mfd: mc13xxx-core: Initialize the lock prior to using it

From: Fabio Estevam <[email protected]>

After request_threaded_irq() is called there is a chance that an interrupt may occur before the 'mc13xxx->lock' is initialized, which will trigger a kernel warning.

In order to prevent that, move the initialization of 'mc13xxx->lock' prior to requesting the interrupts.

---
 drivers/mfd/mc13xxx-core.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

index 06e64b6..9e296c4 100644
@@ -660,13 +660,13 @@ int mc13xxx_common_init(struct device *dev)
 	if (ret)
 		return ret;
 
+	mutex_init(&mc13xxx->lock);
+
 	ret = request_threaded_irq(mc13xxx->irq, NULL, mc13xxx_irq_thread,
 			IRQF_ONESHOT | IRQF_TRIGGER_HIGH, "mc13xxx", mc13xxx);
 	if (ret)
 		return ret;
 
-	mutex_init(&mc13xxx->lock);
-
 	if (mc13xxx_probe_flags_dt(mc13xxx) < 0 && pdata)
 		mc13xxx->flags = pdata->flags;
 

1.8.3.2

I'd suggest... "which will trigger a kernel oops" because that's what you got - it killed the IRQ thread, which then caused a few other problems along the way.

Patch looks good, and should fix the reported issue... so with the above description fixed,

Thanks.

FTTC broadband for 0.8mile line: now at 9.7Mbps down 460kbps up... slowly
improving, and getting towards what was expected from it.