Pointer Synchronous Reclamation Beyond Concurrency TS2 – Maged Michael Protector Remover / Reclaimer Hazard pointers protect access to objects that may be removed concurrently. SAFE RECLAMATION Concurrency TS2 std::atomic& src) noexcept; }; hazard_pointer make_hazard_pointer(); Hazard Pointer Synchronous Reclamation Beyond Concurrency TS2 – Maged Michael See N4895 (wg21.link/n4895) for working draft of Concurrency oldptr->retire(); } Shared data, frequently-used, infrequently-updated Hazard Pointer Synchronous Reclamation Beyond Concurrency TS2 – Maged Michael Why use hazard pointers? • Fast (low nanoseconds) and scalable

pointer library for safe reclamation. } std::atomic src_;No need for extending C++26: ● Protection Counting (can be a topic for a future talk) ● Execution of Asynchronous Reclamation Proposed standard standard extensions: ● Synchronous reclamation ● Batch creation and destruction P3135R1: Hazard Pointer Extensions (beyond C++26) P3135R1: Hazard Pointer Extensions (wg21.link/p3135r1)Hazard Pointer Execution ReclamationPossible inline asynchronous reclamation Hazard Pointer Execution of Asynchronous Reclamation void worker() { /* ... */ obj->retire(); // Possible inline reclamation. } Using C++26 Using a separate

These extend the existing shared_ptr and the proposed atomic_shared_ptr which all have safe reclamation facilities. As such we also propose moving shared_ptr and atomic> to this new location Library ○ 33.1 General Concepts ■ 33.1.1 Thread Support ■ 33.1.2 Executor Support ● 33.2 Safe Reclamation ○ 33.2.1 Hazard Pointers ○ 33.2.2 RCU ○ 33.2.3 Latest/Snapshot? ○ 33.2.4 Asymmetric fences 15 ~hazard_pointer_domain(); }; hazard_pointer_domain& hazard_pointer_default_domain() noexcept; // For synchronous reclamation void hazard_pointer_clean_up( hazard_pointer_domain& domain = hazard_pointer_default_domain())

CPU preemption and jitter suppression for online services. Additionally, its innovative memory reclamation algorithm against out of memory (OOM) allows online services to run reliably based on their higher allows the kernel to dynamically change the preemption mode. Memory reclamation against OOM: In the event of OOM, memory reclamation is preferentially performed for process groups with low priorities performance and QoS for online services. openEuler is equipped with its own CPU scheduling and memory reclamation algorithms that are designed for hybrid deployments of cloud native services. These innovative

existing atomic are implemented (the split reference count technique) • Deferred reclamation, i.e., garbage collection in C++ Some assumed knowledge • You’ve heard of lock-free programming existing atomic are implemented (the split reference count technique) • Deferred reclamation, i.e., garbage collection in C++ Some assumed knowledge • You’ve heard of lock-free programming finished reading it Sounds like a textbook use case of deferred reclamation33 Daniel Anderson -- danielanderson.net Deferred reclamation: A quick primer atomic src T* protect(const std::atomic&

StratoVirt allocates and reclaims memory on the real-time memory demand of each workload. The memory reclamation speed can reach up to 3 GB/s through virtio-balloon. • Huge pages: On lightweight VMs, StratoVirt terms of the runtime, GC, and Just-In-Time (JIT) compiler. For example, memory allocation and reclamation in GC are optimized, and redundant memory barriers in JIT code are eliminated. In the benchmark

push(3); } // no more access to `vector`, // lock is unlocked crossbeam • Epoch-based memory reclamation • Easy translation of algorithms that require GC • Work stealing deque • MPMC queues rayon

Deletion Service ● Key Deletion from Active Object store ○ Just check previous snapshot before reclamation ● Snapshot Deletion ○ Check both previous and next snapshot for key claimation ○ No need to

@LykxSassinator • Optimize the compaction trigger mechanism of RocksDB to accelerate disk space reclamation when handling a large number of DELETE versions #17269 @AndreMouche • Support dynamically modifying small might cause memory request speeds to exceed background reclaim speeds, leading to memory reclamation and con- sequent delays in memory allocation. 726 • It is recommended to set vm.min_free_kbytes high-performance SSDs such as NVMe devices, lowering this value helps improve the efficiency of memory reclamation. • dirty_background_ratio percentage ratio. When the total amount of dirty page caches reach this

@LykxSassinator • Optimize the compaction trigger mechanism of RocksDB to accelerate disk space reclamation when handling a large number of DELETE versions #17269 @AndreMouche • Support dynamically modifying small might cause memory request speeds to exceed background reclaim speeds, leading to memory reclamation and con- sequent delays in memory allocation. 721 • It is recommended to set vm.min_free_kbytes high-performance SSDs such as NVMe devices, lowering this value helps improve the efficiency of memory reclamation. • dirty_background_ratio percentage ratio. When the total amount of dirty page caches reach this