mfd: Fix memleak in pcf50633_client_dev_register

Since platform_device_add_data copies the passed data, the allocated
subdev_pdata is never freed. A simple fix would be to either free subdev_pdata
or put it onto the stack. But since the pcf50633 child devices can rely on
beeing children of the pcf50633 core device it's much more elegant to get access
to pcf50633 core structure through that link. This allows to get completly rid
of pcf5033_subdev_pdata.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Paul Fertser <fercerpav@gmail.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
diff --git a/drivers/mfd/pcf50633-adc.c b/drivers/mfd/pcf50633-adc.c
index 3d31e97..6d2e846 100644
--- a/drivers/mfd/pcf50633-adc.c
+++ b/drivers/mfd/pcf50633-adc.c
@@ -209,17 +209,16 @@
 
 static int __devinit pcf50633_adc_probe(struct platform_device *pdev)
 {
-	struct pcf50633_subdev_pdata *pdata = pdev->dev.platform_data;
 	struct pcf50633_adc *adc;
 
 	adc = kzalloc(sizeof(*adc), GFP_KERNEL);
 	if (!adc)
 		return -ENOMEM;
 
-	adc->pcf = pdata->pcf;
+	adc->pcf = dev_to_pcf50633(pdev->dev.parent);
 	platform_set_drvdata(pdev, adc);
 
-	pcf50633_register_irq(pdata->pcf, PCF50633_IRQ_ADCRDY,
+	pcf50633_register_irq(adc->pcf, PCF50633_IRQ_ADCRDY,
 					pcf50633_adc_irq, adc);
 
 	mutex_init(&adc->queue_mutex);