Is there an unshaped version in there for reference? On Dec 25, 2013 3:02 PM, "Toke Høiland-Jørgensen" wrote: > Flashed 3.10.24-8 and ran a bunch of tests against it on my VDSL line. > All the data is here: > > http://files.toke.dk/bufferbloat/data/cerowrt-3.10.24-8/ > > which documents my tests of several different speed settings for the SQM > scripts. Ended up sticking with 46000/9800 Kbit/s corresponding to these > plots: > > > http://files.toke.dk/bufferbloat/data/cerowrt-3.10.24-8/rrul-2013-12-25T232954.622558.3_10_24_8_46000_9800_ethernet_overhead_0_simplest_.json.gz_all_scaled.svg > > http://files.toke.dk/bufferbloat/data/cerowrt-3.10.24-8/rrul-2013-12-25T232954.622558.3_10_24_8_46000_9800_ethernet_overhead_0_simplest_.json.gz_ping_cdf.svg > > http://files.toke.dk/bufferbloat/data/cerowrt-3.10.24-8/rrul-2013-12-25T232954.622558.3_10_24_8_46000_9800_ethernet_overhead_0_simplest_.json.gz_totals.svg > > I ran some tests over wifi, which mostly just show that my laptop's wifi > chipset sucks. Didn't manage to get any TX dma errors, though. > > > Oh, and I got rid of the fixdaemons script in my own build and haven't > had any trouble getting dnsmasq to come up, nor have I seen other averse > effects. > > Also, I fixed the sysupgrade script and pushed the change to git, so > next time Dave spins a build, it should be included. :) > > -Toke > > _______________________________________________ > Cerowrt-devel mailing list > Cerowrt-devel@lists.bufferbloat.net > https://lists.bufferbloat.net/listinfo/cerowrt-devel > >