From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-1.web.codeaurora.org [10.30.226.201]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 9A850817 for ; Mon, 26 Feb 2024 01:55:44 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=10.30.226.201 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1708912544; cv=none; b=DEBC1rGcjWifOqaucLE/6tnLPeSOWvZP4AHc6WMQdluIBScboVLeJi9gVDZ1J6Yzrl8fu6+xmgKv4ZkQEK9dl249as6LI+Op4zRRNaPV6DjPeftyr7DJZtpysCFdST6VACK9aeJkv192cBmBnSsPFEaewp3DUBOErsSDXFjVZvI= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1708912544; c=relaxed/simple; bh=wAXZ4hyJemOApnCphdxxVBJAWRBn3zCVkVMrO0FK4Yw=; h=Date:From:To:Cc:Subject:Message-ID:References:MIME-Version: Content-Type:Content-Disposition:In-Reply-To; b=OJ0Nwhx5OWVk7oAaZYm8eentel0SUEZaLRJzO2zd7aqhCpfXKjbcbvHsvuJYzJiLnPyeiZ9vSxg1S2Q1v3zvx3K3wlUvqHscBwkCa5TeQWCFZPXDQq/XY09TCjk/+6PFlH0fy3B/pyPVNmBNXaaZ/GCNTpZhVlndLIpLVoYgVWA= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=p4b1NiOq; arc=none smtp.client-ip=10.30.226.201 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="p4b1NiOq" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 8E837C433F1; Mon, 26 Feb 2024 01:55:42 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1708912544; bh=wAXZ4hyJemOApnCphdxxVBJAWRBn3zCVkVMrO0FK4Yw=; h=Date:From:To:Cc:Subject:References:In-Reply-To:From; b=p4b1NiOq0deKIz+lcl+sCJRUY/vT2NaIcZLqLW9vSuQP7dsYL5ZGdYY1CtIgdEU+/ LR93kpNBavzko2iI78jsHFj+JYIo4jrFuvGel64YqkGnENoXYotXJB0L1TCcZZD3/l 892MN/fr4NFBcNDOyPAt2dm4FsFw6uFAFKmf+uzBoLekhEIyIboNlt+hnKEILgTskP i1xABZ7rwFS6utw4V1f2rPcMcQ2qlqXC5UQ7Q0v3dsv5MwvZZSl+fp8TAc3Bh9y6e+ IBiLkbrkfpK/vXvuaMwOId9vad0sL5R8r+YhF4Ui2lGGapXdhrO8bkj/ybNWBbKsRi c8YvfXMQQxiQw== Date: Mon, 26 Feb 2024 09:55:37 +0800 From: Geliang Tang To: Matthieu Baerts Cc: mptcp@lists.linux.dev Subject: Re: [PATCH mptcp-net] selftests: mptcp: don't exit when a symbol not found Message-ID: References: <37d9862958848f8f3737eafd1b1ea0bdb63acd91.1708588566.git.tanggeliang@kylinos.cn> <338c256c-1492-4ea8-8d3d-eca7e075b0ef@kernel.org> <2f6096ac-568f-46a6-9938-3d57ad02536b@kernel.org> Precedence: bulk X-Mailing-List: mptcp@lists.linux.dev List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <2f6096ac-568f-46a6-9938-3d57ad02536b@kernel.org> Hi Matt, On Thu, Feb 22, 2024 at 11:49:18AM +0100, Matthieu Baerts wrote: > Hi Geliang, > > On 22/02/2024 11:31 am, Geliang Tang wrote: > > On Thu, Feb 22, 2024 at 09:58:39AM +0100, Matthieu Baerts wrote: > >> Hi Geliang, > >> > >> On 22/02/2024 8:56 am, Geliang Tang wrote: > >>> From: Geliang Tang > >>> > >>> mptcp_lib_kallsyms_has() will always exit when a symbol has not found, it > >>> breaks the test itself. Unexpected errors occur: > >>> > >>> 007 userspace pm add & remove address > >>> syn [ OK ] > >>> synack [ OK ] > >>> ack [ OK ] > >>> add [ OK ] > >>> echo [ OK ] > >>> mptcp_info subflows=2:2 [ OK ] > >>> mptcp_info subflows_total=3:3 [ OK ] > >>> mptcp_info add_addr_signal=2:2 [ OK ] > >>> mptcp_info last_data_sent=191:18 [ OK ] > >>> mptcp_info last_data_recv=40:68 [ OK ] > >>> mptcp_info last_ack_recv=93:74 [ OK ] > >>> dump addrs signal ERROR: missing feature: \ > >>> mptcp_userspace_pm_dump_addr$ symbol not found > >>> Cannot open network namespace "ns1-65d6fb5a-5FviVK": \ > >>> No such file or directory > >>> Cannot open network namespace "ns2-65d6fb5a-5FviVK": \ > >>> No such file or directory > >>> cmp: /tmp/tmp.a7osJs7Nj0: No such file or directory > >>> cmp: /tmp/tmp.f02z6brCQu: No such file or directory > >>> cat: /tmp/tmp.27TzxD2efV: No such file or directory > >>> not ok 1 test: selftest_mptcp_join # FAIL > >>> > >>> To fix this, this patch adds a new argument 'continue' for the helper > >>> mptcp_lib_fail_if_expected_feature() to control whether exit or not. > >>> > >>> Always set this argument to 1 in mptcp_lib_kallsyms_has(). > >>> > >>> Fixes: 83013bdf90a ("selftests: mptcp: connect: skip if MPTCP is not supported") > >> > >> I'm not sure to understand why you sent this :) > > > > Say in the test "signal addresses race test": > > > > if ! mptcp_lib_kallsyms_has "mptcp_pm_subflow_check_next$"; then > > chk_join_nr 3 3 2 > > chk_add_nr 4 4 > > else > > chk_join_nr 3 3 3 > > # the server will not signal the address terminating > > # the MPC subflow > > chk_add_nr 3 3 > > fi > > > > This code will never run to this block: > > > > chk_join_nr 3 3 2 > > chk_add_nr 4 4 > > > > Since if no symbal of mptcp_pm_subflow_check_next, mptcp_lib_kallsyms_has exit. > > This block will never be executed in our CI, because our CI always run > the selftests linked to the same version of the kernel. In this case, > that would not be normal to execute this block in our CI, it has to > complain because we expect to have this symbol in the latest version. > > Other CIs, like LKFT, which are validating stable kernels (e.g. 5.15.x) > using the selftests from the last stable kernel (e.g. 6.7.x) will not > have SELFTESTS_MPTCP_LIB_EXPECT_ALL_FEATURES env var being set to 1 as > we did in our CI. Then it is fine, tests and checks will be skipped if > some features are missing: > > https://tuxapi.tuxsuite.com/v1/groups/linaro/projects/lkft/tests/2cJxC6Sua3VVzRZmpQYKK2jzzNN/logs?format=html > > > You can trigger this by renaming "mptcp_pm_subflow_check_next$" to > > "no_mptcp_pm_subflow_check_next$", an non-exist function name, for test. > > I confirm that if you use 'mptcp-upstream-virtme-docker', this env var > will be set to 1 by default: > > SELFTESTS_MPTCP_LIB_EXPECT_ALL_FEATURES=1 > > You can unset it on your side, but it only makes sense to do that if you > work on stable kernels (and you use selftests from a more recent > version, not the ones attached to the kernel you are validating). > > So yes, in this environment (where we force setting the env var to 1, > not the "default" behaviour), picking a non-existing function name will > cause the test to fail: but that's what we want → we want to be notified > if we picked a non-existing function, like it happened recently: > > https://github.com/multipath-tcp/mptcp_net-next/actions/runs/7988706773 Thanks for your explanation, it took me some time to understand it. Yes, we can use SELFTESTS_MPTCP_LIB_EXPECT_ALL_FEATURES to control whether exit or not. Let's drop this patch. -Geliang > > Cheers, > Matt > -- > Sponsored by the NGI0 Core fund.