VM missing "monitor" CPU flag

[MiLK_MaN]

Hi,

I previously had 3 Intel NUC 10's, running ESX 7 and then having ESX VM's in a nested environment working fine. I have another NUC 8 that is still in this setup now, and I can see that EVC modes are visible on the nested ESX.

All 3 NUC 10's have been converted to use Proxmox as the base hypervisor, and I've installed ESX and selected host for CPU, but ESX is claiming things like EVC and MWAIT is not available on the presented CPU. I've tried everything I can think of to change this, even presenting a Haswell or Skylake CPU and setting "--args "-cpu 'Haswell,hle=off,rtm=off'"" which boots the guest but it still complains that these modes are still not available.

Am I missing something silly?

 

[MiLK_MaN]

So it looks like for some reason that the flag "monitor" is not being passed to the guest. I can see in a "cat /proc/cpuinfo" in proxmox that the flag exists for my CPU, but in the guest it does not show up.

When I try and explicitly set it in the config, I get this error:

root@pve1:/etc/pve/qemu-server# qm start 108
kvm: warning: host doesn't support requested feature: CPUID.01H:ECX.monitor [bit 3]
kvm: warning: host doesn't support requested feature: CPUID.01H:ECX.monitor [bit 3]

root@pve1:/etc/pve/qemu-server# cat /proc/cpuinfo | grep monitor
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb invpcid_single ssbd ibrs ibpb stibp ibrs_enhanced tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid mpx rdseed adx smap clflushopt intel_pt xsaveopt xsavec xgetbv1 xsaves dtherm ida arat pln pts hwp hwp_notify hwp_act_window hwp_epp md_clear flush_l1d arch_capabilities

 

[Christoph_]

Same here on pve-manager/7.1-10/6ddebafe (running kernel: 5.15.12-1-pve) running on AMD Epyc CPU.

VM is set to "host" CPU Type but ESXi 7 claims no monitor/mwait support available.

PVE Host cpuinfo lists support.

 

[nitrocb]

Hello everybody,

I also experience the same problem running ESXi on proxmox on a AMD ryzen cpu.
Is this a known issue now or is this by intent?

Best regards

 

[patvdv]

Hi,

If you are looking to use EVC for a vSAN setup (with vCLS clustering services), you can work around this by disabling EVC for the individual vCLS VMs. You need to upgrade the HW compatability of the VM to level 14 first at least. After that you will find an additional menu item VMware EVC in the Configure section of the VM profile. You can then disable EVC and the vCLS will start just fine.

 

[mario156090]

hello, this problem is not fixed yed? I mean about a KVM fix or linux kernel fix and not workaround.

I can see this problem is with AMD processors.

 

[bkinigadner]

I found a way to pass the monitor flag to a host vm.

KVM does not pass the monitor flag to the host vm by default, as it once was not capable of doing so, and didn't want to break anything, so you have to manually add it.

1. go to the vms config file you want to have monitor pass through, in my case it is the VM with the VMID 101:

vim /etc/pve/qemu-server/<VMID>.conf

vim /etc/pve/qemu-server/101.conf

2. You have to add arguments, that will be passed directly to the qemu, the argument you have to pass are

-cpu host,monitor=on -overcommit cpu-pm=on

Add this line to your VM config, so it looks something like this:

...

agent: 1
args: -cpu host,monitor=on -overcommit cpu-pm=on
boot: order=scsi0;ide2;net0
cores: 16
cpu: host
ide2: local:iso/phoenix-ce2.0-fraser-6.5.2-stable-fnd-5.3.4-x86_64.iso,media=cdrom,size=5448096K
memory: 64000
...

This should pass through the monitor flag.