Re: [PATCH v6 5/5] x86/kvm: Avoid dynamic allocation of pvclock data when SEV is active

From: Sean Christopherson
Date: Mon Sep 10 2018 - 09:29:15 EST


On Mon, 2018-09-10 at 08:15 -0500, Brijesh Singh wrote:
>
> On 9/10/18 7:27 AM, Borislav Petkov wrote:
> >
> > On Fri, Sep 07, 2018 at 12:57:30PM -0500, Brijesh Singh wrote:
> > >
> > > diff --git a/arch/x86/kernel/kvmclock.c b/arch/x86/kernel/kvmclock.c
> > > index 376fd3a..6086b56 100644
> > > --- a/arch/x86/kernel/kvmclock.c
> > > +++ b/arch/x86/kernel/kvmclock.c
> > > @@ -65,6 +65,15 @@ static struct pvclock_vsyscall_time_info
> > > Âstatic struct pvclock_wall_clock wall_clock __decrypted;
> > > Âstatic DEFINE_PER_CPU(struct pvclock_vsyscall_time_info *, hv_clock_per_cpu);
> > > Â
> > > +#ifdef CONFIG_AMD_MEM_ENCRYPT
> > > +/*
> > > + * The auxiliary array will be used when SEV is active. In non-SEV case,
> > > + * it will be freed by free_decrypted_mem().
> > > + */
> > > +static struct pvclock_vsyscall_time_info
> > > + hv_clock_aux[NR_CPUS] __decrypted_aux;
> > Hmm, so worst case that's 64 4K pages:
> >
> > (8192*32)/4096 = 64 4K pages.
> We can minimize the worst case memory usage. The number of VCPUs
> supported by KVM maybe less than NR_CPUS. e.g Currently KVM_MAX_VCPUS is
> set to 288

KVM_MAX_VCPUS is a property of the host, whereas this code runs in the
guest, e.g.ÂKVM_MAX_VCPUS could be 2048 in the host for all we know.

> (288 * 64)/4096 = 4 4K pages.
>
> (pvclock_vsyscall_time_info is cache aligned so it will be 64 bytes)

Ah, I was wondering why my calculations were always different than
yours. ÂI was looking at struct pvclock_vcpu_time_info, which is 32
bytes.

> #if NR_CPUS > KVM_MAX_VCPUS
> #define HV_AUX_ARRAY_SIZEÂ KVM_MAX_VCPUS
> #else
> #define HV_AUX_ARRAY_SIZE NR_CPUS
> #endif
>
> static struct pvclock_vsyscall_time_info
> ÂÂÂ ÂÂÂ ÂÂÂ ÂÂÂ ÂÂÂ ÂÂÂ hv_clock_aux[HV_AUX_ARRAY_SIZE] __decrypted_aux;