| ugexe | i can't put my finger on it but it still seems a bit redundant | 00:09 | |
| Like without having thought about it for more than 30 seconds I think I would have conditionally set a `new_status` variable and then have a single MVM_cas call outside of the conditionals, something like `MVM_cas(&other->gc_status, status_bits | suspend_bits, new_status | suspend_bits) == (status_bits | suspend_bits))`. Although having typed out this thought I'm not sure that would be equivalent | 00:15 | ||
|
00:42
timo left
00:55
leont left,
leont joined
10:15
librasteve_ joined
10:59
vrurg joined,
vrurg_ left
11:15
timo joined
12:01
Geth left,
Geth joined
13:05
librasteve_ left
|
|||
| patrickb | Are frames (on the context of the debugserver) completely anologous to the `caller` chain? | 13:07 | |
| lizmat | afaik, yes, that's at least the mental model I've been using timo? | 13:24 | |
| but note, something like ".say for ^10" may introduce a frame that is not obvious from looking at the code | 13:25 | ||
| otoh, static optimization may have removed frames already | 13:30 | ||
| is also my understanding | |||
| patrickb | Context: I'm pondering how to present the different views of data in the debugger. Now I wonder if I should allow introspecting both, frames and the caller chain or if I could just as well drop one of the two concepts. Given frames don't have a strong presence in Raku I wonder if I could just present the debugger frames under the name "callers". But that's only ok to do if they match perfectly. | 13:33 | |
| lizmat | there *is* a class CallFrame in Rakudo | 13:43 | |
| docs.raku.org/type/CallFrame | 13:44 | ||
| patrickb | Good point! | 14:16 | |
| timo | i have been meaning to double-check how exactly especially inlines are reflected (or not) in that output, but what the debugserver gives you is, as far as i can tell, exactly what you can get a "context" object for, which is also a moarvm "thing" that we use inside rakudo to do stuff | 14:35 | |
|
15:32
librasteve_ joined
19:02
Sario left,
Sario joined
20:15
librasteve_ left
23:27
gfldex left,
gfldex joined
|
|||