From: "Chen, Kenneth W" <kenneth.w.chen@intel.com>
x86-64 may have to allocate a bunch of upper levels of pagetables, and those
allocations may fail. When they do, unmap_hugepage_range() needs to be able
to clean up after them.
Acked-by: William Lee Irwin III <wli@holomorphy.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
---
25-akpm/arch/i386/mm/hugetlbpage.c | 7 +++++--
1 files changed, 5 insertions(+), 2 deletions(-)
diff -puN arch/i386/mm/hugetlbpage.c~x86_64-hugetlb-fix arch/i386/mm/hugetlbpage.c
--- 25/arch/i386/mm/hugetlbpage.c~x86_64-hugetlb-fix 2005-02-28 16:41:32.000000000 -0800
+++ 25-akpm/arch/i386/mm/hugetlbpage.c 2005-02-28 16:41:32.000000000 -0800
@@ -209,14 +209,17 @@ void unmap_hugepage_range(struct vm_area
{
struct mm_struct *mm = vma->vm_mm;
unsigned long address;
- pte_t pte;
+ pte_t pte, *ptep;
struct page *page;
BUG_ON(start & (HPAGE_SIZE - 1));
BUG_ON(end & (HPAGE_SIZE - 1));
for (address = start; address < end; address += HPAGE_SIZE) {
- pte = ptep_get_and_clear(huge_pte_offset(mm, address));
+ ptep = huge_pte_offset(mm, address);
+ if (!ptep)
+ continue;
+ pte = ptep_get_and_clear(ptep);
if (pte_none(pte))
continue;
page = pte_page(pte);
_