Closed janvrany closed 1 year ago
Let's dedicate some time perhaps this coming Thursday (or any day shortly after the Workshop) to this. I haven't yet paid detailed attention to the AArch64 model, and am intrigued by your findings about "empty decoder" etc.
This PR contains a bunch of unrelated changes that we done when working on AArch64 support and on alternative bitvector implementation.
However, I think they're worth having merged - they make things more consistent and also reduce the number of changes other branches (which will probably live as branches for little longer)
Feel free to cherry-pick what looks good and discuss what does not look so good.