gpiolib: remove irq_to_gpio() definition
Subsystems (5)
RFC 1/3 gpiolib: remove irq_to_gpio() definition
From: Arnd Bergmann <[email protected]>
All implementations other than coldfire have returned an error since the avr32 and blackfin architectures got removed, and the last user in driver code was removed in 2016, so just remove this old interface.
The only reference is now in the Chinese documentation, which should be changed to remove this reference as well.
--- Documentation/driver-api/gpio/legacy.rst | 20 +++++--------------- arch/m68k/include/asm/gpio.h | 7 ------- arch/sh/include/asm/gpio.h | 5 ----- include/linux/gpio.h | 12 ------------ 4 files changed, 5 insertions(+), 39 deletions(-)
Documentation/driver-api/gpio/legacy.rst +5 -15
@@ -382,22 +382,18 @@ GPIOs mapped to IRQs -------------------- GPIO numbers are unsigned integers; so are IRQ numbers. These make up two logically distinct namespaces (GPIO 0 need not use IRQ 0). You can -map between them using calls like:: +map between them using:: /* map GPIO numbers to IRQ numbers */ int gpio_to_irq(unsigned gpio); - /* map IRQ numbers to GPIO numbers (avoid using this) */ - int irq_to_gpio(unsigned irq); - -Those return either the corresponding number in the other namespace, or +This returns an irq number corresponding to the gpio number, or else a negative errno code if the mapping can't be done. (For example, some GPIOs can't be used as IRQs.) It is an unchecked error to use a GPIO -number that wasn't set up as an input using gpio_direction_input(), or -to use an IRQ number that didn't originally come from gpio_to_irq(). +number that wasn't set up as an input using gpio_direction_input(). -These two mapping calls are expected to cost on the order of a single -addition or subtraction. They're not allowed to sleep. +The mapping call is expected to cost on the order of a single +addition or subtraction. It is not allowed to sleep. Non-error values returned from gpio_to_irq() can be passed to request_irq() or free_irq(). They will often be stored into IRQ resources for platform
@@ -405,12 +401,6 @@ devices, by the board-specific initialization code. Note that IRQ trigger options are part of the IRQ interface, e.g. IRQF_TRIGGER_FALLING, as are system wakeup capabilities. -Non-error values returned from irq_to_gpio() would most commonly be used -with gpio_get_value(), for example to initialize or update driver state -when the IRQ is edge-triggered. Note that some platforms don't support -this reverse mapping, so you should avoid using it. - - Emulating Open Drain Signals ---------------------------- Sometimes shared signals need to use "open drain" signaling, where only the
arch/m68k/include/asm/gpio.h +0 -7
@@ -66,13 +66,6 @@ static inline int gpio_to_irq(unsigned gpio)return __gpio_to_irq (gpio); } -static inline int irq_to_gpio (unsigned irq) -{ -return (irq >=MCFGPIO_IRQ_VECBASE && - irq < (MCFGPIO_IRQ_VECBASE +MCFGPIO_IRQ_MAX )) ? - irq -MCFGPIO_IRQ_VECBASE : -ENXIO ; -} -static inline int gpio_cansleep (unsigned gpio) {return gpio <MCFGPIO_PIN_MAX ?0 :__gpio_cansleep (gpio);
arch/sh/include/asm/gpio.h +0 -5
@@ -40,11 +40,6 @@ static inline int gpio_to_irq(unsigned gpio)return __gpio_to_irq (gpio); } -static inline int irq_to_gpio (unsigned int irq) -{ -return -ENOSYS ; -} -#endif /* CONFIG_GPIOLIB */ #endif /* __ASM_SH_GPIO_H */
include/linux/gpio.h +0 -12
@@ -81,11 +81,6 @@ static inline int gpio_to_irq(unsigned int gpio) return __gpio_to_irq(gpio); } -static inline int irq_to_gpio (unsigned int irq) -{ -return -EINVAL ; -} -#endif /* ! CONFIG_ARCH_HAVE_CUSTOM_GPIO_H */ /* CONFIG_GPIOLIB: bindings for managed devices that want to request gpios */
@@ -219,13 +214,6 @@ static inline int gpio_to_irq(unsigned gpio)return -EINVAL ; } -static inline int irq_to_gpio (unsigned irq) -{ -/* irq can never have been returned from gpio_to_irq() */ -WARN_ON (1 ); -return -EINVAL ; -} -static inline int devm_gpio_request (struct device *dev,unsigned gpio,const char *label) {
Signature · 1 line
2.29.2
RFC 2/3 gpiolib: remove empty asm/gpio.h files
From: Arnd Bergmann <[email protected]>
The arm and sh versions of this file are identical to the generic versions and can just be removed.
The drivers that actually use the sh3 specific version also include cpu/gpio.h directly. This leaves coldfire as the only gpio driver that needs something custom for gpiolib.
--- arch/arm/Kconfig | 1 - arch/arm/include/asm/gpio.h | 22 ------------------ arch/sh/Kconfig | 1 - arch/sh/include/asm/gpio.h | 45 ------------------------------------- 4 files changed, 69 deletions(-) delete mode 100644 arch/arm/include/asm/gpio.h delete mode 100644 arch/sh/include/asm/gpio.h
arch/arm/Kconfig +0 -1
@@ -23,7 +23,6 @@ config ARM select ARCH_HAS_SYNC_DMA_FOR_CPU if SWIOTLB || !MMU select ARCH_HAS_TEARDOWN_DMA_OPS if MMU select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST - select ARCH_HAVE_CUSTOM_GPIO_H select ARCH_HAVE_NMI_SAFE_CMPXCHG if CPU_V7 || CPU_V7M || CPU_V6K select ARCH_HAS_GCOV_PROFILE_ALL select ARCH_KEEP_MEMBLOCK
arch/arm/include/asm/gpio.h deleted +0 -22
@@ -1,22 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef _ARCH_ARM_GPIO_H -#define _ARCH_ARM_GPIO_H - -/* Note: this may rely upon the value of ARCH_NR_GPIOS set in mach/gpio.h */ -#include <asm-generic/gpio.h> - -/* The trivial gpiolib dispatchers */ -#define gpio_get_value __gpio_get_value -#define gpio_set_value __gpio_set_value -#define gpio_cansleep __gpio_cansleep - -/* -* Provide a default gpio_to_irq() which should satisfy every case. -* However, some platforms want to do this differently, so allow them -* to override it. -*/ -#ifndef gpio_to_irq -#define gpio_to_irq __gpio_to_irq -#endif - -#endif /* _ARCH_ARM_GPIO_H */
arch/sh/Kconfig +0 -1
@@ -4,7 +4,6 @@ config SUPERH select ARCH_32BIT_OFF_T select ARCH_ENABLE_MEMORY_HOTPLUG if SPARSEMEM && MMU select ARCH_ENABLE_MEMORY_HOTREMOVE if SPARSEMEM && MMU - select ARCH_HAVE_CUSTOM_GPIO_H select ARCH_HAVE_NMI_SAFE_CMPXCHG if (GUSA_RB || CPU_SH4A) select ARCH_HAS_BINFMT_FLAT if !MMU select ARCH_HAS_GIGANTIC_PAGE
arch/sh/include/asm/gpio.h deleted +0 -45
45 lines not on this page · raw message
Signature · 1 line
2.29.2
Re: [RFC 2/3] gpiolib: remove empty asm/gpio.h files
Hi Arnd,
Quoted 7 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
The arm and sh versions of this file are identical to the generic versions and can just be removed.
The drivers that actually use the sh3 specific version also include
cpu/gpio.hdirectly. This leaves coldfire as the only gpio driver
All but arch/sh/boards/board-magicpanelr2.c do. So that one needs a direct inclusion of <cpu/gpio.h>.
Quoted 3 lines
that needs something custom for gpiolib.
Signed-off-by Arnd Bergmann <[email protected]>
With the above fixed:
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]
In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
Re: [RFC 2/3] gpiolib: remove empty asm/gpio.h files
In reply to Arnd BergmannQuoted 10 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
The arm and sh versions of this file are identical to the generic versions and can just be removed.
The drivers that actually use the sh3 specific version also include
cpu/gpio.hdirectly. This leaves coldfire as the only gpio driver that needs something custom for gpiolib.Signed-off-by Arnd Bergmann <[email protected]>
Excellent!
I suggest Bartosz can queue this in the GPIO tree.
Yours,
Linus Walleij
Re: [RFC 2/3] gpiolib: remove empty asm/gpio.h files
Quoted 12 lines · Linus Walleij
Quoted 9 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
The arm and sh versions of this file are identical to the generic versions and can just be removed.
The drivers that actually use the sh3 specific version also include
cpu/gpio.hdirectly. This leaves coldfire as the only gpio driver that needs something custom for gpiolib.
Quoted 6 lines
Quoted 1 line
Signed-off-by Arnd Bergmann <[email protected]>Excellent!
Reviewed-by Linus Walleij <[email protected]>I suggest Bartosz can queue this in the GPIO tree.
I just sent v2 with slightly updated changelog texts and additional patches but without your Reviewed-by:. I suppose I'll have a v3 after the merge window, let's use that.
Arnd
RFC 3/3 gpiolib: coldfire: remove custom asm/gpio.h
In reply to Arnd BergmannFrom: Arnd Bergmann <[email protected]>
Now that coldfire is the only user of a custom asm/gpio.h, it seems better to remove this as well, and have the same interface everywhere.
For the gpio_get_value()/gpio_set_value()/gpio_to_irq(), gpio_cansleep() functions, the custom version is only a micro-optimization to inline the function for constant GPIO numbers. However, in the coldfire defconfigs, I was unable to find a single instance where this micro-optimization was even used, so to my best knowledge removing this has no downsides.
The custom gpio_request_one() function is even less useful, as it is guarded by an #ifdef that is never true.
--- arch/m68k/Kconfig.cpu | 1 - arch/m68k/include/asm/gpio.h | 95 ------------------------------------ drivers/gpio/Kconfig | 8 --- include/linux/gpio.h | 7 --- 4 files changed, 111 deletions(-) delete mode 100644 arch/m68k/include/asm/gpio.h
arch/m68k/Kconfig.cpu +0 -1
@@ -24,7 +24,6 @@ config M68KCLASSIC config COLDFIRE bool "Coldfire CPU family support" - select ARCH_HAVE_CUSTOM_GPIO_H select CPU_HAS_NO_BITFIELDS select CPU_HAS_NO_CAS select CPU_HAS_NO_MULDIV64
arch/m68k/include/asm/gpio.h deleted +0 -95
95 lines not on this page · raw message
drivers/gpio/Kconfig +0 -8
@@ -3,14 +3,6 @@ # GPIO infrastructure and drivers # -config ARCH_HAVE_CUSTOM_GPIO_H - bool - help - Selecting this config option from the architecture Kconfig allows - the architecture to provide a custom asm/gpio.h implementation - overriding the default implementations. New uses of this are - strongly discouraged. - menuconfig GPIOLIB bool "GPIO Support" help
include/linux/gpio.h +0 -7
@@ -54,11 +54,6 @@ struct gpio { };#ifdef CONFIG_GPIOLIB - -#ifdef CONFIG_ARCH_HAVE_CUSTOM_GPIO_H -#include <asm/gpio.h> -#else -#include <asm-generic/gpio.h> static inline int gpio_get_value (unsigned int gpio)
@@ -81,8 +76,6 @@ static inline int gpio_to_irq(unsigned int gpio)return __gpio_to_irq (gpio); } -#endif /* ! CONFIG_ARCH_HAVE_CUSTOM_GPIO_H */ -/* CONFIG_GPIOLIB: bindings for managed devices that want to request gpios */ struct device ;
Signature · 1 line
2.29.2
Re: [RFC 3/3] gpiolib: coldfire: remove custom asm/gpio.h
Hi Arnd,
Quoted 10 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
Now that coldfire is the only user of a custom
asm/gpio.h, it seems better to remove this as well, and have the same interface everywhere.For the gpio_get_value()/gpio_set_value()/gpio_to_irq(),
gpio_cansleep()functions, the custom version is only a micro-optimization to inline the function for constant GPIO numbers. However, in the coldfire defconfigs, I was unable to find a single instance where this micro-optimization was even used, so to my best knowledge removing this has no downsides.
The only user seems to be QSPI chip select handling (not bit-banged data transfer) in arch/m68k/coldfire/device.c, but that indeed depends on CONFIG_SPI_COLDFIRE_QSPI, which is not set in any of the defconfigs. That doesn't mean there were/are no real users, though ;-)
Quoted 4 lines
The custom
gpio_request_one()function is even less useful, as it is guarded by an #ifdef that is never true.Signed-off-by Arnd Bergmann <[email protected]>
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]
In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
Greg Ungerer
Reviewed-byRe: [RFC 3/3] gpiolib: coldfire: remove custom asm/gpio.h
Hi arnd, Geert,
Quoted 14 lines · Geert Uytterhoeven
Hi Arnd,
Quoted 10 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
Now that coldfire is the only user of a custom
asm/gpio.h, it seems better to remove this as well, and have the same interface everywhere.For the gpio_get_value()/gpio_set_value()/gpio_to_irq(),
gpio_cansleep()functions, the custom version is only a micro-optimization to inline the function for constant GPIO numbers. However, in the coldfire defconfigs, I was unable to find a single instance where this micro-optimization was even used, so to my best knowledge removing this has no downsides.
Quoted 4 lines
The only user seems to be QSPI chip select handling (not bit-banged data transfer) in
arch/m68k/coldfire/device.c, but that indeed depends onCONFIG_SPI_COLDFIRE_QSPI, which is not set in any of the defconfigs. That doesn't mean there were/are no real users, though ;-)
That is definitely used by some.
But the generalization and removal of the special casing seems like a win to me.
Quoted 6 lines
Quoted 4 lines
The custom
gpio_request_one()function is even less useful, as it is guarded by an #ifdef that is never true.Signed-off-by Arnd Bergmann <[email protected]>Reviewed-by Geert Uytterhoeven <[email protected]>
Regards
Greg
Quoted 11 lines
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
Re: [RFC 3/3] gpiolib: coldfire: remove custom asm/gpio.h
In reply to Arnd BergmannQuoted 11 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
Now that coldfire is the only user of a custom
asm/gpio.h, it seems better to remove this as well, and have the same interface everywhere.For the gpio_get_value()/gpio_set_value()/gpio_to_irq(),
gpio_cansleep()functions, the custom version is only a micro-optimization to inline the function for constant GPIO numbers. However, in the coldfire defconfigs, I was unable to find a single instance where this micro-optimization was even used, so to my best knowledge removing this has no downsides.
Quoted 4 lines
The custom
gpio_request_one()function is even less useful, as it is guarded by an #ifdef that is never true.Signed-off-by Arnd Bergmann <[email protected]>
Yours,
Linus Walleij
Quoted 8 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
All implementations other than coldfire have returned an error since the avr32 and blackfin architectures got removed, and the last user in driver code was removed in 2016, so just remove this old interface.
The only reference is now in the Chinese documentation, which should be changed to remove this reference as well.
FWIW,
Thanks, Arnd, for cleaning this up!
Quoted 124 lines
Cc Fu Wei <[email protected]>Cc Alex Shi <[email protected]>Cc Hu Haowen <[email protected]>Signed-off-by Arnd Bergmann <[email protected]>--- Documentation/driver-api/gpio/legacy.rst | 20 +++++--------------- arch/m68k/include/asm/gpio.h | 7 ------- arch/sh/include/asm/gpio.h | 5 ----- include/linux/gpio.h | 12 ------------ 4 files changed, 5 insertions(+), 39 deletions(-)Documentation/driver-api/gpio/legacy.rst +5
-15@@ -382,22 +382,18 @@ GPIOs mapped to IRQs -------------------- GPIO numbers are unsigned integers; so are IRQ numbers. These make up two logically distinct namespaces (GPIO 0 need not use IRQ 0). You can -map between them using calls like:: +map between them using:: /* map GPIO numbers to IRQ numbers */ int gpio_to_irq(unsigned gpio); - /* map IRQ numbers to GPIO numbers (avoid using this) */ - int irq_to_gpio(unsigned irq); - -Those return either the corresponding number in the other namespace, or +This returns an irq number corresponding to the gpio number, or else a negative errno code if the mapping can't be done. (For example, some GPIOs can't be used as IRQs.) It is an unchecked error to use a GPIO -number that wasn't set up as an input using gpio_direction_input(), or -to use an IRQ number that didn't originally come from gpio_to_irq(). +number that wasn't set up as an input using gpio_direction_input(). -These two mapping calls are expected to cost on the order of a single -addition or subtraction. They're not allowed to sleep. +The mapping call is expected to cost on the order of a single +addition or subtraction. It is not allowed to sleep. Non-error values returned from gpio_to_irq() can be passed to request_irq() or free_irq(). They will often be stored into IRQ resources for platform@@ -405,12 +401,6 @@ devices, by the board-specific initialization code. Note that IRQ trigger options are part of the IRQ interface, e.g. IRQF_TRIGGER_FALLING, as are system wakeup capabilities. -Non-error values returned from irq_to_gpio() would most commonly be used -with gpio_get_value(), for example to initialize or update driver state -when the IRQ is edge-triggered. Note that some platforms don't support -this reverse mapping, so you should avoid using it. - - Emulating Open Drain Signals ---------------------------- Sometimes shared signals need to use "open drain" signaling, where only thearch/m68k/include/asm/gpio.h +0
-7@@ -66,13 +66,6 @@ static inline int gpio_to_irq(unsigned gpio)return __gpio_to_irq (gpio); } -static inline int irq_to_gpio (unsigned irq) -{ -return (irq >=MCFGPIO_IRQ_VECBASE && - irq < (MCFGPIO_IRQ_VECBASE +MCFGPIO_IRQ_MAX )) ? - irq -MCFGPIO_IRQ_VECBASE : -ENXIO ; -} -static inline int gpio_cansleep (unsigned gpio) {return gpio <MCFGPIO_PIN_MAX ?0 :__gpio_cansleep (gpio);arch/sh/include/asm/gpio.h +0
-5@@ -40,11 +40,6 @@ static inline int gpio_to_irq(unsigned gpio)return __gpio_to_irq (gpio); } -static inline int irq_to_gpio (unsigned int irq) -{ -return -ENOSYS ; -} -#endif /* CONFIG_GPIOLIB */ #endif /* __ASM_SH_GPIO_H */ include/linux/gpio.h +0
-12@@ -81,11 +81,6 @@ static inline int gpio_to_irq(unsigned int gpio) return __gpio_to_irq(gpio); } -static inline int irq_to_gpio (unsigned int irq) -{ -return -EINVAL ; -} -#endif /* ! CONFIG_ARCH_HAVE_CUSTOM_GPIO_H */ /* CONFIG_GPIOLIB: bindings for managed devices that want to request gpios */ @@ -219,13 +214,6 @@ static inline int gpio_to_irq(unsigned gpio)return -EINVAL ; } -static inline int irq_to_gpio (unsigned irq) -{ -/* irq can never have been returned from gpio_to_irq() */ -WARN_ON (1 ); -return -EINVAL ; -} -static inline int devm_gpio_request (struct device *dev,unsigned gpio,const char *label) {Signature · 1 line
2.29.2
Signature · 2 lines
With Best Regards, Andy Shevchenko
Quoted 7 lines · Andy Shevchenko
Quoted 6 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
All implementations other than coldfire have returned an error since the avr32 and blackfin architectures got removed, and the last user in driver code was removed in 2016, so just remove this old interface.
Quoted 5 lines
Quoted 2 lines
The only reference is now in the Chinese documentation, which should be changed to remove this reference as well.
FWIW,
Reviewed-by Andy Shevchenko <[email protected]>
This is for the entire series.
Quoted 1 line
Thanks, Arnd, for cleaning this up!
Signature · 2 lines
With Best Regards, Andy Shevchenko
Quoted 6 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
All implementations other than coldfire have returned an error since the avr32 and blackfin architectures got removed, and the last user in driver code was removed in 2016, so just remove this old interface.
Quoted 8 lines
The only reference is now in the Chinese documentation, which should be changed to remove this reference as well.
Cc Fu Wei <[email protected]>Cc Alex Shi <[email protected]>Cc Hu Haowen <[email protected]>Signed-off-by Arnd Bergmann <[email protected]>
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]
In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
Quoted 6 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
All implementations other than coldfire have returned an error since the avr32 and blackfin architectures got removed, and the last user in driver code was removed in 2016, so just remove this old interface.
Quoted 8 lines
The only reference is now in the Chinese documentation, which should be changed to remove this reference as well.
Cc Fu Wei <[email protected]>Cc Alex Shi <[email protected]>Cc Hu Haowen <[email protected]>Signed-off-by Arnd Bergmann <[email protected]>
Thanks for doing this Arnd. I had it on my mind some years ago but I think Coldfire or something was still using it and didn't think about it as the last user disappeared.
Yours,
Linus Walleij