Re: [PATCH 1/3] kvm-s390: Implement the directed yield (diag 9c) hypervisor call for KVM

[Christian Borntraeger <borntraeger@de.ibm.com>]

On 24/04/12 14:04, Avi Kivity wrote:
[...]

>> +	int tid;
>> +	int i;
>> +
>> +	tid = vcpu->run->s.regs.gprs[(vcpu->arch.sie_block->ipa & 0xf0) >> 4];
>> +	vcpu->stat.diagnose_9c++;
>> +	VCPU_EVENT(vcpu, 5, "diag time slice end directed to %d", tid);
>> +
>> +	if (tid == vcpu->vcpu_id)
>> +		return 0;
>> +
>> +	kvm_for_each_vcpu(i, tcpu, kvm) {
>> +		if (tcpu->vcpu_id == tid) {
> 
> Can two vcpus match the same tid?

Dont think so. tid (target id) is the cpu number of the CPU holding the
spinlock and vcpu_id is checked for in kvm_vcpu_init to be unique.

[...]
>> +			vcpu_put(vcpu);
>> +			if (yield_to(task, 1)) {
>> +				vcpu_load(vcpu);
> 
> vcpu_put()/_load()s are unneeded here, preempt notifiers will
> automatically call kvm_arch_vcpu_put()/_load() if we schedule.

Right. Will send an updated patch.

[...]
>> @@ -74,6 +74,7 @@ struct kvm_stats_debugfs_item debugfs_entries[] = {
>>  	{ "instruction_sigp_restart", VCPU_STAT(instruction_sigp_restart) },
>>  	{ "diagnose_10", VCPU_STAT(diagnose_10) },
>>  	{ "diagnose_44", VCPU_STAT(diagnose_44) },
>> +	{ "diagnose_9c", VCPU_STAT(diagnose_9c) },
>>  	{ NULL }
>>  };
>>  
> 
> We're switching to tracepoints instead of homebrewed statistics.  It's
> in feature-removal-schedule.txt already.  It's okay to add them now, but
> please prepare for their removal.

Yes, we have started with some tracepoint code conversion. We will try to speed
up a bit.

Christian