Commit a04cd160 authored by Zi Yan's avatar Zi Yan Committed by Linus Torvalds
Browse files

mm: migrate: use thp_order instead of HPAGE_PMD_ORDER for new page allocation. 

Fix a VM_BUG_ON_FOLIO(folio_nr_pages(old) != nr_pages) crash.

With folios support, it is possible to have other than HPAGE_PMD_ORDER
THPs, in the form of folios, in the system.  Use thp_order() to correctly
determine the source page order during migration.

Link: https://lkml.kernel.org/r/20220404165325.1883267-1-zi.yan@sent.com
Link: https://lore.kernel.org/linux-mm/20220404132908.GA785673@u2004/


Fixes: d68eccad ("mm/filemap: Allow large folios to be added to the page cache")
Reported-by: default avatarNaoya Horiguchi <naoya.horiguchi@linux.dev>
Signed-off-by: default avatarZi Yan <ziy@nvidia.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Michal Hocko <mhocko@kernel.org>
Signed-off-by: default avatarAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: default avatarLinus Torvalds <torvalds@linux-foundation.org>
parent d00c50b3

mm/mempolicy.c

+1 −1
Original line number Diff line number Diff line
@@ -1209,7 +1209,7 @@ static struct page *new_page(struct page *page, unsigned long start) 
		struct page *thp;



		thp = alloc_hugepage_vma(GFP_TRANSHUGE, vma, address,

					 HPAGE_PMD_ORDER);

					 thp_order(page));

		if (!thp)

			return NULL;

		prep_transhuge_page(thp);

mm/migrate.c

+1 −1
Original line number Diff line number Diff line
@@ -1547,7 +1547,7 @@ struct page *alloc_migration_target(struct page *page, unsigned long private) 
		 */

		gfp_mask &= ~__GFP_RECLAIM;

		gfp_mask |= GFP_TRANSHUGE;

		order = HPAGE_PMD_ORDER;

		order = thp_order(page);

	}

	zidx = zone_idx(page_zone(page));

	if (is_highmem_idx(zidx) || zidx == ZONE_MOVABLE)