| Message ID | 1468231586-104167-1-git-send-email-andriy.shevchenko@linux.intel.com |
|---|---|
| State | New |
| Headers | show |
On Mon, Jul 11, 2016 at 12:06 PM, Andy Shevchenko <andriy.shevchenko@linux.intel.com> wrote: > There is a potential race when two threads do the writes to the same register > in parallel. > > Prevent out of order in such case by protecting I/O access by spin lock. > > Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Patch applied. May I convince you to send a patch adding yourself as maintainer for this driver in the MAINTAINERS file? Yours, Linus Walleij -- To unsubscribe from this list: send the line "unsubscribe linux-gpio" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, 2016-07-11 at 14:22 +0200, Linus Walleij wrote: > On Mon, Jul 11, 2016 at 12:06 PM, Andy Shevchenko > <andriy.shevchenko@linux.intel.com> wrote: > > > There is a potential race when two threads do the writes to the same > > register > > in parallel. > > > > Prevent out of order in such case by protecting I/O access by spin > > lock. > > > > Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> > > Patch applied. > > May I convince you to send a patch adding yourself as maintainer > for this driver in the MAINTAINERS file? NP. And pinctrl perhaps?
diff --git a/drivers/gpio/gpio-merrifield.c b/drivers/gpio/gpio-merrifield.c index 11066f6..45b5127 100644 --- a/drivers/gpio/gpio-merrifield.c +++ b/drivers/gpio/gpio-merrifield.c @@ -105,7 +105,11 @@ static int mrfld_gpio_get(struct gpio_chip *chip, unsigned int offset) static void mrfld_gpio_set(struct gpio_chip *chip, unsigned int offset, int value) { + struct mrfld_gpio *priv = gpiochip_get_data(chip); void __iomem *gpsr, *gpcr; + unsigned long flags; + + raw_spin_lock_irqsave(&priv->lock, flags); if (value) { gpsr = gpio_reg(chip, offset, GPSR); @@ -114,6 +118,8 @@ static void mrfld_gpio_set(struct gpio_chip *chip, unsigned int offset, gpcr = gpio_reg(chip, offset, GPCR); writel(BIT(offset % 32), gpcr); } + + raw_spin_unlock_irqrestore(&priv->lock, flags); } static int mrfld_gpio_direction_input(struct gpio_chip *chip, @@ -160,8 +166,13 @@ static void mrfld_irq_ack(struct irq_data *d) struct mrfld_gpio *priv = irq_data_get_irq_chip_data(d); u32 gpio = irqd_to_hwirq(d); void __iomem *gisr = gpio_reg(&priv->chip, gpio, GISR); + unsigned long flags; + + raw_spin_lock_irqsave(&priv->lock, flags); writel(BIT(gpio % 32), gisr); + + raw_spin_unlock_irqrestore(&priv->lock, flags); } static void mrfld_irq_unmask_mask(struct irq_data *d, bool unmask)
There is a potential race when two threads do the writes to the same register in parallel. Prevent out of order in such case by protecting I/O access by spin lock. Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> --- drivers/gpio/gpio-merrifield.c | 11 +++++++++++ 1 file changed, 11 insertions(+)