This PR changes the way actions are generated to prevent hitting an LLVM corner case when compiling action function and also improve the optimizations applied to each action, by not giving up.
This would become much more important as FilterStates, Unwind and Replay actions are added and they increase the number of states which are jumping to each others.
This PR changes the way actions are generated to prevent hitting an LLVM corner case when compiling action function and also improve the optimizations applied to each action, by not giving up.
This would become much more important as
FilterStates
,Unwind
andReplay
actions are added and they increase the number of states which are jumping to each others.