On Wed, Jul 28, 2010 at 08:53:37AM +0200, Sascha Hauer wrote: > On Tue, Jul 27, 2010 at 03:06:08PM +0300, Baruch Siach wrote: > > Signed-off-by: Baruch Siach <baruch@xxxxxxxxxx> > > --- > > drivers/media/video/mx2_camera.c | 8 +++++--- > > 1 files changed, 5 insertions(+), 3 deletions(-) > > > > diff --git a/drivers/media/video/mx2_camera.c b/drivers/media/video/mx2_camera.c > > index 1536bd4..b42ad8d 100644 > > --- a/drivers/media/video/mx2_camera.c > > +++ b/drivers/media/video/mx2_camera.c > > @@ -420,15 +420,17 @@ static irqreturn_t mx25_camera_irq(int irq_csi, void *data) > > struct mx2_camera_dev *pcdev = data; > > u32 status = readl(pcdev->base_csi + CSISR); > > > > - if (status & CSISR_DMA_TSF_FB1_INT) > > + writel(status, pcdev->base_csi + CSISR); > > + > > + if (!(status & (CSISR_DMA_TSF_FB1_INT | CSISR_DMA_TSF_FB2_INT))) > > + return IRQ_NONE; > > I'm not sure this is correct. When we get here, the interrupt definitely > is from the camera, it's not a shared interrupt. So this only provokes a > 'nobody cared' message from the kernel (if it's still present, I don't > know). You'll only get the 'nobody cared' message if it's happened many times in a short space of time. The odd spurious IRQ_NONE has little effect. It is good practice to return IRQ_NONE if there's nothing pending - it allows stuck IRQs to be detected and disabled without taking the system down. In other words, it should make the system more robust. -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html