Re: [Intel-wired-lan] [PATCH net-next v1 4/7] selftests: net: add a test for shared UDP tunnel info tables

[Jakub Kicinski <kuba@kernel.org>]

On Sat, 19 Sep 2020 07:23:58 +0000 Brown, Aaron F wrote:
> > From: Intel-wired-lan <intel-wired-lan-bounces@osuosl.org> On Behalf Of Jakub
> > Kicinski
> > Sent: Tuesday, July 21, 2020 6:27 PM
> > To: davem@davemloft.net
> > Cc: netdev@vger.kernel.org; intel-wired-lan@lists.osuosl.org; Jakub Kicinski
> > <kuba@kernel.org>
> > Subject: [Intel-wired-lan] [PATCH net-next v1 4/7] selftests: net: add a test for
> > shared UDP tunnel info tables
> > 
> > Add a test run of checks validating the shared UDP tunnel port
> > tables function as we expect.
> > 
> > Signed-off-by: Jakub Kicinski <kuba@kernel.org>
> > ---
> >  .../drivers/net/netdevsim/udp_tunnel_nic.sh   | 109 ++++++++++++++++++
> >  1 file changed, 109 insertions(+)
> >   
> I ran into two things while running this script.
> 1. The script as it exists in the git tree (Jeff Kirshers next-queue) is not executable.  I don't know if that's a patch issue or translation into the tree.  Easy enough to get around, but should probably be executable to start.

Ah, good catch, thanks! Please adjust in your tree or I can send a
follow up with other patches I still have queued.

> 2. The script runs into a handful of errors,7 to be precise.  I'm not sure if they are real failures, incorrect expectations or maybe something in my kernel .config (I have been using a minimal .config and enabling modules as needed.)

Can you share the .config? Do you have tunnels enabled? 
I just retested and it passes :(

> The output I get from it is:
> ----------------------------------------------------------------------------------------------------
> u1518:[0]/usr/src/kernels/next-queue> cat ~/udp_tunnel-sh-outut.txt                                                                     
> ERROR: table 0 on port 1: basic - VxLAN v4 devices
>        check_table: wrong entry 0
>        expected: port: 4789     type: 1
>        have:     port: 0        type: 0
> ERROR: table 0 on port 1: basic - VxLAN v4 devices
>        check_table: wrong entry 0
>        expected: port: 4789     type: 1
>        have:     port: 0        type: 0
> ERROR: table 0 on port 1: basic - VxLAN v6 devices
>        check_table: wrong entry 0
>        expected: port: 4789     type: 1
>        have:     port: 0        type: 0
> ERROR: table 0 on port 1: basic - VxLAN v6 devices
>        check_table: wrong entry 0
>        expected: port: 4789     type: 1
>        have:     port: 0        type: 0
> ERROR: table 0 on port 1: basic - another VxLAN v6 devices
>        check_table: wrong entry 0
>        expected: port: 4789     type: 1
>        have:     port: 0        type: 0
> ERROR: table 0 on port 1: basic - Geneve device
>        check_table: wrong entry 0
>        expected: port: 4789     type: 1
>        have:     port: 0        type: 0
> ERROR: table 1 on port 1: basic - Geneve device
>        check_table: wrong entry 0
>        expected: port: 6081     type: 2
>        have:     port: 0        type: 0
> FAILED 7/435 checks
> u1518:[0]/usr/src/kernels/next-queue>
> ----------------------------------------------------------------------------------------------------
> The script sends messages to dmesg, most look to be informative "set" and "unset" messages, but I do get a handful of failed messages.  The dmesg queue was cleared before the run so only contains the udp_tunnel-sh messages:
> ----------------------------------------------------------------------------------------------------
> u1518:[0]/usr/src/kernels/next-queue> dmesg|grep -i fail
> [ 8909.179462] netdevsim netdevsim386 eth4: UDP tunnel port sync failed port 10000 type vxlan: -110
> [ 8909.328763] netdevsim netdevsim386 eth4: UDP tunnel port sync failed port 20000 type geneve: -2
> [ 8909.444028] netdevsim netdevsim386 eth4: UDP tunnel port sync failed port 10000 type vxlan: -110
> [ 8909.592049] netdevsim netdevsim386 eth4: UDP tunnel port sync failed port 20000 type geneve: -2
> u1518:[0]/usr/src/kernels/next-queue>

These are error injection, the test does. They are expected.