Parrot 6.4.0 "Double-eyed Fig Parrot" | parrot.org/ | Log: irclog.perlgeek.de/parrot | #parrotsketch meeting Tuesday 19:30 UTC
Set by moderator on 25 May 2014.
00:42 rurban joined 01:14 sorear joined 01:19 rurban joined 01:23 d4l3k_ joined 01:34 FROGGS_ joined 01:37 Util_ joined, pmichaud joined 02:01 sorear joined 02:15 sorear joined 02:21 rurban joined 03:22 rurban joined 03:23 rurban1 joined 03:53 rurban joined 03:55 zeitue joined
zeitue How does the parrot VM implement registers? 03:56
04:34 FROGGS_ joined 04:37 particle joined 05:03 FROGGS_ joined 05:26 rurban joined 05:28 rurban1 joined
Util zeitue: I am not sure at what depth you are asking your question. 05:58
For the basics, search for "register" in:
docs.parrot.org/parrot/devel/html/d...o.pod.html
docs.parrot.org/parrot/devel/html/d...w.pod.html
For the lowest-level details, see lines 15-57 of:
include/parrot/context.h
(I meant to say:) 05:59
github.com/parrot/parrot/blob/mast.../context.h
Util sleeps, but will check the backscroll for more questions, in 8 hours.
06:29 rurban joined
zeitue Util, thanks that's what I wanted to know 06:46
06:55 noize joined 07:30 rurban joined 08:06 FROGGS_ joined 08:27 basiliscos joined 08:31 rurban joined 08:33 rurban1 joined 09:34 rurban joined 10:35 rurban joined 11:35 rurban joined 12:16 rurban joined
Coke wonders who cottoo is. 12:18
12:19 bighugedog joined 12:56 rurban joined 13:11 rurban joined
rurban_ my overnight benchmark run brought overall better results and no improvement (0%) in the WB branch 13:14
13:16 Chirag joined
Chirag rurban_ : I mailed mdk .. waiting for a response.. Also, Upasana talked to perigrin for me.. she got a reply that they don't have any more rooms and that there is too less time to arrange sponsorship for me.. 13:24
rurban_ yes. looks bad. too late 13:26
I didn't think about it
Chirag I have mailed my university to give me details so that I can try for Bulgaria .. but that has got too many moving parts for me 13:28
13:28 ivan joined
rurban_ yes, for me too 13:30
13:35 Chirag_ joined 13:37 Chirag_ joined 13:38 Chirag_ joined, Chirag joined
Chirag rurban_ : Can u please take a look at os.pmc ?.. Not sure if it requires changes.. 14:38
14:53 bluescreen joined
rurban_ i haven't looked at dynpmc's yet at all. they all need improvements, yes 15:10
Chirag actually this one just has METHODs and no VTABLE .. not sure about some.. will leave it for now 15:14
rurban_ METHOD"s should be fine for now. pmc2c didn't complain 15:21
more important would be to find the remaining GC bug. I'll try to add some diagbnostic code to the GC to find the offending PMC 15:22
Chirag ok..
ptr.pmc has *get_pointer() with manual_wb .. It returns a void* ptr .. Do we really require a WB ?? 15:24
rurban_ lunch break 15:37
yes. it was there before 15:40
rurban commenting your recent changes on github 15:58
on exceptions which do not change SELF we don't need a WB 15:59
Chirag please give an instance.. 16:01
i ll correct these 16:05
rurban see github.com/ZYROz/parrot/commit/fec...f4ebe64e53 16:06
16:11 FROGGS joined
Chirag I am pretty sure i messed up in [a-i] as well.. will redo 16:49
16:57 davidfetter joined
Chirag rurban: destroy() is a non writer.. it should compulsorily be a no_wb.. right?? 17:27
rurban_ destroy is a special case. it really needs a WB 17:56
Chirag oh..
18:23 rurban joined 18:40 Chirag joined
Chirag VTABLE void push_string(STRING *v) { 18:41
VTABLE_push_string(INTERP, PARROT_IMAGEIOSTRINGS(SELF)->list, v);
}
rurban: should this have a manual_wb?
self is not being passed.. i think 18:42
rurban_ only a subitem of SELF is WB'ed here, buut since subitem (->list) is WB'ed I think :manual_wb should be safe here. In doubt let it WB'ed 18:45
we still have to find this gc.t bug 18:46
so you can prove t/op/gc.t with or without
Chirag alright.. I ll note this one 18:47
rurban_ Object.morph misses {} 19:00
Chirag oops .. corrected 19:05
rurban_ hmm, worse than before 19:07
Chirag huh?
rurban_ t/compilers/opsc/02-parse-all-ops.t SEGV's now at parsing src/ops/var.ops
Chirag my recent push misses lots of UNUSED .. will fix them in a while 19:09
rurban_ yes, no prob. I saw
but first I want to find this missing WB 19:10
rebuild, now the same old op/gc.t bug only 19:13
SEGV in Class_instantiate
now' I'll find it 19:14
19:27 sorear joined
Chirag I have to re-fix p*.pmc .. But this one had come back after I changed [a-i] 19:29
rurban_ t/op/gc.t passes fine with another GC type: ms2 instead of gms 19:46
Chirag I dont know the difference between them... But day before it did pass for us.. I think a change I am making in [a-d]* might be the one.. 19:54
more specifically c*.pmc or default.pmc 19:55
rurban_ --gc=ms passes also 19:59
I'm just trying those GCs because they are easier to debug 20:00
20:02 wagle joined
dalek rrot: 923e016 | rurban++ | / (3 files):
--gc=inf: fix t/op/gc.t and mention gms in Configure --help

gc_type = INF should skip the first gc tests		 20:26
rrot: 6dfbb64 | rurban++ | / (2 files):
add local .perldb helper to .gitignore/MANIFEST.SKIP

so that git clean -dxf will not wipe it.
rrot: 8e173af | rurban++ | config/auto/pmc.pm:
Improve pmc2c deps for PMC/$class.pm libs

depends only on the really used deps. also sort the deps to generate comparable Makefile
Heuristic branch merge: pushed 34 commits to parrot/rurban/pmc2c_orig2-gh1069 by rurban 20:40
rurban_ rebased and cleaned up rurban/pmc2c_orig2-gh1069 20:41
20:51 travis-ci joined
travis-ci [travis-ci] parrot/parrot#1033 (master - 8e173af : Reini Urban): The build passed. 20:51
[travis-ci] Change view : github.com/parrot/parrot/compare/a...173af06203
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/26888165
20:51 travis-ci left 21:55 bluescreen_ joined 22:14 rurban joined 22:57 basiliscos 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!