josh-k has quit [Ping timeout: 244 seconds]
josh-k has joined #rubinius
josh-k_ has quit [Ping timeout: 245 seconds]
josh-k_ has joined #rubinius
josh-k has quit [Ping timeout: 252 seconds]
amsi has quit [Quit: Leaving]
josh-k_ has quit [Remote host closed the connection]
kfpratt has joined #rubinius
kfpratt has quit [Remote host closed the connection]
johnmuhl has joined #rubinius
Bwild has quit [Ping timeout: 250 seconds]
meh` has quit [Ping timeout: 256 seconds]
|jemc| has quit [Ping timeout: 265 seconds]
kfpratt has joined #rubinius
kfpratt has quit [Ping timeout: 258 seconds]
josh-k has joined #rubinius
josh-k_ has joined #rubinius
josh-k has quit [Ping timeout: 245 seconds]
josh-k has joined #rubinius
josh-k_ has quit [Ping timeout: 250 seconds]
max96at|off is now known as max96at
havenwood has quit [Remote host closed the connection]
josh-k has quit [Remote host closed the connection]
<yopp> morning and merry christmas!
<yorickpeterse> ho ho ho
goyox86 has joined #rubinius
<goyox86> Merry Christmas!
noop has joined #rubinius
goyox86 has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
goyox86 has joined #rubinius
cremes has joined #rubinius
cremes has quit [Quit: cremes]
cremes has joined #rubinius
cremes has quit [Client Quit]
noop has quit [Read error: Connection reset by peer]
noop has joined #rubinius
kfpratt has joined #rubinius
max96at is now known as max96at|off
johnmuhl has quit [Quit: Connection closed for inactivity]
kfpratt has quit [Remote host closed the connection]
cypher23 has quit [Ping timeout: 250 seconds]
havenwood has joined #rubinius
cypher23 has joined #rubinius
Bwild has joined #rubinius
meh` has joined #rubinius
|jemc| has joined #rubinius
havenwood has quit [Remote host closed the connection]
havenwood has joined #rubinius
goyox86 has quit [Read error: Connection reset by peer]
cremes has joined #rubinius
noop has quit [Ping timeout: 240 seconds]
havenwood has quit []
havenwood has joined #rubinius
<jc00ke> Are there plans to GC symbols in rbx too?
<yxhuvud> they won't have a choice if they want to be able to run rails5.
<jc00ke> yxhuvud: I'm not entirely sure that's true, if we handle symbols efficiently (which I'm guess we do)
<yopp> Just to be sure: in 2.2 symbols will change their object_id, and that's and issue with that?
<yopp> And then, what the problem there?
<yopp> I'm not quite getting this. :(
<jc00ke> I was addressing your question about object_id changing for a symbol. It doesn't change after a GC run in CRuby 2.2
<jc00ke> No problem, per se, just wondering what this looks like for other implementations. Maybe I missed something and Symbol GC is an OK thing?
diegoviola has joined #rubinius
<yopp> Yeah, and I'm still not quite understand how this feature might break rails, if not implemented in rbx?
<jc00ke> I don't know either. My guess is that Rails is going to have a better memory profile on 2.2+ because they use symbols to avoid DoS attacks
<yopp> From dhh's reply to your question seems like that, yeah: dhh: @jc00ke Because we have to convert anything that could have been a symbol from outside data to a string lest we open DoS hole.
<yopp> Seems like not the memory profile is the issue, but the conversion from string to symb
darix has joined #rubinius
<darix> ok
<darix> which which llvm version does llvm/Config/config.h exist?
<darix> llvm 3.5.0 doesnt seems to have it
<yopp> Yeah, I heard something about that later, someone raised a concern about DDoSing rails app with symbols. And rails-core patched this a bit later
<jc00ke> darix: what OS are you on, and how did you install LLVM?
<darix> jc00ke: opensuse. rpm.
<darix> jc00ke: i just try to upgrade my rubinius package to 2.4
<jc00ke> yopp: yeah, I guess that's the cruft he was talking about. My guess is that the messaging around this requirement is incorrect. An implementation should be able to handle things like GC however it likes.
<jc00ke> I think this is a soft req beyond kwargs and a version number
<jc00ke> Anyway, nbd
<yopp> hum
<jc00ke> darix: hmm, I haven't seen much regarding rbx on OpenSuse via RPM. Are you able to compile it manually from a clone of the repo?
<yopp> ObjectSpace.each_object(Symbol).count
<yopp> => 0
<yopp> oh my, it was rubinius
<darix> jc00ke: i am one of the main ruby packagers for opensuse. so if it fails in the package build env it will also fail when I build from the tarball
<darix> because i would still be using the llvm packages
<jc00ke> darix: OK, I'm not familiar with how RPMs actually work.
<yopp> huh
<yopp> now symbols are frozen
<yopp> :s.instance_variable_set(:@moo, 1)
<darix> jc00ke: from the build time ... they arent much different
<yopp> since 2.1
<yopp> so it's not a big deal anyway
<darix> brixen yorickpeterse: what do you need llvm/Config/config.h for?
<yopp> you can't modify Symbol instance, you can re-create it anytime because it's only carrying the object_id
<darix> --llvm-shared and it looks for static libraries
diegoviola has quit [Quit: WeeChat 1.0.1]
<darix> ah configure check failed
<darix> configure --help should *not* overwrite configure.log :p
<jc00ke> darix: whoa, you're right!
<darix> ok
<darix> the configure check for llvm-shared fails with:
<darix> /usr/include/c++/4.8/bits/c++0x_warning.h:32:2: error: #error This file requires compiler and library support for the ISO C++ 2011 standard. This support is currently experimental, and must be enabled with the -std=c++11 or -std=gnu++11 compiler options.
<yopp> jc00ke, actually they change object_id
<darix> fixed by setting -std=c++11 in the CXXFLAGS
diegoviola has joined #rubinius
<yopp> a) there is a mistake in code: lvar s references :sym1 when GC.start is called
<diegoviola> sad
<darix> got it to compile but it fails later in the build with some segv in rubinius
<diegoviola> I mean, really?
<|jemc|> zenspider muckraking and sowing discord in the community? I'm shocked.
diegoviola has quit [Quit: WeeChat 1.0.1]
<yorickpeterse> darix: argh yeah the c++11 crap
<yorickpeterse> darix: LLVM 3.5 decided to use some c++11 all of a sudden
<yorickpeterse> I think we've enabled that since though
<yorickpeterse> darix: Yeah, -std=c++11 is added for LLVM 3.0.5 or newer
<yorickpeterse> errr
<yorickpeterse> 3.5.0
diegoviola has joined #rubinius
RealMarc has quit [Ping timeout: 250 seconds]
<yorickpeterse> hmm...we've seen more segfaults like that recently
diegoviola has quit [Client Quit]
<yorickpeterse> interesting that it crashes in the console
<yorickpeterse> darix: could you report this as a bug?
<yorickpeterse> (with the output in a gist)
<darix> just the crash related or whole build log?
<yorickpeterse> I'd gist the whole thing
<yorickpeterse> gives some context to it
<yorickpeterse> I'm pretty sure somebody else reported similar problems but I can't find the issue atm
<yorickpeterse> It's not threading related though, at least I'm not certain of that
<yorickpeterse> it seems to be a problem in the Rbx console during shutdown/fork()
<darix> yorickpeterse: the last 2 steps are in pthread :)
<darix> smells a bit like threading related tbh
Bwild has quit [Ping timeout: 250 seconds]
<yorickpeterse> that's because the console runs in a separate thread
<yorickpeterse> basically the pthread stuff almost always shows up in stacktraces
<yorickpeterse> since threads are pretty much always involved
<yorickpeterse> I'll fix the title a bit
<darix> yorickpeterse: checked in the source also so you can see the patches
<darix> but I doubt they affect that
Bwild has joined #rubinius
<darix> yorickpeterse: JFYI the crash also happens with older distros (13.1 e.g.)
<darix> newer distro is still building
drdanmaku has joined #rubinius
<yorickpeterse> hmm
diegoviola has joined #rubinius
drdanmaku has left #rubinius [#rubinius]
centrx has joined #rubinius
zacts has joined #rubinius
diegoviola has quit [Remote host closed the connection]
elia has joined #rubinius
Bwild has quit [Ping timeout: 255 seconds]
Dyrim has joined #rubinius