[PATCH] D157118: [NFC][Clang] Fix static analyzer concerns

[Aaron Ballman via Phabricator via cfe-commits]

aaron.ballman accepted this revision.
aaron.ballman added a comment.
This revision is now accepted and ready to land.

LGTM though there may be some additional changes worth considering; I don't insist on changes though.



================
Comment at: clang/lib/AST/StmtPrinter.cpp:178
 void StmtPrinter::PrintRawCompoundStmt(CompoundStmt *Node) {
+  assert(Node && "Compound statement cannot be null");
   OS << "{" << NL;
----------------
Hmmm, I think this is what we're effectively already hoping is the case today, but I don't think that's a safe assertion by itself. Consider: https://github.com/llvm/llvm-project/blob/cd29ebb862b5c7a81c9f39c8c493f9246d6e5f0b/clang/lib/AST/StmtPrinter.cpp#L602

It might be worth looking at all the calls to `PrintRawCompoundStmt()` to see which ones potentially can pass null in, and decide if there are additional changes to make. e.g., should that `dyn_cast` be a `cast` instead so it cannot return nullptr and we get the assertion a little bit earlier when calling `cast<>`?


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D157118/new/

https://reviews.llvm.org/D157118