01:34 ggoebel114 joined 03:09 cognominal joined 06:01 ggoebel114 joined 06:25 Woodi joined 07:32 FROGGS[mobile] joined 07:51 FROGGS joined 10:31 dolmen joined 10:32 dolmen left 12:32 tgt joined 12:51 tgt joined 14:46 ggoebel114 joined 15:27 jnap joined 16:38 rblackwe joined, tgt joined, Woodi joined 16:42 rblackwe joined, odc joined 16:50 Woodi joined 17:02 colomon joined 17:11 FROGGS joined 17:21 retupmoca joined 17:33 ssutch joined 18:12 tgt joined
dalek arVM: 7496b56 | (Tobias Leich)++ | Configure.pl:
make -O1 the default instead of -O3 as long as we have that gc bug

There is a gc bug about an MVMStaticFrame that explodes on linuxes under -O3. Hopefully it builts on TimToady++'s machine now, too.
18:54
TimToady we'll see...
FROGGS yeah :S 18:55
lue remembers something about premature optimization :) 18:56
FROGGS hehe 18:57
premature -O3 ?
TimToady is O1 default? because there seems to be no -O1
FROGGS TimToady: -O1 should be the default when --optimize=\d is not supplied 18:58
TimToady ain't
course it doesn't tell me most of the commands 18:59
FROGGS hmmm
ohh, hmmm
well... 19:00
TimToady but there's no -O in the Makefile
FROGGS but then I don't understand why it explodes on your box at all
(not that I understand why it explodes on my box using -O3)
TimToady are you exploding in m-BOOTSTRAP? 19:01
FROGGS yes
gist.github.com/FROGGS/8efa32fce9e288e9f7f0 19:02
this way 19:03
TimToady looks about the same
trying with an explicit --optimize=1
but my nqp wants to recompile everything now... 19:04
FROGGS you could skip that if you did not use --gen-nqp/--gen-moar 19:06
I am decreasing the nursery size now... maybe I discover something 19:07
TimToady -O1 doesn't help 19:17
FROGGS damn
Object does not exist in serialization context 19:49
at gen/moar/stage2/QAST.nqp:3828 (/home/froggs/dev/rakudo/../nqp/install/languages/nqp/lib/QAST.moarvm:deserialization_code:24)
O.o
diakopter FROGGS: that's an interesting one 22:55
FROGGS I just added a BEGIN { nqp::say(nqp::x("a", 1024)) } in src/gen/m-BOOTSTRAP.nqp line 44 23:03