<html><head></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; "><br><div><div>On May 31, 2012, at 11:35 AM, David Blaikie wrote:</div><br class="Apple-interchange-newline"><blockquote type="cite"><br><br><div class="gmail_quote">On Thu, May 31, 2012 at 11:07 AM, Anna Zaks <span dir="ltr"><<a href="mailto:ganna@apple.com" target="_blank">ganna@apple.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Author: zaks<br>
Date: Thu May 31 13:07:55 2012<br>
New Revision: 157762<br>
<br>
URL: <a href="http://llvm.org/viewvc/llvm-project?rev=157762&view=rev" target="_blank">http://llvm.org/viewvc/llvm-project?rev=157762&view=rev</a><br>
Log:<br>
[analyzer] Cleanup for r157721.<br>
We should lock the number of elements after the initial parsing is<br>
complete. Recursive AST visitors in AnalyzesConsumer and CallGarph can<br>
trigger lazy pch deserialization resulting in more calls to<br>
HandleTopLevelDecl and appending to the LocalTUDecls list. We should<br>
ignore those.<br></blockquote><div><br></div><div>Is it only a perf impact to visit them unnecessarily? or could this affect (in)correctness as well? (in the latter case, a test case might be nice, if it's practical to write one)</div>
<div><br></div></div></blockquote><div><br></div>Performance issue. Very tricky to write a test case.</div><div><br><blockquote type="cite"><div class="gmail_quote"><div>Thanks!</div><div>- David</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
Modified:<br>
cfe/trunk/lib/StaticAnalyzer/Frontend/AnalysisConsumer.cpp<br>
<br>
Modified: cfe/trunk/lib/StaticAnalyzer/Frontend/AnalysisConsumer.cpp<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/StaticAnalyzer/Frontend/AnalysisConsumer.cpp?rev=157762&r1=157761&r2=157762&view=diff" target="_blank">http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/StaticAnalyzer/Frontend/AnalysisConsumer.cpp?rev=157762&r1=157761&r2=157762&view=diff</a><br>
==============================================================================<br>
--- cfe/trunk/lib/StaticAnalyzer/Frontend/AnalysisConsumer.cpp (original)<br>
+++ cfe/trunk/lib/StaticAnalyzer/Frontend/AnalysisConsumer.cpp Thu May 31 13:07:55 2012<br>
@@ -230,7 +230,7 @@<br>
<br>
/// \brief Build the call graph for all the top level decls of this TU and<br>
/// use it to define the order in which the functions should be visited.<br>
- void HandleDeclsGallGraph();<br>
+ void HandleDeclsGallGraph(const unsigned LocalTUDeclsSize);<br>
<br>
/// \brief Run analyzes(syntax or path sensitive) on the given function.<br>
/// \param Mode - determines if we are requesting syntax only or path<br>
@@ -311,18 +311,16 @@<br>
}<br>
}<br>
<br>
-void AnalysisConsumer::HandleDeclsGallGraph() {<br>
+void AnalysisConsumer::HandleDeclsGallGraph(const unsigned LocalTUDeclsSize) {<br>
// Otherwise, use the Callgraph to derive the order.<br>
// Build the Call Graph.<br>
CallGraph CG;<br>
<br>
// Add all the top level declarations to the graph.<br>
- // Note: TraverseDecl may modify LocalTUDecls, but only by appending more<br>
- // entries. Thus we don't use an iterator, but rely on LocalTUDecls<br>
- // random access. By doing so, we automatically compensate for iterators<br>
- // possibly being invalidated, although this is a bit slower.<br>
- const unsigned n = LocalTUDecls.size();<br>
- for (unsigned i = 0 ; i < n ; ++i) {<br>
+ // Note: CallGraph can trigger deserialization of more items from a pch<br>
+ // (though HandleInterestingDecl); triggering additions to LocalTUDecls.<br>
+ // We rely on random access to add the initially processed Decls to CG.<br>
+ for (unsigned i = 0 ; i < LocalTUDeclsSize ; ++i) {<br>
CG.addToCallGraph(LocalTUDecls[i]);<br>
}<br>
<br>
@@ -414,13 +412,13 @@<br>
// entries. Thus we don't use an iterator, but rely on LocalTUDecls<br>
// random access. By doing so, we automatically compensate for iterators<br>
// possibly being invalidated, although this is a bit slower.<br>
- const unsigned n = LocalTUDecls.size();<br>
- for (unsigned i = 0 ; i < n ; ++i) {<br>
+ const unsigned LocalTUDeclsSize = LocalTUDecls.size();<br>
+ for (unsigned i = 0 ; i < LocalTUDeclsSize ; ++i) {<br>
TraverseDecl(LocalTUDecls[i]);<br>
}<br>
<br>
if (Mgr->shouldInlineCall())<br>
- HandleDeclsGallGraph();<br>
+ HandleDeclsGallGraph(LocalTUDeclsSize);<br>
<br>
// After all decls handled, run checkers on the entire TranslationUnit.<br>
checkerMgr->runCheckersOnEndOfTranslationUnit(TU, *Mgr, BR);<br>
<br>
<br>
_______________________________________________<br>
cfe-commits mailing list<br>
<a href="mailto:cfe-commits@cs.uiuc.edu">cfe-commits@cs.uiuc.edu</a><br>
<a href="http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits" target="_blank">http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits</a><br>
</blockquote></div><br>
</blockquote></div><br></body></html>