Turning on PROXY Protocol forces clients to perform ALPN, otherwise connections are dropped

ATS: 8.0.5
OpenSSL: 1.0.2r

```
CONFIG proxy.config.http.server_ports STRING 8080:pp:proto=http 8443:ssl:pp:proto=http
```

Running Nginx on localhost, proxying TCP connections on 80/443 to 8080/8443 with PROXY protocol:
```
stream {
upstream internal {
server 127.0.0.1:8443;
}
server {
listen 443;
proxy_protocol on;
proxy_pass internal;
}
}
```

Working via cURL:
```bash
$ curl https://localhost -I
HTTP/1.1 200 OK
Server: ATS/8.0.5
Date: Tue, 01 Oct 2019 00:09:30 GMT
Content-Type: text/html
Age: 0
Connection: keep-alive
```

Not working via cURL with ALPN disabled:
```bash
$ curl https://localhost -I --no-alpn
curl: (52) Empty reply from server
```

Not working via Python `requests`:
```python
import requests
requests.get('https://localhost')
#
# urllib3.exceptions.ProtocolError: ('Connection aborted.', RemoteDisconnected('Remote end closed connection without response'))
```

Working via Python `requests` after forcing ALPN:
```python
from requests import Session
from requests.adapters import HTTPAdapter
from requests.packages.urllib3.util.ssl_ import create_urllib3_context

class HTTP11ALPNAdapter(HTTPAdapter):
def init_poolmanager(self, *args, **kwargs):
ctx = create_urllib3_context()
ctx.set_alpn_protocols(['http/1.1'])
kwargs['ssl_context'] = ctx
return super().init_poolmanager(*args, **kwargs)

s = requests.Session()
s.mount('https://localhost', HTTP11ALPNAdapter())
s.get('https://localhost')
#
```

@masaori335 was able to reproduce this and track down that it was the `pp` in the `server_ports` config line that was triggering this behavior. When `pp` was not specified, clients received successful responses even when not performing ALPN