Re: [PATCH] mm, memory_hotplug: teach has_unmovable_pages about of LRU migrateable pages

[Michal Hocko]

On Tue 06-11-18 08:22:16, Baoquan He wrote:
> On 11/05/18 at 06:10pm, Michal Hocko wrote:
> > On Mon 05-11-18 22:23:08, Baoquan He wrote:
> > > On 11/05/18 at 01:38pm, Michal Hocko wrote:
> > > > On Mon 05-11-18 18:25:20, Baoquan He wrote:
> > > > > Hi Michal,
> > > > > 
> > > > > On 11/05/18 at 10:28am, Michal Hocko wrote:
> > > > > > 
> > > > > > Or something like this. Ugly as hell, no question about that. I also
> > > > > > have to think about this some more to convince myself this will not
> > > > > > result in an endless loop under some situations.
> > > > > 
> > > > > It failed. Paste the log and patch diff here, please help check if I made
> > > > > any mistake on manual code change. The log is at bottom.
> > > > 
> > > > The retry patch is obviously still racy, it just makes the race window
> > > > slightly smaller and I hoped it would catch most of those races but this
> > > > is obviously not the case.
> > > > 
> > > > I was thinking about your MIGRATE_MOVABLE check some more and I still do
> > > > not like it much, we just change migrate type at many places and I have
> > > > hard time to actually see this is always safe wrt. to what we need here.
> > > > 
> > > > We should be able to restore the zone type check though. The
> > > > primary problem fixed by 15c30bc09085 ("mm, memory_hotplug: make
> > > > has_unmovable_pages more robust") was that early allocations made it to
> > > > the zone_movable range. If we add the check _after_ the PageReserved()
> > > > check then we should be able to rule all bootmem allocation out.
> > > > 
> > > > So what about the following (on top of the previous patch which makes
> > > > sense on its own I believe).
> > > 
> > > Yes, I think this looks very reasonable and should be robust.
> > > 
> > > Have tested it, hot removing 4 hotpluggable nodes continusously
> > > succeeds, and then hot adding them back, still works well.
> > > 
> > > So please feel free to add my Tested-by or Acked-by.
> > > 
> > > Tested-by: Baoquan He <bhe@xxxxxxxxxx>
> > > or
> > > Acked-by: Baoquan He <bhe@xxxxxxxxxx>
> > 
> > Thanks for retesting! Does this apply to both patches?
> 
> Sorry, don't get it. I just applied this on top of linus's tree and
> tested. Do you mean applying it on top of previous code change?

Yes. While the first patch will obviously not help for movable zone
because the movable check will override any later check it
seems still useful to reduce false positives on normal zones.

Or do you think this is not worth it?

-- 
Michal Hocko
SUSE Labs