On Thu, Mar 15, 2018 at 11:04:53PM +0000, Andrew Morton wrote: > On Thu, 15 Mar 2018 18:23:53 +0300 "Kirill A. Shutemov" > <[email protected]> wrote: > > > khugepaged is not yet able to convert PTE-mapped huge pages back to PMD > > mapped. We do not collapse such pages. See check khugepaged_scan_pmd(). > > > > But if between khugepaged_scan_pmd() and __collapse_huge_page_isolate() > > somebody managed to instantiate THP in the range and then split the PMD > > back to PTEs we would have a problem -- VM_BUG_ON_PAGE(PageCompound(page)) > > will get triggered. > > > > It's possible since we drop mmap_sem during collapse to re-take for > > write. > > > > Replace the VM_BUG_ON() with graceful collapse fail. > > > > Signed-off-by: Kirill A. Shutemov <[email protected]> > > Fixes: b1caa957ae6d ("khugepaged: ignore pmd tables with THP mapped with > > ptes") > > Jan 2016. Do we need a cc:stable?
Yes, please. I forgot to put it. -- Kirill A. Shutemov
