| |||||||||||||
A suspend blockers post-mortemA suspend blockers post-mortemPosted Jun 3, 2010 11:57 UTC (Thu) by jmspeex (subscriber, #51639)Parent article: A suspend blockers post-mortem
I don't know what the best solution is, but the issue here is that the call for embedded developers to merge their stuff is really in contradiction with how suspend blockers were handled. This means one should change, though I don't know which one. Either the developers accept this sort of functionality with the understanding that API/ABI *will* be unstable (and may be removed when something else comes), or else you just tell developers that "submit your patches upstream" only really applies for drivers.
(Log in to post comments)
A suspend blockers post-mortem Posted Jun 3, 2010 13:26 UTC (Thu) by tglx (subscriber, #31301) [Link]
> Either the developers accept this sort of functionality with the understanding that API/ABI *will* be unstable (and may be removed when something else comes) ...
That's the main problem. Once we have an user space visible ABI/API we can not break it. We are stuck to it.
In kernel APIs are known to be subject of change, but that's a totally different playground.
|
|||||||||||||