<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN"
"http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head><meta http-equiv="content-type" content="text/html; charset=utf-8" />
<title>[181456] trunk/Source/JavaScriptCore</title>
</head>
<body>

<style type="text/css"><!--
#msg dl.meta { border: 1px #006 solid; background: #369; padding: 6px; color: #fff; }
#msg dl.meta dt { float: left; width: 6em; font-weight: bold; }
#msg dt:after { content:':';}
#msg dl, #msg dt, #msg ul, #msg li, #header, #footer, #logmsg { font-family: verdana,arial,helvetica,sans-serif; font-size: 10pt;  }
#msg dl a { font-weight: bold}
#msg dl a:link    { color:#fc3; }
#msg dl a:active  { color:#ff0; }
#msg dl a:visited { color:#cc6; }
h3 { font-family: verdana,arial,helvetica,sans-serif; font-size: 10pt; font-weight: bold; }
#msg pre { overflow: auto; background: #ffc; border: 1px #fa0 solid; padding: 6px; }
#logmsg { background: #ffc; border: 1px #fa0 solid; padding: 1em 1em 0 1em; }
#logmsg p, #logmsg pre, #logmsg blockquote { margin: 0 0 1em 0; }
#logmsg p, #logmsg li, #logmsg dt, #logmsg dd { line-height: 14pt; }
#logmsg h1, #logmsg h2, #logmsg h3, #logmsg h4, #logmsg h5, #logmsg h6 { margin: .5em 0; }
#logmsg h1:first-child, #logmsg h2:first-child, #logmsg h3:first-child, #logmsg h4:first-child, #logmsg h5:first-child, #logmsg h6:first-child { margin-top: 0; }
#logmsg ul, #logmsg ol { padding: 0; list-style-position: inside; margin: 0 0 0 1em; }
#logmsg ul { text-indent: -1em; padding-left: 1em; }#logmsg ol { text-indent: -1.5em; padding-left: 1.5em; }
#logmsg > ul, #logmsg > ol { margin: 0 0 1em 0; }
#logmsg pre { background: #eee; padding: 1em; }
#logmsg blockquote { border: 1px solid #fa0; border-left-width: 10px; padding: 1em 1em 0 1em; background: white;}
#logmsg dl { margin: 0; }
#logmsg dt { font-weight: bold; }
#logmsg dd { margin: 0; padding: 0 0 0.5em 0; }
#logmsg dd:before { content:'\00bb';}
#logmsg table { border-spacing: 0px; border-collapse: collapse; border-top: 4px solid #fa0; border-bottom: 1px solid #fa0; background: #fff; }
#logmsg table th { text-align: left; font-weight: normal; padding: 0.2em 0.5em; border-top: 1px dotted #fa0; }
#logmsg table td { text-align: right; border-top: 1px dotted #fa0; padding: 0.2em 0.5em; }
#logmsg table thead th { text-align: center; border-bottom: 1px solid #fa0; }
#logmsg table th.Corner { text-align: left; }
#logmsg hr { border: none 0; border-top: 2px dashed #fa0; height: 1px; }
#header, #footer { color: #fff; background: #636; border: 1px #300 solid; padding: 6px; }
#patch { width: 100%; }
#patch h4 {font-family: verdana,arial,helvetica,sans-serif;font-size:10pt;padding:8px;background:#369;color:#fff;margin:0;}
#patch .propset h4, #patch .binary h4 {margin:0;}
#patch pre {padding:0;line-height:1.2em;margin:0;}
#patch .diff {width:100%;background:#eee;padding: 0 0 10px 0;overflow:auto;}
#patch .propset .diff, #patch .binary .diff  {padding:10px 0;}
#patch span {display:block;padding:0 10px;}
#patch .modfile, #patch .addfile, #patch .delfile, #patch .propset, #patch .binary, #patch .copfile {border:1px solid #ccc;margin:10px 0;}
#patch ins {background:#dfd;text-decoration:none;display:block;padding:0 10px;}
#patch del {background:#fdd;text-decoration:none;display:block;padding:0 10px;}
#patch .lines, .info {color:#888;background:#fff;}
--></style>
<div id="msg">
<dl class="meta">
<dt>Revision</dt> <dd><a href="http://trac.webkit.org/projects/webkit/changeset/181456">181456</a></dd>
<dt>Author</dt> <dd>mark.lam@apple.com</dd>
<dt>Date</dt> <dd>2015-03-12 15:14:32 -0700 (Thu, 12 Mar 2015)</dd>
</dl>

<h3>Log Message</h3>
<pre>Use std::atomic for CodeBlock::m_visitAggregateHasBeenCalled.
&lt;https://webkit.org/b/142640&gt;

Reviewed by Mark Hahnenberg.

We used to spin our own compare and swap on a uint8_t.  Now that we can
use C++11, let's use std::atomic instead.

* bytecode/CodeBlock.cpp:
(JSC::CodeBlock::visitAggregate):
- The CAS here needs std::memory_order_acquire ordering because it
  requires lock acquisition semantics to visit the CodeBlock.

* bytecode/CodeBlock.h:
(JSC::CodeBlockSet::mark):
* heap/CodeBlockSet.cpp:
(JSC::CodeBlockSet::clearMarksForFullCollection):
(JSC::CodeBlockSet::clearMarksForEdenCollection):
- These can go with relaxed ordering because they are all done before
  the GC starts parallel marking.</pre>

<h3>Modified Paths</h3>
<ul>
<li><a href="#trunkSourceJavaScriptCoreChangeLog">trunk/Source/JavaScriptCore/ChangeLog</a></li>
<li><a href="#trunkSourceJavaScriptCorebytecodeCodeBlockcpp">trunk/Source/JavaScriptCore/bytecode/CodeBlock.cpp</a></li>
<li><a href="#trunkSourceJavaScriptCorebytecodeCodeBlockh">trunk/Source/JavaScriptCore/bytecode/CodeBlock.h</a></li>
<li><a href="#trunkSourceJavaScriptCoreheapCodeBlockSetcpp">trunk/Source/JavaScriptCore/heap/CodeBlockSet.cpp</a></li>
</ul>

</div>
<div id="patch">
<h3>Diff</h3>
<a id="trunkSourceJavaScriptCoreChangeLog"></a>
<div class="modfile"><h4>Modified: trunk/Source/JavaScriptCore/ChangeLog (181455 => 181456)</h4>
<pre class="diff"><span>
<span class="info">--- trunk/Source/JavaScriptCore/ChangeLog        2015-03-12 21:01:37 UTC (rev 181455)
+++ trunk/Source/JavaScriptCore/ChangeLog        2015-03-12 22:14:32 UTC (rev 181456)
</span><span class="lines">@@ -1,3 +1,26 @@
</span><ins>+2015-03-12  Mark Lam  &lt;mark.lam@apple.com&gt;
+
+        Use std::atomic for CodeBlock::m_visitAggregateHasBeenCalled.
+        &lt;https://webkit.org/b/142640&gt;
+
+        Reviewed by Mark Hahnenberg.
+
+        We used to spin our own compare and swap on a uint8_t.  Now that we can
+        use C++11, let's use std::atomic instead.
+
+        * bytecode/CodeBlock.cpp:
+        (JSC::CodeBlock::visitAggregate):
+        - The CAS here needs std::memory_order_acquire ordering because it
+          requires lock acquisition semantics to visit the CodeBlock.
+
+        * bytecode/CodeBlock.h:
+        (JSC::CodeBlockSet::mark):
+        * heap/CodeBlockSet.cpp:
+        (JSC::CodeBlockSet::clearMarksForFullCollection):
+        (JSC::CodeBlockSet::clearMarksForEdenCollection):
+        - These can go with relaxed ordering because they are all done before
+          the GC starts parallel marking.
+
</ins><span class="cx"> 2015-03-12  Csaba Osztrogonác  &lt;ossy@webkit.org&gt;
</span><span class="cx"> 
</span><span class="cx">         [cmake] Fix the incremental build issue revealed by r181419
</span></span></pre></div>
<a id="trunkSourceJavaScriptCorebytecodeCodeBlockcpp"></a>
<div class="modfile"><h4>Modified: trunk/Source/JavaScriptCore/bytecode/CodeBlock.cpp (181455 => 181456)</h4>
<pre class="diff"><span>
<span class="info">--- trunk/Source/JavaScriptCore/bytecode/CodeBlock.cpp        2015-03-12 21:01:37 UTC (rev 181455)
+++ trunk/Source/JavaScriptCore/bytecode/CodeBlock.cpp        2015-03-12 22:14:32 UTC (rev 181456)
</span><span class="lines">@@ -2195,26 +2195,12 @@
</span><span class="cx"> {
</span><span class="cx"> #if ENABLE(PARALLEL_GC)
</span><span class="cx">     // I may be asked to scan myself more than once, and it may even happen concurrently.
</span><del>-    // To this end, use a CAS loop to check if I've been called already. Only one thread
-    // may proceed past this point - whichever one wins the CAS race.
-    unsigned oldValue;
-    do {
-        oldValue = m_visitAggregateHasBeenCalled;
-        if (oldValue) {
-            // Looks like someone else won! Return immediately to ensure that we don't
-            // trace the same CodeBlock concurrently. Doing so is hazardous since we will
-            // be mutating the state of ValueProfiles, which contain JSValues, which can
-            // have word-tearing on 32-bit, leading to awesome timing-dependent crashes
-            // that are nearly impossible to track down.
-            
-            // Also note that it must be safe to return early as soon as we see the
-            // value true (well, (unsigned)1), since once a GC thread is in this method
-            // and has won the CAS race (i.e. was responsible for setting the value true)
-            // it will definitely complete the rest of this method before declaring
-            // termination.
-            return;
-        }
-    } while (!WTF::weakCompareAndSwap(&amp;m_visitAggregateHasBeenCalled, 0, 1));
</del><ins>+    // To this end, use an atomic operation to check (and set) if I've been called already.
+    // Only one thread may proceed past this point - whichever one wins the atomic set race.
+    bool expected = false;
+    bool setByMe = m_visitAggregateHasBeenCalled.compare_exchange_strong(expected, true, std::memory_order_acquire);
+    if (!setByMe)
+        return;
</ins><span class="cx"> #endif // ENABLE(PARALLEL_GC)
</span><span class="cx">     
</span><span class="cx">     if (!!m_alternative)
</span></span></pre></div>
<a id="trunkSourceJavaScriptCorebytecodeCodeBlockh"></a>
<div class="modfile"><h4>Modified: trunk/Source/JavaScriptCore/bytecode/CodeBlock.h (181455 => 181456)</h4>
<pre class="diff"><span>
<span class="info">--- trunk/Source/JavaScriptCore/bytecode/CodeBlock.h        2015-03-12 21:01:37 UTC (rev 181455)
+++ trunk/Source/JavaScriptCore/bytecode/CodeBlock.h        2015-03-12 22:14:32 UTC (rev 181456)
</span><span class="lines">@@ -1065,7 +1065,7 @@
</span><span class="cx">     bool m_isStrictMode;
</span><span class="cx">     bool m_needsActivation;
</span><span class="cx">     bool m_mayBeExecuting;
</span><del>-    uint8_t m_visitAggregateHasBeenCalled;
</del><ins>+    std::atomic&lt;bool&gt; m_visitAggregateHasBeenCalled;
</ins><span class="cx"> 
</span><span class="cx">     RefPtr&lt;SourceProvider&gt; m_source;
</span><span class="cx">     unsigned m_sourceOffset;
</span><span class="lines">@@ -1301,7 +1301,7 @@
</span><span class="cx">     
</span><span class="cx">     codeBlock-&gt;m_mayBeExecuting = true;
</span><span class="cx">     // We might not have cleared the marks for this CodeBlock, but we need to visit it.
</span><del>-    codeBlock-&gt;m_visitAggregateHasBeenCalled = false;
</del><ins>+    codeBlock-&gt;m_visitAggregateHasBeenCalled.store(false, std::memory_order_relaxed);
</ins><span class="cx"> #if ENABLE(GGC)
</span><span class="cx">     m_currentlyExecuting.append(codeBlock);
</span><span class="cx"> #endif
</span></span></pre></div>
<a id="trunkSourceJavaScriptCoreheapCodeBlockSetcpp"></a>
<div class="modfile"><h4>Modified: trunk/Source/JavaScriptCore/heap/CodeBlockSet.cpp (181455 => 181456)</h4>
<pre class="diff"><span>
<span class="info">--- trunk/Source/JavaScriptCore/heap/CodeBlockSet.cpp        2015-03-12 21:01:37 UTC (rev 181455)
+++ trunk/Source/JavaScriptCore/heap/CodeBlockSet.cpp        2015-03-12 22:14:32 UTC (rev 181456)
</span><span class="lines">@@ -65,7 +65,7 @@
</span><span class="cx"> {
</span><span class="cx">     for (CodeBlock* codeBlock : m_oldCodeBlocks) {
</span><span class="cx">         codeBlock-&gt;m_mayBeExecuting = false;
</span><del>-        codeBlock-&gt;m_visitAggregateHasBeenCalled = false;
</del><ins>+        codeBlock-&gt;m_visitAggregateHasBeenCalled.store(false, std::memory_order_relaxed);
</ins><span class="cx">     }
</span><span class="cx"> 
</span><span class="cx">     // We promote after we clear marks on the old generation CodeBlocks because
</span><span class="lines">@@ -82,7 +82,7 @@
</span><span class="cx">             continue;
</span><span class="cx">         executable-&gt;forEachCodeBlock([](CodeBlock* codeBlock) {
</span><span class="cx">             codeBlock-&gt;m_mayBeExecuting = false;
</span><del>-            codeBlock-&gt;m_visitAggregateHasBeenCalled = false;
</del><ins>+            codeBlock-&gt;m_visitAggregateHasBeenCalled.store(false, std::memory_order_relaxed);
</ins><span class="cx">         });
</span><span class="cx">     }
</span><span class="cx"> }
</span></span></pre>
</div>
</div>

</body>
</html>