2008-02-05 07:28:25 +01:00
|
|
|
/*
|
mcp23s08: add i2c support
Add i2c bindings for the mcp230xx devices. This is quite a lot simpler
than the spi one as there's no funky sub addressing done (one struct
i2c_client per struct gpio_chip).
The mcp23s08_platform_data structure is reused for i2c, even though
only a single mcp23s08_chip_info structure is needed.
To use, simply fill out a platform_data structure and pass it in
i2c_board_info, E.G.:
static const struct mcp23s08_platform_data mcp23017_data = {
.chip[0] = {
.pullups = 0x00ff,
},
.base = 240,
};
static struct i2c_board_info __initdata i2c_devs[] = {
{ I2C_BOARD_INFO("mcp23017", 0x20),
.platform_data = &smartview_mcp23017_data, },
...
};
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2011-07-15 10:25:32 +02:00
|
|
|
* MCP23S08 SPI/GPIO gpio expander driver
|
2008-02-05 07:28:25 +01:00
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/kernel.h>
|
|
|
|
#include <linux/device.h>
|
|
|
|
#include <linux/mutex.h>
|
2011-07-03 19:38:09 +02:00
|
|
|
#include <linux/module.h>
|
2009-09-23 01:46:37 +02:00
|
|
|
#include <linux/gpio.h>
|
mcp23s08: add i2c support
Add i2c bindings for the mcp230xx devices. This is quite a lot simpler
than the spi one as there's no funky sub addressing done (one struct
i2c_client per struct gpio_chip).
The mcp23s08_platform_data structure is reused for i2c, even though
only a single mcp23s08_chip_info structure is needed.
To use, simply fill out a platform_data structure and pass it in
i2c_board_info, E.G.:
static const struct mcp23s08_platform_data mcp23017_data = {
.chip[0] = {
.pullups = 0x00ff,
},
.base = 240,
};
static struct i2c_board_info __initdata i2c_devs[] = {
{ I2C_BOARD_INFO("mcp23017", 0x20),
.platform_data = &smartview_mcp23017_data, },
...
};
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2011-07-15 10:25:32 +02:00
|
|
|
#include <linux/i2c.h>
|
2008-02-05 07:28:25 +01:00
|
|
|
#include <linux/spi/spi.h>
|
|
|
|
#include <linux/spi/mcp23s08.h>
|
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 09:04:11 +01:00
|
|
|
#include <linux/slab.h>
|
2011-03-09 17:56:30 +01:00
|
|
|
#include <asm/byteorder.h>
|
2008-02-05 07:28:25 +01:00
|
|
|
|
2011-03-09 17:56:30 +01:00
|
|
|
/**
|
|
|
|
* MCP types supported by driver
|
|
|
|
*/
|
|
|
|
#define MCP_TYPE_S08 0
|
|
|
|
#define MCP_TYPE_S17 1
|
mcp23s08: add i2c support
Add i2c bindings for the mcp230xx devices. This is quite a lot simpler
than the spi one as there's no funky sub addressing done (one struct
i2c_client per struct gpio_chip).
The mcp23s08_platform_data structure is reused for i2c, even though
only a single mcp23s08_chip_info structure is needed.
To use, simply fill out a platform_data structure and pass it in
i2c_board_info, E.G.:
static const struct mcp23s08_platform_data mcp23017_data = {
.chip[0] = {
.pullups = 0x00ff,
},
.base = 240,
};
static struct i2c_board_info __initdata i2c_devs[] = {
{ I2C_BOARD_INFO("mcp23017", 0x20),
.platform_data = &smartview_mcp23017_data, },
...
};
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2011-07-15 10:25:32 +02:00
|
|
|
#define MCP_TYPE_008 2
|
|
|
|
#define MCP_TYPE_017 3
|
2008-02-05 07:28:25 +01:00
|
|
|
|
|
|
|
/* Registers are all 8 bits wide.
|
|
|
|
*
|
|
|
|
* The mcp23s17 has twice as many bits, and can be configured to work
|
|
|
|
* with either 16 bit registers or with two adjacent 8 bit banks.
|
|
|
|
*/
|
|
|
|
#define MCP_IODIR 0x00 /* init/reset: all ones */
|
|
|
|
#define MCP_IPOL 0x01
|
|
|
|
#define MCP_GPINTEN 0x02
|
|
|
|
#define MCP_DEFVAL 0x03
|
|
|
|
#define MCP_INTCON 0x04
|
|
|
|
#define MCP_IOCON 0x05
|
|
|
|
# define IOCON_SEQOP (1 << 5)
|
|
|
|
# define IOCON_HAEN (1 << 3)
|
|
|
|
# define IOCON_ODR (1 << 2)
|
|
|
|
# define IOCON_INTPOL (1 << 1)
|
|
|
|
#define MCP_GPPU 0x06
|
|
|
|
#define MCP_INTF 0x07
|
|
|
|
#define MCP_INTCAP 0x08
|
|
|
|
#define MCP_GPIO 0x09
|
|
|
|
#define MCP_OLAT 0x0a
|
|
|
|
|
2011-03-09 17:56:30 +01:00
|
|
|
struct mcp23s08;
|
|
|
|
|
|
|
|
struct mcp23s08_ops {
|
|
|
|
int (*read)(struct mcp23s08 *mcp, unsigned reg);
|
|
|
|
int (*write)(struct mcp23s08 *mcp, unsigned reg, unsigned val);
|
|
|
|
int (*read_regs)(struct mcp23s08 *mcp, unsigned reg,
|
|
|
|
u16 *vals, unsigned n);
|
|
|
|
};
|
|
|
|
|
2008-02-05 07:28:25 +01:00
|
|
|
struct mcp23s08 {
|
|
|
|
u8 addr;
|
|
|
|
|
2011-03-09 17:56:30 +01:00
|
|
|
u16 cache[11];
|
2008-02-05 07:28:25 +01:00
|
|
|
/* lock protects the cached values */
|
|
|
|
struct mutex lock;
|
|
|
|
|
|
|
|
struct gpio_chip chip;
|
|
|
|
|
2011-03-09 17:56:30 +01:00
|
|
|
const struct mcp23s08_ops *ops;
|
2011-07-15 10:25:31 +02:00
|
|
|
void *data; /* ops specific data */
|
2008-02-05 07:28:25 +01:00
|
|
|
};
|
|
|
|
|
2011-03-09 17:56:30 +01:00
|
|
|
/* A given spi_device can represent up to eight mcp23sxx chips
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
* sharing the same chipselect but using different addresses
|
|
|
|
* (e.g. chips #0 and #3 might be populated, but not #1 or $2).
|
|
|
|
* Driver data holds all the per-chip data.
|
|
|
|
*/
|
|
|
|
struct mcp23s08_driver_data {
|
|
|
|
unsigned ngpio;
|
2011-03-09 17:56:30 +01:00
|
|
|
struct mcp23s08 *mcp[8];
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
struct mcp23s08 chip[];
|
|
|
|
};
|
|
|
|
|
mcp23s08: add i2c support
Add i2c bindings for the mcp230xx devices. This is quite a lot simpler
than the spi one as there's no funky sub addressing done (one struct
i2c_client per struct gpio_chip).
The mcp23s08_platform_data structure is reused for i2c, even though
only a single mcp23s08_chip_info structure is needed.
To use, simply fill out a platform_data structure and pass it in
i2c_board_info, E.G.:
static const struct mcp23s08_platform_data mcp23017_data = {
.chip[0] = {
.pullups = 0x00ff,
},
.base = 240,
};
static struct i2c_board_info __initdata i2c_devs[] = {
{ I2C_BOARD_INFO("mcp23017", 0x20),
.platform_data = &smartview_mcp23017_data, },
...
};
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2011-07-15 10:25:32 +02:00
|
|
|
/*----------------------------------------------------------------------*/
|
|
|
|
|
|
|
|
#ifdef CONFIG_I2C
|
|
|
|
|
|
|
|
static int mcp23008_read(struct mcp23s08 *mcp, unsigned reg)
|
|
|
|
{
|
|
|
|
return i2c_smbus_read_byte_data(mcp->data, reg);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int mcp23008_write(struct mcp23s08 *mcp, unsigned reg, unsigned val)
|
|
|
|
{
|
|
|
|
return i2c_smbus_write_byte_data(mcp->data, reg, val);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mcp23008_read_regs(struct mcp23s08 *mcp, unsigned reg, u16 *vals, unsigned n)
|
|
|
|
{
|
|
|
|
while (n--) {
|
|
|
|
int ret = mcp23008_read(mcp, reg++);
|
|
|
|
if (ret < 0)
|
|
|
|
return ret;
|
|
|
|
*vals++ = ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int mcp23017_read(struct mcp23s08 *mcp, unsigned reg)
|
|
|
|
{
|
|
|
|
return i2c_smbus_read_word_data(mcp->data, reg << 1);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int mcp23017_write(struct mcp23s08 *mcp, unsigned reg, unsigned val)
|
|
|
|
{
|
|
|
|
return i2c_smbus_write_word_data(mcp->data, reg << 1, val);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mcp23017_read_regs(struct mcp23s08 *mcp, unsigned reg, u16 *vals, unsigned n)
|
|
|
|
{
|
|
|
|
while (n--) {
|
|
|
|
int ret = mcp23017_read(mcp, reg++);
|
|
|
|
if (ret < 0)
|
|
|
|
return ret;
|
|
|
|
*vals++ = ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct mcp23s08_ops mcp23008_ops = {
|
|
|
|
.read = mcp23008_read,
|
|
|
|
.write = mcp23008_write,
|
|
|
|
.read_regs = mcp23008_read_regs,
|
|
|
|
};
|
|
|
|
|
|
|
|
static const struct mcp23s08_ops mcp23017_ops = {
|
|
|
|
.read = mcp23017_read,
|
|
|
|
.write = mcp23017_write,
|
|
|
|
.read_regs = mcp23017_read_regs,
|
|
|
|
};
|
|
|
|
|
|
|
|
#endif /* CONFIG_I2C */
|
|
|
|
|
|
|
|
/*----------------------------------------------------------------------*/
|
|
|
|
|
2011-07-15 10:25:31 +02:00
|
|
|
#ifdef CONFIG_SPI_MASTER
|
|
|
|
|
2008-02-05 07:28:25 +01:00
|
|
|
static int mcp23s08_read(struct mcp23s08 *mcp, unsigned reg)
|
|
|
|
{
|
|
|
|
u8 tx[2], rx[1];
|
|
|
|
int status;
|
|
|
|
|
|
|
|
tx[0] = mcp->addr | 0x01;
|
|
|
|
tx[1] = reg;
|
2011-07-15 10:25:31 +02:00
|
|
|
status = spi_write_then_read(mcp->data, tx, sizeof tx, rx, sizeof rx);
|
2008-02-05 07:28:25 +01:00
|
|
|
return (status < 0) ? status : rx[0];
|
|
|
|
}
|
|
|
|
|
2011-03-09 17:56:30 +01:00
|
|
|
static int mcp23s08_write(struct mcp23s08 *mcp, unsigned reg, unsigned val)
|
2008-02-05 07:28:25 +01:00
|
|
|
{
|
|
|
|
u8 tx[3];
|
|
|
|
|
|
|
|
tx[0] = mcp->addr;
|
|
|
|
tx[1] = reg;
|
|
|
|
tx[2] = val;
|
2011-07-15 10:25:31 +02:00
|
|
|
return spi_write_then_read(mcp->data, tx, sizeof tx, NULL, 0);
|
2008-02-05 07:28:25 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
2011-03-09 17:56:30 +01:00
|
|
|
mcp23s08_read_regs(struct mcp23s08 *mcp, unsigned reg, u16 *vals, unsigned n)
|
2008-02-05 07:28:25 +01:00
|
|
|
{
|
2011-03-09 17:56:30 +01:00
|
|
|
u8 tx[2], *tmp;
|
|
|
|
int status;
|
2008-02-05 07:28:25 +01:00
|
|
|
|
|
|
|
if ((n + reg) > sizeof mcp->cache)
|
|
|
|
return -EINVAL;
|
|
|
|
tx[0] = mcp->addr | 0x01;
|
|
|
|
tx[1] = reg;
|
2011-03-09 17:56:30 +01:00
|
|
|
|
|
|
|
tmp = (u8 *)vals;
|
2011-07-15 10:25:31 +02:00
|
|
|
status = spi_write_then_read(mcp->data, tx, sizeof tx, tmp, n);
|
2011-03-09 17:56:30 +01:00
|
|
|
if (status >= 0) {
|
|
|
|
while (n--)
|
|
|
|
vals[n] = tmp[n]; /* expand to 16bit */
|
|
|
|
}
|
|
|
|
return status;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int mcp23s17_read(struct mcp23s08 *mcp, unsigned reg)
|
|
|
|
{
|
|
|
|
u8 tx[2], rx[2];
|
|
|
|
int status;
|
|
|
|
|
|
|
|
tx[0] = mcp->addr | 0x01;
|
|
|
|
tx[1] = reg << 1;
|
2011-07-15 10:25:31 +02:00
|
|
|
status = spi_write_then_read(mcp->data, tx, sizeof tx, rx, sizeof rx);
|
2011-03-09 17:56:30 +01:00
|
|
|
return (status < 0) ? status : (rx[0] | (rx[1] << 8));
|
|
|
|
}
|
|
|
|
|
|
|
|
static int mcp23s17_write(struct mcp23s08 *mcp, unsigned reg, unsigned val)
|
|
|
|
{
|
|
|
|
u8 tx[4];
|
|
|
|
|
|
|
|
tx[0] = mcp->addr;
|
|
|
|
tx[1] = reg << 1;
|
|
|
|
tx[2] = val;
|
|
|
|
tx[3] = val >> 8;
|
2011-07-15 10:25:31 +02:00
|
|
|
return spi_write_then_read(mcp->data, tx, sizeof tx, NULL, 0);
|
2011-03-09 17:56:30 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mcp23s17_read_regs(struct mcp23s08 *mcp, unsigned reg, u16 *vals, unsigned n)
|
|
|
|
{
|
|
|
|
u8 tx[2];
|
|
|
|
int status;
|
|
|
|
|
|
|
|
if ((n + reg) > sizeof mcp->cache)
|
|
|
|
return -EINVAL;
|
|
|
|
tx[0] = mcp->addr | 0x01;
|
|
|
|
tx[1] = reg << 1;
|
|
|
|
|
2011-07-15 10:25:31 +02:00
|
|
|
status = spi_write_then_read(mcp->data, tx, sizeof tx,
|
2011-03-09 17:56:30 +01:00
|
|
|
(u8 *)vals, n * 2);
|
|
|
|
if (status >= 0) {
|
|
|
|
while (n--)
|
|
|
|
vals[n] = __le16_to_cpu((__le16)vals[n]);
|
|
|
|
}
|
|
|
|
|
|
|
|
return status;
|
2008-02-05 07:28:25 +01:00
|
|
|
}
|
|
|
|
|
2011-03-09 17:56:30 +01:00
|
|
|
static const struct mcp23s08_ops mcp23s08_ops = {
|
|
|
|
.read = mcp23s08_read,
|
|
|
|
.write = mcp23s08_write,
|
|
|
|
.read_regs = mcp23s08_read_regs,
|
|
|
|
};
|
|
|
|
|
|
|
|
static const struct mcp23s08_ops mcp23s17_ops = {
|
|
|
|
.read = mcp23s17_read,
|
|
|
|
.write = mcp23s17_write,
|
|
|
|
.read_regs = mcp23s17_read_regs,
|
|
|
|
};
|
|
|
|
|
2011-07-15 10:25:31 +02:00
|
|
|
#endif /* CONFIG_SPI_MASTER */
|
2011-03-09 17:56:30 +01:00
|
|
|
|
2008-02-05 07:28:25 +01:00
|
|
|
/*----------------------------------------------------------------------*/
|
|
|
|
|
|
|
|
static int mcp23s08_direction_input(struct gpio_chip *chip, unsigned offset)
|
|
|
|
{
|
|
|
|
struct mcp23s08 *mcp = container_of(chip, struct mcp23s08, chip);
|
|
|
|
int status;
|
|
|
|
|
|
|
|
mutex_lock(&mcp->lock);
|
|
|
|
mcp->cache[MCP_IODIR] |= (1 << offset);
|
2011-03-09 17:56:30 +01:00
|
|
|
status = mcp->ops->write(mcp, MCP_IODIR, mcp->cache[MCP_IODIR]);
|
2008-02-05 07:28:25 +01:00
|
|
|
mutex_unlock(&mcp->lock);
|
|
|
|
return status;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int mcp23s08_get(struct gpio_chip *chip, unsigned offset)
|
|
|
|
{
|
|
|
|
struct mcp23s08 *mcp = container_of(chip, struct mcp23s08, chip);
|
|
|
|
int status;
|
|
|
|
|
|
|
|
mutex_lock(&mcp->lock);
|
|
|
|
|
|
|
|
/* REVISIT reading this clears any IRQ ... */
|
2011-03-09 17:56:30 +01:00
|
|
|
status = mcp->ops->read(mcp, MCP_GPIO);
|
2008-02-05 07:28:25 +01:00
|
|
|
if (status < 0)
|
|
|
|
status = 0;
|
|
|
|
else {
|
|
|
|
mcp->cache[MCP_GPIO] = status;
|
|
|
|
status = !!(status & (1 << offset));
|
|
|
|
}
|
|
|
|
mutex_unlock(&mcp->lock);
|
|
|
|
return status;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int __mcp23s08_set(struct mcp23s08 *mcp, unsigned mask, int value)
|
|
|
|
{
|
2011-03-09 17:56:30 +01:00
|
|
|
unsigned olat = mcp->cache[MCP_OLAT];
|
2008-02-05 07:28:25 +01:00
|
|
|
|
|
|
|
if (value)
|
|
|
|
olat |= mask;
|
|
|
|
else
|
|
|
|
olat &= ~mask;
|
|
|
|
mcp->cache[MCP_OLAT] = olat;
|
2011-03-09 17:56:30 +01:00
|
|
|
return mcp->ops->write(mcp, MCP_OLAT, olat);
|
2008-02-05 07:28:25 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
static void mcp23s08_set(struct gpio_chip *chip, unsigned offset, int value)
|
|
|
|
{
|
|
|
|
struct mcp23s08 *mcp = container_of(chip, struct mcp23s08, chip);
|
2011-03-09 17:56:30 +01:00
|
|
|
unsigned mask = 1 << offset;
|
2008-02-05 07:28:25 +01:00
|
|
|
|
|
|
|
mutex_lock(&mcp->lock);
|
|
|
|
__mcp23s08_set(mcp, mask, value);
|
|
|
|
mutex_unlock(&mcp->lock);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
mcp23s08_direction_output(struct gpio_chip *chip, unsigned offset, int value)
|
|
|
|
{
|
|
|
|
struct mcp23s08 *mcp = container_of(chip, struct mcp23s08, chip);
|
2011-03-09 17:56:30 +01:00
|
|
|
unsigned mask = 1 << offset;
|
2008-02-05 07:28:25 +01:00
|
|
|
int status;
|
|
|
|
|
|
|
|
mutex_lock(&mcp->lock);
|
|
|
|
status = __mcp23s08_set(mcp, mask, value);
|
|
|
|
if (status == 0) {
|
|
|
|
mcp->cache[MCP_IODIR] &= ~mask;
|
2011-03-09 17:56:30 +01:00
|
|
|
status = mcp->ops->write(mcp, MCP_IODIR, mcp->cache[MCP_IODIR]);
|
2008-02-05 07:28:25 +01:00
|
|
|
}
|
|
|
|
mutex_unlock(&mcp->lock);
|
|
|
|
return status;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*----------------------------------------------------------------------*/
|
|
|
|
|
|
|
|
#ifdef CONFIG_DEBUG_FS
|
|
|
|
|
|
|
|
#include <linux/seq_file.h>
|
|
|
|
|
|
|
|
/*
|
|
|
|
* This shows more info than the generic gpio dump code:
|
|
|
|
* pullups, deglitching, open drain drive.
|
|
|
|
*/
|
|
|
|
static void mcp23s08_dbg_show(struct seq_file *s, struct gpio_chip *chip)
|
|
|
|
{
|
|
|
|
struct mcp23s08 *mcp;
|
|
|
|
char bank;
|
2008-05-23 22:04:43 +02:00
|
|
|
int t;
|
2008-02-05 07:28:25 +01:00
|
|
|
unsigned mask;
|
|
|
|
|
|
|
|
mcp = container_of(chip, struct mcp23s08, chip);
|
|
|
|
|
|
|
|
/* NOTE: we only handle one bank for now ... */
|
2011-03-09 17:56:30 +01:00
|
|
|
bank = '0' + ((mcp->addr >> 1) & 0x7);
|
2008-02-05 07:28:25 +01:00
|
|
|
|
|
|
|
mutex_lock(&mcp->lock);
|
2011-03-09 17:56:30 +01:00
|
|
|
t = mcp->ops->read_regs(mcp, 0, mcp->cache, ARRAY_SIZE(mcp->cache));
|
2008-02-05 07:28:25 +01:00
|
|
|
if (t < 0) {
|
|
|
|
seq_printf(s, " I/O ERROR %d\n", t);
|
|
|
|
goto done;
|
|
|
|
}
|
|
|
|
|
2011-03-09 17:56:30 +01:00
|
|
|
for (t = 0, mask = 1; t < chip->ngpio; t++, mask <<= 1) {
|
2008-02-05 07:28:25 +01:00
|
|
|
const char *label;
|
|
|
|
|
|
|
|
label = gpiochip_is_requested(chip, t);
|
|
|
|
if (!label)
|
|
|
|
continue;
|
|
|
|
|
|
|
|
seq_printf(s, " gpio-%-3d P%c.%d (%-12s) %s %s %s",
|
|
|
|
chip->base + t, bank, t, label,
|
|
|
|
(mcp->cache[MCP_IODIR] & mask) ? "in " : "out",
|
|
|
|
(mcp->cache[MCP_GPIO] & mask) ? "hi" : "lo",
|
2012-04-25 11:51:53 +02:00
|
|
|
(mcp->cache[MCP_GPPU] & mask) ? "up" : " ");
|
2008-02-05 07:28:25 +01:00
|
|
|
/* NOTE: ignoring the irq-related registers */
|
|
|
|
seq_printf(s, "\n");
|
|
|
|
}
|
|
|
|
done:
|
|
|
|
mutex_unlock(&mcp->lock);
|
|
|
|
}
|
|
|
|
|
|
|
|
#else
|
|
|
|
#define mcp23s08_dbg_show NULL
|
|
|
|
#endif
|
|
|
|
|
|
|
|
/*----------------------------------------------------------------------*/
|
|
|
|
|
2011-07-15 10:25:31 +02:00
|
|
|
static int mcp23s08_probe_one(struct mcp23s08 *mcp, struct device *dev,
|
|
|
|
void *data, unsigned addr,
|
2011-03-09 17:56:30 +01:00
|
|
|
unsigned type, unsigned base, unsigned pullups)
|
2008-02-05 07:28:25 +01:00
|
|
|
{
|
2011-07-15 10:25:31 +02:00
|
|
|
int status;
|
2008-02-05 07:28:25 +01:00
|
|
|
|
|
|
|
mutex_init(&mcp->lock);
|
|
|
|
|
2011-07-15 10:25:31 +02:00
|
|
|
mcp->data = data;
|
|
|
|
mcp->addr = addr;
|
2008-02-05 07:28:25 +01:00
|
|
|
|
|
|
|
mcp->chip.direction_input = mcp23s08_direction_input;
|
|
|
|
mcp->chip.get = mcp23s08_get;
|
|
|
|
mcp->chip.direction_output = mcp23s08_direction_output;
|
|
|
|
mcp->chip.set = mcp23s08_set;
|
|
|
|
mcp->chip.dbg_show = mcp23s08_dbg_show;
|
|
|
|
|
2011-07-15 10:25:31 +02:00
|
|
|
switch (type) {
|
|
|
|
#ifdef CONFIG_SPI_MASTER
|
|
|
|
case MCP_TYPE_S08:
|
2011-03-09 17:56:30 +01:00
|
|
|
mcp->ops = &mcp23s08_ops;
|
|
|
|
mcp->chip.ngpio = 8;
|
|
|
|
mcp->chip.label = "mcp23s08";
|
2011-07-15 10:25:31 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
case MCP_TYPE_S17:
|
|
|
|
mcp->ops = &mcp23s17_ops;
|
|
|
|
mcp->chip.ngpio = 16;
|
|
|
|
mcp->chip.label = "mcp23s17";
|
|
|
|
break;
|
|
|
|
#endif /* CONFIG_SPI_MASTER */
|
|
|
|
|
mcp23s08: add i2c support
Add i2c bindings for the mcp230xx devices. This is quite a lot simpler
than the spi one as there's no funky sub addressing done (one struct
i2c_client per struct gpio_chip).
The mcp23s08_platform_data structure is reused for i2c, even though
only a single mcp23s08_chip_info structure is needed.
To use, simply fill out a platform_data structure and pass it in
i2c_board_info, E.G.:
static const struct mcp23s08_platform_data mcp23017_data = {
.chip[0] = {
.pullups = 0x00ff,
},
.base = 240,
};
static struct i2c_board_info __initdata i2c_devs[] = {
{ I2C_BOARD_INFO("mcp23017", 0x20),
.platform_data = &smartview_mcp23017_data, },
...
};
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2011-07-15 10:25:32 +02:00
|
|
|
#ifdef CONFIG_I2C
|
|
|
|
case MCP_TYPE_008:
|
|
|
|
mcp->ops = &mcp23008_ops;
|
|
|
|
mcp->chip.ngpio = 8;
|
|
|
|
mcp->chip.label = "mcp23008";
|
|
|
|
break;
|
|
|
|
|
|
|
|
case MCP_TYPE_017:
|
|
|
|
mcp->ops = &mcp23017_ops;
|
|
|
|
mcp->chip.ngpio = 16;
|
|
|
|
mcp->chip.label = "mcp23017";
|
|
|
|
break;
|
|
|
|
#endif /* CONFIG_I2C */
|
|
|
|
|
2011-07-15 10:25:31 +02:00
|
|
|
default:
|
|
|
|
dev_err(dev, "invalid device type (%d)\n", type);
|
|
|
|
return -EINVAL;
|
2011-03-09 17:56:30 +01:00
|
|
|
}
|
2011-07-15 10:25:31 +02:00
|
|
|
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
mcp->chip.base = base;
|
2008-02-05 07:28:25 +01:00
|
|
|
mcp->chip.can_sleep = 1;
|
2011-07-15 10:25:31 +02:00
|
|
|
mcp->chip.dev = dev;
|
2008-04-28 11:14:45 +02:00
|
|
|
mcp->chip.owner = THIS_MODULE;
|
2008-02-05 07:28:25 +01:00
|
|
|
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
/* verify MCP_IOCON.SEQOP = 0, so sequential reads work,
|
|
|
|
* and MCP_IOCON.HAEN = 1, so we work with all chips.
|
|
|
|
*/
|
2011-03-09 17:56:30 +01:00
|
|
|
status = mcp->ops->read(mcp, MCP_IOCON);
|
2008-02-05 07:28:25 +01:00
|
|
|
if (status < 0)
|
|
|
|
goto fail;
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
if ((status & IOCON_SEQOP) || !(status & IOCON_HAEN)) {
|
2011-03-09 17:56:30 +01:00
|
|
|
/* mcp23s17 has IOCON twice, make sure they are in sync */
|
|
|
|
status &= ~(IOCON_SEQOP | (IOCON_SEQOP << 8));
|
|
|
|
status |= IOCON_HAEN | (IOCON_HAEN << 8);
|
|
|
|
status = mcp->ops->write(mcp, MCP_IOCON, status);
|
2008-02-05 07:28:25 +01:00
|
|
|
if (status < 0)
|
|
|
|
goto fail;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* configure ~100K pullups */
|
2011-03-09 17:56:30 +01:00
|
|
|
status = mcp->ops->write(mcp, MCP_GPPU, pullups);
|
2008-02-05 07:28:25 +01:00
|
|
|
if (status < 0)
|
|
|
|
goto fail;
|
|
|
|
|
2011-03-09 17:56:30 +01:00
|
|
|
status = mcp->ops->read_regs(mcp, 0, mcp->cache, ARRAY_SIZE(mcp->cache));
|
2008-02-05 07:28:25 +01:00
|
|
|
if (status < 0)
|
|
|
|
goto fail;
|
|
|
|
|
|
|
|
/* disable inverter on input */
|
|
|
|
if (mcp->cache[MCP_IPOL] != 0) {
|
|
|
|
mcp->cache[MCP_IPOL] = 0;
|
2011-03-09 17:56:30 +01:00
|
|
|
status = mcp->ops->write(mcp, MCP_IPOL, 0);
|
|
|
|
if (status < 0)
|
|
|
|
goto fail;
|
2008-02-05 07:28:25 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/* disable irqs */
|
|
|
|
if (mcp->cache[MCP_GPINTEN] != 0) {
|
|
|
|
mcp->cache[MCP_GPINTEN] = 0;
|
2011-03-09 17:56:30 +01:00
|
|
|
status = mcp->ops->write(mcp, MCP_GPINTEN, 0);
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
if (status < 0)
|
|
|
|
goto fail;
|
2008-02-05 07:28:25 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
status = gpiochip_add(&mcp->chip);
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
fail:
|
|
|
|
if (status < 0)
|
2011-07-15 10:25:31 +02:00
|
|
|
dev_dbg(dev, "can't setup chip %d, --> %d\n",
|
|
|
|
addr, status);
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
return status;
|
|
|
|
}
|
|
|
|
|
mcp23s08: add i2c support
Add i2c bindings for the mcp230xx devices. This is quite a lot simpler
than the spi one as there's no funky sub addressing done (one struct
i2c_client per struct gpio_chip).
The mcp23s08_platform_data structure is reused for i2c, even though
only a single mcp23s08_chip_info structure is needed.
To use, simply fill out a platform_data structure and pass it in
i2c_board_info, E.G.:
static const struct mcp23s08_platform_data mcp23017_data = {
.chip[0] = {
.pullups = 0x00ff,
},
.base = 240,
};
static struct i2c_board_info __initdata i2c_devs[] = {
{ I2C_BOARD_INFO("mcp23017", 0x20),
.platform_data = &smartview_mcp23017_data, },
...
};
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2011-07-15 10:25:32 +02:00
|
|
|
/*----------------------------------------------------------------------*/
|
|
|
|
|
|
|
|
#ifdef CONFIG_I2C
|
|
|
|
|
|
|
|
static int __devinit mcp230xx_probe(struct i2c_client *client,
|
|
|
|
const struct i2c_device_id *id)
|
|
|
|
{
|
|
|
|
struct mcp23s08_platform_data *pdata;
|
|
|
|
struct mcp23s08 *mcp;
|
|
|
|
int status;
|
|
|
|
|
|
|
|
pdata = client->dev.platform_data;
|
|
|
|
if (!pdata || !gpio_is_valid(pdata->base)) {
|
|
|
|
dev_dbg(&client->dev, "invalid or missing platform data\n");
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
|
|
|
mcp = kzalloc(sizeof *mcp, GFP_KERNEL);
|
|
|
|
if (!mcp)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
status = mcp23s08_probe_one(mcp, &client->dev, client, client->addr,
|
|
|
|
id->driver_data, pdata->base,
|
|
|
|
pdata->chip[0].pullups);
|
|
|
|
if (status)
|
|
|
|
goto fail;
|
|
|
|
|
|
|
|
i2c_set_clientdata(client, mcp);
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
fail:
|
|
|
|
kfree(mcp);
|
|
|
|
|
|
|
|
return status;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int __devexit mcp230xx_remove(struct i2c_client *client)
|
|
|
|
{
|
|
|
|
struct mcp23s08 *mcp = i2c_get_clientdata(client);
|
|
|
|
int status;
|
|
|
|
|
|
|
|
status = gpiochip_remove(&mcp->chip);
|
|
|
|
if (status == 0)
|
|
|
|
kfree(mcp);
|
|
|
|
|
|
|
|
return status;
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct i2c_device_id mcp230xx_id[] = {
|
|
|
|
{ "mcp23008", MCP_TYPE_008 },
|
|
|
|
{ "mcp23017", MCP_TYPE_017 },
|
|
|
|
{ },
|
|
|
|
};
|
|
|
|
MODULE_DEVICE_TABLE(i2c, mcp230xx_id);
|
|
|
|
|
|
|
|
static struct i2c_driver mcp230xx_driver = {
|
|
|
|
.driver = {
|
|
|
|
.name = "mcp230xx",
|
|
|
|
.owner = THIS_MODULE,
|
|
|
|
},
|
|
|
|
.probe = mcp230xx_probe,
|
2012-11-19 19:20:08 +01:00
|
|
|
.remove = mcp230xx_remove,
|
mcp23s08: add i2c support
Add i2c bindings for the mcp230xx devices. This is quite a lot simpler
than the spi one as there's no funky sub addressing done (one struct
i2c_client per struct gpio_chip).
The mcp23s08_platform_data structure is reused for i2c, even though
only a single mcp23s08_chip_info structure is needed.
To use, simply fill out a platform_data structure and pass it in
i2c_board_info, E.G.:
static const struct mcp23s08_platform_data mcp23017_data = {
.chip[0] = {
.pullups = 0x00ff,
},
.base = 240,
};
static struct i2c_board_info __initdata i2c_devs[] = {
{ I2C_BOARD_INFO("mcp23017", 0x20),
.platform_data = &smartview_mcp23017_data, },
...
};
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2011-07-15 10:25:32 +02:00
|
|
|
.id_table = mcp230xx_id,
|
|
|
|
};
|
|
|
|
|
|
|
|
static int __init mcp23s08_i2c_init(void)
|
|
|
|
{
|
|
|
|
return i2c_add_driver(&mcp230xx_driver);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void mcp23s08_i2c_exit(void)
|
|
|
|
{
|
|
|
|
i2c_del_driver(&mcp230xx_driver);
|
|
|
|
}
|
|
|
|
|
|
|
|
#else
|
|
|
|
|
|
|
|
static int __init mcp23s08_i2c_init(void) { return 0; }
|
|
|
|
static void mcp23s08_i2c_exit(void) { }
|
|
|
|
|
|
|
|
#endif /* CONFIG_I2C */
|
|
|
|
|
|
|
|
/*----------------------------------------------------------------------*/
|
|
|
|
|
2011-07-15 10:25:31 +02:00
|
|
|
#ifdef CONFIG_SPI_MASTER
|
|
|
|
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
static int mcp23s08_probe(struct spi_device *spi)
|
|
|
|
{
|
|
|
|
struct mcp23s08_platform_data *pdata;
|
|
|
|
unsigned addr;
|
|
|
|
unsigned chips = 0;
|
|
|
|
struct mcp23s08_driver_data *data;
|
2011-03-09 17:56:30 +01:00
|
|
|
int status, type;
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
unsigned base;
|
|
|
|
|
2011-03-09 17:56:30 +01:00
|
|
|
type = spi_get_device_id(spi)->driver_data;
|
|
|
|
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
pdata = spi->dev.platform_data;
|
2009-01-15 22:50:45 +01:00
|
|
|
if (!pdata || !gpio_is_valid(pdata->base)) {
|
|
|
|
dev_dbg(&spi->dev, "invalid or missing platform data\n");
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
|
2011-03-09 17:56:30 +01:00
|
|
|
for (addr = 0; addr < ARRAY_SIZE(pdata->chip); addr++) {
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
if (!pdata->chip[addr].is_present)
|
|
|
|
continue;
|
|
|
|
chips++;
|
2011-03-09 17:56:30 +01:00
|
|
|
if ((type == MCP_TYPE_S08) && (addr > 3)) {
|
|
|
|
dev_err(&spi->dev,
|
|
|
|
"mcp23s08 only supports address 0..3\n");
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
}
|
|
|
|
if (!chips)
|
|
|
|
return -ENODEV;
|
|
|
|
|
|
|
|
data = kzalloc(sizeof *data + chips * sizeof(struct mcp23s08),
|
|
|
|
GFP_KERNEL);
|
|
|
|
if (!data)
|
|
|
|
return -ENOMEM;
|
|
|
|
spi_set_drvdata(spi, data);
|
|
|
|
|
|
|
|
base = pdata->base;
|
2011-03-09 17:56:30 +01:00
|
|
|
for (addr = 0; addr < ARRAY_SIZE(pdata->chip); addr++) {
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
if (!pdata->chip[addr].is_present)
|
|
|
|
continue;
|
|
|
|
chips--;
|
|
|
|
data->mcp[addr] = &data->chip[chips];
|
2011-07-15 10:25:31 +02:00
|
|
|
status = mcp23s08_probe_one(data->mcp[addr], &spi->dev, spi,
|
|
|
|
0x40 | (addr << 1), type, base,
|
2011-03-09 17:56:30 +01:00
|
|
|
pdata->chip[addr].pullups);
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
if (status < 0)
|
|
|
|
goto fail;
|
2011-03-09 17:56:30 +01:00
|
|
|
|
|
|
|
base += (type == MCP_TYPE_S17) ? 16 : 8;
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
}
|
|
|
|
data->ngpio = base - pdata->base;
|
2008-02-05 07:28:25 +01:00
|
|
|
|
|
|
|
/* NOTE: these chips have a relatively sane IRQ framework, with
|
|
|
|
* per-signal masking and level/edge triggering. It's not yet
|
|
|
|
* handled here...
|
|
|
|
*/
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
fail:
|
2011-03-09 17:56:30 +01:00
|
|
|
for (addr = 0; addr < ARRAY_SIZE(data->mcp); addr++) {
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
int tmp;
|
|
|
|
|
|
|
|
if (!data->mcp[addr])
|
|
|
|
continue;
|
|
|
|
tmp = gpiochip_remove(&data->mcp[addr]->chip);
|
|
|
|
if (tmp < 0)
|
|
|
|
dev_err(&spi->dev, "%s --> %d\n", "remove", tmp);
|
|
|
|
}
|
|
|
|
kfree(data);
|
2008-02-05 07:28:25 +01:00
|
|
|
return status;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int mcp23s08_remove(struct spi_device *spi)
|
|
|
|
{
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
struct mcp23s08_driver_data *data = spi_get_drvdata(spi);
|
|
|
|
unsigned addr;
|
2008-02-05 07:28:25 +01:00
|
|
|
int status = 0;
|
|
|
|
|
2011-03-09 17:56:30 +01:00
|
|
|
for (addr = 0; addr < ARRAY_SIZE(data->mcp); addr++) {
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
int tmp;
|
|
|
|
|
|
|
|
if (!data->mcp[addr])
|
|
|
|
continue;
|
|
|
|
|
|
|
|
tmp = gpiochip_remove(&data->mcp[addr]->chip);
|
|
|
|
if (tmp < 0) {
|
|
|
|
dev_err(&spi->dev, "%s --> %d\n", "remove", tmp);
|
|
|
|
status = tmp;
|
|
|
|
}
|
|
|
|
}
|
2008-02-05 07:28:25 +01:00
|
|
|
if (status == 0)
|
gpio: mcp23s08 handles multiple chips per chipselect
Teach the mcp23s08 driver about a curious feature of these chips: up to
four of them can share the same chipselect, with the SPI signals wired in
parallel, by matching two bits in the first protocol byte against two
address lines on the chip.
This is handled by three software changes:
* Platform data now holds an array of per-chip structs, not
just one chip's address and pullup configuration.
* Probe() and remove() now use another level of structure,
wrapping an instance of the original structure for each
mcp23s08 chip sharing that chipselect.
* The HAEN bit is set, so that the hardware address bits can no
longer be ignored (boot firmware may not have enabled them).
The "one struct per chip" preserves the guts of the current code,
but platform_data will need minor changes.
OLD:
/* incorrect "slave" ID may not have mattered */
.slave = 3,
.pullups = BIT(3) | BIT(1) | BIT(0),
NEW:
/* slave address _must_ match chip's wiring */
.chip[3] = {
.is_present = true,
.pullups = BIT(3) | BIT(1) | BIT(0),
},
There's no change in how things _behave_ for spi_device nodes with a
single mcp23s08 chip. New multi-chip configurations assign GPIOs in
sequence, without holes. The spi_device just resembles a bigger
controller, but internally it has multiple gpio_chip instances.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-07-25 10:46:09 +02:00
|
|
|
kfree(data);
|
2008-02-05 07:28:25 +01:00
|
|
|
return status;
|
|
|
|
}
|
|
|
|
|
2011-03-09 17:56:30 +01:00
|
|
|
static const struct spi_device_id mcp23s08_ids[] = {
|
|
|
|
{ "mcp23s08", MCP_TYPE_S08 },
|
|
|
|
{ "mcp23s17", MCP_TYPE_S17 },
|
|
|
|
{ },
|
|
|
|
};
|
|
|
|
MODULE_DEVICE_TABLE(spi, mcp23s08_ids);
|
|
|
|
|
2008-02-05 07:28:25 +01:00
|
|
|
static struct spi_driver mcp23s08_driver = {
|
|
|
|
.probe = mcp23s08_probe,
|
|
|
|
.remove = mcp23s08_remove,
|
2011-03-09 17:56:30 +01:00
|
|
|
.id_table = mcp23s08_ids,
|
2008-02-05 07:28:25 +01:00
|
|
|
.driver = {
|
|
|
|
.name = "mcp23s08",
|
|
|
|
.owner = THIS_MODULE,
|
|
|
|
},
|
|
|
|
};
|
|
|
|
|
2011-07-15 10:25:31 +02:00
|
|
|
static int __init mcp23s08_spi_init(void)
|
|
|
|
{
|
|
|
|
return spi_register_driver(&mcp23s08_driver);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void mcp23s08_spi_exit(void)
|
|
|
|
{
|
|
|
|
spi_unregister_driver(&mcp23s08_driver);
|
|
|
|
}
|
|
|
|
|
|
|
|
#else
|
|
|
|
|
|
|
|
static int __init mcp23s08_spi_init(void) { return 0; }
|
|
|
|
static void mcp23s08_spi_exit(void) { }
|
|
|
|
|
|
|
|
#endif /* CONFIG_SPI_MASTER */
|
|
|
|
|
2008-02-05 07:28:25 +01:00
|
|
|
/*----------------------------------------------------------------------*/
|
|
|
|
|
|
|
|
static int __init mcp23s08_init(void)
|
|
|
|
{
|
mcp23s08: add i2c support
Add i2c bindings for the mcp230xx devices. This is quite a lot simpler
than the spi one as there's no funky sub addressing done (one struct
i2c_client per struct gpio_chip).
The mcp23s08_platform_data structure is reused for i2c, even though
only a single mcp23s08_chip_info structure is needed.
To use, simply fill out a platform_data structure and pass it in
i2c_board_info, E.G.:
static const struct mcp23s08_platform_data mcp23017_data = {
.chip[0] = {
.pullups = 0x00ff,
},
.base = 240,
};
static struct i2c_board_info __initdata i2c_devs[] = {
{ I2C_BOARD_INFO("mcp23017", 0x20),
.platform_data = &smartview_mcp23017_data, },
...
};
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2011-07-15 10:25:32 +02:00
|
|
|
int ret;
|
|
|
|
|
|
|
|
ret = mcp23s08_spi_init();
|
|
|
|
if (ret)
|
|
|
|
goto spi_fail;
|
|
|
|
|
|
|
|
ret = mcp23s08_i2c_init();
|
|
|
|
if (ret)
|
|
|
|
goto i2c_fail;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
i2c_fail:
|
|
|
|
mcp23s08_spi_exit();
|
|
|
|
spi_fail:
|
|
|
|
return ret;
|
2008-02-05 07:28:25 +01:00
|
|
|
}
|
mcp23s08: add i2c support
Add i2c bindings for the mcp230xx devices. This is quite a lot simpler
than the spi one as there's no funky sub addressing done (one struct
i2c_client per struct gpio_chip).
The mcp23s08_platform_data structure is reused for i2c, even though
only a single mcp23s08_chip_info structure is needed.
To use, simply fill out a platform_data structure and pass it in
i2c_board_info, E.G.:
static const struct mcp23s08_platform_data mcp23017_data = {
.chip[0] = {
.pullups = 0x00ff,
},
.base = 240,
};
static struct i2c_board_info __initdata i2c_devs[] = {
{ I2C_BOARD_INFO("mcp23017", 0x20),
.platform_data = &smartview_mcp23017_data, },
...
};
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2011-07-15 10:25:32 +02:00
|
|
|
/* register after spi/i2c postcore initcall and before
|
2008-10-16 07:02:46 +02:00
|
|
|
* subsys initcalls that may rely on these GPIOs
|
|
|
|
*/
|
|
|
|
subsys_initcall(mcp23s08_init);
|
2008-02-05 07:28:25 +01:00
|
|
|
|
|
|
|
static void __exit mcp23s08_exit(void)
|
|
|
|
{
|
2011-07-15 10:25:31 +02:00
|
|
|
mcp23s08_spi_exit();
|
mcp23s08: add i2c support
Add i2c bindings for the mcp230xx devices. This is quite a lot simpler
than the spi one as there's no funky sub addressing done (one struct
i2c_client per struct gpio_chip).
The mcp23s08_platform_data structure is reused for i2c, even though
only a single mcp23s08_chip_info structure is needed.
To use, simply fill out a platform_data structure and pass it in
i2c_board_info, E.G.:
static const struct mcp23s08_platform_data mcp23017_data = {
.chip[0] = {
.pullups = 0x00ff,
},
.base = 240,
};
static struct i2c_board_info __initdata i2c_devs[] = {
{ I2C_BOARD_INFO("mcp23017", 0x20),
.platform_data = &smartview_mcp23017_data, },
...
};
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2011-07-15 10:25:32 +02:00
|
|
|
mcp23s08_i2c_exit();
|
2008-02-05 07:28:25 +01:00
|
|
|
}
|
|
|
|
module_exit(mcp23s08_exit);
|
|
|
|
|
|
|
|
MODULE_LICENSE("GPL");
|