00:13:00  * rmichnikquit (Quit: Connection closed for inactivity)
00:33:28  <aklein>caitp: yeah, if this patch is all about perf I think the right thing to attack would be the hole checks, not the scoping (if VAR is wrong)
00:34:18  * bnoordhuisjoined
00:37:41  <caitp>aklein: it's not spelled out in the cl, not clear where it makes sense to put it other than tests. anyways, redeclaring parameters with var is always ok, even if under the hood they're implemented as let. (re cl comment)
00:37:47  <caitp>hard to answer comments on rietveld from mobile
00:38:49  * bnoordhuisquit (Ping timeout: 264 seconds)
00:40:11  <aklein>caitp: ok, so the behavior's getting "more correct" with this patch? as I said, this part of the spec is one that I don't (yet?) have internalized
00:41:14  <caitp>Well, it's more correct per spec
00:41:40  <caitp>I'm not sure there's any observable difference other than perf though
00:45:08  <aklein>caitp: I guess I'd be inclined to hold off for now on this patch; it's a good chunk of added code
00:47:43  <caitp>i do want to fix the perf of rest params before shipping though, it's pretty bad. slower than the old impl
00:48:03  * trungl-botquit (Remote host closed the connection)
00:48:42  * scottmgquit (Ping timeout: 244 seconds)
00:48:56  <aklein>caitp: that's fair
00:49:53  * guorquit (Ping timeout: 246 seconds)
00:49:53  * rmcilroyquit (Ping timeout: 246 seconds)
00:50:35  * WakiMikoquit (Ping timeout: 246 seconds)
00:53:15  * ncthom91joined
00:53:57  * rmcilroyjoined
00:54:08  * guorjoined
00:54:18  * trungl-botjoined
00:55:11  * WakiMikojoined
01:12:09  <caitp>aklein: I've updated the bugfix with your suggestion, I'll split the refactoring into another cl if you think it's worth cleaning it up a bit
01:12:32  * seventhjoined
01:16:13  * jgiquit (Quit: jgi)
01:29:31  * plutoniixjoined
02:33:58  * seventhquit (Ping timeout: 268 seconds)
03:00:29  * ncthom91quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
03:44:12  * esasjoined
03:57:25  * plutoniixquit (Read error: Connection reset by peer)
03:58:55  * plutoniixjoined
04:25:07  * jgijoined
04:32:35  * ncthom91joined
04:52:24  * ncthom91quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
04:55:27  * jgiquit (Quit: jgi)
04:57:58  * thefourtheyejoined
05:03:45  * esasquit (Read error: No route to host)
05:04:05  * scottmgjoined
05:08:09  * jgijoined
05:17:55  * jgiquit (Quit: jgi)
05:18:20  * jgijoined
05:22:34  <trungl-bot>Tree closed by [email protected]: Tree is closed (Automatic: "webkit_unit_tests" on http://build.chromium.org/p/client.v8.fyi/builders/V8-Blink%20Linux%2064%20%28dbg%29/builds/1356 "V8-Blink Linux 64 (dbg)" from b37907ff7f866873ddfbfc97670b43c19a5fc7f9: [email protected])
05:22:49  * jgiquit (Client Quit)
05:33:54  * ncthom91joined
05:38:02  * scottmgquit (Ping timeout: 244 seconds)
05:38:19  * enaqxquit (Remote host closed the connection)
05:38:42  <trungl-bot>Tree opened by [email protected]: Tree is open (Blink broken? Chromium tree also red)
05:44:59  * enaqxjoined
05:51:18  * ncthom91quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
06:45:47  * JoWiejoined
07:06:59  * mostynbjoined
07:48:45  * BobGneuquit (Ping timeout: 244 seconds)
07:49:50  * Bob_Gneujoined
08:07:04  * aperezdcjoined
08:55:35  * enaqxquit (Remote host closed the connection)
09:04:44  * enaqxjoined
09:14:11  * aperezdcquit (Ping timeout: 250 seconds)
09:16:13  * aperezdcjoined
09:25:06  * bnoordhuisjoined
10:06:12  * plutoniixquit (Read error: Connection reset by peer)
10:06:47  * plutoniixjoined
10:26:19  * plutoniixquit (Quit: จรลี จรลา)
10:45:45  * juanjosanchezjoined
11:02:43  * Net147quit (Ping timeout: 268 seconds)
11:03:56  * Net147joined
11:08:42  * Net147quit (Ping timeout: 250 seconds)
11:09:27  * bnoordhuisquit (Ping timeout: 260 seconds)
11:11:19  * Net147joined
11:15:49  * enaqxquit (Remote host closed the connection)
11:23:42  * enaqxjoined
11:33:41  * juanjosanchezquit (Read error: Connection reset by peer)
11:39:07  * danielijoined
11:41:13  * enaqxquit (Remote host closed the connection)
11:44:09  * enaqxjoined
11:45:03  * danieliquit (Read error: Connection reset by peer)
11:46:13  * enaqxquit (Remote host closed the connection)
11:47:09  * bnoordhuisjoined
12:21:35  <trungl-bot>Tree closed by [email protected]: Tree is closed (Automatic: "bot_update" on http://build.chromium.org/p/client.v8.fyi/builders/Chromium%20Win%20SyzyASAN/builds/231 "Chromium Win SyzyASAN" from ba7b64139886ab7987406a92bf0e801b807c9fe4: [email protected],[email protected])
12:30:21  * enaqxjoined
12:35:14  * enaqxquit (Remote host closed the connection)
12:35:41  <trungl-bot>Tree opened by [email protected]: Tree is open
12:38:43  * bnoordhuisquit (Ping timeout: 250 seconds)
12:48:02  * danielijoined
13:45:05  * bnoordhuisjoined
13:46:08  * bobmcwjoined
13:49:57  * bnoordhuisquit (Ping timeout: 272 seconds)
13:51:19  * davijoined
14:05:23  * plutoniixjoined
14:08:36  * C-Manjoined
14:08:55  * danieliquit (Ping timeout: 265 seconds)
14:12:26  * aperezdcquit (Quit: aperezdc)
14:13:13  * jgijoined
14:27:31  * danielijoined
14:29:52  * danieliquit (Read error: Connection reset by peer)
14:32:56  * jgiquit (Quit: jgi)
14:45:29  * dchermanjoined
14:52:12  * enaqxjoined
14:52:36  * bradleymeckjoined
14:58:42  * enaqxquit (Remote host closed the connection)
15:04:47  * mostynbquit (Quit: Leaving)
15:07:57  * enaqxjoined
15:08:02  * enaqxquit (Remote host closed the connection)
15:08:14  * enaqxjoined
15:43:27  * aperezdcjoined
15:44:43  * bnoordhuisjoined
15:48:34  * RT|Chatzillaquit (Quit: ChatZilla 0.9.86.1 [Firefox 2.0.0.22pre/2009081014])
16:08:06  * jgijoined
16:13:27  * Bob_Gneuquit (Ping timeout: 240 seconds)
16:14:38  * ncthom91joined
16:14:49  * Bob_Gneujoined
16:28:28  * ncthom91quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
16:32:13  <caitp>littledan: --print-scopes seems to indicate that it's not context allocation. Something in my "conditionally ignore TDZ" patch (which is mostly just using `var` instead of `let` for desugared formals, and getting rid of an extra scope --- both if HasParameterExpressions() from the spec is true for a function) causes the rest benchmark to outperform master by a factor of 10
16:32:40  <caitp>it's bizarre since it doesn't look like it's crawling up the context chain or anything
16:34:16  <caitp>s/outperform/outscore/?
16:51:14  * esasjoined
17:01:00  * ncthom91joined
17:04:32  * bradleymeckquit (Quit: bradleymeck)
17:20:51  * scottmgjoined
18:07:22  * bradleymeckjoined
18:16:26  * jgiquit (Quit: jgi)
18:22:17  * JoWiequit (Quit: Connection closed for inactivity)
18:22:49  * jgijoined
18:27:50  * aperezdcquit (Quit: aperezdc)
18:28:19  * aperezdcjoined
18:39:49  * daviquit (Ping timeout: 246 seconds)
18:41:55  * plutoniixquit (Ping timeout: 246 seconds)
18:42:53  * plutoniixjoined
18:55:45  * ncthom91quit (Quit: Textual IRC Client: www.textualapp.com)
19:06:02  * davijoined
19:06:30  * bradleymeckquit (Quit: bradleymeck)
19:07:19  <caitp>does anyone know what specific chain of events can cause the compiler of a code object to be reported as "unknown" (eg in https://github.com/v8/v8-git-mirror/blob/master/src/objects.cc#L11792-L11797)?
19:07:29  <caitp>trying to figure out what that means here.
19:42:08  * bnoordhuisquit (Ping timeout: 252 seconds)
19:45:11  * bradleymeckjoined
19:47:44  * jgiquit (Quit: jgi)
19:50:10  * daviquit (Ping timeout: 246 seconds)
20:07:57  * bobmcwquit (Remote host closed the connection)
20:30:32  * bnoordhuisjoined
20:32:07  * C-Manquit (Quit: Connection reset by beer)
20:51:21  <littledan>caitp, what if you hack fullcodegen to eliminate the hole checks?
20:52:44  <caitp>I think I've done that, although it's hard to tell from the generated code since the hole value isn't obvious
20:52:54  <caitp>being a pointer and all
20:59:42  * microcol1nelquit (Changing host)
20:59:42  * microcol1neljoined
20:59:51  * microcol1nelchanged nick to microcolonel
21:01:44  * jgijoined
21:13:53  <caitp>maybe the use of "let" isn't really the major botleneck? it's probably just the spec-incorrect assignment that babel uses that gets it 10x faster, maybe helped a tiny bit by manual licm and preallocation of the array, but those probably don't make a huge difference
21:14:23  <caitp>but it sounds like the compiler people don't want to add a way to get the speed of assignment with the spec correctness that is needed
21:24:53  * enaqx_joined
21:27:10  * dchermanquit (Ping timeout: 244 seconds)
21:27:30  * enaqxquit (Ping timeout: 250 seconds)
21:29:34  * enaqxjoined
21:29:53  * enaqx_quit (Ping timeout: 268 seconds)
21:32:57  * rendarjoined
21:34:07  * enaqxquit (Ping timeout: 240 seconds)
21:34:57  * enaqxjoined
21:39:20  * dchermanjoined
21:43:25  * enaqx_joined
21:45:22  * enaqx__joined
21:46:28  * enaqxquit (Ping timeout: 255 seconds)
21:47:46  * enaqx_quit (Ping timeout: 246 seconds)
21:47:50  * bradleymeckquit (Quit: bradleymeck)
21:50:59  * enaqx__quit (Ping timeout: 265 seconds)
21:51:21  * bradleymeckjoined
21:56:09  * bradleymeckquit (Client Quit)
22:02:17  * aperezdcquit (Ping timeout: 244 seconds)
22:04:02  * enaqxjoined
22:08:18  * bradleymeckjoined
22:25:55  * RT|Chatzillajoined
22:26:00  * dchermanquit (Ping timeout: 268 seconds)
22:37:38  * enaqxquit (Remote host closed the connection)
22:39:18  * enaqxjoined
22:41:59  * bradleymeckquit (Quit: bradleymeck)
23:10:44  * scottmgquit (Ping timeout: 265 seconds)
23:11:43  * scottmg_joined
23:12:11  * scottmgjoined
23:21:34  * seventhjoined
23:33:00  * xiinotulpjoined
23:36:25  * plutoniixquit (Ping timeout: 264 seconds)
23:39:05  * rendarquit
23:49:51  * bnoordhuisquit (Ping timeout: 240 seconds)
23:52:17  * enaqxquit (Remote host closed the connection)
23:55:01  * enaqxjoined