timotimo jnthn: you know how we recently got param_sp and param_sn into the profiler's allocation logging thingie? 15:10
jnthn: i'm looking at a profile where "pull-one" is allocating a million BOOTHashes, pretty much all of them from jitted code, but the spesh output doesn't show a param_sn at all 15:11
i'm not sure what's going on there
timotimo the afl-tmin processes right now take about 50 minutes to minimize each test case 18:46
timotimo does the bytecode actually know the number of arguments a frame wants to accept statically? 22:46