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), or a KVM_CAP_xyz constant, which
72 means availability needs to be checked with KVM_CHECK_EXTENSION
75 Architectures: which instruction set architectures provide this ioctl.
76 x86 includes both i386 and x86_64.
78 Type: system, vm, or vcpu.
80 Parameters: what parameters are accepted by the ioctl.
82 Returns: the return value. General error numbers (EBADF, ENOMEM, EINVAL)
83 are not detailed, but errors with specific meanings are.
86 4.1 KVM_GET_API_VERSION
92 Returns: the constant KVM_API_VERSION (=12)
94 This identifies the API version as the stable kvm API. It is not
95 expected that this number will change. However, Linux 2.6.20 and
96 2.6.21 report earlier versions; these are not documented and not
97 supported. Applications should refuse to run if KVM_GET_API_VERSION
98 returns a value other than 12. If this check passes, all ioctls
99 described as 'basic' will be available.
107 Parameters: machine type identifier (KVM_VM_*)
108 Returns: a VM fd that can be used to control the new virtual machine.
110 The new VM has no virtual cpus and no memory. An mmap() of a VM fd
111 will access the virtual machine's physical address space; offset zero
112 corresponds to guest physical address zero. Use of mmap() on a VM fd
113 is discouraged if userspace memory allocation (KVM_CAP_USER_MEMORY) is
115 You most certainly want to use 0 as machine type.
117 In order to create user controlled virtual machines on S390, check
118 KVM_CAP_S390_UCONTROL and use the flag KVM_VM_S390_UCONTROL as
119 privileged user (CAP_SYS_ADMIN).
122 4.3 KVM_GET_MSR_INDEX_LIST
127 Parameters: struct kvm_msr_list (in/out)
128 Returns: 0 on success; -1 on error
130 E2BIG: the msr index list is to be to fit in the array specified by
133 struct kvm_msr_list {
134 __u32 nmsrs; /* number of msrs in entries */
138 This ioctl returns the guest msrs that are supported. The list varies
139 by kvm version and host processor, but does not change otherwise. The
140 user fills in the size of the indices array in nmsrs, and in return
141 kvm adjusts nmsrs to reflect the actual number of msrs and fills in
142 the indices array with their numbers.
144 Note: if kvm indicates supports MCE (KVM_CAP_MCE), then the MCE bank MSRs are
145 not returned in the MSR list, as different vcpus can have a different number
146 of banks, as set via the KVM_X86_SETUP_MCE ioctl.
149 4.4 KVM_CHECK_EXTENSION
154 Parameters: extension identifier (KVM_CAP_*)
155 Returns: 0 if unsupported; 1 (or some other positive integer) if supported
157 The API allows the application to query about extensions to the core
158 kvm API. Userspace passes an extension identifier (an integer) and
159 receives an integer that describes the extension availability.
160 Generally 0 means no and 1 means yes, but some extensions may report
161 additional information in the integer return value.
164 4.5 KVM_GET_VCPU_MMAP_SIZE
170 Returns: size of vcpu mmap area, in bytes
172 The KVM_RUN ioctl (cf.) communicates with userspace via a shared
173 memory region. This ioctl returns the size of that region. See the
174 KVM_RUN documentation for details.
177 4.6 KVM_SET_MEMORY_REGION
182 Parameters: struct kvm_memory_region (in)
183 Returns: 0 on success, -1 on error
185 This ioctl is obsolete and has been removed.
193 Parameters: vcpu id (apic id on x86)
194 Returns: vcpu fd on success, -1 on error
196 This API adds a vcpu to a virtual machine. The vcpu id is a small integer
197 in the range [0, max_vcpus).
199 The recommended max_vcpus value can be retrieved using the KVM_CAP_NR_VCPUS of
200 the KVM_CHECK_EXTENSION ioctl() at run-time.
201 The maximum possible value for max_vcpus can be retrieved using the
202 KVM_CAP_MAX_VCPUS of the KVM_CHECK_EXTENSION ioctl() at run-time.
204 If the KVM_CAP_NR_VCPUS does not exist, you should assume that max_vcpus is 4
206 If the KVM_CAP_MAX_VCPUS does not exist, you should assume that max_vcpus is
207 same as the value returned from KVM_CAP_NR_VCPUS.
209 On powerpc using book3s_hv mode, the vcpus are mapped onto virtual
210 threads in one or more virtual CPU cores. (This is because the
211 hardware requires all the hardware threads in a CPU core to be in the
212 same partition.) The KVM_CAP_PPC_SMT capability indicates the number
213 of vcpus per virtual core (vcore). The vcore id is obtained by
214 dividing the vcpu id by the number of vcpus per vcore. The vcpus in a
215 given vcore will always be in the same physical core as each other
216 (though that might be a different physical core from time to time).
217 Userspace can control the threading (SMT) mode of the guest by its
218 allocation of vcpu ids. For example, if userspace wants
219 single-threaded guest vcpus, it should make all vcpu ids be a multiple
220 of the number of vcpus per vcore.
222 For virtual cpus that have been created with S390 user controlled virtual
223 machines, the resulting vcpu fd can be memory mapped at page offset
224 KVM_S390_SIE_PAGE_OFFSET in order to obtain a memory map of the virtual
225 cpu's hardware control block.
228 4.8 KVM_GET_DIRTY_LOG (vm ioctl)
233 Parameters: struct kvm_dirty_log (in/out)
234 Returns: 0 on success, -1 on error
236 /* for KVM_GET_DIRTY_LOG */
237 struct kvm_dirty_log {
241 void __user *dirty_bitmap; /* one bit per page */
246 Given a memory slot, return a bitmap containing any pages dirtied
247 since the last call to this ioctl. Bit 0 is the first page in the
248 memory slot. Ensure the entire structure is cleared to avoid padding
252 4.9 KVM_SET_MEMORY_ALIAS
257 Parameters: struct kvm_memory_alias (in)
258 Returns: 0 (success), -1 (error)
260 This ioctl is obsolete and has been removed.
269 Returns: 0 on success, -1 on error
271 EINTR: an unmasked signal is pending
273 This ioctl is used to run a guest virtual cpu. While there are no
274 explicit parameters, there is an implicit parameter block that can be
275 obtained by mmap()ing the vcpu fd at offset 0, with the size given by
276 KVM_GET_VCPU_MMAP_SIZE. The parameter block is formatted as a 'struct
277 kvm_run' (see below).
283 Architectures: all except ARM, arm64
285 Parameters: struct kvm_regs (out)
286 Returns: 0 on success, -1 on error
288 Reads the general purpose registers from the vcpu.
292 /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */
293 __u64 rax, rbx, rcx, rdx;
294 __u64 rsi, rdi, rsp, rbp;
295 __u64 r8, r9, r10, r11;
296 __u64 r12, r13, r14, r15;
304 Architectures: all except ARM, arm64
306 Parameters: struct kvm_regs (in)
307 Returns: 0 on success, -1 on error
309 Writes the general purpose registers into the vcpu.
311 See KVM_GET_REGS for the data structure.
317 Architectures: x86, ppc
319 Parameters: struct kvm_sregs (out)
320 Returns: 0 on success, -1 on error
322 Reads special registers from the vcpu.
326 struct kvm_segment cs, ds, es, fs, gs, ss;
327 struct kvm_segment tr, ldt;
328 struct kvm_dtable gdt, idt;
329 __u64 cr0, cr2, cr3, cr4, cr8;
332 __u64 interrupt_bitmap[(KVM_NR_INTERRUPTS + 63) / 64];
335 /* ppc -- see arch/powerpc/include/uapi/asm/kvm.h */
337 interrupt_bitmap is a bitmap of pending external interrupts. At most
338 one bit may be set. This interrupt has been acknowledged by the APIC
339 but not yet injected into the cpu core.
345 Architectures: x86, ppc
347 Parameters: struct kvm_sregs (in)
348 Returns: 0 on success, -1 on error
350 Writes special registers into the vcpu. See KVM_GET_SREGS for the
359 Parameters: struct kvm_translation (in/out)
360 Returns: 0 on success, -1 on error
362 Translates a virtual address according to the vcpu's current address
365 struct kvm_translation {
367 __u64 linear_address;
370 __u64 physical_address;
381 Architectures: x86, ppc
383 Parameters: struct kvm_interrupt (in)
384 Returns: 0 on success, -1 on error
386 Queues a hardware interrupt vector to be injected. This is only
387 useful if in-kernel local APIC or equivalent is not used.
389 /* for KVM_INTERRUPT */
390 struct kvm_interrupt {
397 Note 'irq' is an interrupt vector, not an interrupt pin or line.
401 Queues an external interrupt to be injected. This ioctl is overleaded
402 with 3 different irq values:
406 This injects an edge type external interrupt into the guest once it's ready
407 to receive interrupts. When injected, the interrupt is done.
409 b) KVM_INTERRUPT_UNSET
411 This unsets any pending interrupt.
413 Only available with KVM_CAP_PPC_UNSET_IRQ.
415 c) KVM_INTERRUPT_SET_LEVEL
417 This injects a level type external interrupt into the guest context. The
418 interrupt stays pending until a specific ioctl with KVM_INTERRUPT_UNSET
421 Only available with KVM_CAP_PPC_IRQ_LEVEL.
423 Note that any value for 'irq' other than the ones stated above is invalid
424 and incurs unexpected behavior.
435 Support for this has been removed. Use KVM_SET_GUEST_DEBUG instead.
443 Parameters: struct kvm_msrs (in/out)
444 Returns: 0 on success, -1 on error
446 Reads model-specific registers from the vcpu. Supported msr indices can
447 be obtained using KVM_GET_MSR_INDEX_LIST.
450 __u32 nmsrs; /* number of msrs in entries */
453 struct kvm_msr_entry entries[0];
456 struct kvm_msr_entry {
462 Application code should set the 'nmsrs' member (which indicates the
463 size of the entries array) and the 'index' member of each array entry.
464 kvm will fill in the 'data' member.
472 Parameters: struct kvm_msrs (in)
473 Returns: 0 on success, -1 on error
475 Writes model-specific registers to the vcpu. See KVM_GET_MSRS for the
478 Application code should set the 'nmsrs' member (which indicates the
479 size of the entries array), and the 'index' and 'data' members of each
488 Parameters: struct kvm_cpuid (in)
489 Returns: 0 on success, -1 on error
491 Defines the vcpu responses to the cpuid instruction. Applications
492 should use the KVM_SET_CPUID2 ioctl if available.
495 struct kvm_cpuid_entry {
504 /* for KVM_SET_CPUID */
508 struct kvm_cpuid_entry entries[0];
512 4.21 KVM_SET_SIGNAL_MASK
517 Parameters: struct kvm_signal_mask (in)
518 Returns: 0 on success, -1 on error
520 Defines which signals are blocked during execution of KVM_RUN. This
521 signal mask temporarily overrides the threads signal mask. Any
522 unblocked signal received (except SIGKILL and SIGSTOP, which retain
523 their traditional behaviour) will cause KVM_RUN to return with -EINTR.
525 Note the signal will only be delivered if not blocked by the original
528 /* for KVM_SET_SIGNAL_MASK */
529 struct kvm_signal_mask {
540 Parameters: struct kvm_fpu (out)
541 Returns: 0 on success, -1 on error
543 Reads the floating point state from the vcpu.
545 /* for KVM_GET_FPU and KVM_SET_FPU */
550 __u8 ftwx; /* in fxsave format */
566 Parameters: struct kvm_fpu (in)
567 Returns: 0 on success, -1 on error
569 Writes the floating point state to the vcpu.
571 /* for KVM_GET_FPU and KVM_SET_FPU */
576 __u8 ftwx; /* in fxsave format */
587 4.24 KVM_CREATE_IRQCHIP
589 Capability: KVM_CAP_IRQCHIP
590 Architectures: x86, ia64, ARM, arm64
593 Returns: 0 on success, -1 on error
595 Creates an interrupt controller model in the kernel. On x86, creates a virtual
596 ioapic, a virtual PIC (two PICs, nested), and sets up future vcpus to have a
597 local APIC. IRQ routing for GSIs 0-15 is set to both PIC and IOAPIC; GSI 16-23
598 only go to the IOAPIC. On ia64, a IOSAPIC is created. On ARM/arm64, a GIC is
604 Capability: KVM_CAP_IRQCHIP
605 Architectures: x86, ia64, arm, arm64
607 Parameters: struct kvm_irq_level
608 Returns: 0 on success, -1 on error
610 Sets the level of a GSI input to the interrupt controller model in the kernel.
611 On some architectures it is required that an interrupt controller model has
612 been previously created with KVM_CREATE_IRQCHIP. Note that edge-triggered
613 interrupts require the level to be set to 1 and then back to 0.
615 ARM/arm64 can signal an interrupt either at the CPU level, or at the
616 in-kernel irqchip (GIC), and for in-kernel irqchip can tell the GIC to
617 use PPIs designated for specific cpus. The irq field is interpreted
620 Â bits: | 31 ... 24 | 23 ... 16 | 15 ... 0 |
621 field: | irq_type | vcpu_index | irq_id |
623 The irq_type field has the following values:
624 - irq_type[0]: out-of-kernel GIC: irq_id 0 is IRQ, irq_id 1 is FIQ
625 - irq_type[1]: in-kernel GIC: SPI, irq_id between 32 and 1019 (incl.)
626 (the vcpu_index field is ignored)
627 - irq_type[2]: in-kernel GIC: PPI, irq_id between 16 and 31 (incl.)
629 (The irq_id field thus corresponds nicely to the IRQ ID in the ARM GIC specs)
631 In both cases, level is used to raise/lower the line.
633 struct kvm_irq_level {
636 __s32 status; /* not used for KVM_IRQ_LEVEL */
638 __u32 level; /* 0 or 1 */
644 Capability: KVM_CAP_IRQCHIP
645 Architectures: x86, ia64
647 Parameters: struct kvm_irqchip (in/out)
648 Returns: 0 on success, -1 on error
650 Reads the state of a kernel interrupt controller created with
651 KVM_CREATE_IRQCHIP into a buffer provided by the caller.
654 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */
657 char dummy[512]; /* reserving space */
658 struct kvm_pic_state pic;
659 struct kvm_ioapic_state ioapic;
666 Capability: KVM_CAP_IRQCHIP
667 Architectures: x86, ia64
669 Parameters: struct kvm_irqchip (in)
670 Returns: 0 on success, -1 on error
672 Sets the state of a kernel interrupt controller created with
673 KVM_CREATE_IRQCHIP from a buffer provided by the caller.
676 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */
679 char dummy[512]; /* reserving space */
680 struct kvm_pic_state pic;
681 struct kvm_ioapic_state ioapic;
686 4.28 KVM_XEN_HVM_CONFIG
688 Capability: KVM_CAP_XEN_HVM
691 Parameters: struct kvm_xen_hvm_config (in)
692 Returns: 0 on success, -1 on error
694 Sets the MSR that the Xen HVM guest uses to initialize its hypercall
695 page, and provides the starting address and size of the hypercall
696 blobs in userspace. When the guest writes the MSR, kvm copies one
697 page of a blob (32- or 64-bit, depending on the vcpu mode) to guest
700 struct kvm_xen_hvm_config {
713 Capability: KVM_CAP_ADJUST_CLOCK
716 Parameters: struct kvm_clock_data (out)
717 Returns: 0 on success, -1 on error
719 Gets the current timestamp of kvmclock as seen by the current guest. In
720 conjunction with KVM_SET_CLOCK, it is used to ensure monotonicity on scenarios
723 struct kvm_clock_data {
724 __u64 clock; /* kvmclock current value */
732 Capability: KVM_CAP_ADJUST_CLOCK
735 Parameters: struct kvm_clock_data (in)
736 Returns: 0 on success, -1 on error
738 Sets the current timestamp of kvmclock to the value specified in its parameter.
739 In conjunction with KVM_GET_CLOCK, it is used to ensure monotonicity on scenarios
742 struct kvm_clock_data {
743 __u64 clock; /* kvmclock current value */
749 4.31 KVM_GET_VCPU_EVENTS
751 Capability: KVM_CAP_VCPU_EVENTS
752 Extended by: KVM_CAP_INTR_SHADOW
755 Parameters: struct kvm_vcpu_event (out)
756 Returns: 0 on success, -1 on error
758 Gets currently pending exceptions, interrupts, and NMIs as well as related
761 struct kvm_vcpu_events {
785 KVM_VCPUEVENT_VALID_SHADOW may be set in the flags field to signal that
786 interrupt.shadow contains a valid state. Otherwise, this field is undefined.
789 4.32 KVM_SET_VCPU_EVENTS
791 Capability: KVM_CAP_VCPU_EVENTS
792 Extended by: KVM_CAP_INTR_SHADOW
795 Parameters: struct kvm_vcpu_event (in)
796 Returns: 0 on success, -1 on error
798 Set pending exceptions, interrupts, and NMIs as well as related states of the
801 See KVM_GET_VCPU_EVENTS for the data structure.
803 Fields that may be modified asynchronously by running VCPUs can be excluded
804 from the update. These fields are nmi.pending and sipi_vector. Keep the
805 corresponding bits in the flags field cleared to suppress overwriting the
806 current in-kernel state. The bits are:
808 KVM_VCPUEVENT_VALID_NMI_PENDING - transfer nmi.pending to the kernel
809 KVM_VCPUEVENT_VALID_SIPI_VECTOR - transfer sipi_vector
811 If KVM_CAP_INTR_SHADOW is available, KVM_VCPUEVENT_VALID_SHADOW can be set in
812 the flags field to signal that interrupt.shadow contains a valid state and
813 shall be written into the VCPU.
816 4.33 KVM_GET_DEBUGREGS
818 Capability: KVM_CAP_DEBUGREGS
821 Parameters: struct kvm_debugregs (out)
822 Returns: 0 on success, -1 on error
824 Reads debug registers from the vcpu.
826 struct kvm_debugregs {
835 4.34 KVM_SET_DEBUGREGS
837 Capability: KVM_CAP_DEBUGREGS
840 Parameters: struct kvm_debugregs (in)
841 Returns: 0 on success, -1 on error
843 Writes debug registers into the vcpu.
845 See KVM_GET_DEBUGREGS for the data structure. The flags field is unused
846 yet and must be cleared on entry.
849 4.35 KVM_SET_USER_MEMORY_REGION
851 Capability: KVM_CAP_USER_MEM
854 Parameters: struct kvm_userspace_memory_region (in)
855 Returns: 0 on success, -1 on error
857 struct kvm_userspace_memory_region {
860 __u64 guest_phys_addr;
861 __u64 memory_size; /* bytes */
862 __u64 userspace_addr; /* start of the userspace allocated memory */
865 /* for kvm_memory_region::flags */
866 #define KVM_MEM_LOG_DIRTY_PAGES (1UL << 0)
867 #define KVM_MEM_READONLY (1UL << 1)
869 This ioctl allows the user to create or modify a guest physical memory
870 slot. When changing an existing slot, it may be moved in the guest
871 physical memory space, or its flags may be modified. It may not be
872 resized. Slots may not overlap in guest physical address space.
874 Memory for the region is taken starting at the address denoted by the
875 field userspace_addr, which must point at user addressable memory for
876 the entire memory slot size. Any object may back this memory, including
877 anonymous memory, ordinary files, and hugetlbfs.
879 It is recommended that the lower 21 bits of guest_phys_addr and userspace_addr
880 be identical. This allows large pages in the guest to be backed by large
883 The flags field supports two flags: KVM_MEM_LOG_DIRTY_PAGES and
884 KVM_MEM_READONLY. The former can be set to instruct KVM to keep track of
885 writes to memory within the slot. See KVM_GET_DIRTY_LOG ioctl to know how to
886 use it. The latter can be set, if KVM_CAP_READONLY_MEM capability allows it,
887 to make a new slot read-only. In this case, writes to this memory will be
888 posted to userspace as KVM_EXIT_MMIO exits.
890 When the KVM_CAP_SYNC_MMU capability is available, changes in the backing of
891 the memory region are automatically reflected into the guest. For example, an
892 mmap() that affects the region will be made visible immediately. Another
893 example is madvise(MADV_DROP).
895 It is recommended to use this API instead of the KVM_SET_MEMORY_REGION ioctl.
896 The KVM_SET_MEMORY_REGION does not allow fine grained control over memory
897 allocation and is deprecated.
900 4.36 KVM_SET_TSS_ADDR
902 Capability: KVM_CAP_SET_TSS_ADDR
905 Parameters: unsigned long tss_address (in)
906 Returns: 0 on success, -1 on error
908 This ioctl defines the physical address of a three-page region in the guest
909 physical address space. The region must be within the first 4GB of the
910 guest physical address space and must not conflict with any memory slot
911 or any mmio address. The guest may malfunction if it accesses this memory
914 This ioctl is required on Intel-based hosts. This is needed on Intel hardware
915 because of a quirk in the virtualization implementation (see the internals
916 documentation when it pops into existence).
921 Capability: KVM_CAP_ENABLE_CAP
922 Architectures: ppc, s390
924 Parameters: struct kvm_enable_cap (in)
925 Returns: 0 on success; -1 on error
927 +Not all extensions are enabled by default. Using this ioctl the application
928 can enable an extension, making it available to the guest.
930 On systems that do not support this ioctl, it always fails. On systems that
931 do support it, it only works for extensions that are supported for enablement.
933 To check if a capability can be enabled, the KVM_CHECK_EXTENSION ioctl should
936 struct kvm_enable_cap {
940 The capability that is supposed to get enabled.
944 A bitfield indicating future enhancements. Has to be 0 for now.
948 Arguments for enabling a feature. If a feature needs initial values to
949 function properly, this is the place to put them.
955 4.38 KVM_GET_MP_STATE
957 Capability: KVM_CAP_MP_STATE
958 Architectures: x86, ia64
960 Parameters: struct kvm_mp_state (out)
961 Returns: 0 on success; -1 on error
963 struct kvm_mp_state {
967 Returns the vcpu's current "multiprocessing state" (though also valid on
968 uniprocessor guests).
972 - KVM_MP_STATE_RUNNABLE: the vcpu is currently running
973 - KVM_MP_STATE_UNINITIALIZED: the vcpu is an application processor (AP)
974 which has not yet received an INIT signal
975 - KVM_MP_STATE_INIT_RECEIVED: the vcpu has received an INIT signal, and is
977 - KVM_MP_STATE_HALTED: the vcpu has executed a HLT instruction and
978 is waiting for an interrupt
979 - KVM_MP_STATE_SIPI_RECEIVED: the vcpu has just received a SIPI (vector
980 accessible via KVM_GET_VCPU_EVENTS)
982 This ioctl is only useful after KVM_CREATE_IRQCHIP. Without an in-kernel
983 irqchip, the multiprocessing state must be maintained by userspace.
986 4.39 KVM_SET_MP_STATE
988 Capability: KVM_CAP_MP_STATE
989 Architectures: x86, ia64
991 Parameters: struct kvm_mp_state (in)
992 Returns: 0 on success; -1 on error
994 Sets the vcpu's current "multiprocessing state"; see KVM_GET_MP_STATE for
997 This ioctl is only useful after KVM_CREATE_IRQCHIP. Without an in-kernel
998 irqchip, the multiprocessing state must be maintained by userspace.
1001 4.40 KVM_SET_IDENTITY_MAP_ADDR
1003 Capability: KVM_CAP_SET_IDENTITY_MAP_ADDR
1006 Parameters: unsigned long identity (in)
1007 Returns: 0 on success, -1 on error
1009 This ioctl defines the physical address of a one-page region in the guest
1010 physical address space. The region must be within the first 4GB of the
1011 guest physical address space and must not conflict with any memory slot
1012 or any mmio address. The guest may malfunction if it accesses this memory
1015 This ioctl is required on Intel-based hosts. This is needed on Intel hardware
1016 because of a quirk in the virtualization implementation (see the internals
1017 documentation when it pops into existence).
1020 4.41 KVM_SET_BOOT_CPU_ID
1022 Capability: KVM_CAP_SET_BOOT_CPU_ID
1023 Architectures: x86, ia64
1025 Parameters: unsigned long vcpu_id
1026 Returns: 0 on success, -1 on error
1028 Define which vcpu is the Bootstrap Processor (BSP). Values are the same
1029 as the vcpu id in KVM_CREATE_VCPU. If this ioctl is not called, the default
1035 Capability: KVM_CAP_XSAVE
1038 Parameters: struct kvm_xsave (out)
1039 Returns: 0 on success, -1 on error
1045 This ioctl would copy current vcpu's xsave struct to the userspace.
1050 Capability: KVM_CAP_XSAVE
1053 Parameters: struct kvm_xsave (in)
1054 Returns: 0 on success, -1 on error
1060 This ioctl would copy userspace's xsave struct to the kernel.
1065 Capability: KVM_CAP_XCRS
1068 Parameters: struct kvm_xcrs (out)
1069 Returns: 0 on success, -1 on error
1080 struct kvm_xcr xcrs[KVM_MAX_XCRS];
1084 This ioctl would copy current vcpu's xcrs to the userspace.
1089 Capability: KVM_CAP_XCRS
1092 Parameters: struct kvm_xcrs (in)
1093 Returns: 0 on success, -1 on error
1104 struct kvm_xcr xcrs[KVM_MAX_XCRS];
1108 This ioctl would set vcpu's xcr to the value userspace specified.
1111 4.46 KVM_GET_SUPPORTED_CPUID
1113 Capability: KVM_CAP_EXT_CPUID
1116 Parameters: struct kvm_cpuid2 (in/out)
1117 Returns: 0 on success, -1 on error
1122 struct kvm_cpuid_entry2 entries[0];
1125 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX 1
1126 #define KVM_CPUID_FLAG_STATEFUL_FUNC 2
1127 #define KVM_CPUID_FLAG_STATE_READ_NEXT 4
1129 struct kvm_cpuid_entry2 {
1140 This ioctl returns x86 cpuid features which are supported by both the hardware
1141 and kvm. Userspace can use the information returned by this ioctl to
1142 construct cpuid information (for KVM_SET_CPUID2) that is consistent with
1143 hardware, kernel, and userspace capabilities, and with user requirements (for
1144 example, the user may wish to constrain cpuid to emulate older hardware,
1145 or for feature consistency across a cluster).
1147 Userspace invokes KVM_GET_SUPPORTED_CPUID by passing a kvm_cpuid2 structure
1148 with the 'nent' field indicating the number of entries in the variable-size
1149 array 'entries'. If the number of entries is too low to describe the cpu
1150 capabilities, an error (E2BIG) is returned. If the number is too high,
1151 the 'nent' field is adjusted and an error (ENOMEM) is returned. If the
1152 number is just right, the 'nent' field is adjusted to the number of valid
1153 entries in the 'entries' array, which is then filled.
1155 The entries returned are the host cpuid as returned by the cpuid instruction,
1156 with unknown or unsupported features masked out. Some features (for example,
1157 x2apic), may not be present in the host cpu, but are exposed by kvm if it can
1158 emulate them efficiently. The fields in each entry are defined as follows:
1160 function: the eax value used to obtain the entry
1161 index: the ecx value used to obtain the entry (for entries that are
1163 flags: an OR of zero or more of the following:
1164 KVM_CPUID_FLAG_SIGNIFCANT_INDEX:
1165 if the index field is valid
1166 KVM_CPUID_FLAG_STATEFUL_FUNC:
1167 if cpuid for this function returns different values for successive
1168 invocations; there will be several entries with the same function,
1169 all with this flag set
1170 KVM_CPUID_FLAG_STATE_READ_NEXT:
1171 for KVM_CPUID_FLAG_STATEFUL_FUNC entries, set if this entry is
1172 the first entry to be read by a cpu
1173 eax, ebx, ecx, edx: the values returned by the cpuid instruction for
1174 this function/index combination
1176 The TSC deadline timer feature (CPUID leaf 1, ecx[24]) is always returned
1177 as false, since the feature depends on KVM_CREATE_IRQCHIP for local APIC
1178 support. Instead it is reported via
1180 ioctl(KVM_CHECK_EXTENSION, KVM_CAP_TSC_DEADLINE_TIMER)
1182 if that returns true and you use KVM_CREATE_IRQCHIP, or if you emulate the
1183 feature in userspace, then you can enable the feature for KVM_SET_CPUID2.
1186 4.47 KVM_PPC_GET_PVINFO
1188 Capability: KVM_CAP_PPC_GET_PVINFO
1191 Parameters: struct kvm_ppc_pvinfo (out)
1192 Returns: 0 on success, !0 on error
1194 struct kvm_ppc_pvinfo {
1200 This ioctl fetches PV specific information that need to be passed to the guest
1201 using the device tree or other means from vm context.
1203 The hcall array defines 4 instructions that make up a hypercall.
1205 If any additional field gets added to this structure later on, a bit for that
1206 additional piece of information will be set in the flags bitmap.
1208 The flags bitmap is defined as:
1210 /* the host supports the ePAPR idle hcall
1211 #define KVM_PPC_PVINFO_FLAGS_EV_IDLE (1<<0)
1213 4.48 KVM_ASSIGN_PCI_DEVICE
1215 Capability: KVM_CAP_DEVICE_ASSIGNMENT
1216 Architectures: x86 ia64
1218 Parameters: struct kvm_assigned_pci_dev (in)
1219 Returns: 0 on success, -1 on error
1221 Assigns a host PCI device to the VM.
1223 struct kvm_assigned_pci_dev {
1224 __u32 assigned_dev_id;
1234 The PCI device is specified by the triple segnr, busnr, and devfn.
1235 Identification in succeeding service requests is done via assigned_dev_id. The
1236 following flags are specified:
1238 /* Depends on KVM_CAP_IOMMU */
1239 #define KVM_DEV_ASSIGN_ENABLE_IOMMU (1 << 0)
1240 /* The following two depend on KVM_CAP_PCI_2_3 */
1241 #define KVM_DEV_ASSIGN_PCI_2_3 (1 << 1)
1242 #define KVM_DEV_ASSIGN_MASK_INTX (1 << 2)
1244 If KVM_DEV_ASSIGN_PCI_2_3 is set, the kernel will manage legacy INTx interrupts
1245 via the PCI-2.3-compliant device-level mask, thus enable IRQ sharing with other
1246 assigned devices or host devices. KVM_DEV_ASSIGN_MASK_INTX specifies the
1247 guest's view on the INTx mask, see KVM_ASSIGN_SET_INTX_MASK for details.
1249 The KVM_DEV_ASSIGN_ENABLE_IOMMU flag is a mandatory option to ensure
1250 isolation of the device. Usages not specifying this flag are deprecated.
1252 Only PCI header type 0 devices with PCI BAR resources are supported by
1253 device assignment. The user requesting this ioctl must have read/write
1254 access to the PCI sysfs resource files associated with the device.
1257 4.49 KVM_DEASSIGN_PCI_DEVICE
1259 Capability: KVM_CAP_DEVICE_DEASSIGNMENT
1260 Architectures: x86 ia64
1262 Parameters: struct kvm_assigned_pci_dev (in)
1263 Returns: 0 on success, -1 on error
1265 Ends PCI device assignment, releasing all associated resources.
1267 See KVM_CAP_DEVICE_ASSIGNMENT for the data structure. Only assigned_dev_id is
1268 used in kvm_assigned_pci_dev to identify the device.
1271 4.50 KVM_ASSIGN_DEV_IRQ
1273 Capability: KVM_CAP_ASSIGN_DEV_IRQ
1274 Architectures: x86 ia64
1276 Parameters: struct kvm_assigned_irq (in)
1277 Returns: 0 on success, -1 on error
1279 Assigns an IRQ to a passed-through device.
1281 struct kvm_assigned_irq {
1282 __u32 assigned_dev_id;
1283 __u32 host_irq; /* ignored (legacy field) */
1291 The following flags are defined:
1293 #define KVM_DEV_IRQ_HOST_INTX (1 << 0)
1294 #define KVM_DEV_IRQ_HOST_MSI (1 << 1)
1295 #define KVM_DEV_IRQ_HOST_MSIX (1 << 2)
1297 #define KVM_DEV_IRQ_GUEST_INTX (1 << 8)
1298 #define KVM_DEV_IRQ_GUEST_MSI (1 << 9)
1299 #define KVM_DEV_IRQ_GUEST_MSIX (1 << 10)
1301 It is not valid to specify multiple types per host or guest IRQ. However, the
1302 IRQ type of host and guest can differ or can even be null.
1305 4.51 KVM_DEASSIGN_DEV_IRQ
1307 Capability: KVM_CAP_ASSIGN_DEV_IRQ
1308 Architectures: x86 ia64
1310 Parameters: struct kvm_assigned_irq (in)
1311 Returns: 0 on success, -1 on error
1313 Ends an IRQ assignment to a passed-through device.
1315 See KVM_ASSIGN_DEV_IRQ for the data structure. The target device is specified
1316 by assigned_dev_id, flags must correspond to the IRQ type specified on
1317 KVM_ASSIGN_DEV_IRQ. Partial deassignment of host or guest IRQ is allowed.
1320 4.52 KVM_SET_GSI_ROUTING
1322 Capability: KVM_CAP_IRQ_ROUTING
1323 Architectures: x86 ia64
1325 Parameters: struct kvm_irq_routing (in)
1326 Returns: 0 on success, -1 on error
1328 Sets the GSI routing table entries, overwriting any previously set entries.
1330 struct kvm_irq_routing {
1333 struct kvm_irq_routing_entry entries[0];
1336 No flags are specified so far, the corresponding field must be set to zero.
1338 struct kvm_irq_routing_entry {
1344 struct kvm_irq_routing_irqchip irqchip;
1345 struct kvm_irq_routing_msi msi;
1350 /* gsi routing entry types */
1351 #define KVM_IRQ_ROUTING_IRQCHIP 1
1352 #define KVM_IRQ_ROUTING_MSI 2
1354 No flags are specified so far, the corresponding field must be set to zero.
1356 struct kvm_irq_routing_irqchip {
1361 struct kvm_irq_routing_msi {
1369 4.53 KVM_ASSIGN_SET_MSIX_NR
1371 Capability: KVM_CAP_DEVICE_MSIX
1372 Architectures: x86 ia64
1374 Parameters: struct kvm_assigned_msix_nr (in)
1375 Returns: 0 on success, -1 on error
1377 Set the number of MSI-X interrupts for an assigned device. The number is
1378 reset again by terminating the MSI-X assignment of the device via
1379 KVM_DEASSIGN_DEV_IRQ. Calling this service more than once at any earlier
1382 struct kvm_assigned_msix_nr {
1383 __u32 assigned_dev_id;
1388 #define KVM_MAX_MSIX_PER_DEV 256
1391 4.54 KVM_ASSIGN_SET_MSIX_ENTRY
1393 Capability: KVM_CAP_DEVICE_MSIX
1394 Architectures: x86 ia64
1396 Parameters: struct kvm_assigned_msix_entry (in)
1397 Returns: 0 on success, -1 on error
1399 Specifies the routing of an MSI-X assigned device interrupt to a GSI. Setting
1400 the GSI vector to zero means disabling the interrupt.
1402 struct kvm_assigned_msix_entry {
1403 __u32 assigned_dev_id;
1405 __u16 entry; /* The index of entry in the MSI-X table */
1410 4.55 KVM_SET_TSC_KHZ
1412 Capability: KVM_CAP_TSC_CONTROL
1415 Parameters: virtual tsc_khz
1416 Returns: 0 on success, -1 on error
1418 Specifies the tsc frequency for the virtual machine. The unit of the
1422 4.56 KVM_GET_TSC_KHZ
1424 Capability: KVM_CAP_GET_TSC_KHZ
1428 Returns: virtual tsc-khz on success, negative value on error
1430 Returns the tsc frequency of the guest. The unit of the return value is
1431 KHz. If the host has unstable tsc this ioctl returns -EIO instead as an
1437 Capability: KVM_CAP_IRQCHIP
1440 Parameters: struct kvm_lapic_state (out)
1441 Returns: 0 on success, -1 on error
1443 #define KVM_APIC_REG_SIZE 0x400
1444 struct kvm_lapic_state {
1445 char regs[KVM_APIC_REG_SIZE];
1448 Reads the Local APIC registers and copies them into the input argument. The
1449 data format and layout are the same as documented in the architecture manual.
1454 Capability: KVM_CAP_IRQCHIP
1457 Parameters: struct kvm_lapic_state (in)
1458 Returns: 0 on success, -1 on error
1460 #define KVM_APIC_REG_SIZE 0x400
1461 struct kvm_lapic_state {
1462 char regs[KVM_APIC_REG_SIZE];
1465 Copies the input argument into the the Local APIC registers. The data format
1466 and layout are the same as documented in the architecture manual.
1471 Capability: KVM_CAP_IOEVENTFD
1474 Parameters: struct kvm_ioeventfd (in)
1475 Returns: 0 on success, !0 on error
1477 This ioctl attaches or detaches an ioeventfd to a legal pio/mmio address
1478 within the guest. A guest write in the registered address will signal the
1479 provided event instead of triggering an exit.
1481 struct kvm_ioeventfd {
1483 __u64 addr; /* legal pio/mmio address */
1484 __u32 len; /* 1, 2, 4, or 8 bytes */
1490 For the special case of virtio-ccw devices on s390, the ioevent is matched
1491 to a subchannel/virtqueue tuple instead.
1493 The following flags are defined:
1495 #define KVM_IOEVENTFD_FLAG_DATAMATCH (1 << kvm_ioeventfd_flag_nr_datamatch)
1496 #define KVM_IOEVENTFD_FLAG_PIO (1 << kvm_ioeventfd_flag_nr_pio)
1497 #define KVM_IOEVENTFD_FLAG_DEASSIGN (1 << kvm_ioeventfd_flag_nr_deassign)
1498 #define KVM_IOEVENTFD_FLAG_VIRTIO_CCW_NOTIFY \
1499 (1 << kvm_ioeventfd_flag_nr_virtio_ccw_notify)
1501 If datamatch flag is set, the event will be signaled only if the written value
1502 to the registered address is equal to datamatch in struct kvm_ioeventfd.
1504 For virtio-ccw devices, addr contains the subchannel id and datamatch the
1510 Capability: KVM_CAP_SW_TLB
1513 Parameters: struct kvm_dirty_tlb (in)
1514 Returns: 0 on success, -1 on error
1516 struct kvm_dirty_tlb {
1521 This must be called whenever userspace has changed an entry in the shared
1522 TLB, prior to calling KVM_RUN on the associated vcpu.
1524 The "bitmap" field is the userspace address of an array. This array
1525 consists of a number of bits, equal to the total number of TLB entries as
1526 determined by the last successful call to KVM_CONFIG_TLB, rounded up to the
1527 nearest multiple of 64.
1529 Each bit corresponds to one TLB entry, ordered the same as in the shared TLB
1532 The array is little-endian: the bit 0 is the least significant bit of the
1533 first byte, bit 8 is the least significant bit of the second byte, etc.
1534 This avoids any complications with differing word sizes.
1536 The "num_dirty" field is a performance hint for KVM to determine whether it
1537 should skip processing the bitmap and just invalidate everything. It must
1538 be set to the number of set bits in the bitmap.
1541 4.61 KVM_ASSIGN_SET_INTX_MASK
1543 Capability: KVM_CAP_PCI_2_3
1546 Parameters: struct kvm_assigned_pci_dev (in)
1547 Returns: 0 on success, -1 on error
1549 Allows userspace to mask PCI INTx interrupts from the assigned device. The
1550 kernel will not deliver INTx interrupts to the guest between setting and
1551 clearing of KVM_ASSIGN_SET_INTX_MASK via this interface. This enables use of
1552 and emulation of PCI 2.3 INTx disable command register behavior.
1554 This may be used for both PCI 2.3 devices supporting INTx disable natively and
1555 older devices lacking this support. Userspace is responsible for emulating the
1556 read value of the INTx disable bit in the guest visible PCI command register.
1557 When modifying the INTx disable state, userspace should precede updating the
1558 physical device command register by calling this ioctl to inform the kernel of
1559 the new intended INTx mask state.
1561 Note that the kernel uses the device INTx disable bit to internally manage the
1562 device interrupt state for PCI 2.3 devices. Reads of this register may
1563 therefore not match the expected value. Writes should always use the guest
1564 intended INTx disable value rather than attempting to read-copy-update the
1565 current physical device state. Races between user and kernel updates to the
1566 INTx disable bit are handled lazily in the kernel. It's possible the device
1567 may generate unintended interrupts, but they will not be injected into the
1570 See KVM_ASSIGN_DEV_IRQ for the data structure. The target device is specified
1571 by assigned_dev_id. In the flags field, only KVM_DEV_ASSIGN_MASK_INTX is
1575 4.62 KVM_CREATE_SPAPR_TCE
1577 Capability: KVM_CAP_SPAPR_TCE
1578 Architectures: powerpc
1580 Parameters: struct kvm_create_spapr_tce (in)
1581 Returns: file descriptor for manipulating the created TCE table
1583 This creates a virtual TCE (translation control entry) table, which
1584 is an IOMMU for PAPR-style virtual I/O. It is used to translate
1585 logical addresses used in virtual I/O into guest physical addresses,
1586 and provides a scatter/gather capability for PAPR virtual I/O.
1588 /* for KVM_CAP_SPAPR_TCE */
1589 struct kvm_create_spapr_tce {
1594 The liobn field gives the logical IO bus number for which to create a
1595 TCE table. The window_size field specifies the size of the DMA window
1596 which this TCE table will translate - the table will contain one 64
1597 bit TCE entry for every 4kiB of the DMA window.
1599 When the guest issues an H_PUT_TCE hcall on a liobn for which a TCE
1600 table has been created using this ioctl(), the kernel will handle it
1601 in real mode, updating the TCE table. H_PUT_TCE calls for other
1602 liobns will cause a vm exit and must be handled by userspace.
1604 The return value is a file descriptor which can be passed to mmap(2)
1605 to map the created TCE table into userspace. This lets userspace read
1606 the entries written by kernel-handled H_PUT_TCE calls, and also lets
1607 userspace update the TCE table directly which is useful in some
1611 4.63 KVM_ALLOCATE_RMA
1613 Capability: KVM_CAP_PPC_RMA
1614 Architectures: powerpc
1616 Parameters: struct kvm_allocate_rma (out)
1617 Returns: file descriptor for mapping the allocated RMA
1619 This allocates a Real Mode Area (RMA) from the pool allocated at boot
1620 time by the kernel. An RMA is a physically-contiguous, aligned region
1621 of memory used on older POWER processors to provide the memory which
1622 will be accessed by real-mode (MMU off) accesses in a KVM guest.
1623 POWER processors support a set of sizes for the RMA that usually
1624 includes 64MB, 128MB, 256MB and some larger powers of two.
1626 /* for KVM_ALLOCATE_RMA */
1627 struct kvm_allocate_rma {
1631 The return value is a file descriptor which can be passed to mmap(2)
1632 to map the allocated RMA into userspace. The mapped area can then be
1633 passed to the KVM_SET_USER_MEMORY_REGION ioctl to establish it as the
1634 RMA for a virtual machine. The size of the RMA in bytes (which is
1635 fixed at host kernel boot time) is returned in the rma_size field of
1636 the argument structure.
1638 The KVM_CAP_PPC_RMA capability is 1 or 2 if the KVM_ALLOCATE_RMA ioctl
1639 is supported; 2 if the processor requires all virtual machines to have
1640 an RMA, or 1 if the processor can use an RMA but doesn't require it,
1641 because it supports the Virtual RMA (VRMA) facility.
1646 Capability: KVM_CAP_USER_NMI
1650 Returns: 0 on success, -1 on error
1652 Queues an NMI on the thread's vcpu. Note this is well defined only
1653 when KVM_CREATE_IRQCHIP has not been called, since this is an interface
1654 between the virtual cpu core and virtual local APIC. After KVM_CREATE_IRQCHIP
1655 has been called, this interface is completely emulated within the kernel.
1657 To use this to emulate the LINT1 input with KVM_CREATE_IRQCHIP, use the
1658 following algorithm:
1661 - read the local APIC's state (KVM_GET_LAPIC)
1662 - check whether changing LINT1 will queue an NMI (see the LVT entry for LINT1)
1663 - if so, issue KVM_NMI
1666 Some guests configure the LINT1 NMI input to cause a panic, aiding in
1670 4.65 KVM_S390_UCAS_MAP
1672 Capability: KVM_CAP_S390_UCONTROL
1675 Parameters: struct kvm_s390_ucas_mapping (in)
1676 Returns: 0 in case of success
1678 The parameter is defined like this:
1679 struct kvm_s390_ucas_mapping {
1685 This ioctl maps the memory at "user_addr" with the length "length" to
1686 the vcpu's address space starting at "vcpu_addr". All parameters need to
1687 be aligned by 1 megabyte.
1690 4.66 KVM_S390_UCAS_UNMAP
1692 Capability: KVM_CAP_S390_UCONTROL
1695 Parameters: struct kvm_s390_ucas_mapping (in)
1696 Returns: 0 in case of success
1698 The parameter is defined like this:
1699 struct kvm_s390_ucas_mapping {
1705 This ioctl unmaps the memory in the vcpu's address space starting at
1706 "vcpu_addr" with the length "length". The field "user_addr" is ignored.
1707 All parameters need to be aligned by 1 megabyte.
1710 4.67 KVM_S390_VCPU_FAULT
1712 Capability: KVM_CAP_S390_UCONTROL
1715 Parameters: vcpu absolute address (in)
1716 Returns: 0 in case of success
1718 This call creates a page table entry on the virtual cpu's address space
1719 (for user controlled virtual machines) or the virtual machine's address
1720 space (for regular virtual machines). This only works for minor faults,
1721 thus it's recommended to access subject memory page via the user page
1722 table upfront. This is useful to handle validity intercepts for user
1723 controlled virtual machines to fault in the virtual cpu's lowcore pages
1724 prior to calling the KVM_RUN ioctl.
1727 4.68 KVM_SET_ONE_REG
1729 Capability: KVM_CAP_ONE_REG
1732 Parameters: struct kvm_one_reg (in)
1733 Returns: 0 on success, negative value on failure
1735 struct kvm_one_reg {
1740 Using this ioctl, a single vcpu register can be set to a specific value
1741 defined by user space with the passed in struct kvm_one_reg, where id
1742 refers to the register identifier as described below and addr is a pointer
1743 to a variable with the respective size. There can be architecture agnostic
1744 and architecture specific registers. Each have their own range of operation
1745 and their own constants and width. To keep track of the implemented
1746 registers, find a list below:
1748 Arch | Register | Width (bits)
1750 PPC | KVM_REG_PPC_HIOR | 64
1751 PPC | KVM_REG_PPC_IAC1 | 64
1752 PPC | KVM_REG_PPC_IAC2 | 64
1753 PPC | KVM_REG_PPC_IAC3 | 64
1754 PPC | KVM_REG_PPC_IAC4 | 64
1755 PPC | KVM_REG_PPC_DAC1 | 64
1756 PPC | KVM_REG_PPC_DAC2 | 64
1757 PPC | KVM_REG_PPC_DABR | 64
1758 PPC | KVM_REG_PPC_DSCR | 64
1759 PPC | KVM_REG_PPC_PURR | 64
1760 PPC | KVM_REG_PPC_SPURR | 64
1761 PPC | KVM_REG_PPC_DAR | 64
1762 PPC | KVM_REG_PPC_DSISR | 32
1763 PPC | KVM_REG_PPC_AMR | 64
1764 PPC | KVM_REG_PPC_UAMOR | 64
1765 PPC | KVM_REG_PPC_MMCR0 | 64
1766 PPC | KVM_REG_PPC_MMCR1 | 64
1767 PPC | KVM_REG_PPC_MMCRA | 64
1768 PPC | KVM_REG_PPC_PMC1 | 32
1769 PPC | KVM_REG_PPC_PMC2 | 32
1770 PPC | KVM_REG_PPC_PMC3 | 32
1771 PPC | KVM_REG_PPC_PMC4 | 32
1772 PPC | KVM_REG_PPC_PMC5 | 32
1773 PPC | KVM_REG_PPC_PMC6 | 32
1774 PPC | KVM_REG_PPC_PMC7 | 32
1775 PPC | KVM_REG_PPC_PMC8 | 32
1776 PPC | KVM_REG_PPC_FPR0 | 64
1778 PPC | KVM_REG_PPC_FPR31 | 64
1779 PPC | KVM_REG_PPC_VR0 | 128
1781 PPC | KVM_REG_PPC_VR31 | 128
1782 PPC | KVM_REG_PPC_VSR0 | 128
1784 PPC | KVM_REG_PPC_VSR31 | 128
1785 PPC | KVM_REG_PPC_FPSCR | 64
1786 PPC | KVM_REG_PPC_VSCR | 32
1787 PPC | KVM_REG_PPC_VPA_ADDR | 64
1788 PPC | KVM_REG_PPC_VPA_SLB | 128
1789 PPC | KVM_REG_PPC_VPA_DTL | 128
1790 PPC | KVM_REG_PPC_EPCR | 32
1791 PPC | KVM_REG_PPC_EPR | 32
1792 PPC | KVM_REG_PPC_TCR | 32
1793 PPC | KVM_REG_PPC_TSR | 32
1794 PPC | KVM_REG_PPC_OR_TSR | 32
1795 PPC | KVM_REG_PPC_CLEAR_TSR | 32
1796 PPC | KVM_REG_PPC_MAS0 | 32
1797 PPC | KVM_REG_PPC_MAS1 | 32
1798 PPC | KVM_REG_PPC_MAS2 | 64
1799 PPC | KVM_REG_PPC_MAS7_3 | 64
1800 PPC | KVM_REG_PPC_MAS4 | 32
1801 PPC | KVM_REG_PPC_MAS6 | 32
1802 PPC | KVM_REG_PPC_MMUCFG | 32
1803 PPC | KVM_REG_PPC_TLB0CFG | 32
1804 PPC | KVM_REG_PPC_TLB1CFG | 32
1805 PPC | KVM_REG_PPC_TLB2CFG | 32
1806 PPC | KVM_REG_PPC_TLB3CFG | 32
1807 PPC | KVM_REG_PPC_TLB0PS | 32
1808 PPC | KVM_REG_PPC_TLB1PS | 32
1809 PPC | KVM_REG_PPC_TLB2PS | 32
1810 PPC | KVM_REG_PPC_TLB3PS | 32
1811 PPC | KVM_REG_PPC_EPTCFG | 32
1812 PPC | KVM_REG_PPC_ICP_STATE | 64
1813 PPC | KVM_REG_PPC_TB_OFFSET | 64
1814 PPC | KVM_REG_PPC_SPMC1 | 32
1815 PPC | KVM_REG_PPC_SPMC2 | 32
1816 PPC | KVM_REG_PPC_IAMR | 64
1817 PPC | KVM_REG_PPC_TFHAR | 64
1818 PPC | KVM_REG_PPC_TFIAR | 64
1819 PPC | KVM_REG_PPC_TEXASR | 64
1820 PPC | KVM_REG_PPC_FSCR | 64
1821 PPC | KVM_REG_PPC_PSPB | 32
1822 PPC | KVM_REG_PPC_EBBHR | 64
1823 PPC | KVM_REG_PPC_EBBRR | 64
1824 PPC | KVM_REG_PPC_BESCR | 64
1825 PPC | KVM_REG_PPC_TAR | 64
1826 PPC | KVM_REG_PPC_DPDES | 64
1827 PPC | KVM_REG_PPC_DAWR | 64
1828 PPC | KVM_REG_PPC_DAWRX | 64
1829 PPC | KVM_REG_PPC_CIABR | 64
1830 PPC | KVM_REG_PPC_IC | 64
1831 PPC | KVM_REG_PPC_VTB | 64
1832 PPC | KVM_REG_PPC_CSIGR | 64
1833 PPC | KVM_REG_PPC_TACR | 64
1834 PPC | KVM_REG_PPC_TCSCR | 64
1835 PPC | KVM_REG_PPC_PID | 64
1836 PPC | KVM_REG_PPC_ACOP | 64
1837 PPC | KVM_REG_PPC_VRSAVE | 32
1838 PPC | KVM_REG_PPC_TM_GPR0 | 64
1840 PPC | KVM_REG_PPC_TM_GPR31 | 64
1841 PPC | KVM_REG_PPC_TM_VSR0 | 128
1843 PPC | KVM_REG_PPC_TM_VSR63 | 128
1844 PPC | KVM_REG_PPC_TM_CR | 64
1845 PPC | KVM_REG_PPC_TM_LR | 64
1846 PPC | KVM_REG_PPC_TM_CTR | 64
1847 PPC | KVM_REG_PPC_TM_FPSCR | 64
1848 PPC | KVM_REG_PPC_TM_AMR | 64
1849 PPC | KVM_REG_PPC_TM_PPR | 64
1850 PPC | KVM_REG_PPC_TM_VRSAVE | 64
1851 PPC | KVM_REG_PPC_TM_VSCR | 32
1852 PPC | KVM_REG_PPC_TM_DSCR | 64
1853 PPC | KVM_REG_PPC_TM_TAR | 64
1855 ARM registers are mapped using the lower 32 bits. The upper 16 of that
1856 is the register group type, or coprocessor number:
1858 ARM core registers have the following id bit patterns:
1859 0x4020 0000 0010 <index into the kvm_regs struct:16>
1861 ARM 32-bit CP15 registers have the following id bit patterns:
1862 0x4020 0000 000F <zero:1> <crn:4> <crm:4> <opc1:4> <opc2:3>
1864 ARM 64-bit CP15 registers have the following id bit patterns:
1865 0x4030 0000 000F <zero:1> <zero:4> <crm:4> <opc1:4> <zero:3>
1867 ARM CCSIDR registers are demultiplexed by CSSELR value:
1868 0x4020 0000 0011 00 <csselr:8>
1870 ARM 32-bit VFP control registers have the following id bit patterns:
1871 0x4020 0000 0012 1 <regno:12>
1873 ARM 64-bit FP registers have the following id bit patterns:
1874 0x4030 0000 0012 0 <regno:12>
1877 arm64 registers are mapped using the lower 32 bits. The upper 16 of
1878 that is the register group type, or coprocessor number:
1880 arm64 core/FP-SIMD registers have the following id bit patterns. Note
1881 that the size of the access is variable, as the kvm_regs structure
1882 contains elements ranging from 32 to 128 bits. The index is a 32bit
1883 value in the kvm_regs structure seen as a 32bit array.
1884 0x60x0 0000 0010 <index into the kvm_regs struct:16>
1886 arm64 CCSIDR registers are demultiplexed by CSSELR value:
1887 0x6020 0000 0011 00 <csselr:8>
1889 arm64 system registers have the following id bit patterns:
1890 0x6030 0000 0013 <op0:2> <op1:3> <crn:4> <crm:4> <op2:3>
1892 4.69 KVM_GET_ONE_REG
1894 Capability: KVM_CAP_ONE_REG
1897 Parameters: struct kvm_one_reg (in and out)
1898 Returns: 0 on success, negative value on failure
1900 This ioctl allows to receive the value of a single register implemented
1901 in a vcpu. The register to read is indicated by the "id" field of the
1902 kvm_one_reg struct passed in. On success, the register value can be found
1903 at the memory location pointed to by "addr".
1905 The list of registers accessible using this interface is identical to the
1909 4.70 KVM_KVMCLOCK_CTRL
1911 Capability: KVM_CAP_KVMCLOCK_CTRL
1912 Architectures: Any that implement pvclocks (currently x86 only)
1915 Returns: 0 on success, -1 on error
1917 This signals to the host kernel that the specified guest is being paused by
1918 userspace. The host will set a flag in the pvclock structure that is checked
1919 from the soft lockup watchdog. The flag is part of the pvclock structure that
1920 is shared between guest and host, specifically the second bit of the flags
1921 field of the pvclock_vcpu_time_info structure. It will be set exclusively by
1922 the host and read/cleared exclusively by the guest. The guest operation of
1923 checking and clearing the flag must an atomic operation so
1924 load-link/store-conditional, or equivalent must be used. There are two cases
1925 where the guest will clear the flag: when the soft lockup watchdog timer resets
1926 itself or when a soft lockup is detected. This ioctl can be called any time
1927 after pausing the vcpu, but before it is resumed.
1932 Capability: KVM_CAP_SIGNAL_MSI
1935 Parameters: struct kvm_msi (in)
1936 Returns: >0 on delivery, 0 if guest blocked the MSI, and -1 on error
1938 Directly inject a MSI message. Only valid with in-kernel irqchip that handles
1949 No flags are defined so far. The corresponding field must be 0.
1952 4.71 KVM_CREATE_PIT2
1954 Capability: KVM_CAP_PIT2
1957 Parameters: struct kvm_pit_config (in)
1958 Returns: 0 on success, -1 on error
1960 Creates an in-kernel device model for the i8254 PIT. This call is only valid
1961 after enabling in-kernel irqchip support via KVM_CREATE_IRQCHIP. The following
1962 parameters have to be passed:
1964 struct kvm_pit_config {
1971 #define KVM_PIT_SPEAKER_DUMMY 1 /* emulate speaker port stub */
1973 PIT timer interrupts may use a per-VM kernel thread for injection. If it
1974 exists, this thread will have a name of the following pattern:
1976 kvm-pit/<owner-process-pid>
1978 When running a guest with elevated priorities, the scheduling parameters of
1979 this thread may have to be adjusted accordingly.
1981 This IOCTL replaces the obsolete KVM_CREATE_PIT.
1986 Capability: KVM_CAP_PIT_STATE2
1989 Parameters: struct kvm_pit_state2 (out)
1990 Returns: 0 on success, -1 on error
1992 Retrieves the state of the in-kernel PIT model. Only valid after
1993 KVM_CREATE_PIT2. The state is returned in the following structure:
1995 struct kvm_pit_state2 {
1996 struct kvm_pit_channel_state channels[3];
2003 /* disable PIT in HPET legacy mode */
2004 #define KVM_PIT_FLAGS_HPET_LEGACY 0x00000001
2006 This IOCTL replaces the obsolete KVM_GET_PIT.
2011 Capability: KVM_CAP_PIT_STATE2
2014 Parameters: struct kvm_pit_state2 (in)
2015 Returns: 0 on success, -1 on error
2017 Sets the state of the in-kernel PIT model. Only valid after KVM_CREATE_PIT2.
2018 See KVM_GET_PIT2 for details on struct kvm_pit_state2.
2020 This IOCTL replaces the obsolete KVM_SET_PIT.
2023 4.74 KVM_PPC_GET_SMMU_INFO
2025 Capability: KVM_CAP_PPC_GET_SMMU_INFO
2026 Architectures: powerpc
2029 Returns: 0 on success, -1 on error
2031 This populates and returns a structure describing the features of
2032 the "Server" class MMU emulation supported by KVM.
2033 This can in turn be used by userspace to generate the appropriate
2034 device-tree properties for the guest operating system.
2036 The structure contains some global informations, followed by an
2037 array of supported segment page sizes:
2039 struct kvm_ppc_smmu_info {
2043 struct kvm_ppc_one_seg_page_size sps[KVM_PPC_PAGE_SIZES_MAX_SZ];
2046 The supported flags are:
2048 - KVM_PPC_PAGE_SIZES_REAL:
2049 When that flag is set, guest page sizes must "fit" the backing
2050 store page sizes. When not set, any page size in the list can
2051 be used regardless of how they are backed by userspace.
2053 - KVM_PPC_1T_SEGMENTS
2054 The emulated MMU supports 1T segments in addition to the
2057 The "slb_size" field indicates how many SLB entries are supported
2059 The "sps" array contains 8 entries indicating the supported base
2060 page sizes for a segment in increasing order. Each entry is defined
2063 struct kvm_ppc_one_seg_page_size {
2064 __u32 page_shift; /* Base page shift of segment (or 0) */
2065 __u32 slb_enc; /* SLB encoding for BookS */
2066 struct kvm_ppc_one_page_size enc[KVM_PPC_PAGE_SIZES_MAX_SZ];
2069 An entry with a "page_shift" of 0 is unused. Because the array is
2070 organized in increasing order, a lookup can stop when encoutering
2073 The "slb_enc" field provides the encoding to use in the SLB for the
2074 page size. The bits are in positions such as the value can directly
2075 be OR'ed into the "vsid" argument of the slbmte instruction.
2077 The "enc" array is a list which for each of those segment base page
2078 size provides the list of supported actual page sizes (which can be
2079 only larger or equal to the base page size), along with the
2080 corresponding encoding in the hash PTE. Similarly, the array is
2081 8 entries sorted by increasing sizes and an entry with a "0" shift
2082 is an empty entry and a terminator:
2084 struct kvm_ppc_one_page_size {
2085 __u32 page_shift; /* Page shift (or 0) */
2086 __u32 pte_enc; /* Encoding in the HPTE (>>12) */
2089 The "pte_enc" field provides a value that can OR'ed into the hash
2090 PTE's RPN field (ie, it needs to be shifted left by 12 to OR it
2091 into the hash PTE second double word).
2095 Capability: KVM_CAP_IRQFD
2098 Parameters: struct kvm_irqfd (in)
2099 Returns: 0 on success, -1 on error
2101 Allows setting an eventfd to directly trigger a guest interrupt.
2102 kvm_irqfd.fd specifies the file descriptor to use as the eventfd and
2103 kvm_irqfd.gsi specifies the irqchip pin toggled by this event. When
2104 an event is tiggered on the eventfd, an interrupt is injected into
2105 the guest using the specified gsi pin. The irqfd is removed using
2106 the KVM_IRQFD_FLAG_DEASSIGN flag, specifying both kvm_irqfd.fd
2109 With KVM_CAP_IRQFD_RESAMPLE, KVM_IRQFD supports a de-assert and notify
2110 mechanism allowing emulation of level-triggered, irqfd-based
2111 interrupts. When KVM_IRQFD_FLAG_RESAMPLE is set the user must pass an
2112 additional eventfd in the kvm_irqfd.resamplefd field. When operating
2113 in resample mode, posting of an interrupt through kvm_irq.fd asserts
2114 the specified gsi in the irqchip. When the irqchip is resampled, such
2115 as from an EOI, the gsi is de-asserted and the user is notifed via
2116 kvm_irqfd.resamplefd. It is the user's responsibility to re-queue
2117 the interrupt if the device making use of it still requires service.
2118 Note that closing the resamplefd is not sufficient to disable the
2119 irqfd. The KVM_IRQFD_FLAG_RESAMPLE is only necessary on assignment
2120 and need not be specified with KVM_IRQFD_FLAG_DEASSIGN.
2122 4.76 KVM_PPC_ALLOCATE_HTAB
2124 Capability: KVM_CAP_PPC_ALLOC_HTAB
2125 Architectures: powerpc
2127 Parameters: Pointer to u32 containing hash table order (in/out)
2128 Returns: 0 on success, -1 on error
2130 This requests the host kernel to allocate an MMU hash table for a
2131 guest using the PAPR paravirtualization interface. This only does
2132 anything if the kernel is configured to use the Book 3S HV style of
2133 virtualization. Otherwise the capability doesn't exist and the ioctl
2134 returns an ENOTTY error. The rest of this description assumes Book 3S
2137 There must be no vcpus running when this ioctl is called; if there
2138 are, it will do nothing and return an EBUSY error.
2140 The parameter is a pointer to a 32-bit unsigned integer variable
2141 containing the order (log base 2) of the desired size of the hash
2142 table, which must be between 18 and 46. On successful return from the
2143 ioctl, it will have been updated with the order of the hash table that
2146 If no hash table has been allocated when any vcpu is asked to run
2147 (with the KVM_RUN ioctl), the host kernel will allocate a
2148 default-sized hash table (16 MB).
2150 If this ioctl is called when a hash table has already been allocated,
2151 the kernel will clear out the existing hash table (zero all HPTEs) and
2152 return the hash table order in the parameter. (If the guest is using
2153 the virtualized real-mode area (VRMA) facility, the kernel will
2154 re-create the VMRA HPTEs on the next KVM_RUN of any vcpu.)
2156 4.77 KVM_S390_INTERRUPT
2160 Type: vm ioctl, vcpu ioctl
2161 Parameters: struct kvm_s390_interrupt (in)
2162 Returns: 0 on success, -1 on error
2164 Allows to inject an interrupt to the guest. Interrupts can be floating
2165 (vm ioctl) or per cpu (vcpu ioctl), depending on the interrupt type.
2167 Interrupt parameters are passed via kvm_s390_interrupt:
2169 struct kvm_s390_interrupt {
2175 type can be one of the following:
2177 KVM_S390_SIGP_STOP (vcpu) - sigp restart
2178 KVM_S390_PROGRAM_INT (vcpu) - program check; code in parm
2179 KVM_S390_SIGP_SET_PREFIX (vcpu) - sigp set prefix; prefix address in parm
2180 KVM_S390_RESTART (vcpu) - restart
2181 KVM_S390_INT_VIRTIO (vm) - virtio external interrupt; external interrupt
2182 parameters in parm and parm64
2183 KVM_S390_INT_SERVICE (vm) - sclp external interrupt; sclp parameter in parm
2184 KVM_S390_INT_EMERGENCY (vcpu) - sigp emergency; source cpu in parm
2185 KVM_S390_INT_EXTERNAL_CALL (vcpu) - sigp external call; source cpu in parm
2186 KVM_S390_INT_IO(ai,cssid,ssid,schid) (vm) - compound value to indicate an
2187 I/O interrupt (ai - adapter interrupt; cssid,ssid,schid - subchannel);
2188 I/O interruption parameters in parm (subchannel) and parm64 (intparm,
2189 interruption subclass)
2190 KVM_S390_MCHK (vm, vcpu) - machine check interrupt; cr 14 bits in parm,
2191 machine check interrupt code in parm64 (note that
2192 machine checks needing further payload are not
2193 supported by this ioctl)
2195 Note that the vcpu ioctl is asynchronous to vcpu execution.
2197 4.78 KVM_PPC_GET_HTAB_FD
2199 Capability: KVM_CAP_PPC_HTAB_FD
2200 Architectures: powerpc
2202 Parameters: Pointer to struct kvm_get_htab_fd (in)
2203 Returns: file descriptor number (>= 0) on success, -1 on error
2205 This returns a file descriptor that can be used either to read out the
2206 entries in the guest's hashed page table (HPT), or to write entries to
2207 initialize the HPT. The returned fd can only be written to if the
2208 KVM_GET_HTAB_WRITE bit is set in the flags field of the argument, and
2209 can only be read if that bit is clear. The argument struct looks like
2212 /* For KVM_PPC_GET_HTAB_FD */
2213 struct kvm_get_htab_fd {
2219 /* Values for kvm_get_htab_fd.flags */
2220 #define KVM_GET_HTAB_BOLTED_ONLY ((__u64)0x1)
2221 #define KVM_GET_HTAB_WRITE ((__u64)0x2)
2223 The `start_index' field gives the index in the HPT of the entry at
2224 which to start reading. It is ignored when writing.
2226 Reads on the fd will initially supply information about all
2227 "interesting" HPT entries. Interesting entries are those with the
2228 bolted bit set, if the KVM_GET_HTAB_BOLTED_ONLY bit is set, otherwise
2229 all entries. When the end of the HPT is reached, the read() will
2230 return. If read() is called again on the fd, it will start again from
2231 the beginning of the HPT, but will only return HPT entries that have
2232 changed since they were last read.
2234 Data read or written is structured as a header (8 bytes) followed by a
2235 series of valid HPT entries (16 bytes) each. The header indicates how
2236 many valid HPT entries there are and how many invalid entries follow
2237 the valid entries. The invalid entries are not represented explicitly
2238 in the stream. The header format is:
2240 struct kvm_get_htab_header {
2246 Writes to the fd create HPT entries starting at the index given in the
2247 header; first `n_valid' valid entries with contents from the data
2248 written, then `n_invalid' invalid entries, invalidating any previously
2249 valid entries found.
2251 4.79 KVM_CREATE_DEVICE
2253 Capability: KVM_CAP_DEVICE_CTRL
2255 Parameters: struct kvm_create_device (in/out)
2256 Returns: 0 on success, -1 on error
2258 ENODEV: The device type is unknown or unsupported
2259 EEXIST: Device already created, and this type of device may not
2260 be instantiated multiple times
2262 Other error conditions may be defined by individual device types or
2263 have their standard meanings.
2265 Creates an emulated device in the kernel. The file descriptor returned
2266 in fd can be used with KVM_SET/GET/HAS_DEVICE_ATTR.
2268 If the KVM_CREATE_DEVICE_TEST flag is set, only test whether the
2269 device type is supported (not necessarily whether it can be created
2272 Individual devices should not define flags. Attributes should be used
2273 for specifying any behavior that is not implied by the device type
2276 struct kvm_create_device {
2277 __u32 type; /* in: KVM_DEV_TYPE_xxx */
2278 __u32 fd; /* out: device handle */
2279 __u32 flags; /* in: KVM_CREATE_DEVICE_xxx */
2282 4.80 KVM_SET_DEVICE_ATTR/KVM_GET_DEVICE_ATTR
2284 Capability: KVM_CAP_DEVICE_CTRL
2286 Parameters: struct kvm_device_attr
2287 Returns: 0 on success, -1 on error
2289 ENXIO: The group or attribute is unknown/unsupported for this device
2290 EPERM: The attribute cannot (currently) be accessed this way
2291 (e.g. read-only attribute, or attribute that only makes
2292 sense when the device is in a different state)
2294 Other error conditions may be defined by individual device types.
2296 Gets/sets a specified piece of device configuration and/or state. The
2297 semantics are device-specific. See individual device documentation in
2298 the "devices" directory. As with ONE_REG, the size of the data
2299 transferred is defined by the particular attribute.
2301 struct kvm_device_attr {
2302 __u32 flags; /* no flags currently defined */
2303 __u32 group; /* device-defined */
2304 __u64 attr; /* group-defined */
2305 __u64 addr; /* userspace address of attr data */
2308 4.81 KVM_HAS_DEVICE_ATTR
2310 Capability: KVM_CAP_DEVICE_CTRL
2312 Parameters: struct kvm_device_attr
2313 Returns: 0 on success, -1 on error
2315 ENXIO: The group or attribute is unknown/unsupported for this device
2317 Tests whether a device supports a particular attribute. A successful
2318 return indicates the attribute is implemented. It does not necessarily
2319 indicate that the attribute can be read or written in the device's
2320 current state. "addr" is ignored.
2322 4.82 KVM_ARM_VCPU_INIT
2325 Architectures: arm, arm64
2327 Parameters: struct struct kvm_vcpu_init (in)
2328 Returns: 0 on success; -1 on error
2330 Â EINVAL: Â Â Â the target is unknown, or the combination of features is invalid.
2331 Â ENOENT: Â Â Â a features bit specified is unknown.
2333 This tells KVM what type of CPU to present to the guest, and what
2334 optional features it should have. Â This will cause a reset of the cpu
2335 registers to their initial values. Â If this is not called, KVM_RUN will
2336 return ENOEXEC for that vcpu.
2338 Note that because some registers reflect machine topology, all vcpus
2339 should be created before this ioctl is invoked.
2342 - KVM_ARM_VCPU_POWER_OFF: Starts the CPU in a power-off state.
2343 Depends on KVM_CAP_ARM_PSCI.
2344 - KVM_ARM_VCPU_EL1_32BIT: Starts the CPU in a 32bit mode.
2345 Depends on KVM_CAP_ARM_EL1_32BIT (arm64 only).
2348 4.83 KVM_ARM_PREFERRED_TARGET
2351 Architectures: arm, arm64
2353 Parameters: struct struct kvm_vcpu_init (out)
2354 Returns: 0 on success; -1 on error
2356 ENODEV: no preferred target available for the host
2358 This queries KVM for preferred CPU target type which can be emulated
2359 by KVM on underlying host.
2361 The ioctl returns struct kvm_vcpu_init instance containing information
2362 about preferred CPU target type and recommended features for it. The
2363 kvm_vcpu_init->features bitmap returned will have feature bits set if
2364 the preferred target recommends setting these features, but this is
2367 The information returned by this ioctl can be used to prepare an instance
2368 of struct kvm_vcpu_init for KVM_ARM_VCPU_INIT ioctl which will result in
2369 in VCPU matching underlying host.
2372 4.84 KVM_GET_REG_LIST
2375 Architectures: arm, arm64
2377 Parameters: struct kvm_reg_list (in/out)
2378 Returns: 0 on success; -1 on error
2380 Â E2BIG: Â Â Â Â the reg index list is too big to fit in the array specified by
2381 Â Â Â Â Â Â Â Â Â Â Â Â the user (the number required will be written into n).
2383 struct kvm_reg_list {
2384 __u64 n; /* number of registers in reg[] */
2388 This ioctl returns the guest registers that are supported for the
2389 KVM_GET_ONE_REG/KVM_SET_ONE_REG calls.
2391 4.85 KVM_ARM_SET_DEVICE_ADDR
2393 Capability: KVM_CAP_ARM_SET_DEVICE_ADDR
2394 Architectures: arm, arm64
2396 Parameters: struct kvm_arm_device_address (in)
2397 Returns: 0 on success, -1 on error
2399 ENODEV: The device id is unknown
2400 ENXIO: Device not supported on current system
2401 EEXIST: Address already set
2402 E2BIG: Address outside guest physical address space
2403 EBUSY: Address overlaps with other device range
2405 struct kvm_arm_device_addr {
2410 Specify a device address in the guest's physical address space where guests
2411 can access emulated or directly exposed devices, which the host kernel needs
2412 to know about. The id field is an architecture specific identifier for a
2415 ARM/arm64 divides the id field into two parts, a device id and an
2416 address type id specific to the individual device.
2418 Â bits: | 63 ... 32 | 31 ... 16 | 15 ... 0 |
2419 field: | 0x00000000 | device id | addr type id |
2421 ARM/arm64 currently only require this when using the in-kernel GIC
2422 support for the hardware VGIC features, using KVM_ARM_DEVICE_VGIC_V2
2423 as the device id. When setting the base address for the guest's
2424 mapping of the VGIC virtual CPU and distributor interface, the ioctl
2425 must be called after calling KVM_CREATE_IRQCHIP, but before calling
2426 KVM_RUN on any of the VCPUs. Calling this ioctl twice for any of the
2427 base addresses will return -EEXIST.
2429 4.86 KVM_PPC_RTAS_DEFINE_TOKEN
2431 Capability: KVM_CAP_PPC_RTAS
2434 Parameters: struct kvm_rtas_token_args
2435 Returns: 0 on success, -1 on error
2437 Defines a token value for a RTAS (Run Time Abstraction Services)
2438 service in order to allow it to be handled in the kernel. The
2439 argument struct gives the name of the service, which must be the name
2440 of a service that has a kernel-side implementation. If the token
2441 value is non-zero, it will be associated with that service, and
2442 subsequent RTAS calls by the guest specifying that token will be
2443 handled by the kernel. If the token value is 0, then any token
2444 associated with the service will be forgotten, and subsequent RTAS
2445 calls by the guest for that service will be passed to userspace to be
2449 5. The kvm_run structure
2450 ------------------------
2452 Application code obtains a pointer to the kvm_run structure by
2453 mmap()ing a vcpu fd. From that point, application code can control
2454 execution by changing fields in kvm_run prior to calling the KVM_RUN
2455 ioctl, and obtain information about the reason KVM_RUN returned by
2456 looking up structure members.
2460 __u8 request_interrupt_window;
2462 Request that KVM_RUN return when it becomes possible to inject external
2463 interrupts into the guest. Useful in conjunction with KVM_INTERRUPT.
2470 When KVM_RUN has returned successfully (return value 0), this informs
2471 application code why KVM_RUN has returned. Allowable values for this
2472 field are detailed below.
2474 __u8 ready_for_interrupt_injection;
2476 If request_interrupt_window has been specified, this field indicates
2477 an interrupt can be injected now with KVM_INTERRUPT.
2481 The value of the current interrupt flag. Only valid if in-kernel
2482 local APIC is not used.
2486 /* in (pre_kvm_run), out (post_kvm_run) */
2489 The value of the cr8 register. Only valid if in-kernel local APIC is
2490 not used. Both input and output.
2494 The value of the APIC BASE msr. Only valid if in-kernel local
2495 APIC is not used. Both input and output.
2498 /* KVM_EXIT_UNKNOWN */
2500 __u64 hardware_exit_reason;
2503 If exit_reason is KVM_EXIT_UNKNOWN, the vcpu has exited due to unknown
2504 reasons. Further architecture-specific information is available in
2505 hardware_exit_reason.
2507 /* KVM_EXIT_FAIL_ENTRY */
2509 __u64 hardware_entry_failure_reason;
2512 If exit_reason is KVM_EXIT_FAIL_ENTRY, the vcpu could not be run due
2513 to unknown reasons. Further architecture-specific information is
2514 available in hardware_entry_failure_reason.
2516 /* KVM_EXIT_EXCEPTION */
2526 #define KVM_EXIT_IO_IN 0
2527 #define KVM_EXIT_IO_OUT 1
2529 __u8 size; /* bytes */
2532 __u64 data_offset; /* relative to kvm_run start */
2535 If exit_reason is KVM_EXIT_IO, then the vcpu has
2536 executed a port I/O instruction which could not be satisfied by kvm.
2537 data_offset describes where the data is located (KVM_EXIT_IO_OUT) or
2538 where kvm expects application code to place the data for the next
2539 KVM_RUN invocation (KVM_EXIT_IO_IN). Data format is a packed array.
2542 struct kvm_debug_exit_arch arch;
2555 If exit_reason is KVM_EXIT_MMIO, then the vcpu has
2556 executed a memory-mapped I/O instruction which could not be satisfied
2557 by kvm. The 'data' member contains the written data if 'is_write' is
2558 true, and should be filled by application code otherwise.
2560 NOTE: For KVM_EXIT_IO, KVM_EXIT_MMIO, KVM_EXIT_OSI, KVM_EXIT_DCR,
2561 KVM_EXIT_PAPR and KVM_EXIT_EPR the corresponding
2562 operations are complete (and guest state is consistent) only after userspace
2563 has re-entered the kernel with KVM_RUN. The kernel side will first finish
2564 incomplete operations and then check for pending signals. Userspace
2565 can re-enter the guest with an unmasked signal pending to complete
2568 /* KVM_EXIT_HYPERCALL */
2577 Unused. This was once used for 'hypercall to userspace'. To implement
2578 such functionality, use KVM_EXIT_IO (x86) or KVM_EXIT_MMIO (all except s390).
2579 Note KVM_EXIT_IO is significantly faster than KVM_EXIT_MMIO.
2581 /* KVM_EXIT_TPR_ACCESS */
2588 To be documented (KVM_TPR_ACCESS_REPORTING).
2590 /* KVM_EXIT_S390_SIEIC */
2593 __u64 mask; /* psw upper half */
2594 __u64 addr; /* psw lower half */
2601 /* KVM_EXIT_S390_RESET */
2602 #define KVM_S390_RESET_POR 1
2603 #define KVM_S390_RESET_CLEAR 2
2604 #define KVM_S390_RESET_SUBSYSTEM 4
2605 #define KVM_S390_RESET_CPU_INIT 8
2606 #define KVM_S390_RESET_IPL 16
2607 __u64 s390_reset_flags;
2611 /* KVM_EXIT_S390_UCONTROL */
2613 __u64 trans_exc_code;
2617 s390 specific. A page fault has occurred for a user controlled virtual
2618 machine (KVM_VM_S390_UNCONTROL) on it's host page table that cannot be
2619 resolved by the kernel.
2620 The program code and the translation exception code that were placed
2621 in the cpu's lowcore are presented here as defined by the z Architecture
2622 Principles of Operation Book in the Chapter for Dynamic Address Translation
2639 MOL uses a special hypercall interface it calls 'OSI'. To enable it, we catch
2640 hypercalls and exit with this exit struct that contains all the guest gprs.
2642 If exit_reason is KVM_EXIT_OSI, then the vcpu has triggered such a hypercall.
2643 Userspace can now handle the hypercall and when it's done modify the gprs as
2644 necessary. Upon guest entry all guest GPRs will then be replaced by the values
2647 /* KVM_EXIT_PAPR_HCALL */
2654 This is used on 64-bit PowerPC when emulating a pSeries partition,
2655 e.g. with the 'pseries' machine type in qemu. It occurs when the
2656 guest does a hypercall using the 'sc 1' instruction. The 'nr' field
2657 contains the hypercall number (from the guest R3), and 'args' contains
2658 the arguments (from the guest R4 - R12). Userspace should put the
2659 return code in 'ret' and any extra returned values in args[].
2660 The possible hypercalls are defined in the Power Architecture Platform
2661 Requirements (PAPR) document available from www.power.org (free
2662 developer registration required to access it).
2664 /* KVM_EXIT_S390_TSCH */
2666 __u16 subchannel_id;
2667 __u16 subchannel_nr;
2674 s390 specific. This exit occurs when KVM_CAP_S390_CSS_SUPPORT has been enabled
2675 and TEST SUBCHANNEL was intercepted. If dequeued is set, a pending I/O
2676 interrupt for the target subchannel has been dequeued and subchannel_id,
2677 subchannel_nr, io_int_parm and io_int_word contain the parameters for that
2678 interrupt. ipb is needed for instruction parameter decoding.
2685 On FSL BookE PowerPC chips, the interrupt controller has a fast patch
2686 interrupt acknowledge path to the core. When the core successfully
2687 delivers an interrupt, it automatically populates the EPR register with
2688 the interrupt vector number and acknowledges the interrupt inside
2689 the interrupt controller.
2691 In case the interrupt controller lives in user space, we need to do
2692 the interrupt acknowledge cycle through it to fetch the next to be
2693 delivered interrupt vector using this exit.
2695 It gets triggered whenever both KVM_CAP_PPC_EPR are enabled and an
2696 external interrupt has just been delivered into the guest. User space
2697 should put the acknowledged interrupt vector into the 'epr' field.
2699 /* Fix the size of the union. */
2704 * shared registers between kvm and userspace.
2705 * kvm_valid_regs specifies the register classes set by the host
2706 * kvm_dirty_regs specified the register classes dirtied by userspace
2707 * struct kvm_sync_regs is architecture specific, as well as the
2708 * bits for kvm_valid_regs and kvm_dirty_regs
2710 __u64 kvm_valid_regs;
2711 __u64 kvm_dirty_regs;
2713 struct kvm_sync_regs regs;
2717 If KVM_CAP_SYNC_REGS is defined, these fields allow userspace to access
2718 certain guest registers without having to call SET/GET_*REGS. Thus we can
2719 avoid some system call overhead if userspace has to handle the exit.
2720 Userspace can query the validity of the structure by checking
2721 kvm_valid_regs for specific bits. These bits are architecture specific
2722 and usually define the validity of a groups of registers. (e.g. one bit
2723 for general purpose registers)
2728 6. Capabilities that can be enabled
2729 -----------------------------------
2731 There are certain capabilities that change the behavior of the virtual CPU when
2732 enabled. To enable them, please see section 4.37. Below you can find a list of
2733 capabilities and what their effect on the vCPU is when enabling them.
2735 The following information is provided along with the description:
2737 Architectures: which instruction set architectures provide this ioctl.
2738 x86 includes both i386 and x86_64.
2740 Parameters: what parameters are accepted by the capability.
2742 Returns: the return value. General error numbers (EBADF, ENOMEM, EINVAL)
2743 are not detailed, but errors with specific meanings are.
2750 Returns: 0 on success; -1 on error
2752 This capability enables interception of OSI hypercalls that otherwise would
2753 be treated as normal system calls to be injected into the guest. OSI hypercalls
2754 were invented by Mac-on-Linux to have a standardized communication mechanism
2755 between the guest and the host.
2757 When this capability is enabled, KVM_EXIT_OSI can occur.
2760 6.2 KVM_CAP_PPC_PAPR
2764 Returns: 0 on success; -1 on error
2766 This capability enables interception of PAPR hypercalls. PAPR hypercalls are
2767 done using the hypercall instruction "sc 1".
2769 It also sets the guest privilege level to "supervisor" mode. Usually the guest
2770 runs in "hypervisor" privilege mode with a few missing features.
2772 In addition to the above, it changes the semantics of SDR1. In this mode, the
2773 HTAB address part of SDR1 contains an HVA instead of a GPA, as PAPR keeps the
2774 HTAB invisible to the guest.
2776 When this capability is enabled, KVM_EXIT_PAPR_HCALL can occur.
2782 Parameters: args[0] is the address of a struct kvm_config_tlb
2783 Returns: 0 on success; -1 on error
2785 struct kvm_config_tlb {
2792 Configures the virtual CPU's TLB array, establishing a shared memory area
2793 between userspace and KVM. The "params" and "array" fields are userspace
2794 addresses of mmu-type-specific data structures. The "array_len" field is an
2795 safety mechanism, and should be set to the size in bytes of the memory that
2796 userspace has reserved for the array. It must be at least the size dictated
2797 by "mmu_type" and "params".
2799 While KVM_RUN is active, the shared region is under control of KVM. Its
2800 contents are undefined, and any modification by userspace results in
2801 boundedly undefined behavior.
2803 On return from KVM_RUN, the shared region will reflect the current state of
2804 the guest's TLB. If userspace makes any changes, it must call KVM_DIRTY_TLB
2805 to tell KVM which entries have been changed, prior to calling KVM_RUN again
2808 For mmu types KVM_MMU_FSL_BOOKE_NOHV and KVM_MMU_FSL_BOOKE_HV:
2809 - The "params" field is of type "struct kvm_book3e_206_tlb_params".
2810 - The "array" field points to an array of type "struct
2811 kvm_book3e_206_tlb_entry".
2812 - The array consists of all entries in the first TLB, followed by all
2813 entries in the second TLB.
2814 - Within a TLB, entries are ordered first by increasing set number. Within a
2815 set, entries are ordered by way (increasing ESEL).
2816 - The hash for determining set number in TLB0 is: (MAS2 >> 12) & (num_sets - 1)
2817 where "num_sets" is the tlb_sizes[] value divided by the tlb_ways[] value.
2818 - The tsize field of mas1 shall be set to 4K on TLB0, even though the
2819 hardware ignores this value for TLB0.
2821 6.4 KVM_CAP_S390_CSS_SUPPORT
2825 Returns: 0 on success; -1 on error
2827 This capability enables support for handling of channel I/O instructions.
2829 TEST PENDING INTERRUPTION and the interrupt portion of TEST SUBCHANNEL are
2830 handled in-kernel, while the other I/O instructions are passed to userspace.
2832 When this capability is enabled, KVM_EXIT_S390_TSCH will occur on TEST
2833 SUBCHANNEL intercepts.
2838 Parameters: args[0] defines whether the proxy facility is active
2839 Returns: 0 on success; -1 on error
2841 This capability enables or disables the delivery of interrupts through the
2842 external proxy facility.
2844 When enabled (args[0] != 0), every time the guest gets an external interrupt
2845 delivered, it automatically exits into user space with a KVM_EXIT_EPR exit
2846 to receive the topmost interrupt vector.
2848 When disabled (args[0] == 0), behavior is as if this facility is unsupported.
2850 When this capability is enabled, KVM_EXIT_EPR can occur.
2852 6.6 KVM_CAP_IRQ_MPIC
2855 Parameters: args[0] is the MPIC device fd
2856 args[1] is the MPIC CPU number for this vcpu
2858 This capability connects the vcpu to an in-kernel MPIC device.
2860 6.7 KVM_CAP_IRQ_XICS
2863 Parameters: args[0] is the XICS device fd
2864 args[1] is the XICS CPU number (server ID) for this vcpu
2866 This capability connects the vcpu to an in-kernel XICS device.