This paper presents the most exhaustive study of synchronization to date. We span multiple layers, from hardware cache-coherence protocols up to high-level concurrent software. We do so on different types of architectures, from single-socket -uniform and nonuniform -to multi-socket -directory and broadcastbased -many-cores. We draw a set of observations that, roughly speaking, imply that scalability of synchronization is mainly a property of the hardware.

A priori, locking seems easy: To protect shared data from concurrent accesses, it is sufficient to lock before accessing the data and unlock after. Nevertheless, making locking efficient requires finetuning (a) the granularity of locks and (b) the locking strategy for each lock and possibly each workload. As a result, locking can become very complicated to design and debug.We present GLS, a middleware that makes lock-based programming simple and effective. GLS offers the classic lock-unlock interface of locks. However, in contrast to classic lock libraries, GLS does not require any effort from the programmer for allocating and initializing locks, nor for selecting the appropriate locking strategy. With GLS, all these intricacies of locking are hidden from the programmer. GLS is based on GLK, a generic lock algorithm that dynamically adapts to the contention level on the lock object. GLK is able to deliver the best performance among simple spinlocks, scalable queue-based locks, and blocking locks. Furthermore, GLS offers several debugging options for easily detecting various lockrelated issues, such as deadlocks.We evaluate GLS and GLK on two modern hardware platforms, using several software systems (i.e., HamsterDB, Kyoto Cabinet, Memcached, MySQL, SQLite) and show how GLK improves their performance by 23% on average, compared to their default locking strategies. We illustrate the simplicity of using GLS and its debugging facilities by rewriting the synchronization code for Memcached and detecting two potential correctness issues. CCS Concepts•Computing methodologies → Shared memory algorithms; Concurrent algorithms; •Computer systems organization → Multicore architectures; Keywords Locking; Adaptive Locking; Locking Middleware; Locking Runtime; Synchronization; Multi-cores; Performance * Work done while the author was at EPFL. Currently at Google. † Author names appear in alphabetical order.Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from permissions@acm.org.

We introduce "asynchronized concurrency (ASCY)," a paradigm consisting of four complementary programming patterns. ASCY calls for the design of concurrent search data structures (CSDSs) to resemble that of their sequential counterparts. We argue that ASCY leads to implementations which are portably scalable: they scale across different types of hardware platforms, including single and multisocket ones, for various classes of workloads, such as readonly and read-write, and according to different performance metrics, including throughput, latency, and energy. We substantiate our thesis through the most exhaustive evaluation of CSDSs to date, involving 6 platforms, 22 state-of-the-art CSDS algorithms, 10 re-engineered state-of-the-art CSDS algorithms following the ASCY patterns, and 2 new CSDS algorithms designed with ASCY in mind. We observe up to 30% improvements in throughput in the re-engineered algorithms, while our new algorithms out-perform the state-ofthe-art alternatives.

No abstract

No abstract

A plethora of optimized mutex lock algorithms have been designed over the past 25 years to mitigate performance bottlenecks related to critical sections and locks. Unfortunately, there is currently no broad study of the behavior of these optimized lock algorithms on realistic applications that consider different performance metrics, such as energy efficiency and tail latency. In this article, we perform a thorough and practical analysis of synchronization, with the goal of providing software developers with enough information to design fast, scalable, and energy-efficient synchronization in their systems. First, we perform a performance study of 28 state-of-the-art mutex lock algorithms, on 40 applications, on four different multicore machines. We consider not only throughput (traditionally the main performance metric) but also energy efficiency and tail latency, which are becoming increasingly important. Second, we present an in-depth analysis in which we summarize our findings for all the studied applications. In particular, we describe nine different lock-related performance bottlenecks, and we propose six guidelines helping software developers with their choice of a lock algorithm according to the different lock properties and the application characteristics. From our detailed analysis, we make several observations regarding locking algorithms and application behaviors, several of which have not been previously discovered: (i) applications stress not only the lock–unlock interface but also the full locking API (e.g., trylocks, condition variables); (ii) the memory footprint of a lock can directly affect the application performance; (iii) for many applications, the interaction between locks and scheduling is an important application performance factor; (vi) lock tail latencies may or may not affect application tail latency; (v) no single lock is systematically the best; (vi) choosing the best lock is difficult; and (vii) energy efficiency and throughput go hand in hand in the context of lock algorithms. These findings highlight that locking involves more considerations than the simple lock/unlock interface and call for further research on designing low-memory footprint adaptive locks that fully and efficiently support the full lock interface, and consider all performance metrics.

Portability and efficiency are usually antagonists in multicore computing. In order to develop efficient code, one needs to take into account the topology of the target multi-cores (e.g., for locality). This clearly hampers code portability. In this paper, we show that you can have the cake and eat it too.We introduce MCTOP, an abstraction of multi-core topologies augmented with important low-level hardware information, such as memory bandwidths and communication latencies. We show how to automatically generate MCTOP using libmctop, our library that leverages the determinism of cache-coherence protocols to infer the topology of multi-cores using only latency measurements.MCTOP enables developers to accurately and portably define high-level performance optimization policies. We illustrate several such policies through four examples: (i-ii) thread placement in OpenMP and in a MapReduce library, (iii) a topology-aware mergesort algorithm, as well as (iv) automatic backoff schemes for locks. We illustrate the portability of these optimizations on five processors from Intel, AMD, and Oracle, with low effort.

We introduce OPTIK, a new practical design pattern for designing and implementing fast and scalable concurrent data structures. OPTIK relies on the commonly-used technique of version numbers for detecting conflicting concurrent operations. We show how to implement the OPTIK pattern using the novel concept of OPTIK locks. These locks enable the use of version numbers for implementing very efficient optimistic concurrent data structures. Existing state-of-the-art lock-based data structures acquire the lock and then check for conflicts. In contrast, with OPTIK locks, we merge the lock acquisition with the detection of conflicting concurrency in a single atomic step, similarly to lock-free algorithms. We illustrate the power of our OPTIK pattern and its implementation by introducing four new algorithms and by optimizing four state-of-the-art algorithms for linked lists, skip lists, hash tables, and queues. Our results show that concurrent data structures built using OPTIK are more scalable than the state of the art.