[PATCH] intel_idle: Fix false positive RCU splats due to incorrect hardirqs state

[PATCH] intel_idle: Fix false positive RCU splats due to incorrect hardirqs state

[Waiman Long]

Commit 32d4fd5751ea ("cpuidle,intel_idle: Fix CPUIDLE_FLAG_IRQ_ENABLE")
uses raw_local_irq_enable/local_irq_disable() around call to
__intel_idle() in intel_idle_irq().

With interrupt enabled, timer tick interrupt can happen and a
subsequently call to __do_softirq() may change the lockdep hardirqs state
of a debug kernel back to 'on'. This will result in a mismatch between
the cpu hardirqs state (off) and the lockdep hardirqs state (on) causing
a number of false positive "WARNING: suspicious RCU usage" splats.

Fix that by using local_irq_disable() to disable interrupt in
intel_idle_irq().

Fixes: 32d4fd5751ea ("cpuidle,intel_idle: Fix CPUIDLE_FLAG_IRQ_ENABLE")
Signed-off-by: Waiman Long <longman@redhat.com>
---
 drivers/idle/intel_idle.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c
index f5c6802aa6c3..907700d1e78e 100644
--- a/drivers/idle/intel_idle.c
+++ b/drivers/idle/intel_idle.c
@@ -162,7 +162,13 @@ static __cpuidle int intel_idle_irq(struct cpuidle_device *dev,
 
 	raw_local_irq_enable();
 	ret = __intel_idle(dev, drv, index);
-	raw_local_irq_disable();
+
+	/*
+	 * The lockdep hardirqs state may be changed to 'on' with timer
+	 * tick interrupt followed by __do_softirq(). Use local_irq_disable()
+	 * to keep the hardirqs state correct.
+	 */
+	local_irq_disable();
 
 	return ret;
 }
-- 
2.31.1

Re: [PATCH] intel_idle: Fix false positive RCU splats due to incorrect hardirqs state

[Rafael J. Wysocki]

On Sat, Jul 23, 2022 at 10:00 PM Waiman Long <longman@redhat.com> wrote:
>
> Commit 32d4fd5751ea ("cpuidle,intel_idle: Fix CPUIDLE_FLAG_IRQ_ENABLE")
> uses raw_local_irq_enable/local_irq_disable() around call to
> __intel_idle() in intel_idle_irq().
>
> With interrupt enabled, timer tick interrupt can happen and a
> subsequently call to __do_softirq() may change the lockdep hardirqs state
> of a debug kernel back to 'on'. This will result in a mismatch between
> the cpu hardirqs state (off) and the lockdep hardirqs state (on) causing
> a number of false positive "WARNING: suspicious RCU usage" splats.
>
> Fix that by using local_irq_disable() to disable interrupt in
> intel_idle_irq().
>
> Fixes: 32d4fd5751ea ("cpuidle,intel_idle: Fix CPUIDLE_FLAG_IRQ_ENABLE")
> Signed-off-by: Waiman Long <longman@redhat.com>

Applied as 5.19-rc material.

Peter, please let me know if you have concerns.

> ---
>  drivers/idle/intel_idle.c | 8 +++++++-
>  1 file changed, 7 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c
> index f5c6802aa6c3..907700d1e78e 100644
> --- a/drivers/idle/intel_idle.c
> +++ b/drivers/idle/intel_idle.c
> @@ -162,7 +162,13 @@ static __cpuidle int intel_idle_irq(struct cpuidle_device *dev,
>
>         raw_local_irq_enable();
>         ret = __intel_idle(dev, drv, index);
> -       raw_local_irq_disable();
> +
> +       /*
> +        * The lockdep hardirqs state may be changed to 'on' with timer
> +        * tick interrupt followed by __do_softirq(). Use local_irq_disable()
> +        * to keep the hardirqs state correct.
> +        */
> +       local_irq_disable();
>
>         return ret;
>  }
> --
> 2.31.1
>