[06:03] -queuebot:#ubuntu-release- Packageset: 101 entries have been added or removed [06:43] vorlon: I'm not disappointed, but want to make sure: this is just i386 removal or would you want/need something changed in DPDK for this? [07:17] cpaelzer: just i386 removal === andrewc is now known as Guest19848 [08:26] sil2100, could you please accept wslu srus? [08:53] -queuebot:#ubuntu-release- New: accepted linux-signed [amd64] (focal-proposed) [5.4.0-8.11] [08:53] -queuebot:#ubuntu-release- New: accepted linux-signed [ppc64el] (focal-proposed) [5.4.0-8.11] [08:53] -queuebot:#ubuntu-release- New: accepted linux-signed [arm64] (focal-proposed) [5.4.0-8.11] [08:53] -queuebot:#ubuntu-release- New: accepted linux-signed [s390x] (focal-proposed) [5.4.0-8.11] [09:01] -queuebot:#ubuntu-release- Unapproved: landscape-client (eoan-proposed/main) [18.01-0ubuntu9 => 18.01-0ubuntu9.1] (ubuntu-server) [09:20] hey there [09:20] SRU team, we would like https://launchpad.net/ubuntu/+source/pulseaudio/1:8.0-0ubuntu3.11 to be pulled out of xenial-updates due to bug #1856054 [09:20] bug 1856054 in pulseaudio (Ubuntu) "pulseaudio now depends on libsnapd-glib1 which recommends snapd" [High,Confirmed] https://launchpad.net/bugs/1856054 [09:21] rbasak, tjaalton, sil2100, ^ can one of you help with that? [09:21] popey, ^ fyi [09:21] xenial-updates? [09:22] yes [09:22] it builds with libsnapd-glib now [09:22] I don't think we have power to do that [09:22] the bug mentions bionic [09:22] Yeah I was about to say [09:22] right, but that impacts xenial as well [09:22] Need an AA to pull it [09:23] SRU team can only accept something on top I believe [09:23] I'm AA and can delete [09:23] so should I just go ahead and delete from updates? [09:23] unsure what's the process there [09:23] +1 wearing my SRU hat to treat it as a regression and in principle pull it [09:23] rbasak, tjaalton, the problem is that libsnapd-glib which got added as a depends is a small glue lib, but it Recommends snapd [09:24] As I assume it was an unintended consequence? [09:24] rbasak, tjaalton, it's less an issue on bionic since we do install snapd by default there [09:24] indeed [09:24] ok [09:24] people who did the SRU didn't take into account that not everyone is using snapd [09:24] especially not on xenial [09:24] was just going to ask why it's an issue because it's not clear from the bug [09:24] lol ;) [09:25] it basically means any xenial user having pulseaudio is going to have snapd pulled in [09:25] which is probably going to annoy a lot of people [09:25] seb128: I think you should delete it from updates assuming it hasn't been there very long [09:25] yes, go for ti [09:25] it [09:25] rbasak, tjaalton, thanks [09:25] seb128: but we need to upload a replacement soon to avoid confusion with people having an unpublished version installed [09:25] I wonder if we should delete from bionic as well? [09:25] right [09:25] seb128: +1 from me as well, the procedures are generally: remove the version from -updates, and copy-package the previous version so that it's re-published, and then quickly look for a revert/fix [09:26] or restore that one once we have fixed snapd-glib to lower the Recommends to a Suggests [09:26] If the effect is the same, then we should do the same thing on all affected supported releases [09:26] well, bionic has snapd installed by default on Ubuntu Desktop at least [09:26] but it's still going to pull snapd for e.g xubuntu users who might not want it [09:26] There are also a contingent of users who deliberately remove snapd on installation [09:27] should that lib really be recommending? [09:27] I wouldn't say that we "support" that, but if we reinstante snapd as an unintended consequence, rather than a deliberate choice, then I think that's something we should accept as a regression and pull it for the benefit of that contingent too. [09:27] That's probably the fix, but in the interim it's correct to remove the SRUs [09:35] k, SRU deleted from now, I'm going to republish the old one to updates [09:36] the new one is in proposed [09:36] we need to fix snapd-glib to lower the recommends to a suggest then we can restore the pulseaudio one [09:36] thanks rbasak tjaalton sil2100 [09:36] popey, ^ [09:37] Nice one, thanks. [09:37] popey, thx for pointing it out! [09:41] np [09:41] someone lobbed an angry youtube video my way. i will thank them :) [09:42] hahah [09:42] sincerely, outraged of youtube [09:43] "sincerely". This is a new alternate universe youtube I would quite like. [09:49] popey: the video by 'Learn Linux'? [09:49] yes [09:50] -queuebot:#ubuntu-release- Unapproved: pulseaudio (xenial-updates/main) [1:8.0-0ubuntu3.11 => 1:8.0-0ubuntu3.10] (core) (sync) [09:50] ok, I commented with the bug link [09:50] as did i [09:50] :D [09:58] rbasak, tjaalton, ^ I copied back the previous version but it hit unapproved, should I just accept it from there? [09:59] fine by me [09:59] +1 [10:00] thx [10:00] -queuebot:#ubuntu-release- New binary: zinnia [s390x] (focal-proposed/universe) [0.06-5ubuntu1] (input-methods) [10:01] -queuebot:#ubuntu-release- Unapproved: pulseaudio (bionic-updates/main) [1:11.1-1ubuntu7.5 => 1:11.1-1ubuntu7.4] (desktop-core, ubuntu-server) (sync) [10:02] -queuebot:#ubuntu-release- New binary: zinnia [amd64] (focal-proposed/universe) [0.06-5ubuntu1] (input-methods) [10:02] -queuebot:#ubuntu-release- New binary: zinnia [armhf] (focal-proposed/universe) [0.06-5ubuntu1] (input-methods) [10:02] -queuebot:#ubuntu-release- New binary: zinnia [arm64] (focal-proposed/universe) [0.06-5ubuntu1] (input-methods) [10:02] -queuebot:#ubuntu-release- New binary: zinnia [ppc64el] (focal-proposed/universe) [0.06-5ubuntu1] (input-methods) [10:02] -queuebot:#ubuntu-release- Unapproved: accepted pulseaudio [sync] (bionic-updates) [1:11.1-1ubuntu7.4] [10:02] -queuebot:#ubuntu-release- Unapproved: accepted pulseaudio [sync] (xenial-updates) [1:8.0-0ubuntu3.10] [10:20] sil2100: I'm just reviewing open-vm-tools in the SRU queues now - carry over from yesterday [10:32] rbasak: thanks for the heads up, I didn't start my SRU shift yet if anything anyway [10:42] -queuebot:#ubuntu-release- Unapproved: accepted open-vm-tools [source] (eoan-proposed) [2:11.0.1-2ubuntu0.19.10.2] [10:43] -queuebot:#ubuntu-release- Unapproved: accepted open-vm-tools [source] (disco-proposed) [2:11.0.1-2ubuntu0.19.04.2] [10:44] -queuebot:#ubuntu-release- Unapproved: accepted open-vm-tools [source] (bionic-proposed) [2:11.0.1-2ubuntu0.18.04.2] [10:45] thanks rbasak [11:07] -queuebot:#ubuntu-release- New: accepted zinnia [amd64] (focal-proposed) [0.06-5ubuntu1] [11:07] -queuebot:#ubuntu-release- New: accepted zinnia [armhf] (focal-proposed) [0.06-5ubuntu1] [11:07] -queuebot:#ubuntu-release- New: accepted zinnia [s390x] (focal-proposed) [0.06-5ubuntu1] [11:07] -queuebot:#ubuntu-release- New: accepted zinnia [arm64] (focal-proposed) [0.06-5ubuntu1] [11:07] -queuebot:#ubuntu-release- New: accepted zinnia [ppc64el] (focal-proposed) [0.06-5ubuntu1] [12:10] I fixed https://people.canonical.com/~ubuntu-archive/germinate-output/ubuntu.focal/ to run on amd64 as of the last run; that was the clear intent of the code vorlon wrote in update-germinate, but there was a bug that meant it was still running for i386 [12:10] (Similarly the other *.focal other than i386.focal) [13:02] -queuebot:#ubuntu-release- Unapproved: valgrind (bionic-proposed/main) [1:3.13.0-2ubuntu2.1 => 1:3.13.0-2ubuntu2.2] (ubuntu-desktop) [13:33] sil2100, could you please accept this 'valgrind' ^ upload if your time permit today ? Thanks in advance [13:49] sil2100, could you also please check landscape-client for eoan? [14:13] -queuebot:#ubuntu-release- Unapproved: linux-firmware (bionic-proposed/main) [1.173.13 => 1.173.14] (core, kernel) [14:14] -queuebot:#ubuntu-release- Unapproved: linux-firmware (disco-proposed/main) [1.178.6 => 1.178.7] (core, kernel) [14:16] -queuebot:#ubuntu-release- Unapproved: linux-firmware (eoan-proposed/main) [1.183.2 => 1.183.3] (core, kernel) [14:17] -queuebot:#ubuntu-release- Unapproved: xen (disco-proposed/main) [4.9.2-0ubuntu2 => 4.9.2-0ubuntu4] (ubuntu-desktop, ubuntu-server, virt) [14:28] rbalint: as an fyi simpoir and I are currently working at bumping lds-client on eoan to 19.12 [14:29] does your fix included in 19.12 release ? [14:32] rbalint, would you be okay if we add your fix on top of 19.12 (If not already there ?) instead ? [14:34] sil2100: hey do you know who did this: 03:50 -queuebot:#ubuntu-release- Unapproved: pulseaudio (xenial-updates/main) [1:8.0-0ubuntu3.11 => 1:8.0-0ubuntu3.10] (core) (sync) [14:35] sil2100: and this: 04:01 -queuebot:#ubuntu-release- Unapproved: pulseaudio (bionic-updates/main) [1:11.1-1ubuntu7.5 => 1:11.1-1ubuntu7.4] (desktop-core, ubuntu-server) (sync) [14:35] jdstrand: hey! I think it was seb128, an emergency update pull as it was causing some regressions for users [14:35] Pulling in snapd via a recommends chain I guess [14:35] sil2100: notice, whoever did that downgraded pulseaudio from an accepted SRU [14:35] https://launchpad.net/ubuntu/+source/pulseaudio/+publishinghistory [14:36] jdstrand: yes, it's an intentional downgrade [14:36] what was the issue? [14:36] this broke snapd tests (cc mborzecki) [14:37] ok, I see backscroll [14:38] jdstrand: I think they're working on pushing a new, fixed version [14:46] seb128: are you using a new bug to track the pulseaudio issue or are you going to move https://bugs.launchpad.net/ubuntu/+source/pulseaudio/+bug/1781428 back to Fix Committed? It is currently Fix Released, which is wrong with the revert [14:46] Ubuntu bug 1781428 in pulseaudio (Ubuntu Bionic) "please enable snap mediation support" [Medium,Fix released] [14:49] seb128: also, when do you expect the updated pulseaudio to land? I'm guessing this is a new full SRU cycle so not until after the new year? [14:52] jdstrand, hey, @bug should be set back to fix commited indeed [14:53] jdstrand, I think we just need to make snapd-glib lower its recommends to a suggest, that can probably be done today/tomorrow and might not need a full week staging since it's just a packaging change, but that's up to the SRU team to decide [14:54] it should be fine to aim for after the w.e moving back to updates imho, but again it's a SRU team call [14:58] seb128: ok, I'll adjust the bug and make a comment. is there a snapd-glib bug? [14:59] jdstrand, bug #1856054 [14:59] bug 1856054 in pulseaudio (Ubuntu) "pulseaudio now depends on libsnapd-glib1 which recommends snapd" [High,Confirmed] https://launchpad.net/bugs/1856054 [15:00] kenvandine seems to be on it [15:02] seb128: yeah, jamesh is going to talk to robert then prepare an upload [15:04] kenvandine, jdstrand, I opened https://bugs.launchpad.net/ubuntu/+source/snapd-glib/+bug/1856196 for the SRU with testcase, etc. Going to be easier than conflicting problems on the same reports [15:04] Ubuntu bug 1856196 in snapd-glib (Ubuntu) "Lower the snapd recommends to a suggests" [Undecided,New] [15:04] seb128: thanks [15:04] kenvandine, it's a simple s/Recommends/Suggests afaik, you could probably just do that change and upload today if we want to aim at getting the SRU back in before holidays [15:05] seb128: sure the snapd-glib bug being separate is fine. fyi, https://bugs.launchpad.net/ubuntu/+source/pulseaudio/+bug/1781428/comments/21 [15:05] Ubuntu bug 1781428 in pulseaudio (Ubuntu Bionic) "please enable snap mediation support" [Medium,Fix committed] [15:06] vorlon: could i386 builds use the nocheck profile? that would get rid of the build-depends on xvfb (and xauth) [15:11] jdstrand, ack, it's a bit unfortunate that the test can't adapt/check the capability. Does it mean the tests need to hardcode the environment they are used on? e.g they will fail on other distributions? [15:13] seb128: the test is specifically meant to catch if the mediation patches are dropped in Ubuntu [15:13] seb128: they are working as designed [15:14] the test is run on other distros (but could if other distros picked up the patch) [15:14] stepping into a meeting [15:14] jdstrand, k, it seems a bit unfortunate that the test is another component and not in pulseaudio then :/ [15:15] seb128: there are all kinds of functional tests in snapd [15:15] anyway... [15:15] jdstrand, right, thanks for the headsup and enjoy the meeting! [15:16] thanks! [15:19] vorlon: there are 28 packages that build-depend on xvfb, and four of them don't support nocheck atm, but a fixed libglvnd is in experimental and I fixed libxkbcommon in git. it leaves apport and qtwebsockets-opensource-src to fix [15:20] seb128: "That's only a depends change, snapd isn't going to be pulled in when installing the library anymore but that's wanted" -> are you sure that there's nothing relying on that? [15:20] For example, is there another path to get snapd installed by default where it gets installed by default? [15:22] kenvandine, ^ to check with robert I guess? [15:22] rbasak, xenial didn't even have that libary, it got added as a SRU, but yeah we should check that gnome-software (which is the only user of the lib there afaik) doesn pull snapd if/when needed [15:23] yeah, that's why i'd rather have robert in the loop [15:24] seb128: and i think neon uses snapd-glib as well [15:25] for discover [15:25] kenvandine, doesn't seem to on xenial according to rdepends on my xenial system [15:25] but that might be an issue on bionic [15:25] yeah [15:25] anyway we should review how snapd gets pulled in [15:29] -queuebot:#ubuntu-release- Unapproved: accepted valgrind [source] (bionic-proposed) [1:3.13.0-2ubuntu2.2] [15:56] slashd, re: landscape-client no, the fix is not in 19.12, i cherry-picked it to focal [15:58] slashd, i'd like to get this fix out before christmas, do you have a similar time-frame? [15:58] rbalint, I can upload 19.12 next week [15:58] including your fix [15:58] simpoir, ^ [16:00] slashd, this sounds like getting it to release next year 6 Jan, the earliest [16:01] rbalint, more or less yeah, if you need it before that, feel free to pursue with your upload [16:03] slashd, this is a small fix, maybe it could go out after like 5-6 days, then the 19.12 backport can enter -proposed next wednesday/thursday [16:19] rbalint, lgtm [16:25] now we just have to find someone to accept the small fix today sil2100 ;-) [16:26] Which small fix? ;) [16:27] I'm reviewing wslu for now [16:39] -queuebot:#ubuntu-release- Unapproved: accepted wslu [source] (eoan-proposed) [2.3.2-0ubuntu2~19.10.3] [16:41] -queuebot:#ubuntu-release- Unapproved: accepted wslu [source] (disco-proposed) [2.3.2-0ubuntu2~19.04.3] [16:43] can somewhat tell britney to ignore virtualbox/amd64 which is blocking libvncserver to migrate? it's a kernel/virtualbox issue, not due to the vnclib [16:43] -queuebot:#ubuntu-release- Unapproved: accepted wslu [source] (bionic-proposed) [2.3.2-0ubuntu2~18.04.3] [16:43] sil2100, the landscape-client one in eoan [16:47] -queuebot:#ubuntu-release- Unapproved: accepted landscape-client [source] (eoan-proposed) [18.01-0ubuntu9.1] [16:58] -queuebot:#ubuntu-release- Unapproved: accepted wslu [source] (xenial-proposed) [2.3.2-0ubuntu2~16.04.3] [17:07] vorlon, LocutusOfBorg: where are we at on getting virtualbox 6.0.14-dfsg-4 to migrate for focal? We have a 5.4 kernel in -proposed now [18:29] sforshee, one single test failure [18:37] sforshee, https://launchpad.net/ubuntu/+source/kopanocore [18:38] python3 /usr/sbin/kopano-search --help --> works [18:38] /usr/sbin/kopano-search --help --> doesn't work [18:38] this is the reason for the test failure [18:38] [pid 3172] openat(AT_FDCWD, "/sbin/ldconfig", O_RDONLY) = -1 EACCES (Permission denied) [18:38] and this is what actually happens with strace [18:39] so, I don't know why python3 behaves in such way, if launched with the ./ or python3 interpreter [18:42] of course, under sbuild it fails, under pbuilder it works -.-' [18:43] unless somebody has a better idea, I don't have anything more to do for debugging [18:55] tjaalton: nocheck> that would not generally be correct, no [19:06] -queuebot:#ubuntu-release- Unapproved: tomcat-native (bionic-proposed/universe) [1.2.21-1~18.04.1 => 1.2.21-1~18.04.1build1] (no packageset) [19:10] tdaitx you're the java guy, right? can you review this upload to make sure i got the facts right? it's a simple no-change rebuild, but i don't have experience with tomcat or static JNI libs so would be good to confirm the explanation of why the rebuild is needed is correct [19:10] by 'this upload' i meant tomcat-native ^ [19:10] from lp #1854072 [19:10] Launchpad bug 1854072 in tomcat-native (Ubuntu Bionic) "tomcat-native needs recompile to use TLSv1.3 from openssl 1.1" [Medium,In progress] https://launchpad.net/bugs/1854072 [19:11] locutus_: soooo kopanocore 8.7.0-5 introduces changes related to apparmor profiles, isn't this probably a regression caused by that? [19:12] locutus_: or possibly 8.7.0-4 [20:15] vorlon: then I don't see other way than adding libdmx and libxres so xorg-server can build [20:18] tjaalton: that's probably the parsimonious solution right now [20:53] vorlon, you are a genius, this explains why calling with python3 works [21:14] -queuebot:#ubuntu-release- Unapproved: accepted linux-firmware [source] (eoan-proposed) [1.183.3] [21:18] -queuebot:#ubuntu-release- Unapproved: accepted linux-firmware [source] (disco-proposed) [1.178.7] [21:21] -queuebot:#ubuntu-release- Unapproved: accepted linux-firmware [source] (bionic-proposed) [1.173.14] [21:36] -queuebot:#ubuntu-release- Unapproved: nova (eoan-proposed/main) [2:20.0.0-0ubuntu1 => 2:20.0.1-0ubuntu1] (openstack, ubuntu-server) [21:46] vorlon: binutils now ftbfs on i386, same as the autopkg test. I don't like investiing time on "fixing" this in the toolchain, so please could you restore the cross compiler packages? [22:01] doko: link to this build failure? [22:01] LocutusOfBorg: what I don't know is why intrigieri didn't run into this in Debian [22:03] doko: because the autopkgtest failure is due to failing to resolve dependencies on python3* cross-architecture, so that's not going to be the same in a ftbfs in launchpad [22:09] https://launchpad.net/~doko/+archive/ubuntu/toolchain/+build/18241328 [22:12] doko: so I see gcc-9-cross* still in the archive, which package are you missing? [22:13] and gcc-defaults is certainly in the archive [22:14] doko: so it's quite possible that something needs changing in the archive, but it's not obvious to me what that is, can you dig into it and let me know? [22:18] vorlon: as I said, it's the missing cross compilers [22:24] what cross compilers are missing? [22:24] I just told you gcc-9-cross* are still in the archive [22:24] "the cross compilers" is not a package name I can fix [22:27] sbuild-build-depends-binutils-dummy : Depends: g++-aarch64-linux-gnu but it is not going to be installed [22:27] Depends: g++-arm-linux-gnueabi but it is not going to be installed [22:28] [...] [22:28] $ rmadison -s focal -a i386 g++-aarch64-linux-gnu [22:28] g++-aarch64-linux-gnu | 4:9.2.1-3.1ubuntu1 | focal | i386 [22:28] $ [22:28] your move. [22:29] apparently the binutils 2.33.1-5 built, so what changed after that? [22:30] anyway, afk now [22:36] vorlon, apparmor is not installed in debian autopkgtests? [22:36] LocutusOfBorg: but somebody should have been testing while preparing the changes to the apparmor profile [22:37] LocutusOfBorg: anyway, I found the answer, it's because /usr/sbin/ldconfig is a wrapper *script* in Ubuntu and not in Debian [22:38] so it needs r permission on ldconfig because it's a script, and an additional x permission on ldconfig.real [22:40] /{,usr/}sbin/ldconfig.real rx, ? [22:46] Pixr rather [22:46] (to retain the other semantics) [22:46] and I merged it all on one line with ldconfig [22:53] seb128: heh I just beat you to doing https://code.launchpad.net/~seb128/britney/gtk-version-updated/+merge/376734 [22:53] vorlon, yeah, I saw, I deleted my mp when I saw the conflict markers on the generated diff :) [22:54] vorlon, can you review https://code.launchpad.net/~seb128/britney/distroinfo-i386-badtest/+merge/376689 ? [22:55] seb128: was next on my list - done (but I moved the hint up with the rest of the ones related to python modules) [22:55] vorlon, k, thanks! [22:56] hmm you know this one might be fixable [22:56] because the python module is arch: all [22:57] though really a better solution is to make it skippable when doing cross-testing [22:57] so I'll leave it for now [23:01] k [23:23] -queuebot:#ubuntu-release- New binary: python-pbcommand [amd64] (focal-proposed/universe) [1.1.1+git20191122.ec024c3-1] (no packageset)