Re: [PATCH v5 18/18] kvm: arm64: Allow tuning the physical address size for VM

[Auger Eric]

Hi Suzuki,
On 9/17/18 12:41 PM, Suzuki K Poulose wrote:
> Allow specifying the physical address size limit for a new
> VM via the kvm_type argument for the KVM_CREATE_VM ioctl. This
> allows us to finalise the stage2 page table as early as possible
> and hence perform the right checks on the memory slots
> without complication. The size is ecnoded as Log2(PA_Size) in
encoded
> bits[7:0] of the type field. For backward compatibility the
> value 0 is reserved and implies 40bits. Also, lift the limit
> of the IPA to host limit and allow lower IPA sizes (e.g, 32).
> 
> The userspace could check the extension KVM_CAP_ARM_VM_PHYS_SHIFT
> for the availability of this feature. The cap check returns the
> maximum limit for the physical address shift supported by the host.
> 
> Cc: Marc Zyngier <marc.zyngier@xxxxxxx>
> Cc: Christoffer Dall <cdall@xxxxxxxxxx>
> Cc: Peter Maydel <peter.maydell@xxxxxxxxxx>
> Cc: Paolo Bonzini <pbonzini@xxxxxxxxxx>
> Cc: Radim KrÄmÃÅ <rkrcmar@xxxxxxxxxx>
> Signed-off-by: Suzuki K Poulose <suzuki.poulose@xxxxxxx>
> ---
> Changes since v4:
>  - Fold the introduction of the KVM_CAP_ARM_VM_PHYS_SHIFT to this
>    patch to allow detection of the availability of the feature for
>    userspace.
>  - Document the API
>  - Restrict the feature only to arm64.
> Changes since V3:
>  - Switch to a CAP, that can be checkd via EXTENSIONS on KVM device
>    fd, rather than a dedicated ioctl.
> ---
>  Documentation/virtual/kvm/api.txt       |  8 ++++++++
>  arch/arm64/include/asm/stage2_pgtable.h | 20 --------------------
>  arch/arm64/kvm/reset.c                  | 20 ++++++++++++++++----
>  include/uapi/linux/kvm.h                | 10 ++++++++++
>  4 files changed, 34 insertions(+), 24 deletions(-)
> 
> diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt
> index c664064f76fb..f860251ff27c 100644
> --- a/Documentation/virtual/kvm/api.txt
> +++ b/Documentation/virtual/kvm/api.txt
> @@ -122,6 +122,14 @@ the default trap & emulate implementation (which changes the virtual
>  memory layout to fit in user mode), check KVM_CAP_MIPS_VZ and use the
>  flag KVM_VM_MIPS_VZ.
>  
> +To configure the physical address space size for a VM (IPA size) on arm64,
> +check KVM_CAP_ARM_VM_PHYS_SHIFT (which returns the maximum limit for the
> +IPA shift) and use KVM_VM_TYPE_ARM_PHYS_SHIFT(PHYS_SHIFT). Bits[7-0] of the
> +machine type has been reserved for specifying the PHYS_SHIFT.
are reserved to pass the PHYS_SHIFT?
> +The supported range is [32...IPA_LIMIT], where IPA_LIMIT could be
s/could be/is
> +identified by checking KVM_CAP_ARM_VM_PHYS_SHIFT. For backward compatibility
> +a value of 0 selects 40bits.
> +
>  
>  4.3 KVM_GET_MSR_INDEX_LIST, KVM_GET_MSR_FEATURE_INDEX_LIST
>  
> diff --git a/arch/arm64/include/asm/stage2_pgtable.h b/arch/arm64/include/asm/stage2_pgtable.h
> index 6a56fdff0823..0b339f5a4a7c 100644
> --- a/arch/arm64/include/asm/stage2_pgtable.h
> +++ b/arch/arm64/include/asm/stage2_pgtable.h
> @@ -42,28 +42,8 @@
>   * the range (IPA_SHIFT, IPA_SHIFT - 4).
>   */
>  #define stage2_pgtable_levels(ipa)	ARM64_HW_PGTABLE_LEVELS((ipa) - 4)
> -#define STAGE2_PGTABLE_LEVELS		stage2_pgtable_levels(KVM_PHYS_SHIFT)
>  #define kvm_stage2_levels(kvm)		VTCR_EL2_LVLS(kvm->arch.vtcr)
>  
> -/*
> - * With all the supported VA_BITs and 40bit guest IPA, the following condition
> - * is always true:
> - *
> - *       STAGE2_PGTABLE_LEVELS <= CONFIG_PGTABLE_LEVELS
> - *
> - * We base our stage-2 page table walker helpers on this assumption and
> - * fall back to using the host version of the helper wherever possible.
> - * i.e, if a particular level is not folded (e.g, PUD) at stage2, we fall back
> - * to using the host version, since it is guaranteed it is not folded at host.
> - *
> - * If the condition breaks in the future, we can rearrange the host level
> - * definitions and reuse them for stage2. Till then...
> - */
> -#if STAGE2_PGTABLE_LEVELS > CONFIG_PGTABLE_LEVELS
> -#error "Unsupported combination of guest IPA and host VA_BITS."
> -#endif
> -
> -
>  /* stage2_pgdir_shift() is the size mapped by top-level stage2 entry for the VM */
>  #define stage2_pgdir_shift(kvm)		pt_levels_pgdir_shift(kvm_stage2_levels(kvm))
>  #define stage2_pgdir_size(kvm)		(1ULL << stage2_pgdir_shift(kvm))
> diff --git a/arch/arm64/kvm/reset.c b/arch/arm64/kvm/reset.c
> index 0393bb974b23..c9640159e11f 100644
> --- a/arch/arm64/kvm/reset.c
> +++ b/arch/arm64/kvm/reset.c
> @@ -89,6 +89,9 @@ int kvm_arch_dev_ioctl_check_extension(struct kvm *kvm, long ext)
>  	case KVM_CAP_VCPU_EVENTS:
>  		r = 1;
>  		break;
> +	case KVM_CAP_ARM_VM_PHYS_SHIFT:
> +		r = kvm_ipa_limit;
> +		break;
>  	default:
>  		r = 0;
>  	}
> @@ -190,16 +193,25 @@ int kvm_arm_config_vm(struct kvm *kvm, unsigned long type)
>  {
>  	u64 vtcr = VTCR_EL2_FLAGS;
>  	u64 parange;
> -	u8 lvls;
> +	u8 lvls, ipa_shift;
>  
> -	if (type)
> +	if (type & ~KVM_VM_TYPE_ARM_PHYS_SHIFT_MASK)
>  		return -EINVAL;
>  
> +	ipa_shift = KVM_VM_TYPE_ARM_PHYS_SHIFT(type);
> +	if (ipa_shift) {
> +		if (ipa_shift > kvm_ipa_limit ||
> +		    ipa_shift < 32)
> +			return -EINVAL;
> +	} else {
> +		ipa_shift = KVM_PHYS_SHIFT;
> +	}
> +
>  	/*
>  	 * Use a minimum 2 level page table to prevent splitting
>  	 * host PMD huge pages at stage2.
>  	 */
> -	lvls = stage2_pgtable_levels(KVM_PHYS_SHIFT);
> +	lvls = stage2_pgtable_levels(ipa_shift);
>  	if (lvls < 2)
>  		lvls = 2;
>  
> @@ -221,7 +233,7 @@ int kvm_arm_config_vm(struct kvm *kvm, unsigned long type)
>  		VTCR_EL2_VS_8BIT;
>  
>  	vtcr |= VTCR_EL2_LVLS_TO_SL0(lvls);
> -	vtcr |= VTCR_EL2_T0SZ(KVM_PHYS_SHIFT);
> +	vtcr |= VTCR_EL2_T0SZ(ipa_shift);
>  
>  	kvm->arch.vtcr = vtcr;
>  	return 0;
> diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h
> index 07548de5c988..2a6b29c446db 100644
> --- a/include/uapi/linux/kvm.h
> +++ b/include/uapi/linux/kvm.h
> @@ -750,6 +750,15 @@ struct kvm_ppc_resize_hpt {
>  
>  #define KVM_S390_SIE_PAGE_OFFSET 1
>  
> +/*
> + * On arm64, machine type can be used to request the physical
> + * address size for the VM. Bits[7-0] has been reserved for the PA
s/has been reserved/are?

Thanks

Eric
> + * size shift (i.e, log2(PA_Size)). For backward compatibility,
> + * value 0 implies the default IPA size, 40bits.
> + */
> +#define KVM_VM_TYPE_ARM_PHYS_SHIFT_MASK	0xffULL
> +#define KVM_VM_TYPE_ARM_PHYS_SHIFT(x)		\
> +	((x) & KVM_VM_TYPE_ARM_PHYS_SHIFT_MASK)
>  /*
>   * ioctls for /dev/kvm fds:
>   */
> @@ -952,6 +961,7 @@ struct kvm_ppc_resize_hpt {
>  #define KVM_CAP_S390_HPAGE_1M 156
>  #define KVM_CAP_NESTED_STATE 157
>  #define KVM_CAP_ARM_INJECT_SERROR_ESR 158
> +#define KVM_CAP_ARM_VM_PHYS_SHIFT 159 /* returns maximum PA shift for a VM */
>  
>  #ifdef KVM_CAP_IRQ_ROUTING
>  
>