00:28:36  * bnoordhuisquit (Read error: Operation timed out)
00:38:17  * vikingquit (Remote host closed the connection)
00:48:24  * xaqjoined
01:04:42  * C-Manquit (Quit: Connection reset by beer)
01:47:34  * joshthecoderquit (Quit: Leaving...)
02:57:36  * xaqquit (Remote host closed the connection)
03:20:27  * xaqjoined
03:38:28  * joshthecoderjoined
04:35:35  * xaq_joined
04:36:03  * xaqquit (Read error: Connection reset by peer)
04:37:47  * xaq_quit (Remote host closed the connection)
04:53:40  * [[zzz]]joined
04:56:35  * [[zz]]quit (Ping timeout: 245 seconds)
05:10:22  * [[zzz]]changed nick to [[zz]]
05:11:36  * Deathzorquit (Ping timeout: 240 seconds)
05:19:23  * Deathzorjoined
06:53:59  * rendarjoined
07:07:17  * joshthecoderquit (Quit: Leaving...)
07:29:27  * tmcguirejoined
07:31:29  * xan_joined
08:03:26  * Net147joined
10:36:45  * decoderquit (Ping timeout: 256 seconds)
10:37:58  * decoderjoined
11:24:18  * bnoordhuisjoined
11:28:39  * Net147quit (Quit: HydraIRC -> http://www.hydrairc.com <- Nine out of ten l33t h4x0rz prefer it)
12:36:56  * C-Manjoined
15:09:35  * jamespage_joined
15:09:44  * jamespage_quit (Read error: Connection reset by peer)
15:19:41  * RT|Chatzillaquit (Quit: ChatZilla 0.9.86.1 [Firefox 2.0.0.22pre/2009081014])
17:48:26  * joshthecoderjoined
17:52:40  * ranmanjoined
19:06:57  * xaqjoined
19:12:56  * ricky26joined
19:14:32  <ricky26>Does anyone know anything about the stack-checking functionality of v8?
19:14:43  <ricky26>I am trying to get it to work with my micro-threading library
19:14:49  <ricky26>but it just returns null from Context::New();
19:15:11  <ricky26>I've had a poke around, and it looks like I could use an isolate to get at the functionality
19:15:22  <ricky26>but so far my searches have drawn up naught.
19:24:24  * joshthecoderquit (Quit: Leaving...)
20:08:00  * joshthecoderjoined
20:24:22  * mralephjoined
21:00:36  * joshthecoderquit (Read error: Connection reset by peer)
21:00:46  * joshthecoderjoined
21:22:07  <sanjoyd>ricky26: you can call Isolate::Current() to get the current Isolate.
21:22:16  <sanjoyd>But that will incur a TLS lookup, I think.
21:22:55  <sanjoyd>ricky26: what exactly are you trying to do?
21:44:02  * xaqquit (Remote host closed the connection)
21:46:36  * xaqjoined
21:50:43  * rendarquit
22:04:43  <ricky26>sanjoyd: Use V8 with userland threads.
22:04:57  <ricky26>sanjoyd: I'm currently using swapcontext & co.
22:05:15  <sanjoyd>ricky26: only one thread can be executing code in an isolate at one time.
22:05:32  <ricky26>sanjoyd: yeah, these would all run under one kernel-level thread.
22:05:40  <ricky26>they can't actually execute concurrently
22:05:44  <ricky26>they're like co-routines.
22:05:54  <sanjoyd>Umm, V8 uses TLS in some places.
22:06:07  <ricky26>Which should still work, as they are running in the same thread. :<
22:06:32  <ricky26>Theoretically, I'm only messing with the GP registers and the stack pointer
22:06:43  <ricky26>it's the stack pointer bit which is upsetting V8.
22:06:47  <sanjoyd>Are you adjusting the stack limits?
22:06:53  <ricky26>sanjoyd: How do I do this?
22:06:58  <sanjoyd>Yeah, V8 inserts a stack check after at every back edge and function call.
22:07:03  <ricky26>:<
22:07:04  <sanjoyd>Let me see check.
22:07:09  <ricky26>that sounds like exactly what I need.
22:16:59  <sanjoyd>ricky26: the real limit is a heap root (StackLimit) in heap.h
22:17:11  <ricky26>huh?
22:17:14  <sanjoyd>As in the limit that the compiled JS fetches and compares the SP against.
22:17:43  <sanjoyd>But this is set / modified in a lot of places -- for instance when threads enter and exit isolates.
22:18:05  <sanjoyd>grep for SetStackLimits
22:26:57  * RT|Chatzillajoined
22:55:57  * xaqquit (Remote host closed the connection)
23:08:11  * superjoejoined
23:08:45  <superjoe>hey, since what version of v8 has there been generator support?
23:08:52  <superjoe>`yield` and such
23:09:24  <superjoe>oh, looks like maybe it's not out yet.
23:10:01  <mraleph>superjoe: there is no such version of v8 AFAIK
23:10:32  * bnoordhuisquit (Ping timeout: 276 seconds)
23:11:20  <superjoe>how sad
23:11:22  <superjoe>thanks
23:13:00  <ranman>Why did you guys move off of v8?
23:13:53  <superjoe>who guys?
23:15:30  <mraleph>ranman: are you asking me?
23:15:44  <ranman>anyone who works on v8
23:15:44  <mraleph>ranman: what do you mean by "you"?
23:15:47  <ranman>errr
23:15:49  <ranman>wait
23:15:55  <ranman>wow I asked the wrong question sorry
23:16:01  <ranman>why did you guys move off of scons?
23:16:23  <mraleph>ah, because Chromium uses gyp to build itself.
23:16:35  <mraleph>so we essentially were maintaining two build systems.
23:16:43  <mraleph>which is not so nice.
23:16:51  <ranman>as I understand it -- you're still maintaing gyp and make
23:17:10  <ranman>also it seems to only look at timestamps for determining whether or not to rebuild files now
23:17:10  <mraleph>gyp generates makefile
23:17:21  <mraleph>probably.
23:17:30  <mraleph>gyp is "generate your projects"
23:17:32  <ranman>so yeah… that's not good :/
23:17:36  <mraleph>so it can generate many things.
23:17:50  <mraleph>ninja, makefiles, xcode projects etc.
23:17:59  <sanjoyd>ranman: how else to determine if you should rebuild?
23:18:09  <mraleph>sanjoyd: hash contents.
23:18:17  <mraleph>in addition to dates.
23:18:23  <sanjoyd>Hmm, but will that save much time?
23:18:30  <ranman>@sanjoyd: a ton of time.
23:18:39  <ranman>compiling takes much longer than checking/making a hash
23:18:40  <mraleph>dunno, probably not on my machine ;-)
23:18:58  <sanjoyd>ranman: yeah, that's true; but what I'm asking is that does it happen often
23:19:05  <sanjoyd>that the date on a file changes w/o the contents changing?
23:19:26  <ranman>@sanjoyd: all the time if people are copying source trees to and from machines (rsync)
23:19:36  <sanjoyd>Oh, okay.
23:19:39  <ranman>although I'm sure you can also preserve timestamps on rsync
23:19:44  <ranman>I just don't think most people think about that
23:19:56  <sanjoyd>Or use igt.
23:19:57  <sanjoyd>*git
23:20:06  <mraleph>or just use zip :-)
23:20:42  <ranman>well git would have the same problems right ?
23:22:43  <sanjoyd>Hmm, it seems git doesn't preserve modification times either.
23:23:02  <sanjoyd>But ideally you wouldn't version the binaries anyway, so I guess it usually doesn't matter.
23:45:45  <ranman>it doesn't usually
23:45:52  <ranman>I just have an awful use case
23:54:02  * xaqjoined