From 30510387a5e45bfcf8190e03ec7aa15b295828e2 Mon Sep 17 00:00:00 2001 From: Wei Wang Date: Mon, 12 Nov 2018 12:23:14 +0000 Subject: svm: Add mutex_lock to protect apic_access_page_done on AMD systems There is a race condition when accessing kvm->arch.apic_access_page_done. Due to it, x86_set_memory_region will fail when creating the second vcpu for a svm guest. Add a mutex_lock to serialize the accesses to apic_access_page_done. This lock is also used by vmx for the same purpose. Signed-off-by: Wei Wang Signed-off-by: Amadeusz Juskowiak Signed-off-by: Julian Stecklina Signed-off-by: Suravee Suthikulpanit Reviewed-by: Joerg Roedel Cc: stable@vger.kernel.org Signed-off-by: Paolo Bonzini --- arch/x86/kvm/svm.c | 19 +++++++++++-------- 1 file changed, 11 insertions(+), 8 deletions(-) (limited to 'arch/x86/kvm/svm.c') diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c index 0e21ccc46792..033ec01661b9 100644 --- a/arch/x86/kvm/svm.c +++ b/arch/x86/kvm/svm.c @@ -1664,20 +1664,23 @@ static u64 *avic_get_physical_id_entry(struct kvm_vcpu *vcpu, static int avic_init_access_page(struct kvm_vcpu *vcpu) { struct kvm *kvm = vcpu->kvm; - int ret; + int ret = 0; + mutex_lock(&kvm->slots_lock); if (kvm->arch.apic_access_page_done) - return 0; + goto out; - ret = x86_set_memory_region(kvm, - APIC_ACCESS_PAGE_PRIVATE_MEMSLOT, - APIC_DEFAULT_PHYS_BASE, - PAGE_SIZE); + ret = __x86_set_memory_region(kvm, + APIC_ACCESS_PAGE_PRIVATE_MEMSLOT, + APIC_DEFAULT_PHYS_BASE, + PAGE_SIZE); if (ret) - return ret; + goto out; kvm->arch.apic_access_page_done = true; - return 0; +out: + mutex_unlock(&kvm->slots_lock); + return ret; } static int avic_init_backing_page(struct kvm_vcpu *vcpu) -- cgit From 326e742533bf0a23f0127d8ea62fb558ba665f08 Mon Sep 17 00:00:00 2001 From: Leonid Shatz Date: Tue, 6 Nov 2018 12:14:25 +0200 Subject: KVM: nVMX/nSVM: Fix bug which sets vcpu->arch.tsc_offset to L1 tsc_offset Since commit e79f245ddec1 ("X86/KVM: Properly update 'tsc_offset' to represent the running guest"), vcpu->arch.tsc_offset meaning was changed to always reflect the tsc_offset value set on active VMCS. Regardless if vCPU is currently running L1 or L2. However, above mentioned commit failed to also change kvm_vcpu_write_tsc_offset() to set vcpu->arch.tsc_offset correctly. This is because vmx_write_tsc_offset() could set the tsc_offset value in active VMCS to given offset parameter *plus vmcs12->tsc_offset*. However, kvm_vcpu_write_tsc_offset() just sets vcpu->arch.tsc_offset to given offset parameter. Without taking into account the possible addition of vmcs12->tsc_offset. (Same is true for SVM case). Fix this issue by changing kvm_x86_ops->write_tsc_offset() to return actually set tsc_offset in active VMCS and modify kvm_vcpu_write_tsc_offset() to set returned value in vcpu->arch.tsc_offset. In addition, rename write_tsc_offset() callback to write_l1_tsc_offset() to make it clear that it is meant to set L1 TSC offset. Fixes: e79f245ddec1 ("X86/KVM: Properly update 'tsc_offset' to represent the running guest") Reviewed-by: Liran Alon Reviewed-by: Mihai Carabas Reviewed-by: Krish Sadhukhan Signed-off-by: Leonid Shatz Cc: stable@vger.kernel.org Signed-off-by: Paolo Bonzini --- arch/x86/include/asm/kvm_host.h | 3 ++- arch/x86/kvm/svm.c | 5 +++-- arch/x86/kvm/vmx.c | 21 +++++++++------------ arch/x86/kvm/x86.c | 6 +++--- 4 files changed, 17 insertions(+), 18 deletions(-) (limited to 'arch/x86/kvm/svm.c') diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h index 55e51ff7e421..fbda5a917c5b 100644 --- a/arch/x86/include/asm/kvm_host.h +++ b/arch/x86/include/asm/kvm_host.h @@ -1094,7 +1094,8 @@ struct kvm_x86_ops { bool (*has_wbinvd_exit)(void); u64 (*read_l1_tsc_offset)(struct kvm_vcpu *vcpu); - void (*write_tsc_offset)(struct kvm_vcpu *vcpu, u64 offset); + /* Returns actual tsc_offset set in active VMCS */ + u64 (*write_l1_tsc_offset)(struct kvm_vcpu *vcpu, u64 offset); void (*get_exit_info)(struct kvm_vcpu *vcpu, u64 *info1, u64 *info2); diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c index 033ec01661b9..a24733aade4c 100644 --- a/arch/x86/kvm/svm.c +++ b/arch/x86/kvm/svm.c @@ -1446,7 +1446,7 @@ static u64 svm_read_l1_tsc_offset(struct kvm_vcpu *vcpu) return vcpu->arch.tsc_offset; } -static void svm_write_tsc_offset(struct kvm_vcpu *vcpu, u64 offset) +static u64 svm_write_l1_tsc_offset(struct kvm_vcpu *vcpu, u64 offset) { struct vcpu_svm *svm = to_svm(vcpu); u64 g_tsc_offset = 0; @@ -1464,6 +1464,7 @@ static void svm_write_tsc_offset(struct kvm_vcpu *vcpu, u64 offset) svm->vmcb->control.tsc_offset = offset + g_tsc_offset; mark_dirty(svm->vmcb, VMCB_INTERCEPTS); + return svm->vmcb->control.tsc_offset; } static void avic_init_vmcb(struct vcpu_svm *svm) @@ -7152,7 +7153,7 @@ static struct kvm_x86_ops svm_x86_ops __ro_after_init = { .has_wbinvd_exit = svm_has_wbinvd_exit, .read_l1_tsc_offset = svm_read_l1_tsc_offset, - .write_tsc_offset = svm_write_tsc_offset, + .write_l1_tsc_offset = svm_write_l1_tsc_offset, .set_tdp_cr3 = set_tdp_cr3, diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c index 5f43fcfc225b..764c23dc444f 100644 --- a/arch/x86/kvm/vmx.c +++ b/arch/x86/kvm/vmx.c @@ -3464,11 +3464,9 @@ static u64 vmx_read_l1_tsc_offset(struct kvm_vcpu *vcpu) return vcpu->arch.tsc_offset; } -/* - * writes 'offset' into guest's timestamp counter offset register - */ -static void vmx_write_tsc_offset(struct kvm_vcpu *vcpu, u64 offset) +static u64 vmx_write_l1_tsc_offset(struct kvm_vcpu *vcpu, u64 offset) { + u64 active_offset = offset; if (is_guest_mode(vcpu)) { /* * We're here if L1 chose not to trap WRMSR to TSC. According @@ -3476,17 +3474,16 @@ static void vmx_write_tsc_offset(struct kvm_vcpu *vcpu, u64 offset) * set for L2 remains unchanged, and still needs to be added * to the newly set TSC to get L2's TSC. */ - struct vmcs12 *vmcs12; - /* recalculate vmcs02.TSC_OFFSET: */ - vmcs12 = get_vmcs12(vcpu); - vmcs_write64(TSC_OFFSET, offset + - (nested_cpu_has(vmcs12, CPU_BASED_USE_TSC_OFFSETING) ? - vmcs12->tsc_offset : 0)); + struct vmcs12 *vmcs12 = get_vmcs12(vcpu); + if (nested_cpu_has(vmcs12, CPU_BASED_USE_TSC_OFFSETING)) + active_offset += vmcs12->tsc_offset; } else { trace_kvm_write_tsc_offset(vcpu->vcpu_id, vmcs_read64(TSC_OFFSET), offset); - vmcs_write64(TSC_OFFSET, offset); } + + vmcs_write64(TSC_OFFSET, active_offset); + return active_offset; } /* @@ -15074,7 +15071,7 @@ static struct kvm_x86_ops vmx_x86_ops __ro_after_init = { .has_wbinvd_exit = cpu_has_vmx_wbinvd_exit, .read_l1_tsc_offset = vmx_read_l1_tsc_offset, - .write_tsc_offset = vmx_write_tsc_offset, + .write_l1_tsc_offset = vmx_write_l1_tsc_offset, .set_tdp_cr3 = vmx_set_cr3, diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 7e4be1f2f253..d02937760c3b 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -1665,8 +1665,7 @@ EXPORT_SYMBOL_GPL(kvm_read_l1_tsc); static void kvm_vcpu_write_tsc_offset(struct kvm_vcpu *vcpu, u64 offset) { - kvm_x86_ops->write_tsc_offset(vcpu, offset); - vcpu->arch.tsc_offset = offset; + vcpu->arch.tsc_offset = kvm_x86_ops->write_l1_tsc_offset(vcpu, offset); } static inline bool kvm_check_tsc_unstable(void) @@ -1794,7 +1793,8 @@ EXPORT_SYMBOL_GPL(kvm_write_tsc); static inline void adjust_tsc_offset_guest(struct kvm_vcpu *vcpu, s64 adjustment) { - kvm_vcpu_write_tsc_offset(vcpu, vcpu->arch.tsc_offset + adjustment); + u64 tsc_offset = kvm_x86_ops->read_l1_tsc_offset(vcpu); + kvm_vcpu_write_tsc_offset(vcpu, tsc_offset + adjustment); } static inline void adjust_tsc_offset_host(struct kvm_vcpu *vcpu, s64 adjustment) -- cgit From fd65d3142f734bc4376053c8d75670041903134d Mon Sep 17 00:00:00 2001 From: Jim Mattson Date: Tue, 22 May 2018 09:54:20 -0700 Subject: kvm: svm: Ensure an IBPB on all affected CPUs when freeing a vmcb Previously, we only called indirect_branch_prediction_barrier on the logical CPU that freed a vmcb. This function should be called on all logical CPUs that last loaded the vmcb in question. Fixes: 15d45071523d ("KVM/x86: Add IBPB support") Reported-by: Neel Natu Signed-off-by: Jim Mattson Reviewed-by: Konrad Rzeszutek Wilk Cc: stable@vger.kernel.org Signed-off-by: Paolo Bonzini --- arch/x86/kvm/svm.c | 20 +++++++++++++++----- 1 file changed, 15 insertions(+), 5 deletions(-) (limited to 'arch/x86/kvm/svm.c') diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c index a24733aade4c..cc6467b35a85 100644 --- a/arch/x86/kvm/svm.c +++ b/arch/x86/kvm/svm.c @@ -2193,21 +2193,31 @@ out: return ERR_PTR(err); } +static void svm_clear_current_vmcb(struct vmcb *vmcb) +{ + int i; + + for_each_online_cpu(i) + cmpxchg(&per_cpu(svm_data, i)->current_vmcb, vmcb, NULL); +} + static void svm_free_vcpu(struct kvm_vcpu *vcpu) { struct vcpu_svm *svm = to_svm(vcpu); + /* + * The vmcb page can be recycled, causing a false negative in + * svm_vcpu_load(). So, ensure that no logical CPU has this + * vmcb page recorded as its current vmcb. + */ + svm_clear_current_vmcb(svm->vmcb); + __free_page(pfn_to_page(__sme_clr(svm->vmcb_pa) >> PAGE_SHIFT)); __free_pages(virt_to_page(svm->msrpm), MSRPM_ALLOC_ORDER); __free_page(virt_to_page(svm->nested.hsave)); __free_pages(virt_to_page(svm->nested.msrpm), MSRPM_ALLOC_ORDER); kvm_vcpu_uninit(vcpu); kmem_cache_free(kvm_vcpu_cache, svm); - /* - * The vmcb page can be recycled, causing a false negative in - * svm_vcpu_load(). So do a full IBPB now. - */ - indirect_branch_prediction_barrier(); } static void svm_vcpu_load(struct kvm_vcpu *vcpu, int cpu) -- cgit