netdev - Re: [RFC PATCH net-next 1/1] tc-testing: Restore original behaviour for namespaces in tdc

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAMDBHY+DSWg5GbHEHbEh4H96DfxuZ4Vv6Lx=LRHz4jxCH6esRg@mail.gmail.com>
Date:   Fri, 7 Jun 2019 12:09:49 -0400
From:   Lucas Bates <lucasb@...atatu.com>
To:     Davide Caratti <dcaratti@...hat.com>
Cc:     Linux Kernel Network Developers <netdev@...r.kernel.org>,
        Nicolas Dichtel <nicolas.dichtel@...nd.com>,
        David Miller <davem@...emloft.net>,
        Jamal Hadi Salim <jhs@...atatu.com>, kernel@...atatu.com,
        Cong Wang <xiyou.wangcong@...il.com>,
        Jiri Pirko <jiri@...nulli.us>,
        Marcelo Ricardo Leitner <mleitner@...hat.com>,
        Vlad Buslov <vladbu@...lanox.com>,
        Hangbin Liu <haliu@...hat.com>
Subject: Re: [RFC PATCH net-next 1/1] tc-testing: Restore original behaviour
 for namespaces in tdc

On Fri, Jun 7, 2019 at 8:18 AM Davide Caratti <dcaratti@...hat.com> wrote:
> From what I see, it is a fix for the reported problem (e.g. tests failing
> because of 'nsPlugin' uninstalled). And, I want to followup fixing the
> bpf.json in tc-actions, so that
>
> # ./tdc.py  -l -c bpf | grep eBPF
>  e939: (actions, bpf) Add eBPF action with valid object-file
>  282d: (actions, bpf) Add eBPF action with invalid object-file
>
> require the buildebpfPlugin (unless anybody disagrees, I will also revert
> the meaning of '-B' also, like you did for '-n')
Yes, those should definitely be tagged.  I probably should have
included them in the patch.

As for the -B option, that would probably be a good idea.  Required
plugins, I think, should always be on and require a user to explicitly
not use them.

So in that case, nsPlugin and buildepfPlugin, being required, should
have options to explicitly disable their features.  valgrindPlugin,
since it's not explicitly required to run any given test, doesn't need
that.  Maybe we should separate the plugins into different directories
based on this?

>
> few comments after a preliminary test:
> 1) the patch still does not cover the two categories that use $DEV2 (i.e.
> flower and concurrency still fail in my environment)
I will have to try that out and see what is going on.  I didn't try
the $DEV2 tests against this.

> 2) I've been reported, and reproduced with latest fedora, a problem in
> nsPlugin.py. All tests in the 'filter' category still fail, unless I do
>
> # sed -i "s#ip#/sbin/ip#g" nsPlugin.py
>
> otherwise, the 'prepare' stage fails:
>
> # ./tdc.py  -e  5339
>  -- ns/SubPlugin.__init__
> Test 5339: Del entire fw filter
>
> -----> prepare stage *** Could not execute: "$TC qdisc add dev $DEV1 ingress"
>
> -----> prepare stage *** Error message: "/bin/sh: ip: command not found
> "
> returncode 127; expected [0]
>
> -----> prepare stage *** Aborting test run.
>
> (maybe we should use a variable for that, instead of hardcoded command
> name, like we do for $TC ?)

Ooh...  Yes, yes we should.  Sounds like the new version of Fedora
isn't including sbin on the path anymore?  That's going to require
another minor change that I think I'll submit in a separate patch.

Thanks, Davide!