github.com/moarvm/moarvm | IRC logs at colabti.org/irclogger/irclogger_logs/moarvm
Set by AlexDaniel on 12 June 2018.
nwc10 jnthn: paste.scsys.co.uk/582237 - that test case, wtih ASAN's excitement 08:43
I needed to set MVM_SPESH_BLOCKING=1
I can't currerntly build rakudo with MVM_SPESH_BLOCKING=1
and NQP fails 3 tests
jnthn nwc10: Thanks 09:37
nwc10 I hope that the cause for both the bug tested here and the rakudo build failure turns out to be the same 09:45
brrt \o 10:23
yoleaux 11 Nov 2018 10:59Z <nine> brrt: t/02-rakudo/12-proto-arity-count.t is failing occasionally with 'Register types do not match between value and node' at the end after all tests passed. Error goes away completely with MVM_SPESH_BLOCKING=1 which makes me believe that we're trying to JIT compile while already shutting down the VM.
brrt nine: noted
that is weird though
dogbert17 brrt: there's a tad more info in M#992 but you've probably already seen that 10:41
synopsebot_ M#992 [open]: github.com/MoarVM/MoarVM/issues/992 [⚠ blocker ⚠] MVM_panic in t/02-rakudo/12-proto-arity-count.t
brrt I guess I'll have to actually take a look one of these days 12:32
nwc10 brrt: or bribe jnthn -- nakup.itesco.cz/groceries/en-GB/sh.../beers/all 12:38
brrt or that. But I'm not sure it's ethical 12:44
brrt Hmm. I kind of want to debug the issue you're all sending me. But I'm tired and have a bunch of $work. So it's not likely to be today 17:00
nine .tell brrt scratch that idea of JITing while interp shutdown. We don't reach the end of MVM_interp_run and not even the FUDGED! message in t/spec/S06-advanced/callframe.rakudo.moar. But in all cases, we finish all tests. 20:16
yoleaux nine: I'll pass your message to brrt.