On Mon, Dec 17, 2012 at 10:49:00AM +0800, Amos Kong wrote: > Hi all, > > I have total four patches for updating netperf to 2.6.0 > > Two patches for Client-tests: > [Virt-test-devel] [Client-Tests PATCH 0/2] update netperf to 2.6.0 > ├─>[Client-Tests PATCH 1/2] Update netperf to 2.6.0 > ├─>[Client-Tests PATCH 2/2] replace netperf tar package > > Two patches for virt-tests: > ├─>[PATCH] netperf: update config and path > └─>[PATCH] netperf: correct send size parameters for TCP_MAERTS > > > The issue wenli found was already fixed in netperf upstream. > We have '-C -c' parameters in current autotest code, and I also posted > a patch (last one) to correct send size parameters. So it's ok to use > 2.6.0
Sorry, we have '-C -c' parameters in current autotest code', so the output of send-size (TCP_MAERTS) will be wrong. We can fix it by remove '-C -c' parameter, or add a patch as wait_for_read.patch, it's not necessary, because the send-size isn't used in parsing result. We can ignore it. What I need to do next is updating Windows guest's netperf to 2.6.0, or still using 2.4.5 for Windows guest. I will update those four patches when I finished it. > parameter issue: > http://www.netperf.org/pipermail/netperf-dev/2012-December/000870.html > man page: > http://www.netperf.org/pipermail/netperf-dev/2012-December/000871.html > > Do you have other issue with those (4) update patches? > -- Amos. _______________________________________________ Autotest-kernel mailing list [email protected] https://www.redhat.com/mailman/listinfo/autotest-kernel
