#archlinux-ports | Logs for 2018-01-08
Back
[00:14:09] <buildmaster> glew-wayland is broken (says buildknecht2).
[00:23:50] -!- graueRitter has joined #archlinux-ports
[00:23:52] <buildmaster> glew-wayland is broken (says buildknecht2).
[00:26:43] <guys> schroot had a false source url, fixed
[00:27:17] <deep42thought> thanks!
[00:27:20] <deep42thought> the others are mostly these qt5-issues
[00:37:03] -!- graueRitter has parted #archlinux-ports
[00:41:21] -!- p71 has quit [Ping timeout: 246 seconds]
[00:44:51] <guys> curious, what are the ones in (parentheses)
[00:45:12] <deep42thought> these are packages which have not all dependencies met, currently
[00:45:44] <deep42thought> their name should be a link to a graph of their dependencies
[00:46:03] <guys> Ah
[00:46:06] <guys> > /usr/bin/ldmd2: error while loading shared libraries: libLLVM-4.0.so: cannot open shared object file: No such file or directory
[00:46:11] <guys> What you do!
[00:46:31] <deep42thought> break things?
[00:46:37] -!- p71 has joined #archlinux-ports
[00:47:41] <guys> haskell-gauge is another "my tests are bork" failure
[00:52:26] <deep42thought> which package does ldmd2 belong to?
[00:59:37] -!- ztrawhcse has joined #archlinux-ports
[01:01:10] -!- guys has quit [Ping timeout: 260 seconds]
[01:05:55] -!- eschwartz[m] has quit [Remote host closed the connection]
[01:08:27] -!- cypa[m] has quit [Read error: Connection reset by peer]
[01:08:27] -!- FSMaxB has quit [Remote host closed the connection]
[01:08:27] -!- z3ntu has quit [Read error: Connection reset by peer]
[01:08:27] -!- brtln has quit [Remote host closed the connection]
[01:08:27] -!- titus_livius has quit [Ping timeout: 258 seconds]
[01:08:59] -!- titus_livius has joined #archlinux-ports
[01:14:40] -!- z3ntu has joined #archlinux-ports
[01:22:01] ztrawhcse is now known as guys
[01:29:21] -!- FSMaxB has joined #archlinux-ports
[01:29:21] -!- brtln has joined #archlinux-ports
[01:29:21] -!- cypa[m] has joined #archlinux-ports
[01:29:21] -!- eschwartz[m] has joined #archlinux-ports
[02:12:35] -!- tyler has joined #archlinux-ports
[02:12:52] -!- tyler has quit [Client Quit]
[02:31:23] -!- deep42thought has quit [Quit: Leaving.]
[02:38:20] <buildmaster> ldc is broken (says buildknecht2).
[03:02:39] -!- guys has quit [Ping timeout: 248 seconds]
[03:16:56] -!- guys has joined #archlinux-ports
[04:27:24] -!- sunshavi has joined #archlinux-ports
[05:00:44] -!- ztrawhcse has joined #archlinux-ports
[05:01:00] -!- guys has quit [Killed (orwell.freenode.net (Nickname regained by services))]
[05:01:00] ztrawhcse is now known as guys
[05:23:49] -!- hringriin_ has joined #archlinux-ports
[05:26:39] -!- hringriin has quit [Ping timeout: 248 seconds]
[05:26:40] hringriin_ is now known as hringriin
[05:59:07] -!- oaken-source has joined #archlinux-ports
[06:33:27] -!- guys has quit [Ping timeout: 240 seconds]
[06:38:22] -!- guys has joined #archlinux-ports
[06:42:24] -!- deep42thought has joined #archlinux-ports
[06:44:01] <deep42thought> ldc seems to have massive circular dependency issues
[06:44:54] <deep42thought> I compiled it yesterday by manually installing old llvm-libs, but this package can't compile itself now due to some wrong libphobos2-ldc linking ...
[06:55:32] -!- deep42thought has quit [Quit: Leaving.]
[07:12:13] -!- Vollzornbrot has joined #archlinux-ports
[07:12:18] <Vollzornbrot> Moin
[07:17:01] -!- deep42thought has joined #archlinux-ports
[07:17:13] <deep42thought> Hi Vollzornbrot
[07:24:09] -!- isacdaavid has quit [Remote host closed the connection]
[07:47:40] -!- deep42thought has quit [Quit: Leaving.]
[08:09:39] <buildmaster> fcitx-libpinyin is broken (says nlopc46).
[08:27:47] <guys> I did it again!
[08:27:59] <guys> more gl.h issues
[08:42:00] -!- guys has quit [Ping timeout: 260 seconds]
[08:50:48] -!- speak has joined #archlinux-ports
[08:51:07] <speak> Hi! Just wanted to say thanks for the folks making archlinux32 happen, saved my butt this weekend!
[08:51:27] <speak> The migration to the new repos was incredibly smooth, cheers!
[08:54:11] -!- deep42thought has joined #archlinux-ports
[08:55:51] -!- guys has joined #archlinux-ports
[08:56:22] <deep42thought> speak: you're welcome. btw: you mean to say, the transition was "unexpectedly smooth" ? ;-)
[08:57:13] <speak> Maybe >_>
[08:57:15] <speak> :D
[08:59:18] <speak> On top it was an ancient installation already (not updated for like a year), so even on a normal Arch update there's normally some issues
[09:06:45] <deep42thought> Thanks to Vollzornbrot for donating another build slave!
[09:07:39] <deep42thought> maybe we should set up some credit system and show some high scores on the buildmasters website ;-)
[09:09:35] <deep42thought> much like boinc does
[09:28:10] <buildmaster> haskell-hedgehog is broken (says buildknecht3).
[09:38:46] -!- Vollzornbrot has quit [Quit: Bye bye!]
[10:24:30] <buildmaster> shadowsocks-qt5 is broken (says buildknecht2).
[10:41:53] -!- guys has quit [Ping timeout: 248 seconds]
[10:55:46] -!- guys has joined #archlinux-ports
[11:02:46] <deep42thought> /usr/lib/libphobos2-ldc.so.73 is installed along with liblphobos-1:1.3.0-1, but it's not recognized by "pacman -Qo"
[11:02:50] <deep42thought> How can that be???
[11:06:47] <deep42thought> looks, like a pacman hook does that
[11:30:46] -!- deep42thought has quit [Remote host closed the connection]
[11:35:51] -!- deep42thought has joined #archlinux-ports
[11:53:22] oof is now known as bq
[12:04:11] <brtln> ldconfig creates missing symlinks
[12:04:24] <brtln> that should be created in PKGBUILD
[12:04:37] <brtln> (my first guess)
[12:16:47] <deep42thought> first part of your statement looks right, I can't comment on the second one
[12:22:05] -!- thorald has joined #archlinux-ports
[12:41:49] -!- ztrawhcse has joined #archlinux-ports
[12:42:21] -!- guys has quit [Ping timeout: 264 seconds]
[12:43:23] ztrawhcse is now known as guys
[12:59:34] -!- krumelmonster has joined #archlinux-ports
[13:17:14] -!- krumelmonster has quit [Quit: Leaving.]
[13:17:15] -!- krumelmonster1 has joined #archlinux-ports
[13:32:51] -!- thorald has quit [Read error: Connection reset by peer]
[13:44:54] <deep42thought> general question is, how we compile compilers+libraries, which have these circular dependencies (e.g. update of libraries breaks compiler, but working compiler is needed to compile itself)
[13:46:53] <guys> I wonder how *we* handle it, if it comes to that.
[13:47:18] <guys> I know ordinary circular dependencies just need to be bootstrapped.
[13:47:33] <deep42thought> I guess, the maintainers manually install / extract the old libraries next to the new ones
[13:47:55] <deep42thought> at least, that's what I tried with ldc (and which partially worked)
[13:50:14] <deep42thought> or one could use a statically linked bootstrap-compiler
[13:50:16] <deep42thought> :-/
[13:51:03] <deep42thought> abaumann_: made the proposal to have a statically linked set of the crucial binaries (pacman, systemd, gcc, ...)
[13:51:35] <deep42thought> so if we'd get _that_ working (which sounds like a big effort), then everything should be fine :-)
[13:59:33] -!- krumelmonster1 has quit [Read error: No route to host]
[13:59:34] -!- krumelmonster has joined #archlinux-ports
[14:17:11] -!- krumelmonster has quit [Ping timeout: 276 seconds]
[14:20:21] <brtln> yes, we modify chroots
[14:21:17] <brtln> we haven't had a chance to bootstrap anything recently though, but we would probably package binary first and then use it to build
[14:27:29] -!- krumelmonster has joined #archlinux-ports
[14:44:22] -!- guys has quit [Ping timeout: 268 seconds]
[14:44:53] -!- guys has joined #archlinux-ports
[15:46:55] -!- Polichronucci has quit [Ping timeout: 248 seconds]
[15:48:31] -!- Polichronucci has joined #archlinux-ports
[15:49:00] -!- hringriin has quit [Quit: ^Cc12^CbSee ^Cb^Cc04^Cya^Cc!]
[15:49:51] -!- hringriin has joined #archlinux-ports
[15:52:34] <buildmaster> recoll is broken (says buildknecht).
[15:57:49] -!- hringriin_ has joined #archlinux-ports
[16:01:53] -!- hringriin_ has quit [Client Quit]
[16:03:56] <buildmaster> kcm-fcitx is broken (says buildknecht2).
[16:09:05] -!- krumelmonster has quit [Ping timeout: 240 seconds]
[16:40:22] -!- deep42thought has quit [Quit: Leaving.]
[16:42:38] -!- ztrawhcse has joined #archlinux-ports
[16:42:56] guys is now known as Guest6715
[16:42:56] -!- Guest6715 has quit [Killed (kornbluth.freenode.net (Nickname regained by services))]
[16:42:56] ztrawhcse is now known as guys
[16:50:44] -!- krumelmonster has joined #archlinux-ports
[17:02:56] -!- oaken-source has quit [Ping timeout: 276 seconds]
[17:33:50] -!- ztrawhcse has joined #archlinux-ports
[17:34:08] -!- guys has quit [Killed (barjavel.freenode.net (Nickname regained by services))]
[17:34:08] ztrawhcse is now known as guys
[18:00:35] -!- guys has quit [Ping timeout: 240 seconds]
[18:06:01] -!- ztrawhcse has joined #archlinux-ports
[18:06:23] ztrawhcse is now known as guys
[18:19:56] -!- abaumann has joined #archlinux-ports
[18:24:58] -!- oaken-source has joined #archlinux-ports
[18:33:03] <buildmaster> fcitx-qt5 is broken (says buildknecht3).
[18:44:31] -!- oaken-source has quit [Ping timeout: 248 seconds]
[18:47:15] -!- krumelmonster has quit [Remote host closed the connection]
[19:30:46] -!- oaken-source has joined #archlinux-ports
[19:31:20] -!- deep42thought has joined #archlinux-ports
[19:33:09] <deep42thought> abaumann: Have you got some time to fix (part of) our mess?
[19:33:14] <deep42thought> :-D
[19:55:13] <abaumann> deep42thought: I see no mess :-)
[19:55:19] <deep42thought> lol
[19:55:38] <abaumann> the mesa-beast is back :-(
[19:55:45] <deep42thought> well, I see some qt5-errors looking for GL/gl.h
[19:55:48] <deep42thought> yup
[19:56:01] <deep42thought> and ldc fails due to its circular dependency
[19:56:21] <abaumann> there seems to be something around the 'D' language.. *sigh*
[19:56:33] <deep42thought> ldc <--> llvm/liblphobos (if I'm not mistaken)
[19:57:09] <abaumann> ah. yeah. the libLLVM. so major release hopped, and lcd cannot compile itself again with itself - or something along those lines
[19:57:33] <abaumann> valgrind fails it's tests.
[19:57:33] <deep42thought> yes
[19:57:54] <abaumann> I still have issue with ELF text relro flags breaking keyboard layout loading.
[19:58:11] <deep42thought> I compiled ldc with the new libraries by manually extracting the old package next to the new one, but apparently, the ldc produced is broken :-/
[19:58:20] <deep42thought> hmmm
[19:58:22] <abaumann> I want at least see if it happens also on 64-bit, because if yes, there is a major issue lurking.
[19:59:10] <deep42thought> the keyboard stuff?
[19:59:20] <abaumann> compilers and interpreters/instrumenters breaking is usually an indicator to a toochain issue.
[19:59:23] <abaumann> yes.
[19:59:36] <abaumann> https://bbs.archlinux32.org
[19:59:38] <phrik> Title: [Solved] Trouble setting the propper keymap layout / Installation / Arch Linux 32 Forums (at bbs.archlinux32.org)
[19:59:55] <abaumann> Solved == a workaround exists for this specific problem for users.
[20:00:06] <deep42thought> yeah, I read that
[20:00:16] <abaumann> But not being able to call gzip in loadkeys when systemd-vconsole loads a keymap during boot is not a good sign.
[20:01:25] <abaumann> and then there is the haskell-hslua showstopper (literaly)
[20:01:51] <deep42thought> :-D
[20:02:05] -!- guys has quit [Ping timeout: 240 seconds]
[20:02:25] -!- oaken-source has quit [Ping timeout: 248 seconds]
[20:04:43] -!- guys has joined #archlinux-ports
[20:15:26] <deep42thought> I thought, we had solved that mesa problem
[20:15:36] <deep42thought> do you remember what we did / changed?
[20:16:06] <abaumann> revert to an old libgndvd library in the build slave?
[20:17:08] <abaumann> libglvnd
[20:17:12] <abaumann> I meant.
[20:22:02] <deep42thought> ah, problem might be, that there are multiple providers for libgl
[20:22:29] <deep42thought> we put an explicite dependency on libglvnd on extra/nvidia
[20:22:31] <abaumann> yeah. I think so. But I though liblvnd is there to be able to do a runtime switch.
[20:23:12] <abaumann> isn't libglvnd supposed to be vendor netural?
[20:23:46] <deep42thought> to me it looks the other way round: libgl id provided by libglvnd
[20:23:52] <deep42thought> and probably some other package, too
[20:23:56] <abaumann> ah
[20:24:37] <deep42thought> ah, no, I'm wrong
[20:32:59] <deep42thought> great, kphotoalbum is now also broken with that qt5-issue (additionally to the now-fixed race condition)
[20:35:02] <deep42thought> adding libglvnd as makedepends seems to fix it
[20:35:17] <deep42thought> :-/
[20:35:21] <abaumann> yeah. but this is the wrong place.
[20:35:27] <deep42thought> I thought so
[20:35:29] <abaumann> qt5-base is the one depending on mesa
[20:35:52] <abaumann> I was even considering adding mesa directly there to fix everything upstream using Qt
[20:36:02] <abaumann> but this feels also wrong. :-)
[20:36:23] <abaumann> Exception: <stdout>: hPutChar: invalid argument (invalid character)
[20:36:31] <abaumann> argh. I start to hate Haskell.
[20:36:52] <abaumann> and tests running only in well-prepared environments
[20:37:29] <deep42thought> qt5-base depends on libgl
[20:37:41] <deep42thought> which is provided by libglvnd and nvidia-340xx-utils
[20:37:42] <abaumann> but the current provider of libgl is what?
[20:37:46] <abaumann> ah.
[20:37:54] <abaumann> both have mesa header files?
[20:37:59] <abaumann> Gl/gh.h
[20:38:01] <abaumann> gl.h
[20:38:14] <deep42thought> hah!
[20:38:19] <deep42thought> nvidia* is in testing
[20:38:23] <abaumann> because I have seen mesa hopping in and out upstream.
[20:38:25] <deep42thought> so it's prefered over libglvnd
[20:38:28] <abaumann> aha.
[20:38:33] <deep42thought> (despite alphabetical!! order)
[20:38:57] <deep42thought> buildmaster: why dont you stabilize nvidia-340xx-utils
[20:38:59] <buildmaster> "nvidia-340xx-utils" is not tested yet!
[20:39:55] <deep42thought> buildmaster: why dont you stabilize nvidia-340xx-utils
[20:39:57] <buildmaster> The following packages are dependent on "nvidia-340xx-utils", but cannot be stabilized:
[20:39:57] <buildmaster> 0ad.d08fcd46dcd3fe53d2321c07b771cc9e8c0c92cf.af5c73427d55fff444e94570c37ac4f43fade962.community (not tested yet for 1 days)
[20:39:57] <buildmaster> acpi_call.661104d5b79720d0357ca2a17da176551882553d.af5c73427d55fff444e94570c37ac4f43fade962.community (not tested yet for 1 days)
[20:39:57] <buildmaster> aegisub.d08fcd46dcd3fe53d2321c07b771cc9e8c0c92cf.af5c73427d55fff444e94570c37ac4f43fade962.community (not tested yet for 1 days)
[20:39:57] <buildmaster> amule.d08fcd46dcd3fe53d2321c07b771cc9e8c0c92cf.af5c73427d55fff444e94570c37ac4f43fade962.community (not tested yet for 1 days)
[20:39:57] <buildmaster> ... (146 lines total)
[20:40:00] <deep42thought> uff
[20:40:16] <deep42thought> we should put libglvnd into build-support, I think
[20:40:21] <deep42thought> this seems like the cleaner solution
[20:40:35] <abaumann> yepp
[20:40:48] <deep42thought> done
[20:41:03] <abaumann> this solves half of the problems in the list :-)
[20:41:20] <deep42thought> hopefully
[20:41:34] <deep42thought> we should keep libglvnd around in build-support
[20:41:48] <deep42thought> just in case ;-)
[20:43:01] <abaumann> aha: setenv LANG en_US.UTF-8
[20:43:07] <deep42thought> huh?
[20:43:15] <deep42thought> who does that?
[20:43:17] <abaumann> tests breaking if you don't set it.
[20:43:21] <abaumann> in haskell-guage
[20:43:22] <deep42thought> oh
[20:43:38] <deep42thought> "portable" also means "across the atlantic"?
[20:43:48] <abaumann> I don't get the point why it's not set on my 64-bit build machine..
[20:44:06] <abaumann> "portable" as in "weighs less than 30 kilograms" ;-)
[20:44:15] <deep42thought> "has handles"
[20:44:41] <deep42thought> a ship might also be portable - at least if it fits into a port
[20:46:21] <abaumann> all unit tests cramped into one Sanity.hs, nice :-)
[20:49:03] <deep42thought> I'm optimistic and remove the broken-tag from all those GL/gl.h packages - so don't be surprised when we're flooded with broken-packages messages in a few minutes :-)
[20:50:58] <abaumann> ah
[20:53:11] <abaumann> the haskell-gauge package fails because the gauge fib22 test takes to long. :-)
[20:53:29] <abaumann> doing absolute benchmarking in unit tests.
[20:54:19] <deep42thought> but this would avoid these endless loops :-)
[20:54:50] <abaumann> I'm not against benchmarking. Just do it relativ to previous runs and measure relative changes per platform.
[20:59:06] <deep42thought> so a workaround will/can/should be to increase the allowed timeß
[20:59:07] <deep42thought> ?
[20:59:19] <abaumann> how is your Haskell lately? ;-)
[20:59:39] <abaumann> let tooLong = 30
[20:59:43] <abaumann> is a good candidate
[21:01:15] <deep42thought> um, haven't seen a single haskell program so far
[21:01:35] <abaumann> same here :-)
[21:17:23] -!- charwhee has joined #archlinux-ports
[21:41:01] -!- charwhee has quit [Quit: Page closed]
[21:52:13] <abaumann> ah. ok. I just disabled the one failing test and reported upstream. so haskell-gauge passes..
[21:52:30] <deep42thought> yippie!
[21:52:50] <deep42thought> btw: libgl-dependent packages seem fine, too
[21:52:50] <abaumann> aeh.. it's not such an important package..?
[21:52:55] <abaumann> ah. :-)
[21:53:10] <abaumann> yeah. down to 17 failures.
[21:53:15] <deep42thought> well ,every broken package bothers me
[21:53:25] <abaumann> true.
[21:53:30] <deep42thought> because it potentially blocks unstaging other packages
[21:54:49] -!- shentey has joined #archlinux-ports
[22:01:51] -!- guys has quit [Quit: A random quit message]
[22:01:55] -!- ztrawhcse has joined #archlinux-ports
[22:02:35] ztrawhcse is now known as guys
[22:03:09] <deep42thought> should we simply disable the offending valgrind checks, too?
[22:03:53] <abaumann> where they tested upstream before?
[22:04:04] <deep42thought> no
[22:04:09] <abaumann> ah. :-)
[22:04:38] <deep42thought> not even for x86_64
[22:04:46] <abaumann> yeah. maybe an idea to disable them for now.
[22:06:27] <abaumann> I am pretty sure some tests of valgrind fail becuase of the way software gets compiled on the system at hand.
[22:06:44] <abaumann> so, if you enable certain flags in compilers and linkers valrgind will start to fail.
[22:07:13] <abaumann> we can build it without tests and then I can do some manual test on some of my software to see if it works in principle
[22:07:24] <deep42thought> ok
[22:16:35] -!- abaumann has quit [Quit: leaving]
[22:32:28] -!- shentey has quit [Quit: Konversation terminated!]
[22:34:21] -!- guys has quit [Ping timeout: 256 seconds]
[22:46:00] <deep42thought> tyzoid: I get a 502 on bugs.archlinux32.org - are you up to something?
[22:46:53] -!- guys has joined #archlinux-ports
[23:08:11] -!- deep42thought has quit [Quit: Leaving.]
[23:25:01] -!- fhdrin has joined #archlinux-ports
[23:32:29] -!- isacdaavid has joined #archlinux-ports
[23:58:27] -!- fhdrin has quit [Ping timeout: 240 seconds]