[fedora-arm] Fedora 25 Alpha blocker status mail #1

Hi folks! We're in Alpha freeze and Alpha go/no-go is scheduled for next week (2016-08-18), so it's time for the first blocker status mail. tl;dr action summary ==================== 1. https://bugzilla.redhat.com/show_bug.cgi?id=1365661 ACTION: releng/Cloud teams to figure out the problem, if it persists. 2. https://bugzilla.redhat.com/show_bug.cgi?id=1352680 ACTION: pjones to do a new fwupdate build and create an update with efivar, efibootmgr and all the other efivar-dependent packages. 3. https://bugzilla.redhat.com/show_bug.cgi?id=1324623 ACTION: FESCo to decide on whatever they've been asked to decide, glibc maintainers (fweimer), ARM team and releng to agree on a fix. 4. https://bugzilla.redhat.com/show_bug.cgi?id=1353054 ACTION: QA and FreeIPA folks to help test and karma the update for pushing. 5. https://bugzilla.redhat.com/show_bug.cgi?id=1366403 ACTION: blocker reviewers to vote on blocker status (will happen at Monday's blocker meeting), sssd devs to investigate and fix. Also note that some validation tests have still not been run, QA folks we should be aiming to fill out all the Alpha validation tests ASAP :) bug-by-bug detail ================= 1. https://bugzilla.redhat.com/show_bug.cgi?id=1365661 - distribution - NEW Cloud images fail to compose from 20160809.n.0 Cloud images stopped showing up with the 20160809 compose. We are not yet sure why. We *did* get cloud images for today's Rawhide compose, so it's possible they'll show up again with the next successful 25 compose. ACTION: releng/Cloud teams to figure out the problem, if it persists. 2. https://bugzilla.redhat.com/show_bug.cgi?id=1352680 - efibootmgr - NEW efibootmgr calls in anaconda crashing since efivar-0.24-1.fc25 landed efivar API/ABI changed and broke efibootmgr. As of yesterday there are new builds of efivar and efibootmgr and they do appear to solve the bug, but there's a problem: another package which depends on efivar - fwupdate - has not been rebuilt, and it is part of the Workstation package set. We need fwupdate to be built and included in the update submitted to fix this bug. ACTION: pjones to do a new fwupdate build and create an update with efivar, efibootmgr and all the other efivar-dependent packages. 3. https://bugzilla.redhat.com/show_bug.cgi?id=1324623 - glibc - ASSIGNED glibc: split libcrypt into separate package and package libcrypt-nss with NSS-based implementation A consequence of this package split has been to break the ARM Xfce image build, which is a release-blocking image. Relevant folks are broadly aware of the issue, but it doesn't seem like a fix has been agreed yet, and the issue has been referred to FESCo for some reason. ACTION: FESCo to decide on whatever they've been asked to decide, glibc maintainers (fweimer), ARM team and releng to agree on a fix. 4. https://bugzilla.redhat.com/show_bug.cgi?id=1353054 - pki-core - MODIFIED FreeIPA server deployment fails due to pki-core dangling symlinks This bug has now been fixed, with this update: https://admin.fedoraproject.org/updates/FEDORA-2016-0a384ead60 I have tested and confirmed the fix, the update ideally needs a couple more karma points to be pushed stable. ACTION: QA and FreeIPA folks to help test and karma the update for pushing. 5. https://bugzilla.redhat.com/show_bug.cgi?id=1366403 - sssd - NEW [abrt] sssd-common: ipa_dyndns_update_send(): sssd_be killed by SIGSEGV This is a proposed blocker, but it does seem quite worrying; we run into it with the openQA test that enrols a FreeIPA client via an anaconda kickstart file. The enrolment works, but on boot of the installed system, sssd_be crashes and this seems to prevent login as a FreeIPA user. sssd team is looking into this one. ACTION: blocker reviewers to vote on blocker status (will happen at Monday's blocker meeting), sssd devs to investigate and fix.