l4-hackers November 2003

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

9 participants
5 discussions

microkernel and realtime
by Yu Neil 24 Nov '03

24 Nov '03

hello: Recently, I consider implementation heritical scheduler on top of the l4 microkernel, But we meet a problem, the Pistachio handled the interrupt in microkernel internal, In that time, I remember a project in TU-Dresden is build on top of microkernel, But it`s implementation in the old l4 interface, just like the Hazelnut that not handled the interrupt (?), The suitation is some microkernel abstract the interrupt and some microkernel not abstract the interrupt. Can we implementation a real time soft mechanism with interrupt abstracted microkernel, the real time ipc, feature as: interrupt generated then a ipc send to handler and switch to the handler at once Neil _________________________________________________________________ 与联机的朋友进行交流，请使用 MSN Messenger: http://messenger.msn.com/cn

3 2

Various L4-related questions...
by Michael Engel 17 Nov '03

17 Nov '03

Hi, it's quite some time since I last had a look at L4 and a lot seems to have happened since then. As I'm planning to use a L4-derived microkernel as basis for a research project here at Marburg University, I have some questions for the experts... The currently maintained L4-compatible microkernels seem to be Fiasco (Dresden) and L4KA::Hazelnut/Pistachio (Karlsruhe). Is there a document describing the differences/compatibility between the various versions, probably comparing these implementations to the original L4 kernel? The platform for my project will probably be a AMD "Hammer"-based system (e.g. Athlon 64 or Opteron). Is there already work underway to run Fiasco or L4KA in 64-bit mode on the x86-64 CPUs? If not, I'm thinking about starting an effort to port one of the kernels (based on the x86-32 version). Which do you think would be most appropriate? Btw., the squirrel on the L4KA web pages is quite cute ;-) - is there a high-res bitmap or vector graphics version of the logo available? My last question concerns L4Linux. The most recent version seems to be based on kernel 2.2.25. Are there any plans to adapt a more recent Linux kernel version, like 2.4 or the upcoming 2.6, as a L4 personality? (I'm aware that this might involve a lot of work...) Best regards, Michael -- Dipl.-Ing. Michael Engel - mengel(a)informatik.uni-marburg.de Philipps-Universit=E4t Marburg - FB Mathematik und Informatik Hans-Meerwein-Str. - 35032 Marburg Tel. 06421 / 28 21562 - Fax: 06421 / 28 21573

4 3

Fiasco 1.1 microkernel released
by Udo A. Steinberg 11 Nov '03

11 Nov '03

Hello L4 hackers, The Fiasco Team is pleased to announce version 1.1 of the Fiasco L4 microkernel developed at the University of Technology, Dresden and released under the GNU General Public License (GPL). This Fiasco release is mostly a bugfix release. Among the new features added in version 1.1 is support for the ARM processor architecture, the integration of the powerful Fiasco kernel debugger (JDB) into the Fiasco-UX usermode kernel and a completely revised "tagged TLB" emulation (also known as small address spaces). For a more verbose list of changes since the last release see the CHANGES file. The Fiasco microkernel currently supports the stable version V.2 and the experimental version X.0 of the L4 interface specification and runs on x86 (i486 and above) and ARM SA1110 CPUs. The included Fiasco-UX port runs on any x86-based Linux system as a normal usermode application. For further information please visit the Fiasco homepage at: http://os.inf.tu-dresden.de/fiasco/ Happy hacking, The Fiasco Team

1 0

RE: l4-ka, c++!
by M. Edward Borasky 08 Nov '03

08 Nov '03

Thanks!! I'm just getting into this L4 / Hurd game. My main concern isn't efficiency -- I leave the speed issues to the folks who give us a 2X clock frequency boost every 18 months. :) My concerns are correctness and stability of the OS, and various project management issues like predictability and stability of dates, interfaces and specifications, production of accurate and usable documentation, etc. I'm by no stretch of the imagination a competent C++ programmer; I can just barely read C++ once the object-oriented features show up. In my trolling about, I discovered that the Fiasco folks have put some effort into "correctness proofs" of Fiasco. See the Vfiasco site http://os.inf.tu-dresden.de/vfiasco/ How difficult would it be to shift this effort from Fiasco to Pistachio? How difficult would it be to port DROPS and friends http://os.inf.tu-dresden.de/drops/ from Fiasco to Pistachio? I'm interested in both tracks -- a real-time operating system and a general-purpose Linux-like desktop and server system. "One OS to rule them all ..." :) -- M. Edward (Ed) Borasky mailto:znmeb@borasky-research.net http://www.borasky-research.net "Suppose that tonight, while you sleep, a miracle happens - you wake up tomorrow with what you have longed for! How will you discover that a miracle happened? How will your loved ones? What will be different? What will you notice? What do you need to explode into tomorrow with grace, power, love, passion and confidence?" -- L. Michael Hall, PhD

1 0

some newbie questions
by cyberspas＠narod.ru 05 Nov '03

05 Nov '03

Hello l4-hackers, I run my small program as root task on top of Pistachio 0.3 microkernel and receive some errors when it run. All works fine, but when in KDB i try to use change verbose level of sigma0 once and again (about 5-6 times) I receive Assertion: Assertion !tcb->queue_state.is_set(queue_state_t::ready) failed in file /home/tosha/pistachio-0.3/kernel/include/api/v4/schedule.h, line 296 (fn=f010099b2) Can you explain me where is the problem ? Code: #include <l4io.h> #include <l4/kip.h> #include <l4/ipc.h> #include <l4/message.h> #include <l4/kdebug.h> #include <l4/thread.h> int main (void) { printf("Myself: %x\n", L4_Myself().raw); L4_Msg_t msg; L4_ThreadId_t tid; tid = L4_Pager(); L4_MsgClear(&msg); L4_MsgAppendWord(&msg, 1); L4_MsgAppendWord(&msg, 9); L4_Set_Label(&msg.tag, -1001<<4); L4_MsgLoad(&msg); L4_Send(tid); printf("aaa\n"); L4_MsgClear(&msg); L4_MsgAppendWord(&msg, 1); L4_MsgAppendWord(&msg, 9); L4_Set_Label(&msg.tag, -1001<<4); L4_MsgLoad(&msg); L4_Send(tid); while(true) { printf("root\n"); L4_KDB_Enter("enter kdb"); } } -- Best regards, Anton mailto:cyberspas@narod.ru

1 0

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 November 2003