If we have already calculated the incoming state before, use that
as our starting point to ensure we are conservative.
Details
Details
Diff Detail
Diff Detail
- Repository
- rG LLVM Github Monorepo
Event Timeline
Comment Actions
LGTM
For context, Craig and I debugged this earlier today. We had an infinite loop issue which came down to the fact we allowed two waves of updates to propagate through a loop, and the merge points allowed us to toggle back and forth between states instead of falling to the bottom value of the lattice. No known small reproducer at this time unfortunately.