On Thu, 1 Mar 2007, Ilpo Järvinen wrote: > Previously the prior_ssthresh was only used in the undo functions > that were called by fastretrans_alert. FRTO processing, however, > happens before that and one of the responses is doing undo too. > > I think that after this patch, FRTO should be ECN-safe; only the > undo_cwr response is a suspect anyway. The possible cases: > [...snip...] > > This problem could be solved locally in the undo_cwr response of > FRTO by falling back to ratehalving response when FLAG_ECE is > enabled.
I think that doing it in the response is better that this approach, since it knows that the ssthresh has been halved already within that round-trip, so there is no need to do that again... I'll submit the patch tomorrow... With this prior_ssthresh clearing move alone, the ssthresh ends up being halved twice if I tought it right (first in tcp_enter_frto and then again in tcp_enter_cwr that is called from fastretrans_alert)... So please, drop this patch. -- i.
