Perl 6 - the future is here, just unevenly distributed

IRC log for #moarvm, 2014-11-20

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

All times shown according to UTC.

Time Nick Message
00:27 vendethiel joined #moarvm
00:58 vendethiel joined #moarvm
01:39 zakharyas joined #moarvm
02:03 vendethiel joined #moarvm
02:50 jimmy_ joined #moarvm
02:50 vendethiel joined #moarvm
03:35 vendethiel joined #moarvm
04:48 vendethiel joined #moarvm
05:15 vendethiel joined #moarvm
05:41 vendethiel joined #moarvm
06:27 vendethiel joined #moarvm
06:38 kjs_ joined #moarvm
07:45 FROGGS joined #moarvm
07:58 zakharyas joined #moarvm
08:03 vendethiel joined #moarvm
08:55 nwc10 "All tests successful." - ASAN still doesn't spot anything
08:55 FROGGS I'd like to get these OSX failures nailed down :/
08:57 JimmyZ FROGGS: https://github.com/MoarVM/MoarVM/commit/33970c6dfdddb5f1d3d667bf1e294ee8f09185d0#diff-0cb6ca157d7b79b7d7e5ace869130f8cR879
08:59 FROGGS JimmyZ: I replied today via phone... did you get my message?
09:01 JimmyZ FROGGS: I didn't see it anywhere
09:01 FROGGS JimmyZ: k... I said that reverting it is fine by me, in exchange that you'd take a look at issue #154 :o)
09:03 FROGGS my gut feeling is that the string allocated here: https://github.com/MoarVM/MoarVM/blob/master/src/strings/ops.c#L826
09:04 * JimmyZ doesn't have OSX :(
09:05 FROGGS causes trouble when we do not use it because of: https://github.com/MoarVM/MoarVM/blob/master/src/strings/ops.c#L883
09:06 FROGGS JimmyZ: but you have brains! :D
09:07 JimmyZ FROGGS: it's not a problem with allocating unused MVMString.
09:07 FROGGS but it is about trying to free that union
09:08 FROGGS maybe clang+osx needs special treatment when it comes to unions?
09:08 FROGGS ... or onions
09:11 JimmyZ and I don't think the bt give me any info :(. I think it's about GC movement which causes *any pointing to wrong place :P
09:14 FROGGS hmmm
09:14 FROGGS but any points to memory managed by C
09:18 JimmyZ FROGGS: yeah, but GC maybe move its address to wrong place, something like off-by-one on OSX
09:19 JimmyZ MVM_gc_allocate_object allocates zero memory be default
09:19 JimmyZ *zeroed
09:22 FROGGS yes yes, I know now :o)
09:23 lizmat The MoarVM build is filled with warnings... maybe that can be a source of inspiration ?
09:23 * JimmyZ saw some warnings with NFC.c
09:24 JimmyZ *NFA.c
09:34 lizmat afk for a few hours&
09:47 FROGGS JimmyZ: these warnings are about debug message printing only AFAIK... but wants fixing anyway
09:47 FROGGS lizmat: would be nice if you could provide a build log then :o)
09:47 JimmyZ yeah, I didn't see any other warnings :P
09:48 vendethiel joined #moarvm
10:15 dalek joined #moarvm
11:01 vendethiel joined #moarvm
11:32 lizmat FROGGS: here you go: a copy of a new build log on OS X: https://gist.github.com/lizmat/862c5a8868a916122c42
11:34 lizmat it's only 5.6K lines and 370K bytes
11:58 vendethiel joined #moarvm
12:22 vendethiel joined #moarvm
13:14 vendethiel joined #moarvm
13:56 brrt joined #moarvm
14:11 dalek MoarVM: 2508b40 | (Elizabeth Mattijsen)++ | src/core/frame.c:
14:11 dalek MoarVM: Fix warnings
14:11 dalek MoarVM:
14:11 dalek MoarVM: "using the result of an assignment as a condition without parentheses" on OS X
14:11 dalek MoarVM: (clang)
14:11 dalek MoarVM: review: https://github.com/MoarVM/MoarVM/commit/2508b40bd5
14:27 hoelzro joined #moarvm
14:34 JimmyZ joined #moarvm
14:34 dalek MoarVM: 6cd9478 | (Elizabeth Mattijsen)++ | src/ (4 files):
14:34 dalek MoarVM: Fix "'&&' within '||'" warnings on OS X (clang)
14:34 dalek MoarVM: review: https://github.com/MoarVM/MoarVM/commit/6cd9478aec
14:46 dalek MoarVM: 85b1a50 | (Elizabeth Mattijsen)++ | src/6model/serialization.c:
14:46 dalek MoarVM: Fix suspect warning on OS X (clang)
14:46 dalek MoarVM:
14:46 dalek MoarVM: type specifier missing, defaults to 'int'.  Since all of the similar functions
14:46 dalek MoarVM: where void as well, I made this one also void (instead of the default int).
14:46 dalek MoarVM: This also fixed the related "control reaches end of non-void function" warning.
14:46 dalek MoarVM: review: https://github.com/MoarVM/MoarVM/commit/85b1a50f4a
14:50 brrt lizmat++
14:50 brrt now if you could speed up clang in the first place.. :-P
14:50 lizmat I would much rather prefer a stable rakudo on OS X  :-(
14:53 brrt isn't it?
14:53 brrt hmmm
14:53 kjs_ joined #moarvm
15:00 [Coke] brrt: not for a month now.
15:01 [Coke] lots of segfaults (jit) and invalid memory usage errors (nonjit)
15:01 brrt ok, that's odd
15:01 brrt in spectests?
15:01 [Coke] one of the latest tickets has a backtrace on a memory issue.
15:01 [Coke] brrt: yes.
15:01 brrt ok
15:01 brrt let me see if i can repeat that a bit
15:01 brrt things are rarely memory issues though
15:02 [Coke] brrt: Just telling you what rakudo is telling us. :)
15:03 brrt :-) i know, just telling you my gut feeling
15:03 brrt memory issues are fairly rare
15:03 [Coke] note that the backtrace I have there, I only got when running under the test_summary test harness. (was definitely coming from rakudo, though, not the perl5 harness)
15:11 dalek MoarVM: 910a06c | (Elizabeth Mattijsen)++ | src/mast/driver.c:
15:11 dalek MoarVM: fix another assignment in condition warning
15:11 dalek MoarVM: review: https://github.com/MoarVM/MoarVM/commit/910a06cca9
15:12 lizmat timotimo: looking at src/spesh/optimize.c, line 310
15:12 lizmat if ((safety_cur->info->operands[0] & MVM_operand_rw_mask == MVM_operand_write_reg)
15:13 lizmat I think we need parens here (suggestion by clang)
15:13 lizmat clang is doing the == before the &
15:14 lizmat brrt: don't you agree?
15:16 brrt lett me see
15:16 jnthn Not jut in Clang, it's that way in the C standard... http://en.cppreference.com/w/c/language/operator_precedence
15:16 brrt wait
15:16 brrt that seems... wrong
15:16 brrt i /think/ we should have two &'s there
15:16 jnthn No, & is right; it's a bitmask
15:17 lizmat from the wording of the wording of the variables, I think we just need parens
15:17 brrt but a == b isn't guaranteed to set 1
15:17 brrt or is it
15:17 lizmat that's why I think there's () missing
15:17 lizmat feels to me it should first mask, and then compare
15:18 jnthn lizmat: Yes, that's the intent. Mask then compare.
15:18 jnthn lizmat: So I think the code is wrong.
15:18 lizmat ok, that's one bug fixed then...
15:18 * lizmat goes to test effect on rakudo
15:25 lizmat still flaky spectest  :-(
15:25 zakharyas joined #moarvm
15:27 dalek MoarVM: 018a260 | (Elizabeth Mattijsen)++ | src/spesh/optimize.c:
15:27 dalek MoarVM: First mask bitmap, then compare
15:27 dalek MoarVM: review: https://github.com/MoarVM/MoarVM/commit/018a260f38
15:29 lizmat src/spesh/inline.c:96 gives "add explicit braces to avoid dangling else" warning
15:30 lizmat not sure what to make of that...
15:37 dalek MoarVM: 434de0c | (Elizabeth Mattijsen)++ | src/spesh/inline.c:
15:37 dalek MoarVM: Fix dangling else (hopefully the right way)
15:37 dalek MoarVM: review: https://github.com/MoarVM/MoarVM/commit/434de0c265
15:38 * brrt is running spectests
15:38 brrt hmmm
15:39 lizmat ?
15:39 brrt for so far as these are clang issues, i could test them at home
15:39 brrt my mac-access is limited to $work
15:39 lizmat that would be great...
15:39 brrt :-)
15:39 lizmat I mean, the noise of the clang build is deafening, drowning out any other squeaks
15:39 * brrt agrees
15:40 brrt which is why i always build with gcc ;-)
15:42 brrt otherwise... i'd fear to think what could possibly be so mac-os-x specific that moar would be sensitive for
15:46 dalek MoarVM: cf39e6c | jimmy++ | src/strings/ops.c:
15:46 dalek MoarVM: Revert "attempt to fix a non-alloc+free #154"
15:46 dalek MoarVM:
15:46 dalek MoarVM: This reverts commit 33970c6dfdddb5f1d3d667bf1e294ee8f09185d0.
15:46 dalek MoarVM: review: https://github.com/MoarVM/MoarVM/commit/cf39e6c885
15:46 dalek MoarVM: c8b0b97 | jimmy++ | src/gc/debug.h:
15:46 dalek MoarVM: remove an unused semicolon
15:46 dalek MoarVM: review: https://github.com/MoarVM/MoarVM/commit/c8b0b979fa
15:53 jnthn sleep &
15:53 lizmat gnight jnthn !
15:54 brrt tests are flaky indeed
15:54 brrt sleep well jnthn :-)
15:55 JimmyZ good night too
16:14 lizmat implicit declaration of function 'MVM_panic'
16:14 lizmat guess we forgot to include a .h somewhere ?
16:17 FROGGS[mobile] joined #moarvm
16:18 dalek MoarVM: 4f1ba61 | (Elizabeth Mattijsen)++ | src/platform/posix/mmap.c:
16:18 dalek MoarVM: Add missing include
16:18 dalek MoarVM: review: https://github.com/MoarVM/MoarVM/commit/4f1ba61af2
16:28 dalek MoarVM: 8f4d53f | (Elizabeth Mattijsen)++ | src/jit/graph.c:
16:28 dalek MoarVM: Fix another assignment as condition warning
16:28 dalek MoarVM: review: https://github.com/MoarVM/MoarVM/commit/8f4d53f5af
16:36 vendethiel joined #moarvm
16:36 dalek MoarVM: 933a2a2 | (Elizabeth Mattijsen)++ | tools/ucd2c.pl:
16:36 dalek MoarVM: Fix a *lot* of warnings in strings/unicode.c
16:36 dalek MoarVM: review: https://github.com/MoarVM/MoarVM/commit/933a2a22a8
16:47 FROGGS joined #moarvm
16:59 [Coke] no love on the daily run, though it predates a bunch of the warnings commits.
17:03 [Coke] sounds like brrt was able to reproduce some flakiness, that's good.
17:12 [Coke] lizmat: btw, I'm not running the spec tests in parallel.
17:12 [Coke] and I'm still seeing the fails. (I'm using test_summary, which is currently single-threaded)
17:12 lizmat ah?  and still flakiness?
17:12 lizmat hmmm.... lemme try that as well
17:13 [Coke] but same thing; fails only when running through a harness; run by hand, most of the time ok.
17:13 [Coke] I ended up making it reproducable by making a single line spectest.data, and always using test summary to launch.
17:13 lizmat does anybody know how and when to run tools/ucd2c.pl ?
17:14 lizmat [Coke]: is there a difference between "make spectest" and "make t/spec/foo.t" ?
17:14 lizmat wrt to harnessing ?
17:16 [Coke] I never use make .../foo.t, I unno off the top of my head.
17:16 [Coke] (if I'm running a single file, I almost always use prove -v -e t/fudgeandrun t/spec/foo.t
17:19 [Coke] looking at the makefile, looks like make *.t uses the harness, yes.
17:19 lizmat I *rarely* get an error testing like that
17:20 [Coke] right. any of the memory issues, if you run it by itself, it's fine.
17:20 lizmat fwiw, in the past weeks I also got errors *in* the perl5 harness
17:20 [Coke] I suspect *some* kind of memory pressure, but I have 16G, not sure that's really teh issue.
17:20 lizmat only twice, but still
17:20 * lizmat has 16G as well
17:20 lizmat maybe we haz 2much
17:33 [Coke] OHNOES
17:37 lizmat [Coke]: datapoint with TEST_JOBS=1: only 1 flapper
17:37 lizmat but not none  :-(
17:51 FROGGS_ joined #moarvm
18:47 vendethiel joined #moarvm
22:13 kjs_ joined #moarvm
23:00 kjs_ joined #moarvm
23:08 kjs_ joined #moarvm
23:33 vendethiel joined #moarvm

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