github.com/moarvm/moarvm | IRC logs at colabti.org/irclogger/irclogger_logs/moarvm
Set by AlexDaniel on 12 June 2018.
MasterDuke also, guess this isn't really a big deal at all, but we create 3 temps, but only ever use two 00:16
timotimo it looks like it's specifically built to use the temp slots, since it takes that idx parameter 00:58
and it does look like all functions that use force_bigint use it only(?) to provide a bigint into an mp op 00:59
scovit MasterDuke: it feels to me that libtom should provide a macro to reserve an initialized mp_int on the stack, and that would be a saner approach 10:30
tellable6 scovit, I'll pass your message to MasterDuke
MasterDuke timotimo: ah, i hadn't checked that far, good to know 15:01
tellable6 2020-01-09T10:30:06Z #moarvm <scovit> MasterDuke: it feels to me that libtom should provide a macro to reserve an initialized mp_int on the stack, and that would be a saner approach
MasterDuke anybody want to help me debug `This representation (P6num) cannot unbox to other types (for type TestNum)` thrown here github.com/perl6/nqp/blob/master/t...int.t#L137 when using this moarvm branch github.com/MoarVM/MoarVM/pull/1221 ? 15:08
why doesn't nqp show the right line number in its backtrace? it says the line number of the str() sub, but then immediately after that it just says line 1 of the mainline, not the actual line it gets to 15:32
scovit Hi MasterDuke, about your message of yesterday, why MoarVM does not allocate thos mp_int on the stack? I've looked at libtommath docs, there is mp_init_set and mp_init_set_int in that library. 15:48
scovit Basically it would be a trivial change to force_bigint static function, and you free that array on the global tc 15:49
MasterDuke scovit: good question, i don't know
my first goal is just to get moarvm/nqp/raku working with libtommath v1.2.0 with the fewest possible changes 15:50
scovit me neither. I do not think it matters in terms of performance particularly
but code clarity for sure
Ok :)
MasterDuke but after that there are a bunch of possible cleanups/simplifications. e.g., ltm 1.2 introduces some functions that we had created ourselves, so now wee can switch to theirs 15:52
so i'll keep your idea in mind for that secondary work, thanks 15:53
scovit you are welcome, thank to you to take care of upgrading the library 15:54
brrt \o 18:40
nwc10 o/ 18:51
brrt probably needs to check out the raku/perl side of things a bit more often... 18:55
MasterDuke timotimo: you around? this error i'm getting seems like the sort of thing you've helped me out with before (if only the lessons had stuck...) 19:52
timotimo oh? 20:20

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!