1 The Definitive KVM (Kernel-based Virtual Machine) API Documentation
2 ===================================================================
7 The kvm API is a set of ioctls that are issued to control various aspects
8 of a virtual machine. The ioctls belong to three classes
10 - System ioctls: These query and set global attributes which affect the
11 whole kvm subsystem. In addition a system ioctl is used to create
14 - VM ioctls: These query and set attributes that affect an entire virtual
15 machine, for example memory layout. In addition a VM ioctl is used to
16 create virtual cpus (vcpus).
18 Only run VM ioctls from the same process (address space) that was used
21 - vcpu ioctls: These query and set attributes that control the operation
22 of a single virtual cpu.
24 Only run vcpu ioctls from the same thread that was used to create the
31 The kvm API is centered around file descriptors. An initial
32 open("/dev/kvm") obtains a handle to the kvm subsystem; this handle
33 can be used to issue system ioctls. A KVM_CREATE_VM ioctl on this
34 handle will create a VM file descriptor which can be used to issue VM
35 ioctls. A KVM_CREATE_VCPU ioctl on a VM fd will create a virtual cpu
36 and return a file descriptor pointing to it. Finally, ioctls on a vcpu
37 fd can be used to control the vcpu, including the important task of
38 actually running guest code.
40 In general file descriptors can be migrated among processes by means
41 of fork() and the SCM_RIGHTS facility of unix domain socket. These
42 kinds of tricks are explicitly not supported by kvm. While they will
43 not cause harm to the host, their actual behavior is not guaranteed by
44 the API. The only supported use is one virtual machine per process,
45 and one vcpu per thread.
51 As of Linux 2.6.22, the KVM ABI has been stabilized: no backward
52 incompatible change are allowed. However, there is an extension
53 facility that allows backward-compatible extensions to the API to be
56 The extension mechanism is not based on the Linux version number.
57 Instead, kvm defines extension identifiers and a facility to query
58 whether a particular extension identifier is available. If it is, a
59 set of ioctls is available for application use.
65 This section describes ioctls that can be used to control kvm guests.
66 For each ioctl, the following information is provided along with a
69 Capability: which KVM extension provides this ioctl. Can be 'basic',
70 which means that is will be provided by any kernel that supports
71 API version 12 (see section 4.1), a KVM_CAP_xyz constant, which
72 means availability needs to be checked with KVM_CHECK_EXTENSION
73 (see section 4.4), or 'none' which means that while not all kernels
74 support this ioctl, there's no capability bit to check its
75 availability: for kernels that don't support the ioctl,
76 the ioctl returns -ENOTTY.
78 Architectures: which instruction set architectures provide this ioctl.
79 x86 includes both i386 and x86_64.
81 Type: system, vm, or vcpu.
83 Parameters: what parameters are accepted by the ioctl.
85 Returns: the return value. General error numbers (EBADF, ENOMEM, EINVAL)
86 are not detailed, but errors with specific meanings are.
89 4.1 KVM_GET_API_VERSION
95 Returns: the constant KVM_API_VERSION (=12)
97 This identifies the API version as the stable kvm API. It is not
98 expected that this number will change. However, Linux 2.6.20 and
99 2.6.21 report earlier versions; these are not documented and not
100 supported. Applications should refuse to run if KVM_GET_API_VERSION
101 returns a value other than 12. If this check passes, all ioctls
102 described as 'basic' will be available.
110 Parameters: machine type identifier (KVM_VM_*)
111 Returns: a VM fd that can be used to control the new virtual machine.
113 The new VM has no virtual cpus and no memory.
114 You probably want to use 0 as machine type.
116 In order to create user controlled virtual machines on S390, check
117 KVM_CAP_S390_UCONTROL and use the flag KVM_VM_S390_UCONTROL as
118 privileged user (CAP_SYS_ADMIN).
120 To use hardware assisted virtualization on MIPS (VZ ASE) rather than
121 the default trap & emulate implementation (which changes the virtual
122 memory layout to fit in user mode), check KVM_CAP_MIPS_VZ and use the
126 On arm64, the physical address size for a VM (IPA Size limit) is limited
127 to 40bits by default. The limit can be configured if the host supports the
128 extension KVM_CAP_ARM_VM_IPA_SIZE. When supported, use
129 KVM_VM_TYPE_ARM_IPA_SIZE(IPA_Bits) to set the size in the machine type
130 identifier, where IPA_Bits is the maximum width of any physical
131 address used by the VM. The IPA_Bits is encoded in bits[7-0] of the
132 machine type identifier.
134 e.g, to configure a guest to use 48bit physical address size :
136 vm_fd = ioctl(dev_fd, KVM_CREATE_VM, KVM_VM_TYPE_ARM_IPA_SIZE(48));
138 The requested size (IPA_Bits) must be :
139 0 - Implies default size, 40bits (for backward compatibility)
143 N - Implies N bits, where N is a positive integer such that,
144 32 <= N <= Host_IPA_Limit
146 Host_IPA_Limit is the maximum possible value for IPA_Bits on the host and
147 is dependent on the CPU capability and the kernel configuration. The limit can
148 be retrieved using KVM_CAP_ARM_VM_IPA_SIZE of the KVM_CHECK_EXTENSION
151 Please note that configuring the IPA size does not affect the capability
152 exposed by the guest CPUs in ID_AA64MMFR0_EL1[PARange]. It only affects
153 size of the address translated by the stage2 level (guest physical to
154 host physical address translations).
157 4.3 KVM_GET_MSR_INDEX_LIST, KVM_GET_MSR_FEATURE_INDEX_LIST
159 Capability: basic, KVM_CAP_GET_MSR_FEATURES for KVM_GET_MSR_FEATURE_INDEX_LIST
162 Parameters: struct kvm_msr_list (in/out)
163 Returns: 0 on success; -1 on error
165 EFAULT: the msr index list cannot be read from or written to
166 E2BIG: the msr index list is to be to fit in the array specified by
169 struct kvm_msr_list {
170 __u32 nmsrs; /* number of msrs in entries */
174 The user fills in the size of the indices array in nmsrs, and in return
175 kvm adjusts nmsrs to reflect the actual number of msrs and fills in the
176 indices array with their numbers.
178 KVM_GET_MSR_INDEX_LIST returns the guest msrs that are supported. The list
179 varies by kvm version and host processor, but does not change otherwise.
181 Note: if kvm indicates supports MCE (KVM_CAP_MCE), then the MCE bank MSRs are
182 not returned in the MSR list, as different vcpus can have a different number
183 of banks, as set via the KVM_X86_SETUP_MCE ioctl.
185 KVM_GET_MSR_FEATURE_INDEX_LIST returns the list of MSRs that can be passed
186 to the KVM_GET_MSRS system ioctl. This lets userspace probe host capabilities
187 and processor features that are exposed via MSRs (e.g., VMX capabilities).
188 This list also varies by kvm version and host processor, but does not change
192 4.4 KVM_CHECK_EXTENSION
194 Capability: basic, KVM_CAP_CHECK_EXTENSION_VM for vm ioctl
196 Type: system ioctl, vm ioctl
197 Parameters: extension identifier (KVM_CAP_*)
198 Returns: 0 if unsupported; 1 (or some other positive integer) if supported
200 The API allows the application to query about extensions to the core
201 kvm API. Userspace passes an extension identifier (an integer) and
202 receives an integer that describes the extension availability.
203 Generally 0 means no and 1 means yes, but some extensions may report
204 additional information in the integer return value.
206 Based on their initialization different VMs may have different capabilities.
207 It is thus encouraged to use the vm ioctl to query for capabilities (available
208 with KVM_CAP_CHECK_EXTENSION_VM on the vm fd)
210 4.5 KVM_GET_VCPU_MMAP_SIZE
216 Returns: size of vcpu mmap area, in bytes
218 The KVM_RUN ioctl (cf.) communicates with userspace via a shared
219 memory region. This ioctl returns the size of that region. See the
220 KVM_RUN documentation for details.
223 4.6 KVM_SET_MEMORY_REGION
228 Parameters: struct kvm_memory_region (in)
229 Returns: 0 on success, -1 on error
231 This ioctl is obsolete and has been removed.
239 Parameters: vcpu id (apic id on x86)
240 Returns: vcpu fd on success, -1 on error
242 This API adds a vcpu to a virtual machine. No more than max_vcpus may be added.
243 The vcpu id is an integer in the range [0, max_vcpu_id).
245 The recommended max_vcpus value can be retrieved using the KVM_CAP_NR_VCPUS of
246 the KVM_CHECK_EXTENSION ioctl() at run-time.
247 The maximum possible value for max_vcpus can be retrieved using the
248 KVM_CAP_MAX_VCPUS of the KVM_CHECK_EXTENSION ioctl() at run-time.
250 If the KVM_CAP_NR_VCPUS does not exist, you should assume that max_vcpus is 4
252 If the KVM_CAP_MAX_VCPUS does not exist, you should assume that max_vcpus is
253 same as the value returned from KVM_CAP_NR_VCPUS.
255 The maximum possible value for max_vcpu_id can be retrieved using the
256 KVM_CAP_MAX_VCPU_ID of the KVM_CHECK_EXTENSION ioctl() at run-time.
258 If the KVM_CAP_MAX_VCPU_ID does not exist, you should assume that max_vcpu_id
259 is the same as the value returned from KVM_CAP_MAX_VCPUS.
261 On powerpc using book3s_hv mode, the vcpus are mapped onto virtual
262 threads in one or more virtual CPU cores. (This is because the
263 hardware requires all the hardware threads in a CPU core to be in the
264 same partition.) The KVM_CAP_PPC_SMT capability indicates the number
265 of vcpus per virtual core (vcore). The vcore id is obtained by
266 dividing the vcpu id by the number of vcpus per vcore. The vcpus in a
267 given vcore will always be in the same physical core as each other
268 (though that might be a different physical core from time to time).
269 Userspace can control the threading (SMT) mode of the guest by its
270 allocation of vcpu ids. For example, if userspace wants
271 single-threaded guest vcpus, it should make all vcpu ids be a multiple
272 of the number of vcpus per vcore.
274 For virtual cpus that have been created with S390 user controlled virtual
275 machines, the resulting vcpu fd can be memory mapped at page offset
276 KVM_S390_SIE_PAGE_OFFSET in order to obtain a memory map of the virtual
277 cpu's hardware control block.
280 4.8 KVM_GET_DIRTY_LOG (vm ioctl)
285 Parameters: struct kvm_dirty_log (in/out)
286 Returns: 0 on success, -1 on error
288 /* for KVM_GET_DIRTY_LOG */
289 struct kvm_dirty_log {
293 void __user *dirty_bitmap; /* one bit per page */
298 Given a memory slot, return a bitmap containing any pages dirtied
299 since the last call to this ioctl. Bit 0 is the first page in the
300 memory slot. Ensure the entire structure is cleared to avoid padding
303 If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 specifies
304 the address space for which you want to return the dirty bitmap.
305 They must be less than the value that KVM_CHECK_EXTENSION returns for
306 the KVM_CAP_MULTI_ADDRESS_SPACE capability.
309 4.9 KVM_SET_MEMORY_ALIAS
314 Parameters: struct kvm_memory_alias (in)
315 Returns: 0 (success), -1 (error)
317 This ioctl is obsolete and has been removed.
326 Returns: 0 on success, -1 on error
328 EINTR: an unmasked signal is pending
330 This ioctl is used to run a guest virtual cpu. While there are no
331 explicit parameters, there is an implicit parameter block that can be
332 obtained by mmap()ing the vcpu fd at offset 0, with the size given by
333 KVM_GET_VCPU_MMAP_SIZE. The parameter block is formatted as a 'struct
334 kvm_run' (see below).
340 Architectures: all except ARM, arm64
342 Parameters: struct kvm_regs (out)
343 Returns: 0 on success, -1 on error
345 Reads the general purpose registers from the vcpu.
349 /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */
350 __u64 rax, rbx, rcx, rdx;
351 __u64 rsi, rdi, rsp, rbp;
352 __u64 r8, r9, r10, r11;
353 __u64 r12, r13, r14, r15;
359 /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */
370 Architectures: all except ARM, arm64
372 Parameters: struct kvm_regs (in)
373 Returns: 0 on success, -1 on error
375 Writes the general purpose registers into the vcpu.
377 See KVM_GET_REGS for the data structure.
383 Architectures: x86, ppc
385 Parameters: struct kvm_sregs (out)
386 Returns: 0 on success, -1 on error
388 Reads special registers from the vcpu.
392 struct kvm_segment cs, ds, es, fs, gs, ss;
393 struct kvm_segment tr, ldt;
394 struct kvm_dtable gdt, idt;
395 __u64 cr0, cr2, cr3, cr4, cr8;
398 __u64 interrupt_bitmap[(KVM_NR_INTERRUPTS + 63) / 64];
401 /* ppc -- see arch/powerpc/include/uapi/asm/kvm.h */
403 interrupt_bitmap is a bitmap of pending external interrupts. At most
404 one bit may be set. This interrupt has been acknowledged by the APIC
405 but not yet injected into the cpu core.
411 Architectures: x86, ppc
413 Parameters: struct kvm_sregs (in)
414 Returns: 0 on success, -1 on error
416 Writes special registers into the vcpu. See KVM_GET_SREGS for the
425 Parameters: struct kvm_translation (in/out)
426 Returns: 0 on success, -1 on error
428 Translates a virtual address according to the vcpu's current address
431 struct kvm_translation {
433 __u64 linear_address;
436 __u64 physical_address;
447 Architectures: x86, ppc, mips
449 Parameters: struct kvm_interrupt (in)
450 Returns: 0 on success, negative on failure.
452 Queues a hardware interrupt vector to be injected.
454 /* for KVM_INTERRUPT */
455 struct kvm_interrupt {
462 Returns: 0 on success,
463 -EEXIST if an interrupt is already enqueued
464 -EINVAL the the irq number is invalid
465 -ENXIO if the PIC is in the kernel
466 -EFAULT if the pointer is invalid
468 Note 'irq' is an interrupt vector, not an interrupt pin or line. This
469 ioctl is useful if the in-kernel PIC is not used.
473 Queues an external interrupt to be injected. This ioctl is overleaded
474 with 3 different irq values:
478 This injects an edge type external interrupt into the guest once it's ready
479 to receive interrupts. When injected, the interrupt is done.
481 b) KVM_INTERRUPT_UNSET
483 This unsets any pending interrupt.
485 Only available with KVM_CAP_PPC_UNSET_IRQ.
487 c) KVM_INTERRUPT_SET_LEVEL
489 This injects a level type external interrupt into the guest context. The
490 interrupt stays pending until a specific ioctl with KVM_INTERRUPT_UNSET
493 Only available with KVM_CAP_PPC_IRQ_LEVEL.
495 Note that any value for 'irq' other than the ones stated above is invalid
496 and incurs unexpected behavior.
500 Queues an external interrupt to be injected into the virtual CPU. A negative
501 interrupt number dequeues the interrupt.
512 Support for this has been removed. Use KVM_SET_GUEST_DEBUG instead.
517 Capability: basic (vcpu), KVM_CAP_GET_MSR_FEATURES (system)
519 Type: system ioctl, vcpu ioctl
520 Parameters: struct kvm_msrs (in/out)
521 Returns: number of msrs successfully returned;
524 When used as a system ioctl:
525 Reads the values of MSR-based features that are available for the VM. This
526 is similar to KVM_GET_SUPPORTED_CPUID, but it returns MSR indices and values.
527 The list of msr-based features can be obtained using KVM_GET_MSR_FEATURE_INDEX_LIST
530 When used as a vcpu ioctl:
531 Reads model-specific registers from the vcpu. Supported msr indices can
532 be obtained using KVM_GET_MSR_INDEX_LIST in a system ioctl.
535 __u32 nmsrs; /* number of msrs in entries */
538 struct kvm_msr_entry entries[0];
541 struct kvm_msr_entry {
547 Application code should set the 'nmsrs' member (which indicates the
548 size of the entries array) and the 'index' member of each array entry.
549 kvm will fill in the 'data' member.
557 Parameters: struct kvm_msrs (in)
558 Returns: 0 on success, -1 on error
560 Writes model-specific registers to the vcpu. See KVM_GET_MSRS for the
563 Application code should set the 'nmsrs' member (which indicates the
564 size of the entries array), and the 'index' and 'data' members of each
573 Parameters: struct kvm_cpuid (in)
574 Returns: 0 on success, -1 on error
576 Defines the vcpu responses to the cpuid instruction. Applications
577 should use the KVM_SET_CPUID2 ioctl if available.
580 struct kvm_cpuid_entry {
589 /* for KVM_SET_CPUID */
593 struct kvm_cpuid_entry entries[0];
597 4.21 KVM_SET_SIGNAL_MASK
602 Parameters: struct kvm_signal_mask (in)
603 Returns: 0 on success, -1 on error
605 Defines which signals are blocked during execution of KVM_RUN. This
606 signal mask temporarily overrides the threads signal mask. Any
607 unblocked signal received (except SIGKILL and SIGSTOP, which retain
608 their traditional behaviour) will cause KVM_RUN to return with -EINTR.
610 Note the signal will only be delivered if not blocked by the original
613 /* for KVM_SET_SIGNAL_MASK */
614 struct kvm_signal_mask {
625 Parameters: struct kvm_fpu (out)
626 Returns: 0 on success, -1 on error
628 Reads the floating point state from the vcpu.
630 /* for KVM_GET_FPU and KVM_SET_FPU */
635 __u8 ftwx; /* in fxsave format */
651 Parameters: struct kvm_fpu (in)
652 Returns: 0 on success, -1 on error
654 Writes the floating point state to the vcpu.
656 /* for KVM_GET_FPU and KVM_SET_FPU */
661 __u8 ftwx; /* in fxsave format */
672 4.24 KVM_CREATE_IRQCHIP
674 Capability: KVM_CAP_IRQCHIP, KVM_CAP_S390_IRQCHIP (s390)
675 Architectures: x86, ARM, arm64, s390
678 Returns: 0 on success, -1 on error
680 Creates an interrupt controller model in the kernel.
681 On x86, creates a virtual ioapic, a virtual PIC (two PICs, nested), and sets up
682 future vcpus to have a local APIC. IRQ routing for GSIs 0-15 is set to both
683 PIC and IOAPIC; GSI 16-23 only go to the IOAPIC.
684 On ARM/arm64, a GICv2 is created. Any other GIC versions require the usage of
685 KVM_CREATE_DEVICE, which also supports creating a GICv2. Using
686 KVM_CREATE_DEVICE is preferred over KVM_CREATE_IRQCHIP for GICv2.
687 On s390, a dummy irq routing table is created.
689 Note that on s390 the KVM_CAP_S390_IRQCHIP vm capability needs to be enabled
690 before KVM_CREATE_IRQCHIP can be used.
695 Capability: KVM_CAP_IRQCHIP
696 Architectures: x86, arm, arm64
698 Parameters: struct kvm_irq_level
699 Returns: 0 on success, -1 on error
701 Sets the level of a GSI input to the interrupt controller model in the kernel.
702 On some architectures it is required that an interrupt controller model has
703 been previously created with KVM_CREATE_IRQCHIP. Note that edge-triggered
704 interrupts require the level to be set to 1 and then back to 0.
706 On real hardware, interrupt pins can be active-low or active-high. This
707 does not matter for the level field of struct kvm_irq_level: 1 always
708 means active (asserted), 0 means inactive (deasserted).
710 x86 allows the operating system to program the interrupt polarity
711 (active-low/active-high) for level-triggered interrupts, and KVM used
712 to consider the polarity. However, due to bitrot in the handling of
713 active-low interrupts, the above convention is now valid on x86 too.
714 This is signaled by KVM_CAP_X86_IOAPIC_POLARITY_IGNORED. Userspace
715 should not present interrupts to the guest as active-low unless this
716 capability is present (or unless it is not using the in-kernel irqchip,
720 ARM/arm64 can signal an interrupt either at the CPU level, or at the
721 in-kernel irqchip (GIC), and for in-kernel irqchip can tell the GIC to
722 use PPIs designated for specific cpus. The irq field is interpreted
725 Â bits: | 31 ... 24 | 23 ... 16 | 15 ... 0 |
726 field: | irq_type | vcpu_index | irq_id |
728 The irq_type field has the following values:
729 - irq_type[0]: out-of-kernel GIC: irq_id 0 is IRQ, irq_id 1 is FIQ
730 - irq_type[1]: in-kernel GIC: SPI, irq_id between 32 and 1019 (incl.)
731 (the vcpu_index field is ignored)
732 - irq_type[2]: in-kernel GIC: PPI, irq_id between 16 and 31 (incl.)
734 (The irq_id field thus corresponds nicely to the IRQ ID in the ARM GIC specs)
736 In both cases, level is used to assert/deassert the line.
738 struct kvm_irq_level {
741 __s32 status; /* not used for KVM_IRQ_LEVEL */
743 __u32 level; /* 0 or 1 */
749 Capability: KVM_CAP_IRQCHIP
752 Parameters: struct kvm_irqchip (in/out)
753 Returns: 0 on success, -1 on error
755 Reads the state of a kernel interrupt controller created with
756 KVM_CREATE_IRQCHIP into a buffer provided by the caller.
759 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */
762 char dummy[512]; /* reserving space */
763 struct kvm_pic_state pic;
764 struct kvm_ioapic_state ioapic;
771 Capability: KVM_CAP_IRQCHIP
774 Parameters: struct kvm_irqchip (in)
775 Returns: 0 on success, -1 on error
777 Sets the state of a kernel interrupt controller created with
778 KVM_CREATE_IRQCHIP from a buffer provided by the caller.
781 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */
784 char dummy[512]; /* reserving space */
785 struct kvm_pic_state pic;
786 struct kvm_ioapic_state ioapic;
791 4.28 KVM_XEN_HVM_CONFIG
793 Capability: KVM_CAP_XEN_HVM
796 Parameters: struct kvm_xen_hvm_config (in)
797 Returns: 0 on success, -1 on error
799 Sets the MSR that the Xen HVM guest uses to initialize its hypercall
800 page, and provides the starting address and size of the hypercall
801 blobs in userspace. When the guest writes the MSR, kvm copies one
802 page of a blob (32- or 64-bit, depending on the vcpu mode) to guest
805 struct kvm_xen_hvm_config {
818 Capability: KVM_CAP_ADJUST_CLOCK
821 Parameters: struct kvm_clock_data (out)
822 Returns: 0 on success, -1 on error
824 Gets the current timestamp of kvmclock as seen by the current guest. In
825 conjunction with KVM_SET_CLOCK, it is used to ensure monotonicity on scenarios
828 When KVM_CAP_ADJUST_CLOCK is passed to KVM_CHECK_EXTENSION, it returns the
829 set of bits that KVM can return in struct kvm_clock_data's flag member.
831 The only flag defined now is KVM_CLOCK_TSC_STABLE. If set, the returned
832 value is the exact kvmclock value seen by all VCPUs at the instant
833 when KVM_GET_CLOCK was called. If clear, the returned value is simply
834 CLOCK_MONOTONIC plus a constant offset; the offset can be modified
835 with KVM_SET_CLOCK. KVM will try to make all VCPUs follow this clock,
836 but the exact value read by each VCPU could differ, because the host
839 struct kvm_clock_data {
840 __u64 clock; /* kvmclock current value */
848 Capability: KVM_CAP_ADJUST_CLOCK
851 Parameters: struct kvm_clock_data (in)
852 Returns: 0 on success, -1 on error
854 Sets the current timestamp of kvmclock to the value specified in its parameter.
855 In conjunction with KVM_GET_CLOCK, it is used to ensure monotonicity on scenarios
858 struct kvm_clock_data {
859 __u64 clock; /* kvmclock current value */
865 4.31 KVM_GET_VCPU_EVENTS
867 Capability: KVM_CAP_VCPU_EVENTS
868 Extended by: KVM_CAP_INTR_SHADOW
869 Architectures: x86, arm, arm64
871 Parameters: struct kvm_vcpu_event (out)
872 Returns: 0 on success, -1 on error
876 Gets currently pending exceptions, interrupts, and NMIs as well as related
879 struct kvm_vcpu_events {
908 __u8 exception_has_payload;
909 __u64 exception_payload;
912 The following bits are defined in the flags field:
914 - KVM_VCPUEVENT_VALID_SHADOW may be set to signal that
915 interrupt.shadow contains a valid state.
917 - KVM_VCPUEVENT_VALID_SMM may be set to signal that smi contains a
920 - KVM_VCPUEVENT_VALID_PAYLOAD may be set to signal that the
921 exception_has_payload, exception_payload, and exception.pending
922 fields contain a valid state. This bit will be set whenever
923 KVM_CAP_EXCEPTION_PAYLOAD is enabled.
927 If the guest accesses a device that is being emulated by the host kernel in
928 such a way that a real device would generate a physical SError, KVM may make
929 a virtual SError pending for that VCPU. This system error interrupt remains
930 pending until the guest takes the exception by unmasking PSTATE.A.
932 Running the VCPU may cause it to take a pending SError, or make an access that
933 causes an SError to become pending. The event's description is only valid while
934 the VPCU is not running.
936 This API provides a way to read and write the pending 'event' state that is not
937 visible to the guest. To save, restore or migrate a VCPU the struct representing
938 the state can be read then written using this GET/SET API, along with the other
939 guest-visible registers. It is not possible to 'cancel' an SError that has been
942 A device being emulated in user-space may also wish to generate an SError. To do
943 this the events structure can be populated by user-space. The current state
944 should be read first, to ensure no existing SError is pending. If an existing
945 SError is pending, the architecture's 'Multiple SError interrupts' rules should
946 be followed. (2.5.3 of DDI0587.a "ARM Reliability, Availability, and
947 Serviceability (RAS) Specification").
949 SError exceptions always have an ESR value. Some CPUs have the ability to
950 specify what the virtual SError's ESR value should be. These systems will
951 advertise KVM_CAP_ARM_INJECT_SERROR_ESR. In this case exception.has_esr will
952 always have a non-zero value when read, and the agent making an SError pending
953 should specify the ISS field in the lower 24 bits of exception.serror_esr. If
954 the system supports KVM_CAP_ARM_INJECT_SERROR_ESR, but user-space sets the events
955 with exception.has_esr as zero, KVM will choose an ESR.
957 Specifying exception.has_esr on a system that does not support it will return
958 -EINVAL. Setting anything other than the lower 24bits of exception.serror_esr
961 struct kvm_vcpu_events {
965 /* Align it to 8 bytes */
972 4.32 KVM_SET_VCPU_EVENTS
974 Capability: KVM_CAP_VCPU_EVENTS
975 Extended by: KVM_CAP_INTR_SHADOW
976 Architectures: x86, arm, arm64
978 Parameters: struct kvm_vcpu_event (in)
979 Returns: 0 on success, -1 on error
983 Set pending exceptions, interrupts, and NMIs as well as related states of the
986 See KVM_GET_VCPU_EVENTS for the data structure.
988 Fields that may be modified asynchronously by running VCPUs can be excluded
989 from the update. These fields are nmi.pending, sipi_vector, smi.smm,
990 smi.pending. Keep the corresponding bits in the flags field cleared to
991 suppress overwriting the current in-kernel state. The bits are:
993 KVM_VCPUEVENT_VALID_NMI_PENDING - transfer nmi.pending to the kernel
994 KVM_VCPUEVENT_VALID_SIPI_VECTOR - transfer sipi_vector
995 KVM_VCPUEVENT_VALID_SMM - transfer the smi sub-struct.
997 If KVM_CAP_INTR_SHADOW is available, KVM_VCPUEVENT_VALID_SHADOW can be set in
998 the flags field to signal that interrupt.shadow contains a valid state and
999 shall be written into the VCPU.
1001 KVM_VCPUEVENT_VALID_SMM can only be set if KVM_CAP_X86_SMM is available.
1003 If KVM_CAP_EXCEPTION_PAYLOAD is enabled, KVM_VCPUEVENT_VALID_PAYLOAD
1004 can be set in the flags field to signal that the
1005 exception_has_payload, exception_payload, and exception.pending fields
1006 contain a valid state and shall be written into the VCPU.
1010 Set the pending SError exception state for this VCPU. It is not possible to
1011 'cancel' an Serror that has been made pending.
1013 See KVM_GET_VCPU_EVENTS for the data structure.
1016 4.33 KVM_GET_DEBUGREGS
1018 Capability: KVM_CAP_DEBUGREGS
1021 Parameters: struct kvm_debugregs (out)
1022 Returns: 0 on success, -1 on error
1024 Reads debug registers from the vcpu.
1026 struct kvm_debugregs {
1035 4.34 KVM_SET_DEBUGREGS
1037 Capability: KVM_CAP_DEBUGREGS
1040 Parameters: struct kvm_debugregs (in)
1041 Returns: 0 on success, -1 on error
1043 Writes debug registers into the vcpu.
1045 See KVM_GET_DEBUGREGS for the data structure. The flags field is unused
1046 yet and must be cleared on entry.
1049 4.35 KVM_SET_USER_MEMORY_REGION
1051 Capability: KVM_CAP_USER_MEM
1054 Parameters: struct kvm_userspace_memory_region (in)
1055 Returns: 0 on success, -1 on error
1057 struct kvm_userspace_memory_region {
1060 __u64 guest_phys_addr;
1061 __u64 memory_size; /* bytes */
1062 __u64 userspace_addr; /* start of the userspace allocated memory */
1065 /* for kvm_memory_region::flags */
1066 #define KVM_MEM_LOG_DIRTY_PAGES (1UL << 0)
1067 #define KVM_MEM_READONLY (1UL << 1)
1069 This ioctl allows the user to create or modify a guest physical memory
1070 slot. When changing an existing slot, it may be moved in the guest
1071 physical memory space, or its flags may be modified. It may not be
1072 resized. Slots may not overlap in guest physical address space.
1073 Bits 0-15 of "slot" specifies the slot id and this value should be
1074 less than the maximum number of user memory slots supported per VM.
1075 The maximum allowed slots can be queried using KVM_CAP_NR_MEMSLOTS,
1076 if this capability is supported by the architecture.
1078 If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of "slot"
1079 specifies the address space which is being modified. They must be
1080 less than the value that KVM_CHECK_EXTENSION returns for the
1081 KVM_CAP_MULTI_ADDRESS_SPACE capability. Slots in separate address spaces
1082 are unrelated; the restriction on overlapping slots only applies within
1085 Memory for the region is taken starting at the address denoted by the
1086 field userspace_addr, which must point at user addressable memory for
1087 the entire memory slot size. Any object may back this memory, including
1088 anonymous memory, ordinary files, and hugetlbfs.
1090 It is recommended that the lower 21 bits of guest_phys_addr and userspace_addr
1091 be identical. This allows large pages in the guest to be backed by large
1094 The flags field supports two flags: KVM_MEM_LOG_DIRTY_PAGES and
1095 KVM_MEM_READONLY. The former can be set to instruct KVM to keep track of
1096 writes to memory within the slot. See KVM_GET_DIRTY_LOG ioctl to know how to
1097 use it. The latter can be set, if KVM_CAP_READONLY_MEM capability allows it,
1098 to make a new slot read-only. In this case, writes to this memory will be
1099 posted to userspace as KVM_EXIT_MMIO exits.
1101 When the KVM_CAP_SYNC_MMU capability is available, changes in the backing of
1102 the memory region are automatically reflected into the guest. For example, an
1103 mmap() that affects the region will be made visible immediately. Another
1104 example is madvise(MADV_DROP).
1106 It is recommended to use this API instead of the KVM_SET_MEMORY_REGION ioctl.
1107 The KVM_SET_MEMORY_REGION does not allow fine grained control over memory
1108 allocation and is deprecated.
1111 4.36 KVM_SET_TSS_ADDR
1113 Capability: KVM_CAP_SET_TSS_ADDR
1116 Parameters: unsigned long tss_address (in)
1117 Returns: 0 on success, -1 on error
1119 This ioctl defines the physical address of a three-page region in the guest
1120 physical address space. The region must be within the first 4GB of the
1121 guest physical address space and must not conflict with any memory slot
1122 or any mmio address. The guest may malfunction if it accesses this memory
1125 This ioctl is required on Intel-based hosts. This is needed on Intel hardware
1126 because of a quirk in the virtualization implementation (see the internals
1127 documentation when it pops into existence).
1132 Capability: KVM_CAP_ENABLE_CAP
1133 Architectures: mips, ppc, s390
1135 Parameters: struct kvm_enable_cap (in)
1136 Returns: 0 on success; -1 on error
1138 Capability: KVM_CAP_ENABLE_CAP_VM
1141 Parameters: struct kvm_enable_cap (in)
1142 Returns: 0 on success; -1 on error
1144 +Not all extensions are enabled by default. Using this ioctl the application
1145 can enable an extension, making it available to the guest.
1147 On systems that do not support this ioctl, it always fails. On systems that
1148 do support it, it only works for extensions that are supported for enablement.
1150 To check if a capability can be enabled, the KVM_CHECK_EXTENSION ioctl should
1153 struct kvm_enable_cap {
1157 The capability that is supposed to get enabled.
1161 A bitfield indicating future enhancements. Has to be 0 for now.
1165 Arguments for enabling a feature. If a feature needs initial values to
1166 function properly, this is the place to put them.
1171 The vcpu ioctl should be used for vcpu-specific capabilities, the vm ioctl
1172 for vm-wide capabilities.
1174 4.38 KVM_GET_MP_STATE
1176 Capability: KVM_CAP_MP_STATE
1177 Architectures: x86, s390, arm, arm64
1179 Parameters: struct kvm_mp_state (out)
1180 Returns: 0 on success; -1 on error
1182 struct kvm_mp_state {
1186 Returns the vcpu's current "multiprocessing state" (though also valid on
1187 uniprocessor guests).
1189 Possible values are:
1191 - KVM_MP_STATE_RUNNABLE: the vcpu is currently running [x86,arm/arm64]
1192 - KVM_MP_STATE_UNINITIALIZED: the vcpu is an application processor (AP)
1193 which has not yet received an INIT signal [x86]
1194 - KVM_MP_STATE_INIT_RECEIVED: the vcpu has received an INIT signal, and is
1195 now ready for a SIPI [x86]
1196 - KVM_MP_STATE_HALTED: the vcpu has executed a HLT instruction and
1197 is waiting for an interrupt [x86]
1198 - KVM_MP_STATE_SIPI_RECEIVED: the vcpu has just received a SIPI (vector
1199 accessible via KVM_GET_VCPU_EVENTS) [x86]
1200 - KVM_MP_STATE_STOPPED: the vcpu is stopped [s390,arm/arm64]
1201 - KVM_MP_STATE_CHECK_STOP: the vcpu is in a special error state [s390]
1202 - KVM_MP_STATE_OPERATING: the vcpu is operating (running or halted)
1204 - KVM_MP_STATE_LOAD: the vcpu is in a special load/startup state
1207 On x86, this ioctl is only useful after KVM_CREATE_IRQCHIP. Without an
1208 in-kernel irqchip, the multiprocessing state must be maintained by userspace on
1209 these architectures.
1213 The only states that are valid are KVM_MP_STATE_STOPPED and
1214 KVM_MP_STATE_RUNNABLE which reflect if the vcpu is paused or not.
1216 4.39 KVM_SET_MP_STATE
1218 Capability: KVM_CAP_MP_STATE
1219 Architectures: x86, s390, arm, arm64
1221 Parameters: struct kvm_mp_state (in)
1222 Returns: 0 on success; -1 on error
1224 Sets the vcpu's current "multiprocessing state"; see KVM_GET_MP_STATE for
1227 On x86, this ioctl is only useful after KVM_CREATE_IRQCHIP. Without an
1228 in-kernel irqchip, the multiprocessing state must be maintained by userspace on
1229 these architectures.
1233 The only states that are valid are KVM_MP_STATE_STOPPED and
1234 KVM_MP_STATE_RUNNABLE which reflect if the vcpu should be paused or not.
1236 4.40 KVM_SET_IDENTITY_MAP_ADDR
1238 Capability: KVM_CAP_SET_IDENTITY_MAP_ADDR
1241 Parameters: unsigned long identity (in)
1242 Returns: 0 on success, -1 on error
1244 This ioctl defines the physical address of a one-page region in the guest
1245 physical address space. The region must be within the first 4GB of the
1246 guest physical address space and must not conflict with any memory slot
1247 or any mmio address. The guest may malfunction if it accesses this memory
1250 Setting the address to 0 will result in resetting the address to its default
1253 This ioctl is required on Intel-based hosts. This is needed on Intel hardware
1254 because of a quirk in the virtualization implementation (see the internals
1255 documentation when it pops into existence).
1257 Fails if any VCPU has already been created.
1259 4.41 KVM_SET_BOOT_CPU_ID
1261 Capability: KVM_CAP_SET_BOOT_CPU_ID
1264 Parameters: unsigned long vcpu_id
1265 Returns: 0 on success, -1 on error
1267 Define which vcpu is the Bootstrap Processor (BSP). Values are the same
1268 as the vcpu id in KVM_CREATE_VCPU. If this ioctl is not called, the default
1274 Capability: KVM_CAP_XSAVE
1277 Parameters: struct kvm_xsave (out)
1278 Returns: 0 on success, -1 on error
1284 This ioctl would copy current vcpu's xsave struct to the userspace.
1289 Capability: KVM_CAP_XSAVE
1292 Parameters: struct kvm_xsave (in)
1293 Returns: 0 on success, -1 on error
1299 This ioctl would copy userspace's xsave struct to the kernel.
1304 Capability: KVM_CAP_XCRS
1307 Parameters: struct kvm_xcrs (out)
1308 Returns: 0 on success, -1 on error
1319 struct kvm_xcr xcrs[KVM_MAX_XCRS];
1323 This ioctl would copy current vcpu's xcrs to the userspace.
1328 Capability: KVM_CAP_XCRS
1331 Parameters: struct kvm_xcrs (in)
1332 Returns: 0 on success, -1 on error
1343 struct kvm_xcr xcrs[KVM_MAX_XCRS];
1347 This ioctl would set vcpu's xcr to the value userspace specified.
1350 4.46 KVM_GET_SUPPORTED_CPUID
1352 Capability: KVM_CAP_EXT_CPUID
1355 Parameters: struct kvm_cpuid2 (in/out)
1356 Returns: 0 on success, -1 on error
1361 struct kvm_cpuid_entry2 entries[0];
1364 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX BIT(0)
1365 #define KVM_CPUID_FLAG_STATEFUL_FUNC BIT(1)
1366 #define KVM_CPUID_FLAG_STATE_READ_NEXT BIT(2)
1368 struct kvm_cpuid_entry2 {
1379 This ioctl returns x86 cpuid features which are supported by both the
1380 hardware and kvm in its default configuration. Userspace can use the
1381 information returned by this ioctl to construct cpuid information (for
1382 KVM_SET_CPUID2) that is consistent with hardware, kernel, and
1383 userspace capabilities, and with user requirements (for example, the
1384 user may wish to constrain cpuid to emulate older hardware, or for
1385 feature consistency across a cluster).
1387 Note that certain capabilities, such as KVM_CAP_X86_DISABLE_EXITS, may
1388 expose cpuid features (e.g. MONITOR) which are not supported by kvm in
1389 its default configuration. If userspace enables such capabilities, it
1390 is responsible for modifying the results of this ioctl appropriately.
1392 Userspace invokes KVM_GET_SUPPORTED_CPUID by passing a kvm_cpuid2 structure
1393 with the 'nent' field indicating the number of entries in the variable-size
1394 array 'entries'. If the number of entries is too low to describe the cpu
1395 capabilities, an error (E2BIG) is returned. If the number is too high,
1396 the 'nent' field is adjusted and an error (ENOMEM) is returned. If the
1397 number is just right, the 'nent' field is adjusted to the number of valid
1398 entries in the 'entries' array, which is then filled.
1400 The entries returned are the host cpuid as returned by the cpuid instruction,
1401 with unknown or unsupported features masked out. Some features (for example,
1402 x2apic), may not be present in the host cpu, but are exposed by kvm if it can
1403 emulate them efficiently. The fields in each entry are defined as follows:
1405 function: the eax value used to obtain the entry
1406 index: the ecx value used to obtain the entry (for entries that are
1408 flags: an OR of zero or more of the following:
1409 KVM_CPUID_FLAG_SIGNIFCANT_INDEX:
1410 if the index field is valid
1411 KVM_CPUID_FLAG_STATEFUL_FUNC:
1412 if cpuid for this function returns different values for successive
1413 invocations; there will be several entries with the same function,
1414 all with this flag set
1415 KVM_CPUID_FLAG_STATE_READ_NEXT:
1416 for KVM_CPUID_FLAG_STATEFUL_FUNC entries, set if this entry is
1417 the first entry to be read by a cpu
1418 eax, ebx, ecx, edx: the values returned by the cpuid instruction for
1419 this function/index combination
1421 The TSC deadline timer feature (CPUID leaf 1, ecx[24]) is always returned
1422 as false, since the feature depends on KVM_CREATE_IRQCHIP for local APIC
1423 support. Instead it is reported via
1425 ioctl(KVM_CHECK_EXTENSION, KVM_CAP_TSC_DEADLINE_TIMER)
1427 if that returns true and you use KVM_CREATE_IRQCHIP, or if you emulate the
1428 feature in userspace, then you can enable the feature for KVM_SET_CPUID2.
1431 4.47 KVM_PPC_GET_PVINFO
1433 Capability: KVM_CAP_PPC_GET_PVINFO
1436 Parameters: struct kvm_ppc_pvinfo (out)
1437 Returns: 0 on success, !0 on error
1439 struct kvm_ppc_pvinfo {
1445 This ioctl fetches PV specific information that need to be passed to the guest
1446 using the device tree or other means from vm context.
1448 The hcall array defines 4 instructions that make up a hypercall.
1450 If any additional field gets added to this structure later on, a bit for that
1451 additional piece of information will be set in the flags bitmap.
1453 The flags bitmap is defined as:
1455 /* the host supports the ePAPR idle hcall
1456 #define KVM_PPC_PVINFO_FLAGS_EV_IDLE (1<<0)
1458 4.52 KVM_SET_GSI_ROUTING
1460 Capability: KVM_CAP_IRQ_ROUTING
1461 Architectures: x86 s390 arm arm64
1463 Parameters: struct kvm_irq_routing (in)
1464 Returns: 0 on success, -1 on error
1466 Sets the GSI routing table entries, overwriting any previously set entries.
1468 On arm/arm64, GSI routing has the following limitation:
1469 - GSI routing does not apply to KVM_IRQ_LINE but only to KVM_IRQFD.
1471 struct kvm_irq_routing {
1474 struct kvm_irq_routing_entry entries[0];
1477 No flags are specified so far, the corresponding field must be set to zero.
1479 struct kvm_irq_routing_entry {
1485 struct kvm_irq_routing_irqchip irqchip;
1486 struct kvm_irq_routing_msi msi;
1487 struct kvm_irq_routing_s390_adapter adapter;
1488 struct kvm_irq_routing_hv_sint hv_sint;
1493 /* gsi routing entry types */
1494 #define KVM_IRQ_ROUTING_IRQCHIP 1
1495 #define KVM_IRQ_ROUTING_MSI 2
1496 #define KVM_IRQ_ROUTING_S390_ADAPTER 3
1497 #define KVM_IRQ_ROUTING_HV_SINT 4
1500 - KVM_MSI_VALID_DEVID: used along with KVM_IRQ_ROUTING_MSI routing entry
1501 type, specifies that the devid field contains a valid value. The per-VM
1502 KVM_CAP_MSI_DEVID capability advertises the requirement to provide
1503 the device ID. If this capability is not available, userspace should
1504 never set the KVM_MSI_VALID_DEVID flag as the ioctl might fail.
1507 struct kvm_irq_routing_irqchip {
1512 struct kvm_irq_routing_msi {
1522 If KVM_MSI_VALID_DEVID is set, devid contains a unique device identifier
1523 for the device that wrote the MSI message. For PCI, this is usually a
1524 BFD identifier in the lower 16 bits.
1526 On x86, address_hi is ignored unless the KVM_X2APIC_API_USE_32BIT_IDS
1527 feature of KVM_CAP_X2APIC_API capability is enabled. If it is enabled,
1528 address_hi bits 31-8 provide bits 31-8 of the destination id. Bits 7-0 of
1529 address_hi must be zero.
1531 struct kvm_irq_routing_s390_adapter {
1535 __u32 summary_offset;
1539 struct kvm_irq_routing_hv_sint {
1545 4.55 KVM_SET_TSC_KHZ
1547 Capability: KVM_CAP_TSC_CONTROL
1550 Parameters: virtual tsc_khz
1551 Returns: 0 on success, -1 on error
1553 Specifies the tsc frequency for the virtual machine. The unit of the
1557 4.56 KVM_GET_TSC_KHZ
1559 Capability: KVM_CAP_GET_TSC_KHZ
1563 Returns: virtual tsc-khz on success, negative value on error
1565 Returns the tsc frequency of the guest. The unit of the return value is
1566 KHz. If the host has unstable tsc this ioctl returns -EIO instead as an
1572 Capability: KVM_CAP_IRQCHIP
1575 Parameters: struct kvm_lapic_state (out)
1576 Returns: 0 on success, -1 on error
1578 #define KVM_APIC_REG_SIZE 0x400
1579 struct kvm_lapic_state {
1580 char regs[KVM_APIC_REG_SIZE];
1583 Reads the Local APIC registers and copies them into the input argument. The
1584 data format and layout are the same as documented in the architecture manual.
1586 If KVM_X2APIC_API_USE_32BIT_IDS feature of KVM_CAP_X2APIC_API is
1587 enabled, then the format of APIC_ID register depends on the APIC mode
1588 (reported by MSR_IA32_APICBASE) of its VCPU. x2APIC stores APIC ID in
1589 the APIC_ID register (bytes 32-35). xAPIC only allows an 8-bit APIC ID
1590 which is stored in bits 31-24 of the APIC register, or equivalently in
1591 byte 35 of struct kvm_lapic_state's regs field. KVM_GET_LAPIC must then
1592 be called after MSR_IA32_APICBASE has been set with KVM_SET_MSR.
1594 If KVM_X2APIC_API_USE_32BIT_IDS feature is disabled, struct kvm_lapic_state
1595 always uses xAPIC format.
1600 Capability: KVM_CAP_IRQCHIP
1603 Parameters: struct kvm_lapic_state (in)
1604 Returns: 0 on success, -1 on error
1606 #define KVM_APIC_REG_SIZE 0x400
1607 struct kvm_lapic_state {
1608 char regs[KVM_APIC_REG_SIZE];
1611 Copies the input argument into the Local APIC registers. The data format
1612 and layout are the same as documented in the architecture manual.
1614 The format of the APIC ID register (bytes 32-35 of struct kvm_lapic_state's
1615 regs field) depends on the state of the KVM_CAP_X2APIC_API capability.
1616 See the note in KVM_GET_LAPIC.
1621 Capability: KVM_CAP_IOEVENTFD
1624 Parameters: struct kvm_ioeventfd (in)
1625 Returns: 0 on success, !0 on error
1627 This ioctl attaches or detaches an ioeventfd to a legal pio/mmio address
1628 within the guest. A guest write in the registered address will signal the
1629 provided event instead of triggering an exit.
1631 struct kvm_ioeventfd {
1633 __u64 addr; /* legal pio/mmio address */
1634 __u32 len; /* 0, 1, 2, 4, or 8 bytes */
1640 For the special case of virtio-ccw devices on s390, the ioevent is matched
1641 to a subchannel/virtqueue tuple instead.
1643 The following flags are defined:
1645 #define KVM_IOEVENTFD_FLAG_DATAMATCH (1 << kvm_ioeventfd_flag_nr_datamatch)
1646 #define KVM_IOEVENTFD_FLAG_PIO (1 << kvm_ioeventfd_flag_nr_pio)
1647 #define KVM_IOEVENTFD_FLAG_DEASSIGN (1 << kvm_ioeventfd_flag_nr_deassign)
1648 #define KVM_IOEVENTFD_FLAG_VIRTIO_CCW_NOTIFY \
1649 (1 << kvm_ioeventfd_flag_nr_virtio_ccw_notify)
1651 If datamatch flag is set, the event will be signaled only if the written value
1652 to the registered address is equal to datamatch in struct kvm_ioeventfd.
1654 For virtio-ccw devices, addr contains the subchannel id and datamatch the
1657 With KVM_CAP_IOEVENTFD_ANY_LENGTH, a zero length ioeventfd is allowed, and
1658 the kernel will ignore the length of guest write and may get a faster vmexit.
1659 The speedup may only apply to specific architectures, but the ioeventfd will
1664 Capability: KVM_CAP_SW_TLB
1667 Parameters: struct kvm_dirty_tlb (in)
1668 Returns: 0 on success, -1 on error
1670 struct kvm_dirty_tlb {
1675 This must be called whenever userspace has changed an entry in the shared
1676 TLB, prior to calling KVM_RUN on the associated vcpu.
1678 The "bitmap" field is the userspace address of an array. This array
1679 consists of a number of bits, equal to the total number of TLB entries as
1680 determined by the last successful call to KVM_CONFIG_TLB, rounded up to the
1681 nearest multiple of 64.
1683 Each bit corresponds to one TLB entry, ordered the same as in the shared TLB
1686 The array is little-endian: the bit 0 is the least significant bit of the
1687 first byte, bit 8 is the least significant bit of the second byte, etc.
1688 This avoids any complications with differing word sizes.
1690 The "num_dirty" field is a performance hint for KVM to determine whether it
1691 should skip processing the bitmap and just invalidate everything. It must
1692 be set to the number of set bits in the bitmap.
1695 4.62 KVM_CREATE_SPAPR_TCE
1697 Capability: KVM_CAP_SPAPR_TCE
1698 Architectures: powerpc
1700 Parameters: struct kvm_create_spapr_tce (in)
1701 Returns: file descriptor for manipulating the created TCE table
1703 This creates a virtual TCE (translation control entry) table, which
1704 is an IOMMU for PAPR-style virtual I/O. It is used to translate
1705 logical addresses used in virtual I/O into guest physical addresses,
1706 and provides a scatter/gather capability for PAPR virtual I/O.
1708 /* for KVM_CAP_SPAPR_TCE */
1709 struct kvm_create_spapr_tce {
1714 The liobn field gives the logical IO bus number for which to create a
1715 TCE table. The window_size field specifies the size of the DMA window
1716 which this TCE table will translate - the table will contain one 64
1717 bit TCE entry for every 4kiB of the DMA window.
1719 When the guest issues an H_PUT_TCE hcall on a liobn for which a TCE
1720 table has been created using this ioctl(), the kernel will handle it
1721 in real mode, updating the TCE table. H_PUT_TCE calls for other
1722 liobns will cause a vm exit and must be handled by userspace.
1724 The return value is a file descriptor which can be passed to mmap(2)
1725 to map the created TCE table into userspace. This lets userspace read
1726 the entries written by kernel-handled H_PUT_TCE calls, and also lets
1727 userspace update the TCE table directly which is useful in some
1731 4.63 KVM_ALLOCATE_RMA
1733 Capability: KVM_CAP_PPC_RMA
1734 Architectures: powerpc
1736 Parameters: struct kvm_allocate_rma (out)
1737 Returns: file descriptor for mapping the allocated RMA
1739 This allocates a Real Mode Area (RMA) from the pool allocated at boot
1740 time by the kernel. An RMA is a physically-contiguous, aligned region
1741 of memory used on older POWER processors to provide the memory which
1742 will be accessed by real-mode (MMU off) accesses in a KVM guest.
1743 POWER processors support a set of sizes for the RMA that usually
1744 includes 64MB, 128MB, 256MB and some larger powers of two.
1746 /* for KVM_ALLOCATE_RMA */
1747 struct kvm_allocate_rma {
1751 The return value is a file descriptor which can be passed to mmap(2)
1752 to map the allocated RMA into userspace. The mapped area can then be
1753 passed to the KVM_SET_USER_MEMORY_REGION ioctl to establish it as the
1754 RMA for a virtual machine. The size of the RMA in bytes (which is
1755 fixed at host kernel boot time) is returned in the rma_size field of
1756 the argument structure.
1758 The KVM_CAP_PPC_RMA capability is 1 or 2 if the KVM_ALLOCATE_RMA ioctl
1759 is supported; 2 if the processor requires all virtual machines to have
1760 an RMA, or 1 if the processor can use an RMA but doesn't require it,
1761 because it supports the Virtual RMA (VRMA) facility.
1766 Capability: KVM_CAP_USER_NMI
1770 Returns: 0 on success, -1 on error
1772 Queues an NMI on the thread's vcpu. Note this is well defined only
1773 when KVM_CREATE_IRQCHIP has not been called, since this is an interface
1774 between the virtual cpu core and virtual local APIC. After KVM_CREATE_IRQCHIP
1775 has been called, this interface is completely emulated within the kernel.
1777 To use this to emulate the LINT1 input with KVM_CREATE_IRQCHIP, use the
1778 following algorithm:
1781 - read the local APIC's state (KVM_GET_LAPIC)
1782 - check whether changing LINT1 will queue an NMI (see the LVT entry for LINT1)
1783 - if so, issue KVM_NMI
1786 Some guests configure the LINT1 NMI input to cause a panic, aiding in
1790 4.65 KVM_S390_UCAS_MAP
1792 Capability: KVM_CAP_S390_UCONTROL
1795 Parameters: struct kvm_s390_ucas_mapping (in)
1796 Returns: 0 in case of success
1798 The parameter is defined like this:
1799 struct kvm_s390_ucas_mapping {
1805 This ioctl maps the memory at "user_addr" with the length "length" to
1806 the vcpu's address space starting at "vcpu_addr". All parameters need to
1807 be aligned by 1 megabyte.
1810 4.66 KVM_S390_UCAS_UNMAP
1812 Capability: KVM_CAP_S390_UCONTROL
1815 Parameters: struct kvm_s390_ucas_mapping (in)
1816 Returns: 0 in case of success
1818 The parameter is defined like this:
1819 struct kvm_s390_ucas_mapping {
1825 This ioctl unmaps the memory in the vcpu's address space starting at
1826 "vcpu_addr" with the length "length". The field "user_addr" is ignored.
1827 All parameters need to be aligned by 1 megabyte.
1830 4.67 KVM_S390_VCPU_FAULT
1832 Capability: KVM_CAP_S390_UCONTROL
1835 Parameters: vcpu absolute address (in)
1836 Returns: 0 in case of success
1838 This call creates a page table entry on the virtual cpu's address space
1839 (for user controlled virtual machines) or the virtual machine's address
1840 space (for regular virtual machines). This only works for minor faults,
1841 thus it's recommended to access subject memory page via the user page
1842 table upfront. This is useful to handle validity intercepts for user
1843 controlled virtual machines to fault in the virtual cpu's lowcore pages
1844 prior to calling the KVM_RUN ioctl.
1847 4.68 KVM_SET_ONE_REG
1849 Capability: KVM_CAP_ONE_REG
1852 Parameters: struct kvm_one_reg (in)
1853 Returns: 0 on success, negative value on failure
1855 struct kvm_one_reg {
1860 Using this ioctl, a single vcpu register can be set to a specific value
1861 defined by user space with the passed in struct kvm_one_reg, where id
1862 refers to the register identifier as described below and addr is a pointer
1863 to a variable with the respective size. There can be architecture agnostic
1864 and architecture specific registers. Each have their own range of operation
1865 and their own constants and width. To keep track of the implemented
1866 registers, find a list below:
1868 Arch | Register | Width (bits)
1870 PPC | KVM_REG_PPC_HIOR | 64
1871 PPC | KVM_REG_PPC_IAC1 | 64
1872 PPC | KVM_REG_PPC_IAC2 | 64
1873 PPC | KVM_REG_PPC_IAC3 | 64
1874 PPC | KVM_REG_PPC_IAC4 | 64
1875 PPC | KVM_REG_PPC_DAC1 | 64
1876 PPC | KVM_REG_PPC_DAC2 | 64
1877 PPC | KVM_REG_PPC_DABR | 64
1878 PPC | KVM_REG_PPC_DSCR | 64
1879 PPC | KVM_REG_PPC_PURR | 64
1880 PPC | KVM_REG_PPC_SPURR | 64
1881 PPC | KVM_REG_PPC_DAR | 64
1882 PPC | KVM_REG_PPC_DSISR | 32
1883 PPC | KVM_REG_PPC_AMR | 64
1884 PPC | KVM_REG_PPC_UAMOR | 64
1885 PPC | KVM_REG_PPC_MMCR0 | 64
1886 PPC | KVM_REG_PPC_MMCR1 | 64
1887 PPC | KVM_REG_PPC_MMCRA | 64
1888 PPC | KVM_REG_PPC_MMCR2 | 64
1889 PPC | KVM_REG_PPC_MMCRS | 64
1890 PPC | KVM_REG_PPC_SIAR | 64
1891 PPC | KVM_REG_PPC_SDAR | 64
1892 PPC | KVM_REG_PPC_SIER | 64
1893 PPC | KVM_REG_PPC_PMC1 | 32
1894 PPC | KVM_REG_PPC_PMC2 | 32
1895 PPC | KVM_REG_PPC_PMC3 | 32
1896 PPC | KVM_REG_PPC_PMC4 | 32
1897 PPC | KVM_REG_PPC_PMC5 | 32
1898 PPC | KVM_REG_PPC_PMC6 | 32
1899 PPC | KVM_REG_PPC_PMC7 | 32
1900 PPC | KVM_REG_PPC_PMC8 | 32
1901 PPC | KVM_REG_PPC_FPR0 | 64
1903 PPC | KVM_REG_PPC_FPR31 | 64
1904 PPC | KVM_REG_PPC_VR0 | 128
1906 PPC | KVM_REG_PPC_VR31 | 128
1907 PPC | KVM_REG_PPC_VSR0 | 128
1909 PPC | KVM_REG_PPC_VSR31 | 128
1910 PPC | KVM_REG_PPC_FPSCR | 64
1911 PPC | KVM_REG_PPC_VSCR | 32
1912 PPC | KVM_REG_PPC_VPA_ADDR | 64
1913 PPC | KVM_REG_PPC_VPA_SLB | 128
1914 PPC | KVM_REG_PPC_VPA_DTL | 128
1915 PPC | KVM_REG_PPC_EPCR | 32
1916 PPC | KVM_REG_PPC_EPR | 32
1917 PPC | KVM_REG_PPC_TCR | 32
1918 PPC | KVM_REG_PPC_TSR | 32
1919 PPC | KVM_REG_PPC_OR_TSR | 32
1920 PPC | KVM_REG_PPC_CLEAR_TSR | 32
1921 PPC | KVM_REG_PPC_MAS0 | 32
1922 PPC | KVM_REG_PPC_MAS1 | 32
1923 PPC | KVM_REG_PPC_MAS2 | 64
1924 PPC | KVM_REG_PPC_MAS7_3 | 64
1925 PPC | KVM_REG_PPC_MAS4 | 32
1926 PPC | KVM_REG_PPC_MAS6 | 32
1927 PPC | KVM_REG_PPC_MMUCFG | 32
1928 PPC | KVM_REG_PPC_TLB0CFG | 32
1929 PPC | KVM_REG_PPC_TLB1CFG | 32
1930 PPC | KVM_REG_PPC_TLB2CFG | 32
1931 PPC | KVM_REG_PPC_TLB3CFG | 32
1932 PPC | KVM_REG_PPC_TLB0PS | 32
1933 PPC | KVM_REG_PPC_TLB1PS | 32
1934 PPC | KVM_REG_PPC_TLB2PS | 32
1935 PPC | KVM_REG_PPC_TLB3PS | 32
1936 PPC | KVM_REG_PPC_EPTCFG | 32
1937 PPC | KVM_REG_PPC_ICP_STATE | 64
1938 PPC | KVM_REG_PPC_TB_OFFSET | 64
1939 PPC | KVM_REG_PPC_SPMC1 | 32
1940 PPC | KVM_REG_PPC_SPMC2 | 32
1941 PPC | KVM_REG_PPC_IAMR | 64
1942 PPC | KVM_REG_PPC_TFHAR | 64
1943 PPC | KVM_REG_PPC_TFIAR | 64
1944 PPC | KVM_REG_PPC_TEXASR | 64
1945 PPC | KVM_REG_PPC_FSCR | 64
1946 PPC | KVM_REG_PPC_PSPB | 32
1947 PPC | KVM_REG_PPC_EBBHR | 64
1948 PPC | KVM_REG_PPC_EBBRR | 64
1949 PPC | KVM_REG_PPC_BESCR | 64
1950 PPC | KVM_REG_PPC_TAR | 64
1951 PPC | KVM_REG_PPC_DPDES | 64
1952 PPC | KVM_REG_PPC_DAWR | 64
1953 PPC | KVM_REG_PPC_DAWRX | 64
1954 PPC | KVM_REG_PPC_CIABR | 64
1955 PPC | KVM_REG_PPC_IC | 64
1956 PPC | KVM_REG_PPC_VTB | 64
1957 PPC | KVM_REG_PPC_CSIGR | 64
1958 PPC | KVM_REG_PPC_TACR | 64
1959 PPC | KVM_REG_PPC_TCSCR | 64
1960 PPC | KVM_REG_PPC_PID | 64
1961 PPC | KVM_REG_PPC_ACOP | 64
1962 PPC | KVM_REG_PPC_VRSAVE | 32
1963 PPC | KVM_REG_PPC_LPCR | 32
1964 PPC | KVM_REG_PPC_LPCR_64 | 64
1965 PPC | KVM_REG_PPC_PPR | 64
1966 PPC | KVM_REG_PPC_ARCH_COMPAT | 32
1967 PPC | KVM_REG_PPC_DABRX | 32
1968 PPC | KVM_REG_PPC_WORT | 64
1969 PPC | KVM_REG_PPC_SPRG9 | 64
1970 PPC | KVM_REG_PPC_DBSR | 32
1971 PPC | KVM_REG_PPC_TIDR | 64
1972 PPC | KVM_REG_PPC_PSSCR | 64
1973 PPC | KVM_REG_PPC_DEC_EXPIRY | 64
1974 PPC | KVM_REG_PPC_PTCR | 64
1975 PPC | KVM_REG_PPC_TM_GPR0 | 64
1977 PPC | KVM_REG_PPC_TM_GPR31 | 64
1978 PPC | KVM_REG_PPC_TM_VSR0 | 128
1980 PPC | KVM_REG_PPC_TM_VSR63 | 128
1981 PPC | KVM_REG_PPC_TM_CR | 64
1982 PPC | KVM_REG_PPC_TM_LR | 64
1983 PPC | KVM_REG_PPC_TM_CTR | 64
1984 PPC | KVM_REG_PPC_TM_FPSCR | 64
1985 PPC | KVM_REG_PPC_TM_AMR | 64
1986 PPC | KVM_REG_PPC_TM_PPR | 64
1987 PPC | KVM_REG_PPC_TM_VRSAVE | 64
1988 PPC | KVM_REG_PPC_TM_VSCR | 32
1989 PPC | KVM_REG_PPC_TM_DSCR | 64
1990 PPC | KVM_REG_PPC_TM_TAR | 64
1991 PPC | KVM_REG_PPC_TM_XER | 64
1993 MIPS | KVM_REG_MIPS_R0 | 64
1995 MIPS | KVM_REG_MIPS_R31 | 64
1996 MIPS | KVM_REG_MIPS_HI | 64
1997 MIPS | KVM_REG_MIPS_LO | 64
1998 MIPS | KVM_REG_MIPS_PC | 64
1999 MIPS | KVM_REG_MIPS_CP0_INDEX | 32
2000 MIPS | KVM_REG_MIPS_CP0_ENTRYLO0 | 64
2001 MIPS | KVM_REG_MIPS_CP0_ENTRYLO1 | 64
2002 MIPS | KVM_REG_MIPS_CP0_CONTEXT | 64
2003 MIPS | KVM_REG_MIPS_CP0_CONTEXTCONFIG| 32
2004 MIPS | KVM_REG_MIPS_CP0_USERLOCAL | 64
2005 MIPS | KVM_REG_MIPS_CP0_XCONTEXTCONFIG| 64
2006 MIPS | KVM_REG_MIPS_CP0_PAGEMASK | 32
2007 MIPS | KVM_REG_MIPS_CP0_PAGEGRAIN | 32
2008 MIPS | KVM_REG_MIPS_CP0_SEGCTL0 | 64
2009 MIPS | KVM_REG_MIPS_CP0_SEGCTL1 | 64
2010 MIPS | KVM_REG_MIPS_CP0_SEGCTL2 | 64
2011 MIPS | KVM_REG_MIPS_CP0_PWBASE | 64
2012 MIPS | KVM_REG_MIPS_CP0_PWFIELD | 64
2013 MIPS | KVM_REG_MIPS_CP0_PWSIZE | 64
2014 MIPS | KVM_REG_MIPS_CP0_WIRED | 32
2015 MIPS | KVM_REG_MIPS_CP0_PWCTL | 32
2016 MIPS | KVM_REG_MIPS_CP0_HWRENA | 32
2017 MIPS | KVM_REG_MIPS_CP0_BADVADDR | 64
2018 MIPS | KVM_REG_MIPS_CP0_BADINSTR | 32
2019 MIPS | KVM_REG_MIPS_CP0_BADINSTRP | 32
2020 MIPS | KVM_REG_MIPS_CP0_COUNT | 32
2021 MIPS | KVM_REG_MIPS_CP0_ENTRYHI | 64
2022 MIPS | KVM_REG_MIPS_CP0_COMPARE | 32
2023 MIPS | KVM_REG_MIPS_CP0_STATUS | 32
2024 MIPS | KVM_REG_MIPS_CP0_INTCTL | 32
2025 MIPS | KVM_REG_MIPS_CP0_CAUSE | 32
2026 MIPS | KVM_REG_MIPS_CP0_EPC | 64
2027 MIPS | KVM_REG_MIPS_CP0_PRID | 32
2028 MIPS | KVM_REG_MIPS_CP0_EBASE | 64
2029 MIPS | KVM_REG_MIPS_CP0_CONFIG | 32
2030 MIPS | KVM_REG_MIPS_CP0_CONFIG1 | 32
2031 MIPS | KVM_REG_MIPS_CP0_CONFIG2 | 32
2032 MIPS | KVM_REG_MIPS_CP0_CONFIG3 | 32
2033 MIPS | KVM_REG_MIPS_CP0_CONFIG4 | 32
2034 MIPS | KVM_REG_MIPS_CP0_CONFIG5 | 32
2035 MIPS | KVM_REG_MIPS_CP0_CONFIG7 | 32
2036 MIPS | KVM_REG_MIPS_CP0_XCONTEXT | 64
2037 MIPS | KVM_REG_MIPS_CP0_ERROREPC | 64
2038 MIPS | KVM_REG_MIPS_CP0_KSCRATCH1 | 64
2039 MIPS | KVM_REG_MIPS_CP0_KSCRATCH2 | 64
2040 MIPS | KVM_REG_MIPS_CP0_KSCRATCH3 | 64
2041 MIPS | KVM_REG_MIPS_CP0_KSCRATCH4 | 64
2042 MIPS | KVM_REG_MIPS_CP0_KSCRATCH5 | 64
2043 MIPS | KVM_REG_MIPS_CP0_KSCRATCH6 | 64
2044 MIPS | KVM_REG_MIPS_CP0_MAAR(0..63) | 64
2045 MIPS | KVM_REG_MIPS_COUNT_CTL | 64
2046 MIPS | KVM_REG_MIPS_COUNT_RESUME | 64
2047 MIPS | KVM_REG_MIPS_COUNT_HZ | 64
2048 MIPS | KVM_REG_MIPS_FPR_32(0..31) | 32
2049 MIPS | KVM_REG_MIPS_FPR_64(0..31) | 64
2050 MIPS | KVM_REG_MIPS_VEC_128(0..31) | 128
2051 MIPS | KVM_REG_MIPS_FCR_IR | 32
2052 MIPS | KVM_REG_MIPS_FCR_CSR | 32
2053 MIPS | KVM_REG_MIPS_MSA_IR | 32
2054 MIPS | KVM_REG_MIPS_MSA_CSR | 32
2056 ARM registers are mapped using the lower 32 bits. The upper 16 of that
2057 is the register group type, or coprocessor number:
2059 ARM core registers have the following id bit patterns:
2060 0x4020 0000 0010 <index into the kvm_regs struct:16>
2062 ARM 32-bit CP15 registers have the following id bit patterns:
2063 0x4020 0000 000F <zero:1> <crn:4> <crm:4> <opc1:4> <opc2:3>
2065 ARM 64-bit CP15 registers have the following id bit patterns:
2066 0x4030 0000 000F <zero:1> <zero:4> <crm:4> <opc1:4> <zero:3>
2068 ARM CCSIDR registers are demultiplexed by CSSELR value:
2069 0x4020 0000 0011 00 <csselr:8>
2071 ARM 32-bit VFP control registers have the following id bit patterns:
2072 0x4020 0000 0012 1 <regno:12>
2074 ARM 64-bit FP registers have the following id bit patterns:
2075 0x4030 0000 0012 0 <regno:12>
2077 ARM firmware pseudo-registers have the following bit pattern:
2078 0x4030 0000 0014 <regno:16>
2081 arm64 registers are mapped using the lower 32 bits. The upper 16 of
2082 that is the register group type, or coprocessor number:
2084 arm64 core/FP-SIMD registers have the following id bit patterns. Note
2085 that the size of the access is variable, as the kvm_regs structure
2086 contains elements ranging from 32 to 128 bits. The index is a 32bit
2087 value in the kvm_regs structure seen as a 32bit array.
2088 0x60x0 0000 0010 <index into the kvm_regs struct:16>
2090 arm64 CCSIDR registers are demultiplexed by CSSELR value:
2091 0x6020 0000 0011 00 <csselr:8>
2093 arm64 system registers have the following id bit patterns:
2094 0x6030 0000 0013 <op0:2> <op1:3> <crn:4> <crm:4> <op2:3>
2096 arm64 firmware pseudo-registers have the following bit pattern:
2097 0x6030 0000 0014 <regno:16>
2100 MIPS registers are mapped using the lower 32 bits. The upper 16 of that is
2101 the register group type:
2103 MIPS core registers (see above) have the following id bit patterns:
2104 0x7030 0000 0000 <reg:16>
2106 MIPS CP0 registers (see KVM_REG_MIPS_CP0_* above) have the following id bit
2107 patterns depending on whether they're 32-bit or 64-bit registers:
2108 0x7020 0000 0001 00 <reg:5> <sel:3> (32-bit)
2109 0x7030 0000 0001 00 <reg:5> <sel:3> (64-bit)
2111 Note: KVM_REG_MIPS_CP0_ENTRYLO0 and KVM_REG_MIPS_CP0_ENTRYLO1 are the MIPS64
2112 versions of the EntryLo registers regardless of the word size of the host
2113 hardware, host kernel, guest, and whether XPA is present in the guest, i.e.
2114 with the RI and XI bits (if they exist) in bits 63 and 62 respectively, and
2115 the PFNX field starting at bit 30.
2117 MIPS MAARs (see KVM_REG_MIPS_CP0_MAAR(*) above) have the following id bit
2119 0x7030 0000 0001 01 <reg:8>
2121 MIPS KVM control registers (see above) have the following id bit patterns:
2122 0x7030 0000 0002 <reg:16>
2124 MIPS FPU registers (see KVM_REG_MIPS_FPR_{32,64}() above) have the following
2125 id bit patterns depending on the size of the register being accessed. They are
2126 always accessed according to the current guest FPU mode (Status.FR and
2127 Config5.FRE), i.e. as the guest would see them, and they become unpredictable
2128 if the guest FPU mode is changed. MIPS SIMD Architecture (MSA) vector
2129 registers (see KVM_REG_MIPS_VEC_128() above) have similar patterns as they
2130 overlap the FPU registers:
2131 0x7020 0000 0003 00 <0:3> <reg:5> (32-bit FPU registers)
2132 0x7030 0000 0003 00 <0:3> <reg:5> (64-bit FPU registers)
2133 0x7040 0000 0003 00 <0:3> <reg:5> (128-bit MSA vector registers)
2135 MIPS FPU control registers (see KVM_REG_MIPS_FCR_{IR,CSR} above) have the
2136 following id bit patterns:
2137 0x7020 0000 0003 01 <0:3> <reg:5>
2139 MIPS MSA control registers (see KVM_REG_MIPS_MSA_{IR,CSR} above) have the
2140 following id bit patterns:
2141 0x7020 0000 0003 02 <0:3> <reg:5>
2144 4.69 KVM_GET_ONE_REG
2146 Capability: KVM_CAP_ONE_REG
2149 Parameters: struct kvm_one_reg (in and out)
2150 Returns: 0 on success, negative value on failure
2152 This ioctl allows to receive the value of a single register implemented
2153 in a vcpu. The register to read is indicated by the "id" field of the
2154 kvm_one_reg struct passed in. On success, the register value can be found
2155 at the memory location pointed to by "addr".
2157 The list of registers accessible using this interface is identical to the
2161 4.70 KVM_KVMCLOCK_CTRL
2163 Capability: KVM_CAP_KVMCLOCK_CTRL
2164 Architectures: Any that implement pvclocks (currently x86 only)
2167 Returns: 0 on success, -1 on error
2169 This signals to the host kernel that the specified guest is being paused by
2170 userspace. The host will set a flag in the pvclock structure that is checked
2171 from the soft lockup watchdog. The flag is part of the pvclock structure that
2172 is shared between guest and host, specifically the second bit of the flags
2173 field of the pvclock_vcpu_time_info structure. It will be set exclusively by
2174 the host and read/cleared exclusively by the guest. The guest operation of
2175 checking and clearing the flag must an atomic operation so
2176 load-link/store-conditional, or equivalent must be used. There are two cases
2177 where the guest will clear the flag: when the soft lockup watchdog timer resets
2178 itself or when a soft lockup is detected. This ioctl can be called any time
2179 after pausing the vcpu, but before it is resumed.
2184 Capability: KVM_CAP_SIGNAL_MSI
2185 Architectures: x86 arm arm64
2187 Parameters: struct kvm_msi (in)
2188 Returns: >0 on delivery, 0 if guest blocked the MSI, and -1 on error
2190 Directly inject a MSI message. Only valid with in-kernel irqchip that handles
2202 flags: KVM_MSI_VALID_DEVID: devid contains a valid value. The per-VM
2203 KVM_CAP_MSI_DEVID capability advertises the requirement to provide
2204 the device ID. If this capability is not available, userspace
2205 should never set the KVM_MSI_VALID_DEVID flag as the ioctl might fail.
2207 If KVM_MSI_VALID_DEVID is set, devid contains a unique device identifier
2208 for the device that wrote the MSI message. For PCI, this is usually a
2209 BFD identifier in the lower 16 bits.
2211 On x86, address_hi is ignored unless the KVM_X2APIC_API_USE_32BIT_IDS
2212 feature of KVM_CAP_X2APIC_API capability is enabled. If it is enabled,
2213 address_hi bits 31-8 provide bits 31-8 of the destination id. Bits 7-0 of
2214 address_hi must be zero.
2217 4.71 KVM_CREATE_PIT2
2219 Capability: KVM_CAP_PIT2
2222 Parameters: struct kvm_pit_config (in)
2223 Returns: 0 on success, -1 on error
2225 Creates an in-kernel device model for the i8254 PIT. This call is only valid
2226 after enabling in-kernel irqchip support via KVM_CREATE_IRQCHIP. The following
2227 parameters have to be passed:
2229 struct kvm_pit_config {
2236 #define KVM_PIT_SPEAKER_DUMMY 1 /* emulate speaker port stub */
2238 PIT timer interrupts may use a per-VM kernel thread for injection. If it
2239 exists, this thread will have a name of the following pattern:
2241 kvm-pit/<owner-process-pid>
2243 When running a guest with elevated priorities, the scheduling parameters of
2244 this thread may have to be adjusted accordingly.
2246 This IOCTL replaces the obsolete KVM_CREATE_PIT.
2251 Capability: KVM_CAP_PIT_STATE2
2254 Parameters: struct kvm_pit_state2 (out)
2255 Returns: 0 on success, -1 on error
2257 Retrieves the state of the in-kernel PIT model. Only valid after
2258 KVM_CREATE_PIT2. The state is returned in the following structure:
2260 struct kvm_pit_state2 {
2261 struct kvm_pit_channel_state channels[3];
2268 /* disable PIT in HPET legacy mode */
2269 #define KVM_PIT_FLAGS_HPET_LEGACY 0x00000001
2271 This IOCTL replaces the obsolete KVM_GET_PIT.
2276 Capability: KVM_CAP_PIT_STATE2
2279 Parameters: struct kvm_pit_state2 (in)
2280 Returns: 0 on success, -1 on error
2282 Sets the state of the in-kernel PIT model. Only valid after KVM_CREATE_PIT2.
2283 See KVM_GET_PIT2 for details on struct kvm_pit_state2.
2285 This IOCTL replaces the obsolete KVM_SET_PIT.
2288 4.74 KVM_PPC_GET_SMMU_INFO
2290 Capability: KVM_CAP_PPC_GET_SMMU_INFO
2291 Architectures: powerpc
2294 Returns: 0 on success, -1 on error
2296 This populates and returns a structure describing the features of
2297 the "Server" class MMU emulation supported by KVM.
2298 This can in turn be used by userspace to generate the appropriate
2299 device-tree properties for the guest operating system.
2301 The structure contains some global information, followed by an
2302 array of supported segment page sizes:
2304 struct kvm_ppc_smmu_info {
2308 struct kvm_ppc_one_seg_page_size sps[KVM_PPC_PAGE_SIZES_MAX_SZ];
2311 The supported flags are:
2313 - KVM_PPC_PAGE_SIZES_REAL:
2314 When that flag is set, guest page sizes must "fit" the backing
2315 store page sizes. When not set, any page size in the list can
2316 be used regardless of how they are backed by userspace.
2318 - KVM_PPC_1T_SEGMENTS
2319 The emulated MMU supports 1T segments in addition to the
2323 This flag indicates that HPT guests are not supported by KVM,
2324 thus all guests must use radix MMU mode.
2326 The "slb_size" field indicates how many SLB entries are supported
2328 The "sps" array contains 8 entries indicating the supported base
2329 page sizes for a segment in increasing order. Each entry is defined
2332 struct kvm_ppc_one_seg_page_size {
2333 __u32 page_shift; /* Base page shift of segment (or 0) */
2334 __u32 slb_enc; /* SLB encoding for BookS */
2335 struct kvm_ppc_one_page_size enc[KVM_PPC_PAGE_SIZES_MAX_SZ];
2338 An entry with a "page_shift" of 0 is unused. Because the array is
2339 organized in increasing order, a lookup can stop when encoutering
2342 The "slb_enc" field provides the encoding to use in the SLB for the
2343 page size. The bits are in positions such as the value can directly
2344 be OR'ed into the "vsid" argument of the slbmte instruction.
2346 The "enc" array is a list which for each of those segment base page
2347 size provides the list of supported actual page sizes (which can be
2348 only larger or equal to the base page size), along with the
2349 corresponding encoding in the hash PTE. Similarly, the array is
2350 8 entries sorted by increasing sizes and an entry with a "0" shift
2351 is an empty entry and a terminator:
2353 struct kvm_ppc_one_page_size {
2354 __u32 page_shift; /* Page shift (or 0) */
2355 __u32 pte_enc; /* Encoding in the HPTE (>>12) */
2358 The "pte_enc" field provides a value that can OR'ed into the hash
2359 PTE's RPN field (ie, it needs to be shifted left by 12 to OR it
2360 into the hash PTE second double word).
2364 Capability: KVM_CAP_IRQFD
2365 Architectures: x86 s390 arm arm64
2367 Parameters: struct kvm_irqfd (in)
2368 Returns: 0 on success, -1 on error
2370 Allows setting an eventfd to directly trigger a guest interrupt.
2371 kvm_irqfd.fd specifies the file descriptor to use as the eventfd and
2372 kvm_irqfd.gsi specifies the irqchip pin toggled by this event. When
2373 an event is triggered on the eventfd, an interrupt is injected into
2374 the guest using the specified gsi pin. The irqfd is removed using
2375 the KVM_IRQFD_FLAG_DEASSIGN flag, specifying both kvm_irqfd.fd
2378 With KVM_CAP_IRQFD_RESAMPLE, KVM_IRQFD supports a de-assert and notify
2379 mechanism allowing emulation of level-triggered, irqfd-based
2380 interrupts. When KVM_IRQFD_FLAG_RESAMPLE is set the user must pass an
2381 additional eventfd in the kvm_irqfd.resamplefd field. When operating
2382 in resample mode, posting of an interrupt through kvm_irq.fd asserts
2383 the specified gsi in the irqchip. When the irqchip is resampled, such
2384 as from an EOI, the gsi is de-asserted and the user is notified via
2385 kvm_irqfd.resamplefd. It is the user's responsibility to re-queue
2386 the interrupt if the device making use of it still requires service.
2387 Note that closing the resamplefd is not sufficient to disable the
2388 irqfd. The KVM_IRQFD_FLAG_RESAMPLE is only necessary on assignment
2389 and need not be specified with KVM_IRQFD_FLAG_DEASSIGN.
2391 On arm/arm64, gsi routing being supported, the following can happen:
2392 - in case no routing entry is associated to this gsi, injection fails
2393 - in case the gsi is associated to an irqchip routing entry,
2394 irqchip.pin + 32 corresponds to the injected SPI ID.
2395 - in case the gsi is associated to an MSI routing entry, the MSI
2396 message and device ID are translated into an LPI (support restricted
2397 to GICv3 ITS in-kernel emulation).
2399 4.76 KVM_PPC_ALLOCATE_HTAB
2401 Capability: KVM_CAP_PPC_ALLOC_HTAB
2402 Architectures: powerpc
2404 Parameters: Pointer to u32 containing hash table order (in/out)
2405 Returns: 0 on success, -1 on error
2407 This requests the host kernel to allocate an MMU hash table for a
2408 guest using the PAPR paravirtualization interface. This only does
2409 anything if the kernel is configured to use the Book 3S HV style of
2410 virtualization. Otherwise the capability doesn't exist and the ioctl
2411 returns an ENOTTY error. The rest of this description assumes Book 3S
2414 There must be no vcpus running when this ioctl is called; if there
2415 are, it will do nothing and return an EBUSY error.
2417 The parameter is a pointer to a 32-bit unsigned integer variable
2418 containing the order (log base 2) of the desired size of the hash
2419 table, which must be between 18 and 46. On successful return from the
2420 ioctl, the value will not be changed by the kernel.
2422 If no hash table has been allocated when any vcpu is asked to run
2423 (with the KVM_RUN ioctl), the host kernel will allocate a
2424 default-sized hash table (16 MB).
2426 If this ioctl is called when a hash table has already been allocated,
2427 with a different order from the existing hash table, the existing hash
2428 table will be freed and a new one allocated. If this is ioctl is
2429 called when a hash table has already been allocated of the same order
2430 as specified, the kernel will clear out the existing hash table (zero
2431 all HPTEs). In either case, if the guest is using the virtualized
2432 real-mode area (VRMA) facility, the kernel will re-create the VMRA
2433 HPTEs on the next KVM_RUN of any vcpu.
2435 4.77 KVM_S390_INTERRUPT
2439 Type: vm ioctl, vcpu ioctl
2440 Parameters: struct kvm_s390_interrupt (in)
2441 Returns: 0 on success, -1 on error
2443 Allows to inject an interrupt to the guest. Interrupts can be floating
2444 (vm ioctl) or per cpu (vcpu ioctl), depending on the interrupt type.
2446 Interrupt parameters are passed via kvm_s390_interrupt:
2448 struct kvm_s390_interrupt {
2454 type can be one of the following:
2456 KVM_S390_SIGP_STOP (vcpu) - sigp stop; optional flags in parm
2457 KVM_S390_PROGRAM_INT (vcpu) - program check; code in parm
2458 KVM_S390_SIGP_SET_PREFIX (vcpu) - sigp set prefix; prefix address in parm
2459 KVM_S390_RESTART (vcpu) - restart
2460 KVM_S390_INT_CLOCK_COMP (vcpu) - clock comparator interrupt
2461 KVM_S390_INT_CPU_TIMER (vcpu) - CPU timer interrupt
2462 KVM_S390_INT_VIRTIO (vm) - virtio external interrupt; external interrupt
2463 parameters in parm and parm64
2464 KVM_S390_INT_SERVICE (vm) - sclp external interrupt; sclp parameter in parm
2465 KVM_S390_INT_EMERGENCY (vcpu) - sigp emergency; source cpu in parm
2466 KVM_S390_INT_EXTERNAL_CALL (vcpu) - sigp external call; source cpu in parm
2467 KVM_S390_INT_IO(ai,cssid,ssid,schid) (vm) - compound value to indicate an
2468 I/O interrupt (ai - adapter interrupt; cssid,ssid,schid - subchannel);
2469 I/O interruption parameters in parm (subchannel) and parm64 (intparm,
2470 interruption subclass)
2471 KVM_S390_MCHK (vm, vcpu) - machine check interrupt; cr 14 bits in parm,
2472 machine check interrupt code in parm64 (note that
2473 machine checks needing further payload are not
2474 supported by this ioctl)
2476 Note that the vcpu ioctl is asynchronous to vcpu execution.
2478 4.78 KVM_PPC_GET_HTAB_FD
2480 Capability: KVM_CAP_PPC_HTAB_FD
2481 Architectures: powerpc
2483 Parameters: Pointer to struct kvm_get_htab_fd (in)
2484 Returns: file descriptor number (>= 0) on success, -1 on error
2486 This returns a file descriptor that can be used either to read out the
2487 entries in the guest's hashed page table (HPT), or to write entries to
2488 initialize the HPT. The returned fd can only be written to if the
2489 KVM_GET_HTAB_WRITE bit is set in the flags field of the argument, and
2490 can only be read if that bit is clear. The argument struct looks like
2493 /* For KVM_PPC_GET_HTAB_FD */
2494 struct kvm_get_htab_fd {
2500 /* Values for kvm_get_htab_fd.flags */
2501 #define KVM_GET_HTAB_BOLTED_ONLY ((__u64)0x1)
2502 #define KVM_GET_HTAB_WRITE ((__u64)0x2)
2504 The `start_index' field gives the index in the HPT of the entry at
2505 which to start reading. It is ignored when writing.
2507 Reads on the fd will initially supply information about all
2508 "interesting" HPT entries. Interesting entries are those with the
2509 bolted bit set, if the KVM_GET_HTAB_BOLTED_ONLY bit is set, otherwise
2510 all entries. When the end of the HPT is reached, the read() will
2511 return. If read() is called again on the fd, it will start again from
2512 the beginning of the HPT, but will only return HPT entries that have
2513 changed since they were last read.
2515 Data read or written is structured as a header (8 bytes) followed by a
2516 series of valid HPT entries (16 bytes) each. The header indicates how
2517 many valid HPT entries there are and how many invalid entries follow
2518 the valid entries. The invalid entries are not represented explicitly
2519 in the stream. The header format is:
2521 struct kvm_get_htab_header {
2527 Writes to the fd create HPT entries starting at the index given in the
2528 header; first `n_valid' valid entries with contents from the data
2529 written, then `n_invalid' invalid entries, invalidating any previously
2530 valid entries found.
2532 4.79 KVM_CREATE_DEVICE
2534 Capability: KVM_CAP_DEVICE_CTRL
2536 Parameters: struct kvm_create_device (in/out)
2537 Returns: 0 on success, -1 on error
2539 ENODEV: The device type is unknown or unsupported
2540 EEXIST: Device already created, and this type of device may not
2541 be instantiated multiple times
2543 Other error conditions may be defined by individual device types or
2544 have their standard meanings.
2546 Creates an emulated device in the kernel. The file descriptor returned
2547 in fd can be used with KVM_SET/GET/HAS_DEVICE_ATTR.
2549 If the KVM_CREATE_DEVICE_TEST flag is set, only test whether the
2550 device type is supported (not necessarily whether it can be created
2553 Individual devices should not define flags. Attributes should be used
2554 for specifying any behavior that is not implied by the device type
2557 struct kvm_create_device {
2558 __u32 type; /* in: KVM_DEV_TYPE_xxx */
2559 __u32 fd; /* out: device handle */
2560 __u32 flags; /* in: KVM_CREATE_DEVICE_xxx */
2563 4.80 KVM_SET_DEVICE_ATTR/KVM_GET_DEVICE_ATTR
2565 Capability: KVM_CAP_DEVICE_CTRL, KVM_CAP_VM_ATTRIBUTES for vm device,
2566 KVM_CAP_VCPU_ATTRIBUTES for vcpu device
2567 Type: device ioctl, vm ioctl, vcpu ioctl
2568 Parameters: struct kvm_device_attr
2569 Returns: 0 on success, -1 on error
2571 ENXIO: The group or attribute is unknown/unsupported for this device
2572 or hardware support is missing.
2573 EPERM: The attribute cannot (currently) be accessed this way
2574 (e.g. read-only attribute, or attribute that only makes
2575 sense when the device is in a different state)
2577 Other error conditions may be defined by individual device types.
2579 Gets/sets a specified piece of device configuration and/or state. The
2580 semantics are device-specific. See individual device documentation in
2581 the "devices" directory. As with ONE_REG, the size of the data
2582 transferred is defined by the particular attribute.
2584 struct kvm_device_attr {
2585 __u32 flags; /* no flags currently defined */
2586 __u32 group; /* device-defined */
2587 __u64 attr; /* group-defined */
2588 __u64 addr; /* userspace address of attr data */
2591 4.81 KVM_HAS_DEVICE_ATTR
2593 Capability: KVM_CAP_DEVICE_CTRL, KVM_CAP_VM_ATTRIBUTES for vm device,
2594 KVM_CAP_VCPU_ATTRIBUTES for vcpu device
2595 Type: device ioctl, vm ioctl, vcpu ioctl
2596 Parameters: struct kvm_device_attr
2597 Returns: 0 on success, -1 on error
2599 ENXIO: The group or attribute is unknown/unsupported for this device
2600 or hardware support is missing.
2602 Tests whether a device supports a particular attribute. A successful
2603 return indicates the attribute is implemented. It does not necessarily
2604 indicate that the attribute can be read or written in the device's
2605 current state. "addr" is ignored.
2607 4.82 KVM_ARM_VCPU_INIT
2610 Architectures: arm, arm64
2612 Parameters: struct kvm_vcpu_init (in)
2613 Returns: 0 on success; -1 on error
2615 Â EINVAL: Â Â Â the target is unknown, or the combination of features is invalid.
2616 Â ENOENT: Â Â Â a features bit specified is unknown.
2618 This tells KVM what type of CPU to present to the guest, and what
2619 optional features it should have. Â This will cause a reset of the cpu
2620 registers to their initial values. Â If this is not called, KVM_RUN will
2621 return ENOEXEC for that vcpu.
2623 Note that because some registers reflect machine topology, all vcpus
2624 should be created before this ioctl is invoked.
2626 Userspace can call this function multiple times for a given vcpu, including
2627 after the vcpu has been run. This will reset the vcpu to its initial
2628 state. All calls to this function after the initial call must use the same
2629 target and same set of feature flags, otherwise EINVAL will be returned.
2632 - KVM_ARM_VCPU_POWER_OFF: Starts the CPU in a power-off state.
2633 Depends on KVM_CAP_ARM_PSCI. If not set, the CPU will be powered on
2634 and execute guest code when KVM_RUN is called.
2635 - KVM_ARM_VCPU_EL1_32BIT: Starts the CPU in a 32bit mode.
2636 Depends on KVM_CAP_ARM_EL1_32BIT (arm64 only).
2637 - KVM_ARM_VCPU_PSCI_0_2: Emulate PSCI v0.2 (or a future revision
2638 backward compatible with v0.2) for the CPU.
2639 Depends on KVM_CAP_ARM_PSCI_0_2.
2640 - KVM_ARM_VCPU_PMU_V3: Emulate PMUv3 for the CPU.
2641 Depends on KVM_CAP_ARM_PMU_V3.
2644 4.83 KVM_ARM_PREFERRED_TARGET
2647 Architectures: arm, arm64
2649 Parameters: struct struct kvm_vcpu_init (out)
2650 Returns: 0 on success; -1 on error
2652 ENODEV: no preferred target available for the host
2654 This queries KVM for preferred CPU target type which can be emulated
2655 by KVM on underlying host.
2657 The ioctl returns struct kvm_vcpu_init instance containing information
2658 about preferred CPU target type and recommended features for it. The
2659 kvm_vcpu_init->features bitmap returned will have feature bits set if
2660 the preferred target recommends setting these features, but this is
2663 The information returned by this ioctl can be used to prepare an instance
2664 of struct kvm_vcpu_init for KVM_ARM_VCPU_INIT ioctl which will result in
2665 in VCPU matching underlying host.
2668 4.84 KVM_GET_REG_LIST
2671 Architectures: arm, arm64, mips
2673 Parameters: struct kvm_reg_list (in/out)
2674 Returns: 0 on success; -1 on error
2676 Â E2BIG: Â Â Â Â the reg index list is too big to fit in the array specified by
2677 Â Â Â Â Â Â Â Â Â Â Â Â the user (the number required will be written into n).
2679 struct kvm_reg_list {
2680 __u64 n; /* number of registers in reg[] */
2684 This ioctl returns the guest registers that are supported for the
2685 KVM_GET_ONE_REG/KVM_SET_ONE_REG calls.
2688 4.85 KVM_ARM_SET_DEVICE_ADDR (deprecated)
2690 Capability: KVM_CAP_ARM_SET_DEVICE_ADDR
2691 Architectures: arm, arm64
2693 Parameters: struct kvm_arm_device_address (in)
2694 Returns: 0 on success, -1 on error
2696 ENODEV: The device id is unknown
2697 ENXIO: Device not supported on current system
2698 EEXIST: Address already set
2699 E2BIG: Address outside guest physical address space
2700 EBUSY: Address overlaps with other device range
2702 struct kvm_arm_device_addr {
2707 Specify a device address in the guest's physical address space where guests
2708 can access emulated or directly exposed devices, which the host kernel needs
2709 to know about. The id field is an architecture specific identifier for a
2712 ARM/arm64 divides the id field into two parts, a device id and an
2713 address type id specific to the individual device.
2715 Â bits: | 63 ... 32 | 31 ... 16 | 15 ... 0 |
2716 field: | 0x00000000 | device id | addr type id |
2718 ARM/arm64 currently only require this when using the in-kernel GIC
2719 support for the hardware VGIC features, using KVM_ARM_DEVICE_VGIC_V2
2720 as the device id. When setting the base address for the guest's
2721 mapping of the VGIC virtual CPU and distributor interface, the ioctl
2722 must be called after calling KVM_CREATE_IRQCHIP, but before calling
2723 KVM_RUN on any of the VCPUs. Calling this ioctl twice for any of the
2724 base addresses will return -EEXIST.
2726 Note, this IOCTL is deprecated and the more flexible SET/GET_DEVICE_ATTR API
2727 should be used instead.
2730 4.86 KVM_PPC_RTAS_DEFINE_TOKEN
2732 Capability: KVM_CAP_PPC_RTAS
2735 Parameters: struct kvm_rtas_token_args
2736 Returns: 0 on success, -1 on error
2738 Defines a token value for a RTAS (Run Time Abstraction Services)
2739 service in order to allow it to be handled in the kernel. The
2740 argument struct gives the name of the service, which must be the name
2741 of a service that has a kernel-side implementation. If the token
2742 value is non-zero, it will be associated with that service, and
2743 subsequent RTAS calls by the guest specifying that token will be
2744 handled by the kernel. If the token value is 0, then any token
2745 associated with the service will be forgotten, and subsequent RTAS
2746 calls by the guest for that service will be passed to userspace to be
2749 4.87 KVM_SET_GUEST_DEBUG
2751 Capability: KVM_CAP_SET_GUEST_DEBUG
2752 Architectures: x86, s390, ppc, arm64
2754 Parameters: struct kvm_guest_debug (in)
2755 Returns: 0 on success; -1 on error
2757 struct kvm_guest_debug {
2760 struct kvm_guest_debug_arch arch;
2763 Set up the processor specific debug registers and configure vcpu for
2764 handling guest debug events. There are two parts to the structure, the
2765 first a control bitfield indicates the type of debug events to handle
2766 when running. Common control bits are:
2768 - KVM_GUESTDBG_ENABLE: guest debugging is enabled
2769 - KVM_GUESTDBG_SINGLESTEP: the next run should single-step
2771 The top 16 bits of the control field are architecture specific control
2772 flags which can include the following:
2774 - KVM_GUESTDBG_USE_SW_BP: using software breakpoints [x86, arm64]
2775 - KVM_GUESTDBG_USE_HW_BP: using hardware breakpoints [x86, s390, arm64]
2776 - KVM_GUESTDBG_INJECT_DB: inject DB type exception [x86]
2777 - KVM_GUESTDBG_INJECT_BP: inject BP type exception [x86]
2778 - KVM_GUESTDBG_EXIT_PENDING: trigger an immediate guest exit [s390]
2780 For example KVM_GUESTDBG_USE_SW_BP indicates that software breakpoints
2781 are enabled in memory so we need to ensure breakpoint exceptions are
2782 correctly trapped and the KVM run loop exits at the breakpoint and not
2783 running off into the normal guest vector. For KVM_GUESTDBG_USE_HW_BP
2784 we need to ensure the guest vCPUs architecture specific registers are
2785 updated to the correct (supplied) values.
2787 The second part of the structure is architecture specific and
2788 typically contains a set of debug registers.
2790 For arm64 the number of debug registers is implementation defined and
2791 can be determined by querying the KVM_CAP_GUEST_DEBUG_HW_BPS and
2792 KVM_CAP_GUEST_DEBUG_HW_WPS capabilities which return a positive number
2793 indicating the number of supported registers.
2795 When debug events exit the main run loop with the reason
2796 KVM_EXIT_DEBUG with the kvm_debug_exit_arch part of the kvm_run
2797 structure containing architecture specific debug information.
2799 4.88 KVM_GET_EMULATED_CPUID
2801 Capability: KVM_CAP_EXT_EMUL_CPUID
2804 Parameters: struct kvm_cpuid2 (in/out)
2805 Returns: 0 on success, -1 on error
2810 struct kvm_cpuid_entry2 entries[0];
2813 The member 'flags' is used for passing flags from userspace.
2815 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX BIT(0)
2816 #define KVM_CPUID_FLAG_STATEFUL_FUNC BIT(1)
2817 #define KVM_CPUID_FLAG_STATE_READ_NEXT BIT(2)
2819 struct kvm_cpuid_entry2 {
2830 This ioctl returns x86 cpuid features which are emulated by
2831 kvm.Userspace can use the information returned by this ioctl to query
2832 which features are emulated by kvm instead of being present natively.
2834 Userspace invokes KVM_GET_EMULATED_CPUID by passing a kvm_cpuid2
2835 structure with the 'nent' field indicating the number of entries in
2836 the variable-size array 'entries'. If the number of entries is too low
2837 to describe the cpu capabilities, an error (E2BIG) is returned. If the
2838 number is too high, the 'nent' field is adjusted and an error (ENOMEM)
2839 is returned. If the number is just right, the 'nent' field is adjusted
2840 to the number of valid entries in the 'entries' array, which is then
2843 The entries returned are the set CPUID bits of the respective features
2844 which kvm emulates, as returned by the CPUID instruction, with unknown
2845 or unsupported feature bits cleared.
2847 Features like x2apic, for example, may not be present in the host cpu
2848 but are exposed by kvm in KVM_GET_SUPPORTED_CPUID because they can be
2849 emulated efficiently and thus not included here.
2851 The fields in each entry are defined as follows:
2853 function: the eax value used to obtain the entry
2854 index: the ecx value used to obtain the entry (for entries that are
2856 flags: an OR of zero or more of the following:
2857 KVM_CPUID_FLAG_SIGNIFCANT_INDEX:
2858 if the index field is valid
2859 KVM_CPUID_FLAG_STATEFUL_FUNC:
2860 if cpuid for this function returns different values for successive
2861 invocations; there will be several entries with the same function,
2862 all with this flag set
2863 KVM_CPUID_FLAG_STATE_READ_NEXT:
2864 for KVM_CPUID_FLAG_STATEFUL_FUNC entries, set if this entry is
2865 the first entry to be read by a cpu
2866 eax, ebx, ecx, edx: the values returned by the cpuid instruction for
2867 this function/index combination
2869 4.89 KVM_S390_MEM_OP
2871 Capability: KVM_CAP_S390_MEM_OP
2874 Parameters: struct kvm_s390_mem_op (in)
2875 Returns: = 0 on success,
2876 < 0 on generic error (e.g. -EFAULT or -ENOMEM),
2877 > 0 if an exception occurred while walking the page tables
2879 Read or write data from/to the logical (virtual) memory of a VCPU.
2881 Parameters are specified via the following structure:
2883 struct kvm_s390_mem_op {
2884 __u64 gaddr; /* the guest address */
2885 __u64 flags; /* flags */
2886 __u32 size; /* amount of bytes */
2887 __u32 op; /* type of operation */
2888 __u64 buf; /* buffer in userspace */
2889 __u8 ar; /* the access register number */
2890 __u8 reserved[31]; /* should be set to 0 */
2893 The type of operation is specified in the "op" field. It is either
2894 KVM_S390_MEMOP_LOGICAL_READ for reading from logical memory space or
2895 KVM_S390_MEMOP_LOGICAL_WRITE for writing to logical memory space. The
2896 KVM_S390_MEMOP_F_CHECK_ONLY flag can be set in the "flags" field to check
2897 whether the corresponding memory access would create an access exception
2898 (without touching the data in the memory at the destination). In case an
2899 access exception occurred while walking the MMU tables of the guest, the
2900 ioctl returns a positive error number to indicate the type of exception.
2901 This exception is also raised directly at the corresponding VCPU if the
2902 flag KVM_S390_MEMOP_F_INJECT_EXCEPTION is set in the "flags" field.
2904 The start address of the memory region has to be specified in the "gaddr"
2905 field, and the length of the region in the "size" field. "buf" is the buffer
2906 supplied by the userspace application where the read data should be written
2907 to for KVM_S390_MEMOP_LOGICAL_READ, or where the data that should be written
2908 is stored for a KVM_S390_MEMOP_LOGICAL_WRITE. "buf" is unused and can be NULL
2909 when KVM_S390_MEMOP_F_CHECK_ONLY is specified. "ar" designates the access
2910 register number to be used.
2912 The "reserved" field is meant for future extensions. It is not used by
2913 KVM with the currently defined set of flags.
2915 4.90 KVM_S390_GET_SKEYS
2917 Capability: KVM_CAP_S390_SKEYS
2920 Parameters: struct kvm_s390_skeys
2921 Returns: 0 on success, KVM_S390_GET_KEYS_NONE if guest is not using storage
2922 keys, negative value on error
2924 This ioctl is used to get guest storage key values on the s390
2925 architecture. The ioctl takes parameters via the kvm_s390_skeys struct.
2927 struct kvm_s390_skeys {
2930 __u64 skeydata_addr;
2935 The start_gfn field is the number of the first guest frame whose storage keys
2938 The count field is the number of consecutive frames (starting from start_gfn)
2939 whose storage keys to get. The count field must be at least 1 and the maximum
2940 allowed value is defined as KVM_S390_SKEYS_ALLOC_MAX. Values outside this range
2941 will cause the ioctl to return -EINVAL.
2943 The skeydata_addr field is the address to a buffer large enough to hold count
2944 bytes. This buffer will be filled with storage key data by the ioctl.
2946 4.91 KVM_S390_SET_SKEYS
2948 Capability: KVM_CAP_S390_SKEYS
2951 Parameters: struct kvm_s390_skeys
2952 Returns: 0 on success, negative value on error
2954 This ioctl is used to set guest storage key values on the s390
2955 architecture. The ioctl takes parameters via the kvm_s390_skeys struct.
2956 See section on KVM_S390_GET_SKEYS for struct definition.
2958 The start_gfn field is the number of the first guest frame whose storage keys
2961 The count field is the number of consecutive frames (starting from start_gfn)
2962 whose storage keys to get. The count field must be at least 1 and the maximum
2963 allowed value is defined as KVM_S390_SKEYS_ALLOC_MAX. Values outside this range
2964 will cause the ioctl to return -EINVAL.
2966 The skeydata_addr field is the address to a buffer containing count bytes of
2967 storage keys. Each byte in the buffer will be set as the storage key for a
2968 single frame starting at start_gfn for count frames.
2970 Note: If any architecturally invalid key value is found in the given data then
2971 the ioctl will return -EINVAL.
2975 Capability: KVM_CAP_S390_INJECT_IRQ
2978 Parameters: struct kvm_s390_irq (in)
2979 Returns: 0 on success, -1 on error
2981 EINVAL: interrupt type is invalid
2982 type is KVM_S390_SIGP_STOP and flag parameter is invalid value
2983 type is KVM_S390_INT_EXTERNAL_CALL and code is bigger
2984 than the maximum of VCPUs
2985 EBUSY: type is KVM_S390_SIGP_SET_PREFIX and vcpu is not stopped
2986 type is KVM_S390_SIGP_STOP and a stop irq is already pending
2987 type is KVM_S390_INT_EXTERNAL_CALL and an external call interrupt
2990 Allows to inject an interrupt to the guest.
2992 Using struct kvm_s390_irq as a parameter allows
2993 to inject additional payload which is not
2994 possible via KVM_S390_INTERRUPT.
2996 Interrupt parameters are passed via kvm_s390_irq:
2998 struct kvm_s390_irq {
3001 struct kvm_s390_io_info io;
3002 struct kvm_s390_ext_info ext;
3003 struct kvm_s390_pgm_info pgm;
3004 struct kvm_s390_emerg_info emerg;
3005 struct kvm_s390_extcall_info extcall;
3006 struct kvm_s390_prefix_info prefix;
3007 struct kvm_s390_stop_info stop;
3008 struct kvm_s390_mchk_info mchk;
3013 type can be one of the following:
3015 KVM_S390_SIGP_STOP - sigp stop; parameter in .stop
3016 KVM_S390_PROGRAM_INT - program check; parameters in .pgm
3017 KVM_S390_SIGP_SET_PREFIX - sigp set prefix; parameters in .prefix
3018 KVM_S390_RESTART - restart; no parameters
3019 KVM_S390_INT_CLOCK_COMP - clock comparator interrupt; no parameters
3020 KVM_S390_INT_CPU_TIMER - CPU timer interrupt; no parameters
3021 KVM_S390_INT_EMERGENCY - sigp emergency; parameters in .emerg
3022 KVM_S390_INT_EXTERNAL_CALL - sigp external call; parameters in .extcall
3023 KVM_S390_MCHK - machine check interrupt; parameters in .mchk
3026 Note that the vcpu ioctl is asynchronous to vcpu execution.
3028 4.94 KVM_S390_GET_IRQ_STATE
3030 Capability: KVM_CAP_S390_IRQ_STATE
3033 Parameters: struct kvm_s390_irq_state (out)
3034 Returns: >= number of bytes copied into buffer,
3035 -EINVAL if buffer size is 0,
3036 -ENOBUFS if buffer size is too small to fit all pending interrupts,
3037 -EFAULT if the buffer address was invalid
3039 This ioctl allows userspace to retrieve the complete state of all currently
3040 pending interrupts in a single buffer. Use cases include migration
3041 and introspection. The parameter structure contains the address of a
3042 userspace buffer and its length:
3044 struct kvm_s390_irq_state {
3046 __u32 flags; /* will stay unused for compatibility reasons */
3048 __u32 reserved[4]; /* will stay unused for compatibility reasons */
3051 Userspace passes in the above struct and for each pending interrupt a
3052 struct kvm_s390_irq is copied to the provided buffer.
3054 The structure contains a flags and a reserved field for future extensions. As
3055 the kernel never checked for flags == 0 and QEMU never pre-zeroed flags and
3056 reserved, these fields can not be used in the future without breaking
3059 If -ENOBUFS is returned the buffer provided was too small and userspace
3060 may retry with a bigger buffer.
3062 4.95 KVM_S390_SET_IRQ_STATE
3064 Capability: KVM_CAP_S390_IRQ_STATE
3067 Parameters: struct kvm_s390_irq_state (in)
3068 Returns: 0 on success,
3069 -EFAULT if the buffer address was invalid,
3070 -EINVAL for an invalid buffer length (see below),
3071 -EBUSY if there were already interrupts pending,
3072 errors occurring when actually injecting the
3073 interrupt. See KVM_S390_IRQ.
3075 This ioctl allows userspace to set the complete state of all cpu-local
3076 interrupts currently pending for the vcpu. It is intended for restoring
3077 interrupt state after a migration. The input parameter is a userspace buffer
3078 containing a struct kvm_s390_irq_state:
3080 struct kvm_s390_irq_state {
3082 __u32 flags; /* will stay unused for compatibility reasons */
3084 __u32 reserved[4]; /* will stay unused for compatibility reasons */
3087 The restrictions for flags and reserved apply as well.
3088 (see KVM_S390_GET_IRQ_STATE)
3090 The userspace memory referenced by buf contains a struct kvm_s390_irq
3091 for each interrupt to be injected into the guest.
3092 If one of the interrupts could not be injected for some reason the
3095 len must be a multiple of sizeof(struct kvm_s390_irq). It must be > 0
3096 and it must not exceed (max_vcpus + 32) * sizeof(struct kvm_s390_irq),
3097 which is the maximum number of possibly pending cpu-local interrupts.
3101 Capability: KVM_CAP_X86_SMM
3105 Returns: 0 on success, -1 on error
3107 Queues an SMI on the thread's vcpu.
3109 4.97 KVM_CAP_PPC_MULTITCE
3111 Capability: KVM_CAP_PPC_MULTITCE
3115 This capability means the kernel is capable of handling hypercalls
3116 H_PUT_TCE_INDIRECT and H_STUFF_TCE without passing those into the user
3117 space. This significantly accelerates DMA operations for PPC KVM guests.
3118 User space should expect that its handlers for these hypercalls
3119 are not going to be called if user space previously registered LIOBN
3120 in KVM (via KVM_CREATE_SPAPR_TCE or similar calls).
3122 In order to enable H_PUT_TCE_INDIRECT and H_STUFF_TCE use in the guest,
3123 user space might have to advertise it for the guest. For example,
3124 IBM pSeries (sPAPR) guest starts using them if "hcall-multi-tce" is
3125 present in the "ibm,hypertas-functions" device-tree property.
3127 The hypercalls mentioned above may or may not be processed successfully
3128 in the kernel based fast path. If they can not be handled by the kernel,
3129 they will get passed on to user space. So user space still has to have
3130 an implementation for these despite the in kernel acceleration.
3132 This capability is always enabled.
3134 4.98 KVM_CREATE_SPAPR_TCE_64
3136 Capability: KVM_CAP_SPAPR_TCE_64
3137 Architectures: powerpc
3139 Parameters: struct kvm_create_spapr_tce_64 (in)
3140 Returns: file descriptor for manipulating the created TCE table
3142 This is an extension for KVM_CAP_SPAPR_TCE which only supports 32bit
3143 windows, described in 4.62 KVM_CREATE_SPAPR_TCE
3145 This capability uses extended struct in ioctl interface:
3147 /* for KVM_CAP_SPAPR_TCE_64 */
3148 struct kvm_create_spapr_tce_64 {
3152 __u64 offset; /* in pages */
3153 __u64 size; /* in pages */
3156 The aim of extension is to support an additional bigger DMA window with
3157 a variable page size.
3158 KVM_CREATE_SPAPR_TCE_64 receives a 64bit window size, an IOMMU page shift and
3159 a bus offset of the corresponding DMA window, @size and @offset are numbers
3162 @flags are not used at the moment.
3164 The rest of functionality is identical to KVM_CREATE_SPAPR_TCE.
3166 4.99 KVM_REINJECT_CONTROL
3168 Capability: KVM_CAP_REINJECT_CONTROL
3171 Parameters: struct kvm_reinject_control (in)
3172 Returns: 0 on success,
3173 -EFAULT if struct kvm_reinject_control cannot be read,
3174 -ENXIO if KVM_CREATE_PIT or KVM_CREATE_PIT2 didn't succeed earlier.
3176 i8254 (PIT) has two modes, reinject and !reinject. The default is reinject,
3177 where KVM queues elapsed i8254 ticks and monitors completion of interrupt from
3178 vector(s) that i8254 injects. Reinject mode dequeues a tick and injects its
3179 interrupt whenever there isn't a pending interrupt from i8254.
3180 !reinject mode injects an interrupt as soon as a tick arrives.
3182 struct kvm_reinject_control {
3187 pit_reinject = 0 (!reinject mode) is recommended, unless running an old
3188 operating system that uses the PIT for timing (e.g. Linux 2.4.x).
3190 4.100 KVM_PPC_CONFIGURE_V3_MMU
3192 Capability: KVM_CAP_PPC_RADIX_MMU or KVM_CAP_PPC_HASH_MMU_V3
3195 Parameters: struct kvm_ppc_mmuv3_cfg (in)
3196 Returns: 0 on success,
3197 -EFAULT if struct kvm_ppc_mmuv3_cfg cannot be read,
3198 -EINVAL if the configuration is invalid
3200 This ioctl controls whether the guest will use radix or HPT (hashed
3201 page table) translation, and sets the pointer to the process table for
3204 struct kvm_ppc_mmuv3_cfg {
3206 __u64 process_table;
3209 There are two bits that can be set in flags; KVM_PPC_MMUV3_RADIX and
3210 KVM_PPC_MMUV3_GTSE. KVM_PPC_MMUV3_RADIX, if set, configures the guest
3211 to use radix tree translation, and if clear, to use HPT translation.
3212 KVM_PPC_MMUV3_GTSE, if set and if KVM permits it, configures the guest
3213 to be able to use the global TLB and SLB invalidation instructions;
3214 if clear, the guest may not use these instructions.
3216 The process_table field specifies the address and size of the guest
3217 process table, which is in the guest's space. This field is formatted
3218 as the second doubleword of the partition table entry, as defined in
3219 the Power ISA V3.00, Book III section 5.7.6.1.
3221 4.101 KVM_PPC_GET_RMMU_INFO
3223 Capability: KVM_CAP_PPC_RADIX_MMU
3226 Parameters: struct kvm_ppc_rmmu_info (out)
3227 Returns: 0 on success,
3228 -EFAULT if struct kvm_ppc_rmmu_info cannot be written,
3229 -EINVAL if no useful information can be returned
3231 This ioctl returns a structure containing two things: (a) a list
3232 containing supported radix tree geometries, and (b) a list that maps
3233 page sizes to put in the "AP" (actual page size) field for the tlbie
3234 (TLB invalidate entry) instruction.
3236 struct kvm_ppc_rmmu_info {
3237 struct kvm_ppc_radix_geom {
3242 __u32 ap_encodings[8];
3245 The geometries[] field gives up to 8 supported geometries for the
3246 radix page table, in terms of the log base 2 of the smallest page
3247 size, and the number of bits indexed at each level of the tree, from
3248 the PTE level up to the PGD level in that order. Any unused entries
3249 will have 0 in the page_shift field.
3251 The ap_encodings gives the supported page sizes and their AP field
3252 encodings, encoded with the AP value in the top 3 bits and the log
3253 base 2 of the page size in the bottom 6 bits.
3255 4.102 KVM_PPC_RESIZE_HPT_PREPARE
3257 Capability: KVM_CAP_SPAPR_RESIZE_HPT
3258 Architectures: powerpc
3260 Parameters: struct kvm_ppc_resize_hpt (in)
3261 Returns: 0 on successful completion,
3262 >0 if a new HPT is being prepared, the value is an estimated
3263 number of milliseconds until preparation is complete
3264 -EFAULT if struct kvm_reinject_control cannot be read,
3265 -EINVAL if the supplied shift or flags are invalid
3266 -ENOMEM if unable to allocate the new HPT
3267 -ENOSPC if there was a hash collision when moving existing
3268 HPT entries to the new HPT
3269 -EIO on other error conditions
3271 Used to implement the PAPR extension for runtime resizing of a guest's
3272 Hashed Page Table (HPT). Specifically this starts, stops or monitors
3273 the preparation of a new potential HPT for the guest, essentially
3274 implementing the H_RESIZE_HPT_PREPARE hypercall.
3276 If called with shift > 0 when there is no pending HPT for the guest,
3277 this begins preparation of a new pending HPT of size 2^(shift) bytes.
3278 It then returns a positive integer with the estimated number of
3279 milliseconds until preparation is complete.
3281 If called when there is a pending HPT whose size does not match that
3282 requested in the parameters, discards the existing pending HPT and
3283 creates a new one as above.
3285 If called when there is a pending HPT of the size requested, will:
3286 * If preparation of the pending HPT is already complete, return 0
3287 * If preparation of the pending HPT has failed, return an error
3288 code, then discard the pending HPT.
3289 * If preparation of the pending HPT is still in progress, return an
3290 estimated number of milliseconds until preparation is complete.
3292 If called with shift == 0, discards any currently pending HPT and
3293 returns 0 (i.e. cancels any in-progress preparation).
3295 flags is reserved for future expansion, currently setting any bits in
3296 flags will result in an -EINVAL.
3298 Normally this will be called repeatedly with the same parameters until
3299 it returns <= 0. The first call will initiate preparation, subsequent
3300 ones will monitor preparation until it completes or fails.
3302 struct kvm_ppc_resize_hpt {
3308 4.103 KVM_PPC_RESIZE_HPT_COMMIT
3310 Capability: KVM_CAP_SPAPR_RESIZE_HPT
3311 Architectures: powerpc
3313 Parameters: struct kvm_ppc_resize_hpt (in)
3314 Returns: 0 on successful completion,
3315 -EFAULT if struct kvm_reinject_control cannot be read,
3316 -EINVAL if the supplied shift or flags are invalid
3317 -ENXIO is there is no pending HPT, or the pending HPT doesn't
3318 have the requested size
3319 -EBUSY if the pending HPT is not fully prepared
3320 -ENOSPC if there was a hash collision when moving existing
3321 HPT entries to the new HPT
3322 -EIO on other error conditions
3324 Used to implement the PAPR extension for runtime resizing of a guest's
3325 Hashed Page Table (HPT). Specifically this requests that the guest be
3326 transferred to working with the new HPT, essentially implementing the
3327 H_RESIZE_HPT_COMMIT hypercall.
3329 This should only be called after KVM_PPC_RESIZE_HPT_PREPARE has
3330 returned 0 with the same parameters. In other cases
3331 KVM_PPC_RESIZE_HPT_COMMIT will return an error (usually -ENXIO or
3332 -EBUSY, though others may be possible if the preparation was started,
3335 This will have undefined effects on the guest if it has not already
3336 placed itself in a quiescent state where no vcpu will make MMU enabled
3339 On succsful completion, the pending HPT will become the guest's active
3340 HPT and the previous HPT will be discarded.
3342 On failure, the guest will still be operating on its previous HPT.
3344 struct kvm_ppc_resize_hpt {
3350 4.104 KVM_X86_GET_MCE_CAP_SUPPORTED
3352 Capability: KVM_CAP_MCE
3355 Parameters: u64 mce_cap (out)
3356 Returns: 0 on success, -1 on error
3358 Returns supported MCE capabilities. The u64 mce_cap parameter
3359 has the same format as the MSR_IA32_MCG_CAP register. Supported
3360 capabilities will have the corresponding bits set.
3362 4.105 KVM_X86_SETUP_MCE
3364 Capability: KVM_CAP_MCE
3367 Parameters: u64 mcg_cap (in)
3368 Returns: 0 on success,
3369 -EFAULT if u64 mcg_cap cannot be read,
3370 -EINVAL if the requested number of banks is invalid,
3371 -EINVAL if requested MCE capability is not supported.
3373 Initializes MCE support for use. The u64 mcg_cap parameter
3374 has the same format as the MSR_IA32_MCG_CAP register and
3375 specifies which capabilities should be enabled. The maximum
3376 supported number of error-reporting banks can be retrieved when
3377 checking for KVM_CAP_MCE. The supported capabilities can be
3378 retrieved with KVM_X86_GET_MCE_CAP_SUPPORTED.
3380 4.106 KVM_X86_SET_MCE
3382 Capability: KVM_CAP_MCE
3385 Parameters: struct kvm_x86_mce (in)
3386 Returns: 0 on success,
3387 -EFAULT if struct kvm_x86_mce cannot be read,
3388 -EINVAL if the bank number is invalid,
3389 -EINVAL if VAL bit is not set in status field.
3391 Inject a machine check error (MCE) into the guest. The input
3394 struct kvm_x86_mce {
3404 If the MCE being reported is an uncorrected error, KVM will
3405 inject it as an MCE exception into the guest. If the guest
3406 MCG_STATUS register reports that an MCE is in progress, KVM
3407 causes an KVM_EXIT_SHUTDOWN vmexit.
3409 Otherwise, if the MCE is a corrected error, KVM will just
3410 store it in the corresponding bank (provided this bank is
3411 not holding a previously reported uncorrected error).
3413 4.107 KVM_S390_GET_CMMA_BITS
3415 Capability: KVM_CAP_S390_CMMA_MIGRATION
3418 Parameters: struct kvm_s390_cmma_log (in, out)
3419 Returns: 0 on success, a negative value on error
3421 This ioctl is used to get the values of the CMMA bits on the s390
3422 architecture. It is meant to be used in two scenarios:
3423 - During live migration to save the CMMA values. Live migration needs
3424 to be enabled via the KVM_REQ_START_MIGRATION VM property.
3425 - To non-destructively peek at the CMMA values, with the flag
3426 KVM_S390_CMMA_PEEK set.
3428 The ioctl takes parameters via the kvm_s390_cmma_log struct. The desired
3429 values are written to a buffer whose location is indicated via the "values"
3430 member in the kvm_s390_cmma_log struct. The values in the input struct are
3431 also updated as needed.
3432 Each CMMA value takes up one byte.
3434 struct kvm_s390_cmma_log {
3445 start_gfn is the number of the first guest frame whose CMMA values are
3448 count is the length of the buffer in bytes,
3450 values points to the buffer where the result will be written to.
3452 If count is greater than KVM_S390_SKEYS_MAX, then it is considered to be
3453 KVM_S390_SKEYS_MAX. KVM_S390_SKEYS_MAX is re-used for consistency with
3456 The result is written in the buffer pointed to by the field values, and
3457 the values of the input parameter are updated as follows.
3459 Depending on the flags, different actions are performed. The only
3460 supported flag so far is KVM_S390_CMMA_PEEK.
3462 The default behaviour if KVM_S390_CMMA_PEEK is not set is:
3463 start_gfn will indicate the first page frame whose CMMA bits were dirty.
3464 It is not necessarily the same as the one passed as input, as clean pages
3467 count will indicate the number of bytes actually written in the buffer.
3468 It can (and very often will) be smaller than the input value, since the
3469 buffer is only filled until 16 bytes of clean values are found (which
3470 are then not copied in the buffer). Since a CMMA migration block needs
3471 the base address and the length, for a total of 16 bytes, we will send
3472 back some clean data if there is some dirty data afterwards, as long as
3473 the size of the clean data does not exceed the size of the header. This
3474 allows to minimize the amount of data to be saved or transferred over
3475 the network at the expense of more roundtrips to userspace. The next
3476 invocation of the ioctl will skip over all the clean values, saving
3477 potentially more than just the 16 bytes we found.
3479 If KVM_S390_CMMA_PEEK is set:
3480 the existing storage attributes are read even when not in migration
3481 mode, and no other action is performed;
3483 the output start_gfn will be equal to the input start_gfn,
3485 the output count will be equal to the input count, except if the end of
3486 memory has been reached.
3489 the field "remaining" will indicate the total number of dirty CMMA values
3490 still remaining, or 0 if KVM_S390_CMMA_PEEK is set and migration mode is
3495 values points to the userspace buffer where the result will be stored.
3497 This ioctl can fail with -ENOMEM if not enough memory can be allocated to
3498 complete the task, with -ENXIO if CMMA is not enabled, with -EINVAL if
3499 KVM_S390_CMMA_PEEK is not set but migration mode was not enabled, with
3500 -EFAULT if the userspace address is invalid or if no page table is
3501 present for the addresses (e.g. when using hugepages).
3503 4.108 KVM_S390_SET_CMMA_BITS
3505 Capability: KVM_CAP_S390_CMMA_MIGRATION
3508 Parameters: struct kvm_s390_cmma_log (in)
3509 Returns: 0 on success, a negative value on error
3511 This ioctl is used to set the values of the CMMA bits on the s390
3512 architecture. It is meant to be used during live migration to restore
3513 the CMMA values, but there are no restrictions on its use.
3514 The ioctl takes parameters via the kvm_s390_cmma_values struct.
3515 Each CMMA value takes up one byte.
3517 struct kvm_s390_cmma_log {
3528 start_gfn indicates the starting guest frame number,
3530 count indicates how many values are to be considered in the buffer,
3532 flags is not used and must be 0.
3534 mask indicates which PGSTE bits are to be considered.
3536 remaining is not used.
3538 values points to the buffer in userspace where to store the values.
3540 This ioctl can fail with -ENOMEM if not enough memory can be allocated to
3541 complete the task, with -ENXIO if CMMA is not enabled, with -EINVAL if
3542 the count field is too large (e.g. more than KVM_S390_CMMA_SIZE_MAX) or
3543 if the flags field was not 0, with -EFAULT if the userspace address is
3544 invalid, if invalid pages are written to (e.g. after the end of memory)
3545 or if no page table is present for the addresses (e.g. when using
3548 4.109 KVM_PPC_GET_CPU_CHAR
3550 Capability: KVM_CAP_PPC_GET_CPU_CHAR
3551 Architectures: powerpc
3553 Parameters: struct kvm_ppc_cpu_char (out)
3554 Returns: 0 on successful completion
3555 -EFAULT if struct kvm_ppc_cpu_char cannot be written
3557 This ioctl gives userspace information about certain characteristics
3558 of the CPU relating to speculative execution of instructions and
3559 possible information leakage resulting from speculative execution (see
3560 CVE-2017-5715, CVE-2017-5753 and CVE-2017-5754). The information is
3561 returned in struct kvm_ppc_cpu_char, which looks like this:
3563 struct kvm_ppc_cpu_char {
3564 __u64 character; /* characteristics of the CPU */
3565 __u64 behaviour; /* recommended software behaviour */
3566 __u64 character_mask; /* valid bits in character */
3567 __u64 behaviour_mask; /* valid bits in behaviour */
3570 For extensibility, the character_mask and behaviour_mask fields
3571 indicate which bits of character and behaviour have been filled in by
3572 the kernel. If the set of defined bits is extended in future then
3573 userspace will be able to tell whether it is running on a kernel that
3574 knows about the new bits.
3576 The character field describes attributes of the CPU which can help
3577 with preventing inadvertent information disclosure - specifically,
3578 whether there is an instruction to flash-invalidate the L1 data cache
3579 (ori 30,30,0 or mtspr SPRN_TRIG2,rN), whether the L1 data cache is set
3580 to a mode where entries can only be used by the thread that created
3581 them, whether the bcctr[l] instruction prevents speculation, and
3582 whether a speculation barrier instruction (ori 31,31,0) is provided.
3584 The behaviour field describes actions that software should take to
3585 prevent inadvertent information disclosure, and thus describes which
3586 vulnerabilities the hardware is subject to; specifically whether the
3587 L1 data cache should be flushed when returning to user mode from the
3588 kernel, and whether a speculation barrier should be placed between an
3589 array bounds check and the array access.
3591 These fields use the same bit definitions as the new
3592 H_GET_CPU_CHARACTERISTICS hypercall.
3594 4.110 KVM_MEMORY_ENCRYPT_OP
3599 Parameters: an opaque platform specific structure (in/out)
3600 Returns: 0 on success; -1 on error
3602 If the platform supports creating encrypted VMs then this ioctl can be used
3603 for issuing platform-specific memory encryption commands to manage those
3606 Currently, this ioctl is used for issuing Secure Encrypted Virtualization
3607 (SEV) commands on AMD Processors. The SEV commands are defined in
3608 Documentation/virtual/kvm/amd-memory-encryption.rst.
3610 4.111 KVM_MEMORY_ENCRYPT_REG_REGION
3615 Parameters: struct kvm_enc_region (in)
3616 Returns: 0 on success; -1 on error
3618 This ioctl can be used to register a guest memory region which may
3619 contain encrypted data (e.g. guest RAM, SMRAM etc).
3621 It is used in the SEV-enabled guest. When encryption is enabled, a guest
3622 memory region may contain encrypted data. The SEV memory encryption
3623 engine uses a tweak such that two identical plaintext pages, each at
3624 different locations will have differing ciphertexts. So swapping or
3625 moving ciphertext of those pages will not result in plaintext being
3626 swapped. So relocating (or migrating) physical backing pages for the SEV
3627 guest will require some additional steps.
3629 Note: The current SEV key management spec does not provide commands to
3630 swap or migrate (move) ciphertext pages. Hence, for now we pin the guest
3631 memory region registered with the ioctl.
3633 4.112 KVM_MEMORY_ENCRYPT_UNREG_REGION
3638 Parameters: struct kvm_enc_region (in)
3639 Returns: 0 on success; -1 on error
3641 This ioctl can be used to unregister the guest memory region registered
3642 with KVM_MEMORY_ENCRYPT_REG_REGION ioctl above.
3644 4.113 KVM_HYPERV_EVENTFD
3646 Capability: KVM_CAP_HYPERV_EVENTFD
3649 Parameters: struct kvm_hyperv_eventfd (in)
3651 This ioctl (un)registers an eventfd to receive notifications from the guest on
3652 the specified Hyper-V connection id through the SIGNAL_EVENT hypercall, without
3653 causing a user exit. SIGNAL_EVENT hypercall with non-zero event flag number
3654 (bits 24-31) still triggers a KVM_EXIT_HYPERV_HCALL user exit.
3656 struct kvm_hyperv_eventfd {
3663 The conn_id field should fit within 24 bits:
3665 #define KVM_HYPERV_CONN_ID_MASK 0x00ffffff
3667 The acceptable values for the flags field are:
3669 #define KVM_HYPERV_EVENTFD_DEASSIGN (1 << 0)
3671 Returns: 0 on success,
3672 -EINVAL if conn_id or flags is outside the allowed range
3673 -ENOENT on deassign if the conn_id isn't registered
3674 -EEXIST on assign if the conn_id is already registered
3676 4.114 KVM_GET_NESTED_STATE
3678 Capability: KVM_CAP_NESTED_STATE
3681 Parameters: struct kvm_nested_state (in/out)
3682 Returns: 0 on success, -1 on error
3684 E2BIG: the total state size (including the fixed-size part of struct
3685 kvm_nested_state) exceeds the value of 'size' specified by
3686 the user; the size required will be written into size.
3688 struct kvm_nested_state {
3693 struct kvm_vmx_nested_state vmx;
3694 struct kvm_svm_nested_state svm;
3700 #define KVM_STATE_NESTED_GUEST_MODE 0x00000001
3701 #define KVM_STATE_NESTED_RUN_PENDING 0x00000002
3703 #define KVM_STATE_NESTED_SMM_GUEST_MODE 0x00000001
3704 #define KVM_STATE_NESTED_SMM_VMXON 0x00000002
3706 struct kvm_vmx_nested_state {
3715 This ioctl copies the vcpu's nested virtualization state from the kernel to
3718 The maximum size of the state, including the fixed-size part of struct
3719 kvm_nested_state, can be retrieved by passing KVM_CAP_NESTED_STATE to
3720 the KVM_CHECK_EXTENSION ioctl().
3722 4.115 KVM_SET_NESTED_STATE
3724 Capability: KVM_CAP_NESTED_STATE
3727 Parameters: struct kvm_nested_state (in)
3728 Returns: 0 on success, -1 on error
3730 This copies the vcpu's kvm_nested_state struct from userspace to the kernel. For
3731 the definition of struct kvm_nested_state, see KVM_GET_NESTED_STATE.
3733 4.116 KVM_(UN)REGISTER_COALESCED_MMIO
3735 Capability: KVM_CAP_COALESCED_MMIO (for coalesced mmio)
3736 KVM_CAP_COALESCED_PIO (for coalesced pio)
3739 Parameters: struct kvm_coalesced_mmio_zone
3740 Returns: 0 on success, < 0 on error
3742 Coalesced I/O is a performance optimization that defers hardware
3743 register write emulation so that userspace exits are avoided. It is
3744 typically used to reduce the overhead of emulating frequently accessed
3747 When a hardware register is configured for coalesced I/O, write accesses
3748 do not exit to userspace and their value is recorded in a ring buffer
3749 that is shared between kernel and userspace.
3751 Coalesced I/O is used if one or more write accesses to a hardware
3752 register can be deferred until a read or a write to another hardware
3753 register on the same device. This last access will cause a vmexit and
3754 userspace will process accesses from the ring buffer before emulating
3755 it. That will avoid exiting to userspace on repeated writes.
3757 Coalesced pio is based on coalesced mmio. There is little difference
3758 between coalesced mmio and pio except that coalesced pio records accesses
3761 5. The kvm_run structure
3762 ------------------------
3764 Application code obtains a pointer to the kvm_run structure by
3765 mmap()ing a vcpu fd. From that point, application code can control
3766 execution by changing fields in kvm_run prior to calling the KVM_RUN
3767 ioctl, and obtain information about the reason KVM_RUN returned by
3768 looking up structure members.
3772 __u8 request_interrupt_window;
3774 Request that KVM_RUN return when it becomes possible to inject external
3775 interrupts into the guest. Useful in conjunction with KVM_INTERRUPT.
3777 __u8 immediate_exit;
3779 This field is polled once when KVM_RUN starts; if non-zero, KVM_RUN
3780 exits immediately, returning -EINTR. In the common scenario where a
3781 signal is used to "kick" a VCPU out of KVM_RUN, this field can be used
3782 to avoid usage of KVM_SET_SIGNAL_MASK, which has worse scalability.
3783 Rather than blocking the signal outside KVM_RUN, userspace can set up
3784 a signal handler that sets run->immediate_exit to a non-zero value.
3786 This field is ignored if KVM_CAP_IMMEDIATE_EXIT is not available.
3793 When KVM_RUN has returned successfully (return value 0), this informs
3794 application code why KVM_RUN has returned. Allowable values for this
3795 field are detailed below.
3797 __u8 ready_for_interrupt_injection;
3799 If request_interrupt_window has been specified, this field indicates
3800 an interrupt can be injected now with KVM_INTERRUPT.
3804 The value of the current interrupt flag. Only valid if in-kernel
3805 local APIC is not used.
3809 More architecture-specific flags detailing state of the VCPU that may
3810 affect the device's behavior. The only currently defined flag is
3811 KVM_RUN_X86_SMM, which is valid on x86 machines and is set if the
3812 VCPU is in system management mode.
3814 /* in (pre_kvm_run), out (post_kvm_run) */
3817 The value of the cr8 register. Only valid if in-kernel local APIC is
3818 not used. Both input and output.
3822 The value of the APIC BASE msr. Only valid if in-kernel local
3823 APIC is not used. Both input and output.
3826 /* KVM_EXIT_UNKNOWN */
3828 __u64 hardware_exit_reason;
3831 If exit_reason is KVM_EXIT_UNKNOWN, the vcpu has exited due to unknown
3832 reasons. Further architecture-specific information is available in
3833 hardware_exit_reason.
3835 /* KVM_EXIT_FAIL_ENTRY */
3837 __u64 hardware_entry_failure_reason;
3840 If exit_reason is KVM_EXIT_FAIL_ENTRY, the vcpu could not be run due
3841 to unknown reasons. Further architecture-specific information is
3842 available in hardware_entry_failure_reason.
3844 /* KVM_EXIT_EXCEPTION */
3854 #define KVM_EXIT_IO_IN 0
3855 #define KVM_EXIT_IO_OUT 1
3857 __u8 size; /* bytes */
3860 __u64 data_offset; /* relative to kvm_run start */
3863 If exit_reason is KVM_EXIT_IO, then the vcpu has
3864 executed a port I/O instruction which could not be satisfied by kvm.
3865 data_offset describes where the data is located (KVM_EXIT_IO_OUT) or
3866 where kvm expects application code to place the data for the next
3867 KVM_RUN invocation (KVM_EXIT_IO_IN). Data format is a packed array.
3869 /* KVM_EXIT_DEBUG */
3871 struct kvm_debug_exit_arch arch;
3874 If the exit_reason is KVM_EXIT_DEBUG, then a vcpu is processing a debug event
3875 for which architecture specific information is returned.
3885 If exit_reason is KVM_EXIT_MMIO, then the vcpu has
3886 executed a memory-mapped I/O instruction which could not be satisfied
3887 by kvm. The 'data' member contains the written data if 'is_write' is
3888 true, and should be filled by application code otherwise.
3890 The 'data' member contains, in its first 'len' bytes, the value as it would
3891 appear if the VCPU performed a load or store of the appropriate width directly
3894 NOTE: For KVM_EXIT_IO, KVM_EXIT_MMIO, KVM_EXIT_OSI, KVM_EXIT_PAPR and
3895 KVM_EXIT_EPR the corresponding
3896 operations are complete (and guest state is consistent) only after userspace
3897 has re-entered the kernel with KVM_RUN. The kernel side will first finish
3898 incomplete operations and then check for pending signals. Userspace
3899 can re-enter the guest with an unmasked signal pending to complete
3902 /* KVM_EXIT_HYPERCALL */
3911 Unused. This was once used for 'hypercall to userspace'. To implement
3912 such functionality, use KVM_EXIT_IO (x86) or KVM_EXIT_MMIO (all except s390).
3913 Note KVM_EXIT_IO is significantly faster than KVM_EXIT_MMIO.
3915 /* KVM_EXIT_TPR_ACCESS */
3922 To be documented (KVM_TPR_ACCESS_REPORTING).
3924 /* KVM_EXIT_S390_SIEIC */
3927 __u64 mask; /* psw upper half */
3928 __u64 addr; /* psw lower half */
3935 /* KVM_EXIT_S390_RESET */
3936 #define KVM_S390_RESET_POR 1
3937 #define KVM_S390_RESET_CLEAR 2
3938 #define KVM_S390_RESET_SUBSYSTEM 4
3939 #define KVM_S390_RESET_CPU_INIT 8
3940 #define KVM_S390_RESET_IPL 16
3941 __u64 s390_reset_flags;
3945 /* KVM_EXIT_S390_UCONTROL */
3947 __u64 trans_exc_code;
3951 s390 specific. A page fault has occurred for a user controlled virtual
3952 machine (KVM_VM_S390_UNCONTROL) on it's host page table that cannot be
3953 resolved by the kernel.
3954 The program code and the translation exception code that were placed
3955 in the cpu's lowcore are presented here as defined by the z Architecture
3956 Principles of Operation Book in the Chapter for Dynamic Address Translation
3966 Deprecated - was used for 440 KVM.
3973 MOL uses a special hypercall interface it calls 'OSI'. To enable it, we catch
3974 hypercalls and exit with this exit struct that contains all the guest gprs.
3976 If exit_reason is KVM_EXIT_OSI, then the vcpu has triggered such a hypercall.
3977 Userspace can now handle the hypercall and when it's done modify the gprs as
3978 necessary. Upon guest entry all guest GPRs will then be replaced by the values
3981 /* KVM_EXIT_PAPR_HCALL */
3988 This is used on 64-bit PowerPC when emulating a pSeries partition,
3989 e.g. with the 'pseries' machine type in qemu. It occurs when the
3990 guest does a hypercall using the 'sc 1' instruction. The 'nr' field
3991 contains the hypercall number (from the guest R3), and 'args' contains
3992 the arguments (from the guest R4 - R12). Userspace should put the
3993 return code in 'ret' and any extra returned values in args[].
3994 The possible hypercalls are defined in the Power Architecture Platform
3995 Requirements (PAPR) document available from www.power.org (free
3996 developer registration required to access it).
3998 /* KVM_EXIT_S390_TSCH */
4000 __u16 subchannel_id;
4001 __u16 subchannel_nr;
4008 s390 specific. This exit occurs when KVM_CAP_S390_CSS_SUPPORT has been enabled
4009 and TEST SUBCHANNEL was intercepted. If dequeued is set, a pending I/O
4010 interrupt for the target subchannel has been dequeued and subchannel_id,
4011 subchannel_nr, io_int_parm and io_int_word contain the parameters for that
4012 interrupt. ipb is needed for instruction parameter decoding.
4019 On FSL BookE PowerPC chips, the interrupt controller has a fast patch
4020 interrupt acknowledge path to the core. When the core successfully
4021 delivers an interrupt, it automatically populates the EPR register with
4022 the interrupt vector number and acknowledges the interrupt inside
4023 the interrupt controller.
4025 In case the interrupt controller lives in user space, we need to do
4026 the interrupt acknowledge cycle through it to fetch the next to be
4027 delivered interrupt vector using this exit.
4029 It gets triggered whenever both KVM_CAP_PPC_EPR are enabled and an
4030 external interrupt has just been delivered into the guest. User space
4031 should put the acknowledged interrupt vector into the 'epr' field.
4033 /* KVM_EXIT_SYSTEM_EVENT */
4035 #define KVM_SYSTEM_EVENT_SHUTDOWN 1
4036 #define KVM_SYSTEM_EVENT_RESET 2
4037 #define KVM_SYSTEM_EVENT_CRASH 3
4042 If exit_reason is KVM_EXIT_SYSTEM_EVENT then the vcpu has triggered
4043 a system-level event using some architecture specific mechanism (hypercall
4044 or some special instruction). In case of ARM/ARM64, this is triggered using
4045 HVC instruction based PSCI call from the vcpu. The 'type' field describes
4046 the system-level event type. The 'flags' field describes architecture
4047 specific flags for the system-level event.
4049 Valid values for 'type' are:
4050 KVM_SYSTEM_EVENT_SHUTDOWN -- the guest has requested a shutdown of the
4051 VM. Userspace is not obliged to honour this, and if it does honour
4052 this does not need to destroy the VM synchronously (ie it may call
4053 KVM_RUN again before shutdown finally occurs).
4054 KVM_SYSTEM_EVENT_RESET -- the guest has requested a reset of the VM.
4055 As with SHUTDOWN, userspace can choose to ignore the request, or
4056 to schedule the reset to occur in the future and may call KVM_RUN again.
4057 KVM_SYSTEM_EVENT_CRASH -- the guest crash occurred and the guest
4058 has requested a crash condition maintenance. Userspace can choose
4059 to ignore the request, or to gather VM memory core dump and/or
4060 reset/shutdown of the VM.
4062 /* KVM_EXIT_IOAPIC_EOI */
4067 Indicates that the VCPU's in-kernel local APIC received an EOI for a
4068 level-triggered IOAPIC interrupt. This exit only triggers when the
4069 IOAPIC is implemented in userspace (i.e. KVM_CAP_SPLIT_IRQCHIP is enabled);
4070 the userspace IOAPIC should process the EOI and retrigger the interrupt if
4071 it is still asserted. Vector is the LAPIC interrupt vector for which the
4074 struct kvm_hyperv_exit {
4075 #define KVM_EXIT_HYPERV_SYNIC 1
4076 #define KVM_EXIT_HYPERV_HCALL 2
4092 /* KVM_EXIT_HYPERV */
4093 struct kvm_hyperv_exit hyperv;
4094 Indicates that the VCPU exits into userspace to process some tasks
4095 related to Hyper-V emulation.
4096 Valid values for 'type' are:
4097 KVM_EXIT_HYPERV_SYNIC -- synchronously notify user-space about
4098 Hyper-V SynIC state change. Notification is used to remap SynIC
4099 event/message pages and to enable/disable SynIC messages/events processing
4102 /* Fix the size of the union. */
4107 * shared registers between kvm and userspace.
4108 * kvm_valid_regs specifies the register classes set by the host
4109 * kvm_dirty_regs specified the register classes dirtied by userspace
4110 * struct kvm_sync_regs is architecture specific, as well as the
4111 * bits for kvm_valid_regs and kvm_dirty_regs
4113 __u64 kvm_valid_regs;
4114 __u64 kvm_dirty_regs;
4116 struct kvm_sync_regs regs;
4117 char padding[SYNC_REGS_SIZE_BYTES];
4120 If KVM_CAP_SYNC_REGS is defined, these fields allow userspace to access
4121 certain guest registers without having to call SET/GET_*REGS. Thus we can
4122 avoid some system call overhead if userspace has to handle the exit.
4123 Userspace can query the validity of the structure by checking
4124 kvm_valid_regs for specific bits. These bits are architecture specific
4125 and usually define the validity of a groups of registers. (e.g. one bit
4126 for general purpose registers)
4128 Please note that the kernel is allowed to use the kvm_run structure as the
4129 primary storage for certain register types. Therefore, the kernel may use the
4130 values in kvm_run even if the corresponding bit in kvm_dirty_regs is not set.
4136 6. Capabilities that can be enabled on vCPUs
4137 --------------------------------------------
4139 There are certain capabilities that change the behavior of the virtual CPU or
4140 the virtual machine when enabled. To enable them, please see section 4.37.
4141 Below you can find a list of capabilities and what their effect on the vCPU or
4142 the virtual machine is when enabling them.
4144 The following information is provided along with the description:
4146 Architectures: which instruction set architectures provide this ioctl.
4147 x86 includes both i386 and x86_64.
4149 Target: whether this is a per-vcpu or per-vm capability.
4151 Parameters: what parameters are accepted by the capability.
4153 Returns: the return value. General error numbers (EBADF, ENOMEM, EINVAL)
4154 are not detailed, but errors with specific meanings are.
4162 Returns: 0 on success; -1 on error
4164 This capability enables interception of OSI hypercalls that otherwise would
4165 be treated as normal system calls to be injected into the guest. OSI hypercalls
4166 were invented by Mac-on-Linux to have a standardized communication mechanism
4167 between the guest and the host.
4169 When this capability is enabled, KVM_EXIT_OSI can occur.
4172 6.2 KVM_CAP_PPC_PAPR
4177 Returns: 0 on success; -1 on error
4179 This capability enables interception of PAPR hypercalls. PAPR hypercalls are
4180 done using the hypercall instruction "sc 1".
4182 It also sets the guest privilege level to "supervisor" mode. Usually the guest
4183 runs in "hypervisor" privilege mode with a few missing features.
4185 In addition to the above, it changes the semantics of SDR1. In this mode, the
4186 HTAB address part of SDR1 contains an HVA instead of a GPA, as PAPR keeps the
4187 HTAB invisible to the guest.
4189 When this capability is enabled, KVM_EXIT_PAPR_HCALL can occur.
4196 Parameters: args[0] is the address of a struct kvm_config_tlb
4197 Returns: 0 on success; -1 on error
4199 struct kvm_config_tlb {
4206 Configures the virtual CPU's TLB array, establishing a shared memory area
4207 between userspace and KVM. The "params" and "array" fields are userspace
4208 addresses of mmu-type-specific data structures. The "array_len" field is an
4209 safety mechanism, and should be set to the size in bytes of the memory that
4210 userspace has reserved for the array. It must be at least the size dictated
4211 by "mmu_type" and "params".
4213 While KVM_RUN is active, the shared region is under control of KVM. Its
4214 contents are undefined, and any modification by userspace results in
4215 boundedly undefined behavior.
4217 On return from KVM_RUN, the shared region will reflect the current state of
4218 the guest's TLB. If userspace makes any changes, it must call KVM_DIRTY_TLB
4219 to tell KVM which entries have been changed, prior to calling KVM_RUN again
4222 For mmu types KVM_MMU_FSL_BOOKE_NOHV and KVM_MMU_FSL_BOOKE_HV:
4223 - The "params" field is of type "struct kvm_book3e_206_tlb_params".
4224 - The "array" field points to an array of type "struct
4225 kvm_book3e_206_tlb_entry".
4226 - The array consists of all entries in the first TLB, followed by all
4227 entries in the second TLB.
4228 - Within a TLB, entries are ordered first by increasing set number. Within a
4229 set, entries are ordered by way (increasing ESEL).
4230 - The hash for determining set number in TLB0 is: (MAS2 >> 12) & (num_sets - 1)
4231 where "num_sets" is the tlb_sizes[] value divided by the tlb_ways[] value.
4232 - The tsize field of mas1 shall be set to 4K on TLB0, even though the
4233 hardware ignores this value for TLB0.
4235 6.4 KVM_CAP_S390_CSS_SUPPORT
4240 Returns: 0 on success; -1 on error
4242 This capability enables support for handling of channel I/O instructions.
4244 TEST PENDING INTERRUPTION and the interrupt portion of TEST SUBCHANNEL are
4245 handled in-kernel, while the other I/O instructions are passed to userspace.
4247 When this capability is enabled, KVM_EXIT_S390_TSCH will occur on TEST
4248 SUBCHANNEL intercepts.
4250 Note that even though this capability is enabled per-vcpu, the complete
4251 virtual machine is affected.
4257 Parameters: args[0] defines whether the proxy facility is active
4258 Returns: 0 on success; -1 on error
4260 This capability enables or disables the delivery of interrupts through the
4261 external proxy facility.
4263 When enabled (args[0] != 0), every time the guest gets an external interrupt
4264 delivered, it automatically exits into user space with a KVM_EXIT_EPR exit
4265 to receive the topmost interrupt vector.
4267 When disabled (args[0] == 0), behavior is as if this facility is unsupported.
4269 When this capability is enabled, KVM_EXIT_EPR can occur.
4271 6.6 KVM_CAP_IRQ_MPIC
4274 Parameters: args[0] is the MPIC device fd
4275 args[1] is the MPIC CPU number for this vcpu
4277 This capability connects the vcpu to an in-kernel MPIC device.
4279 6.7 KVM_CAP_IRQ_XICS
4283 Parameters: args[0] is the XICS device fd
4284 args[1] is the XICS CPU number (server ID) for this vcpu
4286 This capability connects the vcpu to an in-kernel XICS device.
4288 6.8 KVM_CAP_S390_IRQCHIP
4294 This capability enables the in-kernel irqchip for s390. Please refer to
4295 "4.24 KVM_CREATE_IRQCHIP" for details.
4297 6.9 KVM_CAP_MIPS_FPU
4301 Parameters: args[0] is reserved for future use (should be 0).
4303 This capability allows the use of the host Floating Point Unit by the guest. It
4304 allows the Config1.FP bit to be set to enable the FPU in the guest. Once this is
4305 done the KVM_REG_MIPS_FPR_* and KVM_REG_MIPS_FCR_* registers can be accessed
4306 (depending on the current guest FPU register mode), and the Status.FR,
4307 Config5.FRE bits are accessible via the KVM API and also from the guest,
4308 depending on them being supported by the FPU.
4310 6.10 KVM_CAP_MIPS_MSA
4314 Parameters: args[0] is reserved for future use (should be 0).
4316 This capability allows the use of the MIPS SIMD Architecture (MSA) by the guest.
4317 It allows the Config3.MSAP bit to be set to enable the use of MSA by the guest.
4318 Once this is done the KVM_REG_MIPS_VEC_* and KVM_REG_MIPS_MSA_* registers can be
4319 accessed, and the Config5.MSAEn bit is accessible via the KVM API and also from
4322 6.74 KVM_CAP_SYNC_REGS
4323 Architectures: s390, x86
4324 Target: s390: always enabled, x86: vcpu
4326 Returns: x86: KVM_CHECK_EXTENSION returns a bit-array indicating which register
4327 sets are supported (bitfields defined in arch/x86/include/uapi/asm/kvm.h).
4329 As described above in the kvm_sync_regs struct info in section 5 (kvm_run):
4330 KVM_CAP_SYNC_REGS "allow[s] userspace to access certain guest registers
4331 without having to call SET/GET_*REGS". This reduces overhead by eliminating
4332 repeated ioctl calls for setting and/or getting register values. This is
4333 particularly important when userspace is making synchronous guest state
4334 modifications, e.g. when emulating and/or intercepting instructions in
4337 For s390 specifics, please refer to the source code.
4340 - the register sets to be copied out to kvm_run are selectable
4341 by userspace (rather that all sets being copied out for every exit).
4342 - vcpu_events are available in addition to regs and sregs.
4344 For x86, the 'kvm_valid_regs' field of struct kvm_run is overloaded to
4345 function as an input bit-array field set by userspace to indicate the
4346 specific register sets to be copied out on the next exit.
4348 To indicate when userspace has modified values that should be copied into
4349 the vCPU, the all architecture bitarray field, 'kvm_dirty_regs' must be set.
4350 This is done using the same bitflags as for the 'kvm_valid_regs' field.
4351 If the dirty bit is not set, then the register set values will not be copied
4352 into the vCPU even if they've been modified.
4354 Unused bitfields in the bitarrays must be set to zero.
4356 struct kvm_sync_regs {
4357 struct kvm_regs regs;
4358 struct kvm_sregs sregs;
4359 struct kvm_vcpu_events events;
4362 7. Capabilities that can be enabled on VMs
4363 ------------------------------------------
4365 There are certain capabilities that change the behavior of the virtual
4366 machine when enabled. To enable them, please see section 4.37. Below
4367 you can find a list of capabilities and what their effect on the VM
4368 is when enabling them.
4370 The following information is provided along with the description:
4372 Architectures: which instruction set architectures provide this ioctl.
4373 x86 includes both i386 and x86_64.
4375 Parameters: what parameters are accepted by the capability.
4377 Returns: the return value. General error numbers (EBADF, ENOMEM, EINVAL)
4378 are not detailed, but errors with specific meanings are.
4381 7.1 KVM_CAP_PPC_ENABLE_HCALL
4384 Parameters: args[0] is the sPAPR hcall number
4385 args[1] is 0 to disable, 1 to enable in-kernel handling
4387 This capability controls whether individual sPAPR hypercalls (hcalls)
4388 get handled by the kernel or not. Enabling or disabling in-kernel
4389 handling of an hcall is effective across the VM. On creation, an
4390 initial set of hcalls are enabled for in-kernel handling, which
4391 consists of those hcalls for which in-kernel handlers were implemented
4392 before this capability was implemented. If disabled, the kernel will
4393 not to attempt to handle the hcall, but will always exit to userspace
4394 to handle it. Note that it may not make sense to enable some and
4395 disable others of a group of related hcalls, but KVM does not prevent
4396 userspace from doing that.
4398 If the hcall number specified is not one that has an in-kernel
4399 implementation, the KVM_ENABLE_CAP ioctl will fail with an EINVAL
4402 7.2 KVM_CAP_S390_USER_SIGP
4407 This capability controls which SIGP orders will be handled completely in user
4408 space. With this capability enabled, all fast orders will be handled completely
4414 - CONDITIONAL EMERGENCY SIGNAL
4416 All other orders will be handled completely in user space.
4418 Only privileged operation exceptions will be checked for in the kernel (or even
4419 in the hardware prior to interception). If this capability is not enabled, the
4420 old way of handling SIGP orders is used (partially in kernel and user space).
4422 7.3 KVM_CAP_S390_VECTOR_REGISTERS
4426 Returns: 0 on success, negative value on error
4428 Allows use of the vector registers introduced with z13 processor, and
4429 provides for the synchronization between host and user space. Will
4430 return -EINVAL if the machine does not support vectors.
4432 7.4 KVM_CAP_S390_USER_STSI
4437 This capability allows post-handlers for the STSI instruction. After
4438 initial handling in the kernel, KVM exits to user space with
4439 KVM_EXIT_S390_STSI to allow user space to insert further data.
4441 Before exiting to userspace, kvm handlers should fill in s390_stsi field of
4452 @addr - guest address of STSI SYSIB
4456 @ar - access register number
4458 KVM handlers should exit to userspace with rc = -EREMOTE.
4460 7.5 KVM_CAP_SPLIT_IRQCHIP
4463 Parameters: args[0] - number of routes reserved for userspace IOAPICs
4464 Returns: 0 on success, -1 on error
4466 Create a local apic for each processor in the kernel. This can be used
4467 instead of KVM_CREATE_IRQCHIP if the userspace VMM wishes to emulate the
4468 IOAPIC and PIC (and also the PIT, even though this has to be enabled
4471 This capability also enables in kernel routing of interrupt requests;
4472 when KVM_CAP_SPLIT_IRQCHIP only routes of KVM_IRQ_ROUTING_MSI type are
4473 used in the IRQ routing table. The first args[0] MSI routes are reserved
4474 for the IOAPIC pins. Whenever the LAPIC receives an EOI for these routes,
4475 a KVM_EXIT_IOAPIC_EOI vmexit will be reported to userspace.
4477 Fails if VCPU has already been created, or if the irqchip is already in the
4478 kernel (i.e. KVM_CREATE_IRQCHIP has already been called).
4485 Allows use of runtime-instrumentation introduced with zEC12 processor.
4486 Will return -EINVAL if the machine does not support runtime-instrumentation.
4487 Will return -EBUSY if a VCPU has already been created.
4489 7.7 KVM_CAP_X2APIC_API
4492 Parameters: args[0] - features that should be enabled
4493 Returns: 0 on success, -EINVAL when args[0] contains invalid features
4495 Valid feature flags in args[0] are
4497 #define KVM_X2APIC_API_USE_32BIT_IDS (1ULL << 0)
4498 #define KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK (1ULL << 1)
4500 Enabling KVM_X2APIC_API_USE_32BIT_IDS changes the behavior of
4501 KVM_SET_GSI_ROUTING, KVM_SIGNAL_MSI, KVM_SET_LAPIC, and KVM_GET_LAPIC,
4502 allowing the use of 32-bit APIC IDs. See KVM_CAP_X2APIC_API in their
4503 respective sections.
4505 KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK must be enabled for x2APIC to work
4506 in logical mode or with more than 255 VCPUs. Otherwise, KVM treats 0xff
4507 as a broadcast even in x2APIC mode in order to support physical x2APIC
4508 without interrupt remapping. This is undesirable in logical mode,
4509 where 0xff represents CPUs 0-7 in cluster 0.
4511 7.8 KVM_CAP_S390_USER_INSTR0
4516 With this capability enabled, all illegal instructions 0x0000 (2 bytes) will
4517 be intercepted and forwarded to user space. User space can use this
4518 mechanism e.g. to realize 2-byte software breakpoints. The kernel will
4519 not inject an operating exception for these instructions, user space has
4520 to take care of that.
4522 This capability can be enabled dynamically even if VCPUs were already
4523 created and are running.
4529 Returns: 0 on success; -EINVAL if the machine does not support
4530 guarded storage; -EBUSY if a VCPU has already been created.
4532 Allows use of guarded storage for the KVM guest.
4534 7.10 KVM_CAP_S390_AIS
4539 Allow use of adapter-interruption suppression.
4540 Returns: 0 on success; -EBUSY if a VCPU has already been created.
4542 7.11 KVM_CAP_PPC_SMT
4545 Parameters: vsmt_mode, flags
4547 Enabling this capability on a VM provides userspace with a way to set
4548 the desired virtual SMT mode (i.e. the number of virtual CPUs per
4549 virtual core). The virtual SMT mode, vsmt_mode, must be a power of 2
4550 between 1 and 8. On POWER8, vsmt_mode must also be no greater than
4551 the number of threads per subcore for the host. Currently flags must
4552 be 0. A successful call to enable this capability will result in
4553 vsmt_mode being returned when the KVM_CAP_PPC_SMT capability is
4554 subsequently queried for the VM. This capability is only supported by
4555 HV KVM, and can only be set before any VCPUs have been created.
4556 The KVM_CAP_PPC_SMT_POSSIBLE capability indicates which virtual SMT
4557 modes are available.
4559 7.12 KVM_CAP_PPC_FWNMI
4564 With this capability a machine check exception in the guest address
4565 space will cause KVM to exit the guest with NMI exit reason. This
4566 enables QEMU to build error log and branch to guest kernel registered
4567 machine check handling routine. Without this capability KVM will
4568 branch to guests' 0x200 interrupt vector.
4570 7.13 KVM_CAP_X86_DISABLE_EXITS
4573 Parameters: args[0] defines which exits are disabled
4574 Returns: 0 on success, -EINVAL when args[0] contains invalid exits
4576 Valid bits in args[0] are
4578 #define KVM_X86_DISABLE_EXITS_MWAIT (1 << 0)
4579 #define KVM_X86_DISABLE_EXITS_HLT (1 << 1)
4581 Enabling this capability on a VM provides userspace with a way to no
4582 longer intercept some instructions for improved latency in some
4583 workloads, and is suggested when vCPUs are associated to dedicated
4584 physical CPUs. More bits can be added in the future; userspace can
4585 just pass the KVM_CHECK_EXTENSION result to KVM_ENABLE_CAP to disable
4588 Do not enable KVM_FEATURE_PV_UNHALT if you disable HLT exits.
4590 7.14 KVM_CAP_S390_HPAGE_1M
4594 Returns: 0 on success, -EINVAL if hpage module parameter was not set
4595 or cmma is enabled, or the VM has the KVM_VM_S390_UCONTROL
4598 With this capability the KVM support for memory backing with 1m pages
4599 through hugetlbfs can be enabled for a VM. After the capability is
4600 enabled, cmma can't be enabled anymore and pfmfi and the storage key
4601 interpretation are disabled. If cmma has already been enabled or the
4602 hpage module parameter is not set to 1, -EINVAL is returned.
4604 While it is generally possible to create a huge page backed VM without
4605 this capability, the VM will not be able to run.
4607 7.15 KVM_CAP_MSR_PLATFORM_INFO
4610 Parameters: args[0] whether feature should be enabled or not
4612 With this capability, a guest may read the MSR_PLATFORM_INFO MSR. Otherwise,
4613 a #GP would be raised when the guest tries to access. Currently, this
4614 capability does not enable write permissions of this MSR for the guest.
4616 7.16 KVM_CAP_PPC_NESTED_HV
4620 Returns: 0 on success, -EINVAL when the implementation doesn't support
4621 nested-HV virtualization.
4623 HV-KVM on POWER9 and later systems allows for "nested-HV"
4624 virtualization, which provides a way for a guest VM to run guests that
4625 can run using the CPU's supervisor mode (privileged non-hypervisor
4626 state). Enabling this capability on a VM depends on the CPU having
4627 the necessary functionality and on the facility being enabled with a
4628 kvm-hv module parameter.
4630 7.17 KVM_CAP_EXCEPTION_PAYLOAD
4633 Parameters: args[0] whether feature should be enabled or not
4635 With this capability enabled, CR2 will not be modified prior to the
4636 emulated VM-exit when L1 intercepts a #PF exception that occurs in
4637 L2. Similarly, for kvm-intel only, DR6 will not be modified prior to
4638 the emulated VM-exit when L1 intercepts a #DB exception that occurs in
4639 L2. As a result, when KVM_GET_VCPU_EVENTS reports a pending #PF (or
4640 #DB) exception for L2, exception.has_payload will be set and the
4641 faulting address (or the new DR6 bits*) will be reported in the
4642 exception_payload field. Similarly, when userspace injects a #PF (or
4643 #DB) into L2 using KVM_SET_VCPU_EVENTS, it is expected to set
4644 exception.has_payload and to put the faulting address (or the new DR6
4645 bits*) in the exception_payload field.
4647 This capability also enables exception.pending in struct
4648 kvm_vcpu_events, which allows userspace to distinguish between pending
4649 and injected exceptions.
4652 * For the new DR6 bits, note that bit 16 is set iff the #DB exception
4655 8. Other capabilities.
4656 ----------------------
4658 This section lists capabilities that give information about other
4659 features of the KVM implementation.
4661 8.1 KVM_CAP_PPC_HWRNG
4665 This capability, if KVM_CHECK_EXTENSION indicates that it is
4666 available, means that that the kernel has an implementation of the
4667 H_RANDOM hypercall backed by a hardware random-number generator.
4668 If present, the kernel H_RANDOM handler can be enabled for guest use
4669 with the KVM_CAP_PPC_ENABLE_HCALL capability.
4671 8.2 KVM_CAP_HYPERV_SYNIC
4674 This capability, if KVM_CHECK_EXTENSION indicates that it is
4675 available, means that that the kernel has an implementation of the
4676 Hyper-V Synthetic interrupt controller(SynIC). Hyper-V SynIC is
4677 used to support Windows Hyper-V based guest paravirt drivers(VMBus).
4679 In order to use SynIC, it has to be activated by setting this
4680 capability via KVM_ENABLE_CAP ioctl on the vcpu fd. Note that this
4681 will disable the use of APIC hardware virtualization even if supported
4682 by the CPU, as it's incompatible with SynIC auto-EOI behavior.
4684 8.3 KVM_CAP_PPC_RADIX_MMU
4688 This capability, if KVM_CHECK_EXTENSION indicates that it is
4689 available, means that that the kernel can support guests using the
4690 radix MMU defined in Power ISA V3.00 (as implemented in the POWER9
4693 8.4 KVM_CAP_PPC_HASH_MMU_V3
4697 This capability, if KVM_CHECK_EXTENSION indicates that it is
4698 available, means that that the kernel can support guests using the
4699 hashed page table MMU defined in Power ISA V3.00 (as implemented in
4700 the POWER9 processor), including in-memory segment tables.
4706 This capability, if KVM_CHECK_EXTENSION on the main kvm handle indicates that
4707 it is available, means that full hardware assisted virtualization capabilities
4708 of the hardware are available for use through KVM. An appropriate
4709 KVM_VM_MIPS_* type must be passed to KVM_CREATE_VM to create a VM which
4712 If KVM_CHECK_EXTENSION on a kvm VM handle indicates that this capability is
4713 available, it means that the VM is using full hardware assisted virtualization
4714 capabilities of the hardware. This is useful to check after creating a VM with
4715 KVM_VM_MIPS_DEFAULT.
4717 The value returned by KVM_CHECK_EXTENSION should be compared against known
4718 values (see below). All other values are reserved. This is to allow for the
4719 possibility of other hardware assisted virtualization implementations which
4720 may be incompatible with the MIPS VZ ASE.
4722 0: The trap & emulate implementation is in use to run guest code in user
4723 mode. Guest virtual memory segments are rearranged to fit the guest in the
4724 user mode address space.
4726 1: The MIPS VZ ASE is in use, providing full hardware assisted
4727 virtualization, including standard guest virtual memory segments.
4733 This capability, if KVM_CHECK_EXTENSION on the main kvm handle indicates that
4734 it is available, means that the trap & emulate implementation is available to
4735 run guest code in user mode, even if KVM_CAP_MIPS_VZ indicates that hardware
4736 assisted virtualisation is also available. KVM_VM_MIPS_TE (0) must be passed
4737 to KVM_CREATE_VM to create a VM which utilises it.
4739 If KVM_CHECK_EXTENSION on a kvm VM handle indicates that this capability is
4740 available, it means that the VM is using trap & emulate.
4742 8.7 KVM_CAP_MIPS_64BIT
4746 This capability indicates the supported architecture type of the guest, i.e. the
4747 supported register and address width.
4749 The values returned when this capability is checked by KVM_CHECK_EXTENSION on a
4750 kvm VM handle correspond roughly to the CP0_Config.AT register field, and should
4751 be checked specifically against known values (see below). All other values are
4754 0: MIPS32 or microMIPS32.
4755 Both registers and addresses are 32-bits wide.
4756 It will only be possible to run 32-bit guest code.
4758 1: MIPS64 or microMIPS64 with access only to 32-bit compatibility segments.
4759 Registers are 64-bits wide, but addresses are 32-bits wide.
4760 64-bit guest code may run but cannot access MIPS64 memory segments.
4761 It will also be possible to run 32-bit guest code.
4763 2: MIPS64 or microMIPS64 with access to all address segments.
4764 Both registers and addresses are 64-bits wide.
4765 It will be possible to run 64-bit or 32-bit guest code.
4767 8.9 KVM_CAP_ARM_USER_IRQ
4769 Architectures: arm, arm64
4770 This capability, if KVM_CHECK_EXTENSION indicates that it is available, means
4771 that if userspace creates a VM without an in-kernel interrupt controller, it
4772 will be notified of changes to the output level of in-kernel emulated devices,
4773 which can generate virtual interrupts, presented to the VM.
4774 For such VMs, on every return to userspace, the kernel
4775 updates the vcpu's run->s.regs.device_irq_level field to represent the actual
4776 output level of the device.
4778 Whenever kvm detects a change in the device output level, kvm guarantees at
4779 least one return to userspace before running the VM. This exit could either
4780 be a KVM_EXIT_INTR or any other exit event, like KVM_EXIT_MMIO. This way,
4781 userspace can always sample the device output level and re-compute the state of
4782 the userspace interrupt controller. Userspace should always check the state
4783 of run->s.regs.device_irq_level on every kvm exit.
4784 The value in run->s.regs.device_irq_level can represent both level and edge
4785 triggered interrupt signals, depending on the device. Edge triggered interrupt
4786 signals will exit to userspace with the bit in run->s.regs.device_irq_level
4787 set exactly once per edge signal.
4789 The field run->s.regs.device_irq_level is available independent of
4790 run->kvm_valid_regs or run->kvm_dirty_regs bits.
4792 If KVM_CAP_ARM_USER_IRQ is supported, the KVM_CHECK_EXTENSION ioctl returns a
4793 number larger than 0 indicating the version of this capability is implemented
4794 and thereby which bits in in run->s.regs.device_irq_level can signal values.
4796 Currently the following bits are defined for the device_irq_level bitmap:
4798 KVM_CAP_ARM_USER_IRQ >= 1:
4800 KVM_ARM_DEV_EL1_VTIMER - EL1 virtual timer
4801 KVM_ARM_DEV_EL1_PTIMER - EL1 physical timer
4802 KVM_ARM_DEV_PMU - ARM PMU overflow interrupt signal
4804 Future versions of kvm may implement additional events. These will get
4805 indicated by returning a higher number from KVM_CHECK_EXTENSION and will be
4808 8.10 KVM_CAP_PPC_SMT_POSSIBLE
4812 Querying this capability returns a bitmap indicating the possible
4813 virtual SMT modes that can be set using KVM_CAP_PPC_SMT. If bit N
4814 (counting from the right) is set, then a virtual SMT mode of 2^N is
4817 8.11 KVM_CAP_HYPERV_SYNIC2
4821 This capability enables a newer version of Hyper-V Synthetic interrupt
4822 controller (SynIC). The only difference with KVM_CAP_HYPERV_SYNIC is that KVM
4823 doesn't clear SynIC message and event flags pages when they are enabled by
4824 writing to the respective MSRs.
4826 8.12 KVM_CAP_HYPERV_VP_INDEX
4830 This capability indicates that userspace can load HV_X64_MSR_VP_INDEX msr. Its
4831 value is used to denote the target vcpu for a SynIC interrupt. For
4832 compatibilty, KVM initializes this msr to KVM's internal vcpu index. When this
4833 capability is absent, userspace can still query this msr's value.
4835 8.13 KVM_CAP_S390_AIS_MIGRATION
4840 This capability indicates if the flic device will be able to get/set the
4841 AIS states for migration via the KVM_DEV_FLIC_AISM_ALL attribute and allows
4842 to discover this without having to create a flic device.
4844 8.14 KVM_CAP_S390_PSW
4848 This capability indicates that the PSW is exposed via the kvm_run structure.
4850 8.15 KVM_CAP_S390_GMAP
4854 This capability indicates that the user space memory used as guest mapping can
4855 be anywhere in the user memory address space, as long as the memory slots are
4856 aligned and sized to a segment (1MB) boundary.
4858 8.16 KVM_CAP_S390_COW
4862 This capability indicates that the user space memory used as guest mapping can
4863 use copy-on-write semantics as well as dirty pages tracking via read-only page
4866 8.17 KVM_CAP_S390_BPB
4870 This capability indicates that kvm will implement the interfaces to handle
4871 reset, migration and nested KVM for branch prediction blocking. The stfle
4872 facility 82 should not be provided to the guest without this capability.
4874 8.18 KVM_CAP_HYPERV_TLBFLUSH
4878 This capability indicates that KVM supports paravirtualized Hyper-V TLB Flush
4880 HvFlushVirtualAddressSpace, HvFlushVirtualAddressSpaceEx,
4881 HvFlushVirtualAddressList, HvFlushVirtualAddressListEx.
4883 8.19 KVM_CAP_ARM_INJECT_SERROR_ESR
4885 Architectures: arm, arm64
4887 This capability indicates that userspace can specify (via the
4888 KVM_SET_VCPU_EVENTS ioctl) the syndrome value reported to the guest when it
4889 takes a virtual SError interrupt exception.
4890 If KVM advertises this capability, userspace can only specify the ISS field for
4891 the ESR syndrome. Other parts of the ESR, such as the EC are generated by the
4892 CPU when the exception is taken. If this virtual SError is taken to EL1 using
4893 AArch64, this value will be reported in the ISS field of ESR_ELx.
4895 See KVM_CAP_VCPU_EVENTS for more details.
4896 8.20 KVM_CAP_HYPERV_SEND_IPI
4900 This capability indicates that KVM supports paravirtualized Hyper-V IPI send
4902 HvCallSendSyntheticClusterIpi, HvCallSendSyntheticClusterIpiEx.