Re: [PATCH] target/i386: Remove monitor from some CPU model

[Eduardo Habkost]

On Fri, Nov 15, 2019 at 04:33:45PM +0800, Tao Xu wrote:
> Add new version of Snowridge, Denverton, Opteron_G3, EPYC, and Dhyana
> CPU model to remove MONITOR/MWAIT feature.
> 
> After QEMU/KVM use "-overcommit cpu-pm=on" to expose MONITOR/MWAIT
> (commit id 6f131f13e68d648a8e4f083c667ab1acd88ce4cd), the MONITOR/MWAIT
> feature in these CPU model is unused.
> 
> Signed-off-by: Tao Xu <address@hidden>
> ---
>  target/i386/cpu.c | 58 +++++++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 58 insertions(+)
> 
> diff --git a/target/i386/cpu.c b/target/i386/cpu.c
> index a624163ac2..7c5f1e8fe0 100644
> --- a/target/i386/cpu.c
> +++ b/target/i386/cpu.c
> @@ -2770,6 +2770,19 @@ static X86CPUDefinition builtin_x86_defs[] = {
>              MSR_ARCH_CAP_RDCL_NO | MSR_ARCH_CAP_SKIP_L1DFL_VMENTRY,
>          .xlevel = 0x80000008,
>          .model_id = "Intel Atom Processor (Denverton)",
> +        .versions = (X86CPUVersionDefinition[]) {
> +            { .version = 1 },
> +            {
> +                .version = 2,
> +                .props = (PropValue[]) {
> +                    { "monitor", "off" },
> +                    { "model-id",
> +                      "Intel Atom Processor (Denverton, no MONITOR)" },

We never changed model-id when adding/removing features in
machine-type code, and I don't see why we should start doing that
now.  This info might be helpful on "-cpu help", but probably
confusing for people looking at /proc/cpuinfo inside the VM.

If you think it is important to add extra info to "-cpu help", I
suggest you send a patch adding a separate field instead of
changing model ID on CPUID.

> [...]

-- 
Eduardo