Jan Midtgaard (Tarides)
Olivier Nicole (Tarides)
Nicolas Osborne (Tarides)

Parallel and concurrent code is notoriously hard to test because of the involved non-determinism, yet it is facing OCaml programmers with the coming OCaml 5.0 multicore release. We present two related testing libraries to improve upon the situation: - Lin – a library to test for linearizability - STM – a state-machine testing library
Both libraries build on QCheck, a black-box, property-based testing library in the style of QuickCheck. The two libraries represent different trade-offs between required user effort and provided guarantees and thereby supplement each other.

Deepali Ande (IIT Madras)
KC Sivaramakrishnan (IIT Madras and Tarides)

OCaml 5 introduces effect handlers as a mechanism for con- current programming. With effect handlers, concurrency can be expressed in direct-style rather than in monadic-style as in Lwt and Async. Rather than baking in the notion of a thread scheduler, the compiler exposes delimited continuations, with the idea that different libraries may implement their own concurrency primitive, with their own schedulers. Under this setting, given that the notion of a concurrent task is tied to a particular scheduler, it is challenging to allow tasks from different schedulers to interact with each other. If this problem is not solved, the ecosystem runs the risk of repeating the schism between Lwt and Async in OCaml today.
In this paper, we observe that the composability of effect handlers permits composability of schedulers. The key idea is that we can use effect handlers to define a uniform interface for suspending and resuming tasks, which is implemented by each of the schedulers. On top of this, we define scheduler-agnostic synchronization structures that allow tasks from different schedulers to interact. We also report on how this mechanism can be extended to capture the notion of task cancellation that extends across different schedulers

Artur Miguel Dias (NOVA University of Lisbon and NOVA LINCS)
Simão Melo de Sousa (Universidade da Beira Interior and NOVA LINCS)
Antonio Ravara (NOVA LINCS & FCT, NOVA University of Lisbon)

We present the motivation and principles that lead to the development of support for exercises on Formal Languages and Automata Theory in Learn-OCaml, its integration with OFLAT, our web based teaching environment (mostly implemented in OCaml), its uses in the classroom in two Portuguese Universities, and in supporting students independent work.

David Allsopp (Tarides UK)

OCaml’s ecosystem centres around compilation from source code, rather than precompiled binaries. The compiler itself is only provided as a source distribution and OCaml’s official package repository in opam also manages source code packages only.
Even on a high spec. machine, compiling OCaml from source code takes a minute, for many users, it’s slower. The recommended workflows1 for OCaml development focus on having a compiler installation for each individual project.
Various solutions are available downstream for trying to reduce the set-up required to get a working OCaml compiler for a project ranging from predistributed binaries to rewriting binary artefacts.
This talk presents an unexpected journey which began as a series of fixes to eliminate some esoteric failures of OCaml programs to launch, yet ended with a series of relatively simple changes to OCaml which allow various downstream workarounds to be brought home to upstream OCaml.

Oleg Kiselyov (Tohoku University, Japan)
Tomoaki Kobayashi (Tohoku University)
Aggelos Biboudis (Oracle)
Nick Palladinos (Nessos Information Technologies, SA)

We present the stream processing library that achieves the highest performance of existing OCaml streaming libraries, attaining the speed and memory efficiency of hand-written state machines. It supports finite and infinite streams with the familiar declarative interface, of any combination of map, filter, take(while), drop(while), zip, flatmap combinators and tupling. Experienced users may use the lower-level interface of stateful streams and implement accumulating maps, compression and windowing. The library is based on assured code generation (at present, of OCaml and C) and guarantees in all cases complete fusion.

Haochen M. Kotoi-Xie (Kotoi-Xie Consultancy, Inc.)
Hirotetsu Hongo (Kotoi-Xie Consultancy, Inc.)
Yuta Sato (Kotoi-Xie Consultancy, Inc.)
Shinya Yamaguchi (Kotoi-Xie Consultancy, Inc.)

We propose to give a talk at the OCaml Users and Developers Workshop 2022, about our to-be-open-sourced project Bindoj being developed at KXC (Kotoi-Xie Consultancy, Inc.) Bindoj is a generative programming tooling library written in OCaml that centering around descriptions of datatypes which are mainly used in business / application logics.

David Allsopp (Tarides UK)
Raja Boujbel (OCamlPro)
Kate Deplaix (Tarides UK)
Louis Gesbert (OCamlPro)

OPAM 1.2 was released in 2014. It was four years before opam 2.0 succeeded it, and another three for opam 2.1. The release of opam 2.2 is imminent, and will have arrived just a year after its predecessor. This talk presents features added to opam 2.1 intended to make dealing with multiple versions of opam easier. We see how small additional metadata, both stored in .opam “roots” and also added to opam’s command line provide opam’s developers with the chance to add new features more easily, with confidence that users will be able to upgrade safely.
In presenting this, we’ll also trumpet some of what we think are the amazing new features of opam 2.1 (and 2.2), and why we think you should keenly watch for new releases and upgrade immediately on release!

Darius Foo (National University of Singapore)
Wei-Ngan Chin (National University of Singapore)

This presentation will cover a framework for application-level tracing of OCaml programs. We outline a solution to the main technical challenge, which is being able to log typed values with lower overhead and maintenance burden than existing approaches. We then demonstrate the tools we have built around this for visualizing and exploring executions.

Tim McGilchrist (Tarides)
David Allsopp (Tarides UK)
Patrick Ferris (Tarides)
Antonin Décimo (Tarides)
Thomas Leonard (Tarides UK)
Anil Madhavapeddy (University of Cambridge, UK)
Kate Deplaix (Tarides UK)

OBuilder - Homogeneous builds with OBuilder and OCaml
This talk will present a lightweight sandboxing solution (OBuilder) that works beyond the usual Linux containerisation solutions, providing support for macOS, Windows and the BSDs without requiring full (expensive) virtualisation. We will cover the implementation for macOS and Windows, the challenges encountered with providing sandboxes on such different platforms, and how this work is being used to provide cross-platform builds to the OCaml community.
We previously introduced OCaml-CI (OCaml Workshop 2020 “OCaml-CI : A Zero-Configuration CI”) providing an opionated, fast-feedback CI system for OCaml projects. Since the end of 2020, opam-repo-ci has provided a similar service for testing pull requests to opam-repository. Both of these systems use OBuilder to provide support for multiple operating systems and hardware architectures.

Andrey Mokhov (Jane Street)
Arseniy Alekseyev (Jane Street)

We present Memo, an incremental computation library that supports a new, faster and more scalable, file-watching build mode in Dune 3.0. The requirements from the build systems domain make Memo a unique point in the design space of incremental computation libraries. Specifically, Memo needs to cope with concurrency, dynamic dependencies, dependency cycles, and non-determinism, provide support for efficiently collecting and reporting user-friendly errors, and scale to computation graphs containing tens of millions of incremental nodes.

Guillaume Munch-Maccagnoni (INRIA)

This paper reports an experiment with a large pages allocator for the OCaml runtime, with measured performance improvements. A large pages allocator (also known in the literature under other names: superpages, etc.) is a standard component of a memory allocator that stands between the OS and user-facing allocators (e.g. minor and major heaps) and which reserves and manages large chunks of contiguous memory. The OCaml runtime currently gives up a good amount of control by assigning this role to the system allocator. Other languages have a simple implementation from which practical lessons can be learnt (especially in terms of portability), such as the one from the Go runtime.
Control over large pages affect (traditionally) the components implementing virtual address translation (hardware-level page table, TLB), in particular by leveraging (hardware-level) huge pages. It also enables (more specifically) efficient implementations of the OCaml page table, a data structure used in various parts of the OCaml runtime to classify pointers, to distinguish for instance which blocks belong to the heaps and which ones are “out-of-heap” during the marking phase of the GC.
One goal was to evaluate the possible performance of a page table for multicore OCaml. While I did not use original techniques, some of the results are unexpected a priori based on expressed beliefs in the OCaml community.
In essence, this paper reports the good hypothetical performance, in rigorous practical terms, of embedding (borrowing) linearly-allocated values inside garbage-collected values. A companion submission to the ML workshop reports a symmetrical result: how to efficiently embed (own) garbage-collected values inside linearly-allocated values. Taken together, the broader motivation is to show the feasibility of basing linear allocation with re-use (Lafont 1988, Baker 1992) in languages that would still leverage state-of-art garbage collection for non-linear values.

Stephen Dolan (Jane Street)
Leo White (Jane Street)

Allocating values on a stack instead of the garbage collected heap can improve performance by improving cache locality and avoiding GC pauses. However, it requires that the values do not escape the lifetime of their associated stack frame. We describe an extension to OCaml that allows values to be allocated on a stack and ensures through the type system that they do not escape their stack frame.

The demo repository is online.