On 02/11/2016 06:21 AM, Kirill A. Shutemov wrote: > vma_addjust_trans_huge() splits pmd if it's crossing VMA boundary. > During split we munlock the huge page which requires rmap walk. > rmap wants to take the lock on its own.
Ahhh, ... so we $SUBJECT in order to fix it. Now it all makes sense. Maybe I'm old fashioned, but I tend to have forgotten $SUBJECT by the time I start to read the patch body text. It's really handy for me when the body text stands on its own.
