1 The Definitive KVM (Kernel-based Virtual Machine) API Documentation
2 ===================================================================
6 The kvm API is a set of ioctls that are issued to control various aspects
7 of a virtual machine. The ioctls belong to three classes
9 - System ioctls: These query and set global attributes which affect the
10 whole kvm subsystem. In addition a system ioctl is used to create
13 - VM ioctls: These query and set attributes that affect an entire virtual
14 machine, for example memory layout. In addition a VM ioctl is used to
15 create virtual cpus (vcpus).
17 Only run VM ioctls from the same process (address space) that was used
20 - vcpu ioctls: These query and set attributes that control the operation
21 of a single virtual cpu.
23 Only run vcpu ioctls from the same thread that was used to create the
28 The kvm API is centered around file descriptors. An initial
29 open("/dev/kvm") obtains a handle to the kvm subsystem; this handle
30 can be used to issue system ioctls. A KVM_CREATE_VM ioctl on this
31 handle will create a VM file descriptor which can be used to issue VM
32 ioctls. A KVM_CREATE_VCPU ioctl on a VM fd will create a virtual cpu
33 and return a file descriptor pointing to it. Finally, ioctls on a vcpu
34 fd can be used to control the vcpu, including the important task of
35 actually running guest code.
37 In general file descriptors can be migrated among processes by means
38 of fork() and the SCM_RIGHTS facility of unix domain socket. These
39 kinds of tricks are explicitly not supported by kvm. While they will
40 not cause harm to the host, their actual behavior is not guaranteed by
41 the API. The only supported use is one virtual machine per process,
42 and one vcpu per thread.
46 As of Linux 2.6.22, the KVM ABI has been stabilized: no backward
47 incompatible change are allowed. However, there is an extension
48 facility that allows backward-compatible extensions to the API to be
51 The extension mechanism is not based on on the Linux version number.
52 Instead, kvm defines extension identifiers and a facility to query
53 whether a particular extension identifier is available. If it is, a
54 set of ioctls is available for application use.
58 This section describes ioctls that can be used to control kvm guests.
59 For each ioctl, the following information is provided along with a
62 Capability: which KVM extension provides this ioctl. Can be 'basic',
63 which means that is will be provided by any kernel that supports
64 API version 12 (see section 4.1), or a KVM_CAP_xyz constant, which
65 means availability needs to be checked with KVM_CHECK_EXTENSION
68 Architectures: which instruction set architectures provide this ioctl.
69 x86 includes both i386 and x86_64.
71 Type: system, vm, or vcpu.
73 Parameters: what parameters are accepted by the ioctl.
75 Returns: the return value. General error numbers (EBADF, ENOMEM, EINVAL)
76 are not detailed, but errors with specific meanings are.
78 4.1 KVM_GET_API_VERSION
84 Returns: the constant KVM_API_VERSION (=12)
86 This identifies the API version as the stable kvm API. It is not
87 expected that this number will change. However, Linux 2.6.20 and
88 2.6.21 report earlier versions; these are not documented and not
89 supported. Applications should refuse to run if KVM_GET_API_VERSION
90 returns a value other than 12. If this check passes, all ioctls
91 described as 'basic' will be available.
99 Returns: a VM fd that can be used to control the new virtual machine.
101 The new VM has no virtual cpus and no memory. An mmap() of a VM fd
102 will access the virtual machine's physical address space; offset zero
103 corresponds to guest physical address zero. Use of mmap() on a VM fd
104 is discouraged if userspace memory allocation (KVM_CAP_USER_MEMORY) is
107 4.3 KVM_GET_MSR_INDEX_LIST
112 Parameters: struct kvm_msr_list (in/out)
113 Returns: 0 on success; -1 on error
115 E2BIG: the msr index list is to be to fit in the array specified by
118 struct kvm_msr_list {
119 __u32 nmsrs; /* number of msrs in entries */
123 This ioctl returns the guest msrs that are supported. The list varies
124 by kvm version and host processor, but does not change otherwise. The
125 user fills in the size of the indices array in nmsrs, and in return
126 kvm adjusts nmsrs to reflect the actual number of msrs and fills in
127 the indices array with their numbers.
129 Note: if kvm indicates supports MCE (KVM_CAP_MCE), then the MCE bank MSRs are
130 not returned in the MSR list, as different vcpus can have a different number
131 of banks, as set via the KVM_X86_SETUP_MCE ioctl.
133 4.4 KVM_CHECK_EXTENSION
138 Parameters: extension identifier (KVM_CAP_*)
139 Returns: 0 if unsupported; 1 (or some other positive integer) if supported
141 The API allows the application to query about extensions to the core
142 kvm API. Userspace passes an extension identifier (an integer) and
143 receives an integer that describes the extension availability.
144 Generally 0 means no and 1 means yes, but some extensions may report
145 additional information in the integer return value.
147 4.5 KVM_GET_VCPU_MMAP_SIZE
153 Returns: size of vcpu mmap area, in bytes
155 The KVM_RUN ioctl (cf.) communicates with userspace via a shared
156 memory region. This ioctl returns the size of that region. See the
157 KVM_RUN documentation for details.
159 4.6 KVM_SET_MEMORY_REGION
164 Parameters: struct kvm_memory_region (in)
165 Returns: 0 on success, -1 on error
167 This ioctl is obsolete and has been removed.
174 Parameters: vcpu id (apic id on x86)
175 Returns: vcpu fd on success, -1 on error
177 This API adds a vcpu to a virtual machine. The vcpu id is a small integer
178 in the range [0, max_vcpus).
180 4.8 KVM_GET_DIRTY_LOG (vm ioctl)
185 Parameters: struct kvm_dirty_log (in/out)
186 Returns: 0 on success, -1 on error
188 /* for KVM_GET_DIRTY_LOG */
189 struct kvm_dirty_log {
193 void __user *dirty_bitmap; /* one bit per page */
198 Given a memory slot, return a bitmap containing any pages dirtied
199 since the last call to this ioctl. Bit 0 is the first page in the
200 memory slot. Ensure the entire structure is cleared to avoid padding
203 4.9 KVM_SET_MEMORY_ALIAS
208 Parameters: struct kvm_memory_alias (in)
209 Returns: 0 (success), -1 (error)
211 This ioctl is obsolete and has been removed.
219 Returns: 0 on success, -1 on error
221 EINTR: an unmasked signal is pending
223 This ioctl is used to run a guest virtual cpu. While there are no
224 explicit parameters, there is an implicit parameter block that can be
225 obtained by mmap()ing the vcpu fd at offset 0, with the size given by
226 KVM_GET_VCPU_MMAP_SIZE. The parameter block is formatted as a 'struct
227 kvm_run' (see below).
234 Parameters: struct kvm_regs (out)
235 Returns: 0 on success, -1 on error
237 Reads the general purpose registers from the vcpu.
241 /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */
242 __u64 rax, rbx, rcx, rdx;
243 __u64 rsi, rdi, rsp, rbp;
244 __u64 r8, r9, r10, r11;
245 __u64 r12, r13, r14, r15;
254 Parameters: struct kvm_regs (in)
255 Returns: 0 on success, -1 on error
257 Writes the general purpose registers into the vcpu.
259 See KVM_GET_REGS for the data structure.
264 Architectures: x86, ppc
266 Parameters: struct kvm_sregs (out)
267 Returns: 0 on success, -1 on error
269 Reads special registers from the vcpu.
273 struct kvm_segment cs, ds, es, fs, gs, ss;
274 struct kvm_segment tr, ldt;
275 struct kvm_dtable gdt, idt;
276 __u64 cr0, cr2, cr3, cr4, cr8;
279 __u64 interrupt_bitmap[(KVM_NR_INTERRUPTS + 63) / 64];
282 /* ppc -- see arch/powerpc/include/asm/kvm.h */
284 interrupt_bitmap is a bitmap of pending external interrupts. At most
285 one bit may be set. This interrupt has been acknowledged by the APIC
286 but not yet injected into the cpu core.
291 Architectures: x86, ppc
293 Parameters: struct kvm_sregs (in)
294 Returns: 0 on success, -1 on error
296 Writes special registers into the vcpu. See KVM_GET_SREGS for the
304 Parameters: struct kvm_translation (in/out)
305 Returns: 0 on success, -1 on error
307 Translates a virtual address according to the vcpu's current address
310 struct kvm_translation {
312 __u64 linear_address;
315 __u64 physical_address;
325 Architectures: x86, ppc
327 Parameters: struct kvm_interrupt (in)
328 Returns: 0 on success, -1 on error
330 Queues a hardware interrupt vector to be injected. This is only
331 useful if in-kernel local APIC or equivalent is not used.
333 /* for KVM_INTERRUPT */
334 struct kvm_interrupt {
341 Note 'irq' is an interrupt vector, not an interrupt pin or line.
345 Queues an external interrupt to be injected. This ioctl is overleaded
346 with 3 different irq values:
350 This injects an edge type external interrupt into the guest once it's ready
351 to receive interrupts. When injected, the interrupt is done.
353 b) KVM_INTERRUPT_UNSET
355 This unsets any pending interrupt.
357 Only available with KVM_CAP_PPC_UNSET_IRQ.
359 c) KVM_INTERRUPT_SET_LEVEL
361 This injects a level type external interrupt into the guest context. The
362 interrupt stays pending until a specific ioctl with KVM_INTERRUPT_UNSET
365 Only available with KVM_CAP_PPC_IRQ_LEVEL.
367 Note that any value for 'irq' other than the ones stated above is invalid
368 and incurs unexpected behavior.
378 Support for this has been removed. Use KVM_SET_GUEST_DEBUG instead.
385 Parameters: struct kvm_msrs (in/out)
386 Returns: 0 on success, -1 on error
388 Reads model-specific registers from the vcpu. Supported msr indices can
389 be obtained using KVM_GET_MSR_INDEX_LIST.
392 __u32 nmsrs; /* number of msrs in entries */
395 struct kvm_msr_entry entries[0];
398 struct kvm_msr_entry {
404 Application code should set the 'nmsrs' member (which indicates the
405 size of the entries array) and the 'index' member of each array entry.
406 kvm will fill in the 'data' member.
413 Parameters: struct kvm_msrs (in)
414 Returns: 0 on success, -1 on error
416 Writes model-specific registers to the vcpu. See KVM_GET_MSRS for the
419 Application code should set the 'nmsrs' member (which indicates the
420 size of the entries array), and the 'index' and 'data' members of each
428 Parameters: struct kvm_cpuid (in)
429 Returns: 0 on success, -1 on error
431 Defines the vcpu responses to the cpuid instruction. Applications
432 should use the KVM_SET_CPUID2 ioctl if available.
435 struct kvm_cpuid_entry {
444 /* for KVM_SET_CPUID */
448 struct kvm_cpuid_entry entries[0];
451 4.21 KVM_SET_SIGNAL_MASK
456 Parameters: struct kvm_signal_mask (in)
457 Returns: 0 on success, -1 on error
459 Defines which signals are blocked during execution of KVM_RUN. This
460 signal mask temporarily overrides the threads signal mask. Any
461 unblocked signal received (except SIGKILL and SIGSTOP, which retain
462 their traditional behaviour) will cause KVM_RUN to return with -EINTR.
464 Note the signal will only be delivered if not blocked by the original
467 /* for KVM_SET_SIGNAL_MASK */
468 struct kvm_signal_mask {
478 Parameters: struct kvm_fpu (out)
479 Returns: 0 on success, -1 on error
481 Reads the floating point state from the vcpu.
483 /* for KVM_GET_FPU and KVM_SET_FPU */
488 __u8 ftwx; /* in fxsave format */
503 Parameters: struct kvm_fpu (in)
504 Returns: 0 on success, -1 on error
506 Writes the floating point state to the vcpu.
508 /* for KVM_GET_FPU and KVM_SET_FPU */
513 __u8 ftwx; /* in fxsave format */
523 4.24 KVM_CREATE_IRQCHIP
525 Capability: KVM_CAP_IRQCHIP
526 Architectures: x86, ia64
529 Returns: 0 on success, -1 on error
531 Creates an interrupt controller model in the kernel. On x86, creates a virtual
532 ioapic, a virtual PIC (two PICs, nested), and sets up future vcpus to have a
533 local APIC. IRQ routing for GSIs 0-15 is set to both PIC and IOAPIC; GSI 16-23
534 only go to the IOAPIC. On ia64, a IOSAPIC is created.
538 Capability: KVM_CAP_IRQCHIP
539 Architectures: x86, ia64
541 Parameters: struct kvm_irq_level
542 Returns: 0 on success, -1 on error
544 Sets the level of a GSI input to the interrupt controller model in the kernel.
545 Requires that an interrupt controller model has been previously created with
546 KVM_CREATE_IRQCHIP. Note that edge-triggered interrupts require the level
547 to be set to 1 and then back to 0.
549 struct kvm_irq_level {
552 __s32 status; /* not used for KVM_IRQ_LEVEL */
554 __u32 level; /* 0 or 1 */
559 Capability: KVM_CAP_IRQCHIP
560 Architectures: x86, ia64
562 Parameters: struct kvm_irqchip (in/out)
563 Returns: 0 on success, -1 on error
565 Reads the state of a kernel interrupt controller created with
566 KVM_CREATE_IRQCHIP into a buffer provided by the caller.
569 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */
572 char dummy[512]; /* reserving space */
573 struct kvm_pic_state pic;
574 struct kvm_ioapic_state ioapic;
580 Capability: KVM_CAP_IRQCHIP
581 Architectures: x86, ia64
583 Parameters: struct kvm_irqchip (in)
584 Returns: 0 on success, -1 on error
586 Sets the state of a kernel interrupt controller created with
587 KVM_CREATE_IRQCHIP from a buffer provided by the caller.
590 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */
593 char dummy[512]; /* reserving space */
594 struct kvm_pic_state pic;
595 struct kvm_ioapic_state ioapic;
599 4.28 KVM_XEN_HVM_CONFIG
601 Capability: KVM_CAP_XEN_HVM
604 Parameters: struct kvm_xen_hvm_config (in)
605 Returns: 0 on success, -1 on error
607 Sets the MSR that the Xen HVM guest uses to initialize its hypercall
608 page, and provides the starting address and size of the hypercall
609 blobs in userspace. When the guest writes the MSR, kvm copies one
610 page of a blob (32- or 64-bit, depending on the vcpu mode) to guest
613 struct kvm_xen_hvm_config {
625 Capability: KVM_CAP_ADJUST_CLOCK
628 Parameters: struct kvm_clock_data (out)
629 Returns: 0 on success, -1 on error
631 Gets the current timestamp of kvmclock as seen by the current guest. In
632 conjunction with KVM_SET_CLOCK, it is used to ensure monotonicity on scenarios
635 struct kvm_clock_data {
636 __u64 clock; /* kvmclock current value */
643 Capability: KVM_CAP_ADJUST_CLOCK
646 Parameters: struct kvm_clock_data (in)
647 Returns: 0 on success, -1 on error
649 Sets the current timestamp of kvmclock to the value specified in its parameter.
650 In conjunction with KVM_GET_CLOCK, it is used to ensure monotonicity on scenarios
653 struct kvm_clock_data {
654 __u64 clock; /* kvmclock current value */
659 4.31 KVM_GET_VCPU_EVENTS
661 Capability: KVM_CAP_VCPU_EVENTS
662 Extended by: KVM_CAP_INTR_SHADOW
665 Parameters: struct kvm_vcpu_event (out)
666 Returns: 0 on success, -1 on error
668 Gets currently pending exceptions, interrupts, and NMIs as well as related
671 struct kvm_vcpu_events {
695 KVM_VCPUEVENT_VALID_SHADOW may be set in the flags field to signal that
696 interrupt.shadow contains a valid state. Otherwise, this field is undefined.
698 4.32 KVM_SET_VCPU_EVENTS
700 Capability: KVM_CAP_VCPU_EVENTS
701 Extended by: KVM_CAP_INTR_SHADOW
704 Parameters: struct kvm_vcpu_event (in)
705 Returns: 0 on success, -1 on error
707 Set pending exceptions, interrupts, and NMIs as well as related states of the
710 See KVM_GET_VCPU_EVENTS for the data structure.
712 Fields that may be modified asynchronously by running VCPUs can be excluded
713 from the update. These fields are nmi.pending and sipi_vector. Keep the
714 corresponding bits in the flags field cleared to suppress overwriting the
715 current in-kernel state. The bits are:
717 KVM_VCPUEVENT_VALID_NMI_PENDING - transfer nmi.pending to the kernel
718 KVM_VCPUEVENT_VALID_SIPI_VECTOR - transfer sipi_vector
720 If KVM_CAP_INTR_SHADOW is available, KVM_VCPUEVENT_VALID_SHADOW can be set in
721 the flags field to signal that interrupt.shadow contains a valid state and
722 shall be written into the VCPU.
724 4.33 KVM_GET_DEBUGREGS
726 Capability: KVM_CAP_DEBUGREGS
729 Parameters: struct kvm_debugregs (out)
730 Returns: 0 on success, -1 on error
732 Reads debug registers from the vcpu.
734 struct kvm_debugregs {
742 4.34 KVM_SET_DEBUGREGS
744 Capability: KVM_CAP_DEBUGREGS
747 Parameters: struct kvm_debugregs (in)
748 Returns: 0 on success, -1 on error
750 Writes debug registers into the vcpu.
752 See KVM_GET_DEBUGREGS for the data structure. The flags field is unused
753 yet and must be cleared on entry.
755 4.35 KVM_SET_USER_MEMORY_REGION
757 Capability: KVM_CAP_USER_MEM
760 Parameters: struct kvm_userspace_memory_region (in)
761 Returns: 0 on success, -1 on error
763 struct kvm_userspace_memory_region {
766 __u64 guest_phys_addr;
767 __u64 memory_size; /* bytes */
768 __u64 userspace_addr; /* start of the userspace allocated memory */
771 /* for kvm_memory_region::flags */
772 #define KVM_MEM_LOG_DIRTY_PAGES 1UL
774 This ioctl allows the user to create or modify a guest physical memory
775 slot. When changing an existing slot, it may be moved in the guest
776 physical memory space, or its flags may be modified. It may not be
777 resized. Slots may not overlap in guest physical address space.
779 Memory for the region is taken starting at the address denoted by the
780 field userspace_addr, which must point at user addressable memory for
781 the entire memory slot size. Any object may back this memory, including
782 anonymous memory, ordinary files, and hugetlbfs.
784 It is recommended that the lower 21 bits of guest_phys_addr and userspace_addr
785 be identical. This allows large pages in the guest to be backed by large
788 The flags field supports just one flag, KVM_MEM_LOG_DIRTY_PAGES, which
789 instructs kvm to keep track of writes to memory within the slot. See
790 the KVM_GET_DIRTY_LOG ioctl.
792 When the KVM_CAP_SYNC_MMU capability, changes in the backing of the memory
793 region are automatically reflected into the guest. For example, an mmap()
794 that affects the region will be made visible immediately. Another example
795 is madvise(MADV_DROP).
797 It is recommended to use this API instead of the KVM_SET_MEMORY_REGION ioctl.
798 The KVM_SET_MEMORY_REGION does not allow fine grained control over memory
799 allocation and is deprecated.
801 4.36 KVM_SET_TSS_ADDR
803 Capability: KVM_CAP_SET_TSS_ADDR
806 Parameters: unsigned long tss_address (in)
807 Returns: 0 on success, -1 on error
809 This ioctl defines the physical address of a three-page region in the guest
810 physical address space. The region must be within the first 4GB of the
811 guest physical address space and must not conflict with any memory slot
812 or any mmio address. The guest may malfunction if it accesses this memory
815 This ioctl is required on Intel-based hosts. This is needed on Intel hardware
816 because of a quirk in the virtualization implementation (see the internals
817 documentation when it pops into existence).
821 Capability: KVM_CAP_ENABLE_CAP
824 Parameters: struct kvm_enable_cap (in)
825 Returns: 0 on success; -1 on error
827 +Not all extensions are enabled by default. Using this ioctl the application
828 can enable an extension, making it available to the guest.
830 On systems that do not support this ioctl, it always fails. On systems that
831 do support it, it only works for extensions that are supported for enablement.
833 To check if a capability can be enabled, the KVM_CHECK_EXTENSION ioctl should
836 struct kvm_enable_cap {
840 The capability that is supposed to get enabled.
844 A bitfield indicating future enhancements. Has to be 0 for now.
848 Arguments for enabling a feature. If a feature needs initial values to
849 function properly, this is the place to put them.
854 4.38 KVM_GET_MP_STATE
856 Capability: KVM_CAP_MP_STATE
857 Architectures: x86, ia64
859 Parameters: struct kvm_mp_state (out)
860 Returns: 0 on success; -1 on error
862 struct kvm_mp_state {
866 Returns the vcpu's current "multiprocessing state" (though also valid on
867 uniprocessor guests).
871 - KVM_MP_STATE_RUNNABLE: the vcpu is currently running
872 - KVM_MP_STATE_UNINITIALIZED: the vcpu is an application processor (AP)
873 which has not yet received an INIT signal
874 - KVM_MP_STATE_INIT_RECEIVED: the vcpu has received an INIT signal, and is
876 - KVM_MP_STATE_HALTED: the vcpu has executed a HLT instruction and
877 is waiting for an interrupt
878 - KVM_MP_STATE_SIPI_RECEIVED: the vcpu has just received a SIPI (vector
879 accessible via KVM_GET_VCPU_EVENTS)
881 This ioctl is only useful after KVM_CREATE_IRQCHIP. Without an in-kernel
882 irqchip, the multiprocessing state must be maintained by userspace.
884 4.39 KVM_SET_MP_STATE
886 Capability: KVM_CAP_MP_STATE
887 Architectures: x86, ia64
889 Parameters: struct kvm_mp_state (in)
890 Returns: 0 on success; -1 on error
892 Sets the vcpu's current "multiprocessing state"; see KVM_GET_MP_STATE for
895 This ioctl is only useful after KVM_CREATE_IRQCHIP. Without an in-kernel
896 irqchip, the multiprocessing state must be maintained by userspace.
898 4.40 KVM_SET_IDENTITY_MAP_ADDR
900 Capability: KVM_CAP_SET_IDENTITY_MAP_ADDR
903 Parameters: unsigned long identity (in)
904 Returns: 0 on success, -1 on error
906 This ioctl defines the physical address of a one-page region in the guest
907 physical address space. The region must be within the first 4GB of the
908 guest physical address space and must not conflict with any memory slot
909 or any mmio address. The guest may malfunction if it accesses this memory
912 This ioctl is required on Intel-based hosts. This is needed on Intel hardware
913 because of a quirk in the virtualization implementation (see the internals
914 documentation when it pops into existence).
916 4.41 KVM_SET_BOOT_CPU_ID
918 Capability: KVM_CAP_SET_BOOT_CPU_ID
919 Architectures: x86, ia64
921 Parameters: unsigned long vcpu_id
922 Returns: 0 on success, -1 on error
924 Define which vcpu is the Bootstrap Processor (BSP). Values are the same
925 as the vcpu id in KVM_CREATE_VCPU. If this ioctl is not called, the default
930 Capability: KVM_CAP_XSAVE
933 Parameters: struct kvm_xsave (out)
934 Returns: 0 on success, -1 on error
940 This ioctl would copy current vcpu's xsave struct to the userspace.
944 Capability: KVM_CAP_XSAVE
947 Parameters: struct kvm_xsave (in)
948 Returns: 0 on success, -1 on error
954 This ioctl would copy userspace's xsave struct to the kernel.
958 Capability: KVM_CAP_XCRS
961 Parameters: struct kvm_xcrs (out)
962 Returns: 0 on success, -1 on error
973 struct kvm_xcr xcrs[KVM_MAX_XCRS];
977 This ioctl would copy current vcpu's xcrs to the userspace.
981 Capability: KVM_CAP_XCRS
984 Parameters: struct kvm_xcrs (in)
985 Returns: 0 on success, -1 on error
996 struct kvm_xcr xcrs[KVM_MAX_XCRS];
1000 This ioctl would set vcpu's xcr to the value userspace specified.
1002 4.46 KVM_GET_SUPPORTED_CPUID
1004 Capability: KVM_CAP_EXT_CPUID
1007 Parameters: struct kvm_cpuid2 (in/out)
1008 Returns: 0 on success, -1 on error
1013 struct kvm_cpuid_entry2 entries[0];
1016 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX 1
1017 #define KVM_CPUID_FLAG_STATEFUL_FUNC 2
1018 #define KVM_CPUID_FLAG_STATE_READ_NEXT 4
1020 struct kvm_cpuid_entry2 {
1031 This ioctl returns x86 cpuid features which are supported by both the hardware
1032 and kvm. Userspace can use the information returned by this ioctl to
1033 construct cpuid information (for KVM_SET_CPUID2) that is consistent with
1034 hardware, kernel, and userspace capabilities, and with user requirements (for
1035 example, the user may wish to constrain cpuid to emulate older hardware,
1036 or for feature consistency across a cluster).
1038 Userspace invokes KVM_GET_SUPPORTED_CPUID by passing a kvm_cpuid2 structure
1039 with the 'nent' field indicating the number of entries in the variable-size
1040 array 'entries'. If the number of entries is too low to describe the cpu
1041 capabilities, an error (E2BIG) is returned. If the number is too high,
1042 the 'nent' field is adjusted and an error (ENOMEM) is returned. If the
1043 number is just right, the 'nent' field is adjusted to the number of valid
1044 entries in the 'entries' array, which is then filled.
1046 The entries returned are the host cpuid as returned by the cpuid instruction,
1047 with unknown or unsupported features masked out. Some features (for example,
1048 x2apic), may not be present in the host cpu, but are exposed by kvm if it can
1049 emulate them efficiently. The fields in each entry are defined as follows:
1051 function: the eax value used to obtain the entry
1052 index: the ecx value used to obtain the entry (for entries that are
1054 flags: an OR of zero or more of the following:
1055 KVM_CPUID_FLAG_SIGNIFCANT_INDEX:
1056 if the index field is valid
1057 KVM_CPUID_FLAG_STATEFUL_FUNC:
1058 if cpuid for this function returns different values for successive
1059 invocations; there will be several entries with the same function,
1060 all with this flag set
1061 KVM_CPUID_FLAG_STATE_READ_NEXT:
1062 for KVM_CPUID_FLAG_STATEFUL_FUNC entries, set if this entry is
1063 the first entry to be read by a cpu
1064 eax, ebx, ecx, edx: the values returned by the cpuid instruction for
1065 this function/index combination
1067 4.47 KVM_PPC_GET_PVINFO
1069 Capability: KVM_CAP_PPC_GET_PVINFO
1072 Parameters: struct kvm_ppc_pvinfo (out)
1073 Returns: 0 on success, !0 on error
1075 struct kvm_ppc_pvinfo {
1081 This ioctl fetches PV specific information that need to be passed to the guest
1082 using the device tree or other means from vm context.
1084 For now the only implemented piece of information distributed here is an array
1085 of 4 instructions that make up a hypercall.
1087 If any additional field gets added to this structure later on, a bit for that
1088 additional piece of information will be set in the flags bitmap.
1090 4.48 KVM_ASSIGN_PCI_DEVICE
1092 Capability: KVM_CAP_DEVICE_ASSIGNMENT
1093 Architectures: x86 ia64
1095 Parameters: struct kvm_assigned_pci_dev (in)
1096 Returns: 0 on success, -1 on error
1098 Assigns a host PCI device to the VM.
1100 struct kvm_assigned_pci_dev {
1101 __u32 assigned_dev_id;
1111 The PCI device is specified by the triple segnr, busnr, and devfn.
1112 Identification in succeeding service requests is done via assigned_dev_id. The
1113 following flags are specified:
1115 /* Depends on KVM_CAP_IOMMU */
1116 #define KVM_DEV_ASSIGN_ENABLE_IOMMU (1 << 0)
1118 4.49 KVM_DEASSIGN_PCI_DEVICE
1120 Capability: KVM_CAP_DEVICE_DEASSIGNMENT
1121 Architectures: x86 ia64
1123 Parameters: struct kvm_assigned_pci_dev (in)
1124 Returns: 0 on success, -1 on error
1126 Ends PCI device assignment, releasing all associated resources.
1128 See KVM_CAP_DEVICE_ASSIGNMENT for the data structure. Only assigned_dev_id is
1129 used in kvm_assigned_pci_dev to identify the device.
1131 4.50 KVM_ASSIGN_DEV_IRQ
1133 Capability: KVM_CAP_ASSIGN_DEV_IRQ
1134 Architectures: x86 ia64
1136 Parameters: struct kvm_assigned_irq (in)
1137 Returns: 0 on success, -1 on error
1139 Assigns an IRQ to a passed-through device.
1141 struct kvm_assigned_irq {
1142 __u32 assigned_dev_id;
1156 The following flags are defined:
1158 #define KVM_DEV_IRQ_HOST_INTX (1 << 0)
1159 #define KVM_DEV_IRQ_HOST_MSI (1 << 1)
1160 #define KVM_DEV_IRQ_HOST_MSIX (1 << 2)
1162 #define KVM_DEV_IRQ_GUEST_INTX (1 << 8)
1163 #define KVM_DEV_IRQ_GUEST_MSI (1 << 9)
1164 #define KVM_DEV_IRQ_GUEST_MSIX (1 << 10)
1166 It is not valid to specify multiple types per host or guest IRQ. However, the
1167 IRQ type of host and guest can differ or can even be null.
1169 4.51 KVM_DEASSIGN_DEV_IRQ
1171 Capability: KVM_CAP_ASSIGN_DEV_IRQ
1172 Architectures: x86 ia64
1174 Parameters: struct kvm_assigned_irq (in)
1175 Returns: 0 on success, -1 on error
1177 Ends an IRQ assignment to a passed-through device.
1179 See KVM_ASSIGN_DEV_IRQ for the data structure. The target device is specified
1180 by assigned_dev_id, flags must correspond to the IRQ type specified on
1181 KVM_ASSIGN_DEV_IRQ. Partial deassignment of host or guest IRQ is allowed.
1183 4.52 KVM_SET_GSI_ROUTING
1185 Capability: KVM_CAP_IRQ_ROUTING
1186 Architectures: x86 ia64
1188 Parameters: struct kvm_irq_routing (in)
1189 Returns: 0 on success, -1 on error
1191 Sets the GSI routing table entries, overwriting any previously set entries.
1193 struct kvm_irq_routing {
1196 struct kvm_irq_routing_entry entries[0];
1199 No flags are specified so far, the corresponding field must be set to zero.
1201 struct kvm_irq_routing_entry {
1207 struct kvm_irq_routing_irqchip irqchip;
1208 struct kvm_irq_routing_msi msi;
1213 /* gsi routing entry types */
1214 #define KVM_IRQ_ROUTING_IRQCHIP 1
1215 #define KVM_IRQ_ROUTING_MSI 2
1217 No flags are specified so far, the corresponding field must be set to zero.
1219 struct kvm_irq_routing_irqchip {
1224 struct kvm_irq_routing_msi {
1231 4.53 KVM_ASSIGN_SET_MSIX_NR
1233 Capability: KVM_CAP_DEVICE_MSIX
1234 Architectures: x86 ia64
1236 Parameters: struct kvm_assigned_msix_nr (in)
1237 Returns: 0 on success, -1 on error
1239 Set the number of MSI-X interrupts for an assigned device. This service can
1240 only be called once in the lifetime of an assigned device.
1242 struct kvm_assigned_msix_nr {
1243 __u32 assigned_dev_id;
1248 #define KVM_MAX_MSIX_PER_DEV 256
1250 4.54 KVM_ASSIGN_SET_MSIX_ENTRY
1252 Capability: KVM_CAP_DEVICE_MSIX
1253 Architectures: x86 ia64
1255 Parameters: struct kvm_assigned_msix_entry (in)
1256 Returns: 0 on success, -1 on error
1258 Specifies the routing of an MSI-X assigned device interrupt to a GSI. Setting
1259 the GSI vector to zero means disabling the interrupt.
1261 struct kvm_assigned_msix_entry {
1262 __u32 assigned_dev_id;
1264 __u16 entry; /* The index of entry in the MSI-X table */
1268 4.54 KVM_SET_TSC_KHZ
1270 Capability: KVM_CAP_TSC_CONTROL
1273 Parameters: virtual tsc_khz
1274 Returns: 0 on success, -1 on error
1276 Specifies the tsc frequency for the virtual machine. The unit of the
1279 4.55 KVM_GET_TSC_KHZ
1281 Capability: KVM_CAP_GET_TSC_KHZ
1285 Returns: virtual tsc-khz on success, negative value on error
1287 Returns the tsc frequency of the guest. The unit of the return value is
1288 KHz. If the host has unstable tsc this ioctl returns -EIO instead as an
1291 5. The kvm_run structure
1293 Application code obtains a pointer to the kvm_run structure by
1294 mmap()ing a vcpu fd. From that point, application code can control
1295 execution by changing fields in kvm_run prior to calling the KVM_RUN
1296 ioctl, and obtain information about the reason KVM_RUN returned by
1297 looking up structure members.
1301 __u8 request_interrupt_window;
1303 Request that KVM_RUN return when it becomes possible to inject external
1304 interrupts into the guest. Useful in conjunction with KVM_INTERRUPT.
1311 When KVM_RUN has returned successfully (return value 0), this informs
1312 application code why KVM_RUN has returned. Allowable values for this
1313 field are detailed below.
1315 __u8 ready_for_interrupt_injection;
1317 If request_interrupt_window has been specified, this field indicates
1318 an interrupt can be injected now with KVM_INTERRUPT.
1322 The value of the current interrupt flag. Only valid if in-kernel
1323 local APIC is not used.
1327 /* in (pre_kvm_run), out (post_kvm_run) */
1330 The value of the cr8 register. Only valid if in-kernel local APIC is
1331 not used. Both input and output.
1335 The value of the APIC BASE msr. Only valid if in-kernel local
1336 APIC is not used. Both input and output.
1339 /* KVM_EXIT_UNKNOWN */
1341 __u64 hardware_exit_reason;
1344 If exit_reason is KVM_EXIT_UNKNOWN, the vcpu has exited due to unknown
1345 reasons. Further architecture-specific information is available in
1346 hardware_exit_reason.
1348 /* KVM_EXIT_FAIL_ENTRY */
1350 __u64 hardware_entry_failure_reason;
1353 If exit_reason is KVM_EXIT_FAIL_ENTRY, the vcpu could not be run due
1354 to unknown reasons. Further architecture-specific information is
1355 available in hardware_entry_failure_reason.
1357 /* KVM_EXIT_EXCEPTION */
1367 #define KVM_EXIT_IO_IN 0
1368 #define KVM_EXIT_IO_OUT 1
1370 __u8 size; /* bytes */
1373 __u64 data_offset; /* relative to kvm_run start */
1376 If exit_reason is KVM_EXIT_IO, then the vcpu has
1377 executed a port I/O instruction which could not be satisfied by kvm.
1378 data_offset describes where the data is located (KVM_EXIT_IO_OUT) or
1379 where kvm expects application code to place the data for the next
1380 KVM_RUN invocation (KVM_EXIT_IO_IN). Data format is a packed array.
1383 struct kvm_debug_exit_arch arch;
1396 If exit_reason is KVM_EXIT_MMIO, then the vcpu has
1397 executed a memory-mapped I/O instruction which could not be satisfied
1398 by kvm. The 'data' member contains the written data if 'is_write' is
1399 true, and should be filled by application code otherwise.
1401 NOTE: For KVM_EXIT_IO, KVM_EXIT_MMIO and KVM_EXIT_OSI, the corresponding
1402 operations are complete (and guest state is consistent) only after userspace
1403 has re-entered the kernel with KVM_RUN. The kernel side will first finish
1404 incomplete operations and then check for pending signals. Userspace
1405 can re-enter the guest with an unmasked signal pending to complete
1408 /* KVM_EXIT_HYPERCALL */
1417 Unused. This was once used for 'hypercall to userspace'. To implement
1418 such functionality, use KVM_EXIT_IO (x86) or KVM_EXIT_MMIO (all except s390).
1419 Note KVM_EXIT_IO is significantly faster than KVM_EXIT_MMIO.
1421 /* KVM_EXIT_TPR_ACCESS */
1428 To be documented (KVM_TPR_ACCESS_REPORTING).
1430 /* KVM_EXIT_S390_SIEIC */
1433 __u64 mask; /* psw upper half */
1434 __u64 addr; /* psw lower half */
1441 /* KVM_EXIT_S390_RESET */
1442 #define KVM_S390_RESET_POR 1
1443 #define KVM_S390_RESET_CLEAR 2
1444 #define KVM_S390_RESET_SUBSYSTEM 4
1445 #define KVM_S390_RESET_CPU_INIT 8
1446 #define KVM_S390_RESET_IPL 16
1447 __u64 s390_reset_flags;
1465 MOL uses a special hypercall interface it calls 'OSI'. To enable it, we catch
1466 hypercalls and exit with this exit struct that contains all the guest gprs.
1468 If exit_reason is KVM_EXIT_OSI, then the vcpu has triggered such a hypercall.
1469 Userspace can now handle the hypercall and when it's done modify the gprs as
1470 necessary. Upon guest entry all guest GPRs will then be replaced by the values
1473 /* Fix the size of the union. */