Parrot 4.10.0 "[Red-eared Parakeet]" parrot.org/ | Log: irclog.perlgeek.de/parrot
Set by moderator on 21 November 2012.
00:11 Reini joined 00:25 kid51 joined
dalek p/gh67-threads: 2032137 | rurban++ | src/pmc/nqplexpad.pmc:
mark enum_class_Proxy sections with PARROT_THREAD_H_GUARD

Provide an experimental fallback "static int enum_class_Proxy = 0" for non-threaded code.
00:35
00:50 Reini_ joined 01:11 woosley joined 01:12 Mike-PerlRecruiter_ joined 01:56 Reini joined 02:43 aloha joined 03:51 dngor joined 04:00 dngor_ joined 04:04 dngor joined 04:32 dngor_ joined 04:36 Psyche^ joined 04:42 dngor joined 05:14 Reini joined 08:58 Khisanth joined 09:26 bouncy joined, Patterner joined 09:27 Mike-PerlRecruiter_ joined, perlite joined, l3l1p joined, sivoais joined, pjcj joined, arnsholt joined, Tene joined, sorear joined, janus joined, elmex joined, eternaleye joined, moritz joined, Infinoid joined, he_ joined, dukeleto joined, cotto joined, TonyC joined, sri joined, p6eval joined, simcop2387 joined, Maddingue joined, Timbus joined, tokuhirom joined, smash joined, patspam joined, allison joined, cxreg joined 10:00 woosley left
dalek p: 39ab731 | moritz++ | src/pmc/ (2 files):
move UNUSED macro call after declarations

makes -Werror=declaration-after-statement happy
11:55
p: 124bf55 | moritz++ | src/QRegex/P6Regex/Grammar.nqp:
improve Null pattern detection a bit
12:01
13:15 Mike-PerlRecruiter_ joined 13:23 PacoAir joined 13:35 PacoAir joined 14:15 mtk joined 14:21 Reini joined 15:26 davidfetter joined 15:28 Reini joined 15:44 brrt joined 15:46 bluescreen joined
compiler I'll be merging threads and rurban/gh870-pmc_class now. Both are stable 15:51
Coke ooh, sentient compiler! 15:52
compiler nqp and rakudo both work fine with threads, just code which will use threads/Task/Proxy needs to be changed.
rurban/gh870-pmc_class makes it easier on the C-level to check for dynpmc class types 15:53
and it sanifies the dynpmc_group API and headers
uh, what's my name now? I'm rurban 15:54
rurban wonder who changed my nick. it was not me. 15:55
16:08 dmalcolm joined 16:24 alester joined 16:35 brrt left
dalek rrot: 1888b30 | rurban++ | lib/Parrot/Test/Pod.pm:
fix t/codingstd/pod_todo.t

docs/pdds/|docs/dev/pmc are not Pod::Simple parsable. They fail wrapping overlong links. workaround other Pod::Simple parser failures and detect the other TODO/FIXME/XXX in pods.
17:06
rrot: a9b1ada | rurban++ | docs/pdds/pdd25_concurrency.pod:
[pdd] Update the pdd25_concurrency.pod links

Google uses now a shorter format. Added Stefan Seiferts announcement and Master Thesis about our new Hybrid Threads
rrot/rurban/gh870-pmc_class: 230342a | fperrad++ | tools/dev/mk_inno_language.pl:
[win32/inno] follow nqp
rrot/rurban/gh870-pmc_class: 2e66846 | fperrad++ | tools/dev/fetch_languages.pl:
[languages] skip update when missing directory
rrot/rurban/gh870-pmc_class: f69accc | rurban++ | tools/dev/ (2 files):
Merge remote-tracking branch 'origin/master' into rurban/gh870-pmc_class
rurban Proposed threads Changelog line: 17:12
parrot supports now fast and lightweight OS threads, based on Nat Tucks's initial work
on lightweight green threads and then Stefan Seifert's extension to OS threads as
hybrid threads. 17:13
See niner.name/Hybrid_Threads_for_the_Parrot_VM.pdf
I guess we have to fix docs/pdds/pdd25_concurrency.pod to add our current hybrid approach 17:14
dalek rrot/threads: 36ae677 | rurban++ | lib/Parrot/Test/Pod.pm:
fix t/codingstd/pod_todo.t

docs/pdds/|docs/dev/pmc are not Pod::Simple parsable. They fail wrapping overlong links. workaround other Pod::Simple parser failures and detect the other TODO/FIXME/XXX in pods.
17:19
rrot/threads: 0396004 | rurban++ | docs/pdds/pdd25_concurrency.pod:
[pdd] Update the pdd25_concurrency.pod links

Google uses now a shorter format. Added Stefan Seiferts announcement and Master Thesis about our new Hybrid Threads
rrot/threads: 0bc6d9f | rurban++ | / (2 files):
[pdd] add GSoC Threads: Chandons Results

green threads implemented by Nat Tuck URL
rrot/threads: f7f6ca4 | rurban++ | / (2 files):
[doc] Document --numthreads and add ChangeLog entry
17:32
17:48 schm00ster joined
dalek rrot/threads: 5190ab1 | rurban++ | docs/pdds/pdd25_concurrency.pod:
[codingstd] fix pdd_format.t formatting
17:54
17:58 wagle joined
dalek rrot: 7c8eb67 | rurban++ | src/pmc/schedulermessage.pmc:
[doc] Better schedulermessage title

Make schedulermessage distinguishable from scheduler at
  docs.parrot.org/parrot/latest/html/pmc.html.
18:06
rrot: af53504 | rurban++ | MANIFEST (2 files):
[install] add myconfig parrot_nci_thunk_gen parrotbug to [devel]
rrot: c5ac3f1 | rurban++ | / (2 files):
Merge branch 'master' of github.com:parrot/parrot
18:39 tuxit joined
tadzik is it now? Is it merged? 18:55
awesome! 18:56
rurban not yet fully. Still some tiny bits missing 18:58
In 1-2 hrs 18:59
I had to bump PBC_COMPAT and am now regenerating the native_pbc's
linux, darwin and powerpc look fine 19:00
dalek rrot/threads: ca52886 | rurban++ | / (14 files):
PBC_COMPAT bump to 13.0 for threads and Proxy, updated native_pbcs
19:13
rurban threads are now merged 19:17
tadzik hooray
rurban++
thank you
19:18 davidfetter joined
rurban long live hybrid os + green threads, Thanks to Nat Tuck and Stefan Seifert. They did all the work 19:18
dalek Heuristic branch merge: pushed 257 commits to parrot by rurban
cotto nine++ 19:21
dalek rrot: 0c01639 | rurban++ | tools/dev/pbc_to_exe.winxed:
pbc_to_exe: do not link -s as --strip/-s is ld specific, deprecated and breaks debugging

strip should rather be done later at installation.
19:30
rrot/rurban/gh870-pmc_class: aed7e71 | rurban++ | lib/Parrot/Pmc2c/ (3 files):
[GH #870] Export INTVAL pmc_class_Classname types for dynpmc

Each pmc2c library header or single dynpmc header contains now the declaration of a global type, initialized dynamically.
rrot/rurban/gh870-pmc_class: 3e89695 | rurban++ | lib/Parrot/Pmc2c/ (2 files):
[GH #870] fix c->h confusion

Emit H now into the header actually, not the c
rrot/rurban/gh870-pmc_class: b099203 | rurban++ | lib/Parrot/Pmc2c/PMC.pm:
[GH #870] seperate extern int from local int.

This works now okay
rrot/rurban/gh870-pmc_class: 9bf3cac | rurban++ | lib/Parrot/Pmc2c/ (3 files):
[GH #870] rename pmc_class_ to dynpmc_class_
rrot/rurban/gh870-pmc_class: 8a5771a | rurban++ | lib/Parrot/Pmc2c/ (3 files):
[GH #870] refactor exports and headers
rrot/rurban/gh870-pmc_class: d71d0bc | rurban++ | lib/Parrot/Pmc2c/ (2 files):
[GH #870] library loaders only load Parrot_lib_$name_load

darwin detected a problem with multiple exports of dynpmc_class_Foo symbols. We cannot provide duplicate object code to load dynpmc groups. The group calls the seperate Parrot_lib_$name_load funcs, which by them export the dynpmc_class_$name symbols.
rrot/rurban/gh870-pmc_class: fe13fd4 | rurban++ | lib/Parrot/Pmc2c/ (2 files):
[GH #870] sort the pmcs, lc Parrot_lib_*_load name
rrot/rurban/gh870-pmc_class: f45bcf7 | rurban++ | lib/Parrot/Pmc2c/PMC.pm:
[GH #870][codingstd] hard tab Pmc2c/PMC.pm line 587
19:52 travis-ci joined
travis-ci [travis-ci] parrot/parrot#647 (threads - 0bc6d9f : Reini Urban): The build was broken. 19:52
[travis-ci] Change view : github.com/parrot/parrot/compare/2...c6d9ff94a1
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/3480037
19:52 travis-ci left
dalek rrot: 9c32ec0 | rurban++ | tools/dev/install_dev_files.pl:
[install] install parrotbug [devel]bin asis
19:59
rurban The travis-ci breakage was expected and is fixed later with build #649 20:00
dalek rrot: b09c161 | rurban++ | .travis.yml:
travis should not smoke branch threads anymore
20:02
rrot/rurban/gh870-pmc_class: 04d2ffc | rurban++ | lib/Parrot/Pmc2c/ (3 files):
[GH #870] Export INTVAL dynpmc_class_<ClassName> types for each dynpmc

Each pmc2c library header or single dynpmc header contains now the declaration of a global type, initialized dynamically.
library loaders only load Parrot_lib_$name_load: darwin detected a problem with multiple exports of dynpmc_class_Foo symbols. We cannot provide duplicate object code to load dynpmc groups. The group calls the seperate Parrot_lib_$name_load funcs, which by them export the dynpmc_class_$name symbols. Sort the grouped dynpmc's by lc Parrot_lib_*_load name
20:17
rrot: 04d2ffc | rurban++ | lib/Parrot/Pmc2c/ (3 files):
[GH #870] Export INTVAL dynpmc_class_<ClassName> types for each dynpmc

Each pmc2c library header or single dynpmc header contains now the declaration of a global type, initialized dynamically.
library loaders only load Parrot_lib_$name_load: darwin detected a problem with multiple exports of dynpmc_class_Foo symbols. We cannot provide duplicate object code to load dynpmc groups. The group calls the seperate Parrot_lib_$name_load funcs, which by them export the dynpmc_class_$name symbols. Sort the grouped dynpmc's by lc Parrot_lib_*_load name
20:20
rrot: 4fa3d40 | rurban++ | ChangeLog:
ChangeLog entry for rurban/gh870-pmc_class
rrot: 9e2d21d | rurban++ | / (4 files):
Merge branch 'rurban/gh870-pmc_class'
rrot/rurban/gh870-pmc_class: 4fa3d40 | rurban++ | ChangeLog:
ChangeLog entry for rurban/gh870-pmc_class
rurban dukeleto: I'm adding now your t/pmc/task-nqp.t as ext/nqp-rx/t/nqp/43-task.t 20:33
nope, as ext/nqp-rx/t/nqp/67-threads.t 20:34
dukeleto rurban: works for me 20:47
rurban++ # merging threads
rurban I'm enabling now the parrot-nqp tests. They were disabled and broken for a while 20:52
dalek rrot: a6111e4 | rurban++ | ext/nqp-rx/t/nqp/67-threads.t:
test parrot-nqp thread safety with lexpads (nqp #67)

test_more is not yet thread-safe
20:53
rrot: 21519b4 | rurban++ | ext/nqp-rx/t/nqp/67-threads.t:
[test] fix ext/nqp-rx/t/nqp/67-threads.t thread-safety

do not use test_more.pir yet
rrot: 11492d6 | rurban++ | ext/nqp-rx/t/nqp/ (42 files):
[test] fix ext/nqp-rx/t/nqp/ codas

nqp tests pass now all
rurban Next would be fixing test+more.pir to be thread-safe
tadzik oh oh
does my testcase now work?
rurban tadzik: no, not yet 20:54
that's now a nqp branch: gh67-threads
but the parrot-nqp counterpart works ok
I justed added it
tadzik yeah, parrot-nqp did work for some time, iirc 20:55
rurban dukeleto: you can now remove your smoke-me/threads_tests branch
dukeleto rurban: hokey dokey 20:56
rurban perl t/harness ext/nqp-rx/t/nqp/*.t passes now all tests
dukeleto rurban: done 20:57
i am interested to see how the runtime of our travis tests changes post-threads-merge: travis-ci.org/parrot/parrot/builds 20:58
rurban the last two tests are somewhat broken on travis. e.g. 648.6 fails for the pdd_format tests, ok. But 648.7 is still running 21:15
It were probably too many commits for travis, so it works down the queue. 21:16
I want to add EXTRA_TEST_ARGS += ext/nqp-rx/t/nqp/*.t in ext/nqp-rx/Defines.mak 21:25
Nope. I'd rarher add nqp tests to the compilers testset 21:31
dalek rrot: 61c3d66 | rurban++ | ext/nqp-rx/t/p6regex/ (12 files):
[tests] remove duplicate ext/nqp-rx/t/p6regex tests

These tests do not work anymore (bitrot), and are a duplicate of the tests at t/compilers/pge/perl6regex/
21:40
rrot: 4f78e5c | rurban++ | lib/Parrot/Harness/TestSets.pm:
[test] add ext/nqp-rx/t/nqp/*t to the compilers tests

These tests were left out since nqp-rx was moved to ext
rrot: ed84556 | rurban++ | ChangeLog:
[test] ChangeLog lines for nqp-rx test changes
21:42
kudo/stdier: 6b01881 | jnthn++ | src/ (2 files):
Start showing context, with eject marker.
21:44
kudo/stdier: a109e73 | jnthn++ | src/Perl6/Grammar.pm:
Make panic always throw X::Comp::AdHoc.

This also means all the panics get the location included in the error.
rrot: dc58001 | rurban++ | t/codingstd/filenames.t:
[codingstd] TODO a failing filenames.t test for windows

Fails for Windows installable_parrot_nci_thunk_gen.exe:36 only, and Windows does allow filesnames length > 36 chars
21:58
kudo/stdier: cff96f1 | jnthn++ | src/core/Exception.pm:
Color output of compile time errors.
22:02
kudo/stdier: 8dbb9a8 | jnthn++ | src/core/Exception.pm:
Only show location if we have it.

Just hardening for the case we don't for some reason.
22:05
22:12 dngor joined
dalek rrot: c3f5458 | rurban++ | t/codingstd/filenames.t:
[test] get rid of the failing TODO test as it is logically wrong

The filenames are taken from MANIFEST and installable_parrot_nci_thunk_gen.exe is always there, regardless the OS. .exe is only on Windows, where it is safe.
22:17
22:17 dngor_ joined 22:21 travis-ci joined
travis-ci [travis-ci] parrot/parrot#649 (threads - 5190ab1 : Reini Urban): The build was fixed. 22:21
[travis-ci] Change view : github.com/parrot/parrot/compare/f...90ab1bd5ce
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/3480578
22:21 travis-ci left 22:26 dngor joined 22:46 bluescreen joined 23:21 travis-ci joined
travis-ci [travis-ci] parrot/parrot#650 (master - c5ac3f1 : Reini Urban): The build was broken. 23:21
[travis-ci] Change view : github.com/parrot/parrot/compare/a...ac3f199722
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/3480746
23:21 travis-ci left
dalek p: 9808c64 | rurban++ | src/pmc/sixmodelobject.pmc:
untabify src/pmc/sixmodelobject.pmc

windows 4-char tabs made it hard to read on unix. Added coda
23:46
p/gh67-threads: 98cc8bd | rurban++ | src/pmc/nqplexpad.pmc:
nqplexpad.pmc: add coda
23:48
p/gh67-threads: 35cff96 | rurban++ | t/nqp/67-threads.t:
t/nqp/67-threads.t: fix test - print ok 1
23:51 diakopter joined
diakopter hi; is someone around who is knowledgeable in the new threads system? 23:52