[clang-tools-extra] [clang-tidy] fix false positive that floating point variable only used in increment expr in cert-flp30-c (PR #108706)

[Julian Schmidt via cfe-commits]

================
@@ -9,22 +9,33 @@
 #include "FloatLoopCounter.h"
 #include "clang/AST/ASTContext.h"
 #include "clang/ASTMatchers/ASTMatchFinder.h"
+#include "clang/ASTMatchers/ASTMatchers.h"
 
 using namespace clang::ast_matchers;
 
 namespace clang::tidy::cert {
 
 void FloatLoopCounter::registerMatchers(MatchFinder *Finder) {
-  Finder->addMatcher(
-      forStmt(hasIncrement(expr(hasType(realFloatingPointType())))).bind("for"),
-      this);
+  Finder->addMatcher(forStmt(hasIncrement(forEachDescendant(
+                                 declRefExpr(hasType(realFloatingPointType()),
+                                             to(varDecl().bind("var"))))),
+                             hasCondition(forEachDescendant(declRefExpr(
+                                 hasType(realFloatingPointType()),
+                                 to(varDecl(equalsBoundNode("var")))))))
+                         .bind("for"),
+                     this);
 }
 
 void FloatLoopCounter::check(const MatchFinder::MatchResult &Result) {
   const auto *FS = Result.Nodes.getNodeAs<ForStmt>("for");
 
-  diag(FS->getInc()->getExprLoc(), "loop induction expression should not have "
-                                   "floating-point type");
+  diag(FS->getInc()->getBeginLoc(), "loop induction expression should not have "
+                                    "floating-point type");
----------------
5chmidti wrote:

Thoughts on binding both `declRefExpr`s and streaming their source ranges into the diagnostic?

```
for (float val = 0.1f; val <= 1.0f; val += 0.1f) {}
                       ~~~          ^~~
```
(probably also changing the diag location to the `declRefExpr` in the increment as well)

You could also add the name of the bound variable to the diagnostic

https://github.com/llvm/llvm-project/pull/108706