From: Li Xinhai <lixinhai.lxh@xxxxxxxxx>
mbind() is required to report EFAULT if range, specified by addr and len,
contains unmapped holes. In current implementation, below rules are applied
for this checking:
1 Unmapped holes at any part of the specified range should be reported as
EFAULT if mbind() for none MPOL_DEFAULT cases;
2 Unmapped holes at any part of the specified range should be ignored (do
not reprot EFAULT) if mbind() for MPOL_DEFAULT case;
3 The whole range in an unmapped hole should be reported as EFAULT;
Note that rule 2 does not fullfill the mbind() API definition, but since
that behavior has existed for long days (refer the usage of internal flag
MPOL_MF_DISCONTIG_OK), this patch does not plan to change it.
Cases do not follow those rules and been fixed by this patch are:
case_1: unmapped hole at tail side of the sepcified range when mbind() for
non MPOL_DEFAULT cases, EFAULT is not reported (conflicts rule 1).
[ hole ][ vma ][ hole ]
[ range ]
case_2: unmapped hole at head side of the specified range when mbind() for
MPOL_DEFAULT case, EFAULT is reported (conflicts rule 2).
[ hole ][ vma ][ hole ]
[ range ]
Fixes: 9d8cebd4bcd7 ("mm: fix mbind vma merge problem")
Fixes: 6f4576e3687b ("mempolicy: apply page table walker on queue_pages_range()")
Fixes: 48684a65b4e3 ("mm: pagewalk: fix misbehavior of walk_page_range for vma(VM_PFNMAP)")
Signed-off-by: Li Xinhai <lixinhai.lxh@xxxxxxxxx>
Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
Cc: Naoya Horiguchi<n-horiguchi@xxxxxxxxxxxxx>
Cc: Michal Hocko <mhocko@xxxxxxxx>
Cc: Vlastimil Babka <vbabka@xxxxxxx>
Cc: linux-man <linux-man@xxxxxxxxxxxxxxx>
---
Changes v3->v4:
- fix my email address;
Changes v2->v3:
- Add more details in change log;
- Check holes in .test_walk() and after call walk_page_range();
mm/mempolicy.c | 52 ++++++++++++++++++++++++++++++++++++----------------
1 file changed, 36 insertions(+), 16 deletions(-)
diff --git a/mm/mempolicy.c b/mm/mempolicy.c
index 4ae967b..b2e10bf 100644
--- a/mm/mempolicy.c
+++ b/mm/mempolicy.c
@@ -410,7 +410,9 @@ struct queue_pages {
struct list_head *pagelist;
unsigned long flags;
nodemask_t *nmask;
- struct vm_area_struct *prev;
+ unsigned long start;
+ struct vm_area_struct *first;
+ struct vm_area_struct *last;
};
/*
@@ -618,6 +620,21 @@ static int queue_pages_test_walk(unsigned long start, unsigned long end,
unsigned long endvma = vma->vm_end;
unsigned long flags = qp->flags;
+ /* range check first */
+ VM_BUG_ON((vma->vm_start > start) || (vma->vm_end < end));
+
+ if (!qp->first) {
+ qp->first = vma;
+ if (!(flags & MPOL_MF_DISCONTIG_OK) &&
+ (qp->start < vma->vm_start))
+ /* hole at head side of range */
+ return -EFAULT;
+ } else if (!(flags & MPOL_MF_DISCONTIG_OK) &&
+ (vma->vm_prev->vm_end < vma->vm_start))
+ /* hole at middle of range */
+ return -EFAULT;
+ qp->last = vma;
+
/*
* Need check MPOL_MF_STRICT to return -EIO if possible
* regardless of vma_migratable
@@ -628,17 +645,6 @@ static int queue_pages_test_walk(unsigned long start, unsigned long end,
if (endvma > end)
endvma = end;
- if (vma->vm_start > start)
- start = vma->vm_start;
-
- if (!(flags & MPOL_MF_DISCONTIG_OK)) {
- if (!vma->vm_next && vma->vm_end < end)
- return -EFAULT;
- if (qp->prev && qp->prev->vm_end < vma->vm_start)
- return -EFAULT;
- }
-
- qp->prev = vma;
if (flags & MPOL_MF_LAZY) {
/* Similar to task_numa_work, skip inaccessible VMAs */
@@ -679,14 +685,29 @@ static int queue_pages_test_walk(unsigned long start, unsigned long end,
nodemask_t *nodes, unsigned long flags,
struct list_head *pagelist)
{
+ int err;
struct queue_pages qp = {
.pagelist = pagelist,
.flags = flags,
.nmask = nodes,
- .prev = NULL,
+ .start = start,
+ .first = NULL,
+ .last = NULL,
};
- return walk_page_range(mm, start, end, &queue_pages_walk_ops, &qp);
+ err = walk_page_range(mm, start, end, &queue_pages_walk_ops, &qp);
+
+ if (err != -EFAULT) {
+ if (!qp.first)
+ /* whole range in hole */
+ err = -EFAULT;
+ else if (!(flags & MPOL_MF_DISCONTIG_OK) &&
+ (qp.last->vm_end < end))
+ /* hole at tail side of range */
+ err = -EFAULT;
+ }
+
+ return err;
}
/*
@@ -738,8 +759,7 @@ static int mbind_range(struct mm_struct *mm, unsigned long start,
unsigned long vmend;
vma = find_vma(mm, start);
- if (!vma || vma->vm_start > start)
- return -EFAULT;
+ VM_BUG_ON(!vma);
prev = vma->vm_prev;
if (start > vma->vm_start)
--
1.8.3.1
