My DSP hints
contain information on using DSP compiler tools under WINE (target systems
are DSK6211 and DSK6711, and also the DSP56303).
People experimenting with the C6x DSKs may be interested in:
The beagle board contains an
OMAP3530 processor with superscalar ARM Cortex-A8 and TMS320C64x+ DSP
components (with embedded linux port).
Mity DSP have announced the MityDSP-Pro which integrates the TI
C6455 DSP with a Xilinx XC3S2000 Spartan 3 FPGA, flash memory, SDRAM, and
Serial Rapid I/O links (also available with lower-cost TMS320C6454
processor and/or with larger FPGAs);
A TMS320C3x emulator is
available from the Embedded Signal Processing Lab of the Electrical and
Computer Engineering Department at the University of Texas at Austin
(project titled WETICS).
SciLAB is a comprehensive matrix mathematics package developed by
France's INRIA, available from the SciLAB Web Site.
The Scicos real-time system
design tools allow SciLab based design of models and controllers to be
implemented in real-time within an RTAI framework - see also Modeling and Simulation in
Scilab/Scicos, Stephen L. Campbell, Jean-Philippe Chancelier, and
Ramine Nikoukhah (ISBN 0-387-27802-8).
Octave is another very powerful matrix mathematics package
developed by John W. Eaton, Department of Chemical Engineering, University
of Texas, and which is available from the OCTAVE Web Site.
RLaB is a matrix maths tool for signal processing or other typical
"EE" work, created by Ian Searle from Boeing --- it is available from the
RLaB Web Site where source code
and contributed material is available. The local site http://mirriwinni.it.jcu.edu.au/~ftp/rlab
also provides RLaB source/binary distributions. You can also obtain the
RLaB Primer that Ian Searle and I wrote at this site (pdf, 300dpi
ps, 600dpi
ps). RLaB is no longer under active development but ...
New! RLaB lives on, in the form of RLaB Plus,
being developed by Marijan Kostrun.
FFTW = fast platform-independent portable (!) FFT from a team at
MIT. It is available at www.fftw.org.
Some assembler tools are also available as options in microprocessor
cross assemblers kept in http://mirriwinni.it.jcu.edu.au/~ftp/uP
(the Arnold cross assembler suite handles a number of DSP devices).
Directory matlab contains
the latex source code for the matlab primer, a recent matcom-to-C++
announcement, and some old matlab 3.5 audio IO C mex files that I wrote in
a previous life (a copy of the original communications toolbox which used
these routines is available but newer versions should be retrieved from The Mathswork's WWW site).
R.I.P. - the original Mathtools folks did some interesting early
work on a matlab-to-C compiler. They sold their technology to Mathworks
who now operate URL http://www.mathtools.net