On Mon, Oct 16, 2017 at 06:01:22PM -0700, Vinicius Costa Gomes wrote: > Hi,
Vinicius; this looks really good to me, testing shows no errors and behavior is as expected. I see some differences in received bw for offload=1 and 0 though Just using tsn_talker and tsn_listener: tc qdisc add dev eth2 parent 8001:1 cbs idleslope 20000 sendslope -980000 \ hicredit 30 locredit -1470 offload 1 with offload=1 (20Mbps BW) Receiving data rate: 19420 kbps Receiving data rate: 19410 kbps Similarly, but with offload=0 (SW, 20Mbps BW) Receiving data rate: 19980 kbps Receiving data rate: 20170 kbps So slightly higer variance in rate when doing SW, which is as expected. I plan to run this through some more tests tomorrow and measure variance on the receiver etc, I'll keep you posted on my findings. Feel free to add a tested-by though Tested-by: Henrik Austad <hen...@austad.us> and a Vehemently-supported-by: Henrik .. :) -Henrik
signature.asc
Description: PGP signature
