fix-an-endianness-bug-in-ext4_encrypted-zeroout

   1 ext4: fix an endianness bug in ext4_encrypted_zeroout()
   2
   3 From: Al Viro <viro@ZenIV.linux.org.uk>
   4
   5 ex->ee_block is not host-endian (note that accesses of other fields
   6 of *ex right next to that line go through the helpers that do proper
   7 conversion from little-endian to host-endian; it might make sense
   8 to add similar for ->ee_block to avoid reintroducing that kind of
   9 bugs...)
  10
  11 Cc: stable@vger.kernel.org # v4.1+
  12 Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
  13 Signed-off-by: Theodore Ts'o <tytso@mit.edu>
  14 ---
  15
  16 diff --git a/fs/ext4/crypto.c b/fs/ext4/crypto.c
  17 index af06830..1a08350 100644
  18 --- a/fs/ext4/crypto.c
  19 +++ b/fs/ext4/crypto.c
  20 @@ -389,7 +389,7 @@ int ext4_encrypted_zeroout(struct inode *inode, struct ext4_extent *ex)
  21         struct ext4_crypto_ctx  *ctx;
  22         struct page             *ciphertext_page = NULL;
  23         struct bio              *bio;
  24 -       ext4_lblk_t             lblk = ex->ee_block;
  25 +       ext4_lblk_t             lblk = le32_to_cpu(ex->ee_block);
  26         ext4_fsblk_t            pblk = ext4_ext_pblock(ex);
  27         unsigned int            len = ext4_ext_get_actual_len(ex);
  28         int                     ret, err = 0;
  29 --
  30 To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
  31 the body of a message to majordomo@vger.kernel.org
  32 More majordomo info at  http://vger.kernel.org/majordomo-info.html
  33 Please read the FAQ at  http://www.tux.org/lkml/
  34