Thread (1 message) 1 message, 1 author, 2020-08-11

Re: [PATCH] genirq/PM: Always unlock IRQ descriptor in rearm_wake_irq

From: Rafael J. Wysocki <hidden>
Date: 2020-08-11 19:21:55
Also in: lkml

On 8/11/2020 8:00 PM, Guenter Roeck wrote:
rearm_wake_irq() does not unlock the irq descriptor if the interrupt
is not suspended or if wakeup is not enabled on it. Fix it.

Fixes: 3a79bc63d9075 ("PCI: irq: Introduce rearm_wake_irq()")
Cc: Rafael J. Wysocki <redacted>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Acked-by: Rafael J. Wysocki <redacted>

And it needs to go to -stable (even though the bug is latent now, 
because this function is called for suspended IRQs only AFAICS).

Or I can apply this as the mistake was in my commit.  Please let me know 
what you prefer.

quoted hunk
---
  kernel/irq/pm.c | 8 ++++++--
  1 file changed, 6 insertions(+), 2 deletions(-)
diff --git a/kernel/irq/pm.c b/kernel/irq/pm.c
index 8f557fa1f4fe..c6c7e187ae74 100644
--- a/kernel/irq/pm.c
+++ b/kernel/irq/pm.c
@@ -185,14 +185,18 @@ void rearm_wake_irq(unsigned int irq)
  	unsigned long flags;
  	struct irq_desc *desc = irq_get_desc_buslock(irq, &flags, IRQ_GET_DESC_CHECK_GLOBAL);
  
-	if (!desc || !(desc->istate & IRQS_SUSPENDED) ||
-	    !irqd_is_wakeup_set(&desc->irq_data))
+	if (!desc)
  		return;
  
+	if (!(desc->istate & IRQS_SUSPENDED) ||
+	    !irqd_is_wakeup_set(&desc->irq_data))
+		goto unlock;
+
  	desc->istate &= ~IRQS_SUSPENDED;
  	irqd_set(&desc->irq_data, IRQD_WAKEUP_ARMED);
  	__enable_irq(desc);
  
+unlock:
  	irq_put_desc_busunlock(desc, flags);
  }
  
  
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help