github.com/moarvm/moarvm | IRC logs at colabti.org/irclogger/irclogger_logs/moarvm
Set by AlexDaniel on 12 June 2018.
MasterDuke timotimo: you're right, MVM_string_latin1_decode doesn't see nearly as many duplicates, the most are 26 12:34
timotimo are there duplicates when you include which compunit the strings come from? 20:12
MasterDuke hm, don't think that's available inside MVM_string_latin1_decode (where i was doing my printing)? 20:33
timotimo if you're collecting the data from gdb using breakpoints and commands, it could be stolen from a stack frame further up 21:02
but that might be a bit annoying
you should be able to use at least the caller of latin1_decode, which is potentially MVM_cu_string_obtain or something 21:03
right now i'm trying vscode with their c/c++ extension 21:05
seems all right so far
MasterDuke i've been experimenting with vscode at work. don't like it 21:08
double clicking to select text and copy/paste is incredibly frustrating. it keeps not selecting, so then i past right in the middle instead of replacing 21:09
timotimo that does sound very annoying 21:51
MasterDuke huh, compunits don't have an id or name? 22:28
at least, i don't see anything in MVMCompUnit or MVMCompUnitBody
if the idx passed to MVM_cu_obtain_string is the right thing to use then duplication drops considerably. i.e., highest count is 23 for idx = 1, string = '<mainline>' 22:34