The first line of text would be in the canvas subdom, at the same position as the rendered canvas text; the second line of text is rendered by canvas. The user agent would render the cursor/selection over the canvas element.