On Fri, Jan 28, 2022 at 06:44:24PM +0200, Laurent Pinchart wrote: > Hi Sakari, > > On Fri, Jan 28, 2022 at 06:35:59PM +0200, Sakari Ailus wrote: > > On Fri, Jan 28, 2022 at 06:30:33PM +0200, Laurent Pinchart wrote: > > > On Fri, Jan 28, 2022 at 06:07:36PM +0200, Sakari Ailus wrote: > > > > On Fri, Jan 28, 2022 at 04:58:40PM +0200, Laurent Pinchart wrote: > > > > > Hi Sakari, > > > > > > > > > > Thank you for the patch. > > > > > > > > And thanks for the review! > > > > > > > > > On Fri, Jan 28, 2022 at 10:33:09AM +0200, Sakari Ailus wrote: > > > > > > The bus_info field is set by most drivers based on the type of the device > > > > > > bus as well as the name of the device. Do this in v4l_querycap() so > > > > > > drivers don't need to. This keeps compatibility with non-default and silly > > > > > > bus_info. > > > > > > > > > > Do we have many PCI or platform drivers with non-default bus_info ? > > > > > > > > Quite a few actually. Most have something static in bus_info and there are > > > > also PCI Express devices that use "PCIe:" prefix. These are mostly old > > > > drivers though. I've left them as-is. > > > > > > > > New drivers shouldn't bother changing the default unless there's a good > > > > reason to do that. > > > > > > Let's make it very clear in the documentation then. > > > > Right now there's no place for documenting this. Basically this is about > > what IOCTL argument struct fields are set in the framework and which ones > > the driver is responsible for. I don't think addressing that belongs to > > this set. > > As this sets tries to fix the mess caused by drivers using semi-random > bus_info, I'd like to disagree :-) It can be done in a patch on top. Well I think starting to document where the responsibility of filling IOCTL argument structs fields is definitely does not belong here. Maybe Hans has an opinion? :-) -- Sakari Ailus
