linux-kernel - Re: [PATCH v3 net-next 09/10] selftests/tc-testing: test that taprio can only be attached as root

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: <87il9w0xx7.fsf@intel.com>
Date:   Thu, 03 Aug 2023 11:43:16 -0700
From:   Vinicius Costa Gomes <vinicius.gomes@...el.com>
To:     Vladimir Oltean <vladimir.oltean@....com>
Cc:     netdev@...r.kernel.org, "David S. Miller" <davem@...emloft.net>,
        Eric Dumazet <edumazet@...gle.com>,
        Jakub Kicinski <kuba@...nel.org>,
        Paolo Abeni <pabeni@...hat.com>,
        Jamal Hadi Salim <jhs@...atatu.com>,
        Cong Wang <xiyou.wangcong@...il.com>,
        Jiri Pirko <jiri@...nulli.us>, linux-kernel@...r.kernel.org,
        intel-wired-lan@...ts.osuosl.org,
        Muhammad Husaini Zulkifli <muhammad.husaini.zulkifli@...el.com>,
        Peilin Ye <yepeilin.cs@...il.com>,
        Pedro Tammela <pctammela@...atatu.com>,
        Richard Cochran <richardcochran@...il.com>,
        Zhengchao Shao <shaozhengchao@...wei.com>,
        Maxim Georgiev <glipus@...il.com>
Subject: Re: [PATCH v3 net-next 09/10] selftests/tc-testing: test that
 taprio can only be attached as root

Vladimir Oltean <vladimir.oltean@....com> writes:

> Hi Vinicius,
>
> On Wed, Aug 02, 2023 at 04:29:55PM -0700, Vinicius Costa Gomes wrote:
>> Vladimir Oltean <vladimir.oltean@....com> writes:
>> This test is somehow flaky (all others are fine), 1 in ~4 times, it fails.
>> 
>> Taking a look at the test I couldn't quickly find out the reason for the
>> flakyness.
>> 
>> Here's the verbose output of one of the failures:
>> 
>> vcgomes@...-cfl-clr-30 ~/src/net-next/tools/testing/selftests/tc-testing $ sudo ./tdc.py -e 39b4 -v
>> All test results:
>> 
>> 1..1
>> not ok 1 39b4 - Reject grafting taprio as child qdisc of software taprio
>> 	Could not match regex pattern. Verify command output:
>> parse error: Objects must consist of key:value pairs at line 1, column 334
>
> Interesting. I'm not seeing this, and I re-ran it a few times. The error
> message seems to come from jq, as if it's not able to parse something.
>
> Sorry, I only have caveman debugging techniques. Could you remove the
> pipe into jq and rerun a few times, see what it prints when it fails?
>
> diff --git a/tools/testing/selftests/tc-testing/tc-tests/qdiscs/taprio.json b/tools/testing/selftests/tc-testing/tc-tests/qdiscs/taprio.json
> index de51408544e2..bb6be1f78e31 100644
> --- a/tools/testing/selftests/tc-testing/tc-tests/qdiscs/taprio.json
> +++ b/tools/testing/selftests/tc-testing/tc-tests/qdiscs/taprio.json
> @@ -148,8 +148,8 @@
>          ],
>          "cmdUnderTest": "$TC qdisc replace dev $ETH parent 8001:7 taprio num_tc 8 map 0 1 2 3 4 5 6 7 queues 1@0 1@1 1@2 1@3 1@4 1@5 1@6 1@7 base-time 200 sched-entry S ff 20000000 clockid CLOCK_TAI",
>          "expExitCode": "2",
> -        "verifyCmd": "$TC -j qdisc show dev $ETH root | jq '.[].options.base_time'",
> -        "matchPattern": "0",
> +        "verifyCmd": "$TC -j qdisc show dev $ETH root",
> +        "matchPattern": "\\[{\"kind\":\"taprio\",\"handle\":\"8001:\",\"root\":true,\"refcnt\":9,\"options\":{\"tc\":0,\"map\":\\[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0\\],\"queues\":\\[\\],\"clockid\":\"TAI\",\"base_time\":0,\"cycle_time\":20000000,\"cycle_time_extension\":0,\"schedule\":\\[{\"index\":0,\"cmd\":\"S\",\"gatemask\":\"0xff\",\"interval\":20000000}\\],\"max-sdu\":\\[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0\\],\"fp\":\\[\"E\",\"E\",\"E\",\"E\",\"E\",\"E\",\"E\",\"E\",\"E\",\"E\",\"E\",\"E\",\"E\",\"E\",\"E\",\"E\"\\]}}\\]",
>          "matchCount": "1",
>          "teardown": [
>              "$TC qdisc del dev $ETH root",

Hmmm, I think that this test discovered another bug (perhaps even two).
When it fails here's the json I get (edited for clarity):

[{"kind":"taprio","handle":"8001:","root":true,"refcnt":9,
  "options":{
        "tc":0,
        "map":[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],
        "queues":[],
        "clockid":"TAI",
        "base_time":0,
        "cycle_time":0,
        "cycle_time_extension":0,
        {
                "base_time":0,
                "cycle_time":20000000,
                "cycle_time_extension":0,
                "schedule":[{"index":0,"cmd":"S","gatemask":"0xff","interval":20000000}]
        }}}]

Thinking out loud: If I am reading this right, there's no "oper"
schedule, only an "admin" schedule. So the first bug is probably a
taprio bug when deciding if it should create an "open" vs. "admin"
schedule.

The second bug seems to be in the way that q_taprio in iproute2
handles the admin schedule, is just an object inside another, which
seems to be invalid.

Does it make sense?


Cheers,
-- 
Vinicius