[clang] [analyzer][NFC] Add some docs for LazyCompoundValue (PR #97407)

DonĂ¡t Nagy via cfe-commits cfe-commits at lists.llvm.org
Wed Jul 3 00:45:11 PDT 2024


================
@@ -346,6 +352,39 @@ class CompoundVal : public NonLoc {
   static bool classof(SVal V) { return V.getKind() == CompoundValKind; }
 };
 
+/// The simplest example of a concrete compound value is nonloc::CompoundVal,
+/// which represents a concrete r-value of an initializer-list or a string.
+/// Internally, it contains an llvm::ImmutableList of SVal's stored inside the
+/// literal.
+///
+/// However, there is another compound value used in the analyzer, which appears
+/// much more often during analysis, which is nonloc::LazyCompoundVal. This
+/// value is an r-value that represents a snapshot of any structure "as a whole"
+/// at a given moment during the analysis. Such value is already quite far from
+/// being re- ferred to as "concrete", as many fields inside it would be unknown
+/// or symbolic. nonloc::LazyCompoundVal operates by storing two things:
+///   * a reference to the TypedValueRegion being snapshotted (yes, it is always
+///     typed), and also
+///   * a copy of the whole Store object, obtained from the ProgramState in
----------------
NagyDonat wrote:

I do understand functional programming (I had a phase when I learnt Haskell and for a few years I thought that it's the absolutely perfect language :star_struck: -- since then I'm more realistic but I still like it), I was just confused by the fact that this line emphasizes "a copy of the whole Store object" as opposed to the "reference to" in the previous line.

In a functional language the word "copy" is practically non-existent (we speak about values, not objects and their identity, so there is no reason to say that this is _a copy of_ e.g. that list -- we simply say that "this is that list"), while in C++ saying "copy" instead of "move" or "pointer/reference" is an important synonym of "we allocate lots of new memory", so I'd still prefer tweaking this line.

https://github.com/llvm/llvm-project/pull/97407


More information about the cfe-commits mailing list