On 29 June 2017 at 13:35, Philippe Mathieu-Daudé <address@hidden> wrote:
This is true it is better to abort here than risking a deadlock.
However it seems to me they are 3 issues here:
- the deadlock pattern is caused by using a global variable,
- stellaris:ssys_calculate_system_clock() no checking RCC.SYSDIV and
RCC2.SYSDIV2 values <= 2 are reserved (GUEST_ERROR)
- stellaris:ssys_write(RCC2) not overrides correctly RCC
Stellaris works fine. It's other ARMv7M boards (which might forget
to set system_clock_scale) which don't work.
I'd rather take this opportunity to fix the deadlock pattern using a
getter/setter on system_clock_scale, doing the zero check in the setter and
eventually aborting in the getter, what do you think?
We should be using a clock property on the CPU instead of system_clock_scale.
Unfortunately Konrad's series attempting to add that infrastructure
is still in the "trying to sort out the right API in code review"
stage. I don't think it's worth trying to fiddle with the API
for it before we have the right eventual infrastructure in place.
(What system_clock_scale is actually doing is setting the
emulated frequency of the CPU, since that affects the
frequency of the timer.)
thanks
-- PMM