On Thu, 2017-12-28 at 08:26 +0900, Minchan Kim wrote: > Hello James, > > On Wed, Dec 27, 2017 at 12:50:17PM -0800, James Bottomley wrote: > > > > Reverting these three patches fixes the problem: > > > > commit aa8d22a11da933dbf880b4933b58931f4aefe91c > > Author: Minchan Kim <minchan@xxxxxxxxxx> > > Date: Wed Nov 15 17:33:11 2017 -0800 > > > > mm: swap: SWP_SYNCHRONOUS_IO: skip swapcache only if swapped > > page > > has no other reference > > > > commit 0bcac06f27d7528591c27ac2b093ccd71c5d0168 > > Author: Minchan Kim <minchan@xxxxxxxxxx> > > Date: Wed Nov 15 17:33:07 2017 -0800 > > > > mm, swap: skip swapcache for swapin of synchronous device > > > > Also need to revert: > > > > commit e9a6effa500526e2a19d5ad042cb758b55b1ef93 > > Author: Huang Ying <huang.ying.caritas@xxxxxxxxx> > > Date: Wed Nov 15 17:33:15 2017 -0800 > > > > mm, swap: fix false error message in __swp_swapcount() > > > > (The latter is simply because it used a function that is eliminated > > by > > one of the other reversions). They came into the merge window via > > the > > -mm tree as part of a 4 part series: > > > > Subject: [PATCH v2 0/4] skip swapcache for super fast device > > Message-Id: <1505886205-9671-1-git-send-email-minchan@kernel > > .org > > > > > > > > > > James > > Thanks for the report. > Patches are related to synchronous swap devices like brd, zram, > nvdimm so > > 1. What swap device do you use among them? I've reproduced on nvme and sata spinning rust. > 2. Could you tell me how you can reproduce it? The way to reproduce is to force something to swap and then get it to try to touch the page again. I do this on my systems by using a large virtual machine, as I said in the email. There isn't really any definitive reproduction method beyond that. James -- 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>