mm/mmap: Fix VM_LOCKED check in do_vmi_align_munmap()

Origin: https://git.kernel.org/pub/scm/linux/kernel/git/stable/stable-queue.git/tree/queue-6.1/mm-mmap-fix-vm_locked-check-in-do_vmi_align_munmap.patch

6.1 backport of the patch [1] uses 'next' vma instead of 'split' vma.
Fix the mistake.

[1] commit 606c812eb1d5 ("mm/mmap: Fix error path in do_vmi_align_munmap()")

Fixes: a149174ff8bb ("mm/mmap: Fix error path in do_vmi_align_munmap()")
Signed-off-by: Suren Baghdasaryan <surenb@google.com>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Gbp-Pq: Topic bugfix/all
Gbp-Pq: Name mm-mmap-fix-vm_locked-check-in-do_vmi_align_munmap.patch

diff --git a/mm/mmap.c b/mm/mmap.c
index b8af52db3bbe045e940f9a8f63f17f3860c7e74e..1597a96b667fd33fd73930418f776867173f32cb 100644 (file)
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -2484,7 +2484,7 @@ do_mas_align_munmap(struct ma_state *mas, struct vm_area_struct *vma,
                        error = mas_store_gfp(&mas_detach, split, GFP_KERNEL);
                        if (error)
                                goto munmap_gather_failed;
-                       if (next->vm_flags & VM_LOCKED)
+                       if (split->vm_flags & VM_LOCKED)
                                locked_vm += vma_pages(split);
 
                        count++;