Am 17.06.2013 23:08, schrieb Richard Henderson:
> On 06/16/2013 08:57 AM, Andreas Färber wrote:
>> Since converting first_cpu to CPUState and making CPU_INTERRUPT_HARD
>> available through qom/cpu.h, it no longer depends on CPUSH4State.
>
> Re 25/, how is CPU_INTERRUPT_HARD any less target dependent than
> CPU_INTERRUPT_TGT_EXT_0? Both require knowledge of how the system delivers
> interrupts.
CPU_INTERRUPT_HARD is semantically well-defined for all targets, whereas
CPU_INTERRUPT_TGT_{INT,EXT}_* can mean pretty much anything.
That said, I don't insist on this patch. It compiles the file one time
less than without but still rebuilds whenever I change qom/cpu.h.
Andreas
--
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg
