#archlinux32 | Logs for 2019-08-15

Back
[02:36:39] -!- MrBIOS has quit [Read error: Connection reset by peer]
[03:22:29] -!- isacdaavid has joined #archlinux32
[03:59:43] -!- bill-auger has quit [Read error: Connection reset by peer]
[04:00:17] -!- bill-auger has joined #archlinux32
[04:15:28] -!- thePiGrepper has quit [Ping timeout: 258 seconds]
[08:32:00] -!- isacdaavid has quit [Ping timeout: 252 seconds]
[08:43:09] -!- abaumann has joined #archlinux32
[08:43:10] <buildmaster> Hi abaumann!
[08:43:10] <buildmaster> !rq abaumann
[08:43:11] <phrik> buildmaster: <abaumann> bleep. I'm human.
[08:43:26] <abaumann> The following packages appear on the build- and deletion-list:
[08:43:27] <abaumann> liquid-dsp-sse4.1
[08:43:40] <abaumann> seems to be an easy one, I can fiddle in the database. :-)
[08:44:37] <abaumann> it exists upstream, so it shouldn't be on the deletion list
[09:07:25] * buildmaster resumes sanity.
[09:07:36] <abaumann> \o/
[09:12:31] <abaumann> liquid-dsp-sse4.1 is a subpackage of liquid-dsp, so maybe there is a problem in the sub-package logic?
[09:12:43] <abaumann> anyways: I want to build java 8 now. :-)
[09:24:24] <buildmaster> pentium4/java8-openjdk is broken (says rechenknecht): https://archlinux32.org
[09:32:50] <abaumann> yes, this is because my buildslave has a bootstrap with hand-compiled jdk8
[09:33:16] <abaumann> still, forced builds get exectuted on the forced slave _and_ on a random other build slave.
[09:33:28] <abaumann> I just hope, the second build result gets ignored
[09:38:21] <abaumann> so, java8 looks good
[09:47:25] <buildmaster> pentium4/libreoffice-fresh-i18n is broken (says eurobuild6-4): https://archlinux32.org
[09:47:40] <abaumann> really?
[09:48:06] <abaumann> mmh. maybe it takes a modern java environement, which is still broken
[09:54:00] -!- infides has joined #archlinux32
[10:17:11] <abaumann> elibrokeit: looking forward to new pacman, tons of fixes and features. :-)
[10:17:18] * buildmaster failed to execute a mysql query - can you have a look at "tmp.mysql-functions.query.2019-08-15T10:06:41.iQnTjg.stdin"?.
[10:17:32] <abaumann> oh no. deadlocks again.
[10:17:56] <abaumann> ERROR 1205 (HY000) at line 3: Lock wait timeout exceeded; try restarting transaction
[10:18:29] <abaumann> where is the purpose of transactions when all scripts have to be locked with a global lock?
[10:19:10] <abaumann> this was a seed-build-list colliding with something
[10:30:37] <abaumann> db-update colides with almost everything else in terms of deadlocks
[10:31:00] <abaumann> which is quite logical, as it changes the state of the db quite heavily
[10:32:16] <abaumann> deadlocks are dependency cycles of two transactions. so, if a write on a row depends on a read on another row and a second transation depends in reversed order on those two rows, you get a deadlock
[10:32:38] <abaumann> what can help is to establish deterministic orders how records are selected or updated, using sorts.
[10:51:50] <abaumann> mmh. why was jdk9 removed upstream?
[11:24:43] <abaumann> now a sanity-check and a db-update are blocking each other..
[11:26:57] * buildmaster resumes sanity.
[11:27:08] <abaumann> mmh. this is really curious..
[11:34:31] <abaumann> ..I'll check all the locking in place, maybe more agressive locking helps against deadlock issues..
[11:59:32] <buildmaster> i486/conky is broken (says eurobuild6-7-i486): https://archlinux32.org
[12:02:31] <buildmaster> i686/go-tools are broken (says eurobuild6-1): https://archlinux32.org
[12:03:40] <buildmaster> i486/harfbuzz is broken (says tyzoid-srv0-vm486): https://archlinux32.org
[12:07:18] <buildmaster> i486/liquid-dsp is broken (says eurobuild6-7-i486): https://archlinux32.org
[12:14:20] <buildmaster> pentium4/go-tools are broken (says nlopc46): https://archlinux32.org
[13:01:42] <buildmaster> i486/go is broken (says tyzoid-srv0-vm486): https://archlinux32.org
[13:19:54] <abaumann> Removing '-fno-plt' from CFLAGS and CXXFLAGS to prevent build fail with this architecture
[13:20:02] <abaumann> *grmpf* java 10, should have checked that.
[13:20:15] <abaumann> of course, otherwise we get PLT/GOT linking errors
[13:20:51] <buildmaster> i486/aarch64-linux-gnu-gcc is broken (says tyzoid-srv0-vm486): https://archlinux32.org
[13:23:05] <finsternis> I installed libwapcaplet manually from testing repo, now I get: warning: libwapcaplet: local (0.4.2-1.0) is newer than community (0.4.1-2.0) when I do pacman -Syu
[13:23:09] <abaumann> lessons learned: we should have kept more attention to all 32-bit related changes before the drop of 32-bit happened.
[13:23:25] <abaumann> and we should grep for i686 in the PKGBUILDs when creating a pentium4 architecture :-)
[13:24:34] <abaumann> finsternis: you installed it on a stable release, I suppose..
[13:25:01] <finsternis> yes, I guess that is not a problem
[13:25:51] <abaumann> yeah. the downgrade from build 2.0 to 1.0 is worriesome
[13:26:12] <abaumann> the problem is upstream:
[13:26:13] <abaumann> https://git.archlinux.org
[13:26:15] <phrik> Title: svntogit/community.git - Git clone of the 'community' repository (at git.archlinux.org)
[13:26:31] <abaumann> or not?
[13:26:33] <abaumann> no
[13:27:21] <finsternis> there was a software that required that latest version of libwapcaplet, so I picked that version from the testing repo
[13:27:30] <abaumann> on a stable repo.
[13:27:37] <abaumann> on a stable installation
[13:27:45] <abaumann> naturally pacman -Syyu downgrades again
[13:28:00] <abaumann> you have to put that package into IgnorePkg in /etc/pacman.conf
[13:28:08] <abaumann> for the time being..
[13:28:47] <abaumann> mmh. let me check the differences reports
[13:28:50] <abaumann> https://www.archlinux32.org
[13:28:56] <phrik> Title: Arch Linux 32 - Package Differences Reports (at www.archlinux32.org)
[13:29:07] <finsternis> libcss Depends On : libparserutils>=0.2.4 libwapcaplet>=0.4.1
[13:29:14] <abaumann> oh. penium4 is missing there :-)
[13:29:36] <abaumann> pacman -S libcss
[13:29:36] <abaumann> resolving dependencies...
[13:29:36] <abaumann> warning: cannot resolve "libwapcaplet>=0.4.2", a dependency of "libcss"
[13:29:37] <abaumann> :: The following package cannot be upgraded due to unresolvable dependencies: libcss
[13:29:42] <abaumann> yep.
[13:29:45] <abaumann> this sounds bad
[13:30:00] <abaumann> so libcss hit stable without libwacaplet
[13:30:07] <abaumann> let me nudge manually :-)
[13:32:47] <abaumann> also 0.4.1-2.0 is not marked red, though it's out of date
[13:39:08] <abaumann> ./db-update -w -f pentium4/community-testing/libwapcaplet -f i686/community-testing/libwacaplet
[13:39:11] <abaumann> is blocked..
[13:39:26] * abaumann is twiddling his thumbs
[13:45:43] <abaumann> so, libwacaplet is trapped in staging. This is unusual..
[14:04:52] <abaumann> db-update seems to take hours blocking everything else
[14:05:04] <abaumann> something feels wrong here..
[14:07:32] <abaumann> mmh, the arch32 patches for java10-openjdk don't seem to work at all..
[14:35:27] * buildmaster failed to execute a mysql query - can you have a look at "tmp.mysql-functions.query.2019-08-15T14:24:47.tdEdFs.stdin"?.
[14:38:32] <abaumann> of course..
[14:59:40] -!- abaumann has quit [Remote host closed the connection]
[15:29:49] * buildmaster resumes sanity.
[16:10:00] <buildmaster> i486/lollypop is broken (says eurobuild6-7-i486): https://archlinux32.org
[16:11:14] -!- abaumann has joined #archlinux32
[16:11:14] <buildmaster> Hi abaumann!
[16:11:14] <buildmaster> !rq abaumann
[16:11:15] <phrik> buildmaster: <abaumann> We should really not build a package more than 99 times. ;-)
[16:18:44] -!- thePiGrepper has joined #archlinux32
[16:18:56] <abaumann> finsternis: netsurf and libwaocaplet are in stable.
[16:20:53] <abaumann> though for me netsurf segfaults :-(
[16:21:55] <abaumann> I guess some gtk3 ABI mismatch
[16:41:21] -!- MrBIOS has joined #archlinux32
[16:42:42] <abaumann> well, netsurf work, till you press some flashy things on my web page, then it segfaults :-)
[16:44:24] -!- MrBIOS_ has joined #archlinux32
[16:45:52] -!- MrBIOS has quit [Ping timeout: 245 seconds]
[16:45:52] MrBIOS_ is now known as MrBIOS
[17:04:49] * buildmaster failed to execute a mysql query - can you have a look at "tmp.mysql-functions.query.2019-08-15T16:56:32.6urCde.stdin"?.
[17:24:43] -!- isacdaavid has joined #archlinux32
[17:40:54] <abaumann> basically the mysql database starts to tap on my nerves!
[17:41:02] <abaumann> almost nothing can be executed without hitting a deadlock
[18:13:03] -!- thePiGrepper has quit [Ping timeout: 258 seconds]
[18:13:59] * buildmaster resumes sanity.
[18:15:17] * abaumann is restarting ./get-package-updates -w
[18:49:40] <abaumann> oh fun. rebuilding java 11 installs java 12. :-)
[18:49:55] <abaumann> bootstrapping java back in time or what.. ;-)
[18:56:24] <abaumann> https://buildmaster.archlinux32.org
[18:56:25] <phrik> Title: Munin :: buildmaster :: buildmaster :: buildmaster combined (at buildmaster.archlinux32.org)
[18:56:35] <abaumann> shows quite an increase in runtime of db-update..
[19:36:05] -!- thePiGrepper has joined #archlinux32
[19:45:13] -!- abaumann has quit [Quit: leaving]
[20:43:02] -!- soalokin has joined #archlinux32
[20:57:42] <buildmaster> pentium4/java10-openjdk is broken (says eurobuild6-6): https://archlinux32.org
[21:07:16] -!- isacdaavid has quit [Ping timeout: 246 seconds]
[21:07:39] -!- eschwartz has quit [Ping timeout: 250 seconds]
[21:16:55] -!- eschwartz has joined #archlinux32
[21:59:12] -!- MrBIOS has quit [Quit: MrBIOS]
[22:22:55] -!- MrBIOS has joined #archlinux32
[22:33:16] -!- MrBIOS has quit [Quit: MrBIOS]
[22:33:36] -!- MrBIOS has joined #archlinux32
[22:37:57] -!- MrBIOS has quit [Ping timeout: 245 seconds]
[23:19:31] -!- infides has quit [Ping timeout: 248 seconds]