[ECOS] ARM7 stack issue (was ARM vectors.S hang issue)

[Jesper Skov]

>>>>> "Bill" == Bill Diehls <billabloke@yahoo.com> writes:

Bill> Summarizing from the previous email, we are porting eCos to the
Bill> Gameboy Advance.  We are debugging timer interrupt code that
Bill> causes the platform to hang if our user program makes calls to
Bill> some routines such as diag_printf or sprintf. (see thread below)

Just FYI, I had problems with the AEB platform rev C (using a Sharp
ARM7DI core). When enabling the cache, I would occasionally see
corruption of application stack space, IIRC. I've not had time to
investigate so it's presently running with the cache disabled.

There might be a generic bug in the ARM HAL, which is only triggered
when the timing is just right - and that could be the case both on the
AEB rev C and the Gameboy.

Or there could be a HW bug in the CPU cores, of course. Have you
checked for CPU erratas?

Jesper