David Rhodus has checked in a new version of BIND, which may correct the recently discovered possible DoS. (commit notes don’t specify.)
Month: November 2003
So, it looks like this:
From an offhand comment Matt Dillon made, the first ‘release’ of Dragonfly is planned at least several months from now. (No definite date yet.) FreeBSD-5 should be entering a stable branch (5.3) around the same time, and the stable 2.6 Linux kernel should be appearing around mid-December, close to the same time as FreeBSD 5.2. KDE 3.2 should also be out in February. Gnome 2.6 is due in March.
So, early 2004 looks like it will have much to play with.
Not much to report, in part because of the U.S. holiday, so here’s a link to a recently-announced OpenBSD variant: ekkoBSD.
Matt Dillon has added in a new ‘libcaps’ library, mostly for experimentation, which appears to be the base for userland threading.
Probably in part because of a long-running off-topic SCO thread in dragonfly.kernel, ‘trent’ suggested an advocacy or general group. Matt Dillon said “in a few months, just prior to the first release”.
Emiel Kollof plans to change the Heimdal utilities to a format more similar to other platforms, unless someone else wants it.
In a conversation about updating
bind, several people noted that changes other than vendor updates to
contrib can wait until there’s a complete packaging system in place. (i.e. things like bind and sendmail are staying in place, for now.)
There’s been a number of people reporting various breakages during install or boot time. To weed out problems caused by old data right away, remember to:
rm -rf /usr/obj
Matt Dillon added an “upcall mechanism to support userland LWKT”. You can look at the man page for upc_register in source.
In a followup to the K42 post, Max Laier pointed at the L4Ka project, which is somewhat similar in scope.
Michal Ostrowski, a developer with the K42 project at IBM Research, posted to dragonfly.kernel and noted that DragonFly and K42 are very similar in design and could benefit from interaction. He brought up several papers located on their site: a K42 overview, K42 and traditional UNIX APIs, and K42’s threading and scheduling infrastructure. The closely-related Tornado operating system also has some good information.
Jeroen Ruigrok pointed out this Newsforge article that says SCO will probably be bringing the AT&T settlement over BSD into court. According to people’s opinions, however, this is a last legal flailing, rather than anything that will significantly affect BSD-derived systems like DragonFly.
Daemonnews also has a mention of this, where Marc Rassbach points out the article’s author has a ‘poison pen’ history vs. BSD products, and Peter Hansteen notes a Forbes article on SCO, this Byte article, and the ever-bearded Greg Lehey’s excellent appraisal of this mess.
I’ve corrected my how-to-install writeup, based on what several people said in dragonfly.kernel. Several steps are now removed, but it still has the same flavor.
According to Matt Dillon, DragonFly will skip the multiple branch style of FreeBSD (STABLE, CURRENT branches) and tag the single main branch, slipping as needed for security fixes.
Aaaand it’s done – rebuilding your system with current sources will give you a
uname that reports “DragonFly”. Through some trickery, most ports are apparently not broken by this.
Matt Dillon’s bringing in the ‘DragonFly’ name to replace ‘FreeBSD’ in the source, which may break a number of things over the next few days, including all ports.
Joshua Coombs asked about good books for BSD kernel/network programming. Jeroen Ruigrok listed: “The Design and Implementation of 4.4BSD”, “Unix Internals: The New Frontiers”, and “The Design of the Unix Operating System”.
From Joshua Coombs:
The make release step no longer depends on perl, thanks to Jeroen Ruigrok.
Yet more __P() macros have been removed. Check the commits record around the 13th to see. __P() should now be gone most everywhere except
We had some power outages/surges here because of a windstorm, and my UPS didn’t handle it well, along with other local network equipment. So, I’ll be slow with news posts until I get my internal network in better working order.
Matt Dillon noted he would be creating a port override for gcc33, as the port from FreeBSD’s port tree doesn’t seem to work just right.
I’ll quote Matt Dillon’s entry cause I’m working late:
” The MBWTest program (/tmp/mbw1) attempts to figure out the L1 and L2 cache sizes and measures L1, L2, and non-cached linear memory bandwidth.”
Joshua Coombs is kicking around the idea of modeling – or even implementing – his new routing model, in Perl.
This week Matt Dillon is doing:
- lwkt_token and IPI code optimization
- GCC 3.x (just for support of the next item)
- 64 bit AMD64 support
For those of you late to the party and wondering why his work schedule is spotlighted, Matt Dillon is the originator of the DragonFly project, and is doing much heavy lifting.
Not directly about DragonFly, but chances are good you are using it. Release 4.4 of XFree86 is due by the end of the month, and you can look at the changelog.
Matt Dillon’s added boot code from FreeBSD 5 – this allows AMD64 and ELF64 support. He also pushed in new linker code and some (not yet enabled) support for UFS2.
Use installkernel and installworld as part of your build process, and you should be fine with these changes. However, you will manually have to copy
There’s been a lot of new code lately – that’s good!
Variant symlinks are possible now, though you currently have to set the sysctl vfs.varsym_enable. ‘varsym’ can be used to mess with them at any time, however.
Joshua Coombs has updated his writeup of a new routing model, based on feedback.
Matt Dillon’s reorganizing some of the header files; if you build a new kernel anytime soon, make sure you build from scratch using ‘config -r’, as some of the old header files have now vanished.
Congratulations are due to committer Jeroen Ruigrok, who is apparently getting hitched.
Changes to systat by Kenneth Culver gives it the ability to show each network interface and its throughput.
Jeffrey Hsu and Matt Dillon’s network changes are being committed – the first third is in, according to a commit by Matt.
Matt describes the plan thusly:
“Basically the goal of this work is to isolate and serialize PCBs in specific threads in order to (A) not have to lock them and (B) improve cache locality for ISR processing loops as well as for data. Isolating a network PCB means dealing with the points where the PCB talks to other parts of the system. There are three points where this happens:
- incoming packets go through preprocessing (e.g. IP) before
being routed to the target protocol & PCB (e.g. TCP and UDP).
- user syscalls operate on PCBs
- timers and such initiate work related to particular PCBs”
I wish I knew what a PCB was.
Matt Dillon posted an interesting bit about what’s needed/planned for non-emulated use of the AMD64:
Matt Dillon’s putting in some material from FreeBSD-5, so if PNPBIOS is defined, you may not produce a working kernel during the next few days.
update: PNPBIOS should not be enabled if you have a AMD64 machine, and the K8V motherboard, as that appears to be broken.
Jeroen Ruigrok is setting up Bugzilla for tracking bugs/requests for DragonFly. As part of the process, there’s some 6,000 (!) items brought over from FreeBSD-4. No link yet…
The 3c940 ethernet driver (found on ASUS K8V motherboards) is now supported. It’s the ‘sk’ device. Matt Dillon’s AMD64 machine apparently arrived.
I was pointed at the gmane.org site to find old dragonfly.kernel postings. I moved in what was there, and so the local kernel archive has several months of history added in, now. Docs doesn’t seem to be working yet…
David Rhodus has been commiting a whole bunch of fixes from FreeBSD today. You can tell just by looking at the mailing list archive for commits@, freshly set up, and updated hourly along with the other archives.
Anyone have old list traffic sitting around, perhaps in mbox format? I’d like to backfill.
send-pr, the strange commandline utility for making a bug report, now goes to the bugs mailing list.
Byron Schlemmer put up Matt Dillon’s slides from his Berkley talk as HTML pages.
Eirik Nygaard has also cleaned out all the __P() in usr.bin and usr.sbin, and incidentally gained a commit bit. Congrabulations.
In a thread about booting media, Matt Dillon noted he was going to:
- Look at Jeff Hsu’s TCP thread code
- implement variant symlinks
- start VFS messaging and environments
VFS will make a number of other things – especially a new port system – possible.
James Frazer found, and David Rhodus corroborated, that NTFS support tain’t working.
till noted in a comment here that www.dragonflybsd.de is live, and is using the RDF feed from this site.