Ryan Dooley brought up The CITRUS Project as a way to assist internationalization in DragonFly.
According to this benchmark, linked to by Xin LI, there’s file descriptor allocation code in FreeBSD 5 that may be worth the effort to port.
This is the 200th post here – neat! I’m averaging just under 1.5 posts per day, which is good: healthy posting, healthy project.
In a discussion about benchmarks, it was noted that
/etc/malloc.conf changes can help benchmarks tremendously. Rahul Siddharthan suggested ‘
/etc/malloc.conf -> H' and Jeremy Messenger suggested '
/etc/malloc.conf -> aj‘
Also, Matt Dillon made a number of suggestions on what to check when benchmarking DragonFly vs. FreeBSD (4 or 5)
Matt Dillon quote follows:
Continue reading “Benchmark setups”
Simon ‘corecode’ Schubert pointed out prelinking should be mentioned on the 2003 report; here’s the text that will soon show up on that report:
Prelinking capability was added to DragonFly by Simon ‘corecode’ Schubert, which allows faster loading of applications that use a large number of dynamic libraries while running, like Qt/KDE. It is not currently hooked into the system or any port building process.
Matt Dillon fixed an apparently long-term problem in OpenSSH where a server can hang because it has a lot of data to send, but no immediate resources to do it with.
Reproduce it like so:
limit filesize 64k
ssh remotebox -n cat /usr/share/dict/words | cat > junkfile
The IBM ServeRAID controller is now supported, thanks to TONETANI Tomokazu. That would be the “ips” device.
Eirik Nygaard was looking for something to do; Max Laier pointed out removal of
#if defined(__FreeBSD__) / #if __FreeBSD_version > 5 would help, and Jeffrey Hsu indicated backporting the UFS2 size extensions would also be good.
I’ll quote my own followup to say there’s plenty of non-coding tasks available, too.
Port-fixing fiend Joerg Sonnenberger has committed a dfports override for OpenOffice.
David Rhodus imported Hyperthreading changes from FreeBSD which allow you to automatically use Hyperthreading on supported CPUs with just the regular multiprocessor options turned on in your kernel; e.g.
options SMP, options APIC_IO.
However, the DragonFly version has no idling loops in it to reduce CPU resource contention. Because of the way DragonFly schedules per-CPU/sends IPI messages, there’s no performance issue caused by multiple CPUS HLTing. Already, a benefit.
Dave Leimbach has added changes to KDE in CVS to allow kdebase to compile on DragonFly. This is in the actual KDE source code, not a ports override.
The dragonflybsd.org website now contains a FAQ, which is mostly my fault.
I’ve cleaned up my local archive of the DragonFly discussion groups so that start and end dates are correct; they are available at http://www.shiningsilence.com/mailarchive/.
Paul Herman, Senior Researcher of the Kitchen Refrigerator, has a really nice writeup about time and clocks in operating systems (with graphs, even!), and work he wants to bring into DragonFly.
A bunch of changes came in:
hostname now takes a -r option that will set the hostname based on reverse lookup of an IP address, or -i which does the same using the computer’s primary IP. It also works on IPv4 or IPv6. This very good idea comes from Kent Ibbetson.
mixer has had FreeBSD-5 changes added in. It now can take relative volume changes, thanks to Craig Dooley.
Jeroen Ruigrok added in support for the SoundBlaster Audigy and Audigy 2, apparently sourced from “patches by Orlando Bassotto, which were taken from the ALSA Project and the SoundBlaster OSS repository”.