[PATCH] D130906: [clang] format string checking for conpile-time evaluated str literal

[Timm Bäder via Phabricator via cfe-commits]

tbaeder added inline comments.


================
Comment at: clang/lib/Sema/SemaChecking.cpp:8504
     return SLCT_UncheckedLiteral;
+  Expr::EvalResult Result;
+  if (E->EvaluateAsRValue(Result, S.Context)) {
----------------
A comment above this line would be helpful. Would also visually separate it from the `return` above, which just confused me.


================
Comment at: clang/lib/Sema/SemaChecking.cpp:8507
+    if (Result.Val.isLValue()) {
+      auto *LVE = Result.Val.getLValueBase().dyn_cast<const Expr *>();
+      if (LVE && LVE->getStmtClass() == Stmt::StringLiteralClass) {
----------------
I think you should be able to unify the two `if` statements.

Can you not `dyn_cast_or_null<StringLiteral>(Result.Val.getLValueBase())` here instead of casting to `Expr*` and checking the `StmtClass`?


================
Comment at: clang/test/Sema/format-strings-scanf.c:235
   scanf(0 ? "%s" : "%d", i); // no warning
-  scanf(1 ? "%s" : "%d", i); // expected-warning{{format specifies type 'char *'}}
+  scanf(1 ? "%s" : "%d", i); // expected-warning{{format specifies type 'char *'}} \
+                             // expected-note{{format string is defined here}}
----------------
inclyc wrote:
> These new notes are FixIt hints, looks much better than before.
Can you show some sample output?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D130906/new/

https://reviews.llvm.org/D130906