KVM: Support Intel Key Locker
cover KVM: Support Intel Key Locker
This patch set is to support KVM virtualization of Key Locker feature [1][2].
Key Locker provides a mechanism to encrypt and decrypt data without directly providing AES key but a "handle" instead.
Handles are essentially an encrypted form of those underlying real AES keys. "IWKey (Internal Wrapping Key)", loaded inside CPU, inaccessible from outside, is the key used to seal real AES Keys into handles. Thus, a real AES Key exists in memory for only a short period of time, when user is requesting a 'handle' from it. After that, the real AES Key can be erased, user then uses handle, with new Key Locker instructions, to perform AES encryption/decryption. By OS policy, usually, handles will be revoked after reboot, then any handles that may have been stolen should no longer be useful to the attacker after the reboot.
IWKey, is the core of this framework. It is loaded into CPU by LOADIWKEY instruction, then inaccessible from CPU-outside anymore. LOADIWKEY is the only Key Locker instruction that will cause VM-exit (if we set so in VM Execution Control). When load IWKey into CPU, we can ask CPU further randomize it, if HW supports this.
The IWKey can also be distributed among CPUs, rather than LOADIWKEY on each CPU, by: first backup IWKey to platform specific storage, then copy it on target CPU. The backup process is triggered by writing to MSR IA32_COPY_LOCAL_TO_PLATFORM. The restore process is triggered by writing to MSR IA32_COPY_PLATFORM_LOCAL.
Virtualization Design
Key Locker Spec [2] indicates virtualization limitations by current HW implementation.
1) IWKey cannot be read from CPU after it's loaded (this is the nature of this feature) and only 1 copy of IWKey inside 1 CPU.
2) Initial implementations may take a significant amount of time to perform a copy of IWKeyBackup to IWKey (via a write to MSR IA32_COPY_PLATFORM_LOCAL) so it may cause a significant performance impact to reload IWKey after each VM exit.
Due to above reasons, virtualization design makes below decisions
1) don't expose HW randomize IWKey capability (CPUID.0x19.ECX[1]) to guest. As such, guest IWKey cannot be preserved by VMM across vCPU switch. (VMM cannot know what IWKey is set in physical CPU if HW randomized.)
2) guests and host can only use Key Locker feature exclusively. [4]
The virtualization implementation is generally straight forward
1) On VM-Exit of guest 'LOADIWKEY', VMM stores the IWKey of that vCPU, and set in physical CPU on behalf.
2) On each vCPU load/put, VMM is responsible to set IWKEY to pCPU or clear it.
3) On guest backup local to platform operation, VMM traps the write to MSR, and simulate the IWKey store process by store it in a KVM scope area (kvm_arch), mark the success status in the shadow msr_ia32_iwkey_backup_status and msr_ia32_copy_status.
4) On guest copy from platform to local operation, VMM traps the write to MSR and simulate the process by load kvm_arch.iwkey_backup to vCPU.iwkey; and simulate the success status in the
shadow msr_ia32_copy_status.
5) Guest read the 2 status MSRs will also be trapped and return the shadow value.
6) Other Key Locker instructions can run without VM-Exit in non-root mode.
To the end, we don't suggest this feature to be migratable, as if so, IWKey would have to be exposed to user space, which would further weaken this feature's security significance.
P.S. this patch set is based on Chang's Kernel Key Locker upstream-v2
https://lore.kernel.org/lkml/[email protected]/
Patch 1 ~ 3, x86 code, which lay the foundation of following KVM patches. Have been reviewed by Tony Luck.
Patch 4 ~ 9, KVM enabling of Key Locker.
Patch 10 ~ 15, nested VMX support for Key Locker.
[1] Intel Architecture Instruction Set Extensions Programming Reference:
https://software.intel.com/content/www/us/en/develop/download/intel-architecture-instruction-set-extensions-programming-reference.html
[2] Intel Key Locker Specification:
https://software.intel.com/content/www/us/en/develop/download/intel-key-locker-specification.html
[3] Kernel enablement patch:
https://lore.kernel.org/lkml/[email protected]/
[4] It's possible to use Key Locker by both the guest and host, albeit with reduced security benefits. I.e., store host IWKey in VMM scoped place (memory/register), VMM switches host-IWKey and guest-IWKey between VM-{Exit/Entry} by LOADIWKEY instruction.
But in this case, an adversary that can observe arbitrary VMM memory may be able to steal both the handles and IWKey. And this case also require the VMM to be running before the first IWKey load.
---
Change log since last RFC patch set:
* Don't loadiwkey every VM-Exit/Entry. (Sean Christopherson)
* Don't enable Tertiary VM-Exec Control if using eVMCS, as its support in enlightened VMCS isn't ready yet. (Vitaly Kuznetsov)
* Extended BUILD_CONTROLS_SHADOW to support 64-bit variation. (Sean Christopherson)
* Refactored patch set and other changes.
Hu, Robert (2):
kvm/vmx: Detect Tertiary VM-Execution control when setup VMCS config kvm/vmx: dump_vmcs() reports tertiary_exec_control field as well
Robert Hoo (13):
x86/keylocker: Move KEYSRC_{SW,HW}RAND to keylocker.h x86/cpufeatures: Define Key Locker sub feature flags x86/feat_ctl: Add new VMX feature, Tertiary VM-Execution control and
LOADIWKEY Exiting
kvm/vmx: Extend BUILD_CONTROLS_SHADOW macro to support 64-bit
variation
kvm/vmx: Set Tertiary VM-Execution control field When init vCPU's VMCS kvm/vmx: Add KVM support on guest Key Locker operations kvm/cpuid: Enumerate Key Locker feature in KVM kvm/vmx/nested: Support new IA32_VMX_PROCBASED_CTLS3 vmx capability
MSR
kvm/vmx: Implement vmx_compute_tertiary_exec_control() kvm/vmx/vmcs12: Add Tertiary VM-Exec control field in vmcs12 kvm/vmx/nested: Support Tertiary VM-Exec control in vmcs02 kvm/vmx/nested: Support CR4.KL in nested
kvm/vmx/nested: Enable nested LOADIWKEY VM-exit
arch/x86/include/asm/cpufeatures.h | 5 + arch/x86/include/asm/keylocker.h | 3 + arch/x86/include/asm/kvm_host.h | 24 ++- arch/x86/include/asm/msr-index.h | 1 + arch/x86/include/asm/vmx.h | 9 ++ arch/x86/include/asm/vmxfeatures.h | 6 +- arch/x86/include/uapi/asm/vmx.h | 2 + arch/x86/kernel/cpu/feat_ctl.c | 9 ++ arch/x86/kernel/cpu/scattered.c | 5 + arch/x86/kernel/keylocker.c | 2 - arch/x86/kvm/cpuid.c | 26 +++- arch/x86/kvm/reverse_cpuid.h | 32 +++- arch/x86/kvm/vmx/capabilities.h | 9 ++ arch/x86/kvm/vmx/evmcs.c | 2 + arch/x86/kvm/vmx/evmcs.h | 1 + arch/x86/kvm/vmx/nested.c | 38 ++++- arch/x86/kvm/vmx/nested.h | 7 + arch/x86/kvm/vmx/vmcs.h | 1 + arch/x86/kvm/vmx/vmcs12.c | 1 + arch/x86/kvm/vmx/vmcs12.h | 4 +- arch/x86/kvm/vmx/vmx.c | 290 ++++++++++++++++++++++++++++++++++++- arch/x86/kvm/vmx/vmx.h | 24 +-- arch/x86/kvm/x86.c | 2 + arch/x86/kvm/x86.h | 2 + 24 files changed, 475 insertions(+), 30 deletions(-)
Signature · 1 line
1.8.3.1
PATCH 1/15 x86/keylocker: Move KEYSRC_{SW,HW}RAND to keylocker.h
KVM needs the KEYSRC_SWRAND and KEYSRC_HWRAND macro definitions. Move them to <asm/keylocker.h>
--- arch/x86/include/asm/keylocker.h | 3 +++ arch/x86/kernel/keylocker.c | 2 -- 2 files changed, 3 insertions(+), 2 deletions(-)
arch/x86/include/asm/keylocker.h +3 -0
@@ -9,6 +9,9 @@#include <linux/bits.h> #include <asm/msr.h> +#define KEYSRC_SWRAND 0 +#define KEYSRC_HWRAND BIT(1) +#define KEYLOCKER_CPUID 0x019#define KEYLOCKER_CPUID_EAX_SUPERVISOR BIT(0)#define KEYLOCKER_CPUID_EBX_AESKLE BIT(0)
arch/x86/kernel/keylocker.c +0 -2
@@ -66,8 +66,6 @@ void flush_keylocker_data(void) keydata.valid = false; } -#define KEYSRC_SWRAND 0 -#define KEYSRC_HWRAND BIT(1)#define KEYSRC_HWRAND_RETRY 10/**
Signature · 1 line
1.8.3.1
PATCH 2/15 x86/cpufeatures: Define Key Locker sub feature flags
In reply to Robert HooThough KeyLocker is generally enumerated by CPUID.(07H,0):ECX.KL[bit23], CPUID.19H:{EBX,ECX} enumerate more details of KeyLocker supporting status. Define them in scattered cpuid bits.
CPUID.19H:EBX
bit0 enumerates if OS (CR4.KeyLocker) and BIOS have enabled KeyLocker. bit2 enumerates if wide Key Locker instructions are supported. bit4 enumerates if IWKey backup is supported.
CPUID.19H:ECX
bit0 enumerates if the NoBackup parameter to LOADIWKEY is supported. bit1 enumerates if IWKey randomization is supported.
Most of above features don't necessarily appear in /proc/cpuinfo, except "iwkey_rand", which we think might be interesting to indicate that the system supports randomized IWKey.
--- arch/x86/include/asm/cpufeatures.h | 5 +++++ arch/x86/kernel/cpu/scattered.c | 5 +++++ 2 files changed, 10 insertions(+)
arch/x86/include/asm/cpufeatures.h +5 -0
@@ -294,6 +294,11 @@#define X86_FEATURE_PER_THREAD_MBA (11*32+ 7)/* "" Per-thread Memory Bandwidth Allocation */ #define X86_FEATURE_SGX1 (11*32+ 8)/* "" Basic SGX */ #define X86_FEATURE_SGX2 (11*32+ 9)/* "" SGX Enclave Dynamic Memory Management (EDMM) */ +#define X86_FEATURE_KL_INS_ENABLED (11*32 + 10)/* "" Key Locker instructions */ +#define X86_FEATURE_KL_WIDE (11*32 + 11)/* "" Wide Key Locker instructions */ +#define X86_FEATURE_IWKEY_BACKUP (11*32 + 12)/* "" IWKey backup */ +#define X86_FEATURE_IWKEY_NOBACKUP (11*32 + 13)/* "" NoBackup parameter to LOADIWKEY */ +#define X86_FEATURE_IWKEY_RAND (11*32 + 14)/* IWKey Randomization */ /* Intel-defined CPU features, CPUID level 0x00000007:1 (EAX), word 12 */ #define X86_FEATURE_AVX_VNNI (12*32+ 4)/* AVX VNNI instructions */
arch/x86/kernel/cpu/scattered.c +5 -0
@@ -38,6 +38,11 @@ struct cpuid_bit { { X86_FEATURE_PER_THREAD_MBA, CPUID_ECX, 0, 0x00000010, 3 }, { X86_FEATURE_SGX1, CPUID_EAX, 0, 0x00000012, 0 }, { X86_FEATURE_SGX2, CPUID_EAX, 1, 0x00000012, 0 }, + { X86_FEATURE_KL_INS_ENABLED, CPUID_EBX, 0, 0x00000019, 0 }, + { X86_FEATURE_KL_WIDE, CPUID_EBX, 2, 0x00000019, 0 }, + { X86_FEATURE_IWKEY_BACKUP, CPUID_EBX, 4, 0x00000019, 0 }, + { X86_FEATURE_IWKEY_NOBACKUP, CPUID_ECX, 0, 0x00000019, 0 }, + { X86_FEATURE_IWKEY_RAND, CPUID_ECX, 1, 0x00000019, 0 }, { X86_FEATURE_HW_PSTATE, CPUID_EDX, 7, 0x80000007, 0 }, { X86_FEATURE_CPB, CPUID_EDX, 9, 0x80000007, 0 }, { X86_FEATURE_PROC_FEEDBACK, CPUID_EDX, 11, 0x80000007, 0 },
Signature · 1 line
1.8.3.1
PATCH 3/15 x86/feat_ctl: Add new VMX feature, Tertiary VM-Execution control and LOADIWKEY Exiting
In reply to Robert HooThere is a new VMX capability MSR IA32_VMX_PROCBASED_CTLS3.
All 64 bits of this MSR define capability bits for the new tertiary VM-Exec control so two new 32-bit vmx_feature leaves are needed to record all the capabilities.
The 2 new VMX features:
Tertiary VM-Execution control is enumerated by bit 17 of existing Primary VM-Execution control.
LOADIWKEY Exiting is enumerated by bit 0 of this new tertiary VM-Exec control, which designates if guest running 'loadiwkey' instruction will cause VM-Exit.
--- arch/x86/include/asm/msr-index.h | 1 + arch/x86/include/asm/vmxfeatures.h | 6 +++++- arch/x86/kernel/cpu/feat_ctl.c | 9 +++++++++ 3 files changed, 15 insertions(+), 1 deletion(-)
arch/x86/include/asm/msr-index.h +1 -0
@@ -915,6 +915,7 @@#define MSR_IA32_VMX_TRUE_EXIT_CTLS 0x0000048f#define MSR_IA32_VMX_TRUE_ENTRY_CTLS 0x00000490#define MSR_IA32_VMX_VMFUNC 0x00000491 +#define MSR_IA32_VMX_PROCBASED_CTLS3 0x00000492/* VMX_BASIC bits and bitmasks */ #define VMX_BASIC_VMCS_SIZE_SHIFT 32
arch/x86/include/asm/vmxfeatures.h +5 -1
@@ -5,7 +5,7 @@/* * Defines VMX CPU feature bits */ -#define NVMXINTS 3/* N 32-bit words worth of info */ +#define NVMXINTS 5/* N 32-bit words worth of info */ /* * Note: If the comment begins with a quoted string, that string is used
@@ -43,6 +43,7 @@#define VMX_FEATURE_RDTSC_EXITING ( 1*32+ 12)/* "" VM-Exit on RDTSC */ #define VMX_FEATURE_CR3_LOAD_EXITING ( 1*32+ 15)/* "" VM-Exit on writes to CR3 */ #define VMX_FEATURE_CR3_STORE_EXITING ( 1*32+ 16)/* "" VM-Exit on reads from CR3 */ +#define VMX_FEATURE_TER_CONTROLS (1*32 + 17)/* "" Enable Tertiary VM-Execution Controls */ #define VMX_FEATURE_CR8_LOAD_EXITING ( 1*32+ 19)/* "" VM-Exit on writes to CR8 */ #define VMX_FEATURE_CR8_STORE_EXITING ( 1*32+ 20)/* "" VM-Exit on reads from CR8 */ #define VMX_FEATURE_VIRTUAL_TPR ( 1*32+ 21)/* "vtpr" TPR virtualization, a.k.a. TPR shadow */
@@ -85,4 +86,7 @@#define VMX_FEATURE_ENCLV_EXITING ( 2*32+ 28)/* "" VM-Exit on ENCLV (leaf dependent) */ #define VMX_FEATURE_BUS_LOCK_DETECTION ( 2*32+ 30)/* "" VM-Exit when bus lock caused */ +/* Tertiary Processor-Based VM-Execution Controls, word 3 */ +#define VMX_FEATURE_LOADIWKEY_EXITING (3*32 + 0)/* "" VM-Exit on LOADIWKey */ +#endif /* _ASM_X86_VMXFEATURES_H */
arch/x86/kernel/cpu/feat_ctl.c +9 -0
@@ -15,6 +15,8 @@ enum vmx_feature_leafs {MISC_FEATURES =0 ,PRIMARY_CTLS ,SECONDARY_CTLS , +TERTIARY_CTLS_LOW , +TERTIARY_CTLS_HIGH , NR_VMX_FEATURE_WORDS, };
@@ -42,6 +44,13 @@ static void init_vmx_capabilities(struct cpuinfo_x86 *c)rdmsr_safe (MSR_IA32_VMX_PROCBASED_CTLS2 , &ign, &supported); c->vmx_capability[SECONDARY_CTLS ] = supported; +/* +* For tertiary execution controls MSR, it's actually a 64bit allowed-1. +*/ +rdmsr_safe (MSR_IA32_VMX_PROCBASED_CTLS3 , &ign, &supported); + c->vmx_capability[TERTIARY_CTLS_LOW ] = ign; + c->vmx_capability[TERTIARY_CTLS_HIGH ] = supported; +rdmsr (MSR_IA32_VMX_PINBASED_CTLS , ign, supported);rdmsr_safe (MSR_IA32_VMX_VMFUNC , &ign, &funcs);
Signature · 1 line
1.8.3.1
PATCH 4/15 kvm/vmx: Detect Tertiary VM-Execution control when setup VMCS config
In reply to Robert HooFrom: "Hu, Robert" <[email protected]>
Add new Tertiary VM-Exec Control field for vmcs_config and related functions.
And when eVMCS in use, filter it out.
--- arch/x86/include/asm/vmx.h | 1 + arch/x86/kvm/vmx/capabilities.h | 7 +++++++ arch/x86/kvm/vmx/evmcs.c | 2 ++ arch/x86/kvm/vmx/evmcs.h | 1 + arch/x86/kvm/vmx/vmx.c | 5 ++++- 5 files changed, 15 insertions(+), 1 deletion(-)
arch/x86/include/asm/vmx.h +1 -0
@@ -31,6 +31,7 @@#define CPU_BASED_RDTSC_EXITING VMCS_CONTROL_BIT(RDTSC_EXITING)#define CPU_BASED_CR3_LOAD_EXITING VMCS_CONTROL_BIT(CR3_LOAD_EXITING)#define CPU_BASED_CR3_STORE_EXITING VMCS_CONTROL_BIT(CR3_STORE_EXITING) +#define CPU_BASED_ACTIVATE_TERTIARY_CONTROLS VMCS_CONTROL_BIT(TER_CONTROLS)#define CPU_BASED_CR8_LOAD_EXITING VMCS_CONTROL_BIT(CR8_LOAD_EXITING)#define CPU_BASED_CR8_STORE_EXITING VMCS_CONTROL_BIT(CR8_STORE_EXITING)#define CPU_BASED_TPR_SHADOW VMCS_CONTROL_BIT(VIRTUAL_TPR)
arch/x86/kvm/vmx/capabilities.h +7 -0
@@ -60,6 +60,7 @@ struct vmcs_config {u32 pin_based_exec_ctrl;u32 cpu_based_exec_ctrl;u32 cpu_based_2nd_exec_ctrl; +u64 cpu_based_3rd_exec_ctrl;u32 vmexit_ctrl;u32 vmentry_ctrl;struct nested_vmx_msrs nested;
@@ -133,6 +134,12 @@ static inline bool cpu_has_secondary_exec_ctrls(void)CPU_BASED_ACTIVATE_SECONDARY_CONTROLS ; } +static inline bool cpu_has_tertiary_exec_ctrls (void ) +{ +return vmcs_config.cpu_based_exec_ctrl & +CPU_BASED_ACTIVATE_TERTIARY_CONTROLS ; +} +static inline bool cpu_has_vmx_virtualize_apic_accesses (void ) { return vmcs_config.cpu_based_2nd_exec_ctrl &
arch/x86/kvm/vmx/evmcs.c +2 -0
@@ -299,8 +299,10 @@__init void evmcs_sanitize_exec_ctrls (struct vmcs_config *vmcs_conf) { + vmcs_conf->cpu_based_exec_ctrl &= ~EVMCS1_UNSUPPORTED_EXEC_CTRL ; vmcs_conf->pin_based_exec_ctrl &= ~EVMCS1_UNSUPPORTED_PINCTRL ; vmcs_conf->cpu_based_2nd_exec_ctrl &= ~EVMCS1_UNSUPPORTED_2NDEXEC ; + vmcs_conf->cpu_based_3rd_exec_ctrl =0 ; vmcs_conf->vmexit_ctrl &= ~EVMCS1_UNSUPPORTED_VMEXIT_CTRL ; vmcs_conf->vmentry_ctrl &= ~EVMCS1_UNSUPPORTED_VMENTRY_CTRL ;
arch/x86/kvm/vmx/evmcs.h +1 -0
@@ -50,6 +50,7 @@*/ #define EVMCS1_UNSUPPORTED_PINCTRL (PIN_BASED_POSTED_INTR | \ PIN_BASED_VMX_PREEMPTION_TIMER) +#define EVMCS1_UNSUPPORTED_EXEC_CTRL (CPU_BASED_ACTIVATE_TERTIARY_CONTROLS)#define EVMCS1_UNSUPPORTED_2NDEXEC \ (SECONDARY_EXEC_VIRTUAL_INTR_DELIVERY | \ SECONDARY_EXEC_VIRTUALIZE_APIC_ACCESSES | \
arch/x86/kvm/vmx/vmx.c +4 -1
@@ -2506,6 +2506,7 @@ static __init int setup_vmcs_config(struct vmcs_config *vmcs_conf,u32 _pin_based_exec_control =0 ;u32 _cpu_based_exec_control =0 ;u32 _cpu_based_2nd_exec_control =0 ; +u64 _cpu_based_3rd_exec_control =0 ;u32 _vmexit_control =0 ;u32 _vmentry_control =0 ;
@@ -2527,7 +2528,8 @@ static __init int setup_vmcs_config(struct vmcs_config *vmcs_conf, opt =CPU_BASED_TPR_SHADOW |CPU_BASED_USE_MSR_BITMAPS | -CPU_BASED_ACTIVATE_SECONDARY_CONTROLS ; +CPU_BASED_ACTIVATE_SECONDARY_CONTROLS | +CPU_BASED_ACTIVATE_TERTIARY_CONTROLS ;if (adjust_vmx_controls (min, opt,MSR_IA32_VMX_PROCBASED_CTLS , &_cpu_based_exec_control) <0 )return -EIO ;
@@ -2688,6 +2690,7 @@ static __init int setup_vmcs_config(struct vmcs_config *vmcs_conf, vmcs_conf->pin_based_exec_ctrl = _pin_based_exec_control; vmcs_conf->cpu_based_exec_ctrl = _cpu_based_exec_control; vmcs_conf->cpu_based_2nd_exec_ctrl = _cpu_based_2nd_exec_control; + vmcs_conf->cpu_based_3rd_exec_ctrl = _cpu_based_3rd_exec_control; vmcs_conf->vmexit_ctrl = _vmexit_control; vmcs_conf->vmentry_ctrl = _vmentry_control;
Signature · 1 line
1.8.3.1
PATCH 5/15 kvm/vmx: Extend BUILD_CONTROLS_SHADOW macro to support 64-bit variation
In reply to Robert HooThe Tertiary VM-Exec Control, different from previous control fields, is 64 bit. So extend BUILD_CONTROLS_SHADOW() by adding a 'bit' parameter, to support both 32 bit and 64 bit fields' auxiliary functions building. Also, define the auxiliary functions for Tertiary control field here, using the new BUILD_CONTROLS_SHADOW().
--- arch/x86/kvm/vmx/vmx.h | 23 ++++++++++++----------- 1 file changed, 12 insertions(+), 11 deletions(-)
arch/x86/kvm/vmx/vmx.h +12 -11
@@ -412,31 +412,32 @@ static inline u8 vmx_get_rvi(void)return vmcs_read16 (GUEST_INTR_STATUS ) &0xff ; } -#define BUILD_CONTROLS_SHADOW (lname, uname) \ -static inline void lname##_controls_set(struct vcpu_vmx *vmx, u32 val) \ +#define BUILD_CONTROLS_SHADOW (lname, uname, bits) \ +static inline void lname##_controls_set(struct vcpu_vmx *vmx, u##bits val) \ { \ if (vmx->loaded_vmcs->controls_shadow.lname != val) { \ - vmcs_write32(uname, val); \ + vmcs_write##bits(uname, val); \ vmx->loaded_vmcs->controls_shadow.lname = val; \ } \ } \ -static inline u32 lname##_controls_get(struct vcpu_vmx *vmx) \ +static inline u##bits lname##_controls_get(struct vcpu_vmx *vmx) \ { \ return vmx->loaded_vmcs->controls_shadow.lname; \ } \ -static inline void lname##_controls_setbit(struct vcpu_vmx *vmx, u32 val) \ +static inline void lname##_controls_setbit(struct vcpu_vmx *vmx, u##bits val) \ { \ lname##_controls_set(vmx, lname##_controls_get(vmx) | val); \ } \ -static inline void lname##_controls_clearbit(struct vcpu_vmx *vmx, u32 val) \ +static inline void lname##_controls_clearbit(struct vcpu_vmx *vmx, u##bits val) \ { \ lname##_controls_set(vmx, lname##_controls_get(vmx) & ~val); \ } -BUILD_CONTROLS_SHADOW(vm_entry, VM_ENTRY_CONTROLS) -BUILD_CONTROLS_SHADOW(vm_exit, VM_EXIT_CONTROLS) -BUILD_CONTROLS_SHADOW(pin, PIN_BASED_VM_EXEC_CONTROL) -BUILD_CONTROLS_SHADOW(exec, CPU_BASED_VM_EXEC_CONTROL) -BUILD_CONTROLS_SHADOW(secondary_exec, SECONDARY_VM_EXEC_CONTROL) +BUILD_CONTROLS_SHADOW(vm_entry, VM_ENTRY_CONTROLS, 32) +BUILD_CONTROLS_SHADOW(vm_exit, VM_EXIT_CONTROLS, 32) +BUILD_CONTROLS_SHADOW(pin, PIN_BASED_VM_EXEC_CONTROL, 32) +BUILD_CONTROLS_SHADOW(exec, CPU_BASED_VM_EXEC_CONTROL, 32) +BUILD_CONTROLS_SHADOW(secondary_exec, SECONDARY_VM_EXEC_CONTROL, 32) +BUILD_CONTROLS_SHADOW(tertiary_exec, TERTIARY_VM_EXEC_CONTROL, 64)static inline void vmx_register_cache_reset (struct kvm_vcpu *vcpu) {
Signature · 1 line
1.8.3.1
PATCH 6/15 kvm/vmx: Set Tertiary VM-Execution control field When init vCPU's VMCS
In reply to Robert HooDefine the new 64bit VMCS field, as well as its auxiliary set function. In init_vmcs(), set this field to the value previously configured. But in vmx_exec_control(), at this moment, returning vmcs_config.primary_exec_control with tertiary-exec artifially disabled, as till now no real feature on it is enabled. This will be removed later in Key Locker enablement patch.
--- arch/x86/include/asm/vmx.h | 2 ++ arch/x86/kvm/vmx/vmcs.h | 1 + arch/x86/kvm/vmx/vmx.c | 6 ++++++ 3 files changed, 9 insertions(+)
arch/x86/include/asm/vmx.h +2 -0
@@ -222,6 +222,8 @@ enum vmcs_field {ENCLS_EXITING_BITMAP_HIGH =0x0000202F ,TSC_MULTIPLIER =0x00002032 ,TSC_MULTIPLIER_HIGH =0x00002033 , +TERTIARY_VM_EXEC_CONTROL =0x00002034 , +TERTIARY_VM_EXEC_CONTROL_HIGH =0x00002035 ,GUEST_PHYSICAL_ADDRESS =0x00002400 ,GUEST_PHYSICAL_ADDRESS_HIGH =0x00002401 , VMCS_LINK_POINTER = 0x00002800,
arch/x86/kvm/vmx/vmcs.h +1 -0
@@ -48,6 +48,7 @@ struct vmcs_controls_shadow {u32 pin;u32 exec;u32 secondary_exec; +u64 tertiary_exec; };/*
arch/x86/kvm/vmx/vmx.c +6 -0
@@ -4255,6 +4255,9 @@ u32 vmx_exec_control(struct vcpu_vmx *vmx) CPU_BASED_MONITOR_EXITING);if (kvm_hlt_in_guest (vmx->vcpu.kvm)) exec_control &= ~CPU_BASED_HLT_EXITING ; + +/* Disable Tertiary-Exec Control at this moment, as no feature used yet */ + exec_control &= ~CPU_BASED_ACTIVATE_TERTIARY_CONTROLS ;return exec_control; }
@@ -4413,6 +4416,9 @@ static void init_vmcs(struct vcpu_vmx *vmx)secondary_exec_controls_set (vmx, vmx->secondary_exec_control); } +if (cpu_has_tertiary_exec_ctrls ()) +tertiary_exec_controls_set (vmx, vmcs_config.cpu_based_3rd_exec_ctrl); +if (kvm_vcpu_apicv_active (&vmx->vcpu)) {vmcs_write64 (EOI_EXIT_BITMAP0 ,0 );vmcs_write64 (EOI_EXIT_BITMAP1 ,0 );
Signature · 1 line
1.8.3.1
PATCH 7/15 kvm/vmx: dump_vmcs() reports tertiary_exec_control field as well
In reply to Robert HooFrom: "Hu, Robert" <[email protected]>
--- arch/x86/kvm/vmx/vmx.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)
arch/x86/kvm/vmx/vmx.c +7 -2
@@ -5808,6 +5808,7 @@ void dump_vmcs(struct kvm_vcpu *vcpu)struct vcpu_vmx *vmx =to_vmx (vcpu);u32 vmentry_ctl, vmexit_ctl;u32 cpu_based_exec_ctrl, pin_based_exec_ctrl, secondary_exec_control; +u64 tertiary_exec_control =0 ;unsigned long cr4;int efer_slot;
@@ -5825,6 +5826,9 @@ void dump_vmcs(struct kvm_vcpu *vcpu)if (cpu_has_secondary_exec_ctrls ()) secondary_exec_control =vmcs_read32 (SECONDARY_VM_EXEC_CONTROL ); +if (cpu_has_tertiary_exec_ctrls ()) + tertiary_exec_control =vmcs_read64 (TERTIARY_VM_EXEC_CONTROL ); +pr_err ("*** Guest State ***\n" ); pr_err("CR0: actual=0x%016lx, shadow=0x%016lx, gh_mask=%016lx\n", vmcs_readl(GUEST_CR0), vmcs_readl(CR0_READ_SHADOW),
@@ -5921,8 +5925,9 @@ void dump_vmcs(struct kvm_vcpu *vcpu)vmx_dump_msrs ("host autoload" , &vmx->msr_autoload.host);pr_err ("*** Control State ***\n" ); -pr_err ("PinBased=%08x CPUBased=%08x SecondaryExec=%08x\n" , - pin_based_exec_ctrl, cpu_based_exec_ctrl, secondary_exec_control); +pr_err ("PinBased=0x%08x CPUBased=0x%08x SecondaryExec=0x%08x TertiaryExec=0x%016llx\n" , + pin_based_exec_ctrl, cpu_based_exec_ctrl, secondary_exec_control, + tertiary_exec_control);pr_err ("EntryControls=%08x ExitControls=%08x\n" , vmentry_ctl, vmexit_ctl); pr_err("ExceptionBitmap=%08x PFECmask=%08x PFECmatch=%08x\n", vmcs_read32(EXCEPTION_BITMAP),
Signature · 1 line
1.8.3.1
PATCH 8/15 kvm/vmx: Add KVM support on guest Key Locker operations
In reply to Robert HooDon't clear CPU_BASED_ACTIVATE_TERTIARY_CONTROLS in vmx_exec_control(), as we really need it now.
Enable and implement handle_loadiwkey() VM-Exit handler, which fetches guest IWKey and do it onbehalf. Other Key Locker instructions can execute in non-root mode.
(Note: till this patch, we haven't expose Key Locker feature to guest yet, guest Kernel won't set CR4.KL, if guest deliberately execute Key Locker instructions, it will get #UD, instead of VM-Exit.)
We load guest's IWKey when load vcpu, even if it is NULL (guest doesn't support/enable Key Locker), to flush last vcpu's IWKey, which is possibly another VM's.
We flush guest's IWKey (loadiwkey with all 0) when put vcpu.
Trap guest write on MSRs of IA32_COPY_LOCAL_TO_PLATFORM and IA32_COPY_PLATFORM_TO_LOCAL_TO_PLATFORM, emulate IWKey save and restore operations.
Trap guest read on MSRs of IA32_COPY_STATUS and IA32_IWKEYBACKUP_STATUS, return their shadow values.
Analogous to adjust_vmx_controls(), we define the adjust_vmx_controls_64() auxiliary function, for MSR_IA32_VMX_PROCBASED_CTLS3 is 64bit allow-1 semantics, different from previous VMX capability MSRs, which were 32bit allow-0 and 32bit allow-1.
Also, define a helper get_xmm(), which per input index fetches an xmm value. VM-Exit of LOADIWKEY saves IWKey.encryption_key in some 2 xmm regs, and LOADIWKEY itself implicitly uses xmm0~2 as input. This helper facilitates xmm value's save/restore.
--- arch/x86/include/asm/kvm_host.h | 22 ++++ arch/x86/include/asm/vmx.h | 6 + arch/x86/include/uapi/asm/vmx.h | 2 + arch/x86/kvm/vmx/vmx.c | 249 +++++++++++++++++++++++++++++++++++++++- 4 files changed, 276 insertions(+), 3 deletions(-)
arch/x86/include/asm/kvm_host.h +22 -0
@@ -558,6 +558,19 @@ struct kvm_vcpu_xen {u64 runstate_times[4 ]; }; +#if defined(CONFIG_ARCH_SUPPORTS_INT128 ) && defined(CONFIG_CC_HAS_INT128 ) +typedef unsigned __int128 u128 ; +#else +typedef struct { +u64 reg64[2 ]; +}u128 ; +#endif + +struct iwkey { +u128 encryption_key[2 ];/* 256bit encryption key */ +u128 integrity_key;/* 128bit integration key */ +}; +struct kvm_vcpu_arch {/* * rip and regs accesses must go through
@@ -849,6 +862,11 @@ struct kvm_vcpu_arch {/* Protected Guests */ bool guest_state_protected; + +/* Intel KeyLocker */ +bool iwkey_loaded; +struct iwkey iwkey; +u32 msr_ia32_copy_status; };struct kvm_lpage_info {
@@ -1003,6 +1021,10 @@ struct kvm_arch {bool apic_access_page_done;unsigned long apicv_inhibit_reasons; +bool iwkey_backup_valid; +u32 msr_ia32_iwkey_backup_status; +struct iwkey iwkey_backup; +gpa_t wall_clock;bool mwait_in_guest;
arch/x86/include/asm/vmx.h +6 -0
@@ -76,6 +76,12 @@#define SECONDARY_EXEC_ENABLE_USR_WAIT_PAUSE VMCS_CONTROL_BIT(USR_WAIT_PAUSE)#define SECONDARY_EXEC_BUS_LOCK_DETECTION VMCS_CONTROL_BIT(BUS_LOCK_DETECTION) +/* +* Definitions of Tertiary Processor-Based VM-Execution Controls. +*/ +#define TERTIARY_EXEC_LOADIWKEY_EXITING VMCS_CONTROL_BIT(LOADIWKEY_EXITING) + +#define PIN_BASED_EXT_INTR_MASK VMCS_CONTROL_BIT(INTR_EXITING)#define PIN_BASED_NMI_EXITING VMCS_CONTROL_BIT(NMI_EXITING)#define PIN_BASED_VIRTUAL_NMIS VMCS_CONTROL_BIT(VIRTUAL_NMIS)
arch/x86/include/uapi/asm/vmx.h +2 -0
@@ -90,6 +90,7 @@#define EXIT_REASON_XRSTORS 64#define EXIT_REASON_UMWAIT 67#define EXIT_REASON_TPAUSE 68 +#define EXIT_REASON_LOADIWKEY 69#define EXIT_REASON_BUS_LOCK 74#define VMX_EXIT_REASONS \
@@ -153,6 +154,7 @@ { EXIT_REASON_XRSTORS, "XRSTORS" }, \ { EXIT_REASON_UMWAIT, "UMWAIT" }, \ { EXIT_REASON_TPAUSE, "TPAUSE" }, \ + { EXIT_REASON_LOADIWKEY, "LOADIWKEY" }, \ { EXIT_REASON_BUS_LOCK, "BUS_LOCK" } #define VMX_EXIT_REASON_FLAGS \
arch/x86/kvm/vmx/vmx.c +246 -3
315 lines not on this page · raw message
Signature · 1 line
1.8.3.1
PATCH 9/15 kvm/cpuid: Enumerate Key Locker feature in KVM
In reply to Robert HooIn kvm_set_cpu_caps(), add Key Locker feature enumeration, under the condition that 1) HW supports this feature 2) host Kernel isn't enabled with this feature.
Among its sub-features, filter out randomization support bit (CPUID.0x19.ECX[1]), as by design it cannot be supported at this moment. (Refer to Intel Key Locker Spec.)
Also, CPUID.0x19.EBX[0] (Key Locker instructions is enabled) is dynamic, based on CR4.KL status, thus reserve CR4.KL, and update this CPUID bit dynamically.
--- arch/x86/include/asm/kvm_host.h | 2 +- arch/x86/kvm/cpuid.c | 26 ++++++++++++++++++++++++-- arch/x86/kvm/reverse_cpuid.h | 32 ++++++++++++++++++++++++++++---- arch/x86/kvm/vmx/vmx.c | 2 +- arch/x86/kvm/x86.h | 2 ++ 5 files changed, 56 insertions(+), 8 deletions(-)
arch/x86/include/asm/kvm_host.h +1 -1
@@ -103,7 +103,7 @@ | X86_CR4_PGE | X86_CR4_PCE | X86_CR4_OSFXSR | X86_CR4_PCIDE \ | X86_CR4_OSXSAVE | X86_CR4_SMEP | X86_CR4_FSGSBASE \ | X86_CR4_OSXMMEXCPT | X86_CR4_LA57 | X86_CR4_VMXE \ - | X86_CR4_SMAP | X86_CR4_PKE | X86_CR4_UMIP)) + | X86_CR4_SMAP | X86_CR4_PKE | X86_CR4_UMIP | X86_CR4_KEYLOCKER))#define CR8_RESERVED_BITS (~(unsigned long)X86_CR8_TPR)
arch/x86/kvm/cpuid.c +24 -2
@@ -135,6 +135,12 @@ void kvm_update_cpuid_runtime(struct kvm_vcpu *vcpu) cpuid_entry_has(best, X86_FEATURE_XSAVEC))) best->ebx = xstate_required_size(vcpu->arch.xcr0, true); +/* update CPUID.0x19.EBX[0], depends on CR4.KL */ + best =kvm_find_cpuid_entry (vcpu,0x19 ,0 ); +if (best) +cpuid_entry_change (best,X86_FEATURE_KL_INS_ENABLED , +kvm_read_cr4_bits (vcpu,X86_CR4_KEYLOCKER )); + best =kvm_find_cpuid_entry (vcpu,KVM_CPUID_FEATURES ,0 );if (kvm_hlt_in_guest (vcpu->kvm) && best && (best->eax & (1 <<KVM_FEATURE_PV_UNHALT )))
@@ -456,14 +462,20 @@ void kvm_set_cpu_caps(void)kvm_cpu_cap_mask (CPUID_7_ECX ,F (AVX512VBMI ) |F (LA57 ) |F (PKU ) |0 /*OSPKE*/ |F (RDPID ) |F (AVX512_VPOPCNTDQ ) |F (UMIP ) |F (AVX512_VBMI2 ) |F (GFNI ) | -F (VAES ) |F (VPCLMULQDQ ) |F (AVX512_VNNI ) |F (AVX512_BITALG ) | -F (CLDEMOTE ) |F (MOVDIRI ) |F (MOVDIR64B ) |0 /*WAITPKG*/ | +F (VAES ) |0 /*KEYLOCKER*/ |F (VPCLMULQDQ ) |F (AVX512_VNNI ) | +F (AVX512_BITALG ) |F (CLDEMOTE ) |F (MOVDIRI ) |F (MOVDIR64B ) |0 /*WAITPKG*/ |F (SGX_LC ) ); +/* Set LA57 based on hardware capability. */ if (cpuid_ecx (7 ) &F (LA57 ))kvm_cpu_cap_set (X86_FEATURE_LA57 ); +/* At present, host and guest can only exclusively use KeyLocker */ +if (!boot_cpu_has (X86_FEATURE_KEYLOCKER ) && (cpuid_ecx (0x7 ) & +feature_bit (KEYLOCKER ))) +kvm_cpu_cap_set (X86_FEATURE_KEYLOCKER ); +/* * PKU not yet implemented for shadow paging and requires OSPKE * to be set on the host. Clear it if that is not the case
@@ -500,6 +512,10 @@ void kvm_set_cpu_caps(void)kvm_cpu_cap_init_scattered (CPUID_12_EAX ,SF (SGX1 ) |SF (SGX2 ) ); +kvm_cpu_cap_init_scattered (CPUID_19_EBX ,SF (KL_INS_ENABLED ) |SF (KL_WIDE ) | +SF (IWKEY_BACKUP )); +/* No randomize exposed to guest */ +kvm_cpu_cap_init_scattered (CPUID_19_ECX ,SF (IWKEY_NOBACKUP )); kvm_cpu_cap_mask(CPUID_8000_0001_ECX, F(LAHF_LM) | F(CMP_LEGACY) | 0 /*SVM*/ | 0 /* ExtApicSpace */ |
@@ -870,6 +886,12 @@ static inline int __do_cpuid_func(struct kvm_cpuid_array *array, u32 function) goto out; }break ; +/* KeyLocker */ +case 0x19 : +cpuid_entry_override (entry,CPUID_19_EBX ); +cpuid_entry_override (entry,CPUID_19_ECX ); +break ; +case KVM_CPUID_SIGNATURE : {static const char signature[12 ] ="KVMKVMKVM\0\0" ;const u32 *sigptr = (const u32 *)signature;
arch/x86/kvm/reverse_cpuid.h +28 -4
@@ -13,6 +13,9 @@*/ enum kvm_only_cpuid_leafs {CPUID_12_EAX =NCAPINTS , +CPUID_19_EBX , +CPUID_19_ECX , +NR_KVM_CPU_CAPS ,NKVMCAPINTS =NR_KVM_CPU_CAPS -NCAPINTS ,
@@ -24,6 +27,13 @@ enum kvm_only_cpuid_leafs {#define KVM_X86_FEATURE_SGX1 KVM_X86_FEATURE(CPUID_12_EAX, 0)#define KVM_X86_FEATURE_SGX2 KVM_X86_FEATURE(CPUID_12_EAX, 1) +/* Intel-defined Key Locker sub-features, CPUID level 0x19 (EBX). */ +#define KVM_X86_FEATURE_KL_INS_ENABLED KVM_X86_FEATURE(CPUID_19_EBX, 0) +#define KVM_X86_FEATURE_KL_WIDE KVM_X86_FEATURE(CPUID_19_EBX, 2) +#define KVM_X86_FEATURE_IWKEY_BACKUP KVM_X86_FEATURE(CPUID_19_EBX, 4) +#define KVM_X86_FEATURE_IWKEY_NOBACKUP KVM_X86_FEATURE(CPUID_19_ECX, 0) +#define KVM_X86_FEATURE_IWKEY_RAND KVM_X86_FEATURE(CPUID_19_ECX, 1) +struct cpuid_reg {u32 function;u32 index;
@@ -48,6 +58,8 @@ struct cpuid_reg { [CPUID_7_1_EAX] = { 7, 1, CPUID_EAX}, [CPUID_12_EAX] = {0x00000012, 0, CPUID_EAX}, [CPUID_8000_001F_EAX] = {0x8000001f, 0, CPUID_EAX}, + [CPUID_19_EBX] = { 0x19, 0, CPUID_EBX}, + [CPUID_19_ECX] = { 0x19, 0, CPUID_ECX}, };/*
@@ -74,12 +86,24 @@ static __always_inline void reverse_cpuid_check(unsigned int x86_leaf)*/ static __always_inline u32 __feature_translate(int x86_feature) { -if (x86_feature ==X86_FEATURE_SGX1 ) +switch (x86_feature) { +case X86_FEATURE_SGX1 :return KVM_X86_FEATURE_SGX1 ; -else if (x86_feature ==X86_FEATURE_SGX2 ) +case X86_FEATURE_SGX2 :return KVM_X86_FEATURE_SGX2 ; - -return x86_feature; +case X86_FEATURE_KL_INS_ENABLED : +return KVM_X86_FEATURE_KL_INS_ENABLED ; +case X86_FEATURE_KL_WIDE : +return KVM_X86_FEATURE_KL_WIDE ; +case X86_FEATURE_IWKEY_BACKUP : +return KVM_X86_FEATURE_IWKEY_BACKUP ; +case X86_FEATURE_IWKEY_NOBACKUP : +return KVM_X86_FEATURE_IWKEY_NOBACKUP ; +case X86_FEATURE_IWKEY_RAND : +return KVM_X86_FEATURE_IWKEY_RAND ; +default : +return x86_feature; + } } static __always_inline u32 __feature_leaf(int x86_feature)
arch/x86/kvm/vmx/vmx.c +1 -1
@@ -3507,7 +3507,7 @@ void vmx_set_cr4(struct kvm_vcpu *vcpu, unsigned long cr4)vmcs_writel (CR4_READ_SHADOW , cr4);vmcs_writel (GUEST_CR4 , hw_cr4); -if ((cr4 ^ old_cr4) & (X86_CR4_OSXSAVE |X86_CR4_PKE )) +if ((cr4 ^ old_cr4) & (X86_CR4_OSXSAVE |X86_CR4_PKE |X86_CR4_KEYLOCKER ))kvm_update_cpuid_runtime (vcpu); }
arch/x86/kvm/x86.h +2 -0
@@ -485,6 +485,8 @@ int kvm_handle_memory_failure(struct kvm_vcpu *vcpu, int r, __reserved_bits |=X86_CR4_VMXE ; \if (!__cpu_has (__c,X86_FEATURE_PCID )) \ __reserved_bits |=X86_CR4_PCIDE ; \ +if (!__cpu_has (__c,X86_FEATURE_KEYLOCKER )) \ + __reserved_bits |=X86_CR4_KEYLOCKER ; \ __reserved_bits; \ })
Signature · 1 line
1.8.3.1
PATCH 10/15 kvm/vmx/nested: Support new IA32_VMX_PROCBASED_CTLS3 vmx capability MSR
In reply to Robert HooAdd this new VMX capability MSR in nested_vmx_msrs, and related functions for its nested support.
Don't set its LOADIWKEY VM-Exit bit at present. It will be enabled in last patch when everything's prepared.
--- arch/x86/kvm/vmx/capabilities.h | 2 ++ arch/x86/kvm/vmx/nested.c | 22 +++++++++++++++++++++- arch/x86/kvm/vmx/vmx.c | 6 +++--- arch/x86/kvm/x86.c | 2 ++ 4 files changed, 28 insertions(+), 4 deletions(-)
arch/x86/kvm/vmx/capabilities.h +2 -0
@@ -33,6 +33,8 @@ struct nested_vmx_msrs {u32 procbased_ctls_high;u32 secondary_ctls_low;u32 secondary_ctls_high; +/* Tertiary Controls is 64bit allow-1 semantics */ +u64 tertiary_ctls;u32 pinbased_ctls_low;u32 pinbased_ctls_high;u32 exit_ctls_low;
arch/x86/kvm/vmx/nested.c +21 -1
@@ -1272,6 +1272,18 @@ static int vmx_restore_vmx_basic(struct vcpu_vmx *vmx, u64 data) lowp = &vmx->nested.msrs.secondary_ctls_low; highp = &vmx->nested.msrs.secondary_ctls_high;break ; +/* +* MSR_IA32_VMX_PROCBASED_CTLS3 is 64bit, all allow-1 (must-be-0) +* semantics. +*/ +case MSR_IA32_VMX_PROCBASED_CTLS3 : +/* Check must-be-0 bits are still 0. */ +if (!is_bitwise_subset (vmx->nested.msrs.tertiary_ctls, + data,GENMASK_ULL (63 ,0 ))) +return -EINVAL ; + + vmx->nested.msrs.tertiary_ctls = data; +return 0 ;default :BUG (); }
@@ -1408,6 +1420,7 @@ int vmx_set_vmx_msr(struct kvm_vcpu *vcpu, u32 msr_index, u64 data)case MSR_IA32_VMX_TRUE_EXIT_CTLS :case MSR_IA32_VMX_TRUE_ENTRY_CTLS :case MSR_IA32_VMX_PROCBASED_CTLS2 : +case MSR_IA32_VMX_PROCBASED_CTLS3 :return vmx_restore_control_msr (vmx, msr_index, data);case MSR_IA32_VMX_MISC :return vmx_restore_vmx_misc (vmx, data);
@@ -1503,6 +1516,9 @@ int vmx_get_vmx_msr(struct nested_vmx_msrs *msrs, u32 msr_index, u64 *pdata) msrs->secondary_ctls_low, msrs->secondary_ctls_high);break ; +case MSR_IA32_VMX_PROCBASED_CTLS3 : + *pdata = msrs->tertiary_ctls; +break ;case MSR_IA32_VMX_EPT_VPID_CAP : *pdata = msrs->ept_caps | ((u64 )msrs->vpid_caps <<32 );
@@ -6429,7 +6445,8 @@ void nested_vmx_setup_ctls_msrs(struct nested_vmx_msrs *msrs, u32 ept_caps)CPU_BASED_USE_IO_BITMAPS |CPU_BASED_MONITOR_TRAP_FLAG |CPU_BASED_MONITOR_EXITING |CPU_BASED_RDPMC_EXITING |CPU_BASED_RDTSC_EXITING |CPU_BASED_PAUSE_EXITING | -CPU_BASED_TPR_SHADOW |CPU_BASED_ACTIVATE_SECONDARY_CONTROLS ; +CPU_BASED_TPR_SHADOW |CPU_BASED_ACTIVATE_SECONDARY_CONTROLS | +CPU_BASED_ACTIVATE_TERTIARY_CONTROLS ;/* * We can allow some features even when not supported by the * hardware. For example, L1 can specify an MSR bitmap - and we
@@ -6467,6 +6484,9 @@ void nested_vmx_setup_ctls_msrs(struct nested_vmx_msrs *msrs, u32 ept_caps)SECONDARY_EXEC_RDSEED_EXITING |SECONDARY_EXEC_XSAVES ; +if (msrs->procbased_ctls_high &CPU_BASED_ACTIVATE_TERTIARY_CONTROLS ) +rdmsrl (MSR_IA32_VMX_PROCBASED_CTLS3 , msrs->tertiary_ctls); + msrs->tertiary_ctls &=0 ;/* * We can emulate "VMCS shadowing," even if the hardware * doesn't support it.
arch/x86/kvm/vmx/vmx.c +3 -3
@@ -1981,7 +1981,7 @@ static inline bool vmx_feature_control_msr_valid(struct kvm_vcpu *vcpu,static int vmx_get_msr_feature (struct kvm_msr_entry *msr) {switch (msr->index) { - case MSR_IA32_VMX_BASIC ... MSR_IA32_VMX_VMFUNC: + case MSR_IA32_VMX_BASIC ... MSR_IA32_VMX_PROCBASED_CTLS3:if (!nested)return 1 ;return vmx_get_vmx_msr (&vmcs_config.nested, msr->index, &msr->data);
@@ -2069,7 +2069,7 @@ static int vmx_get_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info) msr_info->data =to_vmx (vcpu)->msr_ia32_sgxlepubkeyhash [msr_info->index -MSR_IA32_SGXLEPUBKEYHASH0 ];break ; - case MSR_IA32_VMX_BASIC ... MSR_IA32_VMX_VMFUNC: + case MSR_IA32_VMX_BASIC ... MSR_IA32_VMX_PROCBASED_CTLS3:if (!nested_vmx_allowed (vcpu))return 1 ; if (vmx_get_vmx_msr(&vmx->nested.msrs, msr_info->index,
@@ -2399,7 +2399,7 @@ static int vmx_set_msr(struct kvm_vcpu *vcpu, struct msr_data *msr_info) vmx->msr_ia32_sgxlepubkeyhash [msr_index -MSR_IA32_SGXLEPUBKEYHASH0 ] = data;break ; - case MSR_IA32_VMX_BASIC ... MSR_IA32_VMX_VMFUNC: + case MSR_IA32_VMX_BASIC ... MSR_IA32_VMX_PROCBASED_CTLS3:if (!msr_info->host_initiated)return 1 ;/* they are read-only */ if (!nested_vmx_allowed (vcpu))
arch/x86/kvm/x86.c +2 -0
@@ -1332,6 +1332,7 @@ int kvm_emulate_rdpmc(struct kvm_vcpu *vcpu)MSR_IA32_VMX_PROCBASED_CTLS2 ,MSR_IA32_VMX_EPT_VPID_CAP ,MSR_IA32_VMX_VMFUNC , +MSR_IA32_VMX_PROCBASED_CTLS3 ,MSR_K7_HWCR , MSR_KVM_POLL_CONTROL,
@@ -1363,6 +1364,7 @@ int kvm_emulate_rdpmc(struct kvm_vcpu *vcpu)MSR_IA32_VMX_PROCBASED_CTLS2 ,MSR_IA32_VMX_EPT_VPID_CAP ,MSR_IA32_VMX_VMFUNC , +MSR_IA32_VMX_PROCBASED_CTLS3 ,MSR_F10H_DECFG , MSR_IA32_UCODE_REV,
Signature · 1 line
1.8.3.1
PATCH 11/15 kvm/vmx: Implement vmx_compute_tertiary_exec_control()
In reply to Robert HooLike vmx_compute_secondary_exec_control(), before L1 set VMCS, compute its tertiary control capability MSR's value according to guest CPUID setting.
--- arch/x86/kvm/vmx/vmx.c | 22 ++++++++++++++++++++-- arch/x86/kvm/vmx/vmx.h | 1 + 2 files changed, 21 insertions(+), 2 deletions(-)
arch/x86/kvm/vmx/vmx.c +20 -2
@@ -4523,6 +4523,22 @@ u32 vmx_exec_control(struct vcpu_vmx *vmx)#define vmx_adjust_sec_exec_exiting (vmx, exec_control, lname, uname) \ vmx_adjust_sec_exec_control(vmx, exec_control, lname, uname, uname##_EXITING, true) +static void vmx_compute_tertiary_exec_control (struct vcpu_vmx *vmx) +{ +struct kvm_vcpu *vcpu = &vmx->vcpu; +u32 exec_control = vmcs_config.cpu_based_3rd_exec_ctrl; + +if (nested) { +if (guest_cpuid_has (vcpu,X86_FEATURE_KEYLOCKER )) + vmx->nested.msrs.tertiary_ctls |= +TERTIARY_EXEC_LOADIWKEY_EXITING ; +else + vmx->nested.msrs.tertiary_ctls &= + ~TERTIARY_EXEC_LOADIWKEY_EXITING ; + } + vmx->tertiary_exec_control = exec_control; +} +static void vmx_compute_secondary_exec_control (struct vcpu_vmx *vmx) {struct kvm_vcpu *vcpu = &vmx->vcpu;
@@ -4622,8 +4638,10 @@ static void init_vmcs(struct vcpu_vmx *vmx)secondary_exec_controls_set (vmx, vmx->secondary_exec_control); } -if (cpu_has_tertiary_exec_ctrls ()) -tertiary_exec_controls_set (vmx, vmcs_config.cpu_based_3rd_exec_ctrl); +if (cpu_has_tertiary_exec_ctrls ()) { +vmx_compute_tertiary_exec_control (vmx); +tertiary_exec_controls_set (vmx, vmx->tertiary_exec_control); + }if (kvm_vcpu_apicv_active (&vmx->vcpu)) {vmcs_write64 (EOI_EXIT_BITMAP0 ,0 );
arch/x86/kvm/vmx/vmx.h +1 -0
@@ -258,6 +258,7 @@ struct vcpu_vmx {u32 msr_ia32_umwait_control;u32 secondary_exec_control; +u64 tertiary_exec_control;/* * loaded_vmcs points to the VMCS currently used in this vcpu. For a
Signature · 1 line
1.8.3.1
PATCH 12/15 kvm/vmx/vmcs12: Add Tertiary VM-Exec control field in vmcs12
In reply to Robert Hoo--- arch/x86/kvm/vmx/vmcs12.c | 1 + arch/x86/kvm/vmx/vmcs12.h | 4 +++- 2 files changed, 4 insertions(+), 1 deletion(-)
arch/x86/kvm/vmx/vmcs12.c +1 -0
@@ -51,6 +51,7 @@FIELD64 (VMWRITE_BITMAP , vmwrite_bitmap),FIELD64 (XSS_EXIT_BITMAP , xss_exit_bitmap),FIELD64 (ENCLS_EXITING_BITMAP , encls_exiting_bitmap), +FIELD64 (TERTIARY_VM_EXEC_CONTROL , tertiary_vm_exec_control),FIELD64 (GUEST_PHYSICAL_ADDRESS , guest_physical_address),FIELD64 (VMCS_LINK_POINTER , vmcs_link_pointer), FIELD64(GUEST_IA32_DEBUGCTL, guest_ia32_debugctl),
arch/x86/kvm/vmx/vmcs12.h +3 -1
@@ -70,7 +70,8 @@ struct __packed vmcs12 {u64 eptp_list_address;u64 pml_address;u64 encls_exiting_bitmap; -u64 padding64[2 ];/* room for future expansion */ +u64 tertiary_vm_exec_control; +u64 padding64[1 ];/* room for future expansion */ /* * To allow migration of L1 (complete with its L2 guests) between * machines of different natural widths (32 or 64 bit), we cannot have
@@ -258,6 +259,7 @@ static inline void vmx_check_vmcs12_offsets(void)CHECK_OFFSET (eptp_list_address,304 );CHECK_OFFSET (pml_address,312 );CHECK_OFFSET (encls_exiting_bitmap,320 ); +CHECK_OFFSET (tertiary_vm_exec_control,328 );CHECK_OFFSET (cr0_guest_host_mask,344 );CHECK_OFFSET (cr4_guest_host_mask,352 );CHECK_OFFSET (cr0_read_shadow,360 );
Signature · 1 line
1.8.3.1
PATCH 13/15 kvm/vmx/nested: Support Tertiary VM-Exec control in vmcs02
In reply to Robert Hoo--- arch/x86/kvm/vmx/nested.c | 13 +++++++++++++ 1 file changed, 13 insertions(+)
arch/x86/kvm/vmx/nested.c +13 -0
@@ -2222,6 +2222,7 @@ static void prepare_vmcs02_early_rare(struct vcpu_vmx *vmx,static void prepare_vmcs02_early (struct vcpu_vmx *vmx,struct vmcs12 *vmcs12) {u32 exec_control; +u64 vmcs02_ter_exec_ctrl;u64 guest_efer =nested_vmx_calc_efer (vmx, vmcs12);if (vmx->nested.dirty_vmcs12 || vmx->nested.hv_evmcs)
@@ -2344,6 +2345,18 @@ static void prepare_vmcs02_early(struct vcpu_vmx *vmx, struct vmcs12 *vmcs12)vm_entry_controls_set (vmx, exec_control);/* +* Tertiary EXEC CONTROLS +*/ +if (cpu_has_tertiary_exec_ctrls ()) { + vmcs02_ter_exec_ctrl = vmx->tertiary_exec_control; +if (nested_cpu_has (vmcs12, +CPU_BASED_ACTIVATE_TERTIARY_CONTROLS )) + vmcs02_ter_exec_ctrl |= vmcs12->tertiary_vm_exec_control; + +tertiary_exec_controls_set (vmx, vmcs02_ter_exec_ctrl); + } + +/* * EXIT CONTROLS * * L2->L1 exit controls are emulated - the hardware exit is to L0 so
Signature · 1 line
1.8.3.1
PATCH 14/15 kvm/vmx/nested: Support CR4.KL in nested
In reply to Robert HooAdd CR4.KL in nested.msr.cr4_fixed1 when guest CPUID supports KeyLocker. So that it can pass check when preparing vmcs02.
--- arch/x86/kvm/vmx/vmx.c | 1 + 1 file changed, 1 insertion(+)
arch/x86/kvm/vmx/vmx.c +1 -0
@@ -7440,6 +7440,7 @@ static void nested_vmx_cr_fixed1_bits_update(struct kvm_vcpu *vcpu)cr4_fixed1_update (X86_CR4_PKE , ecx,feature_bit (PKU ));cr4_fixed1_update (X86_CR4_UMIP , ecx,feature_bit (UMIP ));cr4_fixed1_update (X86_CR4_LA57 , ecx,feature_bit (LA57 )); +cr4_fixed1_update (X86_CR4_KEYLOCKER , ecx,feature_bit (KEYLOCKER ));#undef cr4_fixed1_update }
Signature · 1 line
1.8.3.1
PATCH 15/15 kvm/vmx/nested: Enable nested LOADIWKEY VM-exit
In reply to Robert HooSet the LOADIWKEY VM-exit bit in nested vmx ctrl MSR, and let L1 intercept L2's LOADIWKEY VM-Exit.
Add helper nested_cpu_has3(), which returns if some feature in Tertiary VM-Exec Control is set.
--- arch/x86/kvm/vmx/nested.c | 5 ++++- arch/x86/kvm/vmx/nested.h | 7 +++++++ 2 files changed, 11 insertions(+), 1 deletion(-)
arch/x86/kvm/vmx/nested.c +4 -1
@@ -5983,6 +5983,9 @@ static bool nested_vmx_l1_wants_exit(struct kvm_vcpu *vcpu, SECONDARY_EXEC_ENABLE_USR_WAIT_PAUSE);case EXIT_REASON_ENCLS :return nested_vmx_exit_handled_encls (vcpu, vmcs12); +case EXIT_REASON_LOADIWKEY : +return nested_cpu_has3 (vmcs12, +TERTIARY_EXEC_LOADIWKEY_EXITING );default :return true; }
@@ -6499,7 +6502,7 @@ void nested_vmx_setup_ctls_msrs(struct nested_vmx_msrs *msrs, u32 ept_caps)if (msrs->procbased_ctls_high &CPU_BASED_ACTIVATE_TERTIARY_CONTROLS )rdmsrl (MSR_IA32_VMX_PROCBASED_CTLS3 , msrs->tertiary_ctls); - msrs->tertiary_ctls &=0 ; + msrs->tertiary_ctls &=TERTIARY_EXEC_LOADIWKEY_EXITING ;/* * We can emulate "VMCS shadowing," even if the hardware * doesn't support it.
arch/x86/kvm/vmx/nested.h +7 -0
@@ -145,6 +145,13 @@ static inline bool nested_cpu_has2(struct vmcs12 *vmcs12, u32 bit) (vmcs12->secondary_vm_exec_control & bit); } +static inline bool nested_cpu_has3 (struct vmcs12 *vmcs12,u32 bit) +{ +return (vmcs12->cpu_based_vm_exec_control & +CPU_BASED_ACTIVATE_TERTIARY_CONTROLS ) && + (vmcs12->tertiary_vm_exec_control & bit); +} +static inline bool nested_cpu_has_preemption_timer (struct vmcs12 *vmcs12) { return vmcs12->pin_based_vm_exec_control &
Signature · 1 line
1.8.3.1