mm/memory.c: fix mem_cgroup_oom_disable() call missing
authorLaurent Dufour <ldufour@linux.vnet.ibm.com>
Fri, 8 Sep 2017 23:13:12 +0000 (16:13 -0700)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 13 Sep 2017 21:13:36 +0000 (14:13 -0700)
commit de0c799bba2610a8e1e9a50d76a28614520a4cd4 upstream.

Seen while reading the code, in handle_mm_fault(), in the case
arch_vma_access_permitted() is failing the call to
mem_cgroup_oom_disable() is not made.

To fix that, move the call to mem_cgroup_oom_enable() after calling
arch_vma_access_permitted() as it should not have entered the memcg OOM.

Link: http://lkml.kernel.org/r/1504625439-31313-1-git-send-email-ldufour@linux.vnet.ibm.com
Fixes: bae473a423f6 ("mm: introduce fault_env")
Signed-off-by: Laurent Dufour <ldufour@linux.vnet.ibm.com>
Acked-by: Kirill A. Shutemov <kirill@shutemov.name>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
mm/memory.c

index d064caff9d7dfac2f585086a4d032a7d17d86505..1aa63e7dd790808003b2c40785fb44089f834c97 100644 (file)
@@ -3596,6 +3596,11 @@ int handle_mm_fault(struct vm_area_struct *vma, unsigned long address,
        /* do counter updates before entering really critical section. */
        check_sync_rss_stat(current);
 
+       if (!arch_vma_access_permitted(vma, flags & FAULT_FLAG_WRITE,
+                                           flags & FAULT_FLAG_INSTRUCTION,
+                                           flags & FAULT_FLAG_REMOTE))
+               return VM_FAULT_SIGSEGV;
+
        /*
         * Enable the memcg OOM handling for faults triggered in user
         * space.  Kernel faults are handled more gracefully.
@@ -3603,11 +3608,6 @@ int handle_mm_fault(struct vm_area_struct *vma, unsigned long address,
        if (flags & FAULT_FLAG_USER)
                mem_cgroup_oom_enable();
 
-       if (!arch_vma_access_permitted(vma, flags & FAULT_FLAG_WRITE,
-                                           flags & FAULT_FLAG_INSTRUCTION,
-                                           flags & FAULT_FLAG_REMOTE))
-               return VM_FAULT_SIGSEGV;
-
        if (unlikely(is_vm_hugetlb_page(vma)))
                ret = hugetlb_fault(vma->vm_mm, vma, address, flags);
        else