2009-06-20 05:11:00 +04:00
|
|
|
/*
|
|
|
|
|
* Coldfire generic GPIO support.
|
|
|
|
|
*
|
|
|
|
|
* (C) Copyright 2009, Steven King <sfking@fdwdc.com>
|
|
|
|
|
*
|
|
|
|
|
* 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; version 2 of the License.
|
|
|
|
|
*
|
|
|
|
|
* 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.
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
#ifndef mcfgpio_h
|
|
|
|
|
#define mcfgpio_h
|
|
|
|
|
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
#ifdef CONFIG_GPIOLIB
|
2009-06-20 05:11:00 +04:00
|
|
|
#include <asm-generic/gpio.h>
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
#else
|
|
|
|
|
|
|
|
|
|
int __mcfgpio_get_value(unsigned gpio);
|
|
|
|
|
void __mcfgpio_set_value(unsigned gpio, int value);
|
|
|
|
|
int __mcfgpio_direction_input(unsigned gpio);
|
|
|
|
|
int __mcfgpio_direction_output(unsigned gpio, int value);
|
|
|
|
|
int __mcfgpio_request(unsigned gpio);
|
|
|
|
|
void __mcfgpio_free(unsigned gpio);
|
|
|
|
|
|
|
|
|
|
/* our alternate 'gpiolib' functions */
|
|
|
|
|
static inline int __gpio_get_value(unsigned gpio)
|
|
|
|
|
{
|
|
|
|
|
if (gpio < MCFGPIO_PIN_MAX)
|
|
|
|
|
return __mcfgpio_get_value(gpio);
|
|
|
|
|
else
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static inline void __gpio_set_value(unsigned gpio, int value)
|
|
|
|
|
{
|
|
|
|
|
if (gpio < MCFGPIO_PIN_MAX)
|
|
|
|
|
__mcfgpio_set_value(gpio, value);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static inline int __gpio_cansleep(unsigned gpio)
|
|
|
|
|
{
|
|
|
|
|
if (gpio < MCFGPIO_PIN_MAX)
|
|
|
|
|
return 0;
|
|
|
|
|
else
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static inline int __gpio_to_irq(unsigned gpio)
|
|
|
|
|
{
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static inline int gpio_direction_input(unsigned gpio)
|
|
|
|
|
{
|
|
|
|
|
if (gpio < MCFGPIO_PIN_MAX)
|
|
|
|
|
return __mcfgpio_direction_input(gpio);
|
|
|
|
|
else
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static inline int gpio_direction_output(unsigned gpio, int value)
|
|
|
|
|
{
|
|
|
|
|
if (gpio < MCFGPIO_PIN_MAX)
|
|
|
|
|
return __mcfgpio_direction_output(gpio, value);
|
|
|
|
|
else
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static inline int gpio_request(unsigned gpio, const char *label)
|
|
|
|
|
{
|
|
|
|
|
if (gpio < MCFGPIO_PIN_MAX)
|
|
|
|
|
return __mcfgpio_request(gpio);
|
|
|
|
|
else
|
|
|
|
|
return -EINVAL;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static inline void gpio_free(unsigned gpio)
|
|
|
|
|
{
|
|
|
|
|
if (gpio < MCFGPIO_PIN_MAX)
|
|
|
|
|
__mcfgpio_free(gpio);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#endif /* CONFIG_GPIOLIB */
|
2009-06-20 05:11:00 +04:00
|
|
|
|
|
|
|
|
|
2012-04-16 09:59:29 +04:00
|
|
|
/*
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
* The Freescale Coldfire family is quite varied in how they implement GPIO.
|
|
|
|
|
* Some parts have 8 bit ports, some have 16bit and some have 32bit; some have
|
|
|
|
|
* only one port, others have multiple ports; some have a single data latch
|
|
|
|
|
* for both input and output, others have a separate pin data register to read
|
|
|
|
|
* input; some require a read-modify-write access to change an output, others
|
|
|
|
|
* have set and clear registers for some of the outputs; Some have all the
|
|
|
|
|
* GPIOs in a single control area, others have some GPIOs implemented in
|
|
|
|
|
* different modules.
|
2012-04-16 09:59:29 +04:00
|
|
|
*
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
* This implementation attempts accommodate the differences while presenting
|
|
|
|
|
* a generic interface that will optimize to as few instructions as possible.
|
2012-04-16 09:59:29 +04:00
|
|
|
*/
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
#if defined(CONFIG_M5206) || defined(CONFIG_M5206e) || \
|
|
|
|
|
defined(CONFIG_M520x) || defined(CONFIG_M523x) || \
|
|
|
|
|
defined(CONFIG_M527x) || defined(CONFIG_M528x) || \
|
2012-06-07 01:28:31 +04:00
|
|
|
defined(CONFIG_M532x) || defined(CONFIG_M54xx) || \
|
|
|
|
|
defined(CONFIG_M5441x)
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
|
|
|
|
|
/* These parts have GPIO organized by 8 bit ports */
|
|
|
|
|
|
|
|
|
|
#define MCFGPIO_PORTTYPE u8
|
|
|
|
|
#define MCFGPIO_PORTSIZE 8
|
|
|
|
|
#define mcfgpio_read(port) __raw_readb(port)
|
|
|
|
|
#define mcfgpio_write(data, port) __raw_writeb(data, port)
|
|
|
|
|
|
|
|
|
|
#elif defined(CONFIG_M5307) || defined(CONFIG_M5407) || defined(CONFIG_M5272)
|
|
|
|
|
|
|
|
|
|
/* These parts have GPIO organized by 16 bit ports */
|
|
|
|
|
|
|
|
|
|
#define MCFGPIO_PORTTYPE u16
|
|
|
|
|
#define MCFGPIO_PORTSIZE 16
|
|
|
|
|
#define mcfgpio_read(port) __raw_readw(port)
|
|
|
|
|
#define mcfgpio_write(data, port) __raw_writew(data, port)
|
|
|
|
|
|
2012-06-05 19:23:08 +04:00
|
|
|
#elif defined(CONFIG_M5249) || defined(CONFIG_M525x)
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
|
|
|
|
|
/* These parts have GPIO organized by 32 bit ports */
|
|
|
|
|
|
|
|
|
|
#define MCFGPIO_PORTTYPE u32
|
|
|
|
|
#define MCFGPIO_PORTSIZE 32
|
|
|
|
|
#define mcfgpio_read(port) __raw_readl(port)
|
|
|
|
|
#define mcfgpio_write(data, port) __raw_writel(data, port)
|
|
|
|
|
|
|
|
|
|
#endif
|
2012-04-16 09:59:29 +04:00
|
|
|
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
#define mcfgpio_bit(gpio) (1 << ((gpio) % MCFGPIO_PORTSIZE))
|
|
|
|
|
#define mcfgpio_port(gpio) ((gpio) / MCFGPIO_PORTSIZE)
|
|
|
|
|
|
|
|
|
|
#if defined(CONFIG_M520x) || defined(CONFIG_M523x) || \
|
2012-06-07 01:28:31 +04:00
|
|
|
defined(CONFIG_M527x) || defined(CONFIG_M528x) || \
|
|
|
|
|
defined(CONFIG_M532x) || defined(CONFIG_M5441x)
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
/*
|
|
|
|
|
* These parts have an 'Edge' Port module (external interrupt/GPIO) which uses
|
|
|
|
|
* read-modify-write to change an output and a GPIO module which has separate
|
|
|
|
|
* set/clr registers to directly change outputs with a single write access.
|
|
|
|
|
*/
|
|
|
|
|
#if defined(CONFIG_M528x)
|
2012-04-16 09:59:29 +04:00
|
|
|
/*
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
* The 528x also has GPIOs in other modules (GPT, QADC) which use
|
|
|
|
|
* read-modify-write as well as those controlled by the EPORT and GPIO modules.
|
2012-04-16 09:59:29 +04:00
|
|
|
*/
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
#define MCFGPIO_SCR_START 40
|
2012-06-07 01:28:31 +04:00
|
|
|
#elif defined(CONFIGM5441x)
|
|
|
|
|
/* The m5441x EPORT doesn't have its own GPIO port, uses PORT C */
|
|
|
|
|
#define MCFGPIO_SCR_START 0
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
#else
|
|
|
|
|
#define MCFGPIO_SCR_START 8
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
#define MCFGPIO_SETR_PORT(gpio) (MCFGPIO_SETR + \
|
|
|
|
|
mcfgpio_port(gpio - MCFGPIO_SCR_START))
|
|
|
|
|
|
|
|
|
|
#define MCFGPIO_CLRR_PORT(gpio) (MCFGPIO_CLRR + \
|
|
|
|
|
mcfgpio_port(gpio - MCFGPIO_SCR_START))
|
|
|
|
|
#else
|
|
|
|
|
|
|
|
|
|
#define MCFGPIO_SCR_START MCFGPIO_PIN_MAX
|
|
|
|
|
/* with MCFGPIO_SCR == MCFGPIO_PIN_MAX, these will be optimized away */
|
|
|
|
|
#define MCFGPIO_SETR_PORT(gpio) 0
|
|
|
|
|
#define MCFGPIO_CLRR_PORT(gpio) 0
|
2012-04-16 09:59:29 +04:00
|
|
|
|
2009-06-20 05:11:00 +04:00
|
|
|
#endif
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
/*
|
|
|
|
|
* Coldfire specific helper functions
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
/* return the port pin data register for a gpio */
|
|
|
|
|
static inline u32 __mcfgpio_ppdr(unsigned gpio)
|
|
|
|
|
{
|
|
|
|
|
#if defined(CONFIG_M5206) || defined(CONFIG_M5206e) || \
|
|
|
|
|
defined(CONFIG_M5307) || defined(CONFIG_M5407)
|
|
|
|
|
return MCFSIM_PADAT;
|
|
|
|
|
#elif defined(CONFIG_M5272)
|
|
|
|
|
if (gpio < 16)
|
|
|
|
|
return MCFSIM_PADAT;
|
|
|
|
|
else if (gpio < 32)
|
|
|
|
|
return MCFSIM_PBDAT;
|
|
|
|
|
else
|
|
|
|
|
return MCFSIM_PCDAT;
|
2012-06-05 19:23:08 +04:00
|
|
|
#elif defined(CONFIG_M5249) || defined(CONFIG_M525x)
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
if (gpio < 32)
|
|
|
|
|
return MCFSIM2_GPIOREAD;
|
|
|
|
|
else
|
|
|
|
|
return MCFSIM2_GPIO1READ;
|
|
|
|
|
#elif defined(CONFIG_M520x) || defined(CONFIG_M523x) || \
|
2012-06-07 01:28:31 +04:00
|
|
|
defined(CONFIG_M527x) || defined(CONFIG_M528x) || \
|
|
|
|
|
defined(CONFIG_M532x) || defined(CONFIG_M5441x)
|
|
|
|
|
#if !defined(CONFIG_M5441x)
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
if (gpio < 8)
|
|
|
|
|
return MCFEPORT_EPPDR;
|
|
|
|
|
#if defined(CONFIG_M528x)
|
|
|
|
|
else if (gpio < 16)
|
|
|
|
|
return MCFGPTA_GPTPORT;
|
|
|
|
|
else if (gpio < 24)
|
|
|
|
|
return MCFGPTB_GPTPORT;
|
|
|
|
|
else if (gpio < 32)
|
|
|
|
|
return MCFQADC_PORTQA;
|
|
|
|
|
else if (gpio < 40)
|
|
|
|
|
return MCFQADC_PORTQB;
|
2012-06-07 01:28:31 +04:00
|
|
|
#endif /* defined(CONFIG_M528x) */
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
else
|
2012-06-07 01:28:31 +04:00
|
|
|
#endif /* !defined(CONFIG_M5441x) */
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
return MCFGPIO_PPDR + mcfgpio_port(gpio - MCFGPIO_SCR_START);
|
|
|
|
|
#else
|
|
|
|
|
return 0;
|
|
|
|
|
#endif
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* return the port output data register for a gpio */
|
|
|
|
|
static inline u32 __mcfgpio_podr(unsigned gpio)
|
|
|
|
|
{
|
|
|
|
|
#if defined(CONFIG_M5206) || defined(CONFIG_M5206e) || \
|
|
|
|
|
defined(CONFIG_M5307) || defined(CONFIG_M5407)
|
|
|
|
|
return MCFSIM_PADAT;
|
|
|
|
|
#elif defined(CONFIG_M5272)
|
|
|
|
|
if (gpio < 16)
|
|
|
|
|
return MCFSIM_PADAT;
|
|
|
|
|
else if (gpio < 32)
|
|
|
|
|
return MCFSIM_PBDAT;
|
|
|
|
|
else
|
|
|
|
|
return MCFSIM_PCDAT;
|
2012-06-05 19:23:08 +04:00
|
|
|
#elif defined(CONFIG_M5249) || defined(CONFIG_M525x)
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
if (gpio < 32)
|
|
|
|
|
return MCFSIM2_GPIOWRITE;
|
|
|
|
|
else
|
|
|
|
|
return MCFSIM2_GPIO1WRITE;
|
|
|
|
|
#elif defined(CONFIG_M520x) || defined(CONFIG_M523x) || \
|
2012-06-07 01:28:31 +04:00
|
|
|
defined(CONFIG_M527x) || defined(CONFIG_M528x) || \
|
|
|
|
|
defined(CONFIG_M532x) || defined(CONFIG_M5441x)
|
|
|
|
|
#if !defined(CONFIG_M5441x)
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
if (gpio < 8)
|
|
|
|
|
return MCFEPORT_EPDR;
|
|
|
|
|
#if defined(CONFIG_M528x)
|
|
|
|
|
else if (gpio < 16)
|
|
|
|
|
return MCFGPTA_GPTPORT;
|
|
|
|
|
else if (gpio < 24)
|
|
|
|
|
return MCFGPTB_GPTPORT;
|
|
|
|
|
else if (gpio < 32)
|
|
|
|
|
return MCFQADC_PORTQA;
|
|
|
|
|
else if (gpio < 40)
|
|
|
|
|
return MCFQADC_PORTQB;
|
2012-06-07 01:28:31 +04:00
|
|
|
#endif /* defined(CONFIG_M528x) */
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
else
|
2012-06-07 01:28:31 +04:00
|
|
|
#endif /* !defined(CONFIG_M5441x) */
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
return MCFGPIO_PODR + mcfgpio_port(gpio - MCFGPIO_SCR_START);
|
|
|
|
|
#else
|
|
|
|
|
return 0;
|
|
|
|
|
#endif
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* return the port direction data register for a gpio */
|
|
|
|
|
static inline u32 __mcfgpio_pddr(unsigned gpio)
|
|
|
|
|
{
|
|
|
|
|
#if defined(CONFIG_M5206) || defined(CONFIG_M5206e) || \
|
|
|
|
|
defined(CONFIG_M5307) || defined(CONFIG_M5407)
|
|
|
|
|
return MCFSIM_PADDR;
|
|
|
|
|
#elif defined(CONFIG_M5272)
|
|
|
|
|
if (gpio < 16)
|
|
|
|
|
return MCFSIM_PADDR;
|
|
|
|
|
else if (gpio < 32)
|
|
|
|
|
return MCFSIM_PBDDR;
|
|
|
|
|
else
|
|
|
|
|
return MCFSIM_PCDDR;
|
2012-06-05 19:23:08 +04:00
|
|
|
#elif defined(CONFIG_M5249) || defined(CONFIG_M525x)
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
if (gpio < 32)
|
|
|
|
|
return MCFSIM2_GPIOENABLE;
|
|
|
|
|
else
|
|
|
|
|
return MCFSIM2_GPIO1ENABLE;
|
|
|
|
|
#elif defined(CONFIG_M520x) || defined(CONFIG_M523x) || \
|
2012-06-07 01:28:31 +04:00
|
|
|
defined(CONFIG_M527x) || defined(CONFIG_M528x) || \
|
|
|
|
|
defined(CONFIG_M532x) || defined(CONFIG_M5441x)
|
|
|
|
|
#if !defined(CONFIG_M5441x)
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
if (gpio < 8)
|
|
|
|
|
return MCFEPORT_EPDDR;
|
|
|
|
|
#if defined(CONFIG_M528x)
|
|
|
|
|
else if (gpio < 16)
|
|
|
|
|
return MCFGPTA_GPTDDR;
|
|
|
|
|
else if (gpio < 24)
|
|
|
|
|
return MCFGPTB_GPTDDR;
|
|
|
|
|
else if (gpio < 32)
|
|
|
|
|
return MCFQADC_DDRQA;
|
|
|
|
|
else if (gpio < 40)
|
|
|
|
|
return MCFQADC_DDRQB;
|
2012-06-07 01:28:31 +04:00
|
|
|
#endif /* defined(CONFIG_M528x) */
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
else
|
2012-06-07 01:28:31 +04:00
|
|
|
#endif /* !defined(CONFIG_M5441x) */
|
m68knommu: refactor Coldfire GPIO not to require GPIOLIB, eliminate mcf_gpio_chips.
If we're not connecting external GPIO extenders via i2c or spi or whatever, we
probably don't need GPIOLIB. If we provide an alternate implementation of
the GPIOLIB functions to use when only on-chip GPIO is needed, we can change
ARCH_REQUIRE_GPIOLIB to ARCH_WANTS_OPTIONAL_GPIOLIB so that GPIOLIB becomes
optional.
The downside is that in the GPIOLIB=n case, we lose all error checking done by
gpiolib, ie multiply allocating the gpio, free'ing gpio etc., so that the
only checking that can be done is if we reference a gpio on an external part.
Targets that need the extra error checking can still select GPIOLIB=y.
For the case where GPIOLIB=y, we can simplify the table of gpio chips to use a
single chip, eliminating the tables of chips in the 5xxx.c files. The
original motivation for the definition of multiple chips was to match the way
many of the Coldfire variants defined their gpio as a spare array in memory.
However, all this really gains us is some error checking when we request a
gpio, gpiolib can check that it doesn't fall in one of the holes. If thats
important, I think we can still come up with a better way of accomplishing
that.
Also in this patch is some general cleanup and reorganizing of the gpio header
files (I'm sure I must have had a reason why I sometimes used a prefix of
mcf_gpio and other times mcfgpio but for the life of me I can't think of it
now).
Signed-off-by: Steven King <sfking@fdwdc.com>
Signed-off-by: Greg Ungerer <gerg@uclinux.org>
2012-05-22 00:10:19 +04:00
|
|
|
return MCFGPIO_PDDR + mcfgpio_port(gpio - MCFGPIO_SCR_START);
|
|
|
|
|
#else
|
|
|
|
|
return 0;
|
|
|
|
|
#endif
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
#endif /* mcfgpio_h */
|
