--- a/src/hotspot/share/runtime/advancedThresholdPolicy.cpp Wed May 09 07:48:31 2018 +0100
+++ /dev/null Thu Jan 01 00:00:00 1970 +0000
@@ -1,667 +0,0 @@
-/*
- * Copyright (c) 2010, 2018, 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 "code/codeCache.hpp"
-#include "runtime/advancedThresholdPolicy.hpp"
-#include "runtime/handles.inline.hpp"
-#include "runtime/simpleThresholdPolicy.inline.hpp"
-#if INCLUDE_JVMCI
-#include "jvmci/jvmciRuntime.hpp"
-#endif
-
-#ifdef TIERED
-// Print an event.
-void AdvancedThresholdPolicy::print_specific(EventType type, const methodHandle& mh, const methodHandle& imh,
- int bci, CompLevel level) {
- tty->print(" rate=");
- if (mh->prev_time() == 0) tty->print("n/a");
- else tty->print("%f", mh->rate());
-
- tty->print(" k=%.2lf,%.2lf", threshold_scale(CompLevel_full_profile, Tier3LoadFeedback),
- threshold_scale(CompLevel_full_optimization, Tier4LoadFeedback));
-
-}
-
-void AdvancedThresholdPolicy::initialize() {
- int count = CICompilerCount;
-#ifdef _LP64
- // Turn on ergonomic compiler count selection
- if (FLAG_IS_DEFAULT(CICompilerCountPerCPU) && FLAG_IS_DEFAULT(CICompilerCount)) {
- FLAG_SET_DEFAULT(CICompilerCountPerCPU, true);
- }
- if (CICompilerCountPerCPU) {
- // Simple log n seems to grow too slowly for tiered, try something faster: log n * log log n
- int log_cpu = log2_intptr(os::active_processor_count());
- int loglog_cpu = log2_intptr(MAX2(log_cpu, 1));
- count = MAX2(log_cpu * loglog_cpu * 3 / 2, 2);
- FLAG_SET_ERGO(intx, CICompilerCount, count);
- }
-#else
- // On 32-bit systems, the number of compiler threads is limited to 3.
- // On these systems, the virtual address space available to the JVM
- // is usually limited to 2-4 GB (the exact value depends on the platform).
- // As the compilers (especially C2) can consume a large amount of
- // memory, scaling the number of compiler threads with the number of
- // available cores can result in the exhaustion of the address space
- /// available to the VM and thus cause the VM to crash.
- if (FLAG_IS_DEFAULT(CICompilerCount)) {
- count = 3;
- FLAG_SET_ERGO(intx, CICompilerCount, count);
- }
-#endif
-
- if (TieredStopAtLevel < CompLevel_full_optimization) {
- // No C2 compiler thread required
- set_c1_count(count);
- } else {
- set_c1_count(MAX2(count / 3, 1));
- set_c2_count(MAX2(count - c1_count(), 1));
- }
- assert(count == c1_count() + c2_count(), "inconsistent compiler thread count");
-
- // Some inlining tuning
-#ifdef X86
- if (FLAG_IS_DEFAULT(InlineSmallCode)) {
- FLAG_SET_DEFAULT(InlineSmallCode, 2000);
- }
-#endif
-
-#if defined SPARC || defined AARCH64
- if (FLAG_IS_DEFAULT(InlineSmallCode)) {
- FLAG_SET_DEFAULT(InlineSmallCode, 2500);
- }
-#endif
-
- set_increase_threshold_at_ratio();
- set_start_time(os::javaTimeMillis());
-}
-
-// update_rate() is called from select_task() while holding a compile queue lock.
-void AdvancedThresholdPolicy::update_rate(jlong t, Method* m) {
- // Skip update if counters are absent.
- // Can't allocate them since we are holding compile queue lock.
- if (m->method_counters() == NULL) return;
-
- if (is_old(m)) {
- // We don't remove old methods from the queue,
- // so we can just zero the rate.
- m->set_rate(0);
- return;
- }
-
- // We don't update the rate if we've just came out of a safepoint.
- // delta_s is the time since last safepoint in milliseconds.
- jlong delta_s = t - SafepointSynchronize::end_of_last_safepoint();
- jlong delta_t = t - (m->prev_time() != 0 ? m->prev_time() : start_time()); // milliseconds since the last measurement
- // How many events were there since the last time?
- int event_count = m->invocation_count() + m->backedge_count();
- int delta_e = event_count - m->prev_event_count();
-
- // We should be running for at least 1ms.
- if (delta_s >= TieredRateUpdateMinTime) {
- // And we must've taken the previous point at least 1ms before.
- if (delta_t >= TieredRateUpdateMinTime && delta_e > 0) {
- m->set_prev_time(t);
- m->set_prev_event_count(event_count);
- m->set_rate((float)delta_e / (float)delta_t); // Rate is events per millisecond
- } else {
- if (delta_t > TieredRateUpdateMaxTime && delta_e == 0) {
- // If nothing happened for 25ms, zero the rate. Don't modify prev values.
- m->set_rate(0);
- }
- }
- }
-}
-
-// Check if this method has been stale from a given number of milliseconds.
-// See select_task().
-bool AdvancedThresholdPolicy::is_stale(jlong t, jlong timeout, Method* m) {
- jlong delta_s = t - SafepointSynchronize::end_of_last_safepoint();
- jlong delta_t = t - m->prev_time();
- if (delta_t > timeout && delta_s > timeout) {
- int event_count = m->invocation_count() + m->backedge_count();
- int delta_e = event_count - m->prev_event_count();
- // Return true if there were no events.
- return delta_e == 0;
- }
- return false;
-}
-
-// We don't remove old methods from the compile queue even if they have
-// very low activity. See select_task().
-bool AdvancedThresholdPolicy::is_old(Method* method) {
- return method->invocation_count() > 50000 || method->backedge_count() > 500000;
-}
-
-double AdvancedThresholdPolicy::weight(Method* method) {
- return (double)(method->rate() + 1) *
- (method->invocation_count() + 1) * (method->backedge_count() + 1);
-}
-
-// Apply heuristics and return true if x should be compiled before y
-bool AdvancedThresholdPolicy::compare_methods(Method* x, Method* y) {
- if (x->highest_comp_level() > y->highest_comp_level()) {
- // recompilation after deopt
- return true;
- } else
- if (x->highest_comp_level() == y->highest_comp_level()) {
- if (weight(x) > weight(y)) {
- return true;
- }
- }
- return false;
-}
-
-// Is method profiled enough?
-bool AdvancedThresholdPolicy::is_method_profiled(Method* method) {
- MethodData* mdo = method->method_data();
- if (mdo != NULL) {
- int i = mdo->invocation_count_delta();
- int b = mdo->backedge_count_delta();
- return call_predicate_helper<CompLevel_full_profile>(i, b, 1, method);
- }
- return false;
-}
-
-// Called with the queue locked and with at least one element
-CompileTask* AdvancedThresholdPolicy::select_task(CompileQueue* compile_queue) {
- CompileTask *max_blocking_task = NULL;
- CompileTask *max_task = NULL;
- Method* max_method = NULL;
- jlong t = os::javaTimeMillis();
- // Iterate through the queue and find a method with a maximum rate.
- for (CompileTask* task = compile_queue->first(); task != NULL;) {
- CompileTask* next_task = task->next();
- Method* method = task->method();
- update_rate(t, method);
- if (max_task == NULL) {
- max_task = task;
- max_method = method;
- } else {
- // If a method has been stale for some time, remove it from the queue.
- // Blocking tasks and tasks submitted from whitebox API don't become stale
- if (task->can_become_stale() && is_stale(t, TieredCompileTaskTimeout, method) && !is_old(method)) {
- if (PrintTieredEvents) {
- print_event(REMOVE_FROM_QUEUE, method, method, task->osr_bci(), (CompLevel)task->comp_level());
- }
- compile_queue->remove_and_mark_stale(task);
- method->clear_queued_for_compilation();
- task = next_task;
- continue;
- }
-
- // Select a method with a higher rate
- if (compare_methods(method, max_method)) {
- max_task = task;
- max_method = method;
- }
- }
-
- if (task->is_blocking()) {
- if (max_blocking_task == NULL || compare_methods(method, max_blocking_task->method())) {
- max_blocking_task = task;
- }
- }
-
- task = next_task;
- }
-
- if (max_blocking_task != NULL) {
- // In blocking compilation mode, the CompileBroker will make
- // compilations submitted by a JVMCI compiler thread non-blocking. These
- // compilations should be scheduled after all blocking compilations
- // to service non-compiler related compilations sooner and reduce the
- // chance of such compilations timing out.
- max_task = max_blocking_task;
- max_method = max_task->method();
- }
-
- if (max_task->comp_level() == CompLevel_full_profile && TieredStopAtLevel > CompLevel_full_profile
- && is_method_profiled(max_method)) {
- max_task->set_comp_level(CompLevel_limited_profile);
- if (PrintTieredEvents) {
- print_event(UPDATE_IN_QUEUE, max_method, max_method, max_task->osr_bci(), (CompLevel)max_task->comp_level());
- }
- }
-
- return max_task;
-}
-
-double AdvancedThresholdPolicy::threshold_scale(CompLevel level, int feedback_k) {
- double queue_size = CompileBroker::queue_size(level);
- int comp_count = compiler_count(level);
- double k = queue_size / (feedback_k * comp_count) + 1;
-
- // Increase C1 compile threshold when the code cache is filled more
- // than specified by IncreaseFirstTierCompileThresholdAt percentage.
- // The main intention is to keep enough free space for C2 compiled code
- // to achieve peak performance if the code cache is under stress.
- if ((TieredStopAtLevel == CompLevel_full_optimization) && (level != CompLevel_full_optimization)) {
- double current_reverse_free_ratio = CodeCache::reverse_free_ratio(CodeCache::get_code_blob_type(level));
- if (current_reverse_free_ratio > _increase_threshold_at_ratio) {
- k *= exp(current_reverse_free_ratio - _increase_threshold_at_ratio);
- }
- }
- return k;
-}
-
-// Call and loop predicates determine whether a transition to a higher
-// compilation level should be performed (pointers to predicate functions
-// are passed to common()).
-// Tier?LoadFeedback is basically a coefficient that determines of
-// how many methods per compiler thread can be in the queue before
-// the threshold values double.
-bool AdvancedThresholdPolicy::loop_predicate(int i, int b, CompLevel cur_level, Method* method) {
- switch(cur_level) {
- case CompLevel_aot: {
- double k = threshold_scale(CompLevel_full_profile, Tier3LoadFeedback);
- return loop_predicate_helper<CompLevel_aot>(i, b, k, method);
- }
- case CompLevel_none:
- case CompLevel_limited_profile: {
- double k = threshold_scale(CompLevel_full_profile, Tier3LoadFeedback);
- return loop_predicate_helper<CompLevel_none>(i, b, k, method);
- }
- case CompLevel_full_profile: {
- double k = threshold_scale(CompLevel_full_optimization, Tier4LoadFeedback);
- return loop_predicate_helper<CompLevel_full_profile>(i, b, k, method);
- }
- default:
- return true;
- }
-}
-
-bool AdvancedThresholdPolicy::call_predicate(int i, int b, CompLevel cur_level, Method* method) {
- switch(cur_level) {
- case CompLevel_aot: {
- double k = threshold_scale(CompLevel_full_profile, Tier3LoadFeedback);
- return call_predicate_helper<CompLevel_aot>(i, b, k, method);
- }
- case CompLevel_none:
- case CompLevel_limited_profile: {
- double k = threshold_scale(CompLevel_full_profile, Tier3LoadFeedback);
- return call_predicate_helper<CompLevel_none>(i, b, k, method);
- }
- case CompLevel_full_profile: {
- double k = threshold_scale(CompLevel_full_optimization, Tier4LoadFeedback);
- return call_predicate_helper<CompLevel_full_profile>(i, b, k, method);
- }
- default:
- return true;
- }
-}
-
-// If a method is old enough and is still in the interpreter we would want to
-// start profiling without waiting for the compiled method to arrive.
-// We also take the load on compilers into the account.
-bool AdvancedThresholdPolicy::should_create_mdo(Method* method, CompLevel cur_level) {
- if (cur_level == CompLevel_none &&
- CompileBroker::queue_size(CompLevel_full_optimization) <=
- Tier3DelayOn * compiler_count(CompLevel_full_optimization)) {
- int i = method->invocation_count();
- int b = method->backedge_count();
- double k = Tier0ProfilingStartPercentage / 100.0;
- return call_predicate_helper<CompLevel_none>(i, b, k, method) || loop_predicate_helper<CompLevel_none>(i, b, k, method);
- }
- return false;
-}
-
-// Inlining control: if we're compiling a profiled method with C1 and the callee
-// is known to have OSRed in a C2 version, don't inline it.
-bool AdvancedThresholdPolicy::should_not_inline(ciEnv* env, ciMethod* callee) {
- CompLevel comp_level = (CompLevel)env->comp_level();
- if (comp_level == CompLevel_full_profile ||
- comp_level == CompLevel_limited_profile) {
- return callee->highest_osr_comp_level() == CompLevel_full_optimization;
- }
- return false;
-}
-
-// Create MDO if necessary.
-void AdvancedThresholdPolicy::create_mdo(const methodHandle& mh, JavaThread* THREAD) {
- if (mh->is_native() ||
- mh->is_abstract() ||
- mh->is_accessor() ||
- mh->is_constant_getter()) {
- return;
- }
- if (mh->method_data() == NULL) {
- Method::build_interpreter_method_data(mh, CHECK_AND_CLEAR);
- }
-}
-
-
-/*
- * Method states:
- * 0 - interpreter (CompLevel_none)
- * 1 - pure C1 (CompLevel_simple)
- * 2 - C1 with invocation and backedge counting (CompLevel_limited_profile)
- * 3 - C1 with full profiling (CompLevel_full_profile)
- * 4 - C2 (CompLevel_full_optimization)
- *
- * Common state transition patterns:
- * a. 0 -> 3 -> 4.
- * The most common path. But note that even in this straightforward case
- * profiling can start at level 0 and finish at level 3.
- *
- * b. 0 -> 2 -> 3 -> 4.
- * This case occurs when the load on C2 is deemed too high. So, instead of transitioning
- * into state 3 directly and over-profiling while a method is in the C2 queue we transition to
- * level 2 and wait until the load on C2 decreases. This path is disabled for OSRs.
- *
- * c. 0 -> (3->2) -> 4.
- * In this case we enqueue a method for compilation at level 3, but the C1 queue is long enough
- * to enable the profiling to fully occur at level 0. In this case we change the compilation level
- * of the method to 2 while the request is still in-queue, because it'll allow it to run much faster
- * without full profiling while c2 is compiling.
- *
- * d. 0 -> 3 -> 1 or 0 -> 2 -> 1.
- * After a method was once compiled with C1 it can be identified as trivial and be compiled to
- * level 1. These transition can also occur if a method can't be compiled with C2 but can with C1.
- *
- * e. 0 -> 4.
- * This can happen if a method fails C1 compilation (it will still be profiled in the interpreter)
- * or because of a deopt that didn't require reprofiling (compilation won't happen in this case because
- * the compiled version already exists).
- *
- * Note that since state 0 can be reached from any other state via deoptimization different loops
- * are possible.
- *
- */
-
-// Common transition function. Given a predicate determines if a method should transition to another level.
-CompLevel AdvancedThresholdPolicy::common(Predicate p, Method* method, CompLevel cur_level, bool disable_feedback) {
- CompLevel next_level = cur_level;
- int i = method->invocation_count();
- int b = method->backedge_count();
-
- if (is_trivial(method)) {
- next_level = CompLevel_simple;
- } else {
- switch(cur_level) {
- default: break;
- case CompLevel_aot: {
- // If we were at full profile level, would we switch to full opt?
- if (common(p, method, CompLevel_full_profile, disable_feedback) == CompLevel_full_optimization) {
- next_level = CompLevel_full_optimization;
- } else if (disable_feedback || (CompileBroker::queue_size(CompLevel_full_optimization) <=
- Tier3DelayOff * compiler_count(CompLevel_full_optimization) &&
- (this->*p)(i, b, cur_level, method))) {
- next_level = CompLevel_full_profile;
- }
- }
- break;
- case CompLevel_none:
- // If we were at full profile level, would we switch to full opt?
- if (common(p, method, CompLevel_full_profile, disable_feedback) == CompLevel_full_optimization) {
- next_level = CompLevel_full_optimization;
- } else if ((this->*p)(i, b, cur_level, method)) {
-#if INCLUDE_JVMCI
- if (EnableJVMCI && UseJVMCICompiler) {
- // Since JVMCI takes a while to warm up, its queue inevitably backs up during
- // early VM execution. As of 2014-06-13, JVMCI's inliner assumes that the root
- // compilation method and all potential inlinees have mature profiles (which
- // includes type profiling). If it sees immature profiles, JVMCI's inliner
- // can perform pathologically bad (e.g., causing OutOfMemoryErrors due to
- // exploring/inlining too many graphs). Since a rewrite of the inliner is
- // in progress, we simply disable the dialing back heuristic for now and will
- // revisit this decision once the new inliner is completed.
- next_level = CompLevel_full_profile;
- } else
-#endif
- {
- // C1-generated fully profiled code is about 30% slower than the limited profile
- // code that has only invocation and backedge counters. The observation is that
- // if C2 queue is large enough we can spend too much time in the fully profiled code
- // while waiting for C2 to pick the method from the queue. To alleviate this problem
- // we introduce a feedback on the C2 queue size. If the C2 queue is sufficiently long
- // we choose to compile a limited profiled version and then recompile with full profiling
- // when the load on C2 goes down.
- if (!disable_feedback && CompileBroker::queue_size(CompLevel_full_optimization) >
- Tier3DelayOn * compiler_count(CompLevel_full_optimization)) {
- next_level = CompLevel_limited_profile;
- } else {
- next_level = CompLevel_full_profile;
- }
- }
- }
- break;
- case CompLevel_limited_profile:
- if (is_method_profiled(method)) {
- // Special case: we got here because this method was fully profiled in the interpreter.
- next_level = CompLevel_full_optimization;
- } else {
- MethodData* mdo = method->method_data();
- if (mdo != NULL) {
- if (mdo->would_profile()) {
- if (disable_feedback || (CompileBroker::queue_size(CompLevel_full_optimization) <=
- Tier3DelayOff * compiler_count(CompLevel_full_optimization) &&
- (this->*p)(i, b, cur_level, method))) {
- next_level = CompLevel_full_profile;
- }
- } else {
- next_level = CompLevel_full_optimization;
- }
- } else {
- // If there is no MDO we need to profile
- if (disable_feedback || (CompileBroker::queue_size(CompLevel_full_optimization) <=
- Tier3DelayOff * compiler_count(CompLevel_full_optimization) &&
- (this->*p)(i, b, cur_level, method))) {
- next_level = CompLevel_full_profile;
- }
- }
- }
- break;
- case CompLevel_full_profile:
- {
- MethodData* mdo = method->method_data();
- if (mdo != NULL) {
- if (mdo->would_profile()) {
- int mdo_i = mdo->invocation_count_delta();
- int mdo_b = mdo->backedge_count_delta();
- if ((this->*p)(mdo_i, mdo_b, cur_level, method)) {
- next_level = CompLevel_full_optimization;
- }
- } else {
- next_level = CompLevel_full_optimization;
- }
- }
- }
- break;
- }
- }
- return MIN2(next_level, (CompLevel)TieredStopAtLevel);
-}
-
-// Determine if a method should be compiled with a normal entry point at a different level.
-CompLevel AdvancedThresholdPolicy::call_event(Method* method, CompLevel cur_level, JavaThread * thread) {
- CompLevel osr_level = MIN2((CompLevel) method->highest_osr_comp_level(),
- common(&AdvancedThresholdPolicy::loop_predicate, method, cur_level, true));
- CompLevel next_level = common(&AdvancedThresholdPolicy::call_predicate, method, cur_level);
-
- // If OSR method level is greater than the regular method level, the levels should be
- // equalized by raising the regular method level in order to avoid OSRs during each
- // invocation of the method.
- if (osr_level == CompLevel_full_optimization && cur_level == CompLevel_full_profile) {
- MethodData* mdo = method->method_data();
- guarantee(mdo != NULL, "MDO should not be NULL");
- if (mdo->invocation_count() >= 1) {
- next_level = CompLevel_full_optimization;
- }
- } else {
- next_level = MAX2(osr_level, next_level);
- }
-#if INCLUDE_JVMCI
- if (UseJVMCICompiler) {
- next_level = JVMCIRuntime::adjust_comp_level(method, false, next_level, thread);
- }
-#endif
- return next_level;
-}
-
-// Determine if we should do an OSR compilation of a given method.
-CompLevel AdvancedThresholdPolicy::loop_event(Method* method, CompLevel cur_level, JavaThread * thread) {
- CompLevel next_level = common(&AdvancedThresholdPolicy::loop_predicate, method, cur_level, true);
- if (cur_level == CompLevel_none) {
- // If there is a live OSR method that means that we deopted to the interpreter
- // for the transition.
- CompLevel osr_level = MIN2((CompLevel)method->highest_osr_comp_level(), next_level);
- if (osr_level > CompLevel_none) {
- return osr_level;
- }
- }
-#if INCLUDE_JVMCI
- if (UseJVMCICompiler) {
- next_level = JVMCIRuntime::adjust_comp_level(method, true, next_level, thread);
- }
-#endif
- return next_level;
-}
-
-// Update the rate and submit compile
-void AdvancedThresholdPolicy::submit_compile(const methodHandle& mh, int bci, CompLevel level, JavaThread* thread) {
- int hot_count = (bci == InvocationEntryBci) ? mh->invocation_count() : mh->backedge_count();
- update_rate(os::javaTimeMillis(), mh());
- CompileBroker::compile_method(mh, bci, level, mh, hot_count, CompileTask::Reason_Tiered, thread);
-}
-
-bool AdvancedThresholdPolicy::maybe_switch_to_aot(const methodHandle& mh, CompLevel cur_level, CompLevel next_level, JavaThread* thread) {
- if (UseAOT && !delay_compilation_during_startup()) {
- if (cur_level == CompLevel_full_profile || cur_level == CompLevel_none) {
- // If the current level is full profile or interpreter and we're switching to any other level,
- // activate the AOT code back first so that we won't waste time overprofiling.
- compile(mh, InvocationEntryBci, CompLevel_aot, thread);
- // Fall through for JIT compilation.
- }
- if (next_level == CompLevel_limited_profile && cur_level != CompLevel_aot && mh->has_aot_code()) {
- // If the next level is limited profile, use the aot code (if there is any),
- // since it's essentially the same thing.
- compile(mh, InvocationEntryBci, CompLevel_aot, thread);
- // Not need to JIT, we're done.
- return true;
- }
- }
- return false;
-}
-
-
-// Handle the invocation event.
-void AdvancedThresholdPolicy::method_invocation_event(const methodHandle& mh, const methodHandle& imh,
- CompLevel level, CompiledMethod* nm, JavaThread* thread) {
- if (should_create_mdo(mh(), level)) {
- create_mdo(mh, thread);
- }
- CompLevel next_level = call_event(mh(), level, thread);
- if (next_level != level) {
- if (maybe_switch_to_aot(mh, level, next_level, thread)) {
- // No JITting necessary
- return;
- }
- if (is_compilation_enabled() && !CompileBroker::compilation_is_in_queue(mh)) {
- compile(mh, InvocationEntryBci, next_level, thread);
- }
- }
-}
-
-// Handle the back branch event. Notice that we can compile the method
-// with a regular entry from here.
-void AdvancedThresholdPolicy::method_back_branch_event(const methodHandle& mh, const methodHandle& imh,
- int bci, CompLevel level, CompiledMethod* nm, JavaThread* thread) {
- if (should_create_mdo(mh(), level)) {
- create_mdo(mh, thread);
- }
- // Check if MDO should be created for the inlined method
- if (should_create_mdo(imh(), level)) {
- create_mdo(imh, thread);
- }
-
- if (is_compilation_enabled()) {
- CompLevel next_osr_level = loop_event(imh(), level, thread);
- CompLevel max_osr_level = (CompLevel)imh->highest_osr_comp_level();
- // At the very least compile the OSR version
- if (!CompileBroker::compilation_is_in_queue(imh) && (next_osr_level != level)) {
- compile(imh, bci, next_osr_level, thread);
- }
-
- // Use loop event as an opportunity to also check if there's been
- // enough calls.
- CompLevel cur_level, next_level;
- if (mh() != imh()) { // If there is an enclosing method
- if (level == CompLevel_aot) {
- // Recompile the enclosing method to prevent infinite OSRs. Stay at AOT level while it's compiling.
- if (max_osr_level != CompLevel_none && !CompileBroker::compilation_is_in_queue(mh)) {
- compile(mh, InvocationEntryBci, MIN2((CompLevel)TieredStopAtLevel, CompLevel_full_profile), thread);
- }
- } else {
- // Current loop event level is not AOT
- guarantee(nm != NULL, "Should have nmethod here");
- cur_level = comp_level(mh());
- next_level = call_event(mh(), cur_level, thread);
-
- if (max_osr_level == CompLevel_full_optimization) {
- // The inlinee OSRed to full opt, we need to modify the enclosing method to avoid deopts
- bool make_not_entrant = false;
- if (nm->is_osr_method()) {
- // This is an osr method, just make it not entrant and recompile later if needed
- make_not_entrant = true;
- } else {
- if (next_level != CompLevel_full_optimization) {
- // next_level is not full opt, so we need to recompile the
- // enclosing method without the inlinee
- cur_level = CompLevel_none;
- make_not_entrant = true;
- }
- }
- if (make_not_entrant) {
- if (PrintTieredEvents) {
- int osr_bci = nm->is_osr_method() ? nm->osr_entry_bci() : InvocationEntryBci;
- print_event(MAKE_NOT_ENTRANT, mh(), mh(), osr_bci, level);
- }
- nm->make_not_entrant();
- }
- }
- // Fix up next_level if necessary to avoid deopts
- if (next_level == CompLevel_limited_profile && max_osr_level == CompLevel_full_profile) {
- next_level = CompLevel_full_profile;
- }
- if (cur_level != next_level) {
- if (!maybe_switch_to_aot(mh, cur_level, next_level, thread) && !CompileBroker::compilation_is_in_queue(mh)) {
- compile(mh, InvocationEntryBci, next_level, thread);
- }
- }
- }
- } else {
- cur_level = comp_level(mh());
- next_level = call_event(mh(), cur_level, thread);
- if (next_level != cur_level) {
- if (!maybe_switch_to_aot(mh, cur_level, next_level, thread) && !CompileBroker::compilation_is_in_queue(mh)) {
- compile(mh, InvocationEntryBci, next_level, thread);
- }
- }
- }
- }
-}
-
-#endif // TIERED
--- a/src/hotspot/share/runtime/advancedThresholdPolicy.hpp Wed May 09 07:48:31 2018 +0100
+++ /dev/null Thu Jan 01 00:00:00 1970 +0000
@@ -1,235 +0,0 @@
-/*
- * Copyright (c) 2010, 2017, 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.
- *
- */
-
-#ifndef SHARE_VM_RUNTIME_ADVANCEDTHRESHOLDPOLICY_HPP
-#define SHARE_VM_RUNTIME_ADVANCEDTHRESHOLDPOLICY_HPP
-
-#include "runtime/simpleThresholdPolicy.hpp"
-
-#ifdef TIERED
-class CompileTask;
-class CompileQueue;
-
-/*
- * The system supports 5 execution levels:
- * * level 0 - interpreter
- * * level 1 - C1 with full optimization (no profiling)
- * * level 2 - C1 with invocation and backedge counters
- * * level 3 - C1 with full profiling (level 2 + MDO)
- * * level 4 - C2
- *
- * Levels 0, 2 and 3 periodically notify the runtime about the current value of the counters
- * (invocation counters and backedge counters). The frequency of these notifications is
- * different at each level. These notifications are used by the policy to decide what transition
- * to make.
- *
- * Execution starts at level 0 (interpreter), then the policy can decide either to compile the
- * method at level 3 or level 2. The decision is based on the following factors:
- * 1. The length of the C2 queue determines the next level. The observation is that level 2
- * is generally faster than level 3 by about 30%, therefore we would want to minimize the time
- * a method spends at level 3. We should only spend the time at level 3 that is necessary to get
- * adequate profiling. So, if the C2 queue is long enough it is more beneficial to go first to
- * level 2, because if we transitioned to level 3 we would be stuck there until our C2 compile
- * request makes its way through the long queue. When the load on C2 recedes we are going to
- * recompile at level 3 and start gathering profiling information.
- * 2. The length of C1 queue is used to dynamically adjust the thresholds, so as to introduce
- * additional filtering if the compiler is overloaded. The rationale is that by the time a
- * method gets compiled it can become unused, so it doesn't make sense to put too much onto the
- * queue.
- *
- * After profiling is completed at level 3 the transition is made to level 4. Again, the length
- * of the C2 queue is used as a feedback to adjust the thresholds.
- *
- * After the first C1 compile some basic information is determined about the code like the number
- * of the blocks and the number of the loops. Based on that it can be decided that a method
- * is trivial and compiling it with C1 will yield the same code. In this case the method is
- * compiled at level 1 instead of 4.
- *
- * We also support profiling at level 0. If C1 is slow enough to produce the level 3 version of
- * the code and the C2 queue is sufficiently small we can decide to start profiling in the
- * interpreter (and continue profiling in the compiled code once the level 3 version arrives).
- * If the profiling at level 0 is fully completed before level 3 version is produced, a level 2
- * version is compiled instead in order to run faster waiting for a level 4 version.
- *
- * Compile queues are implemented as priority queues - for each method in the queue we compute
- * the event rate (the number of invocation and backedge counter increments per unit of time).
- * When getting an element off the queue we pick the one with the largest rate. Maintaining the
- * rate also allows us to remove stale methods (the ones that got on the queue but stopped
- * being used shortly after that).
-*/
-
-/* Command line options:
- * - Tier?InvokeNotifyFreqLog and Tier?BackedgeNotifyFreqLog control the frequency of method
- * invocation and backedge notifications. Basically every n-th invocation or backedge a mutator thread
- * makes a call into the runtime.
- *
- * - Tier?InvocationThreshold, Tier?CompileThreshold, Tier?BackEdgeThreshold, Tier?MinInvocationThreshold control
- * compilation thresholds.
- * Level 2 thresholds are not used and are provided for option-compatibility and potential future use.
- * Other thresholds work as follows:
- *
- * Transition from interpreter (level 0) to C1 with full profiling (level 3) happens when
- * the following predicate is true (X is the level):
- *
- * i > TierXInvocationThreshold * s || (i > TierXMinInvocationThreshold * s && i + b > TierXCompileThreshold * s),
- *
- * where $i$ is the number of method invocations, $b$ number of backedges and $s$ is the scaling
- * coefficient that will be discussed further.
- * The intuition is to equalize the time that is spend profiling each method.
- * The same predicate is used to control the transition from level 3 to level 4 (C2). It should be
- * noted though that the thresholds are relative. Moreover i and b for the 0->3 transition come
- * from Method* and for 3->4 transition they come from MDO (since profiled invocations are
- * counted separately). Finally, if a method does not contain anything worth profiling, a transition
- * from level 3 to level 4 occurs without considering thresholds (e.g., with fewer invocations than
- * what is specified by Tier4InvocationThreshold).
- *
- * OSR transitions are controlled simply with b > TierXBackEdgeThreshold * s predicates.
- *
- * - Tier?LoadFeedback options are used to automatically scale the predicates described above depending
- * on the compiler load. The scaling coefficients are computed as follows:
- *
- * s = queue_size_X / (TierXLoadFeedback * compiler_count_X) + 1,
- *
- * where queue_size_X is the current size of the compiler queue of level X, and compiler_count_X
- * is the number of level X compiler threads.
- *
- * Basically these parameters describe how many methods should be in the compile queue
- * per compiler thread before the scaling coefficient increases by one.
- *
- * This feedback provides the mechanism to automatically control the flow of compilation requests
- * depending on the machine speed, mutator load and other external factors.
- *
- * - Tier3DelayOn and Tier3DelayOff parameters control another important feedback loop.
- * Consider the following observation: a method compiled with full profiling (level 3)
- * is about 30% slower than a method at level 2 (just invocation and backedge counters, no MDO).
- * Normally, the following transitions will occur: 0->3->4. The problem arises when the C2 queue
- * gets congested and the 3->4 transition is delayed. While the method is the C2 queue it continues
- * executing at level 3 for much longer time than is required by the predicate and at suboptimal speed.
- * The idea is to dynamically change the behavior of the system in such a way that if a substantial
- * load on C2 is detected we would first do the 0->2 transition allowing a method to run faster.
- * And then when the load decreases to allow 2->3 transitions.
- *
- * Tier3Delay* parameters control this switching mechanism.
- * Tier3DelayOn is the number of methods in the C2 queue per compiler thread after which the policy
- * no longer does 0->3 transitions but does 0->2 transitions instead.
- * Tier3DelayOff switches the original behavior back when the number of methods in the C2 queue
- * per compiler thread falls below the specified amount.
- * The hysteresis is necessary to avoid jitter.
- *
- * - TieredCompileTaskTimeout is the amount of time an idle method can spend in the compile queue.
- * Basically, since we use the event rate d(i + b)/dt as a value of priority when selecting a method to
- * compile from the compile queue, we also can detect stale methods for which the rate has been
- * 0 for some time in the same iteration. Stale methods can appear in the queue when an application
- * abruptly changes its behavior.
- *
- * - TieredStopAtLevel, is used mostly for testing. It allows to bypass the policy logic and stick
- * to a given level. For example it's useful to set TieredStopAtLevel = 1 in order to compile everything
- * with pure c1.
- *
- * - Tier0ProfilingStartPercentage allows the interpreter to start profiling when the inequalities in the
- * 0->3 predicate are already exceeded by the given percentage but the level 3 version of the
- * method is still not ready. We can even go directly from level 0 to 4 if c1 doesn't produce a compiled
- * version in time. This reduces the overall transition to level 4 and decreases the startup time.
- * Note that this behavior is also guarded by the Tier3Delay mechanism: when the c2 queue is too long
- * these is not reason to start profiling prematurely.
- *
- * - TieredRateUpdateMinTime and TieredRateUpdateMaxTime are parameters of the rate computation.
- * Basically, the rate is not computed more frequently than TieredRateUpdateMinTime and is considered
- * to be zero if no events occurred in TieredRateUpdateMaxTime.
- */
-
-
-class AdvancedThresholdPolicy : public SimpleThresholdPolicy {
- jlong _start_time;
-
- // Call and loop predicates determine whether a transition to a higher compilation
- // level should be performed (pointers to predicate functions are passed to common().
- // Predicates also take compiler load into account.
- typedef bool (AdvancedThresholdPolicy::*Predicate)(int i, int b, CompLevel cur_level, Method* method);
- bool call_predicate(int i, int b, CompLevel cur_level, Method* method);
- bool loop_predicate(int i, int b, CompLevel cur_level, Method* method);
- // Common transition function. Given a predicate determines if a method should transition to another level.
- CompLevel common(Predicate p, Method* method, CompLevel cur_level, bool disable_feedback = false);
- // Transition functions.
- // call_event determines if a method should be compiled at a different
- // level with a regular invocation entry.
- CompLevel call_event(Method* method, CompLevel cur_level, JavaThread * thread);
- // loop_event checks if a method should be OSR compiled at a different
- // level.
- CompLevel loop_event(Method* method, CompLevel cur_level, JavaThread * thread);
- // Has a method been long around?
- // We don't remove old methods from the compile queue even if they have
- // very low activity (see select_task()).
- inline bool is_old(Method* method);
- // Was a given method inactive for a given number of milliseconds.
- // If it is, we would remove it from the queue (see select_task()).
- inline bool is_stale(jlong t, jlong timeout, Method* m);
- // Compute the weight of the method for the compilation scheduling
- inline double weight(Method* method);
- // Apply heuristics and return true if x should be compiled before y
- inline bool compare_methods(Method* x, Method* y);
- // Compute event rate for a given method. The rate is the number of event (invocations + backedges)
- // per millisecond.
- inline void update_rate(jlong t, Method* m);
- // Compute threshold scaling coefficient
- inline double threshold_scale(CompLevel level, int feedback_k);
- // If a method is old enough and is still in the interpreter we would want to
- // start profiling without waiting for the compiled method to arrive. This function
- // determines whether we should do that.
- inline bool should_create_mdo(Method* method, CompLevel cur_level);
- // Create MDO if necessary.
- void create_mdo(const methodHandle& mh, JavaThread* thread);
- // Is method profiled enough?
- bool is_method_profiled(Method* method);
-
- double _increase_threshold_at_ratio;
-
- bool maybe_switch_to_aot(const methodHandle& mh, CompLevel cur_level, CompLevel next_level, JavaThread* thread);
-
-protected:
- void print_specific(EventType type, const methodHandle& mh, const methodHandle& imh, int bci, CompLevel level);
-
- void set_increase_threshold_at_ratio() { _increase_threshold_at_ratio = 100 / (100 - (double)IncreaseFirstTierCompileThresholdAt); }
- void set_start_time(jlong t) { _start_time = t; }
- jlong start_time() const { return _start_time; }
-
- // Submit a given method for compilation (and update the rate).
- virtual void submit_compile(const methodHandle& mh, int bci, CompLevel level, JavaThread* thread);
- // event() from SimpleThresholdPolicy would call these.
- virtual void method_invocation_event(const methodHandle& method, const methodHandle& inlinee,
- CompLevel level, CompiledMethod* nm, JavaThread* thread);
- virtual void method_back_branch_event(const methodHandle& method, const methodHandle& inlinee,
- int bci, CompLevel level, CompiledMethod* nm, JavaThread* thread);
-public:
- AdvancedThresholdPolicy() : _start_time(0) { }
- // Select task is called by CompileBroker. We should return a task or NULL.
- virtual CompileTask* select_task(CompileQueue* compile_queue);
- virtual void initialize();
- virtual bool should_not_inline(ciEnv* env, ciMethod* callee);
-
-};
-
-#endif // TIERED
-
-#endif // SHARE_VM_RUNTIME_ADVANCEDTHRESHOLDPOLICY_HPP
--- a/src/hotspot/share/runtime/arguments.cpp Wed May 09 07:48:31 2018 +0100
+++ b/src/hotspot/share/runtime/arguments.cpp Wed May 09 09:39:25 2018 +0200
@@ -1603,9 +1603,9 @@
}
void Arguments::set_tiered_flags() {
- // With tiered, set default policy to AdvancedThresholdPolicy, which is 3.
+ // With tiered, set default policy to SimpleThresholdPolicy, which is 2.
if (FLAG_IS_DEFAULT(CompilationPolicyChoice)) {
- FLAG_SET_DEFAULT(CompilationPolicyChoice, 3);
+ FLAG_SET_DEFAULT(CompilationPolicyChoice, 2);
}
if (CompilationPolicyChoice < 2) {
vm_exit_during_initialization(
--- a/src/hotspot/share/runtime/compilationPolicy.cpp Wed May 09 07:48:31 2018 +0100
+++ b/src/hotspot/share/runtime/compilationPolicy.cpp Wed May 09 09:39:25 2018 +0200
@@ -33,7 +33,6 @@
#include "oops/method.inline.hpp"
#include "oops/oop.inline.hpp"
#include "prims/nativeLookup.hpp"
-#include "runtime/advancedThresholdPolicy.hpp"
#include "runtime/compilationPolicy.hpp"
#include "runtime/frame.hpp"
#include "runtime/handles.inline.hpp"
@@ -74,15 +73,8 @@
Unimplemented();
#endif
break;
- case 3:
-#ifdef TIERED
- CompilationPolicy::set_policy(new AdvancedThresholdPolicy());
-#else
- Unimplemented();
-#endif
- break;
default:
- fatal("CompilationPolicyChoice must be in the range: [0-3]");
+ fatal("CompilationPolicyChoice must be in the range: [0-2]");
}
CompilationPolicy::policy()->initialize();
}
--- a/src/hotspot/share/runtime/globals.hpp Wed May 09 07:48:31 2018 +0100
+++ b/src/hotspot/share/runtime/globals.hpp Wed May 09 09:39:25 2018 +0200
@@ -1158,8 +1158,8 @@
"UseDynamicNumberOfCompilerThreads") \
\
product(intx, CompilationPolicyChoice, 0, \
- "which compilation policy (0-3)") \
- range(0, 3) \
+ "which compilation policy (0-2)") \
+ range(0, 2) \
\
develop(bool, UseStackBanging, true, \
"use stack banging for stack overflow checks (required for " \
--- a/src/hotspot/share/runtime/simpleThresholdPolicy.cpp Wed May 09 07:48:31 2018 +0100
+++ b/src/hotspot/share/runtime/simpleThresholdPolicy.cpp Wed May 09 09:39:25 2018 +0200
@@ -140,20 +140,33 @@
}
void SimpleThresholdPolicy::initialize() {
- if (FLAG_IS_DEFAULT(CICompilerCount)) {
- FLAG_SET_DEFAULT(CICompilerCount, 3);
- }
int count = CICompilerCount;
#ifdef _LP64
- // On 64-bit systems, scale the number of compiler threads with
- // the number of cores available on the system. Scaling is not
- // performed on 32-bit systems because it can lead to exhaustion
- // of the virtual memory address space available to the JVM.
+ // Turn on ergonomic compiler count selection
+ if (FLAG_IS_DEFAULT(CICompilerCountPerCPU) && FLAG_IS_DEFAULT(CICompilerCount)) {
+ FLAG_SET_DEFAULT(CICompilerCountPerCPU, true);
+ }
if (CICompilerCountPerCPU) {
- count = MAX2(log2_intptr(os::active_processor_count()) * 3 / 2, 2);
+ // Simple log n seems to grow too slowly for tiered, try something faster: log n * log log n
+ int log_cpu = log2_intptr(os::active_processor_count());
+ int loglog_cpu = log2_intptr(MAX2(log_cpu, 1));
+ count = MAX2(log_cpu * loglog_cpu * 3 / 2, 2);
+ FLAG_SET_ERGO(intx, CICompilerCount, count);
+ }
+#else
+ // On 32-bit systems, the number of compiler threads is limited to 3.
+ // On these systems, the virtual address space available to the JVM
+ // is usually limited to 2-4 GB (the exact value depends on the platform).
+ // As the compilers (especially C2) can consume a large amount of
+ // memory, scaling the number of compiler threads with the number of
+ // available cores can result in the exhaustion of the address space
+ /// available to the VM and thus cause the VM to crash.
+ if (FLAG_IS_DEFAULT(CICompilerCount)) {
+ count = 3;
FLAG_SET_ERGO(intx, CICompilerCount, count);
}
#endif
+
if (TieredStopAtLevel < CompLevel_full_optimization) {
// No C2 compiler thread required
set_c1_count(count);
@@ -162,6 +175,22 @@
set_c2_count(MAX2(count - c1_count(), 1));
}
assert(count == c1_count() + c2_count(), "inconsistent compiler thread count");
+
+ // Some inlining tuning
+#ifdef X86
+ if (FLAG_IS_DEFAULT(InlineSmallCode)) {
+ FLAG_SET_DEFAULT(InlineSmallCode, 2000);
+ }
+#endif
+
+#if defined SPARC || defined AARCH64
+ if (FLAG_IS_DEFAULT(InlineSmallCode)) {
+ FLAG_SET_DEFAULT(InlineSmallCode, 2500);
+ }
+#endif
+
+ set_increase_threshold_at_ratio();
+ set_start_time(os::javaTimeMillis());
}
void SimpleThresholdPolicy::set_carry_if_necessary(InvocationCounter *counter) {
@@ -186,7 +215,66 @@
// Called with the queue locked and with at least one element
CompileTask* SimpleThresholdPolicy::select_task(CompileQueue* compile_queue) {
- return select_task_helper(compile_queue);
+ CompileTask *max_blocking_task = NULL;
+ CompileTask *max_task = NULL;
+ Method* max_method = NULL;
+ jlong t = os::javaTimeMillis();
+ // Iterate through the queue and find a method with a maximum rate.
+ for (CompileTask* task = compile_queue->first(); task != NULL;) {
+ CompileTask* next_task = task->next();
+ Method* method = task->method();
+ update_rate(t, method);
+ if (max_task == NULL) {
+ max_task = task;
+ max_method = method;
+ } else {
+ // If a method has been stale for some time, remove it from the queue.
+ // Blocking tasks and tasks submitted from whitebox API don't become stale
+ if (task->can_become_stale() && is_stale(t, TieredCompileTaskTimeout, method) && !is_old(method)) {
+ if (PrintTieredEvents) {
+ print_event(REMOVE_FROM_QUEUE, method, method, task->osr_bci(), (CompLevel)task->comp_level());
+ }
+ compile_queue->remove_and_mark_stale(task);
+ method->clear_queued_for_compilation();
+ task = next_task;
+ continue;
+ }
+
+ // Select a method with a higher rate
+ if (compare_methods(method, max_method)) {
+ max_task = task;
+ max_method = method;
+ }
+ }
+
+ if (task->is_blocking()) {
+ if (max_blocking_task == NULL || compare_methods(method, max_blocking_task->method())) {
+ max_blocking_task = task;
+ }
+ }
+
+ task = next_task;
+ }
+
+ if (max_blocking_task != NULL) {
+ // In blocking compilation mode, the CompileBroker will make
+ // compilations submitted by a JVMCI compiler thread non-blocking. These
+ // compilations should be scheduled after all blocking compilations
+ // to service non-compiler related compilations sooner and reduce the
+ // chance of such compilations timing out.
+ max_task = max_blocking_task;
+ max_method = max_task->method();
+ }
+
+ if (max_task->comp_level() == CompLevel_full_profile && TieredStopAtLevel > CompLevel_full_profile
+ && is_method_profiled(max_method)) {
+ max_task->set_comp_level(CompLevel_limited_profile);
+ if (PrintTieredEvents) {
+ print_event(UPDATE_IN_QUEUE, max_method, max_method, max_task->osr_bci(), (CompLevel)max_task->comp_level());
+ }
+ }
+
+ return max_task;
}
void SimpleThresholdPolicy::reprofile(ScopeDesc* trap_scope, bool is_osr) {
@@ -284,26 +372,150 @@
}
}
-// Tell the broker to compile the method
+// Update the rate and submit compile
void SimpleThresholdPolicy::submit_compile(const methodHandle& mh, int bci, CompLevel level, JavaThread* thread) {
int hot_count = (bci == InvocationEntryBci) ? mh->invocation_count() : mh->backedge_count();
+ update_rate(os::javaTimeMillis(), mh());
CompileBroker::compile_method(mh, bci, level, mh, hot_count, CompileTask::Reason_Tiered, thread);
}
+// Print an event.
+void SimpleThresholdPolicy::print_specific(EventType type, const methodHandle& mh, const methodHandle& imh,
+ int bci, CompLevel level) {
+ tty->print(" rate=");
+ if (mh->prev_time() == 0) tty->print("n/a");
+ else tty->print("%f", mh->rate());
+
+ tty->print(" k=%.2lf,%.2lf", threshold_scale(CompLevel_full_profile, Tier3LoadFeedback),
+ threshold_scale(CompLevel_full_optimization, Tier4LoadFeedback));
+
+}
+
+// update_rate() is called from select_task() while holding a compile queue lock.
+void SimpleThresholdPolicy::update_rate(jlong t, Method* m) {
+ // Skip update if counters are absent.
+ // Can't allocate them since we are holding compile queue lock.
+ if (m->method_counters() == NULL) return;
+
+ if (is_old(m)) {
+ // We don't remove old methods from the queue,
+ // so we can just zero the rate.
+ m->set_rate(0);
+ return;
+ }
+
+ // We don't update the rate if we've just came out of a safepoint.
+ // delta_s is the time since last safepoint in milliseconds.
+ jlong delta_s = t - SafepointSynchronize::end_of_last_safepoint();
+ jlong delta_t = t - (m->prev_time() != 0 ? m->prev_time() : start_time()); // milliseconds since the last measurement
+ // How many events were there since the last time?
+ int event_count = m->invocation_count() + m->backedge_count();
+ int delta_e = event_count - m->prev_event_count();
+
+ // We should be running for at least 1ms.
+ if (delta_s >= TieredRateUpdateMinTime) {
+ // And we must've taken the previous point at least 1ms before.
+ if (delta_t >= TieredRateUpdateMinTime && delta_e > 0) {
+ m->set_prev_time(t);
+ m->set_prev_event_count(event_count);
+ m->set_rate((float)delta_e / (float)delta_t); // Rate is events per millisecond
+ } else {
+ if (delta_t > TieredRateUpdateMaxTime && delta_e == 0) {
+ // If nothing happened for 25ms, zero the rate. Don't modify prev values.
+ m->set_rate(0);
+ }
+ }
+ }
+}
+
+// Check if this method has been stale from a given number of milliseconds.
+// See select_task().
+bool SimpleThresholdPolicy::is_stale(jlong t, jlong timeout, Method* m) {
+ jlong delta_s = t - SafepointSynchronize::end_of_last_safepoint();
+ jlong delta_t = t - m->prev_time();
+ if (delta_t > timeout && delta_s > timeout) {
+ int event_count = m->invocation_count() + m->backedge_count();
+ int delta_e = event_count - m->prev_event_count();
+ // Return true if there were no events.
+ return delta_e == 0;
+ }
+ return false;
+}
+
+// We don't remove old methods from the compile queue even if they have
+// very low activity. See select_task().
+bool SimpleThresholdPolicy::is_old(Method* method) {
+ return method->invocation_count() > 50000 || method->backedge_count() > 500000;
+}
+
+double SimpleThresholdPolicy::weight(Method* method) {
+ return (double)(method->rate() + 1) *
+ (method->invocation_count() + 1) * (method->backedge_count() + 1);
+}
+
+// Apply heuristics and return true if x should be compiled before y
+bool SimpleThresholdPolicy::compare_methods(Method* x, Method* y) {
+ if (x->highest_comp_level() > y->highest_comp_level()) {
+ // recompilation after deopt
+ return true;
+ } else
+ if (x->highest_comp_level() == y->highest_comp_level()) {
+ if (weight(x) > weight(y)) {
+ return true;
+ }
+ }
+ return false;
+}
+
+// Is method profiled enough?
+bool SimpleThresholdPolicy::is_method_profiled(Method* method) {
+ MethodData* mdo = method->method_data();
+ if (mdo != NULL) {
+ int i = mdo->invocation_count_delta();
+ int b = mdo->backedge_count_delta();
+ return call_predicate_helper<CompLevel_full_profile>(i, b, 1, method);
+ }
+ return false;
+}
+
+double SimpleThresholdPolicy::threshold_scale(CompLevel level, int feedback_k) {
+ double queue_size = CompileBroker::queue_size(level);
+ int comp_count = compiler_count(level);
+ double k = queue_size / (feedback_k * comp_count) + 1;
+
+ // Increase C1 compile threshold when the code cache is filled more
+ // than specified by IncreaseFirstTierCompileThresholdAt percentage.
+ // The main intention is to keep enough free space for C2 compiled code
+ // to achieve peak performance if the code cache is under stress.
+ if ((TieredStopAtLevel == CompLevel_full_optimization) && (level != CompLevel_full_optimization)) {
+ double current_reverse_free_ratio = CodeCache::reverse_free_ratio(CodeCache::get_code_blob_type(level));
+ if (current_reverse_free_ratio > _increase_threshold_at_ratio) {
+ k *= exp(current_reverse_free_ratio - _increase_threshold_at_ratio);
+ }
+ }
+ return k;
+}
+
// Call and loop predicates determine whether a transition to a higher
// compilation level should be performed (pointers to predicate functions
-// are passed to common() transition function).
+// are passed to common()).
+// Tier?LoadFeedback is basically a coefficient that determines of
+// how many methods per compiler thread can be in the queue before
+// the threshold values double.
bool SimpleThresholdPolicy::loop_predicate(int i, int b, CompLevel cur_level, Method* method) {
switch(cur_level) {
case CompLevel_aot: {
- return loop_predicate_helper<CompLevel_aot>(i, b, 1.0, method);
+ double k = threshold_scale(CompLevel_full_profile, Tier3LoadFeedback);
+ return loop_predicate_helper<CompLevel_aot>(i, b, k, method);
}
case CompLevel_none:
case CompLevel_limited_profile: {
- return loop_predicate_helper<CompLevel_none>(i, b, 1.0, method);
+ double k = threshold_scale(CompLevel_full_profile, Tier3LoadFeedback);
+ return loop_predicate_helper<CompLevel_none>(i, b, k, method);
}
case CompLevel_full_profile: {
- return loop_predicate_helper<CompLevel_full_profile>(i, b, 1.0, method);
+ double k = threshold_scale(CompLevel_full_optimization, Tier4LoadFeedback);
+ return loop_predicate_helper<CompLevel_full_profile>(i, b, k, method);
}
default:
return true;
@@ -313,14 +525,17 @@
bool SimpleThresholdPolicy::call_predicate(int i, int b, CompLevel cur_level, Method* method) {
switch(cur_level) {
case CompLevel_aot: {
- return call_predicate_helper<CompLevel_aot>(i, b, 1.0, method);
+ double k = threshold_scale(CompLevel_full_profile, Tier3LoadFeedback);
+ return call_predicate_helper<CompLevel_aot>(i, b, k, method);
}
case CompLevel_none:
case CompLevel_limited_profile: {
- return call_predicate_helper<CompLevel_none>(i, b, 1.0, method);
+ double k = threshold_scale(CompLevel_full_profile, Tier3LoadFeedback);
+ return call_predicate_helper<CompLevel_none>(i, b, k, method);
}
case CompLevel_full_profile: {
- return call_predicate_helper<CompLevel_full_profile>(i, b, 1.0, method);
+ double k = threshold_scale(CompLevel_full_optimization, Tier4LoadFeedback);
+ return call_predicate_helper<CompLevel_full_profile>(i, b, k, method);
}
default:
return true;
@@ -341,31 +556,167 @@
return false;
}
+// If a method is old enough and is still in the interpreter we would want to
+// start profiling without waiting for the compiled method to arrive.
+// We also take the load on compilers into the account.
+bool SimpleThresholdPolicy::should_create_mdo(Method* method, CompLevel cur_level) {
+ if (cur_level == CompLevel_none &&
+ CompileBroker::queue_size(CompLevel_full_optimization) <=
+ Tier3DelayOn * compiler_count(CompLevel_full_optimization)) {
+ int i = method->invocation_count();
+ int b = method->backedge_count();
+ double k = Tier0ProfilingStartPercentage / 100.0;
+ return call_predicate_helper<CompLevel_none>(i, b, k, method) || loop_predicate_helper<CompLevel_none>(i, b, k, method);
+ }
+ return false;
+}
+
+// Inlining control: if we're compiling a profiled method with C1 and the callee
+// is known to have OSRed in a C2 version, don't inline it.
+bool SimpleThresholdPolicy::should_not_inline(ciEnv* env, ciMethod* callee) {
+ CompLevel comp_level = (CompLevel)env->comp_level();
+ if (comp_level == CompLevel_full_profile ||
+ comp_level == CompLevel_limited_profile) {
+ return callee->highest_osr_comp_level() == CompLevel_full_optimization;
+ }
+ return false;
+}
+
+// Create MDO if necessary.
+void SimpleThresholdPolicy::create_mdo(const methodHandle& mh, JavaThread* THREAD) {
+ if (mh->is_native() ||
+ mh->is_abstract() ||
+ mh->is_accessor() ||
+ mh->is_constant_getter()) {
+ return;
+ }
+ if (mh->method_data() == NULL) {
+ Method::build_interpreter_method_data(mh, CHECK_AND_CLEAR);
+ }
+}
+
+
+/*
+ * Method states:
+ * 0 - interpreter (CompLevel_none)
+ * 1 - pure C1 (CompLevel_simple)
+ * 2 - C1 with invocation and backedge counting (CompLevel_limited_profile)
+ * 3 - C1 with full profiling (CompLevel_full_profile)
+ * 4 - C2 (CompLevel_full_optimization)
+ *
+ * Common state transition patterns:
+ * a. 0 -> 3 -> 4.
+ * The most common path. But note that even in this straightforward case
+ * profiling can start at level 0 and finish at level 3.
+ *
+ * b. 0 -> 2 -> 3 -> 4.
+ * This case occurs when the load on C2 is deemed too high. So, instead of transitioning
+ * into state 3 directly and over-profiling while a method is in the C2 queue we transition to
+ * level 2 and wait until the load on C2 decreases. This path is disabled for OSRs.
+ *
+ * c. 0 -> (3->2) -> 4.
+ * In this case we enqueue a method for compilation at level 3, but the C1 queue is long enough
+ * to enable the profiling to fully occur at level 0. In this case we change the compilation level
+ * of the method to 2 while the request is still in-queue, because it'll allow it to run much faster
+ * without full profiling while c2 is compiling.
+ *
+ * d. 0 -> 3 -> 1 or 0 -> 2 -> 1.
+ * After a method was once compiled with C1 it can be identified as trivial and be compiled to
+ * level 1. These transition can also occur if a method can't be compiled with C2 but can with C1.
+ *
+ * e. 0 -> 4.
+ * This can happen if a method fails C1 compilation (it will still be profiled in the interpreter)
+ * or because of a deopt that didn't require reprofiling (compilation won't happen in this case because
+ * the compiled version already exists).
+ *
+ * Note that since state 0 can be reached from any other state via deoptimization different loops
+ * are possible.
+ *
+ */
+
// Common transition function. Given a predicate determines if a method should transition to another level.
-CompLevel SimpleThresholdPolicy::common(Predicate p, Method* method, CompLevel cur_level) {
+CompLevel SimpleThresholdPolicy::common(Predicate p, Method* method, CompLevel cur_level, bool disable_feedback) {
CompLevel next_level = cur_level;
int i = method->invocation_count();
int b = method->backedge_count();
- if (is_trivial(method) && cur_level != CompLevel_aot) {
+ if (is_trivial(method)) {
next_level = CompLevel_simple;
} else {
switch(cur_level) {
- case CompLevel_aot: {
- if ((this->*p)(i, b, cur_level, method)) {
+ default: break;
+ case CompLevel_aot: {
+ // If we were at full profile level, would we switch to full opt?
+ if (common(p, method, CompLevel_full_profile, disable_feedback) == CompLevel_full_optimization) {
+ next_level = CompLevel_full_optimization;
+ } else if (disable_feedback || (CompileBroker::queue_size(CompLevel_full_optimization) <=
+ Tier3DelayOff * compiler_count(CompLevel_full_optimization) &&
+ (this->*p)(i, b, cur_level, method))) {
next_level = CompLevel_full_profile;
}
}
break;
case CompLevel_none:
// If we were at full profile level, would we switch to full opt?
- if (common(p, method, CompLevel_full_profile) == CompLevel_full_optimization) {
+ if (common(p, method, CompLevel_full_profile, disable_feedback) == CompLevel_full_optimization) {
next_level = CompLevel_full_optimization;
} else if ((this->*p)(i, b, cur_level, method)) {
- next_level = CompLevel_full_profile;
+#if INCLUDE_JVMCI
+ if (EnableJVMCI && UseJVMCICompiler) {
+ // Since JVMCI takes a while to warm up, its queue inevitably backs up during
+ // early VM execution. As of 2014-06-13, JVMCI's inliner assumes that the root
+ // compilation method and all potential inlinees have mature profiles (which
+ // includes type profiling). If it sees immature profiles, JVMCI's inliner
+ // can perform pathologically bad (e.g., causing OutOfMemoryErrors due to
+ // exploring/inlining too many graphs). Since a rewrite of the inliner is
+ // in progress, we simply disable the dialing back heuristic for now and will
+ // revisit this decision once the new inliner is completed.
+ next_level = CompLevel_full_profile;
+ } else
+#endif
+ {
+ // C1-generated fully profiled code is about 30% slower than the limited profile
+ // code that has only invocation and backedge counters. The observation is that
+ // if C2 queue is large enough we can spend too much time in the fully profiled code
+ // while waiting for C2 to pick the method from the queue. To alleviate this problem
+ // we introduce a feedback on the C2 queue size. If the C2 queue is sufficiently long
+ // we choose to compile a limited profiled version and then recompile with full profiling
+ // when the load on C2 goes down.
+ if (!disable_feedback && CompileBroker::queue_size(CompLevel_full_optimization) >
+ Tier3DelayOn * compiler_count(CompLevel_full_optimization)) {
+ next_level = CompLevel_limited_profile;
+ } else {
+ next_level = CompLevel_full_profile;
+ }
+ }
}
break;
case CompLevel_limited_profile:
+ if (is_method_profiled(method)) {
+ // Special case: we got here because this method was fully profiled in the interpreter.
+ next_level = CompLevel_full_optimization;
+ } else {
+ MethodData* mdo = method->method_data();
+ if (mdo != NULL) {
+ if (mdo->would_profile()) {
+ if (disable_feedback || (CompileBroker::queue_size(CompLevel_full_optimization) <=
+ Tier3DelayOff * compiler_count(CompLevel_full_optimization) &&
+ (this->*p)(i, b, cur_level, method))) {
+ next_level = CompLevel_full_profile;
+ }
+ } else {
+ next_level = CompLevel_full_optimization;
+ }
+ } else {
+ // If there is no MDO we need to profile
+ if (disable_feedback || (CompileBroker::queue_size(CompLevel_full_optimization) <=
+ Tier3DelayOff * compiler_count(CompLevel_full_optimization) &&
+ (this->*p)(i, b, cur_level, method))) {
+ next_level = CompLevel_full_profile;
+ }
+ }
+ }
+ break;
case CompLevel_full_profile:
{
MethodData* mdo = method->method_data();
@@ -382,17 +733,15 @@
}
}
break;
- default:
- break;
}
}
return MIN2(next_level, (CompLevel)TieredStopAtLevel);
}
// Determine if a method should be compiled with a normal entry point at a different level.
-CompLevel SimpleThresholdPolicy::call_event(Method* method, CompLevel cur_level, JavaThread* thread) {
+CompLevel SimpleThresholdPolicy::call_event(Method* method, CompLevel cur_level, JavaThread * thread) {
CompLevel osr_level = MIN2((CompLevel) method->highest_osr_comp_level(),
- common(&SimpleThresholdPolicy::loop_predicate, method, cur_level));
+ common(&SimpleThresholdPolicy::loop_predicate, method, cur_level, true));
CompLevel next_level = common(&SimpleThresholdPolicy::call_predicate, method, cur_level);
// If OSR method level is greater than the regular method level, the levels should be
@@ -417,7 +766,7 @@
// Determine if we should do an OSR compilation of a given method.
CompLevel SimpleThresholdPolicy::loop_event(Method* method, CompLevel cur_level, JavaThread* thread) {
- CompLevel next_level = common(&SimpleThresholdPolicy::loop_predicate, method, cur_level);
+ CompLevel next_level = common(&SimpleThresholdPolicy::loop_predicate, method, cur_level, true);
if (cur_level == CompLevel_none) {
// If there is a live OSR method that means that we deopted to the interpreter
// for the transition.
@@ -434,13 +783,39 @@
return next_level;
}
+bool SimpleThresholdPolicy::maybe_switch_to_aot(const methodHandle& mh, CompLevel cur_level, CompLevel next_level, JavaThread* thread) {
+ if (UseAOT && !delay_compilation_during_startup()) {
+ if (cur_level == CompLevel_full_profile || cur_level == CompLevel_none) {
+ // If the current level is full profile or interpreter and we're switching to any other level,
+ // activate the AOT code back first so that we won't waste time overprofiling.
+ compile(mh, InvocationEntryBci, CompLevel_aot, thread);
+ // Fall through for JIT compilation.
+ }
+ if (next_level == CompLevel_limited_profile && cur_level != CompLevel_aot && mh->has_aot_code()) {
+ // If the next level is limited profile, use the aot code (if there is any),
+ // since it's essentially the same thing.
+ compile(mh, InvocationEntryBci, CompLevel_aot, thread);
+ // Not need to JIT, we're done.
+ return true;
+ }
+ }
+ return false;
+}
+
// Handle the invocation event.
void SimpleThresholdPolicy::method_invocation_event(const methodHandle& mh, const methodHandle& imh,
- CompLevel level, CompiledMethod* nm, JavaThread* thread) {
- if (is_compilation_enabled() && !CompileBroker::compilation_is_in_queue(mh)) {
- CompLevel next_level = call_event(mh(), level, thread);
- if (next_level != level) {
+ CompLevel level, CompiledMethod* nm, JavaThread* thread) {
+ if (should_create_mdo(mh(), level)) {
+ create_mdo(mh, thread);
+ }
+ CompLevel next_level = call_event(mh(), level, thread);
+ if (next_level != level) {
+ if (maybe_switch_to_aot(mh, level, next_level, thread)) {
+ // No JITting necessary
+ return;
+ }
+ if (is_compilation_enabled() && !CompileBroker::compilation_is_in_queue(mh)) {
compile(mh, InvocationEntryBci, next_level, thread);
}
}
@@ -450,25 +825,77 @@
// with a regular entry from here.
void SimpleThresholdPolicy::method_back_branch_event(const methodHandle& mh, const methodHandle& imh,
int bci, CompLevel level, CompiledMethod* nm, JavaThread* thread) {
- // If the method is already compiling, quickly bail out.
- if (is_compilation_enabled() && !CompileBroker::compilation_is_in_queue(mh)) {
- // Use loop event as an opportunity to also check there's been
- // enough calls.
- CompLevel cur_level = comp_level(mh());
- CompLevel next_level = call_event(mh(), cur_level, thread);
- CompLevel next_osr_level = loop_event(mh(), level, thread);
+ if (should_create_mdo(mh(), level)) {
+ create_mdo(mh, thread);
+ }
+ // Check if MDO should be created for the inlined method
+ if (should_create_mdo(imh(), level)) {
+ create_mdo(imh, thread);
+ }
- next_level = MAX2(next_level,
- next_osr_level < CompLevel_full_optimization ? next_osr_level : cur_level);
- bool is_compiling = false;
- if (next_level != cur_level) {
- compile(mh, InvocationEntryBci, next_level, thread);
- is_compiling = true;
+ if (is_compilation_enabled()) {
+ CompLevel next_osr_level = loop_event(imh(), level, thread);
+ CompLevel max_osr_level = (CompLevel)imh->highest_osr_comp_level();
+ // At the very least compile the OSR version
+ if (!CompileBroker::compilation_is_in_queue(imh) && (next_osr_level != level)) {
+ compile(imh, bci, next_osr_level, thread);
}
- // Do the OSR version
- if (!is_compiling && next_osr_level != level) {
- compile(mh, bci, next_osr_level, thread);
+ // Use loop event as an opportunity to also check if there's been
+ // enough calls.
+ CompLevel cur_level, next_level;
+ if (mh() != imh()) { // If there is an enclosing method
+ if (level == CompLevel_aot) {
+ // Recompile the enclosing method to prevent infinite OSRs. Stay at AOT level while it's compiling.
+ if (max_osr_level != CompLevel_none && !CompileBroker::compilation_is_in_queue(mh)) {
+ compile(mh, InvocationEntryBci, MIN2((CompLevel)TieredStopAtLevel, CompLevel_full_profile), thread);
+ }
+ } else {
+ // Current loop event level is not AOT
+ guarantee(nm != NULL, "Should have nmethod here");
+ cur_level = comp_level(mh());
+ next_level = call_event(mh(), cur_level, thread);
+
+ if (max_osr_level == CompLevel_full_optimization) {
+ // The inlinee OSRed to full opt, we need to modify the enclosing method to avoid deopts
+ bool make_not_entrant = false;
+ if (nm->is_osr_method()) {
+ // This is an osr method, just make it not entrant and recompile later if needed
+ make_not_entrant = true;
+ } else {
+ if (next_level != CompLevel_full_optimization) {
+ // next_level is not full opt, so we need to recompile the
+ // enclosing method without the inlinee
+ cur_level = CompLevel_none;
+ make_not_entrant = true;
+ }
+ }
+ if (make_not_entrant) {
+ if (PrintTieredEvents) {
+ int osr_bci = nm->is_osr_method() ? nm->osr_entry_bci() : InvocationEntryBci;
+ print_event(MAKE_NOT_ENTRANT, mh(), mh(), osr_bci, level);
+ }
+ nm->make_not_entrant();
+ }
+ }
+ // Fix up next_level if necessary to avoid deopts
+ if (next_level == CompLevel_limited_profile && max_osr_level == CompLevel_full_profile) {
+ next_level = CompLevel_full_profile;
+ }
+ if (cur_level != next_level) {
+ if (!maybe_switch_to_aot(mh, cur_level, next_level, thread) && !CompileBroker::compilation_is_in_queue(mh)) {
+ compile(mh, InvocationEntryBci, next_level, thread);
+ }
+ }
+ }
+ } else {
+ cur_level = comp_level(mh());
+ next_level = call_event(mh(), cur_level, thread);
+ if (next_level != cur_level) {
+ if (!maybe_switch_to_aot(mh, cur_level, next_level, thread) && !CompileBroker::compilation_is_in_queue(mh)) {
+ compile(mh, InvocationEntryBci, next_level, thread);
+ }
+ }
}
}
}
--- a/src/hotspot/share/runtime/simpleThresholdPolicy.hpp Wed May 09 07:48:31 2018 +0100
+++ b/src/hotspot/share/runtime/simpleThresholdPolicy.hpp Wed May 09 09:39:25 2018 +0200
@@ -34,8 +34,136 @@
class CompileTask;
class CompileQueue;
+/*
+ * The system supports 5 execution levels:
+ * * level 0 - interpreter
+ * * level 1 - C1 with full optimization (no profiling)
+ * * level 2 - C1 with invocation and backedge counters
+ * * level 3 - C1 with full profiling (level 2 + MDO)
+ * * level 4 - C2
+ *
+ * Levels 0, 2 and 3 periodically notify the runtime about the current value of the counters
+ * (invocation counters and backedge counters). The frequency of these notifications is
+ * different at each level. These notifications are used by the policy to decide what transition
+ * to make.
+ *
+ * Execution starts at level 0 (interpreter), then the policy can decide either to compile the
+ * method at level 3 or level 2. The decision is based on the following factors:
+ * 1. The length of the C2 queue determines the next level. The observation is that level 2
+ * is generally faster than level 3 by about 30%, therefore we would want to minimize the time
+ * a method spends at level 3. We should only spend the time at level 3 that is necessary to get
+ * adequate profiling. So, if the C2 queue is long enough it is more beneficial to go first to
+ * level 2, because if we transitioned to level 3 we would be stuck there until our C2 compile
+ * request makes its way through the long queue. When the load on C2 recedes we are going to
+ * recompile at level 3 and start gathering profiling information.
+ * 2. The length of C1 queue is used to dynamically adjust the thresholds, so as to introduce
+ * additional filtering if the compiler is overloaded. The rationale is that by the time a
+ * method gets compiled it can become unused, so it doesn't make sense to put too much onto the
+ * queue.
+ *
+ * After profiling is completed at level 3 the transition is made to level 4. Again, the length
+ * of the C2 queue is used as a feedback to adjust the thresholds.
+ *
+ * After the first C1 compile some basic information is determined about the code like the number
+ * of the blocks and the number of the loops. Based on that it can be decided that a method
+ * is trivial and compiling it with C1 will yield the same code. In this case the method is
+ * compiled at level 1 instead of 4.
+ *
+ * We also support profiling at level 0. If C1 is slow enough to produce the level 3 version of
+ * the code and the C2 queue is sufficiently small we can decide to start profiling in the
+ * interpreter (and continue profiling in the compiled code once the level 3 version arrives).
+ * If the profiling at level 0 is fully completed before level 3 version is produced, a level 2
+ * version is compiled instead in order to run faster waiting for a level 4 version.
+ *
+ * Compile queues are implemented as priority queues - for each method in the queue we compute
+ * the event rate (the number of invocation and backedge counter increments per unit of time).
+ * When getting an element off the queue we pick the one with the largest rate. Maintaining the
+ * rate also allows us to remove stale methods (the ones that got on the queue but stopped
+ * being used shortly after that).
+*/
+
+/* Command line options:
+ * - Tier?InvokeNotifyFreqLog and Tier?BackedgeNotifyFreqLog control the frequency of method
+ * invocation and backedge notifications. Basically every n-th invocation or backedge a mutator thread
+ * makes a call into the runtime.
+ *
+ * - Tier?InvocationThreshold, Tier?CompileThreshold, Tier?BackEdgeThreshold, Tier?MinInvocationThreshold control
+ * compilation thresholds.
+ * Level 2 thresholds are not used and are provided for option-compatibility and potential future use.
+ * Other thresholds work as follows:
+ *
+ * Transition from interpreter (level 0) to C1 with full profiling (level 3) happens when
+ * the following predicate is true (X is the level):
+ *
+ * i > TierXInvocationThreshold * s || (i > TierXMinInvocationThreshold * s && i + b > TierXCompileThreshold * s),
+ *
+ * where $i$ is the number of method invocations, $b$ number of backedges and $s$ is the scaling
+ * coefficient that will be discussed further.
+ * The intuition is to equalize the time that is spend profiling each method.
+ * The same predicate is used to control the transition from level 3 to level 4 (C2). It should be
+ * noted though that the thresholds are relative. Moreover i and b for the 0->3 transition come
+ * from Method* and for 3->4 transition they come from MDO (since profiled invocations are
+ * counted separately). Finally, if a method does not contain anything worth profiling, a transition
+ * from level 3 to level 4 occurs without considering thresholds (e.g., with fewer invocations than
+ * what is specified by Tier4InvocationThreshold).
+ *
+ * OSR transitions are controlled simply with b > TierXBackEdgeThreshold * s predicates.
+ *
+ * - Tier?LoadFeedback options are used to automatically scale the predicates described above depending
+ * on the compiler load. The scaling coefficients are computed as follows:
+ *
+ * s = queue_size_X / (TierXLoadFeedback * compiler_count_X) + 1,
+ *
+ * where queue_size_X is the current size of the compiler queue of level X, and compiler_count_X
+ * is the number of level X compiler threads.
+ *
+ * Basically these parameters describe how many methods should be in the compile queue
+ * per compiler thread before the scaling coefficient increases by one.
+ *
+ * This feedback provides the mechanism to automatically control the flow of compilation requests
+ * depending on the machine speed, mutator load and other external factors.
+ *
+ * - Tier3DelayOn and Tier3DelayOff parameters control another important feedback loop.
+ * Consider the following observation: a method compiled with full profiling (level 3)
+ * is about 30% slower than a method at level 2 (just invocation and backedge counters, no MDO).
+ * Normally, the following transitions will occur: 0->3->4. The problem arises when the C2 queue
+ * gets congested and the 3->4 transition is delayed. While the method is the C2 queue it continues
+ * executing at level 3 for much longer time than is required by the predicate and at suboptimal speed.
+ * The idea is to dynamically change the behavior of the system in such a way that if a substantial
+ * load on C2 is detected we would first do the 0->2 transition allowing a method to run faster.
+ * And then when the load decreases to allow 2->3 transitions.
+ *
+ * Tier3Delay* parameters control this switching mechanism.
+ * Tier3DelayOn is the number of methods in the C2 queue per compiler thread after which the policy
+ * no longer does 0->3 transitions but does 0->2 transitions instead.
+ * Tier3DelayOff switches the original behavior back when the number of methods in the C2 queue
+ * per compiler thread falls below the specified amount.
+ * The hysteresis is necessary to avoid jitter.
+ *
+ * - TieredCompileTaskTimeout is the amount of time an idle method can spend in the compile queue.
+ * Basically, since we use the event rate d(i + b)/dt as a value of priority when selecting a method to
+ * compile from the compile queue, we also can detect stale methods for which the rate has been
+ * 0 for some time in the same iteration. Stale methods can appear in the queue when an application
+ * abruptly changes its behavior.
+ *
+ * - TieredStopAtLevel, is used mostly for testing. It allows to bypass the policy logic and stick
+ * to a given level. For example it's useful to set TieredStopAtLevel = 1 in order to compile everything
+ * with pure c1.
+ *
+ * - Tier0ProfilingStartPercentage allows the interpreter to start profiling when the inequalities in the
+ * 0->3 predicate are already exceeded by the given percentage but the level 3 version of the
+ * method is still not ready. We can even go directly from level 0 to 4 if c1 doesn't produce a compiled
+ * version in time. This reduces the overall transition to level 4 and decreases the startup time.
+ * Note that this behavior is also guarded by the Tier3Delay mechanism: when the c2 queue is too long
+ * these is not reason to start profiling prematurely.
+ *
+ * - TieredRateUpdateMinTime and TieredRateUpdateMaxTime are parameters of the rate computation.
+ * Basically, the rate is not computed more frequently than TieredRateUpdateMinTime and is considered
+ * to be zero if no events occurred in TieredRateUpdateMaxTime.
+ */
class SimpleThresholdPolicy : public CompilationPolicy {
+ jlong _start_time;
int _c1_count, _c2_count;
// Check if the counter is big enough and set carry (effectively infinity).
@@ -49,7 +177,7 @@
bool call_predicate(int i, int b, CompLevel cur_level, Method* method);
bool loop_predicate(int i, int b, CompLevel cur_level, Method* method);
// Common transition function. Given a predicate determines if a method should transition to another level.
- CompLevel common(Predicate p, Method* method, CompLevel cur_level);
+ CompLevel common(Predicate p, Method* method, CompLevel cur_level, bool disable_feedback = false);
// Transition functions.
// call_event determines if a method should be compiled at a different
// level with a regular invocation entry.
@@ -58,6 +186,35 @@
// level.
CompLevel loop_event(Method* method, CompLevel cur_level, JavaThread* thread);
void print_counters(const char* prefix, const methodHandle& mh);
+ // Has a method been long around?
+ // We don't remove old methods from the compile queue even if they have
+ // very low activity (see select_task()).
+ inline bool is_old(Method* method);
+ // Was a given method inactive for a given number of milliseconds.
+ // If it is, we would remove it from the queue (see select_task()).
+ inline bool is_stale(jlong t, jlong timeout, Method* m);
+ // Compute the weight of the method for the compilation scheduling
+ inline double weight(Method* method);
+ // Apply heuristics and return true if x should be compiled before y
+ inline bool compare_methods(Method* x, Method* y);
+ // Compute event rate for a given method. The rate is the number of event (invocations + backedges)
+ // per millisecond.
+ inline void update_rate(jlong t, Method* m);
+ // Compute threshold scaling coefficient
+ inline double threshold_scale(CompLevel level, int feedback_k);
+ // If a method is old enough and is still in the interpreter we would want to
+ // start profiling without waiting for the compiled method to arrive. This function
+ // determines whether we should do that.
+ inline bool should_create_mdo(Method* method, CompLevel cur_level);
+ // Create MDO if necessary.
+ void create_mdo(const methodHandle& mh, JavaThread* thread);
+ // Is method profiled enough?
+ bool is_method_profiled(Method* method);
+
+ double _increase_threshold_at_ratio;
+
+ bool maybe_switch_to_aot(const methodHandle& mh, CompLevel cur_level, CompLevel next_level, JavaThread* thread);
+
protected:
int c1_count() const { return _c1_count; }
int c2_count() const { return _c2_count; }
@@ -67,7 +224,7 @@
enum EventType { CALL, LOOP, COMPILE, REMOVE_FROM_QUEUE, UPDATE_IN_QUEUE, REPROFILE, MAKE_NOT_ENTRANT };
void print_event(EventType type, const methodHandle& mh, const methodHandle& imh, int bci, CompLevel level);
// Print policy-specific information if necessary
- virtual void print_specific(EventType type, const methodHandle& mh, const methodHandle& imh, int bci, CompLevel level) { }
+ virtual void print_specific(EventType type, const methodHandle& mh, const methodHandle& imh, int bci, CompLevel level);
// Check if the method can be compiled, change level if necessary
void compile(const methodHandle& mh, int bci, CompLevel level, JavaThread* thread);
// Submit a given method for compilation
@@ -87,8 +244,13 @@
CompLevel level, CompiledMethod* nm, JavaThread* thread);
virtual void method_back_branch_event(const methodHandle& method, const methodHandle& inlinee,
int bci, CompLevel level, CompiledMethod* nm, JavaThread* thread);
+
+ void set_increase_threshold_at_ratio() { _increase_threshold_at_ratio = 100 / (100 - (double)IncreaseFirstTierCompileThresholdAt); }
+ void set_start_time(jlong t) { _start_time = t; }
+ jlong start_time() const { return _start_time; }
+
public:
- SimpleThresholdPolicy() : _c1_count(0), _c2_count(0) { }
+ SimpleThresholdPolicy() : _start_time(0), _c1_count(0), _c2_count(0) { }
virtual int compiler_count(CompLevel comp_level) {
if (is_c1_compile(comp_level)) return c1_count();
if (is_c2_compile(comp_level)) return c2_count();
@@ -107,11 +269,7 @@
virtual bool is_mature(Method* method);
// Initialize: set compiler thread count
virtual void initialize();
- virtual bool should_not_inline(ciEnv* env, ciMethod* callee) {
- return (env->comp_level() == CompLevel_limited_profile ||
- env->comp_level() == CompLevel_full_profile) &&
- callee->has_loops();
- }
+ virtual bool should_not_inline(ciEnv* env, ciMethod* callee);
};
#endif // TIERED
--- a/test/hotspot/jtreg/compiler/aot/RecompilationTest.java Wed May 09 07:48:31 2018 +0100
+++ b/test/hotspot/jtreg/compiler/aot/RecompilationTest.java Wed May 09 09:39:25 2018 +0200
@@ -37,26 +37,12 @@
* -extraopt -XX:+UnlockDiagnosticVMOptions -extraopt -XX:+WhiteBoxAPI -extraopt -Xbootclasspath/a:.
* -extraopt -XX:-UseCompressedOops
* -extraopt -XX:CompileCommand=dontinline,compiler.whitebox.SimpleTestCaseHelper::*
- * @run main/othervm -Xmixed -Xbatch -XX:+UseAOT -XX:+TieredCompilation -XX:CompilationPolicyChoice=2
- * -XX:-UseCounterDecay -XX:-UseCompressedOops
- * -XX:-Inline
- * -XX:AOTLibrary=./libRecompilationTest1.so -Xbootclasspath/a:.
- * -XX:+UnlockDiagnosticVMOptions -XX:+WhiteBoxAPI
- * -Dcompiler.aot.RecompilationTest.check_level=1
- * compiler.aot.RecompilationTest
* @run driver compiler.aot.AotCompiler -libname libRecompilationTest2.so
* -class compiler.whitebox.SimpleTestCaseHelper
* -extraopt -Dgraal.TieredAOT=false
* -extraopt -XX:+UnlockDiagnosticVMOptions -extraopt -XX:+WhiteBoxAPI -extraopt -Xbootclasspath/a:.
* -extraopt -XX:-UseCompressedOops
* -extraopt -XX:CompileCommand=dontinline,compiler.whitebox.SimpleTestCaseHelper::*
- * @run main/othervm -Xmixed -Xbatch -XX:+UseAOT -XX:+TieredCompilation -XX:CompilationPolicyChoice=2
- * -XX:-UseCounterDecay -XX:-UseCompressedOops
- * -XX:-Inline
- * -XX:AOTLibrary=./libRecompilationTest2.so -Xbootclasspath/a:.
- * -XX:+UnlockDiagnosticVMOptions -XX:+WhiteBoxAPI
- * -Dcompiler.aot.RecompilationTest.check_level=-1
- * compiler.aot.RecompilationTest
* @run main/othervm -Xmixed -Xbatch -XX:+UseAOT -XX:-TieredCompilation
* -XX:-UseCounterDecay -XX:-UseCompressedOops
* -XX:-Inline
--- a/test/hotspot/jtreg/compiler/tiered/ConstantGettersTransitionsTest.java Wed May 09 07:48:31 2018 +0100
+++ b/test/hotspot/jtreg/compiler/tiered/ConstantGettersTransitionsTest.java Wed May 09 09:39:25 2018 +0200
@@ -34,7 +34,6 @@
* @run main/othervm/timeout=240 -Xmixed -Xbootclasspath/a:. -XX:+UnlockDiagnosticVMOptions
* -XX:+WhiteBoxAPI -XX:+TieredCompilation -XX:-UseCounterDecay
* -XX:CompileCommand=compileonly,compiler.tiered.ConstantGettersTransitionsTest$ConstantGettersTestCase$TrivialMethods::*
- * compiler.tiered.TransitionsTestExecutor
* compiler.tiered.ConstantGettersTransitionsTest
*/
@@ -200,4 +199,4 @@
}
}
}
-}
\ No newline at end of file
+}
--- a/test/hotspot/jtreg/compiler/tiered/LevelTransitionTest.java Wed May 09 07:48:31 2018 +0100
+++ b/test/hotspot/jtreg/compiler/tiered/LevelTransitionTest.java Wed May 09 09:39:25 2018 +0200
@@ -36,7 +36,6 @@
* -XX:+WhiteBoxAPI -XX:+TieredCompilation -XX:-UseCounterDecay
* -XX:CompileCommand=compileonly,compiler.whitebox.SimpleTestCaseHelper::*
* -XX:CompileCommand=compileonly,compiler.tiered.LevelTransitionTest$ExtendedTestCase$CompileMethodHolder::*
- * compiler.tiered.TransitionsTestExecutor
* compiler.tiered.LevelTransitionTest
*/
--- a/test/hotspot/jtreg/compiler/tiered/TransitionsTestExecutor.java Wed May 09 07:48:31 2018 +0100
+++ /dev/null Thu Jan 01 00:00:00 1970 +0000
@@ -1,66 +0,0 @@
-/*
- * Copyright (c) 2014, 2016, 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 compiler.tiered;
-
-import compiler.whitebox.CompilerWhiteBoxTest;
-import jdk.test.lib.process.OutputAnalyzer;
-import jdk.test.lib.process.ProcessTools;
-
-import java.lang.management.ManagementFactory;
-import java.lang.management.RuntimeMXBean;
-import java.util.ArrayList;
-import java.util.Collections;
-import java.util.List;
-
-/**
- * Executes given test in a separate VM with enabled Tiered Compilation for
- * CompilationPolicyChoice 2 and 3
- */
-public class TransitionsTestExecutor {
- public static void main(String[] args) throws Throwable {
- if (CompilerWhiteBoxTest.skipOnTieredCompilation(false)) {
- return;
- }
- if (args.length != 1) {
- throw new Error("TESTBUG: Test name should be specified");
- }
- executeTestFor(2, args[0]);
- executeTestFor(3, args[0]);
- }
-
- private static void executeTestFor(int compilationPolicy, String testName) throws Throwable {
- String policy = "-XX:CompilationPolicyChoice=" + compilationPolicy;
-
- // Get runtime arguments including VM options given to this executor
- RuntimeMXBean runtime = ManagementFactory.getRuntimeMXBean();
- List<String> vmArgs = runtime.getInputArguments();
-
- // Construct execution command with compilation policy choice and test name
- List<String> args = new ArrayList<>(vmArgs);
- Collections.addAll(args, policy, testName);
-
- OutputAnalyzer out = ProcessTools.executeTestJvm(args.toArray(new String[args.size()]));
- out.shouldHaveExitValue(0);
- }
-}