00:03:16  * bradleymeckquit (Quit: bradleymeck)
00:07:53  <jwolfe>can someone help me analyze a test failure? test/debugger/debug/debug-evaluate-no-side-effect-builtins.js is failing on line 89. it looks like `"abCd".toLowerCase()` is supposed to cause side effects? or it isn't supposed to cause side effects? i can't tell what this test is expecting.
00:08:08  <jwolfe>i'm sure i caused this failure, because i made a change to the .toLowerCase() builtin.
00:21:04  * plutoniixjoined
00:21:17  * plutoniixquit (Max SendQ exceeded)
00:21:53  * plutoniixjoined
01:46:12  * unixpicklejoined
01:59:30  * unixpicklequit (Quit: My Mac has gone to sleep. ZZZzzz…)
02:00:16  * bradleymeckjoined
02:05:45  * trevnorrisquit (Ping timeout: 260 seconds)
02:06:52  * trevnorrisjoined
02:47:15  * bradleymeckquit (Quit: bradleymeck)
02:49:53  * bradleymeckjoined
03:14:34  * bradleymeckquit (Quit: bradleymeck)
04:22:56  * bradleymeckjoined
04:34:18  * plutoniixquit (Quit: Leaving)
04:53:49  * chimera3quit (Ping timeout: 240 seconds)
04:54:07  * chimera3joined
05:23:20  <jwolfe>hm. it didn't break any bots in the dry run. i guess it's not really a problem?
05:33:27  * plutoniixjoined
06:07:58  * bradleymeckquit (Quit: bradleymeck)
06:55:21  * plutoniixquit (Quit: Leaving)
08:25:35  * rmcilroyquit (Ping timeout: 264 seconds)
08:37:04  * rmcilroyjoined
09:39:54  * plutoniixjoined
10:18:12  * plutoniixquit (Quit: Leaving)
10:25:09  * mylesborinsquit (Quit: farewell for now)
10:25:40  * mylesborinsjoined
12:55:24  * bradleymeckjoined
13:02:55  * unixpicklejoined
13:58:17  * unixpicklequit (Quit: My Mac has gone to sleep. ZZZzzz…)
14:00:43  * bradleymeckquit (Quit: bradleymeck)
14:08:31  * unixpicklejoined
14:15:02  * unixpicklequit (Quit: My Mac has gone to sleep. ZZZzzz…)
14:24:33  * bradleymeckjoined
14:37:18  * joyeejoined
14:47:37  * unixpicklejoined
14:50:59  * bradleymeckquit (Quit: bradleymeck)
15:18:03  * bradleymeckjoined
15:29:07  * joyeequit (Read error: Connection reset by peer)
15:29:19  * joyeejoined
15:40:10  * joyeequit (Remote host closed the connection)
15:43:20  * joyeejoined
16:19:47  * RT|Chatzillaquit (Quit: ChatZilla 0.9.86.1 [Firefox 2.0.0.22pre/2010030309])
16:31:04  * unixpicklequit (Quit: My Mac has gone to sleep. ZZZzzz…)
17:03:29  * seventhjoined
17:11:18  * joyeequit (Read error: Connection reset by peer)
17:11:49  * joyeejoined
17:12:19  * rwlbuis_joined
17:25:33  * bradleymeckquit (Quit: bradleymeck)
18:09:02  * Amadiro-quit (Excess Flood)
18:12:23  * joyeequit (Ping timeout: 264 seconds)
18:13:37  * Garbeequit (Ping timeout: 260 seconds)
18:14:38  * Amadirojoined
18:17:42  * Garbeejoined
18:18:05  * Garbeechanged nick to Guest85785
18:49:25  <aklein>jwolfe: are you saying it failed locally but not on the bots? still seems bad.
18:55:07  <jwolfe>aklein: yes. only in a noi18n configuration, so maybe i did something weird with args.gn?
18:57:20  <aklein>jwolfe: I don't see those tests run on that build https://build.chromium.org/p/tryserver.v8/builders/v8_linux_noi18n_rel_ng_triggered/builds/49
18:57:56  <aklein>so I suspect it's a real failure
18:59:55  * bradleymeckjoined
19:07:35  <jwolfe>aklein: what does it look like when a bot does run those tests?
19:07:59  <jwolfe>looking at these bot logs make me feel illiterate D_D
19:27:23  * Cube8joined
19:40:43  <aklein>jwolfe: https://build.chromium.org/p/tryserver.v8/builders/v8_linux_nodcheck_rel_ng_triggered/builds/16825 looks like they run as "Check - extra"
19:40:58  <aklein>to see that they didn't run, I clicked on "Check"'s stdio and searched for "debugger"
19:43:38  <jwolfe>but i don't see "debugger" in that second link's "Check - extra"'s "stdio" either
19:44:22  * seventhquit (Ping timeout: 268 seconds)
19:44:37  <aklein>jwolfe: there are two "Check - extra"s there, the one I'm looking at is item 21.
19:45:17  <jwolfe>god damn it. ok i see it.
19:45:57  <jwolfe>so is the process for seeing if a test ran to check all 4 "Check*"'s stdios?
19:48:37  <aklein>I mean, sort of
19:49:35  * bradleymeckquit (Quit: bradleymeck)
19:49:43  <aklein>I think the right thing in this case would be to file an infra bug to say that the noi18n bot should run more tests
19:53:10  * bradleymeckjoined
19:54:26  * bradleymeckquit (Client Quit)
19:59:45  <jwolfe>do the bots use run-tests.py?
20:06:10  <aklein>jwolfe: yep
20:06:43  <jwolfe>is the invocation in the logs? how you invoke run-tests.py determines the set of tests, right?
20:06:43  <aklein>you can see their commandline near the start of those stdio pages
20:07:16  <aklein>e.g., https://build.chromium.org/p/tryserver.v8/builders/v8_linux_nodcheck_rel_ng_triggered/builds/16825/steps/Check/logs/stdio is the check-extra that ran the debugger
20:08:08  <aklein>partly, but the commandline doesn't explicitly list the tests
20:09:07  <aklein>e.g., https://cs.chromium.org/chromium/src/v8/tools/run-tests.py?type=cs&q=options.no_i18n&l=623
20:27:24  * bradleymeckjoined
20:35:13  <jwolfe>so it looks like that bot doesn't have the inspector enabled? is that why it skipped the debugger tests?
20:43:47  <aklein>jwolfe: yup, that's the reason: https://build.chromium.org/p/tryserver.v8/builders/v8_linux_noi18n_rel_ng/builds/78/steps/generate_build_files%20with%20gn%20%28fyi%29/logs/stdio
20:45:11  <aklein>https://cs.chromium.org/chromium/src/v8/infra/mb/mb_config.pyl?rcl=f7036b132f6bcce9312b52a9ba386ae487280713&l=415 is the config
20:47:52  <aklein>and https://codereview.chromium.org/2393863002/#ps20001 is where it happened
20:47:58  <aklein>let me ping kozyatinskiy
20:58:54  * bradleymeckquit (Quit: bradleymeck)
21:06:52  <aklein>jwolfe: this may be historical, I've sent out a change to machenbach to change that config
21:15:02  * unixpicklejoined
21:21:51  <jwolfe>aklein: ok thanks.
21:26:24  <aklein>jwolfe: aha! this is a pre-existing failure, I think: https://chromium-review.googlesource.com/c/456646/
21:26:32  <aklein>see trybot on that change
21:27:14  * Net147quit (Quit: Quit)
21:28:27  <jwolfe>aklein: oh interesting.
21:30:45  <littledan>strange, I thought we ran into that issue when shipping the icu case conversion flag, and changed the test
21:30:56  <littledan>of course, could be another function in the test
21:31:34  <littledan>oh, I see, it's that we asserted that it *did* fail, and that was too strong
21:32:28  * unixpicklequit (Quit: My Mac has gone to sleep. ZZZzzz…)
21:33:32  * unixpicklejoined
21:38:14  <littledan>jwolfe: One crazy option would be to add these builtins to the side-effect free list in src/debug/debug-evaluate.cc. OTOH that wouldn't make the test pass yet, since toLocale{Upper,Lower} case is not a builtin
21:39:45  <jwolfe>littledan: sounds like you understand what that test is doing.
21:39:59  <littledan>I'll upload a patch that fixes the test
21:42:16  <aklein>ah, yeah, the failure on my patch is a different function than the one jwolfe mentioned
21:43:54  * unixpicklequit (Quit: My Mac has gone to sleep. ZZZzzz…)
22:00:30  <aklein>littledan: thanks for hopping on this!
22:04:04  * etnbrdjoined
22:11:35  * Net147joined
22:51:13  * RT|Chatzillajoined
23:12:35  * unixpicklejoined
23:20:02  * Cube8quit (Remote host closed the connection)