MMC/SD, ARM debugging

[Pavel Pisa]

Hello Karel,

thanks much for response.

On Monday 01 August 2005 11:27, Karel Gardas wrote:
> I'm just poor beginner,

I do not count me between expert as well.
I am ARM greenhorn mostly.

> but isn't it possible to let monitor provide GDB 
> stubs and run RTEMS on top of it w/o GDB support? At least I hope this is
> possible and if you find if your board is supported by either RedBoot ot
> U-Boot you might have some success with RTEMS debugging.

I think, that something like should be possible. It is required to map
monitor code into address space right way (I have solved this)
and ensure and ensure, that FIQ is reenabled in RTEMS and FIQ and unhandled
exceptions vectors are set to same value as in the boot code.
I have already planned to do that. I need to find more time for that.

> I'm writting this since I got an impression that your "target" is to debug
> your RTEMS/SMD/MD card support, if this is completely wrong, please excuse
> me.

Again, thanks very much for reply and interrest.
Basically I need to debug mentioned code but on the other hand,
I would be very happy, if ARM debugging could be integrated into RTEMS.
I would be very happy for advises from more experienced developers there.
I hope, that with advice of some RTEMS matador it should not be big problem,
I hope, that I have some needed background knowledge of GDB, stub and ARM
CPU to do and provide that work.

I have some basic image, that two different kinds of RTEMS debugging
support are available. The first is simple integration of stub debugging
and stopping full system, the second one is integration of debugger
into system such way, that it is able to debug individual RTEMS tasks
and can coexist with RTEMS networking support.
Is that image correct? If so, where is best place to start?
Into which files (file names) should be ARM specific part
of GDB stubs added into RTEMS tree for each of these approaches?

Thanks for any reply in advance

               Pavel Pisa