Leslie Hawthorn

We’re excited to announce that Leslie Hawthorn, Community Action and Impact Manager for Red Hat, Inc. will be giving the Wednesday, July 10 keynote at Linaro Connect Europe 2013 (LCE-13).

An internationally known open source advocate, speaker and author, Leslie Hawthorn has spent the last decade engaged in the human-focused bits of technology: community tending, software project management, marketing and public relations. In March 2012, she joined Red Hat, Inc., where she leads the team responsible for Community Action & Impact as part of the company’s Open Source and Standards Team, Office of the CTO. Prior to Red Hat, she served as Outreach Manager at Oregon State University’s Open Source Lab and as a Program Manager for Google’s Open Source Team, where she managed the Google Summer of Code Program, created the contest known as Google Code In and launched the company’s Open Source Developer Blog.

When not focusing on all things open source and community at Red Hat, Leslie is in the “save the world business” and works on side projects that make the world a better place. She serves as a Board Member for CASH Music, an open source platform that allows musician to more easily market their independently produced works, the Privly Foundation, an open source community project that gives users more control over how their data is shared and harvested from social networks, and the Sahana Software Foundation, a software platform that helps NGOs and first responders prepare for and react to humanitarian crises. She Tweets from @lhawthorn, blogs at http://hawthornlandings.org and makes an excellent chicken curry.

In her keynote, titled “Cultivating Community: Best Practices for Overcoming Challenges to Collaboration” Leslie Hawthorne will address the following:

The technical benefits of the open source model are widely known: a wider pool of talent than any one company could bring to bear on a particular problem, the ability to focus more energy on key product differentiators rather than creating and maintaining foundational elements, and higher quality source code. From a technical perspective, open source has triumphed and ushered in a new era of development practices and methodologies that allow disparate groups with different and, often, conflicting agendas to work together effectively, creating better works, faster. Despite its successes, the practice of open source development remains challenging in many of its social aspects: achieving the best possible consensus amongst various parties, navigating the cultural differences inherent in working with a global team and dealing with contentious contributors. To put it another way, software and hardware are ‘easy’, people are difficult.

Drawing on years of experience working with hundreds of open source projects, Leslie Hawthorn will explore best practices for cultivating community cohesion, including:

• Creating consensus among different groups
• Mitigating conflicts that arise due to different cultural norms
• Smoothing the waters after hard conversations and arguments

With open source development as our lens, we will also examine how Negotiation Theory and Collaborative Communication can best be brought to bear in our communities. Attendees will be treated to an examination of real-world scenarios faced by various communities, an understanding of how these communities overcame their challenges and the ongoing work required to maintain community cohesion. Participants will leave with food for discussion on improving their own practices and armed with strategies for improving and strengthening their community relationships.

 More information on the LCE-13 schedule can be found at: http://www.linaro.org/connect/schedule

Bob Monkman

We’re pleased to announce that Bob Monkman of the Enterprise Segment Marketing Team at ARM will be giving the Thursday, July 11th keynote at Linaro Connect Europe 2013 (LCE-13).

As part of the Enterprise Segment Marketing Team at ARM, located in San Jose, CA, Monkman is focused on Enterprise Networking and Software Initiatives for the Enterprise Segment like Linaro Networking Group, SDN and NFV. Bob has been in embedded computing for 25+ years (Let’s just say before the term “world-wide web” , when sending an email beyond your own company was a challenge!), much of it in the commercial RTOS/Middleware world focused on the network communications infrastructure space and did a stint in High Performance Clustered Computing. The bulk of Bob’s career has been in Product Management/Marketing, Strategic Alliances, and Business Development, but he started his career as a hardware/software engineer for Tellabs. Bob’s technology experience includes embedded RTOS, Linux, M2M/IoT, Systems Management Middleware strategy and development tools- now delving deeper into silicon designs. Bob was the Carrier-Grade Linux PM at MontaVista from 2003-2005, where he was active in the Service Availability Forum and OSDL CGL Working Group. Bob holds a BSEE from the University of Illinois.

Monkman is from the Midwestern US (Detroit/Chicago) but has been in the SF Bay area for 16 years. He is married and has an amazing 9-year daughter whom he and his wife adopted from China in 2004. Bob loves watching his favorite sports teams, enjoys running, fitness, fine wine/beer/single malt scotch and classic cars (never simultaneously, however).

His keynote titled “Software Defined Networking- Trends in Networking & Cloud Infrastructure and its Impact on the ARM Ecosystem” will address the following:

If you ask several people for their definition of Software Defined Networking(SDN), you will likely get several different answers. The more general view is that it represents an emerging network virtualization paradigm, based on multiple levels of abstraction, that seeks to centralize, simplify, and bring a degree of automation to the configuration and management of network functions- re-implementing some functions in software virtual machines, on commodity server equipment. The goals are ambitious and seek to lower both CAPEX and OPEX, increase network agility and accelerate new revenue-generating service velocity. The jury is still out as to how far and how successfully these goals will be met, but this session will survey various initiatives such as the mobile operator’s Network Function Virtualization (NFV) working group, relevant open source projects and standards initiatives such as OpenFlow, OpenStack, and OpenDayLight. We will examine what is driving all of this and the potential impact and opportunity for the Linaro community, as well as what role we might play in this significant shift in cloud and networking infrastructure.

More information on the LCE-13 schedule can be found at: http://www.linaro.org/connect/schedule

The only way to predict the future is to have the power to shape it. ~Eric Hofer

The Linaro 13.05 release is now available for download!

The 13.05 Linaro release highlights the focused efforts of all the Linaro Teams – Working Groups, Landing Teams and Platform Teams who have provided all the updates and new features that are integrated on top of Android, OpenEmbedded and Ubuntu baselines during this development cycle.

It is through these monthly releases that Linaro, our members and community do more than predict the future of Linux on ARM; we plan and progress that future with open and collaborative teamwork and success. The 13.05 Linaro release gives a vision of that future as we highlight the accomplishments of each of our teams. Those accomplishments include:

The foundation for the Linaro Stable Kernel (LSK) tree has been prepared. LSK will track the kernel.org community released LTS (Long-Term Stable) kernel tree, and Linaro will add member selected features, platform support and bug fixes. During the 13.05 development cycle, a preview/proof of concept has been delivered, based on linux-3.9.4 stable tree with ARM LT topic branches (TC platform support) and big.LITTLE work (MP and IKS).

The announcement of big.LITTLE in-kernel switcher (IKS) code now available to the public. “This code was also developed for and tested on the Versatile Express TC2 development platform. An MCPM backend and possibly a special cpufreq clock driver are required for this code to be usable on other platforms,” wrote Nicolas Pitre in his announcement to the linaro development mailing list. More information and helpful links can be found at: http://www.linaro.org/linaro-blog/2013/05/02/the-linaro-iks-code-now-publicly-available/

Quality remains a focus at Linaro and on the testing and validation front, members of the LAVA team completed and tested the functionality of the first prototype production run of LAVA LMP. The Platform teams also announced that Texas Instruments Beaglebone Black is now running in LAVA. As part of the effort to improve the developers experience with LAVA, the different client tools were also consolidated into a single package in order to make it easier for the team to provide new features for LAVA users.

The Builds and Baselines team mirrors the quality efforts of the testing and validation team and notes various accomplishments across several of the Linaro teams and Working Groups (WG). The team notes that a detailed Compatibility Test Suite (CTS) report in the standard format is now being generated in LAVA and the QA Services team will include these reports in its weekly/monthly testing reports for platforms in LAVA. Another highlight for this cycle includes the completion of the Galaxy Nexus CI loop and images (without graphics acceleration) can be deployed with a graphics overlay for testing in LAVA. The team also reports that the full Android engineering build for Arndale is now available with virtual framebuffer.

The various releases that help make up the Linaro 13.05 release include:

• Linaro UEFI 2013.05
• Linaro Stable Kernel (LSK) preview 3.9.4-2013.05
• Linux Linaro 3.10-rc2-2013.05
• Linaro GCC 4.8-2013.05 and 4.7-2013.05 (based off the latest GCC releases)
• Linaro Toolchain Binaries 2013.05

A complete list of highlights and accomplishments for LAVA, Builds and Baselines, Graphics, Kernel, Power Management, QA Services, Toolchain, Linaro Enterprise Group (LEG) and Linaro Networking Group (LNG) for the 13.05 Linaro relase can be found at: https://wiki.linaro.org/Cycles/1305/Release

Do you want to help predict the future of Linux on ARM?  Join us in Dublin, Ireland for Linaro Connect Europe 2013 (LCE-13).  Registration is still open! Do you want to highlight the work you or your organization is doing with Linaro code and ARM-processor based boards then be a show off and participate in our Demo Friday event at LCE-13.

USING THIS RELEASE:

We encourage everybody to use the 13.05 release. The download links for all images and components are available on our downloads page:

http://www.linaro.org/downloads/

See the detailed highlights of this release to get an overview of what has been accomplished by the Working Groups, Landing Teams and Platform Teams. The release details are linked from the “Details” column for each released artifact on the release information:

http://wiki.linaro.org/Cycles/1305/Release#Release_Information

USING THE ANDROID-BASED IMAGES

The Android-based images come in three parts: system, userdata and boot. These need to be combined to form a complete Android install. For an explanation of how to do this please see:

http://wiki.linaro.org/Platform/Android/ImageInstallation

If you are interested in getting the source and building these images yourself please see the following pages:

http://wiki.linaro.org/Platform/Android/GetSource

http://wiki.linaro.org/Platform/Android/BuildSource

USING THE UBUNTU-BASED IMAGES

The Ubuntu-based images consist of two parts. The first part is a hardware pack, which can be found under the hwpacks directory and contains hardware specific packages (such as the kernel and bootloader). The second part is the rootfs, which is combined with the hardware pack to create a complete image. For more information on how to create an image please see:

http://wiki.linaro.org/Platform/DevPlatform/Ubuntu/ImageInstallation

USING THE OPEN EMBEDDED-BASED IMAGES

With the Linaro provided downloads and with ARM’s Fast Models virtual platform, you may boot a virtual ARMv8 system and run 64-bit binaries.  For more information please see:

http://www.linaro.org/engineering/armv8

GETTING INVOLVED

More information on Linaro can be found on our websites:

• Homepage: http://www.linaro.org
• Wiki: http://wiki.linaro.org

Also subscribe to the important Linaro mailing lists and join our IRC channels to stay on top of Linaro developments:

• Announcements: http://lists.linaro.org/mailman/listinfo/linaro-announce
• Development: http://lists.linaro.org/mailman/listinfo/linaro-dev
• IRC:
  • #linaro on irc.linaro.org or irc.freenode.net
  • #linaro-android irc.linaro.org or irc.freenode.net

KNOWN ISSUES WITH THIS RELEASE

For any errata issues, please see:

http://wiki.linaro.org/Cycles/1305/Release#Known_Issues

Bug reports for this release should be filed in Launchpad against the individual packages that are affected. If a suitable package cannot be identified, feel free to assign them to:

http://www.launchpad.net/linaro

Andrea Gallo, Director Linaro Enterprise Group (LEG)

For Linaro though, it’s tomorrow, Thursday, 23 May that’s the big day at LinuxTag.  If you are in Berlin, interested in Linux on ARM and want to connect with developers from Linaro then you are in luck as Andrea Gallo, Director of the Linaro Enterprise Group (LEG) is set to open the Linux on ARM/Hardware in Action track. Gallo’s talk will explain how the governance works for both the core Linaro activities and the group specific ones, will describe the latest achievements in the various sub-teams, the mid-term roadmap and the resources available to all community developers to get involved and contribute.

Sessions given in English include:

Linaro Expands its open source collaboration model with the new Enterprise and networking groups – Andrea Gallo (Linaro)

Xen on ARM – Stefano Stabellini (Citrix)

Linux on ARM servers – André Przywara (Calxeda / Linaro)

Open- Source embedded GNU/Linux Plattformen – Benedikt Sauter (embedded projects GmbH)

OpenPhoenux Smartphone: Learnings from the past and ideas for next year – Nikolaus Schaller (Golden Delicious Computers GmbH&Co. KG), Lukas Märdian (OpenPhoenux.org), Christoph Mair (OpenPhoenux.org)

More Information

More information about the full program being offered to attendees of LinuxTag can be found at:  https://www.linuxtag.org/2013/en/program/programm-pdf.html

More information about Linaro can be found at: http://www.linaro.org/

Renato Golin, Linaro Toolchain Engineer

Linaro developers and engineers can be found at various events across the globe.  This week we catch up with Renato Golin of the Linaro Toolchain Team who recently attended the 3rd Annual EuroLLVM event which took place in Paris, France on April 29th and 30th, 2013.  Curious about EuroLLVM, I asked Golin for a few more details about this event and more.

Question: What is EuroLLVM and what was your role in this event?

EuroLLVM is the European version of the LLVM Developers Meeting, in US, where we gather to exchange ideas, present the progress of our projects and discuss the future of the toolchain. This time we had many people from the US, including in the speaker list, which is a good indication that this event is making its mark on the LLVM world.

Back in 2011, I organized the first EuroLLVM in London and, following the success we had a 2012 version was organized by ARM. This year, a team of volunteers in Paris got the lead. I was more on the sidelines than organizing the event, helping with lessons learned, getting the badges, taking the pictures, running around solving last minute problems, etc. The credit this year goes to the core team in Paris, they’ve done an amazing job.

Question: For those who may not know why this event is important can you tell people a little bit about LLVM?

LLVM is an open source toolchain, much like GCC, but with a permissive license, a modern C++ codebase and the possibility of using JIT compilation, which seems to be getting a lot of traction in the graphics community. Another crucial feature, in comparison to GCC, is the ease to add new functionality, even if you keep them private, which makes it very desirable for prototyping or commercializing products on top of it. LLVM has support for a number of architectures, from ARM to SystemZ, and can produce pretty decent code when compared to most modern compilers.

Most people, when talking about why they have chosen LLVM for their personal project, commercial product or academic research, mention how easier it is to get along with the LLVM community. LLVM is young, but it’s moving at a faster pace than possibly any other compiler and part of that comes from how the community works. There is very little protectionism on legacy code, and people are generally welcome to submit patches, which get reviewed quickly and constructive reviews are done. I’ve been following the list since 2009 and don’t remember any major flame or aggressive emails through the list, even when people had horribly messed up the code.

Question: Can you tell readers a little about your role at Linaro?

My current role is to understand how LLVM fits around the ARM ecosystem and what we can do about it. For now, we’re mainly interested in understanding what LLVM can and cannot do, so that we can create a TODO list and present to the community. It’s only when we know that the issues are, how easy it is to implement them and how important it is for the community that we can correctly prioritize and work on it.

Our main focus will be around the Cortex-A* platform, mainly A9 and A15, paying special attention to automatic vectorization (NEON), ABI and cross-compilation issues, as well finding relevant benchmarks for ARM platforms. We want LLVM to perform well where people will use it, not just to look good in the paper.

Question: With Linaro Connect Europe happening in July what can attendees expect to see on the LLVM front?

The Linaro LLVM team has just started, so I can’t promise much. I’m working on the 3.3 release now and will run some benchmarks after that, hopefully before LCE, so that we can get at least some numbers out to share with the other Linaro teams, members and the general ARM community, so that we can focus on the right things up from the start.

The only thing I can promise now is that we’ll listen very carefully to all suggestions, so if anyone has any special requests, please make sure you get them to us, or we’ll never know how important it is for you.

Question: What else would you like people to know about you, EuroLLVM, or your work at Linaro?

That’s about it.

More information about the Linaro Toolchain Team can be found at: https://wiki.linaro.org/WorkingGroups/ToolChain

More information about LLVM can be found at: http://llvm.org/

More information on Linaro Connect Europe 2013 can be found at: http://www.linaro.org/connect

(Side note:  David’s answers are in blue and maddog’s in a dark orange so you can see the interactive exchange that was had as they answered the questions)

At the time of the interview I asked both David and maddog to tell about their current roles and what they do.

David Rusling was at the time of this interview and currently the Chief Technical Officer at Linaro. “2 years, I helped form it. I wave my arms and try to stay ahead of the bow wave,” said Rusling about how long he had been with Linaro and what he does in his role as CTO.

maddog is the Executive Director of Linux International, President of Project Cauã, and Industry Consultant. When asked what he does and how long he has been in those roles he had the following to say.

For Linux International (www.li.org) I have been a voluntary (unpaid) ED since 1995.  LI defended and protected the Linux Trademark against attack, helped to start the Free Standards Base project and helped to start the Linux Professional Institute.  I travel the world helping governments, corporations, companies, universities and people learn how to make or save money with Free Software.   Other than that it is fairly dormant, as the Linux Foundation is doing a good job with most things Linux.  Someday I may start it up again if I see a need that has to be addressed.

Project Cauã (www.projectcaua.org) is a project to help create millions of sustainable, private sector, entrepreneurial jobs in dense urban areas in Latin America.  It is having a few issues about getting to pilot launch, but we are addressing them.

I make my living as an Industry consultant, mostly in Free and Open Source Software.  I am always looking for new business and things to do, so people who are willing to pay are welcome to send me email.  “Free as in Freedom, not as in beer”

[QUESTION] – It’s my understanding that you all worked together at Digital Equipment Corporation together?  If so, is it true that you all worked in the same office and on the same version of Unix?  Did you all know each other before working there?  What is one story from those days you both remember fondly that would help those who are getting stared in Technology?  What was the lessons you learned that you still remember and apply today?

Maddog up St Mary’s church tower in Oxford, about an hour before he got the royal wave (Photo credit: David Rusling)

David: Not only do we go way back, but maddog got me interested into Linux in the first place.  For which I am eternally grateful.

We both worked for Digital; but not even on the same continent.   I worked in Reading, England, maddog in New England.    The catalyst was that Digital decided to build the Alpha processor and sell it around the time that Linux was being created and maddog refused to let the second architecture that Linux ran on be anything other than an Alpha. I had never heard of maddog until I was told about Linux; I met him for the first time just before we went into a meeting with one of the Digital execs.   The rest is history.

As for Digital and Linux; all of us working on it were renegades.  At that time Digital was focused on VMS, Digital Unix and Windows NT.   The Linux work was done with very little management buy in. The thing was that Digital was a great engineering company with some amazing people working for it and you could get away with having a bright idea and following it.   maddog saw what was happening with Linux, believed that it would one day be big and infected the rest of us with his passion.

I was looking for something to own and Linux became that.   It was technically very challenging, fast and fun.   We did things in days or weeks when other teams would take that long to set up a meeting to discuss the possibility of doing something.   I learnt so much, so quickly.

maddog: Yes, we both worked for Digital, but it would have been impossible for David to share my office, as the office was so cluttered that things would have fallen on the poor chap.  I worked for the Digital Unix group, in Nashua, New Hampshire, USA and we had just created a 64-bit version of OSF/1 Unix on the Alpha microprocessor.   At that time the Alpha (a RISC processor) was the fastest in the world, and for a number of years we were registered in the Guinness Book of World Records as being the fastest.

I did not have the pleasure of knowing David before the Alpha/Linux project, but I did get to spend some very nice time with both him and his family in the years afterwards.

I do not know which stories we both remember fondly…perhaps the time that the port was finished and David and I went to a Digital Equipment User’s Society (DECUS) event in Dublin to talk about the project and show it off.  We shared a graduate student flat on the campus that was situated for four students, so had four bedrooms, two baths, a small kitchen and sitting room.  More importantly the campus had three huge pubs that served Guinness….we had a good time.  I am not sure that story will really get people “started in technology”, unless of course you consider that we had fun…and as Linus has often stated, having fun is very important to life.

David: I remember that weekend very well as we ended up going out to an Irish bar with a bunch of folks from Finland.   They gave us pints of Guinness as we got onto the coach to go into town.   I had a huge hangover the next day.   That was my 40th birthday. [maddog sighs:  only 40!]

maddog: Lessons…Hard work, honesty, love for your fellows, kindness towards others, taking care of your family, loving your life partner and children, guiding your children the correct way, earning a decent living, self-dependence, enjoying life and helping others to enjoy it too, these are the things that allow a person to look themselves in the mirror every day.  Do your best, every minute, every hour, every day.

If anyone asks anything more of you than that, it is their problem, not yours.

[QUESTION] - How did each of you move to Linux?  What was the first Linux distribution you can remember using (ok, I am sure you compiled it yourself and probably wrote some of it, too; did you?)?  What year was this?  What distribution do you currently use and if you are comfortable with answering, why do you use that distribution?

David: It’s all maddog’s fault.   The first distribution I ran was Slackware on an old Digital 386 with 120 Mbytes of disk (yes, Mbytes).    That was just before we got it running on Alpha systems.    We first ran a Slackware like home brew (and yes, we did compile it ourselves, but only after we fixed the compiler).    Red Hat were involved early on and we worked with them to create Alpha based distributions, so I ran Red Hat for years.   I got fed up with upgrading systems and moved to Debian for a while.    These days I run Ubuntu, but I occasionally flirt with other distributions.   Ubuntu is flexible enough to let me I use whatever GUI I want, mostly just works and lets me experiment.

maddog: I had been using the equivalent of “Open Source” ever since college in 1969.  I first started with software from DECUS, written by its members and contributed to their library.  This software could be obtained for 5 USD per copy (15 USD if it was long) on paper tape.  This charge was for handling and postal expenses, as the software itself was “free”.  I would then buy new paper tape from the school store and punch off copies that I would sell to my roommates at 1 USD per copy to make back my money.  This was both legal and encouraged by DECUS.

In 1992 when I worked for DEC I worked on a project called “Good Stuff” that compiled and built available Free and Open Source Software (FOSS) on the Digital Unix system, which we handed out to our customers.  It saved them time and money from doing this all themselves, and they loved the work.

In late 1993 I saw an advertisement for “A complete Unix system, source code included, for 99 USD”.  Intrigued, I bought it, only to find out that it only ran on an Intel-compatible device.  Since I only had “real computers” (VAX, MIPS, Alphas), I could not use it, but I mounted it on my workstation and looked at the man(1) pages.  Convinced that it was interesting, I put it in my filing cabinet.

In early 1994 a friend of mine, Kurt Reisler started sending emails to various companies and copying me on them.  Kurt was the chairman of the Unix Special Interest Group for DECUS (UniSIG).  Kurt wanted this person to come from Europe to the Spring DECUS event in New Orleans, and was looking for funding from various small companies.  They all told him that they did not have much money, but they would be happy to send CDs of their software to the event.

I felt bad for Kurt, since he was a good guy and had good ideas, so I went to my management and told them we should fund this, even though I did not really know this speaker nor what he had done, as “Kurt often has good ideas”.

My management funded 5,000 USD to bring this speaker to DECUS and to supply a PC to run the software.

I traveled to DECUS in May of 1994 and found Kurt trying to install some software on the PC and not being very successful.  Along comes this nice young man with sandy, brown hair, white socks and sandals and asks if he can help.  Ten minutes later Linux was running on that PC, for that young man was (of course) Linus.

During that event I saw Linux in operation for the first time, and conceived of the idea of porting Linux as a 64-bit operating system to the Alpha.  While riding a steam-powered riverboat, the Nachez, up and down the Mississippi River I convinced Linus to do the project.  It was symbolic that the decision to port a Free Operating system to the world’s fastest processor was done on board one of the earliest forms of automation.

I returned to my office at Digital and gave a presentation to a small group of managers on what I saw.  The last bullet of the last slide said that “Linux is inevitable”.  When I was asked what that meant I said “nothing is going to stop it”, and they laughed at me.

I sent around an email to all the engineers in Digital Unix about what I saw, and a couple of them wrote back and told me that they had been using Linux for six months on their laptops.  One engineer had written an entire subsystem using Intel Linux on his laptop sitting in his backyard under a tree.  He got the code to compile, then brought the source code to the office and put it on his Alpha system, compiled and linked it, and it worked.

These actions confirmed my resolve, and I started to “pull in favors”.  I had been at DEC for 11 years at that point, done a lot of favors for a lot of people, and now it was time to start “calling them in”. I called a person named Jim Jackson, and convinced him to give me a 30 thousand dollar workstation for a project and person he did not know.  “What will you pay for?”, he asked.

I told him I would pay for the shipping.

I found a group of engineers in the Alpha Technology section of Digital that were also thinking about porting Linux to the Alpha, but a 32-bit port.  I convinced them to join Linus and make it a 64-bit port.  This group was headed by Andry Riebs, and consisted of a group of engineers including Jim Paradis, Jay Estabrook and our own David Rusling.

Left to right:
Donnie Barnes, David Rusling, Erik Troan, Jay Estabrook, David Mosberger-Tang
(Photo Credit: David Rusling)

David: These chaps deserve more recognition.   They invented something called FX!32, technology that translated x86 code to Alpha code on the fly and were very bright and interesting people indeed.   I first saw Linux running on Jim’s Tadpole Alpha laptop; probably the same week that I met maddog.    By a strange coincidence, at that time Tadpole’s VP Engineering was one George Grey, now CEO of Linaro.

maddog: Linus took the rest of 1994 to study the Alpha architecture, finish up V1.2 of the kernel, and plan the change to the source code tree to support multiple hardware architectures.

We started the port in earnest January, 1995 and had a complete Red Hat distribution of Alpha Linux nine months later.

David:  That was really fun.   Everything happened so fast.   I created a bootloader called MILO which used bits of the kernel (block devices, file system, graphics) to load and start the kernel / OS.   The first code that sent to Linus (early PCI support if I remember correctly) was roundly rejected.   I was fuming but went for a walk outside and calmed down.    Linus, it turned out, was perfectly right and accepted the code once I’d rewritten it.

maddog: The rest, as they say, is history.

I joined Linux International in 1995, and became the Executive Director.  In 1999 I left Digital to “Linux” full time.

As a consultant, I use the same Linux distribution that my customer uses.  Recently I finished a job with Red Hat, so currently I am running Fedora.

And that first  “A complete Unix system, source code included, for 99 USD” that I bought in 1993?  It was an Yggdrasil Linux system.

[QUESTION] - Looking back at those days to present day, what amazes you most about the evolution and adoption of Linux and Open Source, both as an Operating System and as a philosophy. What’s been the biggest surprise not only in FOSS, but in technology in general, and what has been the biggest disappointment?

David: These days, if your business has anything to do with software, you will be involved with open source software.    If your business cannot interact with open source software, you will stop having a business. it’s a fact of life.   I wasn’t surprised that Linux was successful; at a certain point in time, it became unstoppable (probably around the time that Red Hat floated on the stock market and IBM backed it).   I am a little surprised that it’s powering the majority of the world’s smartphones, but maybe I shouldn’t be as I helped get Linux running on those early ARM processors (along with many, many others)

It’s not the code that’s the big thing though.   The big thing is the social engineering.    That communities of engineers join together to achieve coherent, robust and practical code bases that can be used in everyday products.    Linux will power the internet of things.

My biggest disappointments are unnecessary secrecy and patents.  Not everything a company does should be secret.  Some things may be, but companies end up with cultures where they do not differentiate between generally useful code and genuinely innovative software.  This means that they don’t share enough information, which doesn’t help the open source community support their hardware.  Innovation is also very overused, it most means ‘more duplicated code’.     Patents are an interesting area and some open source folks are entirely against them.  I’m not, I think that if you invent something original, you should be able to get some commercial advantage from it.   However, many, many patents (particularly software ones) are entirely obvious and the patent system (worldwide) is pretty broken.  The only people making money are the lawyers.

maddog: In those early days the thing that amazed me was the willingness of people I had never met to spend their own money and time to acquire an Alpha processor and help port this little-known operating system to that processor.  Even people who were not DEC customers understood the mission and would beg, borrow or (well hopefully not) steal an Alpha system to do the work.

One that stood out was David Mossburger-Tang, a student at a university in the western part of the United States.  David did a huge amount of work in porting libraries and making sure they were 64-bit clean.  After the porting project was finished David went on to write “SANE”, a back-end for scanners used to this day in Linux.  He also did some very good work in cache utilization, showing how to speed up programs by 40 times with proper cache usage.

I had found David Mossbuger-Tang an Alpha system so he could do his work faster, and shipped it to him on a loan-of-products.  Many years later David contacted me to return the system, but did not know who to return the system to, since Digital was no longer in existence.  David told me that he was currently working for Hewlett Packard.  “Just hold onto it”, I told David, “it will be ‘returned’ soon.”  The next week Hewlett Packard bought Compaq.

David: DavidM was great; wasn’t he at  Linus’ talk at the University of New Hampshire (I think that I have a picture)?   He was one of our first users and was great at testing software, getting new bits working and making things happen.   I think of him each time I use SANE.

maddog: Another super-star was Richard Henderson, who was a college student at Texas A&M University, and did the work for Alpha Linux to have shared libraries.  For his contribution I have sworn that he will never have to buy food or drink in my presence ever.  Later he was the leader of the GNU compiler project and today he works for Red Hat Software.

One final project out of the port was the math library.  Digital had done a huge amount of work on their math library for the Alpha, and while they were willing to have a binary object run on Alpha Linux, they were unwilling to ship the source code.  I was being beaten up by the community, when I turned to them and said:

 “If you are such hot-shot programmers, why don’t you write a better one?”

 Silence for a week.  Then an email:

 “sin(3) is 3% faster”

 another two days:

 “cos(3) is 5% faster

 and so it went until the entire library was re-written, and faster than Digital’s Alpha math library.

Only one routine was never any faster…because “nobody” used it, and nobody cared…it was fast enough.

 [QUESTION] - Now you all are smart guys, and have that “Chuck Norris” type technical knowledge combined with humble and like-able personalities, so how’d you get to be that way?

David: maddog was my mentor and when I stop to consider a philosophical position involving open source, I still think ‘what would maddog think?’.

I’ve been influenced by various great engineers over the time and what they all have in common is humility and respect.  They’re interested in solving problems, not in being right.    The open source world is not about giving orders, it is about influencing people.  The best arguments I’ve seen have been entirely ego free and deeply technical.

I got some email the other day which disagreed with me.  Someone one  the CC list told me that people should not send emails like that to the CTO of a company.  My response is that I’d have failed if that was true.  The guy sending the email was right.

maddog: Mom&Pop(TM) were a gigantic influence.  The steady hand upon the tiller, the safe harbor in the stormy seas of life.  Everyone liked them.

My career was influenced by a lot of people, many of which will be mentioned here for the first time.  Mr. Ralph Rigger, my woodshop teacher in middle school, Mr.  William Roberts, my electronics teacher in high school.  Philosophy came from my 10th and 11th grade English teacher, Mr. Koehler.

David: In that case, I’d better mention Miss Cole.   I’m mildly dyslexic and she helped me learn to read when I was eight.   She started a life long love affair with reading for which I am deeply grateful.

maddog: I owe a lot to my first “boss” and co-workers from co-op at Drexel University (nee Drexel Institute of Technology).  During three terms of co-op (1.5 years total) I worked at Western Electric in Baltimore, Maryland.  John Kammer (my mentor) was “an engineer’s engineer” (think Dilbert), and Bill Collins was his supervisor (fortunately the opposite of the “Pointy Haired Boss”).  In 1969 they encouraged me to sign up for a correspondence course in “How to program the IBM 1130 in FORTRAN”, which was my first exposure to software.

Bill was also the one who encouraged me to grow my beard, which I have not shaved since 1969.

After my first co-op, and while finishing my correspondence course, a salesman from DEC gave me a couple of paperback books on how to program the PDP-8 computers in the physics labs back at Drexel, and by reading those books, and practicing, I learned how to program in assembler and machine language.

 A professor of math and statistics at Drexel, Dr. Richard Haas, who was rumored to be the hardest teacher in the school.  Because of him the class came together and focused as a group, helping each other.  We later learned that his tough exterior was a way to get students to “meet the challenge”.  Several went on to become statistics majors.

 Another big influence was my first job at Aetna Life and Casualty, at that time the largest commercial user of IBM equipment in the “Free World”….a great place to work for a kid right out of college and who wanted to learn.  I also made friends with a lot of the operators in the computer room downstairs and I had direct access to the mainframes.

 My students at Hartford State Technical College, where I taught full time were another influence, where I learned at least as much from them as they learned from me.  I also learned that you do not really know something until you have to teach it to others.

Bell Labs.  What can I say?  Working with incredibly bright people.  Bea Fink, my first female supervisor, and one of the best supervisors I ever had.  I joined Bell at the North Andover facility, hired as their “senior systems administrator for Unix” and I had never touched a Unix system before.  Through my background, the study of books and the help of two great mentors, Bob Wessling and Tom Merrick, I learned Unix quickly, to the point that one day Bob threw up his hands and said “That is it!  I can not teach you anything more!  You have surpassed me!”

David: If I had a time machine, a visit to Bell Labs around that time would be on my list.

maddog: And then there was DEC.  I joined in 1983, the 16th person they had hired to work on their new distribution of Unix.  It was a small family of scrappers with a dream and a goal to create the best Unix system in the world.  I watched as young engineers came in from college to meet with older, more experienced engineers such as Fred Canter and Chet Juszczak.

We called the younger engineers “the boat people”, because they were hired all at once, and it was if they came off the boat at Ellis Island.

We watched and mentored.

Now a lot of those “boat people” work for Red Hat Software….and I am proud of them, what they did and what they continue to do.

Finally, most of the people I have met in FOSS.  As in any group, there are some that are lesser, and some that are greater, but for the most part, I think they are “greater”.  I love most of them.

Sorry, the story of my nickname has been told many times and in many places.  I earned the name at a time when I did not have control of my temper.  I keep using it to remind myself to never lose my temper again.

[QUESTION] - It’s my belief that anyone can become technical and if they really want to they can even become a developer, but there are other things about life in F/LOSS ecosystem that well quite honestly isn’t taught in books. So, lets say I came to each of you asking for mentor-ship and were given the opportunity to have each of you all as a personal FOSS mentor; what would you want to make sure I knew? Why? What resources would you point me to in order to help me ensure I build a strong fundamental cornerstone of F/LOSS knowledge. (ok maybe I am hinting here, just a little, but can you blame me for trying – I mean you all are legends!)

David: Find an itch to scratch.    What problem do you want to help solve? Find a FOSS community that’s doing work around there and have a look at what they’re doing and see how you could contribute.  Try to fit in.  Be useful.  Be patient.

Most communities have pretty good ‘how to join in’ information.    I like LWN as a good source of what’s going on.

maddog: I disagree that “anyone can become technical”, as I have met too many people that despite their best efforts and my own best efforts, they failed.  But this does not mean that they can not contribute, as I believe each person can contribute in their own way.

David: Good point.   When I went to my first UDS, I was amazed at the range of involvement, including testing, documentation, art work and so on.    One of my old bosses in ARM used to say that if all you can do to help is make the tea, then make the tea.

maddog: One of my many interests while growing up was beekeeping.  In a bee hive bees take on several different jobs throughout their lives, ending up (typically) as a “field bee”.   One of the jobs might be making honeycomb.  No bee will build an entire cell, each deposits a little wax, pinches it to shape it a little, then moves on.  In the end is a perfect honeycomb.

The essence of Free Software.

[QUESTION] - maddog I would ask you about general Open Source and Linux Philosophies and how to be a good consumer and contributor to F/LOSS projects.

maddog: Understanding the basic philosophies from a simple perspective is a good place to start:

• I start with Free Software so I do not have to write the whole thing
• I contribute to Free Software so I can work and learn from others
• I do not take away the freedoms of others, as it breaks the chain

It is as close to the proverbial “Golden Rule” as I think that technology can get.

Not all of us can contribute directly to writing code, but all of us can contribute something to make life easier.

At the same time I will point out that even the most ardent supporter of “Free Software” will say that you should be able to earn a living writing code, so there is this balance that is sometimes hard to see.

David: I often talk to companies and communities about open source and I emphasis the moral and philosophical duties beyond what is encoded in the licenses.   In essence, someone gave the code an open source license in order that it could be taken, shared, modified and used.   It is, therefore, your duty to respect that wish.

This can be hard for companies (and individuals) to grok.   It requires a paradigm shift from secret source(s) and NDAs.    Most companies need fairly sophisticated risk strategies in place to handle this interaction with open source well (or their efforts get stranded in their legal departments).    This is not an option, all companies need software and will, to some extent, have to interact with open source software.   It’s not the exception, it’s mainstream.

ARM is a good example.   ARM needs Linux to run well on systems that include its technology and yet the ARM partnership must keep systems under development secret until they’re released.   That’s a difficult balancing act.

[QUESTION] - David, I would ask you about starting on a Technical path and given that both you and I work at Linaro, I would narrow that to ARM what advice would you give me and others about getting started in this area?

David: I like to fix things (or rather, I don’t like broken things) and I like to know how things work.   What else would I be other than an engineer?    Most of the engineers that I know started by doing sciences at school / university.     A lot got into Linux as its free and they like to solve problems and tinker.    The issue, for me, is to create engineers that embrace open source.   I mostly see that outside of the west, in Brazil, Africa and so on.

[QUESTION] – I would ask both of you how do you influence an organization or corporate entity to adopt not only Linux and Open Source for their technical computing needs, but embracing the philosophies of F/LOSS as well. How important do you see this adoption for the future of Linux, F/LOSS, ARM etc.

David: Having spent years doing this at ARM; get the CTO and legal on board, nibble away at the problem.   Support the engineers.    Don’t give up.

maddog: In the beginning I wanted Alpha Linux to be used by universities to study how to use large address spaces to improve sorting and searching techniques.  I would still love to see Donald Knuth update Volume 3 of the “Art of Computer Programming” to include his thoughts on this.

Later I began to embrace the whole FOSS philosophy, which worked well with programmers and students, but was lost on businesses.  Captains of business were not interested in the “Kumbaya philosophies of sandal-wearing hippies.”  Yet I began to see a commercial value of Linux and FOSS that went beyond just “getting the software ‘for free’”

At that time I started talking about saving money and making money with Free Software, and that gained some interest.  I did not talk about software freedom, but talked to business people about regaining control of their business.  I talked about software slavery.  Most business people understand issues of “lack of control” and “slavery” much more than they understand Freedom.

I talked to governments about security, sovereignty, longevity of the solution, balance of trade, and brain-drain.  They responded.

It is my personal belief that for the bulk of software today, the market is too diverse for solutions to be produced in a “closed source” way, and that FOSS will replace closed source models.

As Linus would say “World Domination” is inevitable.

David: I agree.   Open source is mainstream.

[QUESTION] - If you had a magic wand and you could magically fix one area of the F/LOSS ecosystem, what would it be and why?  Since there are no real magic wands other than hard work, perseverance, and a quest for knowledge what do you see as the real world solutions to those problems?

David: I’d get the kernel communities to formally get together every 4 months to make decisions.   The in-kernel communities, server, embedded etc are having separate meetings – it makes heavy engineering in the kernel take too long.    I’d have 3 plumbers a year that they all go to.

Technology wise, Linux is not obsessed enough about power and battery utilisation.     There are also too many subsystems doing similar things.

maddog: With my magic wand I would eliminate software patents and drastically reform copyright laws.

Before the mid-1980s software copyrights and patents did not exist and we had huge amounts of innovation.  After 1990 the only thing truly new and innovative was a talking paper clip, and everyone hated that.

Today most large companies (other than patent trolls) swear they only have software patents to build a pool to defend themselves, yet if software patents did not exist, you would not need to defend yourself.  It is a circular argument.

In the end it is only the consumer who pays the large legal fees, or who is blocked from using a competing product.

Notice that I would eliminate only software patents.  Other types of patents are useful, but I believe the whole patent system does need drastic reform.

David: There’s a clear and present danger from loosely drawn software patents.   I think that this will settle down as the post-PC era finds its feet, and as governments reform patent law.  Sadly, it will probably happen very slowly.

[QUESTION] - Since most of the “Legends in Linux” are your peers and friends, do you have any that you are in awe of and would like to meet that you don’t already know?  If so, who and why?

maddog: I am not a hero worshiper.  I do not follow the lives of “famous people”, or ask for autographs, or have my picture taken with someone famous.  Many times I will ask to have my photo taken with a young person who I think has potential because I want to remember them, not because they are “famous”.

David: I have lost count of the number of pictures of maddog and friends that I see posted on Facebook.

maddog: I have been truly lucky in my life to have met and spent time with some really great people.  Besides the ones I have mentioned before, I have met Rear Admiral Grace Murray Hopper, Maurice Wilkes, Douglas McIlroy, Dennis Ritchie, Ken Thompson, Linus…the list goes on and on.  But some of the most interesting are the ones “behind the scenes” that few people have ever heard of their contributions, and I have been fortunate enough to not only know them, but to be invited into their homes, be with their families and their lives.

This is, for example, my relationship with Linus and his family.  When I do see them (not as often as I would like) we do not talk about Linux.  Instead we talk about mutual interests, and I talk with his wife Tove and his daughters.  I have met their families, who I like immensely, and we have taken sauna together.  Linus and his family went to a county fair with me one time, and to the boardwalk at Santa Cruz beach.

Probably three or four times early in our mutual friendship has the topic turned to Linux, and then only a tiny nudge to get things back on course.  It did not take much effort.

I appreciate the fact that I have known Richard Stallman for over a quarter century, and even though we do not agree all the time on everything, I will admit to having moved closer to his philosophies over the years.

On the wall of my office hangs a plaque with the poem “Desiderata” on it.  I bought the plaque when I was in college.  The poem begins “Go placidly among the noise and haste”….(something I have always found hard to do), and eventually talks about how there will always be people who are greater and lesser than yourself.  No matter how intelligent you are there will always be someone more intelligent than you, or more knowledgeable in some subject or a harder worker, or otherwise gifted where you are not, but “beyond a wholesome discipline, be gentle with yourself”.

The poem goes on to warn you to avoid loud and aggressive persons, “for they are vexations to the spirit”.  So I really prefer to be with small groups of people, to sit down and perhaps have a glass of beer rather than with large groups.  A lot of people might find it surprising to know that I am quite shy.

USENIX, DECUS and the people I met there, were also huge influences on my life, and of course the many thousands and thousands of people I have met around the world in the Free and Open Source movement.

I am sure that after this question a lot of business leaders will dismiss me as a “sandal wearing hippie”.  That is their problem, not mine.

David: I’ll echo Maddog’s comments.    There are, undoubtedly, some very bright individuals writing extraordinary code but, in general, open source is the sum of its many contributors, not of the few.   The influential ones are those that are good are enthusing others with their ideas and in growing capable teams around them.    Open source is distributed intelligence not centralised command and control.  I get a real buzz at open source working sessions, such as those we have at Linaro Connects, where a group of kernel hackers discuss a problem and strategies for solving it.    It’s a very interesting spectacle of social engineering as views are aired and consensus is reached.

That said, I’ve never physically met Ingo Molnar but I’ve watched his work with interest.   As the Linux kernel has got bigger tries to serve many needs, it takes a strength of character to propose wholesale changes in debug and in the scheduler.

I’d never describe Maddog as a ‘sandal wearing hippie’, I usually describe him as ‘Father Christmas on his holidays’.

[QUESTION] - Many people using Linux today don’t know about or even care about the whole GNU/Linux argument. Personally, I think it’s good to know the history of GNU and Linux and the importance of each.  As someone who mentors others, when would you introduce this topic, how would you introduce it, and what are your personally thoughts on the subject?

David: My friends are either in technology and know all about Linux or are entirely ignorant of technology.    With non-technologists you can point at Android (as they can buy that) and talk about open source communities.

maddog: Computers are a huge part of our life, everything from smartphones to supercomputers, yet most people have never even read the license that they blissfully click on.

I would be very happy if at least “computer science” people understood about GNU and Linux in depth, and the rest of the people practiced the “Golden Rule” more…

[QUESTION] - What do you see as the future of Linux?  What areas of this future are you most interested in?  Any predictions about what technology will be in say 1, 5 and 10 years?

maddog: World Domination.  There are currently 1.5 billion desktops in the world, and 7.5 billion people, so 5 billion people have not selected their desktop yet.  These people typically do not speak one of the 50 major languages of the world, do not do business the same way that Western Europe or the USA does business, and represent groups that are too small (and too unprofitable) for closed source countries to address their issues.

David: I don’t think that it will be a desktop (or what we now call a desktop).   For most people, a mobile phone or a tablet together with the infinitely connected internet is the perfect tool.

maddog: We no longer need a multi-million dollar, twenty-ton computer to write software.  I hold in my hand a computer faster than all the computers that Aetna Life and Casualty owned in 1977 when I left, yet it runs off a battery.  Thank you, Dave, ARM and DEC (who did a lot of work on the StrongARM chip and architecture).  It runs an operating system called “Android” based on the work of a college student from Helsinki, Finland.

David: I was listening to BBC Radio 4 on the way to work and an entrepreneur was talking about how easy it is for someone to have an idea and to create a scalable business to support it via the internet.    Creating a web site and writing software has never been easier.

maddog: Linux runs on 98% of the 500 fastest computers in the world today, half of all the servers shipped, and is the most used operating system in embedded system designs.

The timeline of the future?  Sorry, too hard to call.  By now we should all be flying around in rocket belts, according to the magazines of my youth.  But while the timeline is too hard to call, the direction is clear, and Free Software will be there.

David:  In some ways, we’re there already.   Just stop and be amazed.    Power and energy will dominate and something called the internet of things will actually make sense.

[QUESTION] - Is there anything else you would like for readers/viewers to know in regards to your personal journeys in F/LOSS, the companies you work for, or pet projects you are currently working on?

David: Being at Digital when alpha Linux was happening was a happy accident – I would advise people to grab opportunities and to always look for ways to develop, as an engineer and as a person.    My current project is looking at how to use Neon instructions in the kernel for cryptography.

maddog: Getting to know David, meet and know his family, and through a lucky accident, sharing with Dave the experience of a “Royal Wave” from the Queen Mother right before her Diamond Jubilee….but that is another story.

David and I see each other, way too seldom, at Linux events around the world.

David: Sadly true, but sitting on your hotel room balcony drinking your lovely home made Trappist beer and reminiscing was memorable.  I will treasure that Royal Wave forever.

Linaro Connect Europe 2013 Banner

Linaro Connect Europe 2013 (LCE13) is being held at the Burlington Hotel in Dublin, Ireland on 8-12 July.  As we enter this stage in our planning I wanted to remind you to register today!

Still haven’t decided if you should attend?  Do you or others in your organization wonder what is new with GNU Tools? Are you new to Linux on ARM? Maybe Android is your passion? What about those Linaro builds that are built on top of Ubuntu, Fedora, or OpenEmbedded? Do you want to help direct the future of ARM Enterprise Server?  And, let’s not forget about Networking on ARM.  If you have just read the list and thought,  “I need to know more about Linaro and the future of Linux on ARM”, then you definitely need to see how the schedule of events is coming together.

The LCE13 schedule is packed with a wide range of topics from the future looking to the practical hands-on. Tracks for this event focus on Android, Graphics and Multimedia, the Kernel, Platform Development, Power Management, QA and Infrastructure, Tools, Validation and LAVA, and training.

Due to the popularity of the LAVA workshop, which was held at Linaro Connect Asia 2013 (LCA13) in Hong Kong, there will be a Testing and Validation Summit at LCE13.  Sessions for this workshop include: LAVA LMP Showcase, LAVA LMP Features and Integration, LAVA Introduction to Test and Author Tools, and LAVA Engineering Workflows.

Those who attend can expect to participate in sessions such as ARMv8-A Status and Updates, GNU Toolchain Performance, UEFI, Linaro Stable Kernel (LKS), LAVA Introductions and Updates, Overview of Linaro Project Management Methodology, big.LITTLE Upstreaming Status Review and a Technical overview of big.LITTLE Switcher and that’s just a few of the sessions you will find on Monday!

LCE13 in Dublin is going to be a fast-moving, high-energy event where those who attend will be helping to design and build the future of Linux on ARM. Engineers who attend LCE13 will return to their companies with the latest Linux developments around ARM, new ways of optimizing ARM technology, an improved upstreaming support network and more.

You can see the session titles, tracks and place holders on the Linaro Connect website. Keep in mind the schedule is subject to change and the confirmed schedule with session abstracts, keynotes and more will be announced mid-May.

Registration for LCE13 is open, so register today.

Use the Linaro Social Media Channels to stay up to date on the latest announcements, downloads, Blog posts, team news and more.

Follow Linaro on Twitter | Like Linaro on Facebook | Circle Linaro on Google+

Are you currently using a Linaro build, toolchain, or code and want to show off your work?

Then Demo Friday is the place for you. We want to see how you are pairing Linaro’s engineering output and ARM processor-based hardware solutions.

If you or someone you know is interested in the opportunity to demonstrate how you are using Linaro, please email the details of your demonstration to: connect@linaro.org

In your email please include the following:

Please create a poster that lets attendees know all about your demonstration. The template for creating the demonstration poster is available here. All posters must be returned to Linaro by Friday June 7 in order to guarantee printing for Demo Friday event. Poster Size when printed will be A2 (11.7×16.5 inches).

By participating in this Demo Friday event you will also be given the opportunity to have your Demo recorded and promoted via the Linaro YouTube Channel and various Linaro Social Media sites.

If you want to know more about Linaro Connect and Demo Friday you can find highlights and demonstrations from previous Linaro Connect events on the Linaro YouTube Channel and on the Connect Resources page.

So here it is:

Branch big.LITTLE-IKS-snapshot of
git://git.linaro.org/landing-teams/working/arm/kernel.git

Although this is a snapshot of our latest IKS code, it still needs some
minor tidying before it is submitted upstream.  So please consider the
above as a temporary branch for people to look and play with, and not a
branch that we’ll keep stable and maintain.

This code was also developed for and tested on the VExpress TC2
development platform.  An MCPM backend and possibly a special cpufreq
clock driver are required for this code to be usable on other platforms.

The switcher concept is discussed here:
http://lwn.net/Articles/481055/

Porting documentation is available here:
https://wiki.linaro.org/projects/big.LITTLE.MP/Big.Little.Switcher/Docs/porting-guide

The switcher code description is available here (slightly out of date):
https://wiki.linaro.org/projects/big.LITTLE.MP/Big.Little.Switcher/Docs/in-kernel-code

And a review of the MCPM layer upon which IKS and HMP rely:
http://lwn.net/Articles/539082/

The core MCPM patches are queued for inclusion into the v3.10 mainline
kernel.  The MCPM backend for TC2 and the IKS patches will tentatively
be submitted for v3.11.

Originally posted to the linaro-dev mailing list by Nicolas Pitre on Wed May 1 18:33:38 UTC 2013

At first sight the ARM architecture is advancing rapidly because of the low power capability in ARM’s heritage. Developed for the battery driven mobile phone world, ARM cores are designed with low power and high performance/watt as key product objectives. In the server market alone, given that recent estimates show that data center power requirements are already over 2% of the US total energy usage, reducing power consumption is a key objective for both cost and environmental reasons.

The strength of the ARM community is not, however, just around low power; it is also around the ARM business model, which leads to rapid innovation and differentiation. I would argue strongly that the diversity of the ARM SoC vendor community, from strong, established companies to exciting new startups, is what makes the ARM architecture stand out today as the leader in SoC innovation.

Returning to market segments for a moment, it appears that we are in a period of extraordinary disruption in many markets – the rapid move to connected devices and cloud services is driving an explosion in data. The term “big data” is something of an understatement when you consider that 2.5 exabytes (that’s a 1 with 18 trailing zeroes) of data are being created every day. And the rate of increase in that number continues itself to increase.

No single vendor, however large, can innovate fast enough and address the different requirements and needs of these rapidly expanding market segments. I would argue that there are two key enablers that are critical to supporting the rate of innovation in the marketplace – one is the expanding role of the SoC itself, and the other is Open Source Software.

Designer SoCs
The ARM business model lends itself uniquely to product innovation. Through the ARM partnership, multiple vendors can deploy their unique value add and proprietary or custom IP around the ARM Cortex core technology. In most cases this value add comes in the form of adding additional processing elements into the SoC itself – for example advanced I/O or graphics processors, or specialized communications devices. The vibrant community of ARM SoC vendors specializing in single or multiple market segments is creating more choice and innovation in ARM SoCs than for any other microprocessor architecture.

Open Source Software
As this diversification leads to differentiated hardware products tailored for specific market segments, the challenges of delivering high quality software to drive these SoCs becomes more complex. One OS or distribution does not fit all requirements. Extensive changes to the OS kernel may be required to support the hardware innovation or differentiation. Traditional OS development cannot easily address this level of change given the increasing complexity of the underlying hardware. Instead, Open Source operating systems based on Linux have become the software of choice for many markets. Android, Fedora/Red Hat, Montavista, SUSE, Ubuntu and others have become the standard core software for everything from embedded devices to high­end data center servers.

Linaro
Linaro has established itself as the place for the ARM partnership and ecosystem to work together on core open source software. Linaro delivers a very high ROI to participating members by working on the delivery of core software needed by all members – Linaro now has over 150 open source engineers delivering substantial output, for which each member only pays a fraction of the total cost.

Much of this work is carried out by Linaro engineers working with Core and Club member assignees in the Linaro Working Groups. Key areas of current activity include continuing ARMv7 and ARMv8 toolchain work, big.LITTLE IKS and ARM’s MP support, ARMv8 validation and work across Linaro, and improved virtualization support for ARMv7 and ARMv8. Linaro’s Club and Core members define and prioritize the engineering effort, and participate in the development, testing and validation of the software on their SoCs.

Over the past 6 months Linaro has established two segment­focused Groups to deliver further value to existing and new members. The key concept behind Groups is to leverage Linaro’s shared open source software engineering model into segment focused areas, enabling engineering participation from distributions, OEMs and end users.

The Linaro Enterprise Group (LEG) was formed in November 2012. The Group is now fully up and running with 27 engineers focusing on accelerating the software ecosystem for the ARM server market. In addition we have recently announced the formation of the Linaro Networking Group (LNG), which is focused on the networking equipment market. Early work plans include support for Real Time extensions in the Linux kernel for the ARM architecture, work on virtualization for network applications, agreement and development on frameworks for data plane interfaces and support for legacy big­endian software.

Linaro’s membership has now grown to 25 members, each contributing engineering resources and fees to support our expanding engineering effort. The end result is an increasingly powerful software development team, completely focused on delivering the core open source engineering needed by the ARM community. By working on the development of common core open source software and preventing fragmentation, Linaro enables each member to focus more key resources on the development and delivery of their own value­added technology.