16:03 #startmeeting Weekly Ubuntu Foundations team 16:03 Meeting started Thu Jan 30 16:03:05 2020 UTC. The chair is bdmurray. Information about MeetBot at http://wiki.ubuntu.com/meetingology. 16:03 16:03 Available commands: action commands idea info link nick 16:03 #topic Lightning rounds 16:03 echo $(shuf -e vorlon bdmurray xnox tdaitx doko sil2100 rbalint infinity mwhudson juliank waveform) 16:03 tdaitx sil2100 waveform infinity doko bdmurray vorlon mwhudson rbalint juliank xnox 16:03 * xnox win 16:04 tdaitx: 16:04 Short week: out last week 16:04 * openjdk-8 16:04 - investigated os::commit_memory errors on armhf 16:04 - no proper reproducer (yet), happens intermittently, more so with some particular tests, will contact upstream with more data 16:04 - updated test scripts to copy & keep hs_err log files; need to copy this functionality to the other openjdk packages 16:04 * setting up test comparison for jck-11 16:04 * various openjdk blacklists and hinting 16:04 (done) 16:04 sil2100: 16:05 ops, I meant arm64, not armhf, typo! =) 16:05 waveform: 16:05 * Back-ported LP: #1311056 to xenial, bionic, eoan 16:05 * Fixed and LP: #1861292 and back-ported to bionic 16:05 * Lots of work on getting core18 to boot (nothing yet, but some interesting notes from the kernel team on possible incorrect dtb being loaded by the bootloader) 16:05 * More testing on focal and bionic .4 classic images; no issues (other than the flash-kernel one above) 16:05 (done) 16:05 Launchpad bug 1311056 in python-apt (Ubuntu Eoan) "[SRU] apt-add-repository adds duplicate commented/disabled source lines" [Undecided,New] https://launchpad.net/bugs/1311056 16:05 Launchpad bug 1861292 in flash-kernel (Ubuntu Bionic) "flash-kernel failure during kernel upgrade" [Critical,Fix committed] https://launchpad.net/bugs/1861292 16:05 Crap 16:06 * sil2100 not ready anwyway 16:06 doko: 16:06 - finished unversioned python removal, python-defaults migrated 16:06 - started python3.8 defaults transition 16:06 - gcc-10 update 16:06 - binutils & gdb updates 16:06 - more unrelated transition work 16:06 (done) 16:06 bdmurray: its a me! 16:06 updated meta-release files for 19.04 End of Life 16:06 updated daisy code base for 19.04 End of Life 16:06 learned how to use autodeploy with the Error Tracker 16:06 researched retracing queues growing huge 16:07 discovered issue with retracer configuration and ddebs key 16:07 emailed ubuntu-devel regarding retracing backlog 16:07 tested a distribution upgrade from X to B w/ -proposed 16:07 uploaded walinuxagent to F and SRUs (LP: #1851064) 16:07 Launchpad bug 1851064 in walinuxagent (Ubuntu Eoan) "[SRU] Update WALinuxAgent to 2.2.45" [High,In progress] https://launchpad.net/bugs/1851064 16:07 I can go now 16:07 SRU verification of LP: #1856684 16:07 attended Cape Town sprint 16:07 Launchpad bug 1856684 in livecd-rootfs (Ubuntu) "Populate .disk/info in ubuntu-image built images" [Medium,In progress] https://launchpad.net/bugs/1856684 16:07 ✔ done 16:07 * short week, out most of Thursday and Friday afternoon 16:07 * proposed-migration: worked on libffi transition, miscellaneous others 16:07 * disco EOL: running 'swift delete' in a loop for the past 3 days to get rid of the autopkgtest usage 16:07 (done) 16:07 I'll go now o/ 16:07 - SRU reviews and releases 16:07 - 18.04.4 - stuff, lots of stuff, busy busy, no panic yet tho 16:07 (done) 16:08 (didn't have time to write up anything better) 16:08 rbalint is still out 16:08 juliank: 16:08 * Sponsored flash-kernel uploads for waveform 16:08 * dug a bit further into openblas / hypre (LP: #1860601), and other autopkgtest regressions and migrations 16:08 Launchpad bug 1860601 in openblas (Ubuntu) "openblas: causes autopkgtest failure in hypre armhf" [Undecided,New] https://launchpad.net/bugs/1860601 16:08 * could not investigate apport regression as apt just segfaulted in autopkgtest image :/ 16:08 * apt 1.9.8, fixing crash in apt 1.9.7, slightly improving performance, and changing disco mentions to focal 16:08 * update-notifier pep8 -> pycodestyle change 16:08 * synced python-apt 1.9.5 16:08 * merged autopkgtest 5.11 16:08 * merged irssi 1.2.2 16:09 * re-added the pep8 binary package to pep8 to ease the python-defaults transition 16:09 * work on powerpc-utils, reviewing patches 16:09 doko: are you planning to do gcc-9 and gcc-8 uploads to focal? i.e. with just newer stuff from branches that are available? cause there are s390x fixes..... 16:09 * reviewing specs 16:09 * in the progress of reimplementing kernel autoremoval for apt 16:09 (done) 16:09 xnox: at some point, yes 16:09 xnox: your turn 16:09 Back from Cape Town Sprint 16:09 Off to Brussels for FOSDEM / escaping Brexit tomorrow 16:09 Updated s390-tools LP: #1860574 LP: #1860531 LP: #1859018 LP: #1853308 16:09 TODO: AA to approve s390-tools upload https://launchpad.net/ubuntu/focal/+queue?queue_state=1&queue_text=s390-tools 16:09 Launchpad bug 1860574 in Ubuntu on IBM z Systems "[UBUNTU 20.04] zkey: Fix display of clear key size for XTS keys" [High,In progress] https://launchpad.net/bugs/1860574 16:09 Uploaded livecd-rootfs changes to enable SSH access on s390x 16:09 Launchpad bug 1860531 in Ubuntu on IBM z Systems "IPL on z15 always performed regardless of the secure-boot related settings" [High,In progress] https://launchpad.net/bugs/1860531 16:09 Uploaded livecd-rootfs to use the right oem-20.04 flavour, which actually is just oem flavour, with matching ubuntu-cdimage changes, reviewed by Laney 16:09 Launchpad bug 1859018 in Ubuntu on IBM z Systems "[UBUNTU] zkey: Fix listing of keys on file systems reporting DT_UNKNOWN." [High,In progress] https://launchpad.net/bugs/1859018 16:09 Filed bug report that cloud-init crashes with static network configuration 16:09 Error: Could not gather data from Launchpad for bug #1853308 (https://launchpad.net/bugs/1853308). The error has been logged 16:09 Started mailing list discussion with server team that cloud-init does not harvest netplan.yaml that our initrd generates 16:09 Rebasing numpy-less boost on i386 onto boost1.71 builds, in preparation for ABI transition 16:09 Uploaded new qclib LP: #1852718 16:09 Launchpad bug 1852718 in Ubuntu on IBM z Systems "[20.04 FEAT] Upgrade qclib 2.0.1" [High,Fix released] https://launchpad.net/bugs/1852718 16:09 Uploaded casper with finalrd dep into bionic for point release LP: #1840122 16:09 Launchpad bug 1840122 in linux (Ubuntu Eoan) "System fails to reboot from live session or ubiquity-dm - squashfs_read_data failed to read block" [High,Confirmed] https://launchpad.net/bugs/1840122 16:10 Working with frank / ryan / mwhudson on debugging s390x curtin / subiquity / multipath failing to complete the install 16:10 Uploaded xdmf to unbreak really bad py3.8/cmake packaging 16:10 done 16:10 #topic Release incoming bugs 16:10 #link http://reqorts.qa.ubuntu.com/reports/rls-mgr/rls-ff-incoming-bug-tasks.html 16:11 there seem to be quite a few here 16:11 bug 1860531 is targetted and tagged (lame) 16:11 bug 1860531 in Ubuntu on IBM z Systems "IPL on z15 always performed regardless of the secure-boot related settings" [High,In progress] https://launchpad.net/bugs/1860531 16:12 that's in progress 16:12 needs AA to approve mu upload, as mentioned in my status with TODO 16:12 bug 1860432 16:12 bug 1860432 in systemd (Ubuntu) "juju agent fails to start on focal " [Undecided,Confirmed] https://launchpad.net/bugs/1860432 16:12 µ 16:13 so needs forwarding bug report to upstream systemd based on comment #1 16:13 i guess rbalint or I can take it 16:14 it's not critical to fix though 16:14 so let's card that 16:14 because with juju team we did a bit of a "but why are you doing this?!" 16:14 and they are changing code to be more normal. 16:14 bug 1861353 16:14 bug 1861353 in glibc (Ubuntu) "libc6-dev:amd64 is not co-installable with libc6-dev:s390x" [Undecided,New] https://launchpad.net/bugs/1861353 16:14 xnox: but the juju part is fixed, so is this critical for anything to change on the systemd side? 16:14 vorlon: correct juju is going to stop using subdirectories, but it is a potential upstream regression nonetheless 16:15 (worth flagging it up to systemd people, even if they simply release note it) 16:15 ok; so we can take it for further investigation but we might decline to change the behavior, sure 16:16 lets go ahead and target and card this one 16:16 * vorlon nods 16:16 bug 1859858 16:16 bug 1859858 in systemd (Ubuntu Focal) "udev has truncated ID_SERIAL value for scsi disk" [Undecided,New] https://launchpad.net/bugs/1859858 16:18 This one is already targetted so I'll card it. 16:19 bug 1860957 16:19 bug 1860957 in ubiquity (Ubuntu) "ZFS installation crash when ESP already exists" [High,Confirmed] https://launchpad.net/bugs/1860957 16:20 That should go to the desktop team. 16:21 bug 1843063 16:21 bug 1843063 in lshw (Ubuntu) "lshw -json output invalid in Eoan" [High,Confirmed] https://launchpad.net/bugs/1843063 16:22 Let's land their patch, so I'll target and card it. 16:23 https://ezix.org/src/ycheng/lshw/commit/82393aa9b59307a443677df7340a1dd38847a5ef.patch 16:23 Also bug 1860406 is new / undecided 16:23 Error: Launchpad bug 1860406 could not be found 16:23 bug 1850406 16:23 bug 1850406 in ubuntu-release-upgrader (Ubuntu) "python3-distupgrade small misprint bug" [Undecided,New] https://launchpad.net/bugs/1850406 16:24 that's the one I meant 16:24 I've an upload of u-r-u to do anyway so I'll take that and queue jump with it 16:24 nothing new in rls-ee- 16:25 #link http://reqorts.qa.ubuntu.com/reports/rls-mgr/rls-bb-incoming-bug-tasks.html 16:26 BB IS CLOSED 16:26 bug 1854839 still has work for me 16:26 bug 1854839 in debian-installer (Ubuntu) "Fresh Bionic install hangs forever at 66% waiting for os-prober" [Undecided,Confirmed] https://launchpad.net/bugs/1854839 16:26 STOOP 16:26 (just kidding) 16:27 bug 1856560 16:27 bug 1856560 in util-linux (Ubuntu) "ds-identify - stuck in uninterruptible sleep state" [Undecided,New] https://launchpad.net/bugs/1856560 16:27 We seem to have talked about this before. 16:28 okay, xnox will look at the comments 16:28 #topic Team proposed-migration report 16:29 vorlon: can you help here? 16:29 ack 16:30 python3.7, python3.8 are entangled in the libffi transition which I'm working through (libffi itself got its migratability reset by a ruby2.7 autopkgtest regression, which I've hinted away again; we're only about 2 packages away from this getting through) 16:30 libtimedate-perl causing libsoap-wsdl-perl regression, who wants this? 16:30 [LINK] https://people.canonical.com/~ubuntu-archive/proposed-migration/focal/update_excuses_by_team.html#foundations-bugs 16:30 btw 16:31 xnox: ^^ libtimedate-perl? 16:32 bdmurray: can you take libcgi-pm-perl? 16:32 libuv1 is figuring out whether to fix or hint i386, so I'll take that unless someone else is keen 16:32 libffi, glib2.0 are all part of the same deal 16:32 aptdaemon: should force-badtest aptdaemon/all/arm64 IMO 16:33 juliank: ok, I'll assign that to you to follow through on 16:33 vorlon: are you assigning that to me? ok.... 16:33 vorlon: okay 16:33 libgpg-error is another i386 one, would anybody else like to take one of these? 16:34 crickets 16:34 fine, I'll take libgpg-error and libjsoncpp too 16:36 open-iscsi and python-dateutil are in the works 16:37 unattended-upgrades needs an assignee. doko? 16:37 where do you see this? 16:38 doko: the list that we're going through right now 16:38 No 16:38 can you follow up on unattended-upgrades? 16:38 unattended-upgrades we know why it's failing 16:38 It's going to be fixed in a new apt release 16:38 ok, that sounds like juliank is claiming it then 16:38 yes 16:38 ok 16:38 Just finalizing the fixes 16:38 libcap-ng is tied to python3-defaults, which we know the status of 16:39 doko: who is working on open-iscsi and python-dateutil? 16:39 (you said "in the works" but not by who) 16:39 nobody, they are waiting for the tests to finish 16:39 doko: does that mean you retriggered tests? since the tests shouldn't have taken 5 days 16:40 yes 16:40 if so, I'll consider these yours :) 16:40 python3-defaults, doko is obviously working on, if other people have time to pitch in as well that would be good (I'll be helping with it but only after I finish libffi) 16:40 wooohoo, CI for APT fix just passed 16:40 yes, please. 16:40 update-manager? 16:41 update-manager is part of the python3-defaults transition, does it need particular attention? 16:41 can we ignore all the tryton stuff? 16:41 scheduled for removal in Debian anyway 16:41 ignore, no. remove, yes 16:41 we should kill tryton stuff 16:41 and odoo / openerp / tinyerp too 16:41 Odd_Bloke: and xnox no longer care about it =) 16:42 we can follow up on that on #ubuntu-release 16:42 gdb vs python? 16:42 retried 16:42 xnox: :D 16:42 doko: ok 16:42 sl-modem is more i386, so 16:42 what was retried? 16:42 bdmurray: the python autopkgtest failures with new gdb 16:43 okay so update-manager still needs looking at? 16:43 and then p11-kit and talloc are part of the aforementioned transitions 16:43 certainly looks like update-manager needs fixing for python3.8 16:44 so it's worth someone taking that one specifically 16:44 bdmurray: ? 16:44 I can have a look 16:44 thanks 16:44 and that's all 16:44 * juliank just used the big hammer and retried all python-apt triggered-tests that failed 16:44 one more, libreoffice/amd64 ftbfs with out-of-dis-space 16:45 #topic AOB 16:45 ore here: one more, libreoffice/amd64 ftbfs with out-of-disk-space 16:45 I'm probably gonna end up taking a half day to aid my recovery 16:45 desktop is aware 16:45 I'm also taking a swap day on Monday 16:45 ah I am also off tomorrow 16:46 and one more: https://bugs.launchpad.net/auto-package-testing/+bug/1860763 16:46 Launchpad bug 1860763 in Auto Package Testing "the image for the autopkg tests has a big python3 stack installed" [Undecided,New] 16:46 I know Laney has commented on the bug 16:46 please can we fix that, especially now with all the python3 stuff going on? 16:47 yes 16:47 essentially mark everything auto 16:47 pointing out that you'll never have a guarantee that there are no extra packages installed in the test bed 16:47 vorlon: there's no comment 16:47 mark ubuntu-minimal and whitelisted packages manual 16:47 then run autoremove 16:47 oh, it wasn't on the bug 16:47 I saw a reply from him somewher 16:47 I think we can do that 16:47 IRC 16:47 right 16:47 I think my idea should work, just need to figure out the whitelist 16:48 No problem with attempting to minimise the testbed, but the point still stands and it could be another package down the line 16:48 it is a reasonable improvement to try to minimize the contents of the testbed image, if for no other reason than it speeds up upgrades 16:48 sure 16:48 but also what Laney said 16:48 package tests shouldn't misbehave in the presence of extra packages and if they do that's a bug 16:48 alright will y'all follow up in the bug? 16:49 vorlon: I don't think we have test-conflicts yet, might be useful, idk 16:50 juliank: that's one way to ensure tests always behave properly, but in the absence of that there's no test that couldn't be fixed (upstream or whatever) to just do what we want in the presence of extra stuff 16:50 true 16:51 the reverse is also true 16:51 tests could be passing that should be failing because they miss a dependency 16:51 sure, but in both cases you'd have to first discover the difference between testbed implementations and then change the package 16:51 anyway, the bug is valid 16:51 but I don't think it's high priority, it's really corner-case 16:53 okay, anything else? 16:54 alrighty then, thanks everybody 16:54 #endmeeting