Index

Show enters and exits. Hide enters and exits.

00:00:40kronos_vano leaves the room.
00:01:16libc leaves the room.
00:01:59ruejptix: At Death Star, working on world domination. Web/REST framework stuff
00:03:33jptixcool
00:09:18somebody_ leaves the room.
00:10:42joearnold leaves the room.
00:11:00malumalu leaves the room.
00:14:18libc__ enters the room.
00:14:29somebody_ enters the room.
00:22:58libc_ leaves the room.
00:33:19soveran leaves the room.
00:36:48binary42 leaves the room.
00:37:08binary42 enters the room.
00:39:46brynary leaves the room.
00:39:53binary42 leaves the room.
00:40:10binary42 enters the room.
00:43:49binary42 leaves the room.
00:47:15binary42 enters the room.
01:00:37blowmage leaves the room.
01:06:16joearnold enters the room.
01:07:31brynary enters the room.
01:07:49joearnold leaves the room.
01:13:00macournoyer enters the room.
01:15:45joearnold enters the room.
01:15:55nari leaves the room.
01:23:52binary42 leaves the room.
01:29:22benburkert leaves the room.
01:32:56joearnol_ enters the room.
01:33:26joearnold leaves the room.
01:33:53codebrulee enters the room.
01:34:51codebrulee leaves the room.
01:35:45lopex leaves the room.
01:44:18ezmob leaves the room.
01:44:21nari enters the room.
02:14:16macournoyer leaves the room.
02:20:14libc_ enters the room.
02:22:31_VVSiz_ enters the room.
02:23:21flori leaves the room.
02:26:26flori enters the room.
02:28:50libc__ leaves the room.
02:29:28VVSiz_ leaves the room.
02:40:52kamalfariz enters the room.
02:44:14benburkert enters the room.
02:47:33VVSiz_ enters the room.
02:50:13joearnol_ leaves the room.
02:57:39binary42 enters the room.
03:01:41seydar enters the room.
03:01:51seydarbrixen: i used your name again on my math test today
03:01:57seydari honestly think it's helping
03:02:59brixenheh
03:03:16brixenare you using my name in vain?
03:03:48_VVSiz_ leaves the room.
03:05:07seydar leaves the room.
03:13:32macournoyer enters the room.
03:23:27bdimcheff enters the room.
03:26:30nicksieger enters the room.
03:42:21peterhil leaves the room.
03:44:14bdimcheff leaves the room.
03:45:03bdimcheff enters the room.
03:57:43cout enters the room.
04:05:03macournoyer leaves the room.
04:10:27wmoxam enters the room.
04:11:36bdimcheff leaves the room.
04:16:47peterhil enters the room.
04:22:16mdalessio enters the room.
04:40:00wmoxam leaves the room.
04:52:22bdimcheff enters the room.
05:02:43drbrain leaves the room.
05:21:28mdalessio leaves the room.
05:25:55imajes leaves the room.
05:29:27carllerche enters the room.
05:31:50shoe leaves the room.
05:34:56shoe enters the room.
05:37:52bdimcheff leaves the room.
05:52:33somebody__ enters the room.
05:55:26benburkert leaves the room.
06:01:49somebody_ leaves the room.
06:02:24imajes enters the room.
06:04:08benburkert enters the room.
06:08:08asap18 leaves the room.
06:17:59brynary leaves the room.
06:24:44michaellatta enters the room.
06:30:09febeling leaves the room.
06:30:48binary42 leaves the room.
06:33:31blowmage enters the room.
06:51:45headius leaves the room.
06:54:37brynary enters the room.
07:07:16libc__ enters the room.
07:07:20brynary leaves the room.
07:07:32somebody_ enters the room.
07:14:53somebody__ leaves the room.
07:15:34libc_ leaves the room.
07:15:48moconnor enters the room.
07:23:35somebody_ leaves the room.
07:35:43drbrain enters the room.
07:36:23nicksieger leaves the room.
07:50:31benburkert leaves the room.
07:52:21joachimm enters the room.
07:59:26blowmage leaves the room.
08:05:32moconnor leaves the room.
08:09:49benburkert enters the room.
08:35:53joachimm leaves the room.
08:38:20kronos_vano enters the room.
08:44:01boyscoutRe-enabled instrumenting profiler. - 255a070 - Brian Ford
08:44:01boyscoutRemove filtering methods in profiler output. - 1d71a29 - Brian Ford
08:44:01boyscoutFixed responsibilities for creating, registering, removing, deleting profilers. - d815cbb - Brian Ford
08:44:01boyscoutBeginning docs for the profiler. - 63d4267 - Brian Ford
08:44:01boyscoutAdded #printf to IOStub helper. - 3c535b6 - Brian Ford
08:44:02boyscoutProfile FFI and C-API calls. - e3071d0 - Brian Ford
08:44:04boyscoutProfile block/proc calls. - 4804205 - Brian Ford
08:44:06boyscoutAdded specs for the instrumenting profiler. - 38eec19 - Brian Ford
08:44:19scooprgo brian, go! :D
08:44:31brixenhah
08:44:49brixensomeone should probably review that stuff
08:45:34brixenme and native threads aren't exactly bosom buddies
08:45:52brixenrue: in your many sleepless hours ^^^
08:46:32brixenshould probably check it on linux right quick as well
08:52:50boyscoutCI: 38eec19 success. 1444 files, 7122 examples, 23440 expectations, 0 failures, 0 errors
08:54:32kronos_vanoCan anyone explain how to use profile?
08:55:01brixenkronos_vano: take a look at doc/profiling.txt
08:55:07kronos_vanotnx
08:55:08brixenit also works the same as MRI
08:55:19brixenie bin/rbx -rprofile script.rb
08:58:16rueI may or may not be asleep
09:06:56rueprofilenates CI
09:57:32malumalu enters the room.
09:58:49gnufied enters the room.
10:04:45naeu enters the room.
10:11:04kamalfariz leaves the room.
10:14:14wvdschel enters the room.
10:17:22benburkert leaves the room.
10:19:04febeling enters the room.
10:43:22wvdschel leaves the room.
10:43:56botanicus enters the room.
10:59:26drbrain leaves the room.
11:01:16drbrain enters the room.
11:09:12krsh enters the room.
11:09:13wvdschel enters the room.
11:10:41nari leaves the room.
11:38:38chris2 enters the room.
11:49:41nari enters the room.
11:53:55mediogre enters the room.
12:03:08wvdschel leaves the room.
12:06:17botanicus leaves the room.
12:14:29binary42 enters the room.
12:57:54qbproger leaves the room.
13:01:17PhilK enters the room.
13:11:22shoe leaves the room.
13:11:22carllerche leaves the room.
13:11:22scoopr leaves the room.
13:11:22cypher23 leaves the room.
13:11:48shoe enters the room.
13:11:48carllerche enters the room.
13:11:48scoopr enters the room.
13:11:48cypher23 enters the room.
13:56:58febeling enters the room.
14:11:34macournoyer enters the room.
14:12:29soveran enters the room.
14:48:43mediogre leaves the room.
14:52:55gavinstark enters the room.
14:56:40brynary enters the room.
15:05:15febeling_ enters the room.
15:06:58febeling leaves the room.
15:10:06kamalfariz enters the room.
15:12:36binary42 leaves the room.
15:24:00febeling_ leaves the room.
15:43:37blowmage enters the room.
15:52:11nicksieger enters the room.
15:54:53therealadam enters the room.
15:55:06binary42 enters the room.
16:03:03wmoxam enters the room.
16:03:23naeu leaves the room.
16:05:23moconnor enters the room.
16:06:42nicksieger leaves the room.
16:07:36nicksieger enters the room.
16:08:05enebo enters the room.
16:11:45nari leaves the room.
16:13:41benburkert enters the room.
16:13:42naeu enters the room.
16:28:28moconnor leaves the room.
16:35:48benburkert leaves the room.
16:40:55asap18 enters the room.
16:48:02carllerche leaves the room.
16:56:36michaellatta enters the room.
17:00:16kronos_vano leaves the room.
17:07:06blowmage leaves the room.
17:11:11headius enters the room.
17:16:38kamalfariz leaves the room.
17:31:16michaellatta leaves the room.
17:31:17shoe leaves the room.
17:33:39blowmage enters the room.
17:34:56shoe enters the room.
17:35:29krawek enters the room.
17:40:48joachimm enters the room.
17:42:57radarek enters the room.
17:43:09jashmenn enters the room.
17:43:45ezmob enters the room.
17:58:29joearnold enters the room.
18:21:02ezmob_ enters the room.
18:21:03ezmob leaves the room.
18:21:25kronos_vano enters the room.
18:22:29cout leaves the room.
18:24:58jashmenn leaves the room.
18:24:58binary42 leaves the room.
18:24:58imajes leaves the room.
18:24:58skaar_ leaves the room.
18:24:58ice799 leaves the room.
18:26:50ice799 enters the room.
18:34:20skaar enters the room.
18:37:26joachimm leaves the room.
18:59:47naeu leaves the room.
19:06:39naeu enters the room.
19:09:04benburkert enters the room.
19:09:04binary42 enters the room.
19:10:07jashmenn enters the room.
19:19:39dodecaphonic enters the room.
19:23:28botanicus enters the room.
19:38:00joachimm enters the room.
19:43:38jashmenn leaves the room.
19:50:21outerim leaves the room.
19:50:56botanicus leaves the room.
19:51:52botanicus enters the room.
19:53:07soveran leaves the room.
19:53:23soveran enters the room.
20:01:32lopex enters the room.
20:05:00cout enters the room.
20:15:24outerim enters the room.
20:29:50evani'm home!
20:30:05scooprgrats!
20:35:07stepheneb enters the room.
20:37:07brixenwb evan!
20:37:11brixenbags make it?
20:37:26evancourse not!
20:37:31brixen:(
20:37:47evani've got all the critical things
20:37:48evanon me
20:37:52evanso it's fine
20:37:55evanand there is a later flight today
20:38:01evani could have waited at the airport
20:38:12evansupposedly it was on a flight that got in at noon
20:38:15evansupposedly.
20:38:21brixenahh
20:40:23brixenthis is most likely why we are getting the dots from the CI bot on clean builds: http://pastie.org/442099
20:40:38brixenhow a NULL vm gets into vms_ I do not know
20:41:38brixenevan: when you get a chance, pls review the profiler stuff
20:41:43evandon't trust that
20:41:48evanthats gdb on an optimized binary.
20:41:56brixennah, rake build:debug
20:42:15brixenunless that doesn't work on ubuntu
20:42:23evansounds like a thread timing issue
20:42:32soveranhey evan
20:42:42evaneasy enough to have ::prune be a bit more defensive
20:42:47evanand make sure VM isn't NULL
20:43:03evanadd a TODO to figure out why it's NULL.
20:43:06evansoveran: hello.
20:45:22brixenevan: I added an assert on vm
20:45:35brixenI could do if(vm && ...)
20:45:45evanthe assert doesn't solve the crash :)
20:45:46brixenbut when I added that, it seemed to hang for a long time
20:45:50brixenyeah
20:45:53benburkert leaves the room.
20:46:14evancan you get it to happen easy?
20:46:24brixensure
20:46:40brixendelete the .rbc in spec and run bin/mspec ci
20:46:44benburkert enters the room.
20:46:48brixenI'm on ubuntu
20:46:57evanok
20:47:00botanicus leaves the room.
20:47:04evani was going to fire up the new ubuntu
20:47:05brixenbut it has the same behavior as elle afaics
20:47:11evani'll do that and poke it
20:47:16evanit's probably just some thread timing
20:47:31evanit happens at the very end, yes?
20:47:43brixennear the end, yes
20:48:14benburkert leaves the room.
20:50:02joachimm leaves the room.
20:51:47brixenevan: http://pastie.org/442150
20:52:27evanwhats that?
20:52:36brixenthe rb bt when the assert on vm failed
20:52:55brixenso it's around core/time/gmtime
20:53:10brixenit seems to be consistent
20:53:31evanweird
20:53:33brixengrabbing some food, bbiab..
20:53:33evanwhats the whole bt?
20:53:35evank
20:53:48brixenthe whole rb bt?
20:54:36evanactually, nm
20:56:18outerim leaves the room.
20:57:20SoreGums leaves the room.
21:06:39botanicus enters the room.
21:07:19benny enters the room.
21:08:00blowmage_ enters the room.
21:14:11benburkert enters the room.
21:23:49blowmage leaves the room.
21:34:39nicksieger leaves the room.
21:35:10nicksieger enters the room.
21:35:32jashmenn enters the room.
21:46:54radarek leaves the room.
21:48:01bitsweat enters the room.
21:50:48gavinstark leaves the room.
21:51:01benburkert leaves the room.
22:00:33joearnold leaves the room.
22:01:43joearnold enters the room.
22:03:33PhilK leaves the room.
22:03:36joearnold leaves the room.
22:05:43joearnold enters the room.
22:06:21benburkert enters the room.
22:12:27stepheneb leaves the room.
22:13:20brixenbeautiful
22:13:21brixendescribe "Kernel#clone" do it "needs to be reviewed for spec completeness"
22:13:22brixenend
22:13:32brixenno wonder our #clone is broken
22:17:55joearnold leaves the room.
22:18:48macournoyer leaves the room.
22:19:49malumalugood spec :D
22:21:37stepheneb enters the room.
22:23:20imajes enters the room.
22:24:17brixenmalumalu: it's going to be gooder in a sec :)
22:24:47malumaluhm, i don't want to clutter the lighthouse tracker, is it more convenient for you if i just first collect some patches and post them together in an archive?
22:25:39malumalusince most of the patches are quite short and simple
22:25:59brixendidn't you already have a patch accepted?
22:27:02MarkMenard leaves the room.
22:27:29malumaluyes
22:27:48brixenmalumalu: ah yeah 5b480639 right?
22:27:56brixenyou should ping evan for your commit bit!
22:27:58mutle_ enters the room.
22:28:01brixenand you can commit directly
22:28:46evanmalumalu: whats your github username
22:28:50malumaluoh, that would make it easy :D
22:28:54malumaluevan: malu
22:29:01dodecaphonic leaves the room.
22:29:51evanok, you'r added
22:29:53evanbe nice!
22:30:10malumalui'll try to do so :)
22:30:33malumaluand thanks :D
22:34:18malumaluok then, could someone close tickets 759, 760, i'll commit them myself
22:34:47brixenmalumalu: what is your LH email? I'll send you an invite
22:35:44malumalubrixen: simply.malu+rubinius@googlemail.com
22:37:27brixenmalumalu: sent
22:38:26malumaluthank you
22:38:44brixenn/p
22:41:43evanweird
22:42:04evanjust realized that using obj.method(name).call lets you call private methods
22:42:09evanthat makes sense i guess.
22:45:00mutle leaves the room.
22:46:02binary42 leaves the room.
22:46:18blowmage_ leaves the room.
22:46:48blowmage enters the room.
22:47:37therealadamevan: yeah, bruce williams posted about that in something about not using send
22:47:58evangotcha
22:48:19therealadamI think it's an OK idiom
22:49:05stepheneb_ enters the room.
22:49:07stepheneb_ leaves the room.
22:49:07stepheneb leaves the room.
22:49:11stepheneb enters the room.
22:49:54imajes_ enters the room.
22:50:42benburkert leaves the room.
22:51:18imajes_ leaves the room.
22:54:17imajes leaves the room.
22:54:59Keltianotes that rubinius while being llvm-based is not compilable by clang :)
22:55:25headiusas I understand it clang support for C++ is still a little weak, no?
22:55:31Keltiavery weak
22:56:19Keltiabut the error I get may not be related to that
22:56:20Keltiahashtable_itr.c:39:1: error: redefinition of 'hashtable_iterator_key'
22:57:24Keltiaanyway, that was just a very quick test
22:58:13benburkert enters the room.
23:02:41joearnold enters the room.
23:02:44jaribKeltia: rubinius is llvm-based?
23:02:52headiusit's not at the moment
23:03:00headiusthere has been some experimental work, but no LLVM right now
23:03:21jaribright
23:06:21stepheneb leaves the room.
23:08:24Keltiaheadius: have a look in vm/external_libs/llvm... (in master)
23:09:04Keltiathe first version of rubinius was not llvm based but they rewrote it using llvm
23:09:19brixenevan: could you help me sort out some MRI madness
23:11:42joachimm enters the room.
23:12:30benburkert leaves the room.
23:15:13headiusI'm just about positive llvm is not in use right now
23:15:22brixenit's not
23:15:22headiusunless something major changed in the past couple weeks
23:16:29brixenKeltia: it's not rewritten using llvm, but evan has experimented with using llvm to generate machine methods from compiled methods
23:16:39brixenit's not hooked up atm
23:17:11qbproger enters the room.
23:25:56nari enters the room.
23:26:21therealadam leaves the room.
23:29:30wmoxam leaves the room.
23:30:34dodecaphonic enters the room.
23:31:00naeu leaves the room.
23:31:21benburkert enters the room.
23:32:07evanbrixen: whats up?
23:32:14Keltiabrixen: so what's on master?
23:32:36Keltiais confused now
23:33:27rueWho's on third?
23:33:32evanKeltia: whats your question?
23:33:34rueWhat's on second
23:34:04brixenKeltia: go look at rakelib/vm.rake
23:34:15brixenllvm is in the repo, we are not using it atm
23:34:41brixenevan: rb_obj_alloc calls rb_funcall(klass, ID_ALLOCATOR)
23:34:53brixenrb_obj_clone calls rb_obj_alloc
23:35:15brixenrb_define_alloc_func changes ID_ALLOCATOR method for the class
23:35:22Keltiabrixen: ok, understood, thanks
23:35:33brixenwe need a way to repro this either directly or just in c-api
23:36:01brixenbasically, ID_ALLOCATOR == 1, so mri just calls whatever is at hash key 1 in the method table
23:36:04brixenafaics
23:36:44brixenso, unless you've called rb_define_alloc_func, you get the default allocate()
23:36:58brixenand there's no way to change that from Ruby code in MRI that I see
23:37:57evanhm
23:38:26brixenso, I'm wondering about changing Object::clone to call .allocate on obj->klass instead of getting obj->total_size() bytes from the OM
23:38:27evanwhy doesn't it just hook it up as :allocate
23:38:28evani wonder.
23:38:32brixenyeah
23:38:39brixenyou'd wonder that
23:38:55brixenand then recall that MRI is insanity redux :P
23:39:28evanhm.
23:39:39brixenbasically, I think Object::clone should be all in C++
23:39:39evanit makes sense for it to call .allocate
23:39:44brixennot part in ruby
23:40:05evanso if that's all thats needed, sounds good
23:40:06brixenie send .allocate(); fix up; send #initialize_copy
23:40:06evan:)
23:40:13brixenthat's ok?
23:41:06brynary leaves the room.
23:41:06evancan't see why it wouldn't be
23:41:24brixenwell, you mentioned not wanting to send method calls from C++
23:41:31brixenwasn't sure what the guideline was
23:44:39malumalueh, where's StringValue defined?
23:45:14brixenkernel.rb
23:45:21malumaluthanks
23:45:26brixenn/p
23:45:28joachimm leaves the room.
23:47:48rueevan: headius is pruning their repo which reminds me that it is probably high time to get rid of some of the old branches
23:48:07evantrue
23:48:07evantis.
23:48:23headiusif you guys are going to reboot LLVM efforts that would be a huge one to delete
23:48:36evanhuh?
23:48:48joearnold leaves the room.
23:48:50headiusany branches where you have llvm's source checked in
23:48:53joearnold enters the room.
23:48:56evanthey all have it.
23:48:59evanso it doesn't really matter.
23:49:02headiusahh
23:49:11headiusI thought it had been removed...I suppose it's just not building
23:49:29evanyep
23:49:31evanjust not building
23:49:44evanit's preserved in history too
23:49:51evanso any branch that had it an anytime in the past
23:49:53evanwould keep it alive
23:50:03rueWonder how much space it will save..
23:50:15rueAre any of the branches keepers?
23:50:24headiusrue: after my pruning I'll let you know what a new clone size is
23:50:30headiusour repo is a lot older, but it will be interesting to see
23:50:55ruegh-pages, probably
23:54:11evanwell
23:54:45evani'd guess that cmakebuild, cpp, instance_eval, stackfull, and wilson64
23:54:48evancould go, no question
23:54:59rueA full clone is ~100MiB nowadays
23:55:06evanfusion-experiment, i want to see if check is any good code in there
23:55:19evanwell, pruning our branches likely wont make much of a dent in that
23:55:26evanmaster has had almost all files on it at some point
23:56:27bitsweat leaves the room.
23:57:01rueAny little bit helps at this point
23:57:20evantrue enough.
23:57:33evani'm asking the GH admins when they pack a repo
23:57:42evanbecause ours is big enough, it would make a difference now