diff options
| author | Michel Lespinasse <walken@google.com> | 2013-03-28 16:26:23 -0700 |
|---|---|---|
| committer | Linus Torvalds <torvalds@linux-foundation.org> | 2013-03-28 17:45:51 -0700 |
| commit | 09a9f1d27892255cfb9c91203f19476765e2d8d1 (patch) | |
| tree | e6f0985223c87408773c3b81428935d3ce99931b /mm/mmap.c | |
| parent | 0776ce03b1348d39ba3035ea3ee3d268a42912ce (diff) | |
| download | linux-3.10-09a9f1d27892255cfb9c91203f19476765e2d8d1.tar.gz linux-3.10-09a9f1d27892255cfb9c91203f19476765e2d8d1.tar.bz2 linux-3.10-09a9f1d27892255cfb9c91203f19476765e2d8d1.zip | |
Revert "mm: introduce VM_POPULATE flag to better deal with racy userspace programs"
This reverts commit 186930500985 ("mm: introduce VM_POPULATE flag to
better deal with racy userspace programs").
VM_POPULATE only has any effect when userspace plays racy games with
vmas by trying to unmap and remap memory regions that mmap or mlock are
operating on.
Also, the only effect of VM_POPULATE when userspace plays such games is
that it avoids populating new memory regions that get remapped into the
address range that was being operated on by the original mmap or mlock
calls.
Let's remove VM_POPULATE as there isn't any strong argument to mandate a
new vm_flag.
Signed-off-by: Michel Lespinasse <walken@google.com>
Signed-off-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Diffstat (limited to 'mm/mmap.c')
| -rw-r--r-- | mm/mmap.c | 4 |
1 files changed, 3 insertions, 1 deletions
diff --git a/mm/mmap.c b/mm/mmap.c index 2664a47cec9..6466699b16c 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -1306,7 +1306,9 @@ unsigned long do_mmap_pgoff(struct file *file, unsigned long addr, } addr = mmap_region(file, addr, len, vm_flags, pgoff); - if (!IS_ERR_VALUE(addr) && (vm_flags & VM_POPULATE)) + if (!IS_ERR_VALUE(addr) && + ((vm_flags & VM_LOCKED) || + (flags & (MAP_POPULATE | MAP_NONBLOCK)) == MAP_POPULATE)) *populate = len; return addr; } |