NuttX realtime OS ported to run on the Neuros OSD 1.x
Fernando Cassia
this posted here
http://tech.groups.yahoo.com/group/nuttx/message/480
///////////
NuttX v5.9 for the Neuros OSD 1.0 consumer unit builds,
boots, and runs both `nsh' and `ostest' just fine, using
the CodeSourcery 2010q1-188 EABI toolchain on GNU/Linux.
No changes at all were needed (other than the expected
tweaking of the compiler path in `setenv.sh�). Thanks!
///////////
Perhaps with a lighter kernel the problem of running out of memory
will be gone? :)
I still think a Neuros OSD v1.x with 256MB of RAM and an embedded Java
VM would have been great. :)
[If anyone succeeds in replacing the surface mount ram chips to
replace them with a pair of compatible 256MB SDRAM ones let me know
:)]
FC
PS: WRT the memory issue... Is the document below wrong? I fail to see
if it's the development boards which had 64MB vs 32MB on the
production release, or the other way around...
http://wiki.neurostechnology.com/index.php/OSD_Developer_Board_v1
the statements contradict each other:
"Differences between the Developer Samples and the Production Product
8MB Flash and 32MB RAM _instead of 16MB/64MB as in Production_ "
"Development Board Features
DM320 Multimedia SOC with 200 MHz ARM926, 120 MHz C54x DSP
_64 MB of SDRAM_ "
Vladimir Pantelic
> I arrive at these news one year late. But, I don't remember seeing
> this posted here
>
> http://tech.groups.yahoo.com/group/nuttx/message/480
>
> ///////////
> NuttX v5.9 for the Neuros OSD 1.0 consumer unit builds,
> boots, and runs both `nsh' and `ostest' just fine, using
> the CodeSourcery 2010q1-188 EABI toolchain on GNU/Linux.
> No changes at all were needed (other than the expected
> tweaking of the compiler path in `setenv.sh�). Thanks!
> ///////////
>
> Perhaps with a lighter kernel the problem of running out of memory
> will be gone? :)
yeah, just running a kernel and shell hopefully does not need too
much memory :)