[Request]: Support a flag to force tcp DNS

### Feature or enhancement request details

So been seeing where UDP DNS queries are not getting connected back to the bridge network for containers. If I do the same query via TCP using host, nslookup, dig, drill, whatever there are no issues. I am pretty positive it's the device network controls my organization have out there, but can't find the specific one yet. I can just see where a dns response via tcp the next "hop" is mdnsresponder but for udp the flow just ends.

I tried to update resolv.conf with `options use-vc` but it had no effect on the subsequent internal calls. I need to test more, but been doing most tests using musl based images and not glibc which i believe is the issue there.

```
$ host google.com
;; communications error to 192.168.64.1#53: connection refused
;; communications error to 192.168.64.1#53: connection refused
;; no servers could be reached
$ host -T google.com
google.com has address 142.251.179.102
google.com has address 142.251.179.100
google.com has address 142.251.179.113
google.com has address 142.251.179.138
google.com has address 142.251.179.139
google.com has address 142.251.179.101
google.com has IPv6 address 2607:f8b0:4004:c09::71
google.com has IPv6 address 2607:f8b0:4004:c09::64
google.com has IPv6 address 2607:f8b0:4004:c09::8b
google.com has IPv6 address 2607:f8b0:4004:c09::66
google.com mail is handled by 10 smtp.google.com.
google.com has HTTP service bindings 1 . alpn="h2,h3"
```

### Code of Conduct

- [x] I agree to follow this project's Code of Conduct