Bug#1074355: libmina-sshd-java: FTBFS more than 50% of the time

[Pierre Gruet]

Hi Santiago,

Le 04/09/2024 à 12:55, Santiago Vila a écrit :

El 4/9/24 a las 8:38, Pierre Gruet escribió:

I tried to build the package with GRUB_CMDLINE_LINUX="nr_cpus=1" as you suggested, and I got an error I had not met before:

         at com.jcraft.jsch.Session.setPortForwardingR(Session.java:1952)
[...]

Am I correct in assuming this is also the one you encountered?

[ERROR] Tests run: 1, Failures: 0, Errors: 1, Skipped: 0, Time elapsed: 0.134 s <<< FAILURE! - in org.apache.sshd.common.forward.PortForwardingTest [ERROR] testRemoteForwardingSecondTimeInSameSession(org.apache.sshd.common.forward.PortForwardingTest)  Time elapsed: 0.131 s  <<< ERROR! com.jcraft.jsch.JSchException: remote port forwarding failed for listen port 36491          at com.jcraft.jsch.Session._setPortForwardingR(Session.java:2195)

build logs in a more detailed way, I also get a different error
on machines with 2 CPUs:

java.lang.AssertionError: No SSH_MSG_DEBUG
         at org.junit.Assert.fail(Assert.java:89)
         at org.junit.Assert.assertTrue(Assert.java:42)
         at org.junit.Assert.assertNotNull(Assert.java:713)

[...]

For reference, I'm putting all my failed build logs so far here:

https://people.debian.org/~sanvila/build-logs/libmina-sshd-java/

That's the one I found on single-CPU systems, yes, but now that I look at my         at org.apache.sshd.common.kex.extension.StrictKexTest.testConnectionClosedIfSupriousPacketInKex(StrictKexTest.java:163)         at org.apache.sshd.common.kex.extension.StrictKexTest.connectionClosedIfSpuriousPacketFromServerInKex(StrictKexTest.java:134)         at java.base/ jdk.internal.reflect.DirectMethodHandleAccessor.invoke(DirectMethodHandleAccessor.java:103)

Thanks for the build logs. Using them I could list all the problems. I packaged the latest upstream version, addressing some of them, and I workd on the packaging do understand what was going on -- actually upstream has already spotted similar issues. Now the initial problem of this bug thread is still showing up and I can't reproduce it at the moment. However I think the other issues are gone; by chance, would you have recent (> Sep. 22nd) build logs to check this?

Thanks.

All the best,

--
Pierre

OpenPGP_signature.asc
Description: OpenPGP digital signature