github.com/moarvm/moarvm | IRC logs at colabti.org/irclogger/irclogger_logs/moarvm
Set by AlexDaniel on 12 June 2018.
nwc10 good *, #moarvm 06:19
Geth MoarVM/AAAA-A-better-hash: efd104ca44 | (Nicholas Clark)++ | 2 files
oops() with an informative message if hashtable->entry_size has not been set.

If code attempts to perform an insert on a hash table which still has entry_size set to 0, then the size allocation calculation will give 0 bytes, and (without this check) MVM_malloc() (or the FSA) would be called to allocate 0 bytes. malloc() might return NULL, in which case MVM_malloc() would panic (which does not give any useful diagnostics). If not, it will ... (6 more lines)
07:57
nwc10 That, I think, is the only logical thing missing from the branch - an oops if someone manages a code path that would otehrwise panic about 12 lines later 07:58
(or do things that would make ASAN very excited)
merge (fixing that commit message) after the release, but before the next weekly? (and claim that it is fasterand encourage people to test it?) 07:59
I have no coffee - need to be AFK for a bit to fix this. 08:00
sena_kun nwc10, yes, merging after release please. :) 08:09
by the way, mighty #moarvm, are there any immediate release blockers for moarvm release planned for tomorrow?
nine is not aware of any 08:12
nwc10 I now have coffee (and a hot backup) 08:20
I'm not aware of any blockers.
nwc10 good *, brrt 08:20
brrt good * nwc10 08:21
timotimo: I did not get time to respond yesterday but yeah that sound very promising
jnthn sena_kun: Looking over the commits in master since last release, nothing stands out to be as particularly risky, especially if the ecosystem testing looks good. 08:49
[Coke] "merging after release" also "merging after coffee" 13:22
[Coke] is already on to cold caffiene here.