FYI, we noticed the following commit: https://github.com/0day-ci/linux Nicholas-Piggin/optimise-unlock_page-end_page_writeback/20161102-150709 commit 6ab7850933aa7c4859c30045626a86d9cdb5ac61 ("mm: Use owner_priv bit for PageSwapCache, valid when PageSwapBacked") in testcase: fsmark with following parameters: iterations: 1x nr_threads: 1t disk: 8BRD_12G md: RAID6 fs: btrfs filesize: 4M test_size: 60G sync_method: fsyncBeforeClose cpufreq_governor: performance The fsmark is a file system benchmark to test synchronous write workloads, for example, mail servers workload. on test machine: 144 threads Intel(R) Xeon(R) CPU E7-8890 v3 @ 2.50GHz with 128G memory caused below changes: +------------------------------------------------------------------+------------+------------+ | | 0c183d92b2 | 6ab7850933 | +------------------------------------------------------------------+------------+------------+ | boot_successes | 1 | 0 | | boot_failures | 1 | 1 | | invoked_oom-killer:gfp_mask=0x | 1 | | | Mem-Info | 1 | | | Kernel_panic-not_syncing:Out_of_memory_and_no_killable_processes | 1 | | | BUG:kernel_reboot-without-warning_in_test_stage | 0 | 1 | +------------------------------------------------------------------+------------+------------+ user :notice: [ 56.147733] 2016-11-07 14:36:14 fs_mark -d /fs/md0/1 -n 15360 -L 1 -S 1 -s 4194304 kern :crit : [ 56.160360] BTRFS critical (device md0): corrupt leaf, non-root leaf's nritems is 0: block=29540352, root=1, slot=0 user :notice: [ 56.161213] # fs_mark -d /fs/md0/1 -n 15360 -L 1 -S 1 -s 4194304 user :notice: [ 56.162106] # Version 3.3, 1 thread(s) starting at Mon Nov 7 14:36:14 2016 To reproduce: git clone git://git.kernel.org/pub/scm/linux/kernel/git/wfg/lkp-tests.git cd lkp-tests bin/lkp install job.yaml # job file is attached in this email bin/lkp run job.yaml Thanks, Kernel Test Robot