Messages in this thread | | | Date | Thu, 22 May 2014 12:35:06 +0400 | From | Cyrill Gorcunov <> | Subject | Re: [PATCH] mm: /prom/pid/clear_refs: avoid split_huge_page() |
| |
On Thu, May 22, 2014 at 09:32:47AM +0400, Cyrill Gorcunov wrote: > On Thu, May 22, 2014 at 04:11:10AM +0300, Kirill A. Shutemov wrote: > > Andrew Morton wrote: > > > On Wed, 21 May 2014 22:04:22 +0300 "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com> wrote: > > > > > > > Currently we split all THP pages on any clear_refs request. It's not > > > > necessary. We can handle this on PMD level. > > > > > > > > One side effect is that soft dirty will potentially see more dirty > > > > memory, since we will mark whole THP page dirty at once. > > > > > > This clashes pretty badly with > > > http://ozlabs.org/~akpm/mmots/broken-out/clear_refs-redefine-callback-functions-for-page-table-walker.patch > > > > Hm.. For some reason CRIU memory-snapshotting test cases fail on current > > linux-next. I didn't debug why. Mainline works. Folks? > > Thanks for noticing, Kirill! I don't test linux-test regulary will try and > report the results.
OK, I managed to run criu on linux-next. Due to changes in vdso it no longer able to run. I'll handle it in criu and ping you then.
| |