Fix rcv-timeout issue because of Nread timeout

[davidBar-On]

• Version of iperf3 (or development branch, such as master or
  3.1-STABLE) to which this pull request applies:
  master

• Issues fixed (if any): None

• Brief description of code changes (suitable for use as a commit message):

Fix an issue with --rcv-timeout - did not work with value over 10 seconds (my first PR for a problem I found at work as iperf3 user 😄). This is because Nread() timeout after 10 seconds, but iperf_recv_mt() collected statistics even when receiving 0 bytes. Therefore, it was assumed that data blocks where received and so receive did not timeout.

[davidBar-On]

@bmah888, as this is a bug fix (although not critical), just making sure it was not overlooked.

[bmah888]

Thanks for the PR!

Documenting how I reproduced this bug. With a built from tip of the master branch:

server% iperf3 --server --rcv-timeout=2000
client% iperf3 --client server.example.com --time=60

Suspend the client process after a few seconds, observe the timeout on the server side after 2 seconds (approximately).

server% iperf3 --server --rcv-timeout=15000
client% iperf3 --client server.example.com --time=60

Suspend the client process after a few seconds, the server will never timeout and also will never terminate the test.

[bmah888]

With this PR, testing the above scenario results in the server terminating successfully.

Also tested this scenario:

server% iperf3 --server
client% iperf3 --client server.example.com --time=60 --rcv-timeout=15000

After a few seconds, suspend the server side process. Without this PR, the client will not honor the --rcv-timeout parameter, and it runs to the end of the --time parameter and hangs.

With the PR, the --rcv-timeout parameter works correctly.

This looks good for a merge.