1 The Definitive KVM (Kernel-based Virtual Machine) API Documentation
2 ===================================================================
7 The kvm API is a set of ioctls that are issued to control various aspects
8 of a virtual machine. The ioctls belong to three classes
10 - System ioctls: These query and set global attributes which affect the
11 whole kvm subsystem. In addition a system ioctl is used to create
14 - VM ioctls: These query and set attributes that affect an entire virtual
15 machine, for example memory layout. In addition a VM ioctl is used to
16 create virtual cpus (vcpus).
18 Only run VM ioctls from the same process (address space) that was used
21 - vcpu ioctls: These query and set attributes that control the operation
22 of a single virtual cpu.
24 Only run vcpu ioctls from the same thread that was used to create the
31 The kvm API is centered around file descriptors. An initial
32 open("/dev/kvm") obtains a handle to the kvm subsystem; this handle
33 can be used to issue system ioctls. A KVM_CREATE_VM ioctl on this
34 handle will create a VM file descriptor which can be used to issue VM
35 ioctls. A KVM_CREATE_VCPU ioctl on a VM fd will create a virtual cpu
36 and return a file descriptor pointing to it. Finally, ioctls on a vcpu
37 fd can be used to control the vcpu, including the important task of
38 actually running guest code.
40 In general file descriptors can be migrated among processes by means
41 of fork() and the SCM_RIGHTS facility of unix domain socket. These
42 kinds of tricks are explicitly not supported by kvm. While they will
43 not cause harm to the host, their actual behavior is not guaranteed by
44 the API. The only supported use is one virtual machine per process,
45 and one vcpu per thread.
51 As of Linux 2.6.22, the KVM ABI has been stabilized: no backward
52 incompatible change are allowed. However, there is an extension
53 facility that allows backward-compatible extensions to the API to be
56 The extension mechanism is not based on the Linux version number.
57 Instead, kvm defines extension identifiers and a facility to query
58 whether a particular extension identifier is available. If it is, a
59 set of ioctls is available for application use.
65 This section describes ioctls that can be used to control kvm guests.
66 For each ioctl, the following information is provided along with a
69 Capability: which KVM extension provides this ioctl. Can be 'basic',
70 which means that is will be provided by any kernel that supports
71 API version 12 (see section 4.1), a KVM_CAP_xyz constant, which
72 means availability needs to be checked with KVM_CHECK_EXTENSION
73 (see section 4.4), or 'none' which means that while not all kernels
74 support this ioctl, there's no capability bit to check its
75 availability: for kernels that don't support the ioctl,
76 the ioctl returns -ENOTTY.
78 Architectures: which instruction set architectures provide this ioctl.
79 x86 includes both i386 and x86_64.
81 Type: system, vm, or vcpu.
83 Parameters: what parameters are accepted by the ioctl.
85 Returns: the return value. General error numbers (EBADF, ENOMEM, EINVAL)
86 are not detailed, but errors with specific meanings are.
89 4.1 KVM_GET_API_VERSION
95 Returns: the constant KVM_API_VERSION (=12)
97 This identifies the API version as the stable kvm API. It is not
98 expected that this number will change. However, Linux 2.6.20 and
99 2.6.21 report earlier versions; these are not documented and not
100 supported. Applications should refuse to run if KVM_GET_API_VERSION
101 returns a value other than 12. If this check passes, all ioctls
102 described as 'basic' will be available.
110 Parameters: machine type identifier (KVM_VM_*)
111 Returns: a VM fd that can be used to control the new virtual machine.
113 The new VM has no virtual cpus and no memory.
114 You probably want to use 0 as machine type.
116 In order to create user controlled virtual machines on S390, check
117 KVM_CAP_S390_UCONTROL and use the flag KVM_VM_S390_UCONTROL as
118 privileged user (CAP_SYS_ADMIN).
120 To use hardware assisted virtualization on MIPS (VZ ASE) rather than
121 the default trap & emulate implementation (which changes the virtual
122 memory layout to fit in user mode), check KVM_CAP_MIPS_VZ and use the
126 4.3 KVM_GET_MSR_INDEX_LIST, KVM_GET_MSR_FEATURE_INDEX_LIST
128 Capability: basic, KVM_CAP_GET_MSR_FEATURES for KVM_GET_MSR_FEATURE_INDEX_LIST
131 Parameters: struct kvm_msr_list (in/out)
132 Returns: 0 on success; -1 on error
134 EFAULT: the msr index list cannot be read from or written to
135 E2BIG: the msr index list is to be to fit in the array specified by
138 struct kvm_msr_list {
139 __u32 nmsrs; /* number of msrs in entries */
143 The user fills in the size of the indices array in nmsrs, and in return
144 kvm adjusts nmsrs to reflect the actual number of msrs and fills in the
145 indices array with their numbers.
147 KVM_GET_MSR_INDEX_LIST returns the guest msrs that are supported. The list
148 varies by kvm version and host processor, but does not change otherwise.
150 Note: if kvm indicates supports MCE (KVM_CAP_MCE), then the MCE bank MSRs are
151 not returned in the MSR list, as different vcpus can have a different number
152 of banks, as set via the KVM_X86_SETUP_MCE ioctl.
154 KVM_GET_MSR_FEATURE_INDEX_LIST returns the list of MSRs that can be passed
155 to the KVM_GET_MSRS system ioctl. This lets userspace probe host capabilities
156 and processor features that are exposed via MSRs (e.g., VMX capabilities).
157 This list also varies by kvm version and host processor, but does not change
161 4.4 KVM_CHECK_EXTENSION
163 Capability: basic, KVM_CAP_CHECK_EXTENSION_VM for vm ioctl
165 Type: system ioctl, vm ioctl
166 Parameters: extension identifier (KVM_CAP_*)
167 Returns: 0 if unsupported; 1 (or some other positive integer) if supported
169 The API allows the application to query about extensions to the core
170 kvm API. Userspace passes an extension identifier (an integer) and
171 receives an integer that describes the extension availability.
172 Generally 0 means no and 1 means yes, but some extensions may report
173 additional information in the integer return value.
175 Based on their initialization different VMs may have different capabilities.
176 It is thus encouraged to use the vm ioctl to query for capabilities (available
177 with KVM_CAP_CHECK_EXTENSION_VM on the vm fd)
179 4.5 KVM_GET_VCPU_MMAP_SIZE
185 Returns: size of vcpu mmap area, in bytes
187 The KVM_RUN ioctl (cf.) communicates with userspace via a shared
188 memory region. This ioctl returns the size of that region. See the
189 KVM_RUN documentation for details.
192 4.6 KVM_SET_MEMORY_REGION
197 Parameters: struct kvm_memory_region (in)
198 Returns: 0 on success, -1 on error
200 This ioctl is obsolete and has been removed.
208 Parameters: vcpu id (apic id on x86)
209 Returns: vcpu fd on success, -1 on error
211 This API adds a vcpu to a virtual machine. No more than max_vcpus may be added.
212 The vcpu id is an integer in the range [0, max_vcpu_id).
214 The recommended max_vcpus value can be retrieved using the KVM_CAP_NR_VCPUS of
215 the KVM_CHECK_EXTENSION ioctl() at run-time.
216 The maximum possible value for max_vcpus can be retrieved using the
217 KVM_CAP_MAX_VCPUS of the KVM_CHECK_EXTENSION ioctl() at run-time.
219 If the KVM_CAP_NR_VCPUS does not exist, you should assume that max_vcpus is 4
221 If the KVM_CAP_MAX_VCPUS does not exist, you should assume that max_vcpus is
222 same as the value returned from KVM_CAP_NR_VCPUS.
224 The maximum possible value for max_vcpu_id can be retrieved using the
225 KVM_CAP_MAX_VCPU_ID of the KVM_CHECK_EXTENSION ioctl() at run-time.
227 If the KVM_CAP_MAX_VCPU_ID does not exist, you should assume that max_vcpu_id
228 is the same as the value returned from KVM_CAP_MAX_VCPUS.
230 On powerpc using book3s_hv mode, the vcpus are mapped onto virtual
231 threads in one or more virtual CPU cores. (This is because the
232 hardware requires all the hardware threads in a CPU core to be in the
233 same partition.) The KVM_CAP_PPC_SMT capability indicates the number
234 of vcpus per virtual core (vcore). The vcore id is obtained by
235 dividing the vcpu id by the number of vcpus per vcore. The vcpus in a
236 given vcore will always be in the same physical core as each other
237 (though that might be a different physical core from time to time).
238 Userspace can control the threading (SMT) mode of the guest by its
239 allocation of vcpu ids. For example, if userspace wants
240 single-threaded guest vcpus, it should make all vcpu ids be a multiple
241 of the number of vcpus per vcore.
243 For virtual cpus that have been created with S390 user controlled virtual
244 machines, the resulting vcpu fd can be memory mapped at page offset
245 KVM_S390_SIE_PAGE_OFFSET in order to obtain a memory map of the virtual
246 cpu's hardware control block.
249 4.8 KVM_GET_DIRTY_LOG (vm ioctl)
254 Parameters: struct kvm_dirty_log (in/out)
255 Returns: 0 on success, -1 on error
257 /* for KVM_GET_DIRTY_LOG */
258 struct kvm_dirty_log {
262 void __user *dirty_bitmap; /* one bit per page */
267 Given a memory slot, return a bitmap containing any pages dirtied
268 since the last call to this ioctl. Bit 0 is the first page in the
269 memory slot. Ensure the entire structure is cleared to avoid padding
272 If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 specifies
273 the address space for which you want to return the dirty bitmap.
274 They must be less than the value that KVM_CHECK_EXTENSION returns for
275 the KVM_CAP_MULTI_ADDRESS_SPACE capability.
278 4.9 KVM_SET_MEMORY_ALIAS
283 Parameters: struct kvm_memory_alias (in)
284 Returns: 0 (success), -1 (error)
286 This ioctl is obsolete and has been removed.
295 Returns: 0 on success, -1 on error
297 EINTR: an unmasked signal is pending
299 This ioctl is used to run a guest virtual cpu. While there are no
300 explicit parameters, there is an implicit parameter block that can be
301 obtained by mmap()ing the vcpu fd at offset 0, with the size given by
302 KVM_GET_VCPU_MMAP_SIZE. The parameter block is formatted as a 'struct
303 kvm_run' (see below).
309 Architectures: all except ARM, arm64
311 Parameters: struct kvm_regs (out)
312 Returns: 0 on success, -1 on error
314 Reads the general purpose registers from the vcpu.
318 /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */
319 __u64 rax, rbx, rcx, rdx;
320 __u64 rsi, rdi, rsp, rbp;
321 __u64 r8, r9, r10, r11;
322 __u64 r12, r13, r14, r15;
328 /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */
339 Architectures: all except ARM, arm64
341 Parameters: struct kvm_regs (in)
342 Returns: 0 on success, -1 on error
344 Writes the general purpose registers into the vcpu.
346 See KVM_GET_REGS for the data structure.
352 Architectures: x86, ppc
354 Parameters: struct kvm_sregs (out)
355 Returns: 0 on success, -1 on error
357 Reads special registers from the vcpu.
361 struct kvm_segment cs, ds, es, fs, gs, ss;
362 struct kvm_segment tr, ldt;
363 struct kvm_dtable gdt, idt;
364 __u64 cr0, cr2, cr3, cr4, cr8;
367 __u64 interrupt_bitmap[(KVM_NR_INTERRUPTS + 63) / 64];
370 /* ppc -- see arch/powerpc/include/uapi/asm/kvm.h */
372 interrupt_bitmap is a bitmap of pending external interrupts. At most
373 one bit may be set. This interrupt has been acknowledged by the APIC
374 but not yet injected into the cpu core.
380 Architectures: x86, ppc
382 Parameters: struct kvm_sregs (in)
383 Returns: 0 on success, -1 on error
385 Writes special registers into the vcpu. See KVM_GET_SREGS for the
394 Parameters: struct kvm_translation (in/out)
395 Returns: 0 on success, -1 on error
397 Translates a virtual address according to the vcpu's current address
400 struct kvm_translation {
402 __u64 linear_address;
405 __u64 physical_address;
416 Architectures: x86, ppc, mips
418 Parameters: struct kvm_interrupt (in)
419 Returns: 0 on success, negative on failure.
421 Queues a hardware interrupt vector to be injected.
423 /* for KVM_INTERRUPT */
424 struct kvm_interrupt {
431 Returns: 0 on success,
432 -EEXIST if an interrupt is already enqueued
433 -EINVAL the the irq number is invalid
434 -ENXIO if the PIC is in the kernel
435 -EFAULT if the pointer is invalid
437 Note 'irq' is an interrupt vector, not an interrupt pin or line. This
438 ioctl is useful if the in-kernel PIC is not used.
442 Queues an external interrupt to be injected. This ioctl is overleaded
443 with 3 different irq values:
447 This injects an edge type external interrupt into the guest once it's ready
448 to receive interrupts. When injected, the interrupt is done.
450 b) KVM_INTERRUPT_UNSET
452 This unsets any pending interrupt.
454 Only available with KVM_CAP_PPC_UNSET_IRQ.
456 c) KVM_INTERRUPT_SET_LEVEL
458 This injects a level type external interrupt into the guest context. The
459 interrupt stays pending until a specific ioctl with KVM_INTERRUPT_UNSET
462 Only available with KVM_CAP_PPC_IRQ_LEVEL.
464 Note that any value for 'irq' other than the ones stated above is invalid
465 and incurs unexpected behavior.
469 Queues an external interrupt to be injected into the virtual CPU. A negative
470 interrupt number dequeues the interrupt.
481 Support for this has been removed. Use KVM_SET_GUEST_DEBUG instead.
486 Capability: basic (vcpu), KVM_CAP_GET_MSR_FEATURES (system)
488 Type: system ioctl, vcpu ioctl
489 Parameters: struct kvm_msrs (in/out)
490 Returns: number of msrs successfully returned;
493 When used as a system ioctl:
494 Reads the values of MSR-based features that are available for the VM. This
495 is similar to KVM_GET_SUPPORTED_CPUID, but it returns MSR indices and values.
496 The list of msr-based features can be obtained using KVM_GET_MSR_FEATURE_INDEX_LIST
499 When used as a vcpu ioctl:
500 Reads model-specific registers from the vcpu. Supported msr indices can
501 be obtained using KVM_GET_MSR_INDEX_LIST in a system ioctl.
504 __u32 nmsrs; /* number of msrs in entries */
507 struct kvm_msr_entry entries[0];
510 struct kvm_msr_entry {
516 Application code should set the 'nmsrs' member (which indicates the
517 size of the entries array) and the 'index' member of each array entry.
518 kvm will fill in the 'data' member.
526 Parameters: struct kvm_msrs (in)
527 Returns: 0 on success, -1 on error
529 Writes model-specific registers to the vcpu. See KVM_GET_MSRS for the
532 Application code should set the 'nmsrs' member (which indicates the
533 size of the entries array), and the 'index' and 'data' members of each
542 Parameters: struct kvm_cpuid (in)
543 Returns: 0 on success, -1 on error
545 Defines the vcpu responses to the cpuid instruction. Applications
546 should use the KVM_SET_CPUID2 ioctl if available.
549 struct kvm_cpuid_entry {
558 /* for KVM_SET_CPUID */
562 struct kvm_cpuid_entry entries[0];
566 4.21 KVM_SET_SIGNAL_MASK
571 Parameters: struct kvm_signal_mask (in)
572 Returns: 0 on success, -1 on error
574 Defines which signals are blocked during execution of KVM_RUN. This
575 signal mask temporarily overrides the threads signal mask. Any
576 unblocked signal received (except SIGKILL and SIGSTOP, which retain
577 their traditional behaviour) will cause KVM_RUN to return with -EINTR.
579 Note the signal will only be delivered if not blocked by the original
582 /* for KVM_SET_SIGNAL_MASK */
583 struct kvm_signal_mask {
594 Parameters: struct kvm_fpu (out)
595 Returns: 0 on success, -1 on error
597 Reads the floating point state from the vcpu.
599 /* for KVM_GET_FPU and KVM_SET_FPU */
604 __u8 ftwx; /* in fxsave format */
620 Parameters: struct kvm_fpu (in)
621 Returns: 0 on success, -1 on error
623 Writes the floating point state to the vcpu.
625 /* for KVM_GET_FPU and KVM_SET_FPU */
630 __u8 ftwx; /* in fxsave format */
641 4.24 KVM_CREATE_IRQCHIP
643 Capability: KVM_CAP_IRQCHIP, KVM_CAP_S390_IRQCHIP (s390)
644 Architectures: x86, ARM, arm64, s390
647 Returns: 0 on success, -1 on error
649 Creates an interrupt controller model in the kernel.
650 On x86, creates a virtual ioapic, a virtual PIC (two PICs, nested), and sets up
651 future vcpus to have a local APIC. IRQ routing for GSIs 0-15 is set to both
652 PIC and IOAPIC; GSI 16-23 only go to the IOAPIC.
653 On ARM/arm64, a GICv2 is created. Any other GIC versions require the usage of
654 KVM_CREATE_DEVICE, which also supports creating a GICv2. Using
655 KVM_CREATE_DEVICE is preferred over KVM_CREATE_IRQCHIP for GICv2.
656 On s390, a dummy irq routing table is created.
658 Note that on s390 the KVM_CAP_S390_IRQCHIP vm capability needs to be enabled
659 before KVM_CREATE_IRQCHIP can be used.
664 Capability: KVM_CAP_IRQCHIP
665 Architectures: x86, arm, arm64
667 Parameters: struct kvm_irq_level
668 Returns: 0 on success, -1 on error
670 Sets the level of a GSI input to the interrupt controller model in the kernel.
671 On some architectures it is required that an interrupt controller model has
672 been previously created with KVM_CREATE_IRQCHIP. Note that edge-triggered
673 interrupts require the level to be set to 1 and then back to 0.
675 On real hardware, interrupt pins can be active-low or active-high. This
676 does not matter for the level field of struct kvm_irq_level: 1 always
677 means active (asserted), 0 means inactive (deasserted).
679 x86 allows the operating system to program the interrupt polarity
680 (active-low/active-high) for level-triggered interrupts, and KVM used
681 to consider the polarity. However, due to bitrot in the handling of
682 active-low interrupts, the above convention is now valid on x86 too.
683 This is signaled by KVM_CAP_X86_IOAPIC_POLARITY_IGNORED. Userspace
684 should not present interrupts to the guest as active-low unless this
685 capability is present (or unless it is not using the in-kernel irqchip,
689 ARM/arm64 can signal an interrupt either at the CPU level, or at the
690 in-kernel irqchip (GIC), and for in-kernel irqchip can tell the GIC to
691 use PPIs designated for specific cpus. The irq field is interpreted
694 Â bits: | 31 ... 24 | 23 ... 16 | 15 ... 0 |
695 field: | irq_type | vcpu_index | irq_id |
697 The irq_type field has the following values:
698 - irq_type[0]: out-of-kernel GIC: irq_id 0 is IRQ, irq_id 1 is FIQ
699 - irq_type[1]: in-kernel GIC: SPI, irq_id between 32 and 1019 (incl.)
700 (the vcpu_index field is ignored)
701 - irq_type[2]: in-kernel GIC: PPI, irq_id between 16 and 31 (incl.)
703 (The irq_id field thus corresponds nicely to the IRQ ID in the ARM GIC specs)
705 In both cases, level is used to assert/deassert the line.
707 struct kvm_irq_level {
710 __s32 status; /* not used for KVM_IRQ_LEVEL */
712 __u32 level; /* 0 or 1 */
718 Capability: KVM_CAP_IRQCHIP
721 Parameters: struct kvm_irqchip (in/out)
722 Returns: 0 on success, -1 on error
724 Reads the state of a kernel interrupt controller created with
725 KVM_CREATE_IRQCHIP into a buffer provided by the caller.
728 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */
731 char dummy[512]; /* reserving space */
732 struct kvm_pic_state pic;
733 struct kvm_ioapic_state ioapic;
740 Capability: KVM_CAP_IRQCHIP
743 Parameters: struct kvm_irqchip (in)
744 Returns: 0 on success, -1 on error
746 Sets the state of a kernel interrupt controller created with
747 KVM_CREATE_IRQCHIP from a buffer provided by the caller.
750 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */
753 char dummy[512]; /* reserving space */
754 struct kvm_pic_state pic;
755 struct kvm_ioapic_state ioapic;
760 4.28 KVM_XEN_HVM_CONFIG
762 Capability: KVM_CAP_XEN_HVM
765 Parameters: struct kvm_xen_hvm_config (in)
766 Returns: 0 on success, -1 on error
768 Sets the MSR that the Xen HVM guest uses to initialize its hypercall
769 page, and provides the starting address and size of the hypercall
770 blobs in userspace. When the guest writes the MSR, kvm copies one
771 page of a blob (32- or 64-bit, depending on the vcpu mode) to guest
774 struct kvm_xen_hvm_config {
787 Capability: KVM_CAP_ADJUST_CLOCK
790 Parameters: struct kvm_clock_data (out)
791 Returns: 0 on success, -1 on error
793 Gets the current timestamp of kvmclock as seen by the current guest. In
794 conjunction with KVM_SET_CLOCK, it is used to ensure monotonicity on scenarios
797 When KVM_CAP_ADJUST_CLOCK is passed to KVM_CHECK_EXTENSION, it returns the
798 set of bits that KVM can return in struct kvm_clock_data's flag member.
800 The only flag defined now is KVM_CLOCK_TSC_STABLE. If set, the returned
801 value is the exact kvmclock value seen by all VCPUs at the instant
802 when KVM_GET_CLOCK was called. If clear, the returned value is simply
803 CLOCK_MONOTONIC plus a constant offset; the offset can be modified
804 with KVM_SET_CLOCK. KVM will try to make all VCPUs follow this clock,
805 but the exact value read by each VCPU could differ, because the host
808 struct kvm_clock_data {
809 __u64 clock; /* kvmclock current value */
817 Capability: KVM_CAP_ADJUST_CLOCK
820 Parameters: struct kvm_clock_data (in)
821 Returns: 0 on success, -1 on error
823 Sets the current timestamp of kvmclock to the value specified in its parameter.
824 In conjunction with KVM_GET_CLOCK, it is used to ensure monotonicity on scenarios
827 struct kvm_clock_data {
828 __u64 clock; /* kvmclock current value */
834 4.31 KVM_GET_VCPU_EVENTS
836 Capability: KVM_CAP_VCPU_EVENTS
837 Extended by: KVM_CAP_INTR_SHADOW
840 Parameters: struct kvm_vcpu_event (out)
841 Returns: 0 on success, -1 on error
843 Gets currently pending exceptions, interrupts, and NMIs as well as related
846 struct kvm_vcpu_events {
876 Only two fields are defined in the flags field:
878 - KVM_VCPUEVENT_VALID_SHADOW may be set in the flags field to signal that
879 interrupt.shadow contains a valid state.
881 - KVM_VCPUEVENT_VALID_SMM may be set in the flags field to signal that
882 smi contains a valid state.
884 4.32 KVM_SET_VCPU_EVENTS
886 Capability: KVM_CAP_VCPU_EVENTS
887 Extended by: KVM_CAP_INTR_SHADOW
890 Parameters: struct kvm_vcpu_event (in)
891 Returns: 0 on success, -1 on error
893 Set pending exceptions, interrupts, and NMIs as well as related states of the
896 See KVM_GET_VCPU_EVENTS for the data structure.
898 Fields that may be modified asynchronously by running VCPUs can be excluded
899 from the update. These fields are nmi.pending, sipi_vector, smi.smm,
900 smi.pending. Keep the corresponding bits in the flags field cleared to
901 suppress overwriting the current in-kernel state. The bits are:
903 KVM_VCPUEVENT_VALID_NMI_PENDING - transfer nmi.pending to the kernel
904 KVM_VCPUEVENT_VALID_SIPI_VECTOR - transfer sipi_vector
905 KVM_VCPUEVENT_VALID_SMM - transfer the smi sub-struct.
907 If KVM_CAP_INTR_SHADOW is available, KVM_VCPUEVENT_VALID_SHADOW can be set in
908 the flags field to signal that interrupt.shadow contains a valid state and
909 shall be written into the VCPU.
911 KVM_VCPUEVENT_VALID_SMM can only be set if KVM_CAP_X86_SMM is available.
914 4.33 KVM_GET_DEBUGREGS
916 Capability: KVM_CAP_DEBUGREGS
919 Parameters: struct kvm_debugregs (out)
920 Returns: 0 on success, -1 on error
922 Reads debug registers from the vcpu.
924 struct kvm_debugregs {
933 4.34 KVM_SET_DEBUGREGS
935 Capability: KVM_CAP_DEBUGREGS
938 Parameters: struct kvm_debugregs (in)
939 Returns: 0 on success, -1 on error
941 Writes debug registers into the vcpu.
943 See KVM_GET_DEBUGREGS for the data structure. The flags field is unused
944 yet and must be cleared on entry.
947 4.35 KVM_SET_USER_MEMORY_REGION
949 Capability: KVM_CAP_USER_MEM
952 Parameters: struct kvm_userspace_memory_region (in)
953 Returns: 0 on success, -1 on error
955 struct kvm_userspace_memory_region {
958 __u64 guest_phys_addr;
959 __u64 memory_size; /* bytes */
960 __u64 userspace_addr; /* start of the userspace allocated memory */
963 /* for kvm_memory_region::flags */
964 #define KVM_MEM_LOG_DIRTY_PAGES (1UL << 0)
965 #define KVM_MEM_READONLY (1UL << 1)
967 This ioctl allows the user to create or modify a guest physical memory
968 slot. When changing an existing slot, it may be moved in the guest
969 physical memory space, or its flags may be modified. It may not be
970 resized. Slots may not overlap in guest physical address space.
971 Bits 0-15 of "slot" specifies the slot id and this value should be
972 less than the maximum number of user memory slots supported per VM.
973 The maximum allowed slots can be queried using KVM_CAP_NR_MEMSLOTS,
974 if this capability is supported by the architecture.
976 If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of "slot"
977 specifies the address space which is being modified. They must be
978 less than the value that KVM_CHECK_EXTENSION returns for the
979 KVM_CAP_MULTI_ADDRESS_SPACE capability. Slots in separate address spaces
980 are unrelated; the restriction on overlapping slots only applies within
983 Memory for the region is taken starting at the address denoted by the
984 field userspace_addr, which must point at user addressable memory for
985 the entire memory slot size. Any object may back this memory, including
986 anonymous memory, ordinary files, and hugetlbfs.
988 It is recommended that the lower 21 bits of guest_phys_addr and userspace_addr
989 be identical. This allows large pages in the guest to be backed by large
992 The flags field supports two flags: KVM_MEM_LOG_DIRTY_PAGES and
993 KVM_MEM_READONLY. The former can be set to instruct KVM to keep track of
994 writes to memory within the slot. See KVM_GET_DIRTY_LOG ioctl to know how to
995 use it. The latter can be set, if KVM_CAP_READONLY_MEM capability allows it,
996 to make a new slot read-only. In this case, writes to this memory will be
997 posted to userspace as KVM_EXIT_MMIO exits.
999 When the KVM_CAP_SYNC_MMU capability is available, changes in the backing of
1000 the memory region are automatically reflected into the guest. For example, an
1001 mmap() that affects the region will be made visible immediately. Another
1002 example is madvise(MADV_DROP).
1004 It is recommended to use this API instead of the KVM_SET_MEMORY_REGION ioctl.
1005 The KVM_SET_MEMORY_REGION does not allow fine grained control over memory
1006 allocation and is deprecated.
1009 4.36 KVM_SET_TSS_ADDR
1011 Capability: KVM_CAP_SET_TSS_ADDR
1014 Parameters: unsigned long tss_address (in)
1015 Returns: 0 on success, -1 on error
1017 This ioctl defines the physical address of a three-page region in the guest
1018 physical address space. The region must be within the first 4GB of the
1019 guest physical address space and must not conflict with any memory slot
1020 or any mmio address. The guest may malfunction if it accesses this memory
1023 This ioctl is required on Intel-based hosts. This is needed on Intel hardware
1024 because of a quirk in the virtualization implementation (see the internals
1025 documentation when it pops into existence).
1030 Capability: KVM_CAP_ENABLE_CAP, KVM_CAP_ENABLE_CAP_VM
1031 Architectures: x86 (only KVM_CAP_ENABLE_CAP_VM),
1032 mips (only KVM_CAP_ENABLE_CAP), ppc, s390
1033 Type: vcpu ioctl, vm ioctl (with KVM_CAP_ENABLE_CAP_VM)
1034 Parameters: struct kvm_enable_cap (in)
1035 Returns: 0 on success; -1 on error
1037 +Not all extensions are enabled by default. Using this ioctl the application
1038 can enable an extension, making it available to the guest.
1040 On systems that do not support this ioctl, it always fails. On systems that
1041 do support it, it only works for extensions that are supported for enablement.
1043 To check if a capability can be enabled, the KVM_CHECK_EXTENSION ioctl should
1046 struct kvm_enable_cap {
1050 The capability that is supposed to get enabled.
1054 A bitfield indicating future enhancements. Has to be 0 for now.
1058 Arguments for enabling a feature. If a feature needs initial values to
1059 function properly, this is the place to put them.
1064 The vcpu ioctl should be used for vcpu-specific capabilities, the vm ioctl
1065 for vm-wide capabilities.
1067 4.38 KVM_GET_MP_STATE
1069 Capability: KVM_CAP_MP_STATE
1070 Architectures: x86, s390, arm, arm64
1072 Parameters: struct kvm_mp_state (out)
1073 Returns: 0 on success; -1 on error
1075 struct kvm_mp_state {
1079 Returns the vcpu's current "multiprocessing state" (though also valid on
1080 uniprocessor guests).
1082 Possible values are:
1084 - KVM_MP_STATE_RUNNABLE: the vcpu is currently running [x86,arm/arm64]
1085 - KVM_MP_STATE_UNINITIALIZED: the vcpu is an application processor (AP)
1086 which has not yet received an INIT signal [x86]
1087 - KVM_MP_STATE_INIT_RECEIVED: the vcpu has received an INIT signal, and is
1088 now ready for a SIPI [x86]
1089 - KVM_MP_STATE_HALTED: the vcpu has executed a HLT instruction and
1090 is waiting for an interrupt [x86]
1091 - KVM_MP_STATE_SIPI_RECEIVED: the vcpu has just received a SIPI (vector
1092 accessible via KVM_GET_VCPU_EVENTS) [x86]
1093 - KVM_MP_STATE_STOPPED: the vcpu is stopped [s390,arm/arm64]
1094 - KVM_MP_STATE_CHECK_STOP: the vcpu is in a special error state [s390]
1095 - KVM_MP_STATE_OPERATING: the vcpu is operating (running or halted)
1097 - KVM_MP_STATE_LOAD: the vcpu is in a special load/startup state
1100 On x86, this ioctl is only useful after KVM_CREATE_IRQCHIP. Without an
1101 in-kernel irqchip, the multiprocessing state must be maintained by userspace on
1102 these architectures.
1106 The only states that are valid are KVM_MP_STATE_STOPPED and
1107 KVM_MP_STATE_RUNNABLE which reflect if the vcpu is paused or not.
1109 4.39 KVM_SET_MP_STATE
1111 Capability: KVM_CAP_MP_STATE
1112 Architectures: x86, s390, arm, arm64
1114 Parameters: struct kvm_mp_state (in)
1115 Returns: 0 on success; -1 on error
1117 Sets the vcpu's current "multiprocessing state"; see KVM_GET_MP_STATE for
1120 On x86, this ioctl is only useful after KVM_CREATE_IRQCHIP. Without an
1121 in-kernel irqchip, the multiprocessing state must be maintained by userspace on
1122 these architectures.
1126 The only states that are valid are KVM_MP_STATE_STOPPED and
1127 KVM_MP_STATE_RUNNABLE which reflect if the vcpu should be paused or not.
1129 4.40 KVM_SET_IDENTITY_MAP_ADDR
1131 Capability: KVM_CAP_SET_IDENTITY_MAP_ADDR
1134 Parameters: unsigned long identity (in)
1135 Returns: 0 on success, -1 on error
1137 This ioctl defines the physical address of a one-page region in the guest
1138 physical address space. The region must be within the first 4GB of the
1139 guest physical address space and must not conflict with any memory slot
1140 or any mmio address. The guest may malfunction if it accesses this memory
1143 Setting the address to 0 will result in resetting the address to its default
1146 This ioctl is required on Intel-based hosts. This is needed on Intel hardware
1147 because of a quirk in the virtualization implementation (see the internals
1148 documentation when it pops into existence).
1150 Fails if any VCPU has already been created.
1152 4.41 KVM_SET_BOOT_CPU_ID
1154 Capability: KVM_CAP_SET_BOOT_CPU_ID
1157 Parameters: unsigned long vcpu_id
1158 Returns: 0 on success, -1 on error
1160 Define which vcpu is the Bootstrap Processor (BSP). Values are the same
1161 as the vcpu id in KVM_CREATE_VCPU. If this ioctl is not called, the default
1167 Capability: KVM_CAP_XSAVE
1170 Parameters: struct kvm_xsave (out)
1171 Returns: 0 on success, -1 on error
1177 This ioctl would copy current vcpu's xsave struct to the userspace.
1182 Capability: KVM_CAP_XSAVE
1185 Parameters: struct kvm_xsave (in)
1186 Returns: 0 on success, -1 on error
1192 This ioctl would copy userspace's xsave struct to the kernel.
1197 Capability: KVM_CAP_XCRS
1200 Parameters: struct kvm_xcrs (out)
1201 Returns: 0 on success, -1 on error
1212 struct kvm_xcr xcrs[KVM_MAX_XCRS];
1216 This ioctl would copy current vcpu's xcrs to the userspace.
1221 Capability: KVM_CAP_XCRS
1224 Parameters: struct kvm_xcrs (in)
1225 Returns: 0 on success, -1 on error
1236 struct kvm_xcr xcrs[KVM_MAX_XCRS];
1240 This ioctl would set vcpu's xcr to the value userspace specified.
1243 4.46 KVM_GET_SUPPORTED_CPUID
1245 Capability: KVM_CAP_EXT_CPUID
1248 Parameters: struct kvm_cpuid2 (in/out)
1249 Returns: 0 on success, -1 on error
1254 struct kvm_cpuid_entry2 entries[0];
1257 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX BIT(0)
1258 #define KVM_CPUID_FLAG_STATEFUL_FUNC BIT(1)
1259 #define KVM_CPUID_FLAG_STATE_READ_NEXT BIT(2)
1261 struct kvm_cpuid_entry2 {
1272 This ioctl returns x86 cpuid features which are supported by both the hardware
1273 and kvm. Userspace can use the information returned by this ioctl to
1274 construct cpuid information (for KVM_SET_CPUID2) that is consistent with
1275 hardware, kernel, and userspace capabilities, and with user requirements (for
1276 example, the user may wish to constrain cpuid to emulate older hardware,
1277 or for feature consistency across a cluster).
1279 Userspace invokes KVM_GET_SUPPORTED_CPUID by passing a kvm_cpuid2 structure
1280 with the 'nent' field indicating the number of entries in the variable-size
1281 array 'entries'. If the number of entries is too low to describe the cpu
1282 capabilities, an error (E2BIG) is returned. If the number is too high,
1283 the 'nent' field is adjusted and an error (ENOMEM) is returned. If the
1284 number is just right, the 'nent' field is adjusted to the number of valid
1285 entries in the 'entries' array, which is then filled.
1287 The entries returned are the host cpuid as returned by the cpuid instruction,
1288 with unknown or unsupported features masked out. Some features (for example,
1289 x2apic), may not be present in the host cpu, but are exposed by kvm if it can
1290 emulate them efficiently. The fields in each entry are defined as follows:
1292 function: the eax value used to obtain the entry
1293 index: the ecx value used to obtain the entry (for entries that are
1295 flags: an OR of zero or more of the following:
1296 KVM_CPUID_FLAG_SIGNIFCANT_INDEX:
1297 if the index field is valid
1298 KVM_CPUID_FLAG_STATEFUL_FUNC:
1299 if cpuid for this function returns different values for successive
1300 invocations; there will be several entries with the same function,
1301 all with this flag set
1302 KVM_CPUID_FLAG_STATE_READ_NEXT:
1303 for KVM_CPUID_FLAG_STATEFUL_FUNC entries, set if this entry is
1304 the first entry to be read by a cpu
1305 eax, ebx, ecx, edx: the values returned by the cpuid instruction for
1306 this function/index combination
1308 The TSC deadline timer feature (CPUID leaf 1, ecx[24]) is always returned
1309 as false, since the feature depends on KVM_CREATE_IRQCHIP for local APIC
1310 support. Instead it is reported via
1312 ioctl(KVM_CHECK_EXTENSION, KVM_CAP_TSC_DEADLINE_TIMER)
1314 if that returns true and you use KVM_CREATE_IRQCHIP, or if you emulate the
1315 feature in userspace, then you can enable the feature for KVM_SET_CPUID2.
1318 4.47 KVM_PPC_GET_PVINFO
1320 Capability: KVM_CAP_PPC_GET_PVINFO
1323 Parameters: struct kvm_ppc_pvinfo (out)
1324 Returns: 0 on success, !0 on error
1326 struct kvm_ppc_pvinfo {
1332 This ioctl fetches PV specific information that need to be passed to the guest
1333 using the device tree or other means from vm context.
1335 The hcall array defines 4 instructions that make up a hypercall.
1337 If any additional field gets added to this structure later on, a bit for that
1338 additional piece of information will be set in the flags bitmap.
1340 The flags bitmap is defined as:
1342 /* the host supports the ePAPR idle hcall
1343 #define KVM_PPC_PVINFO_FLAGS_EV_IDLE (1<<0)
1345 4.52 KVM_SET_GSI_ROUTING
1347 Capability: KVM_CAP_IRQ_ROUTING
1348 Architectures: x86 s390 arm arm64
1350 Parameters: struct kvm_irq_routing (in)
1351 Returns: 0 on success, -1 on error
1353 Sets the GSI routing table entries, overwriting any previously set entries.
1355 On arm/arm64, GSI routing has the following limitation:
1356 - GSI routing does not apply to KVM_IRQ_LINE but only to KVM_IRQFD.
1358 struct kvm_irq_routing {
1361 struct kvm_irq_routing_entry entries[0];
1364 No flags are specified so far, the corresponding field must be set to zero.
1366 struct kvm_irq_routing_entry {
1372 struct kvm_irq_routing_irqchip irqchip;
1373 struct kvm_irq_routing_msi msi;
1374 struct kvm_irq_routing_s390_adapter adapter;
1375 struct kvm_irq_routing_hv_sint hv_sint;
1380 /* gsi routing entry types */
1381 #define KVM_IRQ_ROUTING_IRQCHIP 1
1382 #define KVM_IRQ_ROUTING_MSI 2
1383 #define KVM_IRQ_ROUTING_S390_ADAPTER 3
1384 #define KVM_IRQ_ROUTING_HV_SINT 4
1387 - KVM_MSI_VALID_DEVID: used along with KVM_IRQ_ROUTING_MSI routing entry
1388 type, specifies that the devid field contains a valid value. The per-VM
1389 KVM_CAP_MSI_DEVID capability advertises the requirement to provide
1390 the device ID. If this capability is not available, userspace should
1391 never set the KVM_MSI_VALID_DEVID flag as the ioctl might fail.
1394 struct kvm_irq_routing_irqchip {
1399 struct kvm_irq_routing_msi {
1409 If KVM_MSI_VALID_DEVID is set, devid contains a unique device identifier
1410 for the device that wrote the MSI message. For PCI, this is usually a
1411 BFD identifier in the lower 16 bits.
1413 On x86, address_hi is ignored unless the KVM_X2APIC_API_USE_32BIT_IDS
1414 feature of KVM_CAP_X2APIC_API capability is enabled. If it is enabled,
1415 address_hi bits 31-8 provide bits 31-8 of the destination id. Bits 7-0 of
1416 address_hi must be zero.
1418 struct kvm_irq_routing_s390_adapter {
1422 __u32 summary_offset;
1426 struct kvm_irq_routing_hv_sint {
1432 4.55 KVM_SET_TSC_KHZ
1434 Capability: KVM_CAP_TSC_CONTROL
1437 Parameters: virtual tsc_khz
1438 Returns: 0 on success, -1 on error
1440 Specifies the tsc frequency for the virtual machine. The unit of the
1444 4.56 KVM_GET_TSC_KHZ
1446 Capability: KVM_CAP_GET_TSC_KHZ
1450 Returns: virtual tsc-khz on success, negative value on error
1452 Returns the tsc frequency of the guest. The unit of the return value is
1453 KHz. If the host has unstable tsc this ioctl returns -EIO instead as an
1459 Capability: KVM_CAP_IRQCHIP
1462 Parameters: struct kvm_lapic_state (out)
1463 Returns: 0 on success, -1 on error
1465 #define KVM_APIC_REG_SIZE 0x400
1466 struct kvm_lapic_state {
1467 char regs[KVM_APIC_REG_SIZE];
1470 Reads the Local APIC registers and copies them into the input argument. The
1471 data format and layout are the same as documented in the architecture manual.
1473 If KVM_X2APIC_API_USE_32BIT_IDS feature of KVM_CAP_X2APIC_API is
1474 enabled, then the format of APIC_ID register depends on the APIC mode
1475 (reported by MSR_IA32_APICBASE) of its VCPU. x2APIC stores APIC ID in
1476 the APIC_ID register (bytes 32-35). xAPIC only allows an 8-bit APIC ID
1477 which is stored in bits 31-24 of the APIC register, or equivalently in
1478 byte 35 of struct kvm_lapic_state's regs field. KVM_GET_LAPIC must then
1479 be called after MSR_IA32_APICBASE has been set with KVM_SET_MSR.
1481 If KVM_X2APIC_API_USE_32BIT_IDS feature is disabled, struct kvm_lapic_state
1482 always uses xAPIC format.
1487 Capability: KVM_CAP_IRQCHIP
1490 Parameters: struct kvm_lapic_state (in)
1491 Returns: 0 on success, -1 on error
1493 #define KVM_APIC_REG_SIZE 0x400
1494 struct kvm_lapic_state {
1495 char regs[KVM_APIC_REG_SIZE];
1498 Copies the input argument into the Local APIC registers. The data format
1499 and layout are the same as documented in the architecture manual.
1501 The format of the APIC ID register (bytes 32-35 of struct kvm_lapic_state's
1502 regs field) depends on the state of the KVM_CAP_X2APIC_API capability.
1503 See the note in KVM_GET_LAPIC.
1508 Capability: KVM_CAP_IOEVENTFD
1511 Parameters: struct kvm_ioeventfd (in)
1512 Returns: 0 on success, !0 on error
1514 This ioctl attaches or detaches an ioeventfd to a legal pio/mmio address
1515 within the guest. A guest write in the registered address will signal the
1516 provided event instead of triggering an exit.
1518 struct kvm_ioeventfd {
1520 __u64 addr; /* legal pio/mmio address */
1521 __u32 len; /* 0, 1, 2, 4, or 8 bytes */
1527 For the special case of virtio-ccw devices on s390, the ioevent is matched
1528 to a subchannel/virtqueue tuple instead.
1530 The following flags are defined:
1532 #define KVM_IOEVENTFD_FLAG_DATAMATCH (1 << kvm_ioeventfd_flag_nr_datamatch)
1533 #define KVM_IOEVENTFD_FLAG_PIO (1 << kvm_ioeventfd_flag_nr_pio)
1534 #define KVM_IOEVENTFD_FLAG_DEASSIGN (1 << kvm_ioeventfd_flag_nr_deassign)
1535 #define KVM_IOEVENTFD_FLAG_VIRTIO_CCW_NOTIFY \
1536 (1 << kvm_ioeventfd_flag_nr_virtio_ccw_notify)
1538 If datamatch flag is set, the event will be signaled only if the written value
1539 to the registered address is equal to datamatch in struct kvm_ioeventfd.
1541 For virtio-ccw devices, addr contains the subchannel id and datamatch the
1544 With KVM_CAP_IOEVENTFD_ANY_LENGTH, a zero length ioeventfd is allowed, and
1545 the kernel will ignore the length of guest write and may get a faster vmexit.
1546 The speedup may only apply to specific architectures, but the ioeventfd will
1551 Capability: KVM_CAP_SW_TLB
1554 Parameters: struct kvm_dirty_tlb (in)
1555 Returns: 0 on success, -1 on error
1557 struct kvm_dirty_tlb {
1562 This must be called whenever userspace has changed an entry in the shared
1563 TLB, prior to calling KVM_RUN on the associated vcpu.
1565 The "bitmap" field is the userspace address of an array. This array
1566 consists of a number of bits, equal to the total number of TLB entries as
1567 determined by the last successful call to KVM_CONFIG_TLB, rounded up to the
1568 nearest multiple of 64.
1570 Each bit corresponds to one TLB entry, ordered the same as in the shared TLB
1573 The array is little-endian: the bit 0 is the least significant bit of the
1574 first byte, bit 8 is the least significant bit of the second byte, etc.
1575 This avoids any complications with differing word sizes.
1577 The "num_dirty" field is a performance hint for KVM to determine whether it
1578 should skip processing the bitmap and just invalidate everything. It must
1579 be set to the number of set bits in the bitmap.
1582 4.62 KVM_CREATE_SPAPR_TCE
1584 Capability: KVM_CAP_SPAPR_TCE
1585 Architectures: powerpc
1587 Parameters: struct kvm_create_spapr_tce (in)
1588 Returns: file descriptor for manipulating the created TCE table
1590 This creates a virtual TCE (translation control entry) table, which
1591 is an IOMMU for PAPR-style virtual I/O. It is used to translate
1592 logical addresses used in virtual I/O into guest physical addresses,
1593 and provides a scatter/gather capability for PAPR virtual I/O.
1595 /* for KVM_CAP_SPAPR_TCE */
1596 struct kvm_create_spapr_tce {
1601 The liobn field gives the logical IO bus number for which to create a
1602 TCE table. The window_size field specifies the size of the DMA window
1603 which this TCE table will translate - the table will contain one 64
1604 bit TCE entry for every 4kiB of the DMA window.
1606 When the guest issues an H_PUT_TCE hcall on a liobn for which a TCE
1607 table has been created using this ioctl(), the kernel will handle it
1608 in real mode, updating the TCE table. H_PUT_TCE calls for other
1609 liobns will cause a vm exit and must be handled by userspace.
1611 The return value is a file descriptor which can be passed to mmap(2)
1612 to map the created TCE table into userspace. This lets userspace read
1613 the entries written by kernel-handled H_PUT_TCE calls, and also lets
1614 userspace update the TCE table directly which is useful in some
1618 4.63 KVM_ALLOCATE_RMA
1620 Capability: KVM_CAP_PPC_RMA
1621 Architectures: powerpc
1623 Parameters: struct kvm_allocate_rma (out)
1624 Returns: file descriptor for mapping the allocated RMA
1626 This allocates a Real Mode Area (RMA) from the pool allocated at boot
1627 time by the kernel. An RMA is a physically-contiguous, aligned region
1628 of memory used on older POWER processors to provide the memory which
1629 will be accessed by real-mode (MMU off) accesses in a KVM guest.
1630 POWER processors support a set of sizes for the RMA that usually
1631 includes 64MB, 128MB, 256MB and some larger powers of two.
1633 /* for KVM_ALLOCATE_RMA */
1634 struct kvm_allocate_rma {
1638 The return value is a file descriptor which can be passed to mmap(2)
1639 to map the allocated RMA into userspace. The mapped area can then be
1640 passed to the KVM_SET_USER_MEMORY_REGION ioctl to establish it as the
1641 RMA for a virtual machine. The size of the RMA in bytes (which is
1642 fixed at host kernel boot time) is returned in the rma_size field of
1643 the argument structure.
1645 The KVM_CAP_PPC_RMA capability is 1 or 2 if the KVM_ALLOCATE_RMA ioctl
1646 is supported; 2 if the processor requires all virtual machines to have
1647 an RMA, or 1 if the processor can use an RMA but doesn't require it,
1648 because it supports the Virtual RMA (VRMA) facility.
1653 Capability: KVM_CAP_USER_NMI
1657 Returns: 0 on success, -1 on error
1659 Queues an NMI on the thread's vcpu. Note this is well defined only
1660 when KVM_CREATE_IRQCHIP has not been called, since this is an interface
1661 between the virtual cpu core and virtual local APIC. After KVM_CREATE_IRQCHIP
1662 has been called, this interface is completely emulated within the kernel.
1664 To use this to emulate the LINT1 input with KVM_CREATE_IRQCHIP, use the
1665 following algorithm:
1668 - read the local APIC's state (KVM_GET_LAPIC)
1669 - check whether changing LINT1 will queue an NMI (see the LVT entry for LINT1)
1670 - if so, issue KVM_NMI
1673 Some guests configure the LINT1 NMI input to cause a panic, aiding in
1677 4.65 KVM_S390_UCAS_MAP
1679 Capability: KVM_CAP_S390_UCONTROL
1682 Parameters: struct kvm_s390_ucas_mapping (in)
1683 Returns: 0 in case of success
1685 The parameter is defined like this:
1686 struct kvm_s390_ucas_mapping {
1692 This ioctl maps the memory at "user_addr" with the length "length" to
1693 the vcpu's address space starting at "vcpu_addr". All parameters need to
1694 be aligned by 1 megabyte.
1697 4.66 KVM_S390_UCAS_UNMAP
1699 Capability: KVM_CAP_S390_UCONTROL
1702 Parameters: struct kvm_s390_ucas_mapping (in)
1703 Returns: 0 in case of success
1705 The parameter is defined like this:
1706 struct kvm_s390_ucas_mapping {
1712 This ioctl unmaps the memory in the vcpu's address space starting at
1713 "vcpu_addr" with the length "length". The field "user_addr" is ignored.
1714 All parameters need to be aligned by 1 megabyte.
1717 4.67 KVM_S390_VCPU_FAULT
1719 Capability: KVM_CAP_S390_UCONTROL
1722 Parameters: vcpu absolute address (in)
1723 Returns: 0 in case of success
1725 This call creates a page table entry on the virtual cpu's address space
1726 (for user controlled virtual machines) or the virtual machine's address
1727 space (for regular virtual machines). This only works for minor faults,
1728 thus it's recommended to access subject memory page via the user page
1729 table upfront. This is useful to handle validity intercepts for user
1730 controlled virtual machines to fault in the virtual cpu's lowcore pages
1731 prior to calling the KVM_RUN ioctl.
1734 4.68 KVM_SET_ONE_REG
1736 Capability: KVM_CAP_ONE_REG
1739 Parameters: struct kvm_one_reg (in)
1740 Returns: 0 on success, negative value on failure
1742 struct kvm_one_reg {
1747 Using this ioctl, a single vcpu register can be set to a specific value
1748 defined by user space with the passed in struct kvm_one_reg, where id
1749 refers to the register identifier as described below and addr is a pointer
1750 to a variable with the respective size. There can be architecture agnostic
1751 and architecture specific registers. Each have their own range of operation
1752 and their own constants and width. To keep track of the implemented
1753 registers, find a list below:
1755 Arch | Register | Width (bits)
1757 PPC | KVM_REG_PPC_HIOR | 64
1758 PPC | KVM_REG_PPC_IAC1 | 64
1759 PPC | KVM_REG_PPC_IAC2 | 64
1760 PPC | KVM_REG_PPC_IAC3 | 64
1761 PPC | KVM_REG_PPC_IAC4 | 64
1762 PPC | KVM_REG_PPC_DAC1 | 64
1763 PPC | KVM_REG_PPC_DAC2 | 64
1764 PPC | KVM_REG_PPC_DABR | 64
1765 PPC | KVM_REG_PPC_DSCR | 64
1766 PPC | KVM_REG_PPC_PURR | 64
1767 PPC | KVM_REG_PPC_SPURR | 64
1768 PPC | KVM_REG_PPC_DAR | 64
1769 PPC | KVM_REG_PPC_DSISR | 32
1770 PPC | KVM_REG_PPC_AMR | 64
1771 PPC | KVM_REG_PPC_UAMOR | 64
1772 PPC | KVM_REG_PPC_MMCR0 | 64
1773 PPC | KVM_REG_PPC_MMCR1 | 64
1774 PPC | KVM_REG_PPC_MMCRA | 64
1775 PPC | KVM_REG_PPC_MMCR2 | 64
1776 PPC | KVM_REG_PPC_MMCRS | 64
1777 PPC | KVM_REG_PPC_SIAR | 64
1778 PPC | KVM_REG_PPC_SDAR | 64
1779 PPC | KVM_REG_PPC_SIER | 64
1780 PPC | KVM_REG_PPC_PMC1 | 32
1781 PPC | KVM_REG_PPC_PMC2 | 32
1782 PPC | KVM_REG_PPC_PMC3 | 32
1783 PPC | KVM_REG_PPC_PMC4 | 32
1784 PPC | KVM_REG_PPC_PMC5 | 32
1785 PPC | KVM_REG_PPC_PMC6 | 32
1786 PPC | KVM_REG_PPC_PMC7 | 32
1787 PPC | KVM_REG_PPC_PMC8 | 32
1788 PPC | KVM_REG_PPC_FPR0 | 64
1790 PPC | KVM_REG_PPC_FPR31 | 64
1791 PPC | KVM_REG_PPC_VR0 | 128
1793 PPC | KVM_REG_PPC_VR31 | 128
1794 PPC | KVM_REG_PPC_VSR0 | 128
1796 PPC | KVM_REG_PPC_VSR31 | 128
1797 PPC | KVM_REG_PPC_FPSCR | 64
1798 PPC | KVM_REG_PPC_VSCR | 32
1799 PPC | KVM_REG_PPC_VPA_ADDR | 64
1800 PPC | KVM_REG_PPC_VPA_SLB | 128
1801 PPC | KVM_REG_PPC_VPA_DTL | 128
1802 PPC | KVM_REG_PPC_EPCR | 32
1803 PPC | KVM_REG_PPC_EPR | 32
1804 PPC | KVM_REG_PPC_TCR | 32
1805 PPC | KVM_REG_PPC_TSR | 32
1806 PPC | KVM_REG_PPC_OR_TSR | 32
1807 PPC | KVM_REG_PPC_CLEAR_TSR | 32
1808 PPC | KVM_REG_PPC_MAS0 | 32
1809 PPC | KVM_REG_PPC_MAS1 | 32
1810 PPC | KVM_REG_PPC_MAS2 | 64
1811 PPC | KVM_REG_PPC_MAS7_3 | 64
1812 PPC | KVM_REG_PPC_MAS4 | 32
1813 PPC | KVM_REG_PPC_MAS6 | 32
1814 PPC | KVM_REG_PPC_MMUCFG | 32
1815 PPC | KVM_REG_PPC_TLB0CFG | 32
1816 PPC | KVM_REG_PPC_TLB1CFG | 32
1817 PPC | KVM_REG_PPC_TLB2CFG | 32
1818 PPC | KVM_REG_PPC_TLB3CFG | 32
1819 PPC | KVM_REG_PPC_TLB0PS | 32
1820 PPC | KVM_REG_PPC_TLB1PS | 32
1821 PPC | KVM_REG_PPC_TLB2PS | 32
1822 PPC | KVM_REG_PPC_TLB3PS | 32
1823 PPC | KVM_REG_PPC_EPTCFG | 32
1824 PPC | KVM_REG_PPC_ICP_STATE | 64
1825 PPC | KVM_REG_PPC_TB_OFFSET | 64
1826 PPC | KVM_REG_PPC_SPMC1 | 32
1827 PPC | KVM_REG_PPC_SPMC2 | 32
1828 PPC | KVM_REG_PPC_IAMR | 64
1829 PPC | KVM_REG_PPC_TFHAR | 64
1830 PPC | KVM_REG_PPC_TFIAR | 64
1831 PPC | KVM_REG_PPC_TEXASR | 64
1832 PPC | KVM_REG_PPC_FSCR | 64
1833 PPC | KVM_REG_PPC_PSPB | 32
1834 PPC | KVM_REG_PPC_EBBHR | 64
1835 PPC | KVM_REG_PPC_EBBRR | 64
1836 PPC | KVM_REG_PPC_BESCR | 64
1837 PPC | KVM_REG_PPC_TAR | 64
1838 PPC | KVM_REG_PPC_DPDES | 64
1839 PPC | KVM_REG_PPC_DAWR | 64
1840 PPC | KVM_REG_PPC_DAWRX | 64
1841 PPC | KVM_REG_PPC_CIABR | 64
1842 PPC | KVM_REG_PPC_IC | 64
1843 PPC | KVM_REG_PPC_VTB | 64
1844 PPC | KVM_REG_PPC_CSIGR | 64
1845 PPC | KVM_REG_PPC_TACR | 64
1846 PPC | KVM_REG_PPC_TCSCR | 64
1847 PPC | KVM_REG_PPC_PID | 64
1848 PPC | KVM_REG_PPC_ACOP | 64
1849 PPC | KVM_REG_PPC_VRSAVE | 32
1850 PPC | KVM_REG_PPC_LPCR | 32
1851 PPC | KVM_REG_PPC_LPCR_64 | 64
1852 PPC | KVM_REG_PPC_PPR | 64
1853 PPC | KVM_REG_PPC_ARCH_COMPAT | 32
1854 PPC | KVM_REG_PPC_DABRX | 32
1855 PPC | KVM_REG_PPC_WORT | 64
1856 PPC | KVM_REG_PPC_SPRG9 | 64
1857 PPC | KVM_REG_PPC_DBSR | 32
1858 PPC | KVM_REG_PPC_TIDR | 64
1859 PPC | KVM_REG_PPC_PSSCR | 64
1860 PPC | KVM_REG_PPC_DEC_EXPIRY | 64
1861 PPC | KVM_REG_PPC_TM_GPR0 | 64
1863 PPC | KVM_REG_PPC_TM_GPR31 | 64
1864 PPC | KVM_REG_PPC_TM_VSR0 | 128
1866 PPC | KVM_REG_PPC_TM_VSR63 | 128
1867 PPC | KVM_REG_PPC_TM_CR | 64
1868 PPC | KVM_REG_PPC_TM_LR | 64
1869 PPC | KVM_REG_PPC_TM_CTR | 64
1870 PPC | KVM_REG_PPC_TM_FPSCR | 64
1871 PPC | KVM_REG_PPC_TM_AMR | 64
1872 PPC | KVM_REG_PPC_TM_PPR | 64
1873 PPC | KVM_REG_PPC_TM_VRSAVE | 64
1874 PPC | KVM_REG_PPC_TM_VSCR | 32
1875 PPC | KVM_REG_PPC_TM_DSCR | 64
1876 PPC | KVM_REG_PPC_TM_TAR | 64
1877 PPC | KVM_REG_PPC_TM_XER | 64
1879 MIPS | KVM_REG_MIPS_R0 | 64
1881 MIPS | KVM_REG_MIPS_R31 | 64
1882 MIPS | KVM_REG_MIPS_HI | 64
1883 MIPS | KVM_REG_MIPS_LO | 64
1884 MIPS | KVM_REG_MIPS_PC | 64
1885 MIPS | KVM_REG_MIPS_CP0_INDEX | 32
1886 MIPS | KVM_REG_MIPS_CP0_ENTRYLO0 | 64
1887 MIPS | KVM_REG_MIPS_CP0_ENTRYLO1 | 64
1888 MIPS | KVM_REG_MIPS_CP0_CONTEXT | 64
1889 MIPS | KVM_REG_MIPS_CP0_CONTEXTCONFIG| 32
1890 MIPS | KVM_REG_MIPS_CP0_USERLOCAL | 64
1891 MIPS | KVM_REG_MIPS_CP0_XCONTEXTCONFIG| 64
1892 MIPS | KVM_REG_MIPS_CP0_PAGEMASK | 32
1893 MIPS | KVM_REG_MIPS_CP0_PAGEGRAIN | 32
1894 MIPS | KVM_REG_MIPS_CP0_SEGCTL0 | 64
1895 MIPS | KVM_REG_MIPS_CP0_SEGCTL1 | 64
1896 MIPS | KVM_REG_MIPS_CP0_SEGCTL2 | 64
1897 MIPS | KVM_REG_MIPS_CP0_PWBASE | 64
1898 MIPS | KVM_REG_MIPS_CP0_PWFIELD | 64
1899 MIPS | KVM_REG_MIPS_CP0_PWSIZE | 64
1900 MIPS | KVM_REG_MIPS_CP0_WIRED | 32
1901 MIPS | KVM_REG_MIPS_CP0_PWCTL | 32
1902 MIPS | KVM_REG_MIPS_CP0_HWRENA | 32
1903 MIPS | KVM_REG_MIPS_CP0_BADVADDR | 64
1904 MIPS | KVM_REG_MIPS_CP0_BADINSTR | 32
1905 MIPS | KVM_REG_MIPS_CP0_BADINSTRP | 32
1906 MIPS | KVM_REG_MIPS_CP0_COUNT | 32
1907 MIPS | KVM_REG_MIPS_CP0_ENTRYHI | 64
1908 MIPS | KVM_REG_MIPS_CP0_COMPARE | 32
1909 MIPS | KVM_REG_MIPS_CP0_STATUS | 32
1910 MIPS | KVM_REG_MIPS_CP0_INTCTL | 32
1911 MIPS | KVM_REG_MIPS_CP0_CAUSE | 32
1912 MIPS | KVM_REG_MIPS_CP0_EPC | 64
1913 MIPS | KVM_REG_MIPS_CP0_PRID | 32
1914 MIPS | KVM_REG_MIPS_CP0_EBASE | 64
1915 MIPS | KVM_REG_MIPS_CP0_CONFIG | 32
1916 MIPS | KVM_REG_MIPS_CP0_CONFIG1 | 32
1917 MIPS | KVM_REG_MIPS_CP0_CONFIG2 | 32
1918 MIPS | KVM_REG_MIPS_CP0_CONFIG3 | 32
1919 MIPS | KVM_REG_MIPS_CP0_CONFIG4 | 32
1920 MIPS | KVM_REG_MIPS_CP0_CONFIG5 | 32
1921 MIPS | KVM_REG_MIPS_CP0_CONFIG7 | 32
1922 MIPS | KVM_REG_MIPS_CP0_XCONTEXT | 64
1923 MIPS | KVM_REG_MIPS_CP0_ERROREPC | 64
1924 MIPS | KVM_REG_MIPS_CP0_KSCRATCH1 | 64
1925 MIPS | KVM_REG_MIPS_CP0_KSCRATCH2 | 64
1926 MIPS | KVM_REG_MIPS_CP0_KSCRATCH3 | 64
1927 MIPS | KVM_REG_MIPS_CP0_KSCRATCH4 | 64
1928 MIPS | KVM_REG_MIPS_CP0_KSCRATCH5 | 64
1929 MIPS | KVM_REG_MIPS_CP0_KSCRATCH6 | 64
1930 MIPS | KVM_REG_MIPS_CP0_MAAR(0..63) | 64
1931 MIPS | KVM_REG_MIPS_COUNT_CTL | 64
1932 MIPS | KVM_REG_MIPS_COUNT_RESUME | 64
1933 MIPS | KVM_REG_MIPS_COUNT_HZ | 64
1934 MIPS | KVM_REG_MIPS_FPR_32(0..31) | 32
1935 MIPS | KVM_REG_MIPS_FPR_64(0..31) | 64
1936 MIPS | KVM_REG_MIPS_VEC_128(0..31) | 128
1937 MIPS | KVM_REG_MIPS_FCR_IR | 32
1938 MIPS | KVM_REG_MIPS_FCR_CSR | 32
1939 MIPS | KVM_REG_MIPS_MSA_IR | 32
1940 MIPS | KVM_REG_MIPS_MSA_CSR | 32
1942 ARM registers are mapped using the lower 32 bits. The upper 16 of that
1943 is the register group type, or coprocessor number:
1945 ARM core registers have the following id bit patterns:
1946 0x4020 0000 0010 <index into the kvm_regs struct:16>
1948 ARM 32-bit CP15 registers have the following id bit patterns:
1949 0x4020 0000 000F <zero:1> <crn:4> <crm:4> <opc1:4> <opc2:3>
1951 ARM 64-bit CP15 registers have the following id bit patterns:
1952 0x4030 0000 000F <zero:1> <zero:4> <crm:4> <opc1:4> <zero:3>
1954 ARM CCSIDR registers are demultiplexed by CSSELR value:
1955 0x4020 0000 0011 00 <csselr:8>
1957 ARM 32-bit VFP control registers have the following id bit patterns:
1958 0x4020 0000 0012 1 <regno:12>
1960 ARM 64-bit FP registers have the following id bit patterns:
1961 0x4030 0000 0012 0 <regno:12>
1963 ARM firmware pseudo-registers have the following bit pattern:
1964 0x4030 0000 0014 <regno:16>
1967 arm64 registers are mapped using the lower 32 bits. The upper 16 of
1968 that is the register group type, or coprocessor number:
1970 arm64 core/FP-SIMD registers have the following id bit patterns. Note
1971 that the size of the access is variable, as the kvm_regs structure
1972 contains elements ranging from 32 to 128 bits. The index is a 32bit
1973 value in the kvm_regs structure seen as a 32bit array.
1974 0x60x0 0000 0010 <index into the kvm_regs struct:16>
1976 arm64 CCSIDR registers are demultiplexed by CSSELR value:
1977 0x6020 0000 0011 00 <csselr:8>
1979 arm64 system registers have the following id bit patterns:
1980 0x6030 0000 0013 <op0:2> <op1:3> <crn:4> <crm:4> <op2:3>
1982 arm64 firmware pseudo-registers have the following bit pattern:
1983 0x6030 0000 0014 <regno:16>
1986 MIPS registers are mapped using the lower 32 bits. The upper 16 of that is
1987 the register group type:
1989 MIPS core registers (see above) have the following id bit patterns:
1990 0x7030 0000 0000 <reg:16>
1992 MIPS CP0 registers (see KVM_REG_MIPS_CP0_* above) have the following id bit
1993 patterns depending on whether they're 32-bit or 64-bit registers:
1994 0x7020 0000 0001 00 <reg:5> <sel:3> (32-bit)
1995 0x7030 0000 0001 00 <reg:5> <sel:3> (64-bit)
1997 Note: KVM_REG_MIPS_CP0_ENTRYLO0 and KVM_REG_MIPS_CP0_ENTRYLO1 are the MIPS64
1998 versions of the EntryLo registers regardless of the word size of the host
1999 hardware, host kernel, guest, and whether XPA is present in the guest, i.e.
2000 with the RI and XI bits (if they exist) in bits 63 and 62 respectively, and
2001 the PFNX field starting at bit 30.
2003 MIPS MAARs (see KVM_REG_MIPS_CP0_MAAR(*) above) have the following id bit
2005 0x7030 0000 0001 01 <reg:8>
2007 MIPS KVM control registers (see above) have the following id bit patterns:
2008 0x7030 0000 0002 <reg:16>
2010 MIPS FPU registers (see KVM_REG_MIPS_FPR_{32,64}() above) have the following
2011 id bit patterns depending on the size of the register being accessed. They are
2012 always accessed according to the current guest FPU mode (Status.FR and
2013 Config5.FRE), i.e. as the guest would see them, and they become unpredictable
2014 if the guest FPU mode is changed. MIPS SIMD Architecture (MSA) vector
2015 registers (see KVM_REG_MIPS_VEC_128() above) have similar patterns as they
2016 overlap the FPU registers:
2017 0x7020 0000 0003 00 <0:3> <reg:5> (32-bit FPU registers)
2018 0x7030 0000 0003 00 <0:3> <reg:5> (64-bit FPU registers)
2019 0x7040 0000 0003 00 <0:3> <reg:5> (128-bit MSA vector registers)
2021 MIPS FPU control registers (see KVM_REG_MIPS_FCR_{IR,CSR} above) have the
2022 following id bit patterns:
2023 0x7020 0000 0003 01 <0:3> <reg:5>
2025 MIPS MSA control registers (see KVM_REG_MIPS_MSA_{IR,CSR} above) have the
2026 following id bit patterns:
2027 0x7020 0000 0003 02 <0:3> <reg:5>
2030 4.69 KVM_GET_ONE_REG
2032 Capability: KVM_CAP_ONE_REG
2035 Parameters: struct kvm_one_reg (in and out)
2036 Returns: 0 on success, negative value on failure
2038 This ioctl allows to receive the value of a single register implemented
2039 in a vcpu. The register to read is indicated by the "id" field of the
2040 kvm_one_reg struct passed in. On success, the register value can be found
2041 at the memory location pointed to by "addr".
2043 The list of registers accessible using this interface is identical to the
2047 4.70 KVM_KVMCLOCK_CTRL
2049 Capability: KVM_CAP_KVMCLOCK_CTRL
2050 Architectures: Any that implement pvclocks (currently x86 only)
2053 Returns: 0 on success, -1 on error
2055 This signals to the host kernel that the specified guest is being paused by
2056 userspace. The host will set a flag in the pvclock structure that is checked
2057 from the soft lockup watchdog. The flag is part of the pvclock structure that
2058 is shared between guest and host, specifically the second bit of the flags
2059 field of the pvclock_vcpu_time_info structure. It will be set exclusively by
2060 the host and read/cleared exclusively by the guest. The guest operation of
2061 checking and clearing the flag must an atomic operation so
2062 load-link/store-conditional, or equivalent must be used. There are two cases
2063 where the guest will clear the flag: when the soft lockup watchdog timer resets
2064 itself or when a soft lockup is detected. This ioctl can be called any time
2065 after pausing the vcpu, but before it is resumed.
2070 Capability: KVM_CAP_SIGNAL_MSI
2071 Architectures: x86 arm arm64
2073 Parameters: struct kvm_msi (in)
2074 Returns: >0 on delivery, 0 if guest blocked the MSI, and -1 on error
2076 Directly inject a MSI message. Only valid with in-kernel irqchip that handles
2088 flags: KVM_MSI_VALID_DEVID: devid contains a valid value. The per-VM
2089 KVM_CAP_MSI_DEVID capability advertises the requirement to provide
2090 the device ID. If this capability is not available, userspace
2091 should never set the KVM_MSI_VALID_DEVID flag as the ioctl might fail.
2093 If KVM_MSI_VALID_DEVID is set, devid contains a unique device identifier
2094 for the device that wrote the MSI message. For PCI, this is usually a
2095 BFD identifier in the lower 16 bits.
2097 On x86, address_hi is ignored unless the KVM_X2APIC_API_USE_32BIT_IDS
2098 feature of KVM_CAP_X2APIC_API capability is enabled. If it is enabled,
2099 address_hi bits 31-8 provide bits 31-8 of the destination id. Bits 7-0 of
2100 address_hi must be zero.
2103 4.71 KVM_CREATE_PIT2
2105 Capability: KVM_CAP_PIT2
2108 Parameters: struct kvm_pit_config (in)
2109 Returns: 0 on success, -1 on error
2111 Creates an in-kernel device model for the i8254 PIT. This call is only valid
2112 after enabling in-kernel irqchip support via KVM_CREATE_IRQCHIP. The following
2113 parameters have to be passed:
2115 struct kvm_pit_config {
2122 #define KVM_PIT_SPEAKER_DUMMY 1 /* emulate speaker port stub */
2124 PIT timer interrupts may use a per-VM kernel thread for injection. If it
2125 exists, this thread will have a name of the following pattern:
2127 kvm-pit/<owner-process-pid>
2129 When running a guest with elevated priorities, the scheduling parameters of
2130 this thread may have to be adjusted accordingly.
2132 This IOCTL replaces the obsolete KVM_CREATE_PIT.
2137 Capability: KVM_CAP_PIT_STATE2
2140 Parameters: struct kvm_pit_state2 (out)
2141 Returns: 0 on success, -1 on error
2143 Retrieves the state of the in-kernel PIT model. Only valid after
2144 KVM_CREATE_PIT2. The state is returned in the following structure:
2146 struct kvm_pit_state2 {
2147 struct kvm_pit_channel_state channels[3];
2154 /* disable PIT in HPET legacy mode */
2155 #define KVM_PIT_FLAGS_HPET_LEGACY 0x00000001
2157 This IOCTL replaces the obsolete KVM_GET_PIT.
2162 Capability: KVM_CAP_PIT_STATE2
2165 Parameters: struct kvm_pit_state2 (in)
2166 Returns: 0 on success, -1 on error
2168 Sets the state of the in-kernel PIT model. Only valid after KVM_CREATE_PIT2.
2169 See KVM_GET_PIT2 for details on struct kvm_pit_state2.
2171 This IOCTL replaces the obsolete KVM_SET_PIT.
2174 4.74 KVM_PPC_GET_SMMU_INFO
2176 Capability: KVM_CAP_PPC_GET_SMMU_INFO
2177 Architectures: powerpc
2180 Returns: 0 on success, -1 on error
2182 This populates and returns a structure describing the features of
2183 the "Server" class MMU emulation supported by KVM.
2184 This can in turn be used by userspace to generate the appropriate
2185 device-tree properties for the guest operating system.
2187 The structure contains some global information, followed by an
2188 array of supported segment page sizes:
2190 struct kvm_ppc_smmu_info {
2194 struct kvm_ppc_one_seg_page_size sps[KVM_PPC_PAGE_SIZES_MAX_SZ];
2197 The supported flags are:
2199 - KVM_PPC_PAGE_SIZES_REAL:
2200 When that flag is set, guest page sizes must "fit" the backing
2201 store page sizes. When not set, any page size in the list can
2202 be used regardless of how they are backed by userspace.
2204 - KVM_PPC_1T_SEGMENTS
2205 The emulated MMU supports 1T segments in addition to the
2208 The "slb_size" field indicates how many SLB entries are supported
2210 The "sps" array contains 8 entries indicating the supported base
2211 page sizes for a segment in increasing order. Each entry is defined
2214 struct kvm_ppc_one_seg_page_size {
2215 __u32 page_shift; /* Base page shift of segment (or 0) */
2216 __u32 slb_enc; /* SLB encoding for BookS */
2217 struct kvm_ppc_one_page_size enc[KVM_PPC_PAGE_SIZES_MAX_SZ];
2220 An entry with a "page_shift" of 0 is unused. Because the array is
2221 organized in increasing order, a lookup can stop when encoutering
2224 The "slb_enc" field provides the encoding to use in the SLB for the
2225 page size. The bits are in positions such as the value can directly
2226 be OR'ed into the "vsid" argument of the slbmte instruction.
2228 The "enc" array is a list which for each of those segment base page
2229 size provides the list of supported actual page sizes (which can be
2230 only larger or equal to the base page size), along with the
2231 corresponding encoding in the hash PTE. Similarly, the array is
2232 8 entries sorted by increasing sizes and an entry with a "0" shift
2233 is an empty entry and a terminator:
2235 struct kvm_ppc_one_page_size {
2236 __u32 page_shift; /* Page shift (or 0) */
2237 __u32 pte_enc; /* Encoding in the HPTE (>>12) */
2240 The "pte_enc" field provides a value that can OR'ed into the hash
2241 PTE's RPN field (ie, it needs to be shifted left by 12 to OR it
2242 into the hash PTE second double word).
2246 Capability: KVM_CAP_IRQFD
2247 Architectures: x86 s390 arm arm64
2249 Parameters: struct kvm_irqfd (in)
2250 Returns: 0 on success, -1 on error
2252 Allows setting an eventfd to directly trigger a guest interrupt.
2253 kvm_irqfd.fd specifies the file descriptor to use as the eventfd and
2254 kvm_irqfd.gsi specifies the irqchip pin toggled by this event. When
2255 an event is triggered on the eventfd, an interrupt is injected into
2256 the guest using the specified gsi pin. The irqfd is removed using
2257 the KVM_IRQFD_FLAG_DEASSIGN flag, specifying both kvm_irqfd.fd
2260 With KVM_CAP_IRQFD_RESAMPLE, KVM_IRQFD supports a de-assert and notify
2261 mechanism allowing emulation of level-triggered, irqfd-based
2262 interrupts. When KVM_IRQFD_FLAG_RESAMPLE is set the user must pass an
2263 additional eventfd in the kvm_irqfd.resamplefd field. When operating
2264 in resample mode, posting of an interrupt through kvm_irq.fd asserts
2265 the specified gsi in the irqchip. When the irqchip is resampled, such
2266 as from an EOI, the gsi is de-asserted and the user is notified via
2267 kvm_irqfd.resamplefd. It is the user's responsibility to re-queue
2268 the interrupt if the device making use of it still requires service.
2269 Note that closing the resamplefd is not sufficient to disable the
2270 irqfd. The KVM_IRQFD_FLAG_RESAMPLE is only necessary on assignment
2271 and need not be specified with KVM_IRQFD_FLAG_DEASSIGN.
2273 On arm/arm64, gsi routing being supported, the following can happen:
2274 - in case no routing entry is associated to this gsi, injection fails
2275 - in case the gsi is associated to an irqchip routing entry,
2276 irqchip.pin + 32 corresponds to the injected SPI ID.
2277 - in case the gsi is associated to an MSI routing entry, the MSI
2278 message and device ID are translated into an LPI (support restricted
2279 to GICv3 ITS in-kernel emulation).
2281 4.76 KVM_PPC_ALLOCATE_HTAB
2283 Capability: KVM_CAP_PPC_ALLOC_HTAB
2284 Architectures: powerpc
2286 Parameters: Pointer to u32 containing hash table order (in/out)
2287 Returns: 0 on success, -1 on error
2289 This requests the host kernel to allocate an MMU hash table for a
2290 guest using the PAPR paravirtualization interface. This only does
2291 anything if the kernel is configured to use the Book 3S HV style of
2292 virtualization. Otherwise the capability doesn't exist and the ioctl
2293 returns an ENOTTY error. The rest of this description assumes Book 3S
2296 There must be no vcpus running when this ioctl is called; if there
2297 are, it will do nothing and return an EBUSY error.
2299 The parameter is a pointer to a 32-bit unsigned integer variable
2300 containing the order (log base 2) of the desired size of the hash
2301 table, which must be between 18 and 46. On successful return from the
2302 ioctl, the value will not be changed by the kernel.
2304 If no hash table has been allocated when any vcpu is asked to run
2305 (with the KVM_RUN ioctl), the host kernel will allocate a
2306 default-sized hash table (16 MB).
2308 If this ioctl is called when a hash table has already been allocated,
2309 with a different order from the existing hash table, the existing hash
2310 table will be freed and a new one allocated. If this is ioctl is
2311 called when a hash table has already been allocated of the same order
2312 as specified, the kernel will clear out the existing hash table (zero
2313 all HPTEs). In either case, if the guest is using the virtualized
2314 real-mode area (VRMA) facility, the kernel will re-create the VMRA
2315 HPTEs on the next KVM_RUN of any vcpu.
2317 4.77 KVM_S390_INTERRUPT
2321 Type: vm ioctl, vcpu ioctl
2322 Parameters: struct kvm_s390_interrupt (in)
2323 Returns: 0 on success, -1 on error
2325 Allows to inject an interrupt to the guest. Interrupts can be floating
2326 (vm ioctl) or per cpu (vcpu ioctl), depending on the interrupt type.
2328 Interrupt parameters are passed via kvm_s390_interrupt:
2330 struct kvm_s390_interrupt {
2336 type can be one of the following:
2338 KVM_S390_SIGP_STOP (vcpu) - sigp stop; optional flags in parm
2339 KVM_S390_PROGRAM_INT (vcpu) - program check; code in parm
2340 KVM_S390_SIGP_SET_PREFIX (vcpu) - sigp set prefix; prefix address in parm
2341 KVM_S390_RESTART (vcpu) - restart
2342 KVM_S390_INT_CLOCK_COMP (vcpu) - clock comparator interrupt
2343 KVM_S390_INT_CPU_TIMER (vcpu) - CPU timer interrupt
2344 KVM_S390_INT_VIRTIO (vm) - virtio external interrupt; external interrupt
2345 parameters in parm and parm64
2346 KVM_S390_INT_SERVICE (vm) - sclp external interrupt; sclp parameter in parm
2347 KVM_S390_INT_EMERGENCY (vcpu) - sigp emergency; source cpu in parm
2348 KVM_S390_INT_EXTERNAL_CALL (vcpu) - sigp external call; source cpu in parm
2349 KVM_S390_INT_IO(ai,cssid,ssid,schid) (vm) - compound value to indicate an
2350 I/O interrupt (ai - adapter interrupt; cssid,ssid,schid - subchannel);
2351 I/O interruption parameters in parm (subchannel) and parm64 (intparm,
2352 interruption subclass)
2353 KVM_S390_MCHK (vm, vcpu) - machine check interrupt; cr 14 bits in parm,
2354 machine check interrupt code in parm64 (note that
2355 machine checks needing further payload are not
2356 supported by this ioctl)
2358 Note that the vcpu ioctl is asynchronous to vcpu execution.
2360 4.78 KVM_PPC_GET_HTAB_FD
2362 Capability: KVM_CAP_PPC_HTAB_FD
2363 Architectures: powerpc
2365 Parameters: Pointer to struct kvm_get_htab_fd (in)
2366 Returns: file descriptor number (>= 0) on success, -1 on error
2368 This returns a file descriptor that can be used either to read out the
2369 entries in the guest's hashed page table (HPT), or to write entries to
2370 initialize the HPT. The returned fd can only be written to if the
2371 KVM_GET_HTAB_WRITE bit is set in the flags field of the argument, and
2372 can only be read if that bit is clear. The argument struct looks like
2375 /* For KVM_PPC_GET_HTAB_FD */
2376 struct kvm_get_htab_fd {
2382 /* Values for kvm_get_htab_fd.flags */
2383 #define KVM_GET_HTAB_BOLTED_ONLY ((__u64)0x1)
2384 #define KVM_GET_HTAB_WRITE ((__u64)0x2)
2386 The `start_index' field gives the index in the HPT of the entry at
2387 which to start reading. It is ignored when writing.
2389 Reads on the fd will initially supply information about all
2390 "interesting" HPT entries. Interesting entries are those with the
2391 bolted bit set, if the KVM_GET_HTAB_BOLTED_ONLY bit is set, otherwise
2392 all entries. When the end of the HPT is reached, the read() will
2393 return. If read() is called again on the fd, it will start again from
2394 the beginning of the HPT, but will only return HPT entries that have
2395 changed since they were last read.
2397 Data read or written is structured as a header (8 bytes) followed by a
2398 series of valid HPT entries (16 bytes) each. The header indicates how
2399 many valid HPT entries there are and how many invalid entries follow
2400 the valid entries. The invalid entries are not represented explicitly
2401 in the stream. The header format is:
2403 struct kvm_get_htab_header {
2409 Writes to the fd create HPT entries starting at the index given in the
2410 header; first `n_valid' valid entries with contents from the data
2411 written, then `n_invalid' invalid entries, invalidating any previously
2412 valid entries found.
2414 4.79 KVM_CREATE_DEVICE
2416 Capability: KVM_CAP_DEVICE_CTRL
2418 Parameters: struct kvm_create_device (in/out)
2419 Returns: 0 on success, -1 on error
2421 ENODEV: The device type is unknown or unsupported
2422 EEXIST: Device already created, and this type of device may not
2423 be instantiated multiple times
2425 Other error conditions may be defined by individual device types or
2426 have their standard meanings.
2428 Creates an emulated device in the kernel. The file descriptor returned
2429 in fd can be used with KVM_SET/GET/HAS_DEVICE_ATTR.
2431 If the KVM_CREATE_DEVICE_TEST flag is set, only test whether the
2432 device type is supported (not necessarily whether it can be created
2435 Individual devices should not define flags. Attributes should be used
2436 for specifying any behavior that is not implied by the device type
2439 struct kvm_create_device {
2440 __u32 type; /* in: KVM_DEV_TYPE_xxx */
2441 __u32 fd; /* out: device handle */
2442 __u32 flags; /* in: KVM_CREATE_DEVICE_xxx */
2445 4.80 KVM_SET_DEVICE_ATTR/KVM_GET_DEVICE_ATTR
2447 Capability: KVM_CAP_DEVICE_CTRL, KVM_CAP_VM_ATTRIBUTES for vm device,
2448 KVM_CAP_VCPU_ATTRIBUTES for vcpu device
2449 Type: device ioctl, vm ioctl, vcpu ioctl
2450 Parameters: struct kvm_device_attr
2451 Returns: 0 on success, -1 on error
2453 ENXIO: The group or attribute is unknown/unsupported for this device
2454 or hardware support is missing.
2455 EPERM: The attribute cannot (currently) be accessed this way
2456 (e.g. read-only attribute, or attribute that only makes
2457 sense when the device is in a different state)
2459 Other error conditions may be defined by individual device types.
2461 Gets/sets a specified piece of device configuration and/or state. The
2462 semantics are device-specific. See individual device documentation in
2463 the "devices" directory. As with ONE_REG, the size of the data
2464 transferred is defined by the particular attribute.
2466 struct kvm_device_attr {
2467 __u32 flags; /* no flags currently defined */
2468 __u32 group; /* device-defined */
2469 __u64 attr; /* group-defined */
2470 __u64 addr; /* userspace address of attr data */
2473 4.81 KVM_HAS_DEVICE_ATTR
2475 Capability: KVM_CAP_DEVICE_CTRL, KVM_CAP_VM_ATTRIBUTES for vm device,
2476 KVM_CAP_VCPU_ATTRIBUTES for vcpu device
2477 Type: device ioctl, vm ioctl, vcpu ioctl
2478 Parameters: struct kvm_device_attr
2479 Returns: 0 on success, -1 on error
2481 ENXIO: The group or attribute is unknown/unsupported for this device
2482 or hardware support is missing.
2484 Tests whether a device supports a particular attribute. A successful
2485 return indicates the attribute is implemented. It does not necessarily
2486 indicate that the attribute can be read or written in the device's
2487 current state. "addr" is ignored.
2489 4.82 KVM_ARM_VCPU_INIT
2492 Architectures: arm, arm64
2494 Parameters: struct kvm_vcpu_init (in)
2495 Returns: 0 on success; -1 on error
2497 Â EINVAL: Â Â Â the target is unknown, or the combination of features is invalid.
2498 Â ENOENT: Â Â Â a features bit specified is unknown.
2500 This tells KVM what type of CPU to present to the guest, and what
2501 optional features it should have. Â This will cause a reset of the cpu
2502 registers to their initial values. Â If this is not called, KVM_RUN will
2503 return ENOEXEC for that vcpu.
2505 Note that because some registers reflect machine topology, all vcpus
2506 should be created before this ioctl is invoked.
2508 Userspace can call this function multiple times for a given vcpu, including
2509 after the vcpu has been run. This will reset the vcpu to its initial
2510 state. All calls to this function after the initial call must use the same
2511 target and same set of feature flags, otherwise EINVAL will be returned.
2514 - KVM_ARM_VCPU_POWER_OFF: Starts the CPU in a power-off state.
2515 Depends on KVM_CAP_ARM_PSCI. If not set, the CPU will be powered on
2516 and execute guest code when KVM_RUN is called.
2517 - KVM_ARM_VCPU_EL1_32BIT: Starts the CPU in a 32bit mode.
2518 Depends on KVM_CAP_ARM_EL1_32BIT (arm64 only).
2519 - KVM_ARM_VCPU_PSCI_0_2: Emulate PSCI v0.2 (or a future revision
2520 backward compatible with v0.2) for the CPU.
2521 Depends on KVM_CAP_ARM_PSCI_0_2.
2522 - KVM_ARM_VCPU_PMU_V3: Emulate PMUv3 for the CPU.
2523 Depends on KVM_CAP_ARM_PMU_V3.
2526 4.83 KVM_ARM_PREFERRED_TARGET
2529 Architectures: arm, arm64
2531 Parameters: struct struct kvm_vcpu_init (out)
2532 Returns: 0 on success; -1 on error
2534 ENODEV: no preferred target available for the host
2536 This queries KVM for preferred CPU target type which can be emulated
2537 by KVM on underlying host.
2539 The ioctl returns struct kvm_vcpu_init instance containing information
2540 about preferred CPU target type and recommended features for it. The
2541 kvm_vcpu_init->features bitmap returned will have feature bits set if
2542 the preferred target recommends setting these features, but this is
2545 The information returned by this ioctl can be used to prepare an instance
2546 of struct kvm_vcpu_init for KVM_ARM_VCPU_INIT ioctl which will result in
2547 in VCPU matching underlying host.
2550 4.84 KVM_GET_REG_LIST
2553 Architectures: arm, arm64, mips
2555 Parameters: struct kvm_reg_list (in/out)
2556 Returns: 0 on success; -1 on error
2558 Â E2BIG: Â Â Â Â the reg index list is too big to fit in the array specified by
2559 Â Â Â Â Â Â Â Â Â Â Â Â the user (the number required will be written into n).
2561 struct kvm_reg_list {
2562 __u64 n; /* number of registers in reg[] */
2566 This ioctl returns the guest registers that are supported for the
2567 KVM_GET_ONE_REG/KVM_SET_ONE_REG calls.
2570 4.85 KVM_ARM_SET_DEVICE_ADDR (deprecated)
2572 Capability: KVM_CAP_ARM_SET_DEVICE_ADDR
2573 Architectures: arm, arm64
2575 Parameters: struct kvm_arm_device_address (in)
2576 Returns: 0 on success, -1 on error
2578 ENODEV: The device id is unknown
2579 ENXIO: Device not supported on current system
2580 EEXIST: Address already set
2581 E2BIG: Address outside guest physical address space
2582 EBUSY: Address overlaps with other device range
2584 struct kvm_arm_device_addr {
2589 Specify a device address in the guest's physical address space where guests
2590 can access emulated or directly exposed devices, which the host kernel needs
2591 to know about. The id field is an architecture specific identifier for a
2594 ARM/arm64 divides the id field into two parts, a device id and an
2595 address type id specific to the individual device.
2597 Â bits: | 63 ... 32 | 31 ... 16 | 15 ... 0 |
2598 field: | 0x00000000 | device id | addr type id |
2600 ARM/arm64 currently only require this when using the in-kernel GIC
2601 support for the hardware VGIC features, using KVM_ARM_DEVICE_VGIC_V2
2602 as the device id. When setting the base address for the guest's
2603 mapping of the VGIC virtual CPU and distributor interface, the ioctl
2604 must be called after calling KVM_CREATE_IRQCHIP, but before calling
2605 KVM_RUN on any of the VCPUs. Calling this ioctl twice for any of the
2606 base addresses will return -EEXIST.
2608 Note, this IOCTL is deprecated and the more flexible SET/GET_DEVICE_ATTR API
2609 should be used instead.
2612 4.86 KVM_PPC_RTAS_DEFINE_TOKEN
2614 Capability: KVM_CAP_PPC_RTAS
2617 Parameters: struct kvm_rtas_token_args
2618 Returns: 0 on success, -1 on error
2620 Defines a token value for a RTAS (Run Time Abstraction Services)
2621 service in order to allow it to be handled in the kernel. The
2622 argument struct gives the name of the service, which must be the name
2623 of a service that has a kernel-side implementation. If the token
2624 value is non-zero, it will be associated with that service, and
2625 subsequent RTAS calls by the guest specifying that token will be
2626 handled by the kernel. If the token value is 0, then any token
2627 associated with the service will be forgotten, and subsequent RTAS
2628 calls by the guest for that service will be passed to userspace to be
2631 4.87 KVM_SET_GUEST_DEBUG
2633 Capability: KVM_CAP_SET_GUEST_DEBUG
2634 Architectures: x86, s390, ppc, arm64
2636 Parameters: struct kvm_guest_debug (in)
2637 Returns: 0 on success; -1 on error
2639 struct kvm_guest_debug {
2642 struct kvm_guest_debug_arch arch;
2645 Set up the processor specific debug registers and configure vcpu for
2646 handling guest debug events. There are two parts to the structure, the
2647 first a control bitfield indicates the type of debug events to handle
2648 when running. Common control bits are:
2650 - KVM_GUESTDBG_ENABLE: guest debugging is enabled
2651 - KVM_GUESTDBG_SINGLESTEP: the next run should single-step
2653 The top 16 bits of the control field are architecture specific control
2654 flags which can include the following:
2656 - KVM_GUESTDBG_USE_SW_BP: using software breakpoints [x86, arm64]
2657 - KVM_GUESTDBG_USE_HW_BP: using hardware breakpoints [x86, s390, arm64]
2658 - KVM_GUESTDBG_INJECT_DB: inject DB type exception [x86]
2659 - KVM_GUESTDBG_INJECT_BP: inject BP type exception [x86]
2660 - KVM_GUESTDBG_EXIT_PENDING: trigger an immediate guest exit [s390]
2662 For example KVM_GUESTDBG_USE_SW_BP indicates that software breakpoints
2663 are enabled in memory so we need to ensure breakpoint exceptions are
2664 correctly trapped and the KVM run loop exits at the breakpoint and not
2665 running off into the normal guest vector. For KVM_GUESTDBG_USE_HW_BP
2666 we need to ensure the guest vCPUs architecture specific registers are
2667 updated to the correct (supplied) values.
2669 The second part of the structure is architecture specific and
2670 typically contains a set of debug registers.
2672 For arm64 the number of debug registers is implementation defined and
2673 can be determined by querying the KVM_CAP_GUEST_DEBUG_HW_BPS and
2674 KVM_CAP_GUEST_DEBUG_HW_WPS capabilities which return a positive number
2675 indicating the number of supported registers.
2677 When debug events exit the main run loop with the reason
2678 KVM_EXIT_DEBUG with the kvm_debug_exit_arch part of the kvm_run
2679 structure containing architecture specific debug information.
2681 4.88 KVM_GET_EMULATED_CPUID
2683 Capability: KVM_CAP_EXT_EMUL_CPUID
2686 Parameters: struct kvm_cpuid2 (in/out)
2687 Returns: 0 on success, -1 on error
2692 struct kvm_cpuid_entry2 entries[0];
2695 The member 'flags' is used for passing flags from userspace.
2697 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX BIT(0)
2698 #define KVM_CPUID_FLAG_STATEFUL_FUNC BIT(1)
2699 #define KVM_CPUID_FLAG_STATE_READ_NEXT BIT(2)
2701 struct kvm_cpuid_entry2 {
2712 This ioctl returns x86 cpuid features which are emulated by
2713 kvm.Userspace can use the information returned by this ioctl to query
2714 which features are emulated by kvm instead of being present natively.
2716 Userspace invokes KVM_GET_EMULATED_CPUID by passing a kvm_cpuid2
2717 structure with the 'nent' field indicating the number of entries in
2718 the variable-size array 'entries'. If the number of entries is too low
2719 to describe the cpu capabilities, an error (E2BIG) is returned. If the
2720 number is too high, the 'nent' field is adjusted and an error (ENOMEM)
2721 is returned. If the number is just right, the 'nent' field is adjusted
2722 to the number of valid entries in the 'entries' array, which is then
2725 The entries returned are the set CPUID bits of the respective features
2726 which kvm emulates, as returned by the CPUID instruction, with unknown
2727 or unsupported feature bits cleared.
2729 Features like x2apic, for example, may not be present in the host cpu
2730 but are exposed by kvm in KVM_GET_SUPPORTED_CPUID because they can be
2731 emulated efficiently and thus not included here.
2733 The fields in each entry are defined as follows:
2735 function: the eax value used to obtain the entry
2736 index: the ecx value used to obtain the entry (for entries that are
2738 flags: an OR of zero or more of the following:
2739 KVM_CPUID_FLAG_SIGNIFCANT_INDEX:
2740 if the index field is valid
2741 KVM_CPUID_FLAG_STATEFUL_FUNC:
2742 if cpuid for this function returns different values for successive
2743 invocations; there will be several entries with the same function,
2744 all with this flag set
2745 KVM_CPUID_FLAG_STATE_READ_NEXT:
2746 for KVM_CPUID_FLAG_STATEFUL_FUNC entries, set if this entry is
2747 the first entry to be read by a cpu
2748 eax, ebx, ecx, edx: the values returned by the cpuid instruction for
2749 this function/index combination
2751 4.89 KVM_S390_MEM_OP
2753 Capability: KVM_CAP_S390_MEM_OP
2756 Parameters: struct kvm_s390_mem_op (in)
2757 Returns: = 0 on success,
2758 < 0 on generic error (e.g. -EFAULT or -ENOMEM),
2759 > 0 if an exception occurred while walking the page tables
2761 Read or write data from/to the logical (virtual) memory of a VCPU.
2763 Parameters are specified via the following structure:
2765 struct kvm_s390_mem_op {
2766 __u64 gaddr; /* the guest address */
2767 __u64 flags; /* flags */
2768 __u32 size; /* amount of bytes */
2769 __u32 op; /* type of operation */
2770 __u64 buf; /* buffer in userspace */
2771 __u8 ar; /* the access register number */
2772 __u8 reserved[31]; /* should be set to 0 */
2775 The type of operation is specified in the "op" field. It is either
2776 KVM_S390_MEMOP_LOGICAL_READ for reading from logical memory space or
2777 KVM_S390_MEMOP_LOGICAL_WRITE for writing to logical memory space. The
2778 KVM_S390_MEMOP_F_CHECK_ONLY flag can be set in the "flags" field to check
2779 whether the corresponding memory access would create an access exception
2780 (without touching the data in the memory at the destination). In case an
2781 access exception occurred while walking the MMU tables of the guest, the
2782 ioctl returns a positive error number to indicate the type of exception.
2783 This exception is also raised directly at the corresponding VCPU if the
2784 flag KVM_S390_MEMOP_F_INJECT_EXCEPTION is set in the "flags" field.
2786 The start address of the memory region has to be specified in the "gaddr"
2787 field, and the length of the region in the "size" field. "buf" is the buffer
2788 supplied by the userspace application where the read data should be written
2789 to for KVM_S390_MEMOP_LOGICAL_READ, or where the data that should be written
2790 is stored for a KVM_S390_MEMOP_LOGICAL_WRITE. "buf" is unused and can be NULL
2791 when KVM_S390_MEMOP_F_CHECK_ONLY is specified. "ar" designates the access
2792 register number to be used.
2794 The "reserved" field is meant for future extensions. It is not used by
2795 KVM with the currently defined set of flags.
2797 4.90 KVM_S390_GET_SKEYS
2799 Capability: KVM_CAP_S390_SKEYS
2802 Parameters: struct kvm_s390_skeys
2803 Returns: 0 on success, KVM_S390_GET_KEYS_NONE if guest is not using storage
2804 keys, negative value on error
2806 This ioctl is used to get guest storage key values on the s390
2807 architecture. The ioctl takes parameters via the kvm_s390_skeys struct.
2809 struct kvm_s390_skeys {
2812 __u64 skeydata_addr;
2817 The start_gfn field is the number of the first guest frame whose storage keys
2820 The count field is the number of consecutive frames (starting from start_gfn)
2821 whose storage keys to get. The count field must be at least 1 and the maximum
2822 allowed value is defined as KVM_S390_SKEYS_ALLOC_MAX. Values outside this range
2823 will cause the ioctl to return -EINVAL.
2825 The skeydata_addr field is the address to a buffer large enough to hold count
2826 bytes. This buffer will be filled with storage key data by the ioctl.
2828 4.91 KVM_S390_SET_SKEYS
2830 Capability: KVM_CAP_S390_SKEYS
2833 Parameters: struct kvm_s390_skeys
2834 Returns: 0 on success, negative value on error
2836 This ioctl is used to set guest storage key values on the s390
2837 architecture. The ioctl takes parameters via the kvm_s390_skeys struct.
2838 See section on KVM_S390_GET_SKEYS for struct definition.
2840 The start_gfn field is the number of the first guest frame whose storage keys
2843 The count field is the number of consecutive frames (starting from start_gfn)
2844 whose storage keys to get. The count field must be at least 1 and the maximum
2845 allowed value is defined as KVM_S390_SKEYS_ALLOC_MAX. Values outside this range
2846 will cause the ioctl to return -EINVAL.
2848 The skeydata_addr field is the address to a buffer containing count bytes of
2849 storage keys. Each byte in the buffer will be set as the storage key for a
2850 single frame starting at start_gfn for count frames.
2852 Note: If any architecturally invalid key value is found in the given data then
2853 the ioctl will return -EINVAL.
2857 Capability: KVM_CAP_S390_INJECT_IRQ
2860 Parameters: struct kvm_s390_irq (in)
2861 Returns: 0 on success, -1 on error
2863 EINVAL: interrupt type is invalid
2864 type is KVM_S390_SIGP_STOP and flag parameter is invalid value
2865 type is KVM_S390_INT_EXTERNAL_CALL and code is bigger
2866 than the maximum of VCPUs
2867 EBUSY: type is KVM_S390_SIGP_SET_PREFIX and vcpu is not stopped
2868 type is KVM_S390_SIGP_STOP and a stop irq is already pending
2869 type is KVM_S390_INT_EXTERNAL_CALL and an external call interrupt
2872 Allows to inject an interrupt to the guest.
2874 Using struct kvm_s390_irq as a parameter allows
2875 to inject additional payload which is not
2876 possible via KVM_S390_INTERRUPT.
2878 Interrupt parameters are passed via kvm_s390_irq:
2880 struct kvm_s390_irq {
2883 struct kvm_s390_io_info io;
2884 struct kvm_s390_ext_info ext;
2885 struct kvm_s390_pgm_info pgm;
2886 struct kvm_s390_emerg_info emerg;
2887 struct kvm_s390_extcall_info extcall;
2888 struct kvm_s390_prefix_info prefix;
2889 struct kvm_s390_stop_info stop;
2890 struct kvm_s390_mchk_info mchk;
2895 type can be one of the following:
2897 KVM_S390_SIGP_STOP - sigp stop; parameter in .stop
2898 KVM_S390_PROGRAM_INT - program check; parameters in .pgm
2899 KVM_S390_SIGP_SET_PREFIX - sigp set prefix; parameters in .prefix
2900 KVM_S390_RESTART - restart; no parameters
2901 KVM_S390_INT_CLOCK_COMP - clock comparator interrupt; no parameters
2902 KVM_S390_INT_CPU_TIMER - CPU timer interrupt; no parameters
2903 KVM_S390_INT_EMERGENCY - sigp emergency; parameters in .emerg
2904 KVM_S390_INT_EXTERNAL_CALL - sigp external call; parameters in .extcall
2905 KVM_S390_MCHK - machine check interrupt; parameters in .mchk
2908 Note that the vcpu ioctl is asynchronous to vcpu execution.
2910 4.94 KVM_S390_GET_IRQ_STATE
2912 Capability: KVM_CAP_S390_IRQ_STATE
2915 Parameters: struct kvm_s390_irq_state (out)
2916 Returns: >= number of bytes copied into buffer,
2917 -EINVAL if buffer size is 0,
2918 -ENOBUFS if buffer size is too small to fit all pending interrupts,
2919 -EFAULT if the buffer address was invalid
2921 This ioctl allows userspace to retrieve the complete state of all currently
2922 pending interrupts in a single buffer. Use cases include migration
2923 and introspection. The parameter structure contains the address of a
2924 userspace buffer and its length:
2926 struct kvm_s390_irq_state {
2928 __u32 flags; /* will stay unused for compatibility reasons */
2930 __u32 reserved[4]; /* will stay unused for compatibility reasons */
2933 Userspace passes in the above struct and for each pending interrupt a
2934 struct kvm_s390_irq is copied to the provided buffer.
2936 The structure contains a flags and a reserved field for future extensions. As
2937 the kernel never checked for flags == 0 and QEMU never pre-zeroed flags and
2938 reserved, these fields can not be used in the future without breaking
2941 If -ENOBUFS is returned the buffer provided was too small and userspace
2942 may retry with a bigger buffer.
2944 4.95 KVM_S390_SET_IRQ_STATE
2946 Capability: KVM_CAP_S390_IRQ_STATE
2949 Parameters: struct kvm_s390_irq_state (in)
2950 Returns: 0 on success,
2951 -EFAULT if the buffer address was invalid,
2952 -EINVAL for an invalid buffer length (see below),
2953 -EBUSY if there were already interrupts pending,
2954 errors occurring when actually injecting the
2955 interrupt. See KVM_S390_IRQ.
2957 This ioctl allows userspace to set the complete state of all cpu-local
2958 interrupts currently pending for the vcpu. It is intended for restoring
2959 interrupt state after a migration. The input parameter is a userspace buffer
2960 containing a struct kvm_s390_irq_state:
2962 struct kvm_s390_irq_state {
2964 __u32 flags; /* will stay unused for compatibility reasons */
2966 __u32 reserved[4]; /* will stay unused for compatibility reasons */
2969 The restrictions for flags and reserved apply as well.
2970 (see KVM_S390_GET_IRQ_STATE)
2972 The userspace memory referenced by buf contains a struct kvm_s390_irq
2973 for each interrupt to be injected into the guest.
2974 If one of the interrupts could not be injected for some reason the
2977 len must be a multiple of sizeof(struct kvm_s390_irq). It must be > 0
2978 and it must not exceed (max_vcpus + 32) * sizeof(struct kvm_s390_irq),
2979 which is the maximum number of possibly pending cpu-local interrupts.
2983 Capability: KVM_CAP_X86_SMM
2987 Returns: 0 on success, -1 on error
2989 Queues an SMI on the thread's vcpu.
2991 4.97 KVM_CAP_PPC_MULTITCE
2993 Capability: KVM_CAP_PPC_MULTITCE
2997 This capability means the kernel is capable of handling hypercalls
2998 H_PUT_TCE_INDIRECT and H_STUFF_TCE without passing those into the user
2999 space. This significantly accelerates DMA operations for PPC KVM guests.
3000 User space should expect that its handlers for these hypercalls
3001 are not going to be called if user space previously registered LIOBN
3002 in KVM (via KVM_CREATE_SPAPR_TCE or similar calls).
3004 In order to enable H_PUT_TCE_INDIRECT and H_STUFF_TCE use in the guest,
3005 user space might have to advertise it for the guest. For example,
3006 IBM pSeries (sPAPR) guest starts using them if "hcall-multi-tce" is
3007 present in the "ibm,hypertas-functions" device-tree property.
3009 The hypercalls mentioned above may or may not be processed successfully
3010 in the kernel based fast path. If they can not be handled by the kernel,
3011 they will get passed on to user space. So user space still has to have
3012 an implementation for these despite the in kernel acceleration.
3014 This capability is always enabled.
3016 4.98 KVM_CREATE_SPAPR_TCE_64
3018 Capability: KVM_CAP_SPAPR_TCE_64
3019 Architectures: powerpc
3021 Parameters: struct kvm_create_spapr_tce_64 (in)
3022 Returns: file descriptor for manipulating the created TCE table
3024 This is an extension for KVM_CAP_SPAPR_TCE which only supports 32bit
3025 windows, described in 4.62 KVM_CREATE_SPAPR_TCE
3027 This capability uses extended struct in ioctl interface:
3029 /* for KVM_CAP_SPAPR_TCE_64 */
3030 struct kvm_create_spapr_tce_64 {
3034 __u64 offset; /* in pages */
3035 __u64 size; /* in pages */
3038 The aim of extension is to support an additional bigger DMA window with
3039 a variable page size.
3040 KVM_CREATE_SPAPR_TCE_64 receives a 64bit window size, an IOMMU page shift and
3041 a bus offset of the corresponding DMA window, @size and @offset are numbers
3044 @flags are not used at the moment.
3046 The rest of functionality is identical to KVM_CREATE_SPAPR_TCE.
3048 4.99 KVM_REINJECT_CONTROL
3050 Capability: KVM_CAP_REINJECT_CONTROL
3053 Parameters: struct kvm_reinject_control (in)
3054 Returns: 0 on success,
3055 -EFAULT if struct kvm_reinject_control cannot be read,
3056 -ENXIO if KVM_CREATE_PIT or KVM_CREATE_PIT2 didn't succeed earlier.
3058 i8254 (PIT) has two modes, reinject and !reinject. The default is reinject,
3059 where KVM queues elapsed i8254 ticks and monitors completion of interrupt from
3060 vector(s) that i8254 injects. Reinject mode dequeues a tick and injects its
3061 interrupt whenever there isn't a pending interrupt from i8254.
3062 !reinject mode injects an interrupt as soon as a tick arrives.
3064 struct kvm_reinject_control {
3069 pit_reinject = 0 (!reinject mode) is recommended, unless running an old
3070 operating system that uses the PIT for timing (e.g. Linux 2.4.x).
3072 4.100 KVM_PPC_CONFIGURE_V3_MMU
3074 Capability: KVM_CAP_PPC_RADIX_MMU or KVM_CAP_PPC_HASH_MMU_V3
3077 Parameters: struct kvm_ppc_mmuv3_cfg (in)
3078 Returns: 0 on success,
3079 -EFAULT if struct kvm_ppc_mmuv3_cfg cannot be read,
3080 -EINVAL if the configuration is invalid
3082 This ioctl controls whether the guest will use radix or HPT (hashed
3083 page table) translation, and sets the pointer to the process table for
3086 struct kvm_ppc_mmuv3_cfg {
3088 __u64 process_table;
3091 There are two bits that can be set in flags; KVM_PPC_MMUV3_RADIX and
3092 KVM_PPC_MMUV3_GTSE. KVM_PPC_MMUV3_RADIX, if set, configures the guest
3093 to use radix tree translation, and if clear, to use HPT translation.
3094 KVM_PPC_MMUV3_GTSE, if set and if KVM permits it, configures the guest
3095 to be able to use the global TLB and SLB invalidation instructions;
3096 if clear, the guest may not use these instructions.
3098 The process_table field specifies the address and size of the guest
3099 process table, which is in the guest's space. This field is formatted
3100 as the second doubleword of the partition table entry, as defined in
3101 the Power ISA V3.00, Book III section 5.7.6.1.
3103 4.101 KVM_PPC_GET_RMMU_INFO
3105 Capability: KVM_CAP_PPC_RADIX_MMU
3108 Parameters: struct kvm_ppc_rmmu_info (out)
3109 Returns: 0 on success,
3110 -EFAULT if struct kvm_ppc_rmmu_info cannot be written,
3111 -EINVAL if no useful information can be returned
3113 This ioctl returns a structure containing two things: (a) a list
3114 containing supported radix tree geometries, and (b) a list that maps
3115 page sizes to put in the "AP" (actual page size) field for the tlbie
3116 (TLB invalidate entry) instruction.
3118 struct kvm_ppc_rmmu_info {
3119 struct kvm_ppc_radix_geom {
3124 __u32 ap_encodings[8];
3127 The geometries[] field gives up to 8 supported geometries for the
3128 radix page table, in terms of the log base 2 of the smallest page
3129 size, and the number of bits indexed at each level of the tree, from
3130 the PTE level up to the PGD level in that order. Any unused entries
3131 will have 0 in the page_shift field.
3133 The ap_encodings gives the supported page sizes and their AP field
3134 encodings, encoded with the AP value in the top 3 bits and the log
3135 base 2 of the page size in the bottom 6 bits.
3137 4.102 KVM_PPC_RESIZE_HPT_PREPARE
3139 Capability: KVM_CAP_SPAPR_RESIZE_HPT
3140 Architectures: powerpc
3142 Parameters: struct kvm_ppc_resize_hpt (in)
3143 Returns: 0 on successful completion,
3144 >0 if a new HPT is being prepared, the value is an estimated
3145 number of milliseconds until preparation is complete
3146 -EFAULT if struct kvm_reinject_control cannot be read,
3147 -EINVAL if the supplied shift or flags are invalid
3148 -ENOMEM if unable to allocate the new HPT
3149 -ENOSPC if there was a hash collision when moving existing
3150 HPT entries to the new HPT
3151 -EIO on other error conditions
3153 Used to implement the PAPR extension for runtime resizing of a guest's
3154 Hashed Page Table (HPT). Specifically this starts, stops or monitors
3155 the preparation of a new potential HPT for the guest, essentially
3156 implementing the H_RESIZE_HPT_PREPARE hypercall.
3158 If called with shift > 0 when there is no pending HPT for the guest,
3159 this begins preparation of a new pending HPT of size 2^(shift) bytes.
3160 It then returns a positive integer with the estimated number of
3161 milliseconds until preparation is complete.
3163 If called when there is a pending HPT whose size does not match that
3164 requested in the parameters, discards the existing pending HPT and
3165 creates a new one as above.
3167 If called when there is a pending HPT of the size requested, will:
3168 * If preparation of the pending HPT is already complete, return 0
3169 * If preparation of the pending HPT has failed, return an error
3170 code, then discard the pending HPT.
3171 * If preparation of the pending HPT is still in progress, return an
3172 estimated number of milliseconds until preparation is complete.
3174 If called with shift == 0, discards any currently pending HPT and
3175 returns 0 (i.e. cancels any in-progress preparation).
3177 flags is reserved for future expansion, currently setting any bits in
3178 flags will result in an -EINVAL.
3180 Normally this will be called repeatedly with the same parameters until
3181 it returns <= 0. The first call will initiate preparation, subsequent
3182 ones will monitor preparation until it completes or fails.
3184 struct kvm_ppc_resize_hpt {
3190 4.103 KVM_PPC_RESIZE_HPT_COMMIT
3192 Capability: KVM_CAP_SPAPR_RESIZE_HPT
3193 Architectures: powerpc
3195 Parameters: struct kvm_ppc_resize_hpt (in)
3196 Returns: 0 on successful completion,
3197 -EFAULT if struct kvm_reinject_control cannot be read,
3198 -EINVAL if the supplied shift or flags are invalid
3199 -ENXIO is there is no pending HPT, or the pending HPT doesn't
3200 have the requested size
3201 -EBUSY if the pending HPT is not fully prepared
3202 -ENOSPC if there was a hash collision when moving existing
3203 HPT entries to the new HPT
3204 -EIO on other error conditions
3206 Used to implement the PAPR extension for runtime resizing of a guest's
3207 Hashed Page Table (HPT). Specifically this requests that the guest be
3208 transferred to working with the new HPT, essentially implementing the
3209 H_RESIZE_HPT_COMMIT hypercall.
3211 This should only be called after KVM_PPC_RESIZE_HPT_PREPARE has
3212 returned 0 with the same parameters. In other cases
3213 KVM_PPC_RESIZE_HPT_COMMIT will return an error (usually -ENXIO or
3214 -EBUSY, though others may be possible if the preparation was started,
3217 This will have undefined effects on the guest if it has not already
3218 placed itself in a quiescent state where no vcpu will make MMU enabled
3221 On succsful completion, the pending HPT will become the guest's active
3222 HPT and the previous HPT will be discarded.
3224 On failure, the guest will still be operating on its previous HPT.
3226 struct kvm_ppc_resize_hpt {
3232 4.104 KVM_X86_GET_MCE_CAP_SUPPORTED
3234 Capability: KVM_CAP_MCE
3237 Parameters: u64 mce_cap (out)
3238 Returns: 0 on success, -1 on error
3240 Returns supported MCE capabilities. The u64 mce_cap parameter
3241 has the same format as the MSR_IA32_MCG_CAP register. Supported
3242 capabilities will have the corresponding bits set.
3244 4.105 KVM_X86_SETUP_MCE
3246 Capability: KVM_CAP_MCE
3249 Parameters: u64 mcg_cap (in)
3250 Returns: 0 on success,
3251 -EFAULT if u64 mcg_cap cannot be read,
3252 -EINVAL if the requested number of banks is invalid,
3253 -EINVAL if requested MCE capability is not supported.
3255 Initializes MCE support for use. The u64 mcg_cap parameter
3256 has the same format as the MSR_IA32_MCG_CAP register and
3257 specifies which capabilities should be enabled. The maximum
3258 supported number of error-reporting banks can be retrieved when
3259 checking for KVM_CAP_MCE. The supported capabilities can be
3260 retrieved with KVM_X86_GET_MCE_CAP_SUPPORTED.
3262 4.106 KVM_X86_SET_MCE
3264 Capability: KVM_CAP_MCE
3267 Parameters: struct kvm_x86_mce (in)
3268 Returns: 0 on success,
3269 -EFAULT if struct kvm_x86_mce cannot be read,
3270 -EINVAL if the bank number is invalid,
3271 -EINVAL if VAL bit is not set in status field.
3273 Inject a machine check error (MCE) into the guest. The input
3276 struct kvm_x86_mce {
3286 If the MCE being reported is an uncorrected error, KVM will
3287 inject it as an MCE exception into the guest. If the guest
3288 MCG_STATUS register reports that an MCE is in progress, KVM
3289 causes an KVM_EXIT_SHUTDOWN vmexit.
3291 Otherwise, if the MCE is a corrected error, KVM will just
3292 store it in the corresponding bank (provided this bank is
3293 not holding a previously reported uncorrected error).
3295 4.107 KVM_S390_GET_CMMA_BITS
3297 Capability: KVM_CAP_S390_CMMA_MIGRATION
3300 Parameters: struct kvm_s390_cmma_log (in, out)
3301 Returns: 0 on success, a negative value on error
3303 This ioctl is used to get the values of the CMMA bits on the s390
3304 architecture. It is meant to be used in two scenarios:
3305 - During live migration to save the CMMA values. Live migration needs
3306 to be enabled via the KVM_REQ_START_MIGRATION VM property.
3307 - To non-destructively peek at the CMMA values, with the flag
3308 KVM_S390_CMMA_PEEK set.
3310 The ioctl takes parameters via the kvm_s390_cmma_log struct. The desired
3311 values are written to a buffer whose location is indicated via the "values"
3312 member in the kvm_s390_cmma_log struct. The values in the input struct are
3313 also updated as needed.
3314 Each CMMA value takes up one byte.
3316 struct kvm_s390_cmma_log {
3327 start_gfn is the number of the first guest frame whose CMMA values are
3330 count is the length of the buffer in bytes,
3332 values points to the buffer where the result will be written to.
3334 If count is greater than KVM_S390_SKEYS_MAX, then it is considered to be
3335 KVM_S390_SKEYS_MAX. KVM_S390_SKEYS_MAX is re-used for consistency with
3338 The result is written in the buffer pointed to by the field values, and
3339 the values of the input parameter are updated as follows.
3341 Depending on the flags, different actions are performed. The only
3342 supported flag so far is KVM_S390_CMMA_PEEK.
3344 The default behaviour if KVM_S390_CMMA_PEEK is not set is:
3345 start_gfn will indicate the first page frame whose CMMA bits were dirty.
3346 It is not necessarily the same as the one passed as input, as clean pages
3349 count will indicate the number of bytes actually written in the buffer.
3350 It can (and very often will) be smaller than the input value, since the
3351 buffer is only filled until 16 bytes of clean values are found (which
3352 are then not copied in the buffer). Since a CMMA migration block needs
3353 the base address and the length, for a total of 16 bytes, we will send
3354 back some clean data if there is some dirty data afterwards, as long as
3355 the size of the clean data does not exceed the size of the header. This
3356 allows to minimize the amount of data to be saved or transferred over
3357 the network at the expense of more roundtrips to userspace. The next
3358 invocation of the ioctl will skip over all the clean values, saving
3359 potentially more than just the 16 bytes we found.
3361 If KVM_S390_CMMA_PEEK is set:
3362 the existing storage attributes are read even when not in migration
3363 mode, and no other action is performed;
3365 the output start_gfn will be equal to the input start_gfn,
3367 the output count will be equal to the input count, except if the end of
3368 memory has been reached.
3371 the field "remaining" will indicate the total number of dirty CMMA values
3372 still remaining, or 0 if KVM_S390_CMMA_PEEK is set and migration mode is
3377 values points to the userspace buffer where the result will be stored.
3379 This ioctl can fail with -ENOMEM if not enough memory can be allocated to
3380 complete the task, with -ENXIO if CMMA is not enabled, with -EINVAL if
3381 KVM_S390_CMMA_PEEK is not set but migration mode was not enabled, with
3382 -EFAULT if the userspace address is invalid or if no page table is
3383 present for the addresses (e.g. when using hugepages).
3385 4.108 KVM_S390_SET_CMMA_BITS
3387 Capability: KVM_CAP_S390_CMMA_MIGRATION
3390 Parameters: struct kvm_s390_cmma_log (in)
3391 Returns: 0 on success, a negative value on error
3393 This ioctl is used to set the values of the CMMA bits on the s390
3394 architecture. It is meant to be used during live migration to restore
3395 the CMMA values, but there are no restrictions on its use.
3396 The ioctl takes parameters via the kvm_s390_cmma_values struct.
3397 Each CMMA value takes up one byte.
3399 struct kvm_s390_cmma_log {
3410 start_gfn indicates the starting guest frame number,
3412 count indicates how many values are to be considered in the buffer,
3414 flags is not used and must be 0.
3416 mask indicates which PGSTE bits are to be considered.
3418 remaining is not used.
3420 values points to the buffer in userspace where to store the values.
3422 This ioctl can fail with -ENOMEM if not enough memory can be allocated to
3423 complete the task, with -ENXIO if CMMA is not enabled, with -EINVAL if
3424 the count field is too large (e.g. more than KVM_S390_CMMA_SIZE_MAX) or
3425 if the flags field was not 0, with -EFAULT if the userspace address is
3426 invalid, if invalid pages are written to (e.g. after the end of memory)
3427 or if no page table is present for the addresses (e.g. when using
3430 4.109 KVM_PPC_GET_CPU_CHAR
3432 Capability: KVM_CAP_PPC_GET_CPU_CHAR
3433 Architectures: powerpc
3435 Parameters: struct kvm_ppc_cpu_char (out)
3436 Returns: 0 on successful completion
3437 -EFAULT if struct kvm_ppc_cpu_char cannot be written
3439 This ioctl gives userspace information about certain characteristics
3440 of the CPU relating to speculative execution of instructions and
3441 possible information leakage resulting from speculative execution (see
3442 CVE-2017-5715, CVE-2017-5753 and CVE-2017-5754). The information is
3443 returned in struct kvm_ppc_cpu_char, which looks like this:
3445 struct kvm_ppc_cpu_char {
3446 __u64 character; /* characteristics of the CPU */
3447 __u64 behaviour; /* recommended software behaviour */
3448 __u64 character_mask; /* valid bits in character */
3449 __u64 behaviour_mask; /* valid bits in behaviour */
3452 For extensibility, the character_mask and behaviour_mask fields
3453 indicate which bits of character and behaviour have been filled in by
3454 the kernel. If the set of defined bits is extended in future then
3455 userspace will be able to tell whether it is running on a kernel that
3456 knows about the new bits.
3458 The character field describes attributes of the CPU which can help
3459 with preventing inadvertent information disclosure - specifically,
3460 whether there is an instruction to flash-invalidate the L1 data cache
3461 (ori 30,30,0 or mtspr SPRN_TRIG2,rN), whether the L1 data cache is set
3462 to a mode where entries can only be used by the thread that created
3463 them, whether the bcctr[l] instruction prevents speculation, and
3464 whether a speculation barrier instruction (ori 31,31,0) is provided.
3466 The behaviour field describes actions that software should take to
3467 prevent inadvertent information disclosure, and thus describes which
3468 vulnerabilities the hardware is subject to; specifically whether the
3469 L1 data cache should be flushed when returning to user mode from the
3470 kernel, and whether a speculation barrier should be placed between an
3471 array bounds check and the array access.
3473 These fields use the same bit definitions as the new
3474 H_GET_CPU_CHARACTERISTICS hypercall.
3476 4.110 KVM_MEMORY_ENCRYPT_OP
3481 Parameters: an opaque platform specific structure (in/out)
3482 Returns: 0 on success; -1 on error
3484 If the platform supports creating encrypted VMs then this ioctl can be used
3485 for issuing platform-specific memory encryption commands to manage those
3488 Currently, this ioctl is used for issuing Secure Encrypted Virtualization
3489 (SEV) commands on AMD Processors. The SEV commands are defined in
3490 Documentation/virtual/kvm/amd-memory-encryption.rst.
3492 4.111 KVM_MEMORY_ENCRYPT_REG_REGION
3497 Parameters: struct kvm_enc_region (in)
3498 Returns: 0 on success; -1 on error
3500 This ioctl can be used to register a guest memory region which may
3501 contain encrypted data (e.g. guest RAM, SMRAM etc).
3503 It is used in the SEV-enabled guest. When encryption is enabled, a guest
3504 memory region may contain encrypted data. The SEV memory encryption
3505 engine uses a tweak such that two identical plaintext pages, each at
3506 different locations will have differing ciphertexts. So swapping or
3507 moving ciphertext of those pages will not result in plaintext being
3508 swapped. So relocating (or migrating) physical backing pages for the SEV
3509 guest will require some additional steps.
3511 Note: The current SEV key management spec does not provide commands to
3512 swap or migrate (move) ciphertext pages. Hence, for now we pin the guest
3513 memory region registered with the ioctl.
3515 4.112 KVM_MEMORY_ENCRYPT_UNREG_REGION
3520 Parameters: struct kvm_enc_region (in)
3521 Returns: 0 on success; -1 on error
3523 This ioctl can be used to unregister the guest memory region registered
3524 with KVM_MEMORY_ENCRYPT_REG_REGION ioctl above.
3526 4.113 KVM_HYPERV_EVENTFD
3528 Capability: KVM_CAP_HYPERV_EVENTFD
3531 Parameters: struct kvm_hyperv_eventfd (in)
3533 This ioctl (un)registers an eventfd to receive notifications from the guest on
3534 the specified Hyper-V connection id through the SIGNAL_EVENT hypercall, without
3535 causing a user exit. SIGNAL_EVENT hypercall with non-zero event flag number
3536 (bits 24-31) still triggers a KVM_EXIT_HYPERV_HCALL user exit.
3538 struct kvm_hyperv_eventfd {
3545 The conn_id field should fit within 24 bits:
3547 #define KVM_HYPERV_CONN_ID_MASK 0x00ffffff
3549 The acceptable values for the flags field are:
3551 #define KVM_HYPERV_EVENTFD_DEASSIGN (1 << 0)
3553 Returns: 0 on success,
3554 -EINVAL if conn_id or flags is outside the allowed range
3555 -ENOENT on deassign if the conn_id isn't registered
3556 -EEXIST on assign if the conn_id is already registered
3559 5. The kvm_run structure
3560 ------------------------
3562 Application code obtains a pointer to the kvm_run structure by
3563 mmap()ing a vcpu fd. From that point, application code can control
3564 execution by changing fields in kvm_run prior to calling the KVM_RUN
3565 ioctl, and obtain information about the reason KVM_RUN returned by
3566 looking up structure members.
3570 __u8 request_interrupt_window;
3572 Request that KVM_RUN return when it becomes possible to inject external
3573 interrupts into the guest. Useful in conjunction with KVM_INTERRUPT.
3575 __u8 immediate_exit;
3577 This field is polled once when KVM_RUN starts; if non-zero, KVM_RUN
3578 exits immediately, returning -EINTR. In the common scenario where a
3579 signal is used to "kick" a VCPU out of KVM_RUN, this field can be used
3580 to avoid usage of KVM_SET_SIGNAL_MASK, which has worse scalability.
3581 Rather than blocking the signal outside KVM_RUN, userspace can set up
3582 a signal handler that sets run->immediate_exit to a non-zero value.
3584 This field is ignored if KVM_CAP_IMMEDIATE_EXIT is not available.
3591 When KVM_RUN has returned successfully (return value 0), this informs
3592 application code why KVM_RUN has returned. Allowable values for this
3593 field are detailed below.
3595 __u8 ready_for_interrupt_injection;
3597 If request_interrupt_window has been specified, this field indicates
3598 an interrupt can be injected now with KVM_INTERRUPT.
3602 The value of the current interrupt flag. Only valid if in-kernel
3603 local APIC is not used.
3607 More architecture-specific flags detailing state of the VCPU that may
3608 affect the device's behavior. The only currently defined flag is
3609 KVM_RUN_X86_SMM, which is valid on x86 machines and is set if the
3610 VCPU is in system management mode.
3612 /* in (pre_kvm_run), out (post_kvm_run) */
3615 The value of the cr8 register. Only valid if in-kernel local APIC is
3616 not used. Both input and output.
3620 The value of the APIC BASE msr. Only valid if in-kernel local
3621 APIC is not used. Both input and output.
3624 /* KVM_EXIT_UNKNOWN */
3626 __u64 hardware_exit_reason;
3629 If exit_reason is KVM_EXIT_UNKNOWN, the vcpu has exited due to unknown
3630 reasons. Further architecture-specific information is available in
3631 hardware_exit_reason.
3633 /* KVM_EXIT_FAIL_ENTRY */
3635 __u64 hardware_entry_failure_reason;
3638 If exit_reason is KVM_EXIT_FAIL_ENTRY, the vcpu could not be run due
3639 to unknown reasons. Further architecture-specific information is
3640 available in hardware_entry_failure_reason.
3642 /* KVM_EXIT_EXCEPTION */
3652 #define KVM_EXIT_IO_IN 0
3653 #define KVM_EXIT_IO_OUT 1
3655 __u8 size; /* bytes */
3658 __u64 data_offset; /* relative to kvm_run start */
3661 If exit_reason is KVM_EXIT_IO, then the vcpu has
3662 executed a port I/O instruction which could not be satisfied by kvm.
3663 data_offset describes where the data is located (KVM_EXIT_IO_OUT) or
3664 where kvm expects application code to place the data for the next
3665 KVM_RUN invocation (KVM_EXIT_IO_IN). Data format is a packed array.
3667 /* KVM_EXIT_DEBUG */
3669 struct kvm_debug_exit_arch arch;
3672 If the exit_reason is KVM_EXIT_DEBUG, then a vcpu is processing a debug event
3673 for which architecture specific information is returned.
3683 If exit_reason is KVM_EXIT_MMIO, then the vcpu has
3684 executed a memory-mapped I/O instruction which could not be satisfied
3685 by kvm. The 'data' member contains the written data if 'is_write' is
3686 true, and should be filled by application code otherwise.
3688 The 'data' member contains, in its first 'len' bytes, the value as it would
3689 appear if the VCPU performed a load or store of the appropriate width directly
3692 NOTE: For KVM_EXIT_IO, KVM_EXIT_MMIO, KVM_EXIT_OSI, KVM_EXIT_PAPR and
3693 KVM_EXIT_EPR the corresponding
3694 operations are complete (and guest state is consistent) only after userspace
3695 has re-entered the kernel with KVM_RUN. The kernel side will first finish
3696 incomplete operations and then check for pending signals. Userspace
3697 can re-enter the guest with an unmasked signal pending to complete
3700 /* KVM_EXIT_HYPERCALL */
3709 Unused. This was once used for 'hypercall to userspace'. To implement
3710 such functionality, use KVM_EXIT_IO (x86) or KVM_EXIT_MMIO (all except s390).
3711 Note KVM_EXIT_IO is significantly faster than KVM_EXIT_MMIO.
3713 /* KVM_EXIT_TPR_ACCESS */
3720 To be documented (KVM_TPR_ACCESS_REPORTING).
3722 /* KVM_EXIT_S390_SIEIC */
3725 __u64 mask; /* psw upper half */
3726 __u64 addr; /* psw lower half */
3733 /* KVM_EXIT_S390_RESET */
3734 #define KVM_S390_RESET_POR 1
3735 #define KVM_S390_RESET_CLEAR 2
3736 #define KVM_S390_RESET_SUBSYSTEM 4
3737 #define KVM_S390_RESET_CPU_INIT 8
3738 #define KVM_S390_RESET_IPL 16
3739 __u64 s390_reset_flags;
3743 /* KVM_EXIT_S390_UCONTROL */
3745 __u64 trans_exc_code;
3749 s390 specific. A page fault has occurred for a user controlled virtual
3750 machine (KVM_VM_S390_UNCONTROL) on it's host page table that cannot be
3751 resolved by the kernel.
3752 The program code and the translation exception code that were placed
3753 in the cpu's lowcore are presented here as defined by the z Architecture
3754 Principles of Operation Book in the Chapter for Dynamic Address Translation
3764 Deprecated - was used for 440 KVM.
3771 MOL uses a special hypercall interface it calls 'OSI'. To enable it, we catch
3772 hypercalls and exit with this exit struct that contains all the guest gprs.
3774 If exit_reason is KVM_EXIT_OSI, then the vcpu has triggered such a hypercall.
3775 Userspace can now handle the hypercall and when it's done modify the gprs as
3776 necessary. Upon guest entry all guest GPRs will then be replaced by the values
3779 /* KVM_EXIT_PAPR_HCALL */
3786 This is used on 64-bit PowerPC when emulating a pSeries partition,
3787 e.g. with the 'pseries' machine type in qemu. It occurs when the
3788 guest does a hypercall using the 'sc 1' instruction. The 'nr' field
3789 contains the hypercall number (from the guest R3), and 'args' contains
3790 the arguments (from the guest R4 - R12). Userspace should put the
3791 return code in 'ret' and any extra returned values in args[].
3792 The possible hypercalls are defined in the Power Architecture Platform
3793 Requirements (PAPR) document available from www.power.org (free
3794 developer registration required to access it).
3796 /* KVM_EXIT_S390_TSCH */
3798 __u16 subchannel_id;
3799 __u16 subchannel_nr;
3806 s390 specific. This exit occurs when KVM_CAP_S390_CSS_SUPPORT has been enabled
3807 and TEST SUBCHANNEL was intercepted. If dequeued is set, a pending I/O
3808 interrupt for the target subchannel has been dequeued and subchannel_id,
3809 subchannel_nr, io_int_parm and io_int_word contain the parameters for that
3810 interrupt. ipb is needed for instruction parameter decoding.
3817 On FSL BookE PowerPC chips, the interrupt controller has a fast patch
3818 interrupt acknowledge path to the core. When the core successfully
3819 delivers an interrupt, it automatically populates the EPR register with
3820 the interrupt vector number and acknowledges the interrupt inside
3821 the interrupt controller.
3823 In case the interrupt controller lives in user space, we need to do
3824 the interrupt acknowledge cycle through it to fetch the next to be
3825 delivered interrupt vector using this exit.
3827 It gets triggered whenever both KVM_CAP_PPC_EPR are enabled and an
3828 external interrupt has just been delivered into the guest. User space
3829 should put the acknowledged interrupt vector into the 'epr' field.
3831 /* KVM_EXIT_SYSTEM_EVENT */
3833 #define KVM_SYSTEM_EVENT_SHUTDOWN 1
3834 #define KVM_SYSTEM_EVENT_RESET 2
3835 #define KVM_SYSTEM_EVENT_CRASH 3
3840 If exit_reason is KVM_EXIT_SYSTEM_EVENT then the vcpu has triggered
3841 a system-level event using some architecture specific mechanism (hypercall
3842 or some special instruction). In case of ARM/ARM64, this is triggered using
3843 HVC instruction based PSCI call from the vcpu. The 'type' field describes
3844 the system-level event type. The 'flags' field describes architecture
3845 specific flags for the system-level event.
3847 Valid values for 'type' are:
3848 KVM_SYSTEM_EVENT_SHUTDOWN -- the guest has requested a shutdown of the
3849 VM. Userspace is not obliged to honour this, and if it does honour
3850 this does not need to destroy the VM synchronously (ie it may call
3851 KVM_RUN again before shutdown finally occurs).
3852 KVM_SYSTEM_EVENT_RESET -- the guest has requested a reset of the VM.
3853 As with SHUTDOWN, userspace can choose to ignore the request, or
3854 to schedule the reset to occur in the future and may call KVM_RUN again.
3855 KVM_SYSTEM_EVENT_CRASH -- the guest crash occurred and the guest
3856 has requested a crash condition maintenance. Userspace can choose
3857 to ignore the request, or to gather VM memory core dump and/or
3858 reset/shutdown of the VM.
3860 /* KVM_EXIT_IOAPIC_EOI */
3865 Indicates that the VCPU's in-kernel local APIC received an EOI for a
3866 level-triggered IOAPIC interrupt. This exit only triggers when the
3867 IOAPIC is implemented in userspace (i.e. KVM_CAP_SPLIT_IRQCHIP is enabled);
3868 the userspace IOAPIC should process the EOI and retrigger the interrupt if
3869 it is still asserted. Vector is the LAPIC interrupt vector for which the
3872 struct kvm_hyperv_exit {
3873 #define KVM_EXIT_HYPERV_SYNIC 1
3874 #define KVM_EXIT_HYPERV_HCALL 2
3890 /* KVM_EXIT_HYPERV */
3891 struct kvm_hyperv_exit hyperv;
3892 Indicates that the VCPU exits into userspace to process some tasks
3893 related to Hyper-V emulation.
3894 Valid values for 'type' are:
3895 KVM_EXIT_HYPERV_SYNIC -- synchronously notify user-space about
3896 Hyper-V SynIC state change. Notification is used to remap SynIC
3897 event/message pages and to enable/disable SynIC messages/events processing
3900 /* Fix the size of the union. */
3905 * shared registers between kvm and userspace.
3906 * kvm_valid_regs specifies the register classes set by the host
3907 * kvm_dirty_regs specified the register classes dirtied by userspace
3908 * struct kvm_sync_regs is architecture specific, as well as the
3909 * bits for kvm_valid_regs and kvm_dirty_regs
3911 __u64 kvm_valid_regs;
3912 __u64 kvm_dirty_regs;
3914 struct kvm_sync_regs regs;
3915 char padding[SYNC_REGS_SIZE_BYTES];
3918 If KVM_CAP_SYNC_REGS is defined, these fields allow userspace to access
3919 certain guest registers without having to call SET/GET_*REGS. Thus we can
3920 avoid some system call overhead if userspace has to handle the exit.
3921 Userspace can query the validity of the structure by checking
3922 kvm_valid_regs for specific bits. These bits are architecture specific
3923 and usually define the validity of a groups of registers. (e.g. one bit
3924 for general purpose registers)
3926 Please note that the kernel is allowed to use the kvm_run structure as the
3927 primary storage for certain register types. Therefore, the kernel may use the
3928 values in kvm_run even if the corresponding bit in kvm_dirty_regs is not set.
3934 6. Capabilities that can be enabled on vCPUs
3935 --------------------------------------------
3937 There are certain capabilities that change the behavior of the virtual CPU or
3938 the virtual machine when enabled. To enable them, please see section 4.37.
3939 Below you can find a list of capabilities and what their effect on the vCPU or
3940 the virtual machine is when enabling them.
3942 The following information is provided along with the description:
3944 Architectures: which instruction set architectures provide this ioctl.
3945 x86 includes both i386 and x86_64.
3947 Target: whether this is a per-vcpu or per-vm capability.
3949 Parameters: what parameters are accepted by the capability.
3951 Returns: the return value. General error numbers (EBADF, ENOMEM, EINVAL)
3952 are not detailed, but errors with specific meanings are.
3960 Returns: 0 on success; -1 on error
3962 This capability enables interception of OSI hypercalls that otherwise would
3963 be treated as normal system calls to be injected into the guest. OSI hypercalls
3964 were invented by Mac-on-Linux to have a standardized communication mechanism
3965 between the guest and the host.
3967 When this capability is enabled, KVM_EXIT_OSI can occur.
3970 6.2 KVM_CAP_PPC_PAPR
3975 Returns: 0 on success; -1 on error
3977 This capability enables interception of PAPR hypercalls. PAPR hypercalls are
3978 done using the hypercall instruction "sc 1".
3980 It also sets the guest privilege level to "supervisor" mode. Usually the guest
3981 runs in "hypervisor" privilege mode with a few missing features.
3983 In addition to the above, it changes the semantics of SDR1. In this mode, the
3984 HTAB address part of SDR1 contains an HVA instead of a GPA, as PAPR keeps the
3985 HTAB invisible to the guest.
3987 When this capability is enabled, KVM_EXIT_PAPR_HCALL can occur.
3994 Parameters: args[0] is the address of a struct kvm_config_tlb
3995 Returns: 0 on success; -1 on error
3997 struct kvm_config_tlb {
4004 Configures the virtual CPU's TLB array, establishing a shared memory area
4005 between userspace and KVM. The "params" and "array" fields are userspace
4006 addresses of mmu-type-specific data structures. The "array_len" field is an
4007 safety mechanism, and should be set to the size in bytes of the memory that
4008 userspace has reserved for the array. It must be at least the size dictated
4009 by "mmu_type" and "params".
4011 While KVM_RUN is active, the shared region is under control of KVM. Its
4012 contents are undefined, and any modification by userspace results in
4013 boundedly undefined behavior.
4015 On return from KVM_RUN, the shared region will reflect the current state of
4016 the guest's TLB. If userspace makes any changes, it must call KVM_DIRTY_TLB
4017 to tell KVM which entries have been changed, prior to calling KVM_RUN again
4020 For mmu types KVM_MMU_FSL_BOOKE_NOHV and KVM_MMU_FSL_BOOKE_HV:
4021 - The "params" field is of type "struct kvm_book3e_206_tlb_params".
4022 - The "array" field points to an array of type "struct
4023 kvm_book3e_206_tlb_entry".
4024 - The array consists of all entries in the first TLB, followed by all
4025 entries in the second TLB.
4026 - Within a TLB, entries are ordered first by increasing set number. Within a
4027 set, entries are ordered by way (increasing ESEL).
4028 - The hash for determining set number in TLB0 is: (MAS2 >> 12) & (num_sets - 1)
4029 where "num_sets" is the tlb_sizes[] value divided by the tlb_ways[] value.
4030 - The tsize field of mas1 shall be set to 4K on TLB0, even though the
4031 hardware ignores this value for TLB0.
4033 6.4 KVM_CAP_S390_CSS_SUPPORT
4038 Returns: 0 on success; -1 on error
4040 This capability enables support for handling of channel I/O instructions.
4042 TEST PENDING INTERRUPTION and the interrupt portion of TEST SUBCHANNEL are
4043 handled in-kernel, while the other I/O instructions are passed to userspace.
4045 When this capability is enabled, KVM_EXIT_S390_TSCH will occur on TEST
4046 SUBCHANNEL intercepts.
4048 Note that even though this capability is enabled per-vcpu, the complete
4049 virtual machine is affected.
4055 Parameters: args[0] defines whether the proxy facility is active
4056 Returns: 0 on success; -1 on error
4058 This capability enables or disables the delivery of interrupts through the
4059 external proxy facility.
4061 When enabled (args[0] != 0), every time the guest gets an external interrupt
4062 delivered, it automatically exits into user space with a KVM_EXIT_EPR exit
4063 to receive the topmost interrupt vector.
4065 When disabled (args[0] == 0), behavior is as if this facility is unsupported.
4067 When this capability is enabled, KVM_EXIT_EPR can occur.
4069 6.6 KVM_CAP_IRQ_MPIC
4072 Parameters: args[0] is the MPIC device fd
4073 args[1] is the MPIC CPU number for this vcpu
4075 This capability connects the vcpu to an in-kernel MPIC device.
4077 6.7 KVM_CAP_IRQ_XICS
4081 Parameters: args[0] is the XICS device fd
4082 args[1] is the XICS CPU number (server ID) for this vcpu
4084 This capability connects the vcpu to an in-kernel XICS device.
4086 6.8 KVM_CAP_S390_IRQCHIP
4092 This capability enables the in-kernel irqchip for s390. Please refer to
4093 "4.24 KVM_CREATE_IRQCHIP" for details.
4095 6.9 KVM_CAP_MIPS_FPU
4099 Parameters: args[0] is reserved for future use (should be 0).
4101 This capability allows the use of the host Floating Point Unit by the guest. It
4102 allows the Config1.FP bit to be set to enable the FPU in the guest. Once this is
4103 done the KVM_REG_MIPS_FPR_* and KVM_REG_MIPS_FCR_* registers can be accessed
4104 (depending on the current guest FPU register mode), and the Status.FR,
4105 Config5.FRE bits are accessible via the KVM API and also from the guest,
4106 depending on them being supported by the FPU.
4108 6.10 KVM_CAP_MIPS_MSA
4112 Parameters: args[0] is reserved for future use (should be 0).
4114 This capability allows the use of the MIPS SIMD Architecture (MSA) by the guest.
4115 It allows the Config3.MSAP bit to be set to enable the use of MSA by the guest.
4116 Once this is done the KVM_REG_MIPS_VEC_* and KVM_REG_MIPS_MSA_* registers can be
4117 accessed, and the Config5.MSAEn bit is accessible via the KVM API and also from
4120 6.74 KVM_CAP_SYNC_REGS
4121 Architectures: s390, x86
4122 Target: s390: always enabled, x86: vcpu
4124 Returns: x86: KVM_CHECK_EXTENSION returns a bit-array indicating which register
4125 sets are supported (bitfields defined in arch/x86/include/uapi/asm/kvm.h).
4127 As described above in the kvm_sync_regs struct info in section 5 (kvm_run):
4128 KVM_CAP_SYNC_REGS "allow[s] userspace to access certain guest registers
4129 without having to call SET/GET_*REGS". This reduces overhead by eliminating
4130 repeated ioctl calls for setting and/or getting register values. This is
4131 particularly important when userspace is making synchronous guest state
4132 modifications, e.g. when emulating and/or intercepting instructions in
4135 For s390 specifics, please refer to the source code.
4138 - the register sets to be copied out to kvm_run are selectable
4139 by userspace (rather that all sets being copied out for every exit).
4140 - vcpu_events are available in addition to regs and sregs.
4142 For x86, the 'kvm_valid_regs' field of struct kvm_run is overloaded to
4143 function as an input bit-array field set by userspace to indicate the
4144 specific register sets to be copied out on the next exit.
4146 To indicate when userspace has modified values that should be copied into
4147 the vCPU, the all architecture bitarray field, 'kvm_dirty_regs' must be set.
4148 This is done using the same bitflags as for the 'kvm_valid_regs' field.
4149 If the dirty bit is not set, then the register set values will not be copied
4150 into the vCPU even if they've been modified.
4152 Unused bitfields in the bitarrays must be set to zero.
4154 struct kvm_sync_regs {
4155 struct kvm_regs regs;
4156 struct kvm_sregs sregs;
4157 struct kvm_vcpu_events events;
4160 7. Capabilities that can be enabled on VMs
4161 ------------------------------------------
4163 There are certain capabilities that change the behavior of the virtual
4164 machine when enabled. To enable them, please see section 4.37. Below
4165 you can find a list of capabilities and what their effect on the VM
4166 is when enabling them.
4168 The following information is provided along with the description:
4170 Architectures: which instruction set architectures provide this ioctl.
4171 x86 includes both i386 and x86_64.
4173 Parameters: what parameters are accepted by the capability.
4175 Returns: the return value. General error numbers (EBADF, ENOMEM, EINVAL)
4176 are not detailed, but errors with specific meanings are.
4179 7.1 KVM_CAP_PPC_ENABLE_HCALL
4182 Parameters: args[0] is the sPAPR hcall number
4183 args[1] is 0 to disable, 1 to enable in-kernel handling
4185 This capability controls whether individual sPAPR hypercalls (hcalls)
4186 get handled by the kernel or not. Enabling or disabling in-kernel
4187 handling of an hcall is effective across the VM. On creation, an
4188 initial set of hcalls are enabled for in-kernel handling, which
4189 consists of those hcalls for which in-kernel handlers were implemented
4190 before this capability was implemented. If disabled, the kernel will
4191 not to attempt to handle the hcall, but will always exit to userspace
4192 to handle it. Note that it may not make sense to enable some and
4193 disable others of a group of related hcalls, but KVM does not prevent
4194 userspace from doing that.
4196 If the hcall number specified is not one that has an in-kernel
4197 implementation, the KVM_ENABLE_CAP ioctl will fail with an EINVAL
4200 7.2 KVM_CAP_S390_USER_SIGP
4205 This capability controls which SIGP orders will be handled completely in user
4206 space. With this capability enabled, all fast orders will be handled completely
4212 - CONDITIONAL EMERGENCY SIGNAL
4214 All other orders will be handled completely in user space.
4216 Only privileged operation exceptions will be checked for in the kernel (or even
4217 in the hardware prior to interception). If this capability is not enabled, the
4218 old way of handling SIGP orders is used (partially in kernel and user space).
4220 7.3 KVM_CAP_S390_VECTOR_REGISTERS
4224 Returns: 0 on success, negative value on error
4226 Allows use of the vector registers introduced with z13 processor, and
4227 provides for the synchronization between host and user space. Will
4228 return -EINVAL if the machine does not support vectors.
4230 7.4 KVM_CAP_S390_USER_STSI
4235 This capability allows post-handlers for the STSI instruction. After
4236 initial handling in the kernel, KVM exits to user space with
4237 KVM_EXIT_S390_STSI to allow user space to insert further data.
4239 Before exiting to userspace, kvm handlers should fill in s390_stsi field of
4250 @addr - guest address of STSI SYSIB
4254 @ar - access register number
4256 KVM handlers should exit to userspace with rc = -EREMOTE.
4258 7.5 KVM_CAP_SPLIT_IRQCHIP
4261 Parameters: args[0] - number of routes reserved for userspace IOAPICs
4262 Returns: 0 on success, -1 on error
4264 Create a local apic for each processor in the kernel. This can be used
4265 instead of KVM_CREATE_IRQCHIP if the userspace VMM wishes to emulate the
4266 IOAPIC and PIC (and also the PIT, even though this has to be enabled
4269 This capability also enables in kernel routing of interrupt requests;
4270 when KVM_CAP_SPLIT_IRQCHIP only routes of KVM_IRQ_ROUTING_MSI type are
4271 used in the IRQ routing table. The first args[0] MSI routes are reserved
4272 for the IOAPIC pins. Whenever the LAPIC receives an EOI for these routes,
4273 a KVM_EXIT_IOAPIC_EOI vmexit will be reported to userspace.
4275 Fails if VCPU has already been created, or if the irqchip is already in the
4276 kernel (i.e. KVM_CREATE_IRQCHIP has already been called).
4283 Allows use of runtime-instrumentation introduced with zEC12 processor.
4284 Will return -EINVAL if the machine does not support runtime-instrumentation.
4285 Will return -EBUSY if a VCPU has already been created.
4287 7.7 KVM_CAP_X2APIC_API
4290 Parameters: args[0] - features that should be enabled
4291 Returns: 0 on success, -EINVAL when args[0] contains invalid features
4293 Valid feature flags in args[0] are
4295 #define KVM_X2APIC_API_USE_32BIT_IDS (1ULL << 0)
4296 #define KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK (1ULL << 1)
4298 Enabling KVM_X2APIC_API_USE_32BIT_IDS changes the behavior of
4299 KVM_SET_GSI_ROUTING, KVM_SIGNAL_MSI, KVM_SET_LAPIC, and KVM_GET_LAPIC,
4300 allowing the use of 32-bit APIC IDs. See KVM_CAP_X2APIC_API in their
4301 respective sections.
4303 KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK must be enabled for x2APIC to work
4304 in logical mode or with more than 255 VCPUs. Otherwise, KVM treats 0xff
4305 as a broadcast even in x2APIC mode in order to support physical x2APIC
4306 without interrupt remapping. This is undesirable in logical mode,
4307 where 0xff represents CPUs 0-7 in cluster 0.
4309 7.8 KVM_CAP_S390_USER_INSTR0
4314 With this capability enabled, all illegal instructions 0x0000 (2 bytes) will
4315 be intercepted and forwarded to user space. User space can use this
4316 mechanism e.g. to realize 2-byte software breakpoints. The kernel will
4317 not inject an operating exception for these instructions, user space has
4318 to take care of that.
4320 This capability can be enabled dynamically even if VCPUs were already
4321 created and are running.
4327 Returns: 0 on success; -EINVAL if the machine does not support
4328 guarded storage; -EBUSY if a VCPU has already been created.
4330 Allows use of guarded storage for the KVM guest.
4332 7.10 KVM_CAP_S390_AIS
4337 Allow use of adapter-interruption suppression.
4338 Returns: 0 on success; -EBUSY if a VCPU has already been created.
4340 7.11 KVM_CAP_PPC_SMT
4343 Parameters: vsmt_mode, flags
4345 Enabling this capability on a VM provides userspace with a way to set
4346 the desired virtual SMT mode (i.e. the number of virtual CPUs per
4347 virtual core). The virtual SMT mode, vsmt_mode, must be a power of 2
4348 between 1 and 8. On POWER8, vsmt_mode must also be no greater than
4349 the number of threads per subcore for the host. Currently flags must
4350 be 0. A successful call to enable this capability will result in
4351 vsmt_mode being returned when the KVM_CAP_PPC_SMT capability is
4352 subsequently queried for the VM. This capability is only supported by
4353 HV KVM, and can only be set before any VCPUs have been created.
4354 The KVM_CAP_PPC_SMT_POSSIBLE capability indicates which virtual SMT
4355 modes are available.
4357 7.12 KVM_CAP_PPC_FWNMI
4362 With this capability a machine check exception in the guest address
4363 space will cause KVM to exit the guest with NMI exit reason. This
4364 enables QEMU to build error log and branch to guest kernel registered
4365 machine check handling routine. Without this capability KVM will
4366 branch to guests' 0x200 interrupt vector.
4368 7.13 KVM_CAP_X86_DISABLE_EXITS
4371 Parameters: args[0] defines which exits are disabled
4372 Returns: 0 on success, -EINVAL when args[0] contains invalid exits
4374 Valid bits in args[0] are
4376 #define KVM_X86_DISABLE_EXITS_MWAIT (1 << 0)
4377 #define KVM_X86_DISABLE_EXITS_HLT (1 << 1)
4379 Enabling this capability on a VM provides userspace with a way to no
4380 longer intercept some instructions for improved latency in some
4381 workloads, and is suggested when vCPUs are associated to dedicated
4382 physical CPUs. More bits can be added in the future; userspace can
4383 just pass the KVM_CHECK_EXTENSION result to KVM_ENABLE_CAP to disable
4386 Do not enable KVM_FEATURE_PV_UNHALT if you disable HLT exits.
4388 8. Other capabilities.
4389 ----------------------
4391 This section lists capabilities that give information about other
4392 features of the KVM implementation.
4394 8.1 KVM_CAP_PPC_HWRNG
4398 This capability, if KVM_CHECK_EXTENSION indicates that it is
4399 available, means that that the kernel has an implementation of the
4400 H_RANDOM hypercall backed by a hardware random-number generator.
4401 If present, the kernel H_RANDOM handler can be enabled for guest use
4402 with the KVM_CAP_PPC_ENABLE_HCALL capability.
4404 8.2 KVM_CAP_HYPERV_SYNIC
4407 This capability, if KVM_CHECK_EXTENSION indicates that it is
4408 available, means that that the kernel has an implementation of the
4409 Hyper-V Synthetic interrupt controller(SynIC). Hyper-V SynIC is
4410 used to support Windows Hyper-V based guest paravirt drivers(VMBus).
4412 In order to use SynIC, it has to be activated by setting this
4413 capability via KVM_ENABLE_CAP ioctl on the vcpu fd. Note that this
4414 will disable the use of APIC hardware virtualization even if supported
4415 by the CPU, as it's incompatible with SynIC auto-EOI behavior.
4417 8.3 KVM_CAP_PPC_RADIX_MMU
4421 This capability, if KVM_CHECK_EXTENSION indicates that it is
4422 available, means that that the kernel can support guests using the
4423 radix MMU defined in Power ISA V3.00 (as implemented in the POWER9
4426 8.4 KVM_CAP_PPC_HASH_MMU_V3
4430 This capability, if KVM_CHECK_EXTENSION indicates that it is
4431 available, means that that the kernel can support guests using the
4432 hashed page table MMU defined in Power ISA V3.00 (as implemented in
4433 the POWER9 processor), including in-memory segment tables.
4439 This capability, if KVM_CHECK_EXTENSION on the main kvm handle indicates that
4440 it is available, means that full hardware assisted virtualization capabilities
4441 of the hardware are available for use through KVM. An appropriate
4442 KVM_VM_MIPS_* type must be passed to KVM_CREATE_VM to create a VM which
4445 If KVM_CHECK_EXTENSION on a kvm VM handle indicates that this capability is
4446 available, it means that the VM is using full hardware assisted virtualization
4447 capabilities of the hardware. This is useful to check after creating a VM with
4448 KVM_VM_MIPS_DEFAULT.
4450 The value returned by KVM_CHECK_EXTENSION should be compared against known
4451 values (see below). All other values are reserved. This is to allow for the
4452 possibility of other hardware assisted virtualization implementations which
4453 may be incompatible with the MIPS VZ ASE.
4455 0: The trap & emulate implementation is in use to run guest code in user
4456 mode. Guest virtual memory segments are rearranged to fit the guest in the
4457 user mode address space.
4459 1: The MIPS VZ ASE is in use, providing full hardware assisted
4460 virtualization, including standard guest virtual memory segments.
4466 This capability, if KVM_CHECK_EXTENSION on the main kvm handle indicates that
4467 it is available, means that the trap & emulate implementation is available to
4468 run guest code in user mode, even if KVM_CAP_MIPS_VZ indicates that hardware
4469 assisted virtualisation is also available. KVM_VM_MIPS_TE (0) must be passed
4470 to KVM_CREATE_VM to create a VM which utilises it.
4472 If KVM_CHECK_EXTENSION on a kvm VM handle indicates that this capability is
4473 available, it means that the VM is using trap & emulate.
4475 8.7 KVM_CAP_MIPS_64BIT
4479 This capability indicates the supported architecture type of the guest, i.e. the
4480 supported register and address width.
4482 The values returned when this capability is checked by KVM_CHECK_EXTENSION on a
4483 kvm VM handle correspond roughly to the CP0_Config.AT register field, and should
4484 be checked specifically against known values (see below). All other values are
4487 0: MIPS32 or microMIPS32.
4488 Both registers and addresses are 32-bits wide.
4489 It will only be possible to run 32-bit guest code.
4491 1: MIPS64 or microMIPS64 with access only to 32-bit compatibility segments.
4492 Registers are 64-bits wide, but addresses are 32-bits wide.
4493 64-bit guest code may run but cannot access MIPS64 memory segments.
4494 It will also be possible to run 32-bit guest code.
4496 2: MIPS64 or microMIPS64 with access to all address segments.
4497 Both registers and addresses are 64-bits wide.
4498 It will be possible to run 64-bit or 32-bit guest code.
4500 8.9 KVM_CAP_ARM_USER_IRQ
4502 Architectures: arm, arm64
4503 This capability, if KVM_CHECK_EXTENSION indicates that it is available, means
4504 that if userspace creates a VM without an in-kernel interrupt controller, it
4505 will be notified of changes to the output level of in-kernel emulated devices,
4506 which can generate virtual interrupts, presented to the VM.
4507 For such VMs, on every return to userspace, the kernel
4508 updates the vcpu's run->s.regs.device_irq_level field to represent the actual
4509 output level of the device.
4511 Whenever kvm detects a change in the device output level, kvm guarantees at
4512 least one return to userspace before running the VM. This exit could either
4513 be a KVM_EXIT_INTR or any other exit event, like KVM_EXIT_MMIO. This way,
4514 userspace can always sample the device output level and re-compute the state of
4515 the userspace interrupt controller. Userspace should always check the state
4516 of run->s.regs.device_irq_level on every kvm exit.
4517 The value in run->s.regs.device_irq_level can represent both level and edge
4518 triggered interrupt signals, depending on the device. Edge triggered interrupt
4519 signals will exit to userspace with the bit in run->s.regs.device_irq_level
4520 set exactly once per edge signal.
4522 The field run->s.regs.device_irq_level is available independent of
4523 run->kvm_valid_regs or run->kvm_dirty_regs bits.
4525 If KVM_CAP_ARM_USER_IRQ is supported, the KVM_CHECK_EXTENSION ioctl returns a
4526 number larger than 0 indicating the version of this capability is implemented
4527 and thereby which bits in in run->s.regs.device_irq_level can signal values.
4529 Currently the following bits are defined for the device_irq_level bitmap:
4531 KVM_CAP_ARM_USER_IRQ >= 1:
4533 KVM_ARM_DEV_EL1_VTIMER - EL1 virtual timer
4534 KVM_ARM_DEV_EL1_PTIMER - EL1 physical timer
4535 KVM_ARM_DEV_PMU - ARM PMU overflow interrupt signal
4537 Future versions of kvm may implement additional events. These will get
4538 indicated by returning a higher number from KVM_CHECK_EXTENSION and will be
4541 8.10 KVM_CAP_PPC_SMT_POSSIBLE
4545 Querying this capability returns a bitmap indicating the possible
4546 virtual SMT modes that can be set using KVM_CAP_PPC_SMT. If bit N
4547 (counting from the right) is set, then a virtual SMT mode of 2^N is
4550 8.11 KVM_CAP_HYPERV_SYNIC2
4554 This capability enables a newer version of Hyper-V Synthetic interrupt
4555 controller (SynIC). The only difference with KVM_CAP_HYPERV_SYNIC is that KVM
4556 doesn't clear SynIC message and event flags pages when they are enabled by
4557 writing to the respective MSRs.
4559 8.12 KVM_CAP_HYPERV_VP_INDEX
4563 This capability indicates that userspace can load HV_X64_MSR_VP_INDEX msr. Its
4564 value is used to denote the target vcpu for a SynIC interrupt. For
4565 compatibilty, KVM initializes this msr to KVM's internal vcpu index. When this
4566 capability is absent, userspace can still query this msr's value.
4568 8.13 KVM_CAP_S390_AIS_MIGRATION
4573 This capability indicates if the flic device will be able to get/set the
4574 AIS states for migration via the KVM_DEV_FLIC_AISM_ALL attribute and allows
4575 to discover this without having to create a flic device.
4577 8.14 KVM_CAP_S390_PSW
4581 This capability indicates that the PSW is exposed via the kvm_run structure.
4583 8.15 KVM_CAP_S390_GMAP
4587 This capability indicates that the user space memory used as guest mapping can
4588 be anywhere in the user memory address space, as long as the memory slots are
4589 aligned and sized to a segment (1MB) boundary.
4591 8.16 KVM_CAP_S390_COW
4595 This capability indicates that the user space memory used as guest mapping can
4596 use copy-on-write semantics as well as dirty pages tracking via read-only page
4599 8.17 KVM_CAP_S390_BPB
4603 This capability indicates that kvm will implement the interfaces to handle
4604 reset, migration and nested KVM for branch prediction blocking. The stfle
4605 facility 82 should not be provided to the guest without this capability.