<!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>[198010] releases/WebKitGTK/webkit-2.12/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/198010">198010</a></dd>
<dt>Author</dt> <dd>carlosgc@webkit.org</dd>
<dt>Date</dt> <dd>2016-03-11 05:49:14 -0800 (Fri, 11 Mar 2016)</dd>
</dl>
<h3>Log Message</h3>
<pre>Merge <a href="http://trac.webkit.org/projects/webkit/changeset/197655">r197655</a> - [JSC] Improve DFG's Int32 ArithMul if one operand is a constant
https://bugs.webkit.org/show_bug.cgi?id=155066
Reviewed by Filip Pizlo.
When multiplying an integer by a constant, DFG was doing quite
a bit worse than baseline JIT.
We were loading the constant into a register, doing the multiply,
the checking the result and both operands for negative zero.
This patch changes:
-Use the multiply-by-immediate form on x86.
-Do as few checks as possible to detect negative-zero.
In most cases, this reduce the negative-zero checks
to zero or one TEST+JUMP.
* assembler/MacroAssembler.h:
(JSC::MacroAssembler::mul32):
* dfg/DFGSpeculativeJIT.cpp:
(JSC::DFG::SpeculativeJIT::compileArithMul):</pre>
<h3>Modified Paths</h3>
<ul>
<li><a href="#releasesWebKitGTKwebkit212SourceJavaScriptCoreChangeLog">releases/WebKitGTK/webkit-2.12/Source/JavaScriptCore/ChangeLog</a></li>
<li><a href="#releasesWebKitGTKwebkit212SourceJavaScriptCoreassemblerMacroAssemblerh">releases/WebKitGTK/webkit-2.12/Source/JavaScriptCore/assembler/MacroAssembler.h</a></li>
<li><a href="#releasesWebKitGTKwebkit212SourceJavaScriptCoredfgDFGSpeculativeJITcpp">releases/WebKitGTK/webkit-2.12/Source/JavaScriptCore/dfg/DFGSpeculativeJIT.cpp</a></li>
</ul>
</div>
<div id="patch">
<h3>Diff</h3>
<a id="releasesWebKitGTKwebkit212SourceJavaScriptCoreChangeLog"></a>
<div class="modfile"><h4>Modified: releases/WebKitGTK/webkit-2.12/Source/JavaScriptCore/ChangeLog (198009 => 198010)</h4>
<pre class="diff"><span>
<span class="info">--- releases/WebKitGTK/webkit-2.12/Source/JavaScriptCore/ChangeLog 2016-03-11 12:59:31 UTC (rev 198009)
+++ releases/WebKitGTK/webkit-2.12/Source/JavaScriptCore/ChangeLog 2016-03-11 13:49:14 UTC (rev 198010)
</span><span class="lines">@@ -1,5 +1,29 @@
</span><span class="cx"> 2016-03-06 Benjamin Poulain <benjamin@webkit.org>
</span><span class="cx">
</span><ins>+ [JSC] Improve DFG's Int32 ArithMul if one operand is a constant
+ https://bugs.webkit.org/show_bug.cgi?id=155066
+
+ Reviewed by Filip Pizlo.
+
+ When multiplying an integer by a constant, DFG was doing quite
+ a bit worse than baseline JIT.
+ We were loading the constant into a register, doing the multiply,
+ the checking the result and both operands for negative zero.
+
+ This patch changes:
+ -Use the multiply-by-immediate form on x86.
+ -Do as few checks as possible to detect negative-zero.
+
+ In most cases, this reduce the negative-zero checks
+ to zero or one TEST+JUMP.
+
+ * assembler/MacroAssembler.h:
+ (JSC::MacroAssembler::mul32):
+ * dfg/DFGSpeculativeJIT.cpp:
+ (JSC::DFG::SpeculativeJIT::compileArithMul):
+
+2016-03-06 Benjamin Poulain <benjamin@webkit.org>
+
</ins><span class="cx"> [JSC] Remove a superfluous Move in front of every double unboxing
</span><span class="cx"> https://bugs.webkit.org/show_bug.cgi?id=155064
</span><span class="cx">
</span></span></pre></div>
<a id="releasesWebKitGTKwebkit212SourceJavaScriptCoreassemblerMacroAssemblerh"></a>
<div class="modfile"><h4>Modified: releases/WebKitGTK/webkit-2.12/Source/JavaScriptCore/assembler/MacroAssembler.h (198009 => 198010)</h4>
<pre class="diff"><span>
<span class="info">--- releases/WebKitGTK/webkit-2.12/Source/JavaScriptCore/assembler/MacroAssembler.h 2016-03-11 12:59:31 UTC (rev 198009)
+++ releases/WebKitGTK/webkit-2.12/Source/JavaScriptCore/assembler/MacroAssembler.h 2016-03-11 13:49:14 UTC (rev 198010)
</span><span class="lines">@@ -115,6 +115,7 @@
</span><span class="cx"> using MacroAssemblerBase::compare32;
</span><span class="cx"> using MacroAssemblerBase::move;
</span><span class="cx"> using MacroAssemblerBase::add32;
</span><ins>+ using MacroAssemblerBase::mul32;
</ins><span class="cx"> using MacroAssemblerBase::and32;
</span><span class="cx"> using MacroAssemblerBase::branchAdd32;
</span><span class="cx"> using MacroAssemblerBase::branchMul32;
</span><span class="lines">@@ -1488,6 +1489,27 @@
</span><span class="cx"> addPtr(imm.asTrustedImm32(), dest);
</span><span class="cx"> }
</span><span class="cx">
</span><ins>+ void mul32(Imm32 imm, RegisterID src, RegisterID dest)
+ {
+ if (shouldBlind(imm)) {
+ if (src != dest || haveScratchRegisterForBlinding()) {
+ if (src == dest) {
+ move(src, scratchRegisterForBlinding());
+ src = scratchRegisterForBlinding();
+ }
+ loadXorBlindedConstant(xorBlindConstant(imm), dest);
+ mul32(src, dest);
+ return;
+ }
+ // If we don't have a scratch register available for use, we'll just
+ // place a random number of nops.
+ uint32_t nopCount = random() & 3;
+ while (nopCount--)
+ nop();
+ }
+ mul32(imm.asTrustedImm32(), src, dest);
+ }
+
</ins><span class="cx"> void and32(Imm32 imm, RegisterID dest)
</span><span class="cx"> {
</span><span class="cx"> if (shouldBlind(imm)) {
</span></span></pre></div>
<a id="releasesWebKitGTKwebkit212SourceJavaScriptCoredfgDFGSpeculativeJITcpp"></a>
<div class="modfile"><h4>Modified: releases/WebKitGTK/webkit-2.12/Source/JavaScriptCore/dfg/DFGSpeculativeJIT.cpp (198009 => 198010)</h4>
<pre class="diff"><span>
<span class="info">--- releases/WebKitGTK/webkit-2.12/Source/JavaScriptCore/dfg/DFGSpeculativeJIT.cpp 2016-03-11 12:59:31 UTC (rev 198009)
+++ releases/WebKitGTK/webkit-2.12/Source/JavaScriptCore/dfg/DFGSpeculativeJIT.cpp 2016-03-11 13:49:14 UTC (rev 198010)
</span><span class="lines">@@ -3658,6 +3658,38 @@
</span><span class="cx"> {
</span><span class="cx"> switch (node->binaryUseKind()) {
</span><span class="cx"> case Int32Use: {
</span><ins>+ if (node->child2()->isInt32Constant()) {
+ SpeculateInt32Operand op1(this, node->child1());
+ GPRTemporary result(this);
+
+ int32_t imm = node->child2()->asInt32();
+ GPRReg op1GPR = op1.gpr();
+ GPRReg resultGPR = result.gpr();
+
+ if (!shouldCheckOverflow(node->arithMode()))
+ m_jit.mul32(Imm32(imm), op1GPR, resultGPR);
+ else {
+ speculationCheck(Overflow, JSValueRegs(), 0,
+ m_jit.branchMul32(MacroAssembler::Overflow, op1GPR, Imm32(imm), resultGPR));
+ }
+
+ // The only way to create negative zero with a constant is:
+ // -negative-op1 * 0.
+ // -zero-op1 * negative constant.
+ if (shouldCheckNegativeZero(node->arithMode())) {
+ if (!imm)
+ speculationCheck(NegativeZero, JSValueRegs(), 0, m_jit.branchTest32(MacroAssembler::Signed, op1GPR));
+ else if (imm < 0) {
+ if (shouldCheckOverflow(node->arithMode()))
+ speculationCheck(NegativeZero, JSValueRegs(), 0, m_jit.branchTest32(MacroAssembler::Zero, resultGPR));
+ else
+ speculationCheck(NegativeZero, JSValueRegs(), 0, m_jit.branchTest32(MacroAssembler::Zero, op1GPR));
+ }
+ }
+
+ int32Result(resultGPR, node);
+ return;
+ }
</ins><span class="cx"> SpeculateInt32Operand op1(this, node->child1());
</span><span class="cx"> SpeculateInt32Operand op2(this, node->child2());
</span><span class="cx"> GPRTemporary result(this);
</span><span class="lines">@@ -3680,16 +3712,16 @@
</span><span class="cx"> // Check for negative zero, if the users of this node care about such things.
</span><span class="cx"> if (shouldCheckNegativeZero(node->arithMode())) {
</span><span class="cx"> MacroAssembler::Jump resultNonZero = m_jit.branchTest32(MacroAssembler::NonZero, result.gpr());
</span><del>- speculationCheck(NegativeZero, JSValueRegs(), 0, m_jit.branch32(MacroAssembler::LessThan, reg1, TrustedImm32(0)));
- speculationCheck(NegativeZero, JSValueRegs(), 0, m_jit.branch32(MacroAssembler::LessThan, reg2, TrustedImm32(0)));
</del><ins>+ speculationCheck(NegativeZero, JSValueRegs(), 0, m_jit.branchTest32(MacroAssembler::Signed, reg1));
+ speculationCheck(NegativeZero, JSValueRegs(), 0, m_jit.branchTest32(MacroAssembler::Signed, reg2));
</ins><span class="cx"> resultNonZero.link(&m_jit);
</span><span class="cx"> }
</span><span class="cx">
</span><span class="cx"> int32Result(result.gpr(), node);
</span><span class="cx"> return;
</span><span class="cx"> }
</span><del>-
-#if USE(JSVALUE64)
</del><ins>+
+#if USE(JSVALUE64)
</ins><span class="cx"> case Int52RepUse: {
</span><span class="cx"> ASSERT(shouldCheckOverflow(node->arithMode()));
</span><span class="cx">
</span></span></pre>
</div>
</div>
</body>
</html>