A Scalable I/O Manager for GHC

Presented by Johan Tibell.

Abstract: The Glasgow Haskell Compiler supports extraordinarily cheap threads. These are implemented using a two-level model, with threads scheduled across a set of OS-level threads. Since the lightweight threads can’t afford to block when performing I/O operations, when a Haskell program starts, it runs an I/O manager thread whose job is to notify other threads when they can safely perform I/O.

The I/O manager manages its file descriptors using the select system call. While select performs well for a small number of file descriptors, it doesn’t scale to a large number of concurrent clients, making GHC less attractive for use in large-scale server development.

This talk will describe a new, more scalable I/O manager that’s currently under development and that hopefully will replace the current I/O manager in a future release of GHC.

Details: Date: January 29th, 2010, Friday Time: 1:30pm Location: Galois Inc., 421 SW 6th Ave. Suite 300 (3rd floor of the Commonwealth building)

Bio: Johan Tibell is a Software Engineer at Google Inc. He received a M.S. in Software Engineering from the Chalmers University of Technology, Sweden, in 2007.

Presented by Borzoo Bonakdarpour.

Design and implementation of distributed systems often involve many subtleties due to their complex structure, non-determinism, and low atomicity as well as occurrence of unanticipated physical events such as faults. Thus, constructing correct distributed systems has always been a challenge and often subject to serious errors. We propose a method for generating distributed implementations from high-level component-based models that only employ simple synchronization primitives. The method is a sequence of three transformations preserving observational equivalence: (1) A transformation from a global state to a partial state model, (2) a transformation which replaces multi-party strong synchronization primitives in atomic components by point-to-point send/receive primitives based on asynchronous message passing, and (3) a final transformation to concrete distributed implementation based on platform and architecture. We study the properties of different transformations, in particular, performance criteria such as degree of parallelism and overhead for coordination.

The second part of the talk will focus on an automated technique for optimal instrumentation of multi-threaded programs for debugging and testing of concurrent data structures. We define a notion of observability that enables debuggers to trace back and locate errors through data-flow instrumentation. Observability in a concurrent program enables a debugger to extract the value of a set of desired variables through instrumenting another (possibly smaller) set of variables. We formulate an optimization problem that aims at minimizing the size of the latter set. Our experimental results on popular concurrent data structures (e.g., linked lists and red-black trees) show significant performance improvement in optimally-instrumented programs using our method as compared to ad-hoc over-instrumented programs.

abstract:

Existing modeled and verified operating systems (OS’s) typically run on uniprocessor platforms and run with interrupts mostly disabled. This makes formal reasoning more tractable: execution is mostly sequential. The eChronos OS is a real-time OS used in tightly constrained devices, running on (uniprocessor) embedded micro-controllers. It is used in the DARPA-funded HACMS program, where it runs the flight control software of a high-assurance quadcopter. To provide low latency, the eChronos OS runs with interrupts enabled and provides a preemptive scheduler. In terms of verification, this means that concurrency reasoning is required, which significantly increases the complexity: application and OS instructions may be interleaved with interrupt handler instructions.

In our work we explicitly model the effect of interrupts and their handling by the hardware and OS. We provide a general formal model of the interleaving between OS code, application code and interrupt handlers. We then instantiate this model to formalise the scheduling behavior of the eChronos OS, and prove the main scheduler property: the running task is always the highest-priority runnable task.

bio:

June Andronick is a Senior Researcher at Data61|CSIRO (formerly NICTA). Her research focuses on increasing the reliability of critical software systems, by mathematically proving that the code behaves as expected and satisfies security and safety requirements. She contributed to the seL4 correctness proof and now focuses on concurrency reasoning for OS code. She leads the concurrency software verification research in Data61, and is deputy leader of the Trustworthy Systems group. She was recognised in 2011 by MIT’s Technology Review as one of the world’s top young innovators (TR35). She holds a PhD in Computer Science from the University of Paris-Sud, France.

abstract:

Hoare monitors, invented by Hansen and Hoare in 1973, are widely used to safely handle concurrent programming in different languages ranging from C++11 to Tower, an EDSL developed by Galois as part of the High-Assurance Cyber Military Systems (HACMS) DARPA program. This talk will explain how basic safety properties are assured using Tower, and how it is possible to improve runtime efficiency and parallelism of Tower-generated C programs by releasing some constraints on the Hoare monitor model. Finally, some test results on SMACCMPilot, a high-assurance autopilot, will be presented.

bio:

Georges-Axel Jaloyan is a CS student at École Normale Supérieure in Paris. He is interning at Galois as part of his Master’s degree. He is interested in embedded systems security and safety-critical systems. He interned previously at NASA Langley Safety Critical Avionics Systems Branch.

Doors open at 6:30. Talks start at 7:00.

Pizza, soda and beer provided by Weedmaps.

October's meeting will host a set of mini talks, up to 20 minutes each.

Talks:

• Paul Rogers:

EntropyString

• Zach:

Elixir Bot Server Frog and Toad

• Moxley Stratton:

Test mocks

Talks start at 7:00pm. Doors open at 6:30pm.

October will host a set of mini talks, up to 20 minutes each. We are requesting proposals for this now. Please send a direct message to Moxley if you're interested.

Pizza, soda and beers sponsored by Weedmaps.

Talks:

• Connor Clay: Testing strategies in Elixir

• Daniel Hedlund: DevDocs

• Andrew Vy: Elixir + Headless Chrome

• Moxley Stratton: Test mocks with "Mox"