Issue #270: Add control for cpu_model and cpu_feature.

[tpodowd]

Add configuration parameters for cpu_model and cpu_feature
which control the setting of the following xml block.

<cpu mode='custom' match='exact'>
    <model fallback='allow'>core2duo</model>
    <feature policy='require' name='vmx'/>
</cpu>

This allows a vagrant KVM guest to provide nested virtualization.

NOTE: The cpu_model parameter was previously was listed to define
the cpu architecture but it actually was not doing anything. This
parameter has been reused to reflect the preferred guest cpu model.

[tpodowd]

Here's a snippet showing the feature is working against a vagrant guest.

$ sudo virsh dumpxml 34 | grep -A 4 '<cpu'
<cpu mode='custom' match='exact'>
     <model fallback='allow'>core2duo</model>
     <feature policy='require' name='vmx'/>
     <feature policy='require' name='ht'/>
</cpu>

Also if I log in to the box, I can see the following.

[vagrant@cloud1 ~]$ egrep '(vmx|svm)' --color=always /proc/cpuinfo
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss syscall nx lm constant_tsc up arch_perfmon rep_good unfair_spinlock pni vmx ssse3 cx16 hypervisor lahf_lm vnmi ept

So, it seems to be working. I can also halt and destroy.

[celebdor]

I tested it on arch linux and it works well for me ;-) Thanks @tpodowd

[tpodowd]

Awesome. Thanks @celebdor. Hopefully it will get merged soon. It seems to be working well for me also.