Long TCP timeout on connection interuption

When TCP protocol is used to exchange data with a remote peer and the connection with that peer is interrupted (e.g. cable pull somewhere on the connection path), the connection will be terminated with a very long delay of about 20 minutes the stack detects this condition during retransmission of unacknowledged data.

This behavior originates from lwIP stack itself which implements long retries only. To overcome this PSLWIP V4.x had been added an implementation of R1 & R2 timeout defined in RFC1122. Unfortunately, this was not migrated to PSLWIP 5.x.