[Pmem] [PATCH 0/3] Map page as PROT_NONE after demotion

Hi, This patch set backported several patches from git branch[1], which helps to reduce the overhead of page table entries scanning and identify pages demoted wrongly by mapping the new page as PROT_NONE when page demotion. With this patch set, there are some slight performance improvement with mysql read and write testing. Testing command and results: sysbench /usr/share/sysbench/oltp_read_write.lua --mysql-user=root --mysql-password=root --tables=200 --table-size=1000000 --threads=16 --time=600 --report-interval=10 prepare With patches: queries performed: transactions: 799824 (1331.06 per sec.) queries: 15996480 (26621.19 per sec.) latency: avg: 12.02 95th percentile: 20.37 Without patches: queries performed: transactions: 791806 (1318.89 per sec.) queries: 15836120 (26377.75 per sec.) latency: avg: 12.13 95th percentile: 20.37 [1] https://git.kernel.org/pub/scm/linux/kernel/git/vishal/tiering.git/log/?h=t… Huang Ying (3): mm, migrate: use flags parameter for remove_migration_ptes() memory tiering: map page as PROT_NONE after demotion memory tiering: control rate limit based on pgpromote_demoted include/linux/mmzone.h | 4 +++ include/linux/page-flags.h | 9 ++++++ include/linux/page_ext.h | 3 ++ include/linux/rmap.h | 8 ++++- include/linux/sched/numa_balancing.h | 61 ++++++++++++++++++++++++++++++++++++ include/trace/events/mmflags.h | 9 +++++- include/trace/events/sched.h | 12 ++++--- kernel/sched/fair.c | 28 +++++++++++++++-- mm/huge_memory.c | 4 +-- mm/mempolicy.c | 2 ++ mm/migrate.c | 59 ++++++++++++++++++++++++++++------ mm/vmstat.c | 1 + 12 files changed, 179 insertions(+), 21 deletions(-) -- 1.8.3.1