Mercurial > hg > graal-jvmci-8
view graal/com.oracle.truffle.sl/src/com/oracle/truffle/sl/nodes/call/SLDispatchNode.java @ 20956:2170de9acab0
SL: use DSL for call dispatches.
author | Christian Humer <christian.humer@gmail.com> |
---|---|
date | Tue, 14 Apr 2015 15:16:14 +0200 |
parents | |
children | 99e3f4c5c853 |
line wrap: on
line source
/* * Copyright (c) 2013, 2015, 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.truffle.sl.nodes.call; import com.oracle.truffle.api.*; import com.oracle.truffle.api.dsl.*; import com.oracle.truffle.api.frame.*; import com.oracle.truffle.api.nodes.*; import com.oracle.truffle.sl.runtime.*; public abstract class SLDispatchNode extends Node { protected static final int INLINE_CACHE_SIZE = 2; public abstract Object executeDispatch(VirtualFrame frame, SLFunction function, Object[] arguments); /** * Inline cached specialization of the dispatch. * * <p> * Since SL is a quite simple language, the benefit of the inline cache is quite small: after * checking that the actual function to be executed is the same as the cachedFuntion, we can * safely execute the cached call target. You can reasonably argue that caching the call target * is overkill, since we could just retrieve it via {@code function.getCallTarget()}. However, * in a more complex language the lookup of the call target is usually much more complicated * than in SL. In addition, caching the call target allows method inlining. * </p> * * <p> * {@code limit = "INLINE_CACHE_SIZE"} Specifies the limit number of inline cache specialization * instantiations. * </p> * <p> * {@code guards = "function == cachedFunction"} The inline cache check. Note that * cachedFunction is a final field so that the compiler can optimize the check. * </p> * <p> * {@code assumptions = "cachedFunction.getCallTargetStable()"} Support for function * redefinition: When a function is redefined, the call target maintained by the SLFunction * object is change. To avoid a check for that, we use an Assumption that is invalidated by the * SLFunction when the change is performed. Since checking an assumption is a no-op in compiled * code, the assumption check performed by the DSL does not add any overhead during optimized * execution. * </p> * * @see Cached * @see Specialization * * @param function the dynamically provided function * @param cachedFunction the cached function of the specialization instance * @param callNode the {@link DirectCallNode} specifically created for the {@link CallTarget} in * cachedFunction. */ @Specialization(limit = "INLINE_CACHE_SIZE", guards = "function == cachedFunction", assumptions = "cachedFunction.getCallTargetStable()") protected static Object doDirect(VirtualFrame frame, SLFunction function, Object[] arguments, // @Cached("function") SLFunction cachedFunction, // @Cached("create(cachedFunction.getCallTarget())") DirectCallNode callNode) { /* Inline cache hit, we are safe to execute the cached call target. */ return callNode.call(frame, arguments); } /** * Slow-path code for a call, used when the polymorphic inline cache exceeded its maximum size * specified in <code>INLINE_CACHE_SIZE</code>. Such calls are not optimized any further, e.g., * no method inlining is performed. */ @Specialization(contains = "doDirect") protected static Object doIndirect(VirtualFrame frame, SLFunction function, Object[] arguments, // @Cached("create()") IndirectCallNode callNode) { /* * SL has a quite simple call lookup: just ask the function for the current call target, and * call it. */ return callNode.call(frame, function.getCallTarget(), arguments); } }