I committed the stuff from the sound project just a few minutes ago. So the SoC 2006 has officially ended for me (I just wait for the T-Shirt now… ).
This doesn’t mean I don’t care about the stuff anymore. I will commit fixes in case problems show up and I’m also responsible in case my ex-mentees have questions or patches. It’s just that the official part is done now.
Happy bughunting to all.
I committed most of Romans work in the linuxolator to current. The new syscalls aren’t used until you run
to switch back (after exiting all linux programs) you just have to run
But you have to do this on i386. Amd64 support is not complete (and besides this, amd64 is still broken and nobody provided the neccessary debugging info to jhb@).
There are some known problems with osrelease=2.6.16, e.g., problems with futexes (visible in acroread, realplay and skype), but some programs already run without obvious problems (linux-firefox, linux-opera).
Any reports about new problems to netchild@ and rdivacky@ please. Reviews, debugging info and patches are welcome too.
I just removed the COMPAT_43 option from the GENERIC kernel in current. This may result in increased performance for some workloads.
In the last days I also “mentored” a little bit my SoC students. Reviewing some changes, suggesting some imrovements, committing some stuff which is ready, discussing various things and so on.
And last but not least, I hope that the last bugs in the update to the new linux base port are ironed out on the ports build cluster (I did some commits in the last days).
The last week has seen some bikesheds. One of them was my commit of the doxygen infrastructure for the kernel subsystems. Some people don’t like the way doxygen requires some markup tags in the comments, some people don’t think such API docs provide additional value and some people fear that 3rd party developers may use some functions which shouldn’t be used. I don’t repeat the counter-arguments of myself and other people here, but there are people out there which already make use of the current unsatisfactory doxygen output and are happy about this infrastructure. Luckily is was superseeded by another bikeshed (and gnn@ wants to work on documenting a subsystem to show the benefits to those people which do not think yet, that this is a good idea). On a related issue, I’m waiting on a repo copy of src/sys/doc to src/tools (it’s one of two repo copies I’m waiting for, ncvs@ seems to be bussy ATM). Some doc@ people think it is more appropriate there.
The FC4 linux base port and the xorg based linux X11 libs port are scheduled for testing in an experimental ports build run, we may see the switch of the default linux base port in the not so distant future. It seems Boris is working on updates to the rest of the linuxolator infrastructure in the Ports Collection (gtk, …), so we may see a lot of updates there after the switch of the default linux base port.
In the last days I also helped/talked with my SoC students. Roman is playing a little bit with an amd64 tinderbox he got access to and as a result he committed support for building the linuxolator on amd64 as a module to perforce (call for testers: he did send a patch to emulation@, please give it a try if you own an amd64 box). Ryan is cataloging the IOCTL’s and their status (implemented, obsolete, …) in the FreeBSD wiki. I already priorized those he did so far, and gave some suggestions how to proceed with the important ones. This way he hasn’t to wait for me or Ariff when he is finished with the cataloging (being a mentor living in a different time zone means you should be ahead of your student… being ahead even before he is able to asks questions is … a boost for your own ego 😉 ).