| 00:07:09
| * plutoniix | joined |
| 00:07:20
| * plutoniix | quit (Max SendQ exceeded) |
| 00:08:07
| * plutoniix | joined |
| 01:03:09
| * xaxxon_ | quit (Ping timeout: 250 seconds) |
| 01:27:51
| * xaxxon | joined |
| 03:03:37
| * xaxxon | quit (Ping timeout: 250 seconds) |
| 03:08:57
| * xaxxon | joined |
| 04:14:09
| * zv | joined |
| 05:07:32
| * xaxxon | quit (Ping timeout: 240 seconds) |
| 05:38:12
| * zv | quit (Quit: WeeChat 1.9) |
| 07:44:16
| <trungl-bot> | Tree closed by [email protected]: closed - maintenance |
| 08:08:27
| <trungl-bot> | Tree opened by [email protected]: open |
| 09:58:18
| * vivekp_ | joined |
| 10:00:45
| * vivekp_ | quit (Client Quit) |
| 10:01:09
| <trungl-bot> | Tree closed by [email protected]: closed (https://build.chromium.org/p/client.v8/builders/V8%20Win64%20-%20msvc/builds/69 from 6c6132880adf636960f47662570766ab424be534) |
| 10:03:18
| * vivekp_ | joined |
| 10:12:42
| * vivekp_ | quit (Quit: This computer has gone to sleep) |
| 10:14:15
| <trungl-bot> | Tree opened by [email protected]: open |
| 10:39:37
| * vivekp_ | joined |
| 10:54:02
| * vivekp_ | quit (Quit: This computer has gone to sleep) |
| 10:59:39
| * vivekp_ | joined |
| 11:02:25
| * vivekp_ | quit (Client Quit) |
| 11:25:09
| * mylesborins | quit (Quit: farewell for now) |
| 11:25:40
| * mylesborins | joined |
| 11:53:38
| * AtumT | joined |
| 12:14:43
| * plutoniix | quit (Quit: Leaving) |
| 12:38:20
| <trungl-bot> | Tree closed by [email protected]: closed (https://build.chromium.org/p/client.v8/builders/V8%20Linux64%20TSAN/builds/18232 from 22932d6b43e9df81c29116f9e4ff1c1e36a0aaa3) |
| 12:55:27
| <trungl-bot> | Tree opened by [email protected]: open (TSAN issue is being fixed) |
| 13:14:28
| * vivekp_ | joined |
| 13:15:34
| * vivekp_ | quit (Client Quit) |
| 13:16:55
| * vivekp_ | joined |
| 13:19:02
| * vivekp_ | quit (Client Quit) |
| 13:20:47
| * vivekp_ | joined |
| 13:30:59
| * AtumT_ | joined |
| 13:32:57
| * AtumT | quit (Ping timeout: 240 seconds) |
| 13:38:59
| * vivekp_ | quit (Quit: This computer has gone to sleep) |
| 13:41:31
| * vivekp_ | joined |
| 13:43:44
| * vivekp_ | quit (Client Quit) |
| 13:44:54
| * vivekp_ | joined |
| 13:51:53
| <trungl-bot> | Tree closed by [email protected]: closed - maintenance |
| 14:02:34
| * vivekp_ | quit (Quit: This computer has gone to sleep) |
| 14:24:04
| <trungl-bot> | Tree opened by [email protected]: open |
| 15:08:37
| * seventh | joined |
| 15:10:10
| * seventh | quit (Remote host closed the connection) |
| 15:53:25
| * vivekp_ | joined |
| 15:54:01
| * vivekp_ | quit (Client Quit) |
| 15:56:26
| * vivekp_ | joined |
| 16:10:02
| * bradleymeck | joined |
| 16:13:28
| * vivekp_ | quit (Quit: This computer has gone to sleep) |
| 16:20:33
| * Guest59 | joined |
| 16:24:44
| * Guest59_ | joined |
| 16:26:11
| * Guest59 | quit (Ping timeout: 240 seconds) |
| 16:27:58
| * Guest59 | joined |
| 16:27:58
| * Guest59_ | quit (Read error: Connection reset by peer) |
| 16:33:52
| * Guest59 | quit (Ping timeout: 240 seconds) |
| 16:34:04
| * Guest59 | joined |
| 16:35:26
| * vivekp_ | joined |
| 17:03:12
| * AtumT | joined |
| 17:05:46
| * AtumT_ | quit (Ping timeout: 264 seconds) |
| 17:09:09
| * seventh | joined |
| 17:21:00
| * plutoniix | joined |
| 17:21:36
| * plutoniix | quit (Max SendQ exceeded) |
| 17:22:15
| * plutoniix | joined |
| 17:24:58
| * RT|Chatzilla | quit (Quit: ChatZilla 0.9.86.1 [Firefox 2.0.0.22pre/2010030309]) |
| 17:27:05
| * bradleymeck | quit (Quit: ) |
| 18:58:30
| * seventh | quit (Remote host closed the connection) |
| 19:13:40
| <caitp> | it would be cool if we didn't emit `ThrowSuperNotCalledIfHole` every time `this` is referenced in subclasses |
| 19:13:56
| <caitp> | I mean it's just an extra byte but it seems unnecessary |
| 19:16:46
| <caitp> | only need to hole-check the first use in each thread of the graph, so |
| 19:17:13
| <caitp> | "it doesn't matter once it gets to turbofan" sure, if it gets to turbofan |
| 19:44:51
| <caitp> | it does seem like the ~1000-line classes-subclass-builtins.js takes a long time to run, with or without my WIP patch (not sure if I'm making it worse or better, I probably slowed down parse time a bit recently, but this seems worse than just parse time), and the super hole-checks show up a lot in that file |
| 19:45:03
| <caitp> | even though they are very simple bytecodes |
| 19:48:57
| * Guest59 | quit (Ping timeout: 240 seconds) |
| 19:57:06
| * Guest59 | joined |
| 20:16:24
| * zv | joined |
| 20:38:29
| * Guest59 | quit (Quit: My Mac has gone to sleep. ZZZzzz…) |
| 20:43:22
| * Guest59 | joined |
| 21:07:06
| * Guest59 | quit (Quit: My Mac has gone to sleep. ZZZzzz…) |
| 21:44:49
| <gsathya> | caitp: +1. we should be able to elide the hole check for most cases in the parser. surprised we dont do that already |
| 21:50:42
| * vivekp_ | quit (Quit: This computer has gone to sleep) |
| 21:57:54
| <caitp> | on that note, come to think of it, we only ever really need a single code copy of the default subclass constructor |
| 21:58:29
| <caitp> | but instead we make a new one for every class, when it could instead just be a new JSFunction referencing a rooted code object |
| 22:00:16
| <caitp> | probably doesn't come up too much in practice, but it's memory and cpu cycles spent when they don't need to be |
| 22:04:49
| <caitp> | I guess each one probably needs its own SharedFunctionInfo, but maybe that's ok |
| 22:40:09
| * Guest59 | joined |
| 23:19:36
| * RT|Chatzilla | joined |
| 23:48:27
| * AtumT | quit (Remote host closed the connection) |
| 23:56:31
| * Guest59 | quit (Quit: My Mac has gone to sleep. ZZZzzz…) |