Parrot 7.0.2 "Crimson Shining Parrot" | parrot.org/ | Log: irclog.perlgeek.de/parrot | #parrotsketch�meeting Fridays 13:30 UTC
Set by moderator on 29 January 2015.
00:28 rurban joined 02:47 ilbot2 joined
moderator Parrot 7.0.2 "Crimson Shining Parrot" | parrot.org/ | Log: irclog.perlgeek.de/parrot | #parrotsketch�meeting Fridays 13:30 UTC
02:55 lauleon joined 03:06 lauleon joined 03:15 lauleon joined 05:21 ggherdov_____ joined 06:11 estrabd_afk joined 06:13 mtj- joined 08:24 estrabd_afk joined 08:32 estrabd_afk joined 08:54 rurban_ joined 08:58 rurban joined, mtj- joined
dalek rrot/smoke-me/stringbuilder-gh1123: 5aa3161 | rurban++ | / (2 files):
[codingstd] for io changes
08:59
09:07 FROGGS joined 09:10 travis-ci joined
travis-ci parrot/parrot#1559 (smoke-me/stringbuilder-gh1123 - 5aa3161 : Reini Urban): The build passed. 09:10
Change view : github.com/parrot/parrot/compare/9...a316106aa9
Build details : travis-ci.org/parrot/parrot/builds/49065858
09:10 travis-ci left 09:13 mtj- joined 09:26 estrabd_afk joined 09:52 FROGGS joined 10:18 rurban_ joined 10:19 rurban1 joined
dalek rrot/smoke-me/slow-mod: f0faeae | rurban++ | / (4 files):
[core] #ifdef PARROT_SLOW_MOD or fast % mod

use optionally the 3.5% faster natural % for int mod, which is "incorrect" or different with negative arguments.
mod is profiled as hot and slow (5.5%) in parrot-bench with linux perf.
11:23
rrot/smoke-me/slow-mod: 8e17e9e | rurban++ | / (4 files):
[core] #ifdef PARROT_SLOW_MOD or fast % mod

use optionally the 3.5% faster natural % for int mod, which is "incorrect" or different with negative arguments.
mod is profiled as hot and slow (5.5%) in parrot-bench with linux perf.
11:27
11:33 travis-ci joined
travis-ci parrot/parrot#1560 (smoke-me/slow-mod - f0faeae : Reini Urban): The build passed. 11:33
Change view : github.com/parrot/parrot/commit/f0faeae534eb
Build details : travis-ci.org/parrot/parrot/builds/49073160
11:33 travis-ci left 11:42 travis-ci joined
travis-ci parrot/parrot#1561 (smoke-me/slow-mod - 8e17e9e : Reini Urban): The build passed. 11:42
Change view : github.com/parrot/parrot/compare/f...17e9ee02cd
Build details : travis-ci.org/parrot/parrot/builds/49073333
11:42 travis-ci left
dalek rrot/smoke-me/stringbuilder-gh1123: ac13b03 | rurban++ | / (3 files):
[cage] even more const io api arguments
12:08
12:20 travis-ci joined
travis-ci parrot/parrot#1562 (smoke-me/stringbuilder-gh1123 - ac13b03 : Reini Urban): The build passed. 12:20
Change view : github.com/parrot/parrot/compare/5...13b03e3939
Build details : travis-ci.org/parrot/parrot/builds/49075294
12:20 travis-ci left 12:53 kid51 joined
dalek rrot/smoke-me/gcdebug-ms: bc5eb0b | rurban++ | / (3 files):
[gc] Support -D1 GC stats with --ccflags=-DMEMORY_DEBUG for --gc=ms
12:53
13:05 travis-ci joined
travis-ci parrot/parrot#1563 (smoke-me/gcdebug-ms - bc5eb0b : Reini Urban): The build passed. 13:05
Change view : github.com/parrot/parrot/commit/bc5eb0b5143e
Build details : travis-ci.org/parrot/parrot/builds/49078051
13:05 travis-ci left 13:12 kid51 joined 13:22 travis-ci joined
travis-ci parrot/parrot#1564 (smoke-me/gcdebug-ms-gh1196 - bc5eb0b : Reini Urban): The build passed. 13:22
Change view : github.com/parrot/parrot/compare/s...-ms-gh1196
Build details : travis-ci.org/parrot/parrot/builds/49079203
13:22 travis-ci left 13:32 FROGGS joined 14:57 rurban joined 15:05 rurban joined
dalek rrot/smoke-me/gcdebug-ms-gh1196: e949ae3 | rurban++ | / (7 files):
Prepare deprecation of the ListChunk GC API, and src/list.c

See GH #1179. See #define PARROT_BUFFERLIKE_LIST in config.h to use it, still. imageiofreeze still uses it.
17:06
17:24 travis-ci joined
travis-ci parrot/parrot#1565 (smoke-me/gcdebug-ms-gh1196 - e949ae3 : Reini Urban): The build passed. 17:24
Change view : github.com/parrot/parrot/compare/b...49ae3c86f2
Build details : travis-ci.org/parrot/parrot/builds/49095657
17:24 travis-ci left
dalek rrot/smoke-me/gcdebug-ms-gh1196: a760eb3 | rurban++ | src/gc/gc_ (2 files):
[gc] Improve -D1 GC stats for --gc=ms
17:26
17:38 travis-ci joined
travis-ci parrot/parrot#1566 (smoke-me/gcdebug-ms-gh1196 - a760eb3 : Reini Urban): The build passed. 17:38
Change view : github.com/parrot/parrot/compare/e...60eb3b0837
Build details : travis-ci.org/parrot/parrot/builds/49097646
17:38 travis-ci left
dalek rrot/smoke-me/gcdebug-ms-gh1196: 882c189 | rurban++ | src/gc/ (2 files):
[gc] add some -D201 support to --gc=ms, for strings
17:52
17:55 rurban_ joined 18:02 travis-ci joined
travis-ci parrot/parrot#1567 (smoke-me/gcdebug-ms-gh1196 - 882c189 : Reini Urban): The build passed. 18:02
Change view : github.com/parrot/parrot/compare/a...2c18973268
Build details : travis-ci.org/parrot/parrot/builds/49099551
18:02 travis-ci left
rurban_ Found the string problems for the old ms, ms2 collectors 18:35
local functions need to block the GC while using a local string var. It may get moved away!
yep, that's it. 30% perf. win! let's cross fingers 18:40
nope, still more... 18:45
dalek rrot/smoke-me/gcdebug-ms-gh1196: d56c942 | rurban++ | src/pmc/ (2 files):
[pmc] add missing fields to StringHandle.clone

read_offset, filename, read_buffer, write_buffer, record_separator also need to get cloned. Probably related to GH #1196
19:18
rrot/smoke-me/gcdebug-ms-gh1196: a88abb9 | rurban++ | src/ (2 files):
[gc] fix a MS segv in Parrot_io_write_s

When certain functions cause a GC, the local string buffers may get moved away, but local STRING* variables or worse, strings inside a StringHandle may still point to the old location.
Lock the GC then for this section, mostly during IO writing to string handles, when the handle needs to realloc its buffer (the stringhandle). But reading into a buffer is also problematic. we really should know the size beforehand.
One way to fix GH #1196, lock the GC sweep in those cases. The other variants would be:
  - lock the whole GC via Parrot_block_GC_mark
  - unset the PObj_is_movable flag in the STRING that it may not be moved,
   as it is still locally referenced.
Remaining ms test failure: t/pmc/fixedstringarray.t
rurban_ hmm, threads problems with ms and only 11% faster now 19:28
19:31 travis-ci joined
travis-ci parrot/parrot#1568 (smoke-me/gcdebug-ms-gh1196 - a88abb9 : Reini Urban): The build passed. 19:31
Change view : github.com/parrot/parrot/compare/8...8abb9c2d21
Build details : travis-ci.org/parrot/parrot/builds/49106488
19:31 travis-ci left
dalek rrot/smoke-me/gc-ms-default: 2484b54 | rurban++ | src/ (2 files):
[gc] fix a MS segv in Parrot_io_write_s

When certain functions cause a GC, the local string buffers may get moved away, but local STRING* variables or worse, strings inside a StringHandle may still point to the old location.
Lock the GC then for this section, mostly during IO writing to string handles, when the handle needs to realloc its buffer (the stringhandle). But reading into a buffer is also problematic. we really should know the size beforehand.
One way to fix GH #1196, lock the GC sweep in those cases. The other variants would be:
  - lock the whole GC via Parrot_block_GC_mark
  - unset the PObj_is_movable flag in the STRING that it may not be moved,
   as it is still locally referenced.
Remaining ms test failure: t/pmc/fixedstringarray.t
19:50
rrot/smoke-me/gc-ms-default: 1cc6cf1 | rurban++ | config/auto/gc.pm:
[config] set MS as new GC default
19:58 rurban left 19:59 travis-ci joined
travis-ci parrot/parrot#1569 (smoke-me/gc-ms-default - 1cc6cf1 : Reini Urban): The build failed. 19:59
Change view : github.com/parrot/parrot/compare/2...c6cf19ef1b
Build details : travis-ci.org/parrot/parrot/builds/49109699
19:59 travis-ci left 23:24 davidfetter joined