<!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>[169787] 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/169787">169787</a></dd>
<dt>Author</dt> <dd>fpizlo@apple.com</dd>
<dt>Date</dt> <dd>2014-06-10 18:53:19 -0700 (Tue, 10 Jun 2014)</dd>
</dl>
<h3>Log Message</h3>
<pre>Prediction propagator should make sure everyone knows that a variable that is in an argument position where other versions of that variable are not MachineInts cannot possibly be flushed as Int52
https://bugs.webkit.org/show_bug.cgi?id=133698
Reviewed by Geoffrey Garen and Mark Hahnenberg.
* dfg/DFGPredictionPropagationPhase.cpp:
(JSC::DFG::PredictionPropagationPhase::propagate): Use the new utility to figure out if a variable could ever represent an Int52.
* dfg/DFGVariableAccessData.cpp:
(JSC::DFG::VariableAccessData::couldRepresentInt52): Add a new utility to detect early on if a variable could possibly be Int52.
(JSC::DFG::VariableAccessData::couldRepresentInt52Impl):
(JSC::DFG::VariableAccessData::flushFormat):
* dfg/DFGVariableAccessData.h:
* tests/stress/int52-inlined-call-argument.js: Added.
(foo):
(bar):</pre>
<h3>Modified Paths</h3>
<ul>
<li><a href="#trunkSourceJavaScriptCoreChangeLog">trunk/Source/JavaScriptCore/ChangeLog</a></li>
<li><a href="#trunkSourceJavaScriptCoredfgDFGPredictionPropagationPhasecpp">trunk/Source/JavaScriptCore/dfg/DFGPredictionPropagationPhase.cpp</a></li>
<li><a href="#trunkSourceJavaScriptCoredfgDFGVariableAccessDatacpp">trunk/Source/JavaScriptCore/dfg/DFGVariableAccessData.cpp</a></li>
<li><a href="#trunkSourceJavaScriptCoredfgDFGVariableAccessDatah">trunk/Source/JavaScriptCore/dfg/DFGVariableAccessData.h</a></li>
</ul>
<h3>Added Paths</h3>
<ul>
<li><a href="#trunkSourceJavaScriptCoretestsstressint52inlinedcallargumentjs">trunk/Source/JavaScriptCore/tests/stress/int52-inlined-call-argument.js</a></li>
</ul>
</div>
<div id="patch">
<h3>Diff</h3>
<a id="trunkSourceJavaScriptCoreChangeLog"></a>
<div class="modfile"><h4>Modified: trunk/Source/JavaScriptCore/ChangeLog (169786 => 169787)</h4>
<pre class="diff"><span>
<span class="info">--- trunk/Source/JavaScriptCore/ChangeLog        2014-06-11 01:51:56 UTC (rev 169786)
+++ trunk/Source/JavaScriptCore/ChangeLog        2014-06-11 01:53:19 UTC (rev 169787)
</span><span class="lines">@@ -1,3 +1,21 @@
</span><ins>+2014-06-10 Filip Pizlo <fpizlo@apple.com>
+
+ Prediction propagator should make sure everyone knows that a variable that is in an argument position where other versions of that variable are not MachineInts cannot possibly be flushed as Int52
+ https://bugs.webkit.org/show_bug.cgi?id=133698
+
+ Reviewed by Geoffrey Garen and Mark Hahnenberg.
+
+ * dfg/DFGPredictionPropagationPhase.cpp:
+ (JSC::DFG::PredictionPropagationPhase::propagate): Use the new utility to figure out if a variable could ever represent an Int52.
+ * dfg/DFGVariableAccessData.cpp:
+ (JSC::DFG::VariableAccessData::couldRepresentInt52): Add a new utility to detect early on if a variable could possibly be Int52.
+ (JSC::DFG::VariableAccessData::couldRepresentInt52Impl):
+ (JSC::DFG::VariableAccessData::flushFormat):
+ * dfg/DFGVariableAccessData.h:
+ * tests/stress/int52-inlined-call-argument.js: Added.
+ (foo):
+ (bar):
+
</ins><span class="cx"> 2014-06-10 Mark Lam <mark.lam@apple.com>
</span><span class="cx">
</span><span class="cx"> Assertion failure at JSC::Structure::checkOffsetConsistency() const + 234.
</span></span></pre></div>
<a id="trunkSourceJavaScriptCoredfgDFGPredictionPropagationPhasecpp"></a>
<div class="modfile"><h4>Modified: trunk/Source/JavaScriptCore/dfg/DFGPredictionPropagationPhase.cpp (169786 => 169787)</h4>
<pre class="diff"><span>
<span class="info">--- trunk/Source/JavaScriptCore/dfg/DFGPredictionPropagationPhase.cpp        2014-06-11 01:51:56 UTC (rev 169786)
+++ trunk/Source/JavaScriptCore/dfg/DFGPredictionPropagationPhase.cpp        2014-06-11 01:53:19 UTC (rev 169787)
</span><span class="lines">@@ -153,8 +153,7 @@
</span><span class="cx"> case GetLocal: {
</span><span class="cx"> VariableAccessData* variable = node->variableAccessData();
</span><span class="cx"> SpeculatedType prediction = variable->prediction();
</span><del>- if ((variable->shouldNeverUnbox() || variable->local().isArgument())
- && (prediction & SpecInt52))
</del><ins>+ if (!variable->couldRepresentInt52() && (prediction & SpecInt52))
</ins><span class="cx"> prediction = (prediction | SpecInt52AsDouble) & ~SpecInt52;
</span><span class="cx"> if (prediction)
</span><span class="cx"> changed |= mergePrediction(prediction);
</span></span></pre></div>
<a id="trunkSourceJavaScriptCoredfgDFGVariableAccessDatacpp"></a>
<div class="modfile"><h4>Modified: trunk/Source/JavaScriptCore/dfg/DFGVariableAccessData.cpp (169786 => 169787)</h4>
<pre class="diff"><span>
<span class="info">--- trunk/Source/JavaScriptCore/dfg/DFGVariableAccessData.cpp        2014-06-11 01:51:56 UTC (rev 169786)
+++ trunk/Source/JavaScriptCore/dfg/DFGVariableAccessData.cpp        2014-06-11 01:53:19 UTC (rev 169787)
</span><span class="lines">@@ -171,6 +171,29 @@
</span><span class="cx"> return checkAndSet(m_prediction, type);
</span><span class="cx"> }
</span><span class="cx">
</span><ins>+bool VariableAccessData::couldRepresentInt52()
+{
+ if (shouldNeverUnbox())
+ return false;
+
+ return couldRepresentInt52Impl();
+}
+
+bool VariableAccessData::couldRepresentInt52Impl()
+{
+ // The hardware has to support it.
+ if (!enableInt52())
+ return false;
+
+ // We punt for machine arguments.
+ if (m_local.isArgument())
+ return false;
+
+ // The argument-aware prediction -- which merges all of an (inlined or machine)
+ // argument's variable access datas' predictions -- must possibly be MachineInt.
+ return !(argumentAwarePrediction() & ~SpecMachineInt);
+}
+
</ins><span class="cx"> FlushFormat VariableAccessData::flushFormat()
</span><span class="cx"> {
</span><span class="cx"> ASSERT(find() == this);
</span><span class="lines">@@ -185,10 +208,16 @@
</span><span class="cx"> return FlushedDouble;
</span><span class="cx">
</span><span class="cx"> SpeculatedType prediction = argumentAwarePrediction();
</span><ins>+
+ // This guard is here to protect the call to couldRepresentInt52(), which will return
+ // true for !prediction.
+ if (!prediction)
+ return FlushedJSValue;
+
</ins><span class="cx"> if (isInt32Speculation(prediction))
</span><span class="cx"> return FlushedInt32;
</span><span class="cx">
</span><del>- if (enableInt52() && !m_local.isArgument() && isMachineIntSpeculation(prediction))
</del><ins>+ if (couldRepresentInt52Impl())
</ins><span class="cx"> return FlushedInt52;
</span><span class="cx">
</span><span class="cx"> if (isCellSpeculation(prediction))
</span></span></pre></div>
<a id="trunkSourceJavaScriptCoredfgDFGVariableAccessDatah"></a>
<div class="modfile"><h4>Modified: trunk/Source/JavaScriptCore/dfg/DFGVariableAccessData.h (169786 => 169787)</h4>
<pre class="diff"><span>
<span class="info">--- trunk/Source/JavaScriptCore/dfg/DFGVariableAccessData.h        2014-06-11 01:51:56 UTC (rev 169786)
+++ trunk/Source/JavaScriptCore/dfg/DFGVariableAccessData.h        2014-06-11 01:53:19 UTC (rev 169787)
</span><span class="lines">@@ -212,12 +212,16 @@
</span><span class="cx">
</span><span class="cx"> FlushFormat flushFormat();
</span><span class="cx">
</span><ins>+ bool couldRepresentInt52();
+
</ins><span class="cx"> FlushedAt flushedAt()
</span><span class="cx"> {
</span><span class="cx"> return FlushedAt(flushFormat(), machineLocal());
</span><span class="cx"> }
</span><span class="cx">
</span><span class="cx"> private:
</span><ins>+ bool couldRepresentInt52Impl();
+
</ins><span class="cx"> // This is slightly space-inefficient, since anything we're unified with
</span><span class="cx"> // will have the same operand and should have the same prediction. But
</span><span class="cx"> // putting them here simplifies the code, and we don't expect DFG space
</span></span></pre></div>
<a id="trunkSourceJavaScriptCoretestsstressint52inlinedcallargumentjs"></a>
<div class="addfile"><h4>Added: trunk/Source/JavaScriptCore/tests/stress/int52-inlined-call-argument.js (0 => 169787)</h4>
<pre class="diff"><span>
<span class="info">--- trunk/Source/JavaScriptCore/tests/stress/int52-inlined-call-argument.js         (rev 0)
+++ trunk/Source/JavaScriptCore/tests/stress/int52-inlined-call-argument.js        2014-06-11 01:53:19 UTC (rev 169787)
</span><span class="lines">@@ -0,0 +1,19 @@
</span><ins>+function foo(a, b, c, p) {
+ a = b + c;
+ if (p) {
+ a++;
+ return a;
+ }
+}
+
+function bar(a, b) {
+ return foo("hello", a, b, a <= b);
+}
+
+noInline(bar);
+
+for (var i = 0; i < 100000; ++i) {
+ var result = bar(2000000000, 2000000000);
+ if (result != 4000000001)
+ throw "Error: bad result: " + result;
+}
</ins></span></pre>
</div>
</div>
</body>
</html>