('x-unknown' encoding is not supported, stored as-is)
(read the subject: this is about Linux on the TI-89/92+)
People have mentioned that a port would be a waste of time and very
difficult because of the lack of an MMU and memory considerations. I
will address both while still being inquisitive about the topic.
AFAIK uClinux is specifically designed for arch's w/out an MMU. Am I
wrong? Anyway, if so, that would be a good place to start.
The TI-89 has a little under 1 MB of total memory (300k in RAM? I
think...) the Ti-92 has about 1 MB in RAM; the 92+ adds additional
archive. I was thinking about having maybe an external memory module on
the linkport, providing maybe 1 or 2 MB. This would be a relatively
simple kernel driver. Idealy, if this becomes a reality, the protocol
should be designed so that the external device could itself have a
linkport, and the kernel driver could encapsulate bytes for it there.
This adds the prospect of even _more_ memory... or calc-to-calc
networking.
But the processor is a limiting factor. Perhaps virtualizing the
processor at the kernel level would be necessary? i.e., doing what
FreeMWare oops Plex86 does, except for the M68k and at the kernel level
where _all_ processes and maybe even some parts of the kernel run in the
virtual environment? Thus basically a simple code pre-scanner would be
needed, and any potential bad instructions breakpointed (e.g., memory
access, so we could do paged memory, etc.) and rewritten when execution
gets there. Process switching can be accomplished by interrupt handlers.
Perhaps once a section of code is scanned it can be cached so that it
does not need to get scanned again? Of course speed would not be
real-time, but I'd sacrifice that for some stability. And Apache on a
graphing calculator is an exciting thought... not really. Now Quake on
the other hand... or xbill... maybe Konqueror... all during math class
;)
Kenneth
__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Tue May 23 2000 - 21:00:14 EST