GSoC proposal, device driver glue code

[Anastassios A. Nanos]

Hi there!
This is my first mail to the list.

I'm interested in submitting an application to GSoC for the Hurd project 
concerning the device driver glue code.
The proposal is @ http://www.cslab.ntua.gr/~ananos/proposal.html

my first try for the abstract produced this:

The GNU/Hurd running on top of GNU/Mach is a free Operating System (free 
as in freedom). The Hurd is a set of "servers" that run over the Mach 
micro-kernel providing support for various tasks needed to complete the 
tasks of a generic OS. It has not been released yet, but it's on a 
growth trajectory - being constantly developed and improved based on a 
stable snapshot. It only supports i386 architecture and a set of 
small-range hardware components. This is due to the way the device 
driver framework is designed. The device drivers used are those of the 
Linux Kernel (versions 1.3 to 2.0) being emulated by means of a 
framework designed in 1996 (and a lot alike ndiswrapper). Glue code has 
been added in GNU/Mach to support various features needed by the Linux 
Kernel and it's drivers. The purpose of this proposal is to update 
gnumach's device support: specifically, I aspire to rewrite and extend 
the glue code residing in GNU/Mach source tree so as to support modern 
hardware with Linux Kernel's device drivers (versions 2.6.x) in the 
bosom of Google's Summer of Code program and the GNU/Hurd project 
developers.

Redesigning the device driver framework of GNU/Mach is an intriguing 
task. However, the project's perspective is to be stable enough to be 
released while being able to support a wide range of common hardware 
used nowadays.

The device driver glue code used currently in GNU/Mach is by and large 
outdated. Using the existing code and design as a starting point, the 
Linux Kernel device drivers as the core and help provided by the 
GNU/hurd project maintainers, I intend to rewrite the glue code needed 
to support updated Linux Kernel's device drivers. The current framework, 
as it stands, necessitates the introduction of the following changes in 
three dimensions. At first, the device drivers have to run unchanged. 
This means that the glue code has to be designed as a layer between the 
real driver code and GNU/Mach's way of interfacing with devices. Next, 
various adjustments have to be introduced so as to implement features 
that the Linux Kernel drivers need to run smoothly. Finally, the code 
has to be fully documented, so as to render the task of maintaining as 
easy as possible.


I would really appreciate some feedback. Sorry if this mail appears as 
spam to you.

Thank you in advance for your time.

A.N.