Fiasco Kernel extended V2 API vs. Pistachio X2 API
adam at os.inf.tu-dresden.de
Thu Jan 27 14:14:25 CET 2011
On Wed Jan 26, 2011 at 14:09:18 -0800, Daniel Waddington - SISA wrote:
> Can someone briefly explain the basic differences between the current
> Fiasco kernel API and the Pistachio X2 API?
The API of Fiasco.OC cannot be associated with V2 anymore. It's
(completely) different, that's also why we gave it the 'OC' suffix.
The main difference is likely that Fiasco (and L4Re) use objects
throughout the system and use capabilities to address those. For the
kernel objects are, for example, tasks (address spaces) and threads, which
you can create through factories (another object). User-level programs
have (or have not) capabilities on those objects, and they can pass
capabilities to other tasks. Objects can also be implemented in
user-level while the kernel provides communication channels. The whole
system on top is built upon those mechanisms.
Adam adam at os.inf.tu-dresden.de
More information about the l4-hackers