Taking from his AHCI work, Matthew Dillon’s working on a Silicon Image 3132 driver. An initial version is available now, though the usual caveats about a brand-new device driver apply.
Update: he’s really moving fast on this.
If you’re one of the few who has seen a ‘no local apic!’ error when booting, Sepherosa Ziehau’s recent commit may have a fix for that. He asks for testers, though he cautions to do it without APIC_IO in your kernel config.
Matthew Dillon has initial support in for port multipliers, along with other AHCI work. It’s not ready for production yet, and he lists the various issues going on, including a need for a different way to mount disks – AHCI changes devicenames from ‘ad’ to ‘da’, which can be a hassle.
Update: hot-swap support, too.
Update update: parallel scans for speed.
The pkg_radd(1) and pkg_search(1) utilities defaulted to pkgbox.dragonflybsd.org. They’ve been switched (by me) to point at avalon.dragonflybsd.org, which has much more bandwidth.
Matthew Dillon’s added AHCI as a kernel module, and has directions for testing. It’s not done, but he has basic hot-plug support in, among other things.
I’ve been posting a lot of “hey test this new technology” items, lately. That’s good. Since I haven’t done it already, here’s a description of AHCI.
Sepherosa Ziehau has added support for various power states on AMD Phenom and Turion-series processors. He has some specific notes that mention there’s more processor family support on the way. Good news for anyone with an AMD-based laptop.
As Hasso Tepper says, please don’t bring in any major changes until after DragonFly’s 2.4 release. This is mostly for the benefit of pkgsrc, so that we can have as complete a working set of packages as possible at release time. 2.4 will probably be in July.
Huawei modems, often available as USB attachments, have been problematic on DragonFly. However, it looks like it’s fixed. (I dont have one to test.) There’s a lot of names involved, so I’ll just point to the commit message.
Sepherosa Ziehau’s recent commit changing how ioapic works may help anyone who has previously had trouble compiling a multiprocessor kernel with IO_APIC enabled. Try it, if that applies to you.
Also, Jordan Gordeev has a potential fix for anyone who has had a failed boot with a ‘BTX Halted’ message; you will have to retrieve it from his Git repo.
Sascha Wildner made two relatively minor commits that solved two long-standing (for me) irritations: a version mismatch in uname for identical versions of DragonFly, and automatic running of newaliases. Both issues have bit me several times in minor but irritating ways over the course of years, and it’s a relief to have them gone. Thanks, Sascha!
Hasso Tepper has added the open source HAL code for ath(4) (old man page), as suggested by Alexander Polakov. I’m not sure if this is related to Dmitry Komissaroff’s work.
Matthew Dillon has added a Makefile in /usr/src/test/vkernel that automates vkernel setup. You can create a virtual system in one step. ‘make help’ in that directory to see all the options.
Aggelos Economopoulos added an interesting feature for virtual kernels: the memory of a given virtual kernel is now accessible directly at /proc/$pid/mem .
Yonetani Tomokazu discovered a permissions problem under Hammer, so Matthew Dillon made a number of commits to fix this and other issues. An update for 2.2 will get them for you, and DragonFly 2.2.2 will be put together very soon so that there’s a release image with these fixed.
Sepherosa Ziehau has added the ability to use High Precision Event Timers (HPET) in DragonFly, based on FreeBSD code. It’s experimental, and he has instructions on how to find if your hardware supports it. It’s apparently a much faster timer than what is used with ACPI, though I do not have details on how that translates into real-world performance.
I had mentioned the new malloc in DragonFly before, but Matthew Dillon has a nice explanation of its performance improvements and the relationship to the existing slab allocator.
Sepherosa Ziehau has added ‘ifpoll’, a feature similar to polling(4). It has to be enabled via a kernel option and so far is only enabled for emx(4); read his commit message for more details.
Hasso Tepper spotted a tty problem that caused a panic in kdesu; Matthew Dillon committed a fix which means the release is delayed until tomorrow. (Thanks, Lazarus, for catching it first)
The 2.2.1 release of DragonFly, rounding up changes since the release (I don’t have a list), should be tomorrow.
Mashing together to make one post:
FreeBSD-SA-09:05.telnet and FreeBSD-SA-09:07.libc have been fixed in DragonFly.
These PC-BSD 7.1 vs. Kubuntu 9.04 Benchmarks are interesting but not that conclusive – different versions of gcc were used. (thanks J. Kanowitz) Here’s a different comparison of performance inside a VM from Ivan Voras.