The strategy is this:
- we compute SCCs once all outlives constraints are known
- we allocate a set of values **per region** for storing liveness
- we allocate a set of values **per SCC** for storing the final values
- when we add a liveness constraint to the region R, we also add it
to the final value of the SCC to which R belongs
- then we can apply the constraints by just walking the DAG for the
SCCs and union'ing the children (which have their liveness
constraints within)
There are a few intermediate refactorings that I really ought to have
broken out into their own commits:
- reverse the constraint graph so that `R1: R2` means `R1 -> R2` and
not `R2 -> R1`. This fits better with the SCC computation and new
style of inference (`->` now means "take value from" and not "push
value into")
- this does affect some of the UI tests, since they traverse the
graph, but mostly the artificial ones and they don't necessarily
seem worse
- put some things (constraint set, etc) into `Rc`. This lets us root
them to permit mutation and iteration. It also guarantees they don't
change, which is critical to the correctness of the algorithm.
- Generalize various helpers that previously operated only on points
to work on any sort of region element.
We used to accumulate a vector of type-check constraints, but now we
generate them as we go, and just store the NLL-style
`OutlivesConstraint` and `TypeTest` information.
This commit modifies the UserAssertTy statement to take a canonicalized
type rather than a regular type so that we can handle the case where the
user provided type contains a inference variable.
This allows us to re-use the `normalize` method on `TypeCheck`, which
is important since normalization may create fresh region
variables. This is not an ideal solution, though, since the current
representation of "liveness constraints" (a vector of (region, point)
pairs) is rather inefficient. Could do somewhat better by converting
to indices, but it'd still be less good than the older code. Unclear
how important this is.
We now visit just the stuff in the CFG, and we add liveness
constraints for all the random types, regions etc that appear within
rvalues and statements.
In particular, if we see a variable is DROP-LIVE, but it is not
MAYBE-INIT, then we can ignore the drop. This leavess attempt to use
more complex refinements of the idea (e.g., for subpaths or subfields)
to future work.
