Hi Luca, thanks for the input On Sat, Mar 16, 2019 at 11:23:42AM +0100, Luca Ceresoli wrote: > Hi Jacopo, Sakari, > > On 15/03/19 11:06, Sakari Ailus wrote: > > Hi Luca, Jacopo, > > > > On Fri, Mar 15, 2019 at 10:45:38AM +0100, Jacopo Mondi wrote: > >> Hi Luca, > >> > >> On Thu, Mar 14, 2019 at 03:45:27PM +0100, Luca Ceresoli wrote: > >>> Hi, > >>> > >>> begging your pardon for the noob question below... > >>> > >> > >> Let a noob help another noob then > >> > >>> On 05/03/19 19:51, Jacopo Mondi wrote: > >>>> From: Niklas Söderlund > >>>> > >>>> Add support to get and set the internal routing between the adv748x > >>>> CSI-2 transmitters sink pad and its multiplexed source pad. This routing > >>>> includes which stream of the multiplexed pad to use, allowing the user > >>>> to select which CSI-2 virtual channel to use when transmitting the > >>>> stream. > >>>> > >>>> Signed-off-by: Niklas Söderlund > >>>> Reviewed-by: Jacopo Mondi > >>>> --- > >>>> drivers/media/i2c/adv748x/adv748x-csi2.c | 65 ++++++++++++++++++++++++ > >>>> 1 file changed, 65 insertions(+) > >>>> > >>>> diff --git a/drivers/media/i2c/adv748x/adv748x-csi2.c b/drivers/media/i2c/adv748x/adv748x-csi2.c > >>>> index d8f7cbee86e7..13454af72c6e 100644 > >>>> --- a/drivers/media/i2c/adv748x/adv748x-csi2.c > >>>> +++ b/drivers/media/i2c/adv748x/adv748x-csi2.c > >>>> @@ -14,6 +14,8 @@ > >>>> > >>>> #include "adv748x.h" > >>>> > >>>> +#define ADV748X_CSI2_ROUTES_MAX 4 > >>>> + > >>>> struct adv748x_csi2_format { > >>>> unsigned int code; > >>>> unsigned int datatype; > >>>> @@ -253,10 +255,73 @@ static int adv748x_csi2_get_frame_desc(struct v4l2_subdev *sd, unsigned int pad, > >>>> return 0; > >>>> } > >>>> > >>>> +static int adv748x_csi2_get_routing(struct v4l2_subdev *sd, > >>>> + struct v4l2_subdev_krouting *routing) > >>>> +{ > >>>> + struct adv748x_csi2 *tx = adv748x_sd_to_csi2(sd); > >>>> + struct v4l2_subdev_route *r = routing->routes; > >>>> + unsigned int vc; > >>>> + > >>>> + if (routing->num_routes < ADV748X_CSI2_ROUTES_MAX) { > >>>> + routing->num_routes = ADV748X_CSI2_ROUTES_MAX; > >>>> + return -ENOSPC; > >>>> + } > >>>> + > >>>> + routing->num_routes = ADV748X_CSI2_ROUTES_MAX; > >>>> + > >>>> + for (vc = 0; vc < ADV748X_CSI2_ROUTES_MAX; vc++) { > >>>> + r->sink_pad = ADV748X_CSI2_SINK; > >>>> + r->sink_stream = 0; > >>>> + r->source_pad = ADV748X_CSI2_SOURCE; > >>>> + r->source_stream = vc; > >>>> + r->flags = vc == tx->vc ? V4L2_SUBDEV_ROUTE_FL_ACTIVE : 0; > >>>> + r++; > >>>> + } > >>>> + > >>>> + return 0; > >>>> +} > >>>> + > >>>> +static int adv748x_csi2_set_routing(struct v4l2_subdev *sd, > >>>> + struct v4l2_subdev_krouting *routing) > >>>> +{ > >>>> + struct adv748x_csi2 *tx = adv748x_sd_to_csi2(sd); > >>>> + struct v4l2_subdev_route *r = routing->routes; > >>>> + unsigned int i; > >>>> + int vc = -1; > >>>> + > >>>> + if (routing->num_routes > ADV748X_CSI2_ROUTES_MAX) > >>>> + return -ENOSPC; > >>>> + > >>>> + for (i = 0; i < routing->num_routes; i++) { > >>>> + if (r->sink_pad != ADV748X_CSI2_SINK || > >>>> + r->sink_stream != 0 || > >>>> + r->source_pad != ADV748X_CSI2_SOURCE || > >>>> + r->source_stream >= ADV748X_CSI2_ROUTES_MAX) > >>>> + return -EINVAL; > >>>> + > >>>> + if (r->flags & V4L2_SUBDEV_ROUTE_FL_ACTIVE) { > >>>> + if (vc != -1) > >>>> + return -EMLINK; > >>>> + > >>>> + vc = r->source_stream; > >>>> + } > >>>> + r++; > >>>> + } > >>>> + > >>>> + if (vc != -1) > >>>> + tx->vc = vc; > >>>> + > >>>> + adv748x_csi2_set_virtual_channel(tx, tx->vc); > >>>> + > >>>> + return 0; > >>>> +} > >>> > >>> Not specific to this patch but rather to the set_routing idea as a > >>> whole: can the set_routing ioctl be called while the stream is running? > >>> > >>> If it cannot, I find it a limiting factor for nowadays use cases. I also > >>> didn't find where the ioctl is rejected. > >>> > >> > >> The framework does not make assumptions about that at the moment. > >> > >>> If it can, then shouldn't this function call s_stream(stop) through the > >>> sink pad whose route becomes disabled, and a s_stream(start) through the > >>> one that gets enabled? > >>> > >> > >> If I got this right, you're here rightfully pointing out that changing > >> the routing between pads in an entity migh impact the pipeline as a > >> whole, and this would require, to activate/deactivate devices that > >> where not part of the pipeline. > > > > I'd say that ultimately this depends on the devices themselves, whether > > they support this or not. In practice I don't think we have any such cases > > at the moment, but it's possible in principle. Changes on the framework may > > well be needed but likely the biggest complications will still be in > > drivers supporting that. > > I understand V4L2 currently does not support changing a pipeline that is > running. However there are many use cases that would require it. > > Most of the use cases that come to my mind involve a multiplexer with > multiple inputs, one of which can be selected to be forwarded. In those > cases s_routing deselects an input and selects another one. How the can > we handle such cases without sending a s_stream on the two upstreams? > Having all possible inputs always running is not a real solution. This seems a very specific use case, I might surely be wrong, but if such a muxing device allows you to switch from one source to another without interruption, its driver should also take care of several other things, such as controlling power and format validations. I don't think that's something that should be considered at framework level, but again without a real use case I can't tell for real. > > > The media links have a dynamic flag for this purpose but I don't think it's > > ever been used. > > > >> > >> This is probably the wrong patch to use an example, as this one is for > >> a multiplexed interface, where there is no need to go through an > >> s_stream() for the two CSI-2 endpoints, but as you pointed out in our > >> brief offline chat, the AFE->TX routing example for this very device > >> is a good one: if we change the analogue source that is internally > >> routed to the CSI-2 output of the adv748x, do we need to s_stream(1) > >> the now routed entity and s_stream(0) on the not not-anymore-routed > >> one? > >> > >> My gut feeling is that this is up to userspace, as it should know > >> what are the requirements of the devices in the system, but this mean > >> going through an s_stream(0)/s_stream(1) sequence on the video device, > >> and that would interrupt the streaming for sure. > >> > >> At the same time, I don't feel too much at ease with the idea of > >> s_routing calling s_stream on the entity' remote subdevices, as this > >> would skip the link format validation that media_pipeline_start() > >> performs. > > > > The link validation must be done in this case as well, it may not be > > simply skipped. > > Agreed. > > The routing VS pipeline validation point is a very important one. The > current proposed workflow is: > > 1. the pipeline is validated as a whole, having knowledge of all the > entities let me specify this to avoid confusions: "all the entities -with an active route in the pipeline-" > 2. streaming is started > 3. s_routing is called on an entity (not on the pipeline!) > > Now the s_routing function in the entity driver is not in a good > position to validate the candidate future pipeline as a whole. > > Naively I'd say there are two possible solutions: > > 1. the s_routing reaches the pipeline first, then the new pipeline is > computed and verified, and if verification succeeds it is applied > 2. a partial pipeline verification mechanism is added, so the entity > receiving a s_routing request to e.g. change the sink pad can invoke > a verification on the part of pipeline that is about to be > activated, and if verification succeeds it is applied > > Somehow I suspect neither is trivial... I would say it is not, but if you have such a device that does not require going through a s_stream(0)/s_stream(1) cycle and all the associated re-negotiation and validations, it seems to me nothing prevents you from handling this in the driver implementation. Maybe it won't look that great, but this seems to be quite a custom design that requires all input sources to be linked to your sink pads, their format validated all at the same time, power, stream activation and internal mux configuration controlled by s_routing. Am I wrong or nothing in this series would prevent your from doing this? tl;dr: I would not make this something the framework should be concerned about, as there's nothing preventing you from implementing support for such a use case. But again, without a real example we can only guess, and I might be overlooking the issue or missing some relevant detail for sure. Thanks j > > -- > Luca > >