Re: [RFC PATCH] memory,memory_hotplug: allow restricting memory blocks to zone movable

[Gregory Price <gourry@gourry.net>]

On Mon, Jan 12, 2026 at 08:28:29AM +0100, Hannes Reinecke wrote:
> On 1/9/26 17:41, Gregory Price wrote:
> > On Thu, Jan 08, 2026 at 03:16:24PM +0100, David Hildenbrand (Red Hat) wrote:
> > > On 1/8/26 08:31, Hannes Reinecke wrote:
> > > > On 1/6/26 21:22, David Hildenbrand (Red Hat) wrote:
> > > > > On 1/6/26 20:59, Gregory Price wrote:
> > > 
> > 
> > I'm starting to think this issue is actually the result of bad patterns
> > in the cxl driver - namely using dax as a path to hotplug sysram.
> > 
> > I suppose either we need a `cxl/dax_region/remove` that handles the
> > whole operation in one go, or
> > 
> > we want `cxl/region/commit` to handle hot(un)plug as a single action.
> > 
> > tl;dr:  Split the dax use case from the sysram use case, and make a
> >          cxl sysram driver directly manage hotplug rather than use dax.
> > 
> 
> Well ... not sure.
> We are doing fine even currently during boot up; we can align policies
> and everything to ensure the system comes up with the 'correct' setting
> Things start to get iffy if one is reconfiguring memory to move from
> daxdev to system ram and vice versa.

Hang tight, i'll post an RFC this morning that shows the idea in code.

region0/memctl/hotplug
  - if not already hotplugged, creates the memory blocks
    unless MHP auto-online or new build config set, do not auto-online

region0/memctl/hotunplug
  - remove memory and hotunplug the blocks
    Done in a single action

region0/memctl/state
  - accepts [online, online_normal, offline]
    does the action on the entire memory region under the hotplug lock
    online - ZONE_MOVABLE
    online_normal - ZONE_NORMAL

I also added the memory event callback to prevent someone from onlining
in zone normal if the last online action was 'online'.  I've tested this
on QEMU and it works fine.

~Gregory