2018-02-02 14:30:08 where if fuse=3.2.0 coming from? 2018-02-02 14:30:16 s390x build pulls in fuse-dev when building ostree 2018-02-02 14:30:24 but it picks fuse 3.2.0 2018-02-02 14:33:38 ok, it is leftovers 2018-02-02 14:33:55 testing repo is not cleaned since it never completed on s390x 2018-02-05 17:56:00 ncopa : wireguard-tools and qemu on s390x are built to latest version. but they are not uploaded to repo. main repo nl.a.o still has older versions. 2018-02-05 18:11:32 where if fuse=3.2.0 coming from? 2018-02-05 18:11:32 s390x build pulls in fuse-dev when building ostree 2018-02-05 18:11:32 but it picks fuse 3.2.0 2018-02-05 18:11:32 testing repo is not cleaned since it never completed on s390x 2018-02-05 18:11:32 ok, it is leftovers 2018-02-05 18:11:43 thing is fuse is not up in s390-edge. interseting 2018-02-05 18:12:21 no build log of fuse, but the package is there in dl-cdn.a.o 2018-02-05 18:12:25 version 2.9.7 2018-02-05 21:58:34 here we go 2018-02-05 21:58:59 hey 2018-02-05 21:59:16 i think ndpi testsuite hanged 2018-02-05 21:59:26 on s390x 2018-02-05 21:59:29 i mchecking it up now 2018-02-05 22:00:09 i'd like to make the builder continue on error on s390x so it can complete everything it can 2018-02-05 22:00:26 and then we could simply disable those who does not work on s390x 2018-02-05 22:04:12 yup, it definitively hangs on s390x 2018-02-05 22:04:15 ../example/ndpiReader -p ../example/protos.txt -q -i pcap/BGP_redist.pcap -w /tmp/reader.out -v 1 2018-02-05 22:04:26 the /tmp/reader.out is empty 2018-02-06 00:09:18 ncopa : aw nice. I just supposed you already made the s390x builder to continue despite the errors :D 2018-02-06 00:10:19 thanks 2018-02-06 00:28:07 that collectd error is unreal. lua-dev has no problem. maybe lua-dev is in queue for next "build cycle". let's wait. 2018-02-06 00:40:55 i wonder if that is an apk bug 2018-02-06 00:41:15 ncopa : nope. there are many packages built okay but not uploaded to repo. 2018-02-06 00:41:19 they are as old as 2017 2018-02-06 00:41:38 in main too? 2018-02-06 00:41:41 yeah 2018-02-06 00:41:51 lua-dev is one of them 2018-02-06 00:42:06 nl.a.o, dl-4.a.o 2018-02-06 00:44:07 maybe tomorrow ? kind of late for you over there :) 2018-02-06 00:44:46 hum 2018-02-06 00:44:46 ok 2018-02-06 00:44:47 i wonder how its supposed to work 2018-02-06 00:44:48 lua5.1 is built 2018-02-06 00:45:07 yes 2018-02-06 00:49:24 i think its a bug in apk-tools 2018-02-06 00:50:08 okay hum 2018-02-06 00:57:47 yup 2018-02-06 00:57:48 if i apk add lua5.1-dev it works 2018-02-06 00:57:50 but `apk add lua-dev` does not work 2018-02-06 00:58:44 it fails to resolve the provides=lua-dev 2018-02-06 01:00:02 but just on s390x ? let me recall if I've done any change on lua on s390x ... 2018-02-06 01:00:50 i cannot reprodue it on my x86_64 machine 2018-02-06 01:01:12 i dont think its related lua itself 2018-02-06 01:01:25 i think its due to there are no aport for "lua", it got renamed to lua5.1 2018-02-06 01:01:31 and has a provides=lua 2018-02-06 01:01:47 the lua5.1-dev has a provides=lua-dev 2018-02-06 01:01:54 apk is supposed to resolve the provides 2018-02-06 01:02:02 it does it on x86_64 apparently 2018-02-06 01:02:07 but not on s390x 2018-02-06 01:02:13 it may have to do with big endian 2018-02-06 01:02:33 alright i am investigating that 2018-02-06 01:02:37 but it could also be due to something else 2018-02-06 01:03:34 maybe recent changes in apk-tools 2018-02-06 01:03:40 i dont think its big-endian issue 2018-02-06 01:03:55 i cannot reproduce it on my s390x dev box 2018-02-06 01:04:02 my apk-tools 2.8.1 works out ok with lua-dev 2018-02-06 01:04:28 aha 2018-02-06 01:04:52 maybe recent changes in apk-tools create this 2018-02-06 01:06:16 what is your dev box version of apk-tools ? and the builder too ? 2018-02-06 01:06:24 2.8.1 seems to be on edge 2018-02-06 01:06:51 $ apk --version 2018-02-06 01:06:51 apk-tools 2.9.1, compiled for s390x. 2018-02-06 01:07:40 how ? the repo nl.a.o only has 2.8.1 2018-02-06 01:07:52 it built locally 2018-02-06 01:07:55 okay ... 2018-02-06 01:07:57 and apk upgrade was run 2018-02-06 01:08:16 let me build 2.9.1 then see 2018-02-06 01:08:38 i think there is a missing pre-requisite to reproduce it 2018-02-06 01:08:46 but i dont know what it is 2018-02-06 01:09:38 ok i think i know what it is 2018-02-06 01:09:59 :D 2018-02-06 01:11:14 this ? https://git.alpinelinux.org/cgit/apk-tools/commit/?id=667cb6bca799e58f58f22ad868761c022c6600dc 2018-02-06 01:12:02 hum 2018-02-06 01:12:04 no i was wrong 2018-02-06 01:12:48 that commits on src/commit.c looks suspicious to me 2018-02-06 01:15:44 yes 2018-02-06 01:15:44 but 2018-02-06 01:15:45 why can i not reproduce it on my dev s390x machine? 2018-02-06 01:16:29 nope. not apk-tools problem. it's lua5.1 problem. apk-tools 2.8.1 on mine fails to add lua-dev at local 2018-02-06 01:17:13 3 last commits on lua5.1 look suspicious 2018-02-06 01:17:24 *4 2018-02-06 01:17:37 yes i think they are related 2018-02-06 01:18:12 let me build older lua5.1 and try to add it at local 2018-02-06 01:18:36 thats it 2018-02-06 01:18:43 must be it 2018-02-06 01:18:47 yep 2018-02-06 01:19:36 hum 2018-02-06 01:19:45 hope 2018-02-06 01:19:49 nope 2018-02-06 01:19:59 i upgraded lua5.1 on my build machine 2018-02-06 01:21:05 well 2018-02-06 01:21:07 i think its the problem 2018-02-06 01:21:40 $ sudo apk add lua-dev | tpaste 2018-02-06 01:21:41 http://tpaste.us/MLk5 2018-02-06 01:22:09 note what version of the lua5.1* packages are installed 2018-02-06 01:22:13 they are not consistent 2018-02-06 01:22:26 hah 2018-02-06 01:22:35 some are installed via http because the older version is available 2018-02-06 01:22:43 yeah gotcha 2018-02-06 01:23:04 gotta install from local only, i suppose 2018-02-06 01:23:10 build-edge-s390x has only local repos 2018-02-06 01:23:19 yes 2018-02-06 01:23:31 i thikn this may be reproduceable on other arches too 2018-02-06 01:24:28 ok i need to sleep now 2018-02-06 01:24:31 yep 2018-02-06 01:24:42 i will restart the builder and let it fail and continue tomorrow 2018-02-06 01:24:59 thanks. i will be here tomorrow whenever you ping me 2018-02-06 01:25:40 im pretty sure i have a few workarounds to make it continue 2018-02-06 01:25:52 but i'd like to make sure its not a bug in apk-tools 2018-02-06 01:26:11 and maybe find a way to improve the error message 2018-02-06 01:26:12 good night 2018-02-06 01:26:15 understood. 2018-02-06 01:26:17 good night 2018-02-06 01:39:04 ncopa : So your discovery of different lua5.1* version is correct. on my build machine, apk 2.8.1, built latest lua5.1, $ apk search lua-dev shows lua5.1-dev-5.1.5-r6. So it's just s390x thing, mixing older and newer lua 2018-02-06 01:40:20 funny thing is, how can older versions of lua get in the way of the build-edge-s390x builder, since it only installs from local. 2018-02-06 01:40:33 which is the latest version. 2018-02-06 01:40:54 guess you can fix it at your side 2018-02-06 11:06:58 weird, every build, different checks fail 2018-02-06 13:17:32 ncopa : can you make a local build of lua on the builder again ? it's the mixture of different lua version that causes the problem as you figured out last night. 2018-02-07 07:22:57 algitbot: retry master 2018-02-07 07:42:08 algitbot: retry master 2018-02-07 07:42:15 oic 2018-02-07 07:42:31 ppc64le machine is behaving actually. 2018-02-07 15:20:46 algitbot: retry master 2018-02-08 15:31:06 ncopa : I think we just should disable the fftw on s390x for now. it affects 2 packages in main, 1 in community and a couple on testing. 2018-02-08 15:31:59 main/alsa-utils is the most important reverse-deps 2018-02-08 15:32:50 good thing is no package uses it as dep 2018-02-08 15:44:30 tmh1999: go on, I’d personally disable s390x builder, but… XD 2018-02-08 15:46:01 jirutka: How so? 2018-02-08 15:52:03 tmh1999: s390x is arch for banks and similar big corporates 2018-02-08 15:53:37 jirutka: oh, well afaik, I haven't received a single dollar from any bank or corporate :D so I guess it's out of my concern. 2018-02-08 15:53:54 tmh1999: exactly 2018-02-08 15:53:55 jirutka: apologize if it cause troubles for you 2018-02-08 15:54:33 tmh1999: no problem, it doesn’t 2018-02-08 15:55:08 tmh1999: I’m more concerned about armhf that is stuck for several days again :( 2018-02-08 15:55:52 jirutka: Yes I also saw that, hum... 2018-02-08 16:04:36 algitbot: retry master 2018-02-08 16:43:03 jirutka: would you mind checking this real quick ? : https://github.com/alpinelinux/aports/pull/3234. I appreciate it. 2018-02-08 16:59:12 algitbot: retry master 2018-02-09 13:23:24 ncopa, hi. it seems the ppc64le machine is back online. 2018-02-09 13:23:30 want to restart the daemon? 2018-02-09 19:42:21 apk-tools:master [passed] |Jakub Jirutka| travis: update alpine-chroot-install to 0.7.0 | https://travis-ci.org/alpinelinux/apk-tools/builds/339612484 2018-02-09 21:19:12 lots of testing fail. working on it... 2018-02-10 04:32:44 nothing new, still frequent faces. submitting... 2018-02-10 08:20:14 wish I hadn't disabled fftw on s390x. should have just disable make test like ncopa suggested. will continue tmr. 2018-02-10 08:20:19 mark here 2018-02-10 14:33:48 what the hell is is this? there are no arch-specific binaries in -tests, I can’t reproduce this failure on my machine 2018-02-10 14:35:22 let me take a look 2018-02-10 14:35:40 kaniini: you have access to the build server? 2018-02-10 14:36:42 i do not, but i can try to build on my own machine 2018-02-10 15:13:32 kaniini: did you find anything about julia? 2018-02-10 18:07:14 not yet 2018-02-11 14:16:54 argh 2018-02-12 21:32:37 ncopa, can you restart the ppc64 builder again? 2018-02-13 14:26:09 strace fails everywhere, aarch64 s390x x86 x86_64 2018-02-13 14:26:19 test failure 2018-02-13 14:26:27 yes i know 2018-02-13 14:26:28 git server died 2018-02-13 14:26:33 before i got it in 2018-02-13 14:27:18 ok this s390x is really annoying. kaniini, if you have a minute :D https://github.com/alpinelinux/aports/pulls/tmh1999 2018-02-13 14:27:27 we are not sure why strace is failing on alpine right now. it is not on adelie. probably due to busybox in the testsuite. 2018-02-13 14:27:58 thing is, testsuite fails differently on differnt arch, at least for s390x and aarch64 2018-02-13 14:28:15 lets see x86 2018-02-13 14:28:37 that is also weird 2018-02-13 14:29:00 rollback ? 2018-02-13 14:29:20 rollback to 4.18 ? 2018-02-13 14:41:49 memcached fails due to intel instruction thing. 2018-02-13 15:21:01 ncopa : while you are at doing s390x, would you mind ? https://github.com/alpinelinux/aports/pulls/tmh1999 2018-02-13 15:23:07 ok 2018-02-13 15:24:47 thanks ! 2018-02-13 15:36:26 ncopa : thanks ! 2018-02-14 03:34:32 aha 2018-02-14 03:34:36 but textrel why 2018-02-14 03:39:56 harumph 2018-02-14 03:40:04 i guess PaX is messing with libucontext still somehow 2018-02-14 03:43:48 kaniini you should implement on s390 is what says ;) 2018-02-14 03:45:09 i do not have access to a s390 machine right now 2018-02-14 03:47:49 kaniini: I can give you one any time. I have kvm here. 2018-02-14 03:48:10 one sec 2018-02-14 03:48:46 ... do i want to know how you have an s390 machine 2018-02-14 03:48:47 :D 2018-02-14 03:49:48 kaniini: I thought you did mentioned you requested a VM from IBM LinuxONE cloud ? 2018-02-14 03:51:02 oh 2018-02-14 03:51:07 that expired like a month ago 2018-02-14 03:51:12 :D 2018-02-14 08:06:22 kaniini, looks like you joined this room via matrix.org instead of alpinelinux.org server. 2018-02-14 08:06:56 yes dunno what the official one is 2018-02-14 08:09:22 isnt it in the room directory? 2018-02-14 08:10:16 #alpine-commits:alpinelinux.org 2018-02-14 08:24:30 . 2018-02-15 15:52:23 omfg 2018-02-15 15:54:58 ok good 2018-02-15 15:55:10 what a garbage buildsystem 2018-02-15 17:57:40 hm, I can't reproduce the pcre2 test failure on my laptop and the build also succeeded on the travis ci builder, any idea what might be causing this? 2018-02-15 17:57:54 do the x86_64 builders still use linux-hardenend? 2018-02-15 18:00:31 yes 2018-02-15 18:09:04 well…that might explain it, i am using -vanilla on my laptop 2018-02-15 18:10:55 I notified the person who submitted the PR, unless he known how to fix this I will probably just revert the upgrade for now 2018-02-15 21:20:16 what the hell is going on?! 2018-02-15 21:21:54 algitbot: retry master 2018-02-15 21:21:56 maybe… 2018-02-15 21:22:30 no, still fking broken with totally useless error message 2018-02-15 23:48:38 jirutka: why is ruby-rugged depending on libgit2-dev => 0.26? 2018-02-15 23:49:28 it’s makedepends and since ruby-rugged is ruby binding for libgit2, it needs specific version, >= 0.26.0, < 0.27.0 2018-02-15 23:49:57 but libgit2 is at 0.25.1 2018-02-15 23:50:06 uh 2018-02-15 23:50:08 I’m idiot 2018-02-15 23:50:12 I know what happend 2018-02-15 23:50:47 few days ago I was preparing upgrade of libgit2, forgot to finish it and also forgot built apks in my ~/packages 2018-02-15 23:51:00 that’s why it builds fine for me, but not on builders 2018-02-15 23:51:23 however, that apk’s error message is really bad, it doesn’t tell anything 2018-02-15 23:51:49 right, i had to look on the arm builder 2018-02-15 23:51:49 so, hm, I guess I should push libgit2 upgrade 2018-02-15 23:52:45 aha, and I see why I didn’t finish it… b/c of f*** Cargo (Rust) 2018-02-15 23:54:44 i guess abuild should be a bit more verbose in this case. not sure why it doesnt list the deps before it tries to install them. 2018-02-15 23:55:30 but a proper msg should probably come from apk 2018-02-16 01:17:51 algitbot: retry master 2018-02-16 01:18:42 clandmeter: okay, it’s fixed now; I’m very sorry for the mess 2018-02-16 01:33:42 jirutka, hm that is what I said but I said it on #alpine-devel. Sorry, I did not know you wouldn't see it there. 2018-02-16 01:34:18 AWilcox[m]: I have 1631 unread messages in #alpine-devel and I’m scared to open it :P :( 2018-02-16 01:35:07 AWilcox[m]: I’m very busy now so I’m minimizing my activity on IRC 2018-02-16 01:35:57 Ah, yeah. I understand. My text at 15:40: https://i.imgur.com/LtTwlaP.png 2018-02-16 01:36:02 I agree APK should give a better error. 2018-02-16 16:44:44 algitbot: retry master 2018-02-17 13:57:18 algitbot: retry master 2018-02-17 13:58:11 algitbot: retry master 2018-02-17 14:10:27 algitbot: retry master 2018-02-17 14:17:43 algitbot: retry master 2018-02-18 14:02:43 mkinitfs:master [failed] |Sören Tempel| initramfs-init: redirect output of brtfs scan to /dev/null | https://travis-ci.org/alpinelinux/mkinitfs/builds/343012184 2018-02-19 09:35:22 algitbot: retry master 2018-02-19 11:53:34 tmh1999: tests fails for main/liboil, main/audit and main/fftw on s390x, can you please help me report those upstream? 2018-02-19 13:45:18 ncopa : sure 2018-02-19 18:27:09 ncopa : I reported liboil upstream but it looks like they are very inactive. for audit, it builds good in kvm, so no need to report. it also sometimes build good sometimes don't, kind of unstable. so I guess we just !check for now :( 2018-02-19 18:27:35 ncopa : for fftw, the same behavior like audit, sometimes pass sometimes fails. so I also suggest !check atm. 2018-02-19 18:27:43 ncopa : how do you think 2018-02-19 18:28:12 *sometimes build good, sometimes fail in kvm 2018-02-19 18:28:18 *in container 2018-02-19 18:28:22 urg ... 2018-02-19 21:24:51 clandmeter2: just a friendly reminder in case you didn't notice so far: the pixz test suite seems to fail on all(?) builders 2018-02-19 21:25:47 hmm thats weird, they didnt fail locally. 2018-02-19 21:31:20 maybe it's missing a check dependency? 2018-02-19 21:31:31 > FAIL: xz-compatibility-c-option.sh 2018-02-19 21:31:36 does it need xz? 2018-02-19 21:34:30 yes it does 2018-02-19 21:34:34 im pusing a fix 2018-02-19 21:35:20 nmeum: thx! 2018-02-19 21:41:50 clandmeter2: your welcome 2018-02-19 21:43:40 the build issue on s390x should also be easy to fix, e.g. by just using the functions from endian.h instead 2018-02-20 10:23:12 mkinitfs:master [failed] |Carlo Landmeter| initramfs: do not relocate mountpoint for netboot | https://travis-ci.org/alpinelinux/mkinitfs/builds/343758886 2018-02-20 11:13:56 jirutka: is the travis ci runner for the mkinitfs repository not using edge? 2018-02-20 11:29:36 nmeum: no, it does not run edge, currently 3.6 2018-02-20 11:37:25 mkinitfs:master [passed] |Jakub Jirutka| travis: update alpine-chroot-install to 0.7.0 and use Alpine edge | https://travis-ci.org/alpinelinux/mkinitfs/builds/343784072 2018-02-20 11:43:33 jirutka: ah, thanks for upgrading it to edge 2018-02-20 11:44:59 nmeum: yw 2018-02-20 22:31:58 apk-tools:master [errored] |William Pitcock| tests: make sure adding a pinning to a child dependency fails | https://travis-ci.org/alpinelinux/apk-tools/builds/344060870 2018-02-20 22:35:54 apk-tools:master [errored] |William Pitcock| tests: make sure adding a pinning to a child dependency fails | https://travis-ci.org/alpinelinux/apk-tools/builds/344060870 2018-02-21 05:09:11 That came out of nowhere o.O 2018-02-21 15:44:57 apk-tools:master [passed] |William Pitcock| solver: allow names with only one provider to be autoselected regardless of priority | https://travis-ci.org/alpinelinux/apk-tools/builds/344373792 2018-02-28 00:51:06 harumph 2018-02-28 00:51:12 the grep testsuite hung all the builders 2018-02-28 19:09:31 algitbot: retry master 2018-02-28 20:54:26 hm, firefox-esr doesn't seem to work with gconf3 2018-02-28 21:12:48 ah 2018-02-28 21:14:58 the real problem is that abuild needs to generate $depends for pc files 2018-02-28 21:15:20 then we could clean up a lot of $depends_dev 2018-02-28 21:15:51 pkgconf --print-depends can be used to generate the lists from each .pc file 2018-02-28 21:15:51 :) 2018-02-28 23:42:44 kaniini: ^ 2018-02-28 23:45:54 hate 2018-02-28 23:46:06 algitbot: retry master 2018-02-28 23:46:07 needs new orbit2 2018-02-28 23:47:36 there we go