On Sat, 27 May 2017, Jeffy Chen wrote:
> If a irq is already disabled & masked, free_irq may cause a unbalanced
> irq shutdown/disable/mask, for example:
No, it's not. irq_shutdown/disable/mask are low level access functions
which can be invoked at any given time.
The only interface which has refcounting is disable/enable_irq().
> devm_request_irq->irq_startup->irq_enable
> disable_irq <-- disabled and masked
> devm_free_irq->irq_shutdown <-- try to disable it again
>
> This would confuse some pinctrl drivers which would control clk in
> irq_enable/irq_disable, for example pinctrl-rockchip/pinctrl-nomadik.
>
> This patch add a state check in irq_shutdown to prevent that.
Please read Documentation/process/SubmittingPatches and search for "this
patch".
> diff --git a/kernel/irq/chip.c b/kernel/irq/chip.c
> index 686be4b..816da03 100644
> --- a/kernel/irq/chip.c
> +++ b/kernel/irq/chip.c
> @@ -206,14 +206,20 @@ int irq_startup(struct irq_desc *desc, bool resend)
>
> void irq_shutdown(struct irq_desc *desc)
> {
> - irq_state_set_disabled(desc);
> desc->depth = 1;
> +
> + if (unlikely(irqd_irq_disabled(&desc->irq_data) &&
> + irqd_irq_masked(&desc->irq_data)))
> + goto out;
This is just wrong.
It's perfectly legit to call disable_irq() and then free_irq(). Still
free_irq() has to be able to invoke chip->irq_shutdown(). Preventing that
will leave some interrupt chips in a half initialized state.
The irq core does not guarntee that the unmask/mask enable/disable
startup/shutdown callbacks are perfectly balanced. irq_shutdown() is only
one place where this can happen. This needs more thought than this 'works
for me' hackery.
Thanks,
tglx
