Loading…
October 31 - November 1 - Co-Located Events
October 28-30 - Conference
Lyon Convention Centre - Lyon, France
More information for Open Source Summit + Embedded Linux Conference Europe 2019
Linux Systems [clear filter]
Monday, October 28
 

11:30 CET

Address Space Separation Inside the Linux Kernel - Mike Rapoport, IBM
Address space isolation has been used to protect the kernel and userspace programs from each other since the invention of the virtual memory. Assuming that kernel bugs and therefore exploits are inevitable
it might be worth isolating parts of the kernel to minimize the damage that these exploits can cause.

Mike is going to present a mechanism for "system call isolation" that allows running a system call with largely reduced page tables and provides the kernel with the ability to inspect the memory accesses and verify their safety based on a pre-defined policy.

Another topic is assigning an address spaces to the Linux namespaces. For instance, by keeping all the objects in a network namespace private, we can achieve levels of isolation equivalent to running a separated network stack.

This idea has already been posted to the linux kernel email list as aset of RFC patches so we'll discuss both the current state of the patchset as well as potential future enhancements.

Speakers
MR

Mike Rapoport

Researcher, IBM
Mike has lots of programming experience in different areas ranging from medical equipment to visual simulation, but most of all he likes hacking on Linux kernel and low level stuff. Throughout his career Mike promoted use of free and open source software and made quite a few contributions... Read More →



Monday October 28, 2019 11:30 - 12:05 CET
Lumiere Auditortium
  Linux Systems
  • Session Slides Included Yes

11:30 CET

Guider: Analyzing All Performance Factors - Peace Lee, Hyundai Motor Company
Peace Lee will introduce the Guider project first. Guider is an integrated runtime performance analyzer.
He will explain various features of Guider to analyze app/system performance. For example, it is possible to monitor and trace performance factors such as sched, lock, malloc, IPC, IRQ, log in real-time from system level to function level easily.
Web-based GUI feature is also supported now.
Finally, with Guider, he will also share know-hows of verifying and improving the performance of software that is being changed and expanded in the development process.

Speakers
avatar for Peace Lee

Peace Lee

Performance Architect, Hyundai Motor Company
Peace Lee is a Linux Performance Specialist. He has been analyzing and improving the performance of apps and system on various platforms based on Linux. He is the owner of Guider (https://github.com/iipeace/guider).



Monday October 28, 2019 11:30 - 12:05 CET
Salon Pasteur
  Linux Systems
  • Session Slides Included Yes

12:20 CET

Bug Introducing Patches - Sasha Levin, Microsoft
The Linux Kernel has spent significant effort on trying to identify commits which fix bugs so that they could be backported to stable branches. This effort includes everything from manual review of patches, to machine learning which attempts to identify the characteristics of a commit which fixes a bug.

What if we would identify commits which introduce bugs instead?

This talk will describe various methods to identify commits which are more likely to introduce a bug, and show a few uses to such algorithm. As a side effect, it will also help folks understand what type of behavior or actions are more likely to introduce bugs and hopefully will show how these can be avoided.

Speakers
SL

Sasha Levin

Software Engineer, Google
Sasha helps maintain the Linux Kernel Stable and LTS trees. He is currently employed by Google where he helps make Linux better. Previously, Sasha was employed by Microsoft and the Ksplice team in Oracle.


Monday October 28, 2019 12:20 - 12:55 CET
Salon Pasteur

12:20 CET

The Unified Tracing Platform - Steven Rostedt, VMware Inc.
One of the complaints about Linux is that there are more than one way to skin the tracing cat. There's ftrace, perf, BPF, bpftrace, LTTng, strace, gdb, and all sorts of ways to accomplish tracing within the Linux system. Some tracers focus on tracing the kernel while others focus on user space and each with their own file format.

Mathieu Desnoyers created babeltrace and the Common Trace Format (CTF) to solve the issue of different tracing formats. But this does not solve the issue of having multiple tools. If one wants to create a tool that requires the work of other tools, it will have a hard time executing them.

This brings up the Unified Tracing Platform (UTP). This effort is to take the functionality of all the tools and create libraries for them. This will allow any tool to utilize the features of the tools incorporated in the UTP. This talk will describe the efforts being made to accomplish the creation of UTP and how it will benefit the Linux ecosystem as a whole.

Speakers
avatar for Steven Rostedt

Steven Rostedt

Software engineer, Google
Steven Rostedt currently works for Google on the ChromeOS baseOS performance team. He is the main developer and maintainer for ftrace, the official tracer of the Linux kernel, as well as the user space tools and libraries that interact with the Linux tracing interface. Steven is also... Read More →



Monday October 28, 2019 12:20 - 12:55 CET
Lumiere Auditortium
  Linux Systems
  • Session Slides Included Yes

12:40 CET

Lightning Talk: Embedded Device Chameleons on Stage - Christian Alexander Sack, Pantacor
This is about how embedded device innovation can be done on software side alone by showing how a device originally coming out of factory using a Debian based OS can be morphed into a device running Android without a factory reinstall.

Speakers
AS

Alexander Sack

CTO & Co-Founder, Pantacor
Alexander is a long term linux and open source leader who tries to make making linux embedded products easier for everyone. During his career he lead various workshops and BoF sessions on a broad set of topics at prominent Linux events such as Ubuntu Developer Summit and Linaro Connect... Read More →



Monday October 28, 2019 12:40 - 12:50 CET
St. Clair 3
  Linux Systems
  • Session Slides Included Yes

14:25 CET

2019 State of U-Boot Development Report - Jagan Teki, Amarula Solutions
The U-Boot bootloader has been evolved for nearly 2 decades and is one of the primary and well-known opensource bootloader choice for embedded industry.

The 2019 State of U-Boot development report describe the key updates, features, issues and challenges faced so far on U-Boot community project.

This talk Jagan Teki start with a brief overview of U-Boot community, TPL, SPL, U-Boot Proper, Build process and Startup sequence and then he traverse how different features has been adopted in U-Boot start from the project beginning to most recent versions till 2019. From this traversing he will address the key features like Image boot, FIT, EFI, Secure Boot, DTS, Driver Model, Device Firmware Upgrade, ATF, OP-TEE and etc.

Once giving enough report, he will also talk about steps to port U-Boot to new hardware. Finally, he will address and review ongoing development work, issues and future development on U-Boot community.

Speakers
avatar for Jagan Teki

Jagan Teki

Co-Founder/Embedded Linux Architect, Amarula Solutions
Jagan is an Embedded Linux Architect and Co-Founder of Amarula Solutions India. His work involves providing Mainline Linux and related ecosystem projects to run on customer hardware devices/boards. He is an active contributor for U-Boot, Linux, Buildroot, Yocto, and maintainer of... Read More →



Monday October 28, 2019 14:25 - 15:00 CET
Lumiere Auditortium
  Linux Systems
  • Session Slides Included Yes

14:25 CET

VM Kernel Tracing with Trace-cmd - Slavomir Kaslev & Tsvetomir Stoyanov, VMware
Many modern workloads run inside VMs making it harder to reason about the overall system performance when faced with a problem caused by VMs competing for shared host resources or issues caused by the interaction between the host and guest kernels.

In this presentation Slavomir and Tzvetomir will discuss a set of recent trace-cmd features allowing for capturing of kernel traces from several running VMs or simultaneously recording of host kernel tracing data alongside the VM kernel traces. The resulting trace streams are automatically synchronized during recording and can later be loaded in KernelShark for visual analysis of the data.

Speakers
avatar for Slavomir Kaslev

Slavomir Kaslev

Open Source Software Engineer, VMware
Slavomir is an open source engineer in the System Engineering team at VMware. Before joining VMware he was Head of R&D at WorldWideFX and part of Google Chrome's GPU team, among other roles. His interests include computer graphics, type theory, math, rock climbing and mountain bi... Read More →
avatar for Tzvetomir Stoyanov

Tzvetomir Stoyanov

Open Source Engineer, VMware
Tzvetomir Stoyanov is an Open Source Engineer at VMware, contributing to a variety of open source projects, including Linux kernel and user space tracing, Edge and IoT, and Machine Learning. Before joining VMware, he worked with Linux and various Unix-like operating systems for twenty... Read More →



Monday October 28, 2019 14:25 - 15:00 CET
Salon Pasteur
  Linux Systems
  • Session Slides Included Yes

15:15 CET

Panel: Outreachy Linux Kernel Internship Report - Vaishali Thakkar, Independent
Come learn about the amazing work our kernel interns have accomplished! Outreachy provides 3 months paid internships for people from groups traditionally underrepresented in tech to work on open source projects. The panel will present the program and this year's Linux kernel projects. Mamta Shukla will talk about adding Alpha blending and Overlay Plane support in VKMS driver. Branden Bonaby will present on Injecting failures into VMBus messages to improve the fuzz testing in the Hyper-V drivers. Shayenne da Luz Moura will talk about her work on VKMS driver and adding IGT tests for the same. Dafna Hirschfeld will talk about her work in vicodec driver of the media subsystem. Nishka Dasgupta will present her work on finding and fixing bugs in the Linux kernel using Coccienlle. Himadri Pandya will talk about her work on improving Linux Kernel for Hyper-V. And finally Vaishali Thakkar will share her experience as a coordinator.

Speakers
HP

Himadri Pandya

Linux Kernel Bug Fixing Mentee, CommunityBridge by The Linux Foundation
Himadri Pandya is pursuing a Masters in Technology in the domain of Information and Communication Technology with a specialization in Machine Intelligence from DAIICT, India. She is currently exploring the field of computer vision. She is fascinated with Operating Systems and has... Read More →
avatar for Nishka Dasgupta

Nishka Dasgupta

Linux Kernel Intern, Outreachy
Nishka Dasgupta completed a bachelor's in computer science from Ashoka University. She followed it up with a Diploma in Advanced Studies andResearch, in the field of access structures. Following her diploma, she worked on the Linux kernel via Outreachy in the summer of 2019. She is... Read More →
avatar for Vaishali Thakkar

Vaishali Thakkar

Freelance linux linux developer, Independent
Vaishali Thakkar works as a Freelance Linux Kernel developer/consultant with clients around the world and has been volunteering as a Linux Kernel coordinator for Outreachy. She has previously worked with Oracle as a Linux Kernel Engineer and Travis Foundation as a Co-organizer for... Read More →
SM

Shayenne Moura

Linux Kernel Intern, Outreachy
Shayenne Moura is a Computer Scientist graduated student at University of Sao Paulo (Brazil). She is interested in many areas, including Machine Learning, Computer Music and Open Source Systems. She was an Outreachy intern on dri-devel (kernel GPU subsystem). Currently, Shayenne is... Read More →



Monday October 28, 2019 15:15 - 15:50 CET
Lumiere Auditortium
  Linux Systems
  • Session Slides Included Yes

15:15 CET

Self Modifying Code in Linux Kernel - What, Where and How - Evgeniy Paltsev, Synopsys
Text segment is READONLY... Wait, what? Who said that? Modern Linux kernel uses self modifying code a lot. And that's not only about modules loading.
During this talk, we'll look at how different kinds of self modifying code are used in Linux kernel to extend functionality, assist in debugging, workaround hardware bugs, mitigate vulnerabilities and simply making kernel faster.

However rewriting code has its price which turns into increasing system complexity. I'll talk about it based on my experience with implementation of static branches for ARC architecture.

PS: don't rewrite yourself.


>>NOTE<<: 'Salon Pasteur' room isn't showed on the map - it is near the main registration area. Just go to the registration and ask them :)

Speakers
avatar for Evgeniy Paltsev

Evgeniy Paltsev

Linux & Zephyr kernel engineer, Synopsys
Linux & Zephyr kernel engineer, Synopsys. Evgenii is a member of Synopsys open-source team which ports and maintains projects for ARC and RISV-V processors architecture. He actively develops Linux & Zephyr kernels with journeys to other projects like U-Boot, Buildroot, uClibc-ng... Read More →



Monday October 28, 2019 15:15 - 15:50 CET
Salon Pasteur
  Linux Systems
  • Session Slides Included Yes

16:20 CET

Adopting Linux on BMW - The Long Road to Integrate Linux as Mainline Platform - Helio Chissini de Castro, BMW CarIT
Automotive industry is in the first stages of adopt fully open source systems on their stack.
Not only the software itself, but constraints on safety, security, compliance, new cultural processes complete different from traditional ones. Is an industry that has a past and simply has no possibility to start from scratch.
At BMW, the adaptation for the new reality was planned in long step plans to achieve the most seamless transition to the new processes.
From the ground build to the main software and OS to the buildsystem and test mechanism, every single detail need to be care in a different way, and integration was the core aspect of all.

Unique aspects as multiple computers, different architectures, strict requirements, certification, a real Linux Brave New World

As OS Team, we want to share how the system was integrated on the middle of the process and the steps to be done to achieve the full platform that will drive our customers for a bright car future.

Speakers
avatar for Helio Chissini de Castro

Helio Chissini de Castro

Senior Software Engineer - Linux OS Domain, BMW CarIT
Helio Chissini de Castro is working as Senior Software Engineer and Linux OS Domain Lead at BMW CarIT software. At BMW CarIT is working on the next base platform project for the assistance vehicles, based on Linux. He is long time KDE contributor as the project was the beginning of... Read More →



Monday October 28, 2019 16:20 - 16:55 CET
Salon Pasteur
  Linux Systems
  • Session Slides Included Yes

16:20 CET

Compact C Type Format Support in the GNU Toolchain - Elena Zannoni, Oracle Corporation
Compact C Type Format (CTF) is a reduced form of debugging information whose main purpose is to describe the type of C entities such as structures, unions, typedefs and function arguments. It originated in
the Solaris kernel and it has been ported to Linux as part of the DTrace for Linux project. It's been used (via libdtrace-ctf) to reduce the size of the debugging information for the Linux kernel and for use in DTrace.

There are many advantages to using CTF, due to its compactness, for many kinds of programs that can't rely on DWARF. DWARF's design strives for generality and expressive power, at the cost of being a rather heavy
format. For example, evaluating DWARF expressions requires an interpreter for a stack-based machine. This, which is not problematic for typical "off-line" debugging programs such as symbolic debuggers (GDB), may be inconvenient for "on-line" debugging programs such as unwinders and stack tracers, due to efficiency and security concerns. CTF is a promising format that helps maintain some level of debuggability, even when the size of the executable is an issue and the DWARF info is being stripped out.

For such reasons we integrated CTF with the GNU toolchain on Linux. This talk will explain how CTF is structured, how we added CTF handling to gcc, gdb binutils and elfutils, and will provide an update on the
upstreaming status.

We believe CTF provides the right foundations for expressing the information needed by "on-line" debugging programs, in a most convenient way.


Speakers
avatar for Elena Zannoni

Elena Zannoni

Director of the Linux Tools and Languages Team, Oracle Corporation
Elena Zannoni is the manager for the Linux Toolchain and Tracing team at Oracle. The team covers the GNU toolchain and DTrace for Linux, among other things. Elena was one of the original GDB global maintainers and has spoken worldwide on topics related to tracing at many conferences... Read More →



Monday October 28, 2019 16:20 - 16:55 CET
Tête d'Or 2
  Linux Systems
  • Session Slides Included Yes

16:20 CET

New Container Kernel Features - Christian Brauner, Canonical Ltd.
Containers have become ubiquitous. Nowadays, the computing landscape is mostly concerned with a higher-level view on the functionality and features of containers. In this talk we will take a step back and focus on the low-level aspects that make containers possible. Specifically, we will look at new kernel features that have been implemented. This includes work such as shiftfs, a filesystem to translate between on-disk ids into a user namespace idmapping, a new seccomp interface which allows to defer decisions about whether or not a syscall is considered successful or not to userspace and other features.

Speakers
avatar for Christian Brauner

Christian Brauner

Principal Software Engineer, Microsoft
Christian Brauner is a kernel developer and maintainer of the LXD and LXC projects currently working at Microsoft. He works mostly upstream on the Linux Kernel maintaining various bits and pieces. He is strongly committed to working in the open, and an avid proponent of Free Software... Read More →



Monday October 28, 2019 16:20 - 16:55 CET
Lumiere Auditortium
  Linux Systems
  • Session Slides Included Yes

17:10 CET

LED Subystem in Kernel - Pavel Machek, Denx
LEDs are very simple, cheap and common devices, and we have easy-to-use subsystem to handle them. Few lines in device tree are enough to drive simple LED. Unfortunately, there's little standartization going on, and there are great differences between devices, so supporting LEDs is not easy: some LEDs can be accessed from interrupts (and we want to use that capability), some take 300msec to update. Most are on/off, but PWM is getting common, and so are three LEDs in one package, allowing display of RGB color. LED subsystem currently treats RGB LED as three separate LEDs, which is not optimal. Then there are triggers, such as "CPU activity", and we have sysfs interface for them. But newer hardware tends to have "this LED is either software controlled or shows ethernet link", and we want good interface for that. Other open issue is LED device naming; currently names are not very consistent.

Speakers
PM

Pavel Machek

Kernel hacker, Denx
Pavel is a long-term kernel hacker. He worked on amd64 kernel port and hibernation for SuSE. Currently he's co-maintaining hibernation and LED subsystem, and works with Denx on various embedded projects.


leds pdf

Monday October 28, 2019 17:10 - 17:45 CET
Lumiere Auditortium
  Linux Systems
  • Session Slides Included Yes

17:10 CET

Once upon an API - Michael Kerrisk, man7.org Training and Consulting
Back in 1997, a new system call, prctl(), was added to the Linux kernel. By now that API does many (arguably, too many) different things to a process. But to begin with, it provided just one feature: to allow a child process to request that the kernel send it a signal when its parent dies. At first, the PR_SET_PDEATHSIG feature seems simple.

However, when one examines its interactions with various other UNIX and Linux API features, such as threads, signals, and exec, the semantics of this feature turn out to be at times both complex and surprising. Some of those semantics were certainly unintended, and are bizarre enough that, as the Linux man-pages
maintainer, I fear documenting them. By looking in detail at this specific example, I’ll explore various pitfalls and lessons we can learn when designing APIs, philosophize a little on the question of who "owns" an API when it comes to defining the semantics of that API, and consider some strategies for improving the API design process.

Speakers
avatar for Michael Kerrisk

Michael Kerrisk

Trainer, man7.org Training and Consulting
Michael Kerrisk is a trainer, author, and programmer who has a passion for investigating and explaining software systems. He is the author of "The Linux Programming Interface", a widely acclaimed book on Linux (and UNIX) system programming. Michael is a New Zealander, living in Munich... Read More →



Monday October 28, 2019 17:10 - 17:45 CET
Salon Pasteur
  Linux Systems
  • Session Slides Included Yes
 
Tuesday, October 29
 

11:30 CET

Debugging Embedded Linux Systems with GDB - Jan Altenberg, Continental Automotive GmbH
For Linux, it doesn’t matter if a program is being run on a server, on a desktop system or on an embedded device. The tools and interfaces are always the same. This portability is one of the key factors for the success of Linux in the industrial sector. But even though the software is portable and the developers are always using the same interfaces, there are special requirements when it comes to embedded software. Most importantly: A program is usually not being developed on the system where it runs on. Therefore we need to generate code for different architectures and (since software always comes with bugs) we need to be able to debug software which is running on a different system. GDB can deal with that situation. It can help you to remote debug a program or to analyze a core file which has been generated on a different system, even if it’s a completely different CPU architecture. This presentation gives an introduction on how to use GDB for debugging an Embedded Linux System.

Speakers
avatar for Jan Altenberg

Jan Altenberg

System Architect / OSS Compliance Officer, Continental Automotive GmbH
Jan Altenberg has 15 years of experience in developing and maintaining Embedded Linux systems. He studied information technologies at the University of Cooperative Education in Stuttgart (Germany). From 2002 - 2006 he was involved in the OCEAN project, a european research project... Read More →



Tuesday October 29, 2019 11:30 - 12:05 CET
Salon Pasteur
  Linux Systems
  • Session Slides Included Yes

11:30 CET

NVMe/TCP for your Data Center - Orit Wasserman, Lightbits Labs
NVMe/TCP is the latest transport added to the of the NVMe over Fabrics protocol and is part of Linux kernel 5.0. One of its great advantages is the use of TCP/IP, the most known network protocol that is implemented in every data center.
It provides performance latencies that are comparable with RDMA transport without requiring any network infrastructure changes.
In this talk, you will get a good understanding of the NVMe/TCP protocol and when to use it. We will also discuss disaggregated storage, and how it can reduce costs and increase usage efficiency, allowing you to get the most of your NVMe storage.

Speakers
avatar for Orit Wasserman

Orit Wasserman

Senior Principal Software Engineer, Red Hat
Orit is a senior principal software engineer at Red Hat, focusing on Container and multi cloud storage. She was a principal architect at Lightbits labs working on NVMe/TCP software-defined storage. At Red Hat, she worked on Ceph object storage (Ceph Rados Gateway), a highly available... Read More →



Tuesday October 29, 2019 11:30 - 12:05 CET
Lumiere Auditortium
  Linux Systems
  • Session Slides Included Yes

12:20 CET

Data Protection on NV-DIMM - Dr. Hannes Reinecke, SUSE Linux GmbH
NV-DIMM has been presented as delivering high-speed, memory-like performance while allowing for large capacity and lower prices than normal DRAM.
One of the most common usecases for NV-DIMM is the so-called DAX (direct access) mode, which allows NV-DIMMs to be used for storing filesystem data but at the same time providing all the benefits of NV-DIMMs.
In order to use these filesystems in mission-critical scenarios some kind of data protection is required. On normal block devices such protection would be guaranteed by RAID systems, and the linux stack does have a full-blown RAID implementation.
However, using this implementation on NV-DIMMs make it impossible to use the DAX capabilities.
This presentation will be focussing on the challenges when implementing data protection for NV-DIMMs, and will be presenting some possible mechanisms for providing data protection while retaining DAX capabilities.

Speakers
HR

Hannes Reinecke

Storage Architect, SUSE Linux
Studied Physics with main focus image processing in Heidelberg from 1990 until 1997, followed by a PhD in Edinburgh 's Heriot-Watt University in 2000. Worked as sysadmin during the studies, mainly in the Mathematical Institute in Heidelberg. Now working at SUSE Labs as Kernel Storage... Read More →


Tuesday October 29, 2019 12:20 - 12:55 CET
Salon Pasteur

12:20 CET

Energy-efficiency and Linux - Rafael Wysocki, Intel
Energy-efficiency is about optimizing the usage of energy with the goal to reduce it to the acceptable minimum.

Some Linux kernel features have been introduced with energy-efficiency in mind. It is the main goal of CPU idle time management, CPU performance scaling, PM-runtime and system-wide PM. Unfortunately, these features are often avoided out of concerns that they may prevent systems from achieving optimum performance or they are underutilized.

Needless to say, it does matter whether or not the kernel's energy-efficiency features are in use, however. The battery life of portable systems is affected by that directly, data center power budgets are not unlimited and, generally speaking, avoidable costs should not be paid, especially if they are environmental as well as financial. In this particular case it should be possible to avoid paying them, by enabling the features in question and controlling them with the help of interfaces provided for that, which I am going to explain.

Speakers
avatar for Rafael Wysocki

Rafael Wysocki

Software Engineer, Intel
Rafael maintains the Linux kernel’s power management infrastructure and the core ACPI support code. He works at Intel and focuses on the mainline Linux kernel development. Rafael has been actively contributing to Linux since 2005, in particular to the kernel’s suspend/hibernate... Read More →



Tuesday October 29, 2019 12:20 - 12:55 CET
Lumiere Auditortium
  Linux Systems
  • Session Slides Included Yes

14:25 CET

RISC-V Boot Process: One Step at a Time - Atish Kumar Patra, Western Digital
A well-supported and standard boot flow is very important for the RISC-V software ecosystem before RISC-V can be a truly competitive alternative to existing mainstream ISAs. However, RISC-V also needs its own trusted firmware to handle RISC-V specific features such as Supervisor Binary Interface (SBI) that allows the operating systems to interact with the supervisor execution environment (SEE). In this talk, Atish will discuss the status of a separate but modular open source SBI implementation (aka OpenSBI) that provides RISC-V specific run time services and how it helps in porting other common boot loaders such as U-Boot, coreboot and EDK2 to RISC-V. He will also discuss how the RISC-V boot process compares to other ISAs and where the community is heading.

Speakers
avatar for ATISH KUMAR PATRA

ATISH KUMAR PATRA

Software Engineer, Rivos
Atish is a Linux kernel engineer working at Rivos. He has contributed to virtualization, early boot code and drivers in Linux kernel and open source firmware for RISC-V.



Tuesday October 29, 2019 14:25 - 15:00 CET
Salon Pasteur
  Linux Systems
  • Session Slides Included Yes

14:25 CET

Systems Tracing and Trace Visualization Lab - Geneviève Bastien, École Polytechnique de Montréal
The Trace Visualization lab introduces participants to system and application tracing and trace visualizations that are an invaluable techniques to understand in-depth system behavior and reach root-cause of problems. The focus of this lab is on post-mortem analysis, for example, when the system or application misbehaved and we want to understand the root cause. It first introduces attendees to system tracing, trace collection and eventually visualization techniques such as flamecharts, flamegraphs, timeline views, critical path view.

This tutorial session will explain why and when tracing is required and its role in supporting related performance analysis techniques, such as distributed tracing, profiling, debugging and service log analysis. Participants are invited to carry out hands on activities that showcase how different "views" can help in scenarios such as resource contention, latency analysis. Use cases presented are both for single machine scenario and distributed systems.

Speakers
avatar for Geneviève Bastien

Geneviève Bastien

Research Associate, École Polytechnique de Montréal
Geneviève Bastien is a research associate at the Dorsal Laboratory of École Polytechnique de Montréal. She is a contributor to the Trace Compass and LTTng projects. Her mission is to make the students' life easier when in comes to prototyping cool new analyses and to make sure... Read More →



Tuesday October 29, 2019 14:25 - 15:50 CET
Lumiere Auditortium
  Linux Systems
  • Session Slides Included Yes

15:15 CET

The Linux Capabilities Model - Michael Kerrisk, man7.org Training and Consulting
Capabilities are an attempt to mitigate the problems that result from the crude granularity of the traditional UNIX/Linux privilege model, by breaking the power of superuser into pieces which can be
individually assigned to executables.

Capabilities have been present on Linux for many years, but they remain poorly understood. And though capabilities are used by many well known pieces of software, it is probably fair to say that they are less used than the original developers may have hoped. Nevertheless, they can be used to make privileged executables that are safer than traditional set-UID-root programs.

In this talk, I'll describe the Linux capabilities model, looking at how capabilities are attached to executable files, and the rules that determine how a process's capabilities transform when it executes a
file. I'll also consider some of the problems of capabilities that have hindered their adoption as well some remaining problems in their implementation.

Speakers
avatar for Michael Kerrisk

Michael Kerrisk

Trainer, man7.org Training and Consulting
Michael Kerrisk is a trainer, author, and programmer who has a passion for investigating and explaining software systems. He is the author of "The Linux Programming Interface", a widely acclaimed book on Linux (and UNIX) system programming. Michael is a New Zealander, living in Munich... Read More →



Tuesday October 29, 2019 15:15 - 15:50 CET
Salon Pasteur
  Linux Systems
  • Session Slides Included Yes

16:20 CET

Advanced Testing using UserModeLinux - Richard Weinberger, sigma star gmbh
UserModeLinux (UML) is the dinosaur of Linux virtualization, it was one of first methods to run multiple Linux systems on the same host.

Over the last years KVM mostly superseded UML, but UML can be still useful for building complex tests for both kernel and userspace application since the kernel runs as regular process on the host side.
In this talk Richard will present in detail on how UML can be (ab)used to build tests which are not possible with hypervisor-based virtualization. This includes an overview of how UML works internally and test examples.

Speakers
avatar for Richard Weinberger

Richard Weinberger

CTO, sigma star gmbh
Richard is co-founder of sigma star gmbh where he offers consulting services around Linux and IT security. Upstream he maintains various subsystems of the Linux kernel such as UserModeLinux and UBIFS. Beside of low level and security aspects of computers he enjoys growing lithops... Read More →


slides pdf

Tuesday October 29, 2019 16:20 - 16:55 CET
Bellecour 3
  Linux Systems
  • Session Slides Included Yes

16:20 CET

Kernel Documentation: What We Have and Where We're Going - Jonathan Corbet, LWN.net
The Linux kernel contains a great deal of documentation, but it has not always been as well cared for as we might like.  In recent years a lot of work has been done to improve our documentation, including the adoption of a new, Sphinx-based toolchain.  This talk from the kernel documentation maintainer will cover what has been done to improve our docs and what is yet to come.

Speakers
avatar for Jonathan Corbet

Jonathan Corbet

Penguin herder, LWN.net
Jonathan Corbet is the kernel documentation maintainer, co-founder of LWN.net (and the author of its Kernel Page), a member of the Linux Foundation's Technical Advisory Board, and the lead author of Linux Device Drivers, Third Edition. He lives in Boulder, Colorado, USA.


Tuesday October 29, 2019 16:20 - 16:55 CET
Salon Pasteur

17:10 CET

Device Tree: Past, Present, and Future - Neil Armstrong, BayLibre
Since the switch of the ARM Linux support from the stable PowerPC Device Tree support, it became an important piece of software used to describe all sorts of devices based on very different hardware architectures.
Currently, BSD* Unixes and even the Zephyr RTOS has switched to Device Tree to describe the hardware. U-boot has also a file format using the Device Tree blob format.
Neil will present you the history of Device Tree from its origins, how it has been used for ARM and now RISC-V from the PowerPC codebase, all the very different current usage and an overview of its future application and evolutions.


Tuesday October 29, 2019 17:10 - 17:45 CET
Lumiere Auditortium
  Linux Systems
  • Session Slides Included Yes

17:10 CET

KernelShark is Now Faster and Stronger and Dives Deeper - Yordan Karadzhov, VMware
KernelShark is an Open Source User interface for Ftrace data. The work on the original GUI was started in 2009 by Steven Rostedt. Two years ago we initiated a complete rewrite, aiming to make The Shark faster and stronger and to allow it to dive much deeper. Finally KernelShark 1.0 has been released. It is a fully functional tool that is much more featureful than its predecessor, but this was just a preparation of the game field and no revolutionary changes have been made visible for the users so far.

All this is about to change. Most of the new key features, that motivated the rewrite, are now ready to be demoed. This includes: merging and visualizing together data from Host and Guest VMs. Abstracted data input allowing to combine data from different tracers. Automated analysis of this data via scripts. Powerful infrastructure for plugins that provides great degree of user customization.
Come to this talk to see this new beast in action.

Speakers
avatar for Yordan Karadzhov

Yordan Karadzhov

Open source engineer, VMware
Yordan Karadzhov has more than 12 years of experience as experimental physicist, includes a Ph.D. in particle physics. During this period Yordan worked in some of the world's largest physics laboratories, like CERN, FermiLab and RAL, developing software for particle physics experiments... Read More →



Tuesday October 29, 2019 17:10 - 17:45 CET
Salon Pasteur
  Linux Systems
  • Session Slides Included Yes
 
Wednesday, October 30
 

11:30 CET

File System Support for Zoned Block Devices - Naohiro Aota & Damien Le Moal, Western Digital
Zoned block device (ZBD) support has been introduced in Linux with kernel
version 4.10. ZBDs have different write constraints than regular block devices.
A ZBD is divided into several zones and each zone must be written sequentially.

The main type of ZBD currently available is SMR HDDs. The NVMe Zoned NameSpace
proposal is also being drafted to add a zone abstraction to the NVMe specifications.

Natively supporting ZBDs in a filesystem is not a trivial change. Some
filesystems must rely on special block layer drivers to ensure sequential writes
(e.g. ext4 and the dm-zoned device mappers). Filesystems using a copy-on-write
design are better candidates for native ZBD support. Examples are F2FS and btrfs.

This talk discusses the principles of ZBD native support in filesystems. Support
in F2FS is discussed and the approach taken with btrfs is next presented. This
is followed with a performance comparison between filesystems with native ZBD
support and regular ones using dm-zoned.

Speakers
NA

Naohiro Aota

Principal Engineer, Western Digital
Naohiro Aota is working at the System Software Group within Western Digital Research. He is working on zoned block device support for file systems like btrfs. He presented the on-going btrfs work at LSFMM 2019.
avatar for Damien Le Moal

Damien Le Moal

Distinguished Engineer, Western Digital
Damien Le Moal manages the System Software Group within Western Digital Research. Damien is a regular contributor to Linux kernel block I/O stack and file system support for the ZBC and ZAC standards (zoned block devices) and is a maintainer of the kernel dm-zoned device mapper and... Read More →



Wednesday October 30, 2019 11:30 - 12:05 CET
Lumiere Auditortium
  Linux Systems
  • Session Slides Included Yes

11:30 CET

How to Make an App Enabled Embedded Linux Product that Fits in 16MB of Flash with Containers - Alexander Sack, Pantacor
The audience will learn how to assemble a Linux embedded product from scratch that includes a basic OS and has the abiilty to install applications as containers using open source OTA technology. Participants will be able to make their own system that they can boot up, connect to the cloud and then deploy new applications to it.

Speakers
AS

Alexander Sack

CTO & Co-Founder, Pantacor
Alexander is a long term linux and open source leader who tries to make making linux embedded products easier for everyone. During his career he lead various workshops and BoF sessions on a broad set of topics at prominent Linux events such as Ubuntu Developer Summit and Linaro Connect... Read More →



Wednesday October 30, 2019 11:30 - 12:55 CET
Salon Pasteur
  Linux Systems
  • Session Slides Included Yes

12:20 CET

Combining WrapFS and eBPF to Provide a Lightweight File System Sandboxing Framework - Ashish Bijlani, Georgia Tech
Filesystem (FS) sandboxing is a useful technique to protect sensitive data from untrusted binaries. However, existing approaches do not allow fine-grained control over policy enforcement (e.g., seccomp), require sudo privileges (e.g., SELinux), incur high performance overhead (e.g., ptrace, FUSE), or are prone to TOCTTOU bugs (e.g., syscall interposition).

We combine eBPF with WrapFS to provide a lightweight, fine-grained FS sandboxing framework called SandFS for unprivileged users and containers. It is a stackable kernel FS that can safely be extended at runtime from user space using eBPF framework to enforce custom security policies in the kernel and offer native performance.

Unprivileged users can use SandFS for protecting private files (e.g., ssh keys) while executing untrusted binaries (e.g., ML models). Web browsers can enforce custom access checks to protect private data from extensions. Containers can be hardened by mounting a separate sandboxing FS layer for each service.

Speakers
AB

Ashish Bijlani

PhD Student, Georgia Tech
Ashish is a senior PhD student at Georgia Tech, Atlanta. His doctoral research focuses on mobile storage and security. He has presented his research at top-tier academic CS conferences and premier conferences, such as OSSNA'18 and LPC'18.



Wednesday October 30, 2019 12:20 - 12:55 CET
Lumiere Auditortium
  Linux Systems
  • Session Slides Included Yes

14:25 CET

Taking Full Advantage of NVME SSDs by Optimizing Page Cache Performance - Christopher Lameter, Jump Trading LLC
Modern SSDs allow transfer rates via NVME of 2-3 Gbytes per second. One can hook up multiple of these amazing drives to a system but sadly most enterprise servers in their default configuration will top out at 3GB/sec and the aggregate performance of multiple of these drives (f.e. via a RAID0 configuration) cannot be effectively utilized. This is a particular problem with Intel processors due to their small 4K page size.

The talk will investigate various ways to deal with this problem and show the advantages and limitations of each approach.

A. Exploiting NUMA affinities to significantly increase performance
B. Direct I/O
C. Huge pages
D. Various knobs in the Kernel configuration that affect performance

There are ongoing efforts to make it easier to get full performance without using any of these techniques by allowing huge page support in the Linux page cache.

In conclusion we will mention current issues to be aware of in current Enterprise distributions that affect performance.

Speakers
avatar for Christoph H Lameter

Christoph H Lameter

Universalist specializing in Computer Science, None
Christoph Lameter retired in January 2020 from High Frequency Trading company in Chicago where he was working as a Team Lead in research and development until the end of January 2020. He was responsible for the R&D on new HPC and HFT hardware and to bring new vendors online as well... Read More →



Wednesday October 30, 2019 14:25 - 15:00 CET
Lumiere Auditortium
  Linux Systems
  • Session Slides Included Yes

14:25 CET

Practical OpenHPC: Cluster Management, HPC Applications, Containers and Cloud - Adrian Reber, Red Hat
Since its inception as a Linux Foundation project in 2015, OpenHPC (https:///openhpc.community) has steadily grown to provide a modern, consistent, reference collection of HPC software.

Although a primary focus of OpenHPC remains in enabling people deploying new HPC clusters to rapidly get their clusters up and running, the OpenHPC software repository itself serves as a reliable, portable, integrated collection of software, libraries, tools and user environment that can be employed in containers and VMs as well as HPC clusters to develop and execute HPC applications.

This workshop will begin with a brief, advanced introduction to OpenHPC. We will then guide attendees through several practical, hands-on exercise modules employing an OpenHPC-based cluster and the OpenHPC software repository to explore real-world activities.

Prerequisites
There are no prerequisites but please bring your fully charged laptop to participate. 

Speakers
avatar for Adrian Reber

Adrian Reber

Principal Software Engineer, Red Hat
Adrian is a Principal Software Engineer at Red Hat and is migrating processes at least since 2010. He started to migrate processes in a high performance computing environment and at some point he migrated so many processes that he got a PhD for that. Occasionally he still migrates... Read More →



Wednesday October 30, 2019 14:25 - 18:00 CET
St. Clair 3
  Linux Systems
  • Session Slides Included Yes

15:15 CET

Demystifying Linux MIPI DSI Subsystem - Jagan Teki, Amarula Solutions
The MIPI Display Serial Interface (MIPI DSI) is a versatile, high-speed interface for a variety of embedded solutions and it is the most common and widely used display interface.

Many users are moving to open source solutions and it becomes a daunting task for them to bring their LCD panels into working and usable state because of lack of technical documentation and guidelines for their vendor specific panels along with associated DSI controllers. So, this talk will address those issues and challenges which are observed while working on Allwinner MIPI DSI controller with variety of associated LCD panels, bridges by validating these interfaces via ARM Mali GPU.

This talk starts with a brief overview of Linux DRM subsystem with bounded display controller interfaces like HDMI, RGB, LVDS and DSI and then the talk will add more details about Linux MIPI DSI controller, DPHY, DSI panel, DSI bridge interfaces drivers along with how these display drivers interact with GPU drivers.

Speakers
avatar for Jagan Teki

Jagan Teki

Co-Founder/Embedded Linux Architect, Amarula Solutions
Jagan is an Embedded Linux Architect and Co-Founder of Amarula Solutions India. His work involves providing Mainline Linux and related ecosystem projects to run on customer hardware devices/boards. He is an active contributor for U-Boot, Linux, Buildroot, Yocto, and maintainer of... Read More →



Wednesday October 30, 2019 15:15 - 15:50 CET
Lumiere Auditortium
  Linux Systems
  • Session Slides Included Yes

16:15 CET

Best of Both Worlds: Building a Linux/Android Hybrid - Bernhard "Bero" Rosenkränzer, LinDev
Haven't we all wished for a "real" Linux device in our pockets? Android is nice, but doesn't make it easy to launch vi, or better yet, an X or Wayland based application - but on the other hand Android is where all the apps are...

This presentation looks into several approaches towards building a working Linux/Android hybrid that gives us the best of both worlds.

Speakers
avatar for Bernhard

Bernhard

Software Engineer, BayLibre
Bernhard "bero" Rosenkränzer has been a Linux developer since the days he saw a stack of 70 floppy disks containing an interesting, totally unknown OS back in the mid-1990s. Before joining BayLibre, he has worked for MandrakeSoft, Red Hat, Linaro and various startups. Outside of... Read More →


Wednesday October 30, 2019 16:15 - 16:50 CET
Lumiere Auditortium
 

Filter sessions
Apply filters to sessions.