On Wed, Aug 10, 2016 at 5:33 PM, Huang, Ying wrote: > > Here it is, Thanks. Appended is a munged "after" list, with the "before" values in parenthesis. It actually looks fairly similar. The biggest difference is that we have "mark_page_accessed()" show up after, and not before. There was also a lot of LRU noise in the non-profile data. I wonder if that is the reason here: the old model of using generic_perform_write/block_page_mkwrite didn't mark the pages accessed, and now with iomap_file_buffered_write() they get marked as active and that screws up the LRU list, and makes us not flush out the dirty pages well (because they are seen as active and not good for writeback), and then you get bad memory use. I'm not seeing anything that looks like locking-related. And I may well have screwed up that list munging. I should have automated it more than I did. Dave, Christoph? Linus --- intel_idle 16.82 (16.88) copy_user_enhanced_fast_string 3.27 (3.94) memset_erms 2.6 (3.26) xfs_bmapi_read 2.24 ___might_sleep 2.04 (2.33) mark_page_accessed 1.93 __block_write_begin_int 1.78 (1.56) up_write 1.72 (1.61) xfs_iext_bno_to_ext 1.7 __block_commit_write.isra.24 1.65 (2.47) down_write 1.51 (1.55) __mark_inode_dirty 1.51 (1.88) unlock_page 1.43 (1.69) xfs_bmap_search_multi_extents 1.25 xfs_bmap_search_extents 1.23 mark_buffer_dirty 1.21 (1.53) xfs_iomap_write_delay 1.19 xfs_iomap_eof_want_preallocate.constprop.8 1.15 iomap_write_actor 1.14 __might_sleep 1.12 (1.26) __radix_tree_lookup 1.08 (1.32) entry_SYSCALL_64_fastpath 1.07 (1.47) pagecache_get_page 0.95 (1.03) _raw_spin_lock 0.95 (1.17) xfs_bmapi_delay 0.93 vfs_write 0.92 (1.14) xfs_file_write_iter 0.86