Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Received-SPF: Pass (protection.outlook.com: domain of nvidia.com designates
 216.228.117.160 as permitted sender) receiver=protection.outlook.com;
 client-ip=216.228.117.160; helo=mail.nvidia.com; pr=C
References: <759fe934-2e43-e9ff-8946-4fd579c09b05@alu.unizg.hr>
User-agent: mu4e 1.8.11; emacs 28.2
From:   Petr Machata <petrm@nvidia.com>
To:     Mirsad Todorovac <mirsad.todorovac@alu.unizg.hr>
CC:     <netdev@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
        "David S. Miller" <davem@davemloft.net>,
        Eric Dumazet <edumazet@google.com>,
        "Jakub Kicinski" <kuba@kernel.org>,
        Paolo Abeni <pabeni@redhat.com>, Shuah Khan <shuah@kernel.org>,
        <linux-kselftest@vger.kernel.org>, Ido Schimmel <idosch@nvidia.com>
Subject: Re: [PROBLEM] seltests: net/forwarding/sch_ets.sh [HANG]
Date:   Thu, 20 Jul 2023 11:43:45 +0200
In-Reply-To: <759fe934-2e43-e9ff-8946-4fd579c09b05@alu.unizg.hr>
Message-ID: <87cz0m9a3n.fsf@nvidia.com>
MIME-Version: 1.0
Content-Type: text/plain
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 20 Jul 2023 10:00:14.0637
 (UTC)
X-MS-Exchange-CrossTenant-Network-Message-Id: 8291cb39-7641-4ae3-03e7-08db89081d05
X-MS-Exchange-CrossTenant-Id: 43083d15-7273-40c1-b7db-39efd9ccc17a
X-MS-Exchange-CrossTenant-OriginalAttributedTenantConnectingIp: TenantId=43083d15-7273-40c1-b7db-39efd9ccc17a;Ip=[216.228.117.160];Helo=[mail.nvidia.com]
X-MS-Exchange-CrossTenant-AuthSource: BN8NAM11FT074.eop-nam11.prod.protection.outlook.com
X-MS-Exchange-CrossTenant-AuthAs: Anonymous
X-MS-Exchange-CrossTenant-FromEntityHeader: HybridOnPrem
X-MS-Exchange-Transport-CrossTenantHeadersStamped: CH3PR12MB7665
X-Spam-Status: No, score=-1.1 required=5.0 tests=BAYES_00,DKIMWL_WL_HIGH,
        DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,DKIM_VALID_EF,FORGED_SPF_HELO,
        RCVD_IN_DNSWL_BLOCKED,RCVD_IN_MSPIKE_H2,SPF_HELO_PASS,SPF_NONE,
        T_SCC_BODY_TEXT_LINE autolearn=no autolearn_force=no version=3.4.6
X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on
        lindbergh.monkeyblade.net
Precedence: bulk
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org


Mirsad Todorovac <mirsad.todorovac@alu.unizg.hr> writes:

> Using the same config for 6.5-rc2 on Ubuntu 22.04 LTS and 22.10, the execution
> stop at the exact same line on both boxes (os I reckon it is more than an
> accident):
>
> # selftests: net/forwarding: sch_ets.sh
> # TEST: ping vlan 10                                                  [ OK ]
> # TEST: ping vlan 11                                                  [ OK ]
> # TEST: ping vlan 12                                                  [ OK ]
> # Running in priomap mode
> # Testing ets bands 3 strict 3, streams 0 1
> # TEST: band 0                                                        [ OK ]
> # INFO: Expected ratio >95% Measured ratio 100.00
> # TEST: band 1                                                        [ OK ]
> # INFO: Expected ratio <5% Measured ratio 0
> # Testing ets bands 3 strict 3, streams 1 2
> # TEST: band 1                                                        [ OK ]
> # INFO: Expected ratio >95% Measured ratio 100.00
> # TEST: band 2                                                        [ OK ]
> # INFO: Expected ratio <5% Measured ratio 0
> # Testing ets bands 4 strict 1 quanta 5000 2500 1500, streams 0 1
> # TEST: band 0                                                        [ OK ]
> # INFO: Expected ratio >95% Measured ratio 100.00
> # TEST: band 1                                                        [ OK ]
> # INFO: Expected ratio <5% Measured ratio 0
> # Testing ets bands 4 strict 1 quanta 5000 2500 1500, streams 1 2
> # TEST: bands 1:2                                                     [ OK ]
> # INFO: Expected ratio 2.00 Measured ratio 1.99
> # Testing ets bands 3 quanta 3300 3300 3300, streams 0 1 2
> # TEST: bands 0:1                                                     [ OK ]
> # INFO: Expected ratio 1.00 Measured ratio .99
> # TEST: bands 0:2                                                     [ OK ]
> # INFO: Expected ratio 1.00 Measured ratio 1.00
> # Testing ets bands 3 quanta 5000 3500 1500, streams 0 1 2
> # TEST: bands 0:1                                                     [ OK ]
> # INFO: Expected ratio 1.42 Measured ratio 1.42
> # TEST: bands 0:2                                                     [ OK ]
> # INFO: Expected ratio 3.33 Measured ratio 3.33
> # Testing ets bands 3 quanta 5000 8000 1500, streams 0 1 2
> # TEST: bands 0:1                                                     [ OK ]
> # INFO: Expected ratio 1.60 Measured ratio 1.59
> # TEST: bands 0:2                                                     [ OK ]
> # INFO: Expected ratio 3.33 Measured ratio 3.33
> # Testing ets bands 2 quanta 5000 2500, streams 0 1
> # TEST: bands 0:1                                                     [ OK ]
> # INFO: Expected ratio 2.00 Measured ratio 1.99
> # Running in classifier mode
> # Testing ets bands 3 strict 3, streams 0 1
> # TEST: band 0                                                        [ OK ]
> # INFO: Expected ratio >95% Measured ratio 100.00
> # TEST: band 1                                                        [ OK ]
> # INFO: Expected ratio <5% Measured ratio 0
> # Testing ets bands 3 strict 3, streams 1 2
> # TEST: band 1                                                        [ OK ]
> # INFO: Expected ratio >95% Measured ratio 100.00
> # TEST: band 2                                                        [ OK ]
> # INFO: Expected ratio <5% Measured ratio 0
> # Testing ets bands 4 strict 1 quanta 5000 2500 1500, streams 0 1
>
> I tried to run 'set -x' enabled version standalone, but that one finished
> correctly (?).
>
> It could be something previous scripts left, but right now I don't have a clue.
> I can attempt to rerun all tests with sch_ets.sh bash 'set -x' enabled later today.

If you run it standalone without set -x, does it finish as well? That
would imply that the reproducer needs to include the previous tests as
well.

It looks like the test is stuck in ets_test_mixed in classifier_mode.
A way to run just this test would be:

    TESTS="classifier_mode ets_test_mixed" ./sch_ets.sh

Looking at the code, the only place that I can see that waits on
anything is the "{ kill %% && wait %%; } 2>/dev/null" line in
stop_traffic() (in lib.sh). Maybe something like this would let
us see if that's the case:

modified   tools/testing/selftests/net/forwarding/lib.sh
@@ -1468,8 +1470,10 @@ start_tcp_traffic()
 
 stop_traffic()
 {
+	echo killing MZ
 	# Suppress noise from killing mausezahn.
 	{ kill %% && wait %%; } 2>/dev/null
+	echo killed MZ
 }