From: Alan Borzeszkowski <alan.borzeszkowski@xxxxxxxxxxxxxxx> commit 0bb18e34abdde7bf58fca8542e2dcf621924ea19 upstream. Interrupt status (GPI_IS) register is cleared by writing 1 to it, not 0. Cc: stable@xxxxxxxxxxxxxxx Signed-off-by: Alan Borzeszkowski <alan.borzeszkowski@xxxxxxxxxxxxxxx> Signed-off-by: Mika Westerberg <mika.westerberg@xxxxxxxxxxxxxxx> Acked-by: Andy Shevchenko <andy@xxxxxxxxxx> Link: https://lore.kernel.org/r/20241204070415.1034449-8-mika.westerberg@xxxxxxxxxxxxxxx Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@xxxxxxxxxx> Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> --- drivers/gpio/gpio-graniterapids.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/drivers/gpio/gpio-graniterapids.c +++ b/drivers/gpio/gpio-graniterapids.c @@ -166,7 +166,7 @@ static void gnr_gpio_irq_ack(struct irq_ guard(raw_spinlock_irqsave)(&priv->lock); reg = readl(addr); - reg &= ~BIT(bit_idx); + reg |= BIT(bit_idx); writel(reg, addr); } Patches currently in stable-queue which might be from alan.borzeszkowski@xxxxxxxxxxxxxxx are queue-6.12/gpio-graniterapids-fix-invalid-rxevcfg-register-bitmask.patch queue-6.12/gpio-graniterapids-check-if-gpio-line-can-be-used-for-irqs.patch queue-6.12/gpio-graniterapids-fix-vgpio-driver-crash.patch queue-6.12/gpio-graniterapids-fix-invalid-gpi_is-register-offset.patch queue-6.12/gpio-graniterapids-determine-if-gpio-pad-can-be-used-by-driver.patch queue-6.12/gpio-graniterapids-fix-gpio-ack-functionality.patch queue-6.12/gpio-graniterapids-fix-incorrect-bar-assignment.patch
