#archlinux-ports | Logs for 2017-09-11
Back
[00:41:00] -!- ztrawhcse has quit [Remote host closed the connection]
[00:42:00] -!- ztrawhcse has joined #archlinux-ports
[02:56:13] -!- Faalagorn has quit [Remote host closed the connection]
[03:03:27] -!- rewbycraft has quit [Ping timeout: 252 seconds]
[03:07:45] -!- rewbycraft has joined #archlinux-ports
[03:49:24] -!- ztrawhcse has quit [Remote host closed the connection]
[03:54:42] -!- ztrawhcse has joined #archlinux-ports
[04:42:06] -!- rmbeer has joined #archlinux-ports
[04:42:18] <rmbeer> hello
[04:42:33] <rmbeer> not understand, what happened with archlinux in 64 and 32 bits?
[04:57:53] -!- rmbeer has parted #archlinux-ports
[05:32:01] <ztrawhcse> wat
[05:33:37] -!- isacdaavid has quit [Quit: isacdaavid]
[05:42:26] -!- isacdaavid has joined #archlinux-ports
[05:58:28] phillid is now known as grillid
[06:01:59] -!- isacdaavid has quit [Quit: isacdaavid]
[06:03:28] -!- isacdaavid has joined #archlinux-ports
[06:06:32] -!- isacdaavid has quit [Client Quit]
[06:06:56] -!- isacdaavid has joined #archlinux-ports
[06:24:16] -!- deep42thought has joined #archlinux-ports
[07:16:46] -!- isacdaavid has quit [Quit: isacdaavid]
[07:23:25] -!- deep42thought has quit [Quit: Leaving.]
[08:45:32] -!- deep42thought has joined #archlinux-ports
[09:17:39] <deep42thought> rewbycraft: I locked myself out from the buildmaster's root, because of some glibc-relatet non-bug: Can you exec "systemctl daemon-reexec" on the buildmaster?
[10:08:03] -!- alyptik has quit [Ping timeout: 252 seconds]
[10:08:36] -!- alyptik has joined #archlinux-ports
[12:18:07] -!- guys has quit [Quit: A random quit message]
[12:18:08] -!- ztrawhcse has quit [Read error: Connection reset by peer]
[13:50:51] -!- p71 has quit [Read error: Connection reset by peer]
[16:03:57] -!- WarheadsSE has quit [Quit: WeeChat 1.6]
[16:04:30] -!- WarheadsSE has joined #archlinux-ports
[16:12:11] -!- p71 has joined #archlinux-ports
[16:14:36] -!- p71 has quit [Excess Flood]
[16:21:13] -!- p71 has joined #archlinux-ports
[16:37:34] -!- deep42thought has quit [Quit: Leaving.]
[19:26:20] -!- isacdaavid has joined #archlinux-ports
[20:05:15] -!- deep42thought has joined #archlinux-ports
[20:27:49] -!- fsckd has quit [Remote host closed the connection]
[20:28:54] -!- fsckd has joined #archlinux-ports
[20:38:49] -!- MST_ has joined #archlinux-ports
[20:39:28] <MST_> https://bugs.archlinux.org
[20:39:29] <phrik> Title: FS#55591 : [clamtk] build fails (wrong source url) (at bugs.archlinux.org)
[20:40:02] <jelle> hmmm
[20:40:52] <deep42thought> funny - shouldn't there be an announcement, when the source changes to some other repository?
[20:47:32] <jelle> and fswebcam cann't be downloaded ;-)
[20:49:07] <deep42thought> what's (with) that?
[20:49:30] <jelle> since you are reporting 404's ;-)
[20:49:52] <deep42thought> usually, I wait a day (or two) before reporting them
[20:50:04] <deep42thought> but: yes :-)
[20:50:06] <jelle> lol
[20:50:07] <jelle> working on it
[20:50:13] <jelle> either getting nuked or fixed
[20:53:19] <deep42thought> Hmm, I wonder if I should put 'export CFLAGS+=" -m32" and 'export CXXFLAGS+=" -m32"' into all PKGBUILD which have options=(!buildflags) ...
[20:56:17] <brtln> what for?
[20:56:28] <brtln> your build env is already 32 bit
[20:56:55] <deep42thought> we had to set these for arrayfire in order to compile
[20:57:02] <jelle> o.o
[20:57:34] <brtln> was there a point in building it all though? looks to be cuda thing
[20:57:48] <brtln> I can imagine that at least once in a while svenstaro does something like that on purpose
[20:58:13] <jelle> brtln: we packge it only for x86_^4
[20:58:19] * jelle shrugs
[20:58:22] <brtln> that's what I'm talking about
[20:58:29] <brtln> "it at all" *, sorry
[20:59:01] <deep42thought> well, it has been built for i686 a few versions before
[20:59:26] <deep42thought> honestly, I have pretty much no idea what 80% of the packages do or how their internals work :-/
[20:59:48] <deep42thought> so you're telling me, we should simply drop it?
[20:59:58] <brtln> same fore me, but rule of thumb is that huge scientific packages aren't designed for 32 bit
[21:00:05] <brtln> yes, I guess that makes more sense
[21:00:20] <brtln> looking at their github, they don't do test builds for any 32 bit arch
[21:00:33] <brtln> either for x86 or arm
[21:00:38] <deep42thought> hmmm
[21:00:50] <deep42thought> but with '-m32', the build succeeded :-/
[21:00:55] <deep42thought> Isn't that all that counts?
[21:00:57] <deep42thought> :-D
[21:01:00] <brtln> :D
[21:01:10] <brtln> sounds hacky though
[21:01:29] <jelle> or upstream is weird
[21:02:07] <deep42thought> well, usually we have '-march=i686' which gets stripped due to '!buildflags'
[21:06:28] <brtln> yeah, but you can imagine the same does happen for us
[21:06:32] <deep42thought> but I'm not really sure, why we'd need it anyway (as long as we compile for i686 and not i486 or i586)
[21:07:23] <deep42thought> I'd expect the build system to get the correct architecture from 'uname -m'
[21:07:34] <deep42thought> but apparently, it doesn't
[21:07:36] <brtln> that should be happening, yes
[21:08:25] <deep42thought> though, I see one problem with 'uname -m': It does not see a difference between i486, i586 and i686 if those are only accomplished by 'setarch ...'
[21:09:14] <deep42thought> however, this should not be an issue for x86_64, because you don't need to "fake" the arch
[21:09:31] <brtln> I think setarch gets called anyway
[21:10:19] <deep42thought> but 'uname -m' still returns the correct arch for you
[21:10:43] <deep42thought> anyway, this discussion is academic, because we're not (yet) dealing with i4/586
[21:21:45] <brtln> sure, because there's nothing higher than x86_64 from this perspective
[21:21:56] <brtln> we would need to require cpu extensions and that's a minefield
[21:22:16] <brtln> we've been discussing doing that for "i686" but that was just silly
[21:22:56] <deep42thought> so you would end up with 2^n "architectures" for n possible cpu extensions?
[21:23:45] <deep42thought> sound like the right point to switch to a compile-all-from-source-distribution ...
[21:25:25] <brtln> no, the idea was to, let's say, require sse2 for 32 bit
[21:25:34] <deep42thought> ah, ok
[21:25:35] <brtln> (theoretically speaking)
[21:25:46] <brtln> and calling that i786
[21:26:04] <brtln> but I failed to see the border of this and x86_64
[21:26:08] <deep42thought> yeah, but there'd be many possible "i786"s, then
[21:26:09] <brtln> so here we are :)
[21:26:14] <deep42thought> :-)
[21:33:14] <WarheadsSE> brtln: don't worry .. sometimes people just assume SSE2 even on ARM ....
[21:33:52] * jelle frowns
[21:36:54] <brtln> heh
[21:48:43] -!- isacdaavid has quit [Quit: isacdaavid]
[22:02:28] -!- MST_ has quit [Remote host closed the connection]
[22:08:26] -!- isacdaavid has joined #archlinux-ports
[22:16:14] -!- isacdaavid_ has joined #archlinux-ports
[22:22:49] -!- Polichronucci_ has joined #archlinux-ports
[22:24:07] -!- isacdaavid has quit [*.net *.split]
[22:24:08] -!- Polichronucci has quit [*.net *.split]
[22:24:09] Polichronucci_ is now known as Polichronucci
[22:26:59] -!- wyre has quit [Ping timeout: 240 seconds]
[22:35:50] -!- wyre has joined #archlinux-ports
[23:26:08] -!- deep42thought has quit [Quit: Leaving.]
[23:27:10] <rewbycraft> deep42thought: sorry for the delay. Re-exec'd systemd
[23:28:27] alyptik is now known as y-
[23:30:11] y- is now known as alyptik