On Wed, 5 Apr 2023 10:40:54 GMT, Roman Kennke <rken...@openjdk.org> wrote:
>> This change adds a fast-locking scheme as an alternative to the current >> stack-locking implementation. It retains the advantages of stack-locking >> (namely fast locking in uncontended code-paths), while avoiding the overload >> of the mark word. That overloading causes massive problems with Lilliput, >> because it means we have to check and deal with this situation when trying >> to access the mark-word. And because of the very racy nature, this turns out >> to be very complex and would involve a variant of the inflation protocol to >> ensure that the object header is stable. (The current implementation of >> setting/fetching the i-hash provides a glimpse into the complexity). >> >> What the original stack-locking does is basically to push a stack-lock onto >> the stack which consists only of the displaced header, and CAS a pointer to >> this stack location into the object header (the lowest two header bits being >> 00 indicate 'stack-locked'). The pointer into the stack can then be used to >> identify which thread currently owns the lock. >> >> This change basically reverses stack-locking: It still CASes the lowest two >> header bits to 00 to indicate 'fast-locked' but does *not* overload the >> upper bits with a stack-pointer. Instead, it pushes the object-reference to >> a thread-local lock-stack. This is a new structure which is basically a >> small array of oops that is associated with each thread. Experience shows >> that this array typcially remains very small (3-5 elements). Using this lock >> stack, it is possible to query which threads own which locks. Most >> importantly, the most common question 'does the current thread own me?' is >> very quickly answered by doing a quick scan of the array. More complex >> queries like 'which thread owns X?' are not performed in very >> performance-critical paths (usually in code like JVMTI or deadlock >> detection) where it is ok to do more complex operations (and we already do). >> The lock-stack is also a new set of GC roots, and would be scanned during >> thread scanning, possibly concurrently, via the normal protocols. >> >> The lock-stack is fixed size, currently with 8 elements. According to my >> experiments with various workloads, this covers the vast majority of >> workloads (in-fact, most workloads seem to never exceed 5 active locks per >> thread at a time). We check for overflow in the fast-paths and when the >> lock-stack is full, we take the slow-path, which would inflate the lock to a >> monitor. That case should be very rare. >> >> In contrast to stack-locking, fast-locking does *not* support recursive >> locking (yet). When that happens, the fast-lock gets inflated to a full >> monitor. It is not clear if it is worth to add support for recursive >> fast-locking. >> >> One trouble is that when a contending thread arrives at a fast-locked >> object, it must inflate the fast-lock to a full monitor. Normally, we need >> to know the current owning thread, and record that in the monitor, so that >> the contending thread can wait for the current owner to properly exit the >> monitor. However, fast-locking doesn't have this information. What we do >> instead is to record a special marker ANONYMOUS_OWNER. When the thread that >> currently holds the lock arrives at monitorexit, and observes >> ANONYMOUS_OWNER, it knows it must be itself, fixes the owner to be itself, >> and then properly exits the monitor, and thus handing over to the contending >> thread. >> >> As an alternative, I considered to remove stack-locking altogether, and only >> use heavy monitors. In most workloads this did not show measurable >> regressions. However, in a few workloads, I have observed severe >> regressions. All of them have been using old synchronized Java collections >> (Vector, Stack), StringBuffer or similar code. The combination of two >> conditions leads to regressions without stack- or fast-locking: 1. The >> workload synchronizes on uncontended locks (e.g. single-threaded use of >> Vector or StringBuffer) and 2. The workload churns such locks. IOW, >> uncontended use of Vector, StringBuffer, etc as such is ok, but creating >> lots of such single-use, single-threaded-locked objects leads to massive >> ObjectMonitor churn, which can lead to a significant performance impact. But >> alas, such code exists, and we probably don't want to punish it if we can >> avoid it. >> >> This change enables to simplify (and speed-up!) a lot of code: >> >> - The inflation protocol is no longer necessary: we can directly CAS the >> (tagged) ObjectMonitor pointer to the object header. >> - Accessing the hashcode could now be done in the fastpath always, if the >> hashcode has been installed. Fast-locked headers can be used directly, for >> monitor-locked objects we can easily reach-through to the displaced header. >> This is safe because Java threads participate in monitor deflation protocol. >> This would be implemented in a separate PR >> >> Also, and I might be mistaken here, this new lightweight locking would make >> synchronized work better with Loom: Because the lock-records are no longer >> scattered across the stack, but instead are densely packed into the >> lock-stack, it should be easy for a vthread to save its lock-stack upon >> unmounting and restore it when re-mounting. However, I am not sure about >> this, and this PR does not attempt to implement that support. >> >> Testing: >> - [x] tier1 x86_64 x aarch64 x +UseFastLocking >> - [x] tier2 x86_64 x aarch64 x +UseFastLocking >> - [x] tier3 x86_64 x aarch64 x +UseFastLocking >> - [x] tier4 x86_64 x aarch64 x +UseFastLocking >> - [x] tier1 x86_64 x aarch64 x -UseFastLocking >> - [x] tier2 x86_64 x aarch64 x -UseFastLocking >> - [x] tier3 x86_64 x aarch64 x -UseFastLocking >> - [x] tier4 x86_64 x aarch64 x -UseFastLocking >> - [x] Several real-world applications have been tested with this change in >> tandem with Lilliput without any problems, yet >> >> ### Performance >> >> #### Simple Microbenchmark >> >> The microbenchmark exercises only the locking primitives for monitorenter >> and monitorexit, without contention. The benchmark can be found >> (here)[https://github.com/rkennke/fastlockbench]. Numbers are in ns/ops. >> >> | | x86_64 | aarch64 | >> | -- | -- | -- | >> | -UseFastLocking | 20.651 | 20.764 | >> | +UseFastLocking | 18.896 | 18.908 | >> >> >> #### Renaissance >> >> | x86_64 | | | | aarch64 | | >> -- | -- | -- | -- | -- | -- | -- | -- >> | stack-locking | fast-locking | | | stack-locking | fast-locking | >> AkkaUct | 841.884 | 836.948 | 0.59% | | 1475.774 | 1465.647 | 0.69% >> Reactors | 11041.427 | 11181.451 | -1.25% | | 11381.751 | 11521.318 | >> -1.21% >> Als | 1367.183 | 1359.358 | 0.58% | | 1678.103 | 1688.067 | -0.59% >> ChiSquare | 577.021 | 577.398 | -0.07% | | 986.619 | 988.063 | -0.15% >> GaussMix | 817.459 | 819.073 | -0.20% | | 1154.293 | 1155.522 | -0.11% >> LogRegression | 598.343 | 603.371 | -0.83% | | 638.052 | 644.306 | -0.97% >> MovieLens | 8248.116 | 8314.576 | -0.80% | | 7569.219 | 7646.828 | -1.01%% >> NaiveBayes | 587.607 | 581.608 | 1.03% | | 541.583 | 550.059 | -1.54% >> PageRank | 3260.553 | 3263.472 | -0.09% | | 4376.405 | 4381.101 | -0.11% >> FjKmeans | 979.978 | 976.122 | 0.40% | | 774.312 | 771.235 | 0.40% >> FutureGenetic | 2187.369 | 2183.271 | 0.19% | | 2685.722 | 2689.056 | >> -0.12% >> ParMnemonics | 2434.551 | 2468.763 | -1.39% | | 4278.225 | 4263.863 | 0.34% >> Scrabble | 111.882 | 111.768 | 0.10% | | 151.796 | 153.959 | -1.40% >> RxScrabble | 210.252 | 211.38 | -0.53% | | 310.116 | 315.594 | -1.74% >> Dotty | 750.415 | 752.658 | -0.30% | | 1033.636 | 1036.168 | -0.24% >> ScalaDoku | 3072.05 | 3051.2 | 0.68% | | 3711.506 | 3690.04 | 0.58% >> ScalaKmeans | 211.427 | 209.957 | 0.70% | | 264.38 | 265.788 | -0.53% >> ScalaStmBench7 | 1017.795 | 1018.869 | -0.11% | | 1088.182 | 1092.266 | >> -0.37% >> Philosophers | 6450.124 | 6565.705 | -1.76% | | 12017.964 | 11902.559 | >> 0.97% >> FinagleChirper | 3953.623 | 3972.647 | -0.48% | | 4750.751 | 4769.274 | >> -0.39% >> FinagleHttp | 3970.526 | 4005.341 | -0.87% | | 5294.125 | 5296.224 | -0.04% > > Roman Kennke has updated the pull request incrementally with one additional > commit since the last revision: > > Named constants for LockingMode v52 is failing quite a few JVM/TI tests with crashes that look like this: # Internal Error (/opt/mach5/mesos/work_dir/slaves/741e9afd-8c02-45c3-b2e2-9db1450d0832-S40935/frameworks/1735e8a2-a1db-478c-8104-60c8b0af87dd-0196/executors/b154597d-2ba7-420d-81c1-ef13f408c137/runs/d52e181d-f011-47c8-a35f-30fcbba5c164/workspace/open/src/hotspot/share/runtime/javaThread.hpp:983), pid=1112738, tid=1112747 # assert(t->is_Java_thread()) failed: incorrect cast to JavaThread # # JRE version: Java(TM) SE Runtime Environment (21.0) (fastdebug build 21-internal-LTS-2023-04-04-2141101.daniel.daugherty.8291555forjdk21.git) # Java VM: Java HotSpot(TM) 64-Bit Server VM (fastdebug 21-internal-LTS-2023-04-04-2141101.daniel.daugherty.8291555forjdk21.git, compiled mode, sharing, tiered, compressed oops, compressed class ptrs, g1 gc, linux-aarch64) # Problematic frame: # V [libjvm.so+0x155baa4] is_lock_owned(Thread*, oop)+0x254 snip Stack: [0x0000fffdb8970000,0x0000fffdb8b70000], sp=0x0000fffdb8b6d480, free space=2037k Native frames: (J=compiled Java code, j=interpreted, Vv=VM code, C=native code) V [libjvm.so+0x155baa4] is_lock_owned(Thread*, oop)+0x254 (javaThread.hpp:983) V [libjvm.so+0x15627d0] ObjectSynchronizer::FastHashCode(Thread*, oop)+0x720 (synchronizer.cpp:956) V [libjvm.so+0x12dbedc] oopDesc::slow_identity_hash()+0x68 (oop.cpp:112) V [libjvm.so+0x105f044] JvmtiTagMapTable::find(oop)+0x1c4 (oop.inline.hpp:364) V [libjvm.so+0x1056744] CallbackWrapper::CallbackWrapper(JvmtiTagMap*, oop)+0xa4 (jvmtiTagMap.cpp:222) V [libjvm.so+0x105c538] CallbackInvoker::invoke_advanced_stack_ref_callback(jvmtiHeapReferenceKind, long, long, int, _jmethodID*, long, int, oop)+0x158 (jvmtiTagMap.cpp:1754) V [libjvm.so+0x105d688] JNILocalRootsClosure::do_oop(oop*)+0x168 (jvmtiTagMap.cpp:2019) V [libjvm.so+0xe8bac8] JNIHandleBlock::oops_do(OopClosure*)+0x68 (jniHandles.cpp:411) V [libjvm.so+0x105ddbc] VM_HeapWalkOperation::collect_stack_roots(JavaThread*, JNILocalRootsClosure*)+0x6dc (jvmtiTagMap.cpp:2725) V [libjvm.so+0x105e368] VM_HeapWalkOperation::collect_stack_roots()+0x138 (jvmtiTagMap.cpp:2772) V [libjvm.so+0x10555f0] VM_HeapWalkOperation::doit()+0x6e0 (jvmtiTagMap.cpp:2827) V [libjvm.so+0x1686ac0] VM_Operation::evaluate()+0x120 (vmOperations.cpp:71) V [libjvm.so+0x16b2710] VMThread::evaluate_operation(VM_Operation*)+0xd0 (vmThread.cpp:281) V [libjvm.so+0x16b3204] VMThread::inner_execute(VM_Operation*)+0x374 (vmThread.cpp:428) V [libjvm.so+0x16b33fc] VMThread::loop()+0x8c (vmThread.cpp:495) V [libjvm.so+0x16b352c] VMThread::run()+0x9c (vmThread.cpp:175) V [libjvm.so+0x15ad4b0] Thread::call_run()+0xac (thread.cpp:224) V [libjvm.so+0x130c0a8] thread_native_entry(Thread*)+0x134 (os_linux.cpp:740) C [libpthread.so.0+0x7908] start_thread+0x188 This code block in `ObjectSynchronizer::FastHashCode()`: // Fall thru so we only have one place that installs the hash in // the ObjectMonitor. } else if (LockingMode == 2 && mark.is_fast_locked() && is_lock_owned(current, obj)) { // This is a fast-lock owned by the calling thread so use the // markWord from the object. hash = mark.hash(); if (hash != 0) { // if it has a hash, just return it return hash; } } else if (LockingMode == 1 && mark.has_locker() && current->is_lock_owned((address)mark.locker())) { is calling this static function: static bool is_lock_owned(Thread* thread, oop obj) { assert(LockingMode == 2, "only call this with new lightweight locking enabled"); return JavaThread::cast(thread)->lock_stack().contains(obj); } and that function used to look like this: static bool is_lock_owned(Thread* thread, oop obj) { assert(LockingMode == 2, "only call this with new lightweight locking enabled"); - return thread->is_Java_thread() ? JavaThread::cast(thread)->lock_stack().contains(obj) : false; + return JavaThread::cast(thread)->lock_stack().contains(obj); } so that `thread->is_Java_thread()` check is needed since the VMThread is the one that's making this`is_lock_owned()` check as part of a hashcode operation. There are 129 test failures in Mach5 Tier4 and 769 test failures in Mach5 Tier5. I don't know yet whether all are due to: # Internal Error (/opt/mach5/mesos/work_dir/slaves/741e9afd-8c02-45c3-b2e2-9db1450d0832-S40935/frameworks/1735e8a2-a1db-478c-8104-60c8b0af87dd-0196/executors/b154597d-2ba7-420d-81c1-ef13f408c137/runs/d52e181d-f011-47c8-a35f-30fcbba5c164/workspace/open/src/hotspot/share/runtime/javaThread.hpp:983), pid=1112738, tid=1112747 # assert(t->is_Java_thread()) failed: incorrect cast to JavaThread ------------- PR Comment: https://git.openjdk.org/jdk/pull/10907#issuecomment-1497720340 PR Comment: https://git.openjdk.org/jdk/pull/10907#issuecomment-1497725623
