Messages in this thread | | | Date | Wed, 14 Aug 2013 16:40:52 -0700 | From | Andrew Morton <> | Subject | Re: [PATCH v5 0/9] extend hugepage migration |
| |
On Fri, 9 Aug 2013 01:21:33 -0400 Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> wrote:
> Here is the 5th version of hugepage migration patchset. > Changes in this version are as follows: > - removed putback_active_hugepages() as a cleanup (1/9) > - added code to check movability of a given hugepage (8/9) > - set GFP MOVABLE flag depending on the movability of hugepage (9/9). > > I feel that 8/9 and 9/9 contain some new things, so need reviews on them. > > TODOs: (likely to be done after this work) > - split page table lock for pmd/pud based hugepage (maybe applicable to thp) > - improve alloc_migrate_target (especially in node choice) > - using page walker in check_range
This is a pretty large and complex patchset. I skimmed the patches (and have one trivial comment) then queued them up for a bit of testing. I've asked Mel if he can find time to review the changes (please).
btw, it would be helpful if this [patch 0/n] had a decent overview of the patch series - what are the objectives, how were they achieved, what value they have to our users, testing results, etc.
mm-prepare-to-remove-proc-sys-vm-hugepages_treat_as_movable.patch had a conflict with http://ozlabs.org/~akpm/mmots/broken-out/mm-hugetlb-move-up-the-code-which-check-availability-of-free-huge-page.patch which I resolved in the obvious manner. Please check that from a runtime perspective.
| |