Messages in this thread | | | Date | Wed, 13 Feb 2019 13:33:39 +0100 | From | Michal Hocko <> | Subject | Re: [PATCH] mm,memory_hotplug: Explicitly pass the head to isolate_huge_page |
| |
On Wed 13-02-19 09:13:14, Oscar Salvador wrote: > On Tue, Feb 12, 2019 at 04:13:05PM -0800, Mike Kravetz wrote: > > Well, commit 94310cbcaa3c ("mm/madvise: enable (soft|hard) offline of > > HugeTLB pages at PGD level") should have allowed migration of gigantic > > pages. I believe it was added for 16GB pages on powerpc. However, due > > to subsequent changes I suspsect this no longer works. > > I will take a look, I am definitely interested in that. > Thanks for pointing it out Mike. > > > > > > This check doesn't make much sense in principle. Why should we bail out > > > based on a section size? We are offlining a pfn range. All that we care > > > about is whether the hugetlb is migrateable. > > > > Yes. Do note that the do_migrate_range is only called from __offline_pages > > with a start_pfn that was returned by scan_movable_pages. scan_movable_pages > > has the hugepage_migration_supported check for PageHuge pages. So, it would > > seem to be redundant to do another check in do_migrate_range. > > Well, the thing is that if the gigantic page does not start at the very beginning > of the memblock, and we do find migrateable pages before it in scan_movable_pages(), > the range that we will pass to do_migrate_ranges() will contain the gigantic page. > So we need the check there to cover that case too, although I agree that the current > check is misleading.
Why isn't our check in has_unmovable_pages sufficient? -- Michal Hocko SUSE Labs
| |