Dave McCracken wrote on Tuesday, August 30, 2005 3:13 PM
> This patch implements page table sharing for all shared memory regions that
> span an entire page table page. It supports sharing at multiple page
> levels, depending on the architecture.
In function pt_share_pte():
> + while ((svma = next_shareable_vma(vma, svma, &iter))) {
> + spgd = pgd_offset(svma->vm_mm, address);
> + if (pgd_none(*spgd))
> + continue;
> +
> + spud = pud_offset(spgd, address);
> + if (pud_none(*spud))
> + continue;
> +
> + spmd = pmd_offset(spud, address);
> + if (pmd_none(*spmd))
> + continue;
....
> + page = pmd_page(*spmd);
> + pt_increment_share(page);
> + pmd_populate(vma->vm_mm, pmd, page);
> + }
Do you really have to iterate through all the vma? Can't you just break
out of the while loop on first successful match and populating the pmd?
I would think you will find them to be the same pte page. Or did I miss
some thing?
--- ./mm/ptshare.c.orig 2005-09-01 21:16:35.311915518 -0700
+++ ./mm/ptshare.c 2005-09-01 21:18:24.629296992 -0700
@@ -200,6 +200,7 @@ pt_share_pte(struct vm_area_struct *vma,
page = pmd_page(*spmd);
pt_increment_share(page);
pmd_populate(vma->vm_mm, pmd, page);
+ break;
}
}
pte = pte_alloc_map(vma->vm_mm, pmd, address);