diff options
Diffstat (limited to 'arch/x86/kvm/vmx.c')
-rw-r--r-- | arch/x86/kvm/vmx.c | 194 |
1 files changed, 158 insertions, 36 deletions
diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c index a0d6bd9ad442..d29216c462b3 100644 --- a/arch/x86/kvm/vmx.c +++ b/arch/x86/kvm/vmx.c @@ -18,6 +18,7 @@ #include "irq.h" #include "mmu.h" +#include "cpuid.h" #include <linux/kvm_host.h> #include <linux/module.h> @@ -39,6 +40,7 @@ #include <asm/mce.h> #include <asm/i387.h> #include <asm/xcr.h> +#include <asm/perf_event.h> #include "trace.h" @@ -49,29 +51,29 @@ MODULE_AUTHOR("Qumranet"); MODULE_LICENSE("GPL"); -static int __read_mostly enable_vpid = 1; +static bool __read_mostly enable_vpid = 1; module_param_named(vpid, enable_vpid, bool, 0444); -static int __read_mostly flexpriority_enabled = 1; +static bool __read_mostly flexpriority_enabled = 1; module_param_named(flexpriority, flexpriority_enabled, bool, S_IRUGO); -static int __read_mostly enable_ept = 1; +static bool __read_mostly enable_ept = 1; module_param_named(ept, enable_ept, bool, S_IRUGO); -static int __read_mostly enable_unrestricted_guest = 1; +static bool __read_mostly enable_unrestricted_guest = 1; module_param_named(unrestricted_guest, enable_unrestricted_guest, bool, S_IRUGO); -static int __read_mostly emulate_invalid_guest_state = 0; +static bool __read_mostly emulate_invalid_guest_state = 0; module_param(emulate_invalid_guest_state, bool, S_IRUGO); -static int __read_mostly vmm_exclusive = 1; +static bool __read_mostly vmm_exclusive = 1; module_param(vmm_exclusive, bool, S_IRUGO); -static int __read_mostly yield_on_hlt = 1; +static bool __read_mostly yield_on_hlt = 1; module_param(yield_on_hlt, bool, S_IRUGO); -static int __read_mostly fasteoi = 1; +static bool __read_mostly fasteoi = 1; module_param(fasteoi, bool, S_IRUGO); /* @@ -79,7 +81,7 @@ module_param(fasteoi, bool, S_IRUGO); * VMX and be a hypervisor for its own guests. If nested=0, guests may not * use VMX instructions. */ -static int __read_mostly nested = 0; +static bool __read_mostly nested = 0; module_param(nested, bool, S_IRUGO); #define KVM_GUEST_CR0_MASK_UNRESTRICTED_GUEST \ @@ -118,7 +120,7 @@ module_param(ple_gap, int, S_IRUGO); static int ple_window = KVM_VMX_DEFAULT_PLE_WINDOW; module_param(ple_window, int, S_IRUGO); -#define NR_AUTOLOAD_MSRS 1 +#define NR_AUTOLOAD_MSRS 8 #define VMCS02_POOL_SIZE 1 struct vmcs { @@ -622,6 +624,7 @@ static unsigned long *vmx_msr_bitmap_legacy; static unsigned long *vmx_msr_bitmap_longmode; static bool cpu_has_load_ia32_efer; +static bool cpu_has_load_perf_global_ctrl; static DECLARE_BITMAP(vmx_vpid_bitmap, VMX_NR_VPIDS); static DEFINE_SPINLOCK(vmx_vpid_lock); @@ -1191,15 +1194,34 @@ static void update_exception_bitmap(struct kvm_vcpu *vcpu) vmcs_write32(EXCEPTION_BITMAP, eb); } +static void clear_atomic_switch_msr_special(unsigned long entry, + unsigned long exit) +{ + vmcs_clear_bits(VM_ENTRY_CONTROLS, entry); + vmcs_clear_bits(VM_EXIT_CONTROLS, exit); +} + static void clear_atomic_switch_msr(struct vcpu_vmx *vmx, unsigned msr) { unsigned i; struct msr_autoload *m = &vmx->msr_autoload; - if (msr == MSR_EFER && cpu_has_load_ia32_efer) { - vmcs_clear_bits(VM_ENTRY_CONTROLS, VM_ENTRY_LOAD_IA32_EFER); - vmcs_clear_bits(VM_EXIT_CONTROLS, VM_EXIT_LOAD_IA32_EFER); - return; + switch (msr) { + case MSR_EFER: + if (cpu_has_load_ia32_efer) { + clear_atomic_switch_msr_special(VM_ENTRY_LOAD_IA32_EFER, + VM_EXIT_LOAD_IA32_EFER); + return; + } + break; + case MSR_CORE_PERF_GLOBAL_CTRL: + if (cpu_has_load_perf_global_ctrl) { + clear_atomic_switch_msr_special( + VM_ENTRY_LOAD_IA32_PERF_GLOBAL_CTRL, + VM_EXIT_LOAD_IA32_PERF_GLOBAL_CTRL); + return; + } + break; } for (i = 0; i < m->nr; ++i) @@ -1215,25 +1237,55 @@ static void clear_atomic_switch_msr(struct vcpu_vmx *vmx, unsigned msr) vmcs_write32(VM_EXIT_MSR_LOAD_COUNT, m->nr); } +static void add_atomic_switch_msr_special(unsigned long entry, + unsigned long exit, unsigned long guest_val_vmcs, + unsigned long host_val_vmcs, u64 guest_val, u64 host_val) +{ + vmcs_write64(guest_val_vmcs, guest_val); + vmcs_write64(host_val_vmcs, host_val); + vmcs_set_bits(VM_ENTRY_CONTROLS, entry); + vmcs_set_bits(VM_EXIT_CONTROLS, exit); +} + static void add_atomic_switch_msr(struct vcpu_vmx *vmx, unsigned msr, u64 guest_val, u64 host_val) { unsigned i; struct msr_autoload *m = &vmx->msr_autoload; - if (msr == MSR_EFER && cpu_has_load_ia32_efer) { - vmcs_write64(GUEST_IA32_EFER, guest_val); - vmcs_write64(HOST_IA32_EFER, host_val); - vmcs_set_bits(VM_ENTRY_CONTROLS, VM_ENTRY_LOAD_IA32_EFER); - vmcs_set_bits(VM_EXIT_CONTROLS, VM_EXIT_LOAD_IA32_EFER); - return; + switch (msr) { + case MSR_EFER: + if (cpu_has_load_ia32_efer) { + add_atomic_switch_msr_special(VM_ENTRY_LOAD_IA32_EFER, + VM_EXIT_LOAD_IA32_EFER, + GUEST_IA32_EFER, + HOST_IA32_EFER, + guest_val, host_val); + return; + } + break; + case MSR_CORE_PERF_GLOBAL_CTRL: + if (cpu_has_load_perf_global_ctrl) { + add_atomic_switch_msr_special( + VM_ENTRY_LOAD_IA32_PERF_GLOBAL_CTRL, + VM_EXIT_LOAD_IA32_PERF_GLOBAL_CTRL, + GUEST_IA32_PERF_GLOBAL_CTRL, + HOST_IA32_PERF_GLOBAL_CTRL, + guest_val, host_val); + return; + } + break; } for (i = 0; i < m->nr; ++i) if (m->guest[i].index == msr) break; - if (i == m->nr) { + if (i == NR_AUTOLOAD_MSRS) { + printk_once(KERN_WARNING"Not enough mst switch entries. " + "Can't add msr %x\n", msr); + return; + } else if (i == m->nr) { ++m->nr; vmcs_write32(VM_ENTRY_MSR_LOAD_COUNT, m->nr); vmcs_write32(VM_EXIT_MSR_LOAD_COUNT, m->nr); @@ -1696,7 +1748,6 @@ static void setup_msrs(struct vcpu_vmx *vmx) int save_nmsrs, index; unsigned long *msr_bitmap; - vmx_load_host_state(vmx); save_nmsrs = 0; #ifdef CONFIG_X86_64 if (is_long_mode(&vmx->vcpu)) { @@ -1905,6 +1956,7 @@ static __init void nested_vmx_setup_ctls_msrs(void) #endif CPU_BASED_MOV_DR_EXITING | CPU_BASED_UNCOND_IO_EXITING | CPU_BASED_USE_IO_BITMAPS | CPU_BASED_MONITOR_EXITING | + CPU_BASED_RDPMC_EXITING | CPU_BASED_ACTIVATE_SECONDARY_CONTROLS; /* * We can allow some features even when not supported by the @@ -2091,12 +2143,10 @@ static int vmx_get_msr(struct kvm_vcpu *vcpu, u32 msr_index, u64 *pdata) return 1; /* Otherwise falls through */ default: - vmx_load_host_state(to_vmx(vcpu)); if (vmx_get_vmx_msr(vcpu, msr_index, pdata)) return 0; msr = find_msr_entry(to_vmx(vcpu), msr_index); if (msr) { - vmx_load_host_state(to_vmx(vcpu)); data = msr->data; break; } @@ -2120,7 +2170,6 @@ static int vmx_set_msr(struct kvm_vcpu *vcpu, u32 msr_index, u64 data) switch (msr_index) { case MSR_EFER: - vmx_load_host_state(vmx); ret = kvm_set_msr_common(vcpu, msr_index, data); break; #ifdef CONFIG_X86_64 @@ -2169,7 +2218,6 @@ static int vmx_set_msr(struct kvm_vcpu *vcpu, u32 msr_index, u64 data) break; msr = find_msr_entry(vmx, msr_index); if (msr) { - vmx_load_host_state(vmx); msr->data = data; break; } @@ -2363,7 +2411,8 @@ static __init int setup_vmcs_config(struct vmcs_config *vmcs_conf) CPU_BASED_USE_TSC_OFFSETING | CPU_BASED_MWAIT_EXITING | CPU_BASED_MONITOR_EXITING | - CPU_BASED_INVLPG_EXITING; + CPU_BASED_INVLPG_EXITING | + CPU_BASED_RDPMC_EXITING; if (yield_on_hlt) min |= CPU_BASED_HLT_EXITING; @@ -2455,6 +2504,42 @@ static __init int setup_vmcs_config(struct vmcs_config *vmcs_conf) && allow_1_setting(MSR_IA32_VMX_EXIT_CTLS, VM_EXIT_LOAD_IA32_EFER); + cpu_has_load_perf_global_ctrl = + allow_1_setting(MSR_IA32_VMX_ENTRY_CTLS, + VM_ENTRY_LOAD_IA32_PERF_GLOBAL_CTRL) + && allow_1_setting(MSR_IA32_VMX_EXIT_CTLS, + VM_EXIT_LOAD_IA32_PERF_GLOBAL_CTRL); + + /* + * Some cpus support VM_ENTRY_(LOAD|SAVE)_IA32_PERF_GLOBAL_CTRL + * but due to arrata below it can't be used. Workaround is to use + * msr load mechanism to switch IA32_PERF_GLOBAL_CTRL. + * + * VM Exit May Incorrectly Clear IA32_PERF_GLOBAL_CTRL [34:32] + * + * AAK155 (model 26) + * AAP115 (model 30) + * AAT100 (model 37) + * BC86,AAY89,BD102 (model 44) + * BA97 (model 46) + * + */ + if (cpu_has_load_perf_global_ctrl && boot_cpu_data.x86 == 0x6) { + switch (boot_cpu_data.x86_model) { + case 26: + case 30: + case 37: + case 44: + case 46: + cpu_has_load_perf_global_ctrl = false; + printk_once(KERN_WARNING"kvm: VM_EXIT_LOAD_IA32_PERF_GLOBAL_CTRL " + "does not work properly. Using workaround\n"); + break; + default: + break; + } + } + return 0; } @@ -2629,11 +2714,13 @@ static gva_t rmode_tss_base(struct kvm *kvm) { if (!kvm->arch.tss_addr) { struct kvm_memslots *slots; + struct kvm_memory_slot *slot; gfn_t base_gfn; slots = kvm_memslots(kvm); - base_gfn = slots->memslots[0].base_gfn + - kvm->memslots->memslots[0].npages - 3; + slot = id_to_memslot(slots, 0); + base_gfn = slot->base_gfn + slot->npages - 3; + return base_gfn << PAGE_SHIFT; } return kvm->arch.tss_addr; @@ -3858,12 +3945,15 @@ static bool nested_exit_on_intr(struct kvm_vcpu *vcpu) static void enable_irq_window(struct kvm_vcpu *vcpu) { u32 cpu_based_vm_exec_control; - if (is_guest_mode(vcpu) && nested_exit_on_intr(vcpu)) - /* We can get here when nested_run_pending caused - * vmx_interrupt_allowed() to return false. In this case, do - * nothing - the interrupt will be injected later. + if (is_guest_mode(vcpu) && nested_exit_on_intr(vcpu)) { + /* + * We get here if vmx_interrupt_allowed() said we can't + * inject to L1 now because L2 must run. Ask L2 to exit + * right after entry, so we can inject to L1 more promptly. */ + kvm_make_request(KVM_REQ_IMMEDIATE_EXIT, vcpu); return; + } cpu_based_vm_exec_control = vmcs_read32(CPU_BASED_VM_EXEC_CONTROL); cpu_based_vm_exec_control |= CPU_BASED_VIRTUAL_INTR_PENDING; @@ -3990,11 +4080,12 @@ static void vmx_set_nmi_mask(struct kvm_vcpu *vcpu, bool masked) static int vmx_interrupt_allowed(struct kvm_vcpu *vcpu) { if (is_guest_mode(vcpu) && nested_exit_on_intr(vcpu)) { - struct vmcs12 *vmcs12; - if (to_vmx(vcpu)->nested.nested_run_pending) + struct vmcs12 *vmcs12 = get_vmcs12(vcpu); + if (to_vmx(vcpu)->nested.nested_run_pending || + (vmcs12->idt_vectoring_info_field & + VECTORING_INFO_VALID_MASK)) return 0; nested_vmx_vmexit(vcpu); - vmcs12 = get_vmcs12(vcpu); vmcs12->vm_exit_reason = EXIT_REASON_EXTERNAL_INTERRUPT; vmcs12->vm_exit_intr_info = 0; /* fall through to normal code, but now in L1, not L2 */ @@ -4524,6 +4615,16 @@ static int handle_invlpg(struct kvm_vcpu *vcpu) return 1; } +static int handle_rdpmc(struct kvm_vcpu *vcpu) +{ + int err; + + err = kvm_rdpmc(vcpu); + kvm_complete_insn_gp(vcpu, err); + + return 1; +} + static int handle_wbinvd(struct kvm_vcpu *vcpu) { skip_emulated_instruction(vcpu); @@ -5474,6 +5575,7 @@ static int (*kvm_vmx_exit_handlers[])(struct kvm_vcpu *vcpu) = { [EXIT_REASON_HLT] = handle_halt, [EXIT_REASON_INVD] = handle_invd, [EXIT_REASON_INVLPG] = handle_invlpg, + [EXIT_REASON_RDPMC] = handle_rdpmc, [EXIT_REASON_VMCALL] = handle_vmcall, [EXIT_REASON_VMCLEAR] = handle_vmclear, [EXIT_REASON_VMLAUNCH] = handle_vmlaunch, @@ -5968,6 +6070,24 @@ static void vmx_cancel_injection(struct kvm_vcpu *vcpu) vmcs_write32(VM_ENTRY_INTR_INFO_FIELD, 0); } +static void atomic_switch_perf_msrs(struct vcpu_vmx *vmx) +{ + int i, nr_msrs; + struct perf_guest_switch_msr *msrs; + + msrs = perf_guest_get_msrs(&nr_msrs); + + if (!msrs) + return; + + for (i = 0; i < nr_msrs; i++) + if (msrs[i].host == msrs[i].guest) + clear_atomic_switch_msr(vmx, msrs[i].msr); + else + add_atomic_switch_msr(vmx, msrs[i].msr, msrs[i].guest, + msrs[i].host); +} + #ifdef CONFIG_X86_64 #define R "r" #define Q "q" @@ -6017,6 +6137,8 @@ static void __noclone vmx_vcpu_run(struct kvm_vcpu *vcpu) if (vcpu->guest_debug & KVM_GUESTDBG_SINGLESTEP) vmx_set_interrupt_shadow(vcpu, 0); + atomic_switch_perf_msrs(vmx); + vmx->__launched = vmx->loaded_vmcs->launched; asm( /* Store host registers */ |