Created attachment 3675 [details] autobuilder log oe-selftest got stuck at oeqa.selftest.signing.Signing.test_signing_packages on the debian-testing autobuilder slave. The reason being gpg which got stuck and had to be manually killed. Reading the log, it seens that gpg experienced some kind of race condition. For ecample these look point in that direction: gpg: waiting for lock (held by 56801) ... gpg: error opening lockfile '/tmp/oeqa-signing-t1axf_rq/gnupg_spawn_agent_sentinel.lock': No such file or directory Full autobuilder log attached
We haven't seen this in seven months (or have we?). What do you think we should do? Just close it as "worksforme" or still leave it open. In the latter case the bug should be re-assigned to somebody else than me.
(In reply to comment #1) > We haven't seen this in seven months (or have we?). > > What do you think we should do? Just close it as "worksforme" or still leave > it open. In the latter case the bug should be re-assigned to somebody else > than me. the problem is still present but for the moment we have the following on meta/classes/sign_rpm.bbclass ---- # Newer versions of gpg (at least 2.1.5 and 2.2.1) have issues when signing occurs in parallel # so unfortunately the signing must be done serially. Once the upstream problem is fixed, # the following line must be removed otherwise we loose all the intrinsic parallelism from # bitbake. For more information, check https://bugzilla.yoctoproject.org/show_bug.cgi?id=12022. do_package_write_rpm[lockfiles] += "${TMPDIR}/gpg.lock" ---
This bug was reported earlier than bug 12022 but the latter has more details on the current status, thus setting this one as duplicated. *** This bug has been marked as a duplicate of bug 12022 ***