On Mon, Feb 27, 2017 at 03:33:15PM +0900, Minchan Kim wrote: > Hi Shaohua, > > On Fri, Feb 24, 2017 at 01:31:47PM -0800, Shaohua Li wrote: > > When memory pressure is high, we free MADV_FREE pages. If the pages are > > not dirty in pte, the pages could be freed immediately. Otherwise we > > can't reclaim them. We put the pages back to anonumous LRU list (by > > setting SwapBacked flag) and the pages will be reclaimed in normal > > swapout way. > > > > We use normal page reclaim policy. Since MADV_FREE pages are put into > > inactive file list, such pages and inactive file pages are reclaimed > > according to their age. This is expected, because we don't want to > > reclaim too many MADV_FREE pages before used once pages. > > > > Based on Minchan's original patch > > > > Cc: Michal Hocko <[email protected]> > > Cc: Minchan Kim <[email protected]> > > Cc: Hugh Dickins <[email protected]> > > Cc: Johannes Weiner <[email protected]> > > Cc: Rik van Riel <[email protected]> > > Cc: Mel Gorman <[email protected]> > > Cc: Andrew Morton <[email protected]> > > Signed-off-by: Shaohua Li <[email protected]> > > --- > > include/linux/rmap.h | 2 +- > > mm/huge_memory.c | 2 ++ > > mm/madvise.c | 1 + > > mm/rmap.c | 40 +++++++++++++++++----------------------- > > mm/vmscan.c | 34 ++++++++++++++++++++++------------ > > 5 files changed, 43 insertions(+), 36 deletions(-) > > > > diff --git a/include/linux/rmap.h b/include/linux/rmap.h > > index 7a39414..fee10d7 100644 > > --- a/include/linux/rmap.h > > +++ b/include/linux/rmap.h > > @@ -298,6 +298,6 @@ static inline int page_mkclean(struct page *page) > > #define SWAP_AGAIN 1 > > #define SWAP_FAIL 2 > > #define SWAP_MLOCK 3 > > -#define SWAP_LZFREE 4 > > +#define SWAP_DIRTY 4 > > I still don't convinced why we should introduce SWAP_DIRTY in try_to_unmap. > https://marc.info/?l=linux-mm&m=148797879123238&w=2 > > We have been SetPageMlocked in there but why cannot we SetPageSwapBacked > in there? It's not a thing to change LRU type but it's just indication > we found the page's status changed in late.
This one I don't have strong preference. Personally I agree with Johannes, handling failure in vmscan sounds better. But since the failure handling is just one statement, this probably doesn't make too much difference. If Johannes and you made an agreement, I'll follow. Thanks, Shaohua
