[PATCH] D56079: Fix invalid test in Guard Widening

[Max Kazantsev via Phabricator via llvm-commits]

mkazantsev added a comment.

Sorry, I do not understand your objection's meaning. The test's comment states that

  ; With guards in loops, we're okay hoisting out the guard into the
  ; containing loop.

There is no containing loop in this test. So the claims of this test are wrong to start with.

In `GenericDomTreeConstruction.h`, there is a comment:

  // Find the furthest away we can get by following successors, then
  // follow them in reverse.  This gives us some reasonable answer about
  // the post-dom tree inside any infinite loop. In particular, it
  // guarantees we get to the farthest away point along *some*
  // path. This also matches the GCC's behavior.
  // If we really wanted a totally complete picture of dominance inside
  // this infinite loop, we could do it with SCC-like algorithms to find
  // the lowest and highest points in the infinite loop.  In theory, it
  // would be nice to give the canonical backedge for the loop, but it's
  // expensive and does not always lead to a minimal set of roots.

So when we are trying to construct PDT for infinite loop, we choose *some* block
as its non-trivial root. No guarantee is given that this block is going to be block `outer`.
I don't remember what was this situation, but there was a case where `inner` was chosen
a root, and therefore no post-dominance between them was inserted.

Could you please elaborate what exactly are you disagree with?


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D56079/new/

https://reviews.llvm.org/D56079