blob: 08883fcdea2c124e4bff9111f90a977a51f086d0 [file] [log] [blame]
/*
* Copyright 2014 The Android Open Source Project
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
#include "jit.h"
#include <dlfcn.h>
#include "art_method-inl.h"
#include "base/enums.h"
#include "base/file_utils.h"
#include "base/logging.h" // For VLOG.
#include "base/memory_tool.h"
#include "base/runtime_debug.h"
#include "base/scoped_flock.h"
#include "base/utils.h"
#include "class_root.h"
#include "debugger.h"
#include "dex/type_lookup_table.h"
#include "entrypoints/runtime_asm_entrypoints.h"
#include "interpreter/interpreter.h"
#include "jit-inl.h"
#include "jit_code_cache.h"
#include "jni/java_vm_ext.h"
#include "mirror/method_handle_impl.h"
#include "mirror/var_handle.h"
#include "oat_file.h"
#include "oat_file_manager.h"
#include "oat_quick_method_header.h"
#include "profile/profile_boot_info.h"
#include "profile/profile_compilation_info.h"
#include "profile_saver.h"
#include "runtime.h"
#include "runtime_options.h"
#include "stack.h"
#include "stack_map.h"
#include "thread-inl.h"
#include "thread_list.h"
namespace art {
namespace jit {
static constexpr bool kEnableOnStackReplacement = true;
// Maximum permitted threshold value.
static constexpr uint32_t kJitMaxThreshold = std::numeric_limits<uint16_t>::max();
// Different compilation threshold constants. These can be overridden on the command line.
// Non-debug default
static constexpr uint32_t kJitDefaultCompileThreshold = 20 * kJitSamplesBatchSize;
// Fast-debug build.
static constexpr uint32_t kJitStressDefaultCompileThreshold = 2 * kJitSamplesBatchSize;
// Slow-debug build.
static constexpr uint32_t kJitSlowStressDefaultCompileThreshold = 2;
// Different warm-up threshold constants. These default to the equivalent compile thresholds divided
// by 2, but can be overridden at the command-line.
static constexpr uint32_t kJitDefaultWarmUpThreshold = kJitDefaultCompileThreshold / 2;
static constexpr uint32_t kJitStressDefaultWarmUpThreshold = kJitStressDefaultCompileThreshold / 2;
static constexpr uint32_t kJitSlowStressDefaultWarmUpThreshold =
kJitSlowStressDefaultCompileThreshold / 2;
DEFINE_RUNTIME_DEBUG_FLAG(Jit, kSlowMode);
// JIT compiler
void* Jit::jit_library_handle_ = nullptr;
JitCompilerInterface* Jit::jit_compiler_ = nullptr;
JitCompilerInterface* (*Jit::jit_load_)(void) = nullptr;
JitOptions* JitOptions::CreateFromRuntimeArguments(const RuntimeArgumentMap& options) {
auto* jit_options = new JitOptions;
jit_options->use_jit_compilation_ = options.GetOrDefault(RuntimeArgumentMap::UseJitCompilation);
jit_options->code_cache_initial_capacity_ =
options.GetOrDefault(RuntimeArgumentMap::JITCodeCacheInitialCapacity);
jit_options->code_cache_max_capacity_ =
options.GetOrDefault(RuntimeArgumentMap::JITCodeCacheMaxCapacity);
jit_options->dump_info_on_shutdown_ =
options.Exists(RuntimeArgumentMap::DumpJITInfoOnShutdown);
jit_options->profile_saver_options_ =
options.GetOrDefault(RuntimeArgumentMap::ProfileSaverOpts);
jit_options->thread_pool_pthread_priority_ =
options.GetOrDefault(RuntimeArgumentMap::JITPoolThreadPthreadPriority);
// Set default compile threshold to aide with sanity checking defaults.
jit_options->compile_threshold_ =
kIsDebugBuild
? (Jit::kSlowMode
? kJitSlowStressDefaultCompileThreshold
: kJitStressDefaultCompileThreshold)
: kJitDefaultCompileThreshold;
// When not running in slow-mode, thresholds are quantized to kJitSamplesbatchsize.
const uint32_t kJitThresholdStep = Jit::kSlowMode ? 1u : kJitSamplesBatchSize;
// Set default warm-up threshold to aide with sanity checking defaults.
jit_options->warmup_threshold_ =
kIsDebugBuild ? (Jit::kSlowMode
? kJitSlowStressDefaultWarmUpThreshold
: kJitStressDefaultWarmUpThreshold)
: kJitDefaultWarmUpThreshold;
// Warmup threshold should be less than compile threshold (so long as compile threshold is not
// zero == JIT-on-first-use).
DCHECK_LT(jit_options->warmup_threshold_, jit_options->compile_threshold_);
DCHECK_EQ(RoundUp(jit_options->warmup_threshold_, kJitThresholdStep),
jit_options->warmup_threshold_);
if (options.Exists(RuntimeArgumentMap::JITCompileThreshold)) {
jit_options->compile_threshold_ = *options.Get(RuntimeArgumentMap::JITCompileThreshold);
}
jit_options->compile_threshold_ = RoundUp(jit_options->compile_threshold_, kJitThresholdStep);
if (options.Exists(RuntimeArgumentMap::JITWarmupThreshold)) {
jit_options->warmup_threshold_ = *options.Get(RuntimeArgumentMap::JITWarmupThreshold);
}
jit_options->warmup_threshold_ = RoundUp(jit_options->warmup_threshold_, kJitThresholdStep);
if (options.Exists(RuntimeArgumentMap::JITOsrThreshold)) {
jit_options->osr_threshold_ = *options.Get(RuntimeArgumentMap::JITOsrThreshold);
} else {
jit_options->osr_threshold_ = jit_options->compile_threshold_ * 2;
if (jit_options->osr_threshold_ > kJitMaxThreshold) {
jit_options->osr_threshold_ =
RoundDown(kJitMaxThreshold, kJitThresholdStep);
}
}
jit_options->osr_threshold_ = RoundUp(jit_options->osr_threshold_, kJitThresholdStep);
// Enforce ordering constraints between thresholds if not jit-on-first-use (when the compile
// threshold is 0).
if (jit_options->compile_threshold_ != 0) {
// Clamp thresholds such that OSR > compile > warm-up (see Jit::MaybeCompileMethod).
jit_options->osr_threshold_ = std::clamp(jit_options->osr_threshold_,
2u * kJitThresholdStep,
RoundDown(kJitMaxThreshold, kJitThresholdStep));
jit_options->compile_threshold_ = std::clamp(jit_options->compile_threshold_,
kJitThresholdStep,
jit_options->osr_threshold_ - kJitThresholdStep);
jit_options->warmup_threshold_ =
std::clamp(jit_options->warmup_threshold_,
0u,
jit_options->compile_threshold_ - kJitThresholdStep);
}
if (options.Exists(RuntimeArgumentMap::JITPriorityThreadWeight)) {
jit_options->priority_thread_weight_ =
*options.Get(RuntimeArgumentMap::JITPriorityThreadWeight);
if (jit_options->priority_thread_weight_ > jit_options->warmup_threshold_) {
LOG(FATAL) << "Priority thread weight is above the warmup threshold.";
} else if (jit_options->priority_thread_weight_ == 0) {
LOG(FATAL) << "Priority thread weight cannot be 0.";
}
} else {
jit_options->priority_thread_weight_ = std::max(
jit_options->warmup_threshold_ / Jit::kDefaultPriorityThreadWeightRatio,
static_cast<size_t>(1));
}
if (options.Exists(RuntimeArgumentMap::JITInvokeTransitionWeight)) {
jit_options->invoke_transition_weight_ =
*options.Get(RuntimeArgumentMap::JITInvokeTransitionWeight);
if (jit_options->invoke_transition_weight_ > jit_options->warmup_threshold_) {
LOG(FATAL) << "Invoke transition weight is above the warmup threshold.";
} else if (jit_options->invoke_transition_weight_ == 0) {
LOG(FATAL) << "Invoke transition weight cannot be 0.";
}
} else {
jit_options->invoke_transition_weight_ = std::max(
jit_options->warmup_threshold_ / Jit::kDefaultInvokeTransitionWeightRatio,
static_cast<size_t>(1));
}
return jit_options;
}
void Jit::DumpInfo(std::ostream& os) {
code_cache_->Dump(os);
cumulative_timings_.Dump(os);
MutexLock mu(Thread::Current(), lock_);
memory_use_.PrintMemoryUse(os);
}
void Jit::DumpForSigQuit(std::ostream& os) {
DumpInfo(os);
ProfileSaver::DumpInstanceInfo(os);
}
void Jit::AddTimingLogger(const TimingLogger& logger) {
cumulative_timings_.AddLogger(logger);
}
Jit::Jit(JitCodeCache* code_cache, JitOptions* options)
: code_cache_(code_cache),
options_(options),
boot_completed_lock_("Jit::boot_completed_lock_"),
cumulative_timings_("JIT timings"),
memory_use_("Memory used for compilation", 16),
lock_("JIT memory use lock") {}
Jit* Jit::Create(JitCodeCache* code_cache, JitOptions* options) {
if (jit_load_ == nullptr) {
LOG(WARNING) << "Not creating JIT: library not loaded";
return nullptr;
}
jit_compiler_ = (jit_load_)();
if (jit_compiler_ == nullptr) {
LOG(WARNING) << "Not creating JIT: failed to allocate a compiler";
return nullptr;
}
std::unique_ptr<Jit> jit(new Jit(code_cache, options));
// If the code collector is enabled, check if that still holds:
// With 'perf', we want a 1-1 mapping between an address and a method.
// We aren't able to keep method pointers live during the instrumentation method entry trampoline
// so we will just disable jit-gc if we are doing that.
if (code_cache->GetGarbageCollectCode()) {
code_cache->SetGarbageCollectCode(!jit_compiler_->GenerateDebugInfo() &&
!Runtime::Current()->GetInstrumentation()->AreExitStubsInstalled());
}
VLOG(jit) << "JIT created with initial_capacity="
<< PrettySize(options->GetCodeCacheInitialCapacity())
<< ", max_capacity=" << PrettySize(options->GetCodeCacheMaxCapacity())
<< ", compile_threshold=" << options->GetCompileThreshold()
<< ", profile_saver_options=" << options->GetProfileSaverOptions();
// Notify native debugger about the classes already loaded before the creation of the jit.
jit->DumpTypeInfoForLoadedTypes(Runtime::Current()->GetClassLinker());
return jit.release();
}
template <typename T>
bool Jit::LoadSymbol(T* address, const char* name, std::string* error_msg) {
*address = reinterpret_cast<T>(dlsym(jit_library_handle_, name));
if (*address == nullptr) {
*error_msg = std::string("JIT couldn't find ") + name + std::string(" entry point");
return false;
}
return true;
}
bool Jit::LoadCompilerLibrary(std::string* error_msg) {
jit_library_handle_ = dlopen(
kIsDebugBuild ? "libartd-compiler.so" : "libart-compiler.so", RTLD_NOW);
if (jit_library_handle_ == nullptr) {
std::ostringstream oss;
oss << "JIT could not load libart-compiler.so: " << dlerror();
*error_msg = oss.str();
return false;
}
if (!LoadSymbol(&jit_load_, "jit_load", error_msg)) {
dlclose(jit_library_handle_);
return false;
}
return true;
}
bool Jit::CompileMethod(ArtMethod* method, Thread* self, bool baseline, bool osr, bool prejit) {
DCHECK(Runtime::Current()->UseJitCompilation());
DCHECK(!method->IsRuntimeMethod());
RuntimeCallbacks* cb = Runtime::Current()->GetRuntimeCallbacks();
// Don't compile the method if it has breakpoints.
if (cb->IsMethodBeingInspected(method) && !cb->IsMethodSafeToJit(method)) {
VLOG(jit) << "JIT not compiling " << method->PrettyMethod()
<< " due to not being safe to jit according to runtime-callbacks. For example, there"
<< " could be breakpoints in this method.";
return false;
}
// Don't compile the method if we are supposed to be deoptimized.
instrumentation::Instrumentation* instrumentation = Runtime::Current()->GetInstrumentation();
if (instrumentation->AreAllMethodsDeoptimized() || instrumentation->IsDeoptimized(method)) {
VLOG(jit) << "JIT not compiling " << method->PrettyMethod() << " due to deoptimization";
return false;
}
JitMemoryRegion* region = GetCodeCache()->GetCurrentRegion();
if (osr && GetCodeCache()->IsSharedRegion(*region)) {
VLOG(jit) << "JIT not osr compiling "
<< method->PrettyMethod()
<< " due to using shared region";
return false;
}
// If we get a request to compile a proxy method, we pass the actual Java method
// of that proxy method, as the compiler does not expect a proxy method.
ArtMethod* method_to_compile = method->GetInterfaceMethodIfProxy(kRuntimePointerSize);
if (!code_cache_->NotifyCompilationOf(method_to_compile, self, osr, prejit, region)) {
return false;
}
VLOG(jit) << "Compiling method "
<< ArtMethod::PrettyMethod(method_to_compile)
<< " osr=" << std::boolalpha << osr;
bool success = jit_compiler_->CompileMethod(self, region, method_to_compile, baseline, osr);
code_cache_->DoneCompiling(method_to_compile, self, osr);
if (!success) {
VLOG(jit) << "Failed to compile method "
<< ArtMethod::PrettyMethod(method_to_compile)
<< " osr=" << std::boolalpha << osr;
}
if (kIsDebugBuild) {
if (self->IsExceptionPending()) {
mirror::Throwable* exception = self->GetException();
LOG(FATAL) << "No pending exception expected after compiling "
<< ArtMethod::PrettyMethod(method)
<< ": "
<< exception->Dump();
}
}
return success;
}
void Jit::WaitForWorkersToBeCreated() {
if (thread_pool_ != nullptr) {
thread_pool_->WaitForWorkersToBeCreated();
}
}
void Jit::DeleteThreadPool() {
Thread* self = Thread::Current();
if (thread_pool_ != nullptr) {
std::unique_ptr<ThreadPool> pool;
{
ScopedSuspendAll ssa(__FUNCTION__);
// Clear thread_pool_ field while the threads are suspended.
// A mutator in the 'AddSamples' method will check against it.
pool = std::move(thread_pool_);
}
// When running sanitized, let all tasks finish to not leak. Otherwise just clear the queue.
if (!kRunningOnMemoryTool) {
pool->StopWorkers(self);
pool->RemoveAllTasks(self);
}
// We could just suspend all threads, but we know those threads
// will finish in a short period, so it's not worth adding a suspend logic
// here. Besides, this is only done for shutdown.
pool->Wait(self, false, false);
}
}
void Jit::StartProfileSaver(const std::string& filename,
const std::vector<std::string>& code_paths) {
if (options_->GetSaveProfilingInfo()) {
ProfileSaver::Start(options_->GetProfileSaverOptions(), filename, code_cache_, code_paths);
}
}
void Jit::StopProfileSaver() {
if (options_->GetSaveProfilingInfo() && ProfileSaver::IsStarted()) {
ProfileSaver::Stop(options_->DumpJitInfoOnShutdown());
}
}
bool Jit::JitAtFirstUse() {
return HotMethodThreshold() == 0;
}
bool Jit::CanInvokeCompiledCode(ArtMethod* method) {
return code_cache_->ContainsPc(method->GetEntryPointFromQuickCompiledCode());
}
Jit::~Jit() {
DCHECK(!options_->GetSaveProfilingInfo() || !ProfileSaver::IsStarted());
if (options_->DumpJitInfoOnShutdown()) {
DumpInfo(LOG_STREAM(INFO));
Runtime::Current()->DumpDeoptimizations(LOG_STREAM(INFO));
}
DeleteThreadPool();
if (jit_compiler_ != nullptr) {
delete jit_compiler_;
jit_compiler_ = nullptr;
}
if (jit_library_handle_ != nullptr) {
dlclose(jit_library_handle_);
jit_library_handle_ = nullptr;
}
}
void Jit::NewTypeLoadedIfUsingJit(mirror::Class* type) {
if (!Runtime::Current()->UseJitCompilation()) {
// No need to notify if we only use the JIT to save profiles.
return;
}
jit::Jit* jit = Runtime::Current()->GetJit();
if (jit->jit_compiler_->GenerateDebugInfo()) {
jit_compiler_->TypesLoaded(&type, 1);
}
}
void Jit::DumpTypeInfoForLoadedTypes(ClassLinker* linker) {
struct CollectClasses : public ClassVisitor {
bool operator()(ObjPtr<mirror::Class> klass) override REQUIRES_SHARED(Locks::mutator_lock_) {
classes_.push_back(klass.Ptr());
return true;
}
std::vector<mirror::Class*> classes_;
};
if (jit_compiler_->GenerateDebugInfo()) {
ScopedObjectAccess so(Thread::Current());
CollectClasses visitor;
linker->VisitClasses(&visitor);
jit_compiler_->TypesLoaded(visitor.classes_.data(), visitor.classes_.size());
}
}
extern "C" void art_quick_osr_stub(void** stack,
size_t stack_size_in_bytes,
const uint8_t* native_pc,
JValue* result,
const char* shorty,
Thread* self);
bool Jit::MaybeDoOnStackReplacement(Thread* thread,
ArtMethod* method,
uint32_t dex_pc,
int32_t dex_pc_offset,
JValue* result) {
if (!kEnableOnStackReplacement) {
return false;
}
Jit* jit = Runtime::Current()->GetJit();
if (jit == nullptr) {
return false;
}
if (UNLIKELY(__builtin_frame_address(0) < thread->GetStackEnd())) {
// Don't attempt to do an OSR if we are close to the stack limit. Since
// the interpreter frames are still on stack, OSR has the potential
// to stack overflow even for a simple loop.
// b/27094810.
return false;
}
// Get the actual Java method if this method is from a proxy class. The compiler
// and the JIT code cache do not expect methods from proxy classes.
method = method->GetInterfaceMethodIfProxy(kRuntimePointerSize);
// Cheap check if the method has been compiled already. That's an indicator that we should
// osr into it.
if (!jit->GetCodeCache()->ContainsPc(method->GetEntryPointFromQuickCompiledCode())) {
return false;
}
// Fetch some data before looking up for an OSR method. We don't want thread
// suspension once we hold an OSR method, as the JIT code cache could delete the OSR
// method while we are being suspended.
CodeItemDataAccessor accessor(method->DexInstructionData());
const size_t number_of_vregs = accessor.RegistersSize();
const char* shorty = method->GetShorty();
std::string method_name(VLOG_IS_ON(jit) ? method->PrettyMethod() : "");
void** memory = nullptr;
size_t frame_size = 0;
ShadowFrame* shadow_frame = nullptr;
const uint8_t* native_pc = nullptr;
{
ScopedAssertNoThreadSuspension sts("Holding OSR method");
const OatQuickMethodHeader* osr_method = jit->GetCodeCache()->LookupOsrMethodHeader(method);
if (osr_method == nullptr) {
// No osr method yet, just return to the interpreter.
return false;
}
CodeInfo code_info(osr_method);
// Find stack map starting at the target dex_pc.
StackMap stack_map = code_info.GetOsrStackMapForDexPc(dex_pc + dex_pc_offset);
if (!stack_map.IsValid()) {
// There is no OSR stack map for this dex pc offset. Just return to the interpreter in the
// hope that the next branch has one.
return false;
}
// Before allowing the jump, make sure no code is actively inspecting the method to avoid
// jumping from interpreter to OSR while e.g. single stepping. Note that we could selectively
// disable OSR when single stepping, but that's currently hard to know at this point.
if (Runtime::Current()->GetRuntimeCallbacks()->IsMethodBeingInspected(method)) {
return false;
}
// We found a stack map, now fill the frame with dex register values from the interpreter's
// shadow frame.
DexRegisterMap vreg_map = code_info.GetDexRegisterMapOf(stack_map);
DCHECK_EQ(vreg_map.size(), number_of_vregs);
frame_size = osr_method->GetFrameSizeInBytes();
// Allocate memory to put shadow frame values. The osr stub will copy that memory to
// stack.
// Note that we could pass the shadow frame to the stub, and let it copy the values there,
// but that is engineering complexity not worth the effort for something like OSR.
memory = reinterpret_cast<void**>(malloc(frame_size));
CHECK(memory != nullptr);
memset(memory, 0, frame_size);
// Art ABI: ArtMethod is at the bottom of the stack.
memory[0] = method;
shadow_frame = thread->PopShadowFrame();
if (vreg_map.empty()) {
// If we don't have a dex register map, then there are no live dex registers at
// this dex pc.
} else {
for (uint16_t vreg = 0; vreg < number_of_vregs; ++vreg) {
DexRegisterLocation::Kind location = vreg_map[vreg].GetKind();
if (location == DexRegisterLocation::Kind::kNone) {
// Dex register is dead or uninitialized.
continue;
}
if (location == DexRegisterLocation::Kind::kConstant) {
// We skip constants because the compiled code knows how to handle them.
continue;
}
DCHECK_EQ(location, DexRegisterLocation::Kind::kInStack);
int32_t vreg_value = shadow_frame->GetVReg(vreg);
int32_t slot_offset = vreg_map[vreg].GetStackOffsetInBytes();
DCHECK_LT(slot_offset, static_cast<int32_t>(frame_size));
DCHECK_GT(slot_offset, 0);
(reinterpret_cast<int32_t*>(memory))[slot_offset / sizeof(int32_t)] = vreg_value;
}
}
native_pc = stack_map.GetNativePcOffset(kRuntimeISA) +
osr_method->GetEntryPoint();
VLOG(jit) << "Jumping to "
<< method_name
<< "@"
<< std::hex << reinterpret_cast<uintptr_t>(native_pc);
}
{
ManagedStack fragment;
thread->PushManagedStackFragment(&fragment);
(*art_quick_osr_stub)(memory,
frame_size,
native_pc,
result,
shorty,
thread);
if (UNLIKELY(thread->GetException() == Thread::GetDeoptimizationException())) {
thread->DeoptimizeWithDeoptimizationException(result);
}
thread->PopManagedStackFragment(fragment);
}
free(memory);
thread->PushShadowFrame(shadow_frame);
VLOG(jit) << "Done running OSR code for " << method_name;
return true;
}
void Jit::AddMemoryUsage(ArtMethod* method, size_t bytes) {
if (bytes > 4 * MB) {
LOG(INFO) << "Compiler allocated "
<< PrettySize(bytes)
<< " to compile "
<< ArtMethod::PrettyMethod(method);
}
MutexLock mu(Thread::Current(), lock_);
memory_use_.AddValue(bytes);
}
class JitCompileTask final : public Task {
public:
enum class TaskKind {
kAllocateProfile,
kCompile,
kCompileBaseline,
kCompileOsr,
kPreCompile,
};
JitCompileTask(ArtMethod* method, TaskKind kind) : method_(method), kind_(kind), klass_(nullptr) {
ScopedObjectAccess soa(Thread::Current());
// For a non-bootclasspath class, add a global ref to the class to prevent class unloading
// until compilation is done.
if (method->GetDeclaringClass()->GetClassLoader() != nullptr) {
klass_ = soa.Vm()->AddGlobalRef(soa.Self(), method_->GetDeclaringClass());
CHECK(klass_ != nullptr);
}
}
~JitCompileTask() {
if (klass_ != nullptr) {
ScopedObjectAccess soa(Thread::Current());
soa.Vm()->DeleteGlobalRef(soa.Self(), klass_);
}
}
void Run(Thread* self) override {
{
ScopedObjectAccess soa(self);
switch (kind_) {
case TaskKind::kPreCompile:
case TaskKind::kCompile:
case TaskKind::kCompileBaseline:
case TaskKind::kCompileOsr: {
Runtime::Current()->GetJit()->CompileMethod(
method_,
self,
/* baseline= */ (kind_ == TaskKind::kCompileBaseline),
/* osr= */ (kind_ == TaskKind::kCompileOsr),
/* prejit= */ (kind_ == TaskKind::kPreCompile));
break;
}
case TaskKind::kAllocateProfile: {
if (ProfilingInfo::Create(self, method_, /* retry_allocation= */ true)) {
VLOG(jit) << "Start profiling " << ArtMethod::PrettyMethod(method_);
}
break;
}
}
}
ProfileSaver::NotifyJitActivity();
}
void Finalize() override {
delete this;
}
private:
ArtMethod* const method_;
const TaskKind kind_;
jobject klass_;
DISALLOW_IMPLICIT_CONSTRUCTORS(JitCompileTask);
};
static std::string GetProfileFile(const std::string& dex_location) {
// Hardcoded assumption where the profile file is.
// TODO(ngeoffray): this is brittle and we would need to change change if we
// wanted to do more eager JITting of methods in a profile. This is
// currently only for system server.
return dex_location + ".prof";
}
static std::string GetBootProfileFile(const std::string& profile) {
// The boot profile can be found next to the compilation profile, with a
// different extension.
return ReplaceFileExtension(profile, "bprof");
}
/**
* A JIT task to madvise DONTNEED dex files after we're done compiling methods.
*/
class JitMadviseDontNeedTask final : public SelfDeletingTask {
public:
explicit JitMadviseDontNeedTask(const std::vector<const DexFile*>& dex_files)
: dex_files_(dex_files) {}
void Run(Thread* self ATTRIBUTE_UNUSED) override {
for (const DexFile* dex_file : dex_files_) {
if (IsAddressKnownBackedByFileOrShared(dex_file->Begin())) {
int result = madvise(const_cast<uint8_t*>(AlignDown(dex_file->Begin(), kPageSize)),
RoundUp(dex_file->Size(), kPageSize),
MADV_DONTNEED);
if (result == -1) {
PLOG(WARNING) << "Madvise failed";
}
}
}
}
private:
std::vector<const DexFile*> dex_files_;
DISALLOW_COPY_AND_ASSIGN(JitMadviseDontNeedTask);
};
class ZygoteTask final : public Task {
public:
ZygoteTask() {}
void Run(Thread* self) override {
Runtime* runtime = Runtime::Current();
std::string profile_file;
for (const std::string& option : runtime->GetImageCompilerOptions()) {
if (android::base::StartsWith(option, "--profile-file=")) {
profile_file = option.substr(strlen("--profile-file="));
break;
}
}
const std::vector<const DexFile*>& boot_class_path =
runtime->GetClassLinker()->GetBootClassPath();
ScopedNullHandle<mirror::ClassLoader> null_handle;
std::string boot_profile = GetBootProfileFile(profile_file);
// We add to the queue for zygote so that we can fork processes in-between
// compilations.
uint32_t added_to_queue = 0;
if (Runtime::Current()->IsPrimaryZygote()) {
// We avoid doing compilation at boot for the secondary zygote, as apps
// forked from it are not critical for boot.
added_to_queue += runtime->GetJit()->CompileMethodsFromBootProfile(
self, boot_class_path, boot_profile, null_handle, /* add_to_queue= */ true);
}
added_to_queue += runtime->GetJit()->CompileMethodsFromProfile(
self, boot_class_path, profile_file, null_handle, /* add_to_queue= */ true);
JitCodeCache* code_cache = runtime->GetJit()->GetCodeCache();
code_cache->GetZygoteMap()->Initialize(added_to_queue);
}
void Finalize() override {
delete this;
}
private:
DISALLOW_COPY_AND_ASSIGN(ZygoteTask);
};
class JitProfileTask final : public Task {
public:
JitProfileTask(const std::vector<std::unique_ptr<const DexFile>>& dex_files,
jobject class_loader) {
ScopedObjectAccess soa(Thread::Current());
StackHandleScope<1> hs(soa.Self());
Handle<mirror::ClassLoader> h_loader(hs.NewHandle(
soa.Decode<mirror::ClassLoader>(class_loader)));
ClassLinker* class_linker = Runtime::Current()->GetClassLinker();
for (const auto& dex_file : dex_files) {
dex_files_.push_back(dex_file.get());
// Register the dex file so that we can guarantee it doesn't get deleted
// while reading it during the task.
class_linker->RegisterDexFile(*dex_file.get(), h_loader.Get());
}
// We also create our own global ref to use this class loader later.
class_loader_ = soa.Vm()->AddGlobalRef(soa.Self(), h_loader.Get());
}
void Run(Thread* self) override {
ScopedObjectAccess soa(self);
StackHandleScope<1> hs(self);
Handle<mirror::ClassLoader> loader = hs.NewHandle<mirror::ClassLoader>(
soa.Decode<mirror::ClassLoader>(class_loader_));
std::string profile = GetProfileFile(dex_files_[0]->GetLocation());
std::string boot_profile = GetBootProfileFile(profile);
Jit* jit = Runtime::Current()->GetJit();
jit->CompileMethodsFromBootProfile(
self,
dex_files_,
boot_profile,
loader,
/* add_to_queue= */ false);
jit->CompileMethodsFromProfile(
self,
dex_files_,
profile,
loader,
/* add_to_queue= */ true);
}
void Finalize() override {
delete this;
}
~JitProfileTask() {
ScopedObjectAccess soa(Thread::Current());
soa.Vm()->DeleteGlobalRef(soa.Self(), class_loader_);
}
private:
std::vector<const DexFile*> dex_files_;
jobject class_loader_;
DISALLOW_COPY_AND_ASSIGN(JitProfileTask);
};
void Jit::CreateThreadPool() {
// There is a DCHECK in the 'AddSamples' method to ensure the tread pool
// is not null when we instrument.
// We need peers as we may report the JIT thread, e.g., in the debugger.
constexpr bool kJitPoolNeedsPeers = true;
thread_pool_.reset(new ThreadPool("Jit thread pool", 1, kJitPoolNeedsPeers));
thread_pool_->SetPthreadPriority(options_->GetThreadPoolPthreadPriority());
Start();
// If we're not using the default boot image location, request a JIT task to
// compile all methods in the boot image profile.
Runtime* runtime = Runtime::Current();
if (runtime->IsZygote() && runtime->IsUsingApexBootImageLocation() && UseJitCompilation()) {
thread_pool_->AddTask(Thread::Current(), new ZygoteTask());
}
}
void Jit::RegisterDexFiles(const std::vector<std::unique_ptr<const DexFile>>& dex_files,
jobject class_loader) {
if (dex_files.empty()) {
return;
}
Runtime* runtime = Runtime::Current();
if (runtime->IsSystemServer() && runtime->IsUsingApexBootImageLocation() && UseJitCompilation()) {
thread_pool_->AddTask(Thread::Current(), new JitProfileTask(dex_files, class_loader));
}
}
bool Jit::CompileMethodFromProfile(Thread* self,
ClassLinker* class_linker,
uint32_t method_idx,
Handle<mirror::DexCache> dex_cache,
Handle<mirror::ClassLoader> class_loader,
bool add_to_queue,
bool compile_after_boot) {
ArtMethod* method = class_linker->ResolveMethodWithoutInvokeType(
method_idx, dex_cache, class_loader);
if (method == nullptr) {
self->ClearException();
return false;
}
if (!method->IsCompilable() || !method->IsInvokable()) {
return false;
}
if (method->IsPreCompiled()) {
// Already seen by another profile.
return false;
}
const void* entry_point = method->GetEntryPointFromQuickCompiledCode();
if (class_linker->IsQuickToInterpreterBridge(entry_point) ||
class_linker->IsQuickGenericJniStub(entry_point) ||
class_linker->IsQuickResolutionStub(entry_point)) {
method->SetPreCompiled();
if (!add_to_queue) {
CompileMethod(method, self, /* baseline= */ false, /* osr= */ false, /* prejit= */ true);
} else {
Task* task = new JitCompileTask(method, JitCompileTask::TaskKind::kPreCompile);
if (compile_after_boot) {
MutexLock mu(Thread::Current(), boot_completed_lock_);
if (!boot_completed_) {
tasks_after_boot_.push_back(task);
return true;
}
DCHECK(tasks_after_boot_.empty());
}
thread_pool_->AddTask(self, task);
return true;
}
}
return false;
}
uint32_t Jit::CompileMethodsFromBootProfile(
Thread* self,
const std::vector<const DexFile*>& dex_files,
const std::string& profile_file,
Handle<mirror::ClassLoader> class_loader,
bool add_to_queue) {
unix_file::FdFile profile(profile_file.c_str(), O_RDONLY, true);
if (profile.Fd() == -1) {
PLOG(WARNING) << "No boot profile: " << profile_file;
return 0u;
}
ProfileBootInfo profile_info;
if (!profile_info.Load(profile.Fd(), dex_files)) {
LOG(ERROR) << "Could not load profile file: " << profile_file;
return 0u;
}
ScopedObjectAccess soa(self);
VariableSizedHandleScope handles(self);
std::vector<Handle<mirror::DexCache>> dex_caches;
ClassLinker* class_linker = Runtime::Current()->GetClassLinker();
for (const DexFile* dex_file : profile_info.GetDexFiles()) {
dex_caches.push_back(handles.NewHandle(class_linker->FindDexCache(self, *dex_file)));
}
uint32_t added_to_queue = 0;
for (const std::pair<uint32_t, uint32_t>& pair : profile_info.GetMethods()) {
if (CompileMethodFromProfile(self,
class_linker,
pair.second,
dex_caches[pair.first],
class_loader,
add_to_queue,
/*compile_after_boot=*/false)) {
++added_to_queue;
}
}
return added_to_queue;
}
uint32_t Jit::CompileMethodsFromProfile(
Thread* self,
const std::vector<const DexFile*>& dex_files,
const std::string& profile_file,
Handle<mirror::ClassLoader> class_loader,
bool add_to_queue) {
if (profile_file.empty()) {
LOG(WARNING) << "Expected a profile file in JIT zygote mode";
return 0u;
}
// We don't generate boot profiles on device, therefore we don't
// need to lock the file.
unix_file::FdFile profile(profile_file.c_str(), O_RDONLY, true);
if (profile.Fd() == -1) {
PLOG(WARNING) << "No profile: " << profile_file;
return 0u;
}
ProfileCompilationInfo profile_info;
if (!profile_info.Load(profile.Fd())) {
LOG(ERROR) << "Could not load profile file";
return 0u;
}
ScopedObjectAccess soa(self);
StackHandleScope<1> hs(self);
MutableHandle<mirror::DexCache> dex_cache = hs.NewHandle<mirror::DexCache>(nullptr);
ClassLinker* class_linker = Runtime::Current()->GetClassLinker();
uint32_t added_to_queue = 0u;
for (const DexFile* dex_file : dex_files) {
if (LocationIsOnArtModule(dex_file->GetLocation().c_str())) {
// The ART module jars are already preopted.
continue;
}
// To speed up class lookups, generate a type lookup table for
// the dex file.
if (dex_file->GetOatDexFile() == nullptr) {
TypeLookupTable type_lookup_table = TypeLookupTable::Create(*dex_file);
type_lookup_tables_.push_back(
std::make_unique<art::OatDexFile>(std::move(type_lookup_table)));
dex_file->SetOatDexFile(type_lookup_tables_.back().get());
}
std::set<dex::TypeIndex> class_types;
std::set<uint16_t> all_methods;
if (!profile_info.GetClassesAndMethods(*dex_file,
&class_types,
&all_methods,
&all_methods,
&all_methods)) {
// This means the profile file did not reference the dex file, which is the case
// if there's no classes and methods of that dex file in the profile.
continue;
}
dex_cache.Assign(class_linker->FindDexCache(self, *dex_file));
CHECK(dex_cache != nullptr) << "Could not find dex cache for " << dex_file->GetLocation();
for (uint16_t method_idx : all_methods) {
if (CompileMethodFromProfile(self,
class_linker,
method_idx,
dex_cache,
class_loader,
add_to_queue,
/*compile_after_boot=*/true)) {
++added_to_queue;
}
}
}
// Add a madvise task to release dex file pages once all compilation is done.
JitMadviseDontNeedTask* task = new JitMadviseDontNeedTask(dex_files);
MutexLock mu(Thread::Current(), boot_completed_lock_);
if (!boot_completed_) {
tasks_after_boot_.push_back(task);
} else {
DCHECK(tasks_after_boot_.empty());
thread_pool_->AddTask(self, task);
}
return added_to_queue;
}
static bool IgnoreSamplesForMethod(ArtMethod* method) REQUIRES_SHARED(Locks::mutator_lock_) {
if (method->IsClassInitializer() || !method->IsCompilable() || method->IsPreCompiled()) {
// We do not want to compile such methods.
return true;
}
if (method->IsNative()) {
ObjPtr<mirror::Class> klass = method->GetDeclaringClass();
if (klass == GetClassRoot<mirror::MethodHandle>() ||
klass == GetClassRoot<mirror::VarHandle>()) {
// MethodHandle and VarHandle invocation methods are required to throw an
// UnsupportedOperationException if invoked reflectively. We achieve this by having native
// implementations that raise the exception. We need to disable JIT compilation of these JNI
// methods as it can lead to transitioning between JIT compiled JNI stubs and generic JNI
// stubs. Since these stubs have different stack representations we can then crash in stack
// walking (b/78151261).
return true;
}
}
return false;
}
bool Jit::MaybeCompileMethod(Thread* self,
ArtMethod* method,
uint32_t old_count,
uint32_t new_count,
bool with_backedges) {
if (thread_pool_ == nullptr) {
return false;
}
if (UNLIKELY(method->IsPreCompiled()) && !with_backedges /* don't check for OSR */) {
const void* entry_point = code_cache_->GetSavedEntryPointOfPreCompiledMethod(method);
if (entry_point != nullptr) {
Runtime::Current()->GetInstrumentation()->UpdateMethodsCode(method, entry_point);
return true;
}
}
if (IgnoreSamplesForMethod(method)) {
return false;
}
if (HotMethodThreshold() == 0) {
// Tests might request JIT on first use (compiled synchronously in the interpreter).
return false;
}
DCHECK_GT(WarmMethodThreshold(), 0);
DCHECK_GT(HotMethodThreshold(), WarmMethodThreshold());
DCHECK_GT(OSRMethodThreshold(), HotMethodThreshold());
DCHECK_GE(PriorityThreadWeight(), 1);
DCHECK_LE(PriorityThreadWeight(), HotMethodThreshold());
if (old_count < WarmMethodThreshold() && new_count >= WarmMethodThreshold()) {
// Note: Native method have no "warm" state or profiling info.
if (!method->IsNative() &&
(method->GetProfilingInfo(kRuntimePointerSize) == nullptr) &&
code_cache_->CanAllocateProfilingInfo()) {
bool success = ProfilingInfo::Create(self, method, /* retry_allocation= */ false);
if (success) {
VLOG(jit) << "Start profiling " << method->PrettyMethod();
}
if (thread_pool_ == nullptr) {
// Calling ProfilingInfo::Create might put us in a suspended state, which could
// lead to the thread pool being deleted when we are shutting down.
return false;
}
if (!success) {
// We failed allocating. Instead of doing the collection on the Java thread, we push
// an allocation to a compiler thread, that will do the collection.
thread_pool_->AddTask(
self, new JitCompileTask(method, JitCompileTask::TaskKind::kAllocateProfile));
}
}
}
if (UseJitCompilation()) {
if (old_count == 0 &&
method->IsNative() &&
Runtime::Current()->IsUsingApexBootImageLocation()) {
// jitzygote: Compile JNI stub on first use to avoid the expensive generic stub.
CompileMethod(method, self, /* baseline= */ false, /* osr= */ false, /* prejit= */ false);
return true;
}
if (old_count < HotMethodThreshold() && new_count >= HotMethodThreshold()) {
if (!code_cache_->ContainsPc(method->GetEntryPointFromQuickCompiledCode())) {
DCHECK(thread_pool_ != nullptr);
thread_pool_->AddTask(self, new JitCompileTask(method, JitCompileTask::TaskKind::kCompile));
}
}
if (old_count < OSRMethodThreshold() && new_count >= OSRMethodThreshold()) {
if (!with_backedges) {
return false;
}
DCHECK(!method->IsNative()); // No back edges reported for native methods.
if (!code_cache_->IsOsrCompiled(method)) {
DCHECK(thread_pool_ != nullptr);
thread_pool_->AddTask(
self, new JitCompileTask(method, JitCompileTask::TaskKind::kCompileOsr));
}
}
}
return true;
}
class ScopedSetRuntimeThread {
public:
explicit ScopedSetRuntimeThread(Thread* self)
: self_(self), was_runtime_thread_(self_->IsRuntimeThread()) {
self_->SetIsRuntimeThread(true);
}
~ScopedSetRuntimeThread() {
self_->SetIsRuntimeThread(was_runtime_thread_);
}
private:
Thread* self_;
bool was_runtime_thread_;
};
void Jit::MethodEntered(Thread* thread, ArtMethod* method) {
Runtime* runtime = Runtime::Current();
if (UNLIKELY(runtime->UseJitCompilation() && JitAtFirstUse())) {
ArtMethod* np_method = method->GetInterfaceMethodIfProxy(kRuntimePointerSize);
if (np_method->IsCompilable()) {
if (!np_method->IsNative() && GetCodeCache()->CanAllocateProfilingInfo()) {
// The compiler requires a ProfilingInfo object for non-native methods.
ProfilingInfo::Create(thread, np_method, /* retry_allocation= */ true);
}
// TODO(ngeoffray): For JIT at first use, use kPreCompile. Currently we don't due to
// conflicts with jitzygote optimizations.
JitCompileTask compile_task(method, JitCompileTask::TaskKind::kCompile);
// Fake being in a runtime thread so that class-load behavior will be the same as normal jit.
ScopedSetRuntimeThread ssrt(thread);
compile_task.Run(thread);
}
return;
}
ProfilingInfo* profiling_info = method->GetProfilingInfo(kRuntimePointerSize);
// Update the entrypoint if the ProfilingInfo has one. The interpreter will call it
// instead of interpreting the method. We don't update it for instrumentation as the entrypoint
// must remain the instrumentation entrypoint.
if ((profiling_info != nullptr) &&
(profiling_info->GetSavedEntryPoint() != nullptr) &&
(method->GetEntryPointFromQuickCompiledCode() != GetQuickInstrumentationEntryPoint())) {
Runtime::Current()->GetInstrumentation()->UpdateMethodsCode(
method, profiling_info->GetSavedEntryPoint());
} else {
AddSamples(thread, method, 1, /* with_backedges= */false);
}
}
void Jit::InvokeVirtualOrInterface(ObjPtr<mirror::Object> this_object,
ArtMethod* caller,
uint32_t dex_pc,
ArtMethod* callee ATTRIBUTE_UNUSED) {
ScopedAssertNoThreadSuspension ants(__FUNCTION__);
DCHECK(this_object != nullptr);
ProfilingInfo* info = caller->GetProfilingInfo(kRuntimePointerSize);
if (info != nullptr) {
info->AddInvokeInfo(dex_pc, this_object->GetClass());
}
}
void Jit::WaitForCompilationToFinish(Thread* self) {
if (thread_pool_ != nullptr) {
thread_pool_->Wait(self, false, false);
}
}
void Jit::Stop() {
Thread* self = Thread::Current();
// TODO(ngeoffray): change API to not require calling WaitForCompilationToFinish twice.
WaitForCompilationToFinish(self);
GetThreadPool()->StopWorkers(self);
WaitForCompilationToFinish(self);
}
void Jit::Start() {
GetThreadPool()->StartWorkers(Thread::Current());
}
ScopedJitSuspend::ScopedJitSuspend() {
jit::Jit* jit = Runtime::Current()->GetJit();
was_on_ = (jit != nullptr) && (jit->GetThreadPool() != nullptr);
if (was_on_) {
jit->Stop();
}
}
ScopedJitSuspend::~ScopedJitSuspend() {
if (was_on_) {
DCHECK(Runtime::Current()->GetJit() != nullptr);
DCHECK(Runtime::Current()->GetJit()->GetThreadPool() != nullptr);
Runtime::Current()->GetJit()->Start();
}
}
void Jit::PostForkChildAction(bool is_system_server, bool is_zygote) {
// Clear the potential boot tasks inherited from the zygote.
{
MutexLock mu(Thread::Current(), boot_completed_lock_);
tasks_after_boot_.clear();
}
if (is_zygote || Runtime::Current()->IsSafeMode()) {
// Delete the thread pool, we are not going to JIT.
thread_pool_.reset(nullptr);
return;
}
// At this point, the compiler options have been adjusted to the particular configuration
// of the forked child. Parse them again.
jit_compiler_->ParseCompilerOptions();
// Adjust the status of code cache collection: the status from zygote was to not collect.
code_cache_->SetGarbageCollectCode(!jit_compiler_->GenerateDebugInfo() &&
!Runtime::Current()->GetInstrumentation()->AreExitStubsInstalled());
if (is_system_server && Runtime::Current()->IsUsingApexBootImageLocation()) {
// Disable garbage collection: we don't want it to delete methods we're compiling
// through boot and system server profiles.
// TODO(ngeoffray): Fix this so we still collect deoptimized and unused code.
code_cache_->SetGarbageCollectCode(false);
}
}
void Jit::PreZygoteFork() {
if (thread_pool_ == nullptr) {
return;
}
thread_pool_->DeleteThreads();
NativeDebugInfoPreFork();
}
void Jit::PostZygoteFork() {
if (thread_pool_ == nullptr) {
return;
}
thread_pool_->CreateThreads();
NativeDebugInfoPostFork();
}
void Jit::BootCompleted() {
Thread* self = Thread::Current();
std::deque<Task*> tasks;
{
MutexLock mu(self, boot_completed_lock_);
tasks = std::move(tasks_after_boot_);
boot_completed_ = true;
}
for (Task* task : tasks) {
thread_pool_->AddTask(self, task);
}
}
bool Jit::CanEncodeMethod(ArtMethod* method, bool is_for_shared_region) const {
return !is_for_shared_region ||
Runtime::Current()->GetHeap()->ObjectIsInBootImageSpace(method->GetDeclaringClass());
}
bool Jit::CanEncodeClass(ObjPtr<mirror::Class> cls, bool is_for_shared_region) const {
return !is_for_shared_region || Runtime::Current()->GetHeap()->ObjectIsInBootImageSpace(cls);
}
bool Jit::CanEncodeString(ObjPtr<mirror::String> string, bool is_for_shared_region) const {
return !is_for_shared_region || Runtime::Current()->GetHeap()->ObjectIsInBootImageSpace(string);
}
bool Jit::CanAssumeInitialized(ObjPtr<mirror::Class> cls, bool is_for_shared_region) const {
if (!is_for_shared_region) {
return cls->IsInitialized();
} else {
// Look up the class status in the oat file.
const DexFile& dex_file = *cls->GetDexCache()->GetDexFile();
const OatDexFile* oat_dex_file = dex_file.GetOatDexFile();
// In case we run without an image there won't be a backing oat file.
if (oat_dex_file == nullptr || oat_dex_file->GetOatFile() == nullptr) {
return false;
}
uint16_t class_def_index = cls->GetDexClassDefIndex();
return oat_dex_file->GetOatClass(class_def_index).GetStatus() >= ClassStatus::kInitialized;
}
}
} // namespace jit
} // namespace art