Guennadi Liakhovetski <g.liakhovetski@xxxxxx> writes: >> - I unload and reload mt9m111 and pxa_camera >> => not any better > > Actually, I think, in this case it should be found again, as long as you > reload pxa-camera while i2c-pxa is already loaded. Damn, you're right. I cross-checked, and reloading pxa_camera rescans the sensor. >> What I'm getting at is that if soc_camera is loaded before the i2c host driver, >> no camera will get any chance to work. Is that normal considering the new driver >> model ? >> I was naively thinking that there would be a "rescan" when the "control" was >> being available for a sensor. > > Yes, unfortunately, it is "normal":-( On the one hand, we shouldn't really > spend _too_ much time on this intermediate version, because, as I said, it > is just a preparatory step for v4l2-subdev. We just have to make sure it > doesn't introduce any significant regressions and doesn't crash too often. OK. So from my side everything is OK (let aside my nitpicking in mioa701.c and mt9m111.c). > OTOH, this is also how it is with v4l2-subdev. With it you first must have > the i2c-adapter driver loaded. Then, when a match between a camera host > and a camera client (sensor) platform device is detected, it is reported > to the v4l2-subdev core, which loads the respective camera i2c driver. OK, why not. > If you then unload the camera-host and i2c adapter drivers, and then you load > the camera-host driver, it then fails to get the adapter, and if you then load > it, nothing else happens. To reprobe you have to unload and reload the camera > host driver. So be it. I'm sure we'll be through it once more in the v4l2-subdev transition, so I'll let aside any objection I could mutter :) Cheers. -- Robert -- 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
