| Message ID | 20211122220423.11256-1-hdegoede@redhat.com (mailing list archive) |
|---|---|
| State | Not Applicable, archived |
| Headers | show |
| Series | pinctrl: baytrail: Set IRQCHIP_SET_TYPE_MASKED flag on the irqchip | expand |
On Mon, Nov 22, 2021 at 11:04:23PM +0100, Hans de Goede wrote: > The byt_irq_type function ends with the IRQ masked, this means that calls > to irq_set_irq_type() while the IRQ is enabled end up masking it, which > is wrong. Add the IRQCHIP_SET_TYPE_MASKED flag to fix this. > > This will make the IRQ core call mask() + unmask() on the IRQ around > a set_type() call when the IRQ is enabled at the type of the call. > > Note in practice irq_set_irq_type() getting called while the IRQ is enabled > almost never happens. I hit this with a buggy DSDT where a wrongly active > (_STA returns 0xf) I2C ACPI devices point to an IRQ already in use by an > _AEI handler, leading to the irq_set_irq_type() call in > acpi_dev_gpio_irq_get_by() getting called while the IRQ is enabled. > > Signed-off-by: Hans de Goede <hdegoede@redhat.com> Acked-by: Mika Westerberg <mika.westerberg@linux.intel.com>
On Tue, Nov 23, 2021 at 01:08:30PM +0200, Mika Westerberg wrote: > On Mon, Nov 22, 2021 at 11:04:23PM +0100, Hans de Goede wrote: > > The byt_irq_type function ends with the IRQ masked, this means that calls > > to irq_set_irq_type() while the IRQ is enabled end up masking it, which > > is wrong. Add the IRQCHIP_SET_TYPE_MASKED flag to fix this. > > > > This will make the IRQ core call mask() + unmask() on the IRQ around > > a set_type() call when the IRQ is enabled at the type of the call. > > > > Note in practice irq_set_irq_type() getting called while the IRQ is enabled > > almost never happens. I hit this with a buggy DSDT where a wrongly active > > (_STA returns 0xf) I2C ACPI devices point to an IRQ already in use by an > > _AEI handler, leading to the irq_set_irq_type() call in > > acpi_dev_gpio_irq_get_by() getting called while the IRQ is enabled. > > > > Signed-off-by: Hans de Goede <hdegoede@redhat.com> > > Acked-by: Mika Westerberg <mika.westerberg@linux.intel.com> Pushed to my review and testing queue, thanks!
diff --git a/drivers/pinctrl/intel/pinctrl-baytrail.c b/drivers/pinctrl/intel/pinctrl-baytrail.c index 744ebe417bff..e3dd105e2f6e 100644 --- a/drivers/pinctrl/intel/pinctrl-baytrail.c +++ b/drivers/pinctrl/intel/pinctrl-baytrail.c @@ -1554,7 +1554,8 @@ static int byt_gpio_probe(struct intel_pinctrl *vg) vg->irqchip.irq_mask = byt_irq_mask, vg->irqchip.irq_unmask = byt_irq_unmask, vg->irqchip.irq_set_type = byt_irq_type, - vg->irqchip.flags = IRQCHIP_SKIP_SET_WAKE, + vg->irqchip.flags = IRQCHIP_SKIP_SET_WAKE | + IRQCHIP_SET_TYPE_MASKED, girq = &gc->irq; girq->chip = &vg->irqchip;
The byt_irq_type function ends with the IRQ masked, this means that calls to irq_set_irq_type() while the IRQ is enabled end up masking it, which is wrong. Add the IRQCHIP_SET_TYPE_MASKED flag to fix this. This will make the IRQ core call mask() + unmask() on the IRQ around a set_type() call when the IRQ is enabled at the type of the call. Note in practice irq_set_irq_type() getting called while the IRQ is enabled almost never happens. I hit this with a buggy DSDT where a wrongly active (_STA returns 0xf) I2C ACPI devices point to an IRQ already in use by an _AEI handler, leading to the irq_set_irq_type() call in acpi_dev_gpio_irq_get_by() getting called while the IRQ is enabled. Signed-off-by: Hans de Goede <hdegoede@redhat.com> --- drivers/pinctrl/intel/pinctrl-baytrail.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)