Re: [GHC] #10149: The argument of mask does not always restore the masking state