users

users@lists.genode.org

2357 discussions

lwIP port to Genode - lwip_recv()
by Amadeo Casas Cuadrado 07 Dec '11

07 Dec '11

Hello, Question about the lwIP port to Genode. Is it possible to receive a large amount of data with lwip_recv()? A quick test shows me the maximum amount of data that is received is 1460 bytes (properly matching TCP_MSS in lwipopts.h). What if I am expecting a larger amount of data? Do I need to perform the defragmentation explicitly in my program? My apologies beforehand for posting to this list if this is a limitation of lwIP stack, and not of its port to Genode. Thanks, Amadeo

2 2

Steps towards an open development process
by Norman Feske 07 Dec '11

07 Dec '11

Hello everyone, over the past four years, the Genode OS Framework has seen rapid progress. Skimming over the release notes of the past releases makes me very proud. It is clear that the project's development is going stronger than ever. However, at the same time, I recognize that progress on a technical level is only one part of a successful Open-Source project. A certainly even more important part is the participation of a diverse community. This is where our project is vastly underdeveloped. By continuing the development at status quo, Genode will continue to steadily improve but it will not be able to capture a significant position in the operating-systems world. Instead, it will possibly remain a curiosity. Therefore, we Genode developers regard the transition of our work to an open and transparent development process as the next big challenge we want to tackle. Review of the situation Even though Genode is an Open-Source project, its development has been pursued largely behind the closed doors of our company Genode Labs. The planning of the road map, most technical discussions, issue tracking, and revision management were used to be done within the company. There had been two reasons for this policy namely the preservation of exclusivity and the efficiency of coordination. Regarding the first reason, when we started our business, we desired to preserve a certain degree of competitive advantage to ourselves by keeping some information "protected" from the public eye. For example, we thought that revealing the detailed history of the over 5000 source-code revisions of the project would enable any outsider or competitor to deeply analyze the way of how our company works. These concerns had been acknowledged by other startup companies with statements in the line of "Open-Source has hurt our business". The second reason is the way of how humans intuitively work together in a non-distributed environment such as a our small company. Instead of discussing technical matters on a mailing list, it appears to be much easier and presumably more efficient to engage in face-to-face conversations. The writeup of our releases notes at regular intervals served us as an instrument to recapture the rationale behind our discussions and document it. This worked exceedingly well. In comparison, discussing every detail on a mailing list seems to be inconvenient and long-winding. Why do we desire a change? The answer to this question can be put quite simply as "to make the project relevant". But it goes deeper than that. First, we see ourself as Free-Software AND Open-Source advocates. I wholeheartedly disagree with statements that suggest that Open Source is incompatible with having a business. For our company, the contrary is true. Without Free and Open-Source Software, there would be no Genode Labs. The great wealth of the GNU software stack forms the basis of all the tools we use every day and it plays a significant role for our passion to develop software. Furthermore, Genode would not be of much value without all the great building blocks in the form of existing Open-Source code that we reuse in our context. Hence, it is our personal desire to contribute to the Free and Open-Source software world and to intensify the collaboration with other projects in the same spirit. We still see the preservation of a certain degree of exclusivity as important for our dual-licensing business model. If we made Genode available under the BSD license, there would be not point in pursuing this model. However, hiding the development process from the public is not only poor-spirited but it creates an artificial barrier for people who want to participate. The book "Producing Open Source Software" (http://producingoss.com) by Karl Fogel was an eye opener to us. Regarding the efficiency of collaboration, I have to admit that the statement above about how great the current way works is really not well-founded - simply because we haven't tried the alternative to discuss everything in public, yet. Obviously, we are risking to spoil our presumed efficiency by changing the mode of collaboration. On the other hand, by documenting the process of solving each problem in the form of public mailing-list postings, we give everyone the chance to contribute to and to learn from our findings. Reaching people outside of Genode Labs is the only way to make our project relevant on a larger scale. The next steps We take the current release cycle as opportunity to execute our plan to open the development process. Our coarse schedule is to migrate our internal issue tracker to the public until mid of December. From this day on, all issues will be tracked publicly. The place for all technical discussions will be the Genode mailing list. Until mid of January, we will prepare a new public source tree in the form of a Git repository that we will then use as mainline development tree. There will be no distinction between our development repository and the public repository anymore. As a further change of policy, we plan to include all documentation as found on the website in the mainline repository and remove the Wiki. This way, contributions to the documentation will be handled in the same fashion as code contributions. Hoping that our designated change is in the interest of all of you, I'm looking forward to see how it will unfold and improve the further progress of Genode. Best regards Norman -- 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

2 1

Announcement: Genode OS Framework version 11.11 released
by Norman Feske 29 Nov '11

29 Nov '11

I am very happy to announce the release of Genode version 11.11. With this version, we explored various opportunities to combine the Genode architecture with virtualization techniques. The topics range from enabling faithful virtualization of x86 PC hardware, over paravirtualized Android, to our custom OS-level and application-level virtualization approaches. The work on these topics materialized in the form of many improvements of the framework. To name a few, the new version features the GDB debugging facility as announced earlier on this mailing list, the integration of the Vancouver virtual machine monitor, the support of running Android based on the work of the L4Android project, and the new ability to execute complex command-line based UNIX software (i.e., VIM) natively on Genode. For regular developers, our new free-standing tool chain significantly improves the situation when developing for different architectures. In particular, this new tool chain alleviates the need for using CodeSourcery tools for ARM and enables the targeting of both x86_32 and x86_64 with a single compiler. The release-notes summary is: * Support for user-level debugging via GDB * Applications and services * Vancouver virtual machine monitor on NOVA * VIM executed on the Noux runtime environment * Terminal emulator implementing the Linux termcap * TCP terminal * Platforms * Fiasco.OC updated to rev 38 * NOVA updated to version 0.4 * New IPC implementation for Linux * Improved support for hybrid Linux/Genode programs * L4Linux on Fiasco.OC * Integration of L4Android (for both x86_32 and ARM) * Stub drivers for block devices, NIC, UART * Device drivers * gPXE device driver environment upgraded to iPXE * Bidirectional UART driver for PL011 and i8250 * PL110 display driver * Support for ARM Versatile Express Cortex-A9x4 * Libc update to FreeBSD 8.2.0 * New unified tool chain based on GCC 4.6.1 As usual, you will find a more elaborative discussion of those topics in our comprehensive release-notes document for version 11.11: http://genode.org/documentation/release-notes/11.11 Have fun with exploring the new version! Norman -- 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

FOSDEM 2012 - Multiserver, microkernel-based operating systems devroom
by Jakub Jermar 22 Nov '11

22 Nov '11

FOSDEM 2012 - Multiserver, microkernel-based operating systems devroom *CALL FOR PARTICIPATION* http://fosdem.org/2012 (apologies for cross-posting) The developers of several free and open-source multiserver and/or microkernel-based operating systems will meet at FOSDEM 2012 in Brussels, Belgium and will share a developer room. The devroom is currently looking for content in the form of talks related to the area of multiserver and/or microkernel-based operating systems. Possible topics include, but are not limited to: OS introduction to developers of other OS Subsystems and architecture Hardware and device drivers Tools and languages Release engineering and testing Experience and learning from mistakes Technical challenges Community and life with an OS-project Academia and education Please send your talk proposal(s) to: microkernel-devroom@...86... no later than by 2011-12-31. Make sure to include the following in your proposal: - title of your talk (will be printed in the FOSDEM booklet) - your full name (will be printed in the FOSDEM booklet) - your short bio giving us some idea of who you are and what is your relation to and qualification for the subject of your talk - number of 15-minute slots that you need for the talk (talks longer than 45 minutes are discouraged) The final devroom schedule (along with accepted talks) will be announced on 2012-01-9 (Jan 9) in: microkernel-devroom@...86... and the speakers will be notified via e-mail. The schedule will be also published on the FOSDEM web and in the FOSDEM booklet after 2012-01-21. If you do not want to or cannot give a talk yourself, you may still send suggestions for what else you would like to see/hear/do in the devroom. Please send your suggestions to the above mailing list. The devroom is scheduled for Sunday of February 5, room K.3.201. The seating capacity of the room is 85 seats. Any changes will be announced in the above mailing list. Jakub Jermar

1 0

Fwd: Multiserver, microkernel-based operating systems devroom at FOSDEM 2012
by Norman Feske 11 Nov '11

11 Nov '11

Hello, I am happy to announce that Genode will participate in FOSDEM 2012 next February. Thanks a lot to Jakub Jermar of the HelenOS project for having invited us to a devroom specifically allocated for microkernel-based and multi-server operating systems such as ours. Please feel hearty invited to join us at FOSDEM and to discuss the planning of the event at the mailing list mentioned in Jakub's official announcement (see below). Cheers Norman -------- Original Message -------- Subject: Multiserver, microkernel-based operating systems devroom at FOSDEM 2012 Date: Fri, 11 Nov 2011 16:57:17 +0100 From: Jakub Jermar The multiserver, microkernel-based operating systems are going to have a dedicated devroom at FOSDEM [0] 2012. The event will take place in Brussels during the weekend of February 4 and 5, 2012. The projects that originally declared interest in the devroom are: GNU Hurd[1] MINIX 3[2] HelenOS[3] Genode[4] If we can get support in the form of devroom content of other related projects, that would make the whole devroom even more worthwile. The reason of having a devroom for interested multiserver and microkernel projects is basically to increase awareness of each other, socialize, talk about the technical problems that our projects have in common and perhaps code together. Or anything else that we can do together. The planning for the devroom is still taking off, but I've already created a mailing list[5]. You are encouraged to subscribe to this mailing list if you would like to participate in the devroom or just come by and say hello to us at FOSDEM 2012. I will send further instructions to that list once the devroom is ready to accept offers of content (e.g. sessions, talks, demos etc.). [0] http://www.fosdem.org [1] http://www.gnu.org/software/hurd/ [2] http://www.minix3.org [3] http://www.helenos.org [4] http://www.genode.org [5] https://lists.fosdem.org/listinfo/microkernel-devroom Please share this message as you think is appropriate. See you in microkernel-devroom@...84... Jakub

1 0

Build-system performance optimization
by Norman Feske 28 Oct '11

28 Oct '11

Hello, since I replaced the HDD of my development machine with a SSD, I experienced quite a boost of build performance. However, at the same time, it became apparent to me that the first stage of our build system consumes too much time considering the simple task it performs (finding inter-library dependencies). This prompted me to spend a relaxing evening to do some profiling, which uncovered a surprising hot spot: The calls of 'echo' for generating the 'var/libdeps' file. As result of my investigation, you can find two optimizations in SVN rev 190, which boosts the performance of the first build stage by factor 3 to 6 (depending on how comprehensive the actual build is) - at least on my machine. The first optimization is the eager detection of already visited libraries to prevent the build system from visiting them again (originally, this detection was performed later when already revisiting the library). This change improves the performance by circa 10%. Second and more importantly, subsequent echo commands are now batched into one command featuring multiple echoes. This way, only one shell is started for executing the whole sequence, which attributes for the significant performance gain mentioned above. This patch makes working with the Genode build system much more enjoyable for me - hopefully for you as well. Cheers Norman -- 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

GDB debugging
by Daniel Waddington 22 Oct '11

22 Oct '11

Hi, I am trying to use GDB. It works fine with test-gdb_monitor, but with my own program I run in to memory quota problems. I tried to fix it through the config file but it did not seem to help. Any clues? Daniel ---- terminal output [init] child "uart_drv" [init] RAM quota: 884736 [init] ELF binary: uart_drv [init] priority: 0 [init] provides service Terminal [init] Warning: Specified quota exceeds available quota. [init] Proceeding with a quota of 103554588 bytes. [init] child "gdb_monitor" [init] RAM quota: 103390748 [init] ELF binary: gdb_monitor [init] priority: 0 [init -> uart_drv] --- i8250 UART driver started --- [init -> gdb_monitor] Starting ldso ... [init] child "timer" announces service "Timer" [init] child "uart_drv" announces service "Terminal" [init -> gdb_monitor] void init_libc_lock_pipe(): init_libc_lock_pipe() [init -> gdb_monitor] using the pipe libc plugin [init -> gdb_monitor] Starting application ... environ: 85004 [init -> gdb_monitor] add_lwp(1, 1, 0) [init -> gdb_monitor] Remote debugging using /dev/terminal [init -> gdb_monitor] Memory model: no memory at address 1009164 [init -> gdb_monitor] Memory model: no memory at address 1009165 [init -> gdb_monitor] Memory model: no memory at address 1009166 [init -> gdb_monitor] Memory model: no memory at address 1009167 [init -> gdb_monitor] Memory model: no memory at address 1009164 [init -> gdb_monitor] Memory model: no memory at address 1009165 [init -> gdb_monitor] Memory model: no memory at address 1009166 [init -> gdb_monitor] Memory model: no memory at address 1009167 [init -> gdb_monitor] Memory model: no memory at address 1009164 [init -> gdb_monitor] Memory model: no memory at address 1009165 [init -> gdb_monitor] Memory model: no memory at address 1009166 [init -> gdb_monitor] Memory model: no memory at address 1009167 [init -> gdb_monitor] linux_resume_one_lwp(step = 0, signal = 0) [init -> gdb_monitor] genode_store_registers() - not yet implemented [init -> gdb_monitor] genode_wait_for_signal_or_gdb_interrupt [init -> gdb_monitor -> thread-migration] Starting ldso ... Quota exceeded: gdb_monitor memory for slab: 4096 used quota: 2175192 ds_size: 2138112 sizeof(Ram_session_component): 216 quota_limit: 3862528 [init -> gdb_monitor] C++ runtime: Genode::Ram_session::Quota_exceeded [init -> gdb_monitor] void* abort(): abort called --------------------------------------------------------------------- CPU 0 [00049d2f]: INT 3 CPU 1 [f0024d6a]: Maskable Interrupt CPU 2 [f0024d6a]: Maskable Interrupt --- config <config verbose="yes"> <parent-provides> <service name="ROM"/> <service name="RAM"/> <service name="IRQ"/> <service name="IO_MEM"/> <service name="IO_PORT"/> <service name="CAP"/> <service name="PD"/> <service name="RM"/> <service name="CPU"/> <service name="LOG"/> <service name="SIGNAL"/> </parent-provides> <default-route> <any-service> <parent/> <any-child/> </any-service> </default-route> <start name="timer"> <resource name="RAM" quantum="512K"/> <provides> <service name="Timer"/> </provides> </start> <start name="uart_drv"> <resource name="RAM" quantum="1M"/> <provides> <service name="Terminal"/> </provides> <config> <policy label="gdb_monitor" uart="1"/> </config> </start> <start name="gdb_monitor"> <resource name="RAM" quantum="1G"/> <config> <target name="thread-migration" /> </config> </start> </config>

3 5

New version of LDSO
by Sam Kerr 18 Oct '11

18 Oct '11

Hi, I have been working on porting some of my dynamic loader work to the newer version of Genode. I thought I would start by compiling the test-ldso program and running it. When I do, I get the following output, but I don't really understand why the errors are there. Can anyone offer guidance? [init -> test-ldso] Starting ldso ... [init -> test-ldso] Starting application ... environ: 83424 [init -> test-ldso] [init -> test-ldso] Static Geekings! [init -> test-ldso] ================ [init -> test-ldso] Test read [init -> test-ldso] no plugin found for read(0) [init -> test-ldso] Static object in funcion ... a: aaaaaaaa b: bbbbbbbb c: cccccccc 6: 00000666 [init -> test-ldso] Shared library function call ...good [init -> test-ldso] Ctor in shared lib ... a: aaaaaaaa b: bbbbbbbb c: cccccccc 6: 00000666 [init -> test-ldso] Exception in shared lib ... good (library) [init -> test-ldso] Cross library exception ... good (cross library) [init -> test-ldso] Exception during RPC: Could not open file "__not_found__" [init -> test-ldso] good [init -> test-ldso] Exception in dynamic binary ... good (binary) [init -> test-ldso] Exception inter-shared library ... good (inter) [init -> test-ldso] Exception from LDSO with explicit catch ... good [init -> test-ldso] Libc test: abs(-10): 10 [init -> test-ldso] ================ [init -> test-ldso] [init] virtual void Genode::Child_policy::exit(int): child exited with exit value 0 -Sam Kerr

2 1

Android on top of Genode
by Stefan Kalkowski 23 Sep '11

23 Sep '11

Dear Genode Community, I'm happy to announce Genode's new facility to drive Google's Android. By porting our changes of L4Linux to the L4Android fork we paved the way to run Android instances in Genode. Beside the support of L4Android, we added a bunch of new virtual drivers for L4Linux + L4Android, that enable the usage of the following Genode device driver interfaces in these para-virtualized Linux clones: * Nic-session * Block-session * Terminal-session The new features are available at our SVN repository. For giving it a quick try, you may have a look at the quickstart section in the updated 'ports-foc/README' document: http://genode.svn.sourceforge.net/viewvc/genode/trunk/ports-foc/README Please be aware that currently we've tested L4Android/Genode on top of the IA32 architecture only. Thanks a lot to Adam Lackorzynski as the main author of L4Linux, and to the whole team around L4Android, as well as to Daniel Waddington for bringing this feature onto our agenda. Best Regards Stefan -- Stefan Kalkowski 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

Error when tried to build manually Hello Demo
by Le Bao TRUONG 22 Sep '11

22 Sep '11

Hi all, I would like build manually the 'Hello Tutorial' demo. But I have an error to execute ./core. I follow the tutorial : 1. Build the server side by make hello/server 2. Build the client side by make drivers/timer make hello/client Then, the tutoriel told me to execute a binary ./core. But I couldn't find core file in my bin/ directory so I build it by make core hello/client Then I execute ./core, my program is stopping at int main(): --- create local services --- int main(): --- start init --- Could not open file "init" ROM module "init" not present int main(): transferred 79 MB to init Would you like give me some advice ? Thanks, Le Bao

2 1

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

users