Xinu arm port, for Raspberry Pi
C Assembly C++ Other
Pull request Compare This branch is 10 commits ahead, 2 commits behind jetcom:pi-hw2.
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.


This project is to port Xinu to the ARM Versatile baseboard. The ARM Versatile
baseboard has the following devices:

- ARM926E, ARM1136 or Cortex-A8 CPU
- PL190 Vectored Interrupt Controller
- Four PL011 UARTs
- SMC 91c111 Ethernet adapter
- PL110 LCD controller
- PL050 KMI with PS/2 keyboard and mouse.
- PCI host bridge. 
- PCI OHCI USB controller.
- LSI53C895A PCI SCSI Host Bus Adapter with hard disk and CD-ROM devices.
- PL181 MultiMedia Card Interface with SD card.

This image can be run with the following command:

    qemu-system-arm -M versatilepb -m 128M -nographic -cpu arm926 -kernel xinu.bin -s

README: information about the directory tree and how to build an image.
Embedded Xinu, Copyright (C) 2008, 2009, 2010, 2012.  All rights reserved.

$Id: README 2157 2010-01-19 00:40:07Z brylow $

Embedded Xinu
Version: 2.10

1. What is Embedded Xinu?
2. Directory Structure
3. Prerequisites
4. Installation Instructions
	4.1 Build Embedded Xinu
	4.2 Make serial and network connections
	4.3 Enter Common Firmware Environment prompt
	4.4 Set IP address
	4.5 Load image over TFTP
4. Links

1. What is Embedded Xinu?

Xinu ("Xinu is not unix", a recursive acronym) is a UNIX-like operating
system originally developed by Douglas Comer for instructional purposes at
Purdue University in the 1980s.

Embedded Xinu is a reimplementation of the original Xinu operating system
on the Arm processors and is suitable on running on inexpensive development 
Arm platforms, such as the Raspberry-pi.  It is suitable for courses and 
research in the areas of Operating Systems, Hardware Systems, Embedded 
Systems, Networking, and Compilers.

2. Directory Structure

Once you have downloaded and extracted the xinu tarball, you will see a
basic directory structure:

	AUTHORS   device/   lib/     mem/      loader/   README  system/
	compile/  include/  LICENSE  mailbox/  network/  shell/  test/

AUTHORS   a brief history of contributors to the Xinu operating system in
          it's varying iterations.

compile/  contains the Makefile and other necessities for building the
          Xinu system once you have a cross-compiler.

device/   contains directories with source for all device drivers in Xinu.

include/  contains all the header files used by Xinu.

lib/      contains library folders (e.g., libxc/) with a Makefile and 
          source for the library

LICENSE   the license under which this project falls.

loader/   contains assembly files and is where the bootloader will begin
          execution of O/S code.

mailbox/  contains source for the mailbox message queuing system.

mem/      contains source for page-based memory protection.

network/  contains code for the TCP/IP networking stack.

README    this document.

RELEASE   release notes for the current version.

shell/    contains the source for all Xinu shell related functions.

system/   contains the source for all Xinu system functions such as the
          nulluser process (initialize.c) as well as code to set up a C
          environment (startup.S).

test/     contains a number of testcases (which can be run using the shell
          command `testsuite`).

3. Prerequisites

3.1 Supported platform with hardware modification

To run Embedded Xinu you need a supported device or virtual machine.
Currently, Embedded Xinu supports Linksys WRT54GL, Linksys WRT160NL,
and the Qemu-mipsel virtual machine for the MIPS target, and is 
currently being build to run on arm-1176 processors for running on a
Raspberry-pi, and Qemu-arm-1176 virtual machine  For an updated list
of supported platforms, visit:

If you are targetting the MIPS platform, and wish to run on a Linksys 
target, you  you need to perform a hardware modification that will 
expose the serial port that exists on the PCB.  For more information 
on this process, see:

3.2 Cross-compiler

To build Embedded Xinu you will need a cross-compiler from your host
computer's architecture to MIPSEL (little endian MIPS for the 54GL router)
or MIPS (big endian for the 160NL router).  Instructions on how to do this
can be found here:

Update for Arm

3.3 Serial communication software

Any serial communication software will do. The Xinu Console Tools include
a program called tty-connect which can serve the purpose for a UNIX 
environment.  More information about the Xinu Console Tools can be found 

3.4 TFTP server software

A TFTP server will provide the router with the ability to download and run
the compiled Embedded Xinu image.  

4. Installation Instructions

4.1 Build Embedded Xinu

Update the MIPS_ROOT and MIPS_PREFIX variables in compile/mipsVars to 
correctly point to the cross-compiler on your machine.

Then, from the compile directory, simply run make, which should leave you
with a xinu.boot file.  This is the binary image you need to transfer to
your router for it to run Embedded Xinu.  The default build is for the
WRT54GL router; change the compile/Makefile PLATFORM variable for other
builds.  See the compile/platforms directory for supported configurations.

4.2 Make serial and network connections

After creating the xinu.boot image you can connect the router's serial port to
your computer and open up a connection using the following settings:

	- 8 data bits, no parity bit, and 1 stop bit (8N1)
	- 115200 bps

4.3 Enter Common Firmware Environment prompt

With the serial connection open, power on the router and immediately start
sending breaks (Control-C) to the device, if your luck holds you will be
greeted with a CFE prompt.


If the router seems to start booting up, you can powercycle and try again.

4.4 Set IP address

By default, the router will have a static IP address of  If you
need to set a different address for your network, run one of the following

ifconfig eth0 -auto                      if you are using a DHCP server 
ifconfig eth0 -addr=*.*.*.*              for a static IP address

4.5 Load image over TFTP

On a computer that is network accessible from the router, start your TFTP
server and place the xinu.boot image in the root directory that the server
makes available.

Then, on the router type the command:

CFE> boot -elf [TFTP server IP]:xinu.boot

If all has gone correctly the router you will be greeted with the Xinu Shell
(xsh$ ), which means you are now running Embedded Xinu!

4. Links

The Embedded Xinu Wiki
	The home of the Embedded Xinu project

Dr. Brylow's Embedded Xinu Lab Infrastructure Page
	More information about the Embedded Xinu Lab at Marquette University