Alan Cox released the 2.2.21-pre3 kernel patch today, building toward a targeted final release of the stable 2.2.21 kernel on March 10th. The changelog for this current -pre patch includes a fairly large set of fixes and updates, catching up with a backlog. Alan goes on to point out that -pre4 will include several more important fixes.

Alan's announcement email follows, including the full changelog.

Zack Brown's Kernel Traffic #155 for February 25th is online.

Robert Love posted a small patch to the lkml, providing an i386 only IRQ-safe atomic counter without locks. The conversation that followed between Robert and Andrew Morton is quite interesting, offering a little perspective into the complexities of dealing with multiple processors and the many Linux ports.

Robert Watson, of the core FreeBSD team, recently sent a message to the 'Current' FreeBSD mailing list, looking to initiate a constructive discussion to develop guidelines for the use of source control software beyond the main CVS repository. The end goal to create "a set of recommendations to maximize communication and acceptance by the broader community".

This is in response to recent lengthy discussion of the many source code repositories used, and the lack/difficulty of communication between all involved as to what's where. Robert's full email follows.

Jeff Garzik recently posted a BitKeeper Kernel Hacking HOWTO (titled Doing the BK Thing, Penguin-Style) to the lkml. Now that Linus is using BitKeeper to manage the 2.5 development kernel tree, many other Linux kernel hackers are beginning to use the tool. The HOWTO, a work in progress, provides an excellent overview of BitKeeper, helping a person use the tool to provide patches to Linus. The actual user's guide can be found here.

GCC 3.0.4, the last of the 3.0 series (the next release will be 3.1, around April 15) has been released. It's a "bugfix release" that (you guessed it) has many bugfixes. Read on for the full announcement.

New in 3.0.4:

• GCC 3.0 now supports newer versions of the NetBSD operating system, which use the ELF object file format, on x86 processors.
• Correct debugging information is generated from functions that have lines from multiple files (e.g. yacc output).
• A fix for whitespace handling in the -traditional preprocessor, which can affect Fortran.
• Fixes to the exception handling runtime.
• More fixes for bad code generation in C++.
• A fix for shared library generation under AIX 4.3.
• Documentation updates.
• Port of GCC to Tensilica's Xtensa processor contributed.
• A fix for compiling the PPC Linux kernel (FAT fs wouldn't link).

Marcelo released 2.4.18-rc3 last night, saying, "So here goes rc3: hopefully the last. I'll only release another -rc in case of really bad problems". This morning 2.4.18-rc4 was released, Marcelo commenting, "Unfortunately something really bad (for some non-x86 archs) [showed] up, so here goes rc4". The changelog for -rc4 described the fix by Tom Gall as "Load code did not set personality for binaries without an interpreter: This was breaking static apps on several archs". Certainly nice to have caught the bug in a release candidate, and not the actual release.

Find the -rc patches here, and the full changelog here.

A recent conversation on the FreeBSD hackers mailing list discussed the possibility of remote kernel debugging, utilizing the polling support of some Ethernet drivers. This is implemented on other systems already, including, it was pointed out, on Apple's Darwin.

The ensuing conversation on the FreeBSD hackers list follows.

LWN.net has posted this week's kernel update, always an interesting read.

This week, KernelTrap has been speaking with Linux kernel hacker Rik van Riel. Rik is most recognized for his impressive rmap VM efforts, available as a patch for the 2.4 kernel here. He's also the founder of kernelnewbies. Living in Brazil, he works for Conectiva.

LVM 1.0.3 is out now.

LVM 1.0.3 supports both version 1 and 2 of the metadata.

There is "*no* need to run any metadata update tools".

Full announcement here.

Guillaume Boissiere announced that a new update on 2.5 kernel status is available on kernelnewbies, following the release of development kernel 2.5.5.

His latest status page includes two new merges:

 o in 2.5.4+   Porting all input devices over to input API  
               (Vojtech Pavlik, James Simmons)
 o in 2.5.5    Add ALSA (Advanced Linux Sound Architecture) 
               (ALSA team)

Murray Stokely announced today that DHCP 3.0.1 RC6 has been imported into the -CURRENT FreeBSD tree. He says, "DHCP 3.x provides DHCP failover support, dynamic DNS updates, and much more".

When the v2.4.0 Linux Kernel was released in January of 2001, it had a new VM designed by Rik van Reil. Nine months later, with the release of 2.4.10, Linus shockingly ripped it out, replacing it with a VM provided by Andrea Arcangeli. Though an impressive feat, to this day the VM is still a sore point with many Linux users. Fortunately, with recent releases (especially in the -aa branch, where Andrea does his testing) the current VM is rapidly stabilizing.

Rik, however, continues to develop his VM, with a strong following. (Alan Cox, for example, continues to include Rik's VM in his branch, as does Red Hat.) The rmap VM is currently up to version 12, with version 11 and beyond being labled "ready for use" by Rik.

Louis Garcia posted some questions to the lkml, wanting to learn more. What follows are his questions and Rik's answers. Ultimately, it appears, much of Rik's efforts are targeted for 2.5 inclusion.

The OpenBeOS team recently had their first internal release. It is described on their project news page with the quote, "Baby steps along the path to glory..."

This first release is intended to be installed on top of an existing 5.03 installation, containing "the most up-to-date versions of OpenTracker and the Mail Daemon Replacement, and several replacement Preferences apps (Keyboard, Menu, Mouse, Screen, VirtualMemory, and Workspaces)."

The release produced a little friction among some of the OpenBeOS mailing list members, as the announcement was found on external websites before being posted internally on the mailing list. A brief sampling of th exchange follows, focusing mainly on project leader Michael Phipps' explanation.

Their project page describes OpenBeOS:

"OpenBeOS is a project dedicated to the re-creation, followed by the extension, of the BeOS."

"Individual servers a