[LWN Logo]
[LWN.net]

Sections:
 Main page
 Linux in the news
 Security
 Kernel
 Distributions
 Development
 Commerce
 Announcements
 Back page
All in one big page

See also: last week's Kernel page.

Kernel development


The current development kernel release is 2.2.0pre8. Developers continue to hammer on the last remaining bugs in the hopes of getting a real 2.2.0 release out soon. That release could conceivably happen any day, though there are enough reports of problems floating around still that it may take a little longer yet. See Alan Cox's 2.2 jobs page for a (shrinking) list of things that need to be done.

Alan has also announced that his "ac" patches are done for now. We are close enough to the 2.2.0 release that any important patches should just go straight to Linus; Alan, instead, is heading back to working on 2.0.37.

A suggestion was made that the 2.2.0 kernel needs a press release. Greg Smart jumped in and threw out a first version. Many comments and versions later, Greg has backed off a bit and the current version seems to be maintained by Ed Lang. His current draft can be found on his web site. Linus has remained silent in the discussion, so there is no word as to whether he will actually be interested in putting out this release under his name or not.

A mailing list has been set up for people interested in improving the kernel configuration scheme. It's controlled by Majordomo, and subscriptions may be had at linux-kbuild-request@torque.net.

Anonymous CVS access to vger.rutgers.edu has been turned back on now that the linux-kernel backlog has stabilized and a bunch of Sparc fixes have been merged in.

A great debugging tool has been languishing, mostly because nobody knew that it existed. Check out Michael Chastain's description of his "trace and replay" system. The trace portion captures all of a process's system calls, much like strace except that it remembers all of the results of the call. When the program is run under the replayer, the system calls are intercepted and the old results stuff in again, allowing an exact replay of the traced execution.

The potential capabilities of this system are pretty amazing. A remote user can send a developer a trace of a failed execution; the developer can then replay the entire thing in a far away location and find the problem. Some great memory access debugging can be supported (much needed in the Linux world). He even claims to be able to implement backwards stepping in the debugger. "Imagine a graphical debugger with a scrollbar for time, where the top is 'beginning of execution' and the bottom is 'end of execution'."

This system apparently worked under 1.3.42, and will need a bit of work to function under a recent version of the OS. A significant amount of interest has been shown this time around; with any luck this tool will soon be fixed up and made available.

A new version of the VESA 3.0 kernel BIOS module has been released. This module provides access to the VESA 3.0 features. This is a low-level access module; integration of its capabilities into higher-level code, such as FBcon or X11, has been left as an exercise for the reader. See the announcement for more information.

Software suspend v4 has been released. This patch allows the system to save a copy of its current running state and halt. On next reboot, the system state is reloaded and (hopefully) picks up where it left off. This patch works entirely within the Linux kernel - APM is not used and is not necessary. Check out the software suspend pagefor more information or to download the patch.

International Kernal Patch 2.2.0-pre7.4 has been released, see the announcement for details. (This patch adds cryptographic capability to the kernel, and is destined to remain forever a patch due to crypto law obnoxiousness).

If you want to work with large (> 2GB) files on Linux, and don't have a nice 64-bit system on your desk, you may want to check out the latest version of Matti Aarnio's Large File Summit patch, which has now been updated to 2.2.0pre8.

A review of the Linux memory management subsystem was posted on the freebsd-hackers list and forwarded on to linux-kernel. It's an interesting look at Linux's memory system from an outside point of view; worth a read.


January 21, 1999

Since we're a weekly publication, chances are we'll be behind a rev or two on the kernel release by the time you read this page. Up-to-the-second information can always be found at LinuxHQ.

 

Next: Distributions

 
Eklektix, Inc. Linux powered! Copyright © 1999 Eklektix, Inc., all rights reserved
Linux ® is a registered trademark of Linus Torvalds