On Tue, Feb 04, 2025 at 07:55:52PM +0100, David Hildenbrand wrote: > Ah, and now I remember where these 3 patches originate from: virtio-mem > handling. > > For virtio-mem I want to register also a remap handler, for example, to > perform the custom preallocation handling. > > So there will be at least two instances getting notified (memory backend, > virtio-mem), and the per-ramblock one would have only allowed to trigger one > (at least with a simple callback as we have today for ->resize). I see, we can put something into commit log with such on decisions, then we'll remember. Said that, this still sounds like a per-ramblock thing, so instead of one hook function we can also have per-ramblock notifier lists. But I agree the perf issue isn't some immediate concern, so I'll leave that to you and William. If so I think we should discuss that in the commit log too, so we decide to not care about perf until necessary (or we just make it per-ramblock..). Thanks, -- Peter Xu
