Mercurial > hg > truffle
view graal/com.oracle.graal.hotspot.hsail/src/com/oracle/graal/hotspot/hsail/HSAILHotSpotLIRGenerator.java @ 13334:39694f3b0ed9
Truffle: set default TruffleInliningMaxCallerSize to 2250.
author | Christian Humer <christian.humer@gmail.com> |
---|---|
date | Sun, 15 Dec 2013 21:43:35 +0100 |
parents | 01fd11e44f73 |
children | bc471f405eb8 |
line wrap: on
line source
/* * Copyright (c) 2013, Oracle and/or its affiliates. All rights reserved. * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER. * * This code is free software; you can redistribute it and/or modify it * under the terms of the GNU General Public License version 2 only, as * published by the Free Software Foundation. * * This code is distributed in the hope that it will be useful, but WITHOUT * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License * version 2 for more details (a copy is included in the LICENSE file that * accompanied this code). * * You should have received a copy of the GNU General Public License version * 2 along with this work; if not, write to the Free Software Foundation, * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. * * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA * or visit www.oracle.com if you need additional information or have any * questions. */ package com.oracle.graal.hotspot.hsail; import sun.misc.*; import com.oracle.graal.api.code.*; import com.oracle.graal.api.meta.*; import com.oracle.graal.compiler.hsail.*; import com.oracle.graal.hotspot.*; import com.oracle.graal.lir.*; import com.oracle.graal.lir.hsail.*; import com.oracle.graal.lir.hsail.HSAILControlFlow.*; import com.oracle.graal.lir.hsail.HSAILMove.*; import com.oracle.graal.nodes.*; import com.oracle.graal.nodes.calc.*; import com.oracle.graal.nodes.extended.*; import com.oracle.graal.nodes.java.*; import com.oracle.graal.phases.util.*; /** * The HotSpot specific portion of the HSAIL LIR generator. */ public class HSAILHotSpotLIRGenerator extends HSAILLIRGenerator { private final HotSpotVMConfig config; public HSAILHotSpotLIRGenerator(StructuredGraph graph, Providers providers, HotSpotVMConfig config, FrameMap frameMap, CallingConvention cc, LIR lir) { super(graph, providers, frameMap, cc, lir); this.config = config; } private int getLogMinObjectAlignment() { return config.logMinObjAlignment(); } private int getNarrowOopShift() { return config.narrowOopShift; } private long getNarrowOopBase() { return config.narrowOopBase; } private int getLogKlassAlignment() { return config.logKlassAlignment; } private int getNarrowKlassShift() { return config.narrowKlassShift; } private long getNarrowKlassBase() { return config.narrowKlassBase; } private static boolean isCompressCandidate(Access access) { return access != null && access.isCompressible(); } /** * Appends either a {@link CompareAndSwapOp} or a {@link CompareAndSwapCompressedOp} depending * on whether the memory location of a given {@link LoweredCompareAndSwapNode} contains a * compressed oop. For the {@link CompareAndSwapCompressedOp} case, allocates a number of * scratch registers. The result {@link #operand(ValueNode) operand} for {@code node} complies * with the API for {@link Unsafe#compareAndSwapInt(Object, long, int, int)}. * * @param address the memory location targeted by the operation */ @Override public void visitCompareAndSwap(LoweredCompareAndSwapNode node, Value address) { Kind kind = node.getNewValue().kind(); assert kind == node.getExpectedValue().kind(); Variable expected = load(operand(node.getExpectedValue())); Variable newValue = load(operand(node.getNewValue())); HSAILAddressValue addressValue = asAddressValue(address); Variable casResult = newVariable(kind); if (config.useCompressedOops && node.isCompressible()) { // make 64-bit scratch variables for expected and new Variable scratchExpected64 = newVariable(Kind.Long); Variable scratchNewValue64 = newVariable(Kind.Long); // make 32-bit scratch variables for expected and new and result Variable scratchExpected32 = newVariable(Kind.Int); Variable scratchNewValue32 = newVariable(Kind.Int); Variable scratchCasResult32 = newVariable(Kind.Int); append(new CompareAndSwapCompressedOp(casResult, addressValue, expected, newValue, scratchExpected64, scratchNewValue64, scratchExpected32, scratchNewValue32, scratchCasResult32, getNarrowOopBase(), getNarrowOopShift(), getLogMinObjectAlignment())); } else { append(new CompareAndSwapOp(casResult, addressValue, expected, newValue)); } Variable nodeResult = newVariable(node.kind()); append(new CondMoveOp(mapKindToCompareOp(kind), casResult, expected, nodeResult, Condition.EQ, Constant.INT_1, Constant.INT_0)); setResult(node, nodeResult); } @Override public Variable emitLoad(Kind kind, Value address, Access access) { HSAILAddressValue loadAddress = asAddressValue(address); Variable result = newVariable(kind); LIRFrameState state = null; if (access instanceof DeoptimizingNode) { state = state((DeoptimizingNode) access); } if (isCompressCandidate(access) && config.useCompressedOops && kind == Kind.Object) { Variable scratch = newVariable(Kind.Long); append(new LoadCompressedPointer(kind, result, scratch, loadAddress, state, getNarrowOopBase(), getNarrowOopShift(), getLogMinObjectAlignment())); } else if (isCompressCandidate(access) && config.useCompressedClassPointers && kind == Kind.Long) { Variable scratch = newVariable(Kind.Long); append(new LoadCompressedPointer(kind, result, scratch, loadAddress, state, getNarrowKlassBase(), getNarrowKlassShift(), getLogKlassAlignment())); } else { append(new LoadOp(kind, result, loadAddress, state)); } return result; } @Override public void emitStore(Kind kind, Value address, Value inputVal, Access access) { HSAILAddressValue storeAddress = asAddressValue(address); LIRFrameState state = null; if (access instanceof DeoptimizingNode) { state = state((DeoptimizingNode) access); } Variable input = load(inputVal); if (isCompressCandidate(access) && config.useCompressedOops && kind == Kind.Object) { Variable scratch = newVariable(Kind.Long); append(new StoreCompressedPointer(kind, storeAddress, input, scratch, state, getNarrowOopBase(), getNarrowOopShift(), getLogMinObjectAlignment())); } else if (isCompressCandidate(access) && config.useCompressedClassPointers && kind == Kind.Long) { Variable scratch = newVariable(Kind.Long); append(new StoreCompressedPointer(kind, storeAddress, input, scratch, state, getNarrowKlassBase(), getNarrowKlassShift(), getLogKlassAlignment())); } else { append(new StoreOp(kind, storeAddress, input, state)); } } /*** * This is a very temporary solution to emitForeignCall. We don't really support foreign calls * yet, but we do want to generate dummy code for them. The ForeignCallXXXOps just end up * emitting a comment as to what Foreign call they would have made. **/ @Override public Variable emitForeignCall(ForeignCallLinkage linkage, DeoptimizingNode info, Value... args) { Variable result = newVariable(Kind.Object); // linkage.getDescriptor().getResultType()); // to make the LIRVerifier happy, we move any constants into registers Value[] argLocations = new Value[args.length]; for (int i = 0; i < args.length; i++) { Value arg = args[i]; AllocatableValue loc = newVariable(arg.getKind()); emitMove(loc, arg); argLocations[i] = loc; } // here we could check the callName if we wanted to only handle certain callnames String callName = linkage.getDescriptor().getName(); switch (argLocations.length) { case 0: append(new ForeignCallNoArgOp(callName, result)); break; case 1: append(new ForeignCall1ArgOp(callName, result, argLocations[0])); break; case 2: append(new ForeignCall2ArgOp(callName, result, argLocations[0], argLocations[1])); break; default: throw new InternalError("NYI emitForeignCall " + callName + ", " + argLocations.length + ", " + linkage); } return result; } @Override protected void emitForeignCall(ForeignCallLinkage linkage, Value result, Value[] arguments, Value[] temps, LIRFrameState info) { // this version of emitForeignCall not used for now } }