================
@@ -555,24 +588,305 @@ std::string
StateUpdateReporter::getMessage(PathSensitiveBugReport &BR) const {
return std::string(Out.str());
}
-bool StateUpdateReporter::providesInformationAboutInteresting(
- SymbolRef Sym, PathSensitiveBugReport &BR) {
- if (!Sym)
- return false;
- for (SymbolRef PartSym : Sym->symbols()) {
- // The interestingess mark may appear on any layer as we're stripping off
- // the SymIntExpr, UnarySymExpr etc. layers...
- if (BR.isInteresting(PartSym))
- return true;
- // ...but if both sides of the expression are symbolic, then there is no
- // practical algorithm to produce separate constraints for the two
- // operands (from the single combined result).
- if (isa<SymSymExpr>(PartSym))
+// If the base expression of `expr` refers to a `ConstantArrayType`,
+// return the element type and the array size.
+// Note that "real" Flexible Array Members are `IncompleteArrayType`.
+// For them, this function returns `std::nullopt`.
+static std::optional<std::pair<QualType, nonloc::ConcreteInt>>
+getArrayTypeInfo(SValBuilder &svb, ArraySubscriptExpr const *expr) {
+ auto const *arrayBaseExpr = expr->getBase()->IgnoreParenImpCasts();
+ auto const arrayQualType = arrayBaseExpr->getType().getCanonicalType();
----------------
NagyDonat wrote:
Here and elsewhere please follow the [LLVM Coding
Standard](https://llvm.org/docs/CodingStandards.html#name-types-functions-variables-and-enumerators-properly)
and use `CamelCase` names for variables.
Moreover, I would suggest avoiding `auto` in situations like this, because IMO
spelling out the type would make the code more readable. (The [coding
standard](https://llvm.org/docs/CodingStandards.html#use-auto-type-deduction-to-make-code-more-readable)
leaves this question for the judgement of the developer. I tend to use `auto`
only when the type name is already spelled out on the same line – e.g. by a
`dyn_cast` – or it is an iterator or some similar implementation detail type.)
https://github.com/llvm/llvm-project/pull/159357
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
