Re: 2.6.38-rc1 problems with khugepaged

[Andrea Arcangeli <aarcange@redhat.com>]

Hello Werner,

this should fix your oops, it's untested still so let me know if you
test it.

It's a noop for x86_64 and it only affected x86 32bit with highpte enabled.

====
Subject: khugepaged: fix pte_unmap for highpte x86_32

From: Andrea Arcangeli <aarcange@redhat.com>

__collapse_huge_page_copy is still dereferencing the pte passed as parameter so
we must pte_unmap after __collapse_huge_page_copy returns, not before.

Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
---

diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index 004c9c2..c4f634b 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -1837,9 +1837,9 @@ static void collapse_huge_page(struct mm_struct *mm,
 	spin_lock(ptl);
 	isolated = __collapse_huge_page_isolate(vma, address, pte);
 	spin_unlock(ptl);
-	pte_unmap(pte);
 
 	if (unlikely(!isolated)) {
+		pte_unmap(pte);
 		spin_lock(&mm->page_table_lock);
 		BUG_ON(!pmd_none(*pmd));
 		set_pmd_at(mm, address, pmd, _pmd);
@@ -1856,6 +1856,7 @@ static void collapse_huge_page(struct mm_struct *mm,
 	anon_vma_unlock(vma->anon_vma);
 
 	__collapse_huge_page_copy(pte, new_page, vma, address, ptl);
+	pte_unmap(pte);
 	__SetPageUptodate(new_page);
 	pgtable = pmd_pgtable(_pmd);
 	VM_BUG_ON(page_count(pgtable) != 1);