l4-hackers

l4-hackers@os.inf.tu-dresden.de

2827 discussions

Samsung R&D - Full-time OS Researcher Position
by Daniel Waddington 10 Mar '11

10 Mar '11

Samsung R&D, San Jose - CA, is recruiting full-time researchers in the area of Operating Systems (OS) for future cyber-physical systems. We are looking for individuals with a passion for systems research and the desire to follow through from concept to pre-transition prototype. Position: Full-time Researcher for the Systems Research Group Summary: The Systems Research Group is developing new OS and software platform designs that can meet the needs of future computing platforms based on multicore, manycore and cloud technologies. We are currently developing a solution based on L4 microkernel technology that will readily scale from sensors-to-servers, whilst providing advanced support for transparent distributed processing, Quality-of-Service (QoS) guarantees for co-located real-time and best-effort applications, and the "first-class" integration of parallel programming technologies. We are looking for highly motivated and creative individuals to join the San Jose research team and to augment existing system research expertise. This is an excellent opportunity to join a new team advancing state-of-the-art software technologies for Samsung's current and future business. Duties and Responsibilities: . Research and develop micro-kernel based solutions for next generation products. . Research and develop innovative technologies to directly address scalability limitations of existing solutions. . Research and develop ideas around the single-system image concept. . Research and develop new ideas in OS-level and language-level support for heterogeneous processor platforms. . Develop ideas around the support of fine-grained resource management and QoS monitoring, policing and adaptation in the context of real-time cyber-physical applications. . Build proof-of-concept prototypes. . Write research papers and technical articles publishing work and results in top-tier conferences and journals. Necessary Skills/Attributes: . Ph.D. in Computer Science or related area with 0-5 years of practical experience. . A track record in OS research pertaining to the above areas of interest with high-quality publications in the field. . Hands-on experience with low-level systems prototyping and development - a true hacker at heart. . Proficiency with C/C++ and Linux-based development including multi-threaded programming. . Excellent communication and team working skills are required with a willingness to work with an international team to deliver an integrated solution. . A willingness to learn new things and take on new challenges. Desirable Skills/Attributes: . Experience with L4 or other similar microkernel technology (e.g., QNX, Integrity) . Experience with embedded systems development. . Experience building OS sub-systems such as memory managers, and task schedulers. . Experience with the use of machine-learning to drive system optimization. . Experience with distributed and parallel computing. . Experience developing on TILE processors and programming with TILE assembly. . Experience developing on ARM processors and programming with ARM assembly. Contact: Please send resumes and enquiries to Dr. Daniel Waddington (d.waddington(a)samsung.com)

1 0

Re: Genode OS framework and kernel-specific features
by Valery V. Sedletski 09 Mar '11

09 Mar '11

On Sun, 06 Mar 2011 09:29:52 +1200 (MSK), Valery V. Sedletski wrote: > >PS: maybe, offtopic, but I'm in trouble a bit. > Already googled the Linux Framebuffer HOWTO and several articles. It seems that vesafb will not work for me (I have Matrox G400 card, on which I get a message "Cannot set vga mode, press Enter..." when booting Linux. Also, it supports VESA 2.0, but GRUB sees only modes not higher than 800x600). I also found some drivers in ./kernel/drivers/video/matrox, will check them.

3 6

Can L4 modules be accessed as files from L4Linux?
by Dan Haim 07 Mar '11

07 Mar '11

Hi, I'm trying to find a solution to the following scenario - I want to have an L4Linux instance running (from a CD) where the root filesystem is several hundred MB's big, so it is not feasible to run it from a ramdisk (the filesystem is larger than the amount of RAM available) - I need to run it directly from the CD. A possible solution I'm thinkning about is to give the L4Linux instance access to the CD-ROM drive, then store a SquashFS image as a file on the drive and UnionFS it with a tmpfs when the system starts. Now here's the catch - While this solution may technically work, it is still not 'good enough' because I can't give the L4Linux instance direct access to the CD-ROM drive - I need to have it completely isolated so a user working on the L4Linux instance will not have access to any files stored on the CD (other than the SquashFS image) even if he/she has root access over the L4Linux instance. I was thinking - is it possible to specify the SquashFS image as an L4 module and then somehow mount it from within L4Linux in-place (i.e. without completely loading it to RAM)? Is there any other way to accomplish what's needed in this scenario? Thank you all, Dan.

2 2

Genode OS framework and kernel-specific features
by Valery V. Sedletski 05 Mar '11

05 Mar '11

Hi all. I am now giving a try to the Genode OS framework. I like it's concepts and ideas and its elegant design, and maybe, will migrate to it eventually. -- I was developing an OS/2 clone based on Fiasco microkernel and l4env. Now they're discontinued. So, I must probably, migrate to something else. The most probable candidates are L4Re or Genode. I like Genode because of its kernel-neutrality (It can work on any L4 flavour, it is very attractive feature) and has interesting ideas regarding the scalability and capability to build a hierarchy of subsystems, minimising the application-specific Trusted Computing Bases. So, as I understood, it abstracts the kernel specifics from a developer. But in some cases, it can be desirable to make some kernel-specific implementations. For example, Fiasco kernel based on l4v2 API has 'segments' feature which, for example, is actively used by L4Linux. We use this feature too. One of our objectives is a binary compatibility with existing OS/2 applications. And OS/2 application ABI includes a convention that FS register for Ring3 application, holds a selector of a structure named Thread Info Block. Though the application can obtain this structure by calling an DosGetInfoBlocks() API, the C startup code generated by many compilers, relies on the convention that FS must contain the TIB selector. So, if we try to implement this feature on L4, we must have the means to allocate GDT selectors and get the allocated selector. On Fiasco, this can be done by filling the registers in a special way and issuing a 'LLDT' instruction, or, by calling a fiasco_gdt_set() function, and get the allocated selector by fiasco_gdt_get_entry_offset() (which is the same, but through a 'l4sys' package). On Fiasco.OC, this can be done similar way (not through LLDT instruction, though). My question is: Can be the underlying kernel determined under Genode, and then try doing kernel-specific things? Or, maybe, there is a more elegant way to do kernel-specific things? Thanks in advance, valery. PS: On Pistachio, as I heard, the application also can operate on GDT selectors, but this possibility has even more limitations. (It seems, only LDT selectors can be altered, but not GDT ones, or something)

2 2

Compiling l4re for BeagleBoard
by Ivan Bludov 04 Mar '11

04 Mar '11

Hello, I'm trying to compile l4re for ARM (beagleboard). I used the cross-toolchain for AMR (from CodeSourcery and toolchain downloaded from your site), but I got errors during making configuration of l4 : =========================================================== set -e; X="tmp.$$$RANDOM.c" ; echo 'int main(void){}'>$X ; \ rm -f $X.out ; LD_PRELOAD=libgendep.so LD_LIBRARY_PATH=/home/ibludov/Downloads/l4re-snapshot-2011022209.0/src/l4/build/tool/gendep/64: GENDEP_SOURCE=$X \ GENDEP_OUTPUT=$X.out arm-linux-gcc -c $X -o $X.o; \ test -e $X.out; echo INT_CPP_NAME=`cat $X.out` \ >>/home/ibludov/Downloads/l4re-snapshot-2011022209.0/src/l4/build/.config.all; \ rm -f $X $X.{o,out}; ERROR: ld.so: object 'libgendep.so' from LD_PRELOAD cannot be preloaded: ignored. ERROR: ld.so: object 'libgendep.so' from LD_PRELOAD cannot be preloaded: ignored. ERROR: ld.so: object 'libgendep.so' from LD_PRELOAD cannot be preloaded: ignored. ============================================================ These errors appear only when I use the arm-linux compiler and do not appear when I use host compiler. Is it a problem of toolchan or some other system libraries? Can you make some advise to fix it? Ivan Bludov

2 1

thread creation help
by Jason Philip 27 Feb '11

27 Feb '11

Can someone tell me how to create a thread (in the same address space as the creating thread)? This very confusing to me, I need help.

3 5

Announcement: Genode 11.02 with support for both NOVA v0.3 and Fiasco.OC
by Norman Feske 24 Feb '11

24 Feb '11

Dear L4 hackers, I am pleased to announce the release of the version 11.02 of the Genode OS Framework. Genode is an OS architecture that promises to align high security and robustness with dynamic application workloads. The Genode OS Framework as the reference implementation of this architecture. With the new version 11.02, we enabled the framework to be used with the Fiasco.OC kernel developed at TU Dresden in addition to the already supported base platforms L4/Fiasco, L4ka::Pistachio, Linux, OKL4v2, Codezero, and NOVA. Furthermore, we upgraded the NOVA support to version 0.3, which was published earlier this month. So now, both current kernel developments at TU Dresden are fully embraced by the framework. We are grateful for the excellent support we got from Alexander Warg and Udo Steinberg. Alongside the expanded platform support, the new release features many improvements, the most prominent being - Fiasco.OC as new base platform - Running Genode natively on MicroBlaze CPUs - Support for version 0.3 of the NOVA hypervisor - Noux execution environment for GNU userland software - Base framework - Core resource accounting using guarded allocators - Protocol enhancements for upgrading session quotas - Qt4 updated to version 4.7.1 - New tooling support for automated integration and testing Please find a more elaborative description of the changes accompanied with a lot of background information in the release notes of version 11.02: http://genode.org/documentation/release-notes/11.02 The new version is ready for download at our download page at Sourceforge and our public SVN repository: http://genode.org/download Best Regards -- Dr.-Ing. Norman Feske Genode Labs http://www.genode-labs.com · http://genode.org Genode Labs GmbH · Amtsgericht Dresden · HRB 28424 · Sitz Dresden Geschäftsführer: Dr.-Ing. Norman Feske, Christian Helmuth

1 0

L4Re: C++ or DICE for IPC
by Valery V. Sedletski 24 Feb '11

24 Feb '11

Hi, all. I checked out a new L4Re userland and a new Fiasco.OC microkernel. I noticed that it seems, no IDL compiler is used for IPC anymore, but found that, like in l4\pkg\examples\clntsrv\ example, IPC is encapsulated in C++ classes. That's nice, C++ is a feature-rich object oriented language. But what can I do if I want to write a program on other languages, like plain C, Ada etc.? I think, such possibility must exist too, and IDL is a means of implementing such a language neutrality. Is DICE discontinued? And how can I do something with communication code if not using IDL? Thanks in advance, WBR, valery

4 6

Boot into video mode with grub2
by Dan Haim 23 Feb '11

23 Feb '11

Hi, I'm trying to boot L4Linux under MAG. I used to use the "vbeset 0x117" GRUB command to set the screen into the correct video mode but now I'm using GRUB2 and I can't find any equivalent command(s). Has anyone successfully booted MAG under GRUB2?

2 2

[L4 on realview Cortex A9] Failing on board check during fiasco bootup
by Naveen Chandrakar 23 Feb '11

23 Feb '11

Hello Guys, I was attempting to run fiasco built for CortexA9/realview-pbx board on qemu. "qemu-system-arm" version used for this experimentation is "QEMU PC emulator version 0.12.5 (qemu-kvm-0.12.5)" L4 Code used "l4re-core-2011020417" with hello world example. Below are the messages which i got when i tried to run fiasco on qemu - 1st set of logs I tried to comment the check inside Board_check::check_board() in the hope that it might boot up, but later i got crash for qemu (hw mismatch ?) - 2nd set of logs *1st set of logs* ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ L4 msgs specifying invalid sys id~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ novino@naveen-linux:~/L4/l4re-core-2011020417/src/l4/mybuild-realview-pbx-a9/images$ qemu-system-arm -M realview-pbx-a9 -cpu cortex-a9 -kernel bootstrap_hello.elf -m 256 -nographic -net nic -net user L4 Bootstrapper Build: #2 Thu Feb 10 18:21:42 MVT 2011, 4.5.1 Scanning up to 256 MB RAM ........some messages ..... Starting kernel fiasco at 00001000 Cache config: ON Realview System ID: Rev=0 HBI=178 Build=0 Arch=5 FPGA=00 Invalid System ID for this kernel config Expected (01780500 & ffffff00) == 1182f500 Stopping. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ end of msgs ~~~~~~~~~~~~~~~~~~~~~ *2nd Set of logs* ~~~~~~~~~~~~~~~~~~~ L4 msgs after modifying Board_check::check_board() ~~~~~~~~~~~~~~~~~~~~~~~~ novino@naveen-linux:~/L4/l4re-core-2011020417/src/l4/mybuild-realview-pbx-a9/images$ qemu-system-arm -M realview-pbx-a9 -cpu cortex-a9 -kernel bootstrap_hello.elf -m 256 -nographic -net nic -net user L4 Bootstrapper Build: #3 Thu Feb 10 19:06:41 MVT 2011, 4.5.1 Scanning up to 256 MB RAM Memory size is 256MB (00000000 - 10000000) RAM: 0000000000000000 - 000000000fffffff: 262144kB Total RAM: 256MB mod04: 010d3000-010f4450: hello mod03: 010b1000-010d2488: l4re mod02: 01073000-010b066c: moe mod01: 01061000-01072304: sigma0 mod00: 01015000-010608f8: fiasco Moving 5 modules to 1100000 with offset eb000 moving module 05 { 10d3000-10f4450 } -> { 11be000-11df450 } moving module 04 { 10b1000-10d2488 } -> { 119c000-11bd488 } moving module 03 { 1073000-10b066c } -> { 115e000-119b66c } moving module 02 { 1061000-1072304 } -> { 114c000-115d304 } moving module 01 { 1015000-10608f8 } -> { 1100000-114b8f8 } Scanning fiasco Scanning sigma0 Scanning moe --init=rom/hello Relocated mbi to [0x100f000-0x100f0d6] Loading fiasco Loading sigma0 Loading moe find kernel info page... found kernel info page at 0x2000 Regions of list regions [ 1000, 19ff] { a00} Kern fiasco [ 2000, 5cfff] { 5b000} Kern fiasco [ 90000, 9e193] { e194} Sigma0 sigma0 [ 140000, 18b48b] { 4b48c} Root moe [ 1000000, 10143eb] { 143ec} Boot bootstrap [ 100f000, 100f1d3] { 1d4} Root Multiboot info [ 119c000, 11df44f] { 43450} Root Modules Memory API Version: (87) experimental Sigma0 config ip:00090000 sp:01013044 Roottask config ip:00140088 sp:00000000 Starting kernel fiasco at 00001000 Cache config: ON Realview System ID: Rev=0 HBI=178 Build=0 Arch=5 FPGA=00 <<Commented>> Invalid System ID for this kernel config Expected (01780500 & ffffff00) == 1182f500 Stopping. Hello from Startup::stage2 Initialize page table Number of IRQs available at this GIC: 96 Number of IRQs available at this GIC: 32 Vmem_alloc::init() qemu: fatal: Unimplemented cp15 register write (c9, c14, {0, 0}) R00=00000000 R01=00000000 R02=80000000 R03=00000001 R04=f00487b0 R05=00054004 R06=00215406 R07=00115406 R08=00015406 R09=00002000 R10=00000000 R11=0100f058 R12=ffffffff R13=f0048670 R14=f005bd3c R15=f005bd28 PSR=a00001d3 N-C- A svc32 Aborted ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~end of msgs ~~~~~~~~~~~~~~~~~~~~~~~~ Warm Regards, Naveen

2 9

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

l4-hackers