On Wed, Oct 25, 2017 at 10:06:36AM +0200, Mark Brown wrote: >On Wed, Oct 25, 2017 at 01:56:02AM +0000, Levin, Alexander (Sasha Levin) wrote: > >> When regulators are successfully registered, we check to see if the >> regulator is a supply for any other registered regulator and if so >> add the new regulator as the supply for the existing regulator(s). > >This had some knock on effects which caused trouble for other things, >for example 3827b64dba (regulator: core: Resolve supplies before >disabling unused regulators). > >In general I would be extremely wary of backporting anything like this >that affects things like device instantiation ordering, even without any >knock on bugs directly in the code there's a good chance that someone >building on stable is going to have some kind of dependency on the >particular behaviour of the kernel version they're using and changes >like this could cause updating to the latest stable to introduce >problems which then discourages people from picking up stable. On the other hand, we do want to be "bug compatible" with upstream, if we force a stable release to behave in a broken way it'll be hard later to move to a newer stable release. I think that we'd rather pull in fixes than avoid breaking broken behaviors. -- Thanks, Sasha