a0715e966d
After removing posixpath.normpath(url) in http.py, the code has a regression bug that the url like 'http://example.com:80/test' can not work. The code urlparse.urljoin() can not work as '%s%s' % (self.endpoint_path, url). Fixes bug #1230032 Change-Id: Ie7266fc3a067b92dfeed169086b4bf6a87dedbd6