Re: [PATCH v5 05/15] KVM: nVMX: Let userspace set nVMX MSR to any _host_ supported value

From: Yu Zhang
Date: Mon Oct 31 2022 - 12:39:42 EST

Next message: Dietmar Eggemann: "Re: [RFC PATCH 07/11] sched: Add proxy execution"
Previous message: Mike Kravetz: "Re: [PATCH 3/9] mm/hugetlb_cgroup: convert set_hugetlb_cgroup*() to folios"
Next in thread: Sean Christopherson: "Re: [PATCH v5 05/15] KVM: nVMX: Let userspace set nVMX MSR to any _host_ supported value"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Hi Sean & Paolo,

On Tue, Jun 07, 2022 at 09:35:54PM +0000, Sean Christopherson wrote:
> Restrict the nVMX MSRs based on KVM's config, not based on the guest's
> current config. Using the guest's config to audit the new config
> prevents userspace from restoring the original config (KVM's config) if
> at any point in the past the guest's config was restricted in any way.

May I ask for an example here, to explain why we use the KVM config
here, instead of the guest's? I mean, the guest's config can be
adjusted after cpuid updates by vmx_vcpu_after_set_cpuid(). Yet the
msr settings in vmcs_config.nested might be outdated by then.

Another question is about the setting of secondary_ctls_high in
nested_vmx_setup_ctls_msrs(). I saw there's a comment saying:
"Do not include those that depend on CPUID bits, they are
added later by vmx_vcpu_after_set_cpuid.".

But since cpuid updates can adjust the vmx->nested.msrs.secondary_ctls_high,
do we really need to clear those flags for secondary_ctls_high in this
global config? Could we just set
msrs->secondary_ctls_high = vmcs_conf->cpu_based_2nd_exec_ctrl?

If yes, code(in nested_vmx_setup_ctls_msrs()) such as
if (enable_ept) {
/* nested EPT: emulate EPT also to L1 */
msrs->secondary_ctls_high |=
SECONDARY_EXEC_ENABLE_EPT;
or
if (cpu_has_vmx_vmfunc()) {
msrs->secondary_ctls_high |=
SECONDARY_EXEC_ENABLE_VMFUNC;
and other similar ones may also be uncessary.

B.R.
Yu

>
> Fixes: 62cc6b9dc61e ("KVM: nVMX: support restore of VMX capability MSRs")
> Cc: stable@xxxxxxxxxxxxxxx
> Cc: David Matlack <dmatlack@xxxxxxxxxx>
> Signed-off-by: Sean Christopherson <seanjc@xxxxxxxxxx>
> ---
> arch/x86/kvm/vmx/nested.c | 100 ++++++++++++++++++++------------------
> 1 file changed, 52 insertions(+), 48 deletions(-)
>
> diff --git a/arch/x86/kvm/vmx/nested.c b/arch/x86/kvm/vmx/nested.c
> index 00c7b00c017a..fca30e79b3a0 100644
> --- a/arch/x86/kvm/vmx/nested.c
> +++ b/arch/x86/kvm/vmx/nested.c
> @@ -1223,7 +1223,7 @@ static int vmx_restore_vmx_basic(struct vcpu_vmx *vmx, u64 data)
> BIT_ULL(49) | BIT_ULL(54) | BIT_ULL(55) |
> /* reserved */
> BIT_ULL(31) | GENMASK_ULL(47, 45) | GENMASK_ULL(63, 56);
> - u64 vmx_basic = vmx->nested.msrs.basic;
> + u64 vmx_basic = vmcs_config.nested.basic;
>
> if (!is_bitwise_subset(vmx_basic, data, feature_and_reserved))
> return -EINVAL;
> @@ -1246,36 +1246,42 @@ static int vmx_restore_vmx_basic(struct vcpu_vmx *vmx, u64 data)
> return 0;
> }
>
> +static void vmx_get_control_msr(struct nested_vmx_msrs *msrs, u32 msr_index,
> + u32 **low, u32 **high)
> +{
> + switch (msr_index) {
> + case MSR_IA32_VMX_TRUE_PINBASED_CTLS:
> + *low = &msrs->pinbased_ctls_low;
> + *high = &msrs->pinbased_ctls_high;
> + break;
> + case MSR_IA32_VMX_TRUE_PROCBASED_CTLS:
> + *low = &msrs->procbased_ctls_low;
> + *high = &msrs->procbased_ctls_high;
> + break;
> + case MSR_IA32_VMX_TRUE_EXIT_CTLS:
> + *low = &msrs->exit_ctls_low;
> + *high = &msrs->exit_ctls_high;
> + break;
> + case MSR_IA32_VMX_TRUE_ENTRY_CTLS:
> + *low = &msrs->entry_ctls_low;
> + *high = &msrs->entry_ctls_high;
> + break;
> + case MSR_IA32_VMX_PROCBASED_CTLS2:
> + *low = &msrs->secondary_ctls_low;
> + *high = &msrs->secondary_ctls_high;
> + break;
> + default:
> + BUG();
> + }
> +}
> +
> static int
> vmx_restore_control_msr(struct vcpu_vmx *vmx, u32 msr_index, u64 data)
> {
> - u64 supported;
> u32 *lowp, *highp;
> + u64 supported;
>
> - switch (msr_index) {
> - case MSR_IA32_VMX_TRUE_PINBASED_CTLS:
> - lowp = &vmx->nested.msrs.pinbased_ctls_low;
> - highp = &vmx->nested.msrs.pinbased_ctls_high;
> - break;
> - case MSR_IA32_VMX_TRUE_PROCBASED_CTLS:
> - lowp = &vmx->nested.msrs.procbased_ctls_low;
> - highp = &vmx->nested.msrs.procbased_ctls_high;
> - break;
> - case MSR_IA32_VMX_TRUE_EXIT_CTLS:
> - lowp = &vmx->nested.msrs.exit_ctls_low;
> - highp = &vmx->nested.msrs.exit_ctls_high;
> - break;
> - case MSR_IA32_VMX_TRUE_ENTRY_CTLS:
> - lowp = &vmx->nested.msrs.entry_ctls_low;
> - highp = &vmx->nested.msrs.entry_ctls_high;
> - break;
> - case MSR_IA32_VMX_PROCBASED_CTLS2:
> - lowp = &vmx->nested.msrs.secondary_ctls_low;
> - highp = &vmx->nested.msrs.secondary_ctls_high;
> - break;
> - default:
> - BUG();
> - }
> + vmx_get_control_msr(&vmcs_config.nested, msr_index, &lowp, &highp);
>
> supported = vmx_control_msr(*lowp, *highp);
>
> @@ -1287,6 +1293,7 @@ vmx_restore_control_msr(struct vcpu_vmx *vmx, u32 msr_index, u64 data)
> if (!is_bitwise_subset(supported, data, GENMASK_ULL(63, 32)))
> return -EINVAL;
>
> + vmx_get_control_msr(&vmx->nested.msrs, msr_index, &lowp, &highp);
> *lowp = data;
> *highp = data >> 32;
> return 0;
> @@ -1300,10 +1307,8 @@ static int vmx_restore_vmx_misc(struct vcpu_vmx *vmx, u64 data)
> BIT_ULL(28) | BIT_ULL(29) | BIT_ULL(30) |
> /* reserved */
> GENMASK_ULL(13, 9) | BIT_ULL(31);
> - u64 vmx_misc;
> -
> - vmx_misc = vmx_control_msr(vmx->nested.msrs.misc_low,
> - vmx->nested.msrs.misc_high);
> + u64 vmx_misc = vmx_control_msr(vmcs_config.nested.misc_low,
> + vmcs_config.nested.misc_high);
>
> if (!is_bitwise_subset(vmx_misc, data, feature_and_reserved_bits))
> return -EINVAL;
> @@ -1331,10 +1336,8 @@ static int vmx_restore_vmx_misc(struct vcpu_vmx *vmx, u64 data)
>
> static int vmx_restore_vmx_ept_vpid_cap(struct vcpu_vmx *vmx, u64 data)
> {
> - u64 vmx_ept_vpid_cap;
> -
> - vmx_ept_vpid_cap = vmx_control_msr(vmx->nested.msrs.ept_caps,
> - vmx->nested.msrs.vpid_caps);
> + u64 vmx_ept_vpid_cap = vmx_control_msr(vmcs_config.nested.ept_caps,
> + vmcs_config.nested.vpid_caps);
>
> /* Every bit is either reserved or a feature bit. */
> if (!is_bitwise_subset(vmx_ept_vpid_cap, data, -1ULL))
> @@ -1345,20 +1348,21 @@ static int vmx_restore_vmx_ept_vpid_cap(struct vcpu_vmx *vmx, u64 data)
> return 0;
> }
>
> +static u64 *vmx_get_fixed0_msr(struct nested_vmx_msrs *msrs, u32 msr_index)
> +{
> + switch (msr_index) {
> + case MSR_IA32_VMX_CR0_FIXED0:
> + return &msrs->cr0_fixed0;
> + case MSR_IA32_VMX_CR4_FIXED0:
> + return &msrs->cr4_fixed0;
> + default:
> + BUG();
> + }
> +}
> +
> static int vmx_restore_fixed0_msr(struct vcpu_vmx *vmx, u32 msr_index, u64 data)
> {
> - u64 *msr;
> -
> - switch (msr_index) {
> - case MSR_IA32_VMX_CR0_FIXED0:
> - msr = &vmx->nested.msrs.cr0_fixed0;
> - break;
> - case MSR_IA32_VMX_CR4_FIXED0:
> - msr = &vmx->nested.msrs.cr4_fixed0;
> - break;
> - default:
> - BUG();
> - }
> + const u64 *msr = vmx_get_fixed0_msr(&vmcs_config.nested, msr_index);
>
> /*
> * 1 bits (which indicates bits which "must-be-1" during VMX operation)
> @@ -1367,7 +1371,7 @@ static int vmx_restore_fixed0_msr(struct vcpu_vmx *vmx, u32 msr_index, u64 data)
> if (!is_bitwise_subset(data, *msr, -1ULL))
> return -EINVAL;
>
> - *msr = data;
> + *vmx_get_fixed0_msr(&vmx->nested.msrs, msr_index) = data;
> return 0;
> }
>
> @@ -1428,7 +1432,7 @@ int vmx_set_vmx_msr(struct kvm_vcpu *vcpu, u32 msr_index, u64 data)
> vmx->nested.msrs.vmcs_enum = data;
> return 0;
> case MSR_IA32_VMX_VMFUNC:
> - if (data & ~vmx->nested.msrs.vmfunc_controls)
> + if (data & ~vmcs_config.nested.vmfunc_controls)
> return -EINVAL;
> vmx->nested.msrs.vmfunc_controls = data;
> return 0;
> --
> 2.36.1.255.ge46751e96f-goog
>

Next message: Dietmar Eggemann: "Re: [RFC PATCH 07/11] sched: Add proxy execution"
Previous message: Mike Kravetz: "Re: [PATCH 3/9] mm/hugetlb_cgroup: convert set_hugetlb_cgroup*() to folios"
Next in thread: Sean Christopherson: "Re: [PATCH v5 05/15] KVM: nVMX: Let userspace set nVMX MSR to any _host_ supported value"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]