act has a lot of code that deals with sanitising and manipulating assembly. This code is from when the project started (as we were previously dealing with herd and needed to compile tests into herd-compatible assembly).
The current direction of act is towards using litmus and doing so in ways that avoid assembly manipulation entirely. As such, the assembly work (sanitising, abstract analysis, obj_splitmus, etc) is basically not being used anymore, and is a maintenance/compilation-time/space burden.
As such, I'm proposing it gets removed entirely from act, for now at least.
Note: In an ideal world, what would happen now is that the act-asm infrastructure would move off into its own separate package. By filing an issue for its removal from act, I'm not suggesting that it's irredeemable or broken or anything like that. I'm just stating that, for now at least, I can't justify its maintenance.
acthas a lot of code that deals with sanitising and manipulating assembly. This code is from when the project started (as we were previously dealing withherdand needed to compile tests intoherd-compatible assembly).The current direction of
actis towards usinglitmusand doing so in ways that avoid assembly manipulation entirely. As such, the assembly work (sanitising, abstract analysis,obj_splitmus, etc) is basically not being used anymore, and is a maintenance/compilation-time/space burden.As such, I'm proposing it gets removed entirely from
act, for now at least.Note: In an ideal world, what would happen now is that the
act-asminfrastructure would move off into its own separate package. By filing an issue for its removal fromact, I'm not suggesting that it's irredeemable or broken or anything like that. I'm just stating that, for now at least, I can't justify its maintenance.