gl* programming (PV112 subject)
Floating point in kernelspace
You may want to compute floating point in kernelspace of Linux. Linux doesn't
support FP operations in some multiplatform way. For x86 architectures you may
kernel_fpu_begin/end function, but RT people won't like you,
because it disables preempt.
In such a situation you need some workaround. I developped floating point module, used in phantom driver. It uses mmaped area, userspace program read that area and compute FP when there are some requests based on circular buffer (userspace program and makefile also included).
To download this module, use
git clone git://decibel.fi.muni.cz/~xslaby/phantom
and snapshots can be taken from snapshots dir (phantom-*).
Nettest module (NetBSD)
Playing with ether interface under NetBSD kernel. You can download it from its dir. There is also a little script (c).
Whole directory containing some other programming stuff.