[llvm] [Analysis] Make LocationSize conversion from uint64_t explicit (PR #133342)

[Philip Reames via llvm-commits]

================
@@ -189,14 +189,16 @@ class LocationSize {
   bool operator==(const LocationSize &Other) const {
     return Value == Other.Value;
   }
-
   bool operator==(const TypeSize &Other) const {
-    return hasValue() && getValue() == Other;
+    return (*this == LocationSize::precise(Other));
----------------
preames wrote:

This change isn't related to the conversion case.  I noticed when writing the other equality operand that this idiom seems unsound - unless I'm misreading the code it allows an imprecise upper bound to compare equal with a typesize.  I can split this off if anyone asks, but I don't have a test case which shows it being actually problematic in practice.  

https://github.com/llvm/llvm-project/pull/133342