lkorenc

Currently, `__remill_undef_*` do not take any operands. However there are places in semantics where it would be desirable to specify a preferred value, since it could reflect the usual behaviour...

Currently if several tests fail, the log is really huge and hard to navigate (especially if tests fail in CI) and will most likely require manual investigation anyway. One way...

After some instructions, value inside some arithmetic flags may be undefined. X86 semantics currently handle this non-uniformly by either using `__remill_undefined_*` intrinsics, storing arbitrary values, or ignoring the undefined value...

DF not handled properly

1

`DF` is not having a correct value after executing some instruction. Best seen with `popf` and can crash program in other cases (by possibly not restoring properly?)

For example `BTS` does not work with greater immediates.

`pop rsp` currently returns result increased by `8` opposed to expected value. One possible explanation is that following happens: ``` new = *rsp store(dst, new) *rsp = *rsp + 8...

`core::SelectOp` and some `void` related fixes.

1

Adds lowering of `core::SelectOp`. Fixes some corner cases related to `void` "values" being returned.

### Prerequisites - [X] I have read the documentation and the proposed feature is not implemented. ### Description Currently conversion tests are an archaic mess with very random directory structue....

### Prerequisites - [X] I have read the documentation and the proposed feature is not implemented. ### Description Currently we turn this feature off as it was introducing breakages on...

Not sure if there are any limitations from MLIR side for this, but for some patterns it make sense to have an overridable methods that users (that link against VAST,...