Camelia, the Perl 6 bug

IRC log for #parrot, 2012-03-20

Parrot | source cross referenced

| Channels | #parrot index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
01:06 ilbot2 joined #parrot
01:06 Topic for #parrot is now Parrot 4.1.0 "Black-headed Parrot" | http://parrot.org | Log: http://irclog.perlgeek.de/parrot | #parrotsketch meeting Tuesday 19:30 UTC
01:34 contingencyplan joined #parrot
04:11 alvis joined #parrot
04:19 dalek parrot: 228daeb | dukeleto++ | ChangeLog:
04:19 dalek parrot: [doc] Add some useful links to the ChangeLog
04:19 dalek parrot: review: https://github.com/parrot/parrot/commit/228daeb3b6
04:29 alvis left #parrot
04:33 travis-ci joined #parrot
04:33 travis-ci [travis-ci] parrot/parrot#177 (master - 228daeb : Jonathan "Duke" Leto): The build is still failing.
04:33 travis-ci [travis-ci] Change view : https://github.com/parrot/par​rot/compare/cd06be8...228daeb
04:33 travis-ci [travis-ci] Build details : http://travis-ci.org/parrot/parrot/builds/904322
04:33 travis-ci left #parrot
05:39 Khisanth joined #parrot
06:09 mdupont joined #parrot
07:25 fperrad joined #parrot
08:29 preflex_ joined #parrot
08:47 mj41 joined #parrot
09:15 lucian joined #parrot
10:16 dalek rakudo/nom: d76065e | moritz++ | src/core/Exception.pm:
10:16 dalek rakudo/nom: fix typo
10:16 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/d76065e777
10:30 jsut joined #parrot
11:47 alvis joined #parrot
12:00 benabik joined #parrot
12:30 benabik Morning, #parrot
12:42 bluescreen joined #parrot
13:51 whiteknight joined #parrot
13:54 hercynium joined #parrot
14:18 whiteknight good morning, #parrot
14:23 benabik o/ whiteknight
14:23 whiteknight hello benabik, how are you doing today?
14:25 benabik whiteknight: meh.  Travelling over the weekend has apparently thrown me way off.  Missed a class yesterday and missed a lot of sleep last night.
14:30 whiteknight ouch. Missing classes is the worst
14:31 dalek rakudo/nom: 704a215 | moritz++ | tools/build/Makefile.in:
14:31 dalek rakudo/nom: try to fix parrot revision reporting for smolder submissions. lumi++
14:31 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/704a215c83
14:34 JimmyZ joined #parrot
14:53 dmalcolm joined #parrot
14:59 PacoAir joined #parrot
15:20 bluescreen joined #parrot
15:45 dukeleto joined #parrot
15:45 dukeleto ~~
15:45 dukeleto blarg.
15:45 dukeleto finding bugs in the release process...
15:46 Psyche^ joined #parrot
15:47 aloha (parrot/parrot) Issues opened : 745 (sh tools/dev/mk_native_pbc fails) by leto : https://github.com/parrot/parrot/issues/745
15:47 dukeleto anybody have tuits on that?
15:49 dalek parrot: c743042 | dukeleto++ | / (10 files):
15:49 dalek parrot: Run perl tools/release/auto_release.pl 4.2.0
15:49 dalek parrot: review: https://github.com/parrot/parrot/commit/c743042572
15:49 dalek parrot: 671468c | dukeleto++ | t/native_pbc/ (6 files):
15:49 dalek parrot: Add some native pbc
15:49 dalek parrot: review: https://github.com/parrot/parrot/commit/671468cdb2
15:51 dukeleto hmm. maybe i need a functional parrot to run that command...
16:07 travis-ci joined #parrot
16:07 travis-ci [travis-ci] parrot/parrot#178 (master - 671468c : Jonathan "Duke" Leto): The build is still failing.
16:07 travis-ci [travis-ci] Change view : https://github.com/parrot/par​rot/compare/228daeb...671468c
16:07 travis-ci [travis-ci] Build details : http://travis-ci.org/parrot/parrot/builds/907485
16:07 travis-ci left #parrot
16:07 whiteknight dukeleto: what kinds of bugs?
16:10 dukeleto whiteknight: making native pbc
16:10 dukeleto whiteknight: i get an ffi error, in the open ticket https://github.com/parrot/parrot/issues/745
16:11 whiteknight blah
16:11 dukeleto lots of interesting warnings on an optimized clang: http://travis-ci.org/#!/parrot/parrot/jobs/907486
16:11 dukeleto questionable casting in our GC
16:11 dukeleto src/gc/alloc_resources.c:722:20: warning: cast from 'char *' to 'PMC *'
16:12 dukeleto so we have t/pmc/big(num|int).t and t/op/64bit.t failing on an optimized clang build on x64 Ubuntu, FYI
16:13 dukeleto which is what Travis is complaining about. Travis is complaining for the right reason!
16:14 dukeleto i pontificate some of our casts are invalid on an optimized x64 clang
16:16 dukeleto whiteknight: my numvalsize is 16
16:16 dukeleto whiteknight: which is unhandled. I think I can fix tis.
16:16 dukeleto this.
16:18 dukeleto whiteknight: and my longdoublesize is 16. This points out fantasies that our codebase thought were true.
16:19 whiteknight long double is 16?
16:19 dukeleto whiteknight:     nv=long double, numvalsize=16, doublesize=8, longdoublesize=16
16:19 dukeleto whiteknight: i am trying a quick fix now
16:21 whiteknight okay
16:21 whiteknight I can't do much testing to help
16:21 whiteknight at least, not until I get home tonight
16:22 contingencyplan joined #parrot
16:28 dukeleto whiteknight: well, i got it building
16:28 dukeleto whiteknight: being release manager is never boring :)
16:28 whiteknight no it is not
16:29 whiteknight it has gotten a heck of a lot more automated in recent months
16:29 dukeleto whiteknight: yes, autorelease.pl helps a bunch
16:29 dukeleto whiteknight: which needs to be documented....
16:29 dukeleto choosing which yak holes to dive into is always the problem...
16:29 whiteknight I started putting together a rosella-based script that does what autorelease.pl did and more, but I haven't finished it yet
16:30 dukeleto whiteknight: autorelease never finished correctly for me, though
16:30 dukeleto whiteknight: it always dies in the pbc stage, but it seems our pbc stuff is fraught with dynamite and chainsaws
16:33 dalek parrot: f7e1647 | dukeleto++ | src/nci/libffi.c:
16:33 dalek parrot: [ffi] Get tools/dev/mk_native_pbc working again by adding logic for when numvalsize=16; LTA
16:33 dalek parrot: review: https://github.com/parrot/parrot/commit/f7e164718c
16:34 whiteknight fun
16:37 dukeleto whiteknight: this begs the question of running pbc stuff on every commit
16:37 dukeleto whiteknight: because pbc stuff is only ran by release managers once a month, it gets bitrotty at the worst times
16:38 dukeleto whiteknight: i.e. i can make travis test it for us
16:39 whiteknight what pbc stuff?
16:43 dukeleto whiteknight: tools/dev/mk*pbc
16:43 dukeleto whiteknight: two scripts which need to recompile parrot to generate pbc
16:43 whiteknight hmmm
16:43 dukeleto whiteknight: and which have almost always failed for me
16:43 dukeleto whiteknight: for one reason or another
16:53 dalek parrot: 9522930 | dukeleto++ | docs/project/release_manager_guide.pod:
16:53 dalek parrot: [doc] Fix line-length failures
16:53 dalek parrot: review: https://github.com/parrot/parrot/commit/9522930869
16:53 travis-ci joined #parrot
16:53 travis-ci [travis-ci] parrot/parrot#179 (master - f7e1647 : Jonathan "Duke" Leto): The build is still failing.
16:53 travis-ci [travis-ci] Change view : https://github.com/parrot/par​rot/compare/671468c...f7e1647
16:53 travis-ci [travis-ci] Build details : http://travis-ci.org/parrot/parrot/builds/907843
16:53 travis-ci left #parrot
16:53 dukeleto blarg
16:53 dukeleto now t/dynoplibs/sysinfo.t is failing because of my nci/ffi hack
16:55 whiteknight heh
16:57 moritz all rakudo spectests pass on parrot master
16:57 moritz parrot RELEASE_4_1_0-105-gd923903 that is
17:04 whiteknight nice
17:12 travis-ci joined #parrot
17:12 travis-ci [travis-ci] parrot/parrot#180 (master - 9522930 : Jonathan "Duke" Leto): The build is still failing.
17:12 travis-ci [travis-ci] Change view : https://github.com/parrot/par​rot/compare/f7e1647...9522930
17:12 travis-ci [travis-ci] Build details : http://travis-ci.org/parrot/parrot/builds/908039
17:12 travis-ci left #parrot
17:13 jsut_ joined #parrot
17:14 davidfetter joined #parrot
17:23 jsut joined #parrot
18:08 whiteknight when is #ps today? I can't keep track of it when daylight savings isn't crapping up my schedule
18:08 benabik My calendar says 3:30 EDT
18:10 whiteknight 3:30? I can do that
18:14 mj41 joined #parrot
18:28 jjore joined #parrot
18:58 bluescreen joined #parrot
19:16 dalek parrot: 0e75eb6 | jkeenan++ | MANIFEST:
19:16 dalek parrot: Rerun mk_manifest_and_skip.pl to pick up one file.
19:16 dalek parrot: review: https://github.com/parrot/parrot/commit/0e75eb61d5
19:26 kid51 joined #parrot
19:31 benabik #ps in now?
19:31 Util now
19:31 travis-ci joined #parrot
19:31 travis-ci [travis-ci] parrot/parrot#181 (master - 0e75eb6 : jkeenan): The build is still failing.
19:31 travis-ci [travis-ci] Change view : https://github.com/parrot/par​rot/compare/9522930...0e75eb6
19:31 travis-ci [travis-ci] Build details : http://travis-ci.org/parrot/parrot/builds/908929
19:31 travis-ci left #parrot
20:30 nn joined #parrot
20:35 sjn left #parrot
20:57 nine msg whiteknight if there were some way to tell the GC to set a flag on any newly created PMCs, this would be enough to decide where a PMC is coming from. Like: Parrot_gc_flag_new_pmcs_on(interp); result = VTABLE_foo(interp, PARROT_PROXY(SELF)->target); return PMC_IS_FLAGGED(result) ? result : Parrot_thread_create_proxy(..., result);
20:57 aloha OK. I'll deliver the message.
20:58 nine msg whiteknight I don't have to implement this right now, but it may be a way to trim back the costs in the future. I'll just go on assuming that there is some way to check the origins of a PMC.
20:58 aloha OK. I'll deliver the message.
21:31 dukeleto ~~
21:37 jsut_ joined #parrot
22:02 hercynium left #parrot
22:35 whiteknight joined #parrot
22:41 perlite joined #parrot
22:42 whiteknight good afternoon, #parrot

| Channels | #parrot index | Today | | Search | Google Search | Plain-Text | summary

Parrot | source cross referenced