Re: [PATCH 01 of 12] Core of mmu notifiers

From: Andrea Arcangeli
Date: Wed Apr 23 2008 - 13:46:08 EST


On Wed, Apr 23, 2008 at 12:09:09PM -0500, Jack Steiner wrote:
>
> You may have spotted this already. If so, just ignore this.
>
> It looks like there is a bug in copy_page_range() around line 667.
> It's possible to do a mmu_notifier_invalidate_range_start(), then
> return -ENOMEM w/o doing a corresponding mmu_notifier_invalidate_range_end().

No I didn't spot it yet, great catch!! ;) Thanks a lot. I think we can
take example by Jack and use our energy to spot any bug in the
mmu-notifier-core like with his above auditing effort (I'm quite
certain you didn't reprouce this with real oom ;) so we get a rock
solid mmu-notifier implementation in 2.6.26 so XPMEM will also benefit
later in 2.6.27 and I hope the last XPMEM internal bugs will also be
fixed by that time.

(for the not going to become mmu-notifier users, nothing to worry
about for you, unless you used KVM or GRU actively with mmu-notifiers
this bug would be entirely harmless with both MMU_NOTIFIER=n and =y,
as previously guaranteed)

Here the still untested fix for review.

diff --git a/mm/memory.c b/mm/memory.c
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -597,6 +597,7 @@
unsigned long next;
unsigned long addr = vma->vm_start;
unsigned long end = vma->vm_end;
+ int ret;

/*
* Don't copy ptes where a page fault will fill them correctly.
@@ -604,33 +605,39 @@
* readonly mappings. The tradeoff is that copy_page_range is more
* efficient than faulting.
*/
+ ret = 0;
if (!(vma->vm_flags & (VM_HUGETLB|VM_NONLINEAR|VM_PFNMAP|VM_INSERTPAGE))) {
if (!vma->anon_vma)
- return 0;
+ goto out;
}

- if (is_vm_hugetlb_page(vma))
- return copy_hugetlb_page_range(dst_mm, src_mm, vma);
+ if (unlikely(is_vm_hugetlb_page(vma))) {
+ ret = copy_hugetlb_page_range(dst_mm, src_mm, vma);
+ goto out;
+ }

if (is_cow_mapping(vma->vm_flags))
mmu_notifier_invalidate_range_start(src_mm, addr, end);

+ ret = 0;
dst_pgd = pgd_offset(dst_mm, addr);
src_pgd = pgd_offset(src_mm, addr);
do {
next = pgd_addr_end(addr, end);
if (pgd_none_or_clear_bad(src_pgd))
continue;
- if (copy_pud_range(dst_mm, src_mm, dst_pgd, src_pgd,
- vma, addr, next))
- return -ENOMEM;
+ if (unlikely(copy_pud_range(dst_mm, src_mm, dst_pgd, src_pgd,
+ vma, addr, next))) {
+ ret = -ENOMEM;
+ break;
+ }
} while (dst_pgd++, src_pgd++, addr = next, addr != end);

if (is_cow_mapping(vma->vm_flags))
mmu_notifier_invalidate_range_end(src_mm,
- vma->vm_start, end);
-
- return 0;
+ vma->vm_start, end);
+out:
+ return ret;
}

static unsigned long zap_pte_range(struct mmu_gather *tlb,
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/