Parrot 4.8.0 "Spix's Macaw" | parrot.org/ | Log: irclog.perlgeek.de/parrot | #parrotsketch meeting Tuesday 19:30 UTC
Set by moderator on 20 September 2012.
00:02 rurban_mobile joined, kid51 joined
whiteknight good evening, #parrot 00:05
00:07 kid51_ joined 00:44 benabik joined 01:03 kid51 joined 01:09 kid51_ joined
rurban gmail is down for me, so irc. 01:14
Did you see my comments on SREG vs the imcc optimizer. At least I could fix the concat bug.
I believe NotFound added encodings 01:15
01:29 Psyche^ joined 01:41 l3l1p joined 01:46 kid51 joined 01:54 kid51_ joined 01:56 MikeFair joined
dalek rtcl-nqp/nqp2: 295de92 | coke++ | src/Partcl/Actions.pm:
More updates for QAST
 02:10
02:30 particle joined 03:20 Psyche^ joined 04:43 l3l1p joined 05:13 ruoso joined 05:18 woosley joined 05:27 rurban_mobile joined 05:39 l3l1p joined 07:09 brrt joined 07:34 Psyche^ joined 08:11 lucian joined 08:42 Khisanth joined
nine Might be worth reading: students.ics.uci.edu/~sbruntha/cgi...key=thesis 12:15
12:51 brrt joined 13:48 bluescreen joined 13:55 ligne joined 14:00 brrt left 14:02 brrt joined 14:10 PacoAir joined 14:19 rurban_mobile joined
rurban nine: Aber nur auf Deutsch 14:59
Oh, it's in english. good. Just the first page is in german. 15:00
brrt deutsch ist super einfach :-) 15:04
nine It even mentions Parrot (though it says Parrot is for executing Perl) 15:25
15:34 contingencyplan joined 15:40 brrt joined 15:53 darbelo joined
dalek rrot/native_pbc2: b6584ae | rurban++ | src/dynpmc/ (2 files):
dynpmc/os: actually build this dummy pmc for loadlib "os" backcompat
 15:59
rrot/native_pbc2: c36a7c5 | rurban++ | MANIFEST.generated:
dynpmc/os: and now actually install it

Sigh
rrot/native_pbc2: 216ce80 | rurban++ | MANIFEST.generated:
[CAGE] remove duplicate os.bundle in MANIFEST.generated
rrot/native_pbc2: 95df0c9 | rurban++ | / (3 files):
[GH# 828] Avoid passsing non-double through va_list

va_list does not support float, long double nor __float128 on most systems. Promote FLOATVAL to double to avoid that.
rrot/native_pbc2: a96c72b | rurban++ | / (7 files):
[GH #828] Fix most tests for non-double floatval

Add more needed precisions to numeric tests, tested with intel long double.
Change a magic number in t/library/test_more.t 888.81<=>888.82 1e-2
   was not stable enough for intel long double
Use t/pmc/testlib/annotations.pbc for binary comparisons of packfiles since
   stringifying non-double numbers is not stable enough.
Fix src/dynoplibs/sys.ops doublesize to numvalsize. This only works for double.
rrot/native_pbc2: ef15cff | rurban++ | src/ops/ (2 files):
[GH #828] Optimize print_n/i and say_n/i

print_n/nc/i/ic and the say variants go through a full Parrot_io_printf with 3 va_start/va_end scenarios and a custom format string parser, and through char* -> STRING->char* conversion until the fixed FLOATVAL_FMT or INTVAL_FMT printf is done. Just to print a single int or number.
This also lost precision for non-double floatval's - if cast to (double). Now avoided.
rrot/native_pbc2: bf3b35c | rurban++ | src/packfile/pf_items.c:
PF_fetch_number: revert union back to cast
rrot/native_pbc2: e925dfc | rurban++ | src/packfile/pf_items.c:
typo in cvt_num16ppc_num10 (d1 uninitialized)
rrot/native_pbc2: d2ed3bb | rurban++ | t/src/extend.t:
skip outdated extend tests with old embed API [GH #829]

I also tried to convert extend_15 to the new API but failed to understand Parrot_runcode()
rrot/native_pbc2: c26e17a | rurban++ | config/gen/makefiles/root.in:
[cage] Makefile cover-clean. Make less noise and use echo

make -s clean listed all dirs to be cleaned up for gcov-generated files. Print only one line
rrot/native_pbc2: 34915f8 | rurban++ | / (5 files):
Merge remote-tracking branch 'origin/master' into native_pbc2
rrot/native_pbc2: 1be8ab8 | rurban++ | lib/Parrot/Test.pm:
Test::Builder::_normalize handle inf/nan
rurban So should I look into vmgen? (as unladen-swallow)
16:12 dmalcolm joined 16:21 woolfy joined 16:22 LizM joined 16:25 zby_home joined 17:03 tuxit joined
dalek rrot/native_pbc2: 6847dd9 | rurban++ | / (5 files):
Add _DIG defines to config.h, fix cvt_num8_num16, add -lquadmath

The __float128 cast does not work, so use snprintf and strtoflt128. Add -lquadmath to gcc when i_quadmath succeeds. Add .c.i makefile suffix rule Fixedr ROUND_NUM_TO 3rd arguments with the new _DIG defines		 19:06
rrot: e923aa3 | rurban++ | tools/dev/pprof2cg.pl:
[cage] Remove windowism from pprof2cg.pl

use /usr/bin/perl code		 19:10
rrot/gh837_concat: 2d37216 | rurban++ | tools/dev/pprof2cg.pl:
[cage] Remove windowism from pprof2cg.pl

use /usr/bin/perl code
chekkaa transferring drupal 6 test site to production now 19:15
done 19:18
enjoy your shiny new drupal 6 site!
if you run into any problems, ping me on #osuosl on freenode or shoot an email to support@osuosl.org 19:19
rurban chekkaa: Many thanks for doing this! 19:38
19:43 tuxit joined
Coke chekkaa: thanks! 19:45
rurban Got now almost most numbers right, just __float128 is hard to set 19:50
Even strtoflt128() does not work for me. 19:51
q 19:52
dalek rrot/native_pbc2: 0552125 | rurban++ | src/packfile/pf_items.c:
Add new 3arg fetch_buf_nn() converters, fix constness

The old fetch_buf_?e_nn() converters took only 2 args, but the nv converters take now 3. So add intermediate inlined special versions.
Fixed a couple of remaining const'ness issues and bswap64 lines.
Hardcode "%.16g" for the strtoflt128 buf. Still does not work yet though.		 19:55
rurban Working now in ByteBuffer resize GH#835 20:14
20:17 vex joined
rurban uh, STRING* source plus char* content is stored in a ByteBuffer, and of course they are not updated together. 20:23
dalek rrot/gh837_concat: cd367ec | rurban++ | t/pmc/bytebuffer (2 files):
Add bytebuffer resize tests for [GH #835]

As it turns out adding the failing test to t/pmc/bytebuffer.t does not reveil any problem, but if called seperately as main it does.		 20:31
rrot/gh837_concat: 203ba2a | rurban++ | t/pmc/bytebuffer (2 files):
Merge branch 'rurban/bytebuffer-resize-gh835' into gh837_concat
rrot/gh837_concat: def5ca9 | rurban++ | / (2 files):
[GH #837] Fix set ByteBuffer from string

set_string_native() did not reset char* content, only the STRING *source.
rrot/rurban/bytebuffer-resize-gh835: 5eddcaa | rurban++ | / (2 files):
[GH #837] Fix set ByteBuffer from string

set_string_native() did not reset char* content, only the STRING *source.		 20:33
p: c458a08 | moritz++ | src/ (2 files):
make Rabin-Miller prime check available as nqp::isprime_I

also special-case 1 not to be a prime		 20:38
kudo/nom: 2e33656 | moritz++ | / (3 files):
add expmod and is-prime
 20:39
rrot/rurban/bytebuffer-resize-gh835: 8da9b50 | rurban++ | / (3 files):
Add bytebuffer resize tests for [GH #835]

As it turns out adding the failing test to t/pmc/bytebuffer.t does not reveil any problem, but if called seperately as main it does. Add t/pmc/bytebuffer2.t to MANIFEST		 20:43
rrot/rurban/bytebuffer-resize-gh835: 196f596 | rurban++ | / (3 files):
[GH #835] Fix set ByteBuffer from string

set_string_native() did not reset char* content, only the STRING *source.
20:44 vex left
rurban I'll merge now rurban/bytebuffer-resize-gh835 and gh837_concat 20:48
dalek rrot/gh837_concat: 93179fc | rurban++ | ChangeLog:
[GH #837] ChangeLog line for imcc optimizer bug
 20:51
rrot/gh837_concat: e82482e | rurban++ | examples/benchmarks/stress_stringsu.pir:
[GH #837] Add benchmark to test encoding slowdown

stress_stringsu.pir tests encoding via sprintf and concat. The fixed version is only a bit slower then the old broken version. 13.666s (broken) vs 13.506s (fixed) vs 3.022s (without encoding).
So fixing the encoding representation as STRING* in SREG and consts 'U' is a worthwile goal.
rrot/gh837_concat: 9c81593 | rurban++ | examples/benchmarks/stress_strings1.pir:
[GH #837] New comparison benchmark against stress_stringsu.pir

Bad news: There is almost no measurable speed advantage from ascii strings to encoded strings. With the fix and without the fix, as sprintf handling slows down the test 4x times.
So converting a SREG and const to encoded STRING* will not gain much performance.
rrot: 5d7aa2d | dukeleto++ | t/pmc/string.t:
[t] Add some failing tests for #837 reported by pmichaud++
 20:57
rrot: b9fc29b | dukeleto++ | t/op/stringu.t:
[t] Move concat tests for #837 to a more appropriate file
rrot: c552eb6 | dukeleto++ | t/pmc/string.t:
Revert "[t] Add some failing tests for #837 reported by pmichaud++"

This reverts commit 5d7aa2d0d7489aadc3bc3037271d919634483b90.
rrot: 8a56a8e | rurban++ | / (3 files):
[GH #837] Honor encoding in imcc optimizer

Context registers unfortunately do not store type information, just the values. We would need at least type 'U" here. So recreate an proper const'ed encoding string from the register value.
rrot: 93179fc | rurban++ | ChangeLog:
[GH #837] ChangeLog line for imcc optimizer bug
rrot: e82482e | rurban++ | examples/benchmarks/stress_stringsu.pir:
[GH #837] Add benchmark to test encoding slowdown

stress_stringsu.pir tests encoding via sprintf and concat. The fixed version is only a bit slower then the old broken version. 13.666s (broken) vs 13.506s (fixed) vs 3.022s (without encoding).
So fixing the encoding representation as STRING* in SREG and consts 'U' is a worthwile goal.
rrot: 9c81593 | rurban++ | examples/benchmarks/stress_strings1.pir:
[GH #837] New comparison benchmark against stress_stringsu.pir

Bad news: There is almost no measurable speed advantage from ascii strings to encoded strings. With the fix and without the fix, as sprintf handling slows down the test 4x times.
So converting a SREG and const to encoded STRING* will not gain much performance.
rrot: 81f3b68 | rurban++ | / (6 files):
Merge branch 'gh837_concat'

Fixes [GH #837], a imcc optimizer problem with SREG's and const not holding STRING*, thus no good encoding information.
rurban nope: rurban/bytebuffer-resize-gh835 is not yet ready to be merged 21:11
dalek rrot/rurban/bytebuffer-resize-gh835: 5d7aa2d | dukeleto++ | t/pmc/string.t:
[t] Add some failing tests for #837 reported by pmichaud++
 21:48
rrot/rurban/bytebuffer-resize-gh835: b9fc29b | dukeleto++ | t/op/stringu.t:
[t] Move concat tests for #837 to a more appropriate file
rrot/rurban/bytebuffer-resize-gh835: c552eb6 | dukeleto++ | t/pmc/string.t:
Revert "[t] Add some failing tests for #837 reported by pmichaud++"

This reverts commit 5d7aa2d0d7489aadc3bc3037271d919634483b90.
rrot/rurban/bytebuffer-resize-gh835: 8a56a8e | rurban++ | / (3 files):
[GH #837] Honor encoding in imcc optimizer

Context registers unfortunately do not store type information, just the values. We would need at least type 'U" here. So recreate an proper const'ed encoding string from the register value.
rrot/rurban/bytebuffer-resize-gh835: 93179fc | rurban++ | ChangeLog:
[GH #837] ChangeLog line for imcc optimizer bug
rrot/rurban/bytebuffer-resize-gh835: e82482e | rurban++ | examples/benchmarks/stress_stringsu.pir:
[GH #837] Add benchmark to test encoding slowdown

stress_stringsu.pir tests encoding via sprintf and concat. The fixed version is only a bit slower then the old broken version. 13.666s (broken) vs 13.506s (fixed) vs 3.022s (without encoding).
So fixing the encoding representation as STRING* in SREG and consts 'U' is a worthwile goal.
rrot/rurban/bytebuffer-resize-gh835: 9c81593 | rurban++ | examples/benchmarks/stress_strings1.pir:
[GH #837] New comparison benchmark against stress_stringsu.pir

Bad news: There is almost no measurable speed advantage from ascii strings to encoded strings. With the fix and without the fix, as sprintf handling slows down the test 4x times.
So converting a SREG and const to encoded STRING* will not gain much performance.
rrot/rurban/bytebuffer-resize-gh835: 81f3b68 | rurban++ | / (6 files):
Merge branch 'gh837_concat'

Fixes [GH #837], a imcc optimizer problem with SREG's and const not holding STRING*, thus no good encoding information.
rrot/rurban/bytebuffer-resize-gh835: 2b95094 | rurban++ | / (5 files):
Merge 'rurban/bytebuffer-resize-gh835'

Fixed conflict in ChangeLog
rrot/rurban/bytebuffer-resize-gh835: 3851cfd | rurban++ | / (3 files):
[GH #835] Rewrite ByteBuffer.set_string_native

Check for all needed content allocation cases. The previous version failed when realloc'ing the content, Now use a fresh copy when needed.
parrot: 8da9b50 | rurban++ | / (3 files):
parrot: Add bytebuffer resize tests for [GH #835]
parrot:
parrot: As it turns out adding the failing test to t/pmc/bytebuffer.t
parrot: does not reveil any problem, but if called seperately as main it does.
parrot: Add t/pmc/bytebuffer2.t to MANIFEST
21:50 dalek joined
dalek rrot/rurban/bytebuffer-resize-gh835: 47c73c6 | rurban++ | t/pmc/bytebuffer.t:
[GH #835] adjust the test plan number
 21:52
rrot/rurban/bytebuffer-resize-gh835: 43b35f7 | rurban++ | t/pmc/bytebuffer2.t:
[codingstd] t/pmc/bytebuffer2.t whitespace
rrot: 49948fa | dukeleto++ | docs/project/release_manager_guide.pod:
[doc] Sign up to release Parrot 5.0.0
 21:55
22:26 travis-ci joined
travis-ci [travis-ci] parrot/parrot#569 (master - 2b95094 : Reini Urban): The build was broken. 22:26
[travis-ci] Change view : github.com/parrot/parrot/compare/8...950944c022
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/2526603
22:26 travis-ci left
chekkaa So, I found out why you guys wanted drupal 6 22:43
there were some modules you wanted 22:44
I'll install those now
cotto thanks 22:52
22:53 lucian joined 23:09 whiteknight joined 23:23 lucian joined
whiteknight good evening, #parrot 23:26
23:50 mdupont joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!