Re: [PATCH v2 2/4] cpu_map: Defined and enable EPYC-Rome model
On Thu, Oct 01, 2020 at 12:22:02 +0200, Markus Schade wrote:
Signed-off-by: Markus Schade <markus.schade(a)hetzner.com>
---
src/cpu_map/index.xml | 1 +
src/cpu_map/meson.build | 1 +
src/cpu_map/x86_EPYC-Rome.xml | 81 +++++++++++++++++++++++++++++++++++
3 files changed, 83 insertions(+)
create mode 100644 src/cpu_map/x86_EPYC-Rome.xml
...
diff --git a/src/cpu_map/x86_EPYC-Rome.xml
b/src/cpu_map/x86_EPYC-Rome.xml
new file mode 100644
index 0000000000..41d4123917
--- /dev/null
+++ b/src/cpu_map/x86_EPYC-Rome.xml
@@ -0,0 +1,81 @@
+<cpus>
+ <model name='EPYC-Rome'>
+ <decode host='on' guest='on'/>
+ <signature family='23' model='49'/>
+ <vendor name='AMD'/>
...
+ <feature name='mtrr'/>
+ <feature name='nx'/>
...
+ <feature name='tsc'/>
+ <feature name='vme'/>
QEMU definition of EPYC-Rome also contains 'npt' (CPUID_SVM_NPT),
'nrip-save' (CPUID_SVM_NRIPSAVE), and 'umip' (CPUID_7_0_ECX_UMIP). Any
specific reason for not including them in libvirt?
+ <feature name='wbnoinvd'/>
+ <feature name='xgetbv1'/>
+ <feature name='xsave'/>
+ <feature name='xsavec'/>
+ <feature name='xsaveerptr'/>
+ <feature name='xsaveopt'/>
+ </model>
+</cpus>
Jirka