github.com/moarvm/moarvm | IRC logs at colabti.org/irclogger/irclogger_logs/moarvm
Set by AlexDaniel on 12 June 2018.
nwc10 good *, #moarvm 15:43
back, half awake, might be more awake tomorrow
[Coke]: read your test results - nothing jumps out at me as an obvious "why", and I can't yet see how to go further to try to dig into them. So, tomorrow 15:44
nine: it occurred to me at some point in the middle of the night - instead of atomic ops, as only 13 flag bits are currently used, could we split flags into two `U8`s, with MVM_CF_HAS_OBJECT_ID in one, MVM_CF_REF_FROM_GEN2 in the other, and the other 11 paritioned in some sane way? 15:46
(likely all the GC related flags in the same "side" of the parition as MVM_CF_REF_FROM_GEN2, and most of the rest in the other side)
I have not tested this thought experiment, other than counting flag bits.