[PATCH] D71279: [DebugInfo][MachineSink] Don't stop copy-propagating DBG_VALUEs that will have a sunk copy created

[Adrian Prantl via Phabricator via llvm-commits]

aprantl added inline comments.


================
Comment at: llvm/lib/CodeGen/MachineSink.cpp:1084
+        ProcessDbgInst(*DbgMI);
+        DbgInstsAlreadySinking.insert(DbgMI);
+      }
----------------
if (!AlreadySinking) ?


================
Comment at: llvm/lib/CodeGen/MachineSink.cpp:1091
+      if (AlreadySinking)
+        continue;
+
----------------
generally, the control flow here isn't quite obvious. It almost looks like a mistake that the instructions inserted into the set on the line above aren't counted here, but I think it's intentional. Is there a way to make it more obvious? Perhaps by renaming the AlreadySinking variable?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D71279/new/

https://reviews.llvm.org/D71279