tbrowder AlexDaniel: nqp PR #414 fixed, passing all checks 01:01
japhb Fresh build of latest rakudo segfaults during zef package install: 01:14
216379 Segmentation fault (core dumped) zef --/tap-harness --/cached install URI File::Temp p6doc Terminal::ANSIColor Term::termios Debugger::UI::CommandLine JSON::Tiny DBIish Linenoise Inline::Perl5 Inline::Python IRC::Client OO::Monitors Grammar::Debugger App::Mi6 cro
(I've not yet made any attempt to wittle that down, but it came after several recursions of the dependency search, right after this line: 01:15
===> Searching for missing dependencies: Digest
AlexDaniel tbrowder: will check once I have time 01:20
japhb: dang it
Geth roast: d63090c18d | (Zoffix Znet)++ | MISC/bug-coverage-stress.t
Cover segfault in a `for` loop + some ops

Closes github.com/rakudo/rakudo/issues/1483 R#1483		 01:23
synopsebot R#1483 [open]: github.com/rakudo/rakudo/issues/1483 [SEGV][regression][testneeded] Cross-HLL inlining segfault
AlexDaniel japhb: if you have any details please comment on github.com/rakudo/rakudo/issues/1527 01:25
also, what is this? travis-ci.org/rakudo/rakudo/jobs/3...4793#L1600 01:43
Zoffix ZOFFLOP: t/spec/S17-supply/batch.t 03:14
ZOFVM: Files=1289, Tests=153218, 148 wallclock secs (20.16 usr 3.30 sys + 3160.22 cusr 157.94 csys = 3341.62 CPU)
Geth rakudo/post-release: 8b7385d810 | (Zoffix Znet)++ | src/core/Uni.pm
Optimize Uni.list

  - Use custom iterator with .push-all and .[count|bool]-only
  - Makes reifying 15x faster and .count-only 653x faster (on 10000-char str)
  - Speeds up Str.ords		 03:30
roast/post-release: d43b90d70c | (Zoffix Znet)++ | S32-str/ords.t
Cover Uni.list optimization

Covers github.com/rakudo/rakudo/commit/8b7385d810		 03:31
roast/post-release: e2b662a4aa | (Zoffix Znet)++ | S32-str/ords.t
Fix up counter in test description
 03:34
AlexDaniel it would be great if someone can merge this: github.com/sergot/http-useragent/pull/188 10:41
because otherwise I seem to get a lot of modules unsucced in toaster
at least that's my current mid-sleep hypothesis 10:45
Geth rakudo/post-release: ba6b84bd0c | (Zoffix Znet)++ | src/core/Rakudo/Internals.pm
Make Str.perl 43x faster for some chars

Those that give +uniprop($ord,'Canonical_Combining_Class') > 0 in particular.
The change from lower-case hex to upper-case hex is intentional. I think it doesn't matter much. We could also skip the hexafication of ords and simply use the `\c[]` escape, but perhaps decimal character codes are less familiar to humans reading the output?		 11:19
lizmat Files=1235, Tests=76220, 312 wallclock secs (14.77 usr 5.40 sys + 2141.29 cusr 214.86 csys = 2376.32 CPU) 11:26
dogbert2 lizmat: welcome back :-) 11:27
lizmat physically, yes
still recovering from my bad cold :-(
dogbert2 aha 11:28
did you get it during your vacation or when you got back?
|Tux| Rakudo version 2018.01-204-gbeaad7b97 - MoarVM version 2018.01-94-gc7774e37d
csv-ip5xs0.784 - 0.805
csv-ip5xs-207.435 - 7.468
csv-parser11.198 - 11.487
csv-test-xs-200.484 - 0.500
test8.460 - 9.120
test-t2.532 - 2.568
test-t --race1.047 - 1.090
test-t-2045.483 - 45.709
test-t-20 --race15.909 - 16.110
 12:16
timotimo releasable6: status 12:40
releasable6 timotimo, Next release in 1 day and ≈6 hours. Blockers: github.com/rakudo/rakudo/issues?q=...%9A%A0%22. Changelog for this release was not started yet
timotimo, Details: gist.github.com/5c0330f577337462e1...74f6447d09
timotimo i should probably post to perl6-users or something to get feedback on the udp datagram api? 12:41
dogbert2 heh, ===SORRY!=== 15:05
Iteration past end of iterator
i get this when I run: MVM_SPESH_NODELAY=1 MVM_SPESH_BLOCKING=1 ./perl6 t/spec/S32-str/windows-1251-windows-1252-encode-decode.t 15:06
the same goes for t/spec/S11-modules/import.t 15:08
AlexDaniel treegrep: t/080-ua.t 15:55
greppable6 AlexDaniel, gist.github.com/f32bf894124f6e371c...01d5ded205
AlexDaniel treegrep: t\/080-ua.t
greppable6 AlexDaniel, gist.github.com/9d78d03982dda078f4...deecec644c 15:56
AlexDaniel treegrep: t\/080\-ua.t
greppable6 AlexDaniel, gist.github.com/1c2fe0cc86aff460a8...30081ce597
AlexDaniel MasterDuke: ok, hmmm… that file hangs without a fix ↑ 15:57
but I thought I tried HTTP::UserAgent tests and they were green?
if anybody is wondering, here's a summary: a recent change to HTTP::UserAgent makes it go into an infinite loop. There are many modules depending on that module, so not only they fail in toaster, but collectively they bog down the cpu when toaster is running… 16:06
dogbert2 AlexDaniel: are your bots still suffering from SEGV's? 16:07
AlexDaniel dogbert2: yes
dogbert2 have you applied brrt's fixes? 16:08
AlexDaniel yes
dogbert2 :(
found some new errors above if you're interested
AlexDaniel dogbert2: it's this, by the way: github.com/rakudo/rakudo/issues/1259 16:09
not going to call that a blocker obviously because it was spotted in November and I'm the only person complaining :) 16:10
dogbert2 :) 16:14
AlexDaniel notable6: release 18:02
notable6 AlexDaniel, No notes for “release”
AlexDaniel fwiw I know about at least one potential blocker, golfing it right now 21:56
(and meanwhile toaster is crunching its stuff with proper HTTP::UserAgent this time) 21:57
dogbert17 AlexDaniel: is it possible to run bisectbot with e.g. MVM_SPESH_DISABLE env var set? 22:03
AlexDaniel sure
dogbert2: I am doing it right now, but… do you have access to the server?
if not please give me your ssh key and you'll be able to play with it yourself 22:04
dogbert17 like, bisect MVM_SPESH_DISABLE=1 t/spec/S32-str/windows-1251-windows-1252-encode-decode.t
AlexDaniel aaaaaaaaaaaaaaaaaaaah you're talking about this…
dogbert17 and I meant MVM_SPESH_NODELAY=1 :-)
AlexDaniel c: MVM_SPESH_NODELAY=1 HEAD say %*ENV<MVM_SPESH_NODELAY> 22:05
committable6 AlexDaniel, ¦HEAD(beaad7b): «1␤»
AlexDaniel IIRC bisectable does not support it, but
I have a trick up my sleeve for that too.
dogbert17 I think there's a regression hiding here 22:06
AlexDaniel c: HEAD run ‘perl6’, ‘sandbox/roast/S32-str/windows-1251-windows-1252-encode-decode.t’ 22:07
committable6 AlexDaniel, ¦HEAD(beaad7b): «1..4␤ok 1 - Test encoding from Unicode to windows-1252␤ok 2 - Test decoding from windows-1252 to Unicode␤ok 3 - Test encoding from Unicode to windows-1251␤ok 4 - Test decoding from windows-1251 to Unicode␤»
AlexDaniel c: HEAD run ENV=>%(MVM_SPESH_DISABLE=>1), ‘perl6’, ‘sandbox/roast/S32-str/windows-1251-windows-1252-encode-decode.t’
committable6 AlexDaniel, ¦HEAD(beaad7b): «Unexpected named argument 'ENV' passed␤ in block <unit> at /tmp/LP9DS4ZRPQ line 1␤␤ «exit code = 1»»
AlexDaniel c: HEAD run env=>%(MVM_SPESH_DISABLE=>1), ‘perl6’, ‘sandbox/roast/S32-str/windows-1251-windows-1252-encode-decode.t’
committable6 AlexDaniel, ¦HEAD(beaad7b): «The spawned command 'perl6' exited unsuccessfully (exit code: 1)␤ in block <unit> at /tmp/B42EdwQPCV line 1␤␤ «exit code = 1»»
dogbert17 MVM_SPESH_NODELAY=1 !!!
AlexDaniel c: HEAD run ENV=>%(%*ENV, MVM_SPESH_NODELAY=>1), ‘perl6’, ‘sandbox/roast/S32-str/windows-1251-windows-1252-encode-decode.t’ 22:08
committable6 AlexDaniel, ¦HEAD(beaad7b): «Unexpected named argument 'ENV' passed␤ in block <unit> at /tmp/TQ8WYP1ulT line 1␤␤ «exit code = 1»»
AlexDaniel c: HEAD run env=>%(%*ENV, MVM_SPESH_NODELAY=>1), ‘perl6’, ‘sandbox/roast/S32-str/windows-1251-windows-1252-encode-decode.t’
committable6 AlexDaniel, ¦HEAD(beaad7b): «===SORRY!===␤Iteration past end of iterator␤The spawned command 'perl6' exited unsuccessfully (exit code: 1)␤ in block <unit> at /tmp/w5q0vV_29w line 1␤␤ «exit code = 1»»
dogbert17 there is the error
AlexDaniel c: 6c run env=>%(%*ENV, MVM_SPESH_NODELAY=>1), ‘perl6’, ‘sandbox/roast/S32-str/windows-1251-windows-1252-encode-decode.t’
committable6 AlexDaniel, gist.github.com/9072719b4b955519cf...82d78c564b 22:09
AlexDaniel dogbert17: now what?
dogbert17: we can bisect this change, but “Error encoding Windows-1252 string: could not encode codepoint 8250” does not look like the right result? 22:10
dogbert17 try this instead t/spec/MISC/bug-coverage.t 22:11
several files failed with the same message 22:12
AlexDaniel btw please ignore this nonsense about writability of precomp files
c: 2018.01,HEAD run env=>%(%*ENV, MVM_SPESH_NODELAY=>1), ‘perl6’, ‘sandbox/roast/MISC/bug-coverage.t’
committable6 AlexDaniel, gist.github.com/040675d3281fecf3a9...763bd3a736
dogbert17 argh’ 22:13
AlexDaniel it's ok, one sec…
dogbert17 no Test::Util
AlexDaniel c: 2018.01,HEAD run env=>%(%*ENV, PERL6LIB=>‘sandbox/roast/packages’, MVM_SPESH_NODELAY=>1), ‘perl6’, ‘sandbox/roast/MISC/bug-coverage.t’
committable6 AlexDaniel, gist.github.com/75de5004739cd4a679...0e98dcf560
AlexDaniel that's ok too 22:14
c: 2018.01,HEAD chdir ‘sandbox/roast’; run env=>%(%*ENV, PERL6LIB=>‘packages’, MVM_SPESH_NODELAY=>1), ‘perl6’, ‘MISC/bug-coverage.t’
committable6 AlexDaniel, gist.github.com/d4d4e8ce8f06c95ded...bb2174e899
AlexDaniel there
c: 6c chdir ‘sandbox/roast’; run env=>%(%*ENV, PERL6LIB=>‘packages’, MVM_SPESH_NODELAY=>1), ‘perl6’, ‘MISC/bug-coverage.t’
dogbert17 i see the error again
AlexDaniel “Unhandled kind 1” isn't that much better :) 22:15
dogbert17 nope
AlexDaniel let's bisect anyway, maybe that will give some clue…
bisect: old=2018.01 chdir ‘sandbox/roast’; run env=>%(%*ENV, PERL6LIB=>‘packages’, MVM_SPESH_NODELAY=>1), ‘perl6’, ‘MISC/bug-coverage.t’
bisectable6 AlexDaniel, Bisecting by output (old=2018.01 new=beaad7b) because on both starting points the exit code is 1 22:16
committable6 AlexDaniel, gist.github.com/1c700696e32be7b7c5...cfd15c76dc
bisectable6 AlexDaniel, bisect log: gist.github.com/9c8efff87c539933b6...8af0147536
AlexDaniel, (2018-01-25) github.com/rakudo/rakudo/commit/d1...d451fd0fb6
AlexDaniel dogbert17: is there a ticket already? 22:17
dogbert17 no, let me write one ...
AlexDaniel c: d1a48efc97^,d1a48efc97 chdir ‘sandbox/roast’; run env=>%(%*ENV, PERL6LIB=>‘packages’, MVM_SPESH_NODELAY=>1), ‘perl6’, ‘MISC/bug-coverage.t’
committable6 AlexDaniel, gist.github.com/71209dcc1127a1519d...b3083970a1 22:18
AlexDaniel oh, that looks different?
c: d1a48efc97^,d1a48efc97 chdir ‘sandbox/roast’; run env=>%(%*ENV, PERL6LIB=>‘packages’, MVM_SPESH_NODELAY=>1), ‘perl6’, ‘S32-str/windows-1251-windows-1252-encode-decode.t’ 22:19
committable6 AlexDaniel, gist.github.com/ec8a15b2ccc7ebe8b8...3be44dfaf8
AlexDaniel bisect: old=2018.01 chdir ‘sandbox/roast’; run env=>%(%*ENV, PERL6LIB=>‘packages’, MVM_SPESH_NODELAY=>1), ‘perl6’, ‘S32-str/windows-1251-windows-1252-encode-decode.t’
bisectable6 AlexDaniel, Bisecting by output (old=2018.01 new=beaad7b) because on both starting points the exit code is 1
AlexDaniel, bisect log: gist.github.com/48deadf085d67157c6...bd34fc2e85 22:20
AlexDaniel, (2018-02-01) github.com/rakudo/rakudo/commit/45...3ce90a1849
AlexDaniel c: 4507a5654108^,4507a5654108 chdir ‘sandbox/roast’; run env=>%(%*ENV, PERL6LIB=>‘packages’, MVM_SPESH_NODELAY=>1), ‘perl6’, ‘S32-str/windows-1251-windows-1252-encode-decode.t’
committable6 AlexDaniel, gist.github.com/83ccbcbf6d40bb084b...a2e679adcd
AlexDaniel c: 4507a5654108^,4507a5654108,HEAD chdir ‘sandbox/roast’; run env=>%(%*ENV, PERL6LIB=>‘packages’, MVM_SPESH_NODELAY=>1), ‘perl6’, ‘S32-str/windows-1251-windows-1252-encode-decode.t’
committable6 AlexDaniel, gist.github.com/8c0b6c4b0d4dbc31ba...e22631915f
AlexDaniel dogbert17: are you following? :) 22:21
dogbert17 AlexDaniel: github.com/MoarVM/MoarVM/issues/804
AlexDaniel look at the last link
bisect: old=4507a5654108 chdir ‘sandbox/roast’; run env=>%(%*ENV, PERL6LIB=>‘packages’, MVM_SPESH_NODELAY=>1), ‘perl6’, ‘S32-str/windows-1251-windows-1252-encode-decode.t’
bisectable6 AlexDaniel, Bisecting by exit code (old=4507a56 new=beaad7b). Old exit code: 0
AlexDaniel, bisect log: gist.github.com/05b65bad784f3c48f4...21ae0fdaa0
AlexDaniel, (2018-02-16) github.com/rakudo/rakudo/commit/be...1e6372d9ae
AlexDaniel oh that's bad 22:22
that's HEAD :D
c: beaad7b979f^,beaad7b979f chdir ‘sandbox/roast’; run env=>%(%*ENV, PERL6LIB=>‘packages’, MVM_SPESH_NODELAY=>1), ‘perl6’, ‘S32-str/windows-1251-windows-1252-encode-decode.t’
committable6 AlexDaniel, gist.github.com/ec3e52f2c72cad3243...951768b2b3
dogbert17 uh oh
AlexDaniel OK! THERE
finally
lol all that pain just to realize that the very HEAD commit is at fault :D 22:23
dogbert17 lol
Geth rakudo/post-release: 6ac56cc081 | (Zoffix Znet)++ | 2 files
Optimize Str.perl and uniprop(Int, Str)

  - Make uniprop(Int, Str) 2.7x faster
   (measured with uniprop(0x308, "Canonical_Combining_Class")
  - It's used by Str.perl for chars with ords above 256, so for
   such Strs, .perl comes out 20% faster
   (measured with unroped `"\c[3000]" x 1000`)
dogbert17 there are a lot of commits in that merge
could it be the SEGV fix? 22:25
AlexDaniel just one commit actually, I think. Yes: github.com/MoarVM/MoarVM/commit/c7...7f5369f982
it could be something else but I doubt it 22:26
dogbert17 oh well :-)
AlexDaniel dogbert17: I filed RT#1534 also 22:29
omg
dogbert17 cool
AlexDaniel GH#1534 I mean
synopsebot GH#1534 [open]: github.com/rakudo/rakudo/issues/1534 [⚠ blocker ⚠] “Iteration past end of iterator” issue
dogbert17 a blocker, that's a shocker :-)
AlexDaniel c: beaad7b979f^,beaad7b979f chdir ‘sandbox/roast’; run env=>%(%*ENV, PERL6LIB=>‘packages’, MVM_SPESH_NODELAY=>1), ‘perl6’, ‘MISC/bug-coverage.t’ 22:31
committable6 AlexDaniel, gist.github.com/9b94475c34efa52cbc...8fb95f3b5f
Zoffix So what constitutes a "regression"? I would've thought it's anything that used to work and then got broken in some release. But looks like only "behaviour changed" qualifier is used instead, which probably covers 90% of bugs 22:32
AlexDaniel Zoffix: I tend to put “regression” label on “behaviour changed” issues when I simply don't know if the current behavior is broken or not 22:34
assuming that someone with more information can remove the label if I'm wrong 22:36
Zoffix: is that wrong? 22:37
.oO( can we eliminate 90% of bug reports if we stop committing? )
 22:38
Zoffix Not wrong. It's just interefering with my idea of me being perfect :) 22:39
ZofBot: a Zoffix never regresses! 22:40
ZofBot Zoffix, 1
Zoffix ZofBot++
AlexDaniel pfffffffffffffft 23:00
Zoffix :)
AlexDaniel got a 100% reproducible segfault when trying to golf something different 23:01
oh… but it could be my fault 23:02
no, does not look like it
Zoffix releasable6: status 23:21
releasable6 Zoffix, Next release in ≈19 hours. Blockers: github.com/rakudo/rakudo/issues?q=...%9A%A0%22. Changelog for this release was not started yet
Zoffix, Details: gist.github.com/6da93f66bee8d9fb66...5f8f5d1734
AlexDaniel eh releasable6, you should learn about github.com/rakudo/rakudo/wiki/ChangeLog-Draft 23:22
it's not finished but saying that it was not started at all is a bit rude :) 23:23
nine: any thoughts? github.com/rakudo/rakudo/issues/1535 23:33
ooooooooooooh 23:37
I see now…
maybe
I think implicit return is so much faster that the issue goes away
no. 23:40
nevermind what I just said

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!