From b96e6506c2ea7c8e3ef0860169b6d0ba2f1aca9f Mon Sep 17 00:00:00 2001 From: Mohammed Gamal Date: Thu, 3 Sep 2020 16:11:22 +0200 Subject: KVM: x86: VMX: Make smaller physical guest address space support user-configurable This patch exposes allow_smaller_maxphyaddr to the user as a module parameter. Since smaller physical address spaces are only supported on VMX, the parameter is only exposed in the kvm_intel module. For now disable support by default, and let the user decide if they want to enable it. Modifications to VMX page fault and EPT violation handling will depend on whether that parameter is enabled. Signed-off-by: Mohammed Gamal Message-Id: <20200903141122.72908-1-mgamal@redhat.com> Signed-off-by: Paolo Bonzini --- arch/x86/kvm/vmx/vmx.c | 15 ++++++++++----- arch/x86/kvm/vmx/vmx.h | 5 ++++- arch/x86/kvm/x86.c | 2 +- 3 files changed, 15 insertions(+), 7 deletions(-) (limited to 'arch') diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c index 75cd720c9e8c..f0384e93548a 100644 --- a/arch/x86/kvm/vmx/vmx.c +++ b/arch/x86/kvm/vmx/vmx.c @@ -129,6 +129,9 @@ static bool __read_mostly enable_preemption_timer = 1; module_param_named(preemption_timer, enable_preemption_timer, bool, S_IRUGO); #endif +extern bool __read_mostly allow_smaller_maxphyaddr; +module_param(allow_smaller_maxphyaddr, bool, S_IRUGO); + #define KVM_VM_CR0_ALWAYS_OFF (X86_CR0_NW | X86_CR0_CD) #define KVM_VM_CR0_ALWAYS_ON_UNRESTRICTED_GUEST X86_CR0_NE #define KVM_VM_CR0_ALWAYS_ON \ @@ -4803,6 +4806,7 @@ static int handle_exception_nmi(struct kvm_vcpu *vcpu) * EPT will cause page fault only if we need to * detect illegal GPAs. */ + WARN_ON_ONCE(!allow_smaller_maxphyaddr); kvm_fixup_and_inject_pf_error(vcpu, cr2, error_code); return 1; } else @@ -5331,7 +5335,7 @@ static int handle_ept_violation(struct kvm_vcpu *vcpu) * would also use advanced VM-exit information for EPT violations to * reconstruct the page fault error code. */ - if (unlikely(kvm_mmu_is_illegal_gpa(vcpu, gpa))) + if (unlikely(allow_smaller_maxphyaddr && kvm_mmu_is_illegal_gpa(vcpu, gpa))) return kvm_emulate_instruction(vcpu, 0); return kvm_mmu_page_fault(vcpu, gpa, error_code, NULL, 0); @@ -8305,11 +8309,12 @@ static int __init vmx_init(void) vmx_check_vmcs12_offsets(); /* - * Intel processors don't have problems with - * GUEST_MAXPHYADDR < HOST_MAXPHYADDR so enable - * it for VMX by default + * Shadow paging doesn't have a (further) performance penalty + * from GUEST_MAXPHYADDR < HOST_MAXPHYADDR so enable it + * by default */ - allow_smaller_maxphyaddr = true; + if (!enable_ept) + allow_smaller_maxphyaddr = true; return 0; } diff --git a/arch/x86/kvm/vmx/vmx.h b/arch/x86/kvm/vmx/vmx.h index a2f82127c170..a0e47720f60c 100644 --- a/arch/x86/kvm/vmx/vmx.h +++ b/arch/x86/kvm/vmx/vmx.h @@ -552,7 +552,10 @@ static inline bool vmx_has_waitpkg(struct vcpu_vmx *vmx) static inline bool vmx_need_pf_intercept(struct kvm_vcpu *vcpu) { - return !enable_ept || cpuid_maxphyaddr(vcpu) < boot_cpu_data.x86_phys_bits; + if (!enable_ept) + return true; + + return allow_smaller_maxphyaddr && cpuid_maxphyaddr(vcpu) < boot_cpu_data.x86_phys_bits; } void dump_vmcs(void); diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index e3de0fe5af37..67362607e396 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -188,7 +188,7 @@ static struct kvm_shared_msrs __percpu *shared_msrs; u64 __read_mostly host_efer; EXPORT_SYMBOL_GPL(host_efer); -bool __read_mostly allow_smaller_maxphyaddr; +bool __read_mostly allow_smaller_maxphyaddr = 0; EXPORT_SYMBOL_GPL(allow_smaller_maxphyaddr); static u64 __read_mostly host_xss; -- cgit v1.2.3 From ee6fa053019478918ca15eadaa93e516ddb9da8d Mon Sep 17 00:00:00 2001 From: Maxim Levitsky Date: Mon, 21 Sep 2020 13:38:05 +0300 Subject: KVM: x86: fix MSR_IA32_TSC read for nested migration MSR reads/writes should always access the L1 state, since the (nested) hypervisor should intercept all the msrs it wants to adjust, and these that it doesn't should be read by the guest as if the host had read it. However IA32_TSC is an exception. Even when not intercepted, guest still reads the value + TSC offset. The write however does not take any TSC offset into account. This is documented in Intel's SDM and seems also to happen on AMD as well. This creates a problem when userspace wants to read the IA32_TSC value and then write it. (e.g for migration) In this case it reads L2 value but write is interpreted as an L1 value. To fix this make the userspace initiated reads of IA32_TSC return L1 value as well. Huge thanks to Dave Gilbert for helping me understand this very confusing semantic of MSR writes. Signed-off-by: Maxim Levitsky Message-Id: <20200921103805.9102-2-mlevitsk@redhat.com> Signed-off-by: Paolo Bonzini --- arch/x86/kvm/x86.c | 17 +++++++++++++++-- 1 file changed, 15 insertions(+), 2 deletions(-) (limited to 'arch') diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 67362607e396..e3fe1d126cc3 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -3224,9 +3224,22 @@ int kvm_get_msr_common(struct kvm_vcpu *vcpu, struct msr_data *msr_info) case MSR_IA32_POWER_CTL: msr_info->data = vcpu->arch.msr_ia32_power_ctl; break; - case MSR_IA32_TSC: - msr_info->data = kvm_scale_tsc(vcpu, rdtsc()) + vcpu->arch.tsc_offset; + case MSR_IA32_TSC: { + /* + * Intel SDM states that MSR_IA32_TSC read adds the TSC offset + * even when not intercepted. AMD manual doesn't explicitly + * state this but appears to behave the same. + * + * On userspace reads and writes, however, we unconditionally + * operate L1's TSC value to ensure backwards-compatible + * behavior for migration. + */ + u64 tsc_offset = msr_info->host_initiated ? vcpu->arch.l1_tsc_offset : + vcpu->arch.tsc_offset; + + msr_info->data = kvm_scale_tsc(vcpu, rdtsc()) + tsc_offset; break; + } case MSR_MTRRcap: case 0x200 ... 0x2ff: return kvm_mtrr_get_msr(vcpu, msr_info->index, &msr_info->data); -- cgit v1.2.3 From 8d214c481611b29458a57913bd786f0ac06f0605 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Wed, 23 Sep 2020 14:53:52 -0700 Subject: KVM: x86: Reset MMU context if guest toggles CR4.SMAP or CR4.PKE Reset the MMU context during kvm_set_cr4() if SMAP or PKE is toggled. Recent commits to (correctly) not reload PDPTRs when SMAP/PKE are toggled inadvertantly skipped the MMU context reset due to the mask of bits that triggers PDPTR loads also being used to trigger MMU context resets. Fixes: 427890aff855 ("kvm: x86: Toggling CR4.SMAP does not load PDPTEs in PAE mode") Fixes: cb957adb4ea4 ("kvm: x86: Toggling CR4.PKE does not load PDPTEs in PAE mode") Cc: Jim Mattson Cc: Peter Shier Cc: Oliver Upton Signed-off-by: Sean Christopherson Message-Id: <20200923215352.17756-1-sean.j.christopherson@intel.com> Signed-off-by: Paolo Bonzini --- arch/x86/kvm/x86.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index e3fe1d126cc3..58fa354b1446 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -976,6 +976,7 @@ int kvm_set_cr4(struct kvm_vcpu *vcpu, unsigned long cr4) unsigned long old_cr4 = kvm_read_cr4(vcpu); unsigned long pdptr_bits = X86_CR4_PGE | X86_CR4_PSE | X86_CR4_PAE | X86_CR4_SMEP; + unsigned long mmu_role_bits = pdptr_bits | X86_CR4_SMAP | X86_CR4_PKE; if (kvm_valid_cr4(vcpu, cr4)) return 1; @@ -1003,7 +1004,7 @@ int kvm_set_cr4(struct kvm_vcpu *vcpu, unsigned long cr4) if (kvm_x86_ops.set_cr4(vcpu, cr4)) return 1; - if (((cr4 ^ old_cr4) & pdptr_bits) || + if (((cr4 ^ old_cr4) & mmu_role_bits) || (!(cr4 & X86_CR4_PCIDE) && (old_cr4 & X86_CR4_PCIDE))) kvm_mmu_reset_context(vcpu); -- cgit v1.2.3 From 4bb05f30483fd21ea5413eaf1182768f251cf625 Mon Sep 17 00:00:00 2001 From: Tom Lendacky Date: Thu, 24 Sep 2020 13:41:57 -0500 Subject: KVM: SVM: Add a dedicated INVD intercept routine The INVD instruction intercept performs emulation. Emulation can't be done on an SEV guest because the guest memory is encrypted. Provide a dedicated intercept routine for the INVD intercept. And since the instruction is emulated as a NOP, just skip it instead. Fixes: 1654efcbc431 ("KVM: SVM: Add KVM_SEV_INIT command") Signed-off-by: Tom Lendacky Message-Id: Signed-off-by: Paolo Bonzini --- arch/x86/kvm/svm/svm.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) (limited to 'arch') diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c index 5764b87379cf..d4e18bda19c7 100644 --- a/arch/x86/kvm/svm/svm.c +++ b/arch/x86/kvm/svm/svm.c @@ -2183,6 +2183,12 @@ static int iret_interception(struct vcpu_svm *svm) return 1; } +static int invd_interception(struct vcpu_svm *svm) +{ + /* Treat an INVD instruction as a NOP and just skip it. */ + return kvm_skip_emulated_instruction(&svm->vcpu); +} + static int invlpg_interception(struct vcpu_svm *svm) { if (!static_cpu_has(X86_FEATURE_DECODEASSISTS)) @@ -2774,7 +2780,7 @@ static int (*const svm_exit_handlers[])(struct vcpu_svm *svm) = { [SVM_EXIT_RDPMC] = rdpmc_interception, [SVM_EXIT_CPUID] = cpuid_interception, [SVM_EXIT_IRET] = iret_interception, - [SVM_EXIT_INVD] = emulate_on_interception, + [SVM_EXIT_INVD] = invd_interception, [SVM_EXIT_PAUSE] = pause_interception, [SVM_EXIT_HLT] = halt_interception, [SVM_EXIT_INVLPG] = invlpg_interception, -- cgit v1.2.3