[PATCH] D84763: DomTree: Make PostDomTree immune to block successors swap

[Jakub Kuderski via Phabricator via llvm-commits]

kuhar added a comment.

I think this is a good direction.

Taking a few steps back, I'd love to see a cfg preservation check like D81558 <https://reviews.llvm.org/D81558> land for both IR and MIR. Without agreeing on the definition of data structure equality/equivalence (CFG, domtree, etc.), it's difficult to make progress if we don't know what exactly to preserve, let alone agreeing on how to preserve things. Historically, I think this assumptions mismatch has been a source of difficult address bug creep.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D84763/new/

https://reviews.llvm.org/D84763