]> err.no Git - linux-2.6/commit
[PATCH] compound page: default destructor
authorHugh Dickins <hugh@veritas.com>
Tue, 14 Feb 2006 21:52:59 +0000 (13:52 -0800)
committerLinus Torvalds <torvalds@g5.osdl.org>
Wed, 15 Feb 2006 00:09:33 +0000 (16:09 -0800)
commitd98c7a09843621f1b145ca5ae8ed03ff04085edb
tree764b943efb177ca506fad61a63bb9e8cbf56a3a7
parent41d78ba55037468e6c86c53e3076d1a74841de39
[PATCH] compound page: default destructor

Somehow I imagined that calling a NULL destructor would free a compound page
rather than oopsing.  No, we must supply a default destructor, __free_pages_ok
using the order noted by prep_compound_page.  hugetlb can still replace this
as before with its own free_huge_page pointer.

The case that needs this is not common: rarely does put_compound_page's
put_page_testzero bring the count down to 0.  But if get_user_pages is applied
to some part of a compound page, without immediate release (e.g.  AIO or
Infiniband), then it's possible for its put_page to come after the containing
vma has been unmapped and the driver done its free_pages.

That's just the kind of case compound pages are supposed to be guarding
against (but Nick points out, nor did PageReserved handle this right).

Signed-off-by: Hugh Dickins <hugh@veritas.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
mm/page_alloc.c