On 2013-12-06 16:57, H Hartley Sweeten wrote:
Do the request_irq() before setting up the subdevices. Only hook up the command support of the irq was sucessfully requested. Note that, because of the IRQF_SHARED flag, nidio_interrupt() _may_ be called before the device is ready and the subdevices are setup. This condition is handled by the (!dev->attached) sanity check. The 'dev->read_subdev' pointer might not be valid at this point so there is an extra check to get the pointer to the comedi_async buffer. Signed-off-by: H Hartley Sweeten <hsweeten@xxxxxxxxxxxxxxxxxxx> Cc: Ian Abbott <abbotti@xxxxxxxxx> Cc: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> --- v3: fix a possible dereference of 's' in nido_interrupt() drivers/staging/comedi/drivers/ni_pcidio.c | 40 +++++++++++++++--------------- 1 file changed, 20 insertions(+), 20 deletions(-) diff --git a/drivers/staging/comedi/drivers/ni_pcidio.c b/drivers/staging/comedi/drivers/ni_pcidio.c index 30c46a3..8f1b64d 100644 --- a/drivers/staging/comedi/drivers/ni_pcidio.c +++ b/drivers/staging/comedi/drivers/ni_pcidio.c @@ -385,7 +385,7 @@ static irqreturn_t nidio_interrupt(int irq, void *d) struct comedi_device *dev = d; struct nidio96_private *devpriv = dev->private; struct comedi_subdevice *s = dev->read_subdev; - struct comedi_async *async = s->async; + struct comedi_async *async = s ? s->async : NULL; struct mite_struct *mite = devpriv->mite;
It should work, but wouldn't it be better to leave `async` unassigned until after the `dev->attached` check?
-- -=( 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
