Re: [Qemu-devel] [PATCH] i386: Enable IA32_MISC_ENABLE MWAIT bit when exposing mwait/monitor

[Eduardo Habkost]

On Mon, May 20, 2019 at 02:59:53PM +0200, Paolo Bonzini wrote:
> On 14/05/19 08:06, Wanpeng Li wrote:
> > From: Wanpeng Li <address@hidden>
> > 
> > The CPUID.01H:ECX[bit 3] ought to mirror the value of the MSR 
> > IA32_MISC_ENABLE MWAIT bit and as userspace has control of them 
> > both, it is userspace's job to configure both bits to match on 
> > the initial setup.
> 
> Queued, thanks.
> 
> Paolo
> 
> > Cc: Eduardo Habkost <address@hidden>
> > Cc: Paolo Bonzini <address@hidden>
> > Cc: Radim Krčmář <address@hidden>
> > Signed-off-by: Wanpeng Li <address@hidden>
> > ---
> >  target/i386/cpu.c | 3 +++
> >  target/i386/cpu.h | 1 +
> >  2 files changed, 4 insertions(+)
> > 
> > diff --git a/target/i386/cpu.c b/target/i386/cpu.c
> > index 722c551..40b6108 100644
> > --- a/target/i386/cpu.c
> > +++ b/target/i386/cpu.c
> > @@ -4729,6 +4729,9 @@ static void x86_cpu_reset(CPUState *s)
> >  
> >      env->pat = 0x0007040600070406ULL;
> >      env->msr_ia32_misc_enable = MSR_IA32_MISC_ENABLE_DEFAULT;
> > +    if (enable_cpu_pm) {
> > +        env->msr_ia32_misc_enable |= MSR_IA32_MISC_ENABLE_MWAIT;
> > +    }

What if enable_cpu_pm is false but we're running TCG, or if
enable_cpu_pm is true but we're not using -cpu host?

Shouldn't this be conditional on
  (env->features[FEAT_1_ECX] & CPUID_EXT_MONITOR)
instead?

> >  
> >      memset(env->dr, 0, sizeof(env->dr));
> >      env->dr[6] = DR6_FIXED_1;
> > diff --git a/target/i386/cpu.h b/target/i386/cpu.h
> > index 0128910..b94c329 100644
> > --- a/target/i386/cpu.h
> > +++ b/target/i386/cpu.h
> > @@ -387,6 +387,7 @@ typedef enum X86Seg {
> >  #define MSR_IA32_MISC_ENABLE            0x1a0
> >  /* Indicates good rep/movs microcode on some processors: */
> >  #define MSR_IA32_MISC_ENABLE_DEFAULT    1
> > +#define MSR_IA32_MISC_ENABLE_MWAIT      (1ULL << 18)
> >  
> >  #define MSR_MTRRphysBase(reg)           (0x200 + 2 * (reg))
> >  #define MSR_MTRRphysMask(reg)           (0x200 + 2 * (reg) + 1)
> > 
> 

-- 
Eduardo