And there's (much) more
Here is a quick run-down of a few more presentations:
- In "A day in the life of a Linux kernel hacker..." (page 185), Wi-Fi maintainer John W. Linville describes some of the background to the Linux kernel development model. He goes into the motivations which get people working on the kernel and explores the different types of collaboration, citing, in addition to programmers, users who report bugs, test the kernel or write documentation. He also discusses the development model, explaining, for example, how a patch finds its way, via various development branches, into the main development tree.
- In his presentation "I/O Topology" (page 235), Martin K. Petersen of Oracle describes the infrastructure for querying I/O topology information, which has been merged into the forthcoming kernel version 2.6.31. Among other things, this is important for allocating storage media with sector sizes other than 512 bytes, or for optimal arrangement of data in RAID arrays.
- A presentation on the use of power saving technologies under Linux and associated problems seems to have become de rigueur at the Linux Symposium. This year's entry was called "Fixing PCI Suspend and Resume" (page 319) and was given by Rafael J. Wysocki. He explains some of the problems caused by activating and deactivating PCI devices when switching into and out of suspend-to-RAM and suspend-to-disk, and how the changes merged into 2.6.29 and 2.6.30 improve the situation.
- In "Autotest – Testing the Untestable" (page 9) two Google employees describe various aspects of using the Autotest framework, which the search engine giant itself uses for its testing, for automatic testing of kernels and hardware.
- The presentation "Putting LTP to test – Validating both the Linux kernel and Test-cases" (page 209) on the Linux Test Project (LTP) deals, not with hardware testing, but with testing the functionality of the Linux kernel and with the programs used for this testing.
- Intel employee A. Leonard (Len) Brown describes the "Simple Firmware Interface" (page 55), abbreviated as SPI. This is an alternative to ACPI attuned to run under Linux, which Intel plans to use in compact devices such as Mobile Internet Devices (MIDs) containing Atom processors. In his presentation Brown makes clear that the forthcoming Moorestown platform for smartphones, Mobile Internet Devices (MIDs) and embedded applications, will support neither a standard PC BIOS nor an ACPI BIOS – in doing so he hints that Intel will be making [ticker:uk_143220 major use of Linux as the operating system] for its forthcoming Atom mobile devices.
- Dominic Duva looks at aspects of programming userland applications for real-time Linux versions in his presentation "From Fast to Predictably Fast" (page 79).
- The title of the presentation on NFS given by a group from HP hints that, for some problems, sometimes, twenty years is never enough: "Twenty Years Later: Still Improving the Correctness of an NFS Server" (page 95).
- Staff from Heinrich Heine University in Düsseldorf gave a presentation entitled "Incremental Checkpointing for Grids" (page 201).
- "Scaling software on multi-core through co-scheduling of related tasks" (page 287) demonstrates that multi-core CPUs do not always make everything faster, but can also throw up problems such as the optimal distribution of processors on CPUs and CPU cores.
All presentation topics at a glance
All the topics covered at the Proceedings of the Linux Symposium 2009:
- Autotest — Testing the Untestable by John Admanski & Steve Howard
- Increasing memory density by using KSM by Andrea Arcangeli
- Sandboxer: Light-Weight Application Isolation in Mobile Internet Devices by R. Banginwar, M. Leibowitz, & T. Tanaka
- Dynamic Debug by Jason Baron
- Measuring Function Duration with Ftrace by Tim Bird
- The Simple Firmware Interface by A. Leonard Brown
- The Corosync High Performance Shared Memory IPC Reusable C Library by Steven C Dake
- GStreamer on Texas Instruments OMAP35x Processors by D. Darling, C. Maupin, & B. Singh
- From Fast to Predictably Fast by Dominic Duval
- Combined Tracing of the Kernel and Applications with LTTng by Pierre-Marc Fournier
- Twenty Years Later: Still Improving the Correctness of an NFS Server by R. Gardner, S. D’Angelo, & M. Sears
- Memory Migration on Next-Touch by Brice Goglin & Nathalie Furmento
- Non Privileged User Package Management: Use Cases, Issues, Proposed Solutions by François-Denis Gonthier & Steven Pigeon
- GeoDNS—Geographically-aware, protocol-agnostic load balancing at the DNS level by John Hawley
- Porting to Linux the Right Way by Neil Horman
- Tracing the HA Cluster of Guests with VESPER by S. Kim, S. Moriya, & S. Oshima
- Hardware Breakpoint (or watchpoint) usage in Linux Kernel by Prasad Krishnan
- Shoot first and stop the OS noise by Christopher Lameter
- Tuning 10Gb network cards on Linux by B.H. Leitao
- A day in the life of a Linux kernel hacker... by John W. Linville
- Transcendent Memory and Linux by Dan Magenheimer
- Incremental Checkpointing for Grids by John Mehnert-Spahn
- Putting LTP to test—Validating both the Linux kernel and Test-cases by Subrata Modak
- Linux-based virtualisation for HPC clusters by L. Nussbaum, F. Anhalt, O. Mornard, & J.-P. Gelas
- I/O Topology by Martin K. Petersen
- Step two in DCCP adoption: The Libraries by L.M. Sales, H. Stuart, H.O. Almeida, & A. Perkusich
- Programmatic Kernel Dump Analysis On Linux by Alex Sidorenko
- Online Hierarchical Storage Manager by S.K. Sinha, R.B. Agrawal, V. Agarwal, R. Vashist, R.K. Sharma, & S. Hendre
- Effect of readahead and file system block reallocation for LBCAS by K. Suzaki, T. Yagi, K. Iijima, N.A. Quynh, & Y. Watanabe
- Scaling software on multi-core through co-scheduling of related tasks by Srivatsa Vaddagiri
- Converged Networking in the Data Center by Peter P. Waskiewicz Jr.
- How to (Not) Lose Your Data by Ric Wheeler
- Testing and verification of cluster file systems by Steven Whitehouse
- Fixing PCI Suspend and Resume by Rafael J. Wysocki
- Real-Time Performance Analysis in Linux-Based Robotic Systems by H. Yoon, J. Song, & J. Lee