[PATCH] D64270: [analyzer][NFC] Prepare visitors for different tracking kinds

Kristóf Umann via Phabricator via cfe-commits cfe-commits at lists.llvm.org
Wed Jul 17 13:22:06 PDT 2019


Szelethus marked an inline comment as done.
Szelethus added inline comments.


================
Comment at: clang/include/clang/StaticAnalyzer/Core/BugReporter/BugReporterVisitors.h:110
+  ConditionTracking
+};
+
----------------
Charusso wrote:
> What about the following?
> ```
> enum class TrackKind {
>   Full,     // comment what it does
>   Condition // comment what it does
> };
> ```
> 
> Please consider the following example by unspoken naming conventions:
> ```
> enum class ColoringKind {
>   RedColoring,
>   BlueColoring
> }
> ```
> would be
> ```
> enum class Color { Red, Blue }
> ```
> where each trivial what is does, by name.
> 
> Please also note that, the thoroughness not working with redecls, assumptions, loop conditions, anything we failed to inline, etc... so it is not really that full tracking. But in our capabilities, it is the full what we could do.
There's a good reasoning behind this -- if you take a look at this, and followup patches, the behavior the different `TrackingKind`s cause may differ from visitor to visitor, so I decided to document each behavior there. That way, you can't avoid the documents even if you didn't glance over this enum.

> Please also note that, the thoroughness not working with redecls, assumptions, loop conditions, anything we failed to inline, etc... so it is not really that full tracking. But in our capabilities, it is the full what we could do.

I don't see what you mean here?


================
Comment at: clang/lib/StaticAnalyzer/Core/BugReporterVisitors.cpp:81
+
+bool bugreporter::trackExpressionValue(const ExplodedNode *InputNode,
+                                       const Expr *E, BugReport &report,
----------------
xazax.hun wrote:
> Do we need this overload? What about a default argument?
I don't want to expose this functionality outside of this file. I don't insist though!


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D64270/new/

https://reviews.llvm.org/D64270





More information about the cfe-commits mailing list