<!-- Some styling for better description lists --><style type='text/css'>dt { font-weight: bold;float: left;display:inline;margin-right: 1em} dd { display:block; margin-left: 2em}</style>

   ***: mszyprow has quit IRC (Ping timeout: 244 seconds)
   sailus: <u>javier__</u>: I'd prefer BUG_ON().
   <br> I think we could remove the model check there.
   <br> It wouldn't hurt to have a second opinion on that though.
   pinchartl: <u>sailus</u>: BUG_ON is harsh
   <br> it's supposed to be used only in situations where continuing execution of the kernel would cause serious issues (security, data corruption, ...)
   <br> as it calls panic()
   javier__: <u>pinchartl</u>: that was my argument as well
   <br> <u>sailus</u>: so I could either remove the checks assuming drivers do the right thing (all drivers do now and is even mentioned in the function doc) or leave it as WARN_ON
   <br> I prefer the former fwiw
   pinchartl: which function are we talking about ?
   sailus: <u>pinchartl</u>: media_device_init()
   <br> This is about javier__'s patches splitting media_device_register() into two.
   <br> media_entity_init() can only fail if the user passes wrong arguments to it, i.e. the dev parameter is NULL.
   <br> Alternatively, the user would have to perform error handling on this, and I don't think it's worth it.
   javier__: <u>pinchartl</u>: sorry I was not online but sailus already explained
   <br> <u>pinchartl</u>: you can look here for contex https://lkml.org/lkml/2015/9/11/72
   <br> *context
   <br> <u>sailus</u>: yes, I agree with you that is not worth it and changed that on v2