[PATCH] D28249: Improve scheduling with branch coalescing

[Lei Huang via Phabricator via llvm-commits]

lei marked 12 inline comments as done.
lei added inline comments.


================
Comment at: lib/CodeGen/BranchCoalescing.cpp:620
+
+  for (MachineBasicBlock::iterator MI = From.BranchBlock->getFirstNonPHI();
+       MI != End;) {
----------------
lei wrote:
> nemanjai wrote:
> > You've already moved all the PHI's from `From.BranchBlock`. Can't this just be a loop over `From.BranchBlock->instrs()`? Or it may need to be a loop from `rbegin()` to `rend()`.
> agreed
changing this to be a range based splice call instead.


https://reviews.llvm.org/D28249