================
@@ -495,6 +503,49 @@ void MoveChecker::checkPostCall(const CallEvent &Call,
assert(!C.isDifferent() && "Should not have made transitions on this path!");
}
+bool MoveChecker::evalCall(const CallEvent &Call, CheckerContext &C) const {
----------------
Xazax-hun wrote:
What is the main reason for using `evalCall` as opposed to `checkPostCall`?
When we use `evalCall` we need to model all aspects of the function, e.g.,
setting up all the bindings in the analysis state to model what the function
would do. And only ever one checker can do `evalCall` for a function. If the
main purpose is to update checker specific state, it is usually better to use
`checkPostCall`.
https://github.com/llvm/llvm-project/pull/196602
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
