[libcamera-devel] [PATCH v5 5/5] libcamera: camera_manager: Enforce unique camera names

[Laurent Pinchart]

Hi Niklas,

On Mon, Aug 03, 2020 at 01:11:16AM +0200, Niklas Söderlund wrote:
> On 2020-08-03 00:19:10 +0300, Laurent Pinchart wrote:
> > On Wed, Jul 29, 2020 at 01:50:55PM +0200, Niklas Söderlund wrote:
> > > The camera name have always been documented that it should be unique but
> > > it has never been enforced. Change this by refusing to add cameras to
> > > the CameraManager that would create two cameras with the exact same
> > > name.
> > > 
> > > Signed-off-by: Niklas Söderlund <niklas.soderlund at ragnatech.se>
> > > Reviewed-by: Jacopo Mondi <jacopo at jmondi.org>
> > > Reviewed-by: Kieran Bingham <kieran.bingham at ideasonboard.com>
> > > ---
> > > * Changes since v4
> > > - Update string in error message.
> > > 
> > > * Changes since v3
> > > - Update commit message.
> > > ---
> > >  src/libcamera/camera_manager.cpp | 6 +++---
> > >  1 file changed, 3 insertions(+), 3 deletions(-)
> > > 
> > > diff --git a/src/libcamera/camera_manager.cpp b/src/libcamera/camera_manager.cpp
> > > index f60491d2c1a7500f..47e0f3129d346183 100644
> > > --- a/src/libcamera/camera_manager.cpp
> > > +++ b/src/libcamera/camera_manager.cpp
> > > @@ -178,10 +178,10 @@ void CameraManager::Private::addCamera(std::shared_ptr<Camera> camera,
> > >  
> > >  	for (std::shared_ptr<Camera> c : cameras_) {
> > >  		if (c->name() == camera->name()) {
> > > -			LOG(Camera, Warning)
> > > -				<< "Registering camera with duplicate name '"
> > > +			LOG(Camera, Error)
> > > +				<< "Skip registration of a camera with a duplicated name '"
> > >  				<< camera->name() << "'";
> > > -			break;
> > > +			return;
> > 
> > Shouldn't the error be propagated up to the pipelien handler ? Other
> > than that, this looks good to me, but I think we need to rename "name"
> > to "id".
> 
> I'm not sure it's useful to propagate the error to the pipeline handler, 
> what will it do with it? The pipeline can't change the ID and try adding 
> the camera again, as this would violate the requirements we have on IDs
> being stable and unique. All it can do is print yet another LOG message 
> saying a duplicated ID is found.

I'm fine not propagating the error as long as we can ensure that no
memory leak or bad memory access will occur. The pipeline handler will
think registration succeeded, and will return true from match(). Can you
check that nothing can crash ?

> I originally had this as a Fatal error as this really should not happen, 
> I'm tempted to switch it back. What do you think?

I think it could make sense, yes.

> > >  		}
> > >  	}
> > >  

-- 
Regards,

Laurent Pinchart