User or Supervisor

[Gerke Kok]

I found my problem: the boot-loader program did not turn of some of its interrupt sources and so I was left handling them. It looked as if it were crashes or access violations but it were un-catched interrupts. Now the hello-world program is working! Up for the next: "ticker"
Thanks for the suggestions,
Gerke Kok

*********** REPLY SEPARATOR  ***********

On 16/11/01 at 17:06 Gerke Kok wrote:

>Hmm, I'l have to find the place that turns it of then... Thanks, I'll
>look into it. I hope my boot-loader does nog turn it of...
>wkr,
>Gerke
>
>
>*********** REPLY SEPARATOR  ***********
>
>On 16/11/01 at 10:43 Eric Norum wrote:
>
>>Gerke Kok wrote:
>>> 
>>> Hello RTEMS-gen68360 users,
>>> I here have a strange question. Does the CPU-context_switch need to be
>>run in Supervisor-mode or can it be run in user-mode as well. It looks
>>like my system has trouble with this. Is traps to an interupt when the
>>instruction movew d1,sr is executed. This is in the one in the m68k-BSP
>>shared code.
>>> My situation:
>>> I use rtems-4.5.0 and a modified version of the gen68360-BSP
>>> Tool-chain nicely provided by OAR (thanks a LOT!)
>>> The board loads the code (hello.exe) via a propriety TFTP-client into
>>memory starting at 0x408000 and starts executing it.
>>
>>RTEMS is superviser-only.
>>As you discovered, lots of things will break if you try to run in user
>>mode.
>>
>>-- 
>>Eric Norum                                 eric.norum at usask.ca
>>Department of Electrical Engineering       Phone: (306) 966-5394
>>University of Saskatchewan                 FAX:   (306) 966-5407
>>Saskatoon, Canada.