> On 08 Jun 2015, at 22:27, Peter Maydell <peter.mayd...@linaro.org> wrote:
>
> On 8 June 2015 at 20:18, Liviu Ionescu <i...@livius.net> wrote:
>>
>> aha, in this case the problem is the two step load, not the reset itself,
>> or even more accurate, it is a problem of making the reset calls in the
>> proper order.
>
> That would also fix this problem, yes.
it did, thank you.
my new reset routine looks like:
static void cortexm_reset(void *opaque)
{
qemu_log_function_name();
CortexMState *cm_state = opaque;
ARMCPU *cpu = cm_state->cpu;
#if defined(CONFIG_VERBOSE)
if (verbosity_level >= VERBOSITY_COMMON) {
printf("Cortex-M core reset.\n");
}
#endif
/* Ensure the image is copied into memory before reset fetches msp & pc */
rom_reset(NULL);
cpu_reset(CPU(cpu));
}
I had to make rom_reset() non static in hw/core/loader.c and add the prototype
at the end of hw/loader.h
if you do the same in armv7m_reset() I guess the stm32f205 device will also
work properly.
regards,
Liviu
