NoQ added a comment.

Yay, that'll make my life a lot easier! I heard you have an automatic tool for 
this sort of stuff, right?



================
Comment at: 
clang/include/clang/StaticAnalyzer/Core/BugReporter/BugReporter.h:408
-public:
-  enum Kind { BasicBRKind, PathSensitiveBRKind };
-
----------------
Hey, i just added that! :D (well, renamed) (rC369320)

I believe we do want a separation between a {bug report, bug reporter} classes 
that's only suitable for path-insensitive reports (which would live in 
libAnalysis and we could handle them to clang-tidy while still being able to 
compile it without CLANG_ENABLE_STATIC_ANALYZER) and all the path-sensitive 
report logic that is pretty huge but only Static Analyzer needs it. For that 
purpose we'd better leave this in. WDYT? See also D66460.

Should i ask on the mailing list whether you're willing to sacrifice building 
clang-tidy without CLANG_ENABLE_STATIC_ANALYZER in order to transition to 
BugReporter? Cause i thought it was obvious that it's not a great idea, even if 
it causes me to do a bit of cleanup work on my end.

That said, i'm surprised that it's deadcode, i.e. that nobody ever dyn_casts 
bug reporters, even though we already have two bug reporter classes. Maybe we 
can indeed remove this facility.


================
Comment at: clang/lib/StaticAnalyzer/Core/BugReporter.cpp:2343
   InterExplodedGraphMap ForwardMap;
-  TrimmedGraph = OriginalGraph->trim(Nodes, &ForwardMap, &InverseMap);
 
----------------
Btw these days we strongly suspect that the whole graph trimming thing is 
useless and should be removed.


================
Comment at: clang/lib/StaticAnalyzer/Core/BugReporterVisitors.cpp:1169
 
-void FindLastStoreBRVisitor::registerStatementVarDecls(
-    BugReport &BR, const Stmt *S, bool EnableNullFPSuppression,
----------------
Woohooo!


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D66473/new/

https://reviews.llvm.org/D66473



_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Reply via email to