Mercurial > hg > icedtea8-forest > hotspot
view src/share/vm/opto/c2compiler.cpp @ 10905:f57189b7648d
8257192: Integrate AArch64 JIT port into 8u
7009641: Don't fail VM when CodeCache is full
8073108: [AArch64] Use x86 and SPARC CPU instructions for GHASH acceleration
8130309: Need to bailout cleanly if creation of stubs fails when codecache is out of space (AArch64 changes)
8131779: AARCH64: add Montgomery multiply intrinsic
8132875: AArch64: Fix error introduced into AArch64 CodeCache by commit for 8130309
8135018: AARCH64: Missing memory barriers for CMS collector
8145320: Create unsafe_arraycopy and generic_arraycopy for AArch64
8148328: aarch64: redundant lsr instructions in stub code.
8148783: aarch64: SEGV running SpecJBB2013
8148948: aarch64: generate_copy_longs calls align() incorrectly
8149080: AArch64: Recognise disjoint array copy in stub code
8149365: aarch64: memory copy does not prefetch on backwards copy
8149907: aarch64: use load/store pair instructions in call_stub
8150038: aarch64: make use of CBZ and CBNZ when comparing narrow pointer with zero
8150045: arraycopy causes segfaults in SATB during garbage collection
8150082: aarch64: optimise small array copy
8150229: aarch64: pipeline class for several instructions is not set correctly
8150313: aarch64: optimise array copy using SIMD instructions
8150394: aarch64: add support for 8.1 LSE CAS instructions
8150652: Remove unused code in AArch64 back end
8151340: aarch64: prefetch the destination word for write prior to ldxr/stxr loops.
8151502: optimize pd_disjoint_words and pd_conjoint_words
8151775: aarch64: add support for 8.1 LSE atomic operations
8152537: aarch64: Make use of CBZ and CBNZ when comparing unsigned values with zero.
8152840: aarch64: improve _unsafe_arraycopy stub routine
8153172: aarch64: hotspot crashes after the 8.1 LSE patch is merged
8153713: aarch64: improve short array clearing using store pair
8153797: aarch64: Add Arrays.fill stub code
8154413: AArch64: Better byte behaviour
8154537: AArch64: some integer rotate instructions are never emitted
8154739: AArch64: TemplateTable::fast_xaccess loads in wrong mode
8155015: Aarch64: bad assert in spill generation code
8155100: AArch64: Relax alignment requirement for byte_map_base
8155612: Aarch64: vector nodes need to support misaligned offset
8155617: aarch64: ClearArray does not use DC ZVA
8155627: Enable SA on AArch64
8155653: TestVectorUnalignedOffset.java not pushed with 8155612
8156731: aarch64: java/util/Arrays/Correct.java fails due to _generic_arraycopy stub routine
8157841: aarch64: prefetch ignores cache line size
8157906: aarch64: some more integer rotate instructions are never emitted
8158913: aarch64: SEGV running Spark terasort
8159052: aarch64: optimise unaligned copies in pd_disjoint_words and pd_conjoint_words
8159063: aarch64: optimise unaligned array copy long
8160748: [AArch64] Inconsistent types for ideal_reg
8161072: AArch64: jtreg compiler/uncommontrap/TestDeoptOOM failure
8161190: AArch64: Fix overflow in immediate cmp instruction
8164113: AArch64: follow-up the fix for 8161598
8165673: AArch64: Fix JNI floating point argument handling
8167200: AArch64: Broken stack pointer adjustment in interpreter
8167421: AArch64: in one core system, fatal error: Illegal threadstate encountered
8167595: AArch64: SEGV in stub code cipherBlockChaining_decryptAESCrypt
8168699: Validate special case invocations [AArch64 support]
8168888: Port 8160591: Improve internal array handling to AArch64.
8170100: AArch64: Crash in C1-compiled code accessing References
8170188: jtreg test compiler/types/TestMeetIncompatibleInterfaceArrays.java causes JVM crash
8170873: PPC64/aarch64: Poor StrictMath performance due to non-optimized compilation
8171537: aarch64: compiler/c1/Test6849574.java generates guarantee failure in C1
8172881: AArch64: assertion failure: the int pressure is incorrect
8173472: AArch64: C1 comparisons with null only use 32-bit instructions
8176100: [AArch64] [REDO][REDO] G1 Needs pre barrier on dereference of weak JNI handles
8177661: Correct ad rule output register types from iRegX to iRegXNoSp
8179954: AArch64: C1 and C2 volatile accesses are not sequentially consistent
8182581: aarch64: fix for crash caused by earlyret of compiled method
8183925: [AArch64] Decouple crash protection from watcher thread
8186325: AArch64: jtreg test hotspot/test/gc/g1/TestJNIWeakG1/TestJNIWeakG1.java SEGV
8187224: aarch64: some inconsistency between aarch64_ad.m4 and aarch64.ad
8189170: [AArch64] Add option to disable stack overflow checking in primordial thread for use with JNI_CreateJavaJVM
8193133: Assertion failure because 0xDEADDEAD can be in-heap
8195685: AArch64 port of 8174962: Better interface invocations
8195859: AArch64: vtableStubs gtest fails after 8174962
8196136: AArch64: Correct register use in patch for JDK-8194686
8196221: AArch64: Mistake in committed patch for JDK-8195859
8199712: [AArch64] Flight Recorder
8203481: Incorrect constraint for unextended_sp in frame:safe_for_sender
8203699: java/lang/invoke/SpecialInterfaceCall fails with SIGILL on aarch64
8205421: AARCH64: StubCodeMark should be placed after alignment
8206163: AArch64: incorrect code generation for StoreCM
8207345: Trampoline generation code reads from uninitialized memory
8207838: AArch64: Float registers incorrectly restored in JNI call
8209413: AArch64: NPE in clhsdb jstack command
8209414: [AArch64] method handle invocation does not respect JVMTI interp_only mode
8209415: Fix JVMTI test failure HS202
8209420: Track membars for volatile accesses so they can be properly optimized
8209835: Aarch64: elide barriers on all volatile operations
8210425: [AArch64] sharedRuntimeTrig/sharedRuntimeTrans compiled without optimization
8211064: [AArch64] Interpreter and c1 don't correctly handle jboolean results in native calls
8211233: MemBarNode::trailing_membar() and MemBarNode::leading_membar() need to handle dying subgraphs better
8213134: AArch64: vector shift failed with MaxVectorSize=8
8213419: [AArch64] C2 may hang in MulLNode::Ideal()/MulINode::Ideal() with gcc 8.2.1
8214857: "bad trailing membar" assert failure at memnode.cpp:3220
8215951: AArch64: jtreg test vmTestbase/nsk/jvmti/PopFrame/popframe005 segfaults
8215961: jdk/jfr/event/os/TestCPUInformation.java fails on AArch64
8216350: AArch64: monitor unlock fast path not called
8216989: CardTableBarrierSetAssembler::gen_write_ref_array_post_barrier() does not check for zero length on AARCH64
8217368: AArch64: C2 recursive stack locking optimisation not triggered
8218185: aarch64: missing LoadStore barrier in TemplateTable::putfield_or_static
8219011: Implement MacroAssembler::warn method on AArch64
8219635: aarch64: missing LoadStore barrier in TemplateTable::fast_storefield
8221220: AArch64: Add StoreStore membar explicitly for Volatile Writes in TemplateTable
8221658: aarch64: add necessary predicate for ubfx patterns
8224671: AArch64: mauve System.arraycopy test failure
8224828: aarch64: rflags is not correct after safepoint poll
8224851: AArch64: fix warnings and errors with Clang and GCC 8.3
8224880: AArch64: java/javac error with AllocatePrefetchDistance
8228400: Remove built-in AArch64 simulator
8228406: Superfluous change in chaitin.hpp
8228593: Revert explicit JDK 7 support additions
8228716: Revert InstanceKlass::print_on debug additions
8228718: Revert incorrect backport of JDK-8129757 to 8-aarch64
8228725: AArch64: Purge method call format support
8228747: Revert "unused" attribute from test_arraycopy_func
8228767: Revert ResourceMark additions
8228770: Revert development hsdis changes
8229123: Revert build fixes for aarch64/zero
8229124: Revert disassembler.cpp changes
8229145: Revert TemplateTable::bytecode() visibility change
8233839: aarch64: missing memory barrier in NewObjectArrayStub and NewTypeArrayStub
8237512: AArch64: aarch64TestHook leaks a BufferBlob
8246482: Build failures with +JFR -PCH
8247979: aarch64: missing side effect of killing flags for clearArray_reg_reg
8248219: aarch64: missing memory barrier in fast_storefield and fast_accessfield
Reviewed-by: shade, aph
author | andrew |
---|---|
date | Mon, 01 Feb 2021 03:48:36 +0000 |
parents | eddd586d1a4c |
children | f79e943d15a7 |
line wrap: on
line source
/* * Copyright (c) 1999, 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. * */ #include "precompiled.hpp" #include "opto/c2compiler.hpp" #include "opto/runtime.hpp" #if defined AD_MD_HPP # include AD_MD_HPP #elif defined TARGET_ARCH_MODEL_x86_32 # include "adfiles/ad_x86_32.hpp" #elif defined TARGET_ARCH_MODEL_x86_64 # include "adfiles/ad_x86_64.hpp" #elif defined TARGET_ARCH_MODEL_aarch64 # include "adfiles/ad_aarch64.hpp" #elif defined TARGET_ARCH_MODEL_sparc # include "adfiles/ad_sparc.hpp" #elif defined TARGET_ARCH_MODEL_zero # include "adfiles/ad_zero.hpp" #elif defined TARGET_ARCH_MODEL_ppc_64 # include "adfiles/ad_ppc_64.hpp" #endif // register information defined by ADLC extern const char register_save_policy[]; extern const int register_save_type[]; const char* C2Compiler::retry_no_subsuming_loads() { return "retry without subsuming loads"; } const char* C2Compiler::retry_no_escape_analysis() { return "retry without escape analysis"; } const char* C2Compiler::retry_class_loading_during_parsing() { return "retry class loading during parsing"; } bool C2Compiler::init_c2_runtime() { // Check assumptions used while running ADLC Compile::adlc_verification(); assert(REG_COUNT <= ConcreteRegisterImpl::number_of_registers, "incompatible register counts"); for (int i = 0; i < ConcreteRegisterImpl::number_of_registers ; i++ ) { OptoReg::vm2opto[i] = OptoReg::Bad; } for( OptoReg::Name i=OptoReg::Name(0); i<OptoReg::Name(REG_COUNT); i = OptoReg::add(i,1) ) { VMReg r = OptoReg::as_VMReg(i); if (r->is_valid()) { OptoReg::vm2opto[r->value()] = i; } } // Check that runtime and architecture description agree on callee-saved-floats bool callee_saved_floats = false; for( OptoReg::Name i=OptoReg::Name(0); i<OptoReg::Name(_last_Mach_Reg); i = OptoReg::add(i,1) ) { // Is there a callee-saved float or double? if( register_save_policy[i] == 'E' /* callee-saved */ && (register_save_type[i] == Op_RegF || register_save_type[i] == Op_RegD) ) { callee_saved_floats = true; } } DEBUG_ONLY( Node::init_NodeProperty(); ) Compile::pd_compiler2_init(); CompilerThread* thread = CompilerThread::current(); HandleMark handle_mark(thread); return OptoRuntime::generate(thread->env()); } void C2Compiler::initialize() { // The first compiler thread that gets here will initialize the // small amount of global state (and runtime stubs) that C2 needs. // There is a race possible once at startup and then we're fine // Note that this is being called from a compiler thread not the // main startup thread. if (should_perform_init()) { bool successful = C2Compiler::init_c2_runtime(); int new_state = (successful) ? initialized : failed; set_state(new_state); } } void C2Compiler::compile_method(ciEnv* env, ciMethod* target, int entry_bci) { assert(is_initialized(), "Compiler thread must be initialized"); bool subsume_loads = SubsumeLoads; bool do_escape_analysis = DoEscapeAnalysis && !env->should_retain_local_variables(); bool eliminate_boxing = EliminateAutoBox; while (!env->failing()) { // Attempt to compile while subsuming loads into machine instructions. Compile C(env, this, target, entry_bci, subsume_loads, do_escape_analysis, eliminate_boxing); // Check result and retry if appropriate. if (C.failure_reason() != NULL) { if (C.failure_reason_is(retry_class_loading_during_parsing())) { env->record_failure(C.failure_reason()); continue; // retry } if (C.failure_reason_is(retry_no_subsuming_loads())) { assert(subsume_loads, "must make progress"); subsume_loads = false; continue; // retry } if (C.failure_reason_is(retry_no_escape_analysis())) { assert(do_escape_analysis, "must make progress"); do_escape_analysis = false; continue; // retry } if (C.has_boxed_value()) { // Recompile without boxing elimination regardless failure reason. assert(eliminate_boxing, "must make progress"); eliminate_boxing = false; continue; // retry } // Pass any other failure reason up to the ciEnv. // Note that serious, irreversible failures are already logged // on the ciEnv via env->record_method_not_compilable(). env->record_failure(C.failure_reason()); } if (StressRecompilation) { if (subsume_loads) { subsume_loads = false; continue; // retry } if (do_escape_analysis) { do_escape_analysis = false; continue; // retry } } // No retry; just break the loop. break; } } void C2Compiler::print_timers() { // do nothing }