Re: [PATCH RFT 1/2] gpio: gpio-ml-ioh: Use spinlock for register access protection
On Sun, Jul 29, 2012 at 4:54 AM, Axel Lin wrote: > gpio_chip.can_sleep is 0, but current code uses mutex in ioh_gpio_set, > ioh_gpio_get and ioh_gpio_direction_input functions. > Thus those functions are not callable from interrupt context. > This patch converts mutex into spinlock. > > Signed-off-by: Axel Lin Applied to my devel branch for the next merge window, it doesn't seem like a regression to me atleast. Yours, Linus Walleij -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH RFT 1/2] gpio: gpio-ml-ioh: Use spinlock for register access protection
On Sun, Jul 29, 2012 at 4:54 AM, Axel Lin axel@gmail.com wrote: gpio_chip.can_sleep is 0, but current code uses mutex in ioh_gpio_set, ioh_gpio_get and ioh_gpio_direction_input functions. Thus those functions are not callable from interrupt context. This patch converts mutex into spinlock. Signed-off-by: Axel Lin axel@gmail.com Applied to my devel branch for the next merge window, it doesn't seem like a regression to me atleast. Yours, Linus Walleij -- To unsubscribe from this list: send the line unsubscribe linux-kernel in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH RFT 1/2] gpio: gpio-ml-ioh: Use spinlock for register access protection
On Mon, 13 Aug 2012 14:01:31 +0200 Linus Walleij wrote: > On Sun, Jul 29, 2012 at 4:54 AM, Axel Lin wrote: > > > gpio_chip.can_sleep is 0, but current code uses mutex in ioh_gpio_set, > > ioh_gpio_get and ioh_gpio_direction_input functions. > > Thus those functions are not callable from interrupt context. > > This patch converts mutex into spinlock. > > > > Signed-off-by: Axel Lin > > So IOH guys what are you saying about this? I'm not the IOH guy, but just happened to work on it for some time :) Anyway, the 2 patches both look good to me. Thanks, Feng -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH RFT 1/2] gpio: gpio-ml-ioh: Use spinlock for register access protection
On Sun, Jul 29, 2012 at 4:54 AM, Axel Lin wrote: > gpio_chip.can_sleep is 0, but current code uses mutex in ioh_gpio_set, > ioh_gpio_get and ioh_gpio_direction_input functions. > Thus those functions are not callable from interrupt context. > This patch converts mutex into spinlock. > > Signed-off-by: Axel Lin So IOH guys what are you saying about this? If you say nothing I will apply these patches. Yours, Linus Walleij -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH RFT 1/2] gpio: gpio-ml-ioh: Use spinlock for register access protection
On Sun, Jul 29, 2012 at 4:54 AM, Axel Lin axel@gmail.com wrote: gpio_chip.can_sleep is 0, but current code uses mutex in ioh_gpio_set, ioh_gpio_get and ioh_gpio_direction_input functions. Thus those functions are not callable from interrupt context. This patch converts mutex into spinlock. Signed-off-by: Axel Lin axel@gmail.com So IOH guys what are you saying about this? If you say nothing I will apply these patches. Yours, Linus Walleij -- To unsubscribe from this list: send the line unsubscribe linux-kernel in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH RFT 1/2] gpio: gpio-ml-ioh: Use spinlock for register access protection
On Mon, 13 Aug 2012 14:01:31 +0200 Linus Walleij linus.wall...@linaro.org wrote: On Sun, Jul 29, 2012 at 4:54 AM, Axel Lin axel@gmail.com wrote: gpio_chip.can_sleep is 0, but current code uses mutex in ioh_gpio_set, ioh_gpio_get and ioh_gpio_direction_input functions. Thus those functions are not callable from interrupt context. This patch converts mutex into spinlock. Signed-off-by: Axel Lin axel@gmail.com So IOH guys what are you saying about this? I'm not the IOH guy, but just happened to work on it for some time :) Anyway, the 2 patches both look good to me. Thanks, Feng -- To unsubscribe from this list: send the line unsubscribe linux-kernel in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
[PATCH RFT 1/2] gpio: gpio-ml-ioh: Use spinlock for register access protection
gpio_chip.can_sleep is 0, but current code uses mutex in ioh_gpio_set, ioh_gpio_get and ioh_gpio_direction_input functions. Thus those functions are not callable from interrupt context. This patch converts mutex into spinlock. Signed-off-by: Axel Lin --- drivers/gpio/gpio-ml-ioh.c | 20 ++-- 1 file changed, 10 insertions(+), 10 deletions(-) diff --git a/drivers/gpio/gpio-ml-ioh.c b/drivers/gpio/gpio-ml-ioh.c index db01f15..6a29ee1 100644 --- a/drivers/gpio/gpio-ml-ioh.c +++ b/drivers/gpio/gpio-ml-ioh.c @@ -87,8 +87,7 @@ struct ioh_gpio_reg_data { * @gpio_use_sel: Save GPIO_USE_SEL1~4 register for PM * @ch:Indicate GPIO channel * @irq_base: Save base of IRQ number for interrupt - * @spinlock: Used for register access protection in - * interrupt context ioh_irq_type and PM; + * @spinlock: Used for register access protection */ struct ioh_gpio { void __iomem *base; @@ -97,7 +96,6 @@ struct ioh_gpio { struct gpio_chip gpio; struct ioh_gpio_reg_data ioh_gpio_reg; u32 gpio_use_sel; - struct mutex lock; int ch; int irq_base; spinlock_t spinlock; @@ -109,8 +107,9 @@ static void ioh_gpio_set(struct gpio_chip *gpio, unsigned nr, int val) { u32 reg_val; struct ioh_gpio *chip = container_of(gpio, struct ioh_gpio, gpio); + unsigned long flags; - mutex_lock(>lock); + spin_lock_irqsave(>spinlock, flags); reg_val = ioread32(>reg->regs[chip->ch].po); if (val) reg_val |= (1 << nr); @@ -118,7 +117,7 @@ static void ioh_gpio_set(struct gpio_chip *gpio, unsigned nr, int val) reg_val &= ~(1 << nr); iowrite32(reg_val, >reg->regs[chip->ch].po); - mutex_unlock(>lock); + spin_unlock_irqrestore(>spinlock, flags); } static int ioh_gpio_get(struct gpio_chip *gpio, unsigned nr) @@ -134,8 +133,9 @@ static int ioh_gpio_direction_output(struct gpio_chip *gpio, unsigned nr, struct ioh_gpio *chip = container_of(gpio, struct ioh_gpio, gpio); u32 pm; u32 reg_val; + unsigned long flags; - mutex_lock(>lock); + spin_lock_irqsave(>spinlock, flags); pm = ioread32(>reg->regs[chip->ch].pm) & ((1 << num_ports[chip->ch]) - 1); pm |= (1 << nr); @@ -148,7 +148,7 @@ static int ioh_gpio_direction_output(struct gpio_chip *gpio, unsigned nr, reg_val &= ~(1 << nr); iowrite32(reg_val, >reg->regs[chip->ch].po); - mutex_unlock(>lock); + spin_unlock_irqrestore(>spinlock, flags); return 0; } @@ -157,13 +157,14 @@ static int ioh_gpio_direction_input(struct gpio_chip *gpio, unsigned nr) { struct ioh_gpio *chip = container_of(gpio, struct ioh_gpio, gpio); u32 pm; + unsigned long flags; - mutex_lock(>lock); + spin_lock_irqsave(>spinlock, flags); pm = ioread32(>reg->regs[chip->ch].pm) & ((1 << num_ports[chip->ch]) - 1); pm &= ~(1 << nr); iowrite32(pm, >reg->regs[chip->ch].pm); - mutex_unlock(>lock); + spin_unlock_irqrestore(>spinlock, flags); return 0; } @@ -447,7 +448,6 @@ static int __devinit ioh_gpio_probe(struct pci_dev *pdev, chip->base = base; chip->reg = chip->base; chip->ch = i; - mutex_init(>lock); spin_lock_init(>spinlock); ioh_gpio_setup(chip, num_ports[i]); ret = gpiochip_add(>gpio); -- 1.7.9.5 -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
[PATCH RFT 1/2] gpio: gpio-ml-ioh: Use spinlock for register access protection
gpio_chip.can_sleep is 0, but current code uses mutex in ioh_gpio_set, ioh_gpio_get and ioh_gpio_direction_input functions. Thus those functions are not callable from interrupt context. This patch converts mutex into spinlock. Signed-off-by: Axel Lin axel@gmail.com --- drivers/gpio/gpio-ml-ioh.c | 20 ++-- 1 file changed, 10 insertions(+), 10 deletions(-) diff --git a/drivers/gpio/gpio-ml-ioh.c b/drivers/gpio/gpio-ml-ioh.c index db01f15..6a29ee1 100644 --- a/drivers/gpio/gpio-ml-ioh.c +++ b/drivers/gpio/gpio-ml-ioh.c @@ -87,8 +87,7 @@ struct ioh_gpio_reg_data { * @gpio_use_sel: Save GPIO_USE_SEL1~4 register for PM * @ch:Indicate GPIO channel * @irq_base: Save base of IRQ number for interrupt - * @spinlock: Used for register access protection in - * interrupt context ioh_irq_type and PM; + * @spinlock: Used for register access protection */ struct ioh_gpio { void __iomem *base; @@ -97,7 +96,6 @@ struct ioh_gpio { struct gpio_chip gpio; struct ioh_gpio_reg_data ioh_gpio_reg; u32 gpio_use_sel; - struct mutex lock; int ch; int irq_base; spinlock_t spinlock; @@ -109,8 +107,9 @@ static void ioh_gpio_set(struct gpio_chip *gpio, unsigned nr, int val) { u32 reg_val; struct ioh_gpio *chip = container_of(gpio, struct ioh_gpio, gpio); + unsigned long flags; - mutex_lock(chip-lock); + spin_lock_irqsave(chip-spinlock, flags); reg_val = ioread32(chip-reg-regs[chip-ch].po); if (val) reg_val |= (1 nr); @@ -118,7 +117,7 @@ static void ioh_gpio_set(struct gpio_chip *gpio, unsigned nr, int val) reg_val = ~(1 nr); iowrite32(reg_val, chip-reg-regs[chip-ch].po); - mutex_unlock(chip-lock); + spin_unlock_irqrestore(chip-spinlock, flags); } static int ioh_gpio_get(struct gpio_chip *gpio, unsigned nr) @@ -134,8 +133,9 @@ static int ioh_gpio_direction_output(struct gpio_chip *gpio, unsigned nr, struct ioh_gpio *chip = container_of(gpio, struct ioh_gpio, gpio); u32 pm; u32 reg_val; + unsigned long flags; - mutex_lock(chip-lock); + spin_lock_irqsave(chip-spinlock, flags); pm = ioread32(chip-reg-regs[chip-ch].pm) ((1 num_ports[chip-ch]) - 1); pm |= (1 nr); @@ -148,7 +148,7 @@ static int ioh_gpio_direction_output(struct gpio_chip *gpio, unsigned nr, reg_val = ~(1 nr); iowrite32(reg_val, chip-reg-regs[chip-ch].po); - mutex_unlock(chip-lock); + spin_unlock_irqrestore(chip-spinlock, flags); return 0; } @@ -157,13 +157,14 @@ static int ioh_gpio_direction_input(struct gpio_chip *gpio, unsigned nr) { struct ioh_gpio *chip = container_of(gpio, struct ioh_gpio, gpio); u32 pm; + unsigned long flags; - mutex_lock(chip-lock); + spin_lock_irqsave(chip-spinlock, flags); pm = ioread32(chip-reg-regs[chip-ch].pm) ((1 num_ports[chip-ch]) - 1); pm = ~(1 nr); iowrite32(pm, chip-reg-regs[chip-ch].pm); - mutex_unlock(chip-lock); + spin_unlock_irqrestore(chip-spinlock, flags); return 0; } @@ -447,7 +448,6 @@ static int __devinit ioh_gpio_probe(struct pci_dev *pdev, chip-base = base; chip-reg = chip-base; chip-ch = i; - mutex_init(chip-lock); spin_lock_init(chip-spinlock); ioh_gpio_setup(chip, num_ports[i]); ret = gpiochip_add(chip-gpio); -- 1.7.9.5 -- To unsubscribe from this list: send the line unsubscribe linux-kernel in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/