[PATCH] Don't pass llvm::DenseSet<> by value.

Alexander Kornienko alexfh at google.com
Wed Apr 23 07:57:16 PDT 2014


Hi jordan_rose,

Fixes massive performance problems on large translation units.

http://reviews.llvm.org/D3471

Files:
  lib/StaticAnalyzer/Frontend/AnalysisConsumer.cpp

Index: lib/StaticAnalyzer/Frontend/AnalysisConsumer.cpp
===================================================================
--- lib/StaticAnalyzer/Frontend/AnalysisConsumer.cpp
+++ lib/StaticAnalyzer/Frontend/AnalysisConsumer.cpp
@@ -307,7 +307,7 @@
   /// analyzed. This allows to redefine the default inlining policies when
   /// analyzing a given function.
   ExprEngine::InliningModes
-  getInliningModeForFunction(const Decl *D, SetOfConstDecls Visited);
+  getInliningModeForFunction(const Decl *D, const SetOfConstDecls &Visited);
 
   /// \brief Build the call graph for all the top level decls of this TU and
   /// use it to define the order in which the functions should be visited.
@@ -414,8 +414,8 @@
 }
 
 static bool shouldSkipFunction(const Decl *D,
-                               SetOfConstDecls Visited,
-                               SetOfConstDecls VisitedAsTopLevel) {
+                               const SetOfConstDecls &Visited,
+                               const SetOfConstDecls &VisitedAsTopLevel) {
   if (VisitedAsTopLevel.count(D))
     return true;
 
@@ -435,7 +435,7 @@
 
 ExprEngine::InliningModes
 AnalysisConsumer::getInliningModeForFunction(const Decl *D,
-                                             SetOfConstDecls Visited) {
+                                             const SetOfConstDecls &Visited) {
   // We want to reanalyze all ObjC methods as top level to report Retain
   // Count naming convention errors more aggressively. But we should tune down
   // inlining when reanalyzing an already inlined function.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: D3471.8775.patch
Type: text/x-patch
Size: 1566 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20140423/b0a00292/attachment.bin>


More information about the cfe-commits mailing list