[00:05] -queuebot:#ubuntu-release- New sync: gnome-shell-extension-caffeine (eoan-proposed/primary) [0~git20190409-2] [06:30] cjwatson: the libreoffice build in proposed has been building on arm64, armhf for ~2.5 days, and seems stuck? can something be done there? [06:30] or maybe vorlon? [06:35] RikMills: All I can do is cancel it (which I've done). Looks like a pkgbinarymangler bug maybe [06:38] cjwatson: thanks! I assume a retry? [06:40] I don't think a retry will help (except for deleting build logs...) [06:41] Er, yeah, don't just stab retry. [06:41] Unlikely to be transient. [06:41] if that is the case, can it be removed from proposed for now? it is blocking tests on other packages [06:41] I'll leave decisions about that to somebody else [06:42] vorlon? [06:43] RikMills, looks like the testsuite is leaving some spurious process... [06:44] a quick and dirty solution might be to not run testsuite on arm* (leaving autopkgtests do their job), 2) see if Debian has patches, or launch gencontrol dh override with some "ps aux" command and see which are the processes [06:44] I'm doing the latter in ppa [06:44] I wouldn't have expected that to cause the pkgbinarymangler errors I was seeing [06:46] oops indeed [06:46] this line Scanning for processes to kill in build PACKAGEBUILD-17762183 [06:46] is because you killed it? [06:46] Yes [06:46] oops my bad [06:47] So please put a bit more thought in rather than stabbing testsuites which don't relate [06:47] pkgstriptranslations: updating translation tarball libreoffice_6.3.1-0ubuntu1_arm64_translations.tar.gz...done [06:47] gzip: stdin: unexpected end of file [06:48] in any case disabling the testsuite in a ppa helps having faster debug... [06:48] that's the first suspicious thing I see, but I'm mostly listening to a presentation so there may well be something before that [06:48] or it could be a parallelisation problem in the packaging machinery [06:48] or something else I haven't worked out [06:48] lol yes [06:48] dh_builddeb.pkgbinarymangler: dpkg-deb --build debian/.debhelper/uno-libs3/dbgsym-root debian/.debhelper/scratch-space/build-uno-libs3 returned exit code 2 [06:49] marcustomlinson: ^- you probably want to look into this one way or another? [07:29] RikMills: did you talk to the uploader of the stuck libreoffice package about what's needed here? (marcustomlinson) [07:30] looks like cjwatson already covered this [07:30] vorlon: yeah. waiting to see what they say. I was a bit hasty earlier [07:31] ok, so you don't want me to delete it yet? [07:32] vorlon: would that lose the logs? if not, then removing would be ok I guess, as this version is going nowhere [07:32] RikMills: it does not remove the logs [07:33] RikMills: I also don't see any other packages in -proposed blocked by libreoffice [07:33] is it causing a build failure of some revdep on arm*? [07:33] vorlon: ok, so if it does no harm that way, then I would like it gone [07:34] it is causing libreoffice test against kio in proposed to fail [07:36] which is holding back 10 other kf5 packages [07:37] RikMills: that doesn't sound right. kio would be tested against the version of the package in release, not proposed, unless someone triggered it differently [07:40] vorlon: test of libreoffice in release against kio in -proposed https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a74ba3b87153c0ac/autopkgtest-eoan/eoan/arm64/libr/libreoffice/20190915_215212_d3f9b@/log.gz [07:41] RikMills: so, it is possible for you to rerun the test using the libreoffice from release + the kde stack, but you have to write the trigger to pull in all of the kde library dependencies together from -proposed [07:41] hmmmm [07:41] I'll give that a go in a bit [07:42] RikMills: yes, that log shows autopkgtest being unable to cherry-pick just kio from -proposed, so it falls back to pulling everything to -proposed, you have to search in the logs to see this [07:42] autopkgtest: WARNING: package ure is not installed though it should be [07:42] autopkgtest: WARNING: Test dependencies are unsatisfiable with using apt pinning. Retrying with using all packages from eoan-proposed [07:42] * RikMills nods [08:00] vorlon: as far as I can tell on last try, all the kf5 components fetched from -proposed have since gone to -release. so I *think* just a plain retry might now work [08:01] RikMills: ok. if that still doesn't work, I'm not opposed to removing libreoffice, but I'm in meetings right now so can't get to it just now [08:02] vorlon: sure. there is no great hurry. :) === pieq_ is now known as pieq [08:18] Laney, juliank: hey, which version of autodep8 is used for generating the debian/tests/control on autopkgtest.u.c? I got lost trying to figure that out, and I want a version that includes the fix for Debian bug #940001 [08:18] Debian bug 940001 in autodep8 "Source: python-foo + Binary: python3-foo + Binary: python-foo-common -> bad" [Normal,Fixed] http://bugs.debian.org/940001 [08:20] vorlon: it's in a checkout in the home directory on the workers [08:20] you need to `make install` it IIRC [08:35] Laney: so where's the hook for that? I didn't find references to it in autopkgtest-cloud [08:35] and you say "home directory on the workers", which are ephemeral? [09:02] the worker controller [09:03] 'git grep autodep8' shows what is done in the install hook [09:38] -queuebot:#ubuntu-release- New binary: pocketsphinx [i386] (eoan-proposed/universe) [0.8.0+real5prealpha+1-5ubuntu1] (no packageset) [09:38] -queuebot:#ubuntu-release- New binary: pocketsphinx [amd64] (eoan-proposed/universe) [0.8.0+real5prealpha+1-5ubuntu1] (no packageset) [09:38] -queuebot:#ubuntu-release- New binary: pocketsphinx [ppc64el] (eoan-proposed/universe) [0.8.0+real5prealpha+1-5ubuntu1] (no packageset) [09:39] please accept it, no rdeps ^^ fixing build failure and sphinxbase migration [09:40] oops doko I copy-pasted your signature, so it seems uploaded by you... sorry for that! [09:40] -queuebot:#ubuntu-release- New binary: pocketsphinx [armhf] (eoan-proposed/universe) [0.8.0+real5prealpha+1-5ubuntu1] (no packageset) [09:41] -queuebot:#ubuntu-release- New binary: pocketsphinx [arm64] (eoan-proposed/universe) [0.8.0+real5prealpha+1-5ubuntu1] (no packageset) [09:42] please remove it from s390x, Debian already removed it ^^ [09:46] Laney: huh. I swear I looked for the autodep8 directory on the cloud worker and did not find it. Ok, thanks [10:07] -queuebot:#ubuntu-release- Unapproved: pulseaudio (bionic-proposed/main) [1:11.1-1ubuntu7.3 => 1:11.1-1ubuntu7.4] (desktop-core, ubuntu-server) [10:12] -queuebot:#ubuntu-release- New binary: pocketsphinx [i386] (eoan-proposed/universe) [0.8.0+real5prealpha+1-5ubuntu2] (no packageset) [10:13] -queuebot:#ubuntu-release- New binary: pocketsphinx [ppc64el] (eoan-proposed/universe) [0.8.0+real5prealpha+1-5ubuntu2] (no packageset) [10:17] -queuebot:#ubuntu-release- New binary: pocketsphinx [arm64] (eoan-proposed/universe) [0.8.0+real5prealpha+1-5ubuntu2] (no packageset) [10:17] -queuebot:#ubuntu-release- New binary: pocketsphinx [armhf] (eoan-proposed/universe) [0.8.0+real5prealpha+1-5ubuntu2] (no packageset) [10:17] -queuebot:#ubuntu-release- New binary: pocketsphinx [amd64] (eoan-proposed/universe) [0.8.0+real5prealpha+1-5ubuntu2] (no packageset) [10:31] doko, FYI readelf is utterly broken on 32 bit [10:32] * LocutusOfBorg opens a bug, wrt abi-monitor segfault and test regression [10:36] for people interested in the abi-monitor regression, https://bugs.launchpad.net/ubuntu/+source/binutils/+bug/1844119 [10:36] Launchpad bug 1844119 in binutils (Ubuntu) "readelf crash on 32bit, leading to abi-monitor testsuite regression" [High,Confirmed] [11:21] doko, does the above bug ring a bell? [11:40] -queuebot:#ubuntu-release- Unapproved: accepted livecd-rootfs [source] (disco-proposed) [2.578.8] [11:42] -queuebot:#ubuntu-release- Unapproved: accepted livecd-rootfs [source] (bionic-proposed) [2.525.30] [11:43] -queuebot:#ubuntu-release- Unapproved: accepted livecd-rootfs [source] (xenial-proposed) [2.408.51] [12:44] LocutusOfBorg: thanks - this is blocking my curl upload too [12:45] -queuebot:#ubuntu-release- New source: nvidia-graphics-drivers-435 (bionic-proposed/primary) [435.21-0ubuntu0.18.04.1] [12:49] -queuebot:#ubuntu-release- Unapproved: nvidia-graphics-drivers-430 (bionic-proposed/restricted) [430.26-0ubuntu0.18.04.2 => 430.50-0ubuntu0.18.04.1] (no packageset) [12:59] -queuebot:#ubuntu-release- Unapproved: accepted dpkg [source] (bionic-proposed) [1.19.0.5ubuntu2.3] [13:03] -queuebot:#ubuntu-release- Unapproved: accepted dpkg [source] (disco-proposed) [1.19.6ubuntu1.2] [13:20] -queuebot:#ubuntu-release- Unapproved: gnome-software (disco-proposed/main) [3.30.6-2ubuntu4.19.04.1 => 3.30.6-2ubuntu4.19.04.2] (ubuntu-desktop) [13:37] I know amurray unfortunately my binutils foo is quite low, I did all my best to at least trace it down to a package [13:39] -queuebot:#ubuntu-release- Unapproved: accepted nvidia-graphics-drivers-430 [source] (bionic-proposed) [430.50-0ubuntu0.18.04.1] [13:43] -queuebot:#ubuntu-release- Unapproved: gnome-software (bionic-proposed/main) [3.28.1-0ubuntu4.18.04.10 => 3.28.1-0ubuntu4.18.04.11] (ubuntu-desktop) [13:51] -queuebot:#ubuntu-release- New source: networking-mlnx (bionic-proposed/primary) [1:12.4.0-0ubuntu0.18.04.1] [13:51] -queuebot:#ubuntu-release- Unapproved: networking-mlnx (disco-proposed/universe) [1:13.1.0-2 => 1:14.0.1-0ubuntu0.19.04.1] (no packageset) === juergh_ is now known as juergh [14:30] -queuebot:#ubuntu-release- New binary: ibus [s390x] (eoan-proposed/main) [1.5.21-1~exp2ubuntu1] (desktop-core, input-methods) [14:31] -queuebot:#ubuntu-release- New binary: ibus [i386] (eoan-proposed/main) [1.5.21-1~exp2ubuntu1] (desktop-core, input-methods) [14:34] -queuebot:#ubuntu-release- New binary: ibus [amd64] (eoan-proposed/main) [1.5.21-1~exp2ubuntu1] (desktop-core, input-methods) [14:35] -queuebot:#ubuntu-release- New binary: ibus [ppc64el] (eoan-proposed/main) [1.5.21-1~exp2ubuntu1] (desktop-core, input-methods) [14:40] -queuebot:#ubuntu-release- New binary: ibus [armhf] (eoan-proposed/main) [1.5.21-1~exp2ubuntu1] (desktop-core, input-methods) [14:43] -queuebot:#ubuntu-release- New binary: ibus [arm64] (eoan-proposed/main) [1.5.21-1~exp2ubuntu1] (desktop-core, input-methods) [14:48] Laney, vorlon: Could I get one of you to upgrade the kernel on the armhf autopkgtest hosts, so I can verify a bug fix? [15:41] infinity: done [15:41] nice [15:41] I was just going to do that [15:48] umm [15:49] the machines look kind of hosed dpkg-wise [16:17] ok fixed [16:24] Laney: hosed dpkg-wise> hmm what happened? unattended-upgrades in progress when I rebooted, and the reboot wasn't adequately blocked? [17:22] -queuebot:#ubuntu-release- New sync: python-hdf5storage (eoan-proposed/primary) [0.1.15-1] [17:47] vorlon: mind if i steal your pandas merge? [18:32] ginggs: has it become a merge now? did Debian fix the ftbfs? [18:41] vorlon: yes, I dropped some of your cherry picks, took some from debian [18:41] ginggs: ok, go for it [18:41] and ended up with something that builds https://launchpad.net/~ginggs/+archive/ubuntu/testing/+sourcepub/10550282/+listing-archive-extra [18:42] vorlon: ack, thanks. I still want to do some testing, so probably only tomorrow [18:49] -queuebot:#ubuntu-release- New binary: spoa [amd64] (eoan-proposed/universe) [3.0.1-1] (no packageset)