2024-04-01 13:10:33 algitbot: retry master 2024-04-01 14:02:40 algitbot: retry master 2024-04-01 20:21:18 spdk passed in gilab ci two days ago 2024-04-01 20:24:23 >>> ERROR: spdk*: Has /home/... in rpath 2024-04-01 20:24:25 hmm 2024-04-01 20:30:50 my mainframe is overheated and I'm currently running on reduced capacity, so I'm currently of not much help to figure that one out 2024-04-01 20:36:04 hmm, locally I get 'undefined reference to `crc64_rocksoft_refl'' 2024-04-01 20:41:23 I made a comment in !63106, that was low-effort enough 2024-04-02 01:29:22 algitbot: retry master 2024-04-02 01:54:31 algitbot: retry master 2024-04-02 02:40:47 algitbot: retry master 2024-04-02 02:46:50 algitbot: retry master 2024-04-02 02:48:54 algitbot: retry master 2024-04-02 02:55:45 algitbot: retry master 2024-04-02 02:58:02 algitbot: retry master 2024-04-02 03:28:55 algitbot: retry master 2024-04-02 03:29:25 algitbot: retry master 2024-04-02 03:51:54 algitbot: retry master 2024-04-02 04:52:20 algitbot: retry master 2024-04-02 04:54:30 algitbot: retry master 2024-04-02 06:25:13 algitbot: retry master 2024-04-02 07:37:09 algitbot: retry master 2024-04-02 07:42:48 \o/ 2024-04-02 07:48:12 algitbot: retry master 2024-04-02 07:58:45 now back to my fever dreams 2024-04-02 09:11:35 algitbot: retry master 2024-04-02 09:45:25 algitbot: retry master 2024-04-02 20:02:43 algitbot: retry master 2024-04-02 20:22:02 ikke: ^? 2024-04-02 20:22:07 bubblewrap-0.9.0-r0: 2024-04-02 20:22:07 ERROR: unable to select packages: 2024-04-02 20:22:07 breaks: world[!bubblewrap] 2024-04-02 20:22:19 probably left over from fixing 2024-04-02 20:22:49 algitbot: retry master 2024-04-03 10:11:21 Why did upstream delete the tag :/ 2024-04-03 10:13:00 The release is still available at https://adishatz.org/lollypop/ but do we want to switch to that... 2024-04-03 10:14:32 Based on https://gitlab.gnome.org/gnumdk the tag has already been deleted twice 2024-04-03 10:35:59 celie: looks like better to pin it to commit instead of tag 2024-04-03 14:04:28 algitbot: retry master 2024-04-03 14:06:09 I think maybe we should just revert the lollypop upgrade, since upstream has removed the tag 2024-04-03 14:08:02 and has apparently forced pushed to master (d8ae916f vs da84abfa for the "data: Fix appdata file for flathub" commit) 2024-04-03 14:10:53 So, without the tag, we can't tell if upstream wants da84abfa to be 1.4.38, or if that will later on be amended again 2024-04-03 14:15:55 done 2024-04-03 14:19:00 Thanks 2024-04-03 14:52:44 algitbot: retry master 2024-04-03 15:39:52 g++: fatal error: Killed signal terminated program cc1plus 2024-04-03 15:39:53 wheeee 2024-04-03 15:39:56 algitbot: retry master 2024-04-03 16:09:38 algitbot: retry master 2024-04-03 16:12:24 algitbot: retry master 2024-04-03 16:51:16 algitbot: retry master 2024-04-03 17:04:23 algitbot: retry master 2024-04-03 22:42:33 algitbot: retry master 2024-04-04 20:23:19 algitbot: retry master 2024-04-05 01:34:17 algitbot: retry master 2024-04-05 01:55:00 algitbot: retry master 2024-04-05 02:07:44 I'm about to crasch but I tried this !63535 2024-04-05 02:09:16 maybe that dependency should be added elsewhere and not just to that aport, idk 2024-04-05 05:13:39 I have !63546 with the error message i get immediately upon running gdbus-codegen in the commit message 2024-04-05 05:14:27 The aarch64 CI is tied up somewhere else though 2024-04-05 05:15:57 but $depends_dev is added to makedepends anyway, so it's only a dependency addition for glib-dev, and shouldn't cause CI to fail 2024-04-05 05:28:55 it's building electron, but apparently just handling one job 2024-04-05 23:56:07 win 32 2024-04-06 01:31:22 algitbot: retry master 2024-04-06 10:07:44 oopsi, typo in the commit message (it's a 1.21.9 rebuild ofc) 2024-04-06 10:07:53 happens 2024-04-06 10:08:54 will do the Go security upgrade on edge shortly as well, but probably want to do the ghc upgrade before that 2024-04-06 11:01:21 algitbot: retry master 2024-04-06 11:21:44 algitbot: retry master 2024-04-06 11:33:01 algitbot: retry master 2024-04-06 11:44:35 algitbot: retry master 2024-04-06 11:51:02 algitbot: retry master 2024-04-06 12:11:18 algitbot: retry 3.19-stable 2024-04-06 12:16:49 algitbot: retry master 2024-04-06 12:42:40 algitbot: retry master 2024-04-06 13:40:42 algitbot: retry master 2024-04-06 13:57:11 should we consider disabling the git-lfs testsuite at some point? I think there hasn't been a single Go rebuild where it hasn't caused failures 2024-04-06 14:47:55 !63644 2024-04-06 15:08:25 git-lfs tests are already disabled for x86, why that one succeeded 2024-04-06 15:09:16 algitbot: retry master 2024-04-06 15:13:45 algitbot: retry master 2024-04-06 15:30:25 algitbot: retry master 2024-04-06 15:36:03 algitbot: retry 3.19-stable 2024-04-06 16:10:45 algitbot: retry master 2024-04-06 16:13:54 algitbot: retry master 2024-04-06 16:16:14 algitbot: retry master 2024-04-06 16:23:45 algitbot: retry master 2024-04-06 18:59:18 algitbot: retry master 2024-04-06 19:01:15 algitbot: retry 3.19-stable 2024-04-06 19:04:28 algitbot: retry master 2024-04-06 19:25:42 algitbot: retry master 2024-04-06 19:35:47 algitbot: retry master 2024-04-06 19:48:59 what's up with raft? 2024-04-06 19:49:30 what about it? 2024-04-06 19:49:40 so:libraft.so.3 (no such package): 2024-04-06 19:49:40 required by: dqlite-1.16.0-r0[so:libraft.so.3] 2024-04-06 19:49:40 ERROR: unable to select packages: 2024-04-06 19:49:41 hm 2024-04-06 19:49:56 https://pkgs.alpinelinux.org/contents?file=libraft.so.3&path=&name=&branch=edge 2024-04-06 19:50:16 is it the wrong repo? as in, community vs testing 2024-04-06 19:50:28 nah, raft is in community, nvm 2024-04-06 19:56:09 p:so:libraft.so.0=0.0.0 2024-04-06 19:56:29 seems the soname is incorrect 2024-04-06 19:59:26 SODIFF: 2024-04-06 19:59:27 -usr/lib/libraft.so.3.0.0: SONAME libraft.so.3 2024-04-06 19:59:29 +usr/lib/libraft.so.0.0.0: SONAME libraft.so.0 2024-04-06 19:59:33 https://gitlab.alpinelinux.org/alpine/aports/-/jobs/1332967 2024-04-06 20:00:33 oh, is that something recent? 2024-04-06 20:00:44 my index shows so:libraft.so.3 2024-04-06 20:01:11 Was merged 6h ago 2024-04-06 20:03:13 libraft_la_LDFLAGS = -version-info 0:0:0 2024-04-06 20:05:53 Previous version had: 2024-04-06 20:05:54 libraft_la_LDFLAGS = -version-info 3:0:0 2024-04-06 20:06:38 ah, switched from canonical to cowsql 2024-04-06 20:08:02 So we have now raft and raft-cowsql, both using the same source 2024-04-06 20:22:41 algitbot: retry master 2024-04-06 20:32:18 fcolista: ^ 2024-04-06 20:32:24 every downgrade needs a pkgrel update 2024-04-06 20:32:44 because otherwise we can get old packages on the mirror ( matching by path/filename ), but new checksums in the indexews 2024-04-06 20:32:46 indexes* 2024-04-06 20:32:49 and it fails to install 2024-04-06 20:33:48 specifically this counts for dl-cdn 2024-04-06 20:34:15 also no fan of revert commits that look like that 2024-04-06 20:34:58 should rather be community/raft: something something 2024-04-06 20:36:02 and ideally tell why it was reverted 2024-04-06 20:38:52 algitbot: retry master 2024-04-06 20:41:14 and !60867 2024-04-06 20:51:57 algitbot: retry master 2024-04-06 20:58:22 algitbot: retry master 2024-04-06 21:18:40 next first of april I will commit community/you-get: BEES!!! 2024-04-06 21:19:25 lol 2024-04-06 21:19:52 will the package actually install bees though 2024-04-06 21:20:03 reminds me 2024-04-06 21:20:06 there was a kernel module 2024-04-06 21:20:09 that was just bees 2024-04-06 21:20:18 but the github page 404s now :/ 2024-04-06 21:20:52 http://web.archive.org/web/20231212094632/https://github.com/driedsquids/bees 2024-04-06 21:22:27 was it upstreamed? 2024-04-06 21:22:49 mainlined, I mean 2024-04-06 21:23:55 oh i wish 2024-04-06 21:33:58 oops. the libyang upgrade was actually a soname bump as well 2024-04-06 21:34:04 and frr doesn't build against it 2024-04-06 21:34:05 sigh 2024-04-06 23:47:43 algitbot: retry master 2024-04-06 23:56:00 algitbot: retry master 2024-04-06 23:58:00 algitbot: retry master 2024-04-07 00:01:02 algitbot: retry master 2024-04-07 00:09:45 dammit! 2024-04-07 00:44:25 algitbot: retry master 2024-04-07 01:33:37 algitbot: retry master 2024-04-07 01:48:14 algitbot: retry master 2024-04-07 01:53:47 Did any of my aports get caught in the Go rebuild? (if you can remember, that is) 2024-04-07 01:58:22 don't think so 2024-04-07 03:27:02 algitbot: retry master 2024-04-07 03:29:29 Wow, lots of downgrades today, raft, py3-astroid, and (very likely) libyang 2024-04-07 04:11:48 algitbot: retry master 2024-04-07 05:40:13 algitbot: retry master 2024-04-07 09:06:14 algitbot: retry master 2024-04-07 12:28:17 ptrc: i found a patch from upstream for the urlwatch Py 3.12 failure: !63687 2024-04-07 12:29:32 seems to work indeed, thanks <3 2024-04-07 12:30:58 You're welcome 2024-04-07 13:10:19 ptrc: Another 2 check failures fixed (both use a similar technique): !63699 and !63700 (but the second one also has a different solution in !63347) 2024-04-07 13:11:36 i'm not sure which one is better out of the latter 2024-04-07 13:11:47 running tests out of build/ seems a bit cursed to me 2024-04-07 13:12:20 Yeah, perhaps that's why that MR hasn't been merged so far 2024-04-07 16:54:24 algitbot: retry master 2024-04-08 05:16:30 fun, checksum failure 2024-04-08 05:25:22 algitbot: retry master 2024-04-08 16:18:46 algitbot: retry 3.18-stable 2024-04-08 16:21:24 algitbot: retry 3.19-stable 2024-04-08 16:34:42 algitbot: retry master 2024-04-08 19:54:24 test 1234 2024-04-09 01:47:23 algitbot: retry master 2024-04-09 06:13:40 riscv64 is building frr against libyang 2.2.8 2024-04-09 06:13:52 algitbot: retry master 2024-04-09 06:33:07 algitbot: retry master 2024-04-09 06:33:10 !63808 2024-04-09 09:37:43 Can i get !63808 so it doesn't block the riscv64 builder? 2024-04-09 10:11:45 algitbot: retry master 2024-04-09 11:59:32 I think another aport that is having test failures on riscv64 is esbuild 2024-04-09 12:25:55 ptrc: ^ Maybe you'd like to see if that fixes the build on Python 3.12 2024-04-09 12:26:05 oki, will try 2024-04-09 12:26:23 I tried the old version on Python 3.11, and got the same error, which does not occur with the new version 2024-04-09 12:31:15 celie: seems like the same issue 2024-04-09 12:32:38 That's too bad, i was hoping having it at the same version as sqlite itself would solve the issue 2024-04-09 12:32:57 Tell me when you upload new logs, so i can go have a look 2024-04-09 12:35:57 uploaded 2024-04-09 12:36:06 but iirc it looks the same as the previous one 2024-04-09 12:37:26 I still see apsw 3.44.0.0 2024-04-09 12:37:46 oh, oops 2024-04-09 12:37:48 one sec 2024-04-09 12:40:13 oh, it worked! 2024-04-09 12:40:15 nice, thanks :3 2024-04-09 12:40:44 You're welcome 2024-04-09 12:45:23 aaand having apsw should also make calibre build, i think 2024-04-09 12:45:57 :) 2024-04-09 12:58:40 algitbot: retry master 2024-04-09 19:28:17 algitbot: retry master 2024-04-09 19:35:15 algitbot: retry master 2024-04-09 20:07:08 algitbot: retry master 2024-04-09 21:59:24 algitbot: retry master 2024-04-10 05:18:28 algitbot: retry master 2024-04-10 07:50:03 algitbot: retry master 2024-04-10 16:59:02 algitbot: kick master 2024-04-10 16:59:12 how does this work 2024-04-10 16:59:20 algitbot: kick build-edge-x86 2024-04-10 16:59:45 it seems stuck on testing/cherrytree 2024-04-10 17:01:14 I think kicking only works on riscv64 2024-04-10 17:01:21 at least that's what i heard the last time i tried it 2024-04-10 19:18:49 https://en.wikipedia.org/wiki/Can_I_Kick_It? 2024-04-10 22:55:37 algitbot: retry master 2024-04-10 23:01:18 algitbot: retry master 2024-04-10 23:06:10 algitbot: retry master 2024-04-10 23:14:13 algitbot: retry master 2024-04-11 00:03:35 algitbot: retry master 2024-04-11 00:05:02 algitbot: retry master 2024-04-11 07:59:27 algitbot: retry master 2024-04-11 08:01:34 algitbot: retry master 2024-04-11 08:05:26 strange OK in CI: https://gitlab.alpinelinux.org/alpine/aports/-/pipelines/225499 2024-04-11 09:05:06 algitbot: retry master 2024-04-11 12:40:12 algitbot: retry master 2024-04-11 13:03:26 algitbot: retry master 2024-04-11 13:51:22 algitbot: retry master 2024-04-11 15:41:23 algitbot: retry master 2024-04-11 15:51:37 algitbot: retry master 2024-04-11 20:04:46 🤔 2024-04-11 20:05:35 PureTryOut: a backport 2024-04-11 20:07:07 I get that, but I don't see why. 3.18 community hasn't been supported for quite a while 2024-04-11 20:17:55 it's a simple fix 2024-04-11 22:48:31 algitbot: retry master 2024-04-11 23:06:33 algitbot: retry master 2024-04-11 23:34:42 algitbot: retry master 2024-04-11 23:38:42 algitbot: retry master 2024-04-11 23:39:49 algitbot: retry master 2024-04-11 23:40:14 algitbot: retry master 2024-04-11 23:43:47 algitbot: retry master 2024-04-11 23:44:46 algitbot: retry master 2024-04-11 23:48:46 algitbot: retry master 2024-04-11 23:49:44 algitbot: retry master 2024-04-11 23:53:34 algitbot: retry master 2024-04-12 00:07:42 shouldn't take more than four hours to build 2024-04-12 00:24:46 algitbot: retry master 2024-04-12 07:14:17 algitbot: retry master 2024-04-12 10:40:39 algitbot: retry master 2024-04-12 11:30:59 algitbot: retry master 2024-04-12 11:44:41 algitbot: retry master 2024-04-12 11:57:34 algitbot: retry master 2024-04-12 12:16:24 algitbot: retry master 2024-04-12 12:17:58 algitbot: retry master 2024-04-12 12:19:18 algitbot: retry master 2024-04-12 12:23:10 algitbot: retry master 2024-04-12 12:28:36 algitbot: retry master 2024-04-12 12:30:29 algitbot: retry master 2024-04-12 12:35:56 algitbot: retry master 2024-04-12 12:37:23 algitbot: retry master 2024-04-12 12:41:07 algitbot: retry master 2024-04-12 13:00:09 algitbot: retry master 2024-04-12 13:15:46 distutils 2024-04-12 13:19:18 !63984 2024-04-12 13:24:38 *sigh* 2024-04-12 13:25:06 Maybe !63984 could go in as CI is unreliable now that it is still using Py 3.11 2024-04-12 13:25:44 Thanks 2024-04-12 13:25:54 The ppc64le thing is failing on a diferent test 2024-04-12 13:25:58 or tests, rather 2024-04-12 13:26:18 yes 2024-04-12 13:26:32 let's see if they're just flakey 2024-04-12 13:27:00 there were a lot of disabled tests 2024-04-12 13:28:52 From 3 to 2 tests failed 2024-04-12 13:28:54 That's an improvement 2024-04-12 13:28:54 one less failed 2024-04-12 13:29:01 algitbot: retry master 2024-04-12 13:29:35 Uh oh, aarch64 has 5 2024-04-12 13:29:58 I guess some of those were the ones you just enabled (and succeeded in CI) 2024-04-12 13:30:06 mhm, three of them the same as the first run on ppc64le 2024-04-12 13:30:51 so, depending on how this goes, maybe disable lsfd/mkfds-netlink-protocol and lsfd/option-inet and possibly lsfd/mkfds-multiplexing 2024-04-12 13:30:57 algitbot: retry master 2024-04-12 13:31:48 only two again on ppc64le 2024-04-12 13:32:33 let's see armv7 2024-04-12 13:33:49 algitbot: retry master 2024-04-12 13:35:32 I want to see on more architectures before I start disabling tests again 2024-04-12 13:36:39 algitbot: retry master 2024-04-12 13:40:11 I wonder if the builder has something in /usr/lib/python3.11 2024-04-12 13:40:42 algitbot: retry master 2024-04-12 13:41:56 For sanlock: !63985 2024-04-12 13:42:54 algitbot: retry master 2024-04-12 13:44:35 cely: why is it on python 3.11? 2024-04-12 13:45:06 algitbot: retry master 2024-04-12 13:46:31 x86_64 passsed building old util-linux 2024-04-12 13:46:44 algitbot: retry master 2024-04-12 13:49:24 algitbot: retry master 2024-04-12 13:49:38 !63986 2024-04-12 13:50:46 Apparently, i was a bit too quick with avahi, and didn't notice that the problem was one level higher (but probably it's still ok, as i grepped avahi source and found references to "distutils" too, just probably protected by an if) 2024-04-12 13:50:54 !63986 2024-04-12 13:50:56 oups 2024-04-12 13:50:59 !63988 2024-04-12 13:50:59 algitbot: retry master 2024-04-12 13:53:28 algitbot: retry master 2024-04-12 13:56:23 algitbot: retry master 2024-04-12 13:56:41 can they build something else? 2024-04-12 13:57:18 I want the others to catch up and fail to build util-linux, or not 2024-04-12 13:57:45 algitbot: retry master 2024-04-12 13:59:14 algitbot: retry master 2024-04-12 13:59:57 Apparently, util-linux is very important 2024-04-12 14:02:36 algitbot: retry master 2024-04-12 14:04:00 algitbot: retry master 2024-04-12 14:05:15 algitbot: retry master 2024-04-12 14:06:36 algitbot: retry master 2024-04-12 14:10:00 algitbot: retry master 2024-04-12 14:10:44 I disabled the failing and the flakey tests for a* and ppc64le 2024-04-12 14:10:59 Btw, Python 3.12.3 is out 2024-04-12 14:11:08 :D 2024-04-12 14:11:20 algitbot: retry master 2024-04-12 14:11:38 any failure on those now should pass on a re-run 2024-04-12 14:11:43 Oh, and it has a CVE 2024-04-12 14:11:47 and we'll see when the others catch up 2024-04-12 14:11:49 ouch 2024-04-12 14:12:17 ncopa: 2024-04-12 14:12:39 Oh but 2024-04-12 14:12:49 It is already available as a patch 2024-04-12 14:13:03 fix-test_xml_etree-expat-2.6.0.patch 2024-04-12 14:13:08 that we have 2024-04-12 14:13:16 ah, ok 2024-04-12 14:13:53 algitbot: retry mastere 2024-04-12 14:14:04 rlye? 2024-04-12 14:14:05 algitbot: retry master 2024-04-12 14:14:31 algitbot: retry maestro 2024-04-12 14:14:36 wat 2024-04-12 14:14:43 but you know mastere 2024-04-12 14:15:08 I guess the dependency chain is util-linux, glib, fontconfig 2024-04-12 14:15:27 algitbot: give omni the branch 2024-04-12 14:15:55 Hmm, i thought it responded to such things 2024-04-12 14:15:57 algitbot: give me 2024-04-12 14:16:03 algitbot: give me the branch 2024-04-12 14:16:45 algitbot: maybe you can 2024-04-12 14:17:50 Anyway gobject-introspection (without py3-setuptools in depends_dev) has built, so let's see if there are any more distutils errors 2024-04-12 14:19:45 lol 2024-04-12 14:20:12 I just looked at git's APKBUILD, and i see "NO_PYTHON=YesPlease" and "NO_TCLTK=YesPlease" 2024-04-12 14:20:35 I'll afk for a bit 2024-04-12 14:20:51 Bye 2024-04-12 14:22:04 So, sanlock passes on ppc64le with the absolute path, but not on s390x 2024-04-12 14:22:35 Hopefully, whatever Python 3.11 thing on the s390x builder is not going to cause problems when it is upgraded to 3.12 2024-04-12 14:23:31 algitbot: retry master 2024-04-12 14:30:28 Hmmm 2024-04-12 14:30:28 algitbot: retry master 2024-04-12 14:33:17 algitbot: retry master 2024-04-12 14:34:07 Oh great 2024-04-12 14:34:26 algitbot: retry master 2024-04-12 15:01:36 algitbot: retry master 2024-04-12 15:03:50 algitbot: retry master 2024-04-12 15:07:08 algitbot: retry master 2024-04-12 15:08:12 algitbot: retry master 2024-04-12 15:08:48 algitbot: retry master 2024-04-12 15:09:05 algitbot: retry master 2024-04-12 15:09:33 algitbot: retry master 2024-04-12 15:10:24 algitbot: retry master 2024-04-12 15:10:55 omni: i think util-linux is still failing on s390x 2024-04-12 15:10:56 algitbot: retry master 2024-04-12 15:11:24 algitbot: retry master 2024-04-12 15:12:29 Probably x86 as well 2024-04-12 15:12:31 algitbot: retry master 2024-04-12 15:13:21 algitbot: retry master 2024-04-12 15:14:13 algitbot: retry master 2024-04-12 15:15:21 algitbot: retry master 2024-04-12 15:15:29 algitbot: retry master 2024-04-12 15:17:43 algitbot: retry master 2024-04-12 15:17:56 algitbot: retry master 2024-04-12 15:19:21 algitbot: retry master 2024-04-12 15:20:01 algitbot: retry master 2024-04-12 15:20:09 algitbot: retry master 2024-04-12 15:21:35 algitbot: retry master 2024-04-12 15:22:23 algitbot: retry master 2024-04-12 15:22:34 algitbot: retry master 2024-04-12 15:24:06 algitbot: retry master 2024-04-12 15:25:09 algitbot: retry master 2024-04-12 15:26:09 algitbot: retry master 2024-04-12 15:26:47 Thanks andypost:matrix.org :) 2024-04-12 15:26:54 algitbot: retry master 2024-04-12 15:27:07 Now i get to see a little bit of community before i go afk 2024-04-12 15:28:08 celeste: thank you a lot! 2024-04-12 15:28:30 algitbot: retry master 2024-04-12 15:29:30 algitbot: retry master 2024-04-12 15:30:28 util-linux probably needs more tests disabled, omni worked on disabling some before going afk (that work has been merged, i think) 2024-04-12 15:30:50 1500+ Python aports in community! This will be interesting.. 2024-04-12 15:31:23 algitbot: retry master 2024-04-12 15:32:02 algitbot: retry master 2024-04-12 15:33:45 I wonder if lsfd/mkfds-netlink-protocol and lsfd/option-inet would need options="net", could that be it? 2024-04-12 15:34:07 No, the builders do not run in rootbld mode yet 2024-04-12 15:34:20 ok 2024-04-12 15:34:41 can x86_64 just get to it already? 2024-04-12 15:35:50 !63995 i would do a gpep517 but CI is unreliable now, so better be safe 2024-04-12 15:36:50 I guess all Python aports will need an explicit py3-setuptools now (unless they use a different backend like flit-core, poetry-core, ...) 2024-04-12 15:37:55 oh, it's finaly done with chromium? 2024-04-12 15:38:04 I guess so 2024-04-12 15:38:07 that may speed things up, I hope 2024-04-12 15:38:13 It finished on edge just in time for Python 3.12 2024-04-12 15:38:57 algitbot: retry master 2024-04-12 15:39:58 algitbot: retry master 2024-04-12 15:40:05 algitbot: retry master 2024-04-12 15:40:19 Circular dep? 2024-04-12 15:40:27 !63996 2024-04-12 15:41:26 No 2024-04-12 15:41:28 algitbot: retry master 2024-04-12 15:42:19 algitbot: retry master 2024-04-12 15:43:18 algitbot: retry master 2024-04-12 15:43:55 !63997 2024-04-12 15:45:40 !63998 2024-04-12 15:45:54 Ok, that will be the last Py3 fix from me today, going afk soon 2024-04-12 15:48:22 algitbot: retry master 2024-04-12 15:50:40 algitbot: retry master 2024-04-12 15:52:53 ok 2024-04-12 15:53:33 algitbot: retry master 2024-04-12 15:56:29 algitbot: retry master 2024-04-12 16:02:20 algitbot: retry master 2024-04-12 16:06:51 algitbot: retry master 2024-04-12 16:09:56 algitbot: retry master 2024-04-12 16:09:59 This probably means the x86 and x86_64 builders have something in the absolute path /usr/lib/python3.11, hopefully that won't cause problems when they are upgraded to Py 3.12 2024-04-12 16:11:03 who can fix/check it? 2024-04-12 16:11:39 Probably ikke 2024-04-12 16:12:22 I'm guessing it could be something like fail2ban 2024-04-12 16:13:07 algitbot: retry master 2024-04-12 16:13:10 your fix wasn't in yet when sanlock failed again 2024-04-12 16:13:41 Yes, i know, but the absolute path matching something 2024-04-12 16:13:46 That means there is something there 2024-04-12 16:13:48 on the builders 2024-04-12 16:14:00 Some archs don't have it, and didn't need the fix 2024-04-12 16:14:03 algitbot: retry master 2024-04-12 16:14:24 algitbot: retry master 2024-04-12 16:15:04 algitbot: retry master 2024-04-12 16:15:29 Uh oh, hopefully this doesn't mean more tests need to be disabled 2024-04-12 16:16:09 no 2024-04-12 16:16:56 I disabled the four tests for all architectures but s390x had already been triggered to try again 2024-04-12 16:16:57 Oh no 2024-04-12 16:17:13 py3-asn1-modules wants py3-asn1<0.6.0 :| 2024-04-12 16:17:55 algitbot: retry master 2024-04-12 16:19:29 Probably needs an upgrade, hopefully that doesn't break something else 2024-04-12 16:20:51 :O 2024-04-12 16:23:34 algitbot: retry master 2024-04-12 16:25:16 !64006 2024-04-12 16:25:22 !64007 2024-04-12 16:25:28 Ok, going afk after this 2024-04-12 16:26:14 I suspect that switching to gpep517 would also silently ignore the version constraint on py3-asn1 2024-04-12 16:27:50 Yeah, my fastjsonschema bump can't succeed in CI 2024-04-12 16:29:14 algitbot: retry master 2024-04-12 16:32:33 algitbot: retry master 2024-04-12 16:34:28 Ok, got to go now, bye 2024-04-12 16:37:20 l8r 2024-04-12 16:42:42 algitbot: retry master 2024-04-12 16:45:32 aight, fixing time i gues 2024-04-12 16:46:42 oh, x86_64 hasn't finished main yet 2024-04-12 16:46:44 sad 2024-04-12 16:48:14 algitbot: retry master 2024-04-12 16:49:35 bumping py3-build to 1.2.1 makes it even worse :( 2024-04-12 16:51:36 algitbot: retry master 2024-04-12 16:52:16 algitbot: retry master 2024-04-12 16:53:01 worst part is that our deps for py3-build are all newer than minimums specified 2024-04-12 16:53:06 maybe too high? 2024-04-12 16:53:39 this could take all weekend 2024-04-12 16:55:00 this *will* take all weekend :p 2024-04-12 16:55:16 sure 2024-04-12 16:55:21 at least :P 2024-04-12 16:55:27 trying to figure out why py3-build tests fail 2024-04-12 16:55:35 algitbot: retry master 2024-04-12 16:55:37 and well 2024-04-12 16:55:47 it's not looking good 2024-04-12 16:56:20 feel free to take over !64010 2024-04-12 16:57:01 i'm currently looking at how gentoo does it 2024-04-12 16:57:15 and they use an ancient version of setuptools 2024-04-12 16:58:57 algitbot: retry master 2024-04-12 16:59:11 > Successfully built test_setuptools-1.tar.gz 2024-04-12 16:59:14 grrr 2024-04-12 16:59:54 algitbot: retry master 2024-04-12 16:59:54 voidlinux then? 2024-04-12 17:00:07 void as in, move all python packages to /dev/null? :p 2024-04-12 17:00:53 yes, but also linux, as I mentioned 2024-04-12 17:00:56 toy kernel 2024-04-12 17:01:46 voidlinux is at python3-setuptools 69.2.0 and python3-build 1.2.1 2024-04-12 17:01:47 algitbot: retry master 2024-04-12 17:02:45 and python3 3.12.3 2024-04-12 17:09:37 omni: moreover it's marked security https://docs.python.org/3/whatsnew/changelog.html#python-3-12-3-final 2024-04-12 17:10:41 yes, but we already have that patched, I was told 2024-04-12 17:11:54 https://git.alpinelinux.org/aports/tree/main/python3/fix-test_xml_etree-expat-2.6.0.patch 2024-04-12 17:12:45 algitbot: retry master 2024-04-12 17:16:42 algitbot: retry master 2024-04-12 17:19:34 algitbot: retry master 2024-04-12 17:20:29 onmi: !64013 2024-04-12 17:22:18 Just dropping by quickly to say, i didn't look at the patch too closely, now i think it's incomplete, and we still need the new release 2024-04-12 17:23:20 algitbot: retry master 2024-04-12 17:24:59 algitbot: retry master 2024-04-12 17:25:21 and there are quite a few other fixes 2024-04-12 17:26:26 also !63990 adds a version constraint on libexpat, probably not required for edge 2024-04-12 17:30:54 algitbot: retry master 2024-04-12 17:34:36 ptrc: is this something? https://git.alpinelinux.org/aports/tree/main/py3-setuptools/APKBUILD#n74 as in that it should be scripts-3.12 not if anything 2024-04-12 17:39:17 algitbot: retry master 2024-04-12 17:40:16 omni: looks like yep https://git.alpinelinux.org/aports/commit/main/py3-setuptools/APKBUILD?id=cc52fd6f1c23fc2e96d59c5aff4b6ce7d62fcf98 2024-04-12 17:47:55 algitbot: retry master 2024-04-12 17:50:06 algitbot: retry master 2024-04-12 17:52:39 eaa152c7e98041cda7987ddaecdcb45233a67843 2024-04-12 17:56:04 algitbot: retry master 2024-04-12 18:03:59 algitbot: retry master 2024-04-12 18:06:27 and filed #15978 2024-04-12 18:17:30 cely: https://tpaste.us/j4Ll 2024-04-12 18:17:33 algitbot: retry master 2024-04-12 18:18:20 2 pyc files not owned by any package 2024-04-12 18:25:24 let's see 2024-04-12 18:36:23 algitbot: retry master 2024-04-12 18:39:50 algitbot: retry master 2024-04-12 18:40:39 let's fail with a newer version 2024-04-12 18:43:40 algitbot: retry master 2024-04-12 18:44:02 algitbot: retry master 2024-04-12 18:46:11 algitbot: retry master 2024-04-12 18:49:35 at least it built maturin 2024-04-12 18:54:54 py3-faker dependency 2024-04-12 18:56:04 algitbot: retry master 2024-04-12 18:58:04 I'll deselect some py3-build tests 2024-04-12 19:00:16 algitbot: retry master 2024-04-12 19:02:53 algitbot: retry master 2024-04-12 19:06:10 I want to just try that before deselecting tests 2024-04-12 19:06:16 algitbot: retry master 2024-04-12 19:07:00 algitbot: retry master 2024-04-12 19:08:43 you people rock! 2024-04-12 19:10:20 algitbot: retry master 2024-04-12 19:14:52 let's see if this does it 2024-04-12 19:14:57 ncopa: :* 2024-04-12 19:15:10 i meeeean 2024-04-12 19:15:23 one of the tests was literally "we're expecting a README to be in the package but it's not" 2024-04-12 19:15:29 idunno if disabling them is such a good idea :p 2024-04-12 19:15:38 but then, we don't exactly use py3-build in aports either 2024-04-12 19:15:40 so let's go 2024-04-12 19:17:18 algitbot: retry master 2024-04-12 19:17:43 saw it in checkdepends in some aports 2024-04-12 19:18:04 algitbot: retry master 2024-04-12 19:19:18 algitbot: retry master 2024-04-12 19:19:33 at this point i'd just do `mosquitto_pub` directly into the retry topic 2024-04-12 19:24:50 algitbot: retry master 2024-04-12 19:36:50 algitbot: retry master 2024-04-12 19:40:52 i'll take this one if nobody wants it 2024-04-12 19:41:10 doesn't sound too bad 2024-04-12 19:41:48 algitbot: retry master 2024-04-12 19:47:20 algitbot: retry master 2024-04-12 19:54:52 algitbot: retry master 2024-04-12 20:00:10 the maturin one is kinda nasty 2024-04-12 20:00:14 > zipfile.BadZipFile: Bad CRC-32 for file '_cffi_backend.cpython-312-aarch64-linux-musl.so' 2024-04-12 20:00:17 really don't like it 2024-04-12 20:01:23 hold on, no 2024-04-12 20:01:25 Using cached cffi-1.16.0-cp312-cp312-linux_aarch64.whl 2024-04-12 20:01:25 Collecting cffi 2024-04-12 20:01:27 what are they doing 2024-04-12 20:01:51 where did they get that wheel huh 2024-04-12 20:06:05 algitbot: retry master 2024-04-12 20:07:37 algitbot: retry master 2024-04-12 20:14:18 algitbot: retry master 2024-04-12 20:15:35 algitbot: retry master 2024-04-12 20:16:25 algitbot: retry master 2024-04-12 20:17:34 ??? 2024-04-12 20:17:40 that's odd 2024-04-12 20:20:51 algitbot: retry master 2024-04-12 20:28:36 algitbot: retry master 2024-04-12 20:31:54 algitbot: retry master 2024-04-12 20:32:38 algitbot: retry master 2024-04-12 20:32:41 shut upp 2024-04-12 20:33:38 algitbot: retry master 2024-04-12 20:37:14 algitbot: retry master 2024-04-12 20:37:54 not sure where did the "just add python3-dev everywhere" trend start 2024-04-12 20:38:00 but it's really odd 2024-04-12 20:38:12 algitbot: retry master 2024-04-12 20:38:23 i'm just gonna set up a loop so it retries every 10 seconds 2024-04-12 20:38:30 algitbot: retry master 2024-04-12 20:40:36 why does it fail building mesa, darn 2024-04-12 20:41:25 at this point we'll just get x86_64 logspam 2024-04-12 20:42:00 algitbot: retry master 2024-04-12 20:42:58 maybe bump mako again 2024-04-12 20:43:17 sure let's try thta 2024-04-12 20:43:54 i wish i could just dump a package from the builders before they get uploaded 2024-04-12 20:43:58 s/dump/fetch/ 2024-04-12 20:44:49 nope 2024-04-12 20:44:51 didn't work 2024-04-12 20:45:58 https://ptrc.gay/FxHlXVmX 2024-04-12 20:45:59 oh, ffs 2024-04-12 20:46:16 obviously that's not gonna work without setuptools or packaging 2024-04-12 20:49:16 ptrc: doesit mean that other builders has this package installed? 2024-04-12 20:49:27 good question 2024-04-12 20:49:34 i don't think other builders had to build mesa? 2024-04-12 20:49:39 or i didn't see it 2024-04-12 20:49:54 otherwise not clear why all passedexcept 8664 2024-04-12 20:50:06 yeah, weird 2024-04-12 20:50:30 ah! 2024-04-12 20:50:35 mesa built everywhere else with python 3.11 2024-04-12 20:50:41 it's just that the x86_64 builder was busy 2024-04-12 20:50:52 and mesa itself doesn't use python3, so it didn't need a pkgrel bump to rebuild 2024-04-12 20:51:34 this one i have no idea why it fails 2024-04-12 20:51:53 works locally for me 2024-04-12 20:52:09 but then, let's see if it fails anywhere else 2024-04-12 20:55:55 ptrc: other arches passed already https://build.alpinelinux.org/buildlogs/build-edge-armv7/community/maturin/maturin-1.5.1-r0.log 2024-04-12 20:56:12 algitbot: retry master 2024-04-12 20:58:10 algitbot: retry master 2024-04-12 21:07:51 algitbot: retry master 2024-04-12 21:11:26 algitbot: retry master 2024-04-12 21:12:19 andypost[m]: i'm running a loop to retry the builders every 30 seconds, so you don't have to ping them every time :p 2024-04-12 21:24:07 alright, heading off to sleep, i really hope the builders don't get stuck on a single package for the entire night 2024-04-12 21:47:40 I don't get why it is failing on aarch64 2024-04-12 21:54:45 bad crc, hmm 2024-04-12 22:09:21 everything else I look at seem to just need something else to be built first 2024-04-12 22:28:48 obviously I don't look at everything ;) 2024-04-12 22:34:00 same issues with 1.4.0-r2 https://build.alpinelinux.org/buildlogs/build-edge-aarch64/community/maturin/maturin-1.4.0-r2.log 2024-04-13 01:01:05 https://github.com/python-cffi/cffi/commit/56f760955c85592b0594895ca626c29174fe8078 2024-04-13 01:34:08 hmmm.. https://git.alpinelinux.org/aports/commit/?id=6bc0c649e2f51429282d7401627e0fe205443665 2024-04-13 01:53:15 !64031 2024-04-13 07:49:27 ok so it seems nothing reached testing yet 2024-04-13 08:28:45 algitbot: kick build-edge-riscv64 2024-04-13 08:28:57 algitbot: kick build-edge-riscv64 2024-04-13 08:33:09 algitbot: kick riscv64 2024-04-13 08:33:14 algitbot: kick edge-riscv64 2024-04-13 08:33:25 algitbot: kick builder-edge-riscv64 2024-04-13 08:33:27 algitbot: kick build-edge-riscv64 2024-04-13 09:02:54 I wonder if we should just disable py3-pytest-tmp-files on armhf as the only thing that uses it, borg-space, is not enabled there as well 2024-04-13 09:17:51 celeste: pyfakefs needs upgrade for py3-pytest-tmp-files 2024-04-13 09:18:44 https://github.com/pytest-dev/pyfakefs/blob/main/CHANGES.md 2024-04-13 09:26:18 filed !64048 2024-04-13 09:47:12 andypost:matrix.org: Probably it needs py3-pandas in checkdepends, but our version of pandas is also quite old 2024-04-13 09:50:26 algitbot: retry master 2024-04-13 09:52:24 algitbot: retry master 2024-04-13 09:54:58 !64044 for py3-django-compression-middleware 2024-04-13 09:58:01 algitbot: retry master 2024-04-13 09:59:07 algitbot: retry master 2024-04-13 10:03:38 algitbot: retry master 2024-04-13 10:03:51 algitbot: retry master 2024-04-13 10:04:18 algitbot: retry master 2024-04-13 10:05:04 algitbot: retry master 2024-04-13 10:05:32 algitbot: retry master 2024-04-13 10:06:01 algitbot: retry master 2024-04-13 10:06:15 algitbot: retry master 2024-04-13 10:15:11 celeste: numpy also needs upgrade https://numpy.org/news/#numpy-1260-released 2024-04-13 10:16:09 Yeah, it will be a never ending upgrade chain 2024-04-13 10:16:43 and they moved to meson) 2024-04-13 10:17:05 Probably better to just disable py3-pytest-tmp-files on armhf and let it upload community (but there's also libpeas that's failing tests) 2024-04-13 10:17:24 it fails tests every where 2024-04-13 10:17:40 better just disable check 2024-04-13 10:18:32 I think to merge !64048 first 2024-04-13 10:20:17 Alright, then maybe after that tmp-files and libpeas can have checks disabled for the time being, so at least one arch will have community uploaded and we can have CI working for that again 2024-04-13 10:20:45 yes, that's the point 2024-04-13 10:21:00 ah it need pandas for tests 2024-04-13 10:22:07 and pandas needs new numpy? 2024-04-13 10:23:50 yes 2024-04-13 10:36:28 py3-fire needs upgrade !64051 2024-04-13 10:41:32 I think the py3-fire thing is a result of the case statement in check() 2024-04-13 10:42:34 algitbot: retry master 2024-04-13 10:59:45 Not sure if i should do such a selective rebuild, but i think !64052 will get the armhf builder idle again 2024-04-13 11:00:26 Let's see if the armhf CI agrees 2024-04-13 11:02:58 Argh, numpy errors again 2024-04-13 11:05:46 algitbot: retry master 2024-04-13 11:08:09 looks mechanize needs to disable ftp test 2024-04-13 11:09:35 Oops, it actually needs !64053 as well 2024-04-13 11:09:52 but now py3-nptyping is failing due to numpy 2024-04-13 11:16:13 filed !64054 2024-04-13 11:20:53 andypost:matrix.org: Try adding py3-meson-python to makedepends 2024-04-13 11:21:09 We have the latest meson-python, so hopefully it helps 2024-04-13 11:21:38 but i'm not so optimistic about tests though, probably they will still fail 2024-04-13 11:21:56 Ha! but tests are disabled for py3-numpy 2024-04-13 11:22:31 Maybe you can also remove py3-nose 2024-04-13 11:22:59 as i think that's for tests (so maybe checkdepends, but they're disabled anyway) 2024-04-13 11:23:54 The thing is, the other archs are complaining about py3-nose being built for 3.11, so with that removed, maybe the CI for other archs will run too 2024-04-13 11:24:16 Oh wait 2024-04-13 11:24:20 py3-meson-python is in community/ 2024-04-13 11:24:25 so they'll complain about that instead 2024-04-13 11:24:40 but at least with py3-meson-python the armhf CI will give some results 2024-04-13 11:27:08 algitbot: retry master 2024-04-13 11:35:46 "ERROR: Executables created by c compiler gcc are not runnable." (from armv7 CI of py3-numpy) 2024-04-13 11:36:56 Maybe you can try cat'ing the log so we can see what it says 2024-04-13 11:43:12 Ugh, i'm juggling so many containers i'm confusing myself 2024-04-13 11:56:18 Grr, the log isn't where meson said it would be 2024-04-13 12:03:05 no there's no log( just checked 2024-04-13 12:17:01 algitbot: retry master 2024-04-13 12:18:05 isn't ptrc stll having their script running retry every 30 seconds? 2024-04-13 12:18:18 Not anymore 2024-04-13 12:18:30 I think she stopped it when arm* was stuck on the last few aports for community/ 2024-04-13 12:18:35 ok 2024-04-13 12:19:36 algitbot: retry master 2024-04-13 12:21:36 Yes, arm* has reached testing/ 2024-04-13 12:22:02 but we had to do #15980 to get there 2024-04-13 12:27:15 algitbot: retry master 2024-04-13 12:27:59 To top it off, very likely, riscv64 is stuck on py3-pexpect, and x86* on py3-fakeredis 2024-04-13 12:28:11 There's been no movement there for hours 2024-04-13 12:28:12 algitbot: retry master 2024-04-13 12:34:50 andypost:matrix.org: How is your progress on py3-numpy? 2024-04-13 12:35:04 celeste: stuck atm 2024-04-13 12:35:55 Still at the executables created by compiler not runnable error? 2024-04-13 12:36:08 yep, and cant get why 2024-04-13 12:36:23 something with flags 2024-04-13 12:37:32 Try to remove export LDFLAGS="$LDFLAGS -shared" 2024-04-13 12:42:07 celeste: maybe just disable failed test for nptyping? 2024-04-13 12:43:46 I'm trying to if tests will pass with new numpy 2024-04-13 12:43:48 to see if* 2024-04-13 12:44:00 Numpy built after i removed LDFLAGS export 2024-04-13 12:45:34 but arm* passed in ci 2024-04-13 12:46:11 Which MR are you talking about? 2024-04-13 12:46:45 https://gitlab.alpinelinux.org/alpine/aports/-/jobs/1341401 2024-04-13 12:47:30 ah tests 2024-04-13 12:47:52 https://gitlab.alpinelinux.org/Celeste/aports/-/jobs/1341410 2024-04-13 12:47:58 nptyping tests still fail with new numpy 2024-04-13 12:49:10 https://github.com/ramonhagenaars/nptyping/pull/114 fixes deprecation but test fails 2024-04-13 12:49:10 and we have latest beartype and nptyping 2024-04-13 12:50:41 I guess we disable those tests then 2024-04-13 12:58:33 algitbot: retry master 2024-04-13 13:00:19 algitbot: retry master 2024-04-13 13:02:07 algitbot: retry master 2024-04-13 13:03:15 I know arm* can't build anything else as py3-sphinx-autodoc-typehints has py3-nptyping in checkdepends 2024-04-13 13:03:24 but does ppc64le have something else? 2024-04-13 13:03:25 algitbot: retry master 2024-04-13 13:04:39 algitbot: retry master 2024-04-13 13:07:18 or rather, s390x, that's what i'm thinking about, and i think no, py3-sport-activities-features has about 5 aports depending on it, plus itself, py3-geotiler, and py3-cytoolz 2024-04-13 13:07:32 That's probably what's left for s390x community 2024-04-13 13:07:37 algitbot: retry master 2024-04-13 13:09:09 algitbot: retry master 2024-04-13 13:10:10 algitbot: retry master 2024-04-13 13:11:00 algitbot: retry master 2024-04-13 13:12:05 py3-dbus-next passed on arm*, not sure why it's failing on ppc64le 2024-04-13 13:14:46 algitbot: retry master 2024-04-13 13:17:05 algitbot: retry master 2024-04-13 13:18:19 algitbot: retry master 2024-04-13 13:19:42 algitbot: retry master 2024-04-13 13:20:53 algitbot: retry master 2024-04-13 13:21:10 I guess there is nothing else for s390x to build 2024-04-13 13:21:21 So, will only retry when ppc64le or aarch64 fails 2024-04-13 13:21:24 Like now 2024-04-13 13:22:09 Hopefully, ppc64le can be the next to reach less than 10 aports left in community/ 2024-04-13 13:22:12 algitbot: retry master 2024-04-13 13:29:19 algitbot: retry master 2024-04-13 13:37:57 algitbot: retry master 2024-04-13 13:43:15 algitbot: retry master 2024-04-13 13:51:10 algitbot: retry master 2024-04-13 14:34:07 algitbot: retry master 2024-04-13 14:34:56 algitbot: retry master 2024-04-13 14:36:34 algitbot: retry master 2024-04-13 14:37:45 Hmmm, so ppc64le is down to the next aport 2024-04-13 14:37:48 in community 2024-04-13 14:37:50 algitbot: retry master 2024-04-13 14:38:22 and py3-dbus-next passed on arm*, so i wonder if it's an issue with the builder 2024-04-13 14:39:20 mhm 2024-04-13 14:42:29 unrelated, just read this https://github.com/pytest-dev/pytest-asyncio/issues/706 2024-04-13 14:43:05 so I guess that's why we stick to py3-pytest-asyncio 0.21 for now 2024-04-13 14:49:08 algitbot: retry master 2024-04-13 14:51:25 build-edge-x86 got unstuck, but build-edge-x86_64 is still stuck 2024-04-13 14:53:15 there you go 2024-04-13 14:53:17 algitbot: retry master 2024-04-13 14:53:30 <3 2024-04-13 14:57:22 !64054 succeeded on arm*, that have uploaded community/ aports, should we..? 2024-04-13 14:59:55 I think moving to meson is a big change 2024-04-13 15:00:22 So, only if it solves a failure somewhere else 2024-04-13 15:01:15 It doesn't move meson? 2024-04-13 15:02:31 it adds py3-meson-python to makedepends 2024-04-13 15:03:38 is it needed? 2024-04-13 15:03:48 for the upgrade, I mean 2024-04-13 15:04:31 Yes it is needed 2024-04-13 15:04:39 This new version moves the gpep517 backend to meson-python 2024-04-13 15:05:40 ok 2024-04-13 15:05:59 !64050 needs a numpy upgrade though 2024-04-13 15:07:39 I think we needed pandas for pytest-tmp-files 2024-04-13 15:08:42 No, it was needed for py3-pyfakefs !64048 2024-04-13 15:09:12 which andypost said was needed for pytest-tmp-files 2024-04-13 15:09:21 but we ended up just disabling tests 2024-04-13 15:09:55 Anyway, have your eyes on !64064 2024-04-13 15:10:19 If it passes i will remove the auxiliary commits (needed as community/ has not been uploaded on all archs) 2024-04-13 15:14:11 Only x86 left 2024-04-13 15:15:11 Ok, all passed 2024-04-13 15:15:17 Rebased, and undrafted 2024-04-13 15:15:35 Hopefully it will allow ppc64le to move on to testing/ 2024-04-13 15:17:43 omni: LDFLAGS causes meson to error out with "executable created by compiler is not runnable" 2024-04-13 15:18:06 Also, !64064 should allow ppc64le to upload community/ and move to testing/ 2024-04-13 15:26:32 ikke: riscv64 may also be stuck on py3-pexpect 2024-04-13 15:26:41 Usually that builds very quickly 2024-04-13 15:28:08 algitbot: kick build-edge-riscv64 2024-04-13 15:28:46 trying !64066 2024-04-13 15:29:14 You are very ambitious to attempt that :) 2024-04-13 15:31:02 but CI will not be of much help now 2024-04-13 15:31:31 \o/ 2024-04-13 15:36:24 pandas also needs meson (and i think it was at this point that lots of errors started appearing, the last time i tried), and pyfakefs needs pandas in checkdepends 2024-04-13 15:37:01 py3-python-meson, to be exact 2024-04-13 15:37:19 py3-meson-python* 2024-04-13 15:37:24 Confusing names :D 2024-04-13 15:39:18 a bunch of releases since yesterday https://github.com/pypa/setuptools/releases 2024-04-13 15:39:27 Please no 2024-04-13 15:39:54 https://setuptools.pypa.io/en/stable/history.html 2024-04-13 15:47:10 algitbot: retry master 2024-04-13 15:49:59 algitbot: retry master 2024-04-13 15:50:51 algitbot: retry master 2024-04-13 15:51:45 cely: don't you want a setuptools upgrade? 2024-04-13 15:52:24 algitbot: retry master 2024-04-13 15:52:30 Only if it doesn't break anything 2024-04-13 15:52:55 breaking things is cool 2024-04-13 15:53:05 Only if it doesn't break anything 2024-04-13 15:53:19 you gotta break things to have things to fix 2024-04-13 15:53:38 Oh, sorry for the repeated line, that was supposed to be a "retry master", haha 2024-04-13 15:53:41 algitbot: retry master 2024-04-13 15:53:46 :D 2024-04-13 15:54:04 maybe it needed repeating too 2024-04-13 15:55:05 algitbot: retry master 2024-04-13 15:55:30 I think we are at the end of this round of rebuild (i know testing hasn't gotten the global rebuild yet) 2024-04-13 15:55:54 s390x is stuck on community/py3-cytoolz, and arm*/ppc64le are stuck on testing/py3-nptyping 2024-04-13 15:56:17 and aarch64 on maturin... *sigh* 2024-04-13 15:56:38 Not to forget riscv64 on main/py3-pexpect 2024-04-13 15:59:31 oh? 2024-04-13 15:59:47 Oh @ riscv64? 2024-04-13 15:59:57 Yes, it's been at pexpect for around half a day 2024-04-13 16:00:01 algitbot: retry master 2024-04-13 16:00:19 ah, I thought it moved earlier 2024-04-13 16:01:49 it should be fairly quick to run, yeah? 2024-04-13 16:02:09 Yes, pexpect is usually fast 2024-04-13 16:03:52 ikke: can you have a look? 2024-04-13 16:08:43 algitbot: retry master 2024-04-13 16:10:39 algitbot: retry master 2024-04-13 16:11:27 Is it time to disable that py3-nptyping test? 2024-04-13 16:11:35 There are quite a number of them disabled already 2024-04-13 16:15:28 algitbot: retry master 2024-04-13 16:15:46 I wonder why aarch64 keeps going back to maturin 2024-04-13 16:16:42 a bunch of aports depend on it 2024-04-13 16:16:58 makedepend at least 2024-04-13 16:30:59 py3-cytoolz fail the same test on x86_64 and s390x 2024-04-13 16:33:10 Seems like it is not a new thing: https://github.com/pytoolz/cytoolz/issues/200 2024-04-13 16:36:34 algitbot: retry master 2024-04-13 16:38:02 !64072 2024-04-13 16:46:23 Maybe it will need to ignore the tests/test_typeguard.py::TypeguardTest::test_success as well 2024-04-13 16:46:38 See riscv64 CI 2024-04-13 16:47:40 but riscv64 is still using Py 3.11 2024-04-13 16:47:49 You could add a bump for py3-beartype (then remove it before merging) 2024-04-13 16:48:08 To see what happens to Py 3.12 on arm* and ppc64le 2024-04-13 16:48:17 s/to/with/ 2024-04-13 16:49:15 mhm, but gotta take a break soon 2024-04-13 16:49:30 No problem, i'm going afk as well 2024-04-13 16:49:44 !64074 2024-04-13 16:49:54 not sure why the test didn't fail for the other architectures 2024-04-13 16:50:56 I think i tried py3-toolz locally, and it failed 2024-04-13 16:51:35 algitbot: retry master 2024-04-13 16:51:51 I think you're deselecting the same test in both cases 2024-04-13 16:53:10 oh, thanks, that was unintentional 2024-04-13 16:53:12 there 2024-04-13 16:54:22 Looks good now 2024-04-13 16:54:54 Yes, it's failing on test_success now 2024-04-13 16:55:02 So, you'll probably have to deselect that as well 2024-04-13 16:55:22 we'll see 2024-04-13 16:55:29 they can be improved later 2024-04-13 16:55:40 algitbot: retry master 2024-04-13 16:56:26 We have the latest typeguard 2024-04-13 16:57:28 and it's failing on riscv64 CI with Py 3.11 anyway, so probably not caused by 3.12 2024-04-13 16:57:56 Ok, s390x is uploading community now 2024-04-13 16:58:02 \o/ 2024-04-13 16:58:05 I've got to go, bye 2024-04-13 16:58:05 at least something 2024-04-13 16:58:16 yeah, me too, a couple of hours ago 2024-04-13 17:05:08 afk *poff* 2024-04-13 19:17:07 algitbot: retry master 2024-04-13 19:22:00 algitbot: retry master 2024-04-13 20:54:04 algitbot: retry master 2024-04-13 21:06:02 algitbot: retry master 2024-04-13 23:00:30 algitbot: retry master 2024-04-13 23:41:21 algitbot: retry master 2024-04-13 23:54:00 algitbot: retry master 2024-04-13 23:54:31 every aport built that is not maturin is an aport built that is not maturin 2024-04-14 00:00:44 algitbot: retry master 2024-04-14 00:10:03 algitbot: retry master 2024-04-14 00:13:15 I'm trying to see what will happen if you just disable tests for maturin in !64082 2024-04-14 00:14:39 otherwise, how do you not use the cached cffi-1.16.0-cp312-cp312-linux_aarch64.whl ? 2024-04-14 00:17:39 Only about 120 aports left for x86 community :) 2024-04-14 00:18:03 Is that downloaded or in the tarball? 2024-04-14 00:18:22 idk how stuff works 2024-04-14 00:28:27 algitbot: retry master 2024-04-14 00:33:18 https://gitlab.alpinelinux.org/alpine/aports/-/jobs/1342055 2024-04-14 00:33:39 I think I'm gonna just disable maturin tests on aarch64 for now 2024-04-14 00:35:10 Alright 2024-04-14 00:39:53 yesh 2024-04-14 00:40:12 aarch64 can finally move on 2024-04-14 00:43:13 and it managed to rebuild py3-pydantic-core with py3-maturin 1.5.1 so there is hope 2024-04-14 00:44:21 algitbot: retry master 2024-04-14 00:45:42 I wonder if we should allow netpbm (and probably libpeas2) tests to fail 2024-04-14 00:50:49 80 more aports for x86 community 2024-04-14 01:03:04 algitbot: retry master 2024-04-14 01:04:18 algitbot: retry master 2024-04-14 01:04:25 algitbot: retry master 2024-04-14 01:05:30 We should get the auto-retry thing going again 2024-04-14 01:06:22 algitbot: retry ptrc 2024-04-14 01:06:32 algitbot: retry master 2024-04-14 01:06:33 Ok, i think we need to give libpeas2 the same treatment as libpeas 2024-04-14 01:16:16 netpbm too 2024-04-14 01:16:21 or else these 2 will block x86 2024-04-14 01:16:24 algitbot: retry master 2024-04-14 01:19:28 "set by middleware" != "set by endpoint" 2024-04-14 01:19:30 Fun 2024-04-14 01:20:39 algitbot: retry master 2024-04-14 01:20:55 60 more aports for x86 2024-04-14 01:37:58 algitbot: retry master 2024-04-14 01:41:19 algitbot: retry master 2024-04-14 01:52:09 algitbot: retry master 2024-04-14 01:56:48 algitbot: retry master 2024-04-14 02:04:44 algitbot: retry master 2024-04-14 02:09:51 algitbot: retry master 2024-04-14 02:17:27 algitbot: retry master 2024-04-14 02:18:23 Less than 40 aports for x86 2024-04-14 02:21:58 algitbot: retry master 2024-04-14 02:25:45 algitbot: retry master 2024-04-14 02:29:10 netbpm on x86 again 2024-04-14 02:30:13 algitbot: retry master 2024-04-14 02:40:11 algitbot: retry master 2024-04-14 02:40:17 !64088 2024-04-14 02:44:40 I am building libpeas2 in an x86 container to see hopefully see what's wrong 2024-04-14 02:45:44 -see 2024-04-14 02:47:30 extension-gjs segfaults in libmozjs 2024-04-14 02:51:56 algitbot: retry master 2024-04-14 03:02:15 What do you think of !64088? 2024-04-14 03:02:18 I am working on libpeas2 2024-04-14 03:02:22 algitbot: retry master 2024-04-14 03:02:30 I remember netpbm tests being flaky, and failing often 2024-04-14 03:02:38 and to top it off, i think it passed in CI 2024-04-14 03:02:42 algitbot: retry master 2024-04-14 03:08:13 There it goes again 2024-04-14 03:08:14 algitbot: retry master 2024-04-14 03:09:19 pkgs.a.o/flagged is gettings lots of py3-* packages now due to the rebuild 2024-04-14 03:09:21 algitbot: retry master 2024-04-14 03:09:24 getting* 2024-04-14 03:10:01 Going to ignore that for the time being, and focus on the failing aports 2024-04-14 03:10:52 Building libpeas 2.0.2 to see if it fixes the segfault 2024-04-14 03:11:44 No, it still fails 2024-04-14 03:11:46 algitbot: retry master 2024-04-14 03:16:36 Building the latest git revision to see if that fixes it 2024-04-14 03:16:38 algitbot: retry master 2024-04-14 03:20:31 algitbot: retry master 2024-04-14 03:20:50 Latest revision does not fix it 2024-04-14 03:24:34 algitbot: retry master 2024-04-14 03:25:20 algitbot: retry master 2024-04-14 03:27:42 algitbot: retry master 2024-04-14 03:27:57 So, there's libpeas2 + 1 other aport in x86 (that isn't netpbm, since it seems that has passed) 2024-04-14 03:35:49 algitbot: retry master 2024-04-14 03:39:19 algitbot: retry master 2024-04-14 03:40:07 algitbot: retry master 2024-04-14 03:40:39 gtg 2024-04-14 03:42:42 Here goes the retries for netpbm on x86_64 2024-04-14 03:42:44 algitbot: retry master 2024-04-14 03:56:28 x86 is down to the last aport (libpeas2) 2024-04-14 03:56:29 algitbot: retry master 2024-04-14 04:01:32 algitbot: retry master 2024-04-14 04:20:37 algitbot: retry master 2024-04-14 04:22:21 I've got it, it's the new glib that's causing libpeas2 gjs tests to fail 2024-04-14 04:23:59 but strangely only for x86*, i'll try to enable the tests for armv7 in CI 2024-04-14 04:30:25 algitbot: retry master 2024-04-14 04:30:55 TestNoRaceSeqSetRelativeSpeed failed 2024-04-14 04:37:50 algitbot: retry master 2024-04-14 04:44:02 algitbot: retry master 2024-04-14 04:47:59 algitbot: retry master 2024-04-14 04:56:08 netpbm again 2024-04-14 04:56:09 algitbot: retry master 2024-04-14 05:50:48 algitbot: retry master 2024-04-14 05:55:52 algitbot: retry master 2024-04-14 06:06:20 algitbot: retry master 2024-04-14 06:06:52 algitbot: retry master 2024-04-14 06:08:44 Maybe !64093 but something else is failing now 2024-04-14 06:14:05 algitbot: retry master 2024-04-14 08:39:12 ^ openvdb has the same absolute path amove problem 2024-04-14 08:42:47 !64097 2024-04-14 08:43:07 algitbot: retry master 2024-04-14 08:53:48 algitbot: retry master 2024-04-14 10:45:44 ^ openvdb !64097 2024-04-14 10:45:57 algitbot: retry master 2024-04-14 11:56:16 algitbot: retry master 2024-04-14 11:58:09 algitbot: retry master 2024-04-14 11:59:12 algitbot: retry master 2024-04-14 12:03:04 algitbot: retry mster 2024-04-14 12:03:06 s/mster/master/ 2024-04-14 12:03:16 algitbot: retry master 2024-04-14 12:11:03 algitbot: retry master 2024-04-14 12:12:15 algitbot: retry master 2024-04-14 12:23:53 algitbot: retry master 2024-04-14 12:33:25 algitbot: retry master 2024-04-14 12:33:58 algitbot: retry master 2024-04-14 12:35:08 algitbot: retry master 2024-04-14 12:35:24 algitbot: retry master 2024-04-14 12:35:38 algitbot: retry master 2024-04-14 12:38:58 algitbot: retry master 2024-04-14 12:39:25 algitbot: retry master 2024-04-14 12:40:05 algitbot: retry master 2024-04-14 12:40:31 algitbot: retry master 2024-04-14 12:40:51 algitbot: retry master 2024-04-14 12:41:07 Are there really so many things depending on py3-dask? 2024-04-14 12:41:07 algitbot: retry master 2024-04-14 12:41:43 algitbot: retry master 2024-04-14 12:41:59 algitbot: retry master 2024-04-14 12:42:21 algitbot: retry master 2024-04-14 12:49:21 omni: There's an example of disabling specific meson tests in community/libqmi and community/nautilus 2024-04-14 12:58:52 algitbot: retry master 2024-04-14 12:59:48 algitbot: retry master 2024-04-14 13:00:49 algitbot: retry master 2024-04-14 13:02:07 algitbot: retry master 2024-04-14 13:06:30 algitbot: retry master 2024-04-14 13:09:32 algitbot: retry master 2024-04-14 13:10:20 algitbot: retry master 2024-04-14 13:11:01 algitbot: retry master 2024-04-14 13:11:22 Wow, 60 more aports, and it keeps picking py3-dask 2024-04-14 13:11:23 algitbot: retry master 2024-04-14 13:11:31 Again! 2024-04-14 13:11:39 algitbot: retry master 2024-04-14 13:11:47 and again! 2024-04-14 13:11:55 algitbot: retry master 2024-04-14 13:12:21 and just 1 aport before it picks py3-dask again 2024-04-14 13:12:22 algitbot: retry master 2024-04-14 13:12:35 Ok, now blender, that will take a while 2024-04-14 13:17:16 nice 2024-04-14 13:18:00 algitbot: retry master 2024-04-14 13:18:27 it worked on x86 but those tests failed on armv7 instead 2024-04-14 13:18:30 algitbot: retry master 2024-04-14 13:19:06 algitbot: retry master 2024-04-14 13:19:59 algitbot: retry master 2024-04-14 13:20:24 algitbot: retry master 2024-04-14 13:20:49 algitbot: retry master 2024-04-14 13:21:52 algitbot: retry master 2024-04-14 13:22:16 algitbot: retry master 2024-04-14 13:22:29 c'mon 2024-04-14 13:23:33 I'm afraid it won't pass, see #15980 2024-04-14 13:32:07 algitbot: retry master 2024-04-14 13:33:03 algitbot: retry master 2024-04-14 13:33:26 algitbot: retry master 2024-04-14 13:33:52 Ok, saga-gis, that should take a while 2024-04-14 13:36:02 algitbot: retry master 2024-04-14 13:39:25 algitbot: retry master 2024-04-14 13:41:47 algitbot: retry master 2024-04-14 13:42:28 I think Python 3.11.9 will break py3-dask as well, though hopefully it is just the tests 2024-04-14 13:49:11 About 12 aports to go for x86 to upload testing 2024-04-14 13:49:21 algitbot: retry master 2024-04-14 13:49:41 algitbot: retry master 2024-04-14 13:50:24 algitbot: retry master 2024-04-14 13:51:23 algitbot: retry master 2024-04-14 13:51:51 algitbot: retry master 2024-04-14 13:52:26 algitbot: retry master 2024-04-14 13:52:52 algitbot: retry master 2024-04-14 13:53:07 algitbot: retry master 2024-04-14 13:53:18 I wonder what resources are temporarily unavailable 2024-04-14 13:56:42 algitbot: retry master 2024-04-14 13:57:32 algitbot: retry master 2024-04-14 13:59:38 algitbot: retry master 2024-04-14 14:00:02 algitbot: retry master 2024-04-14 14:01:18 algitbot: retry master 2024-04-14 14:01:45 algitbot: retry master 2024-04-14 14:02:53 algitbot: retry master 2024-04-14 14:03:26 algitbot: retry master 2024-04-14 14:03:45 algitbot: retry master 2024-04-14 14:04:06 algitbot: retry master 2024-04-14 14:05:21 algitbot: retry master 2024-04-14 14:05:22 algitbot: retry master 2024-04-14 14:05:46 algitbot: retry master 2024-04-14 14:06:19 algitbot: retry master 2024-04-14 14:06:46 algitbot: retry master 2024-04-14 14:07:37 should never have upgraded libpeas2 2024-04-14 14:07:40 algitbot: retry master 2024-04-14 14:07:47 hehe 2024-04-14 14:07:54 The thing is, you enabled tests for armv7 2024-04-14 14:08:02 with that upgrade 2024-04-14 14:08:10 So, it was probably also failing with 2.0.1 2024-04-14 14:08:59 ^ Hopefully that gives a speed boost for x86 testing 2024-04-14 14:09:18 algitbot: retry master 2024-04-14 14:09:53 but it succeeded in CI where x86 was failing that test 2024-04-14 14:10:13 wonder if it is somehow related to build-3-19-armv7 2024-04-14 14:10:43 I think the arm CI is "emulated", as in something Docker something switch to 32-bit mode 2024-04-14 14:11:11 Probably still runs on ARM itself though, that why the double quotes 2024-04-14 14:11:17 algitbot: retry master 2024-04-14 14:11:17 that's* 2024-04-14 14:11:51 algitbot: retry master 2024-04-14 14:13:27 algitbot: retry master 2024-04-14 14:13:56 So, 40 aports left for aarch64 and about 50 for x86_64 2024-04-14 14:14:23 riscv64, on the other hand, juts got started 2024-04-14 14:14:30 algitbot: retry master 2024-04-14 14:15:06 py3-toolz failure looks like the one for py3-cytoolz 2024-04-14 14:15:32 algitbot: retry master 2024-04-14 14:15:49 algitbot: retry master 2024-04-14 14:16:14 algitbot: retry master 2024-04-14 14:16:39 algitbot: retry master 2024-04-14 14:16:45 algitbot: kick build-3-19-armv7 2024-04-14 14:17:06 algitbot: punch build-3-19-armv7 2024-04-14 14:17:37 algitbot: bash build-3-19-armv7 2024-04-14 14:17:58 algitbot: threaten build-3-19-armv7 2024-04-14 14:18:34 algitbot: rebuild build-3-19-armv7 2024-04-14 14:19:27 About 5 aports left for x86 testing! 2024-04-14 14:20:04 3 more! 2024-04-14 14:22:39 Last aport! 2024-04-14 14:22:41 Uploading! 2024-04-14 14:22:43 \o/ 2024-04-14 14:22:56 \o/ 2024-04-14 14:24:09 algitbot: retry master 2024-04-14 14:26:20 algitbot: retry master 2024-04-14 14:29:38 algitbot: retry master 2024-04-14 14:29:49 Ok, will be going semi-afk now, bye 2024-04-14 14:47:49 algitbot: retry master 2024-04-14 15:06:56 algitbot: retry master 2024-04-14 15:20:27 algitbot: retry master 2024-04-14 15:31:43 algitbot: retry master 2024-04-14 15:32:46 algitbot: retry master 2024-04-14 15:41:25 algitbot: retry master 2024-04-14 15:43:28 congrats! only few left 2024-04-14 15:44:09 ikke: build-3-19-armv7 looks like hanging 2nd day 2024-04-14 15:46:09 netpbm may need a few more tests disabled to pass on x86_64 2024-04-14 15:47:07 not really sure what to do about py3-dask !64084 !64105 2024-04-14 15:49:40 and for libpeas2, maybe disable those two tests on armv7, but there was no issues in CI 2024-04-14 15:50:23 omni: dask upgrade screams for https://github.com/dask/dask-expr 2024-04-14 15:50:35 yes 2024-04-14 15:51:00 but I'm not up for maintaining that, so I'd prefer if someone else would 2024-04-14 15:55:27 I suspect current py3-dask would fail also on s390x and ppc64le if rebuilt, perhaps python 3.12.3? 2024-04-14 15:58:33 and it all bound to numpy/pandas( 2024-04-14 16:06:07 omni: can you give a try to patch https://github.com/dask/dask/commit/275a79cef870448e3b9cdac6520bfec839b72b17 2024-04-14 16:11:43 algitbot: retry master 2024-04-14 16:12:16 omni: Maybe you could help me finish up !64088 if netpbm keeps failing? 2024-04-14 16:12:42 The conditional allow failure would have to be x86* now considering both x86 and x86_64 had problems 2024-04-14 16:12:52 I'll think about what i said about CI vs builder later on 2024-04-14 16:14:03 Oh, it passed 2024-04-14 16:14:08 So, x86_64 is uploading community now 2024-04-14 16:14:19 wow 2024-04-14 16:14:28 such upload 2024-04-14 16:14:30 amaze 2024-04-14 16:14:30 Netpbm tests are really flaky 2024-04-14 16:14:36 Maybe the solution would be to pull a community/starship 2024-04-14 16:14:39 which i did for community/librist 2024-04-14 16:14:44 Which is to retry 3 times before failing 2024-04-14 16:15:45 I'll think about updating my MR to do that later next week 2024-04-14 16:17:34 omni: i think py3-dask will fail on 3.19 also ;) 2024-04-14 16:17:35 Alpine 3.19, that is 2024-04-14 16:21:06 algitbot: retry master 2024-04-14 16:21:21 that one keeps on failing 2024-04-14 16:21:34 py3-py-cpuinfo on riscv64 I mean 2024-04-14 16:23:01 Hopefully it is not due to the loongarch patch 2024-04-14 16:25:03 \o/ 2024-04-14 16:25:39 So, issues like #15979 and #15983 can be closed now 2024-04-14 16:25:49 omni: maybe exclude failed tests instead of whole test suite https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/64114#note_394483 2024-04-14 16:28:57 I think dask-expr needs to be included in py3-dask's APKBUILD, like gobject-introspection is included in main/glib now 2024-04-14 16:30:32 algitbot: retry master 2024-04-14 16:31:01 algitbot: retry master 2024-04-14 16:31:31 algitbot: retry master 2024-04-14 16:38:04 gah 2024-04-14 16:38:08 time for a break 2024-04-14 16:38:10 l8r 2024-04-14 16:39:07 omni: later, thanks for all your help (and celie and others as well) 2024-04-14 16:42:06 You're welcome 2024-04-14 16:42:23 I may open an MR to bump slidge deps if i find some time in about 10 minutes or so 2024-04-14 16:42:27 algitbot: retry master 2024-04-14 17:02:09 !64121 2024-04-14 17:03:51 merci 2024-04-14 17:09:26 Thanks 2024-04-14 17:09:51 So, from memory, i think the dependency chains of testing/py3-sphinx-autodoc-typehints, testing/anki, and testing/slidge should be good 2024-04-14 17:09:57 algitbot: retry master 2024-04-14 17:09:58 algitbot: retry master 2024-04-14 17:10:00 first :P 2024-04-14 17:10:13 hehe 2024-04-14 17:36:40 huh 2024-04-14 17:37:30 snapcast needs checksum update 2024-04-14 17:37:40 yup, but x86 got a 404 2024-04-14 17:37:46 like it was _just_ retagged 2024-04-14 17:38:27 Isn't the 404 from distfiles? 2024-04-14 17:38:36 oh yeah 2024-04-14 19:29:44 "Error creating thread: Resource temporarily unavailable" sounds meh 2024-04-14 19:29:57 can't exactly strace it either 2024-04-14 19:31:07 btw omni, celie: thanks for fixing the py3.12 failures through the weekend <3 2024-04-14 20:25:38 btw the py3-dask failure is because python3 got bumped to .3 2024-04-14 20:55:56 algitbot: retry master 2024-04-14 20:59:13 algitbot: retry master 2024-04-14 21:55:50 algitbot: retry master 2024-04-14 22:44:37 !64142 2024-04-14 23:51:39 ... 2024-04-14 23:51:46 @ electron 2024-04-14 23:51:50 ptrc: You're welcome 2024-04-14 23:52:51 omni: Maybe we can do the "downgrade" trick on electron (and firefox-developer-edition, if that is still failing to build) so at least testing can be uploaded? 2024-04-14 23:52:54 I've somewhat made maybe progress with !64084 but I need to clean up 2024-04-14 23:53:16 hmm.. 2024-04-14 23:53:24 algitbot: retry master 2024-04-14 23:54:01 !64138 2024-04-14 23:55:17 should've seen that before I retriggered 2024-04-14 23:57:41 but it failed in CI, I wonder if that was expected 2024-04-14 23:59:23 omni: Have you tried rebuilding py3-dask for Alpine 3.19? 2024-04-15 00:00:15 What i mean is, i think we may need a solution that can work there too 2024-04-15 00:00:52 Unles it is expected that we never touch py3-dask in 3.19 anymore 2024-04-15 00:00:52 Unless* 2024-04-15 00:01:33 no, I have not done that too 2024-04-15 00:03:19 Before you go with the upgrade fix that adds a new aport and so on, maybe better to try out the patch to see if that works, so at least if we need to fix 3.19 too, we can use that (probably won't be able to go the new aport route there) 2024-04-15 00:05:32 no, I will continue to work on the upgrade MR, someone else can look at py3-dask for 3.19 if they want to 2024-04-15 00:07:02 there's !64105 and !64114, the patch might work 2024-04-15 00:09:45 Sure, sure, i probably should've thought of this before you spent all your time on the upgrade 2024-04-15 00:10:12 but as Alpine 3.19 community is about to go out of support, probably it doesn't matter much 2024-04-15 00:10:27 Anyway, the lesson that we can learn from this, probably, is that Python loves breaking changes 2024-04-15 00:10:42 Which is why i said "Please no" to a setuptools upgrade 2024-04-15 00:10:51 At least, in the middle of this Python 3.12 rebuild 2024-04-15 00:12:38 I mean, for Python at least, we should allow its rebuilds to go through with whatever old stuff we know works first, before pulling the foundation out from under it with upgrades in the middle of the rebuild 2024-04-15 00:15:32 Btw, does py3-dask actually need py3-dask-expr when it is actually run, or is it only needed for tests? 2024-04-15 00:16:31 I mean, if it is the latter, then maybe py3-dask-expr doesn't need to be added as a separate aport 2024-04-15 00:17:14 and can just be built in py3-dask's APKBUILD, and installed into the testenv to allow tests to pass 2024-04-15 00:17:55 one of the things to figure out for the clean-up 2024-04-15 00:18:34 I thought about both paths but chose the upgrade one, because things need to move in that direction anyway 2024-04-15 00:19:34 Sure, but it could also mean, you may not have much time to do the clean-up, and need to merge something that works well enough for the time being to move the builder on 2024-04-15 00:20:50 aarch64 CI for the electron fix MR has failed again 2024-04-15 00:21:08 due to community not being uploaded, probably 2024-04-15 00:21:17 or are we only watching the x86_64 CI for this 2024-04-15 00:23:29 Yes, we are probably only watching the x86_64 CI, and it has reached further in the build process compared to the builder (which has no patch), and the first run of the CI (with 1 patch) 2024-04-15 00:24:04 So, maybe you can take a gamble and merge that, since it is failing anyway, and we can do no worse than that 2024-04-15 00:39:05 Anyway, let's see if x86_64 picks one of the other four aports to build this time 2024-04-15 00:39:10 algitbot: retry master 2024-04-15 00:39:28 It did 2024-04-15 00:39:47 So did aarch64 2024-04-15 00:40:03 Though it's back to py3-dask again now 2024-04-15 00:40:06 algitbot: retry master 2024-04-15 00:40:20 algitbot: retry master 2024-04-15 00:40:34 algitbot: retry master 2024-04-15 00:40:50 algitbot: retry master 2024-04-15 00:41:04 algitbot: retry master 2024-04-15 00:41:24 algitbot: retry master 2024-04-15 00:42:08 algitbot: retry master 2024-04-15 00:42:23 algitbot: retry master 2024-04-15 00:42:57 algitbot: retry master 2024-04-15 00:44:06 algitbot: retry master 2024-04-15 00:44:25 algitbot: retry master 2024-04-15 00:44:41 algitbot: retry master 2024-04-15 00:44:58 algitbot: retry master 2024-04-15 00:45:39 algitbot: retry master 2024-04-15 00:48:00 algitbot: retry master 2024-04-15 00:49:50 algitbot: retry master 2024-04-15 00:50:29 algitbot: retry master 2024-04-15 00:52:49 algitbot: retry master 2024-04-15 00:54:10 algitbot: retry master 2024-04-15 00:54:28 algitbot: retry master 2024-04-15 00:54:43 algitbot: retry master 2024-04-15 00:55:32 algitbot: retry master 2024-04-15 00:55:57 algitbot: retry master 2024-04-15 00:57:02 algitbot: retry master 2024-04-15 03:18:56 yoo 2024-04-15 03:19:01 lo 2024-04-15 03:19:13 hah 2024-04-15 03:21:02 this doesn't seem very helpful 2024-04-15 03:24:54 please make it stop 2024-04-15 03:27:30 it's only stressful 2024-04-15 05:07:43 algitbot: retry master 2024-04-15 05:08:32 algitbot: retry master 2024-04-15 05:56:24 15 more aports left for aarch64 testing 2024-04-15 05:56:27 algitbot: retry master 2024-04-15 05:57:13 algitbot: retry master 2024-04-15 05:59:47 algitbot: retry master 2024-04-15 06:10:42 10 more aports 2024-04-15 06:27:42 algitbot: retry master 2024-04-15 06:28:39 algitbot: retry master 2024-04-15 06:29:20 algitbot: retry master 2024-04-15 06:32:31 algitbot: retry 3.19-stable 2024-04-15 06:33:02 algitbot: retry 3.19-stable 2024-04-15 06:48:28 algitbot: retry master 2024-04-15 06:53:47 algitbot: retry master 2024-04-15 07:04:04 algitbot: retry master 2024-04-15 07:15:07 aarch64 is at the last aport for testing! 2024-04-15 07:15:13 Uploading now.. 2024-04-15 07:15:43 So, it seems x86_64 took so long to compile electron that aarch64 caught up in the meantime and uploaded testing before x86_64. 2024-04-15 08:12:47 can't wait for community to fail now 2024-04-15 08:28:00 algitbot: retry master 2024-04-15 09:08:59 algitbot: retry master 2024-04-15 09:10:34 algitbot: retry master 2024-04-15 12:43:04 build-edge-armv7 seems stuck, I was a bit worried that one of the arms would get stuck on some py-dask test, couldn't figure out which one and it didn't seem to happen every time 2024-04-15 12:43:32 also looks like build-edge-x86_64 got stuck on the electron rebuild 2024-04-15 13:00:33 ptrc: did you redo a bunch of the rebuilds? 2024-04-15 13:01:03 i.. took latest edge/community after it finished on x86_64, grepped for `usr/lib/python3.11` and rebuilt those 2024-04-15 13:02:43 ok, because I think I had already rebuilt most of those, or opened MRs 2024-04-15 13:04:02 x86_64 probably hadn't gotten to those yet because it's stuck on electron 2024-04-15 13:10:24 omni: Yeah, i mentioned the duplicated effort to ptrc (in #alpine-rust, off topic, i know..) 2024-04-15 13:27:42 here are some that do need rebuilds https://pkgs.alpinelinux.org/contents?file=%2A&branch=edge&repo=testing&arch=aarch64&page=1&path=%2Fusr%2Flib%2Fpython3.11%2Fsite-packages%2F%2A 2024-04-15 13:30:23 I'll take care of slidge and slidge-matridge, slidge has already been rebuilt on x86_64 and i am waiting on !64189 (a dependency) so i can move them to community 2024-04-15 13:38:04 edge/testing for aarch64 gives me.. 727 packages 2024-04-15 13:38:07 that doesn't sound right 2024-04-15 13:39:36 no? 2024-04-15 13:40:01 check them in git history if they were rebuilt for python 3.12 2024-04-15 13:40:12 https://ptrc.gay/eImrnadd 2024-04-15 13:40:13 oh 2024-04-15 13:40:14 it *is* right 2024-04-15 13:40:15 darn 2024-04-15 13:41:08 would be nice to get MRs of like, 100 packages each 2024-04-15 13:41:15 That's more than what you've been working with before? 2024-04-15 13:41:44 what do you mean? 2024-04-15 13:41:53 ( as in, who, and which before ) 2024-04-15 13:42:06 I mean, you've been rebuilding testing, right? 2024-04-15 13:42:10 locally 2024-04-15 13:42:13 ah 2024-04-15 13:42:22 i've built most of them locally tbf 2024-04-15 13:42:30 i can get a list of which has built successfully for me previously 2024-04-15 13:42:34 and when you said 727 didn't sound right, i was wondering if that was more than what you built locally 2024-04-15 13:42:45 i thought they were bumped by ncopa previously 2024-04-15 13:42:46 but apparently not 2024-04-15 13:43:11 ncopa bumped anki (don't know if there were any more) 2024-04-15 13:43:16 anki + deps 2024-04-15 13:43:29 and then i bumped deps of py3-sphinx-autodoc-typehints and slidge 2024-04-15 13:43:43 As otherwise they would block the upload of testing 2024-04-15 13:44:22 ~/packages/testing/x86_64 ❯ ls | rg -vF -- '-pyc' | wc -l 2024-04-15 13:44:22 867 2024-04-15 13:44:52 i guess that (excluding unrelated stuff) it could be more or less all of them 2024-04-15 13:45:19 I wonder if some packages have more than 1 -pyc 2024-04-15 13:45:28 also, just realised: it would be pretty much impossible to rebuild that with MRs 2024-04-15 13:45:40 we'd have to a massive build dependency tree 2024-04-15 13:48:46 or maybe we just let maintainers rebuild their aports, it would allow us to see who's still around and watching their aports in testing ;) 2024-04-15 13:49:21 i mean 2024-04-15 13:49:28 the dependencies are very tangled 2024-04-15 13:51:23 i don't think i could bump most of my testing/ python stuff without also bumping packages of 10 other people 2024-04-15 13:52:45 You've already got !62405 2024-04-15 13:52:58 !64205 * 2024-04-15 13:54:01 yeah, and that is lacking dependencies 2024-04-15 13:54:11 unless they were bumped and i didn't notice 2024-04-15 13:54:31 ah no, i bumped then 2024-04-15 13:54:34 them* 2024-04-15 13:55:29 This is getting confusing 2024-04-15 13:58:26 hey. seems like i missed to bump a bunch of packages? 2024-04-15 13:58:31 in testing/, yeah 2024-04-15 13:58:37 lol 2024-04-15 13:58:55 in testing i didnt bump anything basically 2024-04-15 13:59:14 fair enough 2024-04-15 13:59:16 the idea was to bump testing while main/community was build 2024-04-15 13:59:21 You missed a bunch of bumps in main and community too 2024-04-15 13:59:35 those worries me more 2024-04-15 13:59:39 idunno if a bunch, around 40? 2024-04-15 14:00:28 Probably victims of rebasing 2024-04-15 14:00:35 we will know when trying to build 3.20 2024-04-15 14:00:49 weird 2024-04-15 14:01:05 $ tpaste < find-python-rdeps.sh 2024-04-15 14:01:05 https://tpaste.us/6MgZ 2024-04-15 14:01:11 thats how i found what packages to rebuild 2024-04-15 14:01:20 i likely missed something 2024-04-15 14:02:28 There were a bunch of aports that had other changes 2024-04-15 14:02:30 but not the pkgrel bump 2024-04-15 14:02:36 So, those were probably due to rebasing 2024-04-15 14:02:40 oh 2024-04-15 14:02:51 thats likely why 2024-04-15 14:03:19 while rebasing i didnt rebump anything that had git commits since the python3 rebuild 2024-04-15 14:03:31 since the python3 3.12 bump commit 2024-04-15 14:03:43 Anyway, now many people are just doing individual rebuilds of testing :) 2024-04-15 14:03:55 good enough for me :) 2024-04-15 14:04:27 thank you so much for helping cleaning up the mess I made 2024-04-15 14:04:31 if we could just get build-edge-armv7 and build-edge-x86_64 unstuck 2024-04-15 14:05:07 i'll have a look 2024-04-15 14:05:28 Everybody chipped in, though omni got the most frustrating py3-dask 2024-04-15 14:05:52 I don't see how it could not be made a mess that needed cleaning up, without maybe freezing everything for a long time while sorting things out 2024-04-15 14:05:58 I think this is as good as it gets 2024-04-15 14:06:36 speaking of py3-dask, lol 2024-04-15 14:06:39 and i think the py3-dask thing was due to Python getting an upgrade in the midst of all this 2024-04-15 14:06:48 > Terminated 2024-04-15 14:06:48 So, some archs had already built that 2024-04-15 14:06:51 what 2024-04-15 14:06:54 but also, yeah 2024-04-15 14:06:54 i killed it 2024-04-15 14:07:09 the 3.12.3 upgrade broke a little 2024-04-15 14:07:12 i killed py3-dask 2024-04-15 14:07:23 So, does py3-dask require more fixing, or just a retry? 2024-04-15 14:07:24 it was stuck, it may get stuck on some test again 2024-04-15 14:07:36 couldn't figure out which one 2024-04-15 14:07:45 and/or was just fed up trying 2024-04-15 14:07:49 You can use pytest -v? 2024-04-15 14:08:10 ah, yes, probably 2024-04-15 14:09:39 i need to get the build-edge-loongarch64 up this week 2024-04-15 14:09:49 and then bootstrap the build-3-20-* machines 2024-04-15 14:10:00 Anyway, ncopa, you bumped a few aports in testing with 647741f7af1 2024-04-15 14:10:42 Though i couldn't figure out why those particular aports, maybe anki probably because it's a complicated one 2024-04-15 14:11:05 yeah i started with it, and pushed those that built 2024-04-15 14:11:15 and I stopped at the first failure 2024-04-15 14:11:29 So, does this mean we'll be getting a Loongarch builder first before CI? 2024-04-15 14:11:41 probably so, yes 2024-04-15 14:11:53 and that is a good question 2024-04-15 14:12:29 i wonder if loongarch can wait til after 3.20 2024-04-15 14:12:43 would be nice if we could do it after 3.20 2024-04-15 14:12:45 Another question would be, do we want the Loongarch CI to be like Riscv64 when it was emulated (failure allowed) 2024-04-15 14:12:58 no, probably now 2024-04-15 14:12:59 but 2024-04-15 14:13:05 those are very good questions 2024-04-15 14:13:50 I wonder if the Loongarch people are waiting for something concrete soon to show their managers or something 2024-04-15 14:14:02 ncopa: can you kill the electron build on x86_64 too? or is it moving? 2024-04-15 14:14:08 thats my impression yes 2024-04-15 14:14:12 is probably moving 2024-04-15 14:14:18 I'll check 2024-04-15 14:14:42 ok, just thought it had been going for a long while, thought it would "just" take 2-3hrs 2024-04-15 14:15:01 The x86_64 builder is slower than aarch64 2024-04-15 14:15:11 Probably because other stuff are also running on the same machine? 2024-04-15 14:15:11 when was that build started? 2024-04-15 14:15:17 its moving 2024-04-15 14:15:35 the x86_64 machine is and old slow machine 2024-04-15 14:15:50 >>> electron: Building testing/electron 29.3.0-r1 (using abuild 3.12.0-r5) started Mon, 15 Apr 2024 10:08:30 +0000 2024-04-15 14:16:00 Mon Apr 15 14:15:52 UTC 2024 2024-04-15 14:16:10 ok 2024-04-15 14:16:21 and there may have been some duplicated effort with electron too 2024-04-15 14:16:33 the electron build took 132 minutes on the x86_64 CI builder 2024-04-15 14:16:35 I think lnl already added patches that removed a dependency on py3-setuptools 2024-04-15 14:16:50 i think it also had to compete with kernel builds in stable branches 2024-04-15 14:17:12 oh, ok 2024-04-15 14:17:54 oh, I thought the current electron build was !64138 ? 2024-04-15 14:18:07 ncopa added py3-setuptools later on, i think 2024-04-15 14:18:16 yeah, c634e9e930ddd11039cc699bfa6cae050d233d39 2024-04-15 14:20:03 i added it because i saw a build failure and looked at the log and though "oh tis is easy to fix..." 2024-04-15 14:20:07 :) 2024-04-15 14:21:03 ncopa: do you remember why llvm14 was disabled on rv64? 2024-04-15 14:21:10 it breaks surprisingly a lot of stuff 2024-04-15 14:21:43 It was stuck, i think 2024-04-15 14:22:08 https://build.alpinelinux.org/buildlogs/build-edge-riscv64/main/llvm14/llvm14-14.0.6-r18.log 2024-04-15 14:22:08 ah 2024-04-15 14:22:09 check failed 2024-04-15 14:22:13 > LLVM ERROR: Relocation type not implemented yet! 2024-04-15 14:23:30 can we.. disable tests on rv64? only 11 of them fail 2024-04-15 14:28:34 no, tests failed 2024-04-15 14:28:51 lets move the discussion to #alpine-devel 2024-04-15 14:55:41 algitbot: retry master 2024-04-15 15:03:12 algitbot: retry master 2024-04-15 15:04:33 oh. 2024-04-15 15:04:41 i forgot that for some of the ovos stuff i disabled check 2024-04-15 15:04:42 damn it 2024-04-15 15:05:18 Time to panic? 2024-04-15 15:07:13 hopefully not panic 2024-04-15 15:07:14 oh wait 2024-04-15 15:07:17 how did x86 finish 2024-04-15 15:07:23 ag 2024-04-15 15:07:23 ah 2024-04-15 15:07:26 it's the previous run 2024-04-15 15:07:29 algitbot: retry master 2024-04-15 15:07:39 Nothing needs retrying 2024-04-15 15:08:40 Hmm, i wonder why ppc64le has more aports to build than aarch64 and x86 2024-04-15 15:08:49 algitbot: retry master 2024-04-15 16:07:58 Bus error, eh? 2024-04-15 17:03:01 https://github.com/xonsh/xonsh/commit/dd4d6f5c26734a568d068aa51d470dcabd0ab5c1 2024-04-15 19:47:52 algitbot: retry master 2024-04-15 19:49:08 to be fair it looks quite good that most builders are at like, 30 packages left 2024-04-15 19:49:26 yeah 2024-04-15 19:50:02 ptrc: I wonder if this could be something https://github.com/xonsh/xonsh/commit/dd4d6f5c26734a568d068aa51d470dcabd0ab5c1 2024-04-15 19:50:19 oh, i didn't see the new failure yet 2024-04-15 19:50:38 oh 2024-04-15 19:50:46 i just disabled the manpage completions 2024-04-15 19:50:52 turns out there's tests depending on it working 2024-04-15 19:50:52 darn it 2024-04-15 19:52:16 also 2024-04-15 19:52:26 omni: afaict the cgi stuff will be useful for the next release 2024-04-15 19:52:31 but for now it's alright 2024-04-15 19:56:22 k 2024-04-15 20:29:44 oh, haha, I just opened an upgrade MR for py3-pyroma 2024-04-15 20:30:18 oh whoops 2024-04-15 20:30:29 if you managed to fix it, sure 2024-04-15 20:30:49 i attempted, and gave up 2024-04-15 20:31:40 not yet 2024-04-15 21:27:18 algitbot: retry master 2024-04-15 21:45:42 algitbot: retry master 2024-04-15 22:47:01 algitbot: retry master 2024-04-15 22:48:19 E PermissionError: [Errno 13] Permission denied: '/tmp/ssh-XXXXBnJpCN/__init__.py' 2024-04-15 22:48:22 huh 2024-04-15 22:51:32 algitbot: retry master 2024-04-15 23:11:43 algitbot: retry master 2024-04-15 23:19:53 algitbot: retry master 2024-04-15 23:23:43 considering disabling gdcm on armv7 and x86 2024-04-16 00:15:35 I guess the Py 3.12 rebuild in testing is mostly over (except for a few more aports that are failing)? 2024-04-16 00:16:10 <3 and thanks to everyone who got this done. 2024-04-16 00:23:28 I think a comparably small number were left out 2024-04-16 00:25:24 I think with how much breakage there's been, that's to be expected 2024-04-16 00:37:10 what timing 2024-04-16 00:37:13 don't retrigger 2024-04-16 00:37:22 I think I'll disable it on x86_64 too 2024-04-16 00:38:41 or maybe try to build without texlive, seems like that is what is failing... 2024-04-16 00:39:07 could be related to the recent tex changes 2024-04-16 00:46:19 !64254 2024-04-16 00:46:46 shouldn't take much longer than five minutes in CI, want to see what it looks like first 2024-04-16 01:53:35 algitbot: retry master 2024-04-16 02:07:03 algitbot: retry master 2024-04-16 02:08:20 hm.. 2024-04-16 02:08:34 algitbot: kick build-edge-riscv64 2024-04-16 02:08:51 algitbot: shake build-edge-riscv64 2024-04-16 02:18:14 algitbot: hello 2024-04-16 02:18:19 algitbot: hi 2024-04-16 03:25:01 algitbot: retry master 2024-04-16 03:25:31 algitbot: retry master 2024-04-16 03:25:53 Hmm, ok so the riscv64 builder doesn't want to continue, that's what you were trying to solve with the "kick" and "shake" 2024-04-16 05:18:17 algitbot: retry master 2024-04-16 05:36:07 algitbot: retry master 2024-04-16 05:37:19 algitbot: retry master 2024-04-16 05:44:43 Is the riscv64 builder still stuck? 2024-04-16 05:45:25 Ah yes, i think the riscv64 CI runner is down too 2024-04-16 05:47:17 yup, cannot connect to the host 2024-04-16 05:49:23 Wow, that's bad 2024-04-16 15:01:36 algitbot: retry master 2024-04-16 15:05:30 ouch, py3-scipy takes such a long time to build, or if it is the tests that take time 2024-04-16 15:06:24 lib/python3.12/site-packages/scipy/linalg/tests/test_matfuncs.py ....FFF [ 19%] 2024-04-16 15:06:25 FFTerminated 2024-04-16 15:08:23 oom killer? 2024-04-16 15:09:35 could be some other killer... ;) 2024-04-16 18:13:46 is there anything python 3.12 related in main/ or community/ we may have missed? 2024-04-16 18:31:32 algitbot: retry master 2024-04-16 18:48:14 I had the disabled and removed md2gemini and py3-language-server packages installed 2024-04-16 23:53:17 algitbot: retry master 2024-04-17 01:33:56 algitbot: retry master 2024-04-17 01:35:28 algitbot: retry master 2024-04-17 01:36:33 algitbot: retry master 2024-04-17 01:38:04 FFplay not found 2024-04-17 02:28:55 !64370 2024-04-17 03:17:36 algitbot: retry master 2024-04-17 04:13:21 algitbot: retry master 2024-04-17 04:32:29 \o/ 2024-04-17 06:57:05 New abuild :) 2024-04-17 08:08:59 algitbot: retry master 2024-04-17 08:10:02 Hmm, i think this mean riscv64 will be the first to try out the new abuild 2024-04-17 08:11:22 Oh wait, x86_64 already did, with homer-app (it is only enabled for 1 arch) 2024-04-17 09:45:42 darn it 2024-04-17 09:45:50 i forgot to bump the release dates 2024-04-17 11:06:44 algitbot: retry master 2024-04-17 11:38:15 algitbot: retry master 2024-04-17 11:56:14 algitbot: retry master 2024-04-17 12:17:01 algitbot: retry master 2024-04-17 12:26:01 algitbot: retry master 2024-04-17 12:26:13 !64396 2024-04-17 13:07:51 algitbot: retry master 2024-04-17 14:06:59 algitbot: retry master 2024-04-17 14:38:00 algitbot: retry master 2024-04-17 14:39:52 "signal: 4, SIGILL: illegal instruction" 2024-04-17 14:40:00 for onefetch 2024-04-17 14:40:09 tests 2024-04-17 14:45:03 \o/ 2024-04-17 14:49:06 \o/ again 2024-04-17 15:11:32 "wget: bad address 'mirror.ibcp.fr'" 2024-04-17 15:11:36 algitbot: retry master 2024-04-17 16:36:39 "mkdio.h: No such file or directory" 2024-04-17 16:37:07 algitbot: retry master 2024-04-17 16:49:10 provided by discount-dev 2024-04-17 17:01:58 algitbot: retry master 2024-04-17 18:51:17 algitbot: retry master 2024-04-17 18:51:20 assertion failures 2024-04-17 18:51:23 (tests) 2024-04-17 18:51:37 ambiguous column name: ROWID (tracker-db-interface-error-quark, 0) 2024-04-17 19:30:23 algitbot: retry master 2024-04-18 01:18:08 algitbot: retry master 2024-04-18 01:35:48 algitbot: retry master 2024-04-18 04:11:44 algitbot: retry master 2024-04-18 07:29:14 algitbot: retry master 2024-04-18 08:26:18 algitbot: retry master 2024-04-18 10:31:25 algitbot: retry master 2024-04-18 12:12:22 ptrc: I think this is the fifth attempt for x86_64 to build firefox-developer-edition-126.0_beta1-r0 2024-04-18 14:37:08 algitbot: retry master 2024-04-18 14:56:13 algitbot: retry master 2024-04-18 15:12:05 algitbot: retry master 2024-04-18 15:59:15 Hmm, so firefox-developer-edition has finally built successfully 2024-04-18 16:00:43 ^^ 2024-04-18 16:05:41 Anyway, i've got to go now, so bye 2024-04-18 16:28:54 algitbot: retry master 2024-04-18 17:24:22 algitbot: retry master 2024-04-18 18:12:32 What's this channel for? 2024-04-18 18:29:30 algitbot: retry master 2024-04-18 18:29:52 lassebq: commits 2024-04-18 18:30:17 hang around and you'll see 2024-04-18 18:30:23 commit log basically? 2024-04-18 18:30:43 with the builders status 2024-04-18 18:30:53 Ok 2024-04-18 18:30:55 https://build.alpinelinux.org/ 2024-04-18 18:32:16 and discussion of build failures 2024-04-18 23:24:48 algitbot: retry master 2024-04-18 23:36:40 So, bind has also failed due to FHS checks, and the MR for bind was made by @Sertonix who also added those checks? 2024-04-18 23:37:17 I wonder if those checks are such a good idea then 2024-04-19 00:19:04 Another victim of the new fhs checks 2024-04-19 01:11:07 One more victim! 2024-04-19 01:26:47 fcolista: i will be merging !64469 to prevent those 2 aports from blocking the riscv64 builder, it cannot pass CI due to the Python rebuild for riscv64 not being uploaded yet, and those 2 aports had pkgrel bumped as part of the Python rebuild anyway 2024-04-19 01:27:30 Maybe later on you will want to see if /run can be created in a way that does not trigger those new FHS checks 2024-04-19 01:28:06 or maybe those FHS checks are unnecessary and just create more problems 2024-04-19 01:30:23 Other archs built those 2 aports before the abuild 3.13.0 with the new checks was merged 2024-04-19 01:30:34 -the 2024-04-19 01:37:06 For reference, the MR that introduced those new checks: https://gitlab.alpinelinux.org/alpine/abuild/-/merge_requests/254 2024-04-19 01:41:35 Even /dev is included in the new checks, luckily we have main/fuse3/dont-mknod-dev-fuse.patch 2024-04-19 01:54:58 algitbot: retry master 2024-04-19 02:01:48 Oh great 2024-04-19 02:01:53 Yet another victim of fhs 2024-04-19 02:31:21 algitbot: retry master 2024-04-19 02:32:15 So, netpbm vendors its own copy of libjasper? 2024-04-19 02:34:01 I mean, there was something that vendored discount a few days ago, and mkdio.h was not found on the first try; and now netpbm fails to find libjasper on first try.. 2024-04-19 04:57:44 algitbot: retry master 2024-04-19 05:54:25 algitbot: retry master 2024-04-19 05:57:44 400 aports to go for riscv64 community 2024-04-19 06:03:18 Never expected s390x to be the first to complete maxima 2024-04-19 08:49:29 algitbot: retry master 2024-04-19 10:14:21 algitbot: retry master 2024-04-19 10:58:17 algitbot: retry master 2024-04-19 14:18:09 algitbot: retry msater 2024-04-19 14:18:12 algitbot: retry master 2024-04-19 14:42:11 Ugh, I accidentally pushed that one... I thought I was on a branch 😒 2024-04-19 14:51:15 algitbot: retry master 2024-04-19 15:27:15 algitbot: retry master 2024-04-19 15:47:40 algitbot: retry master 2024-04-19 15:55:17 What's happening with netpbm? i guess the failure is not getting fixed by retrying 2024-04-19 16:00:20 seems like nothing depends on it, so I guess we could disable it on rv64 2024-04-19 16:11:33 is it the same tests that are failing every time or is it random ones of "Some tests sporadically fail, try to repeat 3 times before failing"? 2024-04-19 16:12:16 I can't even figure out what is failing 2024-04-19 16:40:14 As far as i can tell, it is the same tests that fail 2024-04-19 16:40:20 but i didn't look too deeply 2024-04-19 16:40:27 On x86*, retrying usually solves it 2024-04-19 16:40:32 but apparently, not on riscv64 2024-04-19 16:41:44 So, probably as ikke said, we can probably disable it on riscv64; either that or we start disabling the failing tests, but we'd have to do that conditionally for riscv64...whatever 2024-04-19 17:37:54 algitbot: retry master 2024-04-19 18:20:40 undefined symbol: vp9_block_error_fp_sve 2024-04-19 18:20:48 undefined symbol: vpx_sum_squares_2d_i16_sve 2024-04-19 18:47:03 'ImageComparisonFailure: images not close (RMS 0.002)' 2024-04-19 18:47:05 anoying 2024-04-19 21:43:17 what was that about? 2024-04-19 21:43:48 algitbot: retry master 2024-04-19 21:59:57 algitbot: retry master 2024-04-19 22:00:27 oh-no, I restarted the chromium build that will fail again eventually 2024-04-20 00:15:52 algitbot: retry master 2024-04-20 00:20:18 algitbot: retry master 2024-04-20 00:38:53 algitbot: retry master 2024-04-20 01:07:28 algitbot: retry master 2024-04-20 02:02:16 algitbot: retry master 2024-04-20 02:07:09 algitbot: retry master 2024-04-20 02:30:21 algitbot: retry master 2024-04-20 03:07:40 algitbot: retry master 2024-04-20 03:20:22 algitbot: retry master 2024-04-20 03:43:30 algitbot: retry master 2024-04-20 04:12:06 120 more aports for riscv64 community 2024-04-20 04:15:32 algitbot: retry master 2024-04-20 04:20:49 It seems netpbm has finally built successfully 2024-04-20 04:20:56 algitbot: retry master 2024-04-20 04:25:09 Arghh 2024-04-20 04:25:10 I think community/unit is another victim of the extended FHS checks 2024-04-20 04:25:47 andypost:matrix.org: Would you like to have a look at that? 2024-04-20 04:28:15 I hope we won't be seeing as many aports fail to build in Alpine 3.20 due to the new FHS checks, like how we saw aports fail due to LFS64 in 3.19 2024-04-20 04:29:21 I wonder what's the recommended solution if the FHS checks fail, create the offending directories in a post-install script, or maybe in OpenRC service files (if there are used)? 2024-04-20 04:29:45 I think if we keep doing !fhs, it just defeats the purpose of having those checks in the first place 2024-04-20 04:30:02 better to use revert them and save us the trouble 2024-04-20 04:30:37 algitbot: retry master 2024-04-20 04:30:38 Please build something else 2024-04-20 04:31:01 algitbot: retry master 2024-04-20 04:31:22 s/there are/they are/ 2024-04-20 04:32:03 s/to use/to just/ 2024-04-20 05:13:34 algitbot: retry master 2024-04-20 05:13:57 112 aports left 2024-04-20 05:18:17 algitbot: kick build-edge-riscv64 2024-04-20 05:18:18 algitbot: kick riscv64 2024-04-20 05:18:28 algitbot: kick build-edge-riscv64 2024-04-20 05:23:38 cely: That's no longer functional 2024-04-20 05:33:53 I figured 2024-04-20 05:33:54 algitbot: retry master 2024-04-20 05:39:32 ikke: By the way, has gitlab.a.o been slow since last night? 2024-04-20 05:43:05 algitbot: retry master 2024-04-20 05:58:12 cely: I haven't noticed anything 2024-04-20 06:17:53 Ok 2024-04-20 06:18:16 Probably an issue on my end, other sites seem alright though 2024-04-20 06:28:04 algitbot: retry master 2024-04-20 06:38:16 algitbot: retry master 2024-04-20 06:56:45 algitbot: retry master 2024-04-20 08:14:31 algitbot: retry master 2024-04-20 08:40:01 Quagga wants to put something under /var/run 2024-04-20 08:40:57 algitbot: retry master 2024-04-20 08:46:24 Pingu also wants to put something under /var/run 2024-04-20 08:46:28 I forsee this FHS check thing becoming 3.20's LFS64 2024-04-20 08:46:31 algitbot: retry master 2024-04-20 08:47:38 algitbot: retry master 2024-04-20 09:36:37 did botan fail the samve way on riscv64 as in !63904 ? 2024-04-20 09:38:16 I think so 2024-04-20 10:17:23 I assume it's the same for botan3 2024-04-20 10:37:15 ncopa: see if !63904 and !63903 succeed on riscv64 with the tls_stream_integration removed for it 2024-04-20 10:37:17 gtg 2024-04-20 11:46:31 celeste: any details about changes in FHS and why it affected only RV64 2024-04-20 11:47:18 > Packages must not put anything under /var/run 2024-04-20 11:47:37 You can look at the log for this channel on irclogs.a.o 2024-04-20 11:47:45 I've been ranting about this for the past two days 2024-04-20 11:47:58 https://gitlab.alpinelinux.org/alpine/abuild/-/merge_requests/254 (from memory) 2024-04-20 11:49:02 It only affected riscv64 because of the Python 3.12 rebuild backlog 2024-04-20 11:49:15 The other builders already built that before abuild 3.13.0 with this change was merged 2024-04-20 11:49:39 Well, now it will affect Alpine 3.20, because the builders for that are online 2024-04-20 11:51:18 Too bad the MR that made this change didn't really give an explanation..like the motivation for it, and suggested solutions for packages that now cannot build successfully due to this change 2024-04-20 12:00:22 Oh wow 2024-04-20 12:00:25 sircbot is also caught in this FHS saga 2024-04-20 12:04:10 FHS saga? 2024-04-20 12:04:45 yes > Packages must not put anything under /var/run 2024-04-20 12:04:46 Look at the latest build failed log above (for pingu) 2024-04-20 12:05:36 /run was included too at first, but that was removed/relaxed a few hours ago 2024-04-20 12:05:48 Oh, /var/run? 2024-04-20 12:06:21 The changed was introduced by https://gitlab.alpinelinux.org/alpine/abuild/-/merge_requests/254 2024-04-20 12:06:26 and included in abuild 3.13.0 2024-04-20 12:06:26 for unit I see there's some mix of var/run vs /run usage and technically it savin pid files and sockets to the own dir 2024-04-20 12:07:38 /run has been removed from the FHS check, but /var/run is still there 2024-04-20 12:07:43 Ah thanks for the link 2024-04-20 12:07:52 You're welcome 2024-04-20 12:09:06 I think i've seen maybe around 10 aports affected so far, and there will surely be more 2024-04-20 12:53:26 algitbot: retry master 2024-04-20 12:54:44 algitbot: retry master 2024-04-20 12:55:19 algitbot: retry master 2024-04-20 13:01:32 algitbot: retry master 2024-04-20 13:18:19 celeste: unit can safely remove the dir in package, sadly I'm afk 2024-04-20 13:29:44 andypost:matrix.org: Would a `rmdir "$pkgdir"/var/run/unit "$pkgdir"/var/run` in `package()` of community/unit be alright? 2024-04-20 13:33:44 celeste: yes, that's some unused remains 2024-04-20 13:53:08 andypost:matrix.org: i've opened !64543, please have a look 2024-04-20 14:59:18 algitbot: retry master 2024-04-20 15:02:04 andypost:matrix.org: It seems the the arm* CIs for the unit MR are stuck on something 2024-04-20 15:04:04 Yes, it's always happening after build on arm( I failed to get why celeste 2024-04-20 15:04:43 Like something hangs in memory and locks files 2024-04-20 15:06:16 I wonder if the tests fail to clean up something 2024-04-20 15:20:37 algitbot: retry master 2024-04-20 15:24:29 algitbot: retry master 2024-04-20 15:29:18 algitbot: retry master 2024-04-20 15:29:22 algitbot: retry master 2024-04-20 15:31:18 algitbot: retry master 2024-04-20 15:33:26 algitbot: retry master 2024-04-20 15:34:58 algitbot: retry master 2024-04-20 15:36:23 algitbot: retry master 2024-04-20 15:40:34 algitbot: retry master 2024-04-20 15:49:26 algitbot: retry master 2024-04-20 15:59:05 algitbot: retry master 2024-04-20 16:01:22 openimageio: !64552 2024-04-20 16:03:44 anytun: why would you package anything under /var/run? 2024-04-20 16:04:10 >>> ERROR: anytun*: Packages must not put anything under /var/run 2024-04-20 16:04:15 seems valid 2024-04-20 16:09:07 omni: That's the thing i have been calling the FHS saga 2024-04-20 16:09:09 It has the potential to become the LFS64 of Alpine 3.20 2024-04-20 16:09:24 It all started with https://gitlab.alpinelinux.org/alpine/abuild/-/merge_requests/254 2024-04-20 16:10:11 Even sircbot is affected 2024-04-20 16:10:15 algitbot: Please protest 2024-04-20 16:15:03 Hmm...icu 2024-04-20 16:15:46 for hecks sake? 2024-04-20 16:16:26 ikke: armv7 is also using distfiles.a.o/edge 2024-04-20 16:16:49 I can fix it when I'm home 2024-04-20 16:17:00 Alright 2024-04-20 16:17:46 I'm mentioning it so we don't go the usual route of renaming the file 2024-04-20 16:18:47 I think the checksum was changed to allow the MR touching icu to pass CI, but pkgrel was not bumped 2024-04-20 16:18:52 So, we didn't notice it until now 2024-04-20 16:21:06 algitbot: retry master 2024-04-20 16:21:22 Rust spam! 2024-04-20 16:21:31 algitbot: retry master 2024-04-20 16:23:52 algitbot: retry master 2024-04-20 16:24:27 algitbot: retry master 2024-04-20 16:24:34 algitbot: retry master 2024-04-20 16:27:13 algitbot: retry master 2024-04-20 16:28:07 algitbot: retry master 2024-04-20 16:28:14 algitbot: retry master 2024-04-20 16:29:05 They really love Rust 2024-04-20 16:29:30 I wonder if we can just automate the bootstrapping of Rust 2024-04-20 16:30:00 algitbot: retry master 2024-04-20 16:30:06 algitbot: retry master 2024-04-20 16:30:34 algitbot: retry master 2024-04-20 16:30:52 :) 2024-04-20 16:30:53 algitbot: retry master 2024-04-20 16:31:06 algitbot: retry master 2024-04-20 16:32:33 algitbot: retry master 2024-04-20 16:33:24 algitbot: retry master 2024-04-20 16:34:53 algitbot: retry master 2024-04-20 16:35:20 algitbot: retry master 2024-04-20 16:35:23 Ok, will be going AFK now, bye, hopefully my internet connection gets better tomorrow 2024-04-20 18:14:44 algitbot: retry master 2024-04-20 18:17:54 fcolista: fix or remove anytun? 2024-04-20 18:21:01 algitbot: retry master 2024-04-20 18:22:02 algitbot: retry master 2024-04-20 18:23:35 fyi, if the builders idle, I'm bootstrapping rust 2024-04-20 18:25:43 ikke: thanks for doing that, was gonna ask if someone wanted to 2024-04-20 18:28:40 bootstrapping rust on armv7 now 2024-04-20 18:30:38 algitbot: retry master 2024-04-20 18:30:39 bootstrapping rust on aarch64 2024-04-20 18:31:10 woa, rv64 3-20 builder seems to be tripping 2024-04-20 18:37:33 bootstrapping rust on s390x 2024-04-20 18:39:12 bootstrapping rust on x86_64 2024-04-20 18:57:04 bootstrapping rust on x86 2024-04-20 19:05:41 boostrapping rust on build-3-20-riscv64 2024-04-20 19:06:34 cely: re /var/run the changre is technically correct and we need clean up the the apkbuilds 2024-04-20 19:06:49 do you think we should revert the chance similar to what I did with /run? 2024-04-20 19:07:01 we could revisit it post 3.20 2024-04-20 19:07:59 we could also forward all those issues to sertonix and let they deal with them 2024-04-20 19:09:16 I think that if you propose and do a change, and have strong feelings about the change, that is all fine, as long as you deal with the consequences 2024-04-20 19:09:38 im no fan of person 1 introduce change and person 2 deal with the consequences 2024-04-20 19:10:26 im all ok with changes, and improvements, as long as you take responsibility and deal with the consequences 2024-04-20 19:10:48 is /var/run/things even packaged anyway? 2024-04-20 19:10:56 apparently they are 2024-04-20 19:11:26 at least /run was. openldap. and it blocked bootstrap of 3.20 builders 2024-04-20 19:11:40 whic h is why i reverted it 2024-04-20 19:11:58 cely: let me know if you'd like to revert the /var/run thin in abuild 2024-04-20 19:13:21 in my book the opinion of the people doing the actual work matters more than the opinion of "experts" watching on the side line 2024-04-20 19:13:49 have a nice weekend everyone! 2024-04-20 19:14:10 I mean, I don't find anything here https://pkgs.alpinelinux.org/contents?file=*&path=%2Fvar%2Frun%2F*&name=&branch=edge 2024-04-20 19:14:22 so I'm not sure it's actually put in the package? 2024-04-20 19:16:08 it's directories, not files 2024-04-20 19:16:17 and iirc pkgs.a.o/contents lists only files 2024-04-20 19:16:25 oh, ok 2024-04-20 19:16:42 https://tpaste.us/5Xy9 2024-04-20 19:16:47 there's.. quite a few 2024-04-20 19:16:47 but those directories should rather be created by init files or someting then? 2024-04-20 19:17:00 guess so 2024-04-20 19:17:41 they shouldn't survive a reboot anyway 2024-04-20 19:18:14 my /run is a tmpfs mount and my /var/run is a symlink to it 2024-04-20 19:18:30 ncopa: have a nice weekend you too! 2024-04-20 19:18:37 same here, but then there's stuff like docker containers 2024-04-20 19:19:07 oh, is anyone using that? 2024-04-20 19:19:10 :B 2024-04-20 19:19:33 they can add it in their ContainerFile 2024-04-20 19:20:52 If you install a package, it should be able to run, with or without service 2024-04-20 19:26:17 armv7 done 2024-04-20 19:27:26 aarch64 done 2024-04-20 19:41:05 algitbot: retry master 2024-04-20 20:09:31 x86_64 done 2024-04-20 20:13:48 x86 done 2024-04-20 21:02:28 wth, the failing onefetch test should already be disabled for riscv64 2024-04-20 21:45:30 https://github.com/unicode-org/icu/releases/tag/release-75-1 2024-04-20 21:45:52 but it's checksum issues, right? 2024-04-20 21:50:56 looks like checksum issues indeed 2024-04-20 22:10:02 I'll fix 2024-04-20 22:13:52 but I don't get other checksums... 2024-04-20 23:54:21 omni: icu is due to those archs using distfiles.a.o/edge 2024-04-20 23:55:49 so we need to use temporary names of the files or clean them out from distfiles? 2024-04-20 23:56:28 otherwise we could upgrade !64571 2024-04-20 23:56:48 Does it require a big rebuild? 2024-04-20 23:56:59 I kind of remember something like that for the previous icu 2024-04-20 23:58:34 https://git.alpinelinux.org/aports/log/?qt=grep&q=icu 2024-04-20 23:58:42 includes all the web browsers 2024-04-21 00:00:17 Well, i've raised this distfiles.a.o/edge issue with ikke, so maybe we can wait a bit for those builders to be switched to distfiles.a.o/v3.20 2024-04-21 08:43:43 algitbot: retry master 2024-04-21 10:23:32 algitbot: retry master 2024-04-21 12:51:52 bootstrapping rust on ppc64le 2024-04-21 18:42:02 we could remove rippled, seems like no-one has been interested in upgrading it for a couple of years 2024-04-21 18:42:26 (I'm also biased) 2024-04-21 18:43:54 there seem to be multiple missing symbols 2024-04-21 19:24:36 algitbot: retry master 2024-04-22 12:35:59 kfail 2024-04-22 14:48:20 Ugh, I should making a new branch be the first thing I do so I don't accidentally push to master 😒 2024-04-22 17:38:00 the gnutls failures seem like something to do with botan 2024-04-22 18:19:30 logrotate is sticky bit 2024-04-22 18:54:45 ... s/log/test/ 2024-04-22 18:54:46 well 2024-04-22 18:54:47 whatever 2024-04-22 18:56:53 I tried to look at procps-ng, but it's not that clear what is failing 2024-04-22 18:57:59 strace output: https://tpaste.us/QR9Q 2024-04-22 19:00:54 ikke: i suppose the secret would be to stare at the test for a few hours 2024-04-22 19:01:50 but also, the strace is very verbose 2024-04-22 19:01:57 how about running the test by itself 2024-04-22 19:02:13 ( trying that right now ) 2024-04-22 19:02:37 ptrc: that does not give more details 2024-04-22 19:02:45 oh? 2024-04-22 19:02:56 PASS: procps_pids_new() info=NULL returns -EINVAL 2024-04-22 19:02:58 PASS: procps_pids new then unref 2024-04-22 19:03:00 FAIL: check_fatal_proc_unmounted 2024-04-22 19:03:02 FAIL library/tests/test_pids (exit status: 1) 2024-04-23 03:37:00 3.20 armv7 has moved onto community :) 2024-04-23 03:37:33 Now armhf and aarch64 too 2024-04-23 03:38:51 Probably s390x too, soon 2024-04-23 03:39:20 procps-ng tests are failing without any detail why 2024-04-23 03:39:34 (not even in test-suite.log) 2024-04-23 03:40:17 That's not good 2024-04-23 03:41:12 https://gitlab.com/procps-ng/procps/-/blob/master/library/tests/test_pids.c?ref_type=heads#L57-61 2024-04-23 04:07:37 fixed DISTFILES_MIRROR on rv64 2024-04-23 04:08:06 Thanks 2024-04-23 04:08:29 There also seems to be something about the armhf numbers 2024-04-23 04:08:59 in what sense? 2024-04-23 04:09:07 aports built: 55/5366 | aports total: 60/5372 2024-04-23 04:09:50 I tried to grep the history for this channel to find out if something had failed, and it had retried 2024-04-23 04:10:08 but couldn't find any, and the last error column for that on build.a.o is also empty for me 2024-04-23 04:12:44 Ok, i can't find anything in the channel log for armhf 2024-04-23 04:12:51 So, probably something failed, but it wasn't reported here 2024-04-23 04:13:06 It starts with 1/5312 60/5372 community/msgsl 4.0.0-r0 2024-04-23 04:13:13 so apparently 59 packages have already been built 2024-04-23 04:13:28 (buildrepo -n output) 2024-04-23 04:15:52 Does that mean something is wrong? 2024-04-23 04:22:05 celie: go failed to build (needs bootstrap) 2024-04-23 04:22:34 not sure why it was not reported here though 2024-04-23 04:22:56 I can't find anything for armhf in the past 12 hours in the channel log 2024-04-23 04:23:15 for build-3-20-armhf 2024-04-23 04:24:52 "greetings from ptrc and thanks for all the fish" 2024-04-23 04:24:58 Except for the uploaded main message (which i didn't see for armv7 instead) 2024-04-23 04:25:08 what do you mean? 2024-04-23 04:25:44 regarding that greetings thing 2024-04-23 07:05:41 celie: that message was used to ping the builders every few minutes to retry building stuff 2024-04-23 07:23:39 ptrc: Through MQTT? 2024-04-23 07:25:03 I mean, i thought a very specific message had to be sent, so any message will work? 2024-04-23 07:28:09 any message works 2024-04-23 07:28:16 as long as it's the right topic 2024-04-23 07:28:31 Ok 2024-04-23 08:51:48 maybe py3-scikit-plot too had issues with py3-scipy 1.13.0? 2024-04-23 08:53:10 this I think https://github.com/reiinakano/scikit-plot/pull/120 2024-04-23 09:04:11 !64709 2024-04-23 09:05:24 algitbot: retry master 2024-04-23 09:27:11 algitbot: retry master 2024-04-23 09:31:01 algitbot: retry master 2024-04-23 09:37:13 algitbot: retry master 2024-04-23 09:37:24 no go-bootstrap package 2024-04-23 09:49:20 algitbot: retry master 2024-04-23 10:08:44 finally py3-cytoolz pass on riscv64 2024-04-23 10:18:02 algitbot: retry master 2024-04-23 10:20:23 algitbot: retry master 2024-04-23 10:43:17 ptrc: oh-no 2024-04-23 10:43:33 algitbot: retry master 2024-04-23 10:43:35 that's a classic 2024-04-23 10:43:58 every single CI download from ftp.mozilla.org fails like this maybe 30% of the time 2024-04-23 11:01:38 algitbot: retry master 2024-04-23 11:01:39 that's ten times their web browser market share 2024-04-23 11:01:44 algitbot: retry master 2024-04-23 11:01:44 Here we go with the element-web on armhf retry frenzy 2024-04-23 11:01:46 algitbot: retry master 2024-04-23 11:01:46 algitbot: retry master 2024-04-23 11:01:48 algitbot: retry master 2024-04-23 11:01:48 algitbot: retry master 2024-04-23 11:01:50 algitbot: retry master 2024-04-23 11:01:50 algitbot: retry master 2024-04-23 11:01:52 algitbot: retry master 2024-04-23 11:01:52 So, apparently the libjasper.a undefined reference to jpc_decode/encode/validate in jas_init.c, undefined reference to jpc_decode/encode in jp2_dec/enc.c solved themselves after some retries? 2024-04-23 11:01:54 Oh no, i think 2.4.15.4 and 2.4.15.5 of mod_auth_openidc got yanked... 2024-04-23 11:01:54 algitbot: retry master 2024-04-23 11:01:56 Tags are still on Github, but the APKBUILD uses the release artifacts which are gone 2024-04-23 11:01:58 algitbot: retry master 2024-04-23 11:02:20 algitbot: retry master 2024-04-23 11:02:24 algitbot: retry master 2024-04-23 11:02:26 Ok, a retry fixed it 2024-04-23 11:02:26 Can't connect to static.crates.io, that's not good.. 2024-04-23 11:02:28 without first* 2024-04-23 11:02:28 I wonder if removing -D_GLIBCXX_ASSERTIONS=1 will allow webkit2gtk to build on x86, but that's probably not something that can be removed without looking closer into the issue 2024-04-23 11:02:38 can you do that elsewhere? 2024-04-23 11:02:58 algitbot: retry master 2024-04-23 11:03:00 could be good to have some backlog at least 2024-04-23 11:03:00 algitbot: retry master 2024-04-23 11:03:02 algitbot: retry master 2024-04-23 11:03:06 You're welcome 2024-04-23 11:03:06 nmeum: github-cli also has a ppc64le workaround (it disables CGO for that arch) 2024-04-23 11:03:08 algitbot: retry master 2024-04-23 11:03:12 algitbot: retry master 2024-04-23 11:03:44 algitbot: retry master 2024-04-23 11:03:58 algitbot: retry master 2024-04-23 11:04:00 Rust 1.77.0 is available for ARM \o/ 2024-04-23 11:04:02 That leaves only s390x and riscv64 2024-04-23 11:04:02 and ppc64le 2024-04-23 11:04:03 celie: oh that should no longer be needed, feel free to patch it out :) 2024-04-23 11:04:54 algitbot: retry master 2024-04-23 11:05:16 algitbot: retry master 2024-04-23 11:05:16 algitbot: retry master 2024-04-23 11:05:18 algitbot: retry master 2024-04-23 11:05:18 algitbot: retry master 2024-04-23 11:05:19 celie: ping 2024-04-23 11:05:20 algitbot: retry master 2024-04-23 11:05:22 algitbot: retry master 2024-04-23 11:05:22 algitbot: retry master 2024-04-23 11:05:24 algitbot: retry master 2024-04-23 12:35:29 Sorry, that message about github-cli was my bouncer going haywire and replaying old messages 2024-04-23 13:26:40 algitbot: retry master 2024-04-23 15:52:10 bootstrapping ^ 2024-04-23 15:58:16 About 70 aports left for edge-riscv64 community 2024-04-23 16:12:00 done 2024-04-23 16:30:40 celie: oh, bouncer issues 2024-04-23 16:31:33 The bouncer was bouncing 2024-04-23 16:41:53 algitbot: retry master 2024-04-23 17:53:53 algitbot: retry master 2024-04-23 18:28:49 algitbot: retry master 2024-04-23 22:36:29 algitbot: retry master 2024-04-23 22:36:46 is build-edge-riscv64 stuck? 2024-04-24 00:14:12 algitbot: retry master 2024-04-24 00:14:50 algitbot: retry master 2024-04-24 00:48:46 algitbot: retry master 2024-04-24 01:39:35 algitbot: retry master 2024-04-24 01:42:39 algitbot: retry master 2024-04-24 01:45:42 15 more aports for build-3-20-riscv64 main 2024-04-24 01:51:02 algitbot: retry master 2024-04-24 01:57:17 I think qt5-qtwebengine also needs the chromium Python patches 2024-04-24 01:58:31 Also, isa-l may be failing due to some of the TeX changes, now trying to build isa-l locally to see what the log says 2024-04-24 01:58:43 algitbot: retry master 2024-04-24 02:00:41 "mktexfmt [ERROR]: returning error due to option --strict" 2024-04-24 02:06:56 algitbot: retry master 2024-04-24 02:07:34 algitbot: retry master 2024-04-24 02:09:35 algitbot: retry master 2024-04-24 02:12:08 Hmm, build-3-20-armv7 just started building utf8proc, irssi, emacs..now that's something :) 2024-04-24 02:13:12 Hope emacs doesn't get stuck 2024-04-24 02:13:27 algitbot: retry master 2024-04-24 02:14:51 algitbot: retry master 2024-04-24 02:15:52 algitbot: retry master 2024-04-24 02:20:56 algitbot: retry master 2024-04-24 02:21:08 :O 2024-04-24 02:21:32 build-3-20-aarch64 has built OCaml! 2024-04-24 02:26:16 algitbot: retry master 2024-04-24 02:26:44 7 aports left for riscv64 main, and it seems it keeps wanting to build numactl and ccache 2024-04-24 02:27:27 numactl's test/distance used to be disabled for x86 too, maybe it should now be disabled for riscv64 2024-04-24 02:27:32 -too 2024-04-24 02:28:30 but maybe test/distance's segfault can be looked into first, as i'm not sure if that's the same reason it was disabled for x86 2024-04-24 02:29:25 algitbot: retry master 2024-04-24 02:29:30 Come on riscv64, build something else 2024-04-24 02:29:37 Ugh, numactl again 2024-04-24 02:29:57 algitbot: retry master 2024-04-24 02:30:36 algitbot: retry master 2024-04-24 03:04:12 5 more aports for riscv64 main (including numactl) 2024-04-24 03:35:36 Hmm, we have community/retawq, never noticed that before 2024-04-24 03:35:55 and fossil 2.23 has built successfully on s390x 2024-04-24 03:36:15 Unlike !64721 2024-04-24 03:46:07 Last aport for riscv64 main 2024-04-24 03:51:58 algitbot: retry master 2024-04-24 03:54:45 algitbot: retry master 2024-04-24 04:16:48 algitbot: retry master 2024-04-24 04:24:34 algitbot: retry master 2024-04-24 04:25:02 algitbot: retry master 2024-04-24 04:31:59 algitbot: retry master 2024-04-24 04:40:40 algitbot: retry master 2024-04-24 04:48:33 algitbot: retry master 2024-04-24 05:34:59 algitbot: retry master 2024-04-24 05:35:21 build-3-20-aarch64 is building nim 2024-04-24 06:15:13 Codeberg strikes again with checksum changes (or maybe upstream just retagged the release) 2024-04-24 06:29:29 algitbot: retry master 2024-04-24 07:25:04 algitbot: i'm sorry 2024-04-24 10:16:38 50 more aports for build-edge-riscv64 community 2024-04-24 11:27:09 40 more aports for build-edge-riscv64 community 2024-04-24 13:03:54 py3-corner: Matplotlib is not built with the correct FreeType version to run tests. Rebuild without setting system_freetype=1 in mplsetup.cfg. Expect many image comparison failures below. Expected freetype version 2.12.1. Found freetype version 2.13.2. Freetype build type is not local 2024-04-24 13:04:03 says the log 2024-04-24 13:12:59 About 20 more aports to go for build-edge-riscv64 community 2024-04-24 13:13:43 yes, but now it will fail building tracker 2024-04-24 13:15:55 i wonder why tracker is so broken on riscv64 2024-04-24 13:16:00 also, hi y'all 2024-04-24 13:16:15 it's broken on all since the sqlite upgrade 2024-04-24 13:16:26 ah, makes more sense 2024-04-24 13:16:39 !64777 2024-04-24 13:17:00 I think only the patch is needed, but why not upgrade too? 2024-04-24 13:17:30 oh, that's an awfully simple patch 2024-04-24 13:17:35 i'd say merge it 2024-04-24 13:17:52 especially since the patch MR is merged into master of tracker 2024-04-24 13:17:54 Hi, ptrc, would you like to takeover running the algitbot retry loop? 2024-04-24 13:17:59 go ahead, I dare you :D 2024-04-24 13:17:59 sure :) 2024-04-24 13:18:28 omni: oh? 2024-04-24 13:18:47 is there something i'm missing? 2024-04-24 13:18:55 no, jk 2024-04-24 13:19:16 omni probably means..do you dare to merge it :P 2024-04-24 13:19:22 for a second i wondered if there's some soname bump and it would break half of aports lol 2024-04-24 13:19:39 sure, if all hell breaks loose, that's on me :p 2024-04-24 13:19:58 no, I don't think there's anything, but why not hav someone other than me look at it and merge 2024-04-24 13:20:06 fair enough, yeah 2024-04-24 13:20:16 afaict they moved a lot of stuff from /usr/libexec to /usr/bin 2024-04-24 13:20:19 guess that's fine 2024-04-24 13:20:37 haven't seen any project use libexec of other package really 2024-04-24 13:21:09 as expected, riscv64 failed ocrmypdf tests again 2024-04-24 13:22:19 mopidy looks like it will fail the same tests on all architectures if rebuilt !64778 2024-04-24 13:22:59 Apparently i got so frustrated with !64775, i marked it as Draft twice 2024-04-24 13:23:40 draft of a draft, for when you *really* don't want something to be merged :D 2024-04-24 13:24:57 for riscv64 community/ I think there is mopidy and ocrmypdf left to take care of, maybe some other.. 2024-04-24 13:24:59 the cargo-edit stuff looks like it just needs an update of the expected output 2024-04-24 13:27:01 I need to do some other stuff for a bit, don't hold off from fixing mopidy just because I have an MR open 2024-04-24 13:28:00 nor ocrmypdf !64743 2024-04-24 13:32:14 I wonder if the expected output thing was the same thing that was patched for cargo-auditable 2024-04-24 13:32:48 or closely related 2024-04-24 13:32:57 https://github.com/killercup/cargo-edit/commit/67e110a957d289b3fbd65bcb7bf243b7fe685df0 2024-04-24 13:32:58 looks like this 2024-04-24 13:33:00 i think 2024-04-24 13:33:00 ? 2024-04-24 13:36:59 I'll give it a try in a bit, thanks 2024-04-24 13:40:24 Hopefully the upgrade fixes the afl++ build issues on riscv64 2024-04-24 13:41:53 18 more aports for build-edge-riscv64 community 2024-04-24 13:53:14 algitbot: retry master 2024-04-24 13:53:28 wait what 2024-04-24 13:53:38 i opened the "failed to build" link and it said it succeeded 2024-04-24 13:53:44 puzzled 2024-04-24 13:54:27 Haha 2024-04-24 13:54:50 Your retry loop retried it, and it succeeded before you could have a look at the error 2024-04-24 13:55:55 ohhhhh 2024-04-24 13:56:03 lmao 2024-04-24 14:00:06 For cargo-edit, with the patch you linked to: "1 passed; 35 failed" 2024-04-24 14:00:17 No wonder i made it a draft of a draft 2024-04-24 14:00:37 I wonder why upstream doesn't just make a new release to make things work with new cargo 2024-04-24 14:01:03 or do the tests not matter, and the actual program still works as usual? 2024-04-24 14:01:56 i think the program itself works, it's just that cargo outputs different stuff when handed broken Cargo.toml files 2024-04-24 14:02:13 and barely anybody cares about tests when they just do `cargo install cargo-edit` 2024-04-24 14:02:52 Probably a candidate for having tests disabled (at least for now) then :) 2024-04-24 14:06:56 Oh no, something else that's failing on riscv64 community 2024-04-24 14:08:16 "cannot import name 'interp' from 'scipy'" 2024-04-24 14:08:40 hm, should have failed everywhere 2024-04-24 14:08:44 unless it's the timing 2024-04-24 14:09:40 Scipy was upgraded in the meantime 2024-04-24 14:09:50 it builds for me on x86_64 though 2024-04-24 14:10:13 unless i'm like, a day behind with packages 2024-04-24 14:10:18 Oh, and downgraded again 2024-04-24 14:10:26 Probably will need a version constraint 2024-04-24 14:10:35 like py3-chaospy 2024-04-24 14:13:12 oh hhu 2024-04-24 14:13:17 huh* 2024-04-24 14:13:21 right 2024-04-24 14:13:54 Both versions exist in the index now 2024-04-24 14:18:14 Anyway, i'll let you add that since i'm now fighting with a really lousy internet connection, and will be going AFK shortly 2024-04-24 14:21:36 ^ !64774 2024-04-24 14:31:33 Thanks 2024-04-24 14:31:57 Going AFK now as the connection is not getting any better (Gitlab web UI is virtually unusable) 2024-04-24 14:32:00 Bye 2024-04-24 14:32:30 \o 2024-04-24 14:32:39 thanks for the fixes :) 2024-04-24 14:32:47 You're welcome 2024-04-24 14:47:33 oh, it's that one 2024-04-24 15:01:40 I saw that interp thing the other day 2024-04-24 15:02:02 where did I have it... 2024-04-24 15:02:51 ah, !64709 2024-04-24 15:03:40 keep that in mind for when we upgrade py3-scipy again 2024-04-24 15:14:26 davfs2 so picky 2024-04-24 15:14:34 also yeah i'm getting 3000+ms ping so i'll back soon 2024-04-24 15:28:18 waiting for 3-20 builders to be idle to bootstrap go 2024-04-24 19:11:15 I'd love to be able to bootstrao go 2024-04-24 23:56:52 algitbot: kick build-edge-riscv64 2024-04-25 01:06:42 lua-resty-hmac: !64730 2024-04-25 02:11:07 omni: Thanks for fixing up !64752 for me 2024-04-25 02:20:18 cely: sure, was just surprised by the need of autoconf and automake 2024-04-25 02:20:28 Yes, same here 2024-04-25 02:29:41 I think i have found something, are you still here? 2024-04-25 02:29:59 yea 2024-04-25 02:41:31 !64820 (i will add the commit with what i think is the fix shortly) 2024-04-25 02:58:17 Ok, if you're still here, then i would appreciate it it you took a look at !64820 and see if i've missed anything out 2024-04-25 02:58:28 I am not planning to merge it but assign it to @HRio 2024-04-25 03:45:14 I'm not adding the "type:buildozer-build-failure" label to my MRs that fix issues found by the 3.20 builders 2024-04-25 03:45:18 s/not/now/ 2024-04-25 05:41:16 ptrc: would it be possible to stop the script that spams the builders to retry? 2024-04-25 05:58:19 Oops 2024-04-25 05:58:52 I guess maybe a condition should be added to the script that stops it if some message is sent over MQTT 2024-04-25 06:00:05 Maybe we can agree on a topic for that message to be sent over 2024-04-25 06:01:49 By the way, i think build-3-20-armhf is stuck on rakudo-star 2024-04-25 06:16:42 ikke: stopped 2024-04-25 06:16:59 i hope it didn't cause any trouble 2024-04-25 06:21:07 I think ikke wants to bootstrap Go 2024-04-25 08:36:58 ptrc: no trouble, just to get some idle time to get to bootstrap some things 2024-04-25 08:37:10 openjdk as well 2024-04-25 09:07:26 ikke: is it openjdk8? perhaps Simon F could help? 2024-04-25 09:09:00 and how much idle do you mean by idle? 2024-04-25 09:09:22 omni: openjdk needs to be bootstrapped since a while 2024-04-25 09:11:02 I have postponed merging !64873, thinking I would do that after build-3-19-x86_64 is done with its chromium build, that will take some time and resources on the builders 2024-04-25 09:11:25 algitbot: hello? 2024-04-25 09:11:34 https://gitlab.alpinelinux.org/alpine/aports/-/merge_requests/64783 2024-04-25 09:17:14 !64783 2024-04-25 09:17:46 was it my comma? 2024-04-25 09:18:10 No, you switched the 3rd and 4th digits 2024-04-25 09:18:27 !64873, 2024-04-25 09:18:32 !64783, 2024-04-25 09:19:04 oh, digits, they all look the same to me 2024-04-25 09:50:01 :O 2024-04-25 10:01:48 "No file descriptors available" for ocaml, i wonder if we'll have to add a call to ulimit, or if a re-run will fix it 2024-04-25 10:58:05 algitbot: retry 3.19-stable 2024-04-25 11:42:42 bootstrapping go on s390x 2024-04-25 11:42:57 glhf! 2024-04-25 11:44:40 bootstrapping ^ 2024-04-25 12:10:00 bootstrapping ^ 2024-04-25 12:12:40 ptrc: ^ 2024-04-25 12:13:04 oh, that was jirutka 2024-04-25 12:14:17 to be fair, it did happen most recently to me ( twice ) 2024-04-25 12:14:24 but yeah, wasn't me this time :p 2024-04-25 12:56:49 So, how are we dealing with checksum errors again? (i kinda forgot what we did last time) 2024-04-25 12:56:50 Sure, we have to update the checksum, but do we rename the file as well? 2024-04-25 12:56:55 If we don't bump pkgrel, then the aport won't be rebuilt on edge 2024-04-25 12:57:10 I mean, so technically, we could have the same file name but with the old checksum on distfiles.a.o/edge, and the new checksum on distfiles.a.o/v3.20 2024-04-25 13:19:04 Also, build-3-20-armhf is most definitely stuck on rakudo-star 2024-04-25 13:19:08 my away or the & 2024-04-25 13:19:09 ^ 2024-04-25 13:19:25 Err..what? 2024-04-25 13:19:35 Oh 2024-04-25 13:19:37 The highway failure 2024-04-25 13:20:04 Thanks 2024-04-25 13:20:17 So, are you bootstrapping anything now? 2024-04-25 13:21:01 seems like everything finished 2024-04-25 13:21:26 All the go, openjdk, and ghc bootstrapping are done? 2024-04-25 13:21:45 not globally 2024-04-25 13:21:53 on some arches 2024-04-25 13:22:05 openjdk8 is currently building on armv7 2024-04-25 13:23:38 Ok 2024-04-25 13:25:27 (I typically just bootstrap what it's trying to build, not ahead of time bootstrapping everything) 2024-04-25 13:27:44 I figured that was it when you pointed to the failure, and said "bootstrapping" 2024-04-25 13:27:56 yup 2024-04-25 13:31:57 algitbot: retry master 2024-04-25 13:34:34 bootstrapping go on x86 2024-04-25 13:38:09 ^ :-) 2024-04-25 13:39:31 Does this mean i should avoid pushing anything when you are bootstrapping? 2024-04-25 13:40:33 I stop the builder when I bootstrap 2024-04-25 13:40:53 Ok 2024-04-25 14:17:16 ^ bootstrap 2024-04-25 14:18:17 oh, uhm, hope chromium doesn't get in the way, or at least for too long 2024-04-25 14:19:17 but from what I understand it is currently not usable since a few days #16047 2024-04-25 14:19:31 No chromium on armhf! 2024-04-25 14:19:47 So, what do we do about memtester? 2024-04-25 14:20:16 When abuild depended on curl, pyropus.ca redirecting to pyropus.ca. wasn't a problem 2024-04-25 14:20:28 ah, so go is bootstrapped on the other architectures? 2024-04-25 14:20:45 On most it should be 2024-04-25 14:20:45 Now busybox wget/ssl_client fails to verify the certificate 2024-04-25 14:20:58 or at least aarch64, armv7 and x86_64 2024-04-25 14:21:06 I guesst the quickfix that was suggested should work 2024-04-25 14:21:18 Adding curl back into makedepends? 2024-04-25 14:21:34 yup 2024-04-25 14:34:47 Oh great 2024-04-25 14:35:13 I saw netpbm on build.a.o, and thought "i hope it doesn't fail tests" 2024-04-25 14:35:32 you jinxed it! :D 2024-04-25 14:36:15 jinxable tests will be jinxed 2024-04-25 14:36:52 Ah, the package where the normal build output is filled with errors and failures 2024-04-25 14:41:17 failed: 117 matches, error: 877 matches 2024-04-25 14:41:49 For netpbm? 2024-04-25 14:42:00 (Lousy internet connection will not allow me to finish loading the build log) 2024-04-25 14:42:07 ## This test fails when ghostscript is not available. 2024-04-25 14:42:10 ps-roundtrip.test: FAILURE 2024-04-25 14:43:53 https://tpaste.us/YYyP 2024-04-25 14:43:55 The funny thing is, it eventually passes after enough retries (and maybe when the builder is less busy) 2024-04-25 14:44:18 ugh 2024-04-25 14:45:23 What happened? 2024-04-25 14:45:34 Nothing, just anoyed by flaky tests 2024-04-25 14:46:35 Yeah 2024-04-25 14:46:49 I think the netpbm tests have been brough up from time to time 2024-04-25 14:47:02 but nothing's been done because they eventually pass 2024-04-25 14:47:08 yup 2024-04-25 14:48:27 uhh... 2024-04-25 14:49:22 not super motivated to work on it but I will 2024-04-25 14:50:48 Work on qt5-qtwebengine...or netpbm? 2024-04-25 14:57:49 cely: the web client monster 2024-04-25 14:58:26 62% tests passed, 196 tests failed out of 522 2024-04-25 14:58:42 should we disable highway on rv64? 2024-04-25 14:58:52 The web monster vs the test monster 2024-04-25 15:00:04 We'll have to disable libjxl too, as it depends on highway 2024-04-25 15:00:47 and quite some things depend on libjxl 2024-04-25 15:01:07 and there's a lot of things that (conditionally) have libjxl in makedepends 2024-04-25 15:02:34 I think we should disable main/musl and call it a day 2024-04-25 15:03:15 possibly related https://github.com/google/highway/issues/2078 2024-04-25 15:03:22 omni: would safe us a lot of work :p 2024-04-25 15:03:35 or disable main/busybox 2024-04-25 15:03:35 (not meant as anti-musl, just the first central, essential thing I could think of) 2024-04-25 15:04:04 omni: yeah, figured 2024-04-25 15:09:35 valkey...hmmm 2024-04-25 15:09:54 Need to ping Jirutka about valkey? 2024-04-25 15:10:46 Probably a test failure, but i don't think my internet connection will allow me to finish loading the build log 2024-04-25 15:11:57 I believe it's this: valkey 7.2.5 crashed by signal: 11, si_code: 1 2024-04-25 15:12:39 cely: usually an issue on gitlab would suffice 2024-04-25 15:13:06 Hmm, i think i've seen quite a bit of signal 11's on riscv64 2024-04-25 15:13:28 not sure why 2024-04-25 15:13:43 other than that it's an uncommon architecture 2024-04-25 15:14:02 Someone else will have to make that issue, i'm still fighting with my internet connection over git push 2024-04-25 15:14:21 cely: annoying 2024-04-25 15:14:29 It's been at it for almost 5 minutes 2024-04-25 15:19:55 #16056 2024-04-25 15:20:22 Thanks 2024-04-25 15:29:45 The MR i've been trying to open for the past 20 minutes (can't believe it's taken so long): !64865 2024-04-25 15:32:12 oof 2024-04-25 15:33:29 Yeah, my connection to gitlab.a.o gets worse as the day progresses 2024-04-25 15:33:30 from morning to evening 2024-04-25 15:34:37 Sad to hear 2024-04-25 15:34:49 Hope it gets fixed soon 2024-04-25 15:41:50 Hope it as well 2024-04-25 15:42:13 algitbot: retry master 2024-04-25 17:42:17 algitbot: retry master 2024-04-26 00:19:19 algitbot: retry master 2024-04-26 05:10:20 Hi, apostnikov on a new Matrix homeserver 2024-04-26 05:10:41 yep, experimenting 2024-04-26 05:10:49 Conduit? 2024-04-26 05:12:55 (Some random notes: urfkill will need some autoconf-fu; tau depends on libhandy-dev which has been removed, i think; oneVPL has been renamed libvpl) 2024-04-26 05:15:59 I'm handling onevpl and vips in !64909, among others 2024-04-26 05:18:33 celeste: synapse ATM looks more or less predictable but tons of bugs everywhere( 2024-04-26 05:22:01 Is ghuet (from the same homeserver) in #alpine-devel you as well? 2024-04-26 05:26:46 yes 2024-04-26 05:51:14 cargo-outdated will probably need RUST_MIN_STACK increased as well (due to the p384 in debug mode on 32-bit archs issue) 2024-04-26 06:34:25 tcl-lib building again on ppc64le, let's hope it passes tests this time 2024-04-26 06:41:17 tcl-lib test passed on ppc64le :) 2024-04-26 06:51:03 aarch64 now building ppc64le 2024-04-26 07:14:02 algitbot: retry master 2024-04-26 07:18:12 OCaml failed again :/ 2024-04-26 07:19:22 Same with Netpbm 2024-04-26 09:00:17 algitbot: retry 3.17-stable 2024-04-26 10:21:39 aarch64 was building tcl-lib* (in case anyone reads the log and wonders why i said something so nonsensical 3.5 hours ago) 2024-04-26 13:30:41 algitbot: retry master 2024-04-26 14:53:26 himalaya is also affected by the p384 stack overflow in debug mode on 32-bit archs issue 2024-04-26 15:01:39 what is p384? 2024-04-26 15:01:57 A Rust crate 2024-04-26 15:03:07 Issues with it was discussed in !58129 2024-04-26 15:03:22 and it seems setting RUST_MIN_STACK to a high number solves the stack overflow 2024-04-26 15:12:58 Hmm, riscv64 testing is down to almost 60 aports left 2024-04-26 15:13:14 though there are still some blockers 2024-04-26 15:23:13 algitbot: retry master 2024-04-26 16:17:31 algitbot: retry master 2024-04-26 16:32:12 algitbot: retry master 2024-04-26 16:33:02 algitbot: retry master 2024-04-26 16:34:46 algitbot: retry master 2024-04-26 16:40:47 algitbot: retry master 2024-04-26 16:46:22 algitbot: retry master 2024-04-26 16:52:22 algitbot: retry master 2024-04-26 17:16:22 algitbot: retry master 2024-04-26 21:02:05 algitbot: retry master 2024-04-26 21:13:59 failscale 2024-04-26 21:24:32 algitbot: retry master 2024-04-26 23:38:37 algitbot: retry master 2024-04-26 23:41:29 algitbot: retry master 2024-04-26 23:50:53 algitbot: retry master 2024-04-27 00:02:34 algitbot: retry master 2024-04-27 00:23:03 29 aports left for riscv64 testing 2024-04-27 00:34:56 3 archs have built over 4000 aports for 3.20: aarch64, armv7, and s390x 2024-04-27 00:35:10 over 4000 in community, that is 2024-04-27 00:37:02 mmm 2024-04-27 00:44:53 Hi, omni 2024-04-27 00:49:26 hi 2024-04-27 00:49:53 I didn't manage to get myself to work much more on qt5-qtwebengine 2024-04-27 00:51:37 Hopefully you'll find time later on 2024-04-27 00:55:17 yeah, or energy + motivation + focus 2024-04-27 00:55:31 or just wait long enough for upstream to fix it 2024-04-27 00:55:38 but that may not happen before 3.20 2024-04-27 00:58:13 I wonder if you could justify vendoring Python 3.11 in the QtWebEngine build system 2024-04-27 00:59:21 I think it already vendors the real Ninja 2024-04-27 00:59:48 which i think had some breaking changes in the latest version too 2024-04-27 01:00:10 so if you add ninja-is-really-ninja now you'd run into problems 2024-04-27 01:06:40 I'd rather like to unvendore everything 2024-04-27 01:06:50 but that's probably not possible 2024-04-27 01:08:35 If you can have multiple versions of aports, then you can devendor stuff 2024-04-27 01:09:54 Just like the mbedtls2 thing, i expect that when we do decide to remove it (when upstream pulls support), all the aports that still need it would switch to vendored mbedtls2 (after all, it will no longer receive fixes) 2024-04-27 01:15:07 with 87-based qtwebengine-chromium, I got the impression that some of the vendored stuff would need to be upgraded to versions that we already have packaged 2024-04-27 01:15:24 like six 1.6.0, for example 2024-04-27 01:15:44 for it to build 2024-04-27 01:16:21 Hmm, ok 2024-04-27 01:37:22 I will never understand games with a few hundred MB source tarballs 2024-04-27 01:42:30 So, i downloaded the 400+ MB wesnoth because build logs said it failed checksums 2024-04-27 01:42:37 but the checksum is the same 2024-04-27 01:42:53 Then i looked closer, and it seems the builder downloaded something 859 bytes instead 2024-04-27 01:43:15 Whatever 2024-04-27 01:58:13 Netpbm again :( 2024-04-27 02:03:49 build-3-20-aarch64 now building chromium 2024-04-27 02:14:40 I think armhf has gotten stuck on rebar3 2024-04-27 02:15:03 I've probably seen that happen before 2024-04-27 02:15:10 Oops 2024-04-27 02:15:12 Spoke too soon 2024-04-27 02:15:15 It's moved on now 2024-04-27 02:16:18 Now it will fail on cargo-oudated due to p384 2024-04-27 02:16:21 outdated* 2024-04-27 03:10:07 I think testing/kopia and testing/mimir will need test timeout increased 2024-04-27 03:10:54 but the riscv64 CI seems to be occupied, so i can't test it out there to see if that will actually help it pass, or just uncover some other issue 2024-04-27 03:41:35 18 more aports for riscv64 testing 2024-04-27 04:33:27 Please build something else 2024-04-27 04:33:35 that doesn't fail due to test timeouts 2024-04-27 04:33:37 LOL 2024-04-27 04:33:44 and what does it build? release-plz 2024-04-27 04:34:37 I mean, the name is right on topic, not expecting it to fail 2024-04-27 04:35:25 Hopefully the next one also doesn't fail 2024-04-27 04:35:37 That will get us down to 12 aports 2024-04-27 06:41:56 10 more aports for riscv64 testing, but i have a feeling that the majority of them are failing 2024-04-27 08:59:38 Looks like ansible-lint's checksum has changed again 2024-04-27 12:37:23 ... 2024-04-27 12:38:12 I wonder why it failed and then passed immediately upon retry 2024-04-27 12:45:46 s390x has reached the last 600 aports for 3.20 community 2024-04-27 13:07:02 !64984 2024-04-27 13:08:15 Not sure if i'll be around to see the riscv64 CI run, but if it passes there, then anyone who notices that, please merge it to help riscv64 move along 2024-04-27 13:08:30 There are still a few other failing aports for riscv64 testing 2024-04-27 15:41:32 8 more aports for riscv64 testing! 2024-04-27 15:41:40 Hopefully mimir and kopia's higher test timeout will allow them to pass (just like in the CI), which should then bring the number of aports down to 5 2024-04-27 15:43:09 and maybe the majority of that 5 is the fallout from the Py 3.12 rebuild 2024-04-27 16:19:34 Hmm, mimir is now failing some other thing (that didn't fail in CI) 2024-04-27 16:19:48 I wonder if it's from leftover files on the builder from a previous run 2024-04-27 16:20:37 Oh well, someone else will have to think of what to do to get mimir to pass (if a retry doesn't do it), as i am out of ideas 2024-04-27 17:03:42 algitbot: retry master 2024-04-27 17:23:20 algitbot: retry master 2024-04-27 18:38:59 algitbot: retry master 2024-04-27 19:16:13 algitbot: retry master 2024-04-27 19:48:20 algitbot: retry master 2024-04-27 20:46:14 jirutka needs to decide on !65004 2024-04-27 20:46:38 algitbot: retry master 2024-04-27 21:04:42 algitbot: retry master 2024-04-27 23:33:28 algitbot: rebase master 2024-04-28 00:26:26 algitbot: rebase master 2024-04-28 00:34:55 rebase master? 2024-04-28 00:47:59 uhm 2024-04-28 00:48:21 algitbot: retry master 2024-04-28 00:49:33 Py 3.12 fallout 2024-04-28 00:49:45 algitbot: refurbish mainframe 2024-04-28 01:10:13 algitbot: respin maestro 2024-04-28 01:10:23 algitbot: retry master 2024-04-28 01:23:33 algitbot: retry master 2024-04-28 01:28:15 "main/gdb: enable tui"? 2024-04-28 06:32:53 s390x is down to 120 aports for 3.20 community, so i expect it to start looping on a few failing aports 2024-04-28 06:33:22 There should still be something to build within that 120 though 2024-04-28 06:41:19 109 aports for s390x community now 2024-04-28 06:41:26 I think we'll be seeing 100 aports soon :) 2024-04-28 06:59:30 99 aports for s390x community! 2024-04-28 07:31:05 I wonder if we'll be seeing 50 aports remaining for s390x community by the end of the day 2024-04-28 07:43:03 I'm thinking how to categorize the various build failure this time around, i think i've seen issues due to: libxml2, boost1.84, py3.12, newer rust/cargo 2024-04-28 07:43:38 I think new OpenSSL as well 2024-04-28 10:15:54 algitbot: retry master 2024-04-28 11:44:10 algitbot: retry master 2024-04-28 11:45:29 I wonder if something could be made up so that aports that failed to build get lower weight 2024-04-28 12:24:52 Maybe there should just be a feature for us to queue an aport for building 2024-04-28 12:27:54 algitbot: retry master 2024-04-28 13:49:30 I wonder if the about 60 aports left for s390x community are all failing aports, plus their dependents 2024-04-28 13:50:37 but now we have 4 archs at >= 90% for 3.20 community :) 2024-04-28 13:50:58 >= 90% aports built, that is 2024-04-28 14:37:25 lilypond may be affected by the new texmf-dist changes 2024-04-28 14:39:07 57 aports left for s390x community 2024-04-28 14:47:14 Alright, going AFK now, bye 2024-04-28 15:47:07 algitbot: refry maftew 2024-04-28 15:47:32 algitbot: you need to be more fuzzy 2024-04-28 15:47:47 ACTION hugs algitbot  2024-04-28 15:47:55 algitbot: retry master 2024-04-28 15:52:54 I think i've found the Python blockers for riscv64 testing 2024-04-28 15:56:03 They are: py3-mypy-testing, py3-eventlet, py3-libmdbx, py3-riotctrl, and py3-pytest-mypy 2024-04-28 15:57:28 -py3-mypy-testing 2024-04-28 15:57:48 I was thinking of py3-pytest-mypy but typed that instead 2024-04-28 15:58:57 All 4 seem to be test failures, and this seems to be the first time we have tests enabled for them on the builders 2024-04-28 15:59:06 the riscv64 builders 2024-04-28 16:00:00 Don't know why we didn't catch those failures on the CI, probably these are the aports that were merged without running the riscv64 CI due to community not being uploaded yet 2024-04-28 16:01:40 I wonder if the tests pass on other archs, probably yes, otherwise they would've been fixed or disabled by now (and probably that's why CI was bypassed for riscv64, because it succeeded for all other archs) 2024-04-28 16:02:52 With the exception of py3-eventlet, the other 3 were last rebuilt last year 2024-04-28 16:03:10 and py3-eventlet was built this year before tests were enabled on the builder 2024-04-28 16:06:34 Besides these 4 Python aports, the other blockers for riscv64 testing should be: onnxruntime, and mimir 2024-04-28 16:06:43 Wow 2024-04-28 16:07:14 https://build.alpinelinux.org/buildlogs/build-edge-riscv64/testing/onnxruntime/ 2024-04-28 16:07:31 onnxruntime-1.71.1-r1.log is 4GB?! 2024-04-28 16:10:30 onnxruntime-1.17.1-r1.log* 2024-04-28 16:16:48 algitbot: retry master 2024-04-28 16:18:27 build.a.o lists 7 aports for riscv64 testing 2024-04-28 16:18:41 I think the 7th one is not failing, but just hasn't been built 2024-04-28 16:21:44 Yes, testing/lua-language-server 3.8.3 hasn't been built by the riscv64 builder 2024-04-28 16:23:01 algitbot: retry master 2024-04-28 16:23:09 Alright, back to being AFK now, bye 2024-04-28 18:49:17 algitbot: retry master 2024-04-28 23:54:46 algitbot: retry master 2024-04-29 00:57:44 algitbot: retry master 2024-04-29 02:17:55 algitbot: retry master 2024-04-29 07:20:53 Hmm, "Error calling recv_msg in c_check_version" for community/ts 2024-04-29 11:37:48 algitbot: hi 2024-04-29 11:38:43 algitbot: hi again 2024-04-29 11:40:48 Hmm, armhf has 100 aports left 2024-04-29 13:01:15 algitbot: retry 3.19-stable 2024-04-29 14:03:00 algitbot: retry 3.18-stable 2024-04-29 23:33:23 algitbot: retry master 2024-04-29 23:36:15 algitbot: retry master 2024-04-29 23:56:28 Will the tracker-miners build issue be fixed by !65100? 2024-04-30 01:28:25 ^ octave: "Error occurred during initialization of VM 2024-04-30 01:29:21 Unable to load native library: Error loading shared library libjvm.so: No such file or directory (needed by /usr/lib/jvm/java-1.8-openjdk/jre/lib/amd64/libjava.so)" 2024-04-30 01:30:21 Is there something different with openjdk8 on 3.20? 2024-04-30 03:39:56 algitbot: retry master 2024-04-30 04:04:30 45 aports left for s390x community 2024-04-30 05:03:44 s390x has 40 more aports, and armhf 89 more 2024-04-30 08:58:48 35 aports for s390x community, 70 for armhf 2024-04-30 12:21:10 It seems borgbackup tests want to create dirs under /var/tmp 2024-04-30 14:21:42 Ugh, build-3-20-armv7 is now compiling himalaya, i thought i had merged the fix 2024-04-30 14:22:41 Oh well, it will very likely fail this time 2024-04-30 14:22:45 There 2024-04-30 14:25:26 31 aports left for s390x community, 52 for armhf, and 88 for armv7 2024-04-30 16:17:29 fyi, bootstrapping jdk on a bunch of arches, so they don't show anything building atm