On 2013-10-02 12:17, Ian Abbott wrote:
On 2013-10-01 23:14, H Hartley Sweeten wrote:
Do the request_irq() before setting up the subdevices. Only initialize
the interrupt subdevice if the irq is actually available.
Since it's using IRQ_SHARED, you'll need add a test to the start of
ni6527_interrupt() to check dev->attached before accessing the subdevice:
if (!dev->attached)
return IRQ_NONE;
Actually, the first thing the interrupt does is read the status
register, which should indicate that the device is not asserting a reset
since the hardware has just been initialized by ni6527_reset() to
disable and clear interrupts.
So it should be safe. It may still be worth checking one of
dev->attached or dev->read_subdev for additional safety.
--
-=( Ian Abbott @ MEV Ltd. E-mail: <abbotti@xxxxxxxxx> )=-
-=( Tel: +44 (0)161 477 1898 FAX: +44 (0)161 718 3587 )=-
_______________________________________________
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxx
http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
