December 2011 - microkernel-devroom

[Microkernel-devroom] [TALK-PROPOSAL] The microkernel overhead
by Martin Decky 09 Jan '12

09 Jan '12

Hello folks, please see below my proposal for a talk at our devroom. It would be great to cooperate with anybody who is interested in this topic, because it is not HelenOS-specific. Just let me know. Regards M.D. [Title] The microkernel overhead [Full name] Martin Děcký [Short bio] Martin has been with the HelenOS team for more than 6 years now. A life-long operating systems enthusiast, he enjoys everything from bare metal programming and submitting Linux kernel patches to designing the most progressive microkernel and working on its formal verification as a researcher at Charles University in Prague. [Estimated duration] 60 - 90 min [Abstract] Since the famous Tanenbaum-Torvalds debate [1] the general public sticks to the golden rule of thumb: Microkernel systems, while nice and elegant, are just academic toys. Due to the infamous communication overhead and other self-imposed limitations, they are never going to be as useful for the general use (in terms of performance) as the good old monolithic systems. Since the 1990s many researchers (especially the people around L4) have struggled to lower the overhead using the most extraordinary tricks. Others have tried to apply the microkernel design to mission critical, safety critical and other niche targets, where the benefits of the microkernel design clearly outweights the drawbacks. And other folks spent years creating hybrid systems to get the best (and hopefully not the worst) of the both worlds. But are the drawbacks of the microkernels fundamental? The way computers are designed, the way programmers think and the way the IT economy works have changed profoundly over the last 20 years. We no longer try hard to save every single CPU cycle and every single byte of RAM in every single routine. We acknowledge that spending 20 % more on a faster CPU and more RAM to run inteligebly designed software is a better idea than spending 20 % more each year on maintaing software with tons of ugly performance hacks and quirks. Our machines are massively concurrent and we tend to (or are forced to) think more in terms of effective parallel algorithms than just plain sequential throughput. So perhaps it is time to reconsider the true impact of the microkernel overhead given the present conditions and requirements. Key topics: * Reasons for the microkernel overhead * Qualitative and quantitative analysis of the overhead * Ways to minimize it * Ways to live with it * Ways to embrace it [1] http://en.wikipedia.org/wiki/Tanenbaum–Torvalds_debate

2 2

[Microkernel-devroom] [TALK-PROPOSAL] Dive into HelenOS Device Drivers
by Jiri Svoboda 31 Dec '11

31 Dec '11

Hi folks, here is my proposal for a talk about the specifics of implementing device drivers (device driver framework) in a microkernel operating system, specifically in HelenOS. Cheers -Jiri [Title] Dive into HelenOS Device Drivers [Full name] Jiří Svoboda [Short bio] Jiří is a long-time HelenOS developer who designed or contributed to many areas of the system, including the debugging framework, device driver framework, block layer, networking, the OpenMoko port, as well as functionality related to interfacing with the user. [Estimated duration] 45 min [Abstract] Implementing support for device drivers (device driver frameworks) in a microkernel-based/multiserver operating system presents some unique challenges. In this talk we will go through all the important aspects of device drivers in HelenOS, at the hurdles and the ways they can be overcome. We will discuss topics such as: - Programmed I/O and DMA in user space - Level interrupt handling and interrupt clearing, tasklets - Cooperation of user-space drivers and kernel (debugging) drivers - DDF and Device Manager - automatic driver start, device enumeration and hot (un)plug - Exposing driver services, Location Service - drivers are not created more equal than others, a.k.a Death of the Pseudo Device

1 0

[Microkernel-devroom] Reminder: deadline for talk submissions is tomorrow
by Jakub Jermar 30 Dec '11

30 Dec '11

Hello all, tomorrow is the last day before our devroom's deadline for talk proposal submissions. As of now, there are only four proposals from only two projects (HelenOS and Genode): Jakub Jermar: Welcome! or Why do we meet here today Norman Feske: Introduction of the Genode OS Framework Jiri Svoboda: Introduction to HelenOS Jakub Jermar: Fixing memory management in HelenOS It would be much better if we had talks from all participating projects and in a little bit greater quantity to make the devroom meet its intended purpose. Cheers, Jakub

1 0

[Microkernel-devroom] [TALK-PROPOSAL] Introduction to HelenOS
by Jiri Svoboda 21 Dec '11

21 Dec '11

Hi folks, here is my proposal for introductory talk about HelenOS. Cheers -Jiri [Title] Introduction to HelenOS [Full name] Jiří Svoboda [Short bio] Jiří is a long-time HelenOS developer who designed or contributed to many areas of the system, including the debugging framework, device driver framework, block layer, networking, the OpenMoko port, as well as functionality related to interfacing with the user. [Estimated duration] 30-45 min [Abstract] HelenOS is a general-purpose multiserver operating system written from scratch. It is based on a custom microkernel, SPARTAN, which currently runs on seven different CPU architectures (ARM, IA-32, x86-64, Itanium, MIPS, PowerPC and SPARC V9) . Most system services (e.g. device drivers, file systems, networking) are implemented in regular user-space processes which communicate via asynchronous messaging. First, the talk will briefly introduce the HelenOS project history and community structure. Second, we will briefly talk about the major technical areas and specifics of the system (boot and kernel, IPC mechanisms and naming services, devices and filesystems). This part will be supported by demos of the running system as well as by showing the structure of the source code. Last, we will look at what functionality is already there, what is currently being actively worked on and what are the future plans, which aim toward making the system self-hosting.

1 0

[Microkernel-devroom] [TALK-PROPOSAL] Introduction of the Genode OS Framework
by Norman Feske 19 Dec '11

19 Dec '11

Dear devroom participants, Here is my talk proposal about the Genode project. [Title] Introduction of the Genode OS Framework [Full name] Norman Feske [Short bio] Norman is one of the founders and architects of the Genode project. In 2008, he co-founded Genode Labs - a company with the mission to transform this OS technology from an once obscure research project to a next-generation main-stream OS. [Estimated duration] 45 min [Abstract] Today's operating systems try to find a balance between seemingly conflicting goals. Ease of use is traded against security, resource utilization is traded against resource accountability, and system complexity is traded against scalability. For example, SELinux is ill famed as hard to use and consequently remains widely unused. As another example, isolation kernels minimize the complexity of critical system software but at the cost of limiting these solutions to static applications. The Genode OS architecture shows how these apparently inherent conflicts can be solved by operating-system design. By combining a recursive system structure with capability-based security, mandatory access control becomes easy to deploy. At the same time, the trusted computing base can be minimized for each application individually such that the attack surface for security-critical system functions gets reduced by orders of magnitude compared to existing approaches. Furthermore, a concept for trading physical resources among processes allows for dynamic workloads while maintaining quality of service. That is not just theory - the system is ready for demonstration and its developers are planning to use it as development environment by the end of 2012. After a brief introduction of where Genode comes from, the main part of the talk will be focused on the OS architecture and give a glimpse at the implementation via live demonstrations. Finally, the talk will briefly discuss the planned steps towards using Genode as general-purpose OS.

1 0

[Microkernel-devroom] [TALK-PROPOSAL] Fixing memory management in HelenOS
by Jakub Jermar 18 Dec '11

18 Dec '11

Hi, As there is exactly only 13 days left (including holidays) for submitting talk proposals, I'd like to propose the following talk. At the same time, I don't want to be the only speaker in our devroom, so please, open the floodgates and start submitting proposals too :-) [Title] Fixing memory management in HelenOS [Full name] Jakub Jermar [Short bio] Jakub is a long-time developer of the HelenOS operating system and one of the project founders. Jakub helped to design and develop various parts of the system, including its kernel and the file system layer. [Estimated duration] 45m [Abstract] In this rather technical talk, I will go through the various corners of HelenOS memory management and will present the ongoing effort to fix some of the inherent shortcomings of the original design.

1 0

[Microkernel-devroom] FYI: Fwd: Re: Questions regarding the Multiserver, microkernel operating systems devroom
by Jakub Jermar 03 Dec '11

03 Dec '11

Resending response from FOSDEM's Pascal Bleser, contains interesting points about the presence of public. Btw, there will be 85 seats in our room. Jakub ---- > Second, on our devroom's mailing list, we have been debating the role of > the public that visits our devroom, but can't figure out how much the > public should be involved in the devroom. Is it common for the already > established devrooms to focus their programs rather on the interaction > with the public or the devrooms remain strictly developer-oriented? Depends, and that is interely up to you. We provide projects with the infrastructure, and you do (pretty much) what you want with it. FOSDEM has a huge number of visitors (6000 is a safe estimate), with 90% of them at least being active contributors in open source projects. So you can be sure that the people who will end up landing in your devroom will be highly interested in either hearing more about what you do, get deeply technical information about it, propose some interaction with another project, or just genuinly contribute to what you do. Given the amount of crowd at FOSDEM, you can be sure that the room will be full. You could also plan a mix of both things, have sessions that are more "internal" but also have a few talks that are more "introductory". The complete schedule of your devroom will be on the site and in the booklet which we distribute freely to all visitors.

2 4

[Microkernel-devroom] [TALK-PROPOSAL] Welcome! or Why do we meet here today
by Jakub Jermar 01 Dec '11

01 Dec '11

Hi, As there is exactly 31 days left for submitting talk proposals, I'd like to propose the following talk: [Title] Welcome! or Why do we meet here today [Full name] Jakub Jermar [Short bio] Jakub is a long-time developer of the HelenOS operating system and one of the project founders. Jakub helped to design and develop various parts of the system, including its kernel and the file system layer. [Estimated duration] 15m - 20m [Abstract] This short talk is intended as the opening talk of the entire devroom. After a welcome greeting, I will briefly ponder about the motivation for meeting together and attempt a basic quantitative comparison of participating projects based on available project-metric data. The goal of this quick comparison is to illustrate some apparent similarities and differences between our projects, and to further support our common cause.

1 0