On Tue, Oct 19, 2010 at 06:41:37AM +0800, Andrew Morton wrote: > On Tue, 19 Oct 2010 09:31:42 +1100 > Neil Brown <neilb@xxxxxxx> wrote: > > > On Mon, 18 Oct 2010 14:58:59 -0700 > > Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> wrote: > > > > > On Tue, 19 Oct 2010 00:15:04 +0800 > > > Wu Fengguang <fengguang.wu@xxxxxxxxx> wrote: > > > > > > > Neil find that if too_many_isolated() returns true while performing > > > > direct reclaim we can end up waiting for other threads to complete their > > > > direct reclaim. If those threads are allowed to enter the FS or IO to > > > > free memory, but this thread is not, then it is possible that those > > > > threads will be waiting on this thread and so we get a circular > > > > deadlock. > > > > > > > > some task enters direct reclaim with GFP_KERNEL > > > > => too_many_isolated() false > > > > => vmscan and run into dirty pages > > > > => pageout() > > > > => take some FS lock > > > > => fs/block code does GFP_NOIO allocation > > > > => enter direct reclaim again > > > > => too_many_isolated() true > > > > => waiting for others to progress, however the other > > > > tasks may be circular waiting for the FS lock.. > > I'm assuming that the last four "=>"'s here should have been indented > another stop. Yup. I'll fix it in next post. > > > > The fix is to let !__GFP_IO and !__GFP_FS direct reclaims enjoy higher > > > > priority than normal ones, by honouring them higher throttle threshold. > > > > > > > > Now !GFP_IOFS reclaims won't be waiting for GFP_IOFS reclaims to > > > > progress. They will be blocked only when there are too many concurrent > > > > !GFP_IOFS reclaims, however that's very unlikely because the IO-less > > > > direct reclaims is able to progress much more faster, and they won't > > > > deadlock each other. The threshold is raised high enough for them, so > > > > that there can be sufficient parallel progress of !GFP_IOFS reclaims. > > > > > > I'm not sure that this is really a full fix. Torsten's analysis does > > > appear to point at the real bug: raid1 has code paths which allocate > > > more than a single element from a mempool without starting IO against > > > previous elements. > > > > ... point at "a" real bug. > > > > I think there are two bugs here. > > The raid1 bug that Torsten mentions is certainly real (and has been around > > for an embarrassingly long time). > > The bug that I identified in too_many_isolated is also a real bug and can be > > triggered without md/raid1 in the mix. > > So this is not a 'full fix' for every bug in the kernel :-), > > but it could well be a full fix for this particular bug. Yeah it aims to be a full fix for one bug. > Can we just delete the too_many_isolated() logic? (Crappy comment If the two cond_resched() calls can be removed from shrink_page_list(), the major cause of too many pages being isolated will be gone. However the writeback-waiting logic after should_reclaim_stall() will also block the direct reclaimer for long time with pages isolated, which may bite under pathological conditions. > describes what the code does but not why it does it). Good point. The comment could be improved as follows. Thanks, Fengguang --- Subject: vmscan: comment too_many_isolated() From: Wu Fengguang <fengguang.wu@xxxxxxxxx> Date: Tue Oct 19 09:53:23 CST 2010 Comment "Why it's doing so" rather than "What it does" as proposed by Andrew Morton. Signed-off-by: Wu Fengguang <fengguang.wu@xxxxxxxxx> --- mm/vmscan.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) --- linux-next.orig/mm/vmscan.c 2010-10-19 09:29:44.000000000 +0800 +++ linux-next/mm/vmscan.c 2010-10-19 10:21:41.000000000 +0800 @@ -1142,7 +1142,11 @@ int isolate_lru_page(struct page *page) } /* - * Are there way too many processes in the direct reclaim path already? + * A direct reclaimer may isolate SWAP_CLUSTER_MAX pages from the LRU list and + * then get resheduled. When there are massive number of tasks doing page + * allocation, such sleeping direct reclaimers may keep piling up on each CPU, + * the LRU list will go small and be scanned faster than necessary, leading to + * unnecessary swapping, thrashing and OOM. */ static int too_many_isolated(struct zone *zone, int file, struct scan_control *sc) -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxxx For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>