home *** CD-ROM | disk | FTP | other *** search
- > So there is people using the Brainstorm DSP debugger? Cool. :-)
- > I thought it was more or less useless as it wasn't finished. Atleast that is
-
- Well, I seem to recall using it for debugging my DSP routines as well... ;-)
-
- > Maybe the biggest problem is that the DSP memory is so small?
- > Maybe someone could do a better C-compiler for the 56001?
-
- Better than GCC?
-
- The problem is that another language is needed.
-
- Some reasons why C wouldn't be a good idea:
-
- There's simply no way a compiler could figure out that the strange things
- a C program does to 'unscramble' FFT data can all be handled by switching
- to another addressing mode on the DSP.
-
- It might be possible to recognize circular buffers, but I don't think it
- would be easy (another type of addressing on the DSP).
-
- You'd need a specialized C version to deal reasonably with the fixed point
- numbers. It can be rather tricky to do ordinary integer operations fast on
- the 56k, even if it's possible.
-
- The DSP can not call subroutines to an arbitrary depth without special
- code to remove return addresses from the hardware stack.
-
- > > BTW, do you have a version of DSPDebug with a working send/recieve buffer feature?
-
- I thought I'd used that earlier, but I can't recall getting it to work later.
-
- If the debugger was never released, that might explain things... ;-)
-
- --
- Chalmers University | Why are these | e-mail: rand@cd.chalmers.se
- of Technology | .signatures | johan@rand.thn.htu.se
- | so hard to do | WWW/ftp: rand.thn.htu.se
- Gothenburg, Sweden | well? | (MGIFv5, QLem, BAD MOOD)
-
-