00:00:01  * ircretaryquit (Remote host closed the connection)
00:00:08  * ircretaryjoined
00:02:32  * loladiroquit (Quit: loladiro)
00:07:36  * loladirojoined
01:25:19  * dominictarrquit (Quit: dominictarr)
01:28:52  * dsantiagoquit (Ping timeout: 256 seconds)
01:39:44  * dsantiagojoined
01:43:26  * abraxasjoined
01:44:54  * btraskquit (Quit: btrask)
01:51:29  * dsantiagoquit (Ping timeout: 255 seconds)
01:52:41  * dsantiagojoined
01:57:48  * dsantiagoquit (Ping timeout: 264 seconds)
01:59:40  * dsantiagojoined
02:23:44  * dsantiagoquit (Ping timeout: 258 seconds)
02:25:13  * dsantiagojoined
02:32:29  * dsantiagoquit (Ping timeout: 245 seconds)
02:34:44  * dsantiagojoined
02:34:49  * inolenquit (Quit: Leaving.)
02:35:50  * TooTallNatequit (Quit: ["Textual IRC Client: www.textualapp.com"])
02:54:29  * dsantiagoquit (Ping timeout: 255 seconds)
02:55:11  * dsantiagojoined
03:02:29  * dsantiagoquit (Ping timeout: 248 seconds)
03:05:14  * dsantiagojoined
03:07:16  * perezdjoined
03:24:13  * loladiroquit (Quit: loladiro)
03:26:14  * dsantiagoquit (Ping timeout: 256 seconds)
03:27:43  * dsantiagojoined
03:35:54  * loladirojoined
03:37:00  * inolenjoined
03:39:17  * dsantiagoquit (Ping timeout: 248 seconds)
03:41:12  * dsantiagojoined
03:41:50  * loladiroquit (Quit: loladiro)
03:43:38  * loladirojoined
03:59:04  * loladiroquit (Quit: loladiro)
04:02:42  * loladirojoined
04:04:08  * dsantiagoquit (Ping timeout: 256 seconds)
04:05:15  * dsantiagojoined
04:11:00  * dsantiagoquit (Ping timeout: 264 seconds)
04:12:07  * dsantiagojoined
04:17:08  * dsantiagoquit (Ping timeout: 252 seconds)
04:18:02  * dsantiagojoined
04:30:25  * dsantiagoquit (Ping timeout: 245 seconds)
04:31:51  * dsantiagojoined
04:40:13  * dsantiagoquit (Ping timeout: 240 seconds)
04:45:31  * dsantiagojoined
04:53:30  * perezdquit (Quit: perezd)
05:10:18  * dsantiagoquit (Ping timeout: 264 seconds)
05:12:32  * dsantiagojoined
05:23:22  * paddybyersjoined
05:27:29  * dsantiagoquit (Ping timeout: 255 seconds)
05:29:30  * dsantiagojoined
05:34:36  * dsantiagoquit (Ping timeout: 245 seconds)
05:37:01  * dsantiagojoined
05:43:50  * dsantiagoquit (Remote host closed the connection)
05:44:30  * dsantiagojoined
06:07:03  * dsantiagoquit (Ping timeout: 276 seconds)
06:11:29  * dsantiagojoined
06:25:14  * felixgejoined
06:25:14  * felixgequit (Changing host)
06:25:14  * felixgejoined
06:27:00  * dsantiagoquit (Ping timeout: 256 seconds)
06:30:06  * felixgequit (Ping timeout: 264 seconds)
06:31:02  * dsantiagojoined
06:32:16  * felixgejoined
06:32:16  * felixgequit (Changing host)
06:32:16  * felixgejoined
06:37:31  * `3rdEdenjoined
06:43:00  * paddybyersquit (Remote host closed the connection)
06:43:15  * paddybyersjoined
07:01:45  * wolfeidauquit (Remote host closed the connection)
07:04:56  * benoitcquit (Excess Flood)
07:05:23  * benoitcjoined
07:12:32  * rendarjoined
07:21:05  * wolfeidaujoined
07:29:00  * dsantiagoquit (Ping timeout: 264 seconds)
07:31:03  * dsantiagojoined
07:32:42  * csaohjoined
07:33:34  * paddybyers_joined
07:37:13  * paddybyersquit (Ping timeout: 240 seconds)
07:37:13  * paddybyers_changed nick to paddybyers
08:01:04  * dsantiagoquit (Ping timeout: 256 seconds)
08:04:02  * dsantiagojoined
08:09:12  * dsantiagoquit (Ping timeout: 264 seconds)
08:11:02  * dsantiagojoined
08:19:08  * loladiroquit (Ping timeout: 256 seconds)
08:19:46  * dsantiagoquit (Ping timeout: 256 seconds)
08:20:31  * dsantiagojoined
08:32:06  * dsantiagoquit (Ping timeout: 245 seconds)
08:32:49  * loladirojoined
08:35:01  * dsantiagojoined
08:45:29  * dsantiagoquit (Ping timeout: 255 seconds)
08:47:31  * dsantiagojoined
08:52:29  * dsantiagoquit (Ping timeout: 245 seconds)
08:53:30  * dsantiagojoined
08:59:18  * dsantiagoquit (Ping timeout: 276 seconds)
09:01:55  * dsantiagojoined
09:03:38  * dsantiagoquit (Max SendQ exceeded)
09:04:30  * dsantiagojoined
09:10:03  * csaohquit (Quit: csaoh)
09:17:26  * dsantiagoquit (Ping timeout: 255 seconds)
09:20:00  * dsantiagojoined
09:22:10  * bajtosjoined
09:27:33  * bajtosquit (Quit: bajtos)
09:27:52  * bajtosjoined
09:35:11  * csaohjoined
09:35:44  * dominictarrjoined
09:55:23  * dsantiagoquit (Ping timeout: 252 seconds)
09:58:36  * dsantiagojoined
09:59:04  * bnoordhuisjoined
10:06:01  * dsantiagoquit (Ping timeout: 258 seconds)
10:06:47  * bajtosquit (Ping timeout: 258 seconds)
10:09:01  * dsantiagojoined
10:09:39  * dsantiagoquit (Max SendQ exceeded)
10:10:22  * dsantiagojoined
10:16:47  * hzjoined
10:22:21  * stagasjoined
10:28:58  * dsantiagoquit (Ping timeout: 256 seconds)
10:31:21  * dsantiagojoined
10:38:11  * loladiroquit (Quit: loladiro)
10:57:03  * csaohquit (Quit: csaoh)
11:03:54  * paddybyersquit (Ping timeout: 258 seconds)
11:06:12  * dsantiagoquit (Ping timeout: 264 seconds)
11:06:15  * csaohjoined
11:06:55  * loladirojoined
11:08:24  * dsantiagojoined
11:16:54  * dsantiagoquit (Ping timeout: 264 seconds)
11:17:52  * dsantiagojoined
11:23:56  * abraxasquit (Remote host closed the connection)
11:25:21  * dsantiagoquit (Ping timeout: 256 seconds)
11:27:21  * dsantiagojoined
11:34:39  * dsantiagoquit (Ping timeout: 256 seconds)
11:36:11  * dsantiagojoined
11:41:13  * dsantiagoquit (Ping timeout: 240 seconds)
11:42:42  * dsantiagojoined
11:47:19  * stagasquit (Ping timeout: 246 seconds)
11:56:06  * dsantiagoquit (Ping timeout: 276 seconds)
11:57:19  <bnoordhuis>...
11:57:42  * dsantiagojoined
11:57:45  <bnoordhuis>when a cluster workers calls net.Server#listen() with {fd: n}
11:58:06  <bnoordhuis>it's sent fd n from the _parent_
11:58:49  <bnoordhuis>i mean, why? when is that ever a good idea
12:04:28  * dsantiagoquit (Remote host closed the connection)
12:05:10  * dsantiagojoined
12:32:00  * AvianFlujoined
12:44:01  * dsantiagoquit (Ping timeout: 246 seconds)
12:45:36  * defunctzombie_zzchanged nick to defunctzombie
12:46:35  <indutny>bnoordhuis: hey man
12:46:43  <indutny>have you got a chance to look at new writev stuff?
12:49:17  <bnoordhuis>indutny: sorry, no time
12:50:40  * dsantiagojoined
12:51:38  * bajtosjoined
13:07:33  * dsantiagoquit (Ping timeout: 240 seconds)
13:09:38  * dsantiagojoined
13:17:37  * AvianFluquit (Read error: Connection reset by peer)
13:18:02  * AvianFlujoined
13:25:58  * defunctzombiechanged nick to defunctzombie_zz
13:26:44  * dsantiagoquit (Ping timeout: 255 seconds)
13:29:12  * dsantiagojoined
13:38:13  * defunctzombie_zzchanged nick to defunctzombie
14:01:25  * dsantiagoquit (Ping timeout: 256 seconds)
14:03:40  * dsantiagojoined
14:06:34  * bnoordhuisquit (Ping timeout: 256 seconds)
14:11:36  * felixgequit (Quit: http://www.debuggable.com/)
14:16:02  * AvianFluquit (Remote host closed the connection)
14:17:31  * dsantiagoquit (Ping timeout: 245 seconds)
14:19:40  * dsantiagojoined
14:19:52  * AvianFlujoined
14:32:57  * bnoordhuisjoined
14:41:30  * bnoordhuisquit (Ping timeout: 264 seconds)
14:56:46  * loladiroquit (Quit: loladiro)
15:02:48  * bajtosquit (Quit: bajtos)
15:10:07  <kellabyte>loving libuv, thanks for a fantastically portable library :)
15:12:49  * bnoordhuisjoined
15:21:47  * csaohquit (Remote host closed the connection)
15:22:06  * bnoordhuisquit (Ping timeout: 245 seconds)
15:23:53  * dsantiagoquit (Ping timeout: 240 seconds)
15:25:23  * `3rdEdenchanged nick to `3E|DINNER
15:27:17  * dsantiagojoined
15:35:23  * bnoordhuisjoined
15:40:27  <isaacs>bnoordhuis: yeah, that's weird
15:40:48  <isaacs>bnoordhuis: you'd think that the fd was something you already have (ie, maybe the parent already sent it?)
15:41:53  * bnoordhuisquit (Ping timeout: 252 seconds)
15:54:37  * bajtosjoined
16:02:23  * loladirojoined
16:03:01  * loladiroquit (Client Quit)
16:05:38  * bnoordhuisjoined
16:12:51  * dapjoined
16:16:26  <isaacs>kellabyte: i'm very curious about what you're doing. it seems like it'll be node's cousin, but a database.
16:16:51  * sgallaghjoined
16:16:52  * csaohjoined
16:17:15  * loladirojoined
16:20:49  * dsantiagoquit (Ping timeout: 256 seconds)
16:21:57  * dsantiagojoined
16:24:07  * sgallaghquit (Remote host closed the connection)
16:25:32  * `3E|DINNERquit (Remote host closed the connection)
16:26:06  * isaacsback to TxJS-ing &
16:26:11  <tjfontaine>enjoy
16:29:20  * dsantiagoquit (Ping timeout: 260 seconds)
16:31:33  * dsantiagojoined
16:35:19  * loladiroquit (Quit: loladiro)
16:35:33  * brsonjoined
16:39:53  * qmxchanged nick to qmx|lunch
16:47:41  * loladirojoined
16:50:59  * `3rdEdenjoined
16:53:08  * inolenquit (Quit: Leaving.)
16:54:38  * benoitcquit (Ping timeout: 258 seconds)
17:00:21  * dsantiagoquit (Quit: Computer has gone to sleep.)
17:11:34  * piscisaureus_joined
17:12:23  * TooTallNatejoined
17:15:43  * perezdjoined
17:17:15  * bnoordhuisquit (Ping timeout: 256 seconds)
17:19:09  * rjejoined
17:19:41  * csaohquit (Quit: csaoh)
17:19:57  <bajtos>Hello everyone, I'd like to submit a pull request (744) but some unit-tests are failing.
17:20:08  <bajtos>It seems that the same unit-tests are failing on master branch as well.
17:20:25  <bajtos>Is somebody looking at the problem and working on a fix?
17:20:25  <indutny>you mean failing without your patch?
17:20:29  <bajtos>yes
17:20:46  <tjfontaine>you should also submit prs to help clear out those tests :)
17:20:47  <indutny>on what platform are you testing?
17:20:49  <bajtos>http://jenkins.nodejs.org/job/libuv-master/75/testReport/
17:22:05  * dapquit (Quit: Leaving.)
17:24:03  <kellabyte>isaacs: awhile ago I started experimenting on writing my own database, I've learned a lot, first started in C#, now I'm starting over learning in C++ :)
17:24:23  <indutny>heh
17:24:37  <indutny>kellabyte: what data structure are you going to use?
17:24:38  <indutny>B-tree?
17:25:39  <kellabyte>I started with an LSM SSTable storage engine which was LevelDB but now using an append-only B-Tree storage engine, I can probably easily support both since they are both key/value stores
17:26:10  <indutny>yeah
17:26:14  <indutny>kewl
17:26:25  <indutny>I've built one B+-Tree engine years ago
17:26:30  <kellabyte>oh cool :)
17:26:43  <kellabyte>I'm not good enough to write the storage engines yet, but there are some very good ones out there alreadyd
17:26:43  <indutny>https://github.com/indutny/bplus
17:26:49  <indutny>feel free to borrow anything from it
17:26:57  <indutny>it's MIT licensed
17:27:26  * dapjoined
17:27:43  <kellabyte>ah cool, thanks, so far the one I'm using is this one: http://www.bzero.se/ldapd/btree.html https://github.com/OrangeTide/btree
17:28:13  <indutny>ah, its much better
17:28:13  <kellabyte>lock free modified prefix B+tree append-only
17:28:14  <indutny>I suppose
17:28:35  <indutny>mine isn't lock free :)
17:28:45  <indutny>I wasn't that experienced at the time of implementing it
17:29:00  <kellabyte>still worth reading, I'm not that good to build that kind of stuff yet
17:29:47  * inolenjoined
17:29:53  <bajtos>talking about databases - did you know about this data structure from MSFT? http://research.microsoft.com/pubs/178758/bw-tree-icde2013-final.pdf
17:31:55  <indutny>yep
17:31:55  <indutny>interesting paper
17:33:17  * trevnorrisjoined
17:36:36  <tjfontaine>g'day trevnorris
17:36:52  <trevnorris>tjfontaine: hello hello
17:42:45  * piscisaureus_quit (Read error: Connection reset by peer)
17:44:13  <MI6>nodejs-master: #151 UNSTABLE windows-ia32 (7/580) windows-x64 (8/580) http://jenkins.nodejs.org/job/nodejs-master/151/
17:52:52  * piscisaureus_joined
17:53:45  * dominictarrquit (Quit: dominictarr)
17:54:00  * jmar777joined
18:00:36  <trevnorris>fortunately I realized the memory issue isn't a race condition. it can be captured in debug mode by running the benchmark indefinitely.
18:00:47  <trevnorris>now to make sense of all the debug data
18:05:27  <trevnorris>indutny: have a minute?
18:08:09  <trevnorris>anyone mind verifying a bug for me?
18:08:21  <tjfontaine>what's the issue?
18:08:25  <trevnorris>tjfontaine: https://gist.github.com/trevnorris/5390043
18:08:56  <kellabyte>bajtos: I did read that one, it's used in SQL Server's Hekaton work, not as good as the B-Tree I linked to earlier from my understanding (requires compaction etc)
18:09:14  <trevnorris>tjfontaine: wanted someone else to run the same and see if they could reproduce.
18:09:42  <tjfontaine>trevnorris: building debug now
18:09:46  <trevnorris>tjfontaine: thanks.
18:12:14  <tjfontaine>trevnorris: tripped here as well
18:12:16  * qmx|lunchchanged nick to qmx
18:12:47  <trevnorris>tjfontaine: thanks.
18:13:46  <tjfontaine>this definitely seems like it belongs in indutny's wheelhouse
18:14:35  <trevnorris>tjfontaine: yeah. i'm trying to determine if my new buffer implementation really is crashing, or if it just exposed this bug.
18:16:15  <trevnorris>tjfontaine: ah. interesting. can you run one more thing? "/usr/bin/time ./node benchmark/tls/tls-connect.js concurrency=1 dur=120"
18:16:29  <trevnorris>tjfontaine: mine is crashing w/ "Command exited with non-zero status 8"
18:17:37  * eris0xffjoined
18:19:09  <tjfontaine>no I'm not crashing with that
18:20:44  <trevnorris>tjfontaine: hm. i'm getting the following: https://gist.github.com/trevnorris/5390043#file-non-debug-bench-log
18:23:37  <eris0xff>hi. I'm messages off the wire using libuv. since uv_read_start calls the alloc/read callbacks continuously, is the established parsing pattern to only do one uv_read_start and make the read callback one big state machine (uhh!) ... or ... read_start, malloc and do a read_start/stop for each token collected. I'd prefer the latter because I don't have to track so much state, but that works only if read_start, stop isn't a hugely ex
18:23:37  * bnoordhuisjoined
18:24:05  <trevnorris>bnoordhuis: so that benchmark is crashing for another reason: https://gist.github.com/trevnorris/5390043#file-tls-bench-debug-log
18:24:51  <eris0xff>(it would also save me a lock/unlock)
18:26:07  <tjfontaine>trevnorris: it still crashes, but not with the 'exited non-zero' but that could be just a `time` difference
18:27:10  <trevnorris>tjfontaine: hm. the exit status is delivered from FatalException in node.cc. isaacs recently gave each exit status a unique value so they could be tracked.
18:27:15  <trevnorris>tjfontaine: what error are you getting?
18:28:29  <trevnorris>eris0xff: not sure if it will, but might help to look at node's implementation: https://github.com/joyent/node/blob/master/src/stream_wrap.cc#L147-L165
18:29:18  <tjfontaine>trevnorris: I get the full abort trap like you, just don't have that exited text
18:29:25  <trevnorris>tjfontaine: ah, ok.
18:30:49  <eris0xff>trevnorris: thx
18:31:07  <trevnorris>eris0xff: yup. and come back w/ any more questions. =)
18:31:47  <eris0xff>:-)
18:38:22  <eris0xff>looks like node is using some type of stream object wrapper, but if I'm reading the pattern correctly i think it's setting up separate uv_read_starts per token type or lexeme
18:40:57  * felixgejoined
18:40:58  * felixgequit (Changing host)
18:40:58  * felixgejoined
18:42:25  * perezdquit (Quit: perezd)
18:43:07  <trevnorris>eris0xff: long and short. i'd say the later of the two earlier examples is more correct.
18:43:21  <eris0xff>thx
18:44:00  <eris0xff>read start does a uv__io_start and uv__handle_start
18:44:06  <eris0xff>on the stream
18:44:51  <bnoordhuis>trevnorris: sorry, i have precious little time today and tomorrow
18:45:02  * piscisaureus_quit (Ping timeout: 252 seconds)
18:45:23  <tjfontaine>"You hereby have my consent to fork joyent/node and maintain that patch yourself." haha
18:48:20  <trevnorris>bnoordhuis: no worries. i'm just not going to worry about the failing benchmarks, since it fails anyways.
18:48:27  <trevnorris>until someone else can get a look at it.
18:50:00  <eris0xff>trevnorris: i assume there are multiple reasons for doing double callbacks on stream i/o (one for alloc and one for read)
18:50:05  <eris0xff>(probably to integrate IOCP and reactor models)
18:50:56  <eris0xff>isn't there any particular reason that I should avoid passing a pre-allocated buffer pointer rather than mallocing?
18:51:11  <eris0xff>(I should be able to keep re-using the buffer i have in mind)
18:53:42  * dominictarrjoined
18:54:24  <trevnorris>eris0xff: not that I can think of, but take my advise w/ a grain of salt. I'm only as familiar w/ libuv as I need to be for node.
18:55:28  * eris0xffquit (Ping timeout: 245 seconds)
18:57:42  <trevnorris>tjfontaine: this issue probably isn't related to the "double link-list" error i'm getting from the buffer changes, is it?
18:57:57  <tjfontaine>no probably not
18:58:04  <trevnorris>=/
19:00:52  <trevnorris>tjfontaine: well, this makes it a pain since I hit the other issue in debug mode first.
19:04:21  * piscisaureus_joined
19:15:47  <MI6>joyent/node: Ben Noordhuis v0.10 * afbadde : os: handle 256 character hostnames Fix a (rather academic) buffer overfl (+1 more commits) - http://git.io/Kj40Pg
19:17:00  * hzquit (Ping timeout: 264 seconds)
19:17:45  * loladiroquit (Quit: loladiro)
19:19:27  * bajtosquit (Quit: off to bed)
19:20:55  * Raltquit (Ping timeout: 246 seconds)
19:22:55  * Raltjoined
19:23:42  * brsonquit (Quit: leaving)
19:23:55  * brsonjoined
19:26:39  <MI6>nodejs-v0.10: #135 FAILURE smartos-x64 (1/576) osx-ia32 (1/576) smartos-ia32 (1/576) http://jenkins.nodejs.org/job/nodejs-v0.10/135/
19:35:30  * piscisaureus_quit (Ping timeout: 264 seconds)
19:36:48  * AvianFluquit (Read error: Connection reset by peer)
19:37:24  * AvianFlujoined
19:48:36  * piscisaureus_joined
20:13:49  * eris0xffjoined
20:25:58  <tjfontaine>bnoordhuis broke the windows build
20:26:20  <tjfontaine>src\node_os.cc(56): error C2065: 'MAXHOSTNAMELEN' : undeclared identifier [g:\jenkins\workspace\nodejs-v0.10\DESTCPU\x64\label\windows\node.vcxproj]
20:26:23  <tjfontaine>src\node_os.cc(56): error C2133: 'buf' : unknown size [g:\jenkins\workspace\nodejs-v0.10\DESTCPU\x64\label\windows\node.vcxproj]
20:26:25  <tjfontaine>src\node_os.cc(58): error C2070: 'char []': illegal sizeof operand [g:\jenkins\workspace\nodejs-v0.10\DESTCPU\x64\label\windows\node.vcxproj]
20:28:23  * eris0xffquit (Ping timeout: 245 seconds)
20:33:05  <bnoordhuis>ah
20:36:38  <bnoordhuis>commit afbadde, how poetical
20:36:53  <tjfontaine>heh
20:39:56  <MI6>joyent/node: Ben Noordhuis v0.10 * d58ee7e : os: unbreak windows build Windows doesn't have MAXHOSTNAMELEN. Introduce - http://git.io/0NGleA
20:41:09  <trevnorris>bnoordhuis: i've traced every buffer allocation in node_crypto, and they're only used for return values. never set as part of the class.
20:41:20  * jmar777quit (Remote host closed the connection)
20:41:30  <trevnorris>then removed the SlabBuffer so each allocation was copied, not sliced, to make sure that wasn't an issue either.
20:41:33  <trevnorris>ugh.
20:42:52  * loladirojoined
20:44:37  <trevnorris>since every backtrace always leads back to Alloc or ObjectWrap weak callback, must be some strange interaction going on there.
20:44:48  <trevnorris>just haven't found it yet.
20:59:06  <MI6>nodejs-v0.10: #136 UNSTABLE smartos-x64 (1/576) windows-ia32 (5/576) windows-x64 (5/576) smartos-ia32 (2/576) http://jenkins.nodejs.org/job/nodejs-v0.10/136/
21:01:31  * rendarquit
21:03:09  <trevnorris>tjfontaine: you know of any tools to help me track memory locations? right now i'm just fprintf all the new/delete's and seeing if any match the printed address on error.
21:03:13  <trevnorris>but that's not working.
21:03:29  <tjfontaine>there are other tools, similar to valgrind that will help with that
21:03:33  <tjfontaine>but
21:03:53  <tjfontaine>the fact that you're throwing a VM in the mix, their utility starts to wane slightly
21:06:28  * luigyquit (Ping timeout: 256 seconds)
21:06:29  <indutny>tjfontaine: it still works
21:06:37  <indutny>valgrind
21:06:57  <tjfontaine>it works, it's just sometimes more difficult to track things, it's fine so long as you're malloc'ing outside the vm
21:07:13  <trevnorris>ok. i'll give that a go again.
21:15:36  <trevnorris>indutny: have any feed back on this: https://gist.github.com/trevnorris/5390043
21:15:59  * jmar777joined
21:19:37  <trevnorris>whoot! got some useful data from valgrind: https://gist.github.com/trevnorris/5391372
21:47:49  * `3rdEdenquit (Remote host closed the connection)
21:59:56  <trevnorris>bnoordhuis, tjfontaine: fyi, figured out the issue... I want to kick myself in the face. a lot.
22:00:39  <tjfontaine>hehe, what it was?
22:03:04  <trevnorris>in e501ce4b piscisaureus_ removed the slab backing for buffers, if zero length. I was making it persistent regardless, so two make weak callbacks were being run, and each tried to Dispose the handle.
22:03:36  <tjfontaine>ah
22:04:12  <trevnorris>well, i think that was it.
22:04:33  * loladiroquit (Quit: loladiro)
22:04:50  <trevnorris>it was either that, or i'd allocate an external array data to the object at size 0, then crypto tried to resize it then write to it.
22:05:04  <trevnorris>valgrind was complaining it was writing out of bounds.
22:07:53  <trevnorris>the error messages were cryptic, and it fail at different locations. so i'm still not completely sure.
22:09:21  * perezdjoined
22:10:37  <trevnorris>piscisaureus_: maybe you can enlighten me on the change in e501ce4b. was backing a zero-length buffer causing an issue before?
22:11:07  <piscisaureus_>trevnorris: for someone, yes. It was a micro-opt
22:13:52  <trevnorris>piscisaureus_: hm. well for some reason backing the new buffers caused me 10 hours of debugging. =P
22:18:25  * `3rdEdenjoined
22:19:37  * trippehquit (Ping timeout: 245 seconds)
22:26:58  * `3rdEdenquit (Ping timeout: 258 seconds)
22:30:34  * wolfeidauquit (Remote host closed the connection)
22:44:23  * dsantiagojoined
22:44:37  * piscisaureus_quit (Ping timeout: 245 seconds)
22:51:43  * wolfeidaujoined
23:00:08  <trevnorris>indutny: you still around?
23:03:25  <trevnorris>tjfontaine: if you do a "new char[0]", does that still need to be free'd?
23:06:44  <tjfontaine>trevnorris: off hand I don't know, but a simple int main() { void *f = malloc(0); } run through valgrind will tell you :)
23:07:00  <trevnorris>tjfontaine: ah, good idea. thanks. =)
23:08:26  <trevnorris>so I fixed one problem, and made others.
23:08:47  <trevnorris>swear that if I ever get these buffers working going to take a week off and play halo 4
23:09:37  * mikealjoined
23:10:09  <TooTallNate>trevnorris: +1
23:10:11  <TooTallNate>hahahah
23:10:20  <TooTallNate>actually 4 isn't so great from what I hear :\
23:10:39  <trevnorris>well. I have 4 days and 22 hours on it. =)
23:18:32  <trevnorris>TooTallNate: mind giving a second pair of eyes on this output? https://gist.github.com/trevnorris/5391372/raw/cb623d8cf7de28741fa0849aaf686b3045702060/bb-valgrind.log
23:19:19  <trevnorris>i see where the write is coming from, but I can't figure out why.
23:19:38  <trevnorris>all I know is that for some reason node_crypto makes a lot of zero length buffers.
23:20:50  <TooTallNate>lol
23:21:29  <TooTallNate>sounds like our bindings could use some work
23:21:45  <trevnorris>how do you mean?
23:22:55  <TooTallNate>nothing i'm just grumpy
23:23:12  <trevnorris>heh
23:28:02  <kellabyte>how many IOCP threads does libuv use?
23:43:35  <trevnorris>TooTallNate: know a way to make Undefined() a function argument (e.g. "Local<Value> argv[1] = { Undefined(); }")
23:44:26  <TooTallNate>trevnorris: Local::New(Undefined())
23:44:29  <TooTallNate>i think will work
23:45:44  <trevnorris>TooTallNate: ah, thanks. that was close. Local<Value>::New(Undefined()).
23:45:54  <TooTallNate>ahhh
23:45:54  <TooTallNate>ya that
23:45:55  <TooTallNate>:p
23:52:15  * piscisaureus_joined
23:52:21  * mikealquit (Quit: Leaving.)
23:53:10  <piscisaureus_>kellabyte: re how many IOCP threads does libuv use?
23:53:30  <piscisaureus_>kellabyte: for stuff that goes through the IOCP no threads are needed
23:54:10  <piscisaureus_>kellabyte: except when something runs in the thread pool (like fs ops and getaddrinfo) - that thread pool auto scales up to 512 threads on windows.
23:56:37  * CoverSlidejoined
23:57:03  <trevnorris>TooTallNate: think there's a possibility of a bug in node_cryto that's causing me all this pain?
23:57:34  <TooTallNate>trevnorris: so what is the pain exactly?
23:57:39  <TooTallNate>trevnorris: there's always a chance for bugs
23:57:43  <TooTallNate>no code is perfect ;)
23:58:26  <trevnorris>TooTallNate: first, all tests pass. but benchark/tls/tls-connect.js fails hard.
23:59:21  <trevnorris>TooTallNate: it fails w/ "*** glibc detected *** ./node: double free or corruption"
23:59:40  <trevnorris>and you already saw the valgrind output.
23:59:41  <TooTallNate>that's messed