net: fix the RTO timer retransmitting skb every 1ms if linear option is enabled

commit e4dd0d3a upstream.

In the real workload, I encountered an issue which could cause the RTO
timer to retransmit the skb per 1ms with linear option enabled. The amount
of lost-retransmitted skbs can go up to 1000+ instantly.

The root cause is that if the icsk_rto happens to be zero in the 6th round
(which is the TCP_THIN_LINEAR_RETRIES value), then it will always be zero
due to the changed calculation method in tcp_retransmit_timer() as follows:

icsk->icsk_rto = min(icsk->icsk_rto << 1, TCP_RTO_MAX);

Above line could be converted to
icsk->icsk_rto = min(0 << 1, TCP_RTO_MAX) = 0

Therefore, the timer expires so quickly without any doubt.

I read through the RFC 6298 and found that the RTO value can be rounded
up to a certain value, in Linux, say TCP_RTO_MIN as default, which is
regarded as the lower bound in this patch as suggested by Eric.

Fixes: 36e31b0a ("net: TCP thin linear time...