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;
Date:   Sat, 20 May 2023 07:43:11 +0900 (JST)
Message-Id: <20230520.074311.642413213582621319.yamato@redhat.com>
To:     bruce.dubbs@gmail.com
Cc:     kzak@redhat.com, linux-kernel@vger.kernel.org,
        linux-fsdevel@vger.kernel.org, util-linux@vger.kernel.org,
        renodr2002@gmail.com
Subject: Re: [ANNOUNCE] util-linux v2.39
From:   Masatake YAMATO <yamato@redhat.com>
In-Reply-To: <652d32c5-4b33-ce3a-3de7-9ebc064bbdcb@gmail.com>
References: <20230517112242.3rubpxvxhzsc4kt2@ws.net.home>
        <652d32c5-4b33-ce3a-3de7-9ebc064bbdcb@gmail.com>
Organization: Red Hat Japan, K.K.
Mime-Version: 1.0
Content-Type: Text/Plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Precedence: bulk

Bruce,

> On 5/17/23 06:22, Karel Zak wrote:
>> The util-linux release v2.39 is available at
>>                                       http://www.kernel.org/pub/linux/utils/util-linux/v2.39
>>                                    Feedback and bug reports, as always, are welcomed.
> 
> Karel, I have installed util-linux v2.39 in LFS and have run into a
> problem with one test, test_mkfds.  Actually the test passes, but does
> not clean up after itself. What is left over is:
> 
> tester 32245 1 0 15:43 ?  00:00:00 /sources/util-linux-2.39/test_mkfds
> -q udp 3 4 server-port=34567 client-port=23456 server-do-bind=1
> client-do-bind=1 client-do-connect=1
> tester 32247 1 0 15:43 ?  00:00:00 /sources/util-linux-2.39/test_mkfds
> -q udp6 3 4 lite=1 server-port=34567 client-port=23456
> server-do-bind=1 client-do-bind=1 client-do-connect=1
> 
> It's possible it may be due to something we are doing inside our
> chroot environment, but we've not had this type of problem with
> earlier versions of util-linux.
> 
> In all I do have:
> 
>   All 261 tests PASSED
> 
> but the left over processes interfere later when we try to remove the
> non-root user, tester, that runs the tests.  I can work around the
> problem by disabling test_mkfds, but thought you would like to know.

Thank you for reporting.
Reproduced on my PC. I found two processes were not killed properly.

Could you try the following change?

diff --git a/tests/ts/lsfd/option-inet b/tests/ts/lsfd/option-inet
index 21e66f700..70cc3798d 100755
--- a/tests/ts/lsfd/option-inet
+++ b/tests/ts/lsfd/option-inet
@@ -84,14 +84,10 @@ ts_cd "$TS_OUTDIR"
                   -o ASSOC,TYPE,NAME \
                   -Q "(PID == $PID0) or (PID == $PID1) or (PID == $PID2) or (PID == $PID3) or (PID == $PID4)"
 
-    kill -CONT "${PID0}"
-    wait "${PID0}"
-
-    kill -CONT "${PID1}"
-    wait "${PID1}"
-
-    kill -CONT "${PID2}"
-    wait "${PID2}"
+    for pid in "${PID0}" "${PID1}" "${PID2}" "${PID3}" "${PID4}"; do
+       kill -CONT "${pid}"
+       wait "${pid}"
+    done
 } > "$TS_OUTPUT" 2>&1
 
 ts_finalize