Home Blog Page 417

ELC + OpenIoT: From Cloud Computing to Robot Apocalypse

It’s time once again for that grand gathering of embedded Linux geeks known as the Embedded Linux Conference (ELC), as well as the co-located, non-Linux specific OpenIoT Summit. In this article, we take a closer look at the conference schedule, with keynotes and sessions you won’t want to miss.

ELC + OpenIoT is happening Mar. 12-14 in Portland, Oregon, the home of Linus Torvalds, who created Linux as a desktop OS and has watched it spread throughout the server world. Over the past decade, Linux has found similar success in embedded gear ranging from mobile devices to Internet of Things hubs to industrial equipment to drones.

Torvalds is not a scheduled speaker at the event, but the keynote speakers include the co-founder of another major open source embedded platform. Massimo Banzi is now the clear leader of the Arduino community after Federico Musto, the head of the formerly forked and then reunited Arduino Srl, departed last summer. There are no details as yet on the topic of Banzi’s Tuesday, 10 a.m. keynote, but it will likely be more about OpenIoT than Linux. Arduino appears to have backed away from its experiment in Linux/Arduino hybrid boards, such as the Yun, but it has a renewed emphasis on open source, community, and educational projects. Arduino is now focusing on its Raspberry Pi Zero sized MKR line of wireless, MCU-based Arduino boards.

Keynote lineup: From Google Cloud IoT to the robot apocalypse

Monday’s keynote lineup starts with a welcome from Sony’s Tim Bird, Chair of the Architecture Group of The Linux Foundation’s CE Working Group, and Philip DesAutel, the LF’s IoT Co-Chair. This talk will be followed by Antony Passemard, the Product Management Lead for Google Cloud IoT, who will discuss Google’s growing body of cloud-related IoT and analytics services. A hardware perspective on Linux will be provided by Patricia Florissi, VP and global CTO for Sales at Dell EMC.

Monday’s keynotes will be concluded with an address by Daniel Wilson titled “Sci-Fi Destroys the World, Science Builds It.” You can amp yourself up by reading Wilson’s bestsellers: Robopocalypse, Robogenesis, and How to Survive a Robot Uprising. (Maybe it’s time to keep that TurtleBot under lock and key.)

Tuesday’s lineup includes Banzi’s address, along with a keynote from DesAutels and Kate Stewart, Sr. Director of Strategic Programs at The Linux Foundation. Wednesday’s speakers include Amber Case of the Harvard’s Berkman Klein Center, as well as two longtime Linux leaders: Imad Sousou, VP and GM at Intel’s Open Source Technology Center, and Jonathan Corbet, Executive Editor of LWN.net.

Sign up for ELC/OpenIoT Summit updates to get the latest information:

Sessions: Real-Time Linux, Yocto, Zephyr, and much more

This year’s diverse selection of topics lacks a clearly dominant theme. There are, however, more than the usual sessions on Real-Time Linux. A half dozen ELC presentations include sessions on both the dominant Preempt-RT and its main rival Xenomai, and there’s a report from Ambient Sensors’ Sandra Capri on implementing a pseudo-real-time stack that uses neither of these technologies, with examples detailed for the Raspberry Pi and BeagleBone.

The ELC lineup also includes five sessions on the Yocto Project, which is quietly extending its reach in embedded Linux development on both x86 and ARM platforms, especially when a high degree of customization is required. There’s even a session on porting OpenEmbedded/Yocto to the open source RISC-V architecture (see farther below).

The OpenIoT track, meanwhile, has seven presentations on the lightweight Zephyr OS. This quickly maturing open source distribution for MCUs will also be showcased at Embedded World in Nuremberg, Germany, taking place Feb. 27 to Mar. 1.

The following list is a rather arbitrary selection of sessions that stand out from the pack, along with a few side notes. Other sessions not listed here cover topics such as ROS, GPUs, asymmetric multiprocessing, boot strategies, testing, security, RTS kernel, EAS, SiP, Eclipse, JavaScript, quality assurance, FPGAs, Bluetooth Mesh, NuttX, MQTT, EdgeX Foundry, Android Common Kernel, I3C, Tock, power management, virtualization on the edge, wireless debug, and Azure IoT Edge, among others.

10 sessions to consider for ELC + OpenIoT 2018:

Mozilla’s Iot Framework: Putting People First (Mon. 10:50) — Kathy Giori, Mozilla Corp. Last July, Mozilla unveiled an open source IoT project called Web of Things in collaboration with the W3C. The project has a Things gateway prototype that runs on a Raspberry Pi, and which builds on linkable and discoverable APIs based on standard web technologies.

Zephyr LTS Release, What to Expect and Why are We Doing This (Mon. 11:50) — Anas Nashif, Intel. In its first two years, the Zephyr RTOS has had 11 development releases. This year, we’ll see a major Zephyr LTS release with new features, stable APIs, and a subset of certifications.

Introduction to SoundWire (Mon 2:00) — Vinod Koul, Intel. The MIPI Alliance’s SoundWire audio bus spec, which was recently merged into Linux 4.16, should eventually replace HDA and I2C for low-cost, low-latency audio streaming in embedded devices and PCs.

OpenEmbedded Yocto on RISC-V — New Kid on the Block (Mon. 4:10) — Khem Raj, Comcast. This year we’ll see the first Linux-ready implementation of the open source RISC-V ISA architecture. OpenEmbedded/Yocto is one of first embedded Linux distribution frameworks to support RISC-V, but challenges remain.

Preempt-RT Raspberry Pi Linux (Tue. 2:00) — Tiejun Chen, VMware. Real-time Preempt-RT patches are increasingly being integrated in the Linux kernel, but not the kernel maintained specifically for the Raspberry Pi. Chen explores the potential and challenges of integration.

Secure Containers in Embedded Deployments (Tue. 4:20) — Stefano Stabellini, Xen Project. Stabellini discusses a new approach for enabling embedded hypervisors to run container apps such as minimalist Docker implementations on small embedded Linux devices without sacrificing real-time or safety.

The Salmon Diet: Up-Streaming Drivers as a Form of Optimization (Wed. 11:05) — Gilad Ben-Yossef, Arm. Ben-Yossef describes his experience in up-streaming the previously out-of-tree CryptoCell REE device driver, how the process helped reduce the line count by 30 percent while also decreasing bugs.

Sensor Types and their Uses (Wed. 2:30) — Mike Anderson, The PTR Group. With the increase of IoT and robotics applications, sensors such as gyroscopes, stress gauges, thermocouples, and more are increasingly part of the embedded developer’s toolkit. But what do they do exactly, and how do you integrate them?

Civil Infrastructure Platform: Industrial Grade Open Source Base-Layer (Wed. 3:30) — Yoshitake Kobayashi, Toshiba, and Urs Gleim, Siemens AG. Catch up on one of the Linux Foundation’s latest projects. The CIP project is producing a super long-term supported (SLTS) industrial software layer for civil infrastructure to encourage improved reusability.

3D Printing with Linux and Xenomai (Wed. 3:30) — Kendall Auel, 3D Systems Corp. Auel describes how a dual-kernel architecture using Linux and Xenomai can maintain low and predictable latencies for real time control of a 3D printer, while enabling a complex and resource intensive slicing application to run in parallel.

Registration

Registration is now open for the Embedded Linux Conference and OpenIoT Summit, to be held Mar. 12-14 at the Hilton Portland in Portland, OR. Packages start at $700, rising to $850 on Feb. 18. Academic and hobbyist discounts are available. Linux.com readers can register now with discount code, LINUXRD5, for 5% off the attendee registration.

Your Instant Kubernetes Cluster

This is a condensed and updated version of my previous tutorial Kubernetes in 10 minutes. I’ve removed just about everything I can so this guide still makes sense. Use it when you want to create a cluster on the cloud or on-premises as fast as possible.

1.0 Pick a host

We will be using Ubuntu 16.04 for this guide so that you can copy/paste all the instructions. Here are several environments where I’ve tested this guide. Just pick where you want to run your hosts.

Read more at Alex Ellis blog

What is the IoT? Everything You Need to Know About the Internet of Things Right Now

What is the Internet of Things?

The Internet of Things, or IoT, refers to billions of physical devices around the world that are now connected to the internet, collecting and sharing data. Thanks to cheap processors and wireless networks, it’s possible to turn anything, from a pill to an aeroplane, into part of the IoT. This adds a level of digital intelligence to devices that would be otherwise dumb, enabling them to communicate without a human being involved, and merging the digital and physical worlds.

Pretty much any physical object can be transformed into an IoT device if it can be connected to the internet and controlled that way. A lightbulb that can be switched on using a smartphone app is an IoT device, as is a motion sensor or a smart thermostat in your office or a connected streetlight. 

Read more at ZDNet

Containers, the GPL, and Copyleft: No Reason for Concern

Though open source is thoroughly mainstream, new software technologies and old technologies that get newly popularized sometimes inspire hand-wringing about open source licenses. Most often the concern is about the GNU General Public License (GPL), and specifically the scope of its copyleft requirement, which is often described (somewhat misleadingly) as the GPL’s derivative work issue.

One imperfect way of framing the question is whether GPL-licensed code, when combined in some sense with proprietary code, forms a single modified work such that the proprietary code could be interpreted as being subject to the terms of the GPL. While we haven’t yet seen much of that concern directed to Linux containers, we expect more questions to be raised as adoption of containers continues to grow. But it’s fairly straightforward to show that containers do not raise new or concerning GPL scope issues.

Read more at OpenSource.com

How to Fix the Docker and UFW Security Flaw

If you use Docker on Linux, chances are your system firewall might be relegated to Uncomplicated Firewall (UFW). If that’s the case, you may not know this, but the combination of Docker and UFW poses a bit of a security issue. Why? Because Docker actually bypasses UFW and directly alters iptables, such that a container can bind to a port. This means all those UFW rules you have set won’t apply to Docker containers.

Let me demonstrate this.

I’m going to set up UFW (running on Ubuntu Server 16.04), so that the only thing it will allow through is SSH traffic. To do this, I open a terminal and issue the following commands:

sudo ufw allow ssh
sudo ufw default deny incoming
sudo ufw enable

Read more at TechRepublic

Linux Kernel 4.15: ‘An Unusual Release Cycle’

Linus Torvalds released version 4.15 of the Linux Kernel on Sunday, again, and for a second version in a row, a week later than scheduled. The culprits for the late release were the Meltdown and Spectre bugs, as these two vulnerabilities forced developers to submit major patches well into what should have been the last cycle. Torvalds was not comfortable rushing the release, so he gave it another week.

Unsurprisingly, the first big bunch of patches worth mentioning were those designed to sidestep Meltdown and Spectre. To avoid Meltdown, a problem that affects Intel chips, developers have implemented Page Table Isolation (PTI) for the x86 architecture. If for any reason you want to turn this off, you can use the pti=off kernel boot option.

Spectre v2 affects both Intel and AMD chips and, to avoid it, the kernel now comes with the retpoline mechanism. Retpoline requires a version of GCC that supports the -mindirect-branch=thunk-extern functionality. As with PTI, the Spectre-inhibiting mechanism can be turned of. To do so, use the spectre_v2=off option at boot time. Although developers are working to address Spectre v1, at the moment of writing there is still not a solution, so there is no patch for this bug in 4.15.

The solution for Meltdown on ARM has also been pushed to the next development cycle, but there is a remedy for the bug on PowerPC with the RFI flush of L1-D cachefeature included in this release.

An interesting side affect of all of the above is that new kernels now come with a /sys/devices/system/cpu/vulnerabilities/ virtual directory. This directory shows the vulnerabilities affecting your CPU and the remedies being currently applied.

The issues with buggy chips (and the manufacturers that keep things like this secret) has revived the call for the development of viable open source alternatives. This brings us to the partial support for RISC-V chips that has now been merged into the mainline kernel. RISC-V is an open instruction set architecture that allows manufacturers to create their own implementation of RISC-V chips, and it has resulted in several open sourced chips. While RISC-V chips are currently used mainly in embedded devices, powering things like smart hard disks or Arduino-like development boards, RISC-V proponents argue that the architecture is also well-suited for use on personal computers and even in multi-node supercomputers.

The support for RISC-V, as mentioned above, is still incomplete, and includes the architecture code but no device drivers. This means that, although a Linux kernel will run on RISC-V, there is no significant way to actually interact with the underlying hardware. That said, RISC-V is not vulnerable to any of the bugs that have dogged other closed architectures, and development for its support is progressing at a brisk pace, as the RISC-V Foundation has the support of some of the industries biggest heavyweights.

Other stuff that’s new in kernel 4.15

Torvalds has often declared he likes things boring. Fortunately for him, he says, apart from the Spectre and Meltdown messes, most of the other things that happened in 4.15 were very much run of the mill, such as incremental improvements for drivers, support for new devices, and so on. However, there were a few more things worth pointing out:

  • AMD got support for Secure Encrypted Virtualization. This allows the kernel to fence off the memory a virtual machine is using by encrypting it. The encrypted memory can only be decrypted by the virtual machine that is using it. Not even the hypervisor can see inside it. This means that data being worked on by VMs in the cloud, for example, is safe from being spied on by any other process outside the VM.
  • AMD GPUs get a substantial boost thanks to the inclusion of display code. This gives mainline support to Radeon RX Vega and Raven Ridge cards and also implements HDMI/DP audio for AMD cards.
  • Raspberry Pi aficionados will be glad to know that the 7” touchscreen is now natively supported, which is guaranteed to lead to hundreds of fun projects.

To find out more, you can check out the write-ups at Kernel Newbies and Phoronix.

Learn more about Linux through the free “Introduction to Linux” course from The Linux Foundation and edX.

Index: A Focus on the Future of Code and Community

One of the most significant challenges developers face is keeping up with the increasingly rapid pace of change in our industry. With each new innovation comes a new crop of vendors and best practices, and staying on top of your game can become a second profession in itself.

Cloud, containers, data, analytics, IoT, AI, machine learning, serverless architecture, blockchain: Behind all of these rapidly evolving technologies are the programming languages and developers who are leading the charge into the next era of innovation.

An ideal way for developers to understand all this is through conversations with other developers. We believe conversation about development—like innovation itself—is best when it happens in the open. This idea was the catalyst for Index, a first-of-its-kind, developer-focused event that will take place in San Francisco Feb. 20-22 at Moscone West.

Read more at IBM developerWorks

A Look Inside Facebook’s Open Source Program

Open source becomes more ubiquitous every year, appearing everywhere from government municipalities to universities. Companies of all sizes are also increasingly turning to open source software. In fact, some companies are taking open source a step further by supporting projects financially or working with developers.

Facebook’s open source program, for example, encourages others to release their code as open source, while working and engaging with the community to support open source projects.

Read more at OpenSource.com

Q&A on Machine Learning and Kubernetes with David Aronchick of Google from Kubecon 2017

At the recently concluded Kubecon in Austin, TX, attended by over 4000 engineers, Kubernetes was front, left and center. Due to the nature of workloads and typical heavy compute requirements in training algorithms, Machine Learning topics and its synergy with Kubernetes was discussed in many sessions.

Kubeflow is a platform for making Machine Learning on Kubernetes easy, portable and scalable by providing manifests for creating:

  • A JupyterHub to create and manage Jupyter notebooks
  • Tensorflow training controller to adapt for both CPUs and GPUs, and
  • A Tensorflow serving container

Read more at InfoQ

Why You Should Care About Diversity and Inclusion

Aubrey Blanche, Global Head of Diversity and Inclusion at Atlassian, joins us in this latest edition of The New Stack Makers podcast to talk about the difference between diversity and inclusion and why anyone should care.

“Diversity is being invited to the party,” she said. “Inclusion is being glad you’re there.”

When you create an inclusive culture, Blanche explained, business thrives. Employees who feel comfortable bringing their authentic selves to work perform better and are happier at work, which leads to less turnover, which leads to greater profits.

Read more at The New Stack