Hi Joonsoo, Sorry for late response. On Tue, Aug 26, 2014 at 05:08:14PM +0900, Joonsoo Kim wrote: > This is version 3 patchset which is improved and minimized version of > version 1 to fix freepage accounting problem during memory isolation. > I tried different approach in version 2, but, it looks really complicated > so I change my mind to improve version 1. You can see version 1, 2 in > following links [1] [2], respectively. > > IMO, this v3 is better than v2, because this is simpler than v2 so > better for maintainance and this doesn't change pageblock isolation > logic so it is much easier to backport. > > This problems are found by testing my patchset [3]. There are some race > conditions on pageblock isolation and these race cause incorrect > freepage count. > > Before describing bugs itself, I first explain definition of freepage. > > 1. pages on buddy list are counted as freepage. > 2. pages on isolate migratetype buddy list are *not* counted as freepage. > 3. pages on cma buddy list are counted as CMA freepage, too. > > Now, I describe problems and related patch. > > Patch 1: There is race conditions on getting pageblock migratetype that > it results in misplacement of freepages on buddy list, incorrect > freepage count and un-availability of freepage. > > Patch 2: Freepages on pcp list could have stale cached information to > determine migratetype of buddy list to go. This causes misplacement > of freepages on buddy list and incorrect freepage count. > > Patch 4: Merging between freepages on different migratetype of > pageblocks will cause freepages accouting problem. This patch fixes it. Look though this patchset, it's really simple compared to v2 although it adds some overhead on hotpath but I support this patchset unless other suggest more simple/clean code to fix horrible race bugs. Thanks a lot! > > Without patchset [3], above problem doesn't happens on my CMA allocation > test, because CMA reserved pages aren't used at all. So there is no > chance for above race. > > With patchset [3], I did simple CMA allocation test and get below result. > > - Virtual machine, 4 cpus, 1024 MB memory, 256 MB CMA reservation > - run kernel build (make -j16) on background > - 30 times CMA allocation(8MB * 30 = 240MB) attempts in 5 sec interval > - Result: more than 5000 freepage count are missed > > With patchset [3] and this patchset, I found that no freepage count are > missed so that I conclude that problems are solved. > > These problems can be possible on memory hot remove users, although > I didn't check it further. > > This patchset is based on linux-next-20140826. > Please see individual patches for more information. > > Thanks. > > [1]: https://lkml.org/lkml/2014/7/4/79 > [2]: lkml.org/lkml/2014/8/6/52 > [3]: Aggressively allocate the pages on cma reserved memory > https://lkml.org/lkml/2014/5/30/291 > > Joonsoo Kim (4): > mm/page_alloc: fix incorrect isolation behavior by rechecking > migratetype > mm/page_alloc: add freepage on isolate pageblock to correct buddy > list > mm/page_alloc: move migratetype recheck logic to __free_one_page() > mm/page_alloc: restrict max order of merging on isolated pageblock > > include/linux/mmzone.h | 4 ++++ > include/linux/page-isolation.h | 8 ++++++++ > mm/page_alloc.c | 28 +++++++++++++++++++--------- > mm/page_isolation.c | 2 ++ > 4 files changed, 33 insertions(+), 9 deletions(-) > > -- > 1.7.9.5 > > -- > To unsubscribe, send a message with 'unsubscribe linux-mm' in > the body to majordomo@xxxxxxxxx. For more info on Linux MM, > see: http://www.linux-mm.org/ . > Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a> -- Kind regards, Minchan Kim -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>