On Wed, Nov 14, 2012 at 02:33:44PM -0800, David Rientjes wrote:
> On Wed, 7 Nov 2012, Kirill A. Shutemov wrote:
>
> > diff --git a/mm/huge_memory.c b/mm/huge_memory.c
> > index ff834ea..0d903bf 100644
> > --- a/mm/huge_memory.c
> > +++ b/mm/huge_memory.c
> > @@ -701,6 +701,18 @@ static inline struct page *alloc_hugepage(int defrag)
> > }
> > #endif
> >
> > +static void set_huge_zero_page(pgtable_t pgtable, struct mm_struct *mm,
> > + struct vm_area_struct *vma, unsigned long haddr, pmd_t *pmd)
> > +{
> > + pmd_t entry;
> > + entry = pfn_pmd(huge_zero_pfn, vma->vm_page_prot);
> > + entry = pmd_wrprotect(entry);
> > + entry = pmd_mkhuge(entry);
> > + set_pmd_at(mm, haddr, pmd, entry);
> > + pgtable_trans_huge_deposit(mm, pgtable);
> > + mm->nr_ptes++;
> > +}
> > +
> > int do_huge_pmd_anonymous_page(struct mm_struct *mm, struct vm_area_struct *vma,
> > unsigned long address, pmd_t *pmd,
> > unsigned int flags)
> > @@ -778,6 +790,11 @@ int copy_huge_pmd(struct mm_struct *dst_mm, struct mm_struct *src_mm,
> > pte_free(dst_mm, pgtable);
> > goto out_unlock;
> > }
> > + if (is_huge_zero_pmd(pmd)) {
> > + set_huge_zero_page(pgtable, dst_mm, vma, addr, dst_pmd);
> > + ret = 0;
> > + goto out_unlock;
> > + }
>
> You said in the introduction message in this series that you still allow
> splitting of the pmd, so why no check for pmd_trans_splitting() before
> this?
pmd_trans_splitting() returns true only for pmd which points to a page
under spliiting. It never happens with huge zero page.
We only split a pmd to a page table without touching the page.
mm->page_table_lock is enough to protect against that.
--
Kirill A. Shutemov
Attachment:
signature.asc
Description: Digital signature