December 13, 2012 posted by Antti Kantee
Some years ago I wrote about the possibility to load and use
standard NetBSD kernel modules in rump kernels on i386 and amd64.
With the recent developments in buildrump.sh and the improved
ability to host rump kernels on non-NetBSD platforms, I decided to try
loading a binary NetBSD kernel module into a rump kernel compiled for
and running on Linux. The hypothesis was that the NetBSD kernel modules
should just work since both the NetBSD kernel and Linux processes use
the ELF calling convention, and all platform details are abstracted by
the rump kernel hypercall layer. Sure enough, after two small fixes to
the hypervisor I could mount and access a FFS file system on Linux by
using ffs.kmod as the driver.
December 06, 2012 posted by S.P.Zeidler
The machine normally running www.NetBSD.org and also gnats and mail-index.NetBSD.org had a hardware problem.
It is now working again on a new chassis.
November 07, 2012 posted by Antti Kantee
The unique anykernel capability of NetBSD allows the creation of
rump kernels, which are
partially paravirtualized kernels running on top of a high-level
hypervisor. This technology e.g. enables running the
same file system driver in the monolithic kernel or as a
microkernel style server in userspace. POSIX-compatible
systems have been more or less supported as rump kernel hypervisors for
the past 5 years. A long-time goal has been to extend hypervisor
support further, for example to embedded systems. This would bring the
solid driverbase of NetBSD available to such systems with only the cost of
implementing the hypervisor.
To see how far things can go, last week I started toying with the
manually implement the hypervisor. After some
emscripten, which translates C into
mature, but there is also extensive support for the POSIX API. This meant that
could also compile the existing POSIX hypervisor and have it work.
tree. This contrasts the approach taken by
another similar experiment,
where an x86 Linux is run inside a x86 machine emulator running
I have thrown together a small proof-of-concept demo of how to build a
web service with the capability to access file system images using
kernel with support for the FFS, tmpfs and kernfs file systems. This
rump kernel backend is tied to a lightweight web page which passes
requests from forms to the rump kernel and displays results. When the
bootstraps a rump kernel, and mounts the FFS image r/o at /ffs.
The status can be further manipulated with interactive commands.
The demo is available
I've tested it to work with Firefox and tested it to not work
with Internet Explorer. YMMV with other browsers. Note,
size, so the page may load for a few moments over a slow link --
whitespace removal was the only size reduction technique I used.
sources, you can also look at the
unoptimized version (14MB).
August 07, 2012 posted by Julio Merino
If you are like me and do not like repeating manual tasks over and over again, you probably have written custom scripts to rebuild NetBSD from source and to upgrade your NetBSD installation to a new build. I have had such scripts for years, but they have always been very rudimentary, inflexible and unreliable. What's more: every NetBSD machine I have today has its own custom versions of these scripts, often with little incompatible tweaks to accomodate every installation.
I was now facing the time to rebuild my NetBSD development box (because its VM disk is running out of space), and the thought of having to reconfigure my environment to perform periodic NetBSD builds, and to have to continue using my deficient custom upgrade script gave me the energy to sit down and write decent versions of these two tools. The results of such efforts are sysbuild and sysupgrade, and their goal is to trivialize rebuilds and upgrades.
sysbuild, currently available in pkgsrc/sysutils/sysbuild, automates the process of fetching the NetBSD sources from a CVS repository and performing a build for one or more platforms. The process is defined by a configuration file that controls the parameters of the build, and building a new snapshot from scratch is as easy as issuing one parameter-less command. A companion sysbuild-user package configures an unprivileged system user to perform nightly builds of NetBSD from cron, which is very convenient for developers tracking NetBSD-current. My other post "Introducing sysbuild for NetBSD" and the sysbuild(1) manual page provide more details.
sysupgrade, currently available in pkgsrc/sysutils/sysupgrade, automates the process of bringing a NetBSD system up to date by using a collection of distribution sets (which can come from either sysbuild, your manual builds or a public FTP site). This tool does not have any upgrade knowledge of its own: it is "just" a script for the manual procedure of fetching the distribution sets, upgrading the kernel, unpacking the sets and upgrading /etc with the standard etcupdate and postinstall. My other post "Introducing sysupgrade for NetBSD" and the sysupgrade(8) manual page provide more details.
The tools are released under a typical permissive 3-clause BSD license. I have opted to place them in pkgsrc instead of the base system to give the tools some mileage and to make them immediately available to older releases and operating systems. However, my end goal for these tools (specially sysupgrade) is to end up being part of the NetBSD base system: I find it unfortunate that the base system does not provide a mechanism to upgrade itself to a newer release.
Enjoy the tools and please share your feedback!
... just keep in mind that these are fresh new code so obscure bugs may be lingering around. Hint: Run the included test suites to get some confidence in the implementation before running the tools.
May 28, 2012 posted by Mike M. Volokhov
After weekend, on Monday, I usually warming myself up. It's always good to read an interesting article or meet eminent people to set a pitch for the week. This week such a reading was from Billy Toulas of the unixmen.com
, popular Unix/Linux online magazine who just interviewed Martin Husemann
Martin is one of those people with whom the NetBSD Project is strongly associated. He's former member of the NetBSD Board, and today is acting as sparc64 Portmaster and as part of NetBSD Security and Release Engineering teams. But don't be confused with all those loud names — Martin is very friendly man who fantastically can explain complex things.
The reading is interesting and long, and contains a number of amusing little know facts, so take a big popcorn pack, or start build.sh over NetBSD 6.0 BETA2 (just as I did) and enjoy the interview.
March 18, 2012 posted by Benny Siegert
For informing our peers about NetBSD, for distributing at a conference or meet-up, or simply as a concise introduction – new NetBSD flyers are now available at http://www.NetBSD.org/gallery/advocacy/. They are made to be printed on A4 paper and folded twice. English and German versions are provided.
At the moment, there are three flyers: "What is NetBSD?" gives an overview about the project and the OS, while "Tips and Tricks for NetBSD newcomers" and "NetBSD for Newcomers – Switching from GNU/Linux" are more aimed at people who already know similar operating systems, particularly GNU/Linux, and want to know more about the differences between these systems and NetBSD.
We are very grateful to Julian Fagir, who provided these flyers to the project.
December 15, 2011 posted by Reinoud Zandijk
Article about installing NetBSD on a small nas device LG NAS-N2R1D. This requires some soldering and other black magic but otherwise pretty good example of NetBSD portability.
August 03, 2011 posted by Jörg Sonnenberger
The NetBSD CVS repository has seen a lot of work over the recent months to clean up various historic artefacts. Together with the improvements in cvs2fossil and the Fossil->Git conversion, it is now possible to provide a precise, up-to-date conversion.
The Git repositories can be found on github, checkout pkgsrc and src.
The Fossil repositories can be found at src (mirror) and pkgsrc (mirror). The raw database as faster alternative to cloning can be found on ftp.NetBSD.org.
This repositories are synchronised with a normal latency of 2h. The only differences between a checkout from CVS and Fossil/Git are supposed to be related to historic checkouts of files pulled from a vendor branch. The implemented behaviour is consistent with the changes in the file. RCS IDs are all expanded using the same rules CVS follows.
July 31, 2011 posted by Marc Balmer
More and more modern computers don't come with a "real" keyboard interface anymore, and that makes our lives hard when we have to enter the kernel debugger DDB. Well, not anymore if you run -current. jmcneill@ committed code which lets you use DDB with USB keyboards. See
for the full commit message.
April 25, 2011 posted by S.P.Zeidler
The following projects have been chosen for Google Summer of Code™ this year (sorted by student's last name):
- Implementing the algorithms described in RFC 6056 within the NetBSD
- Student: Vlad Balan
- Mentor: Christos Zoulas
- Socket option to timestamp UDP packets in NetBSD kernel
- Student: Dmitry Cherkassov
- Mentor: Marc Balmer
- pkgsrc cross-building: connecting the dots
- Student: Efstathios Kamperis
- Mentor: Aleksej Saushev
- Add kqueue support to GIO
- Student: Dmitry Matveev
- Mentor: Julio Merino
- Userspace file system and device driver code sharing
- Student: Vyacheslav Matyushin
- Mentor: Alistair Crooks
- IPv6 support for NPF firewall
- Student: Zoltan Arnold Nagy
- Mentor: Martin Husemann, Mindaugas Rasiukevicius
- Add support for FreeBSD package format, RPM and Debian packages(if time allows) to pkgsrc
- Student: Anton Panev
- Mentor: Jeremy C. Reed
- Apropos replacement based on mandoc and SQLite's FTS
- Student: Abhinav Upadhyay
- Mentor: Jörg Sonnenberger
- In-kernel implementation of posix_spawn
- Student: Charles Zhang
- Mentor: Martin Husemann
March 30, 2011 posted by Antti Kantee
The NetBSD core team has announced
a tier system for the hardware architectures supported by NetBSD.
The tier system classifies ports into three tiers.
Summarizing, the tiers consist of ports that NetBSD will support, ports that NetBSD does its best to support,
and ports which may be desupported soon.
The purpose of this classification is to clarify the development roles between old and new architectures.
By making modern architectures an official development priority, the tier system
ensures that NetBSD will continue to be the top choice for a fast, secure and portable free OS on the hardware of today and tomorrow.
March 15, 2011 posted by Antti Kantee
Rump is a componentization of the NetBSD kernel. It lends itself
to multiple uses, such as running kernel code as services in
userspace and for example makes the high-quality NetBSD kernel code
base available for use in multiserver microkernel operating systems.
Running unmodified NetBSD kernel code in standalone userspace
applications has been possible for years. Recently, it also became
possible to use unmodified userland binaries as remote clients for
these lightweight and modular kernel server instances. Things work
straight out of the default NetBSD installation. For example, it
is possible to run an unmodified web browser against a rump TCP/IP
server and restart the TCP/IP server with minimal impact to the
browser. Furthermore, it is possible to run a dedicated TCP/IP
server for every networking application on the system. The combined
flexibility of using proven kernel drivers in lightweight virtual
servers is completely unique to NetBSD.