+static void __exit ti_ssp_exit(void)
+{
+	platform_driver_unregister(&ti_ssp_driver);
+}
+
+arch_initcall_sync(ti_ssp_init);

Keep the initicall line together with the ti_ssp_init definition (goes for the whole series).

Quoted 20 lines
drivers/misc/ti_ssp.c new file +6 -0
+module_exit(ti_ssp_exit);
+
+MODULE_AUTHOR("Cyril Chemparathy");
+MODULE_DESCRIPTION("Sequencer Serial Port (SSP) Driver");
+MODULE_ALIAS("platform:ti_ssp");
+MODULE_LICENSE("GPL");
Signature · 13 lines
1.7.0.4


------------------------------------------------------------------------------
Beautiful is writing same markup. Internet Explorer 9 supports
standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2 & L3.
Spend less time writing and  rewriting code and more time creating great
experiences on the web. Be a part of the beta today.
http://p.sf.net/sfu/beautyoftheweb
_______________________________________________
spi-devel-general mailing list
spi-devel-general-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
https://lists.sourceforge.net/lists/listinfo/spi-devel-general

A message that is not in the archive (sent off-list or not yet received): 20101016050055.GH21170-MrY2KI0G/OVr83L8+7iqerDks+cytr/[email protected]

#6 lore raw

Re: [PATCH 01/12] misc: add driver for sequencer serial port

Hi Grant,

Thanks for the feedback, I will send out an updated v2 with the proposed changes.

[...]

Quoted 19 lines
Quoted 1 line

+EXPORT_SYMBOL(ti_ssp_open);

I'm not thrilled with the ti_ssp_open()/ti_ssp_close() usage model. It appears that the usage model is the board code registers an pile of platform_devices, one for the ssp, and one for each of the behaviours on top of it. Then the various driver instances have to figure out how to find each other and whether or not they are related. Am I correct?

Rather than doing an end-run around the Linux driver model, I strongly recommend using the model to solve your problem. Register only the ssp platform_device in the board support code and pass it data about the intended behaviour (via platform data). When it gets probed, it can then register additional platform devices which will get probed by the requested driver. That way the specific ssp device instance data can be passed reliably to the spi/i2c/whatever driver without any ambiguity, without any uncertainty about whether a port is 'busy', and without the need of these open/close routines.

Quoted 2 lines

(Hint: The trick is to set the platform_device's pdev->dev.parent pointer to make use of the Linux device model's hierarchy).

I am not very thrilled with this approach either :-) I had looked at a few other instances where something similar was being done (spi, i2c, mdio, etc.), but felt that defining a new bus type would be an overkill for this.

For reference, could you please point me to some other place where something similar is being done?

Regards
Cyril.

A message that is not in the archive (sent off-list or not yet received): [email protected]

#8 lore raw

Re: [PATCH 01/12] misc: add driver for sequencer serial port

Quoted 25 lines · Cyril Chemparathy

Hi Grant,

Thanks for the feedback, I will send out an updated v2 with the proposed changes.

[...]

Quoted 19 lines
Quoted 1 line

+EXPORT_SYMBOL(ti_ssp_open);

I'm not thrilled with the ti_ssp_open()/ti_ssp_close() usage model. It appears that the usage model is the board code registers an pile of platform_devices, one for the ssp, and one for each of the behaviours on top of it.  Then the various driver instances have to figure out how to find each other and whether or not they are related. Am I correct?

Rather than doing an end-run around the Linux driver model, I strongly recommend using the model to solve your problem.  Register only the ssp platform_device in the board support code and pass it data about the intended behaviour (via platform data).  When it gets probed, it can then register additional platform devices which will get probed by the requested driver.  That way the specific ssp device instance data can be passed reliably to the spi/i2c/whatever driver without any ambiguity, without any uncertainty about whether a port is 'busy', and without the need of these open/close routines.

Quoted 7 lines
Quoted 2 lines

(Hint: The trick is to set the platform_device's pdev->dev.parent pointer to make use of the Linux device model's hierarchy).

I am not very thrilled with this approach either :-)  I had looked at a few other instances where something similar was being done (spi, i2c, mdio, etc.), but felt that defining a new bus type would be an overkill for this.

Nope! you don't need to define a new bus type. Just continue to use platform bus. If you set the dev->parent pointer before registering the device then the device model with have the appropriate hierarchy.

Quoted 2 lines

For reference, could you please point me to some other place where something similar is being done?

Most of the multifunction devices operating in this way (though those implementations are probably more convoluted than they need to be. Take a look. If you still have trouble with it let me know and I'll draft an example for you.

g.

#9 lore raw

PATCH 2/12 davinci: add tnetv107x ssp platform device

In reply to a message not in the archive

This patch adds an SSP platform device definition for the tnetv107x soc family. The clock lookup entry has also been updated to match.

---
 arch/arm/mach-davinci/devices-tnetv107x.c      |   25 ++++++++++++++++++++++++
 arch/arm/mach-davinci/include/mach/tnetv107x.h |    2 +
 arch/arm/mach-davinci/tnetv107x.c              |    2 +-
 3 files changed, 28 insertions(+), 1 deletions(-)
arch/arm/mach-davinci/devices-tnetv107x.c +25 -0
index c9a86d8..0d0522b 100644
@@ -35,6 +35,7 @@
 #define TNETV107X_SDIO0_BASE			0x08088700
 #define TNETV107X_SDIO1_BASE			0x08088800
 #define TNETV107X_KEYPAD_BASE			0x08088a00
+#define TNETV107X_SSP_BASE			0x08088c00
 #define TNETV107X_ASYNC_EMIF_CNTRL_BASE		0x08200000
 #define TNETV107X_ASYNC_EMIF_DATA_CE0_BASE	0x30000000
 #define TNETV107X_ASYNC_EMIF_DATA_CE1_BASE	0x40000000
@@ -342,6 +343,25 @@ static struct platform_device tsc_device = {
 	.resource	= tsc_resources,
 };
 
+static struct resource ssp_resources [] = {
+	{
+		.start	= TNETV107X_SSP_BASE,
+		.end	= TNETV107X_SSP_BASE + 0x1ff,
+		.flags	= IORESOURCE_MEM,
+	},
+	{
+		.start	= IRQ_TNETV107X_SSP,
+		.flags	= IORESOURCE_IRQ,
+	},
+};
+
+static struct platform_device ssp_device = {
+	.name		= "ti-ssp",
+	.id		= -1,
+	.num_resources	= ARRAY_SIZE(ssp_resources),
+	.resource	= ssp_resources,
+};
+
 void __init tnetv107x_devices_init(struct tnetv107x_device_info *info)
 {
 	int i;
@@ -367,4 +387,9 @@ void __init tnetv107x_devices_init(struct tnetv107x_device_info *info)
 		keypad_device.dev.platform_data = info->keypad_config;
 		platform_device_register(&keypad_device);
 	}
+
+	if (info->ssp_config) {
+		ssp_device.dev.platform_data = info->ssp_config;
+		platform_device_register(&ssp_device);
+	}
 }
arch/arm/mach-davinci/include/mach/tnetv107x.h +2 -0
index 5a681d8..c1df563 100644
@@ -38,12 +38,14 @@
 #include <mach/mmc.h>
 #include <mach/nand.h>
 #include <mach/serial.h>
+#include <mach/ti_ssp.h>
 
 struct tnetv107x_device_info {
 	struct davinci_uart_config	*serial_config;
 	struct davinci_mmc_config	*mmc_config[2];  /* 2 controllers */
 	struct davinci_nand_pdata	*nand_config[4]; /* 4 chipsels */
 	struct matrix_keypad_platform_data *keypad_config;
+	struct ti_ssp_data		*ssp_config;
 };
 
 extern struct platform_device tnetv107x_wdt_device;
arch/arm/mach-davinci/tnetv107x.c +1 -1
index daeae06..fcf30e7 100644
@@ -277,7 +277,7 @@ static struct clk_lookup clks[] = {
 	CLK(NULL,		"timer1",		&clk_timer1),
 	CLK("tnetv107x_wdt.0",	NULL,			&clk_wdt_arm),
 	CLK(NULL,		"clk_wdt_dsp",		&clk_wdt_dsp),
-	CLK("ti-ssp.0",		NULL,			&clk_ssp),
+	CLK("ti-ssp",		NULL,			&clk_ssp),
 	CLK(NULL,		"clk_tdm0",		&clk_tdm0),
 	CLK(NULL,		"clk_vlynq",		&clk_vlynq),
 	CLK(NULL,		"clk_mcdma",		&clk_mcdma),
Signature · 1 line
1.7.0.4
#10 lore raw

PATCH 3/12 davinci: add ssp config for tnetv107x evm board

In reply to a message not in the archive

This patch adds SSP configuration and pin muxing info for tnetv107x evm boards.

---
 arch/arm/mach-davinci/board-tnetv107x-evm.c |   13 +++++++++++++
 1 files changed, 13 insertions(+), 0 deletions(-)
arch/arm/mach-davinci/board-tnetv107x-evm.c +13 -0
index 5afa8fc..c2c467b 100644
@@ -33,6 +33,7 @@
 #include <mach/edma.h>
 #include <mach/mux.h>
 #include <mach/cp_intc.h>
+#include <mach/ti_ssp.h>
 #include <mach/tnetv107x.h>
 
 #define EVM_MMC_WP_GPIO		21
@@ -99,6 +100,12 @@ static const short uart1_pins[] __initdata = {
 	-1
 };
 
+static const short ssp_pins[] __initdata = {
+	TNETV107X_SSP0_0, TNETV107X_SSP0_1, TNETV107X_SSP0_2,
+	TNETV107X_SSP1_0, TNETV107X_SSP1_1, TNETV107X_SSP1_2,
+	TNETV107X_SSP1_3, -1
+};
+
 static struct mtd_partition nand_partitions[] = {
 	/* bootloader (U-Boot, etc) in first 12 sectors */
 	{
@@ -196,17 +203,23 @@ static struct matrix_keypad_platform_data keypad_config = {
 	.no_autorepeat	= 0,
 };
 
+static struct ti_ssp_data ssp_config = {
+	.out_clock = 250 * 1000,
+};
+
 static struct tnetv107x_device_info evm_device_info __initconst = {
 	.serial_config		= &serial_config,
 	.mmc_config[1]		= &mmc_config,	/* controller 1 */
 	.nand_config[0]		= &nand_config,	/* chip select 0 */
 	.keypad_config		= &keypad_config,
+	.ssp_config		= &ssp_config,
 };
 
 static __init void tnetv107x_evm_board_init(void)
 {
 	davinci_cfg_reg_list(sdio1_pins);
 	davinci_cfg_reg_list(uart1_pins);
+	davinci_cfg_reg_list(ssp_pins);
 
 	tnetv107x_devices_init(&evm_device_info);
 }
Signature · 9 lines
1.7.0.4


------------------------------------------------------------------------------
Beautiful is writing same markup. Internet Explorer 9 supports
standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2 & L3.
Spend less time writing and  rewriting code and more time creating great
experiences on the web. Be a part of the beta today.
http://p.sf.net/sfu/beautyoftheweb
#11 lore raw

PATCH 4/12 spi: add ti-ssp spi master driver

In reply to a message not in the archive

This patch adds an SPI master implementation that operates on top of an underlying TI-SSP port.

---
 arch/arm/mach-davinci/include/mach/ti_ssp.h |    6 +
 drivers/spi/Kconfig                         |    6 +
 drivers/spi/Makefile                        |    1 +
 drivers/spi/spi_ti_ssp.c                    |  408 +++++++++++++++++++++++++++
 4 files changed, 421 insertions(+), 0 deletions(-)
 create mode 100644 drivers/spi/spi_ti_ssp.c
arch/arm/mach-davinci/include/mach/ti_ssp.h +6 -0
index 8365101..532b0a3 100644
@@ -32,6 +32,12 @@ struct ti_ssp_port_data {
 	unsigned long	config;
 };
 
+struct ti_ssp_spi_data {
+	struct ti_ssp_port_data	port_data;
+	int			num_cs;
+	void			(*select)(int cs);
+};
+
 struct ti_ssp_port;
 
 /*
drivers/spi/Kconfig +6 -0
index 91c2f4f..c1937e4 100644
@@ -298,6 +298,12 @@ config SPI_STMP3XXX
 	help
 	  SPI driver for Freescale STMP37xx/378x SoC SSP interface
 
+config SPI_TI_SSP
+	tristate "TI SSP Controller SPI Driver"
+	depends on SPI_MASTER && TI_SSP
+	help
+	  SPI master implementation using a TI sequencer serial port.
+
 config SPI_TXX9
 	tristate "Toshiba TXx9 SPI controller"
 	depends on GENERIC_GPIO && CPU_TX49XX
drivers/spi/Makefile +1 -0
index e9cbd18..953aba7 100644
@@ -40,6 +40,7 @@ obj-$(CONFIG_SPI_S3C24XX_GPIO)		+= spi_s3c24xx_gpio.o
 obj-$(CONFIG_SPI_S3C24XX)		+= spi_s3c24xx_hw.o
 obj-$(CONFIG_SPI_S3C64XX)		+= spi_s3c64xx.o
 obj-$(CONFIG_SPI_TXX9)			+= spi_txx9.o
+obj-$(CONFIG_SPI_TI_SSP)		+= spi_ti_ssp.o
 obj-$(CONFIG_SPI_XILINX)		+= xilinx_spi.o
 obj-$(CONFIG_SPI_XILINX_OF)		+= xilinx_spi_of.o
 obj-$(CONFIG_SPI_XILINX_PLTFM)		+= xilinx_spi_pltfm.o
drivers/spi/spi_ti_ssp.c new file +408 -0

408 lines not on this page · raw message

Signature · 9 lines
1.7.0.4


------------------------------------------------------------------------------
Beautiful is writing same markup. Internet Explorer 9 supports
standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2 & L3.
Spend less time writing and  rewriting code and more time creating great
experiences on the web. Be a part of the beta today.
http://p.sf.net/sfu/beautyoftheweb

A message that is not in the archive (sent off-list or not yet received): 1287081535-2864-5-git-send-email-cyril-l0cyMroinI0@public.gmane.org

#13 lore raw

Re: [PATCH 04/12] spi: add ti-ssp spi master driver

Quoted 4 lines · Cyril Chemparathy

This patch adds an SPI master implementation that operates on top of an underlying TI-SSP port.

The driver looks good, but as mentioned in my previous comment, I'd like to see the usage model fixed to make proper use of the linux driver model before I ack it.

g.

Quoted 237 lines
---
 arch/arm/mach-davinci/include/mach/ti_ssp.h |    6 +
 drivers/spi/Kconfig                         |    6 +
 drivers/spi/Makefile                        |    1 +
 drivers/spi/spi_ti_ssp.c                    |  408 +++++++++++++++++++++++++++
 4 files changed, 421 insertions(+), 0 deletions(-)
 create mode 100644 drivers/spi/spi_ti_ssp.c
arch/arm/mach-davinci/include/mach/ti_ssp.h +6 -0
index 8365101..532b0a3 100644
@@ -32,6 +32,12 @@ struct ti_ssp_port_data {
 	unsigned long	config;
 };
 
+struct ti_ssp_spi_data {
+	struct ti_ssp_port_data	port_data;
+	int			num_cs;
+	void			(*select)(int cs);
+};
+
 struct ti_ssp_port;
 
 /*
drivers/spi/Kconfig +6 -0
index 91c2f4f..c1937e4 100644
@@ -298,6 +298,12 @@ config SPI_STMP3XXX
 	help
 	  SPI driver for Freescale STMP37xx/378x SoC SSP interface
 
+config SPI_TI_SSP
+	tristate "TI SSP Controller SPI Driver"
+	depends on SPI_MASTER && TI_SSP
+	help
+	  SPI master implementation using a TI sequencer serial port.
+
 config SPI_TXX9
 	tristate "Toshiba TXx9 SPI controller"
 	depends on GENERIC_GPIO && CPU_TX49XX
drivers/spi/Makefile +1 -0
index e9cbd18..953aba7 100644
@@ -40,6 +40,7 @@ obj-$(CONFIG_SPI_S3C24XX_GPIO)		+= spi_s3c24xx_gpio.o
 obj-$(CONFIG_SPI_S3C24XX)		+= spi_s3c24xx_hw.o
 obj-$(CONFIG_SPI_S3C64XX)		+= spi_s3c64xx.o
 obj-$(CONFIG_SPI_TXX9)			+= spi_txx9.o
+obj-$(CONFIG_SPI_TI_SSP)		+= spi_ti_ssp.o
 obj-$(CONFIG_SPI_XILINX)		+= xilinx_spi.o
 obj-$(CONFIG_SPI_XILINX_OF)		+= xilinx_spi_of.o
 obj-$(CONFIG_SPI_XILINX_PLTFM)		+= xilinx_spi_pltfm.o
drivers/spi/spi_ti_ssp.c new file +177 -0
new file mode 100644
index 0000000..132684c
@@ -0,0 +1,408 @@
+/*
+ * Sequencer Serial Port (SSP) based SPI master driver
+ *
+ * Copyright (C) 2010 Texas Instruments Inc
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * 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., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+
+#include <linux/kernel.h>
+#include <linux/err.h>
+#include <linux/completion.h>
+#include <linux/delay.h>
+#include <linux/platform_device.h>
+#include <linux/spi/spi.h>
+
+#include <mach/ti_ssp.h>
+
+#define MODE_BITS	(SPI_CPHA | SPI_CPOL | SPI_CS_HIGH)
+
+struct ti_ssp_spi {
+	struct spi_master		*master;
+	const struct ti_ssp_spi_data	*pdata;
+	struct device			*dev;
+	spinlock_t			lock;
+	struct list_head		msg_queue;
+	struct completion		complete;
+	int				shutdown:1;
+	struct workqueue_struct		*workqueue;
+	struct work_struct		work;
+	u8				mode, bpw;
+	int				cs_active;
+	struct ti_ssp_port		*ssp;
+	u32				pc_en, pc_dis, pc_wr, pc_rd;
+};
+
+static u32 do_read_data(struct ti_ssp_spi *hw)
+{
+	u32 ret;
+
+	ti_ssp_run(hw->ssp, hw->pc_rd, 0, &ret);
+	return ret;
+}
+
+static void do_write_data(struct ti_ssp_spi *hw, u32 data)
+{
+	ti_ssp_run(hw->ssp, hw->pc_wr, data << (32 - hw->bpw), NULL);
+}
+
+static int do_transfer(struct ti_ssp_spi *hw, struct spi_message *msg,
+		       struct spi_transfer *t)
+{
+	int count;
+
+	if (hw->bpw <= 8) {
+		u8		*rx = t->rx_buf;
+		const u8	*tx = t->tx_buf;
+
+		for (count = 0; count < t->len; count += 1) {
+			if (t->tx_buf)
+				do_write_data(hw, *tx++);
+			if (t->rx_buf)
+				*rx++ = do_read_data(hw);
+		}
+	} else if (hw->bpw <= 16) {
+		u16		*rx = t->rx_buf;
+		const u16	*tx = t->tx_buf;
+
+		for (count = 0; count < t->len; count += 2) {
+			if (t->tx_buf)
+				do_write_data(hw, *tx++);
+			if (t->rx_buf)
+				*rx++ = do_read_data(hw);
+		}
+	} else {
+		u32		*rx = t->rx_buf;
+		const u32	*tx = t->tx_buf;
+
+		for (count = 0; count < t->len; count += 4) {
+			if (t->tx_buf)
+				do_write_data(hw, *tx++);
+			if (t->rx_buf)
+				*rx++ = do_read_data(hw);
+		}
+	}
+
+	msg->actual_length += count; /* bytes transferred */
+
+	dev_dbg(&msg->spi->dev, "xfer %s%s, %d bytes, %d bpw, count %d%s\n",
+		t->tx_buf ? "tx" : "", t->rx_buf ? "rx" : "", t->len,
+		hw->bpw, count, (count < t->len) ? " (under)": "");
+
+	return (count < t->len) ? -EIO : 0; /* left over data */
+}
+
+static void chip_select(struct ti_ssp_spi *hw, int cs_active)
+{
+	cs_active = !!cs_active;
+	if (cs_active == hw->cs_active)
+		return;
+	ti_ssp_run(hw->ssp, cs_active ? hw->pc_en : hw->pc_dis, 0, NULL);
+	hw->cs_active = cs_active;
+}
+
+#define __SHIFT_OUT(bits) 	(SSP_OPCODE_SHIFT | SSP_OUT_MODE | \
+				 cs_en | clk | SSP_COUNT((bits) * 2 - 1))
+#define __SHIFT_IN(bits) 	(SSP_OPCODE_SHIFT | SSP_IN_MODE  | \
+				 cs_en | clk | SSP_COUNT((bits) * 2 - 1))
+
+static int setup_xfer(struct ti_ssp_spi *hw, u8 bpw, u8 mode)
+{
+	int error, idx = 0;
+	u32 seqram[16];
+	u32 cs_en, cs_dis, clk;
+	u32 topbits, botbits;
+
+	mode &= MODE_BITS;
+	if (mode == hw->mode && bpw == hw->bpw)
+		return 0;
+
+	cs_en  = (mode & SPI_CS_HIGH) ? SSP_CS_HIGH : SSP_CS_LOW;
+	cs_dis = (mode & SPI_CS_HIGH) ? SSP_CS_LOW  : SSP_CS_HIGH;
+	clk    = (mode & SPI_CPOL)    ? SSP_CLK_HIGH : SSP_CLK_LOW;
+
+	/* Construct instructions */
+
+	/* Disable Chip Select */
+	hw->pc_dis = idx;
+	seqram[idx++] = SSP_OPCODE_DIRECT | SSP_OUT_MODE | cs_dis | clk;
+	seqram[idx++] = SSP_OPCODE_STOP   | SSP_OUT_MODE | cs_dis | clk;
+
+	/* Enable Chip Select */
+	hw->pc_en = idx;
+	seqram[idx++] = SSP_OPCODE_DIRECT | SSP_OUT_MODE | cs_en | clk;
+	seqram[idx++] = SSP_OPCODE_STOP   | SSP_OUT_MODE | cs_en | clk;
+
+	/* Reads and writes need to be split for bpw > 16 */
+	topbits = (bpw > 16) ? 16 : bpw;
+	botbits = bpw - topbits;
+
+	/* Write */
+	hw->pc_wr = idx;
+	seqram[idx++] = __SHIFT_OUT(topbits) | SSP_ADDR_REG;
+	if (botbits)
+		seqram[idx++] = __SHIFT_OUT(botbits)  | SSP_DATA_REG;
+	seqram[idx++] = SSP_OPCODE_STOP | SSP_OUT_MODE | cs_en | clk;
+
+	/* Read */
+	hw->pc_rd = idx;
+	if (botbits)
+		seqram[idx++] = __SHIFT_IN(botbits) | SSP_ADDR_REG;
+	seqram[idx++] = __SHIFT_IN(topbits) | SSP_DATA_REG;
+	seqram[idx++] = SSP_OPCODE_STOP | SSP_OUT_MODE | cs_en | clk;
+
+	error = ti_ssp_load(hw->ssp, 0, seqram, idx);
+	if (error < 0)
+		return error;
+
+	error = ti_ssp_set_mode(hw->ssp, ((mode & SPI_CPHA) ?
+					  0 : SSP_EARLY_DIN));
+	if (error < 0)
+		return error;
+
+	hw->bpw = bpw;
+	hw->mode = mode;
+

drivers/spi/spi_ti_ssp.c line 187

Quoted 10 lines
drivers/spi/spi_ti_ssp.c new file +10 -0
+	return error;
+}
+
+static void ti_ssp_spi_work(struct work_struct *work)
+{
+	struct ti_ssp_spi *hw = container_of(work, struct ti_ssp_spi, work);
+
+	spin_lock_irq(&hw->lock);
+
+	 while (!list_empty(&hw->msg_queue)) {

Errant space ' ' before while statement.

Quoted 235 lines
drivers/spi/spi_ti_ssp.c new file +221 -0
+		struct spi_message	*m;
+		struct spi_device	*spi;
+		struct spi_transfer	*t = NULL;
+		int			status = 0;
+
+		m = container_of(hw->msg_queue.next, struct spi_message,
+				 queue);
+
+		list_del_init(&m->queue);
+
+		spi = m->spi;
+
+		if (hw->pdata->select)
+			hw->pdata->select(spi->chip_select);
+
+		list_for_each_entry(t, &m->transfers, transfer_list) {
+			int bpw = spi->bits_per_word;
+			int xfer_status;
+
+			if (t->bits_per_word)
+				bpw = t->bits_per_word;
+
+			if (setup_xfer(hw, bpw, spi->mode) < 0)
+				break;
+
+			chip_select(hw, 1);
+
+			spin_unlock_irq(&hw->lock);
+
+			xfer_status = do_transfer(hw, m, t);
+			if (xfer_status < 0)
+				status = xfer_status;
+
+			if (t->delay_usecs)
+				udelay(t->delay_usecs);
+
+			spin_lock_irq(&hw->lock);
+
+			if (t->cs_change)
+				chip_select(hw, 0);
+		}
+
+		chip_select(hw, 0);
+		m->status = status;
+		m->complete(m->context);
+	}
+
+	if (hw->shutdown)
+		complete(&hw->complete);
+
+	spin_unlock_irq(&hw->lock);
+}
+
+static int ti_ssp_spi_setup(struct spi_device *spi)
+{
+	if (spi->bits_per_word > 32)
+		return -EINVAL;
+
+	return 0;
+}
+
+static int ti_ssp_spi_transfer(struct spi_device *spi, struct spi_message *m)
+{
+	struct ti_ssp_spi	*hw;
+	struct spi_transfer	*t;
+	unsigned long		flags;
+	int			error = 0;
+
+	m->actual_length = 0;
+	m->status = -EINPROGRESS;
+
+	hw = spi_master_get_devdata(spi->master);
+
+	if (list_empty(&m->transfers) || !m->complete)
+		return -EINVAL;
+
+	list_for_each_entry(t, &m->transfers, transfer_list) {
+		if (t->len && !(t->rx_buf || t->tx_buf)) {
+			dev_err(&spi->dev, "invalid xfer, no buffer\n");
+			return -EINVAL;
+		}
+
+		if (t->len && t->rx_buf && t->tx_buf) {
+			dev_err(&spi->dev, "invalid xfer, full duplex\n");
+			return -EINVAL;
+		}
+
+		if (t->bits_per_word > 32) {
+			dev_err(&spi->dev, "invalid xfer width %d\n",
+				t->bits_per_word);
+			return -EINVAL;
+		}
+	}
+
+	spin_lock_irqsave(&hw->lock, flags);
+	if (hw->shutdown) {
+		error = -ESHUTDOWN;
+		goto error_unlock;
+	}
+	list_add_tail(&m->queue, &hw->msg_queue);
+	queue_work(hw->workqueue, &hw->work);
+error_unlock:
+	spin_unlock_irqrestore(&hw->lock, flags);
+	return error;
+}
+
+static int __devinit ti_ssp_spi_probe(struct platform_device *pdev)
+{
+	const struct ti_ssp_spi_data *pdata;
+	struct ti_ssp_spi *hw;
+	struct spi_master *master;
+	struct device *dev = &pdev->dev;
+	int error = 0;
+
+	pdata = dev->platform_data;
+	if (!pdata) {
+		dev_err(dev, "platform data not found\n");
+		return -EINVAL;
+	}
+
+	master = spi_alloc_master(dev, sizeof(struct ti_ssp_spi));
+	if (!master) {
+		dev_err(dev, "cannot allocate SPI master\n");
+		return -ENOMEM;
+	}
+
+	hw = spi_master_get_devdata(master);
+	platform_set_drvdata(pdev, hw);
+
+	hw->master = master;
+	hw->dev = dev;
+	hw->pdata = pdata;
+
+	spin_lock_init(&hw->lock);
+	init_completion(&hw->complete);
+	INIT_LIST_HEAD(&hw->msg_queue);
+	INIT_WORK(&hw->work, ti_ssp_spi_work);
+
+	hw->workqueue = create_singlethread_workqueue(dev_name(dev));
+	if (!hw->workqueue) {
+		error = -ENOMEM;
+		dev_err(dev, "work queue creation failed\n");
+		goto error_wq;
+	}
+
+	hw->ssp = ti_ssp_open(&hw->pdata->port_data);
+	if (IS_ERR(hw->ssp)) {
+		error = PTR_ERR(hw->ssp);
+		dev_err(dev, "ssp open failed (%d)\n", error);
+		goto error_open;
+	}
+
+	master->bus_num		= pdev->id;
+	master->num_chipselect	= hw->pdata->num_cs;
+	master->mode_bits	= MODE_BITS;
+	master->flags		= SPI_MASTER_HALF_DUPLEX;
+	master->setup		= ti_ssp_spi_setup;
+	master->transfer	= ti_ssp_spi_transfer;
+
+	error = spi_register_master(master);
+	if (error) {
+		dev_err(dev, "master registration failed\n");
+		goto error_reg;
+	}
+
+	return 0;
+
+error_reg:
+	ti_ssp_close(hw->ssp);
+error_open:
+	destroy_workqueue(hw->workqueue);
+error_wq:
+	spi_master_put(master);
+	return error;
+}
+
+static int __devexit ti_ssp_spi_remove(struct platform_device *pdev)
+{
+	struct ti_ssp_spi *hw = platform_get_drvdata(pdev);
+	int error;
+
+	hw->shutdown = 1;
+	while (!list_empty(&hw->msg_queue)) {
+		error = wait_for_completion_interruptible(&hw->complete);
+		if (error < 0) {
+			hw->shutdown = 0;
+			return error;
+		}
+	}
+
+	ti_ssp_close(hw->ssp);
+	destroy_workqueue(hw->workqueue);
+	spi_unregister_master(hw->master);
+
+	return 0;
+}
+
+static struct platform_driver ti_ssp_spi_driver = {
+	.probe 		= ti_ssp_spi_probe,
+	.remove 	= __devexit_p(ti_ssp_spi_remove),
+	.driver.name	= "ti-ssp-spi",
+	.driver.owner	= THIS_MODULE,
+};
+
+static int __init ti_ssp_spi_init(void)
+{
+	return platform_driver_register(&ti_ssp_spi_driver);
+}
+
+static void __exit ti_ssp_spi_exit(void)
+{
+	platform_driver_unregister(&ti_ssp_spi_driver);
+}
+
+subsys_initcall(ti_ssp_spi_init);
+module_exit(ti_ssp_spi_exit);
+
+MODULE_DESCRIPTION("SSP SPI Master");
+MODULE_AUTHOR("Cyril Chemparathy");
+MODULE_ALIAS("platform:spi_ti_ssp");
+MODULE_LICENSE("GPL");
Signature · 13 lines
1.7.0.4


------------------------------------------------------------------------------
Beautiful is writing same markup. Internet Explorer 9 supports
standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2 & L3.
Spend less time writing and  rewriting code and more time creating great
experiences on the web. Be a part of the beta today.
http://p.sf.net/sfu/beautyoftheweb
_______________________________________________
spi-devel-general mailing list
spi-devel-general-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
https://lists.sourceforge.net/lists/listinfo/spi-devel-general

------------------------------------------------------------------------------
Download new Adobe(R) Flash(R) Builder(TM) 4
The new Adobe(R) Flex(R) 4 and Flash(R) Builder(TM) 4 (formerly Flex(R) Builder(TM)) enable the development of rich applications that run across multiple browsers and platforms. Download your free trials today!
http://p.sf.net/sfu/adobe-dev2dev

#14 lore raw

PATCH 5/12 davinci: add spi devices on tnetv107x evm

In reply to a message not in the archive

This patch adds definitions for spi devices on the tnetv107x evm platform.

---
 arch/arm/mach-davinci/board-tnetv107x-evm.c |   51 +++++++++++++++++++++++++++
 1 files changed, 51 insertions(+), 0 deletions(-)
arch/arm/mach-davinci/board-tnetv107x-evm.c +51 -0
index c2c467b..52378bf 100644
@@ -25,6 +25,7 @@
 #include <linux/mtd/partitions.h>
 #include <linux/input.h>
 #include <linux/input/matrix_keypad.h>
+#include <linux/spi/spi.h>
 
 #include <asm/mach/arch.h>
 #include <asm/mach-types.h>
@@ -38,6 +39,7 @@
 
 #define EVM_MMC_WP_GPIO		21
 #define EVM_MMC_CD_GPIO		24
+#define EVM_SPI_CS_GPIO		54
 
 static int initialize_gpio(int gpio, char *desc)
 {
@@ -215,6 +217,52 @@ static struct tnetv107x_device_info evm_device_info __initconst = {
 	.ssp_config		= &ssp_config,
 };
 
+static void spi_select_device(int cs)
+{
+	static int gpio;
+
+	if (!gpio) {
+		int ret;
+		ret = gpio_request(EVM_SPI_CS_GPIO, "spi chipsel");
+		if (ret < 0) {
+			pr_err("cannot open spi chipsel gpio\n");
+			gpio = -ENOSYS;
+			return;
+		} else {
+			gpio = EVM_SPI_CS_GPIO;
+			gpio_direction_output(gpio, 0);
+		}
+	}
+
+	if (gpio < 0)
+		return;
+
+	return gpio_set_value(gpio, cs ? 1 : 0);
+}
+
+static struct ti_ssp_spi_data spi_master_data = {
+	.port_data = {
+		.ssp_dev_name = "ti-ssp",
+		.port = 1,
+		.iosel = SSP_PIN_SEL(0, SSP_CLOCK)	|
+			 SSP_PIN_SEL(1, SSP_DATA)	|
+			 SSP_PIN_SEL(2, SSP_CHIPSEL)	|
+			 SSP_PIN_SEL(3, SSP_IN)		|
+			 SSP_INPUT_SEL(3),
+	},
+	.num_cs		= 2,
+	.select		= spi_select_device,
+};
+
+static struct platform_device spi_master_device = {
+	.name		= "ti-ssp-spi",
+	.id		= 0,
+	.dev.platform_data = &spi_master_data,
+};
+
+static struct spi_board_info spi_info[] __initconst = {
+};
+
 static __init void tnetv107x_evm_board_init(void)
 {
 	davinci_cfg_reg_list(sdio1_pins);
@@ -222,6 +270,9 @@ static __init void tnetv107x_evm_board_init(void)
 	davinci_cfg_reg_list(ssp_pins);
 
 	tnetv107x_devices_init(&evm_device_info);
+
+	platform_device_register(&spi_master_device);
+	spi_register_board_info(spi_info, ARRAY_SIZE(spi_info));
 }
 
 #ifdef CONFIG_SERIAL_8250_CONSOLE
Signature · 9 lines
1.7.0.4


------------------------------------------------------------------------------
Beautiful is writing same markup. Internet Explorer 9 supports
standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2 & L3.
Spend less time writing and  rewriting code and more time creating great
experiences on the web. Be a part of the beta today.
http://p.sf.net/sfu/beautyoftheweb
#15 lore raw

PATCH 6/12 regulator: add driver for tps6524x regulator

In reply to a message not in the archive

TPS6524X provides three step-down converters and two general-purpose LDO voltage regulators. This device is interfaced using SPI.

---
 drivers/regulator/Kconfig              |   10 +
 drivers/regulator/Makefile             |    1 +
 drivers/regulator/tps6524x-regulator.c |  812 ++++++++++++++++++++++++++++++++
 3 files changed, 823 insertions(+), 0 deletions(-)
 create mode 100644 drivers/regulator/tps6524x-regulator.c
drivers/regulator/Kconfig +10 -0
index 172951b..7875c2e 100644
@@ -235,5 +235,15 @@ config REGULATOR_TPS6586X
 	help
 	  This driver supports TPS6586X voltage regulator chips.
 
+config REGULATOR_TPS6524X
+	tristate "TI TPS6524X Power regulators"
+	depends on TI_SSP
+	help
+	  This driver supports TPS6524X voltage regulator chips. TPS6524X
+	  provides three step-down converters and two general-purpose LDO
+	  voltage regulators.  This device is interfaced using a customized
+	  serial interface currently supported on the sequencer serial
+	  port controller.
+
 endif
 
drivers/regulator/Makefile +1 -0
index 8285fd8..a8e5bc0 100644
@@ -33,6 +33,7 @@ obj-$(CONFIG_REGULATOR_AB3100) += ab3100.o
 
 obj-$(CONFIG_REGULATOR_TPS65023) += tps65023-regulator.o
 obj-$(CONFIG_REGULATOR_TPS6507X) += tps6507x-regulator.o
+obj-$(CONFIG_REGULATOR_TPS6524X) += tps6524x-regulator.o
 obj-$(CONFIG_REGULATOR_88PM8607) += 88pm8607.o
 obj-$(CONFIG_REGULATOR_ISL6271A) += isl6271a-regulator.o
 obj-$(CONFIG_REGULATOR_AB8500)	+= ab8500.o
drivers/regulator/tps6524x-regulator.c new file +812 -0

812 lines not on this page · raw message

Signature · 9 lines
1.7.0.4


------------------------------------------------------------------------------
Beautiful is writing same markup. Internet Explorer 9 supports
standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2 & L3.
Spend less time writing and  rewriting code and more time creating great
experiences on the web. Be a part of the beta today.
http://p.sf.net/sfu/beautyoftheweb

A message that is not in the archive (sent off-list or not yet received): 1287081535-2864-7-git-send-email-cyril-l0cyMroinI0@public.gmane.org

#17 lore raw

Re: [PATCH 06/12] regulator: add driver for tps6524x regulator

Quoted 4 lines · Cyril Chemparathy

TPS6524X provides three step-down converters and two general-purpose LDO voltage regulators. This device is interfaced using SPI.

This looks mostly good but...

Review

Quoted 2 lines
(quoted diff) +2 -0
+static int read(struct tps6524x *hw, int reg)
+{

I suspect you may run into a namespace issue with this in the future.

drivers/regulator/tps6524x-regulator.c line 331

Quoted 9 lines
drivers/regulator/tps6524x-regulator.c new file +9 -0
+static const int dcdc1_voltages[] = {
+	 800000,  825000,  850000,  875000,
+	 900000,  925000,  950000,  975000,
+	1000000, 1025000, 1050000, 1075000,
+	1100000, 1125000, 1150000, 1175000,
+	1200000, 1225000, 1250000, 1275000,
+	1300000, 1325000, 1350000, 1375000,
+	1400000, 1425000, 1450000, 1475000,
+	1500000, 1525000, 1550000, 1575000,

Looks like you could just do the maths for most of these tables, might be a bit clearer/simpler? Not a major issue either way, though.

drivers/regulator/tps6524x-regulator.c line 471

Quoted 13 lines
drivers/regulator/tps6524x-regulator.c new file +13 -0
+static int list_voltage(struct regulator_dev *rdev, unsigned selector)
+{
+	const struct supply_info *info;
+	struct tps6524x *hw;
+
+	hw	= rdev_get_drvdata(rdev);
+	info	= &supply_info[rdev_get_id(rdev)];
+
+	if (info->flags & FIXED_VOLTAGE)
+		return selector ? -EINVAL : info->fixed_voltage;
+	else
+		return (selector < 0 || selector >= info->n_voltages) ?
+			-EINVAL : info->voltages[selector];

I'm not sure the ternery operator is helping legibility here - if statements might be more legible.

drivers/regulator/tps6524x-regulator.c line 513

Quoted 5 lines
drivers/regulator/tps6524x-regulator.c new file +5 -0
+	ret = read_field(hw, &info->voltage);
+	if (ret < 0)
+		return ret;
+	if (ret >= info->n_voltages)
+		return -EINVAL;

Shouldn't this be warning somehow - if we can't understand the register setting that'd be a driver bug, -EINVAL makes it sound like an invalid parameter was supplied. Similarly for the current limit.

Quoted 6 lines
drivers/regulator/tps6524x-regulator.c new file +6 -0
+static ssize_t show_reg(struct device *dev, struct device_attribute *attr,
+			char *buf);
+static ssize_t store_reg(struct device *dev, struct device_attribute *attr,
+			 const char *buf, size_t size);
+
+static DEVICE_ATTR(ldo_set,	0664, show_reg, store_reg);

drivers/regulator/tps6524x-regulator.c line 625

Quoted 10 lines
drivers/regulator/tps6524x-regulator.c new file +10 -0
+static DEVICE_ATTR(block_en,	0664, show_reg, store_reg);
+static DEVICE_ATTR(dcdc_set,	0664, show_reg, store_reg);
+static DEVICE_ATTR(dcdc_en,	0664, show_reg, store_reg);
+static DEVICE_ATTR(usb,		0664, show_reg, store_reg);
+static DEVICE_ATTR(alarm,	0664, show_reg, store_reg);
+static DEVICE_ATTR(int_enable,	0664, show_reg, store_reg);
+static DEVICE_ATTR(int_status,	0664, show_reg, store_reg);
+static DEVICE_ATTR(rev_id,	0664, show_reg, store_reg);
+static DEVICE_ATTR(write_enable,   0664, show_reg, store_reg);
+static DEVICE_ATTR(software_reset, 0664, show_reg, store_reg);

Large chunks of this looks like it's exporting functionality which is part of the regulator API to userspace, replicating functionality which the API provides. Please either remove or replace with standard functionality.

The USB setting should be integrated with the USB subsystem, assuming it's changing a USB current limit.

drivers/regulator/tps6524x-regulator.c line 744

Quoted 5 lines
drivers/regulator/tps6524x-regulator.c new file +5 -0
+	init_data = dev->platform_data;
+	if (!init_data) {
+		dev_err(dev, "could not find regulator platform data\n");
+		return -EIO;
+	}

EIO doesn't look right here...

A message that is not in the archive (sent off-list or not yet received): 20101014210323.GB14479-yzvPICuk2AATkU/[email protected]

#19 lore raw

Re: [PATCH 06/12] regulator: add driver for tps6524x regulator

Hi Mark,

Thanks for the feedback. I agree with your comments, and I will make the necessary changes before posting a v2.

The one exception being...

[...]

Quoted 4 lines
Quoted 4 lines
drivers/regulator/tps6524x-regulator.c new file +4 -0
+static const int dcdc1_voltages[] = {
+	 800000,  825000,  850000,  875000,
+	 900000,  925000,  950000,  975000,
+	1000000, 1025000, 1050000, 1075000,
Quoted 8 lines

drivers/regulator/tps6524x-regulator.c line 331

Quoted 5 lines
drivers/regulator/tps6524x-regulator.c new file +5 -0
+	1100000, 1125000, 1150000, 1175000,
+	1200000, 1225000, 1250000, 1275000,
+	1300000, 1325000, 1350000, 1375000,
+	1400000, 1425000, 1450000, 1475000,
+	1500000, 1525000, 1550000, 1575000,

Looks like you could just do the maths for most of these tables, might be a bit clearer/simpler? Not a major issue either way, though.

... not all of the supplies are linear (e.g. ldo2). Rather than have to deal with both cases (linear and non-linear), enumerating the list of voltages appeared to be simpler.

[...]

Regards
Cyril.

------------------------------------------------------------------------------
Download new Adobe(R) Flash(R) Builder(TM) 4
The new Adobe(R) Flex(R) 4 and Flash(R) Builder(TM) 4 (formerly Flex(R) Builder(TM)) enable the development of rich applications that run across multiple browsers and platforms. Download your free trials today!
http://p.sf.net/sfu/adobe-dev2dev

#21 lore raw

PATCH 7/12 davinci: add tnetv107x evm regulators

In reply to a message not in the archive

This patch adds regulator and spi board info definitions for the tps6524x power management IC found on tnetv107x evm boards.

---
 arch/arm/mach-davinci/board-tnetv107x-evm.c |   91 +++++++++++++++++++++++++++
 1 files changed, 91 insertions(+), 0 deletions(-)
arch/arm/mach-davinci/board-tnetv107x-evm.c +91 -0
index 52378bf..2f13544 100644
@@ -26,6 +26,9 @@
 #include <linux/input.h>
 #include <linux/input/matrix_keypad.h>
 #include <linux/spi/spi.h>
+#include <linux/regulator/machine.h>
+#include <linux/regulator/consumer.h>
+#include <linux/regulator/driver.h>
 
 #include <asm/mach/arch.h>
 #include <asm/mach-types.h>
@@ -260,7 +263,95 @@ static struct platform_device spi_master_device = {
 	.dev.platform_data = &spi_master_data,
 };
 
+static struct regulator_consumer_supply usb_consumers[] = {
+	REGULATOR_SUPPLY("vbus", "musb_hdrc.1"),
+};
+
+static struct regulator_consumer_supply lcd_consumers[] = {
+	REGULATOR_SUPPLY("vlcd", "tps6116x"),
+};
+
+static struct regulator_init_data regulators[] = {
+	{
+		.constraints		= {
+			.name		= "DCDC1",
+			.min_uV		= 1000000,
+			.max_uV		= 1000000,
+			.always_on	= 1,
+			.boot_on	= 1,
+		},
+	},
+	{
+		.constraints		= {
+			.name		= "DCDC2",
+			.min_uV		= 1800000,
+			.max_uV		= 1800000,
+			.always_on	= 1,
+			.boot_on	= 1,
+		},
+	},
+	{
+		.constraints		= {
+			.name		= "DCDC3",
+			.min_uV		= 3300000,
+			.max_uV		= 3300000,
+			.always_on	= 1,
+			.boot_on	= 1,
+		},
+	},
+	{
+		.constraints		= {
+			.name		= "LDO1",
+			.min_uV		= 4800000,
+			.max_uV		= 4800000,
+			.min_uA		= 400000,
+			.max_uA		= 1500000,
+			.always_on	= 1,
+			.boot_on	= 1,
+		},
+	},
+	{
+		.constraints		= {
+			.name		= "LDO1",
+			.min_uV		= 3300000,
+			.max_uV		= 3300000,
+			.always_on	= 1,
+			.boot_on	= 1,
+		},
+	},
+	{
+		.num_consumer_supplies	= ARRAY_SIZE(usb_consumers),
+		.consumer_supplies	= usb_consumers,
+		.constraints		= {
+			.name		= "USB",
+			.min_uV		= 5000000,
+			.max_uV		= 5000000,
+			.min_uA		= 200000,
+			.max_uA		= 1000000,
+			.valid_ops_mask	= REGULATOR_CHANGE_CURRENT |
+					  REGULATOR_CHANGE_STATUS,
+		},
+	},
+	{
+		.num_consumer_supplies	= ARRAY_SIZE(lcd_consumers),
+		.consumer_supplies	= lcd_consumers,
+		.constraints		= {
+			.name		= "LCD",
+			.min_uV		= 5000000,
+			.max_uV		= 5000000,
+			.valid_ops_mask	= REGULATOR_CHANGE_STATUS,
+		},
+	},
+};
+
 static struct spi_board_info spi_info[] __initconst = {
+	{
+		.modalias	= "tps6524x",
+		.bus_num	= 0,
+		.chip_select	= 0,
+		.mode		= SPI_MODE_0,
+		.platform_data	= regulators,
+	},
 };
 
 static __init void tnetv107x_evm_board_init(void)
Signature · 1 line
1.7.0.4

A message that is not in the archive (sent off-list or not yet received): 1287081535-2864-8-git-send-email-cyril-l0cyMroinI0@public.gmane.org

#23 lore raw

Re: [PATCH 07/12] davinci: add tnetv107x evm regulators

arch/arm/mach-davinci/board-tnetv107x-evm.c static struct platform_device spi_master_device = { · line 330

Quoted 9 lines · Cyril Chemparathy
arch/arm/mach-davinci/board-tnetv107x-evm.c +9 -0
+	{
+		.num_consumer_supplies	= ARRAY_SIZE(usb_consumers),
+		.consumer_supplies	= usb_consumers,
+		.constraints		= {
+			.name		= "USB",
+			.min_uV		= 5000000,
+			.max_uV		= 5000000,
+			.min_uA		= 200000,
+			.max_uA		= 1000000,

Are you *sure* about the min_uA here - what would be the goal of enforcing a minimum current draw? For a voltage regulator current limit usually only an upper limit would be specified. The same applies to several other machine drivers in the series.

#24 lore raw

PATCH 8/12 gpio: add ti-ssp virtual gpio driver

In reply to a message not in the archive

This patch adds a GPIO driver based on TI's SSP device. This driver does not support GPIO-IRQs.

---
 arch/arm/mach-davinci/include/mach/ti_ssp.h |    5 +
 drivers/gpio/Kconfig                        |   10 ++
 drivers/gpio/Makefile                       |    1 +
 drivers/gpio/ti-ssp-gpio.c                  |  234 +++++++++++++++++++++++++++
 4 files changed, 250 insertions(+), 0 deletions(-)
 create mode 100644 drivers/gpio/ti-ssp-gpio.c
arch/arm/mach-davinci/include/mach/ti_ssp.h +5 -0
index 532b0a3..4392b43 100644
@@ -38,6 +38,11 @@ struct ti_ssp_spi_data {
 	void			(*select)(int cs);
 };
 
+struct ti_ssp_gpio_data {
+	struct ti_ssp_port_data	port_data;
+	int			start;
+};
+
 struct ti_ssp_port;
 
 /*
drivers/gpio/Kconfig +10 -0
index 510aa20..e400761 100644
@@ -111,6 +111,16 @@ config GPIO_SCH
 	  This driver can also be built as a module. If so, the module
 	  will be called sch-gpio.
 
+config GPIO_TI_SSP
+	tristate "TI SSP Controller GPIO Driver"
+	depends on GPIOLIB && TI_SSP
+	help
+	  Say yes here to support a virtual GPIO interface on TI SSP ports.
+	  Each SSP port translates into 4 GPIOs, with no IRQ support.
+
+	  This driver can also be built as a module. If so, the module
+	  will be called ti-ssp-gpio.
+
 comment "I2C GPIO expanders:"
 
 config GPIO_MAX7300
drivers/gpio/Makefile +1 -0
index fc6019d..98b4551 100644
@@ -23,6 +23,7 @@ obj-$(CONFIG_GPIO_PL061)	+= pl061.o
 obj-$(CONFIG_GPIO_STMPE)	+= stmpe-gpio.o
 obj-$(CONFIG_GPIO_TC35892)	+= tc35892-gpio.o
 obj-$(CONFIG_GPIO_TIMBERDALE)	+= timbgpio.o
+obj-$(CONFIG_GPIO_TI_SSP)	+= ti-ssp-gpio.o
 obj-$(CONFIG_GPIO_TWL4030)	+= twl4030-gpio.o
 obj-$(CONFIG_GPIO_UCB1400)	+= ucb1400_gpio.o
 obj-$(CONFIG_GPIO_XILINX)	+= xilinx_gpio.o
drivers/gpio/ti-ssp-gpio.c new file +234 -0

234 lines not on this page · raw message

Signature · 9 lines
1.7.0.4


------------------------------------------------------------------------------
Beautiful is writing same markup. Internet Explorer 9 supports
standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2 & L3.
Spend less time writing and  rewriting code and more time creating great
experiences on the web. Be a part of the beta today.
http://p.sf.net/sfu/beautyoftheweb
#25 lore raw

PATCH 9/12 davinci: add tnetv107x evm ti-ssp gpio device

In reply to a message not in the archive

This patch adds definitions to hook up one of the ti-ssp ports to the SSP GPIO driver.

---
 arch/arm/mach-davinci/board-tnetv107x-evm.c |   23 +++++++++++++++++++++++
 1 files changed, 23 insertions(+), 0 deletions(-)
arch/arm/mach-davinci/board-tnetv107x-evm.c +23 -0
index 2f13544..7b4f86e 100644
@@ -40,6 +40,8 @@
 #include <mach/ti_ssp.h>
 #include <mach/tnetv107x.h>
 
+#define SSP_GPIO_START		128
+
 #define EVM_MMC_WP_GPIO		21
 #define EVM_MMC_CD_GPIO		24
 #define EVM_SPI_CS_GPIO		54
@@ -263,6 +265,25 @@ static struct platform_device spi_master_device = {
 	.dev.platform_data = &spi_master_data,
 };
 
+static struct ti_ssp_gpio_data ssp_gpio_data = {
+	.port_data = {
+		.ssp_dev_name = "ti-ssp",
+		.port = 0,
+		.iosel = SSP_PIN_SEL(0, SSP_OUT)	|
+			 SSP_PIN_SEL(1, SSP_OUT)	|
+			 SSP_PIN_SEL(2, SSP_OUT)	|
+			 SSP_PIN_SEL(3, SSP_OUT)	|
+			 SSP_INPUT_SEL(3),
+	},
+	.start		= SSP_GPIO_START,
+};
+
+static struct platform_device ssp_gpio_device = {
+	.name		= "ti-ssp-gpio",
+	.id		= 0,
+	.dev.platform_data = &ssp_gpio_data,
+};
+
 static struct regulator_consumer_supply usb_consumers[] = {
 	REGULATOR_SUPPLY("vbus", "musb_hdrc.1"),
 };
@@ -363,6 +384,8 @@ static __init void tnetv107x_evm_board_init(void)
 	tnetv107x_devices_init(&evm_device_info);
 
 	platform_device_register(&spi_master_device);
+	platform_device_register(&ssp_gpio_device);
+
 	spi_register_board_info(spi_info, ARRAY_SIZE(spi_info));
 }
 
Signature · 9 lines
1.7.0.4


------------------------------------------------------------------------------
Beautiful is writing same markup. Internet Explorer 9 supports
standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2 & L3.
Spend less time writing and  rewriting code and more time creating great
experiences on the web. Be a part of the beta today.
http://p.sf.net/sfu/beautyoftheweb
#26 lore raw

PATCH 10/12 backlight: add support for tps6116x controller

In reply to a message not in the archive

TPS6116x is an EasyScale backlight controller device. This driver supports TPS6116x devices connected on a single GPIO.

---
 drivers/video/backlight/Kconfig    |    7 +
 drivers/video/backlight/Makefile   |    2 +-
 drivers/video/backlight/tps6116x.c |  339 ++++++++++++++++++++++++++++++++++++
 3 files changed, 347 insertions(+), 1 deletions(-)
 create mode 100644 drivers/video/backlight/tps6116x.c
drivers/video/backlight/Kconfig +7 -0
index e54a337..06e868e 100644
@@ -307,6 +307,13 @@ config BACKLIGHT_PCF50633
 	  If you have a backlight driven by a NXP PCF50633 MFD, say Y here to
 	  enable its driver.
 
+config BACKLIGHT_TPS6116X
+	tristate "TPS6116X LCD Backlight"
+	depends on GENERIC_GPIO
+	help
+	  This driver controls the LCD backlight level for EasyScale capable
+	  SSP connected backlight controllers.
+
 endif # BACKLIGHT_CLASS_DEVICE
 
 endif # BACKLIGHT_LCD_SUPPORT
drivers/video/backlight/Makefile +1 -1
index 44c0f81..5d407c8 100644
@@ -35,4 +35,4 @@ obj-$(CONFIG_BACKLIGHT_ADP5520)	+= adp5520_bl.o
 obj-$(CONFIG_BACKLIGHT_ADP8860)	+= adp8860_bl.o
 obj-$(CONFIG_BACKLIGHT_88PM860X) += 88pm860x_bl.o
 obj-$(CONFIG_BACKLIGHT_PCF50633)	+= pcf50633-backlight.o
-
+obj-$(CONFIG_BACKLIGHT_TPS6116X)+= tps6116x.o
drivers/video/backlight/tps6116x.c new file +339 -0

339 lines not on this page · raw message

Signature · 9 lines
1.7.0.4


------------------------------------------------------------------------------
Beautiful is writing same markup. Internet Explorer 9 supports
standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2 & L3.
Spend less time writing and  rewriting code and more time creating great
experiences on the web. Be a part of the beta today.
http://p.sf.net/sfu/beautyoftheweb
#27 lore raw

PATCH 11/12 davinci: add tnetv107x evm backlight device

In reply to a message not in the archive

The tnetv107x evm board has a backlight device that is connected on one of the SSP ports. This patch adds the board definitions necessary to plug the backlight driver to the GPIO corresponding to this SSP pin.

---
 arch/arm/mach-davinci/board-tnetv107x-evm.c |    8 ++++++++
 1 files changed, 8 insertions(+), 0 deletions(-)
arch/arm/mach-davinci/board-tnetv107x-evm.c +8 -0
index 7b4f86e..0a1e9c8 100644
@@ -45,6 +45,7 @@
 #define EVM_MMC_WP_GPIO		21
 #define EVM_MMC_CD_GPIO		24
 #define EVM_SPI_CS_GPIO		54
+#define EVM_BACKLIGHT_GPIO	(SSP_GPIO_START + 2)
 
 static int initialize_gpio(int gpio, char *desc)
 {
@@ -375,6 +376,12 @@ static struct spi_board_info spi_info[] __initconst = {
 	},
 };
 
+static struct platform_device backlight_device = {
+	.name		= "tps6116x",
+	.id		= -1,
+	.dev.platform_data = (void *)EVM_BACKLIGHT_GPIO,
+};
+
 static __init void tnetv107x_evm_board_init(void)
 {
 	davinci_cfg_reg_list(sdio1_pins);
@@ -385,6 +392,7 @@ static __init void tnetv107x_evm_board_init(void)
 
 	platform_device_register(&spi_master_device);
 	platform_device_register(&ssp_gpio_device);
+	platform_device_register(&backlight_device);
 
 	spi_register_board_info(spi_info, ARRAY_SIZE(spi_info));
 }
Signature · 9 lines
1.7.0.4


------------------------------------------------------------------------------
Beautiful is writing same markup. Internet Explorer 9 supports
standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2 & L3.
Spend less time writing and  rewriting code and more time creating great
experiences on the web. Be a part of the beta today.
http://p.sf.net/sfu/beautyoftheweb
#28 lore raw

PATCH 12/12 davinci: add tnetv107x evm i2c eeprom device

In reply to a message not in the archive

The tnetv107x evm board has an I2C device connected on one of the SSP ports. This patch adds board definitions for a GPIO based I2C master, as well as definitions for the eeprom device on these boards.

---
 arch/arm/mach-davinci/board-tnetv107x-evm.c |   35 +++++++++++++++++++++++----
 1 files changed, 30 insertions(+), 5 deletions(-)
arch/arm/mach-davinci/board-tnetv107x-evm.c +30 -5
index 0a1e9c8..2e3d842 100644
@@ -29,6 +29,9 @@
 #include <linux/regulator/machine.h>
 #include <linux/regulator/consumer.h>
 #include <linux/regulator/driver.h>
+#include <linux/i2c.h>
+#include <linux/i2c/at24.h>
+#include <linux/i2c-gpio.h>
 
 #include <asm/mach/arch.h>
 #include <asm/mach-types.h>
@@ -45,6 +48,8 @@
 #define EVM_MMC_WP_GPIO		21
 #define EVM_MMC_CD_GPIO		24
 #define EVM_SPI_CS_GPIO		54
+#define EVM_I2C_SDA_GPIO	(SSP_GPIO_START + 0)
+#define EVM_I2C_SCL_GPIO	(SSP_GPIO_START + 1)
 #define EVM_BACKLIGHT_GPIO	(SSP_GPIO_START + 2)
 
 static int initialize_gpio(int gpio, char *desc)
@@ -270,11 +275,6 @@ static struct ti_ssp_gpio_data ssp_gpio_data = {
 	.port_data = {
 		.ssp_dev_name = "ti-ssp",
 		.port = 0,
-		.iosel = SSP_PIN_SEL(0, SSP_OUT)	|
-			 SSP_PIN_SEL(1, SSP_OUT)	|
-			 SSP_PIN_SEL(2, SSP_OUT)	|
-			 SSP_PIN_SEL(3, SSP_OUT)	|
-			 SSP_INPUT_SEL(3),
 	},
 	.start		= SSP_GPIO_START,
 };
@@ -382,6 +382,29 @@ static struct platform_device backlight_device = {
 	.dev.platform_data = (void *)EVM_BACKLIGHT_GPIO,
 };
 
+struct i2c_gpio_platform_data i2c_data = {
+	.sda_pin	= EVM_I2C_SDA_GPIO,
+	.scl_pin	= EVM_I2C_SCL_GPIO,
+};
+
+static struct platform_device i2c_device = {
+	.name		= "i2c-gpio",
+	.id		= 0,
+	.dev.platform_data = &i2c_data,
+};
+
+static struct at24_platform_data at24_config = {
+	.byte_len	= SZ_16K / 8,
+	.page_size	= 16,
+};
+
+static struct i2c_board_info i2c_info[] __initconst =  {
+	{
+		I2C_BOARD_INFO("24c16", 0x50),
+		.platform_data	= &at24_config,
+	},
+};
+
 static __init void tnetv107x_evm_board_init(void)
 {
 	davinci_cfg_reg_list(sdio1_pins);
@@ -393,8 +416,10 @@ static __init void tnetv107x_evm_board_init(void)
 	platform_device_register(&spi_master_device);
 	platform_device_register(&ssp_gpio_device);
 	platform_device_register(&backlight_device);
+	platform_device_register(&i2c_device);
 
 	spi_register_board_info(spi_info, ARRAY_SIZE(spi_info));
+	i2c_register_board_info(0, i2c_info, ARRAY_SIZE(i2c_info));
 }
 
 #ifdef CONFIG_SERIAL_8250_CONSOLE
Signature · 9 lines
1.7.0.4


------------------------------------------------------------------------------
Beautiful is writing same markup. Internet Explorer 9 supports
standards for HTML5, CSS3, SVG 1.1,  ECMAScript5, and DOM L2 & L3.
Spend less time writing and  rewriting code and more time creating great
experiences on the web. Be a part of the beta today.
http://p.sf.net/sfu/beautyoftheweb