LPC214x's BSP for RTEMS

Joel Sherrill joel.sherrill at oarcorp.com
Mon Jun 11 14:04:41 UTC 2007


dave madden wrote:
>>>>>> "RXU" == ray  <xr at trasin.net> writes:
>>>>>>             
>
>     RXU> Hi Joel and I had made some changes in RTEMS CVS.  The
>     RXU> current rtl22xx BSP need 7K bss/data for hello. If you change
>     RXU> the definition for stack size(RTEMS_MINIMUM_STACK_SIZE)from
>     RXU> 4K to 1K. You can also save some memory.  Hope you can put
>     RXU> RTEMS into you lpc214x board.
>
> Hi Ray.
>
> I had some free time, so I copied the rtl22xx BSP and started hacking
> on it.  I got it to boot and prink() some messages, but it dies in
> rtems_termios_initialize with error 5.  I'm not very familiar with GDB
> remote debugging, so it's hard for me to move around efficiently, but
> I think it was trying to allocate a semaphore when it failed.
>   

You have to configure the number of each type of RTEMS object
you want.  You didn't configure any semaphores and probably
didn't link in  the semaphore manager either.

But the bigger question is.. is Ray using the termios console
driver?  I thought he was only using printk() and avoiding the
full featured console device driver and the termios subsystem
entirely.
> Is there a way to get GDB's "next" and "step" commands to use hardware
> breakpoints by default?  That would make my life *much* easier!
> (Can't use normal breakpoints because the code is in flash.  Also,
> there are only 2 hardware breakpoints...talk about tedious!)
>
>   
This is for someone else.
> If you're interested, I can send what I've done.  Mostly, I adjusted
> the linker script and startup code to reflect the 2148 memory map.
>   
When it works, I would like to see another BSP merged.
If the hardware is close enough, you might be able to
share a BSP and use a "variant" to change compile time
options.

--joel
> Regards,
> dave madden
> _______________________________________________
> rtems-users mailing list
> rtems-users at rtems.com
> http://rtems.rtems.org/mailman/listinfo/rtems-users
>   




More information about the users mailing list