Hi, about 3 years ago i firstly write to linux-mm, when i acking about UKSM and make some ugly patches (like that [1]). It's difficult to find conversation (for me at least, i only find my patches). So IIRC: - I trying add support KSM to all process memory for that i patch several memory places to call hooks like madvise. MM folks not like that and at now i'm not only agree, i can fix that another way. First i try look at hugepagesd, but hugepages have a much different usecase. As example it's bad idea add ksm hook to page_fault. So: I use kernel task list in ksm_scan_thread and add logic to allow ksm import VMA from tasks. That behaviour controlled by new attribute: mode I try mimic hugepages attribute, so mode have two states: - normal [old default behaviour] - always [new] - allow ksm to get tasks vma and try working on that. To reduce CPU load & tasklist locking time, ksm try import one VMA per loop. Patch mirror: https://github.com/Nefelim4ag/linux/commit/be6a94e171bf214a26f8186bb76b1d365ccb3b08 Thanks, any comments are appricated! (Yes i can split patch, but it's RFC and it's too small) 1. https://lkml.org/lkml/2014/11/8/208 Timofey Titovets (1): ksm: allow dedup all tasks memory Documentation/vm/ksm.txt | 3 + mm/ksm.c | 139 ++++++++++++++++++++++++++++++++++++++++------- 2 files changed, 121 insertions(+), 21 deletions(-) -- 2.14.1 -- 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>
