github.com/moarvm/moarvm | IRC logs at colabti.org/irclogger/irclogger_logs/moarvm
Set by AlexDaniel on 12 June 2018.
Geth MoarVM/debugserver_invoke_code_support: 6 commits pushed by (Timo Paulssen)++ 13:42
Geth MoarVM/debugserver_invoke_code_support: 6 commits pushed by (Timo Paulssen)++ 15:26
MasterDuke a lot of times `body->elems` is 0 in here github.com/MoarVM/MoarVM/blob/mast...rray.c#L65 any reason not to add an `if (body->elems == 0) return`? 15:38
timotimo not a bad idea i think 15:40
MasterDuke k 15:41
MasterDuke heh, 4m times when compiling CORE.c 15:44
timotimo i wonder if accessing the elems count can be actually slower than accessing the slot kind 15:47
since there's slot kinds that require no marking of the entries, like ints and nums 15:48
MasterDuke 3454595 times it was an obj array with 0 elems, 3734 times a str array with 0 elems 15:53
MasterDuke also 495073 times that `MVM_str_hash_count(tc, hashtable) == 0` 17:06
MasterDuke is Geth down? github.com/MoarVM/MoarVM/pull/1393 21:29
Geth MoarVM: bed0a49f29 | (Daniel Green)++ | 2 files
Early exit gc_mark for VMArray+MVMHash if able

Don't need to go through the setup work to try and mark the held elements if there aren't in fact any elements.
22:26
MoarVM: 18e16b6ecf | MasterDuke17++ (committed using GitHub Web editor) | 2 files
Merge pull request #1393 from MasterDuke17/exit_early_from_gc_mark_if_able

Early exit gc_mark for VMArray+MVMHash if able