On Wed, Jul 05, 2023 at 04:58:47PM +0200, Eelco Chaudron wrote:
>
>
> On 30 Jun 2023, at 21:05, Eric Garver wrote:
> Hi Eric,
>
> I started reviewing the series, and this test was failing every other run for
> me on ‘check-system-userspace’. I ended up making the following additional
> change:
>
> diff --git a/tests/ofproto-macros.at b/tests/ofproto-macros.at
> index d2e6ac768..573ecdd0f 100644
> --- a/tests/ofproto-macros.at
> +++ b/tests/ofproto-macros.at
> @@ -120,7 +120,7 @@ strip_xids () {
>
> # Changes all 'used:...' to say 'used:0.0', to make output easier to compare.
> strip_used () {
> - sed 's/used:[[0-9]]\.[[0-9]]*/used:0.0/'
> + sed 's/used:\([[0-9]]\.[[0-9]]*s\|never\)/used:0.0s/'
> }
In this test case I expect the flow to be used. My local test runs do
not yield "never".
Maybe we need a "ovs-appctl time/warp 5000" before dumping the flow?
This is used in tests/drop-stats.at.
>
> This is also the reason why the intel tests are failing in patchwork;
> https://mail.openvswitch.org/pipermail/ovs-build/2023-June/032011.html
>
> I still need to review the other patches in the series, but some small
> comments while testing the patch set:
>
> We are trying to align all commit messages to start with a Capital and end
> with a dot. So in your case:
>
> ‘tests: system-traffic: Add coverage for drop action.’
ACK. I'll address this in v3.
_______________________________________________
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
