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 on the Linux version number.
57 Instead, kvm defines extension identifiers and a facility to query
58 whether a particular extension identifier is available. If it is, a
59 set of ioctls is available for application use.
65 This section describes ioctls that can be used to control kvm guests.
66 For each ioctl, the following information is provided along with a
69 Capability: which KVM extension provides this ioctl. Can be 'basic',
70 which means that is will be provided by any kernel that supports
71 API version 12 (see section 4.1), or a KVM_CAP_xyz constant, which
72 means availability needs to be checked with KVM_CHECK_EXTENSION
75 Architectures: which instruction set architectures provide this ioctl.
76 x86 includes both i386 and x86_64.
78 Type: system, vm, or vcpu.
80 Parameters: what parameters are accepted by the ioctl.
82 Returns: the return value. General error numbers (EBADF, ENOMEM, EINVAL)
83 are not detailed, but errors with specific meanings are.
86 4.1 KVM_GET_API_VERSION
92 Returns: the constant KVM_API_VERSION (=12)
94 This identifies the API version as the stable kvm API. It is not
95 expected that this number will change. However, Linux 2.6.20 and
96 2.6.21 report earlier versions; these are not documented and not
97 supported. Applications should refuse to run if KVM_GET_API_VERSION
98 returns a value other than 12. If this check passes, all ioctls
99 described as 'basic' will be available.
107 Parameters: machine type identifier (KVM_VM_*)
108 Returns: a VM fd that can be used to control the new virtual machine.
110 The new VM has no virtual cpus and no memory. An mmap() of a VM fd
111 will access the virtual machine's physical address space; offset zero
112 corresponds to guest physical address zero. Use of mmap() on a VM fd
113 is discouraged if userspace memory allocation (KVM_CAP_USER_MEMORY) is
115 You most certainly want to use 0 as machine type.
117 In order to create user controlled virtual machines on S390, check
118 KVM_CAP_S390_UCONTROL and use the flag KVM_VM_S390_UCONTROL as
119 privileged user (CAP_SYS_ADMIN).
122 4.3 KVM_GET_MSR_INDEX_LIST
127 Parameters: struct kvm_msr_list (in/out)
128 Returns: 0 on success; -1 on error
130 E2BIG: the msr index list is to be to fit in the array specified by
133 struct kvm_msr_list {
134 __u32 nmsrs; /* number of msrs in entries */
138 This ioctl returns the guest msrs that are supported. The list varies
139 by kvm version and host processor, but does not change otherwise. The
140 user fills in the size of the indices array in nmsrs, and in return
141 kvm adjusts nmsrs to reflect the actual number of msrs and fills in
142 the indices array with their numbers.
144 Note: if kvm indicates supports MCE (KVM_CAP_MCE), then the MCE bank MSRs are
145 not returned in the MSR list, as different vcpus can have a different number
146 of banks, as set via the KVM_X86_SETUP_MCE ioctl.
149 4.4 KVM_CHECK_EXTENSION
154 Parameters: extension identifier (KVM_CAP_*)
155 Returns: 0 if unsupported; 1 (or some other positive integer) if supported
157 The API allows the application to query about extensions to the core
158 kvm API. Userspace passes an extension identifier (an integer) and
159 receives an integer that describes the extension availability.
160 Generally 0 means no and 1 means yes, but some extensions may report
161 additional information in the integer return value.
164 4.5 KVM_GET_VCPU_MMAP_SIZE
170 Returns: size of vcpu mmap area, in bytes
172 The KVM_RUN ioctl (cf.) communicates with userspace via a shared
173 memory region. This ioctl returns the size of that region. See the
174 KVM_RUN documentation for details.
177 4.6 KVM_SET_MEMORY_REGION
182 Parameters: struct kvm_memory_region (in)
183 Returns: 0 on success, -1 on error
185 This ioctl is obsolete and has been removed.
193 Parameters: vcpu id (apic id on x86)
194 Returns: vcpu fd on success, -1 on error
196 This API adds a vcpu to a virtual machine. The vcpu id is a small integer
197 in the range [0, max_vcpus).
199 The recommended max_vcpus value can be retrieved using the KVM_CAP_NR_VCPUS of
200 the KVM_CHECK_EXTENSION ioctl() at run-time.
201 The maximum possible value for max_vcpus can be retrieved using the
202 KVM_CAP_MAX_VCPUS of the KVM_CHECK_EXTENSION ioctl() at run-time.
204 If the KVM_CAP_NR_VCPUS does not exist, you should assume that max_vcpus is 4
206 If the KVM_CAP_MAX_VCPUS does not exist, you should assume that max_vcpus is
207 same as the value returned from KVM_CAP_NR_VCPUS.
209 On powerpc using book3s_hv mode, the vcpus are mapped onto virtual
210 threads in one or more virtual CPU cores. (This is because the
211 hardware requires all the hardware threads in a CPU core to be in the
212 same partition.) The KVM_CAP_PPC_SMT capability indicates the number
213 of vcpus per virtual core (vcore). The vcore id is obtained by
214 dividing the vcpu id by the number of vcpus per vcore. The vcpus in a
215 given vcore will always be in the same physical core as each other
216 (though that might be a different physical core from time to time).
217 Userspace can control the threading (SMT) mode of the guest by its
218 allocation of vcpu ids. For example, if userspace wants
219 single-threaded guest vcpus, it should make all vcpu ids be a multiple
220 of the number of vcpus per vcore.
222 On powerpc using book3s_hv mode, the vcpus are mapped onto virtual
223 threads in one or more virtual CPU cores. (This is because the
224 hardware requires all the hardware threads in a CPU core to be in the
225 same partition.) The KVM_CAP_PPC_SMT capability indicates the number
226 of vcpus per virtual core (vcore). The vcore id is obtained by
227 dividing the vcpu id by the number of vcpus per vcore. The vcpus in a
228 given vcore will always be in the same physical core as each other
229 (though that might be a different physical core from time to time).
230 Userspace can control the threading (SMT) mode of the guest by its
231 allocation of vcpu ids. For example, if userspace wants
232 single-threaded guest vcpus, it should make all vcpu ids be a multiple
233 of the number of vcpus per vcore.
235 For virtual cpus that have been created with S390 user controlled virtual
236 machines, the resulting vcpu fd can be memory mapped at page offset
237 KVM_S390_SIE_PAGE_OFFSET in order to obtain a memory map of the virtual
238 cpu's hardware control block.
241 4.8 KVM_GET_DIRTY_LOG (vm ioctl)
246 Parameters: struct kvm_dirty_log (in/out)
247 Returns: 0 on success, -1 on error
249 /* for KVM_GET_DIRTY_LOG */
250 struct kvm_dirty_log {
254 void __user *dirty_bitmap; /* one bit per page */
259 Given a memory slot, return a bitmap containing any pages dirtied
260 since the last call to this ioctl. Bit 0 is the first page in the
261 memory slot. Ensure the entire structure is cleared to avoid padding
265 4.9 KVM_SET_MEMORY_ALIAS
270 Parameters: struct kvm_memory_alias (in)
271 Returns: 0 (success), -1 (error)
273 This ioctl is obsolete and has been removed.
282 Returns: 0 on success, -1 on error
284 EINTR: an unmasked signal is pending
286 This ioctl is used to run a guest virtual cpu. While there are no
287 explicit parameters, there is an implicit parameter block that can be
288 obtained by mmap()ing the vcpu fd at offset 0, with the size given by
289 KVM_GET_VCPU_MMAP_SIZE. The parameter block is formatted as a 'struct
290 kvm_run' (see below).
296 Architectures: all except ARM
298 Parameters: struct kvm_regs (out)
299 Returns: 0 on success, -1 on error
301 Reads the general purpose registers from the vcpu.
305 /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */
306 __u64 rax, rbx, rcx, rdx;
307 __u64 rsi, rdi, rsp, rbp;
308 __u64 r8, r9, r10, r11;
309 __u64 r12, r13, r14, r15;
317 Architectures: all except ARM
319 Parameters: struct kvm_regs (in)
320 Returns: 0 on success, -1 on error
322 Writes the general purpose registers into the vcpu.
324 See KVM_GET_REGS for the data structure.
330 Architectures: x86, ppc
332 Parameters: struct kvm_sregs (out)
333 Returns: 0 on success, -1 on error
335 Reads special registers from the vcpu.
339 struct kvm_segment cs, ds, es, fs, gs, ss;
340 struct kvm_segment tr, ldt;
341 struct kvm_dtable gdt, idt;
342 __u64 cr0, cr2, cr3, cr4, cr8;
345 __u64 interrupt_bitmap[(KVM_NR_INTERRUPTS + 63) / 64];
348 /* ppc -- see arch/powerpc/include/asm/kvm.h */
350 interrupt_bitmap is a bitmap of pending external interrupts. At most
351 one bit may be set. This interrupt has been acknowledged by the APIC
352 but not yet injected into the cpu core.
358 Architectures: x86, ppc
360 Parameters: struct kvm_sregs (in)
361 Returns: 0 on success, -1 on error
363 Writes special registers into the vcpu. See KVM_GET_SREGS for the
372 Parameters: struct kvm_translation (in/out)
373 Returns: 0 on success, -1 on error
375 Translates a virtual address according to the vcpu's current address
378 struct kvm_translation {
380 __u64 linear_address;
383 __u64 physical_address;
394 Architectures: x86, ppc
396 Parameters: struct kvm_interrupt (in)
397 Returns: 0 on success, -1 on error
399 Queues a hardware interrupt vector to be injected. This is only
400 useful if in-kernel local APIC or equivalent is not used.
402 /* for KVM_INTERRUPT */
403 struct kvm_interrupt {
410 Note 'irq' is an interrupt vector, not an interrupt pin or line.
414 Queues an external interrupt to be injected. This ioctl is overleaded
415 with 3 different irq values:
419 This injects an edge type external interrupt into the guest once it's ready
420 to receive interrupts. When injected, the interrupt is done.
422 b) KVM_INTERRUPT_UNSET
424 This unsets any pending interrupt.
426 Only available with KVM_CAP_PPC_UNSET_IRQ.
428 c) KVM_INTERRUPT_SET_LEVEL
430 This injects a level type external interrupt into the guest context. The
431 interrupt stays pending until a specific ioctl with KVM_INTERRUPT_UNSET
434 Only available with KVM_CAP_PPC_IRQ_LEVEL.
436 Note that any value for 'irq' other than the ones stated above is invalid
437 and incurs unexpected behavior.
448 Support for this has been removed. Use KVM_SET_GUEST_DEBUG instead.
456 Parameters: struct kvm_msrs (in/out)
457 Returns: 0 on success, -1 on error
459 Reads model-specific registers from the vcpu. Supported msr indices can
460 be obtained using KVM_GET_MSR_INDEX_LIST.
463 __u32 nmsrs; /* number of msrs in entries */
466 struct kvm_msr_entry entries[0];
469 struct kvm_msr_entry {
475 Application code should set the 'nmsrs' member (which indicates the
476 size of the entries array) and the 'index' member of each array entry.
477 kvm will fill in the 'data' member.
485 Parameters: struct kvm_msrs (in)
486 Returns: 0 on success, -1 on error
488 Writes model-specific registers to the vcpu. See KVM_GET_MSRS for the
491 Application code should set the 'nmsrs' member (which indicates the
492 size of the entries array), and the 'index' and 'data' members of each
501 Parameters: struct kvm_cpuid (in)
502 Returns: 0 on success, -1 on error
504 Defines the vcpu responses to the cpuid instruction. Applications
505 should use the KVM_SET_CPUID2 ioctl if available.
508 struct kvm_cpuid_entry {
517 /* for KVM_SET_CPUID */
521 struct kvm_cpuid_entry entries[0];
525 4.21 KVM_SET_SIGNAL_MASK
530 Parameters: struct kvm_signal_mask (in)
531 Returns: 0 on success, -1 on error
533 Defines which signals are blocked during execution of KVM_RUN. This
534 signal mask temporarily overrides the threads signal mask. Any
535 unblocked signal received (except SIGKILL and SIGSTOP, which retain
536 their traditional behaviour) will cause KVM_RUN to return with -EINTR.
538 Note the signal will only be delivered if not blocked by the original
541 /* for KVM_SET_SIGNAL_MASK */
542 struct kvm_signal_mask {
553 Parameters: struct kvm_fpu (out)
554 Returns: 0 on success, -1 on error
556 Reads the floating point state from the vcpu.
558 /* for KVM_GET_FPU and KVM_SET_FPU */
563 __u8 ftwx; /* in fxsave format */
579 Parameters: struct kvm_fpu (in)
580 Returns: 0 on success, -1 on error
582 Writes the floating point state to the vcpu.
584 /* for KVM_GET_FPU and KVM_SET_FPU */
589 __u8 ftwx; /* in fxsave format */
600 4.24 KVM_CREATE_IRQCHIP
602 Capability: KVM_CAP_IRQCHIP
603 Architectures: x86, ia64, ARM
606 Returns: 0 on success, -1 on error
608 Creates an interrupt controller model in the kernel. On x86, creates a virtual
609 ioapic, a virtual PIC (two PICs, nested), and sets up future vcpus to have a
610 local APIC. IRQ routing for GSIs 0-15 is set to both PIC and IOAPIC; GSI 16-23
611 only go to the IOAPIC. On ia64, a IOSAPIC is created. On ARM, a GIC is
617 Capability: KVM_CAP_IRQCHIP
618 Architectures: x86, ia64, arm
620 Parameters: struct kvm_irq_level
621 Returns: 0 on success, -1 on error
623 Sets the level of a GSI input to the interrupt controller model in the kernel.
624 On some architectures it is required that an interrupt controller model has
625 been previously created with KVM_CREATE_IRQCHIP. Note that edge-triggered
626 interrupts require the level to be set to 1 and then back to 0.
628 ARM can signal an interrupt either at the CPU level, or at the in-kernel irqchip
629 (GIC), and for in-kernel irqchip can tell the GIC to use PPIs designated for
630 specific cpus. The irq field is interpreted like this:
632 Â bits: | 31 ... 24 | 23 ... 16 | 15 ... 0 |
633 field: | irq_type | vcpu_index | irq_id |
635 The irq_type field has the following values:
636 - irq_type[0]: out-of-kernel GIC: irq_id 0 is IRQ, irq_id 1 is FIQ
637 - irq_type[1]: in-kernel GIC: SPI, irq_id between 32 and 1019 (incl.)
638 (the vcpu_index field is ignored)
639 - irq_type[2]: in-kernel GIC: PPI, irq_id between 16 and 31 (incl.)
641 (The irq_id field thus corresponds nicely to the IRQ ID in the ARM GIC specs)
643 In both cases, level is used to raise/lower the line.
645 struct kvm_irq_level {
648 __s32 status; /* not used for KVM_IRQ_LEVEL */
650 __u32 level; /* 0 or 1 */
656 Capability: KVM_CAP_IRQCHIP
657 Architectures: x86, ia64
659 Parameters: struct kvm_irqchip (in/out)
660 Returns: 0 on success, -1 on error
662 Reads the state of a kernel interrupt controller created with
663 KVM_CREATE_IRQCHIP into a buffer provided by the caller.
666 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */
669 char dummy[512]; /* reserving space */
670 struct kvm_pic_state pic;
671 struct kvm_ioapic_state ioapic;
678 Capability: KVM_CAP_IRQCHIP
679 Architectures: x86, ia64
681 Parameters: struct kvm_irqchip (in)
682 Returns: 0 on success, -1 on error
684 Sets the state of a kernel interrupt controller created with
685 KVM_CREATE_IRQCHIP from a buffer provided by the caller.
688 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */
691 char dummy[512]; /* reserving space */
692 struct kvm_pic_state pic;
693 struct kvm_ioapic_state ioapic;
698 4.28 KVM_XEN_HVM_CONFIG
700 Capability: KVM_CAP_XEN_HVM
703 Parameters: struct kvm_xen_hvm_config (in)
704 Returns: 0 on success, -1 on error
706 Sets the MSR that the Xen HVM guest uses to initialize its hypercall
707 page, and provides the starting address and size of the hypercall
708 blobs in userspace. When the guest writes the MSR, kvm copies one
709 page of a blob (32- or 64-bit, depending on the vcpu mode) to guest
712 struct kvm_xen_hvm_config {
725 Capability: KVM_CAP_ADJUST_CLOCK
728 Parameters: struct kvm_clock_data (out)
729 Returns: 0 on success, -1 on error
731 Gets the current timestamp of kvmclock as seen by the current guest. In
732 conjunction with KVM_SET_CLOCK, it is used to ensure monotonicity on scenarios
735 struct kvm_clock_data {
736 __u64 clock; /* kvmclock current value */
744 Capability: KVM_CAP_ADJUST_CLOCK
747 Parameters: struct kvm_clock_data (in)
748 Returns: 0 on success, -1 on error
750 Sets the current timestamp of kvmclock to the value specified in its parameter.
751 In conjunction with KVM_GET_CLOCK, it is used to ensure monotonicity on scenarios
754 struct kvm_clock_data {
755 __u64 clock; /* kvmclock current value */
761 4.31 KVM_GET_VCPU_EVENTS
763 Capability: KVM_CAP_VCPU_EVENTS
764 Extended by: KVM_CAP_INTR_SHADOW
767 Parameters: struct kvm_vcpu_event (out)
768 Returns: 0 on success, -1 on error
770 Gets currently pending exceptions, interrupts, and NMIs as well as related
773 struct kvm_vcpu_events {
797 KVM_VCPUEVENT_VALID_SHADOW may be set in the flags field to signal that
798 interrupt.shadow contains a valid state. Otherwise, this field is undefined.
801 4.32 KVM_SET_VCPU_EVENTS
803 Capability: KVM_CAP_VCPU_EVENTS
804 Extended by: KVM_CAP_INTR_SHADOW
807 Parameters: struct kvm_vcpu_event (in)
808 Returns: 0 on success, -1 on error
810 Set pending exceptions, interrupts, and NMIs as well as related states of the
813 See KVM_GET_VCPU_EVENTS for the data structure.
815 Fields that may be modified asynchronously by running VCPUs can be excluded
816 from the update. These fields are nmi.pending and sipi_vector. Keep the
817 corresponding bits in the flags field cleared to suppress overwriting the
818 current in-kernel state. The bits are:
820 KVM_VCPUEVENT_VALID_NMI_PENDING - transfer nmi.pending to the kernel
821 KVM_VCPUEVENT_VALID_SIPI_VECTOR - transfer sipi_vector
823 If KVM_CAP_INTR_SHADOW is available, KVM_VCPUEVENT_VALID_SHADOW can be set in
824 the flags field to signal that interrupt.shadow contains a valid state and
825 shall be written into the VCPU.
828 4.33 KVM_GET_DEBUGREGS
830 Capability: KVM_CAP_DEBUGREGS
833 Parameters: struct kvm_debugregs (out)
834 Returns: 0 on success, -1 on error
836 Reads debug registers from the vcpu.
838 struct kvm_debugregs {
847 4.34 KVM_SET_DEBUGREGS
849 Capability: KVM_CAP_DEBUGREGS
852 Parameters: struct kvm_debugregs (in)
853 Returns: 0 on success, -1 on error
855 Writes debug registers into the vcpu.
857 See KVM_GET_DEBUGREGS for the data structure. The flags field is unused
858 yet and must be cleared on entry.
861 4.35 KVM_SET_USER_MEMORY_REGION
863 Capability: KVM_CAP_USER_MEM
866 Parameters: struct kvm_userspace_memory_region (in)
867 Returns: 0 on success, -1 on error
869 struct kvm_userspace_memory_region {
872 __u64 guest_phys_addr;
873 __u64 memory_size; /* bytes */
874 __u64 userspace_addr; /* start of the userspace allocated memory */
877 /* for kvm_memory_region::flags */
878 #define KVM_MEM_LOG_DIRTY_PAGES (1UL << 0)
879 #define KVM_MEM_READONLY (1UL << 1)
881 This ioctl allows the user to create or modify a guest physical memory
882 slot. When changing an existing slot, it may be moved in the guest
883 physical memory space, or its flags may be modified. It may not be
884 resized. Slots may not overlap in guest physical address space.
886 Memory for the region is taken starting at the address denoted by the
887 field userspace_addr, which must point at user addressable memory for
888 the entire memory slot size. Any object may back this memory, including
889 anonymous memory, ordinary files, and hugetlbfs.
891 It is recommended that the lower 21 bits of guest_phys_addr and userspace_addr
892 be identical. This allows large pages in the guest to be backed by large
895 The flags field supports two flag, KVM_MEM_LOG_DIRTY_PAGES, which instructs
896 kvm to keep track of writes to memory within the slot. See KVM_GET_DIRTY_LOG
897 ioctl. The KVM_CAP_READONLY_MEM capability indicates the availability of the
898 KVM_MEM_READONLY flag. When this flag is set for a memory region, KVM only
899 allows read accesses. Writes will be posted to userspace as KVM_EXIT_MMIO
902 When the KVM_CAP_SYNC_MMU capability is available, changes in the backing of
903 the memory region are automatically reflected into the guest. For example, an
904 mmap() that affects the region will be made visible immediately. Another
905 example is madvise(MADV_DROP).
907 It is recommended to use this API instead of the KVM_SET_MEMORY_REGION ioctl.
908 The KVM_SET_MEMORY_REGION does not allow fine grained control over memory
909 allocation and is deprecated.
912 4.36 KVM_SET_TSS_ADDR
914 Capability: KVM_CAP_SET_TSS_ADDR
917 Parameters: unsigned long tss_address (in)
918 Returns: 0 on success, -1 on error
920 This ioctl defines the physical address of a three-page region in the guest
921 physical address space. The region must be within the first 4GB of the
922 guest physical address space and must not conflict with any memory slot
923 or any mmio address. The guest may malfunction if it accesses this memory
926 This ioctl is required on Intel-based hosts. This is needed on Intel hardware
927 because of a quirk in the virtualization implementation (see the internals
928 documentation when it pops into existence).
933 Capability: KVM_CAP_ENABLE_CAP
936 Parameters: struct kvm_enable_cap (in)
937 Returns: 0 on success; -1 on error
939 +Not all extensions are enabled by default. Using this ioctl the application
940 can enable an extension, making it available to the guest.
942 On systems that do not support this ioctl, it always fails. On systems that
943 do support it, it only works for extensions that are supported for enablement.
945 To check if a capability can be enabled, the KVM_CHECK_EXTENSION ioctl should
948 struct kvm_enable_cap {
952 The capability that is supposed to get enabled.
956 A bitfield indicating future enhancements. Has to be 0 for now.
960 Arguments for enabling a feature. If a feature needs initial values to
961 function properly, this is the place to put them.
967 4.38 KVM_GET_MP_STATE
969 Capability: KVM_CAP_MP_STATE
970 Architectures: x86, ia64
972 Parameters: struct kvm_mp_state (out)
973 Returns: 0 on success; -1 on error
975 struct kvm_mp_state {
979 Returns the vcpu's current "multiprocessing state" (though also valid on
980 uniprocessor guests).
984 - KVM_MP_STATE_RUNNABLE: the vcpu is currently running
985 - KVM_MP_STATE_UNINITIALIZED: the vcpu is an application processor (AP)
986 which has not yet received an INIT signal
987 - KVM_MP_STATE_INIT_RECEIVED: the vcpu has received an INIT signal, and is
989 - KVM_MP_STATE_HALTED: the vcpu has executed a HLT instruction and
990 is waiting for an interrupt
991 - KVM_MP_STATE_SIPI_RECEIVED: the vcpu has just received a SIPI (vector
992 accessible via KVM_GET_VCPU_EVENTS)
994 This ioctl is only useful after KVM_CREATE_IRQCHIP. Without an in-kernel
995 irqchip, the multiprocessing state must be maintained by userspace.
998 4.39 KVM_SET_MP_STATE
1000 Capability: KVM_CAP_MP_STATE
1001 Architectures: x86, ia64
1003 Parameters: struct kvm_mp_state (in)
1004 Returns: 0 on success; -1 on error
1006 Sets the vcpu's current "multiprocessing state"; see KVM_GET_MP_STATE for
1009 This ioctl is only useful after KVM_CREATE_IRQCHIP. Without an in-kernel
1010 irqchip, the multiprocessing state must be maintained by userspace.
1013 4.40 KVM_SET_IDENTITY_MAP_ADDR
1015 Capability: KVM_CAP_SET_IDENTITY_MAP_ADDR
1018 Parameters: unsigned long identity (in)
1019 Returns: 0 on success, -1 on error
1021 This ioctl defines the physical address of a one-page region in the guest
1022 physical address space. The region must be within the first 4GB of the
1023 guest physical address space and must not conflict with any memory slot
1024 or any mmio address. The guest may malfunction if it accesses this memory
1027 This ioctl is required on Intel-based hosts. This is needed on Intel hardware
1028 because of a quirk in the virtualization implementation (see the internals
1029 documentation when it pops into existence).
1032 4.41 KVM_SET_BOOT_CPU_ID
1034 Capability: KVM_CAP_SET_BOOT_CPU_ID
1035 Architectures: x86, ia64
1037 Parameters: unsigned long vcpu_id
1038 Returns: 0 on success, -1 on error
1040 Define which vcpu is the Bootstrap Processor (BSP). Values are the same
1041 as the vcpu id in KVM_CREATE_VCPU. If this ioctl is not called, the default
1047 Capability: KVM_CAP_XSAVE
1050 Parameters: struct kvm_xsave (out)
1051 Returns: 0 on success, -1 on error
1057 This ioctl would copy current vcpu's xsave struct to the userspace.
1062 Capability: KVM_CAP_XSAVE
1065 Parameters: struct kvm_xsave (in)
1066 Returns: 0 on success, -1 on error
1072 This ioctl would copy userspace's xsave struct to the kernel.
1077 Capability: KVM_CAP_XCRS
1080 Parameters: struct kvm_xcrs (out)
1081 Returns: 0 on success, -1 on error
1092 struct kvm_xcr xcrs[KVM_MAX_XCRS];
1096 This ioctl would copy current vcpu's xcrs to the userspace.
1101 Capability: KVM_CAP_XCRS
1104 Parameters: struct kvm_xcrs (in)
1105 Returns: 0 on success, -1 on error
1116 struct kvm_xcr xcrs[KVM_MAX_XCRS];
1120 This ioctl would set vcpu's xcr to the value userspace specified.
1123 4.46 KVM_GET_SUPPORTED_CPUID
1125 Capability: KVM_CAP_EXT_CPUID
1128 Parameters: struct kvm_cpuid2 (in/out)
1129 Returns: 0 on success, -1 on error
1134 struct kvm_cpuid_entry2 entries[0];
1137 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX 1
1138 #define KVM_CPUID_FLAG_STATEFUL_FUNC 2
1139 #define KVM_CPUID_FLAG_STATE_READ_NEXT 4
1141 struct kvm_cpuid_entry2 {
1152 This ioctl returns x86 cpuid features which are supported by both the hardware
1153 and kvm. Userspace can use the information returned by this ioctl to
1154 construct cpuid information (for KVM_SET_CPUID2) that is consistent with
1155 hardware, kernel, and userspace capabilities, and with user requirements (for
1156 example, the user may wish to constrain cpuid to emulate older hardware,
1157 or for feature consistency across a cluster).
1159 Userspace invokes KVM_GET_SUPPORTED_CPUID by passing a kvm_cpuid2 structure
1160 with the 'nent' field indicating the number of entries in the variable-size
1161 array 'entries'. If the number of entries is too low to describe the cpu
1162 capabilities, an error (E2BIG) is returned. If the number is too high,
1163 the 'nent' field is adjusted and an error (ENOMEM) is returned. If the
1164 number is just right, the 'nent' field is adjusted to the number of valid
1165 entries in the 'entries' array, which is then filled.
1167 The entries returned are the host cpuid as returned by the cpuid instruction,
1168 with unknown or unsupported features masked out. Some features (for example,
1169 x2apic), may not be present in the host cpu, but are exposed by kvm if it can
1170 emulate them efficiently. The fields in each entry are defined as follows:
1172 function: the eax value used to obtain the entry
1173 index: the ecx value used to obtain the entry (for entries that are
1175 flags: an OR of zero or more of the following:
1176 KVM_CPUID_FLAG_SIGNIFCANT_INDEX:
1177 if the index field is valid
1178 KVM_CPUID_FLAG_STATEFUL_FUNC:
1179 if cpuid for this function returns different values for successive
1180 invocations; there will be several entries with the same function,
1181 all with this flag set
1182 KVM_CPUID_FLAG_STATE_READ_NEXT:
1183 for KVM_CPUID_FLAG_STATEFUL_FUNC entries, set if this entry is
1184 the first entry to be read by a cpu
1185 eax, ebx, ecx, edx: the values returned by the cpuid instruction for
1186 this function/index combination
1188 The TSC deadline timer feature (CPUID leaf 1, ecx[24]) is always returned
1189 as false, since the feature depends on KVM_CREATE_IRQCHIP for local APIC
1190 support. Instead it is reported via
1192 ioctl(KVM_CHECK_EXTENSION, KVM_CAP_TSC_DEADLINE_TIMER)
1194 if that returns true and you use KVM_CREATE_IRQCHIP, or if you emulate the
1195 feature in userspace, then you can enable the feature for KVM_SET_CPUID2.
1198 4.47 KVM_PPC_GET_PVINFO
1200 Capability: KVM_CAP_PPC_GET_PVINFO
1203 Parameters: struct kvm_ppc_pvinfo (out)
1204 Returns: 0 on success, !0 on error
1206 struct kvm_ppc_pvinfo {
1212 This ioctl fetches PV specific information that need to be passed to the guest
1213 using the device tree or other means from vm context.
1215 The hcall array defines 4 instructions that make up a hypercall.
1217 If any additional field gets added to this structure later on, a bit for that
1218 additional piece of information will be set in the flags bitmap.
1220 The flags bitmap is defined as:
1222 /* the host supports the ePAPR idle hcall
1223 #define KVM_PPC_PVINFO_FLAGS_EV_IDLE (1<<0)
1225 4.48 KVM_ASSIGN_PCI_DEVICE
1227 Capability: KVM_CAP_DEVICE_ASSIGNMENT
1228 Architectures: x86 ia64
1230 Parameters: struct kvm_assigned_pci_dev (in)
1231 Returns: 0 on success, -1 on error
1233 Assigns a host PCI device to the VM.
1235 struct kvm_assigned_pci_dev {
1236 __u32 assigned_dev_id;
1246 The PCI device is specified by the triple segnr, busnr, and devfn.
1247 Identification in succeeding service requests is done via assigned_dev_id. The
1248 following flags are specified:
1250 /* Depends on KVM_CAP_IOMMU */
1251 #define KVM_DEV_ASSIGN_ENABLE_IOMMU (1 << 0)
1252 /* The following two depend on KVM_CAP_PCI_2_3 */
1253 #define KVM_DEV_ASSIGN_PCI_2_3 (1 << 1)
1254 #define KVM_DEV_ASSIGN_MASK_INTX (1 << 2)
1256 If KVM_DEV_ASSIGN_PCI_2_3 is set, the kernel will manage legacy INTx interrupts
1257 via the PCI-2.3-compliant device-level mask, thus enable IRQ sharing with other
1258 assigned devices or host devices. KVM_DEV_ASSIGN_MASK_INTX specifies the
1259 guest's view on the INTx mask, see KVM_ASSIGN_SET_INTX_MASK for details.
1261 The KVM_DEV_ASSIGN_ENABLE_IOMMU flag is a mandatory option to ensure
1262 isolation of the device. Usages not specifying this flag are deprecated.
1264 Only PCI header type 0 devices with PCI BAR resources are supported by
1265 device assignment. The user requesting this ioctl must have read/write
1266 access to the PCI sysfs resource files associated with the device.
1269 4.49 KVM_DEASSIGN_PCI_DEVICE
1271 Capability: KVM_CAP_DEVICE_DEASSIGNMENT
1272 Architectures: x86 ia64
1274 Parameters: struct kvm_assigned_pci_dev (in)
1275 Returns: 0 on success, -1 on error
1277 Ends PCI device assignment, releasing all associated resources.
1279 See KVM_CAP_DEVICE_ASSIGNMENT for the data structure. Only assigned_dev_id is
1280 used in kvm_assigned_pci_dev to identify the device.
1283 4.50 KVM_ASSIGN_DEV_IRQ
1285 Capability: KVM_CAP_ASSIGN_DEV_IRQ
1286 Architectures: x86 ia64
1288 Parameters: struct kvm_assigned_irq (in)
1289 Returns: 0 on success, -1 on error
1291 Assigns an IRQ to a passed-through device.
1293 struct kvm_assigned_irq {
1294 __u32 assigned_dev_id;
1295 __u32 host_irq; /* ignored (legacy field) */
1303 The following flags are defined:
1305 #define KVM_DEV_IRQ_HOST_INTX (1 << 0)
1306 #define KVM_DEV_IRQ_HOST_MSI (1 << 1)
1307 #define KVM_DEV_IRQ_HOST_MSIX (1 << 2)
1309 #define KVM_DEV_IRQ_GUEST_INTX (1 << 8)
1310 #define KVM_DEV_IRQ_GUEST_MSI (1 << 9)
1311 #define KVM_DEV_IRQ_GUEST_MSIX (1 << 10)
1313 It is not valid to specify multiple types per host or guest IRQ. However, the
1314 IRQ type of host and guest can differ or can even be null.
1317 4.51 KVM_DEASSIGN_DEV_IRQ
1319 Capability: KVM_CAP_ASSIGN_DEV_IRQ
1320 Architectures: x86 ia64
1322 Parameters: struct kvm_assigned_irq (in)
1323 Returns: 0 on success, -1 on error
1325 Ends an IRQ assignment to a passed-through device.
1327 See KVM_ASSIGN_DEV_IRQ for the data structure. The target device is specified
1328 by assigned_dev_id, flags must correspond to the IRQ type specified on
1329 KVM_ASSIGN_DEV_IRQ. Partial deassignment of host or guest IRQ is allowed.
1332 4.52 KVM_SET_GSI_ROUTING
1334 Capability: KVM_CAP_IRQ_ROUTING
1335 Architectures: x86 ia64
1337 Parameters: struct kvm_irq_routing (in)
1338 Returns: 0 on success, -1 on error
1340 Sets the GSI routing table entries, overwriting any previously set entries.
1342 struct kvm_irq_routing {
1345 struct kvm_irq_routing_entry entries[0];
1348 No flags are specified so far, the corresponding field must be set to zero.
1350 struct kvm_irq_routing_entry {
1356 struct kvm_irq_routing_irqchip irqchip;
1357 struct kvm_irq_routing_msi msi;
1362 /* gsi routing entry types */
1363 #define KVM_IRQ_ROUTING_IRQCHIP 1
1364 #define KVM_IRQ_ROUTING_MSI 2
1366 No flags are specified so far, the corresponding field must be set to zero.
1368 struct kvm_irq_routing_irqchip {
1373 struct kvm_irq_routing_msi {
1381 4.53 KVM_ASSIGN_SET_MSIX_NR
1383 Capability: KVM_CAP_DEVICE_MSIX
1384 Architectures: x86 ia64
1386 Parameters: struct kvm_assigned_msix_nr (in)
1387 Returns: 0 on success, -1 on error
1389 Set the number of MSI-X interrupts for an assigned device. The number is
1390 reset again by terminating the MSI-X assignment of the device via
1391 KVM_DEASSIGN_DEV_IRQ. Calling this service more than once at any earlier
1394 struct kvm_assigned_msix_nr {
1395 __u32 assigned_dev_id;
1400 #define KVM_MAX_MSIX_PER_DEV 256
1403 4.54 KVM_ASSIGN_SET_MSIX_ENTRY
1405 Capability: KVM_CAP_DEVICE_MSIX
1406 Architectures: x86 ia64
1408 Parameters: struct kvm_assigned_msix_entry (in)
1409 Returns: 0 on success, -1 on error
1411 Specifies the routing of an MSI-X assigned device interrupt to a GSI. Setting
1412 the GSI vector to zero means disabling the interrupt.
1414 struct kvm_assigned_msix_entry {
1415 __u32 assigned_dev_id;
1417 __u16 entry; /* The index of entry in the MSI-X table */
1422 4.55 KVM_SET_TSC_KHZ
1424 Capability: KVM_CAP_TSC_CONTROL
1427 Parameters: virtual tsc_khz
1428 Returns: 0 on success, -1 on error
1430 Specifies the tsc frequency for the virtual machine. The unit of the
1434 4.56 KVM_GET_TSC_KHZ
1436 Capability: KVM_CAP_GET_TSC_KHZ
1440 Returns: virtual tsc-khz on success, negative value on error
1442 Returns the tsc frequency of the guest. The unit of the return value is
1443 KHz. If the host has unstable tsc this ioctl returns -EIO instead as an
1449 Capability: KVM_CAP_IRQCHIP
1452 Parameters: struct kvm_lapic_state (out)
1453 Returns: 0 on success, -1 on error
1455 #define KVM_APIC_REG_SIZE 0x400
1456 struct kvm_lapic_state {
1457 char regs[KVM_APIC_REG_SIZE];
1460 Reads the Local APIC registers and copies them into the input argument. The
1461 data format and layout are the same as documented in the architecture manual.
1466 Capability: KVM_CAP_IRQCHIP
1469 Parameters: struct kvm_lapic_state (in)
1470 Returns: 0 on success, -1 on error
1472 #define KVM_APIC_REG_SIZE 0x400
1473 struct kvm_lapic_state {
1474 char regs[KVM_APIC_REG_SIZE];
1477 Copies the input argument into the the Local APIC registers. The data format
1478 and layout are the same as documented in the architecture manual.
1483 Capability: KVM_CAP_IOEVENTFD
1486 Parameters: struct kvm_ioeventfd (in)
1487 Returns: 0 on success, !0 on error
1489 This ioctl attaches or detaches an ioeventfd to a legal pio/mmio address
1490 within the guest. A guest write in the registered address will signal the
1491 provided event instead of triggering an exit.
1493 struct kvm_ioeventfd {
1495 __u64 addr; /* legal pio/mmio address */
1496 __u32 len; /* 1, 2, 4, or 8 bytes */
1502 The following flags are defined:
1504 #define KVM_IOEVENTFD_FLAG_DATAMATCH (1 << kvm_ioeventfd_flag_nr_datamatch)
1505 #define KVM_IOEVENTFD_FLAG_PIO (1 << kvm_ioeventfd_flag_nr_pio)
1506 #define KVM_IOEVENTFD_FLAG_DEASSIGN (1 << kvm_ioeventfd_flag_nr_deassign)
1508 If datamatch flag is set, the event will be signaled only if the written value
1509 to the registered address is equal to datamatch in struct kvm_ioeventfd.
1514 Capability: KVM_CAP_SW_TLB
1517 Parameters: struct kvm_dirty_tlb (in)
1518 Returns: 0 on success, -1 on error
1520 struct kvm_dirty_tlb {
1525 This must be called whenever userspace has changed an entry in the shared
1526 TLB, prior to calling KVM_RUN on the associated vcpu.
1528 The "bitmap" field is the userspace address of an array. This array
1529 consists of a number of bits, equal to the total number of TLB entries as
1530 determined by the last successful call to KVM_CONFIG_TLB, rounded up to the
1531 nearest multiple of 64.
1533 Each bit corresponds to one TLB entry, ordered the same as in the shared TLB
1536 The array is little-endian: the bit 0 is the least significant bit of the
1537 first byte, bit 8 is the least significant bit of the second byte, etc.
1538 This avoids any complications with differing word sizes.
1540 The "num_dirty" field is a performance hint for KVM to determine whether it
1541 should skip processing the bitmap and just invalidate everything. It must
1542 be set to the number of set bits in the bitmap.
1545 4.61 KVM_ASSIGN_SET_INTX_MASK
1547 Capability: KVM_CAP_PCI_2_3
1550 Parameters: struct kvm_assigned_pci_dev (in)
1551 Returns: 0 on success, -1 on error
1553 Allows userspace to mask PCI INTx interrupts from the assigned device. The
1554 kernel will not deliver INTx interrupts to the guest between setting and
1555 clearing of KVM_ASSIGN_SET_INTX_MASK via this interface. This enables use of
1556 and emulation of PCI 2.3 INTx disable command register behavior.
1558 This may be used for both PCI 2.3 devices supporting INTx disable natively and
1559 older devices lacking this support. Userspace is responsible for emulating the
1560 read value of the INTx disable bit in the guest visible PCI command register.
1561 When modifying the INTx disable state, userspace should precede updating the
1562 physical device command register by calling this ioctl to inform the kernel of
1563 the new intended INTx mask state.
1565 Note that the kernel uses the device INTx disable bit to internally manage the
1566 device interrupt state for PCI 2.3 devices. Reads of this register may
1567 therefore not match the expected value. Writes should always use the guest
1568 intended INTx disable value rather than attempting to read-copy-update the
1569 current physical device state. Races between user and kernel updates to the
1570 INTx disable bit are handled lazily in the kernel. It's possible the device
1571 may generate unintended interrupts, but they will not be injected into the
1574 See KVM_ASSIGN_DEV_IRQ for the data structure. The target device is specified
1575 by assigned_dev_id. In the flags field, only KVM_DEV_ASSIGN_MASK_INTX is
1579 4.62 KVM_CREATE_SPAPR_TCE
1581 Capability: KVM_CAP_SPAPR_TCE
1582 Architectures: powerpc
1584 Parameters: struct kvm_create_spapr_tce (in)
1585 Returns: file descriptor for manipulating the created TCE table
1587 This creates a virtual TCE (translation control entry) table, which
1588 is an IOMMU for PAPR-style virtual I/O. It is used to translate
1589 logical addresses used in virtual I/O into guest physical addresses,
1590 and provides a scatter/gather capability for PAPR virtual I/O.
1592 /* for KVM_CAP_SPAPR_TCE */
1593 struct kvm_create_spapr_tce {
1598 The liobn field gives the logical IO bus number for which to create a
1599 TCE table. The window_size field specifies the size of the DMA window
1600 which this TCE table will translate - the table will contain one 64
1601 bit TCE entry for every 4kiB of the DMA window.
1603 When the guest issues an H_PUT_TCE hcall on a liobn for which a TCE
1604 table has been created using this ioctl(), the kernel will handle it
1605 in real mode, updating the TCE table. H_PUT_TCE calls for other
1606 liobns will cause a vm exit and must be handled by userspace.
1608 The return value is a file descriptor which can be passed to mmap(2)
1609 to map the created TCE table into userspace. This lets userspace read
1610 the entries written by kernel-handled H_PUT_TCE calls, and also lets
1611 userspace update the TCE table directly which is useful in some
1615 4.63 KVM_ALLOCATE_RMA
1617 Capability: KVM_CAP_PPC_RMA
1618 Architectures: powerpc
1620 Parameters: struct kvm_allocate_rma (out)
1621 Returns: file descriptor for mapping the allocated RMA
1623 This allocates a Real Mode Area (RMA) from the pool allocated at boot
1624 time by the kernel. An RMA is a physically-contiguous, aligned region
1625 of memory used on older POWER processors to provide the memory which
1626 will be accessed by real-mode (MMU off) accesses in a KVM guest.
1627 POWER processors support a set of sizes for the RMA that usually
1628 includes 64MB, 128MB, 256MB and some larger powers of two.
1630 /* for KVM_ALLOCATE_RMA */
1631 struct kvm_allocate_rma {
1635 The return value is a file descriptor which can be passed to mmap(2)
1636 to map the allocated RMA into userspace. The mapped area can then be
1637 passed to the KVM_SET_USER_MEMORY_REGION ioctl to establish it as the
1638 RMA for a virtual machine. The size of the RMA in bytes (which is
1639 fixed at host kernel boot time) is returned in the rma_size field of
1640 the argument structure.
1642 The KVM_CAP_PPC_RMA capability is 1 or 2 if the KVM_ALLOCATE_RMA ioctl
1643 is supported; 2 if the processor requires all virtual machines to have
1644 an RMA, or 1 if the processor can use an RMA but doesn't require it,
1645 because it supports the Virtual RMA (VRMA) facility.
1650 Capability: KVM_CAP_USER_NMI
1654 Returns: 0 on success, -1 on error
1656 Queues an NMI on the thread's vcpu. Note this is well defined only
1657 when KVM_CREATE_IRQCHIP has not been called, since this is an interface
1658 between the virtual cpu core and virtual local APIC. After KVM_CREATE_IRQCHIP
1659 has been called, this interface is completely emulated within the kernel.
1661 To use this to emulate the LINT1 input with KVM_CREATE_IRQCHIP, use the
1662 following algorithm:
1665 - read the local APIC's state (KVM_GET_LAPIC)
1666 - check whether changing LINT1 will queue an NMI (see the LVT entry for LINT1)
1667 - if so, issue KVM_NMI
1670 Some guests configure the LINT1 NMI input to cause a panic, aiding in
1674 4.65 KVM_S390_UCAS_MAP
1676 Capability: KVM_CAP_S390_UCONTROL
1679 Parameters: struct kvm_s390_ucas_mapping (in)
1680 Returns: 0 in case of success
1682 The parameter is defined like this:
1683 struct kvm_s390_ucas_mapping {
1689 This ioctl maps the memory at "user_addr" with the length "length" to
1690 the vcpu's address space starting at "vcpu_addr". All parameters need to
1691 be alligned by 1 megabyte.
1694 4.66 KVM_S390_UCAS_UNMAP
1696 Capability: KVM_CAP_S390_UCONTROL
1699 Parameters: struct kvm_s390_ucas_mapping (in)
1700 Returns: 0 in case of success
1702 The parameter is defined like this:
1703 struct kvm_s390_ucas_mapping {
1709 This ioctl unmaps the memory in the vcpu's address space starting at
1710 "vcpu_addr" with the length "length". The field "user_addr" is ignored.
1711 All parameters need to be alligned by 1 megabyte.
1714 4.67 KVM_S390_VCPU_FAULT
1716 Capability: KVM_CAP_S390_UCONTROL
1719 Parameters: vcpu absolute address (in)
1720 Returns: 0 in case of success
1722 This call creates a page table entry on the virtual cpu's address space
1723 (for user controlled virtual machines) or the virtual machine's address
1724 space (for regular virtual machines). This only works for minor faults,
1725 thus it's recommended to access subject memory page via the user page
1726 table upfront. This is useful to handle validity intercepts for user
1727 controlled virtual machines to fault in the virtual cpu's lowcore pages
1728 prior to calling the KVM_RUN ioctl.
1731 4.68 KVM_SET_ONE_REG
1733 Capability: KVM_CAP_ONE_REG
1736 Parameters: struct kvm_one_reg (in)
1737 Returns: 0 on success, negative value on failure
1739 struct kvm_one_reg {
1744 Using this ioctl, a single vcpu register can be set to a specific value
1745 defined by user space with the passed in struct kvm_one_reg, where id
1746 refers to the register identifier as described below and addr is a pointer
1747 to a variable with the respective size. There can be architecture agnostic
1748 and architecture specific registers. Each have their own range of operation
1749 and their own constants and width. To keep track of the implemented
1750 registers, find a list below:
1752 Arch | Register | Width (bits)
1754 PPC | KVM_REG_PPC_HIOR | 64
1755 PPC | KVM_REG_PPC_IAC1 | 64
1756 PPC | KVM_REG_PPC_IAC2 | 64
1757 PPC | KVM_REG_PPC_IAC3 | 64
1758 PPC | KVM_REG_PPC_IAC4 | 64
1759 PPC | KVM_REG_PPC_DAC1 | 64
1760 PPC | KVM_REG_PPC_DAC2 | 64
1761 PPC | KVM_REG_PPC_DABR | 64
1762 PPC | KVM_REG_PPC_DSCR | 64
1763 PPC | KVM_REG_PPC_PURR | 64
1764 PPC | KVM_REG_PPC_SPURR | 64
1765 PPC | KVM_REG_PPC_DAR | 64
1766 PPC | KVM_REG_PPC_DSISR | 32
1767 PPC | KVM_REG_PPC_AMR | 64
1768 PPC | KVM_REG_PPC_UAMOR | 64
1769 PPC | KVM_REG_PPC_MMCR0 | 64
1770 PPC | KVM_REG_PPC_MMCR1 | 64
1771 PPC | KVM_REG_PPC_MMCRA | 64
1772 PPC | KVM_REG_PPC_PMC1 | 32
1773 PPC | KVM_REG_PPC_PMC2 | 32
1774 PPC | KVM_REG_PPC_PMC3 | 32
1775 PPC | KVM_REG_PPC_PMC4 | 32
1776 PPC | KVM_REG_PPC_PMC5 | 32
1777 PPC | KVM_REG_PPC_PMC6 | 32
1778 PPC | KVM_REG_PPC_PMC7 | 32
1779 PPC | KVM_REG_PPC_PMC8 | 32
1780 PPC | KVM_REG_PPC_FPR0 | 64
1782 PPC | KVM_REG_PPC_FPR31 | 64
1783 PPC | KVM_REG_PPC_VR0 | 128
1785 PPC | KVM_REG_PPC_VR31 | 128
1786 PPC | KVM_REG_PPC_VSR0 | 128
1788 PPC | KVM_REG_PPC_VSR31 | 128
1789 PPC | KVM_REG_PPC_FPSCR | 64
1790 PPC | KVM_REG_PPC_VSCR | 32
1791 PPC | KVM_REG_PPC_VPA_ADDR | 64
1792 PPC | KVM_REG_PPC_VPA_SLB | 128
1793 PPC | KVM_REG_PPC_VPA_DTL | 128
1794 PPC | KVM_REG_PPC_EPCR | 32
1796 ARM registers are mapped using the lower 32 bits. The upper 16 of that
1797 is the register group type, or coprocessor number:
1799 ARM core registers have the following id bit patterns:
1800 0x4002 0000 0010 <index into the kvm_regs struct:16>
1802 ARM 32-bit CP15 registers have the following id bit patterns:
1803 0x4002 0000 000F <zero:1> <crn:4> <crm:4> <opc1:4> <opc2:3>
1805 ARM 64-bit CP15 registers have the following id bit patterns:
1806 0x4003 0000 000F <zero:1> <zero:4> <crm:4> <opc1:4> <zero:3>
1808 ARM CCSIDR registers are demultiplexed by CSSELR value:
1809 0x4002 0000 0011 00 <csselr:8>
1811 ARM 32-bit VFP control registers have the following id bit patterns:
1812 0x4002 0000 0012 1 <regno:12>
1814 ARM 64-bit FP registers have the following id bit patterns:
1815 0x4002 0000 0012 0 <regno:12>
1817 4.69 KVM_GET_ONE_REG
1819 Capability: KVM_CAP_ONE_REG
1822 Parameters: struct kvm_one_reg (in and out)
1823 Returns: 0 on success, negative value on failure
1825 This ioctl allows to receive the value of a single register implemented
1826 in a vcpu. The register to read is indicated by the "id" field of the
1827 kvm_one_reg struct passed in. On success, the register value can be found
1828 at the memory location pointed to by "addr".
1830 The list of registers accessible using this interface is identical to the
1834 4.70 KVM_KVMCLOCK_CTRL
1836 Capability: KVM_CAP_KVMCLOCK_CTRL
1837 Architectures: Any that implement pvclocks (currently x86 only)
1840 Returns: 0 on success, -1 on error
1842 This signals to the host kernel that the specified guest is being paused by
1843 userspace. The host will set a flag in the pvclock structure that is checked
1844 from the soft lockup watchdog. The flag is part of the pvclock structure that
1845 is shared between guest and host, specifically the second bit of the flags
1846 field of the pvclock_vcpu_time_info structure. It will be set exclusively by
1847 the host and read/cleared exclusively by the guest. The guest operation of
1848 checking and clearing the flag must an atomic operation so
1849 load-link/store-conditional, or equivalent must be used. There are two cases
1850 where the guest will clear the flag: when the soft lockup watchdog timer resets
1851 itself or when a soft lockup is detected. This ioctl can be called any time
1852 after pausing the vcpu, but before it is resumed.
1857 Capability: KVM_CAP_SIGNAL_MSI
1860 Parameters: struct kvm_msi (in)
1861 Returns: >0 on delivery, 0 if guest blocked the MSI, and -1 on error
1863 Directly inject a MSI message. Only valid with in-kernel irqchip that handles
1874 No flags are defined so far. The corresponding field must be 0.
1877 4.71 KVM_CREATE_PIT2
1879 Capability: KVM_CAP_PIT2
1882 Parameters: struct kvm_pit_config (in)
1883 Returns: 0 on success, -1 on error
1885 Creates an in-kernel device model for the i8254 PIT. This call is only valid
1886 after enabling in-kernel irqchip support via KVM_CREATE_IRQCHIP. The following
1887 parameters have to be passed:
1889 struct kvm_pit_config {
1896 #define KVM_PIT_SPEAKER_DUMMY 1 /* emulate speaker port stub */
1898 PIT timer interrupts may use a per-VM kernel thread for injection. If it
1899 exists, this thread will have a name of the following pattern:
1901 kvm-pit/<owner-process-pid>
1903 When running a guest with elevated priorities, the scheduling parameters of
1904 this thread may have to be adjusted accordingly.
1906 This IOCTL replaces the obsolete KVM_CREATE_PIT.
1911 Capability: KVM_CAP_PIT_STATE2
1914 Parameters: struct kvm_pit_state2 (out)
1915 Returns: 0 on success, -1 on error
1917 Retrieves the state of the in-kernel PIT model. Only valid after
1918 KVM_CREATE_PIT2. The state is returned in the following structure:
1920 struct kvm_pit_state2 {
1921 struct kvm_pit_channel_state channels[3];
1928 /* disable PIT in HPET legacy mode */
1929 #define KVM_PIT_FLAGS_HPET_LEGACY 0x00000001
1931 This IOCTL replaces the obsolete KVM_GET_PIT.
1936 Capability: KVM_CAP_PIT_STATE2
1939 Parameters: struct kvm_pit_state2 (in)
1940 Returns: 0 on success, -1 on error
1942 Sets the state of the in-kernel PIT model. Only valid after KVM_CREATE_PIT2.
1943 See KVM_GET_PIT2 for details on struct kvm_pit_state2.
1945 This IOCTL replaces the obsolete KVM_SET_PIT.
1948 4.74 KVM_PPC_GET_SMMU_INFO
1950 Capability: KVM_CAP_PPC_GET_SMMU_INFO
1951 Architectures: powerpc
1954 Returns: 0 on success, -1 on error
1956 This populates and returns a structure describing the features of
1957 the "Server" class MMU emulation supported by KVM.
1958 This can in turn be used by userspace to generate the appropariate
1959 device-tree properties for the guest operating system.
1961 The structure contains some global informations, followed by an
1962 array of supported segment page sizes:
1964 struct kvm_ppc_smmu_info {
1968 struct kvm_ppc_one_seg_page_size sps[KVM_PPC_PAGE_SIZES_MAX_SZ];
1971 The supported flags are:
1973 - KVM_PPC_PAGE_SIZES_REAL:
1974 When that flag is set, guest page sizes must "fit" the backing
1975 store page sizes. When not set, any page size in the list can
1976 be used regardless of how they are backed by userspace.
1978 - KVM_PPC_1T_SEGMENTS
1979 The emulated MMU supports 1T segments in addition to the
1982 The "slb_size" field indicates how many SLB entries are supported
1984 The "sps" array contains 8 entries indicating the supported base
1985 page sizes for a segment in increasing order. Each entry is defined
1988 struct kvm_ppc_one_seg_page_size {
1989 __u32 page_shift; /* Base page shift of segment (or 0) */
1990 __u32 slb_enc; /* SLB encoding for BookS */
1991 struct kvm_ppc_one_page_size enc[KVM_PPC_PAGE_SIZES_MAX_SZ];
1994 An entry with a "page_shift" of 0 is unused. Because the array is
1995 organized in increasing order, a lookup can stop when encoutering
1998 The "slb_enc" field provides the encoding to use in the SLB for the
1999 page size. The bits are in positions such as the value can directly
2000 be OR'ed into the "vsid" argument of the slbmte instruction.
2002 The "enc" array is a list which for each of those segment base page
2003 size provides the list of supported actual page sizes (which can be
2004 only larger or equal to the base page size), along with the
2005 corresponding encoding in the hash PTE. Similarily, the array is
2006 8 entries sorted by increasing sizes and an entry with a "0" shift
2007 is an empty entry and a terminator:
2009 struct kvm_ppc_one_page_size {
2010 __u32 page_shift; /* Page shift (or 0) */
2011 __u32 pte_enc; /* Encoding in the HPTE (>>12) */
2014 The "pte_enc" field provides a value that can OR'ed into the hash
2015 PTE's RPN field (ie, it needs to be shifted left by 12 to OR it
2016 into the hash PTE second double word).
2020 Capability: KVM_CAP_IRQFD
2023 Parameters: struct kvm_irqfd (in)
2024 Returns: 0 on success, -1 on error
2026 Allows setting an eventfd to directly trigger a guest interrupt.
2027 kvm_irqfd.fd specifies the file descriptor to use as the eventfd and
2028 kvm_irqfd.gsi specifies the irqchip pin toggled by this event. When
2029 an event is tiggered on the eventfd, an interrupt is injected into
2030 the guest using the specified gsi pin. The irqfd is removed using
2031 the KVM_IRQFD_FLAG_DEASSIGN flag, specifying both kvm_irqfd.fd
2034 With KVM_CAP_IRQFD_RESAMPLE, KVM_IRQFD supports a de-assert and notify
2035 mechanism allowing emulation of level-triggered, irqfd-based
2036 interrupts. When KVM_IRQFD_FLAG_RESAMPLE is set the user must pass an
2037 additional eventfd in the kvm_irqfd.resamplefd field. When operating
2038 in resample mode, posting of an interrupt through kvm_irq.fd asserts
2039 the specified gsi in the irqchip. When the irqchip is resampled, such
2040 as from an EOI, the gsi is de-asserted and the user is notifed via
2041 kvm_irqfd.resamplefd. It is the user's responsibility to re-queue
2042 the interrupt if the device making use of it still requires service.
2043 Note that closing the resamplefd is not sufficient to disable the
2044 irqfd. The KVM_IRQFD_FLAG_RESAMPLE is only necessary on assignment
2045 and need not be specified with KVM_IRQFD_FLAG_DEASSIGN.
2047 4.76 KVM_PPC_ALLOCATE_HTAB
2049 Capability: KVM_CAP_PPC_ALLOC_HTAB
2050 Architectures: powerpc
2052 Parameters: Pointer to u32 containing hash table order (in/out)
2053 Returns: 0 on success, -1 on error
2055 This requests the host kernel to allocate an MMU hash table for a
2056 guest using the PAPR paravirtualization interface. This only does
2057 anything if the kernel is configured to use the Book 3S HV style of
2058 virtualization. Otherwise the capability doesn't exist and the ioctl
2059 returns an ENOTTY error. The rest of this description assumes Book 3S
2062 There must be no vcpus running when this ioctl is called; if there
2063 are, it will do nothing and return an EBUSY error.
2065 The parameter is a pointer to a 32-bit unsigned integer variable
2066 containing the order (log base 2) of the desired size of the hash
2067 table, which must be between 18 and 46. On successful return from the
2068 ioctl, it will have been updated with the order of the hash table that
2071 If no hash table has been allocated when any vcpu is asked to run
2072 (with the KVM_RUN ioctl), the host kernel will allocate a
2073 default-sized hash table (16 MB).
2075 If this ioctl is called when a hash table has already been allocated,
2076 the kernel will clear out the existing hash table (zero all HPTEs) and
2077 return the hash table order in the parameter. (If the guest is using
2078 the virtualized real-mode area (VRMA) facility, the kernel will
2079 re-create the VMRA HPTEs on the next KVM_RUN of any vcpu.)
2081 4.77 KVM_S390_INTERRUPT
2085 Type: vm ioctl, vcpu ioctl
2086 Parameters: struct kvm_s390_interrupt (in)
2087 Returns: 0 on success, -1 on error
2089 Allows to inject an interrupt to the guest. Interrupts can be floating
2090 (vm ioctl) or per cpu (vcpu ioctl), depending on the interrupt type.
2092 Interrupt parameters are passed via kvm_s390_interrupt:
2094 struct kvm_s390_interrupt {
2100 type can be one of the following:
2102 KVM_S390_SIGP_STOP (vcpu) - sigp restart
2103 KVM_S390_PROGRAM_INT (vcpu) - program check; code in parm
2104 KVM_S390_SIGP_SET_PREFIX (vcpu) - sigp set prefix; prefix address in parm
2105 KVM_S390_RESTART (vcpu) - restart
2106 KVM_S390_INT_VIRTIO (vm) - virtio external interrupt; external interrupt
2107 parameters in parm and parm64
2108 KVM_S390_INT_SERVICE (vm) - sclp external interrupt; sclp parameter in parm
2109 KVM_S390_INT_EMERGENCY (vcpu) - sigp emergency; source cpu in parm
2110 KVM_S390_INT_EXTERNAL_CALL (vcpu) - sigp external call; source cpu in parm
2112 Note that the vcpu ioctl is asynchronous to vcpu execution.
2114 4.78 KVM_PPC_GET_HTAB_FD
2116 Capability: KVM_CAP_PPC_HTAB_FD
2117 Architectures: powerpc
2119 Parameters: Pointer to struct kvm_get_htab_fd (in)
2120 Returns: file descriptor number (>= 0) on success, -1 on error
2122 This returns a file descriptor that can be used either to read out the
2123 entries in the guest's hashed page table (HPT), or to write entries to
2124 initialize the HPT. The returned fd can only be written to if the
2125 KVM_GET_HTAB_WRITE bit is set in the flags field of the argument, and
2126 can only be read if that bit is clear. The argument struct looks like
2129 /* For KVM_PPC_GET_HTAB_FD */
2130 struct kvm_get_htab_fd {
2136 /* Values for kvm_get_htab_fd.flags */
2137 #define KVM_GET_HTAB_BOLTED_ONLY ((__u64)0x1)
2138 #define KVM_GET_HTAB_WRITE ((__u64)0x2)
2140 The `start_index' field gives the index in the HPT of the entry at
2141 which to start reading. It is ignored when writing.
2143 Reads on the fd will initially supply information about all
2144 "interesting" HPT entries. Interesting entries are those with the
2145 bolted bit set, if the KVM_GET_HTAB_BOLTED_ONLY bit is set, otherwise
2146 all entries. When the end of the HPT is reached, the read() will
2147 return. If read() is called again on the fd, it will start again from
2148 the beginning of the HPT, but will only return HPT entries that have
2149 changed since they were last read.
2151 Data read or written is structured as a header (8 bytes) followed by a
2152 series of valid HPT entries (16 bytes) each. The header indicates how
2153 many valid HPT entries there are and how many invalid entries follow
2154 the valid entries. The invalid entries are not represented explicitly
2155 in the stream. The header format is:
2157 struct kvm_get_htab_header {
2163 Writes to the fd create HPT entries starting at the index given in the
2164 header; first `n_valid' valid entries with contents from the data
2165 written, then `n_invalid' invalid entries, invalidating any previously
2166 valid entries found.
2169 4.77 KVM_ARM_VCPU_INIT
2174 Parameters: struct struct kvm_vcpu_init (in)
2175 Returns: 0 on success; -1 on error
2177 Â EINVAL: Â Â Â the target is unknown, or the combination of features is invalid.
2178 Â ENOENT: Â Â Â a features bit specified is unknown.
2180 This tells KVM what type of CPU to present to the guest, and what
2181 optional features it should have. Â This will cause a reset of the cpu
2182 registers to their initial values. Â If this is not called, KVM_RUN will
2183 return ENOEXEC for that vcpu.
2185 Note that because some registers reflect machine topology, all vcpus
2186 should be created before this ioctl is invoked.
2189 - KVM_ARM_VCPU_POWER_OFF: Starts the CPU in a power-off state.
2190 Depends on KVM_CAP_ARM_PSCI.
2193 4.78 KVM_GET_REG_LIST
2198 Parameters: struct kvm_reg_list (in/out)
2199 Returns: 0 on success; -1 on error
2201 Â E2BIG: Â Â Â Â the reg index list is too big to fit in the array specified by
2202 Â Â Â Â Â Â Â Â Â Â Â Â the user (the number required will be written into n).
2204 struct kvm_reg_list {
2205 __u64 n; /* number of registers in reg[] */
2209 This ioctl returns the guest registers that are supported for the
2210 KVM_GET_ONE_REG/KVM_SET_ONE_REG calls.
2213 5. The kvm_run structure
2214 ------------------------
2216 Application code obtains a pointer to the kvm_run structure by
2217 mmap()ing a vcpu fd. From that point, application code can control
2218 execution by changing fields in kvm_run prior to calling the KVM_RUN
2219 ioctl, and obtain information about the reason KVM_RUN returned by
2220 looking up structure members.
2224 __u8 request_interrupt_window;
2226 Request that KVM_RUN return when it becomes possible to inject external
2227 interrupts into the guest. Useful in conjunction with KVM_INTERRUPT.
2234 When KVM_RUN has returned successfully (return value 0), this informs
2235 application code why KVM_RUN has returned. Allowable values for this
2236 field are detailed below.
2238 __u8 ready_for_interrupt_injection;
2240 If request_interrupt_window has been specified, this field indicates
2241 an interrupt can be injected now with KVM_INTERRUPT.
2245 The value of the current interrupt flag. Only valid if in-kernel
2246 local APIC is not used.
2250 /* in (pre_kvm_run), out (post_kvm_run) */
2253 The value of the cr8 register. Only valid if in-kernel local APIC is
2254 not used. Both input and output.
2258 The value of the APIC BASE msr. Only valid if in-kernel local
2259 APIC is not used. Both input and output.
2262 /* KVM_EXIT_UNKNOWN */
2264 __u64 hardware_exit_reason;
2267 If exit_reason is KVM_EXIT_UNKNOWN, the vcpu has exited due to unknown
2268 reasons. Further architecture-specific information is available in
2269 hardware_exit_reason.
2271 /* KVM_EXIT_FAIL_ENTRY */
2273 __u64 hardware_entry_failure_reason;
2276 If exit_reason is KVM_EXIT_FAIL_ENTRY, the vcpu could not be run due
2277 to unknown reasons. Further architecture-specific information is
2278 available in hardware_entry_failure_reason.
2280 /* KVM_EXIT_EXCEPTION */
2290 #define KVM_EXIT_IO_IN 0
2291 #define KVM_EXIT_IO_OUT 1
2293 __u8 size; /* bytes */
2296 __u64 data_offset; /* relative to kvm_run start */
2299 If exit_reason is KVM_EXIT_IO, then the vcpu has
2300 executed a port I/O instruction which could not be satisfied by kvm.
2301 data_offset describes where the data is located (KVM_EXIT_IO_OUT) or
2302 where kvm expects application code to place the data for the next
2303 KVM_RUN invocation (KVM_EXIT_IO_IN). Data format is a packed array.
2306 struct kvm_debug_exit_arch arch;
2319 If exit_reason is KVM_EXIT_MMIO, then the vcpu has
2320 executed a memory-mapped I/O instruction which could not be satisfied
2321 by kvm. The 'data' member contains the written data if 'is_write' is
2322 true, and should be filled by application code otherwise.
2324 NOTE: For KVM_EXIT_IO, KVM_EXIT_MMIO, KVM_EXIT_OSI, KVM_EXIT_DCR
2325 and KVM_EXIT_PAPR the corresponding
2326 operations are complete (and guest state is consistent) only after userspace
2327 has re-entered the kernel with KVM_RUN. The kernel side will first finish
2328 incomplete operations and then check for pending signals. Userspace
2329 can re-enter the guest with an unmasked signal pending to complete
2332 /* KVM_EXIT_HYPERCALL */
2341 Unused. This was once used for 'hypercall to userspace'. To implement
2342 such functionality, use KVM_EXIT_IO (x86) or KVM_EXIT_MMIO (all except s390).
2343 Note KVM_EXIT_IO is significantly faster than KVM_EXIT_MMIO.
2345 /* KVM_EXIT_TPR_ACCESS */
2352 To be documented (KVM_TPR_ACCESS_REPORTING).
2354 /* KVM_EXIT_S390_SIEIC */
2357 __u64 mask; /* psw upper half */
2358 __u64 addr; /* psw lower half */
2365 /* KVM_EXIT_S390_RESET */
2366 #define KVM_S390_RESET_POR 1
2367 #define KVM_S390_RESET_CLEAR 2
2368 #define KVM_S390_RESET_SUBSYSTEM 4
2369 #define KVM_S390_RESET_CPU_INIT 8
2370 #define KVM_S390_RESET_IPL 16
2371 __u64 s390_reset_flags;
2375 /* KVM_EXIT_S390_UCONTROL */
2377 __u64 trans_exc_code;
2381 s390 specific. A page fault has occurred for a user controlled virtual
2382 machine (KVM_VM_S390_UNCONTROL) on it's host page table that cannot be
2383 resolved by the kernel.
2384 The program code and the translation exception code that were placed
2385 in the cpu's lowcore are presented here as defined by the z Architecture
2386 Principles of Operation Book in the Chapter for Dynamic Address Translation
2403 MOL uses a special hypercall interface it calls 'OSI'. To enable it, we catch
2404 hypercalls and exit with this exit struct that contains all the guest gprs.
2406 If exit_reason is KVM_EXIT_OSI, then the vcpu has triggered such a hypercall.
2407 Userspace can now handle the hypercall and when it's done modify the gprs as
2408 necessary. Upon guest entry all guest GPRs will then be replaced by the values
2411 /* KVM_EXIT_PAPR_HCALL */
2418 This is used on 64-bit PowerPC when emulating a pSeries partition,
2419 e.g. with the 'pseries' machine type in qemu. It occurs when the
2420 guest does a hypercall using the 'sc 1' instruction. The 'nr' field
2421 contains the hypercall number (from the guest R3), and 'args' contains
2422 the arguments (from the guest R4 - R12). Userspace should put the
2423 return code in 'ret' and any extra returned values in args[].
2424 The possible hypercalls are defined in the Power Architecture Platform
2425 Requirements (PAPR) document available from www.power.org (free
2426 developer registration required to access it).
2428 /* Fix the size of the union. */
2433 * shared registers between kvm and userspace.
2434 * kvm_valid_regs specifies the register classes set by the host
2435 * kvm_dirty_regs specified the register classes dirtied by userspace
2436 * struct kvm_sync_regs is architecture specific, as well as the
2437 * bits for kvm_valid_regs and kvm_dirty_regs
2439 __u64 kvm_valid_regs;
2440 __u64 kvm_dirty_regs;
2442 struct kvm_sync_regs regs;
2446 If KVM_CAP_SYNC_REGS is defined, these fields allow userspace to access
2447 certain guest registers without having to call SET/GET_*REGS. Thus we can
2448 avoid some system call overhead if userspace has to handle the exit.
2449 Userspace can query the validity of the structure by checking
2450 kvm_valid_regs for specific bits. These bits are architecture specific
2451 and usually define the validity of a groups of registers. (e.g. one bit
2452 for general purpose registers)
2457 6. Capabilities that can be enabled
2458 -----------------------------------
2460 There are certain capabilities that change the behavior of the virtual CPU when
2461 enabled. To enable them, please see section 4.37. Below you can find a list of
2462 capabilities and what their effect on the vCPU is when enabling them.
2464 The following information is provided along with the description:
2466 Architectures: which instruction set architectures provide this ioctl.
2467 x86 includes both i386 and x86_64.
2469 Parameters: what parameters are accepted by the capability.
2471 Returns: the return value. General error numbers (EBADF, ENOMEM, EINVAL)
2472 are not detailed, but errors with specific meanings are.
2479 Returns: 0 on success; -1 on error
2481 This capability enables interception of OSI hypercalls that otherwise would
2482 be treated as normal system calls to be injected into the guest. OSI hypercalls
2483 were invented by Mac-on-Linux to have a standardized communication mechanism
2484 between the guest and the host.
2486 When this capability is enabled, KVM_EXIT_OSI can occur.
2489 6.2 KVM_CAP_PPC_PAPR
2493 Returns: 0 on success; -1 on error
2495 This capability enables interception of PAPR hypercalls. PAPR hypercalls are
2496 done using the hypercall instruction "sc 1".
2498 It also sets the guest privilege level to "supervisor" mode. Usually the guest
2499 runs in "hypervisor" privilege mode with a few missing features.
2501 In addition to the above, it changes the semantics of SDR1. In this mode, the
2502 HTAB address part of SDR1 contains an HVA instead of a GPA, as PAPR keeps the
2503 HTAB invisible to the guest.
2505 When this capability is enabled, KVM_EXIT_PAPR_HCALL can occur.
2511 Parameters: args[0] is the address of a struct kvm_config_tlb
2512 Returns: 0 on success; -1 on error
2514 struct kvm_config_tlb {
2521 Configures the virtual CPU's TLB array, establishing a shared memory area
2522 between userspace and KVM. The "params" and "array" fields are userspace
2523 addresses of mmu-type-specific data structures. The "array_len" field is an
2524 safety mechanism, and should be set to the size in bytes of the memory that
2525 userspace has reserved for the array. It must be at least the size dictated
2526 by "mmu_type" and "params".
2528 While KVM_RUN is active, the shared region is under control of KVM. Its
2529 contents are undefined, and any modification by userspace results in
2530 boundedly undefined behavior.
2532 On return from KVM_RUN, the shared region will reflect the current state of
2533 the guest's TLB. If userspace makes any changes, it must call KVM_DIRTY_TLB
2534 to tell KVM which entries have been changed, prior to calling KVM_RUN again
2537 For mmu types KVM_MMU_FSL_BOOKE_NOHV and KVM_MMU_FSL_BOOKE_HV:
2538 - The "params" field is of type "struct kvm_book3e_206_tlb_params".
2539 - The "array" field points to an array of type "struct
2540 kvm_book3e_206_tlb_entry".
2541 - The array consists of all entries in the first TLB, followed by all
2542 entries in the second TLB.
2543 - Within a TLB, entries are ordered first by increasing set number. Within a
2544 set, entries are ordered by way (increasing ESEL).
2545 - The hash for determining set number in TLB0 is: (MAS2 >> 12) & (num_sets - 1)
2546 where "num_sets" is the tlb_sizes[] value divided by the tlb_ways[] value.
2547 - The tsize field of mas1 shall be set to 4K on TLB0, even though the
2548 hardware ignores this value for TLB0.