On Tue, 8 Jan 2008, John Heffner wrote:

Andi Kleen wrote:

quoted

David Miller [off-list ref] writes:

quoted

The big problem is that recovery from even a single packet loss in a
window makes us run kfree_skb() for a all the packets in a full
window's worth of data when recovery completes.

Why exactly is it a problem to free them all at once? Are you worried
about kernel preemption latencies?

I also wonder how much of a problem this is (for now, with window sizes of
order 10000 packets.  My understanding is that the biggest problems arise from
O(N^2) time for recovery because every ack was expensive. Have current 
tests shown the final ack to be a major source of problems?

This thread got started because I tried to solve the other latencies but 
realized that it helps very little because this latency spike would 
have remained unsolved and it happens in one of the most common case.

-- 
 i.