Messages in this thread | | | Subject | Re: [PATCH] mm: use managed_zone() for more exact check in zone iteration | From | osalvador <> | Date | Fri, 16 Nov 2018 12:05:04 +0100 |
| |
On Fri, 2018-11-16 at 10:57 +0100, Michal Hocko wrote: > On Thu 15-11-18 13:37:35, Andrew Morton wrote: > [...] > > Worse, the situations in which managed_zone() != populated_zone() > > are > > rare(?), so it will take a long time for problems to be discovered, > > I > > expect. > > We would basically have to deplete the whole zone by the bootmem > allocator or pull out all pages from the page allocator. E.g. memory > hotplug decreases both managed and present counters. I am actually > not > sure that is 100% correct (put on my TODO list to check). There is no > consistency in that regards.
We can only offline non-reserved pages (so, managed pages). Since present pages holds reserved_pages + managed_pages, decreasing both should be fine unless I am mistaken.
Oscar Salvador
| |