On 11/12/18 8:54 AM, Josef Bacik wrote: > On Mon, Nov 12, 2018 at 08:48:35AM -0700, Jens Axboe wrote: >> On 11/12/18 8:45 AM, Paolo Valente wrote: >>> BTW, since this patch series is probably even more useful for older >>> than for future kernels, might it make sense to also propose it for >>> stable/longterm kernels (provided that such a possibility exists)? >> >> That just not how things work, we don't put different things in >> older/stable kernels, it's strictly backports of what we have in >> current/newer kernels. Hence it appears to be a dead end right now. >> > > It may not be useful currently, but my plans are to do a scheduler agnostic > proportional io controller next, so having these interfaces unified would be > nice so I don't have to do a rqos.io.weight or something similar. Thanks, I'm not saying the work isn't useful, I'm saying that we can't go adding different interfaces to stable kernels than what we currently have in tip. I'm all for unified interfaces for this kind of thing, it's much better than having something that's specific to any given implementation. -- Jens Axboe
