<|jemc|>
chrchr: you should know that while rubinius installed via `rbenv install rbx-2.4.1` will work, installing from source into the ~/.rbenv/versions subdir will not work for gem binaries until you run rbenv's internal bash function called "fix_rbx_binstubs"
<|jemc|>
(which rbenv will run for you when you run `rbenv install rbx-...`)
<|jemc|>
or rather, it's ruby-build's internal bash function
<|jemc|>
but that's not the same problem as the problem you're describing - just one that you were sure to run into next after you got building to work ;)
|jemc| has quit [Ping timeout: 245 seconds]
Bwild has quit [Read error: No route to host]
Bwild has joined #rubinius
amsi has quit [Quit: Leaving]
|jemc| has joined #rubinius
tenderlove has joined #rubinius
amclain has joined #rubinius
nirvdrum has quit [Ping timeout: 245 seconds]
Bwild has quit [Ping timeout: 245 seconds]
Bwild has joined #rubinius
tenderlove has quit [Remote host closed the connection]
tenderlove has joined #rubinius
tenderlove has quit [Remote host closed the connection]
tenderlove has joined #rubinius
tenderlove has quit [Remote host closed the connection]
tenderlove has joined #rubinius
tenderlove has quit [Remote host closed the connection]
tenderlove has joined #rubinius
tenderlove has quit [Remote host closed the connection]
tenderlove has joined #rubinius
tenderlove has quit [Remote host closed the connection]
tenderlove has joined #rubinius
tenderlove has quit [Remote host closed the connection]
tenderlove has joined #rubinius
tenderlove has quit [Remote host closed the connection]
tenderlove has joined #rubinius
tenderlove has quit [Remote host closed the connection]
Prathame_ has joined #rubinius
tenderlove has joined #rubinius
tenderlove has quit [Remote host closed the connection]
tenderlove has joined #rubinius
tenderlove has quit [Remote host closed the connection]
meh` has quit [Ping timeout: 256 seconds]
tenderlove has joined #rubinius
tenderlove has quit [Remote host closed the connection]
tenderlove has joined #rubinius
tenderlove has quit [Remote host closed the connection]
tenderlove has joined #rubinius
Prathame_ has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
tenderlove has quit [Ping timeout: 240 seconds]
GitHub184 has joined #rubinius
<GitHub184>
[rubinius] brixen pushed 2 new commits to master: http://git.io/tVYo-w
<GitHub184>
rubinius/master 43f72a7 Brian Shirai: Removed libyaml (again)....
<GitHub184>
rubinius/master 32639db Brian Shirai: Merge branch 'jit-method'
Prathame_ has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
Prathame_ has joined #rubinius
tenderlove has joined #rubinius
tenderlove has quit [Ping timeout: 246 seconds]
Prathame_ has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
Prathame_ has joined #rubinius
Prathame_ has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
tenderlove has joined #rubinius
Prathame_ has joined #rubinius
tenderlove has quit [Ping timeout: 252 seconds]
Prathame_ has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
Prathame_ has joined #rubinius
[spoiler] has joined #rubinius
tenderlove has joined #rubinius
tenderlove has quit [Ping timeout: 256 seconds]
Prathame_ has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
Prathame_ has joined #rubinius
Prathame_ has quit [Ping timeout: 264 seconds]
johnmuhl has quit [Quit: Connection closed for inactivity]
tenderlove has joined #rubinius
tenderlove has quit [Ping timeout: 264 seconds]
JohnBat26 has joined #rubinius
Prathame_ has joined #rubinius
Prathame_ has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
Prathame_ has joined #rubinius
amclain has quit [Quit: Leaving]
tenderlove has joined #rubinius
tenderlove has quit [Ping timeout: 240 seconds]
Prathame_ has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
Prathame_ has joined #rubinius
Prathame_ has quit [Client Quit]
Prathame_ has joined #rubinius
Prathame_ has quit [Client Quit]
Prathame_ has joined #rubinius
tenderlove has joined #rubinius
Prathame_ has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
tenderlove has quit [Ping timeout: 252 seconds]
elia has joined #rubinius
_elia has joined #rubinius
elia has quit [Ping timeout: 240 seconds]
Prathame_ has joined #rubinius
_elia has quit [Quit: Computer has gone to sleep.]
Prathame_ has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
tenderlove has joined #rubinius
tenderlove has quit [Ping timeout: 245 seconds]
goyox86 has joined #rubinius
josh-k has joined #rubinius
<yorickpeterse>
\0/
Prathame_ has joined #rubinius
Prathame_ has quit [Client Quit]
goyox86 has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
<yorickpeterse>
brixen: seems Rubinius::JIT.compile doesn't do anything if you give it a class vs an instance
<yorickpeterse>
brixen: e.g. Rubinius::JIT.compile(SomeClass, SomeClass.instance_method(:foo).executable) doesn't actually seem to improve perf of "foo"
<yorickpeterse>
while using an instance of SomeClass does
<yorickpeterse>
In my totes legit fib() benchmark the latter makes it 2x faster
<yorickpeterse>
the former does nothing
tenderlove has joined #rubinius
tenderlove has quit [Ping timeout: 265 seconds]
Bwild has quit [Remote host closed the connection]
Bwild has joined #rubinius
Bwild has quit [Client Quit]
meh` has joined #rubinius
goyox86 has joined #rubinius
benlovell has joined #rubinius
goyox86 has quit [Client Quit]
johnmuhl has joined #rubinius
goyox86 has joined #rubinius
goyox86_ has joined #rubinius
[spoiler] has quit [Quit: Leaving]
goyox86 has quit [Read error: Connection reset by peer]
goyox86 has joined #rubinius
tenderlove has joined #rubinius
benlovell has quit [Ping timeout: 245 seconds]
benlovell has joined #rubinius
tenderlove has quit [Ping timeout: 245 seconds]
goyox86_ has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
goyox86 has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
Prathame_ has joined #rubinius
<chrisseaton>
yorickpeterse: you're compiling the benchmark before you've ever run it. Can you do that? Does it have any profiling info to go on at all? Wouldn't the first thing the compiled code do deoptimize because all the ICs will be empty and all the branches will never have been executed before.
<chrisseaton>
Not that I know much about how Rubinius works, but in most dynamic language VMs that would be the case.
<chrisseaton>
If you can do it then isn't Rubinius already an AOT compiler for Ruby, and why aren't you dumping that compiled code into elf files?
Pratham__ has joined #rubinius
Prathame_ has quit [Ping timeout: 264 seconds]
josh-k_ has joined #rubinius
josh-k has quit [Ping timeout: 264 seconds]
benlovell has quit [Quit: leaving]
Pratham__ is now known as cha1tanya
<yorickpeterse>
chrisseaton: not sure if I'm following
<yorickpeterse>
the first question: yes, you can now at least tell the JIT to try and compile something
<yorickpeterse>
although it can still decide not to do it from what I gather (e.g. it might just not be able to compile something)
<chrisseaton>
but then all your ICs will be empty in that compiled right? so when it tries to run that compiled code it will stop on the first method call
<yorickpeterse>
how effective it is before calling the method I'm not sure about
<yorickpeterse>
I suppose it depends on what is being compiled
<yorickpeterse>
e.g. you don't strictly need to call something in order to do constant folding (assuming trivial cases such as "10 + 10")
<yorickpeterse>
The gist is mostly meant to show that you can at least now manually instruct the JIT
<chrisseaton>
yorickpeterse: but you can't constant fold that until you've done the method lookup and seen the types actually involved
<chrisseaton>
unless Rubinius tries to optimistically guess types when it generates byte code - I wouldn't know
<yorickpeterse>
Like I said it depends on how trivial it would be, I'm not familiar enough with the JIT to know what assumptions it might make
<brixen>
yorickpeterse: when are you speaking at fosdem again?
<brixen>
well, not again, but the first time
<yorickpeterse>
brixen: 2014-01-31 - 2014-02-01
<yorickpeterse>
my talk is the 31st, first one of the day
<brixen>
ok, cool
<brixen>
yorickpeterse: so, at this point, jitting a method with empty caches won't be able to do anything more than emit calls
<brixen>
it will be able to elide as much other stuff as LLVM can make sense of with the IR we emit
<brixen>
which you can easily inspect with the -Xjit.dump_code options
<brixen>
with the new insn set, a lot of optimistic assumptions will be possible, and functions will be statically inlined since they rely on lexical scope
booly-yam-271 has joined #rubinius
<yorickpeterse>
Yeah I figured it would do not that much
<yorickpeterse>
but nevertheless it's interesting to see that it at least makes a small difference
booly-yam-271 has quit [Remote host closed the connection]
unreal has joined #rubinius
cha1tanya has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
goyox86 has joined #rubinius
Prathame_ has joined #rubinius
Prathame_ is now known as cha1tanya
goyox86_ has joined #rubinius
chrisseaton has quit [Ping timeout: 245 seconds]
jeregrine has quit [Ping timeout: 276 seconds]
cha1tanya has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
<brixen>
ok, JIT metrics output fixed
<brixen>
next up, completely audit metrics
diegoviola has joined #rubinius
Prathame_ has joined #rubinius
<yorickpeterse>
yay, half my stuff moved over to S3/CloudFront
Prathame_ has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz…]
havenwood has quit [Remote host closed the connection]
|jemc| has quit [Ping timeout: 256 seconds]
yorickpeterse has joined #rubinius
yorickpeterse has quit [Client Quit]
dreinull75 has joined #rubinius
<dreinull75>
Thank you for releasing 2.5.0!
<dreinull75>
However, I can't install the skylight gem: checking for rb_thread_call_without_gvl() in ruby/thread.h... no checking for rb_thread_blocking_region()... no Ruby is unexpectedly missing rb_thread_blocking_region. This should not happen.