Category Archives: funded project

Another New Funded Project

Rafal Jaworowski and Semihalf have been awarded a grant to provide FreeBSD with support for the flattened device tree (FDT) technology. This project allows for describing hardware resources of a computer system and their dependencies in a platform-neutral and portable way.

The main consumers of this functionality are embedded systems whose hardware resources assignment cannot be probed or self-discovered.

The FDT idea is inherited from Open Firmware IEEE 1275 device-tree notion (part of the regular Open Firmware implementation), and among other deployments is used as a basis for Power.org's embedded platform reference specification (ePAPR).

"Thanks to this project, embedded FreeBSD platforms will grow in a uniform and extensible way of representing hardware devices, compliant with industry standards (ePAPR, Open Firmware), independent of architecture and platform (portable across ARM, MIPS, PowerPC etc.)," said Rafal Jaworowski, FreeBSD Developer.

Semihalf is a privately owned company, based in Krakow, Poland. They specialize in embedded systems design and development, with expertise in both software and hardware. Among their portfolio are FreeBSD ports to high-end embedded processors (including multi-core) with a wide range of peripheral drivers (storage, networking, pattern matching, security engines etc.); most of this work is publicly available from the FreeBSD repository.

This project will complete by February 2010.

New Console Driver

Ed Schouten has been awarded a grant to write a new console driver for the FreeBSD project. The FreeBSD Foundation is excited to support Ed in providing a more efficient and user-friendly console driver.

This project will allow Ed to add an additional abstraction layer to the kernel. This new terminal layer will sit between the TTY layer, the kernel console and the actual console driver. The existing terminal emulator will be moved into the new terminal layer.

The advantage of the new layer is that the console driver itself will not have to handle any TTY semantics and will just receive a set of character drawing, filling and copying actions. This should make it easier to implement Unicode. It will also be much easier to make the boot process look nice on desktop systems such as PC-BSD).

This project will be completed by the end of December, 2009.