0001 .. SPDX-License-Identifier: GPL-2.0
0002
0003 ===================================================================
0004 The Definitive KVM (Kernel-based Virtual Machine) API Documentation
0005 ===================================================================
0006
0007 1. General description
0008 ======================
0009
0010 The kvm API is a set of ioctls that are issued to control various aspects
0011 of a virtual machine. The ioctls belong to the following classes:
0012
0013 - System ioctls: These query and set global attributes which affect the
0014 whole kvm subsystem. In addition a system ioctl is used to create
0015 virtual machines.
0016
0017 - VM ioctls: These query and set attributes that affect an entire virtual
0018 machine, for example memory layout. In addition a VM ioctl is used to
0019 create virtual cpus (vcpus) and devices.
0020
0021 VM ioctls must be issued from the same process (address space) that was
0022 used to create the VM.
0023
0024 - vcpu ioctls: These query and set attributes that control the operation
0025 of a single virtual cpu.
0026
0027 vcpu ioctls should be issued from the same thread that was used to create
0028 the vcpu, except for asynchronous vcpu ioctl that are marked as such in
0029 the documentation. Otherwise, the first ioctl after switching threads
0030 could see a performance impact.
0031
0032 - device ioctls: These query and set attributes that control the operation
0033 of a single device.
0034
0035 device ioctls must be issued from the same process (address space) that
0036 was used to create the VM.
0037
0038 2. File descriptors
0039 ===================
0040
0041 The kvm API is centered around file descriptors. An initial
0042 open("/dev/kvm") obtains a handle to the kvm subsystem; this handle
0043 can be used to issue system ioctls. A KVM_CREATE_VM ioctl on this
0044 handle will create a VM file descriptor which can be used to issue VM
0045 ioctls. A KVM_CREATE_VCPU or KVM_CREATE_DEVICE ioctl on a VM fd will
0046 create a virtual cpu or device and return a file descriptor pointing to
0047 the new resource. Finally, ioctls on a vcpu or device fd can be used
0048 to control the vcpu or device. For vcpus, this includes the important
0049 task of actually running guest code.
0050
0051 In general file descriptors can be migrated among processes by means
0052 of fork() and the SCM_RIGHTS facility of unix domain socket. These
0053 kinds of tricks are explicitly not supported by kvm. While they will
0054 not cause harm to the host, their actual behavior is not guaranteed by
0055 the API. See "General description" for details on the ioctl usage
0056 model that is supported by KVM.
0057
0058 It is important to note that although VM ioctls may only be issued from
0059 the process that created the VM, a VM's lifecycle is associated with its
0060 file descriptor, not its creator (process). In other words, the VM and
0061 its resources, *including the associated address space*, are not freed
0062 until the last reference to the VM's file descriptor has been released.
0063 For example, if fork() is issued after ioctl(KVM_CREATE_VM), the VM will
0064 not be freed until both the parent (original) process and its child have
0065 put their references to the VM's file descriptor.
0066
0067 Because a VM's resources are not freed until the last reference to its
0068 file descriptor is released, creating additional references to a VM
0069 via fork(), dup(), etc... without careful consideration is strongly
0070 discouraged and may have unwanted side effects, e.g. memory allocated
0071 by and on behalf of the VM's process may not be freed/unaccounted when
0072 the VM is shut down.
0073
0074
0075 3. Extensions
0076 =============
0077
0078 As of Linux 2.6.22, the KVM ABI has been stabilized: no backward
0079 incompatible change are allowed. However, there is an extension
0080 facility that allows backward-compatible extensions to the API to be
0081 queried and used.
0082
0083 The extension mechanism is not based on the Linux version number.
0084 Instead, kvm defines extension identifiers and a facility to query
0085 whether a particular extension identifier is available. If it is, a
0086 set of ioctls is available for application use.
0087
0088
0089 4. API description
0090 ==================
0091
0092 This section describes ioctls that can be used to control kvm guests.
0093 For each ioctl, the following information is provided along with a
0094 description:
0095
0096 Capability:
0097 which KVM extension provides this ioctl. Can be 'basic',
0098 which means that is will be provided by any kernel that supports
0099 API version 12 (see section 4.1), a KVM_CAP_xyz constant, which
0100 means availability needs to be checked with KVM_CHECK_EXTENSION
0101 (see section 4.4), or 'none' which means that while not all kernels
0102 support this ioctl, there's no capability bit to check its
0103 availability: for kernels that don't support the ioctl,
0104 the ioctl returns -ENOTTY.
0105
0106 Architectures:
0107 which instruction set architectures provide this ioctl.
0108 x86 includes both i386 and x86_64.
0109
0110 Type:
0111 system, vm, or vcpu.
0112
0113 Parameters:
0114 what parameters are accepted by the ioctl.
0115
0116 Returns:
0117 the return value. General error numbers (EBADF, ENOMEM, EINVAL)
0118 are not detailed, but errors with specific meanings are.
0119
0120
0121 4.1 KVM_GET_API_VERSION
0122 -----------------------
0123
0124 :Capability: basic
0125 :Architectures: all
0126 :Type: system ioctl
0127 :Parameters: none
0128 :Returns: the constant KVM_API_VERSION (=12)
0129
0130 This identifies the API version as the stable kvm API. It is not
0131 expected that this number will change. However, Linux 2.6.20 and
0132 2.6.21 report earlier versions; these are not documented and not
0133 supported. Applications should refuse to run if KVM_GET_API_VERSION
0134 returns a value other than 12. If this check passes, all ioctls
0135 described as 'basic' will be available.
0136
0137
0138 4.2 KVM_CREATE_VM
0139 -----------------
0140
0141 :Capability: basic
0142 :Architectures: all
0143 :Type: system ioctl
0144 :Parameters: machine type identifier (KVM_VM_*)
0145 :Returns: a VM fd that can be used to control the new virtual machine.
0146
0147 The new VM has no virtual cpus and no memory.
0148 You probably want to use 0 as machine type.
0149
0150 In order to create user controlled virtual machines on S390, check
0151 KVM_CAP_S390_UCONTROL and use the flag KVM_VM_S390_UCONTROL as
0152 privileged user (CAP_SYS_ADMIN).
0153
0154 On arm64, the physical address size for a VM (IPA Size limit) is limited
0155 to 40bits by default. The limit can be configured if the host supports the
0156 extension KVM_CAP_ARM_VM_IPA_SIZE. When supported, use
0157 KVM_VM_TYPE_ARM_IPA_SIZE(IPA_Bits) to set the size in the machine type
0158 identifier, where IPA_Bits is the maximum width of any physical
0159 address used by the VM. The IPA_Bits is encoded in bits[7-0] of the
0160 machine type identifier.
0161
0162 e.g, to configure a guest to use 48bit physical address size::
0163
0164 vm_fd = ioctl(dev_fd, KVM_CREATE_VM, KVM_VM_TYPE_ARM_IPA_SIZE(48));
0165
0166 The requested size (IPA_Bits) must be:
0167
0168 == =========================================================
0169 0 Implies default size, 40bits (for backward compatibility)
0170 N Implies N bits, where N is a positive integer such that,
0171 32 <= N <= Host_IPA_Limit
0172 == =========================================================
0173
0174 Host_IPA_Limit is the maximum possible value for IPA_Bits on the host and
0175 is dependent on the CPU capability and the kernel configuration. The limit can
0176 be retrieved using KVM_CAP_ARM_VM_IPA_SIZE of the KVM_CHECK_EXTENSION
0177 ioctl() at run-time.
0178
0179 Creation of the VM will fail if the requested IPA size (whether it is
0180 implicit or explicit) is unsupported on the host.
0181
0182 Please note that configuring the IPA size does not affect the capability
0183 exposed by the guest CPUs in ID_AA64MMFR0_EL1[PARange]. It only affects
0184 size of the address translated by the stage2 level (guest physical to
0185 host physical address translations).
0186
0187
0188 4.3 KVM_GET_MSR_INDEX_LIST, KVM_GET_MSR_FEATURE_INDEX_LIST
0189 ----------------------------------------------------------
0190
0191 :Capability: basic, KVM_CAP_GET_MSR_FEATURES for KVM_GET_MSR_FEATURE_INDEX_LIST
0192 :Architectures: x86
0193 :Type: system ioctl
0194 :Parameters: struct kvm_msr_list (in/out)
0195 :Returns: 0 on success; -1 on error
0196
0197 Errors:
0198
0199 ====== ============================================================
0200 EFAULT the msr index list cannot be read from or written to
0201 E2BIG the msr index list is too big to fit in the array specified by
0202 the user.
0203 ====== ============================================================
0204
0205 ::
0206
0207 struct kvm_msr_list {
0208 __u32 nmsrs; /* number of msrs in entries */
0209 __u32 indices[0];
0210 };
0211
0212 The user fills in the size of the indices array in nmsrs, and in return
0213 kvm adjusts nmsrs to reflect the actual number of msrs and fills in the
0214 indices array with their numbers.
0215
0216 KVM_GET_MSR_INDEX_LIST returns the guest msrs that are supported. The list
0217 varies by kvm version and host processor, but does not change otherwise.
0218
0219 Note: if kvm indicates supports MCE (KVM_CAP_MCE), then the MCE bank MSRs are
0220 not returned in the MSR list, as different vcpus can have a different number
0221 of banks, as set via the KVM_X86_SETUP_MCE ioctl.
0222
0223 KVM_GET_MSR_FEATURE_INDEX_LIST returns the list of MSRs that can be passed
0224 to the KVM_GET_MSRS system ioctl. This lets userspace probe host capabilities
0225 and processor features that are exposed via MSRs (e.g., VMX capabilities).
0226 This list also varies by kvm version and host processor, but does not change
0227 otherwise.
0228
0229
0230 4.4 KVM_CHECK_EXTENSION
0231 -----------------------
0232
0233 :Capability: basic, KVM_CAP_CHECK_EXTENSION_VM for vm ioctl
0234 :Architectures: all
0235 :Type: system ioctl, vm ioctl
0236 :Parameters: extension identifier (KVM_CAP_*)
0237 :Returns: 0 if unsupported; 1 (or some other positive integer) if supported
0238
0239 The API allows the application to query about extensions to the core
0240 kvm API. Userspace passes an extension identifier (an integer) and
0241 receives an integer that describes the extension availability.
0242 Generally 0 means no and 1 means yes, but some extensions may report
0243 additional information in the integer return value.
0244
0245 Based on their initialization different VMs may have different capabilities.
0246 It is thus encouraged to use the vm ioctl to query for capabilities (available
0247 with KVM_CAP_CHECK_EXTENSION_VM on the vm fd)
0248
0249 4.5 KVM_GET_VCPU_MMAP_SIZE
0250 --------------------------
0251
0252 :Capability: basic
0253 :Architectures: all
0254 :Type: system ioctl
0255 :Parameters: none
0256 :Returns: size of vcpu mmap area, in bytes
0257
0258 The KVM_RUN ioctl (cf.) communicates with userspace via a shared
0259 memory region. This ioctl returns the size of that region. See the
0260 KVM_RUN documentation for details.
0261
0262 Besides the size of the KVM_RUN communication region, other areas of
0263 the VCPU file descriptor can be mmap-ed, including:
0264
0265 - if KVM_CAP_COALESCED_MMIO is available, a page at
0266 KVM_COALESCED_MMIO_PAGE_OFFSET * PAGE_SIZE; for historical reasons,
0267 this page is included in the result of KVM_GET_VCPU_MMAP_SIZE.
0268 KVM_CAP_COALESCED_MMIO is not documented yet.
0269
0270 - if KVM_CAP_DIRTY_LOG_RING is available, a number of pages at
0271 KVM_DIRTY_LOG_PAGE_OFFSET * PAGE_SIZE. For more information on
0272 KVM_CAP_DIRTY_LOG_RING, see section 8.3.
0273
0274
0275 4.6 KVM_SET_MEMORY_REGION
0276 -------------------------
0277
0278 :Capability: basic
0279 :Architectures: all
0280 :Type: vm ioctl
0281 :Parameters: struct kvm_memory_region (in)
0282 :Returns: 0 on success, -1 on error
0283
0284 This ioctl is obsolete and has been removed.
0285
0286
0287 4.7 KVM_CREATE_VCPU
0288 -------------------
0289
0290 :Capability: basic
0291 :Architectures: all
0292 :Type: vm ioctl
0293 :Parameters: vcpu id (apic id on x86)
0294 :Returns: vcpu fd on success, -1 on error
0295
0296 This API adds a vcpu to a virtual machine. No more than max_vcpus may be added.
0297 The vcpu id is an integer in the range [0, max_vcpu_id).
0298
0299 The recommended max_vcpus value can be retrieved using the KVM_CAP_NR_VCPUS of
0300 the KVM_CHECK_EXTENSION ioctl() at run-time.
0301 The maximum possible value for max_vcpus can be retrieved using the
0302 KVM_CAP_MAX_VCPUS of the KVM_CHECK_EXTENSION ioctl() at run-time.
0303
0304 If the KVM_CAP_NR_VCPUS does not exist, you should assume that max_vcpus is 4
0305 cpus max.
0306 If the KVM_CAP_MAX_VCPUS does not exist, you should assume that max_vcpus is
0307 same as the value returned from KVM_CAP_NR_VCPUS.
0308
0309 The maximum possible value for max_vcpu_id can be retrieved using the
0310 KVM_CAP_MAX_VCPU_ID of the KVM_CHECK_EXTENSION ioctl() at run-time.
0311
0312 If the KVM_CAP_MAX_VCPU_ID does not exist, you should assume that max_vcpu_id
0313 is the same as the value returned from KVM_CAP_MAX_VCPUS.
0314
0315 On powerpc using book3s_hv mode, the vcpus are mapped onto virtual
0316 threads in one or more virtual CPU cores. (This is because the
0317 hardware requires all the hardware threads in a CPU core to be in the
0318 same partition.) The KVM_CAP_PPC_SMT capability indicates the number
0319 of vcpus per virtual core (vcore). The vcore id is obtained by
0320 dividing the vcpu id by the number of vcpus per vcore. The vcpus in a
0321 given vcore will always be in the same physical core as each other
0322 (though that might be a different physical core from time to time).
0323 Userspace can control the threading (SMT) mode of the guest by its
0324 allocation of vcpu ids. For example, if userspace wants
0325 single-threaded guest vcpus, it should make all vcpu ids be a multiple
0326 of the number of vcpus per vcore.
0327
0328 For virtual cpus that have been created with S390 user controlled virtual
0329 machines, the resulting vcpu fd can be memory mapped at page offset
0330 KVM_S390_SIE_PAGE_OFFSET in order to obtain a memory map of the virtual
0331 cpu's hardware control block.
0332
0333
0334 4.8 KVM_GET_DIRTY_LOG (vm ioctl)
0335 --------------------------------
0336
0337 :Capability: basic
0338 :Architectures: all
0339 :Type: vm ioctl
0340 :Parameters: struct kvm_dirty_log (in/out)
0341 :Returns: 0 on success, -1 on error
0342
0343 ::
0344
0345 /* for KVM_GET_DIRTY_LOG */
0346 struct kvm_dirty_log {
0347 __u32 slot;
0348 __u32 padding;
0349 union {
0350 void __user *dirty_bitmap; /* one bit per page */
0351 __u64 padding;
0352 };
0353 };
0354
0355 Given a memory slot, return a bitmap containing any pages dirtied
0356 since the last call to this ioctl. Bit 0 is the first page in the
0357 memory slot. Ensure the entire structure is cleared to avoid padding
0358 issues.
0359
0360 If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of slot field specifies
0361 the address space for which you want to return the dirty bitmap. See
0362 KVM_SET_USER_MEMORY_REGION for details on the usage of slot field.
0363
0364 The bits in the dirty bitmap are cleared before the ioctl returns, unless
0365 KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 is enabled. For more information,
0366 see the description of the capability.
0367
0368 Note that the Xen shared info page, if configured, shall always be assumed
0369 to be dirty. KVM will not explicitly mark it such.
0370
0371 4.9 KVM_SET_MEMORY_ALIAS
0372 ------------------------
0373
0374 :Capability: basic
0375 :Architectures: x86
0376 :Type: vm ioctl
0377 :Parameters: struct kvm_memory_alias (in)
0378 :Returns: 0 (success), -1 (error)
0379
0380 This ioctl is obsolete and has been removed.
0381
0382
0383 4.10 KVM_RUN
0384 ------------
0385
0386 :Capability: basic
0387 :Architectures: all
0388 :Type: vcpu ioctl
0389 :Parameters: none
0390 :Returns: 0 on success, -1 on error
0391
0392 Errors:
0393
0394 ======= ==============================================================
0395 EINTR an unmasked signal is pending
0396 ENOEXEC the vcpu hasn't been initialized or the guest tried to execute
0397 instructions from device memory (arm64)
0398 ENOSYS data abort outside memslots with no syndrome info and
0399 KVM_CAP_ARM_NISV_TO_USER not enabled (arm64)
0400 EPERM SVE feature set but not finalized (arm64)
0401 ======= ==============================================================
0402
0403 This ioctl is used to run a guest virtual cpu. While there are no
0404 explicit parameters, there is an implicit parameter block that can be
0405 obtained by mmap()ing the vcpu fd at offset 0, with the size given by
0406 KVM_GET_VCPU_MMAP_SIZE. The parameter block is formatted as a 'struct
0407 kvm_run' (see below).
0408
0409
0410 4.11 KVM_GET_REGS
0411 -----------------
0412
0413 :Capability: basic
0414 :Architectures: all except arm64
0415 :Type: vcpu ioctl
0416 :Parameters: struct kvm_regs (out)
0417 :Returns: 0 on success, -1 on error
0418
0419 Reads the general purpose registers from the vcpu.
0420
0421 ::
0422
0423 /* x86 */
0424 struct kvm_regs {
0425 /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */
0426 __u64 rax, rbx, rcx, rdx;
0427 __u64 rsi, rdi, rsp, rbp;
0428 __u64 r8, r9, r10, r11;
0429 __u64 r12, r13, r14, r15;
0430 __u64 rip, rflags;
0431 };
0432
0433 /* mips */
0434 struct kvm_regs {
0435 /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */
0436 __u64 gpr[32];
0437 __u64 hi;
0438 __u64 lo;
0439 __u64 pc;
0440 };
0441
0442
0443 4.12 KVM_SET_REGS
0444 -----------------
0445
0446 :Capability: basic
0447 :Architectures: all except arm64
0448 :Type: vcpu ioctl
0449 :Parameters: struct kvm_regs (in)
0450 :Returns: 0 on success, -1 on error
0451
0452 Writes the general purpose registers into the vcpu.
0453
0454 See KVM_GET_REGS for the data structure.
0455
0456
0457 4.13 KVM_GET_SREGS
0458 ------------------
0459
0460 :Capability: basic
0461 :Architectures: x86, ppc
0462 :Type: vcpu ioctl
0463 :Parameters: struct kvm_sregs (out)
0464 :Returns: 0 on success, -1 on error
0465
0466 Reads special registers from the vcpu.
0467
0468 ::
0469
0470 /* x86 */
0471 struct kvm_sregs {
0472 struct kvm_segment cs, ds, es, fs, gs, ss;
0473 struct kvm_segment tr, ldt;
0474 struct kvm_dtable gdt, idt;
0475 __u64 cr0, cr2, cr3, cr4, cr8;
0476 __u64 efer;
0477 __u64 apic_base;
0478 __u64 interrupt_bitmap[(KVM_NR_INTERRUPTS + 63) / 64];
0479 };
0480
0481 /* ppc -- see arch/powerpc/include/uapi/asm/kvm.h */
0482
0483 interrupt_bitmap is a bitmap of pending external interrupts. At most
0484 one bit may be set. This interrupt has been acknowledged by the APIC
0485 but not yet injected into the cpu core.
0486
0487
0488 4.14 KVM_SET_SREGS
0489 ------------------
0490
0491 :Capability: basic
0492 :Architectures: x86, ppc
0493 :Type: vcpu ioctl
0494 :Parameters: struct kvm_sregs (in)
0495 :Returns: 0 on success, -1 on error
0496
0497 Writes special registers into the vcpu. See KVM_GET_SREGS for the
0498 data structures.
0499
0500
0501 4.15 KVM_TRANSLATE
0502 ------------------
0503
0504 :Capability: basic
0505 :Architectures: x86
0506 :Type: vcpu ioctl
0507 :Parameters: struct kvm_translation (in/out)
0508 :Returns: 0 on success, -1 on error
0509
0510 Translates a virtual address according to the vcpu's current address
0511 translation mode.
0512
0513 ::
0514
0515 struct kvm_translation {
0516 /* in */
0517 __u64 linear_address;
0518
0519 /* out */
0520 __u64 physical_address;
0521 __u8 valid;
0522 __u8 writeable;
0523 __u8 usermode;
0524 __u8 pad[5];
0525 };
0526
0527
0528 4.16 KVM_INTERRUPT
0529 ------------------
0530
0531 :Capability: basic
0532 :Architectures: x86, ppc, mips, riscv
0533 :Type: vcpu ioctl
0534 :Parameters: struct kvm_interrupt (in)
0535 :Returns: 0 on success, negative on failure.
0536
0537 Queues a hardware interrupt vector to be injected.
0538
0539 ::
0540
0541 /* for KVM_INTERRUPT */
0542 struct kvm_interrupt {
0543 /* in */
0544 __u32 irq;
0545 };
0546
0547 X86:
0548 ^^^^
0549
0550 :Returns:
0551
0552 ========= ===================================
0553 0 on success,
0554 -EEXIST if an interrupt is already enqueued
0555 -EINVAL the irq number is invalid
0556 -ENXIO if the PIC is in the kernel
0557 -EFAULT if the pointer is invalid
0558 ========= ===================================
0559
0560 Note 'irq' is an interrupt vector, not an interrupt pin or line. This
0561 ioctl is useful if the in-kernel PIC is not used.
0562
0563 PPC:
0564 ^^^^
0565
0566 Queues an external interrupt to be injected. This ioctl is overleaded
0567 with 3 different irq values:
0568
0569 a) KVM_INTERRUPT_SET
0570
0571 This injects an edge type external interrupt into the guest once it's ready
0572 to receive interrupts. When injected, the interrupt is done.
0573
0574 b) KVM_INTERRUPT_UNSET
0575
0576 This unsets any pending interrupt.
0577
0578 Only available with KVM_CAP_PPC_UNSET_IRQ.
0579
0580 c) KVM_INTERRUPT_SET_LEVEL
0581
0582 This injects a level type external interrupt into the guest context. The
0583 interrupt stays pending until a specific ioctl with KVM_INTERRUPT_UNSET
0584 is triggered.
0585
0586 Only available with KVM_CAP_PPC_IRQ_LEVEL.
0587
0588 Note that any value for 'irq' other than the ones stated above is invalid
0589 and incurs unexpected behavior.
0590
0591 This is an asynchronous vcpu ioctl and can be invoked from any thread.
0592
0593 MIPS:
0594 ^^^^^
0595
0596 Queues an external interrupt to be injected into the virtual CPU. A negative
0597 interrupt number dequeues the interrupt.
0598
0599 This is an asynchronous vcpu ioctl and can be invoked from any thread.
0600
0601 RISC-V:
0602 ^^^^^^^
0603
0604 Queues an external interrupt to be injected into the virutal CPU. This ioctl
0605 is overloaded with 2 different irq values:
0606
0607 a) KVM_INTERRUPT_SET
0608
0609 This sets external interrupt for a virtual CPU and it will receive
0610 once it is ready.
0611
0612 b) KVM_INTERRUPT_UNSET
0613
0614 This clears pending external interrupt for a virtual CPU.
0615
0616 This is an asynchronous vcpu ioctl and can be invoked from any thread.
0617
0618
0619 4.17 KVM_DEBUG_GUEST
0620 --------------------
0621
0622 :Capability: basic
0623 :Architectures: none
0624 :Type: vcpu ioctl
0625 :Parameters: none)
0626 :Returns: -1 on error
0627
0628 Support for this has been removed. Use KVM_SET_GUEST_DEBUG instead.
0629
0630
0631 4.18 KVM_GET_MSRS
0632 -----------------
0633
0634 :Capability: basic (vcpu), KVM_CAP_GET_MSR_FEATURES (system)
0635 :Architectures: x86
0636 :Type: system ioctl, vcpu ioctl
0637 :Parameters: struct kvm_msrs (in/out)
0638 :Returns: number of msrs successfully returned;
0639 -1 on error
0640
0641 When used as a system ioctl:
0642 Reads the values of MSR-based features that are available for the VM. This
0643 is similar to KVM_GET_SUPPORTED_CPUID, but it returns MSR indices and values.
0644 The list of msr-based features can be obtained using KVM_GET_MSR_FEATURE_INDEX_LIST
0645 in a system ioctl.
0646
0647 When used as a vcpu ioctl:
0648 Reads model-specific registers from the vcpu. Supported msr indices can
0649 be obtained using KVM_GET_MSR_INDEX_LIST in a system ioctl.
0650
0651 ::
0652
0653 struct kvm_msrs {
0654 __u32 nmsrs; /* number of msrs in entries */
0655 __u32 pad;
0656
0657 struct kvm_msr_entry entries[0];
0658 };
0659
0660 struct kvm_msr_entry {
0661 __u32 index;
0662 __u32 reserved;
0663 __u64 data;
0664 };
0665
0666 Application code should set the 'nmsrs' member (which indicates the
0667 size of the entries array) and the 'index' member of each array entry.
0668 kvm will fill in the 'data' member.
0669
0670
0671 4.19 KVM_SET_MSRS
0672 -----------------
0673
0674 :Capability: basic
0675 :Architectures: x86
0676 :Type: vcpu ioctl
0677 :Parameters: struct kvm_msrs (in)
0678 :Returns: number of msrs successfully set (see below), -1 on error
0679
0680 Writes model-specific registers to the vcpu. See KVM_GET_MSRS for the
0681 data structures.
0682
0683 Application code should set the 'nmsrs' member (which indicates the
0684 size of the entries array), and the 'index' and 'data' members of each
0685 array entry.
0686
0687 It tries to set the MSRs in array entries[] one by one. If setting an MSR
0688 fails, e.g., due to setting reserved bits, the MSR isn't supported/emulated
0689 by KVM, etc..., it stops processing the MSR list and returns the number of
0690 MSRs that have been set successfully.
0691
0692
0693 4.20 KVM_SET_CPUID
0694 ------------------
0695
0696 :Capability: basic
0697 :Architectures: x86
0698 :Type: vcpu ioctl
0699 :Parameters: struct kvm_cpuid (in)
0700 :Returns: 0 on success, -1 on error
0701
0702 Defines the vcpu responses to the cpuid instruction. Applications
0703 should use the KVM_SET_CPUID2 ioctl if available.
0704
0705 Caveat emptor:
0706 - If this IOCTL fails, KVM gives no guarantees that previous valid CPUID
0707 configuration (if there is) is not corrupted. Userspace can get a copy
0708 of the resulting CPUID configuration through KVM_GET_CPUID2 in case.
0709 - Using KVM_SET_CPUID{,2} after KVM_RUN, i.e. changing the guest vCPU model
0710 after running the guest, may cause guest instability.
0711 - Using heterogeneous CPUID configurations, modulo APIC IDs, topology, etc...
0712 may cause guest instability.
0713
0714 ::
0715
0716 struct kvm_cpuid_entry {
0717 __u32 function;
0718 __u32 eax;
0719 __u32 ebx;
0720 __u32 ecx;
0721 __u32 edx;
0722 __u32 padding;
0723 };
0724
0725 /* for KVM_SET_CPUID */
0726 struct kvm_cpuid {
0727 __u32 nent;
0728 __u32 padding;
0729 struct kvm_cpuid_entry entries[0];
0730 };
0731
0732
0733 4.21 KVM_SET_SIGNAL_MASK
0734 ------------------------
0735
0736 :Capability: basic
0737 :Architectures: all
0738 :Type: vcpu ioctl
0739 :Parameters: struct kvm_signal_mask (in)
0740 :Returns: 0 on success, -1 on error
0741
0742 Defines which signals are blocked during execution of KVM_RUN. This
0743 signal mask temporarily overrides the threads signal mask. Any
0744 unblocked signal received (except SIGKILL and SIGSTOP, which retain
0745 their traditional behaviour) will cause KVM_RUN to return with -EINTR.
0746
0747 Note the signal will only be delivered if not blocked by the original
0748 signal mask.
0749
0750 ::
0751
0752 /* for KVM_SET_SIGNAL_MASK */
0753 struct kvm_signal_mask {
0754 __u32 len;
0755 __u8 sigset[0];
0756 };
0757
0758
0759 4.22 KVM_GET_FPU
0760 ----------------
0761
0762 :Capability: basic
0763 :Architectures: x86
0764 :Type: vcpu ioctl
0765 :Parameters: struct kvm_fpu (out)
0766 :Returns: 0 on success, -1 on error
0767
0768 Reads the floating point state from the vcpu.
0769
0770 ::
0771
0772 /* for KVM_GET_FPU and KVM_SET_FPU */
0773 struct kvm_fpu {
0774 __u8 fpr[8][16];
0775 __u16 fcw;
0776 __u16 fsw;
0777 __u8 ftwx; /* in fxsave format */
0778 __u8 pad1;
0779 __u16 last_opcode;
0780 __u64 last_ip;
0781 __u64 last_dp;
0782 __u8 xmm[16][16];
0783 __u32 mxcsr;
0784 __u32 pad2;
0785 };
0786
0787
0788 4.23 KVM_SET_FPU
0789 ----------------
0790
0791 :Capability: basic
0792 :Architectures: x86
0793 :Type: vcpu ioctl
0794 :Parameters: struct kvm_fpu (in)
0795 :Returns: 0 on success, -1 on error
0796
0797 Writes the floating point state to the vcpu.
0798
0799 ::
0800
0801 /* for KVM_GET_FPU and KVM_SET_FPU */
0802 struct kvm_fpu {
0803 __u8 fpr[8][16];
0804 __u16 fcw;
0805 __u16 fsw;
0806 __u8 ftwx; /* in fxsave format */
0807 __u8 pad1;
0808 __u16 last_opcode;
0809 __u64 last_ip;
0810 __u64 last_dp;
0811 __u8 xmm[16][16];
0812 __u32 mxcsr;
0813 __u32 pad2;
0814 };
0815
0816
0817 4.24 KVM_CREATE_IRQCHIP
0818 -----------------------
0819
0820 :Capability: KVM_CAP_IRQCHIP, KVM_CAP_S390_IRQCHIP (s390)
0821 :Architectures: x86, arm64, s390
0822 :Type: vm ioctl
0823 :Parameters: none
0824 :Returns: 0 on success, -1 on error
0825
0826 Creates an interrupt controller model in the kernel.
0827 On x86, creates a virtual ioapic, a virtual PIC (two PICs, nested), and sets up
0828 future vcpus to have a local APIC. IRQ routing for GSIs 0-15 is set to both
0829 PIC and IOAPIC; GSI 16-23 only go to the IOAPIC.
0830 On arm64, a GICv2 is created. Any other GIC versions require the usage of
0831 KVM_CREATE_DEVICE, which also supports creating a GICv2. Using
0832 KVM_CREATE_DEVICE is preferred over KVM_CREATE_IRQCHIP for GICv2.
0833 On s390, a dummy irq routing table is created.
0834
0835 Note that on s390 the KVM_CAP_S390_IRQCHIP vm capability needs to be enabled
0836 before KVM_CREATE_IRQCHIP can be used.
0837
0838
0839 4.25 KVM_IRQ_LINE
0840 -----------------
0841
0842 :Capability: KVM_CAP_IRQCHIP
0843 :Architectures: x86, arm64
0844 :Type: vm ioctl
0845 :Parameters: struct kvm_irq_level
0846 :Returns: 0 on success, -1 on error
0847
0848 Sets the level of a GSI input to the interrupt controller model in the kernel.
0849 On some architectures it is required that an interrupt controller model has
0850 been previously created with KVM_CREATE_IRQCHIP. Note that edge-triggered
0851 interrupts require the level to be set to 1 and then back to 0.
0852
0853 On real hardware, interrupt pins can be active-low or active-high. This
0854 does not matter for the level field of struct kvm_irq_level: 1 always
0855 means active (asserted), 0 means inactive (deasserted).
0856
0857 x86 allows the operating system to program the interrupt polarity
0858 (active-low/active-high) for level-triggered interrupts, and KVM used
0859 to consider the polarity. However, due to bitrot in the handling of
0860 active-low interrupts, the above convention is now valid on x86 too.
0861 This is signaled by KVM_CAP_X86_IOAPIC_POLARITY_IGNORED. Userspace
0862 should not present interrupts to the guest as active-low unless this
0863 capability is present (or unless it is not using the in-kernel irqchip,
0864 of course).
0865
0866
0867 arm64 can signal an interrupt either at the CPU level, or at the
0868 in-kernel irqchip (GIC), and for in-kernel irqchip can tell the GIC to
0869 use PPIs designated for specific cpus. The irq field is interpreted
0870 like this::
0871
0872 bits: | 31 ... 28 | 27 ... 24 | 23 ... 16 | 15 ... 0 |
0873 field: | vcpu2_index | irq_type | vcpu_index | irq_id |
0874
0875 The irq_type field has the following values:
0876
0877 - irq_type[0]:
0878 out-of-kernel GIC: irq_id 0 is IRQ, irq_id 1 is FIQ
0879 - irq_type[1]:
0880 in-kernel GIC: SPI, irq_id between 32 and 1019 (incl.)
0881 (the vcpu_index field is ignored)
0882 - irq_type[2]:
0883 in-kernel GIC: PPI, irq_id between 16 and 31 (incl.)
0884
0885 (The irq_id field thus corresponds nicely to the IRQ ID in the ARM GIC specs)
0886
0887 In both cases, level is used to assert/deassert the line.
0888
0889 When KVM_CAP_ARM_IRQ_LINE_LAYOUT_2 is supported, the target vcpu is
0890 identified as (256 * vcpu2_index + vcpu_index). Otherwise, vcpu2_index
0891 must be zero.
0892
0893 Note that on arm64, the KVM_CAP_IRQCHIP capability only conditions
0894 injection of interrupts for the in-kernel irqchip. KVM_IRQ_LINE can always
0895 be used for a userspace interrupt controller.
0896
0897 ::
0898
0899 struct kvm_irq_level {
0900 union {
0901 __u32 irq; /* GSI */
0902 __s32 status; /* not used for KVM_IRQ_LEVEL */
0903 };
0904 __u32 level; /* 0 or 1 */
0905 };
0906
0907
0908 4.26 KVM_GET_IRQCHIP
0909 --------------------
0910
0911 :Capability: KVM_CAP_IRQCHIP
0912 :Architectures: x86
0913 :Type: vm ioctl
0914 :Parameters: struct kvm_irqchip (in/out)
0915 :Returns: 0 on success, -1 on error
0916
0917 Reads the state of a kernel interrupt controller created with
0918 KVM_CREATE_IRQCHIP into a buffer provided by the caller.
0919
0920 ::
0921
0922 struct kvm_irqchip {
0923 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */
0924 __u32 pad;
0925 union {
0926 char dummy[512]; /* reserving space */
0927 struct kvm_pic_state pic;
0928 struct kvm_ioapic_state ioapic;
0929 } chip;
0930 };
0931
0932
0933 4.27 KVM_SET_IRQCHIP
0934 --------------------
0935
0936 :Capability: KVM_CAP_IRQCHIP
0937 :Architectures: x86
0938 :Type: vm ioctl
0939 :Parameters: struct kvm_irqchip (in)
0940 :Returns: 0 on success, -1 on error
0941
0942 Sets the state of a kernel interrupt controller created with
0943 KVM_CREATE_IRQCHIP from a buffer provided by the caller.
0944
0945 ::
0946
0947 struct kvm_irqchip {
0948 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */
0949 __u32 pad;
0950 union {
0951 char dummy[512]; /* reserving space */
0952 struct kvm_pic_state pic;
0953 struct kvm_ioapic_state ioapic;
0954 } chip;
0955 };
0956
0957
0958 4.28 KVM_XEN_HVM_CONFIG
0959 -----------------------
0960
0961 :Capability: KVM_CAP_XEN_HVM
0962 :Architectures: x86
0963 :Type: vm ioctl
0964 :Parameters: struct kvm_xen_hvm_config (in)
0965 :Returns: 0 on success, -1 on error
0966
0967 Sets the MSR that the Xen HVM guest uses to initialize its hypercall
0968 page, and provides the starting address and size of the hypercall
0969 blobs in userspace. When the guest writes the MSR, kvm copies one
0970 page of a blob (32- or 64-bit, depending on the vcpu mode) to guest
0971 memory.
0972
0973 ::
0974
0975 struct kvm_xen_hvm_config {
0976 __u32 flags;
0977 __u32 msr;
0978 __u64 blob_addr_32;
0979 __u64 blob_addr_64;
0980 __u8 blob_size_32;
0981 __u8 blob_size_64;
0982 __u8 pad2[30];
0983 };
0984
0985 If certain flags are returned from the KVM_CAP_XEN_HVM check, they may
0986 be set in the flags field of this ioctl:
0987
0988 The KVM_XEN_HVM_CONFIG_INTERCEPT_HCALL flag requests KVM to generate
0989 the contents of the hypercall page automatically; hypercalls will be
0990 intercepted and passed to userspace through KVM_EXIT_XEN. In this
0991 ase, all of the blob size and address fields must be zero.
0992
0993 The KVM_XEN_HVM_CONFIG_EVTCHN_SEND flag indicates to KVM that userspace
0994 will always use the KVM_XEN_HVM_EVTCHN_SEND ioctl to deliver event
0995 channel interrupts rather than manipulating the guest's shared_info
0996 structures directly. This, in turn, may allow KVM to enable features
0997 such as intercepting the SCHEDOP_poll hypercall to accelerate PV
0998 spinlock operation for the guest. Userspace may still use the ioctl
0999 to deliver events if it was advertised, even if userspace does not
1000 send this indication that it will always do so
1001
1002 No other flags are currently valid in the struct kvm_xen_hvm_config.
1003
1004 4.29 KVM_GET_CLOCK
1005 ------------------
1006
1007 :Capability: KVM_CAP_ADJUST_CLOCK
1008 :Architectures: x86
1009 :Type: vm ioctl
1010 :Parameters: struct kvm_clock_data (out)
1011 :Returns: 0 on success, -1 on error
1012
1013 Gets the current timestamp of kvmclock as seen by the current guest. In
1014 conjunction with KVM_SET_CLOCK, it is used to ensure monotonicity on scenarios
1015 such as migration.
1016
1017 When KVM_CAP_ADJUST_CLOCK is passed to KVM_CHECK_EXTENSION, it returns the
1018 set of bits that KVM can return in struct kvm_clock_data's flag member.
1019
1020 The following flags are defined:
1021
1022 KVM_CLOCK_TSC_STABLE
1023 If set, the returned value is the exact kvmclock
1024 value seen by all VCPUs at the instant when KVM_GET_CLOCK was called.
1025 If clear, the returned value is simply CLOCK_MONOTONIC plus a constant
1026 offset; the offset can be modified with KVM_SET_CLOCK. KVM will try
1027 to make all VCPUs follow this clock, but the exact value read by each
1028 VCPU could differ, because the host TSC is not stable.
1029
1030 KVM_CLOCK_REALTIME
1031 If set, the `realtime` field in the kvm_clock_data
1032 structure is populated with the value of the host's real time
1033 clocksource at the instant when KVM_GET_CLOCK was called. If clear,
1034 the `realtime` field does not contain a value.
1035
1036 KVM_CLOCK_HOST_TSC
1037 If set, the `host_tsc` field in the kvm_clock_data
1038 structure is populated with the value of the host's timestamp counter (TSC)
1039 at the instant when KVM_GET_CLOCK was called. If clear, the `host_tsc` field
1040 does not contain a value.
1041
1042 ::
1043
1044 struct kvm_clock_data {
1045 __u64 clock; /* kvmclock current value */
1046 __u32 flags;
1047 __u32 pad0;
1048 __u64 realtime;
1049 __u64 host_tsc;
1050 __u32 pad[4];
1051 };
1052
1053
1054 4.30 KVM_SET_CLOCK
1055 ------------------
1056
1057 :Capability: KVM_CAP_ADJUST_CLOCK
1058 :Architectures: x86
1059 :Type: vm ioctl
1060 :Parameters: struct kvm_clock_data (in)
1061 :Returns: 0 on success, -1 on error
1062
1063 Sets the current timestamp of kvmclock to the value specified in its parameter.
1064 In conjunction with KVM_GET_CLOCK, it is used to ensure monotonicity on scenarios
1065 such as migration.
1066
1067 The following flags can be passed:
1068
1069 KVM_CLOCK_REALTIME
1070 If set, KVM will compare the value of the `realtime` field
1071 with the value of the host's real time clocksource at the instant when
1072 KVM_SET_CLOCK was called. The difference in elapsed time is added to the final
1073 kvmclock value that will be provided to guests.
1074
1075 Other flags returned by ``KVM_GET_CLOCK`` are accepted but ignored.
1076
1077 ::
1078
1079 struct kvm_clock_data {
1080 __u64 clock; /* kvmclock current value */
1081 __u32 flags;
1082 __u32 pad0;
1083 __u64 realtime;
1084 __u64 host_tsc;
1085 __u32 pad[4];
1086 };
1087
1088
1089 4.31 KVM_GET_VCPU_EVENTS
1090 ------------------------
1091
1092 :Capability: KVM_CAP_VCPU_EVENTS
1093 :Extended by: KVM_CAP_INTR_SHADOW
1094 :Architectures: x86, arm64
1095 :Type: vcpu ioctl
1096 :Parameters: struct kvm_vcpu_event (out)
1097 :Returns: 0 on success, -1 on error
1098
1099 X86:
1100 ^^^^
1101
1102 Gets currently pending exceptions, interrupts, and NMIs as well as related
1103 states of the vcpu.
1104
1105 ::
1106
1107 struct kvm_vcpu_events {
1108 struct {
1109 __u8 injected;
1110 __u8 nr;
1111 __u8 has_error_code;
1112 __u8 pending;
1113 __u32 error_code;
1114 } exception;
1115 struct {
1116 __u8 injected;
1117 __u8 nr;
1118 __u8 soft;
1119 __u8 shadow;
1120 } interrupt;
1121 struct {
1122 __u8 injected;
1123 __u8 pending;
1124 __u8 masked;
1125 __u8 pad;
1126 } nmi;
1127 __u32 sipi_vector;
1128 __u32 flags;
1129 struct {
1130 __u8 smm;
1131 __u8 pending;
1132 __u8 smm_inside_nmi;
1133 __u8 latched_init;
1134 } smi;
1135 __u8 reserved[27];
1136 __u8 exception_has_payload;
1137 __u64 exception_payload;
1138 };
1139
1140 The following bits are defined in the flags field:
1141
1142 - KVM_VCPUEVENT_VALID_SHADOW may be set to signal that
1143 interrupt.shadow contains a valid state.
1144
1145 - KVM_VCPUEVENT_VALID_SMM may be set to signal that smi contains a
1146 valid state.
1147
1148 - KVM_VCPUEVENT_VALID_PAYLOAD may be set to signal that the
1149 exception_has_payload, exception_payload, and exception.pending
1150 fields contain a valid state. This bit will be set whenever
1151 KVM_CAP_EXCEPTION_PAYLOAD is enabled.
1152
1153 - KVM_VCPUEVENT_VALID_TRIPLE_FAULT may be set to signal that the
1154 triple_fault_pending field contains a valid state. This bit will
1155 be set whenever KVM_CAP_X86_TRIPLE_FAULT_EVENT is enabled.
1156
1157 ARM64:
1158 ^^^^^^
1159
1160 If the guest accesses a device that is being emulated by the host kernel in
1161 such a way that a real device would generate a physical SError, KVM may make
1162 a virtual SError pending for that VCPU. This system error interrupt remains
1163 pending until the guest takes the exception by unmasking PSTATE.A.
1164
1165 Running the VCPU may cause it to take a pending SError, or make an access that
1166 causes an SError to become pending. The event's description is only valid while
1167 the VPCU is not running.
1168
1169 This API provides a way to read and write the pending 'event' state that is not
1170 visible to the guest. To save, restore or migrate a VCPU the struct representing
1171 the state can be read then written using this GET/SET API, along with the other
1172 guest-visible registers. It is not possible to 'cancel' an SError that has been
1173 made pending.
1174
1175 A device being emulated in user-space may also wish to generate an SError. To do
1176 this the events structure can be populated by user-space. The current state
1177 should be read first, to ensure no existing SError is pending. If an existing
1178 SError is pending, the architecture's 'Multiple SError interrupts' rules should
1179 be followed. (2.5.3 of DDI0587.a "ARM Reliability, Availability, and
1180 Serviceability (RAS) Specification").
1181
1182 SError exceptions always have an ESR value. Some CPUs have the ability to
1183 specify what the virtual SError's ESR value should be. These systems will
1184 advertise KVM_CAP_ARM_INJECT_SERROR_ESR. In this case exception.has_esr will
1185 always have a non-zero value when read, and the agent making an SError pending
1186 should specify the ISS field in the lower 24 bits of exception.serror_esr. If
1187 the system supports KVM_CAP_ARM_INJECT_SERROR_ESR, but user-space sets the events
1188 with exception.has_esr as zero, KVM will choose an ESR.
1189
1190 Specifying exception.has_esr on a system that does not support it will return
1191 -EINVAL. Setting anything other than the lower 24bits of exception.serror_esr
1192 will return -EINVAL.
1193
1194 It is not possible to read back a pending external abort (injected via
1195 KVM_SET_VCPU_EVENTS or otherwise) because such an exception is always delivered
1196 directly to the virtual CPU).
1197
1198 ::
1199
1200 struct kvm_vcpu_events {
1201 struct {
1202 __u8 serror_pending;
1203 __u8 serror_has_esr;
1204 __u8 ext_dabt_pending;
1205 /* Align it to 8 bytes */
1206 __u8 pad[5];
1207 __u64 serror_esr;
1208 } exception;
1209 __u32 reserved[12];
1210 };
1211
1212 4.32 KVM_SET_VCPU_EVENTS
1213 ------------------------
1214
1215 :Capability: KVM_CAP_VCPU_EVENTS
1216 :Extended by: KVM_CAP_INTR_SHADOW
1217 :Architectures: x86, arm64
1218 :Type: vcpu ioctl
1219 :Parameters: struct kvm_vcpu_event (in)
1220 :Returns: 0 on success, -1 on error
1221
1222 X86:
1223 ^^^^
1224
1225 Set pending exceptions, interrupts, and NMIs as well as related states of the
1226 vcpu.
1227
1228 See KVM_GET_VCPU_EVENTS for the data structure.
1229
1230 Fields that may be modified asynchronously by running VCPUs can be excluded
1231 from the update. These fields are nmi.pending, sipi_vector, smi.smm,
1232 smi.pending. Keep the corresponding bits in the flags field cleared to
1233 suppress overwriting the current in-kernel state. The bits are:
1234
1235 =============================== ==================================
1236 KVM_VCPUEVENT_VALID_NMI_PENDING transfer nmi.pending to the kernel
1237 KVM_VCPUEVENT_VALID_SIPI_VECTOR transfer sipi_vector
1238 KVM_VCPUEVENT_VALID_SMM transfer the smi sub-struct.
1239 =============================== ==================================
1240
1241 If KVM_CAP_INTR_SHADOW is available, KVM_VCPUEVENT_VALID_SHADOW can be set in
1242 the flags field to signal that interrupt.shadow contains a valid state and
1243 shall be written into the VCPU.
1244
1245 KVM_VCPUEVENT_VALID_SMM can only be set if KVM_CAP_X86_SMM is available.
1246
1247 If KVM_CAP_EXCEPTION_PAYLOAD is enabled, KVM_VCPUEVENT_VALID_PAYLOAD
1248 can be set in the flags field to signal that the
1249 exception_has_payload, exception_payload, and exception.pending fields
1250 contain a valid state and shall be written into the VCPU.
1251
1252 If KVM_CAP_X86_TRIPLE_FAULT_EVENT is enabled, KVM_VCPUEVENT_VALID_TRIPLE_FAULT
1253 can be set in flags field to signal that the triple_fault field contains
1254 a valid state and shall be written into the VCPU.
1255
1256 ARM64:
1257 ^^^^^^
1258
1259 User space may need to inject several types of events to the guest.
1260
1261 Set the pending SError exception state for this VCPU. It is not possible to
1262 'cancel' an Serror that has been made pending.
1263
1264 If the guest performed an access to I/O memory which could not be handled by
1265 userspace, for example because of missing instruction syndrome decode
1266 information or because there is no device mapped at the accessed IPA, then
1267 userspace can ask the kernel to inject an external abort using the address
1268 from the exiting fault on the VCPU. It is a programming error to set
1269 ext_dabt_pending after an exit which was not either KVM_EXIT_MMIO or
1270 KVM_EXIT_ARM_NISV. This feature is only available if the system supports
1271 KVM_CAP_ARM_INJECT_EXT_DABT. This is a helper which provides commonality in
1272 how userspace reports accesses for the above cases to guests, across different
1273 userspace implementations. Nevertheless, userspace can still emulate all Arm
1274 exceptions by manipulating individual registers using the KVM_SET_ONE_REG API.
1275
1276 See KVM_GET_VCPU_EVENTS for the data structure.
1277
1278
1279 4.33 KVM_GET_DEBUGREGS
1280 ----------------------
1281
1282 :Capability: KVM_CAP_DEBUGREGS
1283 :Architectures: x86
1284 :Type: vm ioctl
1285 :Parameters: struct kvm_debugregs (out)
1286 :Returns: 0 on success, -1 on error
1287
1288 Reads debug registers from the vcpu.
1289
1290 ::
1291
1292 struct kvm_debugregs {
1293 __u64 db[4];
1294 __u64 dr6;
1295 __u64 dr7;
1296 __u64 flags;
1297 __u64 reserved[9];
1298 };
1299
1300
1301 4.34 KVM_SET_DEBUGREGS
1302 ----------------------
1303
1304 :Capability: KVM_CAP_DEBUGREGS
1305 :Architectures: x86
1306 :Type: vm ioctl
1307 :Parameters: struct kvm_debugregs (in)
1308 :Returns: 0 on success, -1 on error
1309
1310 Writes debug registers into the vcpu.
1311
1312 See KVM_GET_DEBUGREGS for the data structure. The flags field is unused
1313 yet and must be cleared on entry.
1314
1315
1316 4.35 KVM_SET_USER_MEMORY_REGION
1317 -------------------------------
1318
1319 :Capability: KVM_CAP_USER_MEMORY
1320 :Architectures: all
1321 :Type: vm ioctl
1322 :Parameters: struct kvm_userspace_memory_region (in)
1323 :Returns: 0 on success, -1 on error
1324
1325 ::
1326
1327 struct kvm_userspace_memory_region {
1328 __u32 slot;
1329 __u32 flags;
1330 __u64 guest_phys_addr;
1331 __u64 memory_size; /* bytes */
1332 __u64 userspace_addr; /* start of the userspace allocated memory */
1333 };
1334
1335 /* for kvm_memory_region::flags */
1336 #define KVM_MEM_LOG_DIRTY_PAGES (1UL << 0)
1337 #define KVM_MEM_READONLY (1UL << 1)
1338
1339 This ioctl allows the user to create, modify or delete a guest physical
1340 memory slot. Bits 0-15 of "slot" specify the slot id and this value
1341 should be less than the maximum number of user memory slots supported per
1342 VM. The maximum allowed slots can be queried using KVM_CAP_NR_MEMSLOTS.
1343 Slots may not overlap in guest physical address space.
1344
1345 If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of "slot"
1346 specifies the address space which is being modified. They must be
1347 less than the value that KVM_CHECK_EXTENSION returns for the
1348 KVM_CAP_MULTI_ADDRESS_SPACE capability. Slots in separate address spaces
1349 are unrelated; the restriction on overlapping slots only applies within
1350 each address space.
1351
1352 Deleting a slot is done by passing zero for memory_size. When changing
1353 an existing slot, it may be moved in the guest physical memory space,
1354 or its flags may be modified, but it may not be resized.
1355
1356 Memory for the region is taken starting at the address denoted by the
1357 field userspace_addr, which must point at user addressable memory for
1358 the entire memory slot size. Any object may back this memory, including
1359 anonymous memory, ordinary files, and hugetlbfs.
1360
1361 On architectures that support a form of address tagging, userspace_addr must
1362 be an untagged address.
1363
1364 It is recommended that the lower 21 bits of guest_phys_addr and userspace_addr
1365 be identical. This allows large pages in the guest to be backed by large
1366 pages in the host.
1367
1368 The flags field supports two flags: KVM_MEM_LOG_DIRTY_PAGES and
1369 KVM_MEM_READONLY. The former can be set to instruct KVM to keep track of
1370 writes to memory within the slot. See KVM_GET_DIRTY_LOG ioctl to know how to
1371 use it. The latter can be set, if KVM_CAP_READONLY_MEM capability allows it,
1372 to make a new slot read-only. In this case, writes to this memory will be
1373 posted to userspace as KVM_EXIT_MMIO exits.
1374
1375 When the KVM_CAP_SYNC_MMU capability is available, changes in the backing of
1376 the memory region are automatically reflected into the guest. For example, an
1377 mmap() that affects the region will be made visible immediately. Another
1378 example is madvise(MADV_DROP).
1379
1380 It is recommended to use this API instead of the KVM_SET_MEMORY_REGION ioctl.
1381 The KVM_SET_MEMORY_REGION does not allow fine grained control over memory
1382 allocation and is deprecated.
1383
1384
1385 4.36 KVM_SET_TSS_ADDR
1386 ---------------------
1387
1388 :Capability: KVM_CAP_SET_TSS_ADDR
1389 :Architectures: x86
1390 :Type: vm ioctl
1391 :Parameters: unsigned long tss_address (in)
1392 :Returns: 0 on success, -1 on error
1393
1394 This ioctl defines the physical address of a three-page region in the guest
1395 physical address space. The region must be within the first 4GB of the
1396 guest physical address space and must not conflict with any memory slot
1397 or any mmio address. The guest may malfunction if it accesses this memory
1398 region.
1399
1400 This ioctl is required on Intel-based hosts. This is needed on Intel hardware
1401 because of a quirk in the virtualization implementation (see the internals
1402 documentation when it pops into existence).
1403
1404
1405 4.37 KVM_ENABLE_CAP
1406 -------------------
1407
1408 :Capability: KVM_CAP_ENABLE_CAP
1409 :Architectures: mips, ppc, s390, x86
1410 :Type: vcpu ioctl
1411 :Parameters: struct kvm_enable_cap (in)
1412 :Returns: 0 on success; -1 on error
1413
1414 :Capability: KVM_CAP_ENABLE_CAP_VM
1415 :Architectures: all
1416 :Type: vm ioctl
1417 :Parameters: struct kvm_enable_cap (in)
1418 :Returns: 0 on success; -1 on error
1419
1420 .. note::
1421
1422 Not all extensions are enabled by default. Using this ioctl the application
1423 can enable an extension, making it available to the guest.
1424
1425 On systems that do not support this ioctl, it always fails. On systems that
1426 do support it, it only works for extensions that are supported for enablement.
1427
1428 To check if a capability can be enabled, the KVM_CHECK_EXTENSION ioctl should
1429 be used.
1430
1431 ::
1432
1433 struct kvm_enable_cap {
1434 /* in */
1435 __u32 cap;
1436
1437 The capability that is supposed to get enabled.
1438
1439 ::
1440
1441 __u32 flags;
1442
1443 A bitfield indicating future enhancements. Has to be 0 for now.
1444
1445 ::
1446
1447 __u64 args[4];
1448
1449 Arguments for enabling a feature. If a feature needs initial values to
1450 function properly, this is the place to put them.
1451
1452 ::
1453
1454 __u8 pad[64];
1455 };
1456
1457 The vcpu ioctl should be used for vcpu-specific capabilities, the vm ioctl
1458 for vm-wide capabilities.
1459
1460 4.38 KVM_GET_MP_STATE
1461 ---------------------
1462
1463 :Capability: KVM_CAP_MP_STATE
1464 :Architectures: x86, s390, arm64, riscv
1465 :Type: vcpu ioctl
1466 :Parameters: struct kvm_mp_state (out)
1467 :Returns: 0 on success; -1 on error
1468
1469 ::
1470
1471 struct kvm_mp_state {
1472 __u32 mp_state;
1473 };
1474
1475 Returns the vcpu's current "multiprocessing state" (though also valid on
1476 uniprocessor guests).
1477
1478 Possible values are:
1479
1480 ========================== ===============================================
1481 KVM_MP_STATE_RUNNABLE the vcpu is currently running
1482 [x86,arm64,riscv]
1483 KVM_MP_STATE_UNINITIALIZED the vcpu is an application processor (AP)
1484 which has not yet received an INIT signal [x86]
1485 KVM_MP_STATE_INIT_RECEIVED the vcpu has received an INIT signal, and is
1486 now ready for a SIPI [x86]
1487 KVM_MP_STATE_HALTED the vcpu has executed a HLT instruction and
1488 is waiting for an interrupt [x86]
1489 KVM_MP_STATE_SIPI_RECEIVED the vcpu has just received a SIPI (vector
1490 accessible via KVM_GET_VCPU_EVENTS) [x86]
1491 KVM_MP_STATE_STOPPED the vcpu is stopped [s390,arm64,riscv]
1492 KVM_MP_STATE_CHECK_STOP the vcpu is in a special error state [s390]
1493 KVM_MP_STATE_OPERATING the vcpu is operating (running or halted)
1494 [s390]
1495 KVM_MP_STATE_LOAD the vcpu is in a special load/startup state
1496 [s390]
1497 KVM_MP_STATE_SUSPENDED the vcpu is in a suspend state and is waiting
1498 for a wakeup event [arm64]
1499 ========================== ===============================================
1500
1501 On x86, this ioctl is only useful after KVM_CREATE_IRQCHIP. Without an
1502 in-kernel irqchip, the multiprocessing state must be maintained by userspace on
1503 these architectures.
1504
1505 For arm64:
1506 ^^^^^^^^^^
1507
1508 If a vCPU is in the KVM_MP_STATE_SUSPENDED state, KVM will emulate the
1509 architectural execution of a WFI instruction.
1510
1511 If a wakeup event is recognized, KVM will exit to userspace with a
1512 KVM_SYSTEM_EVENT exit, where the event type is KVM_SYSTEM_EVENT_WAKEUP. If
1513 userspace wants to honor the wakeup, it must set the vCPU's MP state to
1514 KVM_MP_STATE_RUNNABLE. If it does not, KVM will continue to await a wakeup
1515 event in subsequent calls to KVM_RUN.
1516
1517 .. warning::
1518
1519 If userspace intends to keep the vCPU in a SUSPENDED state, it is
1520 strongly recommended that userspace take action to suppress the
1521 wakeup event (such as masking an interrupt). Otherwise, subsequent
1522 calls to KVM_RUN will immediately exit with a KVM_SYSTEM_EVENT_WAKEUP
1523 event and inadvertently waste CPU cycles.
1524
1525 Additionally, if userspace takes action to suppress a wakeup event,
1526 it is strongly recommended that it also restores the vCPU to its
1527 original state when the vCPU is made RUNNABLE again. For example,
1528 if userspace masked a pending interrupt to suppress the wakeup,
1529 the interrupt should be unmasked before returning control to the
1530 guest.
1531
1532 For riscv:
1533 ^^^^^^^^^^
1534
1535 The only states that are valid are KVM_MP_STATE_STOPPED and
1536 KVM_MP_STATE_RUNNABLE which reflect if the vcpu is paused or not.
1537
1538 4.39 KVM_SET_MP_STATE
1539 ---------------------
1540
1541 :Capability: KVM_CAP_MP_STATE
1542 :Architectures: x86, s390, arm64, riscv
1543 :Type: vcpu ioctl
1544 :Parameters: struct kvm_mp_state (in)
1545 :Returns: 0 on success; -1 on error
1546
1547 Sets the vcpu's current "multiprocessing state"; see KVM_GET_MP_STATE for
1548 arguments.
1549
1550 On x86, this ioctl is only useful after KVM_CREATE_IRQCHIP. Without an
1551 in-kernel irqchip, the multiprocessing state must be maintained by userspace on
1552 these architectures.
1553
1554 For arm64/riscv:
1555 ^^^^^^^^^^^^^^^^
1556
1557 The only states that are valid are KVM_MP_STATE_STOPPED and
1558 KVM_MP_STATE_RUNNABLE which reflect if the vcpu should be paused or not.
1559
1560 4.40 KVM_SET_IDENTITY_MAP_ADDR
1561 ------------------------------
1562
1563 :Capability: KVM_CAP_SET_IDENTITY_MAP_ADDR
1564 :Architectures: x86
1565 :Type: vm ioctl
1566 :Parameters: unsigned long identity (in)
1567 :Returns: 0 on success, -1 on error
1568
1569 This ioctl defines the physical address of a one-page region in the guest
1570 physical address space. The region must be within the first 4GB of the
1571 guest physical address space and must not conflict with any memory slot
1572 or any mmio address. The guest may malfunction if it accesses this memory
1573 region.
1574
1575 Setting the address to 0 will result in resetting the address to its default
1576 (0xfffbc000).
1577
1578 This ioctl is required on Intel-based hosts. This is needed on Intel hardware
1579 because of a quirk in the virtualization implementation (see the internals
1580 documentation when it pops into existence).
1581
1582 Fails if any VCPU has already been created.
1583
1584 4.41 KVM_SET_BOOT_CPU_ID
1585 ------------------------
1586
1587 :Capability: KVM_CAP_SET_BOOT_CPU_ID
1588 :Architectures: x86
1589 :Type: vm ioctl
1590 :Parameters: unsigned long vcpu_id
1591 :Returns: 0 on success, -1 on error
1592
1593 Define which vcpu is the Bootstrap Processor (BSP). Values are the same
1594 as the vcpu id in KVM_CREATE_VCPU. If this ioctl is not called, the default
1595 is vcpu 0. This ioctl has to be called before vcpu creation,
1596 otherwise it will return EBUSY error.
1597
1598
1599 4.42 KVM_GET_XSAVE
1600 ------------------
1601
1602 :Capability: KVM_CAP_XSAVE
1603 :Architectures: x86
1604 :Type: vcpu ioctl
1605 :Parameters: struct kvm_xsave (out)
1606 :Returns: 0 on success, -1 on error
1607
1608
1609 ::
1610
1611 struct kvm_xsave {
1612 __u32 region[1024];
1613 __u32 extra[0];
1614 };
1615
1616 This ioctl would copy current vcpu's xsave struct to the userspace.
1617
1618
1619 4.43 KVM_SET_XSAVE
1620 ------------------
1621
1622 :Capability: KVM_CAP_XSAVE and KVM_CAP_XSAVE2
1623 :Architectures: x86
1624 :Type: vcpu ioctl
1625 :Parameters: struct kvm_xsave (in)
1626 :Returns: 0 on success, -1 on error
1627
1628 ::
1629
1630
1631 struct kvm_xsave {
1632 __u32 region[1024];
1633 __u32 extra[0];
1634 };
1635
1636 This ioctl would copy userspace's xsave struct to the kernel. It copies
1637 as many bytes as are returned by KVM_CHECK_EXTENSION(KVM_CAP_XSAVE2),
1638 when invoked on the vm file descriptor. The size value returned by
1639 KVM_CHECK_EXTENSION(KVM_CAP_XSAVE2) will always be at least 4096.
1640 Currently, it is only greater than 4096 if a dynamic feature has been
1641 enabled with ``arch_prctl()``, but this may change in the future.
1642
1643 The offsets of the state save areas in struct kvm_xsave follow the
1644 contents of CPUID leaf 0xD on the host.
1645
1646
1647 4.44 KVM_GET_XCRS
1648 -----------------
1649
1650 :Capability: KVM_CAP_XCRS
1651 :Architectures: x86
1652 :Type: vcpu ioctl
1653 :Parameters: struct kvm_xcrs (out)
1654 :Returns: 0 on success, -1 on error
1655
1656 ::
1657
1658 struct kvm_xcr {
1659 __u32 xcr;
1660 __u32 reserved;
1661 __u64 value;
1662 };
1663
1664 struct kvm_xcrs {
1665 __u32 nr_xcrs;
1666 __u32 flags;
1667 struct kvm_xcr xcrs[KVM_MAX_XCRS];
1668 __u64 padding[16];
1669 };
1670
1671 This ioctl would copy current vcpu's xcrs to the userspace.
1672
1673
1674 4.45 KVM_SET_XCRS
1675 -----------------
1676
1677 :Capability: KVM_CAP_XCRS
1678 :Architectures: x86
1679 :Type: vcpu ioctl
1680 :Parameters: struct kvm_xcrs (in)
1681 :Returns: 0 on success, -1 on error
1682
1683 ::
1684
1685 struct kvm_xcr {
1686 __u32 xcr;
1687 __u32 reserved;
1688 __u64 value;
1689 };
1690
1691 struct kvm_xcrs {
1692 __u32 nr_xcrs;
1693 __u32 flags;
1694 struct kvm_xcr xcrs[KVM_MAX_XCRS];
1695 __u64 padding[16];
1696 };
1697
1698 This ioctl would set vcpu's xcr to the value userspace specified.
1699
1700
1701 4.46 KVM_GET_SUPPORTED_CPUID
1702 ----------------------------
1703
1704 :Capability: KVM_CAP_EXT_CPUID
1705 :Architectures: x86
1706 :Type: system ioctl
1707 :Parameters: struct kvm_cpuid2 (in/out)
1708 :Returns: 0 on success, -1 on error
1709
1710 ::
1711
1712 struct kvm_cpuid2 {
1713 __u32 nent;
1714 __u32 padding;
1715 struct kvm_cpuid_entry2 entries[0];
1716 };
1717
1718 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX BIT(0)
1719 #define KVM_CPUID_FLAG_STATEFUL_FUNC BIT(1) /* deprecated */
1720 #define KVM_CPUID_FLAG_STATE_READ_NEXT BIT(2) /* deprecated */
1721
1722 struct kvm_cpuid_entry2 {
1723 __u32 function;
1724 __u32 index;
1725 __u32 flags;
1726 __u32 eax;
1727 __u32 ebx;
1728 __u32 ecx;
1729 __u32 edx;
1730 __u32 padding[3];
1731 };
1732
1733 This ioctl returns x86 cpuid features which are supported by both the
1734 hardware and kvm in its default configuration. Userspace can use the
1735 information returned by this ioctl to construct cpuid information (for
1736 KVM_SET_CPUID2) that is consistent with hardware, kernel, and
1737 userspace capabilities, and with user requirements (for example, the
1738 user may wish to constrain cpuid to emulate older hardware, or for
1739 feature consistency across a cluster).
1740
1741 Dynamically-enabled feature bits need to be requested with
1742 ``arch_prctl()`` before calling this ioctl. Feature bits that have not
1743 been requested are excluded from the result.
1744
1745 Note that certain capabilities, such as KVM_CAP_X86_DISABLE_EXITS, may
1746 expose cpuid features (e.g. MONITOR) which are not supported by kvm in
1747 its default configuration. If userspace enables such capabilities, it
1748 is responsible for modifying the results of this ioctl appropriately.
1749
1750 Userspace invokes KVM_GET_SUPPORTED_CPUID by passing a kvm_cpuid2 structure
1751 with the 'nent' field indicating the number of entries in the variable-size
1752 array 'entries'. If the number of entries is too low to describe the cpu
1753 capabilities, an error (E2BIG) is returned. If the number is too high,
1754 the 'nent' field is adjusted and an error (ENOMEM) is returned. If the
1755 number is just right, the 'nent' field is adjusted to the number of valid
1756 entries in the 'entries' array, which is then filled.
1757
1758 The entries returned are the host cpuid as returned by the cpuid instruction,
1759 with unknown or unsupported features masked out. Some features (for example,
1760 x2apic), may not be present in the host cpu, but are exposed by kvm if it can
1761 emulate them efficiently. The fields in each entry are defined as follows:
1762
1763 function:
1764 the eax value used to obtain the entry
1765
1766 index:
1767 the ecx value used to obtain the entry (for entries that are
1768 affected by ecx)
1769
1770 flags:
1771 an OR of zero or more of the following:
1772
1773 KVM_CPUID_FLAG_SIGNIFCANT_INDEX:
1774 if the index field is valid
1775
1776 eax, ebx, ecx, edx:
1777 the values returned by the cpuid instruction for
1778 this function/index combination
1779
1780 The TSC deadline timer feature (CPUID leaf 1, ecx[24]) is always returned
1781 as false, since the feature depends on KVM_CREATE_IRQCHIP for local APIC
1782 support. Instead it is reported via::
1783
1784 ioctl(KVM_CHECK_EXTENSION, KVM_CAP_TSC_DEADLINE_TIMER)
1785
1786 if that returns true and you use KVM_CREATE_IRQCHIP, or if you emulate the
1787 feature in userspace, then you can enable the feature for KVM_SET_CPUID2.
1788
1789
1790 4.47 KVM_PPC_GET_PVINFO
1791 -----------------------
1792
1793 :Capability: KVM_CAP_PPC_GET_PVINFO
1794 :Architectures: ppc
1795 :Type: vm ioctl
1796 :Parameters: struct kvm_ppc_pvinfo (out)
1797 :Returns: 0 on success, !0 on error
1798
1799 ::
1800
1801 struct kvm_ppc_pvinfo {
1802 __u32 flags;
1803 __u32 hcall[4];
1804 __u8 pad[108];
1805 };
1806
1807 This ioctl fetches PV specific information that need to be passed to the guest
1808 using the device tree or other means from vm context.
1809
1810 The hcall array defines 4 instructions that make up a hypercall.
1811
1812 If any additional field gets added to this structure later on, a bit for that
1813 additional piece of information will be set in the flags bitmap.
1814
1815 The flags bitmap is defined as::
1816
1817 /* the host supports the ePAPR idle hcall
1818 #define KVM_PPC_PVINFO_FLAGS_EV_IDLE (1<<0)
1819
1820 4.52 KVM_SET_GSI_ROUTING
1821 ------------------------
1822
1823 :Capability: KVM_CAP_IRQ_ROUTING
1824 :Architectures: x86 s390 arm64
1825 :Type: vm ioctl
1826 :Parameters: struct kvm_irq_routing (in)
1827 :Returns: 0 on success, -1 on error
1828
1829 Sets the GSI routing table entries, overwriting any previously set entries.
1830
1831 On arm64, GSI routing has the following limitation:
1832
1833 - GSI routing does not apply to KVM_IRQ_LINE but only to KVM_IRQFD.
1834
1835 ::
1836
1837 struct kvm_irq_routing {
1838 __u32 nr;
1839 __u32 flags;
1840 struct kvm_irq_routing_entry entries[0];
1841 };
1842
1843 No flags are specified so far, the corresponding field must be set to zero.
1844
1845 ::
1846
1847 struct kvm_irq_routing_entry {
1848 __u32 gsi;
1849 __u32 type;
1850 __u32 flags;
1851 __u32 pad;
1852 union {
1853 struct kvm_irq_routing_irqchip irqchip;
1854 struct kvm_irq_routing_msi msi;
1855 struct kvm_irq_routing_s390_adapter adapter;
1856 struct kvm_irq_routing_hv_sint hv_sint;
1857 struct kvm_irq_routing_xen_evtchn xen_evtchn;
1858 __u32 pad[8];
1859 } u;
1860 };
1861
1862 /* gsi routing entry types */
1863 #define KVM_IRQ_ROUTING_IRQCHIP 1
1864 #define KVM_IRQ_ROUTING_MSI 2
1865 #define KVM_IRQ_ROUTING_S390_ADAPTER 3
1866 #define KVM_IRQ_ROUTING_HV_SINT 4
1867 #define KVM_IRQ_ROUTING_XEN_EVTCHN 5
1868
1869 flags:
1870
1871 - KVM_MSI_VALID_DEVID: used along with KVM_IRQ_ROUTING_MSI routing entry
1872 type, specifies that the devid field contains a valid value. The per-VM
1873 KVM_CAP_MSI_DEVID capability advertises the requirement to provide
1874 the device ID. If this capability is not available, userspace should
1875 never set the KVM_MSI_VALID_DEVID flag as the ioctl might fail.
1876 - zero otherwise
1877
1878 ::
1879
1880 struct kvm_irq_routing_irqchip {
1881 __u32 irqchip;
1882 __u32 pin;
1883 };
1884
1885 struct kvm_irq_routing_msi {
1886 __u32 address_lo;
1887 __u32 address_hi;
1888 __u32 data;
1889 union {
1890 __u32 pad;
1891 __u32 devid;
1892 };
1893 };
1894
1895 If KVM_MSI_VALID_DEVID is set, devid contains a unique device identifier
1896 for the device that wrote the MSI message. For PCI, this is usually a
1897 BFD identifier in the lower 16 bits.
1898
1899 On x86, address_hi is ignored unless the KVM_X2APIC_API_USE_32BIT_IDS
1900 feature of KVM_CAP_X2APIC_API capability is enabled. If it is enabled,
1901 address_hi bits 31-8 provide bits 31-8 of the destination id. Bits 7-0 of
1902 address_hi must be zero.
1903
1904 ::
1905
1906 struct kvm_irq_routing_s390_adapter {
1907 __u64 ind_addr;
1908 __u64 summary_addr;
1909 __u64 ind_offset;
1910 __u32 summary_offset;
1911 __u32 adapter_id;
1912 };
1913
1914 struct kvm_irq_routing_hv_sint {
1915 __u32 vcpu;
1916 __u32 sint;
1917 };
1918
1919 struct kvm_irq_routing_xen_evtchn {
1920 __u32 port;
1921 __u32 vcpu;
1922 __u32 priority;
1923 };
1924
1925
1926 When KVM_CAP_XEN_HVM includes the KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL bit
1927 in its indication of supported features, routing to Xen event channels
1928 is supported. Although the priority field is present, only the value
1929 KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL is supported, which means delivery by
1930 2 level event channels. FIFO event channel support may be added in
1931 the future.
1932
1933
1934 4.55 KVM_SET_TSC_KHZ
1935 --------------------
1936
1937 :Capability: KVM_CAP_TSC_CONTROL / KVM_CAP_VM_TSC_CONTROL
1938 :Architectures: x86
1939 :Type: vcpu ioctl / vm ioctl
1940 :Parameters: virtual tsc_khz
1941 :Returns: 0 on success, -1 on error
1942
1943 Specifies the tsc frequency for the virtual machine. The unit of the
1944 frequency is KHz.
1945
1946 If the KVM_CAP_VM_TSC_CONTROL capability is advertised, this can also
1947 be used as a vm ioctl to set the initial tsc frequency of subsequently
1948 created vCPUs.
1949
1950 4.56 KVM_GET_TSC_KHZ
1951 --------------------
1952
1953 :Capability: KVM_CAP_GET_TSC_KHZ / KVM_CAP_VM_TSC_CONTROL
1954 :Architectures: x86
1955 :Type: vcpu ioctl / vm ioctl
1956 :Parameters: none
1957 :Returns: virtual tsc-khz on success, negative value on error
1958
1959 Returns the tsc frequency of the guest. The unit of the return value is
1960 KHz. If the host has unstable tsc this ioctl returns -EIO instead as an
1961 error.
1962
1963
1964 4.57 KVM_GET_LAPIC
1965 ------------------
1966
1967 :Capability: KVM_CAP_IRQCHIP
1968 :Architectures: x86
1969 :Type: vcpu ioctl
1970 :Parameters: struct kvm_lapic_state (out)
1971 :Returns: 0 on success, -1 on error
1972
1973 ::
1974
1975 #define KVM_APIC_REG_SIZE 0x400
1976 struct kvm_lapic_state {
1977 char regs[KVM_APIC_REG_SIZE];
1978 };
1979
1980 Reads the Local APIC registers and copies them into the input argument. The
1981 data format and layout are the same as documented in the architecture manual.
1982
1983 If KVM_X2APIC_API_USE_32BIT_IDS feature of KVM_CAP_X2APIC_API is
1984 enabled, then the format of APIC_ID register depends on the APIC mode
1985 (reported by MSR_IA32_APICBASE) of its VCPU. x2APIC stores APIC ID in
1986 the APIC_ID register (bytes 32-35). xAPIC only allows an 8-bit APIC ID
1987 which is stored in bits 31-24 of the APIC register, or equivalently in
1988 byte 35 of struct kvm_lapic_state's regs field. KVM_GET_LAPIC must then
1989 be called after MSR_IA32_APICBASE has been set with KVM_SET_MSR.
1990
1991 If KVM_X2APIC_API_USE_32BIT_IDS feature is disabled, struct kvm_lapic_state
1992 always uses xAPIC format.
1993
1994
1995 4.58 KVM_SET_LAPIC
1996 ------------------
1997
1998 :Capability: KVM_CAP_IRQCHIP
1999 :Architectures: x86
2000 :Type: vcpu ioctl
2001 :Parameters: struct kvm_lapic_state (in)
2002 :Returns: 0 on success, -1 on error
2003
2004 ::
2005
2006 #define KVM_APIC_REG_SIZE 0x400
2007 struct kvm_lapic_state {
2008 char regs[KVM_APIC_REG_SIZE];
2009 };
2010
2011 Copies the input argument into the Local APIC registers. The data format
2012 and layout are the same as documented in the architecture manual.
2013
2014 The format of the APIC ID register (bytes 32-35 of struct kvm_lapic_state's
2015 regs field) depends on the state of the KVM_CAP_X2APIC_API capability.
2016 See the note in KVM_GET_LAPIC.
2017
2018
2019 4.59 KVM_IOEVENTFD
2020 ------------------
2021
2022 :Capability: KVM_CAP_IOEVENTFD
2023 :Architectures: all
2024 :Type: vm ioctl
2025 :Parameters: struct kvm_ioeventfd (in)
2026 :Returns: 0 on success, !0 on error
2027
2028 This ioctl attaches or detaches an ioeventfd to a legal pio/mmio address
2029 within the guest. A guest write in the registered address will signal the
2030 provided event instead of triggering an exit.
2031
2032 ::
2033
2034 struct kvm_ioeventfd {
2035 __u64 datamatch;
2036 __u64 addr; /* legal pio/mmio address */
2037 __u32 len; /* 0, 1, 2, 4, or 8 bytes */
2038 __s32 fd;
2039 __u32 flags;
2040 __u8 pad[36];
2041 };
2042
2043 For the special case of virtio-ccw devices on s390, the ioevent is matched
2044 to a subchannel/virtqueue tuple instead.
2045
2046 The following flags are defined::
2047
2048 #define KVM_IOEVENTFD_FLAG_DATAMATCH (1 << kvm_ioeventfd_flag_nr_datamatch)
2049 #define KVM_IOEVENTFD_FLAG_PIO (1 << kvm_ioeventfd_flag_nr_pio)
2050 #define KVM_IOEVENTFD_FLAG_DEASSIGN (1 << kvm_ioeventfd_flag_nr_deassign)
2051 #define KVM_IOEVENTFD_FLAG_VIRTIO_CCW_NOTIFY \
2052 (1 << kvm_ioeventfd_flag_nr_virtio_ccw_notify)
2053
2054 If datamatch flag is set, the event will be signaled only if the written value
2055 to the registered address is equal to datamatch in struct kvm_ioeventfd.
2056
2057 For virtio-ccw devices, addr contains the subchannel id and datamatch the
2058 virtqueue index.
2059
2060 With KVM_CAP_IOEVENTFD_ANY_LENGTH, a zero length ioeventfd is allowed, and
2061 the kernel will ignore the length of guest write and may get a faster vmexit.
2062 The speedup may only apply to specific architectures, but the ioeventfd will
2063 work anyway.
2064
2065 4.60 KVM_DIRTY_TLB
2066 ------------------
2067
2068 :Capability: KVM_CAP_SW_TLB
2069 :Architectures: ppc
2070 :Type: vcpu ioctl
2071 :Parameters: struct kvm_dirty_tlb (in)
2072 :Returns: 0 on success, -1 on error
2073
2074 ::
2075
2076 struct kvm_dirty_tlb {
2077 __u64 bitmap;
2078 __u32 num_dirty;
2079 };
2080
2081 This must be called whenever userspace has changed an entry in the shared
2082 TLB, prior to calling KVM_RUN on the associated vcpu.
2083
2084 The "bitmap" field is the userspace address of an array. This array
2085 consists of a number of bits, equal to the total number of TLB entries as
2086 determined by the last successful call to KVM_CONFIG_TLB, rounded up to the
2087 nearest multiple of 64.
2088
2089 Each bit corresponds to one TLB entry, ordered the same as in the shared TLB
2090 array.
2091
2092 The array is little-endian: the bit 0 is the least significant bit of the
2093 first byte, bit 8 is the least significant bit of the second byte, etc.
2094 This avoids any complications with differing word sizes.
2095
2096 The "num_dirty" field is a performance hint for KVM to determine whether it
2097 should skip processing the bitmap and just invalidate everything. It must
2098 be set to the number of set bits in the bitmap.
2099
2100
2101 4.62 KVM_CREATE_SPAPR_TCE
2102 -------------------------
2103
2104 :Capability: KVM_CAP_SPAPR_TCE
2105 :Architectures: powerpc
2106 :Type: vm ioctl
2107 :Parameters: struct kvm_create_spapr_tce (in)
2108 :Returns: file descriptor for manipulating the created TCE table
2109
2110 This creates a virtual TCE (translation control entry) table, which
2111 is an IOMMU for PAPR-style virtual I/O. It is used to translate
2112 logical addresses used in virtual I/O into guest physical addresses,
2113 and provides a scatter/gather capability for PAPR virtual I/O.
2114
2115 ::
2116
2117 /* for KVM_CAP_SPAPR_TCE */
2118 struct kvm_create_spapr_tce {
2119 __u64 liobn;
2120 __u32 window_size;
2121 };
2122
2123 The liobn field gives the logical IO bus number for which to create a
2124 TCE table. The window_size field specifies the size of the DMA window
2125 which this TCE table will translate - the table will contain one 64
2126 bit TCE entry for every 4kiB of the DMA window.
2127
2128 When the guest issues an H_PUT_TCE hcall on a liobn for which a TCE
2129 table has been created using this ioctl(), the kernel will handle it
2130 in real mode, updating the TCE table. H_PUT_TCE calls for other
2131 liobns will cause a vm exit and must be handled by userspace.
2132
2133 The return value is a file descriptor which can be passed to mmap(2)
2134 to map the created TCE table into userspace. This lets userspace read
2135 the entries written by kernel-handled H_PUT_TCE calls, and also lets
2136 userspace update the TCE table directly which is useful in some
2137 circumstances.
2138
2139
2140 4.63 KVM_ALLOCATE_RMA
2141 ---------------------
2142
2143 :Capability: KVM_CAP_PPC_RMA
2144 :Architectures: powerpc
2145 :Type: vm ioctl
2146 :Parameters: struct kvm_allocate_rma (out)
2147 :Returns: file descriptor for mapping the allocated RMA
2148
2149 This allocates a Real Mode Area (RMA) from the pool allocated at boot
2150 time by the kernel. An RMA is a physically-contiguous, aligned region
2151 of memory used on older POWER processors to provide the memory which
2152 will be accessed by real-mode (MMU off) accesses in a KVM guest.
2153 POWER processors support a set of sizes for the RMA that usually
2154 includes 64MB, 128MB, 256MB and some larger powers of two.
2155
2156 ::
2157
2158 /* for KVM_ALLOCATE_RMA */
2159 struct kvm_allocate_rma {
2160 __u64 rma_size;
2161 };
2162
2163 The return value is a file descriptor which can be passed to mmap(2)
2164 to map the allocated RMA into userspace. The mapped area can then be
2165 passed to the KVM_SET_USER_MEMORY_REGION ioctl to establish it as the
2166 RMA for a virtual machine. The size of the RMA in bytes (which is
2167 fixed at host kernel boot time) is returned in the rma_size field of
2168 the argument structure.
2169
2170 The KVM_CAP_PPC_RMA capability is 1 or 2 if the KVM_ALLOCATE_RMA ioctl
2171 is supported; 2 if the processor requires all virtual machines to have
2172 an RMA, or 1 if the processor can use an RMA but doesn't require it,
2173 because it supports the Virtual RMA (VRMA) facility.
2174
2175
2176 4.64 KVM_NMI
2177 ------------
2178
2179 :Capability: KVM_CAP_USER_NMI
2180 :Architectures: x86
2181 :Type: vcpu ioctl
2182 :Parameters: none
2183 :Returns: 0 on success, -1 on error
2184
2185 Queues an NMI on the thread's vcpu. Note this is well defined only
2186 when KVM_CREATE_IRQCHIP has not been called, since this is an interface
2187 between the virtual cpu core and virtual local APIC. After KVM_CREATE_IRQCHIP
2188 has been called, this interface is completely emulated within the kernel.
2189
2190 To use this to emulate the LINT1 input with KVM_CREATE_IRQCHIP, use the
2191 following algorithm:
2192
2193 - pause the vcpu
2194 - read the local APIC's state (KVM_GET_LAPIC)
2195 - check whether changing LINT1 will queue an NMI (see the LVT entry for LINT1)
2196 - if so, issue KVM_NMI
2197 - resume the vcpu
2198
2199 Some guests configure the LINT1 NMI input to cause a panic, aiding in
2200 debugging.
2201
2202
2203 4.65 KVM_S390_UCAS_MAP
2204 ----------------------
2205
2206 :Capability: KVM_CAP_S390_UCONTROL
2207 :Architectures: s390
2208 :Type: vcpu ioctl
2209 :Parameters: struct kvm_s390_ucas_mapping (in)
2210 :Returns: 0 in case of success
2211
2212 The parameter is defined like this::
2213
2214 struct kvm_s390_ucas_mapping {
2215 __u64 user_addr;
2216 __u64 vcpu_addr;
2217 __u64 length;
2218 };
2219
2220 This ioctl maps the memory at "user_addr" with the length "length" to
2221 the vcpu's address space starting at "vcpu_addr". All parameters need to
2222 be aligned by 1 megabyte.
2223
2224
2225 4.66 KVM_S390_UCAS_UNMAP
2226 ------------------------
2227
2228 :Capability: KVM_CAP_S390_UCONTROL
2229 :Architectures: s390
2230 :Type: vcpu ioctl
2231 :Parameters: struct kvm_s390_ucas_mapping (in)
2232 :Returns: 0 in case of success
2233
2234 The parameter is defined like this::
2235
2236 struct kvm_s390_ucas_mapping {
2237 __u64 user_addr;
2238 __u64 vcpu_addr;
2239 __u64 length;
2240 };
2241
2242 This ioctl unmaps the memory in the vcpu's address space starting at
2243 "vcpu_addr" with the length "length". The field "user_addr" is ignored.
2244 All parameters need to be aligned by 1 megabyte.
2245
2246
2247 4.67 KVM_S390_VCPU_FAULT
2248 ------------------------
2249
2250 :Capability: KVM_CAP_S390_UCONTROL
2251 :Architectures: s390
2252 :Type: vcpu ioctl
2253 :Parameters: vcpu absolute address (in)
2254 :Returns: 0 in case of success
2255
2256 This call creates a page table entry on the virtual cpu's address space
2257 (for user controlled virtual machines) or the virtual machine's address
2258 space (for regular virtual machines). This only works for minor faults,
2259 thus it's recommended to access subject memory page via the user page
2260 table upfront. This is useful to handle validity intercepts for user
2261 controlled virtual machines to fault in the virtual cpu's lowcore pages
2262 prior to calling the KVM_RUN ioctl.
2263
2264
2265 4.68 KVM_SET_ONE_REG
2266 --------------------
2267
2268 :Capability: KVM_CAP_ONE_REG
2269 :Architectures: all
2270 :Type: vcpu ioctl
2271 :Parameters: struct kvm_one_reg (in)
2272 :Returns: 0 on success, negative value on failure
2273
2274 Errors:
2275
2276 ====== ============================================================
2277 ENOENT no such register
2278 EINVAL invalid register ID, or no such register or used with VMs in
2279 protected virtualization mode on s390
2280 EPERM (arm64) register access not allowed before vcpu finalization
2281 ====== ============================================================
2282
2283 (These error codes are indicative only: do not rely on a specific error
2284 code being returned in a specific situation.)
2285
2286 ::
2287
2288 struct kvm_one_reg {
2289 __u64 id;
2290 __u64 addr;
2291 };
2292
2293 Using this ioctl, a single vcpu register can be set to a specific value
2294 defined by user space with the passed in struct kvm_one_reg, where id
2295 refers to the register identifier as described below and addr is a pointer
2296 to a variable with the respective size. There can be architecture agnostic
2297 and architecture specific registers. Each have their own range of operation
2298 and their own constants and width. To keep track of the implemented
2299 registers, find a list below:
2300
2301 ======= =============================== ============
2302 Arch Register Width (bits)
2303 ======= =============================== ============
2304 PPC KVM_REG_PPC_HIOR 64
2305 PPC KVM_REG_PPC_IAC1 64
2306 PPC KVM_REG_PPC_IAC2 64
2307 PPC KVM_REG_PPC_IAC3 64
2308 PPC KVM_REG_PPC_IAC4 64
2309 PPC KVM_REG_PPC_DAC1 64
2310 PPC KVM_REG_PPC_DAC2 64
2311 PPC KVM_REG_PPC_DABR 64
2312 PPC KVM_REG_PPC_DSCR 64
2313 PPC KVM_REG_PPC_PURR 64
2314 PPC KVM_REG_PPC_SPURR 64
2315 PPC KVM_REG_PPC_DAR 64
2316 PPC KVM_REG_PPC_DSISR 32
2317 PPC KVM_REG_PPC_AMR 64
2318 PPC KVM_REG_PPC_UAMOR 64
2319 PPC KVM_REG_PPC_MMCR0 64
2320 PPC KVM_REG_PPC_MMCR1 64
2321 PPC KVM_REG_PPC_MMCRA 64
2322 PPC KVM_REG_PPC_MMCR2 64
2323 PPC KVM_REG_PPC_MMCRS 64
2324 PPC KVM_REG_PPC_MMCR3 64
2325 PPC KVM_REG_PPC_SIAR 64
2326 PPC KVM_REG_PPC_SDAR 64
2327 PPC KVM_REG_PPC_SIER 64
2328 PPC KVM_REG_PPC_SIER2 64
2329 PPC KVM_REG_PPC_SIER3 64
2330 PPC KVM_REG_PPC_PMC1 32
2331 PPC KVM_REG_PPC_PMC2 32
2332 PPC KVM_REG_PPC_PMC3 32
2333 PPC KVM_REG_PPC_PMC4 32
2334 PPC KVM_REG_PPC_PMC5 32
2335 PPC KVM_REG_PPC_PMC6 32
2336 PPC KVM_REG_PPC_PMC7 32
2337 PPC KVM_REG_PPC_PMC8 32
2338 PPC KVM_REG_PPC_FPR0 64
2339 ...
2340 PPC KVM_REG_PPC_FPR31 64
2341 PPC KVM_REG_PPC_VR0 128
2342 ...
2343 PPC KVM_REG_PPC_VR31 128
2344 PPC KVM_REG_PPC_VSR0 128
2345 ...
2346 PPC KVM_REG_PPC_VSR31 128
2347 PPC KVM_REG_PPC_FPSCR 64
2348 PPC KVM_REG_PPC_VSCR 32
2349 PPC KVM_REG_PPC_VPA_ADDR 64
2350 PPC KVM_REG_PPC_VPA_SLB 128
2351 PPC KVM_REG_PPC_VPA_DTL 128
2352 PPC KVM_REG_PPC_EPCR 32
2353 PPC KVM_REG_PPC_EPR 32
2354 PPC KVM_REG_PPC_TCR 32
2355 PPC KVM_REG_PPC_TSR 32
2356 PPC KVM_REG_PPC_OR_TSR 32
2357 PPC KVM_REG_PPC_CLEAR_TSR 32
2358 PPC KVM_REG_PPC_MAS0 32
2359 PPC KVM_REG_PPC_MAS1 32
2360 PPC KVM_REG_PPC_MAS2 64
2361 PPC KVM_REG_PPC_MAS7_3 64
2362 PPC KVM_REG_PPC_MAS4 32
2363 PPC KVM_REG_PPC_MAS6 32
2364 PPC KVM_REG_PPC_MMUCFG 32
2365 PPC KVM_REG_PPC_TLB0CFG 32
2366 PPC KVM_REG_PPC_TLB1CFG 32
2367 PPC KVM_REG_PPC_TLB2CFG 32
2368 PPC KVM_REG_PPC_TLB3CFG 32
2369 PPC KVM_REG_PPC_TLB0PS 32
2370 PPC KVM_REG_PPC_TLB1PS 32
2371 PPC KVM_REG_PPC_TLB2PS 32
2372 PPC KVM_REG_PPC_TLB3PS 32
2373 PPC KVM_REG_PPC_EPTCFG 32
2374 PPC KVM_REG_PPC_ICP_STATE 64
2375 PPC KVM_REG_PPC_VP_STATE 128
2376 PPC KVM_REG_PPC_TB_OFFSET 64
2377 PPC KVM_REG_PPC_SPMC1 32
2378 PPC KVM_REG_PPC_SPMC2 32
2379 PPC KVM_REG_PPC_IAMR 64
2380 PPC KVM_REG_PPC_TFHAR 64
2381 PPC KVM_REG_PPC_TFIAR 64
2382 PPC KVM_REG_PPC_TEXASR 64
2383 PPC KVM_REG_PPC_FSCR 64
2384 PPC KVM_REG_PPC_PSPB 32
2385 PPC KVM_REG_PPC_EBBHR 64
2386 PPC KVM_REG_PPC_EBBRR 64
2387 PPC KVM_REG_PPC_BESCR 64
2388 PPC KVM_REG_PPC_TAR 64
2389 PPC KVM_REG_PPC_DPDES 64
2390 PPC KVM_REG_PPC_DAWR 64
2391 PPC KVM_REG_PPC_DAWRX 64
2392 PPC KVM_REG_PPC_CIABR 64
2393 PPC KVM_REG_PPC_IC 64
2394 PPC KVM_REG_PPC_VTB 64
2395 PPC KVM_REG_PPC_CSIGR 64
2396 PPC KVM_REG_PPC_TACR 64
2397 PPC KVM_REG_PPC_TCSCR 64
2398 PPC KVM_REG_PPC_PID 64
2399 PPC KVM_REG_PPC_ACOP 64
2400 PPC KVM_REG_PPC_VRSAVE 32
2401 PPC KVM_REG_PPC_LPCR 32
2402 PPC KVM_REG_PPC_LPCR_64 64
2403 PPC KVM_REG_PPC_PPR 64
2404 PPC KVM_REG_PPC_ARCH_COMPAT 32
2405 PPC KVM_REG_PPC_DABRX 32
2406 PPC KVM_REG_PPC_WORT 64
2407 PPC KVM_REG_PPC_SPRG9 64
2408 PPC KVM_REG_PPC_DBSR 32
2409 PPC KVM_REG_PPC_TIDR 64
2410 PPC KVM_REG_PPC_PSSCR 64
2411 PPC KVM_REG_PPC_DEC_EXPIRY 64
2412 PPC KVM_REG_PPC_PTCR 64
2413 PPC KVM_REG_PPC_DAWR1 64
2414 PPC KVM_REG_PPC_DAWRX1 64
2415 PPC KVM_REG_PPC_TM_GPR0 64
2416 ...
2417 PPC KVM_REG_PPC_TM_GPR31 64
2418 PPC KVM_REG_PPC_TM_VSR0 128
2419 ...
2420 PPC KVM_REG_PPC_TM_VSR63 128
2421 PPC KVM_REG_PPC_TM_CR 64
2422 PPC KVM_REG_PPC_TM_LR 64
2423 PPC KVM_REG_PPC_TM_CTR 64
2424 PPC KVM_REG_PPC_TM_FPSCR 64
2425 PPC KVM_REG_PPC_TM_AMR 64
2426 PPC KVM_REG_PPC_TM_PPR 64
2427 PPC KVM_REG_PPC_TM_VRSAVE 64
2428 PPC KVM_REG_PPC_TM_VSCR 32
2429 PPC KVM_REG_PPC_TM_DSCR 64
2430 PPC KVM_REG_PPC_TM_TAR 64
2431 PPC KVM_REG_PPC_TM_XER 64
2432
2433 MIPS KVM_REG_MIPS_R0 64
2434 ...
2435 MIPS KVM_REG_MIPS_R31 64
2436 MIPS KVM_REG_MIPS_HI 64
2437 MIPS KVM_REG_MIPS_LO 64
2438 MIPS KVM_REG_MIPS_PC 64
2439 MIPS KVM_REG_MIPS_CP0_INDEX 32
2440 MIPS KVM_REG_MIPS_CP0_ENTRYLO0 64
2441 MIPS KVM_REG_MIPS_CP0_ENTRYLO1 64
2442 MIPS KVM_REG_MIPS_CP0_CONTEXT 64
2443 MIPS KVM_REG_MIPS_CP0_CONTEXTCONFIG 32
2444 MIPS KVM_REG_MIPS_CP0_USERLOCAL 64
2445 MIPS KVM_REG_MIPS_CP0_XCONTEXTCONFIG 64
2446 MIPS KVM_REG_MIPS_CP0_PAGEMASK 32
2447 MIPS KVM_REG_MIPS_CP0_PAGEGRAIN 32
2448 MIPS KVM_REG_MIPS_CP0_SEGCTL0 64
2449 MIPS KVM_REG_MIPS_CP0_SEGCTL1 64
2450 MIPS KVM_REG_MIPS_CP0_SEGCTL2 64
2451 MIPS KVM_REG_MIPS_CP0_PWBASE 64
2452 MIPS KVM_REG_MIPS_CP0_PWFIELD 64
2453 MIPS KVM_REG_MIPS_CP0_PWSIZE 64
2454 MIPS KVM_REG_MIPS_CP0_WIRED 32
2455 MIPS KVM_REG_MIPS_CP0_PWCTL 32
2456 MIPS KVM_REG_MIPS_CP0_HWRENA 32
2457 MIPS KVM_REG_MIPS_CP0_BADVADDR 64
2458 MIPS KVM_REG_MIPS_CP0_BADINSTR 32
2459 MIPS KVM_REG_MIPS_CP0_BADINSTRP 32
2460 MIPS KVM_REG_MIPS_CP0_COUNT 32
2461 MIPS KVM_REG_MIPS_CP0_ENTRYHI 64
2462 MIPS KVM_REG_MIPS_CP0_COMPARE 32
2463 MIPS KVM_REG_MIPS_CP0_STATUS 32
2464 MIPS KVM_REG_MIPS_CP0_INTCTL 32
2465 MIPS KVM_REG_MIPS_CP0_CAUSE 32
2466 MIPS KVM_REG_MIPS_CP0_EPC 64
2467 MIPS KVM_REG_MIPS_CP0_PRID 32
2468 MIPS KVM_REG_MIPS_CP0_EBASE 64
2469 MIPS KVM_REG_MIPS_CP0_CONFIG 32
2470 MIPS KVM_REG_MIPS_CP0_CONFIG1 32
2471 MIPS KVM_REG_MIPS_CP0_CONFIG2 32
2472 MIPS KVM_REG_MIPS_CP0_CONFIG3 32
2473 MIPS KVM_REG_MIPS_CP0_CONFIG4 32
2474 MIPS KVM_REG_MIPS_CP0_CONFIG5 32
2475 MIPS KVM_REG_MIPS_CP0_CONFIG7 32
2476 MIPS KVM_REG_MIPS_CP0_XCONTEXT 64
2477 MIPS KVM_REG_MIPS_CP0_ERROREPC 64
2478 MIPS KVM_REG_MIPS_CP0_KSCRATCH1 64
2479 MIPS KVM_REG_MIPS_CP0_KSCRATCH2 64
2480 MIPS KVM_REG_MIPS_CP0_KSCRATCH3 64
2481 MIPS KVM_REG_MIPS_CP0_KSCRATCH4 64
2482 MIPS KVM_REG_MIPS_CP0_KSCRATCH5 64
2483 MIPS KVM_REG_MIPS_CP0_KSCRATCH6 64
2484 MIPS KVM_REG_MIPS_CP0_MAAR(0..63) 64
2485 MIPS KVM_REG_MIPS_COUNT_CTL 64
2486 MIPS KVM_REG_MIPS_COUNT_RESUME 64
2487 MIPS KVM_REG_MIPS_COUNT_HZ 64
2488 MIPS KVM_REG_MIPS_FPR_32(0..31) 32
2489 MIPS KVM_REG_MIPS_FPR_64(0..31) 64
2490 MIPS KVM_REG_MIPS_VEC_128(0..31) 128
2491 MIPS KVM_REG_MIPS_FCR_IR 32
2492 MIPS KVM_REG_MIPS_FCR_CSR 32
2493 MIPS KVM_REG_MIPS_MSA_IR 32
2494 MIPS KVM_REG_MIPS_MSA_CSR 32
2495 ======= =============================== ============
2496
2497 ARM registers are mapped using the lower 32 bits. The upper 16 of that
2498 is the register group type, or coprocessor number:
2499
2500 ARM core registers have the following id bit patterns::
2501
2502 0x4020 0000 0010 <index into the kvm_regs struct:16>
2503
2504 ARM 32-bit CP15 registers have the following id bit patterns::
2505
2506 0x4020 0000 000F <zero:1> <crn:4> <crm:4> <opc1:4> <opc2:3>
2507
2508 ARM 64-bit CP15 registers have the following id bit patterns::
2509
2510 0x4030 0000 000F <zero:1> <zero:4> <crm:4> <opc1:4> <zero:3>
2511
2512 ARM CCSIDR registers are demultiplexed by CSSELR value::
2513
2514 0x4020 0000 0011 00 <csselr:8>
2515
2516 ARM 32-bit VFP control registers have the following id bit patterns::
2517
2518 0x4020 0000 0012 1 <regno:12>
2519
2520 ARM 64-bit FP registers have the following id bit patterns::
2521
2522 0x4030 0000 0012 0 <regno:12>
2523
2524 ARM firmware pseudo-registers have the following bit pattern::
2525
2526 0x4030 0000 0014 <regno:16>
2527
2528
2529 arm64 registers are mapped using the lower 32 bits. The upper 16 of
2530 that is the register group type, or coprocessor number:
2531
2532 arm64 core/FP-SIMD registers have the following id bit patterns. Note
2533 that the size of the access is variable, as the kvm_regs structure
2534 contains elements ranging from 32 to 128 bits. The index is a 32bit
2535 value in the kvm_regs structure seen as a 32bit array::
2536
2537 0x60x0 0000 0010 <index into the kvm_regs struct:16>
2538
2539 Specifically:
2540
2541 ======================= ========= ===== =======================================
2542 Encoding Register Bits kvm_regs member
2543 ======================= ========= ===== =======================================
2544 0x6030 0000 0010 0000 X0 64 regs.regs[0]
2545 0x6030 0000 0010 0002 X1 64 regs.regs[1]
2546 ...
2547 0x6030 0000 0010 003c X30 64 regs.regs[30]
2548 0x6030 0000 0010 003e SP 64 regs.sp
2549 0x6030 0000 0010 0040 PC 64 regs.pc
2550 0x6030 0000 0010 0042 PSTATE 64 regs.pstate
2551 0x6030 0000 0010 0044 SP_EL1 64 sp_el1
2552 0x6030 0000 0010 0046 ELR_EL1 64 elr_el1
2553 0x6030 0000 0010 0048 SPSR_EL1 64 spsr[KVM_SPSR_EL1] (alias SPSR_SVC)
2554 0x6030 0000 0010 004a SPSR_ABT 64 spsr[KVM_SPSR_ABT]
2555 0x6030 0000 0010 004c SPSR_UND 64 spsr[KVM_SPSR_UND]
2556 0x6030 0000 0010 004e SPSR_IRQ 64 spsr[KVM_SPSR_IRQ]
2557 0x6060 0000 0010 0050 SPSR_FIQ 64 spsr[KVM_SPSR_FIQ]
2558 0x6040 0000 0010 0054 V0 128 fp_regs.vregs[0] [1]_
2559 0x6040 0000 0010 0058 V1 128 fp_regs.vregs[1] [1]_
2560 ...
2561 0x6040 0000 0010 00d0 V31 128 fp_regs.vregs[31] [1]_
2562 0x6020 0000 0010 00d4 FPSR 32 fp_regs.fpsr
2563 0x6020 0000 0010 00d5 FPCR 32 fp_regs.fpcr
2564 ======================= ========= ===== =======================================
2565
2566 .. [1] These encodings are not accepted for SVE-enabled vcpus. See
2567 KVM_ARM_VCPU_INIT.
2568
2569 The equivalent register content can be accessed via bits [127:0] of
2570 the corresponding SVE Zn registers instead for vcpus that have SVE
2571 enabled (see below).
2572
2573 arm64 CCSIDR registers are demultiplexed by CSSELR value::
2574
2575 0x6020 0000 0011 00 <csselr:8>
2576
2577 arm64 system registers have the following id bit patterns::
2578
2579 0x6030 0000 0013 <op0:2> <op1:3> <crn:4> <crm:4> <op2:3>
2580
2581 .. warning::
2582
2583 Two system register IDs do not follow the specified pattern. These
2584 are KVM_REG_ARM_TIMER_CVAL and KVM_REG_ARM_TIMER_CNT, which map to
2585 system registers CNTV_CVAL_EL0 and CNTVCT_EL0 respectively. These
2586 two had their values accidentally swapped, which means TIMER_CVAL is
2587 derived from the register encoding for CNTVCT_EL0 and TIMER_CNT is
2588 derived from the register encoding for CNTV_CVAL_EL0. As this is
2589 API, it must remain this way.
2590
2591 arm64 firmware pseudo-registers have the following bit pattern::
2592
2593 0x6030 0000 0014 <regno:16>
2594
2595 arm64 SVE registers have the following bit patterns::
2596
2597 0x6080 0000 0015 00 <n:5> <slice:5> Zn bits[2048*slice + 2047 : 2048*slice]
2598 0x6050 0000 0015 04 <n:4> <slice:5> Pn bits[256*slice + 255 : 256*slice]
2599 0x6050 0000 0015 060 <slice:5> FFR bits[256*slice + 255 : 256*slice]
2600 0x6060 0000 0015 ffff KVM_REG_ARM64_SVE_VLS pseudo-register
2601
2602 Access to register IDs where 2048 * slice >= 128 * max_vq will fail with
2603 ENOENT. max_vq is the vcpu's maximum supported vector length in 128-bit
2604 quadwords: see [2]_ below.
2605
2606 These registers are only accessible on vcpus for which SVE is enabled.
2607 See KVM_ARM_VCPU_INIT for details.
2608
2609 In addition, except for KVM_REG_ARM64_SVE_VLS, these registers are not
2610 accessible until the vcpu's SVE configuration has been finalized
2611 using KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE). See KVM_ARM_VCPU_INIT
2612 and KVM_ARM_VCPU_FINALIZE for more information about this procedure.
2613
2614 KVM_REG_ARM64_SVE_VLS is a pseudo-register that allows the set of vector
2615 lengths supported by the vcpu to be discovered and configured by
2616 userspace. When transferred to or from user memory via KVM_GET_ONE_REG
2617 or KVM_SET_ONE_REG, the value of this register is of type
2618 __u64[KVM_ARM64_SVE_VLS_WORDS], and encodes the set of vector lengths as
2619 follows::
2620
2621 __u64 vector_lengths[KVM_ARM64_SVE_VLS_WORDS];
2622
2623 if (vq >= SVE_VQ_MIN && vq <= SVE_VQ_MAX &&
2624 ((vector_lengths[(vq - KVM_ARM64_SVE_VQ_MIN) / 64] >>
2625 ((vq - KVM_ARM64_SVE_VQ_MIN) % 64)) & 1))
2626 /* Vector length vq * 16 bytes supported */
2627 else
2628 /* Vector length vq * 16 bytes not supported */
2629
2630 .. [2] The maximum value vq for which the above condition is true is
2631 max_vq. This is the maximum vector length available to the guest on
2632 this vcpu, and determines which register slices are visible through
2633 this ioctl interface.
2634
2635 (See Documentation/arm64/sve.rst for an explanation of the "vq"
2636 nomenclature.)
2637
2638 KVM_REG_ARM64_SVE_VLS is only accessible after KVM_ARM_VCPU_INIT.
2639 KVM_ARM_VCPU_INIT initialises it to the best set of vector lengths that
2640 the host supports.
2641
2642 Userspace may subsequently modify it if desired until the vcpu's SVE
2643 configuration is finalized using KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE).
2644
2645 Apart from simply removing all vector lengths from the host set that
2646 exceed some value, support for arbitrarily chosen sets of vector lengths
2647 is hardware-dependent and may not be available. Attempting to configure
2648 an invalid set of vector lengths via KVM_SET_ONE_REG will fail with
2649 EINVAL.
2650
2651 After the vcpu's SVE configuration is finalized, further attempts to
2652 write this register will fail with EPERM.
2653
2654 arm64 bitmap feature firmware pseudo-registers have the following bit pattern::
2655
2656 0x6030 0000 0016 <regno:16>
2657
2658 The bitmap feature firmware registers exposes the hypercall services that
2659 are available for userspace to configure. The set bits corresponds to the
2660 services that are available for the guests to access. By default, KVM
2661 sets all the supported bits during VM initialization. The userspace can
2662 discover the available services via KVM_GET_ONE_REG, and write back the
2663 bitmap corresponding to the features that it wishes guests to see via
2664 KVM_SET_ONE_REG.
2665
2666 Note: These registers are immutable once any of the vCPUs of the VM has
2667 run at least once. A KVM_SET_ONE_REG in such a scenario will return
2668 a -EBUSY to userspace.
2669
2670 (See Documentation/virt/kvm/arm/hypercalls.rst for more details.)
2671
2672
2673 MIPS registers are mapped using the lower 32 bits. The upper 16 of that is
2674 the register group type:
2675
2676 MIPS core registers (see above) have the following id bit patterns::
2677
2678 0x7030 0000 0000 <reg:16>
2679
2680 MIPS CP0 registers (see KVM_REG_MIPS_CP0_* above) have the following id bit
2681 patterns depending on whether they're 32-bit or 64-bit registers::
2682
2683 0x7020 0000 0001 00 <reg:5> <sel:3> (32-bit)
2684 0x7030 0000 0001 00 <reg:5> <sel:3> (64-bit)
2685
2686 Note: KVM_REG_MIPS_CP0_ENTRYLO0 and KVM_REG_MIPS_CP0_ENTRYLO1 are the MIPS64
2687 versions of the EntryLo registers regardless of the word size of the host
2688 hardware, host kernel, guest, and whether XPA is present in the guest, i.e.
2689 with the RI and XI bits (if they exist) in bits 63 and 62 respectively, and
2690 the PFNX field starting at bit 30.
2691
2692 MIPS MAARs (see KVM_REG_MIPS_CP0_MAAR(*) above) have the following id bit
2693 patterns::
2694
2695 0x7030 0000 0001 01 <reg:8>
2696
2697 MIPS KVM control registers (see above) have the following id bit patterns::
2698
2699 0x7030 0000 0002 <reg:16>
2700
2701 MIPS FPU registers (see KVM_REG_MIPS_FPR_{32,64}() above) have the following
2702 id bit patterns depending on the size of the register being accessed. They are
2703 always accessed according to the current guest FPU mode (Status.FR and
2704 Config5.FRE), i.e. as the guest would see them, and they become unpredictable
2705 if the guest FPU mode is changed. MIPS SIMD Architecture (MSA) vector
2706 registers (see KVM_REG_MIPS_VEC_128() above) have similar patterns as they
2707 overlap the FPU registers::
2708
2709 0x7020 0000 0003 00 <0:3> <reg:5> (32-bit FPU registers)
2710 0x7030 0000 0003 00 <0:3> <reg:5> (64-bit FPU registers)
2711 0x7040 0000 0003 00 <0:3> <reg:5> (128-bit MSA vector registers)
2712
2713 MIPS FPU control registers (see KVM_REG_MIPS_FCR_{IR,CSR} above) have the
2714 following id bit patterns::
2715
2716 0x7020 0000 0003 01 <0:3> <reg:5>
2717
2718 MIPS MSA control registers (see KVM_REG_MIPS_MSA_{IR,CSR} above) have the
2719 following id bit patterns::
2720
2721 0x7020 0000 0003 02 <0:3> <reg:5>
2722
2723 RISC-V registers are mapped using the lower 32 bits. The upper 8 bits of
2724 that is the register group type.
2725
2726 RISC-V config registers are meant for configuring a Guest VCPU and it has
2727 the following id bit patterns::
2728
2729 0x8020 0000 01 <index into the kvm_riscv_config struct:24> (32bit Host)
2730 0x8030 0000 01 <index into the kvm_riscv_config struct:24> (64bit Host)
2731
2732 Following are the RISC-V config registers:
2733
2734 ======================= ========= =============================================
2735 Encoding Register Description
2736 ======================= ========= =============================================
2737 0x80x0 0000 0100 0000 isa ISA feature bitmap of Guest VCPU
2738 ======================= ========= =============================================
2739
2740 The isa config register can be read anytime but can only be written before
2741 a Guest VCPU runs. It will have ISA feature bits matching underlying host
2742 set by default.
2743
2744 RISC-V core registers represent the general excution state of a Guest VCPU
2745 and it has the following id bit patterns::
2746
2747 0x8020 0000 02 <index into the kvm_riscv_core struct:24> (32bit Host)
2748 0x8030 0000 02 <index into the kvm_riscv_core struct:24> (64bit Host)
2749
2750 Following are the RISC-V core registers:
2751
2752 ======================= ========= =============================================
2753 Encoding Register Description
2754 ======================= ========= =============================================
2755 0x80x0 0000 0200 0000 regs.pc Program counter
2756 0x80x0 0000 0200 0001 regs.ra Return address
2757 0x80x0 0000 0200 0002 regs.sp Stack pointer
2758 0x80x0 0000 0200 0003 regs.gp Global pointer
2759 0x80x0 0000 0200 0004 regs.tp Task pointer
2760 0x80x0 0000 0200 0005 regs.t0 Caller saved register 0
2761 0x80x0 0000 0200 0006 regs.t1 Caller saved register 1
2762 0x80x0 0000 0200 0007 regs.t2 Caller saved register 2
2763 0x80x0 0000 0200 0008 regs.s0 Callee saved register 0
2764 0x80x0 0000 0200 0009 regs.s1 Callee saved register 1
2765 0x80x0 0000 0200 000a regs.a0 Function argument (or return value) 0
2766 0x80x0 0000 0200 000b regs.a1 Function argument (or return value) 1
2767 0x80x0 0000 0200 000c regs.a2 Function argument 2
2768 0x80x0 0000 0200 000d regs.a3 Function argument 3
2769 0x80x0 0000 0200 000e regs.a4 Function argument 4
2770 0x80x0 0000 0200 000f regs.a5 Function argument 5
2771 0x80x0 0000 0200 0010 regs.a6 Function argument 6
2772 0x80x0 0000 0200 0011 regs.a7 Function argument 7
2773 0x80x0 0000 0200 0012 regs.s2 Callee saved register 2
2774 0x80x0 0000 0200 0013 regs.s3 Callee saved register 3
2775 0x80x0 0000 0200 0014 regs.s4 Callee saved register 4
2776 0x80x0 0000 0200 0015 regs.s5 Callee saved register 5
2777 0x80x0 0000 0200 0016 regs.s6 Callee saved register 6
2778 0x80x0 0000 0200 0017 regs.s7 Callee saved register 7
2779 0x80x0 0000 0200 0018 regs.s8 Callee saved register 8
2780 0x80x0 0000 0200 0019 regs.s9 Callee saved register 9
2781 0x80x0 0000 0200 001a regs.s10 Callee saved register 10
2782 0x80x0 0000 0200 001b regs.s11 Callee saved register 11
2783 0x80x0 0000 0200 001c regs.t3 Caller saved register 3
2784 0x80x0 0000 0200 001d regs.t4 Caller saved register 4
2785 0x80x0 0000 0200 001e regs.t5 Caller saved register 5
2786 0x80x0 0000 0200 001f regs.t6 Caller saved register 6
2787 0x80x0 0000 0200 0020 mode Privilege mode (1 = S-mode or 0 = U-mode)
2788 ======================= ========= =============================================
2789
2790 RISC-V csr registers represent the supervisor mode control/status registers
2791 of a Guest VCPU and it has the following id bit patterns::
2792
2793 0x8020 0000 03 <index into the kvm_riscv_csr struct:24> (32bit Host)
2794 0x8030 0000 03 <index into the kvm_riscv_csr struct:24> (64bit Host)
2795
2796 Following are the RISC-V csr registers:
2797
2798 ======================= ========= =============================================
2799 Encoding Register Description
2800 ======================= ========= =============================================
2801 0x80x0 0000 0300 0000 sstatus Supervisor status
2802 0x80x0 0000 0300 0001 sie Supervisor interrupt enable
2803 0x80x0 0000 0300 0002 stvec Supervisor trap vector base
2804 0x80x0 0000 0300 0003 sscratch Supervisor scratch register
2805 0x80x0 0000 0300 0004 sepc Supervisor exception program counter
2806 0x80x0 0000 0300 0005 scause Supervisor trap cause
2807 0x80x0 0000 0300 0006 stval Supervisor bad address or instruction
2808 0x80x0 0000 0300 0007 sip Supervisor interrupt pending
2809 0x80x0 0000 0300 0008 satp Supervisor address translation and protection
2810 ======================= ========= =============================================
2811
2812 RISC-V timer registers represent the timer state of a Guest VCPU and it has
2813 the following id bit patterns::
2814
2815 0x8030 0000 04 <index into the kvm_riscv_timer struct:24>
2816
2817 Following are the RISC-V timer registers:
2818
2819 ======================= ========= =============================================
2820 Encoding Register Description
2821 ======================= ========= =============================================
2822 0x8030 0000 0400 0000 frequency Time base frequency (read-only)
2823 0x8030 0000 0400 0001 time Time value visible to Guest
2824 0x8030 0000 0400 0002 compare Time compare programmed by Guest
2825 0x8030 0000 0400 0003 state Time compare state (1 = ON or 0 = OFF)
2826 ======================= ========= =============================================
2827
2828 RISC-V F-extension registers represent the single precision floating point
2829 state of a Guest VCPU and it has the following id bit patterns::
2830
2831 0x8020 0000 05 <index into the __riscv_f_ext_state struct:24>
2832
2833 Following are the RISC-V F-extension registers:
2834
2835 ======================= ========= =============================================
2836 Encoding Register Description
2837 ======================= ========= =============================================
2838 0x8020 0000 0500 0000 f[0] Floating point register 0
2839 ...
2840 0x8020 0000 0500 001f f[31] Floating point register 31
2841 0x8020 0000 0500 0020 fcsr Floating point control and status register
2842 ======================= ========= =============================================
2843
2844 RISC-V D-extension registers represent the double precision floating point
2845 state of a Guest VCPU and it has the following id bit patterns::
2846
2847 0x8020 0000 06 <index into the __riscv_d_ext_state struct:24> (fcsr)
2848 0x8030 0000 06 <index into the __riscv_d_ext_state struct:24> (non-fcsr)
2849
2850 Following are the RISC-V D-extension registers:
2851
2852 ======================= ========= =============================================
2853 Encoding Register Description
2854 ======================= ========= =============================================
2855 0x8030 0000 0600 0000 f[0] Floating point register 0
2856 ...
2857 0x8030 0000 0600 001f f[31] Floating point register 31
2858 0x8020 0000 0600 0020 fcsr Floating point control and status register
2859 ======================= ========= =============================================
2860
2861
2862 4.69 KVM_GET_ONE_REG
2863 --------------------
2864
2865 :Capability: KVM_CAP_ONE_REG
2866 :Architectures: all
2867 :Type: vcpu ioctl
2868 :Parameters: struct kvm_one_reg (in and out)
2869 :Returns: 0 on success, negative value on failure
2870
2871 Errors include:
2872
2873 ======== ============================================================
2874 ENOENT no such register
2875 EINVAL invalid register ID, or no such register or used with VMs in
2876 protected virtualization mode on s390
2877 EPERM (arm64) register access not allowed before vcpu finalization
2878 ======== ============================================================
2879
2880 (These error codes are indicative only: do not rely on a specific error
2881 code being returned in a specific situation.)
2882
2883 This ioctl allows to receive the value of a single register implemented
2884 in a vcpu. The register to read is indicated by the "id" field of the
2885 kvm_one_reg struct passed in. On success, the register value can be found
2886 at the memory location pointed to by "addr".
2887
2888 The list of registers accessible using this interface is identical to the
2889 list in 4.68.
2890
2891
2892 4.70 KVM_KVMCLOCK_CTRL
2893 ----------------------
2894
2895 :Capability: KVM_CAP_KVMCLOCK_CTRL
2896 :Architectures: Any that implement pvclocks (currently x86 only)
2897 :Type: vcpu ioctl
2898 :Parameters: None
2899 :Returns: 0 on success, -1 on error
2900
2901 This ioctl sets a flag accessible to the guest indicating that the specified
2902 vCPU has been paused by the host userspace.
2903
2904 The host will set a flag in the pvclock structure that is checked from the
2905 soft lockup watchdog. The flag is part of the pvclock structure that is
2906 shared between guest and host, specifically the second bit of the flags
2907 field of the pvclock_vcpu_time_info structure. It will be set exclusively by
2908 the host and read/cleared exclusively by the guest. The guest operation of
2909 checking and clearing the flag must be an atomic operation so
2910 load-link/store-conditional, or equivalent must be used. There are two cases
2911 where the guest will clear the flag: when the soft lockup watchdog timer resets
2912 itself or when a soft lockup is detected. This ioctl can be called any time
2913 after pausing the vcpu, but before it is resumed.
2914
2915
2916 4.71 KVM_SIGNAL_MSI
2917 -------------------
2918
2919 :Capability: KVM_CAP_SIGNAL_MSI
2920 :Architectures: x86 arm64
2921 :Type: vm ioctl
2922 :Parameters: struct kvm_msi (in)
2923 :Returns: >0 on delivery, 0 if guest blocked the MSI, and -1 on error
2924
2925 Directly inject a MSI message. Only valid with in-kernel irqchip that handles
2926 MSI messages.
2927
2928 ::
2929
2930 struct kvm_msi {
2931 __u32 address_lo;
2932 __u32 address_hi;
2933 __u32 data;
2934 __u32 flags;
2935 __u32 devid;
2936 __u8 pad[12];
2937 };
2938
2939 flags:
2940 KVM_MSI_VALID_DEVID: devid contains a valid value. The per-VM
2941 KVM_CAP_MSI_DEVID capability advertises the requirement to provide
2942 the device ID. If this capability is not available, userspace
2943 should never set the KVM_MSI_VALID_DEVID flag as the ioctl might fail.
2944
2945 If KVM_MSI_VALID_DEVID is set, devid contains a unique device identifier
2946 for the device that wrote the MSI message. For PCI, this is usually a
2947 BFD identifier in the lower 16 bits.
2948
2949 On x86, address_hi is ignored unless the KVM_X2APIC_API_USE_32BIT_IDS
2950 feature of KVM_CAP_X2APIC_API capability is enabled. If it is enabled,
2951 address_hi bits 31-8 provide bits 31-8 of the destination id. Bits 7-0 of
2952 address_hi must be zero.
2953
2954
2955 4.71 KVM_CREATE_PIT2
2956 --------------------
2957
2958 :Capability: KVM_CAP_PIT2
2959 :Architectures: x86
2960 :Type: vm ioctl
2961 :Parameters: struct kvm_pit_config (in)
2962 :Returns: 0 on success, -1 on error
2963
2964 Creates an in-kernel device model for the i8254 PIT. This call is only valid
2965 after enabling in-kernel irqchip support via KVM_CREATE_IRQCHIP. The following
2966 parameters have to be passed::
2967
2968 struct kvm_pit_config {
2969 __u32 flags;
2970 __u32 pad[15];
2971 };
2972
2973 Valid flags are::
2974
2975 #define KVM_PIT_SPEAKER_DUMMY 1 /* emulate speaker port stub */
2976
2977 PIT timer interrupts may use a per-VM kernel thread for injection. If it
2978 exists, this thread will have a name of the following pattern::
2979
2980 kvm-pit/<owner-process-pid>
2981
2982 When running a guest with elevated priorities, the scheduling parameters of
2983 this thread may have to be adjusted accordingly.
2984
2985 This IOCTL replaces the obsolete KVM_CREATE_PIT.
2986
2987
2988 4.72 KVM_GET_PIT2
2989 -----------------
2990
2991 :Capability: KVM_CAP_PIT_STATE2
2992 :Architectures: x86
2993 :Type: vm ioctl
2994 :Parameters: struct kvm_pit_state2 (out)
2995 :Returns: 0 on success, -1 on error
2996
2997 Retrieves the state of the in-kernel PIT model. Only valid after
2998 KVM_CREATE_PIT2. The state is returned in the following structure::
2999
3000 struct kvm_pit_state2 {
3001 struct kvm_pit_channel_state channels[3];
3002 __u32 flags;
3003 __u32 reserved[9];
3004 };
3005
3006 Valid flags are::
3007
3008 /* disable PIT in HPET legacy mode */
3009 #define KVM_PIT_FLAGS_HPET_LEGACY 0x00000001
3010 /* speaker port data bit enabled */
3011 #define KVM_PIT_FLAGS_SPEAKER_DATA_ON 0x00000002
3012
3013 This IOCTL replaces the obsolete KVM_GET_PIT.
3014
3015
3016 4.73 KVM_SET_PIT2
3017 -----------------
3018
3019 :Capability: KVM_CAP_PIT_STATE2
3020 :Architectures: x86
3021 :Type: vm ioctl
3022 :Parameters: struct kvm_pit_state2 (in)
3023 :Returns: 0 on success, -1 on error
3024
3025 Sets the state of the in-kernel PIT model. Only valid after KVM_CREATE_PIT2.
3026 See KVM_GET_PIT2 for details on struct kvm_pit_state2.
3027
3028 This IOCTL replaces the obsolete KVM_SET_PIT.
3029
3030
3031 4.74 KVM_PPC_GET_SMMU_INFO
3032 --------------------------
3033
3034 :Capability: KVM_CAP_PPC_GET_SMMU_INFO
3035 :Architectures: powerpc
3036 :Type: vm ioctl
3037 :Parameters: None
3038 :Returns: 0 on success, -1 on error
3039
3040 This populates and returns a structure describing the features of
3041 the "Server" class MMU emulation supported by KVM.
3042 This can in turn be used by userspace to generate the appropriate
3043 device-tree properties for the guest operating system.
3044
3045 The structure contains some global information, followed by an
3046 array of supported segment page sizes::
3047
3048 struct kvm_ppc_smmu_info {
3049 __u64 flags;
3050 __u32 slb_size;
3051 __u32 pad;
3052 struct kvm_ppc_one_seg_page_size sps[KVM_PPC_PAGE_SIZES_MAX_SZ];
3053 };
3054
3055 The supported flags are:
3056
3057 - KVM_PPC_PAGE_SIZES_REAL:
3058 When that flag is set, guest page sizes must "fit" the backing
3059 store page sizes. When not set, any page size in the list can
3060 be used regardless of how they are backed by userspace.
3061
3062 - KVM_PPC_1T_SEGMENTS
3063 The emulated MMU supports 1T segments in addition to the
3064 standard 256M ones.
3065
3066 - KVM_PPC_NO_HASH
3067 This flag indicates that HPT guests are not supported by KVM,
3068 thus all guests must use radix MMU mode.
3069
3070 The "slb_size" field indicates how many SLB entries are supported
3071
3072 The "sps" array contains 8 entries indicating the supported base
3073 page sizes for a segment in increasing order. Each entry is defined
3074 as follow::
3075
3076 struct kvm_ppc_one_seg_page_size {
3077 __u32 page_shift; /* Base page shift of segment (or 0) */
3078 __u32 slb_enc; /* SLB encoding for BookS */
3079 struct kvm_ppc_one_page_size enc[KVM_PPC_PAGE_SIZES_MAX_SZ];
3080 };
3081
3082 An entry with a "page_shift" of 0 is unused. Because the array is
3083 organized in increasing order, a lookup can stop when encoutering
3084 such an entry.
3085
3086 The "slb_enc" field provides the encoding to use in the SLB for the
3087 page size. The bits are in positions such as the value can directly
3088 be OR'ed into the "vsid" argument of the slbmte instruction.
3089
3090 The "enc" array is a list which for each of those segment base page
3091 size provides the list of supported actual page sizes (which can be
3092 only larger or equal to the base page size), along with the
3093 corresponding encoding in the hash PTE. Similarly, the array is
3094 8 entries sorted by increasing sizes and an entry with a "0" shift
3095 is an empty entry and a terminator::
3096
3097 struct kvm_ppc_one_page_size {
3098 __u32 page_shift; /* Page shift (or 0) */
3099 __u32 pte_enc; /* Encoding in the HPTE (>>12) */
3100 };
3101
3102 The "pte_enc" field provides a value that can OR'ed into the hash
3103 PTE's RPN field (ie, it needs to be shifted left by 12 to OR it
3104 into the hash PTE second double word).
3105
3106 4.75 KVM_IRQFD
3107 --------------
3108
3109 :Capability: KVM_CAP_IRQFD
3110 :Architectures: x86 s390 arm64
3111 :Type: vm ioctl
3112 :Parameters: struct kvm_irqfd (in)
3113 :Returns: 0 on success, -1 on error
3114
3115 Allows setting an eventfd to directly trigger a guest interrupt.
3116 kvm_irqfd.fd specifies the file descriptor to use as the eventfd and
3117 kvm_irqfd.gsi specifies the irqchip pin toggled by this event. When
3118 an event is triggered on the eventfd, an interrupt is injected into
3119 the guest using the specified gsi pin. The irqfd is removed using
3120 the KVM_IRQFD_FLAG_DEASSIGN flag, specifying both kvm_irqfd.fd
3121 and kvm_irqfd.gsi.
3122
3123 With KVM_CAP_IRQFD_RESAMPLE, KVM_IRQFD supports a de-assert and notify
3124 mechanism allowing emulation of level-triggered, irqfd-based
3125 interrupts. When KVM_IRQFD_FLAG_RESAMPLE is set the user must pass an
3126 additional eventfd in the kvm_irqfd.resamplefd field. When operating
3127 in resample mode, posting of an interrupt through kvm_irq.fd asserts
3128 the specified gsi in the irqchip. When the irqchip is resampled, such
3129 as from an EOI, the gsi is de-asserted and the user is notified via
3130 kvm_irqfd.resamplefd. It is the user's responsibility to re-queue
3131 the interrupt if the device making use of it still requires service.
3132 Note that closing the resamplefd is not sufficient to disable the
3133 irqfd. The KVM_IRQFD_FLAG_RESAMPLE is only necessary on assignment
3134 and need not be specified with KVM_IRQFD_FLAG_DEASSIGN.
3135
3136 On arm64, gsi routing being supported, the following can happen:
3137
3138 - in case no routing entry is associated to this gsi, injection fails
3139 - in case the gsi is associated to an irqchip routing entry,
3140 irqchip.pin + 32 corresponds to the injected SPI ID.
3141 - in case the gsi is associated to an MSI routing entry, the MSI
3142 message and device ID are translated into an LPI (support restricted
3143 to GICv3 ITS in-kernel emulation).
3144
3145 4.76 KVM_PPC_ALLOCATE_HTAB
3146 --------------------------
3147
3148 :Capability: KVM_CAP_PPC_ALLOC_HTAB
3149 :Architectures: powerpc
3150 :Type: vm ioctl
3151 :Parameters: Pointer to u32 containing hash table order (in/out)
3152 :Returns: 0 on success, -1 on error
3153
3154 This requests the host kernel to allocate an MMU hash table for a
3155 guest using the PAPR paravirtualization interface. This only does
3156 anything if the kernel is configured to use the Book 3S HV style of
3157 virtualization. Otherwise the capability doesn't exist and the ioctl
3158 returns an ENOTTY error. The rest of this description assumes Book 3S
3159 HV.
3160
3161 There must be no vcpus running when this ioctl is called; if there
3162 are, it will do nothing and return an EBUSY error.
3163
3164 The parameter is a pointer to a 32-bit unsigned integer variable
3165 containing the order (log base 2) of the desired size of the hash
3166 table, which must be between 18 and 46. On successful return from the
3167 ioctl, the value will not be changed by the kernel.
3168
3169 If no hash table has been allocated when any vcpu is asked to run
3170 (with the KVM_RUN ioctl), the host kernel will allocate a
3171 default-sized hash table (16 MB).
3172
3173 If this ioctl is called when a hash table has already been allocated,
3174 with a different order from the existing hash table, the existing hash
3175 table will be freed and a new one allocated. If this is ioctl is
3176 called when a hash table has already been allocated of the same order
3177 as specified, the kernel will clear out the existing hash table (zero
3178 all HPTEs). In either case, if the guest is using the virtualized
3179 real-mode area (VRMA) facility, the kernel will re-create the VMRA
3180 HPTEs on the next KVM_RUN of any vcpu.
3181
3182 4.77 KVM_S390_INTERRUPT
3183 -----------------------
3184
3185 :Capability: basic
3186 :Architectures: s390
3187 :Type: vm ioctl, vcpu ioctl
3188 :Parameters: struct kvm_s390_interrupt (in)
3189 :Returns: 0 on success, -1 on error
3190
3191 Allows to inject an interrupt to the guest. Interrupts can be floating
3192 (vm ioctl) or per cpu (vcpu ioctl), depending on the interrupt type.
3193
3194 Interrupt parameters are passed via kvm_s390_interrupt::
3195
3196 struct kvm_s390_interrupt {
3197 __u32 type;
3198 __u32 parm;
3199 __u64 parm64;
3200 };
3201
3202 type can be one of the following:
3203
3204 KVM_S390_SIGP_STOP (vcpu)
3205 - sigp stop; optional flags in parm
3206 KVM_S390_PROGRAM_INT (vcpu)
3207 - program check; code in parm
3208 KVM_S390_SIGP_SET_PREFIX (vcpu)
3209 - sigp set prefix; prefix address in parm
3210 KVM_S390_RESTART (vcpu)
3211 - restart
3212 KVM_S390_INT_CLOCK_COMP (vcpu)
3213 - clock comparator interrupt
3214 KVM_S390_INT_CPU_TIMER (vcpu)
3215 - CPU timer interrupt
3216 KVM_S390_INT_VIRTIO (vm)
3217 - virtio external interrupt; external interrupt
3218 parameters in parm and parm64
3219 KVM_S390_INT_SERVICE (vm)
3220 - sclp external interrupt; sclp parameter in parm
3221 KVM_S390_INT_EMERGENCY (vcpu)
3222 - sigp emergency; source cpu in parm
3223 KVM_S390_INT_EXTERNAL_CALL (vcpu)
3224 - sigp external call; source cpu in parm
3225 KVM_S390_INT_IO(ai,cssid,ssid,schid) (vm)
3226 - compound value to indicate an
3227 I/O interrupt (ai - adapter interrupt; cssid,ssid,schid - subchannel);
3228 I/O interruption parameters in parm (subchannel) and parm64 (intparm,
3229 interruption subclass)
3230 KVM_S390_MCHK (vm, vcpu)
3231 - machine check interrupt; cr 14 bits in parm, machine check interrupt
3232 code in parm64 (note that machine checks needing further payload are not
3233 supported by this ioctl)
3234
3235 This is an asynchronous vcpu ioctl and can be invoked from any thread.
3236
3237 4.78 KVM_PPC_GET_HTAB_FD
3238 ------------------------
3239
3240 :Capability: KVM_CAP_PPC_HTAB_FD
3241 :Architectures: powerpc
3242 :Type: vm ioctl
3243 :Parameters: Pointer to struct kvm_get_htab_fd (in)
3244 :Returns: file descriptor number (>= 0) on success, -1 on error
3245
3246 This returns a file descriptor that can be used either to read out the
3247 entries in the guest's hashed page table (HPT), or to write entries to
3248 initialize the HPT. The returned fd can only be written to if the
3249 KVM_GET_HTAB_WRITE bit is set in the flags field of the argument, and
3250 can only be read if that bit is clear. The argument struct looks like
3251 this::
3252
3253 /* For KVM_PPC_GET_HTAB_FD */
3254 struct kvm_get_htab_fd {
3255 __u64 flags;
3256 __u64 start_index;
3257 __u64 reserved[2];
3258 };
3259
3260 /* Values for kvm_get_htab_fd.flags */
3261 #define KVM_GET_HTAB_BOLTED_ONLY ((__u64)0x1)
3262 #define KVM_GET_HTAB_WRITE ((__u64)0x2)
3263
3264 The 'start_index' field gives the index in the HPT of the entry at
3265 which to start reading. It is ignored when writing.
3266
3267 Reads on the fd will initially supply information about all
3268 "interesting" HPT entries. Interesting entries are those with the
3269 bolted bit set, if the KVM_GET_HTAB_BOLTED_ONLY bit is set, otherwise
3270 all entries. When the end of the HPT is reached, the read() will
3271 return. If read() is called again on the fd, it will start again from
3272 the beginning of the HPT, but will only return HPT entries that have
3273 changed since they were last read.
3274
3275 Data read or written is structured as a header (8 bytes) followed by a
3276 series of valid HPT entries (16 bytes) each. The header indicates how
3277 many valid HPT entries there are and how many invalid entries follow
3278 the valid entries. The invalid entries are not represented explicitly
3279 in the stream. The header format is::
3280
3281 struct kvm_get_htab_header {
3282 __u32 index;
3283 __u16 n_valid;
3284 __u16 n_invalid;
3285 };
3286
3287 Writes to the fd create HPT entries starting at the index given in the
3288 header; first 'n_valid' valid entries with contents from the data
3289 written, then 'n_invalid' invalid entries, invalidating any previously
3290 valid entries found.
3291
3292 4.79 KVM_CREATE_DEVICE
3293 ----------------------
3294
3295 :Capability: KVM_CAP_DEVICE_CTRL
3296 :Type: vm ioctl
3297 :Parameters: struct kvm_create_device (in/out)
3298 :Returns: 0 on success, -1 on error
3299
3300 Errors:
3301
3302 ====== =======================================================
3303 ENODEV The device type is unknown or unsupported
3304 EEXIST Device already created, and this type of device may not
3305 be instantiated multiple times
3306 ====== =======================================================
3307
3308 Other error conditions may be defined by individual device types or
3309 have their standard meanings.
3310
3311 Creates an emulated device in the kernel. The file descriptor returned
3312 in fd can be used with KVM_SET/GET/HAS_DEVICE_ATTR.
3313
3314 If the KVM_CREATE_DEVICE_TEST flag is set, only test whether the
3315 device type is supported (not necessarily whether it can be created
3316 in the current vm).
3317
3318 Individual devices should not define flags. Attributes should be used
3319 for specifying any behavior that is not implied by the device type
3320 number.
3321
3322 ::
3323
3324 struct kvm_create_device {
3325 __u32 type; /* in: KVM_DEV_TYPE_xxx */
3326 __u32 fd; /* out: device handle */
3327 __u32 flags; /* in: KVM_CREATE_DEVICE_xxx */
3328 };
3329
3330 4.80 KVM_SET_DEVICE_ATTR/KVM_GET_DEVICE_ATTR
3331 --------------------------------------------
3332
3333 :Capability: KVM_CAP_DEVICE_CTRL, KVM_CAP_VM_ATTRIBUTES for vm device,
3334 KVM_CAP_VCPU_ATTRIBUTES for vcpu device
3335 KVM_CAP_SYS_ATTRIBUTES for system (/dev/kvm) device (no set)
3336 :Type: device ioctl, vm ioctl, vcpu ioctl
3337 :Parameters: struct kvm_device_attr
3338 :Returns: 0 on success, -1 on error
3339
3340 Errors:
3341
3342 ===== =============================================================
3343 ENXIO The group or attribute is unknown/unsupported for this device
3344 or hardware support is missing.
3345 EPERM The attribute cannot (currently) be accessed this way
3346 (e.g. read-only attribute, or attribute that only makes
3347 sense when the device is in a different state)
3348 ===== =============================================================
3349
3350 Other error conditions may be defined by individual device types.
3351
3352 Gets/sets a specified piece of device configuration and/or state. The
3353 semantics are device-specific. See individual device documentation in
3354 the "devices" directory. As with ONE_REG, the size of the data
3355 transferred is defined by the particular attribute.
3356
3357 ::
3358
3359 struct kvm_device_attr {
3360 __u32 flags; /* no flags currently defined */
3361 __u32 group; /* device-defined */
3362 __u64 attr; /* group-defined */
3363 __u64 addr; /* userspace address of attr data */
3364 };
3365
3366 4.81 KVM_HAS_DEVICE_ATTR
3367 ------------------------
3368
3369 :Capability: KVM_CAP_DEVICE_CTRL, KVM_CAP_VM_ATTRIBUTES for vm device,
3370 KVM_CAP_VCPU_ATTRIBUTES for vcpu device
3371 KVM_CAP_SYS_ATTRIBUTES for system (/dev/kvm) device
3372 :Type: device ioctl, vm ioctl, vcpu ioctl
3373 :Parameters: struct kvm_device_attr
3374 :Returns: 0 on success, -1 on error
3375
3376 Errors:
3377
3378 ===== =============================================================
3379 ENXIO The group or attribute is unknown/unsupported for this device
3380 or hardware support is missing.
3381 ===== =============================================================
3382
3383 Tests whether a device supports a particular attribute. A successful
3384 return indicates the attribute is implemented. It does not necessarily
3385 indicate that the attribute can be read or written in the device's
3386 current state. "addr" is ignored.
3387
3388 4.82 KVM_ARM_VCPU_INIT
3389 ----------------------
3390
3391 :Capability: basic
3392 :Architectures: arm64
3393 :Type: vcpu ioctl
3394 :Parameters: struct kvm_vcpu_init (in)
3395 :Returns: 0 on success; -1 on error
3396
3397 Errors:
3398
3399 ====== =================================================================
3400 EINVAL the target is unknown, or the combination of features is invalid.
3401 ENOENT a features bit specified is unknown.
3402 ====== =================================================================
3403
3404 This tells KVM what type of CPU to present to the guest, and what
3405 optional features it should have. This will cause a reset of the cpu
3406 registers to their initial values. If this is not called, KVM_RUN will
3407 return ENOEXEC for that vcpu.
3408
3409 The initial values are defined as:
3410 - Processor state:
3411 * AArch64: EL1h, D, A, I and F bits set. All other bits
3412 are cleared.
3413 * AArch32: SVC, A, I and F bits set. All other bits are
3414 cleared.
3415 - General Purpose registers, including PC and SP: set to 0
3416 - FPSIMD/NEON registers: set to 0
3417 - SVE registers: set to 0
3418 - System registers: Reset to their architecturally defined
3419 values as for a warm reset to EL1 (resp. SVC)
3420
3421 Note that because some registers reflect machine topology, all vcpus
3422 should be created before this ioctl is invoked.
3423
3424 Userspace can call this function multiple times for a given vcpu, including
3425 after the vcpu has been run. This will reset the vcpu to its initial
3426 state. All calls to this function after the initial call must use the same
3427 target and same set of feature flags, otherwise EINVAL will be returned.
3428
3429 Possible features:
3430
3431 - KVM_ARM_VCPU_POWER_OFF: Starts the CPU in a power-off state.
3432 Depends on KVM_CAP_ARM_PSCI. If not set, the CPU will be powered on
3433 and execute guest code when KVM_RUN is called.
3434 - KVM_ARM_VCPU_EL1_32BIT: Starts the CPU in a 32bit mode.
3435 Depends on KVM_CAP_ARM_EL1_32BIT (arm64 only).
3436 - KVM_ARM_VCPU_PSCI_0_2: Emulate PSCI v0.2 (or a future revision
3437 backward compatible with v0.2) for the CPU.
3438 Depends on KVM_CAP_ARM_PSCI_0_2.
3439 - KVM_ARM_VCPU_PMU_V3: Emulate PMUv3 for the CPU.
3440 Depends on KVM_CAP_ARM_PMU_V3.
3441
3442 - KVM_ARM_VCPU_PTRAUTH_ADDRESS: Enables Address Pointer authentication
3443 for arm64 only.
3444 Depends on KVM_CAP_ARM_PTRAUTH_ADDRESS.
3445 If KVM_CAP_ARM_PTRAUTH_ADDRESS and KVM_CAP_ARM_PTRAUTH_GENERIC are
3446 both present, then both KVM_ARM_VCPU_PTRAUTH_ADDRESS and
3447 KVM_ARM_VCPU_PTRAUTH_GENERIC must be requested or neither must be
3448 requested.
3449
3450 - KVM_ARM_VCPU_PTRAUTH_GENERIC: Enables Generic Pointer authentication
3451 for arm64 only.
3452 Depends on KVM_CAP_ARM_PTRAUTH_GENERIC.
3453 If KVM_CAP_ARM_PTRAUTH_ADDRESS and KVM_CAP_ARM_PTRAUTH_GENERIC are
3454 both present, then both KVM_ARM_VCPU_PTRAUTH_ADDRESS and
3455 KVM_ARM_VCPU_PTRAUTH_GENERIC must be requested or neither must be
3456 requested.
3457
3458 - KVM_ARM_VCPU_SVE: Enables SVE for the CPU (arm64 only).
3459 Depends on KVM_CAP_ARM_SVE.
3460 Requires KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE):
3461
3462 * After KVM_ARM_VCPU_INIT:
3463
3464 - KVM_REG_ARM64_SVE_VLS may be read using KVM_GET_ONE_REG: the
3465 initial value of this pseudo-register indicates the best set of
3466 vector lengths possible for a vcpu on this host.
3467
3468 * Before KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE):
3469
3470 - KVM_RUN and KVM_GET_REG_LIST are not available;
3471
3472 - KVM_GET_ONE_REG and KVM_SET_ONE_REG cannot be used to access
3473 the scalable archietctural SVE registers
3474 KVM_REG_ARM64_SVE_ZREG(), KVM_REG_ARM64_SVE_PREG() or
3475 KVM_REG_ARM64_SVE_FFR;
3476
3477 - KVM_REG_ARM64_SVE_VLS may optionally be written using
3478 KVM_SET_ONE_REG, to modify the set of vector lengths available
3479 for the vcpu.
3480
3481 * After KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE):
3482
3483 - the KVM_REG_ARM64_SVE_VLS pseudo-register is immutable, and can
3484 no longer be written using KVM_SET_ONE_REG.
3485
3486 4.83 KVM_ARM_PREFERRED_TARGET
3487 -----------------------------
3488
3489 :Capability: basic
3490 :Architectures: arm64
3491 :Type: vm ioctl
3492 :Parameters: struct kvm_vcpu_init (out)
3493 :Returns: 0 on success; -1 on error
3494
3495 Errors:
3496
3497 ====== ==========================================
3498 ENODEV no preferred target available for the host
3499 ====== ==========================================
3500
3501 This queries KVM for preferred CPU target type which can be emulated
3502 by KVM on underlying host.
3503
3504 The ioctl returns struct kvm_vcpu_init instance containing information
3505 about preferred CPU target type and recommended features for it. The
3506 kvm_vcpu_init->features bitmap returned will have feature bits set if
3507 the preferred target recommends setting these features, but this is
3508 not mandatory.
3509
3510 The information returned by this ioctl can be used to prepare an instance
3511 of struct kvm_vcpu_init for KVM_ARM_VCPU_INIT ioctl which will result in
3512 VCPU matching underlying host.
3513
3514
3515 4.84 KVM_GET_REG_LIST
3516 ---------------------
3517
3518 :Capability: basic
3519 :Architectures: arm64, mips
3520 :Type: vcpu ioctl
3521 :Parameters: struct kvm_reg_list (in/out)
3522 :Returns: 0 on success; -1 on error
3523
3524 Errors:
3525
3526 ===== ==============================================================
3527 E2BIG the reg index list is too big to fit in the array specified by
3528 the user (the number required will be written into n).
3529 ===== ==============================================================
3530
3531 ::
3532
3533 struct kvm_reg_list {
3534 __u64 n; /* number of registers in reg[] */
3535 __u64 reg[0];
3536 };
3537
3538 This ioctl returns the guest registers that are supported for the
3539 KVM_GET_ONE_REG/KVM_SET_ONE_REG calls.
3540
3541
3542 4.85 KVM_ARM_SET_DEVICE_ADDR (deprecated)
3543 -----------------------------------------
3544
3545 :Capability: KVM_CAP_ARM_SET_DEVICE_ADDR
3546 :Architectures: arm64
3547 :Type: vm ioctl
3548 :Parameters: struct kvm_arm_device_address (in)
3549 :Returns: 0 on success, -1 on error
3550
3551 Errors:
3552
3553 ====== ============================================
3554 ENODEV The device id is unknown
3555 ENXIO Device not supported on current system
3556 EEXIST Address already set
3557 E2BIG Address outside guest physical address space
3558 EBUSY Address overlaps with other device range
3559 ====== ============================================
3560
3561 ::
3562
3563 struct kvm_arm_device_addr {
3564 __u64 id;
3565 __u64 addr;
3566 };
3567
3568 Specify a device address in the guest's physical address space where guests
3569 can access emulated or directly exposed devices, which the host kernel needs
3570 to know about. The id field is an architecture specific identifier for a
3571 specific device.
3572
3573 arm64 divides the id field into two parts, a device id and an
3574 address type id specific to the individual device::
3575
3576 bits: | 63 ... 32 | 31 ... 16 | 15 ... 0 |
3577 field: | 0x00000000 | device id | addr type id |
3578
3579 arm64 currently only require this when using the in-kernel GIC
3580 support for the hardware VGIC features, using KVM_ARM_DEVICE_VGIC_V2
3581 as the device id. When setting the base address for the guest's
3582 mapping of the VGIC virtual CPU and distributor interface, the ioctl
3583 must be called after calling KVM_CREATE_IRQCHIP, but before calling
3584 KVM_RUN on any of the VCPUs. Calling this ioctl twice for any of the
3585 base addresses will return -EEXIST.
3586
3587 Note, this IOCTL is deprecated and the more flexible SET/GET_DEVICE_ATTR API
3588 should be used instead.
3589
3590
3591 4.86 KVM_PPC_RTAS_DEFINE_TOKEN
3592 ------------------------------
3593
3594 :Capability: KVM_CAP_PPC_RTAS
3595 :Architectures: ppc
3596 :Type: vm ioctl
3597 :Parameters: struct kvm_rtas_token_args
3598 :Returns: 0 on success, -1 on error
3599
3600 Defines a token value for a RTAS (Run Time Abstraction Services)
3601 service in order to allow it to be handled in the kernel. The
3602 argument struct gives the name of the service, which must be the name
3603 of a service that has a kernel-side implementation. If the token
3604 value is non-zero, it will be associated with that service, and
3605 subsequent RTAS calls by the guest specifying that token will be
3606 handled by the kernel. If the token value is 0, then any token
3607 associated with the service will be forgotten, and subsequent RTAS
3608 calls by the guest for that service will be passed to userspace to be
3609 handled.
3610
3611 4.87 KVM_SET_GUEST_DEBUG
3612 ------------------------
3613
3614 :Capability: KVM_CAP_SET_GUEST_DEBUG
3615 :Architectures: x86, s390, ppc, arm64
3616 :Type: vcpu ioctl
3617 :Parameters: struct kvm_guest_debug (in)
3618 :Returns: 0 on success; -1 on error
3619
3620 ::
3621
3622 struct kvm_guest_debug {
3623 __u32 control;
3624 __u32 pad;
3625 struct kvm_guest_debug_arch arch;
3626 };
3627
3628 Set up the processor specific debug registers and configure vcpu for
3629 handling guest debug events. There are two parts to the structure, the
3630 first a control bitfield indicates the type of debug events to handle
3631 when running. Common control bits are:
3632
3633 - KVM_GUESTDBG_ENABLE: guest debugging is enabled
3634 - KVM_GUESTDBG_SINGLESTEP: the next run should single-step
3635
3636 The top 16 bits of the control field are architecture specific control
3637 flags which can include the following:
3638
3639 - KVM_GUESTDBG_USE_SW_BP: using software breakpoints [x86, arm64]
3640 - KVM_GUESTDBG_USE_HW_BP: using hardware breakpoints [x86, s390]
3641 - KVM_GUESTDBG_USE_HW: using hardware debug events [arm64]
3642 - KVM_GUESTDBG_INJECT_DB: inject DB type exception [x86]
3643 - KVM_GUESTDBG_INJECT_BP: inject BP type exception [x86]
3644 - KVM_GUESTDBG_EXIT_PENDING: trigger an immediate guest exit [s390]
3645 - KVM_GUESTDBG_BLOCKIRQ: avoid injecting interrupts/NMI/SMI [x86]
3646
3647 For example KVM_GUESTDBG_USE_SW_BP indicates that software breakpoints
3648 are enabled in memory so we need to ensure breakpoint exceptions are
3649 correctly trapped and the KVM run loop exits at the breakpoint and not
3650 running off into the normal guest vector. For KVM_GUESTDBG_USE_HW_BP
3651 we need to ensure the guest vCPUs architecture specific registers are
3652 updated to the correct (supplied) values.
3653
3654 The second part of the structure is architecture specific and
3655 typically contains a set of debug registers.
3656
3657 For arm64 the number of debug registers is implementation defined and
3658 can be determined by querying the KVM_CAP_GUEST_DEBUG_HW_BPS and
3659 KVM_CAP_GUEST_DEBUG_HW_WPS capabilities which return a positive number
3660 indicating the number of supported registers.
3661
3662 For ppc, the KVM_CAP_PPC_GUEST_DEBUG_SSTEP capability indicates whether
3663 the single-step debug event (KVM_GUESTDBG_SINGLESTEP) is supported.
3664
3665 Also when supported, KVM_CAP_SET_GUEST_DEBUG2 capability indicates the
3666 supported KVM_GUESTDBG_* bits in the control field.
3667
3668 When debug events exit the main run loop with the reason
3669 KVM_EXIT_DEBUG with the kvm_debug_exit_arch part of the kvm_run
3670 structure containing architecture specific debug information.
3671
3672 4.88 KVM_GET_EMULATED_CPUID
3673 ---------------------------
3674
3675 :Capability: KVM_CAP_EXT_EMUL_CPUID
3676 :Architectures: x86
3677 :Type: system ioctl
3678 :Parameters: struct kvm_cpuid2 (in/out)
3679 :Returns: 0 on success, -1 on error
3680
3681 ::
3682
3683 struct kvm_cpuid2 {
3684 __u32 nent;
3685 __u32 flags;
3686 struct kvm_cpuid_entry2 entries[0];
3687 };
3688
3689 The member 'flags' is used for passing flags from userspace.
3690
3691 ::
3692
3693 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX BIT(0)
3694 #define KVM_CPUID_FLAG_STATEFUL_FUNC BIT(1) /* deprecated */
3695 #define KVM_CPUID_FLAG_STATE_READ_NEXT BIT(2) /* deprecated */
3696
3697 struct kvm_cpuid_entry2 {
3698 __u32 function;
3699 __u32 index;
3700 __u32 flags;
3701 __u32 eax;
3702 __u32 ebx;
3703 __u32 ecx;
3704 __u32 edx;
3705 __u32 padding[3];
3706 };
3707
3708 This ioctl returns x86 cpuid features which are emulated by
3709 kvm.Userspace can use the information returned by this ioctl to query
3710 which features are emulated by kvm instead of being present natively.
3711
3712 Userspace invokes KVM_GET_EMULATED_CPUID by passing a kvm_cpuid2
3713 structure with the 'nent' field indicating the number of entries in
3714 the variable-size array 'entries'. If the number of entries is too low
3715 to describe the cpu capabilities, an error (E2BIG) is returned. If the
3716 number is too high, the 'nent' field is adjusted and an error (ENOMEM)
3717 is returned. If the number is just right, the 'nent' field is adjusted
3718 to the number of valid entries in the 'entries' array, which is then
3719 filled.
3720
3721 The entries returned are the set CPUID bits of the respective features
3722 which kvm emulates, as returned by the CPUID instruction, with unknown
3723 or unsupported feature bits cleared.
3724
3725 Features like x2apic, for example, may not be present in the host cpu
3726 but are exposed by kvm in KVM_GET_SUPPORTED_CPUID because they can be
3727 emulated efficiently and thus not included here.
3728
3729 The fields in each entry are defined as follows:
3730
3731 function:
3732 the eax value used to obtain the entry
3733 index:
3734 the ecx value used to obtain the entry (for entries that are
3735 affected by ecx)
3736 flags:
3737 an OR of zero or more of the following:
3738
3739 KVM_CPUID_FLAG_SIGNIFCANT_INDEX:
3740 if the index field is valid
3741
3742 eax, ebx, ecx, edx:
3743
3744 the values returned by the cpuid instruction for
3745 this function/index combination
3746
3747 4.89 KVM_S390_MEM_OP
3748 --------------------
3749
3750 :Capability: KVM_CAP_S390_MEM_OP, KVM_CAP_S390_PROTECTED, KVM_CAP_S390_MEM_OP_EXTENSION
3751 :Architectures: s390
3752 :Type: vm ioctl, vcpu ioctl
3753 :Parameters: struct kvm_s390_mem_op (in)
3754 :Returns: = 0 on success,
3755 < 0 on generic error (e.g. -EFAULT or -ENOMEM),
3756 > 0 if an exception occurred while walking the page tables
3757
3758 Read or write data from/to the VM's memory.
3759 The KVM_CAP_S390_MEM_OP_EXTENSION capability specifies what functionality is
3760 supported.
3761
3762 Parameters are specified via the following structure::
3763
3764 struct kvm_s390_mem_op {
3765 __u64 gaddr; /* the guest address */
3766 __u64 flags; /* flags */
3767 __u32 size; /* amount of bytes */
3768 __u32 op; /* type of operation */
3769 __u64 buf; /* buffer in userspace */
3770 union {
3771 struct {
3772 __u8 ar; /* the access register number */
3773 __u8 key; /* access key, ignored if flag unset */
3774 };
3775 __u32 sida_offset; /* offset into the sida */
3776 __u8 reserved[32]; /* ignored */
3777 };
3778 };
3779
3780 The start address of the memory region has to be specified in the "gaddr"
3781 field, and the length of the region in the "size" field (which must not
3782 be 0). The maximum value for "size" can be obtained by checking the
3783 KVM_CAP_S390_MEM_OP capability. "buf" is the buffer supplied by the
3784 userspace application where the read data should be written to for
3785 a read access, or where the data that should be written is stored for
3786 a write access. The "reserved" field is meant for future extensions.
3787 Reserved and unused values are ignored. Future extension that add members must
3788 introduce new flags.
3789
3790 The type of operation is specified in the "op" field. Flags modifying
3791 their behavior can be set in the "flags" field. Undefined flag bits must
3792 be set to 0.
3793
3794 Possible operations are:
3795 * ``KVM_S390_MEMOP_LOGICAL_READ``
3796 * ``KVM_S390_MEMOP_LOGICAL_WRITE``
3797 * ``KVM_S390_MEMOP_ABSOLUTE_READ``
3798 * ``KVM_S390_MEMOP_ABSOLUTE_WRITE``
3799 * ``KVM_S390_MEMOP_SIDA_READ``
3800 * ``KVM_S390_MEMOP_SIDA_WRITE``
3801
3802 Logical read/write:
3803 ^^^^^^^^^^^^^^^^^^^
3804
3805 Access logical memory, i.e. translate the given guest address to an absolute
3806 address given the state of the VCPU and use the absolute address as target of
3807 the access. "ar" designates the access register number to be used; the valid
3808 range is 0..15.
3809 Logical accesses are permitted for the VCPU ioctl only.
3810 Logical accesses are permitted for non-protected guests only.
3811
3812 Supported flags:
3813 * ``KVM_S390_MEMOP_F_CHECK_ONLY``
3814 * ``KVM_S390_MEMOP_F_INJECT_EXCEPTION``
3815 * ``KVM_S390_MEMOP_F_SKEY_PROTECTION``
3816
3817 The KVM_S390_MEMOP_F_CHECK_ONLY flag can be set to check whether the
3818 corresponding memory access would cause an access exception; however,
3819 no actual access to the data in memory at the destination is performed.
3820 In this case, "buf" is unused and can be NULL.
3821
3822 In case an access exception occurred during the access (or would occur
3823 in case of KVM_S390_MEMOP_F_CHECK_ONLY), the ioctl returns a positive
3824 error number indicating the type of exception. This exception is also
3825 raised directly at the corresponding VCPU if the flag
3826 KVM_S390_MEMOP_F_INJECT_EXCEPTION is set.
3827 On protection exceptions, unless specified otherwise, the injected
3828 translation-exception identifier (TEID) indicates suppression.
3829
3830 If the KVM_S390_MEMOP_F_SKEY_PROTECTION flag is set, storage key
3831 protection is also in effect and may cause exceptions if accesses are
3832 prohibited given the access key designated by "key"; the valid range is 0..15.
3833 KVM_S390_MEMOP_F_SKEY_PROTECTION is available if KVM_CAP_S390_MEM_OP_EXTENSION
3834 is > 0.
3835 Since the accessed memory may span multiple pages and those pages might have
3836 different storage keys, it is possible that a protection exception occurs
3837 after memory has been modified. In this case, if the exception is injected,
3838 the TEID does not indicate suppression.
3839
3840 Absolute read/write:
3841 ^^^^^^^^^^^^^^^^^^^^
3842
3843 Access absolute memory. This operation is intended to be used with the
3844 KVM_S390_MEMOP_F_SKEY_PROTECTION flag, to allow accessing memory and performing
3845 the checks required for storage key protection as one operation (as opposed to
3846 user space getting the storage keys, performing the checks, and accessing
3847 memory thereafter, which could lead to a delay between check and access).
3848 Absolute accesses are permitted for the VM ioctl if KVM_CAP_S390_MEM_OP_EXTENSION
3849 is > 0.
3850 Currently absolute accesses are not permitted for VCPU ioctls.
3851 Absolute accesses are permitted for non-protected guests only.
3852
3853 Supported flags:
3854 * ``KVM_S390_MEMOP_F_CHECK_ONLY``
3855 * ``KVM_S390_MEMOP_F_SKEY_PROTECTION``
3856
3857 The semantics of the flags are as for logical accesses.
3858
3859 SIDA read/write:
3860 ^^^^^^^^^^^^^^^^
3861
3862 Access the secure instruction data area which contains memory operands necessary
3863 for instruction emulation for protected guests.
3864 SIDA accesses are available if the KVM_CAP_S390_PROTECTED capability is available.
3865 SIDA accesses are permitted for the VCPU ioctl only.
3866 SIDA accesses are permitted for protected guests only.
3867
3868 No flags are supported.
3869
3870 4.90 KVM_S390_GET_SKEYS
3871 -----------------------
3872
3873 :Capability: KVM_CAP_S390_SKEYS
3874 :Architectures: s390
3875 :Type: vm ioctl
3876 :Parameters: struct kvm_s390_skeys
3877 :Returns: 0 on success, KVM_S390_GET_SKEYS_NONE if guest is not using storage
3878 keys, negative value on error
3879
3880 This ioctl is used to get guest storage key values on the s390
3881 architecture. The ioctl takes parameters via the kvm_s390_skeys struct::
3882
3883 struct kvm_s390_skeys {
3884 __u64 start_gfn;
3885 __u64 count;
3886 __u64 skeydata_addr;
3887 __u32 flags;
3888 __u32 reserved[9];
3889 };
3890
3891 The start_gfn field is the number of the first guest frame whose storage keys
3892 you want to get.
3893
3894 The count field is the number of consecutive frames (starting from start_gfn)
3895 whose storage keys to get. The count field must be at least 1 and the maximum
3896 allowed value is defined as KVM_S390_SKEYS_MAX. Values outside this range
3897 will cause the ioctl to return -EINVAL.
3898
3899 The skeydata_addr field is the address to a buffer large enough to hold count
3900 bytes. This buffer will be filled with storage key data by the ioctl.
3901
3902 4.91 KVM_S390_SET_SKEYS
3903 -----------------------
3904
3905 :Capability: KVM_CAP_S390_SKEYS
3906 :Architectures: s390
3907 :Type: vm ioctl
3908 :Parameters: struct kvm_s390_skeys
3909 :Returns: 0 on success, negative value on error
3910
3911 This ioctl is used to set guest storage key values on the s390
3912 architecture. The ioctl takes parameters via the kvm_s390_skeys struct.
3913 See section on KVM_S390_GET_SKEYS for struct definition.
3914
3915 The start_gfn field is the number of the first guest frame whose storage keys
3916 you want to set.
3917
3918 The count field is the number of consecutive frames (starting from start_gfn)
3919 whose storage keys to get. The count field must be at least 1 and the maximum
3920 allowed value is defined as KVM_S390_SKEYS_MAX. Values outside this range
3921 will cause the ioctl to return -EINVAL.
3922
3923 The skeydata_addr field is the address to a buffer containing count bytes of
3924 storage keys. Each byte in the buffer will be set as the storage key for a
3925 single frame starting at start_gfn for count frames.
3926
3927 Note: If any architecturally invalid key value is found in the given data then
3928 the ioctl will return -EINVAL.
3929
3930 4.92 KVM_S390_IRQ
3931 -----------------
3932
3933 :Capability: KVM_CAP_S390_INJECT_IRQ
3934 :Architectures: s390
3935 :Type: vcpu ioctl
3936 :Parameters: struct kvm_s390_irq (in)
3937 :Returns: 0 on success, -1 on error
3938
3939 Errors:
3940
3941
3942 ====== =================================================================
3943 EINVAL interrupt type is invalid
3944 type is KVM_S390_SIGP_STOP and flag parameter is invalid value,
3945 type is KVM_S390_INT_EXTERNAL_CALL and code is bigger
3946 than the maximum of VCPUs
3947 EBUSY type is KVM_S390_SIGP_SET_PREFIX and vcpu is not stopped,
3948 type is KVM_S390_SIGP_STOP and a stop irq is already pending,
3949 type is KVM_S390_INT_EXTERNAL_CALL and an external call interrupt
3950 is already pending
3951 ====== =================================================================
3952
3953 Allows to inject an interrupt to the guest.
3954
3955 Using struct kvm_s390_irq as a parameter allows
3956 to inject additional payload which is not
3957 possible via KVM_S390_INTERRUPT.
3958
3959 Interrupt parameters are passed via kvm_s390_irq::
3960
3961 struct kvm_s390_irq {
3962 __u64 type;
3963 union {
3964 struct kvm_s390_io_info io;
3965 struct kvm_s390_ext_info ext;
3966 struct kvm_s390_pgm_info pgm;
3967 struct kvm_s390_emerg_info emerg;
3968 struct kvm_s390_extcall_info extcall;
3969 struct kvm_s390_prefix_info prefix;
3970 struct kvm_s390_stop_info stop;
3971 struct kvm_s390_mchk_info mchk;
3972 char reserved[64];
3973 } u;
3974 };
3975
3976 type can be one of the following:
3977
3978 - KVM_S390_SIGP_STOP - sigp stop; parameter in .stop
3979 - KVM_S390_PROGRAM_INT - program check; parameters in .pgm
3980 - KVM_S390_SIGP_SET_PREFIX - sigp set prefix; parameters in .prefix
3981 - KVM_S390_RESTART - restart; no parameters
3982 - KVM_S390_INT_CLOCK_COMP - clock comparator interrupt; no parameters
3983 - KVM_S390_INT_CPU_TIMER - CPU timer interrupt; no parameters
3984 - KVM_S390_INT_EMERGENCY - sigp emergency; parameters in .emerg
3985 - KVM_S390_INT_EXTERNAL_CALL - sigp external call; parameters in .extcall
3986 - KVM_S390_MCHK - machine check interrupt; parameters in .mchk
3987
3988 This is an asynchronous vcpu ioctl and can be invoked from any thread.
3989
3990 4.94 KVM_S390_GET_IRQ_STATE
3991 ---------------------------
3992
3993 :Capability: KVM_CAP_S390_IRQ_STATE
3994 :Architectures: s390
3995 :Type: vcpu ioctl
3996 :Parameters: struct kvm_s390_irq_state (out)
3997 :Returns: >= number of bytes copied into buffer,
3998 -EINVAL if buffer size is 0,
3999 -ENOBUFS if buffer size is too small to fit all pending interrupts,
4000 -EFAULT if the buffer address was invalid
4001
4002 This ioctl allows userspace to retrieve the complete state of all currently
4003 pending interrupts in a single buffer. Use cases include migration
4004 and introspection. The parameter structure contains the address of a
4005 userspace buffer and its length::
4006
4007 struct kvm_s390_irq_state {
4008 __u64 buf;
4009 __u32 flags; /* will stay unused for compatibility reasons */
4010 __u32 len;
4011 __u32 reserved[4]; /* will stay unused for compatibility reasons */
4012 };
4013
4014 Userspace passes in the above struct and for each pending interrupt a
4015 struct kvm_s390_irq is copied to the provided buffer.
4016
4017 The structure contains a flags and a reserved field for future extensions. As
4018 the kernel never checked for flags == 0 and QEMU never pre-zeroed flags and
4019 reserved, these fields can not be used in the future without breaking
4020 compatibility.
4021
4022 If -ENOBUFS is returned the buffer provided was too small and userspace
4023 may retry with a bigger buffer.
4024
4025 4.95 KVM_S390_SET_IRQ_STATE
4026 ---------------------------
4027
4028 :Capability: KVM_CAP_S390_IRQ_STATE
4029 :Architectures: s390
4030 :Type: vcpu ioctl
4031 :Parameters: struct kvm_s390_irq_state (in)
4032 :Returns: 0 on success,
4033 -EFAULT if the buffer address was invalid,
4034 -EINVAL for an invalid buffer length (see below),
4035 -EBUSY if there were already interrupts pending,
4036 errors occurring when actually injecting the
4037 interrupt. See KVM_S390_IRQ.
4038
4039 This ioctl allows userspace to set the complete state of all cpu-local
4040 interrupts currently pending for the vcpu. It is intended for restoring
4041 interrupt state after a migration. The input parameter is a userspace buffer
4042 containing a struct kvm_s390_irq_state::
4043
4044 struct kvm_s390_irq_state {
4045 __u64 buf;
4046 __u32 flags; /* will stay unused for compatibility reasons */
4047 __u32 len;
4048 __u32 reserved[4]; /* will stay unused for compatibility reasons */
4049 };
4050
4051 The restrictions for flags and reserved apply as well.
4052 (see KVM_S390_GET_IRQ_STATE)
4053
4054 The userspace memory referenced by buf contains a struct kvm_s390_irq
4055 for each interrupt to be injected into the guest.
4056 If one of the interrupts could not be injected for some reason the
4057 ioctl aborts.
4058
4059 len must be a multiple of sizeof(struct kvm_s390_irq). It must be > 0
4060 and it must not exceed (max_vcpus + 32) * sizeof(struct kvm_s390_irq),
4061 which is the maximum number of possibly pending cpu-local interrupts.
4062
4063 4.96 KVM_SMI
4064 ------------
4065
4066 :Capability: KVM_CAP_X86_SMM
4067 :Architectures: x86
4068 :Type: vcpu ioctl
4069 :Parameters: none
4070 :Returns: 0 on success, -1 on error
4071
4072 Queues an SMI on the thread's vcpu.
4073
4074 4.97 KVM_X86_SET_MSR_FILTER
4075 ----------------------------
4076
4077 :Capability: KVM_X86_SET_MSR_FILTER
4078 :Architectures: x86
4079 :Type: vm ioctl
4080 :Parameters: struct kvm_msr_filter
4081 :Returns: 0 on success, < 0 on error
4082
4083 ::
4084
4085 struct kvm_msr_filter_range {
4086 #define KVM_MSR_FILTER_READ (1 << 0)
4087 #define KVM_MSR_FILTER_WRITE (1 << 1)
4088 __u32 flags;
4089 __u32 nmsrs; /* number of msrs in bitmap */
4090 __u32 base; /* MSR index the bitmap starts at */
4091 __u8 *bitmap; /* a 1 bit allows the operations in flags, 0 denies */
4092 };
4093
4094 #define KVM_MSR_FILTER_MAX_RANGES 16
4095 struct kvm_msr_filter {
4096 #define KVM_MSR_FILTER_DEFAULT_ALLOW (0 << 0)
4097 #define KVM_MSR_FILTER_DEFAULT_DENY (1 << 0)
4098 __u32 flags;
4099 struct kvm_msr_filter_range ranges[KVM_MSR_FILTER_MAX_RANGES];
4100 };
4101
4102 flags values for ``struct kvm_msr_filter_range``:
4103
4104 ``KVM_MSR_FILTER_READ``
4105
4106 Filter read accesses to MSRs using the given bitmap. A 0 in the bitmap
4107 indicates that a read should immediately fail, while a 1 indicates that
4108 a read for a particular MSR should be handled regardless of the default
4109 filter action.
4110
4111 ``KVM_MSR_FILTER_WRITE``
4112
4113 Filter write accesses to MSRs using the given bitmap. A 0 in the bitmap
4114 indicates that a write should immediately fail, while a 1 indicates that
4115 a write for a particular MSR should be handled regardless of the default
4116 filter action.
4117
4118 ``KVM_MSR_FILTER_READ | KVM_MSR_FILTER_WRITE``
4119
4120 Filter both read and write accesses to MSRs using the given bitmap. A 0
4121 in the bitmap indicates that both reads and writes should immediately fail,
4122 while a 1 indicates that reads and writes for a particular MSR are not
4123 filtered by this range.
4124
4125 flags values for ``struct kvm_msr_filter``:
4126
4127 ``KVM_MSR_FILTER_DEFAULT_ALLOW``
4128
4129 If no filter range matches an MSR index that is getting accessed, KVM will
4130 fall back to allowing access to the MSR.
4131
4132 ``KVM_MSR_FILTER_DEFAULT_DENY``
4133
4134 If no filter range matches an MSR index that is getting accessed, KVM will
4135 fall back to rejecting access to the MSR. In this mode, all MSRs that should
4136 be processed by KVM need to explicitly be marked as allowed in the bitmaps.
4137
4138 This ioctl allows user space to define up to 16 bitmaps of MSR ranges to
4139 specify whether a certain MSR access should be explicitly filtered for or not.
4140
4141 If this ioctl has never been invoked, MSR accesses are not guarded and the
4142 default KVM in-kernel emulation behavior is fully preserved.
4143
4144 Calling this ioctl with an empty set of ranges (all nmsrs == 0) disables MSR
4145 filtering. In that mode, ``KVM_MSR_FILTER_DEFAULT_DENY`` is invalid and causes
4146 an error.
4147
4148 As soon as the filtering is in place, every MSR access is processed through
4149 the filtering except for accesses to the x2APIC MSRs (from 0x800 to 0x8ff);
4150 x2APIC MSRs are always allowed, independent of the ``default_allow`` setting,
4151 and their behavior depends on the ``X2APIC_ENABLE`` bit of the APIC base
4152 register.
4153
4154 .. warning::
4155 MSR accesses coming from nested vmentry/vmexit are not filtered.
4156 This includes both writes to individual VMCS fields and reads/writes
4157 through the MSR lists pointed to by the VMCS.
4158
4159 If a bit is within one of the defined ranges, read and write accesses are
4160 guarded by the bitmap's value for the MSR index if the kind of access
4161 is included in the ``struct kvm_msr_filter_range`` flags. If no range
4162 cover this particular access, the behavior is determined by the flags
4163 field in the kvm_msr_filter struct: ``KVM_MSR_FILTER_DEFAULT_ALLOW``
4164 and ``KVM_MSR_FILTER_DEFAULT_DENY``.
4165
4166 Each bitmap range specifies a range of MSRs to potentially allow access on.
4167 The range goes from MSR index [base .. base+nmsrs]. The flags field
4168 indicates whether reads, writes or both reads and writes are filtered
4169 by setting a 1 bit in the bitmap for the corresponding MSR index.
4170
4171 If an MSR access is not permitted through the filtering, it generates a
4172 #GP inside the guest. When combined with KVM_CAP_X86_USER_SPACE_MSR, that
4173 allows user space to deflect and potentially handle various MSR accesses
4174 into user space.
4175
4176 If a vCPU is in running state while this ioctl is invoked, the vCPU may
4177 experience inconsistent filtering behavior on MSR accesses.
4178
4179 4.98 KVM_CREATE_SPAPR_TCE_64
4180 ----------------------------
4181
4182 :Capability: KVM_CAP_SPAPR_TCE_64
4183 :Architectures: powerpc
4184 :Type: vm ioctl
4185 :Parameters: struct kvm_create_spapr_tce_64 (in)
4186 :Returns: file descriptor for manipulating the created TCE table
4187
4188 This is an extension for KVM_CAP_SPAPR_TCE which only supports 32bit
4189 windows, described in 4.62 KVM_CREATE_SPAPR_TCE
4190
4191 This capability uses extended struct in ioctl interface::
4192
4193 /* for KVM_CAP_SPAPR_TCE_64 */
4194 struct kvm_create_spapr_tce_64 {
4195 __u64 liobn;
4196 __u32 page_shift;
4197 __u32 flags;
4198 __u64 offset; /* in pages */
4199 __u64 size; /* in pages */
4200 };
4201
4202 The aim of extension is to support an additional bigger DMA window with
4203 a variable page size.
4204 KVM_CREATE_SPAPR_TCE_64 receives a 64bit window size, an IOMMU page shift and
4205 a bus offset of the corresponding DMA window, @size and @offset are numbers
4206 of IOMMU pages.
4207
4208 @flags are not used at the moment.
4209
4210 The rest of functionality is identical to KVM_CREATE_SPAPR_TCE.
4211
4212 4.99 KVM_REINJECT_CONTROL
4213 -------------------------
4214
4215 :Capability: KVM_CAP_REINJECT_CONTROL
4216 :Architectures: x86
4217 :Type: vm ioctl
4218 :Parameters: struct kvm_reinject_control (in)
4219 :Returns: 0 on success,
4220 -EFAULT if struct kvm_reinject_control cannot be read,
4221 -ENXIO if KVM_CREATE_PIT or KVM_CREATE_PIT2 didn't succeed earlier.
4222
4223 i8254 (PIT) has two modes, reinject and !reinject. The default is reinject,
4224 where KVM queues elapsed i8254 ticks and monitors completion of interrupt from
4225 vector(s) that i8254 injects. Reinject mode dequeues a tick and injects its
4226 interrupt whenever there isn't a pending interrupt from i8254.
4227 !reinject mode injects an interrupt as soon as a tick arrives.
4228
4229 ::
4230
4231 struct kvm_reinject_control {
4232 __u8 pit_reinject;
4233 __u8 reserved[31];
4234 };
4235
4236 pit_reinject = 0 (!reinject mode) is recommended, unless running an old
4237 operating system that uses the PIT for timing (e.g. Linux 2.4.x).
4238
4239 4.100 KVM_PPC_CONFIGURE_V3_MMU
4240 ------------------------------
4241
4242 :Capability: KVM_CAP_PPC_RADIX_MMU or KVM_CAP_PPC_HASH_MMU_V3
4243 :Architectures: ppc
4244 :Type: vm ioctl
4245 :Parameters: struct kvm_ppc_mmuv3_cfg (in)
4246 :Returns: 0 on success,
4247 -EFAULT if struct kvm_ppc_mmuv3_cfg cannot be read,
4248 -EINVAL if the configuration is invalid
4249
4250 This ioctl controls whether the guest will use radix or HPT (hashed
4251 page table) translation, and sets the pointer to the process table for
4252 the guest.
4253
4254 ::
4255
4256 struct kvm_ppc_mmuv3_cfg {
4257 __u64 flags;
4258 __u64 process_table;
4259 };
4260
4261 There are two bits that can be set in flags; KVM_PPC_MMUV3_RADIX and
4262 KVM_PPC_MMUV3_GTSE. KVM_PPC_MMUV3_RADIX, if set, configures the guest
4263 to use radix tree translation, and if clear, to use HPT translation.
4264 KVM_PPC_MMUV3_GTSE, if set and if KVM permits it, configures the guest
4265 to be able to use the global TLB and SLB invalidation instructions;
4266 if clear, the guest may not use these instructions.
4267
4268 The process_table field specifies the address and size of the guest
4269 process table, which is in the guest's space. This field is formatted
4270 as the second doubleword of the partition table entry, as defined in
4271 the Power ISA V3.00, Book III section 5.7.6.1.
4272
4273 4.101 KVM_PPC_GET_RMMU_INFO
4274 ---------------------------
4275
4276 :Capability: KVM_CAP_PPC_RADIX_MMU
4277 :Architectures: ppc
4278 :Type: vm ioctl
4279 :Parameters: struct kvm_ppc_rmmu_info (out)
4280 :Returns: 0 on success,
4281 -EFAULT if struct kvm_ppc_rmmu_info cannot be written,
4282 -EINVAL if no useful information can be returned
4283
4284 This ioctl returns a structure containing two things: (a) a list
4285 containing supported radix tree geometries, and (b) a list that maps
4286 page sizes to put in the "AP" (actual page size) field for the tlbie
4287 (TLB invalidate entry) instruction.
4288
4289 ::
4290
4291 struct kvm_ppc_rmmu_info {
4292 struct kvm_ppc_radix_geom {
4293 __u8 page_shift;
4294 __u8 level_bits[4];
4295 __u8 pad[3];
4296 } geometries[8];
4297 __u32 ap_encodings[8];
4298 };
4299
4300 The geometries[] field gives up to 8 supported geometries for the
4301 radix page table, in terms of the log base 2 of the smallest page
4302 size, and the number of bits indexed at each level of the tree, from
4303 the PTE level up to the PGD level in that order. Any unused entries
4304 will have 0 in the page_shift field.
4305
4306 The ap_encodings gives the supported page sizes and their AP field
4307 encodings, encoded with the AP value in the top 3 bits and the log
4308 base 2 of the page size in the bottom 6 bits.
4309
4310 4.102 KVM_PPC_RESIZE_HPT_PREPARE
4311 --------------------------------
4312
4313 :Capability: KVM_CAP_SPAPR_RESIZE_HPT
4314 :Architectures: powerpc
4315 :Type: vm ioctl
4316 :Parameters: struct kvm_ppc_resize_hpt (in)
4317 :Returns: 0 on successful completion,
4318 >0 if a new HPT is being prepared, the value is an estimated
4319 number of milliseconds until preparation is complete,
4320 -EFAULT if struct kvm_reinject_control cannot be read,
4321 -EINVAL if the supplied shift or flags are invalid,
4322 -ENOMEM if unable to allocate the new HPT,
4323
4324 Used to implement the PAPR extension for runtime resizing of a guest's
4325 Hashed Page Table (HPT). Specifically this starts, stops or monitors
4326 the preparation of a new potential HPT for the guest, essentially
4327 implementing the H_RESIZE_HPT_PREPARE hypercall.
4328
4329 ::
4330
4331 struct kvm_ppc_resize_hpt {
4332 __u64 flags;
4333 __u32 shift;
4334 __u32 pad;
4335 };
4336
4337 If called with shift > 0 when there is no pending HPT for the guest,
4338 this begins preparation of a new pending HPT of size 2^(shift) bytes.
4339 It then returns a positive integer with the estimated number of
4340 milliseconds until preparation is complete.
4341
4342 If called when there is a pending HPT whose size does not match that
4343 requested in the parameters, discards the existing pending HPT and
4344 creates a new one as above.
4345
4346 If called when there is a pending HPT of the size requested, will:
4347
4348 * If preparation of the pending HPT is already complete, return 0
4349 * If preparation of the pending HPT has failed, return an error
4350 code, then discard the pending HPT.
4351 * If preparation of the pending HPT is still in progress, return an
4352 estimated number of milliseconds until preparation is complete.
4353
4354 If called with shift == 0, discards any currently pending HPT and
4355 returns 0 (i.e. cancels any in-progress preparation).
4356
4357 flags is reserved for future expansion, currently setting any bits in
4358 flags will result in an -EINVAL.
4359
4360 Normally this will be called repeatedly with the same parameters until
4361 it returns <= 0. The first call will initiate preparation, subsequent
4362 ones will monitor preparation until it completes or fails.
4363
4364 4.103 KVM_PPC_RESIZE_HPT_COMMIT
4365 -------------------------------
4366
4367 :Capability: KVM_CAP_SPAPR_RESIZE_HPT
4368 :Architectures: powerpc
4369 :Type: vm ioctl
4370 :Parameters: struct kvm_ppc_resize_hpt (in)
4371 :Returns: 0 on successful completion,
4372 -EFAULT if struct kvm_reinject_control cannot be read,
4373 -EINVAL if the supplied shift or flags are invalid,
4374 -ENXIO is there is no pending HPT, or the pending HPT doesn't
4375 have the requested size,
4376 -EBUSY if the pending HPT is not fully prepared,
4377 -ENOSPC if there was a hash collision when moving existing
4378 HPT entries to the new HPT,
4379 -EIO on other error conditions
4380
4381 Used to implement the PAPR extension for runtime resizing of a guest's
4382 Hashed Page Table (HPT). Specifically this requests that the guest be
4383 transferred to working with the new HPT, essentially implementing the
4384 H_RESIZE_HPT_COMMIT hypercall.
4385
4386 ::
4387
4388 struct kvm_ppc_resize_hpt {
4389 __u64 flags;
4390 __u32 shift;
4391 __u32 pad;
4392 };
4393
4394 This should only be called after KVM_PPC_RESIZE_HPT_PREPARE has
4395 returned 0 with the same parameters. In other cases
4396 KVM_PPC_RESIZE_HPT_COMMIT will return an error (usually -ENXIO or
4397 -EBUSY, though others may be possible if the preparation was started,
4398 but failed).
4399
4400 This will have undefined effects on the guest if it has not already
4401 placed itself in a quiescent state where no vcpu will make MMU enabled
4402 memory accesses.
4403
4404 On succsful completion, the pending HPT will become the guest's active
4405 HPT and the previous HPT will be discarded.
4406
4407 On failure, the guest will still be operating on its previous HPT.
4408
4409 4.104 KVM_X86_GET_MCE_CAP_SUPPORTED
4410 -----------------------------------
4411
4412 :Capability: KVM_CAP_MCE
4413 :Architectures: x86
4414 :Type: system ioctl
4415 :Parameters: u64 mce_cap (out)
4416 :Returns: 0 on success, -1 on error
4417
4418 Returns supported MCE capabilities. The u64 mce_cap parameter
4419 has the same format as the MSR_IA32_MCG_CAP register. Supported
4420 capabilities will have the corresponding bits set.
4421
4422 4.105 KVM_X86_SETUP_MCE
4423 -----------------------
4424
4425 :Capability: KVM_CAP_MCE
4426 :Architectures: x86
4427 :Type: vcpu ioctl
4428 :Parameters: u64 mcg_cap (in)
4429 :Returns: 0 on success,
4430 -EFAULT if u64 mcg_cap cannot be read,
4431 -EINVAL if the requested number of banks is invalid,
4432 -EINVAL if requested MCE capability is not supported.
4433
4434 Initializes MCE support for use. The u64 mcg_cap parameter
4435 has the same format as the MSR_IA32_MCG_CAP register and
4436 specifies which capabilities should be enabled. The maximum
4437 supported number of error-reporting banks can be retrieved when
4438 checking for KVM_CAP_MCE. The supported capabilities can be
4439 retrieved with KVM_X86_GET_MCE_CAP_SUPPORTED.
4440
4441 4.106 KVM_X86_SET_MCE
4442 ---------------------
4443
4444 :Capability: KVM_CAP_MCE
4445 :Architectures: x86
4446 :Type: vcpu ioctl
4447 :Parameters: struct kvm_x86_mce (in)
4448 :Returns: 0 on success,
4449 -EFAULT if struct kvm_x86_mce cannot be read,
4450 -EINVAL if the bank number is invalid,
4451 -EINVAL if VAL bit is not set in status field.
4452
4453 Inject a machine check error (MCE) into the guest. The input
4454 parameter is::
4455
4456 struct kvm_x86_mce {
4457 __u64 status;
4458 __u64 addr;
4459 __u64 misc;
4460 __u64 mcg_status;
4461 __u8 bank;
4462 __u8 pad1[7];
4463 __u64 pad2[3];
4464 };
4465
4466 If the MCE being reported is an uncorrected error, KVM will
4467 inject it as an MCE exception into the guest. If the guest
4468 MCG_STATUS register reports that an MCE is in progress, KVM
4469 causes an KVM_EXIT_SHUTDOWN vmexit.
4470
4471 Otherwise, if the MCE is a corrected error, KVM will just
4472 store it in the corresponding bank (provided this bank is
4473 not holding a previously reported uncorrected error).
4474
4475 4.107 KVM_S390_GET_CMMA_BITS
4476 ----------------------------
4477
4478 :Capability: KVM_CAP_S390_CMMA_MIGRATION
4479 :Architectures: s390
4480 :Type: vm ioctl
4481 :Parameters: struct kvm_s390_cmma_log (in, out)
4482 :Returns: 0 on success, a negative value on error
4483
4484 This ioctl is used to get the values of the CMMA bits on the s390
4485 architecture. It is meant to be used in two scenarios:
4486
4487 - During live migration to save the CMMA values. Live migration needs
4488 to be enabled via the KVM_REQ_START_MIGRATION VM property.
4489 - To non-destructively peek at the CMMA values, with the flag
4490 KVM_S390_CMMA_PEEK set.
4491
4492 The ioctl takes parameters via the kvm_s390_cmma_log struct. The desired
4493 values are written to a buffer whose location is indicated via the "values"
4494 member in the kvm_s390_cmma_log struct. The values in the input struct are
4495 also updated as needed.
4496
4497 Each CMMA value takes up one byte.
4498
4499 ::
4500
4501 struct kvm_s390_cmma_log {
4502 __u64 start_gfn;
4503 __u32 count;
4504 __u32 flags;
4505 union {
4506 __u64 remaining;
4507 __u64 mask;
4508 };
4509 __u64 values;
4510 };
4511
4512 start_gfn is the number of the first guest frame whose CMMA values are
4513 to be retrieved,
4514
4515 count is the length of the buffer in bytes,
4516
4517 values points to the buffer where the result will be written to.
4518
4519 If count is greater than KVM_S390_SKEYS_MAX, then it is considered to be
4520 KVM_S390_SKEYS_MAX. KVM_S390_SKEYS_MAX is re-used for consistency with
4521 other ioctls.
4522
4523 The result is written in the buffer pointed to by the field values, and
4524 the values of the input parameter are updated as follows.
4525
4526 Depending on the flags, different actions are performed. The only
4527 supported flag so far is KVM_S390_CMMA_PEEK.
4528
4529 The default behaviour if KVM_S390_CMMA_PEEK is not set is:
4530 start_gfn will indicate the first page frame whose CMMA bits were dirty.
4531 It is not necessarily the same as the one passed as input, as clean pages
4532 are skipped.
4533
4534 count will indicate the number of bytes actually written in the buffer.
4535 It can (and very often will) be smaller than the input value, since the
4536 buffer is only filled until 16 bytes of clean values are found (which
4537 are then not copied in the buffer). Since a CMMA migration block needs
4538 the base address and the length, for a total of 16 bytes, we will send
4539 back some clean data if there is some dirty data afterwards, as long as
4540 the size of the clean data does not exceed the size of the header. This
4541 allows to minimize the amount of data to be saved or transferred over
4542 the network at the expense of more roundtrips to userspace. The next
4543 invocation of the ioctl will skip over all the clean values, saving
4544 potentially more than just the 16 bytes we found.
4545
4546 If KVM_S390_CMMA_PEEK is set:
4547 the existing storage attributes are read even when not in migration
4548 mode, and no other action is performed;
4549
4550 the output start_gfn will be equal to the input start_gfn,
4551
4552 the output count will be equal to the input count, except if the end of
4553 memory has been reached.
4554
4555 In both cases:
4556 the field "remaining" will indicate the total number of dirty CMMA values
4557 still remaining, or 0 if KVM_S390_CMMA_PEEK is set and migration mode is
4558 not enabled.
4559
4560 mask is unused.
4561
4562 values points to the userspace buffer where the result will be stored.
4563
4564 This ioctl can fail with -ENOMEM if not enough memory can be allocated to
4565 complete the task, with -ENXIO if CMMA is not enabled, with -EINVAL if
4566 KVM_S390_CMMA_PEEK is not set but migration mode was not enabled, with
4567 -EFAULT if the userspace address is invalid or if no page table is
4568 present for the addresses (e.g. when using hugepages).
4569
4570 4.108 KVM_S390_SET_CMMA_BITS
4571 ----------------------------
4572
4573 :Capability: KVM_CAP_S390_CMMA_MIGRATION
4574 :Architectures: s390
4575 :Type: vm ioctl
4576 :Parameters: struct kvm_s390_cmma_log (in)
4577 :Returns: 0 on success, a negative value on error
4578
4579 This ioctl is used to set the values of the CMMA bits on the s390
4580 architecture. It is meant to be used during live migration to restore
4581 the CMMA values, but there are no restrictions on its use.
4582 The ioctl takes parameters via the kvm_s390_cmma_values struct.
4583 Each CMMA value takes up one byte.
4584
4585 ::
4586
4587 struct kvm_s390_cmma_log {
4588 __u64 start_gfn;
4589 __u32 count;
4590 __u32 flags;
4591 union {
4592 __u64 remaining;
4593 __u64 mask;
4594 };
4595 __u64 values;
4596 };
4597
4598 start_gfn indicates the starting guest frame number,
4599
4600 count indicates how many values are to be considered in the buffer,
4601
4602 flags is not used and must be 0.
4603
4604 mask indicates which PGSTE bits are to be considered.
4605
4606 remaining is not used.
4607
4608 values points to the buffer in userspace where to store the values.
4609
4610 This ioctl can fail with -ENOMEM if not enough memory can be allocated to
4611 complete the task, with -ENXIO if CMMA is not enabled, with -EINVAL if
4612 the count field is too large (e.g. more than KVM_S390_CMMA_SIZE_MAX) or
4613 if the flags field was not 0, with -EFAULT if the userspace address is
4614 invalid, if invalid pages are written to (e.g. after the end of memory)
4615 or if no page table is present for the addresses (e.g. when using
4616 hugepages).
4617
4618 4.109 KVM_PPC_GET_CPU_CHAR
4619 --------------------------
4620
4621 :Capability: KVM_CAP_PPC_GET_CPU_CHAR
4622 :Architectures: powerpc
4623 :Type: vm ioctl
4624 :Parameters: struct kvm_ppc_cpu_char (out)
4625 :Returns: 0 on successful completion,
4626 -EFAULT if struct kvm_ppc_cpu_char cannot be written
4627
4628 This ioctl gives userspace information about certain characteristics
4629 of the CPU relating to speculative execution of instructions and
4630 possible information leakage resulting from speculative execution (see
4631 CVE-2017-5715, CVE-2017-5753 and CVE-2017-5754). The information is
4632 returned in struct kvm_ppc_cpu_char, which looks like this::
4633
4634 struct kvm_ppc_cpu_char {
4635 __u64 character; /* characteristics of the CPU */
4636 __u64 behaviour; /* recommended software behaviour */
4637 __u64 character_mask; /* valid bits in character */
4638 __u64 behaviour_mask; /* valid bits in behaviour */
4639 };
4640
4641 For extensibility, the character_mask and behaviour_mask fields
4642 indicate which bits of character and behaviour have been filled in by
4643 the kernel. If the set of defined bits is extended in future then
4644 userspace will be able to tell whether it is running on a kernel that
4645 knows about the new bits.
4646
4647 The character field describes attributes of the CPU which can help
4648 with preventing inadvertent information disclosure - specifically,
4649 whether there is an instruction to flash-invalidate the L1 data cache
4650 (ori 30,30,0 or mtspr SPRN_TRIG2,rN), whether the L1 data cache is set
4651 to a mode where entries can only be used by the thread that created
4652 them, whether the bcctr[l] instruction prevents speculation, and
4653 whether a speculation barrier instruction (ori 31,31,0) is provided.
4654
4655 The behaviour field describes actions that software should take to
4656 prevent inadvertent information disclosure, and thus describes which
4657 vulnerabilities the hardware is subject to; specifically whether the
4658 L1 data cache should be flushed when returning to user mode from the
4659 kernel, and whether a speculation barrier should be placed between an
4660 array bounds check and the array access.
4661
4662 These fields use the same bit definitions as the new
4663 H_GET_CPU_CHARACTERISTICS hypercall.
4664
4665 4.110 KVM_MEMORY_ENCRYPT_OP
4666 ---------------------------
4667
4668 :Capability: basic
4669 :Architectures: x86
4670 :Type: vm
4671 :Parameters: an opaque platform specific structure (in/out)
4672 :Returns: 0 on success; -1 on error
4673
4674 If the platform supports creating encrypted VMs then this ioctl can be used
4675 for issuing platform-specific memory encryption commands to manage those
4676 encrypted VMs.
4677
4678 Currently, this ioctl is used for issuing Secure Encrypted Virtualization
4679 (SEV) commands on AMD Processors. The SEV commands are defined in
4680 Documentation/virt/kvm/x86/amd-memory-encryption.rst.
4681
4682 4.111 KVM_MEMORY_ENCRYPT_REG_REGION
4683 -----------------------------------
4684
4685 :Capability: basic
4686 :Architectures: x86
4687 :Type: system
4688 :Parameters: struct kvm_enc_region (in)
4689 :Returns: 0 on success; -1 on error
4690
4691 This ioctl can be used to register a guest memory region which may
4692 contain encrypted data (e.g. guest RAM, SMRAM etc).
4693
4694 It is used in the SEV-enabled guest. When encryption is enabled, a guest
4695 memory region may contain encrypted data. The SEV memory encryption
4696 engine uses a tweak such that two identical plaintext pages, each at
4697 different locations will have differing ciphertexts. So swapping or
4698 moving ciphertext of those pages will not result in plaintext being
4699 swapped. So relocating (or migrating) physical backing pages for the SEV
4700 guest will require some additional steps.
4701
4702 Note: The current SEV key management spec does not provide commands to
4703 swap or migrate (move) ciphertext pages. Hence, for now we pin the guest
4704 memory region registered with the ioctl.
4705
4706 4.112 KVM_MEMORY_ENCRYPT_UNREG_REGION
4707 -------------------------------------
4708
4709 :Capability: basic
4710 :Architectures: x86
4711 :Type: system
4712 :Parameters: struct kvm_enc_region (in)
4713 :Returns: 0 on success; -1 on error
4714
4715 This ioctl can be used to unregister the guest memory region registered
4716 with KVM_MEMORY_ENCRYPT_REG_REGION ioctl above.
4717
4718 4.113 KVM_HYPERV_EVENTFD
4719 ------------------------
4720
4721 :Capability: KVM_CAP_HYPERV_EVENTFD
4722 :Architectures: x86
4723 :Type: vm ioctl
4724 :Parameters: struct kvm_hyperv_eventfd (in)
4725
4726 This ioctl (un)registers an eventfd to receive notifications from the guest on
4727 the specified Hyper-V connection id through the SIGNAL_EVENT hypercall, without
4728 causing a user exit. SIGNAL_EVENT hypercall with non-zero event flag number
4729 (bits 24-31) still triggers a KVM_EXIT_HYPERV_HCALL user exit.
4730
4731 ::
4732
4733 struct kvm_hyperv_eventfd {
4734 __u32 conn_id;
4735 __s32 fd;
4736 __u32 flags;
4737 __u32 padding[3];
4738 };
4739
4740 The conn_id field should fit within 24 bits::
4741
4742 #define KVM_HYPERV_CONN_ID_MASK 0x00ffffff
4743
4744 The acceptable values for the flags field are::
4745
4746 #define KVM_HYPERV_EVENTFD_DEASSIGN (1 << 0)
4747
4748 :Returns: 0 on success,
4749 -EINVAL if conn_id or flags is outside the allowed range,
4750 -ENOENT on deassign if the conn_id isn't registered,
4751 -EEXIST on assign if the conn_id is already registered
4752
4753 4.114 KVM_GET_NESTED_STATE
4754 --------------------------
4755
4756 :Capability: KVM_CAP_NESTED_STATE
4757 :Architectures: x86
4758 :Type: vcpu ioctl
4759 :Parameters: struct kvm_nested_state (in/out)
4760 :Returns: 0 on success, -1 on error
4761
4762 Errors:
4763
4764 ===== =============================================================
4765 E2BIG the total state size exceeds the value of 'size' specified by
4766 the user; the size required will be written into size.
4767 ===== =============================================================
4768
4769 ::
4770
4771 struct kvm_nested_state {
4772 __u16 flags;
4773 __u16 format;
4774 __u32 size;
4775
4776 union {
4777 struct kvm_vmx_nested_state_hdr vmx;
4778 struct kvm_svm_nested_state_hdr svm;
4779
4780 /* Pad the header to 128 bytes. */
4781 __u8 pad[120];
4782 } hdr;
4783
4784 union {
4785 struct kvm_vmx_nested_state_data vmx[0];
4786 struct kvm_svm_nested_state_data svm[0];
4787 } data;
4788 };
4789
4790 #define KVM_STATE_NESTED_GUEST_MODE 0x00000001
4791 #define KVM_STATE_NESTED_RUN_PENDING 0x00000002
4792 #define KVM_STATE_NESTED_EVMCS 0x00000004
4793
4794 #define KVM_STATE_NESTED_FORMAT_VMX 0
4795 #define KVM_STATE_NESTED_FORMAT_SVM 1
4796
4797 #define KVM_STATE_NESTED_VMX_VMCS_SIZE 0x1000
4798
4799 #define KVM_STATE_NESTED_VMX_SMM_GUEST_MODE 0x00000001
4800 #define KVM_STATE_NESTED_VMX_SMM_VMXON 0x00000002
4801
4802 #define KVM_STATE_VMX_PREEMPTION_TIMER_DEADLINE 0x00000001
4803
4804 struct kvm_vmx_nested_state_hdr {
4805 __u64 vmxon_pa;
4806 __u64 vmcs12_pa;
4807
4808 struct {
4809 __u16 flags;
4810 } smm;
4811
4812 __u32 flags;
4813 __u64 preemption_timer_deadline;
4814 };
4815
4816 struct kvm_vmx_nested_state_data {
4817 __u8 vmcs12[KVM_STATE_NESTED_VMX_VMCS_SIZE];
4818 __u8 shadow_vmcs12[KVM_STATE_NESTED_VMX_VMCS_SIZE];
4819 };
4820
4821 This ioctl copies the vcpu's nested virtualization state from the kernel to
4822 userspace.
4823
4824 The maximum size of the state can be retrieved by passing KVM_CAP_NESTED_STATE
4825 to the KVM_CHECK_EXTENSION ioctl().
4826
4827 4.115 KVM_SET_NESTED_STATE
4828 --------------------------
4829
4830 :Capability: KVM_CAP_NESTED_STATE
4831 :Architectures: x86
4832 :Type: vcpu ioctl
4833 :Parameters: struct kvm_nested_state (in)
4834 :Returns: 0 on success, -1 on error
4835
4836 This copies the vcpu's kvm_nested_state struct from userspace to the kernel.
4837 For the definition of struct kvm_nested_state, see KVM_GET_NESTED_STATE.
4838
4839 4.116 KVM_(UN)REGISTER_COALESCED_MMIO
4840 -------------------------------------
4841
4842 :Capability: KVM_CAP_COALESCED_MMIO (for coalesced mmio)
4843 KVM_CAP_COALESCED_PIO (for coalesced pio)
4844 :Architectures: all
4845 :Type: vm ioctl
4846 :Parameters: struct kvm_coalesced_mmio_zone
4847 :Returns: 0 on success, < 0 on error
4848
4849 Coalesced I/O is a performance optimization that defers hardware
4850 register write emulation so that userspace exits are avoided. It is
4851 typically used to reduce the overhead of emulating frequently accessed
4852 hardware registers.
4853
4854 When a hardware register is configured for coalesced I/O, write accesses
4855 do not exit to userspace and their value is recorded in a ring buffer
4856 that is shared between kernel and userspace.
4857
4858 Coalesced I/O is used if one or more write accesses to a hardware
4859 register can be deferred until a read or a write to another hardware
4860 register on the same device. This last access will cause a vmexit and
4861 userspace will process accesses from the ring buffer before emulating
4862 it. That will avoid exiting to userspace on repeated writes.
4863
4864 Coalesced pio is based on coalesced mmio. There is little difference
4865 between coalesced mmio and pio except that coalesced pio records accesses
4866 to I/O ports.
4867
4868 4.117 KVM_CLEAR_DIRTY_LOG (vm ioctl)
4869 ------------------------------------
4870
4871 :Capability: KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2
4872 :Architectures: x86, arm64, mips
4873 :Type: vm ioctl
4874 :Parameters: struct kvm_clear_dirty_log (in)
4875 :Returns: 0 on success, -1 on error
4876
4877 ::
4878
4879 /* for KVM_CLEAR_DIRTY_LOG */
4880 struct kvm_clear_dirty_log {
4881 __u32 slot;
4882 __u32 num_pages;
4883 __u64 first_page;
4884 union {
4885 void __user *dirty_bitmap; /* one bit per page */
4886 __u64 padding;
4887 };
4888 };
4889
4890 The ioctl clears the dirty status of pages in a memory slot, according to
4891 the bitmap that is passed in struct kvm_clear_dirty_log's dirty_bitmap
4892 field. Bit 0 of the bitmap corresponds to page "first_page" in the
4893 memory slot, and num_pages is the size in bits of the input bitmap.
4894 first_page must be a multiple of 64; num_pages must also be a multiple of
4895 64 unless first_page + num_pages is the size of the memory slot. For each
4896 bit that is set in the input bitmap, the corresponding page is marked "clean"
4897 in KVM's dirty bitmap, and dirty tracking is re-enabled for that page
4898 (for example via write-protection, or by clearing the dirty bit in
4899 a page table entry).
4900
4901 If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of slot field specifies
4902 the address space for which you want to clear the dirty status. See
4903 KVM_SET_USER_MEMORY_REGION for details on the usage of slot field.
4904
4905 This ioctl is mostly useful when KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2
4906 is enabled; for more information, see the description of the capability.
4907 However, it can always be used as long as KVM_CHECK_EXTENSION confirms
4908 that KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 is present.
4909
4910 4.118 KVM_GET_SUPPORTED_HV_CPUID
4911 --------------------------------
4912
4913 :Capability: KVM_CAP_HYPERV_CPUID (vcpu), KVM_CAP_SYS_HYPERV_CPUID (system)
4914 :Architectures: x86
4915 :Type: system ioctl, vcpu ioctl
4916 :Parameters: struct kvm_cpuid2 (in/out)
4917 :Returns: 0 on success, -1 on error
4918
4919 ::
4920
4921 struct kvm_cpuid2 {
4922 __u32 nent;
4923 __u32 padding;
4924 struct kvm_cpuid_entry2 entries[0];
4925 };
4926
4927 struct kvm_cpuid_entry2 {
4928 __u32 function;
4929 __u32 index;
4930 __u32 flags;
4931 __u32 eax;
4932 __u32 ebx;
4933 __u32 ecx;
4934 __u32 edx;
4935 __u32 padding[3];
4936 };
4937
4938 This ioctl returns x86 cpuid features leaves related to Hyper-V emulation in
4939 KVM. Userspace can use the information returned by this ioctl to construct
4940 cpuid information presented to guests consuming Hyper-V enlightenments (e.g.
4941 Windows or Hyper-V guests).
4942
4943 CPUID feature leaves returned by this ioctl are defined by Hyper-V Top Level
4944 Functional Specification (TLFS). These leaves can't be obtained with
4945 KVM_GET_SUPPORTED_CPUID ioctl because some of them intersect with KVM feature
4946 leaves (0x40000000, 0x40000001).
4947
4948 Currently, the following list of CPUID leaves are returned:
4949
4950 - HYPERV_CPUID_VENDOR_AND_MAX_FUNCTIONS
4951 - HYPERV_CPUID_INTERFACE
4952 - HYPERV_CPUID_VERSION
4953 - HYPERV_CPUID_FEATURES
4954 - HYPERV_CPUID_ENLIGHTMENT_INFO
4955 - HYPERV_CPUID_IMPLEMENT_LIMITS
4956 - HYPERV_CPUID_NESTED_FEATURES
4957 - HYPERV_CPUID_SYNDBG_VENDOR_AND_MAX_FUNCTIONS
4958 - HYPERV_CPUID_SYNDBG_INTERFACE
4959 - HYPERV_CPUID_SYNDBG_PLATFORM_CAPABILITIES
4960
4961 Userspace invokes KVM_GET_SUPPORTED_HV_CPUID by passing a kvm_cpuid2 structure
4962 with the 'nent' field indicating the number of entries in the variable-size
4963 array 'entries'. If the number of entries is too low to describe all Hyper-V
4964 feature leaves, an error (E2BIG) is returned. If the number is more or equal
4965 to the number of Hyper-V feature leaves, the 'nent' field is adjusted to the
4966 number of valid entries in the 'entries' array, which is then filled.
4967
4968 'index' and 'flags' fields in 'struct kvm_cpuid_entry2' are currently reserved,
4969 userspace should not expect to get any particular value there.
4970
4971 Note, vcpu version of KVM_GET_SUPPORTED_HV_CPUID is currently deprecated. Unlike
4972 system ioctl which exposes all supported feature bits unconditionally, vcpu
4973 version has the following quirks:
4974
4975 - HYPERV_CPUID_NESTED_FEATURES leaf and HV_X64_ENLIGHTENED_VMCS_RECOMMENDED
4976 feature bit are only exposed when Enlightened VMCS was previously enabled
4977 on the corresponding vCPU (KVM_CAP_HYPERV_ENLIGHTENED_VMCS).
4978 - HV_STIMER_DIRECT_MODE_AVAILABLE bit is only exposed with in-kernel LAPIC.
4979 (presumes KVM_CREATE_IRQCHIP has already been called).
4980
4981 4.119 KVM_ARM_VCPU_FINALIZE
4982 ---------------------------
4983
4984 :Architectures: arm64
4985 :Type: vcpu ioctl
4986 :Parameters: int feature (in)
4987 :Returns: 0 on success, -1 on error
4988
4989 Errors:
4990
4991 ====== ==============================================================
4992 EPERM feature not enabled, needs configuration, or already finalized
4993 EINVAL feature unknown or not present
4994 ====== ==============================================================
4995
4996 Recognised values for feature:
4997
4998 ===== ===========================================
4999 arm64 KVM_ARM_VCPU_SVE (requires KVM_CAP_ARM_SVE)
5000 ===== ===========================================
5001
5002 Finalizes the configuration of the specified vcpu feature.
5003
5004 The vcpu must already have been initialised, enabling the affected feature, by
5005 means of a successful KVM_ARM_VCPU_INIT call with the appropriate flag set in
5006 features[].
5007
5008 For affected vcpu features, this is a mandatory step that must be performed
5009 before the vcpu is fully usable.
5010
5011 Between KVM_ARM_VCPU_INIT and KVM_ARM_VCPU_FINALIZE, the feature may be
5012 configured by use of ioctls such as KVM_SET_ONE_REG. The exact configuration
5013 that should be performaned and how to do it are feature-dependent.
5014
5015 Other calls that depend on a particular feature being finalized, such as
5016 KVM_RUN, KVM_GET_REG_LIST, KVM_GET_ONE_REG and KVM_SET_ONE_REG, will fail with
5017 -EPERM unless the feature has already been finalized by means of a
5018 KVM_ARM_VCPU_FINALIZE call.
5019
5020 See KVM_ARM_VCPU_INIT for details of vcpu features that require finalization
5021 using this ioctl.
5022
5023 4.120 KVM_SET_PMU_EVENT_FILTER
5024 ------------------------------
5025
5026 :Capability: KVM_CAP_PMU_EVENT_FILTER
5027 :Architectures: x86
5028 :Type: vm ioctl
5029 :Parameters: struct kvm_pmu_event_filter (in)
5030 :Returns: 0 on success, -1 on error
5031
5032 ::
5033
5034 struct kvm_pmu_event_filter {
5035 __u32 action;
5036 __u32 nevents;
5037 __u32 fixed_counter_bitmap;
5038 __u32 flags;
5039 __u32 pad[4];
5040 __u64 events[0];
5041 };
5042
5043 This ioctl restricts the set of PMU events that the guest can program.
5044 The argument holds a list of events which will be allowed or denied.
5045 The eventsel+umask of each event the guest attempts to program is compared
5046 against the events field to determine whether the guest should have access.
5047 The events field only controls general purpose counters; fixed purpose
5048 counters are controlled by the fixed_counter_bitmap.
5049
5050 No flags are defined yet, the field must be zero.
5051
5052 Valid values for 'action'::
5053
5054 #define KVM_PMU_EVENT_ALLOW 0
5055 #define KVM_PMU_EVENT_DENY 1
5056
5057 4.121 KVM_PPC_SVM_OFF
5058 ---------------------
5059
5060 :Capability: basic
5061 :Architectures: powerpc
5062 :Type: vm ioctl
5063 :Parameters: none
5064 :Returns: 0 on successful completion,
5065
5066 Errors:
5067
5068 ====== ================================================================
5069 EINVAL if ultravisor failed to terminate the secure guest
5070 ENOMEM if hypervisor failed to allocate new radix page tables for guest
5071 ====== ================================================================
5072
5073 This ioctl is used to turn off the secure mode of the guest or transition
5074 the guest from secure mode to normal mode. This is invoked when the guest
5075 is reset. This has no effect if called for a normal guest.
5076
5077 This ioctl issues an ultravisor call to terminate the secure guest,
5078 unpins the VPA pages and releases all the device pages that are used to
5079 track the secure pages by hypervisor.
5080
5081 4.122 KVM_S390_NORMAL_RESET
5082 ---------------------------
5083
5084 :Capability: KVM_CAP_S390_VCPU_RESETS
5085 :Architectures: s390
5086 :Type: vcpu ioctl
5087 :Parameters: none
5088 :Returns: 0
5089
5090 This ioctl resets VCPU registers and control structures according to
5091 the cpu reset definition in the POP (Principles Of Operation).
5092
5093 4.123 KVM_S390_INITIAL_RESET
5094 ----------------------------
5095
5096 :Capability: none
5097 :Architectures: s390
5098 :Type: vcpu ioctl
5099 :Parameters: none
5100 :Returns: 0
5101
5102 This ioctl resets VCPU registers and control structures according to
5103 the initial cpu reset definition in the POP. However, the cpu is not
5104 put into ESA mode. This reset is a superset of the normal reset.
5105
5106 4.124 KVM_S390_CLEAR_RESET
5107 --------------------------
5108
5109 :Capability: KVM_CAP_S390_VCPU_RESETS
5110 :Architectures: s390
5111 :Type: vcpu ioctl
5112 :Parameters: none
5113 :Returns: 0
5114
5115 This ioctl resets VCPU registers and control structures according to
5116 the clear cpu reset definition in the POP. However, the cpu is not put
5117 into ESA mode. This reset is a superset of the initial reset.
5118
5119
5120 4.125 KVM_S390_PV_COMMAND
5121 -------------------------
5122
5123 :Capability: KVM_CAP_S390_PROTECTED
5124 :Architectures: s390
5125 :Type: vm ioctl
5126 :Parameters: struct kvm_pv_cmd
5127 :Returns: 0 on success, < 0 on error
5128
5129 ::
5130
5131 struct kvm_pv_cmd {
5132 __u32 cmd; /* Command to be executed */
5133 __u16 rc; /* Ultravisor return code */
5134 __u16 rrc; /* Ultravisor return reason code */
5135 __u64 data; /* Data or address */
5136 __u32 flags; /* flags for future extensions. Must be 0 for now */
5137 __u32 reserved[3];
5138 };
5139
5140 **Ultravisor return codes**
5141 The Ultravisor return (reason) codes are provided by the kernel if a
5142 Ultravisor call has been executed to achieve the results expected by
5143 the command. Therefore they are independent of the IOCTL return
5144 code. If KVM changes `rc`, its value will always be greater than 0
5145 hence setting it to 0 before issuing a PV command is advised to be
5146 able to detect a change of `rc`.
5147
5148 **cmd values:**
5149
5150 KVM_PV_ENABLE
5151 Allocate memory and register the VM with the Ultravisor, thereby
5152 donating memory to the Ultravisor that will become inaccessible to
5153 KVM. All existing CPUs are converted to protected ones. After this
5154 command has succeeded, any CPU added via hotplug will become
5155 protected during its creation as well.
5156
5157 Errors:
5158
5159 ===== =============================
5160 EINTR an unmasked signal is pending
5161 ===== =============================
5162
5163 KVM_PV_DISABLE
5164 Deregister the VM from the Ultravisor and reclaim the memory that
5165 had been donated to the Ultravisor, making it usable by the kernel
5166 again. All registered VCPUs are converted back to non-protected
5167 ones.
5168
5169 KVM_PV_VM_SET_SEC_PARMS
5170 Pass the image header from VM memory to the Ultravisor in
5171 preparation of image unpacking and verification.
5172
5173 KVM_PV_VM_UNPACK
5174 Unpack (protect and decrypt) a page of the encrypted boot image.
5175
5176 KVM_PV_VM_VERIFY
5177 Verify the integrity of the unpacked image. Only if this succeeds,
5178 KVM is allowed to start protected VCPUs.
5179
5180 KVM_PV_INFO
5181 :Capability: KVM_CAP_S390_PROTECTED_DUMP
5182
5183 Presents an API that provides Ultravisor related data to userspace
5184 via subcommands. len_max is the size of the user space buffer,
5185 len_written is KVM's indication of how much bytes of that buffer
5186 were actually written to. len_written can be used to determine the
5187 valid fields if more response fields are added in the future.
5188
5189 ::
5190
5191 enum pv_cmd_info_id {
5192 KVM_PV_INFO_VM,
5193 KVM_PV_INFO_DUMP,
5194 };
5195
5196 struct kvm_s390_pv_info_header {
5197 __u32 id;
5198 __u32 len_max;
5199 __u32 len_written;
5200 __u32 reserved;
5201 };
5202
5203 struct kvm_s390_pv_info {
5204 struct kvm_s390_pv_info_header header;
5205 struct kvm_s390_pv_info_dump dump;
5206 struct kvm_s390_pv_info_vm vm;
5207 };
5208
5209 **subcommands:**
5210
5211 KVM_PV_INFO_VM
5212 This subcommand provides basic Ultravisor information for PV
5213 hosts. These values are likely also exported as files in the sysfs
5214 firmware UV query interface but they are more easily available to
5215 programs in this API.
5216
5217 The installed calls and feature_indication members provide the
5218 installed UV calls and the UV's other feature indications.
5219
5220 The max_* members provide information about the maximum number of PV
5221 vcpus, PV guests and PV guest memory size.
5222
5223 ::
5224
5225 struct kvm_s390_pv_info_vm {
5226 __u64 inst_calls_list[4];
5227 __u64 max_cpus;
5228 __u64 max_guests;
5229 __u64 max_guest_addr;
5230 __u64 feature_indication;
5231 };
5232
5233
5234 KVM_PV_INFO_DUMP
5235 This subcommand provides information related to dumping PV guests.
5236
5237 ::
5238
5239 struct kvm_s390_pv_info_dump {
5240 __u64 dump_cpu_buffer_len;
5241 __u64 dump_config_mem_buffer_per_1m;
5242 __u64 dump_config_finalize_len;
5243 };
5244
5245 KVM_PV_DUMP
5246 :Capability: KVM_CAP_S390_PROTECTED_DUMP
5247
5248 Presents an API that provides calls which facilitate dumping a
5249 protected VM.
5250
5251 ::
5252
5253 struct kvm_s390_pv_dmp {
5254 __u64 subcmd;
5255 __u64 buff_addr;
5256 __u64 buff_len;
5257 __u64 gaddr; /* For dump storage state */
5258 };
5259
5260 **subcommands:**
5261
5262 KVM_PV_DUMP_INIT
5263 Initializes the dump process of a protected VM. If this call does
5264 not succeed all other subcommands will fail with -EINVAL. This
5265 subcommand will return -EINVAL if a dump process has not yet been
5266 completed.
5267
5268 Not all PV vms can be dumped, the owner needs to set `dump
5269 allowed` PCF bit 34 in the SE header to allow dumping.
5270
5271 KVM_PV_DUMP_CONFIG_STOR_STATE
5272 Stores `buff_len` bytes of tweak component values starting with
5273 the 1MB block specified by the absolute guest address
5274 (`gaddr`). `buff_len` needs to be `conf_dump_storage_state_len`
5275 aligned and at least >= the `conf_dump_storage_state_len` value
5276 provided by the dump uv_info data. buff_user might be written to
5277 even if an error rc is returned. For instance if we encounter a
5278 fault after writing the first page of data.
5279
5280 KVM_PV_DUMP_COMPLETE
5281 If the subcommand succeeds it completes the dump process and lets
5282 KVM_PV_DUMP_INIT be called again.
5283
5284 On success `conf_dump_finalize_len` bytes of completion data will be
5285 stored to the `buff_addr`. The completion data contains a key
5286 derivation seed, IV, tweak nonce and encryption keys as well as an
5287 authentication tag all of which are needed to decrypt the dump at a
5288 later time.
5289
5290
5291 4.126 KVM_X86_SET_MSR_FILTER
5292 ----------------------------
5293
5294 :Capability: KVM_CAP_X86_MSR_FILTER
5295 :Architectures: x86
5296 :Type: vm ioctl
5297 :Parameters: struct kvm_msr_filter
5298 :Returns: 0 on success, < 0 on error
5299
5300 ::
5301
5302 struct kvm_msr_filter_range {
5303 #define KVM_MSR_FILTER_READ (1 << 0)
5304 #define KVM_MSR_FILTER_WRITE (1 << 1)
5305 __u32 flags;
5306 __u32 nmsrs; /* number of msrs in bitmap */
5307 __u32 base; /* MSR index the bitmap starts at */
5308 __u8 *bitmap; /* a 1 bit allows the operations in flags, 0 denies */
5309 };
5310
5311 #define KVM_MSR_FILTER_MAX_RANGES 16
5312 struct kvm_msr_filter {
5313 #define KVM_MSR_FILTER_DEFAULT_ALLOW (0 << 0)
5314 #define KVM_MSR_FILTER_DEFAULT_DENY (1 << 0)
5315 __u32 flags;
5316 struct kvm_msr_filter_range ranges[KVM_MSR_FILTER_MAX_RANGES];
5317 };
5318
5319 flags values for ``struct kvm_msr_filter_range``:
5320
5321 ``KVM_MSR_FILTER_READ``
5322
5323 Filter read accesses to MSRs using the given bitmap. A 0 in the bitmap
5324 indicates that a read should immediately fail, while a 1 indicates that
5325 a read for a particular MSR should be handled regardless of the default
5326 filter action.
5327
5328 ``KVM_MSR_FILTER_WRITE``
5329
5330 Filter write accesses to MSRs using the given bitmap. A 0 in the bitmap
5331 indicates that a write should immediately fail, while a 1 indicates that
5332 a write for a particular MSR should be handled regardless of the default
5333 filter action.
5334
5335 ``KVM_MSR_FILTER_READ | KVM_MSR_FILTER_WRITE``
5336
5337 Filter both read and write accesses to MSRs using the given bitmap. A 0
5338 in the bitmap indicates that both reads and writes should immediately fail,
5339 while a 1 indicates that reads and writes for a particular MSR are not
5340 filtered by this range.
5341
5342 flags values for ``struct kvm_msr_filter``:
5343
5344 ``KVM_MSR_FILTER_DEFAULT_ALLOW``
5345
5346 If no filter range matches an MSR index that is getting accessed, KVM will
5347 fall back to allowing access to the MSR.
5348
5349 ``KVM_MSR_FILTER_DEFAULT_DENY``
5350
5351 If no filter range matches an MSR index that is getting accessed, KVM will
5352 fall back to rejecting access to the MSR. In this mode, all MSRs that should
5353 be processed by KVM need to explicitly be marked as allowed in the bitmaps.
5354
5355 This ioctl allows user space to define up to 16 bitmaps of MSR ranges to
5356 specify whether a certain MSR access should be explicitly filtered for or not.
5357
5358 If this ioctl has never been invoked, MSR accesses are not guarded and the
5359 default KVM in-kernel emulation behavior is fully preserved.
5360
5361 Calling this ioctl with an empty set of ranges (all nmsrs == 0) disables MSR
5362 filtering. In that mode, ``KVM_MSR_FILTER_DEFAULT_DENY`` is invalid and causes
5363 an error.
5364
5365 As soon as the filtering is in place, every MSR access is processed through
5366 the filtering except for accesses to the x2APIC MSRs (from 0x800 to 0x8ff);
5367 x2APIC MSRs are always allowed, independent of the ``default_allow`` setting,
5368 and their behavior depends on the ``X2APIC_ENABLE`` bit of the APIC base
5369 register.
5370
5371 If a bit is within one of the defined ranges, read and write accesses are
5372 guarded by the bitmap's value for the MSR index if the kind of access
5373 is included in the ``struct kvm_msr_filter_range`` flags. If no range
5374 cover this particular access, the behavior is determined by the flags
5375 field in the kvm_msr_filter struct: ``KVM_MSR_FILTER_DEFAULT_ALLOW``
5376 and ``KVM_MSR_FILTER_DEFAULT_DENY``.
5377
5378 Each bitmap range specifies a range of MSRs to potentially allow access on.
5379 The range goes from MSR index [base .. base+nmsrs]. The flags field
5380 indicates whether reads, writes or both reads and writes are filtered
5381 by setting a 1 bit in the bitmap for the corresponding MSR index.
5382
5383 If an MSR access is not permitted through the filtering, it generates a
5384 #GP inside the guest. When combined with KVM_CAP_X86_USER_SPACE_MSR, that
5385 allows user space to deflect and potentially handle various MSR accesses
5386 into user space.
5387
5388 Note, invoking this ioctl with a vCPU is running is inherently racy. However,
5389 KVM does guarantee that vCPUs will see either the previous filter or the new
5390 filter, e.g. MSRs with identical settings in both the old and new filter will
5391 have deterministic behavior.
5392
5393 4.127 KVM_XEN_HVM_SET_ATTR
5394 --------------------------
5395
5396 :Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO
5397 :Architectures: x86
5398 :Type: vm ioctl
5399 :Parameters: struct kvm_xen_hvm_attr
5400 :Returns: 0 on success, < 0 on error
5401
5402 ::
5403
5404 struct kvm_xen_hvm_attr {
5405 __u16 type;
5406 __u16 pad[3];
5407 union {
5408 __u8 long_mode;
5409 __u8 vector;
5410 struct {
5411 __u64 gfn;
5412 } shared_info;
5413 struct {
5414 __u32 send_port;
5415 __u32 type; /* EVTCHNSTAT_ipi / EVTCHNSTAT_interdomain */
5416 __u32 flags;
5417 union {
5418 struct {
5419 __u32 port;
5420 __u32 vcpu;
5421 __u32 priority;
5422 } port;
5423 struct {
5424 __u32 port; /* Zero for eventfd */
5425 __s32 fd;
5426 } eventfd;
5427 __u32 padding[4];
5428 } deliver;
5429 } evtchn;
5430 __u32 xen_version;
5431 __u64 pad[8];
5432 } u;
5433 };
5434
5435 type values:
5436
5437 KVM_XEN_ATTR_TYPE_LONG_MODE
5438 Sets the ABI mode of the VM to 32-bit or 64-bit (long mode). This
5439 determines the layout of the shared info pages exposed to the VM.
5440
5441 KVM_XEN_ATTR_TYPE_SHARED_INFO
5442 Sets the guest physical frame number at which the Xen "shared info"
5443 page resides. Note that although Xen places vcpu_info for the first
5444 32 vCPUs in the shared_info page, KVM does not automatically do so
5445 and instead requires that KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO be used
5446 explicitly even when the vcpu_info for a given vCPU resides at the
5447 "default" location in the shared_info page. This is because KVM is
5448 not aware of the Xen CPU id which is used as the index into the
5449 vcpu_info[] array, so cannot know the correct default location.
5450
5451 Note that the shared info page may be constantly written to by KVM;
5452 it contains the event channel bitmap used to deliver interrupts to
5453 a Xen guest, amongst other things. It is exempt from dirty tracking
5454 mechanisms — KVM will not explicitly mark the page as dirty each
5455 time an event channel interrupt is delivered to the guest! Thus,
5456 userspace should always assume that the designated GFN is dirty if
5457 any vCPU has been running or any event channel interrupts can be
5458 routed to the guest.
5459
5460 KVM_XEN_ATTR_TYPE_UPCALL_VECTOR
5461 Sets the exception vector used to deliver Xen event channel upcalls.
5462 This is the HVM-wide vector injected directly by the hypervisor
5463 (not through the local APIC), typically configured by a guest via
5464 HVM_PARAM_CALLBACK_IRQ.
5465
5466 KVM_XEN_ATTR_TYPE_EVTCHN
5467 This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates
5468 support for KVM_XEN_HVM_CONFIG_EVTCHN_SEND features. It configures
5469 an outbound port number for interception of EVTCHNOP_send requests
5470 from the guest. A given sending port number may be directed back
5471 to a specified vCPU (by APIC ID) / port / priority on the guest,
5472 or to trigger events on an eventfd. The vCPU and priority can be
5473 changed by setting KVM_XEN_EVTCHN_UPDATE in a subsequent call,
5474 but other fields cannot change for a given sending port. A port
5475 mapping is removed by using KVM_XEN_EVTCHN_DEASSIGN in the flags
5476 field.
5477
5478 KVM_XEN_ATTR_TYPE_XEN_VERSION
5479 This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates
5480 support for KVM_XEN_HVM_CONFIG_EVTCHN_SEND features. It configures
5481 the 32-bit version code returned to the guest when it invokes the
5482 XENVER_version call; typically (XEN_MAJOR << 16 | XEN_MINOR). PV
5483 Xen guests will often use this to as a dummy hypercall to trigger
5484 event channel delivery, so responding within the kernel without
5485 exiting to userspace is beneficial.
5486
5487 4.127 KVM_XEN_HVM_GET_ATTR
5488 --------------------------
5489
5490 :Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO
5491 :Architectures: x86
5492 :Type: vm ioctl
5493 :Parameters: struct kvm_xen_hvm_attr
5494 :Returns: 0 on success, < 0 on error
5495
5496 Allows Xen VM attributes to be read. For the structure and types,
5497 see KVM_XEN_HVM_SET_ATTR above. The KVM_XEN_ATTR_TYPE_EVTCHN
5498 attribute cannot be read.
5499
5500 4.128 KVM_XEN_VCPU_SET_ATTR
5501 ---------------------------
5502
5503 :Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO
5504 :Architectures: x86
5505 :Type: vcpu ioctl
5506 :Parameters: struct kvm_xen_vcpu_attr
5507 :Returns: 0 on success, < 0 on error
5508
5509 ::
5510
5511 struct kvm_xen_vcpu_attr {
5512 __u16 type;
5513 __u16 pad[3];
5514 union {
5515 __u64 gpa;
5516 __u64 pad[4];
5517 struct {
5518 __u64 state;
5519 __u64 state_entry_time;
5520 __u64 time_running;
5521 __u64 time_runnable;
5522 __u64 time_blocked;
5523 __u64 time_offline;
5524 } runstate;
5525 __u32 vcpu_id;
5526 struct {
5527 __u32 port;
5528 __u32 priority;
5529 __u64 expires_ns;
5530 } timer;
5531 __u8 vector;
5532 } u;
5533 };
5534
5535 type values:
5536
5537 KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO
5538 Sets the guest physical address of the vcpu_info for a given vCPU.
5539 As with the shared_info page for the VM, the corresponding page may be
5540 dirtied at any time if event channel interrupt delivery is enabled, so
5541 userspace should always assume that the page is dirty without relying
5542 on dirty logging.
5543
5544 KVM_XEN_VCPU_ATTR_TYPE_VCPU_TIME_INFO
5545 Sets the guest physical address of an additional pvclock structure
5546 for a given vCPU. This is typically used for guest vsyscall support.
5547
5548 KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADDR
5549 Sets the guest physical address of the vcpu_runstate_info for a given
5550 vCPU. This is how a Xen guest tracks CPU state such as steal time.
5551
5552 KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_CURRENT
5553 Sets the runstate (RUNSTATE_running/_runnable/_blocked/_offline) of
5554 the given vCPU from the .u.runstate.state member of the structure.
5555 KVM automatically accounts running and runnable time but blocked
5556 and offline states are only entered explicitly.
5557
5558 KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_DATA
5559 Sets all fields of the vCPU runstate data from the .u.runstate member
5560 of the structure, including the current runstate. The state_entry_time
5561 must equal the sum of the other four times.
5562
5563 KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADJUST
5564 This *adds* the contents of the .u.runstate members of the structure
5565 to the corresponding members of the given vCPU's runstate data, thus
5566 permitting atomic adjustments to the runstate times. The adjustment
5567 to the state_entry_time must equal the sum of the adjustments to the
5568 other four times. The state field must be set to -1, or to a valid
5569 runstate value (RUNSTATE_running, RUNSTATE_runnable, RUNSTATE_blocked
5570 or RUNSTATE_offline) to set the current accounted state as of the
5571 adjusted state_entry_time.
5572
5573 KVM_XEN_VCPU_ATTR_TYPE_VCPU_ID
5574 This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates
5575 support for KVM_XEN_HVM_CONFIG_EVTCHN_SEND features. It sets the Xen
5576 vCPU ID of the given vCPU, to allow timer-related VCPU operations to
5577 be intercepted by KVM.
5578
5579 KVM_XEN_VCPU_ATTR_TYPE_TIMER
5580 This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates
5581 support for KVM_XEN_HVM_CONFIG_EVTCHN_SEND features. It sets the
5582 event channel port/priority for the VIRQ_TIMER of the vCPU, as well
5583 as allowing a pending timer to be saved/restored.
5584
5585 KVM_XEN_VCPU_ATTR_TYPE_UPCALL_VECTOR
5586 This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates
5587 support for KVM_XEN_HVM_CONFIG_EVTCHN_SEND features. It sets the
5588 per-vCPU local APIC upcall vector, configured by a Xen guest with
5589 the HVMOP_set_evtchn_upcall_vector hypercall. This is typically
5590 used by Windows guests, and is distinct from the HVM-wide upcall
5591 vector configured with HVM_PARAM_CALLBACK_IRQ.
5592
5593
5594 4.129 KVM_XEN_VCPU_GET_ATTR
5595 ---------------------------
5596
5597 :Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO
5598 :Architectures: x86
5599 :Type: vcpu ioctl
5600 :Parameters: struct kvm_xen_vcpu_attr
5601 :Returns: 0 on success, < 0 on error
5602
5603 Allows Xen vCPU attributes to be read. For the structure and types,
5604 see KVM_XEN_VCPU_SET_ATTR above.
5605
5606 The KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADJUST type may not be used
5607 with the KVM_XEN_VCPU_GET_ATTR ioctl.
5608
5609 4.130 KVM_ARM_MTE_COPY_TAGS
5610 ---------------------------
5611
5612 :Capability: KVM_CAP_ARM_MTE
5613 :Architectures: arm64
5614 :Type: vm ioctl
5615 :Parameters: struct kvm_arm_copy_mte_tags
5616 :Returns: number of bytes copied, < 0 on error (-EINVAL for incorrect
5617 arguments, -EFAULT if memory cannot be accessed).
5618
5619 ::
5620
5621 struct kvm_arm_copy_mte_tags {
5622 __u64 guest_ipa;
5623 __u64 length;
5624 void __user *addr;
5625 __u64 flags;
5626 __u64 reserved[2];
5627 };
5628
5629 Copies Memory Tagging Extension (MTE) tags to/from guest tag memory. The
5630 ``guest_ipa`` and ``length`` fields must be ``PAGE_SIZE`` aligned. The ``addr``
5631 field must point to a buffer which the tags will be copied to or from.
5632
5633 ``flags`` specifies the direction of copy, either ``KVM_ARM_TAGS_TO_GUEST`` or
5634 ``KVM_ARM_TAGS_FROM_GUEST``.
5635
5636 The size of the buffer to store the tags is ``(length / 16)`` bytes
5637 (granules in MTE are 16 bytes long). Each byte contains a single tag
5638 value. This matches the format of ``PTRACE_PEEKMTETAGS`` and
5639 ``PTRACE_POKEMTETAGS``.
5640
5641 If an error occurs before any data is copied then a negative error code is
5642 returned. If some tags have been copied before an error occurs then the number
5643 of bytes successfully copied is returned. If the call completes successfully
5644 then ``length`` is returned.
5645
5646 4.131 KVM_GET_SREGS2
5647 --------------------
5648
5649 :Capability: KVM_CAP_SREGS2
5650 :Architectures: x86
5651 :Type: vcpu ioctl
5652 :Parameters: struct kvm_sregs2 (out)
5653 :Returns: 0 on success, -1 on error
5654
5655 Reads special registers from the vcpu.
5656 This ioctl (when supported) replaces the KVM_GET_SREGS.
5657
5658 ::
5659
5660 struct kvm_sregs2 {
5661 /* out (KVM_GET_SREGS2) / in (KVM_SET_SREGS2) */
5662 struct kvm_segment cs, ds, es, fs, gs, ss;
5663 struct kvm_segment tr, ldt;
5664 struct kvm_dtable gdt, idt;
5665 __u64 cr0, cr2, cr3, cr4, cr8;
5666 __u64 efer;
5667 __u64 apic_base;
5668 __u64 flags;
5669 __u64 pdptrs[4];
5670 };
5671
5672 flags values for ``kvm_sregs2``:
5673
5674 ``KVM_SREGS2_FLAGS_PDPTRS_VALID``
5675
5676 Indicates thats the struct contain valid PDPTR values.
5677
5678
5679 4.132 KVM_SET_SREGS2
5680 --------------------
5681
5682 :Capability: KVM_CAP_SREGS2
5683 :Architectures: x86
5684 :Type: vcpu ioctl
5685 :Parameters: struct kvm_sregs2 (in)
5686 :Returns: 0 on success, -1 on error
5687
5688 Writes special registers into the vcpu.
5689 See KVM_GET_SREGS2 for the data structures.
5690 This ioctl (when supported) replaces the KVM_SET_SREGS.
5691
5692 4.133 KVM_GET_STATS_FD
5693 ----------------------
5694
5695 :Capability: KVM_CAP_STATS_BINARY_FD
5696 :Architectures: all
5697 :Type: vm ioctl, vcpu ioctl
5698 :Parameters: none
5699 :Returns: statistics file descriptor on success, < 0 on error
5700
5701 Errors:
5702
5703 ====== ======================================================
5704 ENOMEM if the fd could not be created due to lack of memory
5705 EMFILE if the number of opened files exceeds the limit
5706 ====== ======================================================
5707
5708 The returned file descriptor can be used to read VM/vCPU statistics data in
5709 binary format. The data in the file descriptor consists of four blocks
5710 organized as follows:
5711
5712 +-------------+
5713 | Header |
5714 +-------------+
5715 | id string |
5716 +-------------+
5717 | Descriptors |
5718 +-------------+
5719 | Stats Data |
5720 +-------------+
5721
5722 Apart from the header starting at offset 0, please be aware that it is
5723 not guaranteed that the four blocks are adjacent or in the above order;
5724 the offsets of the id, descriptors and data blocks are found in the
5725 header. However, all four blocks are aligned to 64 bit offsets in the
5726 file and they do not overlap.
5727
5728 All blocks except the data block are immutable. Userspace can read them
5729 only one time after retrieving the file descriptor, and then use ``pread`` or
5730 ``lseek`` to read the statistics repeatedly.
5731
5732 All data is in system endianness.
5733
5734 The format of the header is as follows::
5735
5736 struct kvm_stats_header {
5737 __u32 flags;
5738 __u32 name_size;
5739 __u32 num_desc;
5740 __u32 id_offset;
5741 __u32 desc_offset;
5742 __u32 data_offset;
5743 };
5744
5745 The ``flags`` field is not used at the moment. It is always read as 0.
5746
5747 The ``name_size`` field is the size (in byte) of the statistics name string
5748 (including trailing '\0') which is contained in the "id string" block and
5749 appended at the end of every descriptor.
5750
5751 The ``num_desc`` field is the number of descriptors that are included in the
5752 descriptor block. (The actual number of values in the data block may be
5753 larger, since each descriptor may comprise more than one value).
5754
5755 The ``id_offset`` field is the offset of the id string from the start of the
5756 file indicated by the file descriptor. It is a multiple of 8.
5757
5758 The ``desc_offset`` field is the offset of the Descriptors block from the start
5759 of the file indicated by the file descriptor. It is a multiple of 8.
5760
5761 The ``data_offset`` field is the offset of the Stats Data block from the start
5762 of the file indicated by the file descriptor. It is a multiple of 8.
5763
5764 The id string block contains a string which identifies the file descriptor on
5765 which KVM_GET_STATS_FD was invoked. The size of the block, including the
5766 trailing ``'\0'``, is indicated by the ``name_size`` field in the header.
5767
5768 The descriptors block is only needed to be read once for the lifetime of the
5769 file descriptor contains a sequence of ``struct kvm_stats_desc``, each followed
5770 by a string of size ``name_size``.
5771 ::
5772
5773 #define KVM_STATS_TYPE_SHIFT 0
5774 #define KVM_STATS_TYPE_MASK (0xF << KVM_STATS_TYPE_SHIFT)
5775 #define KVM_STATS_TYPE_CUMULATIVE (0x0 << KVM_STATS_TYPE_SHIFT)
5776 #define KVM_STATS_TYPE_INSTANT (0x1 << KVM_STATS_TYPE_SHIFT)
5777 #define KVM_STATS_TYPE_PEAK (0x2 << KVM_STATS_TYPE_SHIFT)
5778 #define KVM_STATS_TYPE_LINEAR_HIST (0x3 << KVM_STATS_TYPE_SHIFT)
5779 #define KVM_STATS_TYPE_LOG_HIST (0x4 << KVM_STATS_TYPE_SHIFT)
5780 #define KVM_STATS_TYPE_MAX KVM_STATS_TYPE_LOG_HIST
5781
5782 #define KVM_STATS_UNIT_SHIFT 4
5783 #define KVM_STATS_UNIT_MASK (0xF << KVM_STATS_UNIT_SHIFT)
5784 #define KVM_STATS_UNIT_NONE (0x0 << KVM_STATS_UNIT_SHIFT)
5785 #define KVM_STATS_UNIT_BYTES (0x1 << KVM_STATS_UNIT_SHIFT)
5786 #define KVM_STATS_UNIT_SECONDS (0x2 << KVM_STATS_UNIT_SHIFT)
5787 #define KVM_STATS_UNIT_CYCLES (0x3 << KVM_STATS_UNIT_SHIFT)
5788 #define KVM_STATS_UNIT_BOOLEAN (0x4 << KVM_STATS_UNIT_SHIFT)
5789 #define KVM_STATS_UNIT_MAX KVM_STATS_UNIT_BOOLEAN
5790
5791 #define KVM_STATS_BASE_SHIFT 8
5792 #define KVM_STATS_BASE_MASK (0xF << KVM_STATS_BASE_SHIFT)
5793 #define KVM_STATS_BASE_POW10 (0x0 << KVM_STATS_BASE_SHIFT)
5794 #define KVM_STATS_BASE_POW2 (0x1 << KVM_STATS_BASE_SHIFT)
5795 #define KVM_STATS_BASE_MAX KVM_STATS_BASE_POW2
5796
5797 struct kvm_stats_desc {
5798 __u32 flags;
5799 __s16 exponent;
5800 __u16 size;
5801 __u32 offset;
5802 __u32 bucket_size;
5803 char name[];
5804 };
5805
5806 The ``flags`` field contains the type and unit of the statistics data described
5807 by this descriptor. Its endianness is CPU native.
5808 The following flags are supported:
5809
5810 Bits 0-3 of ``flags`` encode the type:
5811
5812 * ``KVM_STATS_TYPE_CUMULATIVE``
5813 The statistics reports a cumulative count. The value of data can only be increased.
5814 Most of the counters used in KVM are of this type.
5815 The corresponding ``size`` field for this type is always 1.
5816 All cumulative statistics data are read/write.
5817 * ``KVM_STATS_TYPE_INSTANT``
5818 The statistics reports an instantaneous value. Its value can be increased or
5819 decreased. This type is usually used as a measurement of some resources,
5820 like the number of dirty pages, the number of large pages, etc.
5821 All instant statistics are read only.
5822 The corresponding ``size`` field for this type is always 1.
5823 * ``KVM_STATS_TYPE_PEAK``
5824 The statistics data reports a peak value, for example the maximum number
5825 of items in a hash table bucket, the longest time waited and so on.
5826 The value of data can only be increased.
5827 The corresponding ``size`` field for this type is always 1.
5828 * ``KVM_STATS_TYPE_LINEAR_HIST``
5829 The statistic is reported as a linear histogram. The number of
5830 buckets is specified by the ``size`` field. The size of buckets is specified
5831 by the ``hist_param`` field. The range of the Nth bucket (1 <= N < ``size``)
5832 is [``hist_param``*(N-1), ``hist_param``*N), while the range of the last
5833 bucket is [``hist_param``*(``size``-1), +INF). (+INF means positive infinity
5834 value.)
5835 * ``KVM_STATS_TYPE_LOG_HIST``
5836 The statistic is reported as a logarithmic histogram. The number of
5837 buckets is specified by the ``size`` field. The range of the first bucket is
5838 [0, 1), while the range of the last bucket is [pow(2, ``size``-2), +INF).
5839 Otherwise, The Nth bucket (1 < N < ``size``) covers
5840 [pow(2, N-2), pow(2, N-1)).
5841
5842 Bits 4-7 of ``flags`` encode the unit:
5843
5844 * ``KVM_STATS_UNIT_NONE``
5845 There is no unit for the value of statistics data. This usually means that
5846 the value is a simple counter of an event.
5847 * ``KVM_STATS_UNIT_BYTES``
5848 It indicates that the statistics data is used to measure memory size, in the
5849 unit of Byte, KiByte, MiByte, GiByte, etc. The unit of the data is
5850 determined by the ``exponent`` field in the descriptor.
5851 * ``KVM_STATS_UNIT_SECONDS``
5852 It indicates that the statistics data is used to measure time or latency.
5853 * ``KVM_STATS_UNIT_CYCLES``
5854 It indicates that the statistics data is used to measure CPU clock cycles.
5855 * ``KVM_STATS_UNIT_BOOLEAN``
5856 It indicates that the statistic will always be either 0 or 1. Boolean
5857 statistics of "peak" type will never go back from 1 to 0. Boolean
5858 statistics can be linear histograms (with two buckets) but not logarithmic
5859 histograms.
5860
5861 Note that, in the case of histograms, the unit applies to the bucket
5862 ranges, while the bucket value indicates how many samples fell in the
5863 bucket's range.
5864
5865 Bits 8-11 of ``flags``, together with ``exponent``, encode the scale of the
5866 unit:
5867
5868 * ``KVM_STATS_BASE_POW10``
5869 The scale is based on power of 10. It is used for measurement of time and
5870 CPU clock cycles. For example, an exponent of -9 can be used with
5871 ``KVM_STATS_UNIT_SECONDS`` to express that the unit is nanoseconds.
5872 * ``KVM_STATS_BASE_POW2``
5873 The scale is based on power of 2. It is used for measurement of memory size.
5874 For example, an exponent of 20 can be used with ``KVM_STATS_UNIT_BYTES`` to
5875 express that the unit is MiB.
5876
5877 The ``size`` field is the number of values of this statistics data. Its
5878 value is usually 1 for most of simple statistics. 1 means it contains an
5879 unsigned 64bit data.
5880
5881 The ``offset`` field is the offset from the start of Data Block to the start of
5882 the corresponding statistics data.
5883
5884 The ``bucket_size`` field is used as a parameter for histogram statistics data.
5885 It is only used by linear histogram statistics data, specifying the size of a
5886 bucket in the unit expressed by bits 4-11 of ``flags`` together with ``exponent``.
5887
5888 The ``name`` field is the name string of the statistics data. The name string
5889 starts at the end of ``struct kvm_stats_desc``. The maximum length including
5890 the trailing ``'\0'``, is indicated by ``name_size`` in the header.
5891
5892 The Stats Data block contains an array of 64-bit values in the same order
5893 as the descriptors in Descriptors block.
5894
5895 4.134 KVM_GET_XSAVE2
5896 --------------------
5897
5898 :Capability: KVM_CAP_XSAVE2
5899 :Architectures: x86
5900 :Type: vcpu ioctl
5901 :Parameters: struct kvm_xsave (out)
5902 :Returns: 0 on success, -1 on error
5903
5904
5905 ::
5906
5907 struct kvm_xsave {
5908 __u32 region[1024];
5909 __u32 extra[0];
5910 };
5911
5912 This ioctl would copy current vcpu's xsave struct to the userspace. It
5913 copies as many bytes as are returned by KVM_CHECK_EXTENSION(KVM_CAP_XSAVE2)
5914 when invoked on the vm file descriptor. The size value returned by
5915 KVM_CHECK_EXTENSION(KVM_CAP_XSAVE2) will always be at least 4096.
5916 Currently, it is only greater than 4096 if a dynamic feature has been
5917 enabled with ``arch_prctl()``, but this may change in the future.
5918
5919 The offsets of the state save areas in struct kvm_xsave follow the contents
5920 of CPUID leaf 0xD on the host.
5921
5922 4.135 KVM_XEN_HVM_EVTCHN_SEND
5923 -----------------------------
5924
5925 :Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_EVTCHN_SEND
5926 :Architectures: x86
5927 :Type: vm ioctl
5928 :Parameters: struct kvm_irq_routing_xen_evtchn
5929 :Returns: 0 on success, < 0 on error
5930
5931
5932 ::
5933
5934 struct kvm_irq_routing_xen_evtchn {
5935 __u32 port;
5936 __u32 vcpu;
5937 __u32 priority;
5938 };
5939
5940 This ioctl injects an event channel interrupt directly to the guest vCPU.
5941
5942 4.136 KVM_S390_PV_CPU_COMMAND
5943 -----------------------------
5944
5945 :Capability: KVM_CAP_S390_PROTECTED_DUMP
5946 :Architectures: s390
5947 :Type: vcpu ioctl
5948 :Parameters: none
5949 :Returns: 0 on success, < 0 on error
5950
5951 This ioctl closely mirrors `KVM_S390_PV_COMMAND` but handles requests
5952 for vcpus. It re-uses the kvm_s390_pv_dmp struct and hence also shares
5953 the command ids.
5954
5955 **command:**
5956
5957 KVM_PV_DUMP
5958 Presents an API that provides calls which facilitate dumping a vcpu
5959 of a protected VM.
5960
5961 **subcommand:**
5962
5963 KVM_PV_DUMP_CPU
5964 Provides encrypted dump data like register values.
5965 The length of the returned data is provided by uv_info.guest_cpu_stor_len.
5966
5967 4.137 KVM_S390_ZPCI_OP
5968 ----------------------
5969
5970 :Capability: KVM_CAP_S390_ZPCI_OP
5971 :Architectures: s390
5972 :Type: vm ioctl
5973 :Parameters: struct kvm_s390_zpci_op (in)
5974 :Returns: 0 on success, <0 on error
5975
5976 Used to manage hardware-assisted virtualization features for zPCI devices.
5977
5978 Parameters are specified via the following structure::
5979
5980 struct kvm_s390_zpci_op {
5981 /* in */
5982 __u32 fh; /* target device */
5983 __u8 op; /* operation to perform */
5984 __u8 pad[3];
5985 union {
5986 /* for KVM_S390_ZPCIOP_REG_AEN */
5987 struct {
5988 __u64 ibv; /* Guest addr of interrupt bit vector */
5989 __u64 sb; /* Guest addr of summary bit */
5990 __u32 flags;
5991 __u32 noi; /* Number of interrupts */
5992 __u8 isc; /* Guest interrupt subclass */
5993 __u8 sbo; /* Offset of guest summary bit vector */
5994 __u16 pad;
5995 } reg_aen;
5996 __u64 reserved[8];
5997 } u;
5998 };
5999
6000 The type of operation is specified in the "op" field.
6001 KVM_S390_ZPCIOP_REG_AEN is used to register the VM for adapter event
6002 notification interpretation, which will allow firmware delivery of adapter
6003 events directly to the vm, with KVM providing a backup delivery mechanism;
6004 KVM_S390_ZPCIOP_DEREG_AEN is used to subsequently disable interpretation of
6005 adapter event notifications.
6006
6007 The target zPCI function must also be specified via the "fh" field. For the
6008 KVM_S390_ZPCIOP_REG_AEN operation, additional information to establish firmware
6009 delivery must be provided via the "reg_aen" struct.
6010
6011 The "pad" and "reserved" fields may be used for future extensions and should be
6012 set to 0s by userspace.
6013
6014 5. The kvm_run structure
6015 ========================
6016
6017 Application code obtains a pointer to the kvm_run structure by
6018 mmap()ing a vcpu fd. From that point, application code can control
6019 execution by changing fields in kvm_run prior to calling the KVM_RUN
6020 ioctl, and obtain information about the reason KVM_RUN returned by
6021 looking up structure members.
6022
6023 ::
6024
6025 struct kvm_run {
6026 /* in */
6027 __u8 request_interrupt_window;
6028
6029 Request that KVM_RUN return when it becomes possible to inject external
6030 interrupts into the guest. Useful in conjunction with KVM_INTERRUPT.
6031
6032 ::
6033
6034 __u8 immediate_exit;
6035
6036 This field is polled once when KVM_RUN starts; if non-zero, KVM_RUN
6037 exits immediately, returning -EINTR. In the common scenario where a
6038 signal is used to "kick" a VCPU out of KVM_RUN, this field can be used
6039 to avoid usage of KVM_SET_SIGNAL_MASK, which has worse scalability.
6040 Rather than blocking the signal outside KVM_RUN, userspace can set up
6041 a signal handler that sets run->immediate_exit to a non-zero value.
6042
6043 This field is ignored if KVM_CAP_IMMEDIATE_EXIT is not available.
6044
6045 ::
6046
6047 __u8 padding1[6];
6048
6049 /* out */
6050 __u32 exit_reason;
6051
6052 When KVM_RUN has returned successfully (return value 0), this informs
6053 application code why KVM_RUN has returned. Allowable values for this
6054 field are detailed below.
6055
6056 ::
6057
6058 __u8 ready_for_interrupt_injection;
6059
6060 If request_interrupt_window has been specified, this field indicates
6061 an interrupt can be injected now with KVM_INTERRUPT.
6062
6063 ::
6064
6065 __u8 if_flag;
6066
6067 The value of the current interrupt flag. Only valid if in-kernel
6068 local APIC is not used.
6069
6070 ::
6071
6072 __u16 flags;
6073
6074 More architecture-specific flags detailing state of the VCPU that may
6075 affect the device's behavior. Current defined flags::
6076
6077 /* x86, set if the VCPU is in system management mode */
6078 #define KVM_RUN_X86_SMM (1 << 0)
6079 /* x86, set if bus lock detected in VM */
6080 #define KVM_RUN_BUS_LOCK (1 << 1)
6081 /* arm64, set for KVM_EXIT_DEBUG */
6082 #define KVM_DEBUG_ARCH_HSR_HIGH_VALID (1 << 0)
6083
6084 ::
6085
6086 /* in (pre_kvm_run), out (post_kvm_run) */
6087 __u64 cr8;
6088
6089 The value of the cr8 register. Only valid if in-kernel local APIC is
6090 not used. Both input and output.
6091
6092 ::
6093
6094 __u64 apic_base;
6095
6096 The value of the APIC BASE msr. Only valid if in-kernel local
6097 APIC is not used. Both input and output.
6098
6099 ::
6100
6101 union {
6102 /* KVM_EXIT_UNKNOWN */
6103 struct {
6104 __u64 hardware_exit_reason;
6105 } hw;
6106
6107 If exit_reason is KVM_EXIT_UNKNOWN, the vcpu has exited due to unknown
6108 reasons. Further architecture-specific information is available in
6109 hardware_exit_reason.
6110
6111 ::
6112
6113 /* KVM_EXIT_FAIL_ENTRY */
6114 struct {
6115 __u64 hardware_entry_failure_reason;
6116 __u32 cpu; /* if KVM_LAST_CPU */
6117 } fail_entry;
6118
6119 If exit_reason is KVM_EXIT_FAIL_ENTRY, the vcpu could not be run due
6120 to unknown reasons. Further architecture-specific information is
6121 available in hardware_entry_failure_reason.
6122
6123 ::
6124
6125 /* KVM_EXIT_EXCEPTION */
6126 struct {
6127 __u32 exception;
6128 __u32 error_code;
6129 } ex;
6130
6131 Unused.
6132
6133 ::
6134
6135 /* KVM_EXIT_IO */
6136 struct {
6137 #define KVM_EXIT_IO_IN 0
6138 #define KVM_EXIT_IO_OUT 1
6139 __u8 direction;
6140 __u8 size; /* bytes */
6141 __u16 port;
6142 __u32 count;
6143 __u64 data_offset; /* relative to kvm_run start */
6144 } io;
6145
6146 If exit_reason is KVM_EXIT_IO, then the vcpu has
6147 executed a port I/O instruction which could not be satisfied by kvm.
6148 data_offset describes where the data is located (KVM_EXIT_IO_OUT) or
6149 where kvm expects application code to place the data for the next
6150 KVM_RUN invocation (KVM_EXIT_IO_IN). Data format is a packed array.
6151
6152 ::
6153
6154 /* KVM_EXIT_DEBUG */
6155 struct {
6156 struct kvm_debug_exit_arch arch;
6157 } debug;
6158
6159 If the exit_reason is KVM_EXIT_DEBUG, then a vcpu is processing a debug event
6160 for which architecture specific information is returned.
6161
6162 ::
6163
6164 /* KVM_EXIT_MMIO */
6165 struct {
6166 __u64 phys_addr;
6167 __u8 data[8];
6168 __u32 len;
6169 __u8 is_write;
6170 } mmio;
6171
6172 If exit_reason is KVM_EXIT_MMIO, then the vcpu has
6173 executed a memory-mapped I/O instruction which could not be satisfied
6174 by kvm. The 'data' member contains the written data if 'is_write' is
6175 true, and should be filled by application code otherwise.
6176
6177 The 'data' member contains, in its first 'len' bytes, the value as it would
6178 appear if the VCPU performed a load or store of the appropriate width directly
6179 to the byte array.
6180
6181 .. note::
6182
6183 For KVM_EXIT_IO, KVM_EXIT_MMIO, KVM_EXIT_OSI, KVM_EXIT_PAPR, KVM_EXIT_XEN,
6184 KVM_EXIT_EPR, KVM_EXIT_X86_RDMSR and KVM_EXIT_X86_WRMSR the corresponding
6185 operations are complete (and guest state is consistent) only after userspace
6186 has re-entered the kernel with KVM_RUN. The kernel side will first finish
6187 incomplete operations and then check for pending signals.
6188
6189 The pending state of the operation is not preserved in state which is
6190 visible to userspace, thus userspace should ensure that the operation is
6191 completed before performing a live migration. Userspace can re-enter the
6192 guest with an unmasked signal pending or with the immediate_exit field set
6193 to complete pending operations without allowing any further instructions
6194 to be executed.
6195
6196 ::
6197
6198 /* KVM_EXIT_HYPERCALL */
6199 struct {
6200 __u64 nr;
6201 __u64 args[6];
6202 __u64 ret;
6203 __u32 longmode;
6204 __u32 pad;
6205 } hypercall;
6206
6207 Unused. This was once used for 'hypercall to userspace'. To implement
6208 such functionality, use KVM_EXIT_IO (x86) or KVM_EXIT_MMIO (all except s390).
6209
6210 .. note:: KVM_EXIT_IO is significantly faster than KVM_EXIT_MMIO.
6211
6212 ::
6213
6214 /* KVM_EXIT_TPR_ACCESS */
6215 struct {
6216 __u64 rip;
6217 __u32 is_write;
6218 __u32 pad;
6219 } tpr_access;
6220
6221 To be documented (KVM_TPR_ACCESS_REPORTING).
6222
6223 ::
6224
6225 /* KVM_EXIT_S390_SIEIC */
6226 struct {
6227 __u8 icptcode;
6228 __u64 mask; /* psw upper half */
6229 __u64 addr; /* psw lower half */
6230 __u16 ipa;
6231 __u32 ipb;
6232 } s390_sieic;
6233
6234 s390 specific.
6235
6236 ::
6237
6238 /* KVM_EXIT_S390_RESET */
6239 #define KVM_S390_RESET_POR 1
6240 #define KVM_S390_RESET_CLEAR 2
6241 #define KVM_S390_RESET_SUBSYSTEM 4
6242 #define KVM_S390_RESET_CPU_INIT 8
6243 #define KVM_S390_RESET_IPL 16
6244 __u64 s390_reset_flags;
6245
6246 s390 specific.
6247
6248 ::
6249
6250 /* KVM_EXIT_S390_UCONTROL */
6251 struct {
6252 __u64 trans_exc_code;
6253 __u32 pgm_code;
6254 } s390_ucontrol;
6255
6256 s390 specific. A page fault has occurred for a user controlled virtual
6257 machine (KVM_VM_S390_UNCONTROL) on it's host page table that cannot be
6258 resolved by the kernel.
6259 The program code and the translation exception code that were placed
6260 in the cpu's lowcore are presented here as defined by the z Architecture
6261 Principles of Operation Book in the Chapter for Dynamic Address Translation
6262 (DAT)
6263
6264 ::
6265
6266 /* KVM_EXIT_DCR */
6267 struct {
6268 __u32 dcrn;
6269 __u32 data;
6270 __u8 is_write;
6271 } dcr;
6272
6273 Deprecated - was used for 440 KVM.
6274
6275 ::
6276
6277 /* KVM_EXIT_OSI */
6278 struct {
6279 __u64 gprs[32];
6280 } osi;
6281
6282 MOL uses a special hypercall interface it calls 'OSI'. To enable it, we catch
6283 hypercalls and exit with this exit struct that contains all the guest gprs.
6284
6285 If exit_reason is KVM_EXIT_OSI, then the vcpu has triggered such a hypercall.
6286 Userspace can now handle the hypercall and when it's done modify the gprs as
6287 necessary. Upon guest entry all guest GPRs will then be replaced by the values
6288 in this struct.
6289
6290 ::
6291
6292 /* KVM_EXIT_PAPR_HCALL */
6293 struct {
6294 __u64 nr;
6295 __u64 ret;
6296 __u64 args[9];
6297 } papr_hcall;
6298
6299 This is used on 64-bit PowerPC when emulating a pSeries partition,
6300 e.g. with the 'pseries' machine type in qemu. It occurs when the
6301 guest does a hypercall using the 'sc 1' instruction. The 'nr' field
6302 contains the hypercall number (from the guest R3), and 'args' contains
6303 the arguments (from the guest R4 - R12). Userspace should put the
6304 return code in 'ret' and any extra returned values in args[].
6305 The possible hypercalls are defined in the Power Architecture Platform
6306 Requirements (PAPR) document available from www.power.org (free
6307 developer registration required to access it).
6308
6309 ::
6310
6311 /* KVM_EXIT_S390_TSCH */
6312 struct {
6313 __u16 subchannel_id;
6314 __u16 subchannel_nr;
6315 __u32 io_int_parm;
6316 __u32 io_int_word;
6317 __u32 ipb;
6318 __u8 dequeued;
6319 } s390_tsch;
6320
6321 s390 specific. This exit occurs when KVM_CAP_S390_CSS_SUPPORT has been enabled
6322 and TEST SUBCHANNEL was intercepted. If dequeued is set, a pending I/O
6323 interrupt for the target subchannel has been dequeued and subchannel_id,
6324 subchannel_nr, io_int_parm and io_int_word contain the parameters for that
6325 interrupt. ipb is needed for instruction parameter decoding.
6326
6327 ::
6328
6329 /* KVM_EXIT_EPR */
6330 struct {
6331 __u32 epr;
6332 } epr;
6333
6334 On FSL BookE PowerPC chips, the interrupt controller has a fast patch
6335 interrupt acknowledge path to the core. When the core successfully
6336 delivers an interrupt, it automatically populates the EPR register with
6337 the interrupt vector number and acknowledges the interrupt inside
6338 the interrupt controller.
6339
6340 In case the interrupt controller lives in user space, we need to do
6341 the interrupt acknowledge cycle through it to fetch the next to be
6342 delivered interrupt vector using this exit.
6343
6344 It gets triggered whenever both KVM_CAP_PPC_EPR are enabled and an
6345 external interrupt has just been delivered into the guest. User space
6346 should put the acknowledged interrupt vector into the 'epr' field.
6347
6348 ::
6349
6350 /* KVM_EXIT_SYSTEM_EVENT */
6351 struct {
6352 #define KVM_SYSTEM_EVENT_SHUTDOWN 1
6353 #define KVM_SYSTEM_EVENT_RESET 2
6354 #define KVM_SYSTEM_EVENT_CRASH 3
6355 #define KVM_SYSTEM_EVENT_WAKEUP 4
6356 #define KVM_SYSTEM_EVENT_SUSPEND 5
6357 #define KVM_SYSTEM_EVENT_SEV_TERM 6
6358 __u32 type;
6359 __u32 ndata;
6360 __u64 data[16];
6361 } system_event;
6362
6363 If exit_reason is KVM_EXIT_SYSTEM_EVENT then the vcpu has triggered
6364 a system-level event using some architecture specific mechanism (hypercall
6365 or some special instruction). In case of ARM64, this is triggered using
6366 HVC instruction based PSCI call from the vcpu.
6367
6368 The 'type' field describes the system-level event type.
6369 Valid values for 'type' are:
6370
6371 - KVM_SYSTEM_EVENT_SHUTDOWN -- the guest has requested a shutdown of the
6372 VM. Userspace is not obliged to honour this, and if it does honour
6373 this does not need to destroy the VM synchronously (ie it may call
6374 KVM_RUN again before shutdown finally occurs).
6375 - KVM_SYSTEM_EVENT_RESET -- the guest has requested a reset of the VM.
6376 As with SHUTDOWN, userspace can choose to ignore the request, or
6377 to schedule the reset to occur in the future and may call KVM_RUN again.
6378 - KVM_SYSTEM_EVENT_CRASH -- the guest crash occurred and the guest
6379 has requested a crash condition maintenance. Userspace can choose
6380 to ignore the request, or to gather VM memory core dump and/or
6381 reset/shutdown of the VM.
6382 - KVM_SYSTEM_EVENT_SEV_TERM -- an AMD SEV guest requested termination.
6383 The guest physical address of the guest's GHCB is stored in `data[0]`.
6384 - KVM_SYSTEM_EVENT_WAKEUP -- the exiting vCPU is in a suspended state and
6385 KVM has recognized a wakeup event. Userspace may honor this event by
6386 marking the exiting vCPU as runnable, or deny it and call KVM_RUN again.
6387 - KVM_SYSTEM_EVENT_SUSPEND -- the guest has requested a suspension of
6388 the VM.
6389
6390 If KVM_CAP_SYSTEM_EVENT_DATA is present, the 'data' field can contain
6391 architecture specific information for the system-level event. Only
6392 the first `ndata` items (possibly zero) of the data array are valid.
6393
6394 - for arm64, data[0] is set to KVM_SYSTEM_EVENT_RESET_FLAG_PSCI_RESET2 if
6395 the guest issued a SYSTEM_RESET2 call according to v1.1 of the PSCI
6396 specification.
6397
6398 - for RISC-V, data[0] is set to the value of the second argument of the
6399 ``sbi_system_reset`` call.
6400
6401 Previous versions of Linux defined a `flags` member in this struct. The
6402 field is now aliased to `data[0]`. Userspace can assume that it is only
6403 written if ndata is greater than 0.
6404
6405 For arm/arm64:
6406 --------------
6407
6408 KVM_SYSTEM_EVENT_SUSPEND exits are enabled with the
6409 KVM_CAP_ARM_SYSTEM_SUSPEND VM capability. If a guest invokes the PSCI
6410 SYSTEM_SUSPEND function, KVM will exit to userspace with this event
6411 type.
6412
6413 It is the sole responsibility of userspace to implement the PSCI
6414 SYSTEM_SUSPEND call according to ARM DEN0022D.b 5.19 "SYSTEM_SUSPEND".
6415 KVM does not change the vCPU's state before exiting to userspace, so
6416 the call parameters are left in-place in the vCPU registers.
6417
6418 Userspace is _required_ to take action for such an exit. It must
6419 either:
6420
6421 - Honor the guest request to suspend the VM. Userspace can request
6422 in-kernel emulation of suspension by setting the calling vCPU's
6423 state to KVM_MP_STATE_SUSPENDED. Userspace must configure the vCPU's
6424 state according to the parameters passed to the PSCI function when
6425 the calling vCPU is resumed. See ARM DEN0022D.b 5.19.1 "Intended use"
6426 for details on the function parameters.
6427
6428 - Deny the guest request to suspend the VM. See ARM DEN0022D.b 5.19.2
6429 "Caller responsibilities" for possible return values.
6430
6431 ::
6432
6433 /* KVM_EXIT_IOAPIC_EOI */
6434 struct {
6435 __u8 vector;
6436 } eoi;
6437
6438 Indicates that the VCPU's in-kernel local APIC received an EOI for a
6439 level-triggered IOAPIC interrupt. This exit only triggers when the
6440 IOAPIC is implemented in userspace (i.e. KVM_CAP_SPLIT_IRQCHIP is enabled);
6441 the userspace IOAPIC should process the EOI and retrigger the interrupt if
6442 it is still asserted. Vector is the LAPIC interrupt vector for which the
6443 EOI was received.
6444
6445 ::
6446
6447 struct kvm_hyperv_exit {
6448 #define KVM_EXIT_HYPERV_SYNIC 1
6449 #define KVM_EXIT_HYPERV_HCALL 2
6450 #define KVM_EXIT_HYPERV_SYNDBG 3
6451 __u32 type;
6452 __u32 pad1;
6453 union {
6454 struct {
6455 __u32 msr;
6456 __u32 pad2;
6457 __u64 control;
6458 __u64 evt_page;
6459 __u64 msg_page;
6460 } synic;
6461 struct {
6462 __u64 input;
6463 __u64 result;
6464 __u64 params[2];
6465 } hcall;
6466 struct {
6467 __u32 msr;
6468 __u32 pad2;
6469 __u64 control;
6470 __u64 status;
6471 __u64 send_page;
6472 __u64 recv_page;
6473 __u64 pending_page;
6474 } syndbg;
6475 } u;
6476 };
6477 /* KVM_EXIT_HYPERV */
6478 struct kvm_hyperv_exit hyperv;
6479
6480 Indicates that the VCPU exits into userspace to process some tasks
6481 related to Hyper-V emulation.
6482
6483 Valid values for 'type' are:
6484
6485 - KVM_EXIT_HYPERV_SYNIC -- synchronously notify user-space about
6486
6487 Hyper-V SynIC state change. Notification is used to remap SynIC
6488 event/message pages and to enable/disable SynIC messages/events processing
6489 in userspace.
6490
6491 - KVM_EXIT_HYPERV_SYNDBG -- synchronously notify user-space about
6492
6493 Hyper-V Synthetic debugger state change. Notification is used to either update
6494 the pending_page location or to send a control command (send the buffer located
6495 in send_page or recv a buffer to recv_page).
6496
6497 ::
6498
6499 /* KVM_EXIT_ARM_NISV */
6500 struct {
6501 __u64 esr_iss;
6502 __u64 fault_ipa;
6503 } arm_nisv;
6504
6505 Used on arm64 systems. If a guest accesses memory not in a memslot,
6506 KVM will typically return to userspace and ask it to do MMIO emulation on its
6507 behalf. However, for certain classes of instructions, no instruction decode
6508 (direction, length of memory access) is provided, and fetching and decoding
6509 the instruction from the VM is overly complicated to live in the kernel.
6510
6511 Historically, when this situation occurred, KVM would print a warning and kill
6512 the VM. KVM assumed that if the guest accessed non-memslot memory, it was
6513 trying to do I/O, which just couldn't be emulated, and the warning message was
6514 phrased accordingly. However, what happened more often was that a guest bug
6515 caused access outside the guest memory areas which should lead to a more
6516 meaningful warning message and an external abort in the guest, if the access
6517 did not fall within an I/O window.
6518
6519 Userspace implementations can query for KVM_CAP_ARM_NISV_TO_USER, and enable
6520 this capability at VM creation. Once this is done, these types of errors will
6521 instead return to userspace with KVM_EXIT_ARM_NISV, with the valid bits from
6522 the ESR_EL2 in the esr_iss field, and the faulting IPA in the fault_ipa field.
6523 Userspace can either fix up the access if it's actually an I/O access by
6524 decoding the instruction from guest memory (if it's very brave) and continue
6525 executing the guest, or it can decide to suspend, dump, or restart the guest.
6526
6527 Note that KVM does not skip the faulting instruction as it does for
6528 KVM_EXIT_MMIO, but userspace has to emulate any change to the processing state
6529 if it decides to decode and emulate the instruction.
6530
6531 ::
6532
6533 /* KVM_EXIT_X86_RDMSR / KVM_EXIT_X86_WRMSR */
6534 struct {
6535 __u8 error; /* user -> kernel */
6536 __u8 pad[7];
6537 __u32 reason; /* kernel -> user */
6538 __u32 index; /* kernel -> user */
6539 __u64 data; /* kernel <-> user */
6540 } msr;
6541
6542 Used on x86 systems. When the VM capability KVM_CAP_X86_USER_SPACE_MSR is
6543 enabled, MSR accesses to registers that would invoke a #GP by KVM kernel code
6544 will instead trigger a KVM_EXIT_X86_RDMSR exit for reads and KVM_EXIT_X86_WRMSR
6545 exit for writes.
6546
6547 The "reason" field specifies why the MSR trap occurred. User space will only
6548 receive MSR exit traps when a particular reason was requested during through
6549 ENABLE_CAP. Currently valid exit reasons are:
6550
6551 KVM_MSR_EXIT_REASON_UNKNOWN - access to MSR that is unknown to KVM
6552 KVM_MSR_EXIT_REASON_INVAL - access to invalid MSRs or reserved bits
6553 KVM_MSR_EXIT_REASON_FILTER - access blocked by KVM_X86_SET_MSR_FILTER
6554
6555 For KVM_EXIT_X86_RDMSR, the "index" field tells user space which MSR the guest
6556 wants to read. To respond to this request with a successful read, user space
6557 writes the respective data into the "data" field and must continue guest
6558 execution to ensure the read data is transferred into guest register state.
6559
6560 If the RDMSR request was unsuccessful, user space indicates that with a "1" in
6561 the "error" field. This will inject a #GP into the guest when the VCPU is
6562 executed again.
6563
6564 For KVM_EXIT_X86_WRMSR, the "index" field tells user space which MSR the guest
6565 wants to write. Once finished processing the event, user space must continue
6566 vCPU execution. If the MSR write was unsuccessful, user space also sets the
6567 "error" field to "1".
6568
6569 ::
6570
6571
6572 struct kvm_xen_exit {
6573 #define KVM_EXIT_XEN_HCALL 1
6574 __u32 type;
6575 union {
6576 struct {
6577 __u32 longmode;
6578 __u32 cpl;
6579 __u64 input;
6580 __u64 result;
6581 __u64 params[6];
6582 } hcall;
6583 } u;
6584 };
6585 /* KVM_EXIT_XEN */
6586 struct kvm_hyperv_exit xen;
6587
6588 Indicates that the VCPU exits into userspace to process some tasks
6589 related to Xen emulation.
6590
6591 Valid values for 'type' are:
6592
6593 - KVM_EXIT_XEN_HCALL -- synchronously notify user-space about Xen hypercall.
6594 Userspace is expected to place the hypercall result into the appropriate
6595 field before invoking KVM_RUN again.
6596
6597 ::
6598
6599 /* KVM_EXIT_RISCV_SBI */
6600 struct {
6601 unsigned long extension_id;
6602 unsigned long function_id;
6603 unsigned long args[6];
6604 unsigned long ret[2];
6605 } riscv_sbi;
6606
6607 If exit reason is KVM_EXIT_RISCV_SBI then it indicates that the VCPU has
6608 done a SBI call which is not handled by KVM RISC-V kernel module. The details
6609 of the SBI call are available in 'riscv_sbi' member of kvm_run structure. The
6610 'extension_id' field of 'riscv_sbi' represents SBI extension ID whereas the
6611 'function_id' field represents function ID of given SBI extension. The 'args'
6612 array field of 'riscv_sbi' represents parameters for the SBI call and 'ret'
6613 array field represents return values. The userspace should update the return
6614 values of SBI call before resuming the VCPU. For more details on RISC-V SBI
6615 spec refer, https://github.com/riscv/riscv-sbi-doc.
6616
6617 ::
6618
6619 /* KVM_EXIT_NOTIFY */
6620 struct {
6621 #define KVM_NOTIFY_CONTEXT_INVALID (1 << 0)
6622 __u32 flags;
6623 } notify;
6624
6625 Used on x86 systems. When the VM capability KVM_CAP_X86_NOTIFY_VMEXIT is
6626 enabled, a VM exit generated if no event window occurs in VM non-root mode
6627 for a specified amount of time. Once KVM_X86_NOTIFY_VMEXIT_USER is set when
6628 enabling the cap, it would exit to userspace with the exit reason
6629 KVM_EXIT_NOTIFY for further handling. The "flags" field contains more
6630 detailed info.
6631
6632 The valid value for 'flags' is:
6633
6634 - KVM_NOTIFY_CONTEXT_INVALID -- the VM context is corrupted and not valid
6635 in VMCS. It would run into unknown result if resume the target VM.
6636
6637 ::
6638
6639 /* Fix the size of the union. */
6640 char padding[256];
6641 };
6642
6643 /*
6644 * shared registers between kvm and userspace.
6645 * kvm_valid_regs specifies the register classes set by the host
6646 * kvm_dirty_regs specified the register classes dirtied by userspace
6647 * struct kvm_sync_regs is architecture specific, as well as the
6648 * bits for kvm_valid_regs and kvm_dirty_regs
6649 */
6650 __u64 kvm_valid_regs;
6651 __u64 kvm_dirty_regs;
6652 union {
6653 struct kvm_sync_regs regs;
6654 char padding[SYNC_REGS_SIZE_BYTES];
6655 } s;
6656
6657 If KVM_CAP_SYNC_REGS is defined, these fields allow userspace to access
6658 certain guest registers without having to call SET/GET_*REGS. Thus we can
6659 avoid some system call overhead if userspace has to handle the exit.
6660 Userspace can query the validity of the structure by checking
6661 kvm_valid_regs for specific bits. These bits are architecture specific
6662 and usually define the validity of a groups of registers. (e.g. one bit
6663 for general purpose registers)
6664
6665 Please note that the kernel is allowed to use the kvm_run structure as the
6666 primary storage for certain register types. Therefore, the kernel may use the
6667 values in kvm_run even if the corresponding bit in kvm_dirty_regs is not set.
6668
6669 ::
6670
6671 };
6672
6673
6674
6675 6. Capabilities that can be enabled on vCPUs
6676 ============================================
6677
6678 There are certain capabilities that change the behavior of the virtual CPU or
6679 the virtual machine when enabled. To enable them, please see section 4.37.
6680 Below you can find a list of capabilities and what their effect on the vCPU or
6681 the virtual machine is when enabling them.
6682
6683 The following information is provided along with the description:
6684
6685 Architectures:
6686 which instruction set architectures provide this ioctl.
6687 x86 includes both i386 and x86_64.
6688
6689 Target:
6690 whether this is a per-vcpu or per-vm capability.
6691
6692 Parameters:
6693 what parameters are accepted by the capability.
6694
6695 Returns:
6696 the return value. General error numbers (EBADF, ENOMEM, EINVAL)
6697 are not detailed, but errors with specific meanings are.
6698
6699
6700 6.1 KVM_CAP_PPC_OSI
6701 -------------------
6702
6703 :Architectures: ppc
6704 :Target: vcpu
6705 :Parameters: none
6706 :Returns: 0 on success; -1 on error
6707
6708 This capability enables interception of OSI hypercalls that otherwise would
6709 be treated as normal system calls to be injected into the guest. OSI hypercalls
6710 were invented by Mac-on-Linux to have a standardized communication mechanism
6711 between the guest and the host.
6712
6713 When this capability is enabled, KVM_EXIT_OSI can occur.
6714
6715
6716 6.2 KVM_CAP_PPC_PAPR
6717 --------------------
6718
6719 :Architectures: ppc
6720 :Target: vcpu
6721 :Parameters: none
6722 :Returns: 0 on success; -1 on error
6723
6724 This capability enables interception of PAPR hypercalls. PAPR hypercalls are
6725 done using the hypercall instruction "sc 1".
6726
6727 It also sets the guest privilege level to "supervisor" mode. Usually the guest
6728 runs in "hypervisor" privilege mode with a few missing features.
6729
6730 In addition to the above, it changes the semantics of SDR1. In this mode, the
6731 HTAB address part of SDR1 contains an HVA instead of a GPA, as PAPR keeps the
6732 HTAB invisible to the guest.
6733
6734 When this capability is enabled, KVM_EXIT_PAPR_HCALL can occur.
6735
6736
6737 6.3 KVM_CAP_SW_TLB
6738 ------------------
6739
6740 :Architectures: ppc
6741 :Target: vcpu
6742 :Parameters: args[0] is the address of a struct kvm_config_tlb
6743 :Returns: 0 on success; -1 on error
6744
6745 ::
6746
6747 struct kvm_config_tlb {
6748 __u64 params;
6749 __u64 array;
6750 __u32 mmu_type;
6751 __u32 array_len;
6752 };
6753
6754 Configures the virtual CPU's TLB array, establishing a shared memory area
6755 between userspace and KVM. The "params" and "array" fields are userspace
6756 addresses of mmu-type-specific data structures. The "array_len" field is an
6757 safety mechanism, and should be set to the size in bytes of the memory that
6758 userspace has reserved for the array. It must be at least the size dictated
6759 by "mmu_type" and "params".
6760
6761 While KVM_RUN is active, the shared region is under control of KVM. Its
6762 contents are undefined, and any modification by userspace results in
6763 boundedly undefined behavior.
6764
6765 On return from KVM_RUN, the shared region will reflect the current state of
6766 the guest's TLB. If userspace makes any changes, it must call KVM_DIRTY_TLB
6767 to tell KVM which entries have been changed, prior to calling KVM_RUN again
6768 on this vcpu.
6769
6770 For mmu types KVM_MMU_FSL_BOOKE_NOHV and KVM_MMU_FSL_BOOKE_HV:
6771
6772 - The "params" field is of type "struct kvm_book3e_206_tlb_params".
6773 - The "array" field points to an array of type "struct
6774 kvm_book3e_206_tlb_entry".
6775 - The array consists of all entries in the first TLB, followed by all
6776 entries in the second TLB.
6777 - Within a TLB, entries are ordered first by increasing set number. Within a
6778 set, entries are ordered by way (increasing ESEL).
6779 - The hash for determining set number in TLB0 is: (MAS2 >> 12) & (num_sets - 1)
6780 where "num_sets" is the tlb_sizes[] value divided by the tlb_ways[] value.
6781 - The tsize field of mas1 shall be set to 4K on TLB0, even though the
6782 hardware ignores this value for TLB0.
6783
6784 6.4 KVM_CAP_S390_CSS_SUPPORT
6785 ----------------------------
6786
6787 :Architectures: s390
6788 :Target: vcpu
6789 :Parameters: none
6790 :Returns: 0 on success; -1 on error
6791
6792 This capability enables support for handling of channel I/O instructions.
6793
6794 TEST PENDING INTERRUPTION and the interrupt portion of TEST SUBCHANNEL are
6795 handled in-kernel, while the other I/O instructions are passed to userspace.
6796
6797 When this capability is enabled, KVM_EXIT_S390_TSCH will occur on TEST
6798 SUBCHANNEL intercepts.
6799
6800 Note that even though this capability is enabled per-vcpu, the complete
6801 virtual machine is affected.
6802
6803 6.5 KVM_CAP_PPC_EPR
6804 -------------------
6805
6806 :Architectures: ppc
6807 :Target: vcpu
6808 :Parameters: args[0] defines whether the proxy facility is active
6809 :Returns: 0 on success; -1 on error
6810
6811 This capability enables or disables the delivery of interrupts through the
6812 external proxy facility.
6813
6814 When enabled (args[0] != 0), every time the guest gets an external interrupt
6815 delivered, it automatically exits into user space with a KVM_EXIT_EPR exit
6816 to receive the topmost interrupt vector.
6817
6818 When disabled (args[0] == 0), behavior is as if this facility is unsupported.
6819
6820 When this capability is enabled, KVM_EXIT_EPR can occur.
6821
6822 6.6 KVM_CAP_IRQ_MPIC
6823 --------------------
6824
6825 :Architectures: ppc
6826 :Parameters: args[0] is the MPIC device fd;
6827 args[1] is the MPIC CPU number for this vcpu
6828
6829 This capability connects the vcpu to an in-kernel MPIC device.
6830
6831 6.7 KVM_CAP_IRQ_XICS
6832 --------------------
6833
6834 :Architectures: ppc
6835 :Target: vcpu
6836 :Parameters: args[0] is the XICS device fd;
6837 args[1] is the XICS CPU number (server ID) for this vcpu
6838
6839 This capability connects the vcpu to an in-kernel XICS device.
6840
6841 6.8 KVM_CAP_S390_IRQCHIP
6842 ------------------------
6843
6844 :Architectures: s390
6845 :Target: vm
6846 :Parameters: none
6847
6848 This capability enables the in-kernel irqchip for s390. Please refer to
6849 "4.24 KVM_CREATE_IRQCHIP" for details.
6850
6851 6.9 KVM_CAP_MIPS_FPU
6852 --------------------
6853
6854 :Architectures: mips
6855 :Target: vcpu
6856 :Parameters: args[0] is reserved for future use (should be 0).
6857
6858 This capability allows the use of the host Floating Point Unit by the guest. It
6859 allows the Config1.FP bit to be set to enable the FPU in the guest. Once this is
6860 done the ``KVM_REG_MIPS_FPR_*`` and ``KVM_REG_MIPS_FCR_*`` registers can be
6861 accessed (depending on the current guest FPU register mode), and the Status.FR,
6862 Config5.FRE bits are accessible via the KVM API and also from the guest,
6863 depending on them being supported by the FPU.
6864
6865 6.10 KVM_CAP_MIPS_MSA
6866 ---------------------
6867
6868 :Architectures: mips
6869 :Target: vcpu
6870 :Parameters: args[0] is reserved for future use (should be 0).
6871
6872 This capability allows the use of the MIPS SIMD Architecture (MSA) by the guest.
6873 It allows the Config3.MSAP bit to be set to enable the use of MSA by the guest.
6874 Once this is done the ``KVM_REG_MIPS_VEC_*`` and ``KVM_REG_MIPS_MSA_*``
6875 registers can be accessed, and the Config5.MSAEn bit is accessible via the
6876 KVM API and also from the guest.
6877
6878 6.74 KVM_CAP_SYNC_REGS
6879 ----------------------
6880
6881 :Architectures: s390, x86
6882 :Target: s390: always enabled, x86: vcpu
6883 :Parameters: none
6884 :Returns: x86: KVM_CHECK_EXTENSION returns a bit-array indicating which register
6885 sets are supported
6886 (bitfields defined in arch/x86/include/uapi/asm/kvm.h).
6887
6888 As described above in the kvm_sync_regs struct info in section 5 (kvm_run):
6889 KVM_CAP_SYNC_REGS "allow[s] userspace to access certain guest registers
6890 without having to call SET/GET_*REGS". This reduces overhead by eliminating
6891 repeated ioctl calls for setting and/or getting register values. This is
6892 particularly important when userspace is making synchronous guest state
6893 modifications, e.g. when emulating and/or intercepting instructions in
6894 userspace.
6895
6896 For s390 specifics, please refer to the source code.
6897
6898 For x86:
6899
6900 - the register sets to be copied out to kvm_run are selectable
6901 by userspace (rather that all sets being copied out for every exit).
6902 - vcpu_events are available in addition to regs and sregs.
6903
6904 For x86, the 'kvm_valid_regs' field of struct kvm_run is overloaded to
6905 function as an input bit-array field set by userspace to indicate the
6906 specific register sets to be copied out on the next exit.
6907
6908 To indicate when userspace has modified values that should be copied into
6909 the vCPU, the all architecture bitarray field, 'kvm_dirty_regs' must be set.
6910 This is done using the same bitflags as for the 'kvm_valid_regs' field.
6911 If the dirty bit is not set, then the register set values will not be copied
6912 into the vCPU even if they've been modified.
6913
6914 Unused bitfields in the bitarrays must be set to zero.
6915
6916 ::
6917
6918 struct kvm_sync_regs {
6919 struct kvm_regs regs;
6920 struct kvm_sregs sregs;
6921 struct kvm_vcpu_events events;
6922 };
6923
6924 6.75 KVM_CAP_PPC_IRQ_XIVE
6925 -------------------------
6926
6927 :Architectures: ppc
6928 :Target: vcpu
6929 :Parameters: args[0] is the XIVE device fd;
6930 args[1] is the XIVE CPU number (server ID) for this vcpu
6931
6932 This capability connects the vcpu to an in-kernel XIVE device.
6933
6934 7. Capabilities that can be enabled on VMs
6935 ==========================================
6936
6937 There are certain capabilities that change the behavior of the virtual
6938 machine when enabled. To enable them, please see section 4.37. Below
6939 you can find a list of capabilities and what their effect on the VM
6940 is when enabling them.
6941
6942 The following information is provided along with the description:
6943
6944 Architectures:
6945 which instruction set architectures provide this ioctl.
6946 x86 includes both i386 and x86_64.
6947
6948 Parameters:
6949 what parameters are accepted by the capability.
6950
6951 Returns:
6952 the return value. General error numbers (EBADF, ENOMEM, EINVAL)
6953 are not detailed, but errors with specific meanings are.
6954
6955
6956 7.1 KVM_CAP_PPC_ENABLE_HCALL
6957 ----------------------------
6958
6959 :Architectures: ppc
6960 :Parameters: args[0] is the sPAPR hcall number;
6961 args[1] is 0 to disable, 1 to enable in-kernel handling
6962
6963 This capability controls whether individual sPAPR hypercalls (hcalls)
6964 get handled by the kernel or not. Enabling or disabling in-kernel
6965 handling of an hcall is effective across the VM. On creation, an
6966 initial set of hcalls are enabled for in-kernel handling, which
6967 consists of those hcalls for which in-kernel handlers were implemented
6968 before this capability was implemented. If disabled, the kernel will
6969 not to attempt to handle the hcall, but will always exit to userspace
6970 to handle it. Note that it may not make sense to enable some and
6971 disable others of a group of related hcalls, but KVM does not prevent
6972 userspace from doing that.
6973
6974 If the hcall number specified is not one that has an in-kernel
6975 implementation, the KVM_ENABLE_CAP ioctl will fail with an EINVAL
6976 error.
6977
6978 7.2 KVM_CAP_S390_USER_SIGP
6979 --------------------------
6980
6981 :Architectures: s390
6982 :Parameters: none
6983
6984 This capability controls which SIGP orders will be handled completely in user
6985 space. With this capability enabled, all fast orders will be handled completely
6986 in the kernel:
6987
6988 - SENSE
6989 - SENSE RUNNING
6990 - EXTERNAL CALL
6991 - EMERGENCY SIGNAL
6992 - CONDITIONAL EMERGENCY SIGNAL
6993
6994 All other orders will be handled completely in user space.
6995
6996 Only privileged operation exceptions will be checked for in the kernel (or even
6997 in the hardware prior to interception). If this capability is not enabled, the
6998 old way of handling SIGP orders is used (partially in kernel and user space).
6999
7000 7.3 KVM_CAP_S390_VECTOR_REGISTERS
7001 ---------------------------------
7002
7003 :Architectures: s390
7004 :Parameters: none
7005 :Returns: 0 on success, negative value on error
7006
7007 Allows use of the vector registers introduced with z13 processor, and
7008 provides for the synchronization between host and user space. Will
7009 return -EINVAL if the machine does not support vectors.
7010
7011 7.4 KVM_CAP_S390_USER_STSI
7012 --------------------------
7013
7014 :Architectures: s390
7015 :Parameters: none
7016
7017 This capability allows post-handlers for the STSI instruction. After
7018 initial handling in the kernel, KVM exits to user space with
7019 KVM_EXIT_S390_STSI to allow user space to insert further data.
7020
7021 Before exiting to userspace, kvm handlers should fill in s390_stsi field of
7022 vcpu->run::
7023
7024 struct {
7025 __u64 addr;
7026 __u8 ar;
7027 __u8 reserved;
7028 __u8 fc;
7029 __u8 sel1;
7030 __u16 sel2;
7031 } s390_stsi;
7032
7033 @addr - guest address of STSI SYSIB
7034 @fc - function code
7035 @sel1 - selector 1
7036 @sel2 - selector 2
7037 @ar - access register number
7038
7039 KVM handlers should exit to userspace with rc = -EREMOTE.
7040
7041 7.5 KVM_CAP_SPLIT_IRQCHIP
7042 -------------------------
7043
7044 :Architectures: x86
7045 :Parameters: args[0] - number of routes reserved for userspace IOAPICs
7046 :Returns: 0 on success, -1 on error
7047
7048 Create a local apic for each processor in the kernel. This can be used
7049 instead of KVM_CREATE_IRQCHIP if the userspace VMM wishes to emulate the
7050 IOAPIC and PIC (and also the PIT, even though this has to be enabled
7051 separately).
7052
7053 This capability also enables in kernel routing of interrupt requests;
7054 when KVM_CAP_SPLIT_IRQCHIP only routes of KVM_IRQ_ROUTING_MSI type are
7055 used in the IRQ routing table. The first args[0] MSI routes are reserved
7056 for the IOAPIC pins. Whenever the LAPIC receives an EOI for these routes,
7057 a KVM_EXIT_IOAPIC_EOI vmexit will be reported to userspace.
7058
7059 Fails if VCPU has already been created, or if the irqchip is already in the
7060 kernel (i.e. KVM_CREATE_IRQCHIP has already been called).
7061
7062 7.6 KVM_CAP_S390_RI
7063 -------------------
7064
7065 :Architectures: s390
7066 :Parameters: none
7067
7068 Allows use of runtime-instrumentation introduced with zEC12 processor.
7069 Will return -EINVAL if the machine does not support runtime-instrumentation.
7070 Will return -EBUSY if a VCPU has already been created.
7071
7072 7.7 KVM_CAP_X2APIC_API
7073 ----------------------
7074
7075 :Architectures: x86
7076 :Parameters: args[0] - features that should be enabled
7077 :Returns: 0 on success, -EINVAL when args[0] contains invalid features
7078
7079 Valid feature flags in args[0] are::
7080
7081 #define KVM_X2APIC_API_USE_32BIT_IDS (1ULL << 0)
7082 #define KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK (1ULL << 1)
7083
7084 Enabling KVM_X2APIC_API_USE_32BIT_IDS changes the behavior of
7085 KVM_SET_GSI_ROUTING, KVM_SIGNAL_MSI, KVM_SET_LAPIC, and KVM_GET_LAPIC,
7086 allowing the use of 32-bit APIC IDs. See KVM_CAP_X2APIC_API in their
7087 respective sections.
7088
7089 KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK must be enabled for x2APIC to work
7090 in logical mode or with more than 255 VCPUs. Otherwise, KVM treats 0xff
7091 as a broadcast even in x2APIC mode in order to support physical x2APIC
7092 without interrupt remapping. This is undesirable in logical mode,
7093 where 0xff represents CPUs 0-7 in cluster 0.
7094
7095 7.8 KVM_CAP_S390_USER_INSTR0
7096 ----------------------------
7097
7098 :Architectures: s390
7099 :Parameters: none
7100
7101 With this capability enabled, all illegal instructions 0x0000 (2 bytes) will
7102 be intercepted and forwarded to user space. User space can use this
7103 mechanism e.g. to realize 2-byte software breakpoints. The kernel will
7104 not inject an operating exception for these instructions, user space has
7105 to take care of that.
7106
7107 This capability can be enabled dynamically even if VCPUs were already
7108 created and are running.
7109
7110 7.9 KVM_CAP_S390_GS
7111 -------------------
7112
7113 :Architectures: s390
7114 :Parameters: none
7115 :Returns: 0 on success; -EINVAL if the machine does not support
7116 guarded storage; -EBUSY if a VCPU has already been created.
7117
7118 Allows use of guarded storage for the KVM guest.
7119
7120 7.10 KVM_CAP_S390_AIS
7121 ---------------------
7122
7123 :Architectures: s390
7124 :Parameters: none
7125
7126 Allow use of adapter-interruption suppression.
7127 :Returns: 0 on success; -EBUSY if a VCPU has already been created.
7128
7129 7.11 KVM_CAP_PPC_SMT
7130 --------------------
7131
7132 :Architectures: ppc
7133 :Parameters: vsmt_mode, flags
7134
7135 Enabling this capability on a VM provides userspace with a way to set
7136 the desired virtual SMT mode (i.e. the number of virtual CPUs per
7137 virtual core). The virtual SMT mode, vsmt_mode, must be a power of 2
7138 between 1 and 8. On POWER8, vsmt_mode must also be no greater than
7139 the number of threads per subcore for the host. Currently flags must
7140 be 0. A successful call to enable this capability will result in
7141 vsmt_mode being returned when the KVM_CAP_PPC_SMT capability is
7142 subsequently queried for the VM. This capability is only supported by
7143 HV KVM, and can only be set before any VCPUs have been created.
7144 The KVM_CAP_PPC_SMT_POSSIBLE capability indicates which virtual SMT
7145 modes are available.
7146
7147 7.12 KVM_CAP_PPC_FWNMI
7148 ----------------------
7149
7150 :Architectures: ppc
7151 :Parameters: none
7152
7153 With this capability a machine check exception in the guest address
7154 space will cause KVM to exit the guest with NMI exit reason. This
7155 enables QEMU to build error log and branch to guest kernel registered
7156 machine check handling routine. Without this capability KVM will
7157 branch to guests' 0x200 interrupt vector.
7158
7159 7.13 KVM_CAP_X86_DISABLE_EXITS
7160 ------------------------------
7161
7162 :Architectures: x86
7163 :Parameters: args[0] defines which exits are disabled
7164 :Returns: 0 on success, -EINVAL when args[0] contains invalid exits
7165
7166 Valid bits in args[0] are::
7167
7168 #define KVM_X86_DISABLE_EXITS_MWAIT (1 << 0)
7169 #define KVM_X86_DISABLE_EXITS_HLT (1 << 1)
7170 #define KVM_X86_DISABLE_EXITS_PAUSE (1 << 2)
7171 #define KVM_X86_DISABLE_EXITS_CSTATE (1 << 3)
7172
7173 Enabling this capability on a VM provides userspace with a way to no
7174 longer intercept some instructions for improved latency in some
7175 workloads, and is suggested when vCPUs are associated to dedicated
7176 physical CPUs. More bits can be added in the future; userspace can
7177 just pass the KVM_CHECK_EXTENSION result to KVM_ENABLE_CAP to disable
7178 all such vmexits.
7179
7180 Do not enable KVM_FEATURE_PV_UNHALT if you disable HLT exits.
7181
7182 7.14 KVM_CAP_S390_HPAGE_1M
7183 --------------------------
7184
7185 :Architectures: s390
7186 :Parameters: none
7187 :Returns: 0 on success, -EINVAL if hpage module parameter was not set
7188 or cmma is enabled, or the VM has the KVM_VM_S390_UCONTROL
7189 flag set
7190
7191 With this capability the KVM support for memory backing with 1m pages
7192 through hugetlbfs can be enabled for a VM. After the capability is
7193 enabled, cmma can't be enabled anymore and pfmfi and the storage key
7194 interpretation are disabled. If cmma has already been enabled or the
7195 hpage module parameter is not set to 1, -EINVAL is returned.
7196
7197 While it is generally possible to create a huge page backed VM without
7198 this capability, the VM will not be able to run.
7199
7200 7.15 KVM_CAP_MSR_PLATFORM_INFO
7201 ------------------------------
7202
7203 :Architectures: x86
7204 :Parameters: args[0] whether feature should be enabled or not
7205
7206 With this capability, a guest may read the MSR_PLATFORM_INFO MSR. Otherwise,
7207 a #GP would be raised when the guest tries to access. Currently, this
7208 capability does not enable write permissions of this MSR for the guest.
7209
7210 7.16 KVM_CAP_PPC_NESTED_HV
7211 --------------------------
7212
7213 :Architectures: ppc
7214 :Parameters: none
7215 :Returns: 0 on success, -EINVAL when the implementation doesn't support
7216 nested-HV virtualization.
7217
7218 HV-KVM on POWER9 and later systems allows for "nested-HV"
7219 virtualization, which provides a way for a guest VM to run guests that
7220 can run using the CPU's supervisor mode (privileged non-hypervisor
7221 state). Enabling this capability on a VM depends on the CPU having
7222 the necessary functionality and on the facility being enabled with a
7223 kvm-hv module parameter.
7224
7225 7.17 KVM_CAP_EXCEPTION_PAYLOAD
7226 ------------------------------
7227
7228 :Architectures: x86
7229 :Parameters: args[0] whether feature should be enabled or not
7230
7231 With this capability enabled, CR2 will not be modified prior to the
7232 emulated VM-exit when L1 intercepts a #PF exception that occurs in
7233 L2. Similarly, for kvm-intel only, DR6 will not be modified prior to
7234 the emulated VM-exit when L1 intercepts a #DB exception that occurs in
7235 L2. As a result, when KVM_GET_VCPU_EVENTS reports a pending #PF (or
7236 #DB) exception for L2, exception.has_payload will be set and the
7237 faulting address (or the new DR6 bits*) will be reported in the
7238 exception_payload field. Similarly, when userspace injects a #PF (or
7239 #DB) into L2 using KVM_SET_VCPU_EVENTS, it is expected to set
7240 exception.has_payload and to put the faulting address - or the new DR6
7241 bits\ [#]_ - in the exception_payload field.
7242
7243 This capability also enables exception.pending in struct
7244 kvm_vcpu_events, which allows userspace to distinguish between pending
7245 and injected exceptions.
7246
7247
7248 .. [#] For the new DR6 bits, note that bit 16 is set iff the #DB exception
7249 will clear DR6.RTM.
7250
7251 7.18 KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2
7252
7253 :Architectures: x86, arm64, mips
7254 :Parameters: args[0] whether feature should be enabled or not
7255
7256 Valid flags are::
7257
7258 #define KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE (1 << 0)
7259 #define KVM_DIRTY_LOG_INITIALLY_SET (1 << 1)
7260
7261 With KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE is set, KVM_GET_DIRTY_LOG will not
7262 automatically clear and write-protect all pages that are returned as dirty.
7263 Rather, userspace will have to do this operation separately using
7264 KVM_CLEAR_DIRTY_LOG.
7265
7266 At the cost of a slightly more complicated operation, this provides better
7267 scalability and responsiveness for two reasons. First,
7268 KVM_CLEAR_DIRTY_LOG ioctl can operate on a 64-page granularity rather
7269 than requiring to sync a full memslot; this ensures that KVM does not
7270 take spinlocks for an extended period of time. Second, in some cases a
7271 large amount of time can pass between a call to KVM_GET_DIRTY_LOG and
7272 userspace actually using the data in the page. Pages can be modified
7273 during this time, which is inefficient for both the guest and userspace:
7274 the guest will incur a higher penalty due to write protection faults,
7275 while userspace can see false reports of dirty pages. Manual reprotection
7276 helps reducing this time, improving guest performance and reducing the
7277 number of dirty log false positives.
7278
7279 With KVM_DIRTY_LOG_INITIALLY_SET set, all the bits of the dirty bitmap
7280 will be initialized to 1 when created. This also improves performance because
7281 dirty logging can be enabled gradually in small chunks on the first call
7282 to KVM_CLEAR_DIRTY_LOG. KVM_DIRTY_LOG_INITIALLY_SET depends on
7283 KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE (it is also only available on
7284 x86 and arm64 for now).
7285
7286 KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 was previously available under the name
7287 KVM_CAP_MANUAL_DIRTY_LOG_PROTECT, but the implementation had bugs that make
7288 it hard or impossible to use it correctly. The availability of
7289 KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 signals that those bugs are fixed.
7290 Userspace should not try to use KVM_CAP_MANUAL_DIRTY_LOG_PROTECT.
7291
7292 7.19 KVM_CAP_PPC_SECURE_GUEST
7293 ------------------------------
7294
7295 :Architectures: ppc
7296
7297 This capability indicates that KVM is running on a host that has
7298 ultravisor firmware and thus can support a secure guest. On such a
7299 system, a guest can ask the ultravisor to make it a secure guest,
7300 one whose memory is inaccessible to the host except for pages which
7301 are explicitly requested to be shared with the host. The ultravisor
7302 notifies KVM when a guest requests to become a secure guest, and KVM
7303 has the opportunity to veto the transition.
7304
7305 If present, this capability can be enabled for a VM, meaning that KVM
7306 will allow the transition to secure guest mode. Otherwise KVM will
7307 veto the transition.
7308
7309 7.20 KVM_CAP_HALT_POLL
7310 ----------------------
7311
7312 :Architectures: all
7313 :Target: VM
7314 :Parameters: args[0] is the maximum poll time in nanoseconds
7315 :Returns: 0 on success; -1 on error
7316
7317 This capability overrides the kvm module parameter halt_poll_ns for the
7318 target VM.
7319
7320 VCPU polling allows a VCPU to poll for wakeup events instead of immediately
7321 scheduling during guest halts. The maximum time a VCPU can spend polling is
7322 controlled by the kvm module parameter halt_poll_ns. This capability allows
7323 the maximum halt time to specified on a per-VM basis, effectively overriding
7324 the module parameter for the target VM.
7325
7326 7.21 KVM_CAP_X86_USER_SPACE_MSR
7327 -------------------------------
7328
7329 :Architectures: x86
7330 :Target: VM
7331 :Parameters: args[0] contains the mask of KVM_MSR_EXIT_REASON_* events to report
7332 :Returns: 0 on success; -1 on error
7333
7334 This capability enables trapping of #GP invoking RDMSR and WRMSR instructions
7335 into user space.
7336
7337 When a guest requests to read or write an MSR, KVM may not implement all MSRs
7338 that are relevant to a respective system. It also does not differentiate by
7339 CPU type.
7340
7341 To allow more fine grained control over MSR handling, user space may enable
7342 this capability. With it enabled, MSR accesses that match the mask specified in
7343 args[0] and trigger a #GP event inside the guest by KVM will instead trigger
7344 KVM_EXIT_X86_RDMSR and KVM_EXIT_X86_WRMSR exit notifications which user space
7345 can then handle to implement model specific MSR handling and/or user notifications
7346 to inform a user that an MSR was not handled.
7347
7348 7.22 KVM_CAP_X86_BUS_LOCK_EXIT
7349 -------------------------------
7350
7351 :Architectures: x86
7352 :Target: VM
7353 :Parameters: args[0] defines the policy used when bus locks detected in guest
7354 :Returns: 0 on success, -EINVAL when args[0] contains invalid bits
7355
7356 Valid bits in args[0] are::
7357
7358 #define KVM_BUS_LOCK_DETECTION_OFF (1 << 0)
7359 #define KVM_BUS_LOCK_DETECTION_EXIT (1 << 1)
7360
7361 Enabling this capability on a VM provides userspace with a way to select
7362 a policy to handle the bus locks detected in guest. Userspace can obtain
7363 the supported modes from the result of KVM_CHECK_EXTENSION and define it
7364 through the KVM_ENABLE_CAP.
7365
7366 KVM_BUS_LOCK_DETECTION_OFF and KVM_BUS_LOCK_DETECTION_EXIT are supported
7367 currently and mutually exclusive with each other. More bits can be added in
7368 the future.
7369
7370 With KVM_BUS_LOCK_DETECTION_OFF set, bus locks in guest will not cause vm exits
7371 so that no additional actions are needed. This is the default mode.
7372
7373 With KVM_BUS_LOCK_DETECTION_EXIT set, vm exits happen when bus lock detected
7374 in VM. KVM just exits to userspace when handling them. Userspace can enforce
7375 its own throttling or other policy based mitigations.
7376
7377 This capability is aimed to address the thread that VM can exploit bus locks to
7378 degree the performance of the whole system. Once the userspace enable this
7379 capability and select the KVM_BUS_LOCK_DETECTION_EXIT mode, KVM will set the
7380 KVM_RUN_BUS_LOCK flag in vcpu-run->flags field and exit to userspace. Concerning
7381 the bus lock vm exit can be preempted by a higher priority VM exit, the exit
7382 notifications to userspace can be KVM_EXIT_BUS_LOCK or other reasons.
7383 KVM_RUN_BUS_LOCK flag is used to distinguish between them.
7384
7385 7.23 KVM_CAP_PPC_DAWR1
7386 ----------------------
7387
7388 :Architectures: ppc
7389 :Parameters: none
7390 :Returns: 0 on success, -EINVAL when CPU doesn't support 2nd DAWR
7391
7392 This capability can be used to check / enable 2nd DAWR feature provided
7393 by POWER10 processor.
7394
7395
7396 7.24 KVM_CAP_VM_COPY_ENC_CONTEXT_FROM
7397 -------------------------------------
7398
7399 Architectures: x86 SEV enabled
7400 Type: vm
7401 Parameters: args[0] is the fd of the source vm
7402 Returns: 0 on success; ENOTTY on error
7403
7404 This capability enables userspace to copy encryption context from the vm
7405 indicated by the fd to the vm this is called on.
7406
7407 This is intended to support in-guest workloads scheduled by the host. This
7408 allows the in-guest workload to maintain its own NPTs and keeps the two vms
7409 from accidentally clobbering each other with interrupts and the like (separate
7410 APIC/MSRs/etc).
7411
7412 7.25 KVM_CAP_SGX_ATTRIBUTE
7413 --------------------------
7414
7415 :Architectures: x86
7416 :Target: VM
7417 :Parameters: args[0] is a file handle of a SGX attribute file in securityfs
7418 :Returns: 0 on success, -EINVAL if the file handle is invalid or if a requested
7419 attribute is not supported by KVM.
7420
7421 KVM_CAP_SGX_ATTRIBUTE enables a userspace VMM to grant a VM access to one or
7422 more priveleged enclave attributes. args[0] must hold a file handle to a valid
7423 SGX attribute file corresponding to an attribute that is supported/restricted
7424 by KVM (currently only PROVISIONKEY).
7425
7426 The SGX subsystem restricts access to a subset of enclave attributes to provide
7427 additional security for an uncompromised kernel, e.g. use of the PROVISIONKEY
7428 is restricted to deter malware from using the PROVISIONKEY to obtain a stable
7429 system fingerprint. To prevent userspace from circumventing such restrictions
7430 by running an enclave in a VM, KVM prevents access to privileged attributes by
7431 default.
7432
7433 See Documentation/x86/sgx.rst for more details.
7434
7435 7.26 KVM_CAP_PPC_RPT_INVALIDATE
7436 -------------------------------
7437
7438 :Capability: KVM_CAP_PPC_RPT_INVALIDATE
7439 :Architectures: ppc
7440 :Type: vm
7441
7442 This capability indicates that the kernel is capable of handling
7443 H_RPT_INVALIDATE hcall.
7444
7445 In order to enable the use of H_RPT_INVALIDATE in the guest,
7446 user space might have to advertise it for the guest. For example,
7447 IBM pSeries (sPAPR) guest starts using it if "hcall-rpt-invalidate" is
7448 present in the "ibm,hypertas-functions" device-tree property.
7449
7450 This capability is enabled for hypervisors on platforms like POWER9
7451 that support radix MMU.
7452
7453 7.27 KVM_CAP_EXIT_ON_EMULATION_FAILURE
7454 --------------------------------------
7455
7456 :Architectures: x86
7457 :Parameters: args[0] whether the feature should be enabled or not
7458
7459 When this capability is enabled, an emulation failure will result in an exit
7460 to userspace with KVM_INTERNAL_ERROR (except when the emulator was invoked
7461 to handle a VMware backdoor instruction). Furthermore, KVM will now provide up
7462 to 15 instruction bytes for any exit to userspace resulting from an emulation
7463 failure. When these exits to userspace occur use the emulation_failure struct
7464 instead of the internal struct. They both have the same layout, but the
7465 emulation_failure struct matches the content better. It also explicitly
7466 defines the 'flags' field which is used to describe the fields in the struct
7467 that are valid (ie: if KVM_INTERNAL_ERROR_EMULATION_FLAG_INSTRUCTION_BYTES is
7468 set in the 'flags' field then both 'insn_size' and 'insn_bytes' have valid data
7469 in them.)
7470
7471 7.28 KVM_CAP_ARM_MTE
7472 --------------------
7473
7474 :Architectures: arm64
7475 :Parameters: none
7476
7477 This capability indicates that KVM (and the hardware) supports exposing the
7478 Memory Tagging Extensions (MTE) to the guest. It must also be enabled by the
7479 VMM before creating any VCPUs to allow the guest access. Note that MTE is only
7480 available to a guest running in AArch64 mode and enabling this capability will
7481 cause attempts to create AArch32 VCPUs to fail.
7482
7483 When enabled the guest is able to access tags associated with any memory given
7484 to the guest. KVM will ensure that the tags are maintained during swap or
7485 hibernation of the host; however the VMM needs to manually save/restore the
7486 tags as appropriate if the VM is migrated.
7487
7488 When this capability is enabled all memory in memslots must be mapped as
7489 not-shareable (no MAP_SHARED), attempts to create a memslot with a
7490 MAP_SHARED mmap will result in an -EINVAL return.
7491
7492 When enabled the VMM may make use of the ``KVM_ARM_MTE_COPY_TAGS`` ioctl to
7493 perform a bulk copy of tags to/from the guest.
7494
7495 7.29 KVM_CAP_VM_MOVE_ENC_CONTEXT_FROM
7496 -------------------------------------
7497
7498 Architectures: x86 SEV enabled
7499 Type: vm
7500 Parameters: args[0] is the fd of the source vm
7501 Returns: 0 on success
7502
7503 This capability enables userspace to migrate the encryption context from the VM
7504 indicated by the fd to the VM this is called on.
7505
7506 This is intended to support intra-host migration of VMs between userspace VMMs,
7507 upgrading the VMM process without interrupting the guest.
7508
7509 7.30 KVM_CAP_PPC_AIL_MODE_3
7510 -------------------------------
7511
7512 :Capability: KVM_CAP_PPC_AIL_MODE_3
7513 :Architectures: ppc
7514 :Type: vm
7515
7516 This capability indicates that the kernel supports the mode 3 setting for the
7517 "Address Translation Mode on Interrupt" aka "Alternate Interrupt Location"
7518 resource that is controlled with the H_SET_MODE hypercall.
7519
7520 This capability allows a guest kernel to use a better-performance mode for
7521 handling interrupts and system calls.
7522
7523 7.31 KVM_CAP_DISABLE_QUIRKS2
7524 ----------------------------
7525
7526 :Capability: KVM_CAP_DISABLE_QUIRKS2
7527 :Parameters: args[0] - set of KVM quirks to disable
7528 :Architectures: x86
7529 :Type: vm
7530
7531 This capability, if enabled, will cause KVM to disable some behavior
7532 quirks.
7533
7534 Calling KVM_CHECK_EXTENSION for this capability returns a bitmask of
7535 quirks that can be disabled in KVM.
7536
7537 The argument to KVM_ENABLE_CAP for this capability is a bitmask of
7538 quirks to disable, and must be a subset of the bitmask returned by
7539 KVM_CHECK_EXTENSION.
7540
7541 The valid bits in cap.args[0] are:
7542
7543 =================================== ============================================
7544 KVM_X86_QUIRK_LINT0_REENABLED By default, the reset value for the LVT
7545 LINT0 register is 0x700 (APIC_MODE_EXTINT).
7546 When this quirk is disabled, the reset value
7547 is 0x10000 (APIC_LVT_MASKED).
7548
7549 KVM_X86_QUIRK_CD_NW_CLEARED By default, KVM clears CR0.CD and CR0.NW.
7550 When this quirk is disabled, KVM does not
7551 change the value of CR0.CD and CR0.NW.
7552
7553 KVM_X86_QUIRK_LAPIC_MMIO_HOLE By default, the MMIO LAPIC interface is
7554 available even when configured for x2APIC
7555 mode. When this quirk is disabled, KVM
7556 disables the MMIO LAPIC interface if the
7557 LAPIC is in x2APIC mode.
7558
7559 KVM_X86_QUIRK_OUT_7E_INC_RIP By default, KVM pre-increments %rip before
7560 exiting to userspace for an OUT instruction
7561 to port 0x7e. When this quirk is disabled,
7562 KVM does not pre-increment %rip before
7563 exiting to userspace.
7564
7565 KVM_X86_QUIRK_MISC_ENABLE_NO_MWAIT When this quirk is disabled, KVM sets
7566 CPUID.01H:ECX[bit 3] (MONITOR/MWAIT) if
7567 IA32_MISC_ENABLE[bit 18] (MWAIT) is set.
7568 Additionally, when this quirk is disabled,
7569 KVM clears CPUID.01H:ECX[bit 3] if
7570 IA32_MISC_ENABLE[bit 18] is cleared.
7571
7572 KVM_X86_QUIRK_FIX_HYPERCALL_INSN By default, KVM rewrites guest
7573 VMMCALL/VMCALL instructions to match the
7574 vendor's hypercall instruction for the
7575 system. When this quirk is disabled, KVM
7576 will no longer rewrite invalid guest
7577 hypercall instructions. Executing the
7578 incorrect hypercall instruction will
7579 generate a #UD within the guest.
7580
7581 KVM_X86_QUIRK_MWAIT_NEVER_UD_FAULTS By default, KVM emulates MONITOR/MWAIT (if
7582 they are intercepted) as NOPs regardless of
7583 whether or not MONITOR/MWAIT are supported
7584 according to guest CPUID. When this quirk
7585 is disabled and KVM_X86_DISABLE_EXITS_MWAIT
7586 is not set (MONITOR/MWAIT are intercepted),
7587 KVM will inject a #UD on MONITOR/MWAIT if
7588 they're unsupported per guest CPUID. Note,
7589 KVM will modify MONITOR/MWAIT support in
7590 guest CPUID on writes to MISC_ENABLE if
7591 KVM_X86_QUIRK_MISC_ENABLE_NO_MWAIT is
7592 disabled.
7593 =================================== ============================================
7594
7595 7.32 KVM_CAP_MAX_VCPU_ID
7596 ------------------------
7597
7598 :Architectures: x86
7599 :Target: VM
7600 :Parameters: args[0] - maximum APIC ID value set for current VM
7601 :Returns: 0 on success, -EINVAL if args[0] is beyond KVM_MAX_VCPU_IDS
7602 supported in KVM or if it has been set.
7603
7604 This capability allows userspace to specify maximum possible APIC ID
7605 assigned for current VM session prior to the creation of vCPUs, saving
7606 memory for data structures indexed by the APIC ID. Userspace is able
7607 to calculate the limit to APIC ID values from designated
7608 CPU topology.
7609
7610 The value can be changed only until KVM_ENABLE_CAP is set to a nonzero
7611 value or until a vCPU is created. Upon creation of the first vCPU,
7612 if the value was set to zero or KVM_ENABLE_CAP was not invoked, KVM
7613 uses the return value of KVM_CHECK_EXTENSION(KVM_CAP_MAX_VCPU_ID) as
7614 the maximum APIC ID.
7615
7616 7.33 KVM_CAP_X86_NOTIFY_VMEXIT
7617 ------------------------------
7618
7619 :Architectures: x86
7620 :Target: VM
7621 :Parameters: args[0] is the value of notify window as well as some flags
7622 :Returns: 0 on success, -EINVAL if args[0] contains invalid flags or notify
7623 VM exit is unsupported.
7624
7625 Bits 63:32 of args[0] are used for notify window.
7626 Bits 31:0 of args[0] are for some flags. Valid bits are::
7627
7628 #define KVM_X86_NOTIFY_VMEXIT_ENABLED (1 << 0)
7629 #define KVM_X86_NOTIFY_VMEXIT_USER (1 << 1)
7630
7631 This capability allows userspace to configure the notify VM exit on/off
7632 in per-VM scope during VM creation. Notify VM exit is disabled by default.
7633 When userspace sets KVM_X86_NOTIFY_VMEXIT_ENABLED bit in args[0], VMM will
7634 enable this feature with the notify window provided, which will generate
7635 a VM exit if no event window occurs in VM non-root mode for a specified of
7636 time (notify window).
7637
7638 If KVM_X86_NOTIFY_VMEXIT_USER is set in args[0], upon notify VM exits happen,
7639 KVM would exit to userspace for handling.
7640
7641 This capability is aimed to mitigate the threat that malicious VMs can
7642 cause CPU stuck (due to event windows don't open up) and make the CPU
7643 unavailable to host or other VMs.
7644
7645 8. Other capabilities.
7646 ======================
7647
7648 This section lists capabilities that give information about other
7649 features of the KVM implementation.
7650
7651 8.1 KVM_CAP_PPC_HWRNG
7652 ---------------------
7653
7654 :Architectures: ppc
7655
7656 This capability, if KVM_CHECK_EXTENSION indicates that it is
7657 available, means that the kernel has an implementation of the
7658 H_RANDOM hypercall backed by a hardware random-number generator.
7659 If present, the kernel H_RANDOM handler can be enabled for guest use
7660 with the KVM_CAP_PPC_ENABLE_HCALL capability.
7661
7662 8.2 KVM_CAP_HYPERV_SYNIC
7663 ------------------------
7664
7665 :Architectures: x86
7666
7667 This capability, if KVM_CHECK_EXTENSION indicates that it is
7668 available, means that the kernel has an implementation of the
7669 Hyper-V Synthetic interrupt controller(SynIC). Hyper-V SynIC is
7670 used to support Windows Hyper-V based guest paravirt drivers(VMBus).
7671
7672 In order to use SynIC, it has to be activated by setting this
7673 capability via KVM_ENABLE_CAP ioctl on the vcpu fd. Note that this
7674 will disable the use of APIC hardware virtualization even if supported
7675 by the CPU, as it's incompatible with SynIC auto-EOI behavior.
7676
7677 8.3 KVM_CAP_PPC_RADIX_MMU
7678 -------------------------
7679
7680 :Architectures: ppc
7681
7682 This capability, if KVM_CHECK_EXTENSION indicates that it is
7683 available, means that the kernel can support guests using the
7684 radix MMU defined in Power ISA V3.00 (as implemented in the POWER9
7685 processor).
7686
7687 8.4 KVM_CAP_PPC_HASH_MMU_V3
7688 ---------------------------
7689
7690 :Architectures: ppc
7691
7692 This capability, if KVM_CHECK_EXTENSION indicates that it is
7693 available, means that the kernel can support guests using the
7694 hashed page table MMU defined in Power ISA V3.00 (as implemented in
7695 the POWER9 processor), including in-memory segment tables.
7696
7697 8.5 KVM_CAP_MIPS_VZ
7698 -------------------
7699
7700 :Architectures: mips
7701
7702 This capability, if KVM_CHECK_EXTENSION on the main kvm handle indicates that
7703 it is available, means that full hardware assisted virtualization capabilities
7704 of the hardware are available for use through KVM. An appropriate
7705 KVM_VM_MIPS_* type must be passed to KVM_CREATE_VM to create a VM which
7706 utilises it.
7707
7708 If KVM_CHECK_EXTENSION on a kvm VM handle indicates that this capability is
7709 available, it means that the VM is using full hardware assisted virtualization
7710 capabilities of the hardware. This is useful to check after creating a VM with
7711 KVM_VM_MIPS_DEFAULT.
7712
7713 The value returned by KVM_CHECK_EXTENSION should be compared against known
7714 values (see below). All other values are reserved. This is to allow for the
7715 possibility of other hardware assisted virtualization implementations which
7716 may be incompatible with the MIPS VZ ASE.
7717
7718 == ==========================================================================
7719 0 The trap & emulate implementation is in use to run guest code in user
7720 mode. Guest virtual memory segments are rearranged to fit the guest in the
7721 user mode address space.
7722
7723 1 The MIPS VZ ASE is in use, providing full hardware assisted
7724 virtualization, including standard guest virtual memory segments.
7725 == ==========================================================================
7726
7727 8.6 KVM_CAP_MIPS_TE
7728 -------------------
7729
7730 :Architectures: mips
7731
7732 This capability, if KVM_CHECK_EXTENSION on the main kvm handle indicates that
7733 it is available, means that the trap & emulate implementation is available to
7734 run guest code in user mode, even if KVM_CAP_MIPS_VZ indicates that hardware
7735 assisted virtualisation is also available. KVM_VM_MIPS_TE (0) must be passed
7736 to KVM_CREATE_VM to create a VM which utilises it.
7737
7738 If KVM_CHECK_EXTENSION on a kvm VM handle indicates that this capability is
7739 available, it means that the VM is using trap & emulate.
7740
7741 8.7 KVM_CAP_MIPS_64BIT
7742 ----------------------
7743
7744 :Architectures: mips
7745
7746 This capability indicates the supported architecture type of the guest, i.e. the
7747 supported register and address width.
7748
7749 The values returned when this capability is checked by KVM_CHECK_EXTENSION on a
7750 kvm VM handle correspond roughly to the CP0_Config.AT register field, and should
7751 be checked specifically against known values (see below). All other values are
7752 reserved.
7753
7754 == ========================================================================
7755 0 MIPS32 or microMIPS32.
7756 Both registers and addresses are 32-bits wide.
7757 It will only be possible to run 32-bit guest code.
7758
7759 1 MIPS64 or microMIPS64 with access only to 32-bit compatibility segments.
7760 Registers are 64-bits wide, but addresses are 32-bits wide.
7761 64-bit guest code may run but cannot access MIPS64 memory segments.
7762 It will also be possible to run 32-bit guest code.
7763
7764 2 MIPS64 or microMIPS64 with access to all address segments.
7765 Both registers and addresses are 64-bits wide.
7766 It will be possible to run 64-bit or 32-bit guest code.
7767 == ========================================================================
7768
7769 8.9 KVM_CAP_ARM_USER_IRQ
7770 ------------------------
7771
7772 :Architectures: arm64
7773
7774 This capability, if KVM_CHECK_EXTENSION indicates that it is available, means
7775 that if userspace creates a VM without an in-kernel interrupt controller, it
7776 will be notified of changes to the output level of in-kernel emulated devices,
7777 which can generate virtual interrupts, presented to the VM.
7778 For such VMs, on every return to userspace, the kernel
7779 updates the vcpu's run->s.regs.device_irq_level field to represent the actual
7780 output level of the device.
7781
7782 Whenever kvm detects a change in the device output level, kvm guarantees at
7783 least one return to userspace before running the VM. This exit could either
7784 be a KVM_EXIT_INTR or any other exit event, like KVM_EXIT_MMIO. This way,
7785 userspace can always sample the device output level and re-compute the state of
7786 the userspace interrupt controller. Userspace should always check the state
7787 of run->s.regs.device_irq_level on every kvm exit.
7788 The value in run->s.regs.device_irq_level can represent both level and edge
7789 triggered interrupt signals, depending on the device. Edge triggered interrupt
7790 signals will exit to userspace with the bit in run->s.regs.device_irq_level
7791 set exactly once per edge signal.
7792
7793 The field run->s.regs.device_irq_level is available independent of
7794 run->kvm_valid_regs or run->kvm_dirty_regs bits.
7795
7796 If KVM_CAP_ARM_USER_IRQ is supported, the KVM_CHECK_EXTENSION ioctl returns a
7797 number larger than 0 indicating the version of this capability is implemented
7798 and thereby which bits in run->s.regs.device_irq_level can signal values.
7799
7800 Currently the following bits are defined for the device_irq_level bitmap::
7801
7802 KVM_CAP_ARM_USER_IRQ >= 1:
7803
7804 KVM_ARM_DEV_EL1_VTIMER - EL1 virtual timer
7805 KVM_ARM_DEV_EL1_PTIMER - EL1 physical timer
7806 KVM_ARM_DEV_PMU - ARM PMU overflow interrupt signal
7807
7808 Future versions of kvm may implement additional events. These will get
7809 indicated by returning a higher number from KVM_CHECK_EXTENSION and will be
7810 listed above.
7811
7812 8.10 KVM_CAP_PPC_SMT_POSSIBLE
7813 -----------------------------
7814
7815 :Architectures: ppc
7816
7817 Querying this capability returns a bitmap indicating the possible
7818 virtual SMT modes that can be set using KVM_CAP_PPC_SMT. If bit N
7819 (counting from the right) is set, then a virtual SMT mode of 2^N is
7820 available.
7821
7822 8.11 KVM_CAP_HYPERV_SYNIC2
7823 --------------------------
7824
7825 :Architectures: x86
7826
7827 This capability enables a newer version of Hyper-V Synthetic interrupt
7828 controller (SynIC). The only difference with KVM_CAP_HYPERV_SYNIC is that KVM
7829 doesn't clear SynIC message and event flags pages when they are enabled by
7830 writing to the respective MSRs.
7831
7832 8.12 KVM_CAP_HYPERV_VP_INDEX
7833 ----------------------------
7834
7835 :Architectures: x86
7836
7837 This capability indicates that userspace can load HV_X64_MSR_VP_INDEX msr. Its
7838 value is used to denote the target vcpu for a SynIC interrupt. For
7839 compatibilty, KVM initializes this msr to KVM's internal vcpu index. When this
7840 capability is absent, userspace can still query this msr's value.
7841
7842 8.13 KVM_CAP_S390_AIS_MIGRATION
7843 -------------------------------
7844
7845 :Architectures: s390
7846 :Parameters: none
7847
7848 This capability indicates if the flic device will be able to get/set the
7849 AIS states for migration via the KVM_DEV_FLIC_AISM_ALL attribute and allows
7850 to discover this without having to create a flic device.
7851
7852 8.14 KVM_CAP_S390_PSW
7853 ---------------------
7854
7855 :Architectures: s390
7856
7857 This capability indicates that the PSW is exposed via the kvm_run structure.
7858
7859 8.15 KVM_CAP_S390_GMAP
7860 ----------------------
7861
7862 :Architectures: s390
7863
7864 This capability indicates that the user space memory used as guest mapping can
7865 be anywhere in the user memory address space, as long as the memory slots are
7866 aligned and sized to a segment (1MB) boundary.
7867
7868 8.16 KVM_CAP_S390_COW
7869 ---------------------
7870
7871 :Architectures: s390
7872
7873 This capability indicates that the user space memory used as guest mapping can
7874 use copy-on-write semantics as well as dirty pages tracking via read-only page
7875 tables.
7876
7877 8.17 KVM_CAP_S390_BPB
7878 ---------------------
7879
7880 :Architectures: s390
7881
7882 This capability indicates that kvm will implement the interfaces to handle
7883 reset, migration and nested KVM for branch prediction blocking. The stfle
7884 facility 82 should not be provided to the guest without this capability.
7885
7886 8.18 KVM_CAP_HYPERV_TLBFLUSH
7887 ----------------------------
7888
7889 :Architectures: x86
7890
7891 This capability indicates that KVM supports paravirtualized Hyper-V TLB Flush
7892 hypercalls:
7893 HvFlushVirtualAddressSpace, HvFlushVirtualAddressSpaceEx,
7894 HvFlushVirtualAddressList, HvFlushVirtualAddressListEx.
7895
7896 8.19 KVM_CAP_ARM_INJECT_SERROR_ESR
7897 ----------------------------------
7898
7899 :Architectures: arm64
7900
7901 This capability indicates that userspace can specify (via the
7902 KVM_SET_VCPU_EVENTS ioctl) the syndrome value reported to the guest when it
7903 takes a virtual SError interrupt exception.
7904 If KVM advertises this capability, userspace can only specify the ISS field for
7905 the ESR syndrome. Other parts of the ESR, such as the EC are generated by the
7906 CPU when the exception is taken. If this virtual SError is taken to EL1 using
7907 AArch64, this value will be reported in the ISS field of ESR_ELx.
7908
7909 See KVM_CAP_VCPU_EVENTS for more details.
7910
7911 8.20 KVM_CAP_HYPERV_SEND_IPI
7912 ----------------------------
7913
7914 :Architectures: x86
7915
7916 This capability indicates that KVM supports paravirtualized Hyper-V IPI send
7917 hypercalls:
7918 HvCallSendSyntheticClusterIpi, HvCallSendSyntheticClusterIpiEx.
7919
7920 8.21 KVM_CAP_HYPERV_DIRECT_TLBFLUSH
7921 -----------------------------------
7922
7923 :Architectures: x86
7924
7925 This capability indicates that KVM running on top of Hyper-V hypervisor
7926 enables Direct TLB flush for its guests meaning that TLB flush
7927 hypercalls are handled by Level 0 hypervisor (Hyper-V) bypassing KVM.
7928 Due to the different ABI for hypercall parameters between Hyper-V and
7929 KVM, enabling this capability effectively disables all hypercall
7930 handling by KVM (as some KVM hypercall may be mistakenly treated as TLB
7931 flush hypercalls by Hyper-V) so userspace should disable KVM identification
7932 in CPUID and only exposes Hyper-V identification. In this case, guest
7933 thinks it's running on Hyper-V and only use Hyper-V hypercalls.
7934
7935 8.22 KVM_CAP_S390_VCPU_RESETS
7936 -----------------------------
7937
7938 :Architectures: s390
7939
7940 This capability indicates that the KVM_S390_NORMAL_RESET and
7941 KVM_S390_CLEAR_RESET ioctls are available.
7942
7943 8.23 KVM_CAP_S390_PROTECTED
7944 ---------------------------
7945
7946 :Architectures: s390
7947
7948 This capability indicates that the Ultravisor has been initialized and
7949 KVM can therefore start protected VMs.
7950 This capability governs the KVM_S390_PV_COMMAND ioctl and the
7951 KVM_MP_STATE_LOAD MP_STATE. KVM_SET_MP_STATE can fail for protected
7952 guests when the state change is invalid.
7953
7954 8.24 KVM_CAP_STEAL_TIME
7955 -----------------------
7956
7957 :Architectures: arm64, x86
7958
7959 This capability indicates that KVM supports steal time accounting.
7960 When steal time accounting is supported it may be enabled with
7961 architecture-specific interfaces. This capability and the architecture-
7962 specific interfaces must be consistent, i.e. if one says the feature
7963 is supported, than the other should as well and vice versa. For arm64
7964 see Documentation/virt/kvm/devices/vcpu.rst "KVM_ARM_VCPU_PVTIME_CTRL".
7965 For x86 see Documentation/virt/kvm/x86/msr.rst "MSR_KVM_STEAL_TIME".
7966
7967 8.25 KVM_CAP_S390_DIAG318
7968 -------------------------
7969
7970 :Architectures: s390
7971
7972 This capability enables a guest to set information about its control program
7973 (i.e. guest kernel type and version). The information is helpful during
7974 system/firmware service events, providing additional data about the guest
7975 environments running on the machine.
7976
7977 The information is associated with the DIAGNOSE 0x318 instruction, which sets
7978 an 8-byte value consisting of a one-byte Control Program Name Code (CPNC) and
7979 a 7-byte Control Program Version Code (CPVC). The CPNC determines what
7980 environment the control program is running in (e.g. Linux, z/VM...), and the
7981 CPVC is used for information specific to OS (e.g. Linux version, Linux
7982 distribution...)
7983
7984 If this capability is available, then the CPNC and CPVC can be synchronized
7985 between KVM and userspace via the sync regs mechanism (KVM_SYNC_DIAG318).
7986
7987 8.26 KVM_CAP_X86_USER_SPACE_MSR
7988 -------------------------------
7989
7990 :Architectures: x86
7991
7992 This capability indicates that KVM supports deflection of MSR reads and
7993 writes to user space. It can be enabled on a VM level. If enabled, MSR
7994 accesses that would usually trigger a #GP by KVM into the guest will
7995 instead get bounced to user space through the KVM_EXIT_X86_RDMSR and
7996 KVM_EXIT_X86_WRMSR exit notifications.
7997
7998 8.27 KVM_CAP_X86_MSR_FILTER
7999 ---------------------------
8000
8001 :Architectures: x86
8002
8003 This capability indicates that KVM supports that accesses to user defined MSRs
8004 may be rejected. With this capability exposed, KVM exports new VM ioctl
8005 KVM_X86_SET_MSR_FILTER which user space can call to specify bitmaps of MSR
8006 ranges that KVM should reject access to.
8007
8008 In combination with KVM_CAP_X86_USER_SPACE_MSR, this allows user space to
8009 trap and emulate MSRs that are outside of the scope of KVM as well as
8010 limit the attack surface on KVM's MSR emulation code.
8011
8012 8.28 KVM_CAP_ENFORCE_PV_FEATURE_CPUID
8013 -------------------------------------
8014
8015 Architectures: x86
8016
8017 When enabled, KVM will disable paravirtual features provided to the
8018 guest according to the bits in the KVM_CPUID_FEATURES CPUID leaf
8019 (0x40000001). Otherwise, a guest may use the paravirtual features
8020 regardless of what has actually been exposed through the CPUID leaf.
8021
8022 8.29 KVM_CAP_DIRTY_LOG_RING
8023 ---------------------------
8024
8025 :Architectures: x86
8026 :Parameters: args[0] - size of the dirty log ring
8027
8028 KVM is capable of tracking dirty memory using ring buffers that are
8029 mmaped into userspace; there is one dirty ring per vcpu.
8030
8031 The dirty ring is available to userspace as an array of
8032 ``struct kvm_dirty_gfn``. Each dirty entry it's defined as::
8033
8034 struct kvm_dirty_gfn {
8035 __u32 flags;
8036 __u32 slot; /* as_id | slot_id */
8037 __u64 offset;
8038 };
8039
8040 The following values are defined for the flags field to define the
8041 current state of the entry::
8042
8043 #define KVM_DIRTY_GFN_F_DIRTY BIT(0)
8044 #define KVM_DIRTY_GFN_F_RESET BIT(1)
8045 #define KVM_DIRTY_GFN_F_MASK 0x3
8046
8047 Userspace should call KVM_ENABLE_CAP ioctl right after KVM_CREATE_VM
8048 ioctl to enable this capability for the new guest and set the size of
8049 the rings. Enabling the capability is only allowed before creating any
8050 vCPU, and the size of the ring must be a power of two. The larger the
8051 ring buffer, the less likely the ring is full and the VM is forced to
8052 exit to userspace. The optimal size depends on the workload, but it is
8053 recommended that it be at least 64 KiB (4096 entries).
8054
8055 Just like for dirty page bitmaps, the buffer tracks writes to
8056 all user memory regions for which the KVM_MEM_LOG_DIRTY_PAGES flag was
8057 set in KVM_SET_USER_MEMORY_REGION. Once a memory region is registered
8058 with the flag set, userspace can start harvesting dirty pages from the
8059 ring buffer.
8060
8061 An entry in the ring buffer can be unused (flag bits ``00``),
8062 dirty (flag bits ``01``) or harvested (flag bits ``1X``). The
8063 state machine for the entry is as follows::
8064
8065 dirtied harvested reset
8066 00 -----------> 01 -------------> 1X -------+
8067 ^ |
8068 | |
8069 +------------------------------------------+
8070
8071 To harvest the dirty pages, userspace accesses the mmaped ring buffer
8072 to read the dirty GFNs. If the flags has the DIRTY bit set (at this stage
8073 the RESET bit must be cleared), then it means this GFN is a dirty GFN.
8074 The userspace should harvest this GFN and mark the flags from state
8075 ``01b`` to ``1Xb`` (bit 0 will be ignored by KVM, but bit 1 must be set
8076 to show that this GFN is harvested and waiting for a reset), and move
8077 on to the next GFN. The userspace should continue to do this until the
8078 flags of a GFN have the DIRTY bit cleared, meaning that it has harvested
8079 all the dirty GFNs that were available.
8080
8081 It's not necessary for userspace to harvest the all dirty GFNs at once.
8082 However it must collect the dirty GFNs in sequence, i.e., the userspace
8083 program cannot skip one dirty GFN to collect the one next to it.
8084
8085 After processing one or more entries in the ring buffer, userspace
8086 calls the VM ioctl KVM_RESET_DIRTY_RINGS to notify the kernel about
8087 it, so that the kernel will reprotect those collected GFNs.
8088 Therefore, the ioctl must be called *before* reading the content of
8089 the dirty pages.
8090
8091 The dirty ring can get full. When it happens, the KVM_RUN of the
8092 vcpu will return with exit reason KVM_EXIT_DIRTY_LOG_FULL.
8093
8094 The dirty ring interface has a major difference comparing to the
8095 KVM_GET_DIRTY_LOG interface in that, when reading the dirty ring from
8096 userspace, it's still possible that the kernel has not yet flushed the
8097 processor's dirty page buffers into the kernel buffer (with dirty bitmaps, the
8098 flushing is done by the KVM_GET_DIRTY_LOG ioctl). To achieve that, one
8099 needs to kick the vcpu out of KVM_RUN using a signal. The resulting
8100 vmexit ensures that all dirty GFNs are flushed to the dirty rings.
8101
8102 NOTE: the capability KVM_CAP_DIRTY_LOG_RING and the corresponding
8103 ioctl KVM_RESET_DIRTY_RINGS are mutual exclusive to the existing ioctls
8104 KVM_GET_DIRTY_LOG and KVM_CLEAR_DIRTY_LOG. After enabling
8105 KVM_CAP_DIRTY_LOG_RING with an acceptable dirty ring size, the virtual
8106 machine will switch to ring-buffer dirty page tracking and further
8107 KVM_GET_DIRTY_LOG or KVM_CLEAR_DIRTY_LOG ioctls will fail.
8108
8109 8.30 KVM_CAP_XEN_HVM
8110 --------------------
8111
8112 :Architectures: x86
8113
8114 This capability indicates the features that Xen supports for hosting Xen
8115 PVHVM guests. Valid flags are::
8116
8117 #define KVM_XEN_HVM_CONFIG_HYPERCALL_MSR (1 << 0)
8118 #define KVM_XEN_HVM_CONFIG_INTERCEPT_HCALL (1 << 1)
8119 #define KVM_XEN_HVM_CONFIG_SHARED_INFO (1 << 2)
8120 #define KVM_XEN_HVM_CONFIG_RUNSTATE (1 << 3)
8121 #define KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL (1 << 4)
8122 #define KVM_XEN_HVM_CONFIG_EVTCHN_SEND (1 << 5)
8123
8124 The KVM_XEN_HVM_CONFIG_HYPERCALL_MSR flag indicates that the KVM_XEN_HVM_CONFIG
8125 ioctl is available, for the guest to set its hypercall page.
8126
8127 If KVM_XEN_HVM_CONFIG_INTERCEPT_HCALL is also set, the same flag may also be
8128 provided in the flags to KVM_XEN_HVM_CONFIG, without providing hypercall page
8129 contents, to request that KVM generate hypercall page content automatically
8130 and also enable interception of guest hypercalls with KVM_EXIT_XEN.
8131
8132 The KVM_XEN_HVM_CONFIG_SHARED_INFO flag indicates the availability of the
8133 KVM_XEN_HVM_SET_ATTR, KVM_XEN_HVM_GET_ATTR, KVM_XEN_VCPU_SET_ATTR and
8134 KVM_XEN_VCPU_GET_ATTR ioctls, as well as the delivery of exception vectors
8135 for event channel upcalls when the evtchn_upcall_pending field of a vcpu's
8136 vcpu_info is set.
8137
8138 The KVM_XEN_HVM_CONFIG_RUNSTATE flag indicates that the runstate-related
8139 features KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADDR/_CURRENT/_DATA/_ADJUST are
8140 supported by the KVM_XEN_VCPU_SET_ATTR/KVM_XEN_VCPU_GET_ATTR ioctls.
8141
8142 The KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL flag indicates that IRQ routing entries
8143 of the type KVM_IRQ_ROUTING_XEN_EVTCHN are supported, with the priority
8144 field set to indicate 2 level event channel delivery.
8145
8146 The KVM_XEN_HVM_CONFIG_EVTCHN_SEND flag indicates that KVM supports
8147 injecting event channel events directly into the guest with the
8148 KVM_XEN_HVM_EVTCHN_SEND ioctl. It also indicates support for the
8149 KVM_XEN_ATTR_TYPE_EVTCHN/XEN_VERSION HVM attributes and the
8150 KVM_XEN_VCPU_ATTR_TYPE_VCPU_ID/TIMER/UPCALL_VECTOR vCPU attributes.
8151 related to event channel delivery, timers, and the XENVER_version
8152 interception.
8153
8154 8.31 KVM_CAP_PPC_MULTITCE
8155 -------------------------
8156
8157 :Capability: KVM_CAP_PPC_MULTITCE
8158 :Architectures: ppc
8159 :Type: vm
8160
8161 This capability means the kernel is capable of handling hypercalls
8162 H_PUT_TCE_INDIRECT and H_STUFF_TCE without passing those into the user
8163 space. This significantly accelerates DMA operations for PPC KVM guests.
8164 User space should expect that its handlers for these hypercalls
8165 are not going to be called if user space previously registered LIOBN
8166 in KVM (via KVM_CREATE_SPAPR_TCE or similar calls).
8167
8168 In order to enable H_PUT_TCE_INDIRECT and H_STUFF_TCE use in the guest,
8169 user space might have to advertise it for the guest. For example,
8170 IBM pSeries (sPAPR) guest starts using them if "hcall-multi-tce" is
8171 present in the "ibm,hypertas-functions" device-tree property.
8172
8173 The hypercalls mentioned above may or may not be processed successfully
8174 in the kernel based fast path. If they can not be handled by the kernel,
8175 they will get passed on to user space. So user space still has to have
8176 an implementation for these despite the in kernel acceleration.
8177
8178 This capability is always enabled.
8179
8180 8.32 KVM_CAP_PTP_KVM
8181 --------------------
8182
8183 :Architectures: arm64
8184
8185 This capability indicates that the KVM virtual PTP service is
8186 supported in the host. A VMM can check whether the service is
8187 available to the guest on migration.
8188
8189 8.33 KVM_CAP_HYPERV_ENFORCE_CPUID
8190 ---------------------------------
8191
8192 Architectures: x86
8193
8194 When enabled, KVM will disable emulated Hyper-V features provided to the
8195 guest according to the bits Hyper-V CPUID feature leaves. Otherwise, all
8196 currently implmented Hyper-V features are provided unconditionally when
8197 Hyper-V identification is set in the HYPERV_CPUID_INTERFACE (0x40000001)
8198 leaf.
8199
8200 8.34 KVM_CAP_EXIT_HYPERCALL
8201 ---------------------------
8202
8203 :Capability: KVM_CAP_EXIT_HYPERCALL
8204 :Architectures: x86
8205 :Type: vm
8206
8207 This capability, if enabled, will cause KVM to exit to userspace
8208 with KVM_EXIT_HYPERCALL exit reason to process some hypercalls.
8209
8210 Calling KVM_CHECK_EXTENSION for this capability will return a bitmask
8211 of hypercalls that can be configured to exit to userspace.
8212 Right now, the only such hypercall is KVM_HC_MAP_GPA_RANGE.
8213
8214 The argument to KVM_ENABLE_CAP is also a bitmask, and must be a subset
8215 of the result of KVM_CHECK_EXTENSION. KVM will forward to userspace
8216 the hypercalls whose corresponding bit is in the argument, and return
8217 ENOSYS for the others.
8218
8219 8.35 KVM_CAP_PMU_CAPABILITY
8220 ---------------------------
8221
8222 :Capability KVM_CAP_PMU_CAPABILITY
8223 :Architectures: x86
8224 :Type: vm
8225 :Parameters: arg[0] is bitmask of PMU virtualization capabilities.
8226 :Returns 0 on success, -EINVAL when arg[0] contains invalid bits
8227
8228 This capability alters PMU virtualization in KVM.
8229
8230 Calling KVM_CHECK_EXTENSION for this capability returns a bitmask of
8231 PMU virtualization capabilities that can be adjusted on a VM.
8232
8233 The argument to KVM_ENABLE_CAP is also a bitmask and selects specific
8234 PMU virtualization capabilities to be applied to the VM. This can
8235 only be invoked on a VM prior to the creation of VCPUs.
8236
8237 At this time, KVM_PMU_CAP_DISABLE is the only capability. Setting
8238 this capability will disable PMU virtualization for that VM. Usermode
8239 should adjust CPUID leaf 0xA to reflect that the PMU is disabled.
8240
8241 8.36 KVM_CAP_ARM_SYSTEM_SUSPEND
8242 -------------------------------
8243
8244 :Capability: KVM_CAP_ARM_SYSTEM_SUSPEND
8245 :Architectures: arm64
8246 :Type: vm
8247
8248 When enabled, KVM will exit to userspace with KVM_EXIT_SYSTEM_EVENT of
8249 type KVM_SYSTEM_EVENT_SUSPEND to process the guest suspend request.
8250
8251 8.37 KVM_CAP_S390_PROTECTED_DUMP
8252 --------------------------------
8253
8254 :Capability: KVM_CAP_S390_PROTECTED_DUMP
8255 :Architectures: s390
8256 :Type: vm
8257
8258 This capability indicates that KVM and the Ultravisor support dumping
8259 PV guests. The `KVM_PV_DUMP` command is available for the
8260 `KVM_S390_PV_COMMAND` ioctl and the `KVM_PV_INFO` command provides
8261 dump related UV data. Also the vcpu ioctl `KVM_S390_PV_CPU_COMMAND` is
8262 available and supports the `KVM_PV_DUMP_CPU` subcommand.
8263
8264 8.38 KVM_CAP_VM_DISABLE_NX_HUGE_PAGES
8265 -------------------------------------
8266
8267 :Capability: KVM_CAP_VM_DISABLE_NX_HUGE_PAGES
8268 :Architectures: x86
8269 :Type: vm
8270 :Parameters: arg[0] must be 0.
8271 :Returns: 0 on success, -EPERM if the userspace process does not
8272 have CAP_SYS_BOOT, -EINVAL if args[0] is not 0 or any vCPUs have been
8273 created.
8274
8275 This capability disables the NX huge pages mitigation for iTLB MULTIHIT.
8276
8277 The capability has no effect if the nx_huge_pages module parameter is not set.
8278
8279 This capability may only be set before any vCPUs are created.
8280
8281 8.39 KVM_CAP_S390_CPU_TOPOLOGY
8282 ------------------------------
8283
8284 :Capability: KVM_CAP_S390_CPU_TOPOLOGY
8285 :Architectures: s390
8286 :Type: vm
8287
8288 This capability indicates that KVM will provide the S390 CPU Topology
8289 facility which consist of the interpretation of the PTF instruction for
8290 the function code 2 along with interception and forwarding of both the
8291 PTF instruction with function codes 0 or 1 and the STSI(15,1,x)
8292 instruction to the userland hypervisor.
8293
8294 The stfle facility 11, CPU Topology facility, should not be indicated
8295 to the guest without this capability.
8296
8297 When this capability is present, KVM provides a new attribute group
8298 on vm fd, KVM_S390_VM_CPU_TOPOLOGY.
8299 This new attribute allows to get, set or clear the Modified Change
8300 Topology Report (MTCR) bit of the SCA through the kvm_device_attr
8301 structure.
8302
8303 When getting the Modified Change Topology Report value, the attr->addr
8304 must point to a byte where the value will be stored or retrieved from.
8305
8306 9. Known KVM API problems
8307 =========================
8308
8309 In some cases, KVM's API has some inconsistencies or common pitfalls
8310 that userspace need to be aware of. This section details some of
8311 these issues.
8312
8313 Most of them are architecture specific, so the section is split by
8314 architecture.
8315
8316 9.1. x86
8317 --------
8318
8319 ``KVM_GET_SUPPORTED_CPUID`` issues
8320 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8321
8322 In general, ``KVM_GET_SUPPORTED_CPUID`` is designed so that it is possible
8323 to take its result and pass it directly to ``KVM_SET_CPUID2``. This section
8324 documents some cases in which that requires some care.
8325
8326 Local APIC features
8327 ~~~~~~~~~~~~~~~~~~~
8328
8329 CPU[EAX=1]:ECX[21] (X2APIC) is reported by ``KVM_GET_SUPPORTED_CPUID``,
8330 but it can only be enabled if ``KVM_CREATE_IRQCHIP`` or
8331 ``KVM_ENABLE_CAP(KVM_CAP_IRQCHIP_SPLIT)`` are used to enable in-kernel emulation of
8332 the local APIC.
8333
8334 The same is true for the ``KVM_FEATURE_PV_UNHALT`` paravirtualized feature.
8335
8336 CPU[EAX=1]:ECX[24] (TSC_DEADLINE) is not reported by ``KVM_GET_SUPPORTED_CPUID``.
8337 It can be enabled if ``KVM_CAP_TSC_DEADLINE_TIMER`` is present and the kernel
8338 has enabled in-kernel emulation of the local APIC.
8339
8340 Obsolete ioctls and capabilities
8341 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8342
8343 KVM_CAP_DISABLE_QUIRKS does not let userspace know which quirks are actually
8344 available. Use ``KVM_CHECK_EXTENSION(KVM_CAP_DISABLE_QUIRKS2)`` instead if
8345 available.
8346
8347 Ordering of KVM_GET_*/KVM_SET_* ioctls
8348 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
8349
8350 TBD
