Hi Mauro, On Saturday 29 October 2011 00:30:12 Mauro Carvalho Chehab wrote: > Em 28-10-2011 14:42, Laurent Pinchart escreveu: > > On Friday 28 October 2011 03:31:53 Gilles Gigan wrote: > >> Hi, > >> I would like to know what is the correct way to switch the current > >> video input during capture on a card with a single BT878 chip and 4 > >> inputs > >> (http://store.bluecherry.net/products/PV%252d143-%252d-4-port-video-capt > >> ur e-card-%2830FPS%29-%252d-OEM.html). I tried doing it in two ways: - > >> using VIDIOC_S_INPUT to change the current input. While this works, the > >> next captured frame shows video from the old input in its top half and > >> video from the new input in the bottom half. > > This is is likely easy to fix. The driver has already a logic to prevent > changing the buffer while in the middle of a buffer filling. I suspect > that the BKL removal patches might have broken it somewhat, allowing > things like that. basically, it should be as simple as not allowing > changing the input at the top half. This will work optimally only if the input analog signals are synchronized, right ? If we switch to a new input right when the frame start, can the first frame captured on the new input be corrupted ? > Please try the enclosed patch. > > Regards, > Mauro > > - > > bttv: Avoid switching the video input at the top half. > > Signed-off-by: Mauro Carvalho Chehab <mchehab@xxxxxxxxxx> > > diff --git a/drivers/media/video/bt8xx/bttv-driver.c > b/drivers/media/video/bt8xx/bttv-driver.c index 3dd0660..6a3be6f 100644 > --- a/drivers/media/video/bt8xx/bttv-driver.c > +++ b/drivers/media/video/bt8xx/bttv-driver.c > @@ -3978,7 +3978,7 @@ bttv_irq_switch_video(struct bttv *btv) > bttv_set_dma(btv, 0); > > /* switch input */ > - if (UNSET != btv->new_input) { > + if (! btv->curr.top && UNSET != btv->new_input) { > video_mux(btv,btv->new_input); > btv->new_input = UNSET; > } -- Regards, Laurent Pinchart -- 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
