[PATCH] D20177: [MBP] Factor out the optimizations on branch conditions and unanalyzable branches. NFCI.

[Amaury SECHET via llvm-commits]

deadalnix added a comment.

Overall, looks good. Letting time for other to review.


================
Comment at: lib/CodeGen/MachineBlockPlacement.cpp:1313
@@ -1311,3 +1312,3 @@
     MachineBasicBlock *TBB = nullptr, *FBB = nullptr; // For AnalyzeBranch.
-    if (!TII->AnalyzeBranch(*PrevBB, TBB, FBB, Cond)) {
+    if (!TII->AnalyzeBranch(*PrevBB, TBB, FBB, Cond)) 
       // The "PrevBB" is not yet updated to reflect current code layout, so,
----------------
I'd keep the block here. While not strictly necessary, the comment is quite long and so it makes things clearer IMO.


Repository:
  rL LLVM

http://reviews.llvm.org/D20177