On 04/02/2013 10:56 AM, Alexander Graf wrote: > > On 02.04.2013, at 09:53, Fabien Chouteau wrote: > >> On 04/02/2013 10:09 AM, Alexander Graf wrote: >>> >>> What about the other users of hreset_vector? Can we just remove the >>> variable altogether? >>> >> >> For the CPUs that don't implement MSR[IP], the hreset_prefix is always >> 0x00000000, so hreset_vector will be the reset address (i.e. 0xFFFFFFFC) >> >> I guess all this can be converted to a CPU feature flag, telling if >> reset is an exception or not. > > Hrm, maybe. Not sure which way is cleaner. > > I'll apply your patch for now. It's certainly a step into the right direction. >
OK, thanks, -- Fabien Chouteau
