Although it's unlikely to be an issue, technically we shouldn't be insisting the sha1 field be unique globally, just within each layerbranch, so adjust the constraints. Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
