Cédric Le Goater's on July 17, 2019 10:50 pm:
> On 17/07/2019 07:39, Nicholas Piggin wrote:
>> Implement cpu_exec_enter/exit on ppc which calls into new methods of
>> the same name in PPCVirtualHypervisorClass. These are used by spapr
>> to implement these splpar elements, used in subsequent changes.
>>
>> Signed-off-by: Nicholas Piggin <npig...@gmail.com>
>
> This is nice. I am thinking of using these handlers to push/pull
> the XIVE OS CAM line and escalate to QEMU when a vCPU being notified
> is not dispatched.
Great if it is useful.
>> +static void spapr_cpu_exec_enter(PPCVirtualHypervisor *vhyp, PowerPCCPU
>> *cpu)
>> +{
>> + SpaprCpuState *spapr_cpu = spapr_cpu_state(cpu);
>> +
>> + /* These are only called by TCG, KVM maintains dispatch state */
>> +
>> + spapr_cpu->prod = false;
>
> I would have kept this prod bool for the next patch as we don't use it here.
Okay I'll do that.
>> + if (spapr_cpu->vpa_addr) {
>> + CPUState *cs = CPU(cpu);
>> + unsigned int dispatch;
>> +
>> + dispatch = ldl_be_phys(cs->as,
>> + spapr_cpu->vpa_addr + VPA_DISPATCH_COUNTER);
>> + dispatch++;
>> + if ((dispatch & 1) != 0) /* guest set the "wrong" value */
>> + dispatch++;
>
>
> You might want to add :
>
> qemu_log_mask(LOG_GUEST_ERROR, ...);
>
> when the yield value is wrong.
Hm didn't know about that, good point I can add it.
Thanks,
Nick
