1.. SPDX-License-Identifier: GPL-2.0 2 3=================================================================== 4The Definitive KVM (Kernel-based Virtual Machine) API Documentation 5=================================================================== 6 71. General description 8====================== 9 10The kvm API is a set of ioctls that are issued to control various aspects 11of a virtual machine. The ioctls belong to the following classes: 12 13 - System ioctls: These query and set global attributes which affect the 14 whole kvm subsystem. In addition a system ioctl is used to create 15 virtual machines. 16 17 - VM ioctls: These query and set attributes that affect an entire virtual 18 machine, for example memory layout. In addition a VM ioctl is used to 19 create virtual cpus (vcpus) and devices. 20 21 VM ioctls must be issued from the same process (address space) that was 22 used to create the VM. 23 24 - vcpu ioctls: These query and set attributes that control the operation 25 of a single virtual cpu. 26 27 vcpu ioctls should be issued from the same thread that was used to create 28 the vcpu, except for asynchronous vcpu ioctl that are marked as such in 29 the documentation. Otherwise, the first ioctl after switching threads 30 could see a performance impact. 31 32 - device ioctls: These query and set attributes that control the operation 33 of a single device. 34 35 device ioctls must be issued from the same process (address space) that 36 was used to create the VM. 37 382. File descriptors 39=================== 40 41The kvm API is centered around file descriptors. An initial 42open("/dev/kvm") obtains a handle to the kvm subsystem; this handle 43can be used to issue system ioctls. A KVM_CREATE_VM ioctl on this 44handle will create a VM file descriptor which can be used to issue VM 45ioctls. A KVM_CREATE_VCPU or KVM_CREATE_DEVICE ioctl on a VM fd will 46create a virtual cpu or device and return a file descriptor pointing to 47the new resource. Finally, ioctls on a vcpu or device fd can be used 48to control the vcpu or device. For vcpus, this includes the important 49task of actually running guest code. 50 51In general file descriptors can be migrated among processes by means 52of fork() and the SCM_RIGHTS facility of unix domain socket. These 53kinds of tricks are explicitly not supported by kvm. While they will 54not cause harm to the host, their actual behavior is not guaranteed by 55the API. See "General description" for details on the ioctl usage 56model that is supported by KVM. 57 58It is important to note that althought VM ioctls may only be issued from 59the process that created the VM, a VM's lifecycle is associated with its 60file descriptor, not its creator (process). In other words, the VM and 61its resources, *including the associated address space*, are not freed 62until the last reference to the VM's file descriptor has been released. 63For example, if fork() is issued after ioctl(KVM_CREATE_VM), the VM will 64not be freed until both the parent (original) process and its child have 65put their references to the VM's file descriptor. 66 67Because a VM's resources are not freed until the last reference to its 68file descriptor is released, creating additional references to a VM 69via fork(), dup(), etc... without careful consideration is strongly 70discouraged and may have unwanted side effects, e.g. memory allocated 71by and on behalf of the VM's process may not be freed/unaccounted when 72the VM is shut down. 73 74 753. Extensions 76============= 77 78As of Linux 2.6.22, the KVM ABI has been stabilized: no backward 79incompatible change are allowed. However, there is an extension 80facility that allows backward-compatible extensions to the API to be 81queried and used. 82 83The extension mechanism is not based on the Linux version number. 84Instead, kvm defines extension identifiers and a facility to query 85whether a particular extension identifier is available. If it is, a 86set of ioctls is available for application use. 87 88 894. API description 90================== 91 92This section describes ioctls that can be used to control kvm guests. 93For each ioctl, the following information is provided along with a 94description: 95 96 Capability: 97 which KVM extension provides this ioctl. Can be 'basic', 98 which means that is will be provided by any kernel that supports 99 API version 12 (see section 4.1), a KVM_CAP_xyz constant, which 100 means availability needs to be checked with KVM_CHECK_EXTENSION 101 (see section 4.4), or 'none' which means that while not all kernels 102 support this ioctl, there's no capability bit to check its 103 availability: for kernels that don't support the ioctl, 104 the ioctl returns -ENOTTY. 105 106 Architectures: 107 which instruction set architectures provide this ioctl. 108 x86 includes both i386 and x86_64. 109 110 Type: 111 system, vm, or vcpu. 112 113 Parameters: 114 what parameters are accepted by the ioctl. 115 116 Returns: 117 the return value. General error numbers (EBADF, ENOMEM, EINVAL) 118 are not detailed, but errors with specific meanings are. 119 120 1214.1 KVM_GET_API_VERSION 122----------------------- 123 124:Capability: basic 125:Architectures: all 126:Type: system ioctl 127:Parameters: none 128:Returns: the constant KVM_API_VERSION (=12) 129 130This identifies the API version as the stable kvm API. It is not 131expected that this number will change. However, Linux 2.6.20 and 1322.6.21 report earlier versions; these are not documented and not 133supported. Applications should refuse to run if KVM_GET_API_VERSION 134returns a value other than 12. If this check passes, all ioctls 135described as 'basic' will be available. 136 137 1384.2 KVM_CREATE_VM 139----------------- 140 141:Capability: basic 142:Architectures: all 143:Type: system ioctl 144:Parameters: machine type identifier (KVM_VM_*) 145:Returns: a VM fd that can be used to control the new virtual machine. 146 147The new VM has no virtual cpus and no memory. 148You probably want to use 0 as machine type. 149 150In order to create user controlled virtual machines on S390, check 151KVM_CAP_S390_UCONTROL and use the flag KVM_VM_S390_UCONTROL as 152privileged user (CAP_SYS_ADMIN). 153 154To use hardware assisted virtualization on MIPS (VZ ASE) rather than 155the default trap & emulate implementation (which changes the virtual 156memory layout to fit in user mode), check KVM_CAP_MIPS_VZ and use the 157flag KVM_VM_MIPS_VZ. 158 159 160On arm64, the physical address size for a VM (IPA Size limit) is limited 161to 40bits by default. The limit can be configured if the host supports the 162extension KVM_CAP_ARM_VM_IPA_SIZE. When supported, use 163KVM_VM_TYPE_ARM_IPA_SIZE(IPA_Bits) to set the size in the machine type 164identifier, where IPA_Bits is the maximum width of any physical 165address used by the VM. The IPA_Bits is encoded in bits[7-0] of the 166machine type identifier. 167 168e.g, to configure a guest to use 48bit physical address size:: 169 170 vm_fd = ioctl(dev_fd, KVM_CREATE_VM, KVM_VM_TYPE_ARM_IPA_SIZE(48)); 171 172The requested size (IPA_Bits) must be: 173 174 == ========================================================= 175 0 Implies default size, 40bits (for backward compatibility) 176 N Implies N bits, where N is a positive integer such that, 177 32 <= N <= Host_IPA_Limit 178 == ========================================================= 179 180Host_IPA_Limit is the maximum possible value for IPA_Bits on the host and 181is dependent on the CPU capability and the kernel configuration. The limit can 182be retrieved using KVM_CAP_ARM_VM_IPA_SIZE of the KVM_CHECK_EXTENSION 183ioctl() at run-time. 184 185Creation of the VM will fail if the requested IPA size (whether it is 186implicit or explicit) is unsupported on the host. 187 188Please note that configuring the IPA size does not affect the capability 189exposed by the guest CPUs in ID_AA64MMFR0_EL1[PARange]. It only affects 190size of the address translated by the stage2 level (guest physical to 191host physical address translations). 192 193 1944.3 KVM_GET_MSR_INDEX_LIST, KVM_GET_MSR_FEATURE_INDEX_LIST 195---------------------------------------------------------- 196 197:Capability: basic, KVM_CAP_GET_MSR_FEATURES for KVM_GET_MSR_FEATURE_INDEX_LIST 198:Architectures: x86 199:Type: system ioctl 200:Parameters: struct kvm_msr_list (in/out) 201:Returns: 0 on success; -1 on error 202 203Errors: 204 205 ====== ============================================================ 206 EFAULT the msr index list cannot be read from or written to 207 E2BIG the msr index list is to be to fit in the array specified by 208 the user. 209 ====== ============================================================ 210 211:: 212 213 struct kvm_msr_list { 214 __u32 nmsrs; /* number of msrs in entries */ 215 __u32 indices[0]; 216 }; 217 218The user fills in the size of the indices array in nmsrs, and in return 219kvm adjusts nmsrs to reflect the actual number of msrs and fills in the 220indices array with their numbers. 221 222KVM_GET_MSR_INDEX_LIST returns the guest msrs that are supported. The list 223varies by kvm version and host processor, but does not change otherwise. 224 225Note: if kvm indicates supports MCE (KVM_CAP_MCE), then the MCE bank MSRs are 226not returned in the MSR list, as different vcpus can have a different number 227of banks, as set via the KVM_X86_SETUP_MCE ioctl. 228 229KVM_GET_MSR_FEATURE_INDEX_LIST returns the list of MSRs that can be passed 230to the KVM_GET_MSRS system ioctl. This lets userspace probe host capabilities 231and processor features that are exposed via MSRs (e.g., VMX capabilities). 232This list also varies by kvm version and host processor, but does not change 233otherwise. 234 235 2364.4 KVM_CHECK_EXTENSION 237----------------------- 238 239:Capability: basic, KVM_CAP_CHECK_EXTENSION_VM for vm ioctl 240:Architectures: all 241:Type: system ioctl, vm ioctl 242:Parameters: extension identifier (KVM_CAP_*) 243:Returns: 0 if unsupported; 1 (or some other positive integer) if supported 244 245The API allows the application to query about extensions to the core 246kvm API. Userspace passes an extension identifier (an integer) and 247receives an integer that describes the extension availability. 248Generally 0 means no and 1 means yes, but some extensions may report 249additional information in the integer return value. 250 251Based on their initialization different VMs may have different capabilities. 252It is thus encouraged to use the vm ioctl to query for capabilities (available 253with KVM_CAP_CHECK_EXTENSION_VM on the vm fd) 254 2554.5 KVM_GET_VCPU_MMAP_SIZE 256-------------------------- 257 258:Capability: basic 259:Architectures: all 260:Type: system ioctl 261:Parameters: none 262:Returns: size of vcpu mmap area, in bytes 263 264The KVM_RUN ioctl (cf.) communicates with userspace via a shared 265memory region. This ioctl returns the size of that region. See the 266KVM_RUN documentation for details. 267 268Besides the size of the KVM_RUN communication region, other areas of 269the VCPU file descriptor can be mmap-ed, including: 270 271- if KVM_CAP_COALESCED_MMIO is available, a page at 272 KVM_COALESCED_MMIO_PAGE_OFFSET * PAGE_SIZE; for historical reasons, 273 this page is included in the result of KVM_GET_VCPU_MMAP_SIZE. 274 KVM_CAP_COALESCED_MMIO is not documented yet. 275 276- if KVM_CAP_DIRTY_LOG_RING is available, a number of pages at 277 KVM_DIRTY_LOG_PAGE_OFFSET * PAGE_SIZE. For more information on 278 KVM_CAP_DIRTY_LOG_RING, see section 8.3. 279 280 2814.6 KVM_SET_MEMORY_REGION 282------------------------- 283 284:Capability: basic 285:Architectures: all 286:Type: vm ioctl 287:Parameters: struct kvm_memory_region (in) 288:Returns: 0 on success, -1 on error 289 290This ioctl is obsolete and has been removed. 291 292 2934.7 KVM_CREATE_VCPU 294------------------- 295 296:Capability: basic 297:Architectures: all 298:Type: vm ioctl 299:Parameters: vcpu id (apic id on x86) 300:Returns: vcpu fd on success, -1 on error 301 302This API adds a vcpu to a virtual machine. No more than max_vcpus may be added. 303The vcpu id is an integer in the range [0, max_vcpu_id). 304 305The recommended max_vcpus value can be retrieved using the KVM_CAP_NR_VCPUS of 306the KVM_CHECK_EXTENSION ioctl() at run-time. 307The maximum possible value for max_vcpus can be retrieved using the 308KVM_CAP_MAX_VCPUS of the KVM_CHECK_EXTENSION ioctl() at run-time. 309 310If the KVM_CAP_NR_VCPUS does not exist, you should assume that max_vcpus is 4 311cpus max. 312If the KVM_CAP_MAX_VCPUS does not exist, you should assume that max_vcpus is 313same as the value returned from KVM_CAP_NR_VCPUS. 314 315The maximum possible value for max_vcpu_id can be retrieved using the 316KVM_CAP_MAX_VCPU_ID of the KVM_CHECK_EXTENSION ioctl() at run-time. 317 318If the KVM_CAP_MAX_VCPU_ID does not exist, you should assume that max_vcpu_id 319is the same as the value returned from KVM_CAP_MAX_VCPUS. 320 321On powerpc using book3s_hv mode, the vcpus are mapped onto virtual 322threads in one or more virtual CPU cores. (This is because the 323hardware requires all the hardware threads in a CPU core to be in the 324same partition.) The KVM_CAP_PPC_SMT capability indicates the number 325of vcpus per virtual core (vcore). The vcore id is obtained by 326dividing the vcpu id by the number of vcpus per vcore. The vcpus in a 327given vcore will always be in the same physical core as each other 328(though that might be a different physical core from time to time). 329Userspace can control the threading (SMT) mode of the guest by its 330allocation of vcpu ids. For example, if userspace wants 331single-threaded guest vcpus, it should make all vcpu ids be a multiple 332of the number of vcpus per vcore. 333 334For virtual cpus that have been created with S390 user controlled virtual 335machines, the resulting vcpu fd can be memory mapped at page offset 336KVM_S390_SIE_PAGE_OFFSET in order to obtain a memory map of the virtual 337cpu's hardware control block. 338 339 3404.8 KVM_GET_DIRTY_LOG (vm ioctl) 341-------------------------------- 342 343:Capability: basic 344:Architectures: all 345:Type: vm ioctl 346:Parameters: struct kvm_dirty_log (in/out) 347:Returns: 0 on success, -1 on error 348 349:: 350 351 /* for KVM_GET_DIRTY_LOG */ 352 struct kvm_dirty_log { 353 __u32 slot; 354 __u32 padding; 355 union { 356 void __user *dirty_bitmap; /* one bit per page */ 357 __u64 padding; 358 }; 359 }; 360 361Given a memory slot, return a bitmap containing any pages dirtied 362since the last call to this ioctl. Bit 0 is the first page in the 363memory slot. Ensure the entire structure is cleared to avoid padding 364issues. 365 366If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of slot field specifies 367the address space for which you want to return the dirty bitmap. See 368KVM_SET_USER_MEMORY_REGION for details on the usage of slot field. 369 370The bits in the dirty bitmap are cleared before the ioctl returns, unless 371KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 is enabled. For more information, 372see the description of the capability. 373 3744.9 KVM_SET_MEMORY_ALIAS 375------------------------ 376 377:Capability: basic 378:Architectures: x86 379:Type: vm ioctl 380:Parameters: struct kvm_memory_alias (in) 381:Returns: 0 (success), -1 (error) 382 383This ioctl is obsolete and has been removed. 384 385 3864.10 KVM_RUN 387------------ 388 389:Capability: basic 390:Architectures: all 391:Type: vcpu ioctl 392:Parameters: none 393:Returns: 0 on success, -1 on error 394 395Errors: 396 397 ======= ============================================================== 398 EINTR an unmasked signal is pending 399 ENOEXEC the vcpu hasn't been initialized or the guest tried to execute 400 instructions from device memory (arm64) 401 ENOSYS data abort outside memslots with no syndrome info and 402 KVM_CAP_ARM_NISV_TO_USER not enabled (arm64) 403 EPERM SVE feature set but not finalized (arm64) 404 ======= ============================================================== 405 406This ioctl is used to run a guest virtual cpu. While there are no 407explicit parameters, there is an implicit parameter block that can be 408obtained by mmap()ing the vcpu fd at offset 0, with the size given by 409KVM_GET_VCPU_MMAP_SIZE. The parameter block is formatted as a 'struct 410kvm_run' (see below). 411 412 4134.11 KVM_GET_REGS 414----------------- 415 416:Capability: basic 417:Architectures: all except ARM, arm64 418:Type: vcpu ioctl 419:Parameters: struct kvm_regs (out) 420:Returns: 0 on success, -1 on error 421 422Reads the general purpose registers from the vcpu. 423 424:: 425 426 /* x86 */ 427 struct kvm_regs { 428 /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */ 429 __u64 rax, rbx, rcx, rdx; 430 __u64 rsi, rdi, rsp, rbp; 431 __u64 r8, r9, r10, r11; 432 __u64 r12, r13, r14, r15; 433 __u64 rip, rflags; 434 }; 435 436 /* mips */ 437 struct kvm_regs { 438 /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */ 439 __u64 gpr[32]; 440 __u64 hi; 441 __u64 lo; 442 __u64 pc; 443 }; 444 445 4464.12 KVM_SET_REGS 447----------------- 448 449:Capability: basic 450:Architectures: all except ARM, arm64 451:Type: vcpu ioctl 452:Parameters: struct kvm_regs (in) 453:Returns: 0 on success, -1 on error 454 455Writes the general purpose registers into the vcpu. 456 457See KVM_GET_REGS for the data structure. 458 459 4604.13 KVM_GET_SREGS 461------------------ 462 463:Capability: basic 464:Architectures: x86, ppc 465:Type: vcpu ioctl 466:Parameters: struct kvm_sregs (out) 467:Returns: 0 on success, -1 on error 468 469Reads special registers from the vcpu. 470 471:: 472 473 /* x86 */ 474 struct kvm_sregs { 475 struct kvm_segment cs, ds, es, fs, gs, ss; 476 struct kvm_segment tr, ldt; 477 struct kvm_dtable gdt, idt; 478 __u64 cr0, cr2, cr3, cr4, cr8; 479 __u64 efer; 480 __u64 apic_base; 481 __u64 interrupt_bitmap[(KVM_NR_INTERRUPTS + 63) / 64]; 482 }; 483 484 /* ppc -- see arch/powerpc/include/uapi/asm/kvm.h */ 485 486interrupt_bitmap is a bitmap of pending external interrupts. At most 487one bit may be set. This interrupt has been acknowledged by the APIC 488but not yet injected into the cpu core. 489 490 4914.14 KVM_SET_SREGS 492------------------ 493 494:Capability: basic 495:Architectures: x86, ppc 496:Type: vcpu ioctl 497:Parameters: struct kvm_sregs (in) 498:Returns: 0 on success, -1 on error 499 500Writes special registers into the vcpu. See KVM_GET_SREGS for the 501data structures. 502 503 5044.15 KVM_TRANSLATE 505------------------ 506 507:Capability: basic 508:Architectures: x86 509:Type: vcpu ioctl 510:Parameters: struct kvm_translation (in/out) 511:Returns: 0 on success, -1 on error 512 513Translates a virtual address according to the vcpu's current address 514translation mode. 515 516:: 517 518 struct kvm_translation { 519 /* in */ 520 __u64 linear_address; 521 522 /* out */ 523 __u64 physical_address; 524 __u8 valid; 525 __u8 writeable; 526 __u8 usermode; 527 __u8 pad[5]; 528 }; 529 530 5314.16 KVM_INTERRUPT 532------------------ 533 534:Capability: basic 535:Architectures: x86, ppc, mips 536:Type: vcpu ioctl 537:Parameters: struct kvm_interrupt (in) 538:Returns: 0 on success, negative on failure. 539 540Queues a hardware interrupt vector to be injected. 541 542:: 543 544 /* for KVM_INTERRUPT */ 545 struct kvm_interrupt { 546 /* in */ 547 __u32 irq; 548 }; 549 550X86: 551^^^^ 552 553:Returns: 554 555 ========= =================================== 556 0 on success, 557 -EEXIST if an interrupt is already enqueued 558 -EINVAL the irq number is invalid 559 -ENXIO if the PIC is in the kernel 560 -EFAULT if the pointer is invalid 561 ========= =================================== 562 563Note 'irq' is an interrupt vector, not an interrupt pin or line. This 564ioctl is useful if the in-kernel PIC is not used. 565 566PPC: 567^^^^ 568 569Queues an external interrupt to be injected. This ioctl is overleaded 570with 3 different irq values: 571 572a) KVM_INTERRUPT_SET 573 574 This injects an edge type external interrupt into the guest once it's ready 575 to receive interrupts. When injected, the interrupt is done. 576 577b) KVM_INTERRUPT_UNSET 578 579 This unsets any pending interrupt. 580 581 Only available with KVM_CAP_PPC_UNSET_IRQ. 582 583c) KVM_INTERRUPT_SET_LEVEL 584 585 This injects a level type external interrupt into the guest context. The 586 interrupt stays pending until a specific ioctl with KVM_INTERRUPT_UNSET 587 is triggered. 588 589 Only available with KVM_CAP_PPC_IRQ_LEVEL. 590 591Note that any value for 'irq' other than the ones stated above is invalid 592and incurs unexpected behavior. 593 594This is an asynchronous vcpu ioctl and can be invoked from any thread. 595 596MIPS: 597^^^^^ 598 599Queues an external interrupt to be injected into the virtual CPU. A negative 600interrupt number dequeues the interrupt. 601 602This is an asynchronous vcpu ioctl and can be invoked from any thread. 603 604 6054.17 KVM_DEBUG_GUEST 606-------------------- 607 608:Capability: basic 609:Architectures: none 610:Type: vcpu ioctl 611:Parameters: none) 612:Returns: -1 on error 613 614Support for this has been removed. Use KVM_SET_GUEST_DEBUG instead. 615 616 6174.18 KVM_GET_MSRS 618----------------- 619 620:Capability: basic (vcpu), KVM_CAP_GET_MSR_FEATURES (system) 621:Architectures: x86 622:Type: system ioctl, vcpu ioctl 623:Parameters: struct kvm_msrs (in/out) 624:Returns: number of msrs successfully returned; 625 -1 on error 626 627When used as a system ioctl: 628Reads the values of MSR-based features that are available for the VM. This 629is similar to KVM_GET_SUPPORTED_CPUID, but it returns MSR indices and values. 630The list of msr-based features can be obtained using KVM_GET_MSR_FEATURE_INDEX_LIST 631in a system ioctl. 632 633When used as a vcpu ioctl: 634Reads model-specific registers from the vcpu. Supported msr indices can 635be obtained using KVM_GET_MSR_INDEX_LIST in a system ioctl. 636 637:: 638 639 struct kvm_msrs { 640 __u32 nmsrs; /* number of msrs in entries */ 641 __u32 pad; 642 643 struct kvm_msr_entry entries[0]; 644 }; 645 646 struct kvm_msr_entry { 647 __u32 index; 648 __u32 reserved; 649 __u64 data; 650 }; 651 652Application code should set the 'nmsrs' member (which indicates the 653size of the entries array) and the 'index' member of each array entry. 654kvm will fill in the 'data' member. 655 656 6574.19 KVM_SET_MSRS 658----------------- 659 660:Capability: basic 661:Architectures: x86 662:Type: vcpu ioctl 663:Parameters: struct kvm_msrs (in) 664:Returns: number of msrs successfully set (see below), -1 on error 665 666Writes model-specific registers to the vcpu. See KVM_GET_MSRS for the 667data structures. 668 669Application code should set the 'nmsrs' member (which indicates the 670size of the entries array), and the 'index' and 'data' members of each 671array entry. 672 673It tries to set the MSRs in array entries[] one by one. If setting an MSR 674fails, e.g., due to setting reserved bits, the MSR isn't supported/emulated 675by KVM, etc..., it stops processing the MSR list and returns the number of 676MSRs that have been set successfully. 677 678 6794.20 KVM_SET_CPUID 680------------------ 681 682:Capability: basic 683:Architectures: x86 684:Type: vcpu ioctl 685:Parameters: struct kvm_cpuid (in) 686:Returns: 0 on success, -1 on error 687 688Defines the vcpu responses to the cpuid instruction. Applications 689should use the KVM_SET_CPUID2 ioctl if available. 690 691Note, when this IOCTL fails, KVM gives no guarantees that previous valid CPUID 692configuration (if there is) is not corrupted. Userspace can get a copy of the 693resulting CPUID configuration through KVM_GET_CPUID2 in case. 694 695:: 696 697 struct kvm_cpuid_entry { 698 __u32 function; 699 __u32 eax; 700 __u32 ebx; 701 __u32 ecx; 702 __u32 edx; 703 __u32 padding; 704 }; 705 706 /* for KVM_SET_CPUID */ 707 struct kvm_cpuid { 708 __u32 nent; 709 __u32 padding; 710 struct kvm_cpuid_entry entries[0]; 711 }; 712 713 7144.21 KVM_SET_SIGNAL_MASK 715------------------------ 716 717:Capability: basic 718:Architectures: all 719:Type: vcpu ioctl 720:Parameters: struct kvm_signal_mask (in) 721:Returns: 0 on success, -1 on error 722 723Defines which signals are blocked during execution of KVM_RUN. This 724signal mask temporarily overrides the threads signal mask. Any 725unblocked signal received (except SIGKILL and SIGSTOP, which retain 726their traditional behaviour) will cause KVM_RUN to return with -EINTR. 727 728Note the signal will only be delivered if not blocked by the original 729signal mask. 730 731:: 732 733 /* for KVM_SET_SIGNAL_MASK */ 734 struct kvm_signal_mask { 735 __u32 len; 736 __u8 sigset[0]; 737 }; 738 739 7404.22 KVM_GET_FPU 741---------------- 742 743:Capability: basic 744:Architectures: x86 745:Type: vcpu ioctl 746:Parameters: struct kvm_fpu (out) 747:Returns: 0 on success, -1 on error 748 749Reads the floating point state from the vcpu. 750 751:: 752 753 /* for KVM_GET_FPU and KVM_SET_FPU */ 754 struct kvm_fpu { 755 __u8 fpr[8][16]; 756 __u16 fcw; 757 __u16 fsw; 758 __u8 ftwx; /* in fxsave format */ 759 __u8 pad1; 760 __u16 last_opcode; 761 __u64 last_ip; 762 __u64 last_dp; 763 __u8 xmm[16][16]; 764 __u32 mxcsr; 765 __u32 pad2; 766 }; 767 768 7694.23 KVM_SET_FPU 770---------------- 771 772:Capability: basic 773:Architectures: x86 774:Type: vcpu ioctl 775:Parameters: struct kvm_fpu (in) 776:Returns: 0 on success, -1 on error 777 778Writes the floating point state to the vcpu. 779 780:: 781 782 /* for KVM_GET_FPU and KVM_SET_FPU */ 783 struct kvm_fpu { 784 __u8 fpr[8][16]; 785 __u16 fcw; 786 __u16 fsw; 787 __u8 ftwx; /* in fxsave format */ 788 __u8 pad1; 789 __u16 last_opcode; 790 __u64 last_ip; 791 __u64 last_dp; 792 __u8 xmm[16][16]; 793 __u32 mxcsr; 794 __u32 pad2; 795 }; 796 797 7984.24 KVM_CREATE_IRQCHIP 799----------------------- 800 801:Capability: KVM_CAP_IRQCHIP, KVM_CAP_S390_IRQCHIP (s390) 802:Architectures: x86, ARM, arm64, s390 803:Type: vm ioctl 804:Parameters: none 805:Returns: 0 on success, -1 on error 806 807Creates an interrupt controller model in the kernel. 808On x86, creates a virtual ioapic, a virtual PIC (two PICs, nested), and sets up 809future vcpus to have a local APIC. IRQ routing for GSIs 0-15 is set to both 810PIC and IOAPIC; GSI 16-23 only go to the IOAPIC. 811On ARM/arm64, a GICv2 is created. Any other GIC versions require the usage of 812KVM_CREATE_DEVICE, which also supports creating a GICv2. Using 813KVM_CREATE_DEVICE is preferred over KVM_CREATE_IRQCHIP for GICv2. 814On s390, a dummy irq routing table is created. 815 816Note that on s390 the KVM_CAP_S390_IRQCHIP vm capability needs to be enabled 817before KVM_CREATE_IRQCHIP can be used. 818 819 8204.25 KVM_IRQ_LINE 821----------------- 822 823:Capability: KVM_CAP_IRQCHIP 824:Architectures: x86, arm, arm64 825:Type: vm ioctl 826:Parameters: struct kvm_irq_level 827:Returns: 0 on success, -1 on error 828 829Sets the level of a GSI input to the interrupt controller model in the kernel. 830On some architectures it is required that an interrupt controller model has 831been previously created with KVM_CREATE_IRQCHIP. Note that edge-triggered 832interrupts require the level to be set to 1 and then back to 0. 833 834On real hardware, interrupt pins can be active-low or active-high. This 835does not matter for the level field of struct kvm_irq_level: 1 always 836means active (asserted), 0 means inactive (deasserted). 837 838x86 allows the operating system to program the interrupt polarity 839(active-low/active-high) for level-triggered interrupts, and KVM used 840to consider the polarity. However, due to bitrot in the handling of 841active-low interrupts, the above convention is now valid on x86 too. 842This is signaled by KVM_CAP_X86_IOAPIC_POLARITY_IGNORED. Userspace 843should not present interrupts to the guest as active-low unless this 844capability is present (or unless it is not using the in-kernel irqchip, 845of course). 846 847 848ARM/arm64 can signal an interrupt either at the CPU level, or at the 849in-kernel irqchip (GIC), and for in-kernel irqchip can tell the GIC to 850use PPIs designated for specific cpus. The irq field is interpreted 851like this:: 852 853 bits: | 31 ... 28 | 27 ... 24 | 23 ... 16 | 15 ... 0 | 854 field: | vcpu2_index | irq_type | vcpu_index | irq_id | 855 856The irq_type field has the following values: 857 858- irq_type[0]: 859 out-of-kernel GIC: irq_id 0 is IRQ, irq_id 1 is FIQ 860- irq_type[1]: 861 in-kernel GIC: SPI, irq_id between 32 and 1019 (incl.) 862 (the vcpu_index field is ignored) 863- irq_type[2]: 864 in-kernel GIC: PPI, irq_id between 16 and 31 (incl.) 865 866(The irq_id field thus corresponds nicely to the IRQ ID in the ARM GIC specs) 867 868In both cases, level is used to assert/deassert the line. 869 870When KVM_CAP_ARM_IRQ_LINE_LAYOUT_2 is supported, the target vcpu is 871identified as (256 * vcpu2_index + vcpu_index). Otherwise, vcpu2_index 872must be zero. 873 874Note that on arm/arm64, the KVM_CAP_IRQCHIP capability only conditions 875injection of interrupts for the in-kernel irqchip. KVM_IRQ_LINE can always 876be used for a userspace interrupt controller. 877 878:: 879 880 struct kvm_irq_level { 881 union { 882 __u32 irq; /* GSI */ 883 __s32 status; /* not used for KVM_IRQ_LEVEL */ 884 }; 885 __u32 level; /* 0 or 1 */ 886 }; 887 888 8894.26 KVM_GET_IRQCHIP 890-------------------- 891 892:Capability: KVM_CAP_IRQCHIP 893:Architectures: x86 894:Type: vm ioctl 895:Parameters: struct kvm_irqchip (in/out) 896:Returns: 0 on success, -1 on error 897 898Reads the state of a kernel interrupt controller created with 899KVM_CREATE_IRQCHIP into a buffer provided by the caller. 900 901:: 902 903 struct kvm_irqchip { 904 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */ 905 __u32 pad; 906 union { 907 char dummy[512]; /* reserving space */ 908 struct kvm_pic_state pic; 909 struct kvm_ioapic_state ioapic; 910 } chip; 911 }; 912 913 9144.27 KVM_SET_IRQCHIP 915-------------------- 916 917:Capability: KVM_CAP_IRQCHIP 918:Architectures: x86 919:Type: vm ioctl 920:Parameters: struct kvm_irqchip (in) 921:Returns: 0 on success, -1 on error 922 923Sets the state of a kernel interrupt controller created with 924KVM_CREATE_IRQCHIP from a buffer provided by the caller. 925 926:: 927 928 struct kvm_irqchip { 929 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */ 930 __u32 pad; 931 union { 932 char dummy[512]; /* reserving space */ 933 struct kvm_pic_state pic; 934 struct kvm_ioapic_state ioapic; 935 } chip; 936 }; 937 938 9394.28 KVM_XEN_HVM_CONFIG 940----------------------- 941 942:Capability: KVM_CAP_XEN_HVM 943:Architectures: x86 944:Type: vm ioctl 945:Parameters: struct kvm_xen_hvm_config (in) 946:Returns: 0 on success, -1 on error 947 948Sets the MSR that the Xen HVM guest uses to initialize its hypercall 949page, and provides the starting address and size of the hypercall 950blobs in userspace. When the guest writes the MSR, kvm copies one 951page of a blob (32- or 64-bit, depending on the vcpu mode) to guest 952memory. 953 954:: 955 956 struct kvm_xen_hvm_config { 957 __u32 flags; 958 __u32 msr; 959 __u64 blob_addr_32; 960 __u64 blob_addr_64; 961 __u8 blob_size_32; 962 __u8 blob_size_64; 963 __u8 pad2[30]; 964 }; 965 966If the KVM_XEN_HVM_CONFIG_INTERCEPT_HCALL flag is returned from the 967KVM_CAP_XEN_HVM check, it may be set in the flags field of this ioctl. 968This requests KVM to generate the contents of the hypercall page 969automatically; hypercalls will be intercepted and passed to userspace 970through KVM_EXIT_XEN. In this case, all of the blob size and address 971fields must be zero. 972 973No other flags are currently valid in the struct kvm_xen_hvm_config. 974 9754.29 KVM_GET_CLOCK 976------------------ 977 978:Capability: KVM_CAP_ADJUST_CLOCK 979:Architectures: x86 980:Type: vm ioctl 981:Parameters: struct kvm_clock_data (out) 982:Returns: 0 on success, -1 on error 983 984Gets the current timestamp of kvmclock as seen by the current guest. In 985conjunction with KVM_SET_CLOCK, it is used to ensure monotonicity on scenarios 986such as migration. 987 988When KVM_CAP_ADJUST_CLOCK is passed to KVM_CHECK_EXTENSION, it returns the 989set of bits that KVM can return in struct kvm_clock_data's flag member. 990 991The only flag defined now is KVM_CLOCK_TSC_STABLE. If set, the returned 992value is the exact kvmclock value seen by all VCPUs at the instant 993when KVM_GET_CLOCK was called. If clear, the returned value is simply 994CLOCK_MONOTONIC plus a constant offset; the offset can be modified 995with KVM_SET_CLOCK. KVM will try to make all VCPUs follow this clock, 996but the exact value read by each VCPU could differ, because the host 997TSC is not stable. 998 999:: 1000 1001 struct kvm_clock_data { 1002 __u64 clock; /* kvmclock current value */ 1003 __u32 flags; 1004 __u32 pad[9]; 1005 }; 1006 1007 10084.30 KVM_SET_CLOCK 1009------------------ 1010 1011:Capability: KVM_CAP_ADJUST_CLOCK 1012:Architectures: x86 1013:Type: vm ioctl 1014:Parameters: struct kvm_clock_data (in) 1015:Returns: 0 on success, -1 on error 1016 1017Sets the current timestamp of kvmclock to the value specified in its parameter. 1018In conjunction with KVM_GET_CLOCK, it is used to ensure monotonicity on scenarios 1019such as migration. 1020 1021:: 1022 1023 struct kvm_clock_data { 1024 __u64 clock; /* kvmclock current value */ 1025 __u32 flags; 1026 __u32 pad[9]; 1027 }; 1028 1029 10304.31 KVM_GET_VCPU_EVENTS 1031------------------------ 1032 1033:Capability: KVM_CAP_VCPU_EVENTS 1034:Extended by: KVM_CAP_INTR_SHADOW 1035:Architectures: x86, arm, arm64 1036:Type: vcpu ioctl 1037:Parameters: struct kvm_vcpu_event (out) 1038:Returns: 0 on success, -1 on error 1039 1040X86: 1041^^^^ 1042 1043Gets currently pending exceptions, interrupts, and NMIs as well as related 1044states of the vcpu. 1045 1046:: 1047 1048 struct kvm_vcpu_events { 1049 struct { 1050 __u8 injected; 1051 __u8 nr; 1052 __u8 has_error_code; 1053 __u8 pending; 1054 __u32 error_code; 1055 } exception; 1056 struct { 1057 __u8 injected; 1058 __u8 nr; 1059 __u8 soft; 1060 __u8 shadow; 1061 } interrupt; 1062 struct { 1063 __u8 injected; 1064 __u8 pending; 1065 __u8 masked; 1066 __u8 pad; 1067 } nmi; 1068 __u32 sipi_vector; 1069 __u32 flags; 1070 struct { 1071 __u8 smm; 1072 __u8 pending; 1073 __u8 smm_inside_nmi; 1074 __u8 latched_init; 1075 } smi; 1076 __u8 reserved[27]; 1077 __u8 exception_has_payload; 1078 __u64 exception_payload; 1079 }; 1080 1081The following bits are defined in the flags field: 1082 1083- KVM_VCPUEVENT_VALID_SHADOW may be set to signal that 1084 interrupt.shadow contains a valid state. 1085 1086- KVM_VCPUEVENT_VALID_SMM may be set to signal that smi contains a 1087 valid state. 1088 1089- KVM_VCPUEVENT_VALID_PAYLOAD may be set to signal that the 1090 exception_has_payload, exception_payload, and exception.pending 1091 fields contain a valid state. This bit will be set whenever 1092 KVM_CAP_EXCEPTION_PAYLOAD is enabled. 1093 1094ARM/ARM64: 1095^^^^^^^^^^ 1096 1097If the guest accesses a device that is being emulated by the host kernel in 1098such a way that a real device would generate a physical SError, KVM may make 1099a virtual SError pending for that VCPU. This system error interrupt remains 1100pending until the guest takes the exception by unmasking PSTATE.A. 1101 1102Running the VCPU may cause it to take a pending SError, or make an access that 1103causes an SError to become pending. The event's description is only valid while 1104the VPCU is not running. 1105 1106This API provides a way to read and write the pending 'event' state that is not 1107visible to the guest. To save, restore or migrate a VCPU the struct representing 1108the state can be read then written using this GET/SET API, along with the other 1109guest-visible registers. It is not possible to 'cancel' an SError that has been 1110made pending. 1111 1112A device being emulated in user-space may also wish to generate an SError. To do 1113this the events structure can be populated by user-space. The current state 1114should be read first, to ensure no existing SError is pending. If an existing 1115SError is pending, the architecture's 'Multiple SError interrupts' rules should 1116be followed. (2.5.3 of DDI0587.a "ARM Reliability, Availability, and 1117Serviceability (RAS) Specification"). 1118 1119SError exceptions always have an ESR value. Some CPUs have the ability to 1120specify what the virtual SError's ESR value should be. These systems will 1121advertise KVM_CAP_ARM_INJECT_SERROR_ESR. In this case exception.has_esr will 1122always have a non-zero value when read, and the agent making an SError pending 1123should specify the ISS field in the lower 24 bits of exception.serror_esr. If 1124the system supports KVM_CAP_ARM_INJECT_SERROR_ESR, but user-space sets the events 1125with exception.has_esr as zero, KVM will choose an ESR. 1126 1127Specifying exception.has_esr on a system that does not support it will return 1128-EINVAL. Setting anything other than the lower 24bits of exception.serror_esr 1129will return -EINVAL. 1130 1131It is not possible to read back a pending external abort (injected via 1132KVM_SET_VCPU_EVENTS or otherwise) because such an exception is always delivered 1133directly to the virtual CPU). 1134 1135:: 1136 1137 struct kvm_vcpu_events { 1138 struct { 1139 __u8 serror_pending; 1140 __u8 serror_has_esr; 1141 __u8 ext_dabt_pending; 1142 /* Align it to 8 bytes */ 1143 __u8 pad[5]; 1144 __u64 serror_esr; 1145 } exception; 1146 __u32 reserved[12]; 1147 }; 1148 11494.32 KVM_SET_VCPU_EVENTS 1150------------------------ 1151 1152:Capability: KVM_CAP_VCPU_EVENTS 1153:Extended by: KVM_CAP_INTR_SHADOW 1154:Architectures: x86, arm, arm64 1155:Type: vcpu ioctl 1156:Parameters: struct kvm_vcpu_event (in) 1157:Returns: 0 on success, -1 on error 1158 1159X86: 1160^^^^ 1161 1162Set pending exceptions, interrupts, and NMIs as well as related states of the 1163vcpu. 1164 1165See KVM_GET_VCPU_EVENTS for the data structure. 1166 1167Fields that may be modified asynchronously by running VCPUs can be excluded 1168from the update. These fields are nmi.pending, sipi_vector, smi.smm, 1169smi.pending. Keep the corresponding bits in the flags field cleared to 1170suppress overwriting the current in-kernel state. The bits are: 1171 1172=============================== ================================== 1173KVM_VCPUEVENT_VALID_NMI_PENDING transfer nmi.pending to the kernel 1174KVM_VCPUEVENT_VALID_SIPI_VECTOR transfer sipi_vector 1175KVM_VCPUEVENT_VALID_SMM transfer the smi sub-struct. 1176=============================== ================================== 1177 1178If KVM_CAP_INTR_SHADOW is available, KVM_VCPUEVENT_VALID_SHADOW can be set in 1179the flags field to signal that interrupt.shadow contains a valid state and 1180shall be written into the VCPU. 1181 1182KVM_VCPUEVENT_VALID_SMM can only be set if KVM_CAP_X86_SMM is available. 1183 1184If KVM_CAP_EXCEPTION_PAYLOAD is enabled, KVM_VCPUEVENT_VALID_PAYLOAD 1185can be set in the flags field to signal that the 1186exception_has_payload, exception_payload, and exception.pending fields 1187contain a valid state and shall be written into the VCPU. 1188 1189ARM/ARM64: 1190^^^^^^^^^^ 1191 1192User space may need to inject several types of events to the guest. 1193 1194Set the pending SError exception state for this VCPU. It is not possible to 1195'cancel' an Serror that has been made pending. 1196 1197If the guest performed an access to I/O memory which could not be handled by 1198userspace, for example because of missing instruction syndrome decode 1199information or because there is no device mapped at the accessed IPA, then 1200userspace can ask the kernel to inject an external abort using the address 1201from the exiting fault on the VCPU. It is a programming error to set 1202ext_dabt_pending after an exit which was not either KVM_EXIT_MMIO or 1203KVM_EXIT_ARM_NISV. This feature is only available if the system supports 1204KVM_CAP_ARM_INJECT_EXT_DABT. This is a helper which provides commonality in 1205how userspace reports accesses for the above cases to guests, across different 1206userspace implementations. Nevertheless, userspace can still emulate all Arm 1207exceptions by manipulating individual registers using the KVM_SET_ONE_REG API. 1208 1209See KVM_GET_VCPU_EVENTS for the data structure. 1210 1211 12124.33 KVM_GET_DEBUGREGS 1213---------------------- 1214 1215:Capability: KVM_CAP_DEBUGREGS 1216:Architectures: x86 1217:Type: vm ioctl 1218:Parameters: struct kvm_debugregs (out) 1219:Returns: 0 on success, -1 on error 1220 1221Reads debug registers from the vcpu. 1222 1223:: 1224 1225 struct kvm_debugregs { 1226 __u64 db[4]; 1227 __u64 dr6; 1228 __u64 dr7; 1229 __u64 flags; 1230 __u64 reserved[9]; 1231 }; 1232 1233 12344.34 KVM_SET_DEBUGREGS 1235---------------------- 1236 1237:Capability: KVM_CAP_DEBUGREGS 1238:Architectures: x86 1239:Type: vm ioctl 1240:Parameters: struct kvm_debugregs (in) 1241:Returns: 0 on success, -1 on error 1242 1243Writes debug registers into the vcpu. 1244 1245See KVM_GET_DEBUGREGS for the data structure. The flags field is unused 1246yet and must be cleared on entry. 1247 1248 12494.35 KVM_SET_USER_MEMORY_REGION 1250------------------------------- 1251 1252:Capability: KVM_CAP_USER_MEMORY 1253:Architectures: all 1254:Type: vm ioctl 1255:Parameters: struct kvm_userspace_memory_region (in) 1256:Returns: 0 on success, -1 on error 1257 1258:: 1259 1260 struct kvm_userspace_memory_region { 1261 __u32 slot; 1262 __u32 flags; 1263 __u64 guest_phys_addr; 1264 __u64 memory_size; /* bytes */ 1265 __u64 userspace_addr; /* start of the userspace allocated memory */ 1266 }; 1267 1268 /* for kvm_memory_region::flags */ 1269 #define KVM_MEM_LOG_DIRTY_PAGES (1UL << 0) 1270 #define KVM_MEM_READONLY (1UL << 1) 1271 1272This ioctl allows the user to create, modify or delete a guest physical 1273memory slot. Bits 0-15 of "slot" specify the slot id and this value 1274should be less than the maximum number of user memory slots supported per 1275VM. The maximum allowed slots can be queried using KVM_CAP_NR_MEMSLOTS. 1276Slots may not overlap in guest physical address space. 1277 1278If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of "slot" 1279specifies the address space which is being modified. They must be 1280less than the value that KVM_CHECK_EXTENSION returns for the 1281KVM_CAP_MULTI_ADDRESS_SPACE capability. Slots in separate address spaces 1282are unrelated; the restriction on overlapping slots only applies within 1283each address space. 1284 1285Deleting a slot is done by passing zero for memory_size. When changing 1286an existing slot, it may be moved in the guest physical memory space, 1287or its flags may be modified, but it may not be resized. 1288 1289Memory for the region is taken starting at the address denoted by the 1290field userspace_addr, which must point at user addressable memory for 1291the entire memory slot size. Any object may back this memory, including 1292anonymous memory, ordinary files, and hugetlbfs. 1293 1294On architectures that support a form of address tagging, userspace_addr must 1295be an untagged address. 1296 1297It is recommended that the lower 21 bits of guest_phys_addr and userspace_addr 1298be identical. This allows large pages in the guest to be backed by large 1299pages in the host. 1300 1301The flags field supports two flags: KVM_MEM_LOG_DIRTY_PAGES and 1302KVM_MEM_READONLY. The former can be set to instruct KVM to keep track of 1303writes to memory within the slot. See KVM_GET_DIRTY_LOG ioctl to know how to 1304use it. The latter can be set, if KVM_CAP_READONLY_MEM capability allows it, 1305to make a new slot read-only. In this case, writes to this memory will be 1306posted to userspace as KVM_EXIT_MMIO exits. 1307 1308When the KVM_CAP_SYNC_MMU capability is available, changes in the backing of 1309the memory region are automatically reflected into the guest. For example, an 1310mmap() that affects the region will be made visible immediately. Another 1311example is madvise(MADV_DROP). 1312 1313It is recommended to use this API instead of the KVM_SET_MEMORY_REGION ioctl. 1314The KVM_SET_MEMORY_REGION does not allow fine grained control over memory 1315allocation and is deprecated. 1316 1317 13184.36 KVM_SET_TSS_ADDR 1319--------------------- 1320 1321:Capability: KVM_CAP_SET_TSS_ADDR 1322:Architectures: x86 1323:Type: vm ioctl 1324:Parameters: unsigned long tss_address (in) 1325:Returns: 0 on success, -1 on error 1326 1327This ioctl defines the physical address of a three-page region in the guest 1328physical address space. The region must be within the first 4GB of the 1329guest physical address space and must not conflict with any memory slot 1330or any mmio address. The guest may malfunction if it accesses this memory 1331region. 1332 1333This ioctl is required on Intel-based hosts. This is needed on Intel hardware 1334because of a quirk in the virtualization implementation (see the internals 1335documentation when it pops into existence). 1336 1337 13384.37 KVM_ENABLE_CAP 1339------------------- 1340 1341:Capability: KVM_CAP_ENABLE_CAP 1342:Architectures: mips, ppc, s390 1343:Type: vcpu ioctl 1344:Parameters: struct kvm_enable_cap (in) 1345:Returns: 0 on success; -1 on error 1346 1347:Capability: KVM_CAP_ENABLE_CAP_VM 1348:Architectures: all 1349:Type: vm ioctl 1350:Parameters: struct kvm_enable_cap (in) 1351:Returns: 0 on success; -1 on error 1352 1353.. note:: 1354 1355 Not all extensions are enabled by default. Using this ioctl the application 1356 can enable an extension, making it available to the guest. 1357 1358On systems that do not support this ioctl, it always fails. On systems that 1359do support it, it only works for extensions that are supported for enablement. 1360 1361To check if a capability can be enabled, the KVM_CHECK_EXTENSION ioctl should 1362be used. 1363 1364:: 1365 1366 struct kvm_enable_cap { 1367 /* in */ 1368 __u32 cap; 1369 1370The capability that is supposed to get enabled. 1371 1372:: 1373 1374 __u32 flags; 1375 1376A bitfield indicating future enhancements. Has to be 0 for now. 1377 1378:: 1379 1380 __u64 args[4]; 1381 1382Arguments for enabling a feature. If a feature needs initial values to 1383function properly, this is the place to put them. 1384 1385:: 1386 1387 __u8 pad[64]; 1388 }; 1389 1390The vcpu ioctl should be used for vcpu-specific capabilities, the vm ioctl 1391for vm-wide capabilities. 1392 13934.38 KVM_GET_MP_STATE 1394--------------------- 1395 1396:Capability: KVM_CAP_MP_STATE 1397:Architectures: x86, s390, arm, arm64 1398:Type: vcpu ioctl 1399:Parameters: struct kvm_mp_state (out) 1400:Returns: 0 on success; -1 on error 1401 1402:: 1403 1404 struct kvm_mp_state { 1405 __u32 mp_state; 1406 }; 1407 1408Returns the vcpu's current "multiprocessing state" (though also valid on 1409uniprocessor guests). 1410 1411Possible values are: 1412 1413 ========================== =============================================== 1414 KVM_MP_STATE_RUNNABLE the vcpu is currently running [x86,arm/arm64] 1415 KVM_MP_STATE_UNINITIALIZED the vcpu is an application processor (AP) 1416 which has not yet received an INIT signal [x86] 1417 KVM_MP_STATE_INIT_RECEIVED the vcpu has received an INIT signal, and is 1418 now ready for a SIPI [x86] 1419 KVM_MP_STATE_HALTED the vcpu has executed a HLT instruction and 1420 is waiting for an interrupt [x86] 1421 KVM_MP_STATE_SIPI_RECEIVED the vcpu has just received a SIPI (vector 1422 accessible via KVM_GET_VCPU_EVENTS) [x86] 1423 KVM_MP_STATE_STOPPED the vcpu is stopped [s390,arm/arm64] 1424 KVM_MP_STATE_CHECK_STOP the vcpu is in a special error state [s390] 1425 KVM_MP_STATE_OPERATING the vcpu is operating (running or halted) 1426 [s390] 1427 KVM_MP_STATE_LOAD the vcpu is in a special load/startup state 1428 [s390] 1429 ========================== =============================================== 1430 1431On x86, this ioctl is only useful after KVM_CREATE_IRQCHIP. Without an 1432in-kernel irqchip, the multiprocessing state must be maintained by userspace on 1433these architectures. 1434 1435For arm/arm64: 1436^^^^^^^^^^^^^^ 1437 1438The only states that are valid are KVM_MP_STATE_STOPPED and 1439KVM_MP_STATE_RUNNABLE which reflect if the vcpu is paused or not. 1440 14414.39 KVM_SET_MP_STATE 1442--------------------- 1443 1444:Capability: KVM_CAP_MP_STATE 1445:Architectures: x86, s390, arm, arm64 1446:Type: vcpu ioctl 1447:Parameters: struct kvm_mp_state (in) 1448:Returns: 0 on success; -1 on error 1449 1450Sets the vcpu's current "multiprocessing state"; see KVM_GET_MP_STATE for 1451arguments. 1452 1453On x86, this ioctl is only useful after KVM_CREATE_IRQCHIP. Without an 1454in-kernel irqchip, the multiprocessing state must be maintained by userspace on 1455these architectures. 1456 1457For arm/arm64: 1458^^^^^^^^^^^^^^ 1459 1460The only states that are valid are KVM_MP_STATE_STOPPED and 1461KVM_MP_STATE_RUNNABLE which reflect if the vcpu should be paused or not. 1462 14634.40 KVM_SET_IDENTITY_MAP_ADDR 1464------------------------------ 1465 1466:Capability: KVM_CAP_SET_IDENTITY_MAP_ADDR 1467:Architectures: x86 1468:Type: vm ioctl 1469:Parameters: unsigned long identity (in) 1470:Returns: 0 on success, -1 on error 1471 1472This ioctl defines the physical address of a one-page region in the guest 1473physical address space. The region must be within the first 4GB of the 1474guest physical address space and must not conflict with any memory slot 1475or any mmio address. The guest may malfunction if it accesses this memory 1476region. 1477 1478Setting the address to 0 will result in resetting the address to its default 1479(0xfffbc000). 1480 1481This ioctl is required on Intel-based hosts. This is needed on Intel hardware 1482because of a quirk in the virtualization implementation (see the internals 1483documentation when it pops into existence). 1484 1485Fails if any VCPU has already been created. 1486 14874.41 KVM_SET_BOOT_CPU_ID 1488------------------------ 1489 1490:Capability: KVM_CAP_SET_BOOT_CPU_ID 1491:Architectures: x86 1492:Type: vm ioctl 1493:Parameters: unsigned long vcpu_id 1494:Returns: 0 on success, -1 on error 1495 1496Define which vcpu is the Bootstrap Processor (BSP). Values are the same 1497as the vcpu id in KVM_CREATE_VCPU. If this ioctl is not called, the default 1498is vcpu 0. 1499 1500 15014.42 KVM_GET_XSAVE 1502------------------ 1503 1504:Capability: KVM_CAP_XSAVE 1505:Architectures: x86 1506:Type: vcpu ioctl 1507:Parameters: struct kvm_xsave (out) 1508:Returns: 0 on success, -1 on error 1509 1510 1511:: 1512 1513 struct kvm_xsave { 1514 __u32 region[1024]; 1515 }; 1516 1517This ioctl would copy current vcpu's xsave struct to the userspace. 1518 1519 15204.43 KVM_SET_XSAVE 1521------------------ 1522 1523:Capability: KVM_CAP_XSAVE 1524:Architectures: x86 1525:Type: vcpu ioctl 1526:Parameters: struct kvm_xsave (in) 1527:Returns: 0 on success, -1 on error 1528 1529:: 1530 1531 1532 struct kvm_xsave { 1533 __u32 region[1024]; 1534 }; 1535 1536This ioctl would copy userspace's xsave struct to the kernel. 1537 1538 15394.44 KVM_GET_XCRS 1540----------------- 1541 1542:Capability: KVM_CAP_XCRS 1543:Architectures: x86 1544:Type: vcpu ioctl 1545:Parameters: struct kvm_xcrs (out) 1546:Returns: 0 on success, -1 on error 1547 1548:: 1549 1550 struct kvm_xcr { 1551 __u32 xcr; 1552 __u32 reserved; 1553 __u64 value; 1554 }; 1555 1556 struct kvm_xcrs { 1557 __u32 nr_xcrs; 1558 __u32 flags; 1559 struct kvm_xcr xcrs[KVM_MAX_XCRS]; 1560 __u64 padding[16]; 1561 }; 1562 1563This ioctl would copy current vcpu's xcrs to the userspace. 1564 1565 15664.45 KVM_SET_XCRS 1567----------------- 1568 1569:Capability: KVM_CAP_XCRS 1570:Architectures: x86 1571:Type: vcpu ioctl 1572:Parameters: struct kvm_xcrs (in) 1573:Returns: 0 on success, -1 on error 1574 1575:: 1576 1577 struct kvm_xcr { 1578 __u32 xcr; 1579 __u32 reserved; 1580 __u64 value; 1581 }; 1582 1583 struct kvm_xcrs { 1584 __u32 nr_xcrs; 1585 __u32 flags; 1586 struct kvm_xcr xcrs[KVM_MAX_XCRS]; 1587 __u64 padding[16]; 1588 }; 1589 1590This ioctl would set vcpu's xcr to the value userspace specified. 1591 1592 15934.46 KVM_GET_SUPPORTED_CPUID 1594---------------------------- 1595 1596:Capability: KVM_CAP_EXT_CPUID 1597:Architectures: x86 1598:Type: system ioctl 1599:Parameters: struct kvm_cpuid2 (in/out) 1600:Returns: 0 on success, -1 on error 1601 1602:: 1603 1604 struct kvm_cpuid2 { 1605 __u32 nent; 1606 __u32 padding; 1607 struct kvm_cpuid_entry2 entries[0]; 1608 }; 1609 1610 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX BIT(0) 1611 #define KVM_CPUID_FLAG_STATEFUL_FUNC BIT(1) /* deprecated */ 1612 #define KVM_CPUID_FLAG_STATE_READ_NEXT BIT(2) /* deprecated */ 1613 1614 struct kvm_cpuid_entry2 { 1615 __u32 function; 1616 __u32 index; 1617 __u32 flags; 1618 __u32 eax; 1619 __u32 ebx; 1620 __u32 ecx; 1621 __u32 edx; 1622 __u32 padding[3]; 1623 }; 1624 1625This ioctl returns x86 cpuid features which are supported by both the 1626hardware and kvm in its default configuration. Userspace can use the 1627information returned by this ioctl to construct cpuid information (for 1628KVM_SET_CPUID2) that is consistent with hardware, kernel, and 1629userspace capabilities, and with user requirements (for example, the 1630user may wish to constrain cpuid to emulate older hardware, or for 1631feature consistency across a cluster). 1632 1633Note that certain capabilities, such as KVM_CAP_X86_DISABLE_EXITS, may 1634expose cpuid features (e.g. MONITOR) which are not supported by kvm in 1635its default configuration. If userspace enables such capabilities, it 1636is responsible for modifying the results of this ioctl appropriately. 1637 1638Userspace invokes KVM_GET_SUPPORTED_CPUID by passing a kvm_cpuid2 structure 1639with the 'nent' field indicating the number of entries in the variable-size 1640array 'entries'. If the number of entries is too low to describe the cpu 1641capabilities, an error (E2BIG) is returned. If the number is too high, 1642the 'nent' field is adjusted and an error (ENOMEM) is returned. If the 1643number is just right, the 'nent' field is adjusted to the number of valid 1644entries in the 'entries' array, which is then filled. 1645 1646The entries returned are the host cpuid as returned by the cpuid instruction, 1647with unknown or unsupported features masked out. Some features (for example, 1648x2apic), may not be present in the host cpu, but are exposed by kvm if it can 1649emulate them efficiently. The fields in each entry are defined as follows: 1650 1651 function: 1652 the eax value used to obtain the entry 1653 1654 index: 1655 the ecx value used to obtain the entry (for entries that are 1656 affected by ecx) 1657 1658 flags: 1659 an OR of zero or more of the following: 1660 1661 KVM_CPUID_FLAG_SIGNIFCANT_INDEX: 1662 if the index field is valid 1663 1664 eax, ebx, ecx, edx: 1665 the values returned by the cpuid instruction for 1666 this function/index combination 1667 1668The TSC deadline timer feature (CPUID leaf 1, ecx[24]) is always returned 1669as false, since the feature depends on KVM_CREATE_IRQCHIP for local APIC 1670support. Instead it is reported via:: 1671 1672 ioctl(KVM_CHECK_EXTENSION, KVM_CAP_TSC_DEADLINE_TIMER) 1673 1674if that returns true and you use KVM_CREATE_IRQCHIP, or if you emulate the 1675feature in userspace, then you can enable the feature for KVM_SET_CPUID2. 1676 1677 16784.47 KVM_PPC_GET_PVINFO 1679----------------------- 1680 1681:Capability: KVM_CAP_PPC_GET_PVINFO 1682:Architectures: ppc 1683:Type: vm ioctl 1684:Parameters: struct kvm_ppc_pvinfo (out) 1685:Returns: 0 on success, !0 on error 1686 1687:: 1688 1689 struct kvm_ppc_pvinfo { 1690 __u32 flags; 1691 __u32 hcall[4]; 1692 __u8 pad[108]; 1693 }; 1694 1695This ioctl fetches PV specific information that need to be passed to the guest 1696using the device tree or other means from vm context. 1697 1698The hcall array defines 4 instructions that make up a hypercall. 1699 1700If any additional field gets added to this structure later on, a bit for that 1701additional piece of information will be set in the flags bitmap. 1702 1703The flags bitmap is defined as:: 1704 1705 /* the host supports the ePAPR idle hcall 1706 #define KVM_PPC_PVINFO_FLAGS_EV_IDLE (1<<0) 1707 17084.52 KVM_SET_GSI_ROUTING 1709------------------------ 1710 1711:Capability: KVM_CAP_IRQ_ROUTING 1712:Architectures: x86 s390 arm arm64 1713:Type: vm ioctl 1714:Parameters: struct kvm_irq_routing (in) 1715:Returns: 0 on success, -1 on error 1716 1717Sets the GSI routing table entries, overwriting any previously set entries. 1718 1719On arm/arm64, GSI routing has the following limitation: 1720 1721- GSI routing does not apply to KVM_IRQ_LINE but only to KVM_IRQFD. 1722 1723:: 1724 1725 struct kvm_irq_routing { 1726 __u32 nr; 1727 __u32 flags; 1728 struct kvm_irq_routing_entry entries[0]; 1729 }; 1730 1731No flags are specified so far, the corresponding field must be set to zero. 1732 1733:: 1734 1735 struct kvm_irq_routing_entry { 1736 __u32 gsi; 1737 __u32 type; 1738 __u32 flags; 1739 __u32 pad; 1740 union { 1741 struct kvm_irq_routing_irqchip irqchip; 1742 struct kvm_irq_routing_msi msi; 1743 struct kvm_irq_routing_s390_adapter adapter; 1744 struct kvm_irq_routing_hv_sint hv_sint; 1745 __u32 pad[8]; 1746 } u; 1747 }; 1748 1749 /* gsi routing entry types */ 1750 #define KVM_IRQ_ROUTING_IRQCHIP 1 1751 #define KVM_IRQ_ROUTING_MSI 2 1752 #define KVM_IRQ_ROUTING_S390_ADAPTER 3 1753 #define KVM_IRQ_ROUTING_HV_SINT 4 1754 1755flags: 1756 1757- KVM_MSI_VALID_DEVID: used along with KVM_IRQ_ROUTING_MSI routing entry 1758 type, specifies that the devid field contains a valid value. The per-VM 1759 KVM_CAP_MSI_DEVID capability advertises the requirement to provide 1760 the device ID. If this capability is not available, userspace should 1761 never set the KVM_MSI_VALID_DEVID flag as the ioctl might fail. 1762- zero otherwise 1763 1764:: 1765 1766 struct kvm_irq_routing_irqchip { 1767 __u32 irqchip; 1768 __u32 pin; 1769 }; 1770 1771 struct kvm_irq_routing_msi { 1772 __u32 address_lo; 1773 __u32 address_hi; 1774 __u32 data; 1775 union { 1776 __u32 pad; 1777 __u32 devid; 1778 }; 1779 }; 1780 1781If KVM_MSI_VALID_DEVID is set, devid contains a unique device identifier 1782for the device that wrote the MSI message. For PCI, this is usually a 1783BFD identifier in the lower 16 bits. 1784 1785On x86, address_hi is ignored unless the KVM_X2APIC_API_USE_32BIT_IDS 1786feature of KVM_CAP_X2APIC_API capability is enabled. If it is enabled, 1787address_hi bits 31-8 provide bits 31-8 of the destination id. Bits 7-0 of 1788address_hi must be zero. 1789 1790:: 1791 1792 struct kvm_irq_routing_s390_adapter { 1793 __u64 ind_addr; 1794 __u64 summary_addr; 1795 __u64 ind_offset; 1796 __u32 summary_offset; 1797 __u32 adapter_id; 1798 }; 1799 1800 struct kvm_irq_routing_hv_sint { 1801 __u32 vcpu; 1802 __u32 sint; 1803 }; 1804 1805 18064.55 KVM_SET_TSC_KHZ 1807-------------------- 1808 1809:Capability: KVM_CAP_TSC_CONTROL 1810:Architectures: x86 1811:Type: vcpu ioctl 1812:Parameters: virtual tsc_khz 1813:Returns: 0 on success, -1 on error 1814 1815Specifies the tsc frequency for the virtual machine. The unit of the 1816frequency is KHz. 1817 1818 18194.56 KVM_GET_TSC_KHZ 1820-------------------- 1821 1822:Capability: KVM_CAP_GET_TSC_KHZ 1823:Architectures: x86 1824:Type: vcpu ioctl 1825:Parameters: none 1826:Returns: virtual tsc-khz on success, negative value on error 1827 1828Returns the tsc frequency of the guest. The unit of the return value is 1829KHz. If the host has unstable tsc this ioctl returns -EIO instead as an 1830error. 1831 1832 18334.57 KVM_GET_LAPIC 1834------------------ 1835 1836:Capability: KVM_CAP_IRQCHIP 1837:Architectures: x86 1838:Type: vcpu ioctl 1839:Parameters: struct kvm_lapic_state (out) 1840:Returns: 0 on success, -1 on error 1841 1842:: 1843 1844 #define KVM_APIC_REG_SIZE 0x400 1845 struct kvm_lapic_state { 1846 char regs[KVM_APIC_REG_SIZE]; 1847 }; 1848 1849Reads the Local APIC registers and copies them into the input argument. The 1850data format and layout are the same as documented in the architecture manual. 1851 1852If KVM_X2APIC_API_USE_32BIT_IDS feature of KVM_CAP_X2APIC_API is 1853enabled, then the format of APIC_ID register depends on the APIC mode 1854(reported by MSR_IA32_APICBASE) of its VCPU. x2APIC stores APIC ID in 1855the APIC_ID register (bytes 32-35). xAPIC only allows an 8-bit APIC ID 1856which is stored in bits 31-24 of the APIC register, or equivalently in 1857byte 35 of struct kvm_lapic_state's regs field. KVM_GET_LAPIC must then 1858be called after MSR_IA32_APICBASE has been set with KVM_SET_MSR. 1859 1860If KVM_X2APIC_API_USE_32BIT_IDS feature is disabled, struct kvm_lapic_state 1861always uses xAPIC format. 1862 1863 18644.58 KVM_SET_LAPIC 1865------------------ 1866 1867:Capability: KVM_CAP_IRQCHIP 1868:Architectures: x86 1869:Type: vcpu ioctl 1870:Parameters: struct kvm_lapic_state (in) 1871:Returns: 0 on success, -1 on error 1872 1873:: 1874 1875 #define KVM_APIC_REG_SIZE 0x400 1876 struct kvm_lapic_state { 1877 char regs[KVM_APIC_REG_SIZE]; 1878 }; 1879 1880Copies the input argument into the Local APIC registers. The data format 1881and layout are the same as documented in the architecture manual. 1882 1883The format of the APIC ID register (bytes 32-35 of struct kvm_lapic_state's 1884regs field) depends on the state of the KVM_CAP_X2APIC_API capability. 1885See the note in KVM_GET_LAPIC. 1886 1887 18884.59 KVM_IOEVENTFD 1889------------------ 1890 1891:Capability: KVM_CAP_IOEVENTFD 1892:Architectures: all 1893:Type: vm ioctl 1894:Parameters: struct kvm_ioeventfd (in) 1895:Returns: 0 on success, !0 on error 1896 1897This ioctl attaches or detaches an ioeventfd to a legal pio/mmio address 1898within the guest. A guest write in the registered address will signal the 1899provided event instead of triggering an exit. 1900 1901:: 1902 1903 struct kvm_ioeventfd { 1904 __u64 datamatch; 1905 __u64 addr; /* legal pio/mmio address */ 1906 __u32 len; /* 0, 1, 2, 4, or 8 bytes */ 1907 __s32 fd; 1908 __u32 flags; 1909 __u8 pad[36]; 1910 }; 1911 1912For the special case of virtio-ccw devices on s390, the ioevent is matched 1913to a subchannel/virtqueue tuple instead. 1914 1915The following flags are defined:: 1916 1917 #define KVM_IOEVENTFD_FLAG_DATAMATCH (1 << kvm_ioeventfd_flag_nr_datamatch) 1918 #define KVM_IOEVENTFD_FLAG_PIO (1 << kvm_ioeventfd_flag_nr_pio) 1919 #define KVM_IOEVENTFD_FLAG_DEASSIGN (1 << kvm_ioeventfd_flag_nr_deassign) 1920 #define KVM_IOEVENTFD_FLAG_VIRTIO_CCW_NOTIFY \ 1921 (1 << kvm_ioeventfd_flag_nr_virtio_ccw_notify) 1922 1923If datamatch flag is set, the event will be signaled only if the written value 1924to the registered address is equal to datamatch in struct kvm_ioeventfd. 1925 1926For virtio-ccw devices, addr contains the subchannel id and datamatch the 1927virtqueue index. 1928 1929With KVM_CAP_IOEVENTFD_ANY_LENGTH, a zero length ioeventfd is allowed, and 1930the kernel will ignore the length of guest write and may get a faster vmexit. 1931The speedup may only apply to specific architectures, but the ioeventfd will 1932work anyway. 1933 19344.60 KVM_DIRTY_TLB 1935------------------ 1936 1937:Capability: KVM_CAP_SW_TLB 1938:Architectures: ppc 1939:Type: vcpu ioctl 1940:Parameters: struct kvm_dirty_tlb (in) 1941:Returns: 0 on success, -1 on error 1942 1943:: 1944 1945 struct kvm_dirty_tlb { 1946 __u64 bitmap; 1947 __u32 num_dirty; 1948 }; 1949 1950This must be called whenever userspace has changed an entry in the shared 1951TLB, prior to calling KVM_RUN on the associated vcpu. 1952 1953The "bitmap" field is the userspace address of an array. This array 1954consists of a number of bits, equal to the total number of TLB entries as 1955determined by the last successful call to KVM_CONFIG_TLB, rounded up to the 1956nearest multiple of 64. 1957 1958Each bit corresponds to one TLB entry, ordered the same as in the shared TLB 1959array. 1960 1961The array is little-endian: the bit 0 is the least significant bit of the 1962first byte, bit 8 is the least significant bit of the second byte, etc. 1963This avoids any complications with differing word sizes. 1964 1965The "num_dirty" field is a performance hint for KVM to determine whether it 1966should skip processing the bitmap and just invalidate everything. It must 1967be set to the number of set bits in the bitmap. 1968 1969 19704.62 KVM_CREATE_SPAPR_TCE 1971------------------------- 1972 1973:Capability: KVM_CAP_SPAPR_TCE 1974:Architectures: powerpc 1975:Type: vm ioctl 1976:Parameters: struct kvm_create_spapr_tce (in) 1977:Returns: file descriptor for manipulating the created TCE table 1978 1979This creates a virtual TCE (translation control entry) table, which 1980is an IOMMU for PAPR-style virtual I/O. It is used to translate 1981logical addresses used in virtual I/O into guest physical addresses, 1982and provides a scatter/gather capability for PAPR virtual I/O. 1983 1984:: 1985 1986 /* for KVM_CAP_SPAPR_TCE */ 1987 struct kvm_create_spapr_tce { 1988 __u64 liobn; 1989 __u32 window_size; 1990 }; 1991 1992The liobn field gives the logical IO bus number for which to create a 1993TCE table. The window_size field specifies the size of the DMA window 1994which this TCE table will translate - the table will contain one 64 1995bit TCE entry for every 4kiB of the DMA window. 1996 1997When the guest issues an H_PUT_TCE hcall on a liobn for which a TCE 1998table has been created using this ioctl(), the kernel will handle it 1999in real mode, updating the TCE table. H_PUT_TCE calls for other 2000liobns will cause a vm exit and must be handled by userspace. 2001 2002The return value is a file descriptor which can be passed to mmap(2) 2003to map the created TCE table into userspace. This lets userspace read 2004the entries written by kernel-handled H_PUT_TCE calls, and also lets 2005userspace update the TCE table directly which is useful in some 2006circumstances. 2007 2008 20094.63 KVM_ALLOCATE_RMA 2010--------------------- 2011 2012:Capability: KVM_CAP_PPC_RMA 2013:Architectures: powerpc 2014:Type: vm ioctl 2015:Parameters: struct kvm_allocate_rma (out) 2016:Returns: file descriptor for mapping the allocated RMA 2017 2018This allocates a Real Mode Area (RMA) from the pool allocated at boot 2019time by the kernel. An RMA is a physically-contiguous, aligned region 2020of memory used on older POWER processors to provide the memory which 2021will be accessed by real-mode (MMU off) accesses in a KVM guest. 2022POWER processors support a set of sizes for the RMA that usually 2023includes 64MB, 128MB, 256MB and some larger powers of two. 2024 2025:: 2026 2027 /* for KVM_ALLOCATE_RMA */ 2028 struct kvm_allocate_rma { 2029 __u64 rma_size; 2030 }; 2031 2032The return value is a file descriptor which can be passed to mmap(2) 2033to map the allocated RMA into userspace. The mapped area can then be 2034passed to the KVM_SET_USER_MEMORY_REGION ioctl to establish it as the 2035RMA for a virtual machine. The size of the RMA in bytes (which is 2036fixed at host kernel boot time) is returned in the rma_size field of 2037the argument structure. 2038 2039The KVM_CAP_PPC_RMA capability is 1 or 2 if the KVM_ALLOCATE_RMA ioctl 2040is supported; 2 if the processor requires all virtual machines to have 2041an RMA, or 1 if the processor can use an RMA but doesn't require it, 2042because it supports the Virtual RMA (VRMA) facility. 2043 2044 20454.64 KVM_NMI 2046------------ 2047 2048:Capability: KVM_CAP_USER_NMI 2049:Architectures: x86 2050:Type: vcpu ioctl 2051:Parameters: none 2052:Returns: 0 on success, -1 on error 2053 2054Queues an NMI on the thread's vcpu. Note this is well defined only 2055when KVM_CREATE_IRQCHIP has not been called, since this is an interface 2056between the virtual cpu core and virtual local APIC. After KVM_CREATE_IRQCHIP 2057has been called, this interface is completely emulated within the kernel. 2058 2059To use this to emulate the LINT1 input with KVM_CREATE_IRQCHIP, use the 2060following algorithm: 2061 2062 - pause the vcpu 2063 - read the local APIC's state (KVM_GET_LAPIC) 2064 - check whether changing LINT1 will queue an NMI (see the LVT entry for LINT1) 2065 - if so, issue KVM_NMI 2066 - resume the vcpu 2067 2068Some guests configure the LINT1 NMI input to cause a panic, aiding in 2069debugging. 2070 2071 20724.65 KVM_S390_UCAS_MAP 2073---------------------- 2074 2075:Capability: KVM_CAP_S390_UCONTROL 2076:Architectures: s390 2077:Type: vcpu ioctl 2078:Parameters: struct kvm_s390_ucas_mapping (in) 2079:Returns: 0 in case of success 2080 2081The parameter is defined like this:: 2082 2083 struct kvm_s390_ucas_mapping { 2084 __u64 user_addr; 2085 __u64 vcpu_addr; 2086 __u64 length; 2087 }; 2088 2089This ioctl maps the memory at "user_addr" with the length "length" to 2090the vcpu's address space starting at "vcpu_addr". All parameters need to 2091be aligned by 1 megabyte. 2092 2093 20944.66 KVM_S390_UCAS_UNMAP 2095------------------------ 2096 2097:Capability: KVM_CAP_S390_UCONTROL 2098:Architectures: s390 2099:Type: vcpu ioctl 2100:Parameters: struct kvm_s390_ucas_mapping (in) 2101:Returns: 0 in case of success 2102 2103The parameter is defined like this:: 2104 2105 struct kvm_s390_ucas_mapping { 2106 __u64 user_addr; 2107 __u64 vcpu_addr; 2108 __u64 length; 2109 }; 2110 2111This ioctl unmaps the memory in the vcpu's address space starting at 2112"vcpu_addr" with the length "length". The field "user_addr" is ignored. 2113All parameters need to be aligned by 1 megabyte. 2114 2115 21164.67 KVM_S390_VCPU_FAULT 2117------------------------ 2118 2119:Capability: KVM_CAP_S390_UCONTROL 2120:Architectures: s390 2121:Type: vcpu ioctl 2122:Parameters: vcpu absolute address (in) 2123:Returns: 0 in case of success 2124 2125This call creates a page table entry on the virtual cpu's address space 2126(for user controlled virtual machines) or the virtual machine's address 2127space (for regular virtual machines). This only works for minor faults, 2128thus it's recommended to access subject memory page via the user page 2129table upfront. This is useful to handle validity intercepts for user 2130controlled virtual machines to fault in the virtual cpu's lowcore pages 2131prior to calling the KVM_RUN ioctl. 2132 2133 21344.68 KVM_SET_ONE_REG 2135-------------------- 2136 2137:Capability: KVM_CAP_ONE_REG 2138:Architectures: all 2139:Type: vcpu ioctl 2140:Parameters: struct kvm_one_reg (in) 2141:Returns: 0 on success, negative value on failure 2142 2143Errors: 2144 2145 ====== ============================================================ 2146 ENOENT no such register 2147 EINVAL invalid register ID, or no such register or used with VMs in 2148 protected virtualization mode on s390 2149 EPERM (arm64) register access not allowed before vcpu finalization 2150 ====== ============================================================ 2151 2152(These error codes are indicative only: do not rely on a specific error 2153code being returned in a specific situation.) 2154 2155:: 2156 2157 struct kvm_one_reg { 2158 __u64 id; 2159 __u64 addr; 2160 }; 2161 2162Using this ioctl, a single vcpu register can be set to a specific value 2163defined by user space with the passed in struct kvm_one_reg, where id 2164refers to the register identifier as described below and addr is a pointer 2165to a variable with the respective size. There can be architecture agnostic 2166and architecture specific registers. Each have their own range of operation 2167and their own constants and width. To keep track of the implemented 2168registers, find a list below: 2169 2170 ======= =============================== ============ 2171 Arch Register Width (bits) 2172 ======= =============================== ============ 2173 PPC KVM_REG_PPC_HIOR 64 2174 PPC KVM_REG_PPC_IAC1 64 2175 PPC KVM_REG_PPC_IAC2 64 2176 PPC KVM_REG_PPC_IAC3 64 2177 PPC KVM_REG_PPC_IAC4 64 2178 PPC KVM_REG_PPC_DAC1 64 2179 PPC KVM_REG_PPC_DAC2 64 2180 PPC KVM_REG_PPC_DABR 64 2181 PPC KVM_REG_PPC_DSCR 64 2182 PPC KVM_REG_PPC_PURR 64 2183 PPC KVM_REG_PPC_SPURR 64 2184 PPC KVM_REG_PPC_DAR 64 2185 PPC KVM_REG_PPC_DSISR 32 2186 PPC KVM_REG_PPC_AMR 64 2187 PPC KVM_REG_PPC_UAMOR 64 2188 PPC KVM_REG_PPC_MMCR0 64 2189 PPC KVM_REG_PPC_MMCR1 64 2190 PPC KVM_REG_PPC_MMCRA 64 2191 PPC KVM_REG_PPC_MMCR2 64 2192 PPC KVM_REG_PPC_MMCRS 64 2193 PPC KVM_REG_PPC_MMCR3 64 2194 PPC KVM_REG_PPC_SIAR 64 2195 PPC KVM_REG_PPC_SDAR 64 2196 PPC KVM_REG_PPC_SIER 64 2197 PPC KVM_REG_PPC_SIER2 64 2198 PPC KVM_REG_PPC_SIER3 64 2199 PPC KVM_REG_PPC_PMC1 32 2200 PPC KVM_REG_PPC_PMC2 32 2201 PPC KVM_REG_PPC_PMC3 32 2202 PPC KVM_REG_PPC_PMC4 32 2203 PPC KVM_REG_PPC_PMC5 32 2204 PPC KVM_REG_PPC_PMC6 32 2205 PPC KVM_REG_PPC_PMC7 32 2206 PPC KVM_REG_PPC_PMC8 32 2207 PPC KVM_REG_PPC_FPR0 64 2208 ... 2209 PPC KVM_REG_PPC_FPR31 64 2210 PPC KVM_REG_PPC_VR0 128 2211 ... 2212 PPC KVM_REG_PPC_VR31 128 2213 PPC KVM_REG_PPC_VSR0 128 2214 ... 2215 PPC KVM_REG_PPC_VSR31 128 2216 PPC KVM_REG_PPC_FPSCR 64 2217 PPC KVM_REG_PPC_VSCR 32 2218 PPC KVM_REG_PPC_VPA_ADDR 64 2219 PPC KVM_REG_PPC_VPA_SLB 128 2220 PPC KVM_REG_PPC_VPA_DTL 128 2221 PPC KVM_REG_PPC_EPCR 32 2222 PPC KVM_REG_PPC_EPR 32 2223 PPC KVM_REG_PPC_TCR 32 2224 PPC KVM_REG_PPC_TSR 32 2225 PPC KVM_REG_PPC_OR_TSR 32 2226 PPC KVM_REG_PPC_CLEAR_TSR 32 2227 PPC KVM_REG_PPC_MAS0 32 2228 PPC KVM_REG_PPC_MAS1 32 2229 PPC KVM_REG_PPC_MAS2 64 2230 PPC KVM_REG_PPC_MAS7_3 64 2231 PPC KVM_REG_PPC_MAS4 32 2232 PPC KVM_REG_PPC_MAS6 32 2233 PPC KVM_REG_PPC_MMUCFG 32 2234 PPC KVM_REG_PPC_TLB0CFG 32 2235 PPC KVM_REG_PPC_TLB1CFG 32 2236 PPC KVM_REG_PPC_TLB2CFG 32 2237 PPC KVM_REG_PPC_TLB3CFG 32 2238 PPC KVM_REG_PPC_TLB0PS 32 2239 PPC KVM_REG_PPC_TLB1PS 32 2240 PPC KVM_REG_PPC_TLB2PS 32 2241 PPC KVM_REG_PPC_TLB3PS 32 2242 PPC KVM_REG_PPC_EPTCFG 32 2243 PPC KVM_REG_PPC_ICP_STATE 64 2244 PPC KVM_REG_PPC_VP_STATE 128 2245 PPC KVM_REG_PPC_TB_OFFSET 64 2246 PPC KVM_REG_PPC_SPMC1 32 2247 PPC KVM_REG_PPC_SPMC2 32 2248 PPC KVM_REG_PPC_IAMR 64 2249 PPC KVM_REG_PPC_TFHAR 64 2250 PPC KVM_REG_PPC_TFIAR 64 2251 PPC KVM_REG_PPC_TEXASR 64 2252 PPC KVM_REG_PPC_FSCR 64 2253 PPC KVM_REG_PPC_PSPB 32 2254 PPC KVM_REG_PPC_EBBHR 64 2255 PPC KVM_REG_PPC_EBBRR 64 2256 PPC KVM_REG_PPC_BESCR 64 2257 PPC KVM_REG_PPC_TAR 64 2258 PPC KVM_REG_PPC_DPDES 64 2259 PPC KVM_REG_PPC_DAWR 64 2260 PPC KVM_REG_PPC_DAWRX 64 2261 PPC KVM_REG_PPC_CIABR 64 2262 PPC KVM_REG_PPC_IC 64 2263 PPC KVM_REG_PPC_VTB 64 2264 PPC KVM_REG_PPC_CSIGR 64 2265 PPC KVM_REG_PPC_TACR 64 2266 PPC KVM_REG_PPC_TCSCR 64 2267 PPC KVM_REG_PPC_PID 64 2268 PPC KVM_REG_PPC_ACOP 64 2269 PPC KVM_REG_PPC_VRSAVE 32 2270 PPC KVM_REG_PPC_LPCR 32 2271 PPC KVM_REG_PPC_LPCR_64 64 2272 PPC KVM_REG_PPC_PPR 64 2273 PPC KVM_REG_PPC_ARCH_COMPAT 32 2274 PPC KVM_REG_PPC_DABRX 32 2275 PPC KVM_REG_PPC_WORT 64 2276 PPC KVM_REG_PPC_SPRG9 64 2277 PPC KVM_REG_PPC_DBSR 32 2278 PPC KVM_REG_PPC_TIDR 64 2279 PPC KVM_REG_PPC_PSSCR 64 2280 PPC KVM_REG_PPC_DEC_EXPIRY 64 2281 PPC KVM_REG_PPC_PTCR 64 2282 PPC KVM_REG_PPC_DAWR1 64 2283 PPC KVM_REG_PPC_DAWRX1 64 2284 PPC KVM_REG_PPC_TM_GPR0 64 2285 ... 2286 PPC KVM_REG_PPC_TM_GPR31 64 2287 PPC KVM_REG_PPC_TM_VSR0 128 2288 ... 2289 PPC KVM_REG_PPC_TM_VSR63 128 2290 PPC KVM_REG_PPC_TM_CR 64 2291 PPC KVM_REG_PPC_TM_LR 64 2292 PPC KVM_REG_PPC_TM_CTR 64 2293 PPC KVM_REG_PPC_TM_FPSCR 64 2294 PPC KVM_REG_PPC_TM_AMR 64 2295 PPC KVM_REG_PPC_TM_PPR 64 2296 PPC KVM_REG_PPC_TM_VRSAVE 64 2297 PPC KVM_REG_PPC_TM_VSCR 32 2298 PPC KVM_REG_PPC_TM_DSCR 64 2299 PPC KVM_REG_PPC_TM_TAR 64 2300 PPC KVM_REG_PPC_TM_XER 64 2301 2302 MIPS KVM_REG_MIPS_R0 64 2303 ... 2304 MIPS KVM_REG_MIPS_R31 64 2305 MIPS KVM_REG_MIPS_HI 64 2306 MIPS KVM_REG_MIPS_LO 64 2307 MIPS KVM_REG_MIPS_PC 64 2308 MIPS KVM_REG_MIPS_CP0_INDEX 32 2309 MIPS KVM_REG_MIPS_CP0_ENTRYLO0 64 2310 MIPS KVM_REG_MIPS_CP0_ENTRYLO1 64 2311 MIPS KVM_REG_MIPS_CP0_CONTEXT 64 2312 MIPS KVM_REG_MIPS_CP0_CONTEXTCONFIG 32 2313 MIPS KVM_REG_MIPS_CP0_USERLOCAL 64 2314 MIPS KVM_REG_MIPS_CP0_XCONTEXTCONFIG 64 2315 MIPS KVM_REG_MIPS_CP0_PAGEMASK 32 2316 MIPS KVM_REG_MIPS_CP0_PAGEGRAIN 32 2317 MIPS KVM_REG_MIPS_CP0_SEGCTL0 64 2318 MIPS KVM_REG_MIPS_CP0_SEGCTL1 64 2319 MIPS KVM_REG_MIPS_CP0_SEGCTL2 64 2320 MIPS KVM_REG_MIPS_CP0_PWBASE 64 2321 MIPS KVM_REG_MIPS_CP0_PWFIELD 64 2322 MIPS KVM_REG_MIPS_CP0_PWSIZE 64 2323 MIPS KVM_REG_MIPS_CP0_WIRED 32 2324 MIPS KVM_REG_MIPS_CP0_PWCTL 32 2325 MIPS KVM_REG_MIPS_CP0_HWRENA 32 2326 MIPS KVM_REG_MIPS_CP0_BADVADDR 64 2327 MIPS KVM_REG_MIPS_CP0_BADINSTR 32 2328 MIPS KVM_REG_MIPS_CP0_BADINSTRP 32 2329 MIPS KVM_REG_MIPS_CP0_COUNT 32 2330 MIPS KVM_REG_MIPS_CP0_ENTRYHI 64 2331 MIPS KVM_REG_MIPS_CP0_COMPARE 32 2332 MIPS KVM_REG_MIPS_CP0_STATUS 32 2333 MIPS KVM_REG_MIPS_CP0_INTCTL 32 2334 MIPS KVM_REG_MIPS_CP0_CAUSE 32 2335 MIPS KVM_REG_MIPS_CP0_EPC 64 2336 MIPS KVM_REG_MIPS_CP0_PRID 32 2337 MIPS KVM_REG_MIPS_CP0_EBASE 64 2338 MIPS KVM_REG_MIPS_CP0_CONFIG 32 2339 MIPS KVM_REG_MIPS_CP0_CONFIG1 32 2340 MIPS KVM_REG_MIPS_CP0_CONFIG2 32 2341 MIPS KVM_REG_MIPS_CP0_CONFIG3 32 2342 MIPS KVM_REG_MIPS_CP0_CONFIG4 32 2343 MIPS KVM_REG_MIPS_CP0_CONFIG5 32 2344 MIPS KVM_REG_MIPS_CP0_CONFIG7 32 2345 MIPS KVM_REG_MIPS_CP0_XCONTEXT 64 2346 MIPS KVM_REG_MIPS_CP0_ERROREPC 64 2347 MIPS KVM_REG_MIPS_CP0_KSCRATCH1 64 2348 MIPS KVM_REG_MIPS_CP0_KSCRATCH2 64 2349 MIPS KVM_REG_MIPS_CP0_KSCRATCH3 64 2350 MIPS KVM_REG_MIPS_CP0_KSCRATCH4 64 2351 MIPS KVM_REG_MIPS_CP0_KSCRATCH5 64 2352 MIPS KVM_REG_MIPS_CP0_KSCRATCH6 64 2353 MIPS KVM_REG_MIPS_CP0_MAAR(0..63) 64 2354 MIPS KVM_REG_MIPS_COUNT_CTL 64 2355 MIPS KVM_REG_MIPS_COUNT_RESUME 64 2356 MIPS KVM_REG_MIPS_COUNT_HZ 64 2357 MIPS KVM_REG_MIPS_FPR_32(0..31) 32 2358 MIPS KVM_REG_MIPS_FPR_64(0..31) 64 2359 MIPS KVM_REG_MIPS_VEC_128(0..31) 128 2360 MIPS KVM_REG_MIPS_FCR_IR 32 2361 MIPS KVM_REG_MIPS_FCR_CSR 32 2362 MIPS KVM_REG_MIPS_MSA_IR 32 2363 MIPS KVM_REG_MIPS_MSA_CSR 32 2364 ======= =============================== ============ 2365 2366ARM registers are mapped using the lower 32 bits. The upper 16 of that 2367is the register group type, or coprocessor number: 2368 2369ARM core registers have the following id bit patterns:: 2370 2371 0x4020 0000 0010 <index into the kvm_regs struct:16> 2372 2373ARM 32-bit CP15 registers have the following id bit patterns:: 2374 2375 0x4020 0000 000F <zero:1> <crn:4> <crm:4> <opc1:4> <opc2:3> 2376 2377ARM 64-bit CP15 registers have the following id bit patterns:: 2378 2379 0x4030 0000 000F <zero:1> <zero:4> <crm:4> <opc1:4> <zero:3> 2380 2381ARM CCSIDR registers are demultiplexed by CSSELR value:: 2382 2383 0x4020 0000 0011 00 <csselr:8> 2384 2385ARM 32-bit VFP control registers have the following id bit patterns:: 2386 2387 0x4020 0000 0012 1 <regno:12> 2388 2389ARM 64-bit FP registers have the following id bit patterns:: 2390 2391 0x4030 0000 0012 0 <regno:12> 2392 2393ARM firmware pseudo-registers have the following bit pattern:: 2394 2395 0x4030 0000 0014 <regno:16> 2396 2397 2398arm64 registers are mapped using the lower 32 bits. The upper 16 of 2399that is the register group type, or coprocessor number: 2400 2401arm64 core/FP-SIMD registers have the following id bit patterns. Note 2402that the size of the access is variable, as the kvm_regs structure 2403contains elements ranging from 32 to 128 bits. The index is a 32bit 2404value in the kvm_regs structure seen as a 32bit array:: 2405 2406 0x60x0 0000 0010 <index into the kvm_regs struct:16> 2407 2408Specifically: 2409 2410======================= ========= ===== ======================================= 2411 Encoding Register Bits kvm_regs member 2412======================= ========= ===== ======================================= 2413 0x6030 0000 0010 0000 X0 64 regs.regs[0] 2414 0x6030 0000 0010 0002 X1 64 regs.regs[1] 2415 ... 2416 0x6030 0000 0010 003c X30 64 regs.regs[30] 2417 0x6030 0000 0010 003e SP 64 regs.sp 2418 0x6030 0000 0010 0040 PC 64 regs.pc 2419 0x6030 0000 0010 0042 PSTATE 64 regs.pstate 2420 0x6030 0000 0010 0044 SP_EL1 64 sp_el1 2421 0x6030 0000 0010 0046 ELR_EL1 64 elr_el1 2422 0x6030 0000 0010 0048 SPSR_EL1 64 spsr[KVM_SPSR_EL1] (alias SPSR_SVC) 2423 0x6030 0000 0010 004a SPSR_ABT 64 spsr[KVM_SPSR_ABT] 2424 0x6030 0000 0010 004c SPSR_UND 64 spsr[KVM_SPSR_UND] 2425 0x6030 0000 0010 004e SPSR_IRQ 64 spsr[KVM_SPSR_IRQ] 2426 0x6060 0000 0010 0050 SPSR_FIQ 64 spsr[KVM_SPSR_FIQ] 2427 0x6040 0000 0010 0054 V0 128 fp_regs.vregs[0] [1]_ 2428 0x6040 0000 0010 0058 V1 128 fp_regs.vregs[1] [1]_ 2429 ... 2430 0x6040 0000 0010 00d0 V31 128 fp_regs.vregs[31] [1]_ 2431 0x6020 0000 0010 00d4 FPSR 32 fp_regs.fpsr 2432 0x6020 0000 0010 00d5 FPCR 32 fp_regs.fpcr 2433======================= ========= ===== ======================================= 2434 2435.. [1] These encodings are not accepted for SVE-enabled vcpus. See 2436 KVM_ARM_VCPU_INIT. 2437 2438 The equivalent register content can be accessed via bits [127:0] of 2439 the corresponding SVE Zn registers instead for vcpus that have SVE 2440 enabled (see below). 2441 2442arm64 CCSIDR registers are demultiplexed by CSSELR value:: 2443 2444 0x6020 0000 0011 00 <csselr:8> 2445 2446arm64 system registers have the following id bit patterns:: 2447 2448 0x6030 0000 0013 <op0:2> <op1:3> <crn:4> <crm:4> <op2:3> 2449 2450.. warning:: 2451 2452 Two system register IDs do not follow the specified pattern. These 2453 are KVM_REG_ARM_TIMER_CVAL and KVM_REG_ARM_TIMER_CNT, which map to 2454 system registers CNTV_CVAL_EL0 and CNTVCT_EL0 respectively. These 2455 two had their values accidentally swapped, which means TIMER_CVAL is 2456 derived from the register encoding for CNTVCT_EL0 and TIMER_CNT is 2457 derived from the register encoding for CNTV_CVAL_EL0. As this is 2458 API, it must remain this way. 2459 2460arm64 firmware pseudo-registers have the following bit pattern:: 2461 2462 0x6030 0000 0014 <regno:16> 2463 2464arm64 SVE registers have the following bit patterns:: 2465 2466 0x6080 0000 0015 00 <n:5> <slice:5> Zn bits[2048*slice + 2047 : 2048*slice] 2467 0x6050 0000 0015 04 <n:4> <slice:5> Pn bits[256*slice + 255 : 256*slice] 2468 0x6050 0000 0015 060 <slice:5> FFR bits[256*slice + 255 : 256*slice] 2469 0x6060 0000 0015 ffff KVM_REG_ARM64_SVE_VLS pseudo-register 2470 2471Access to register IDs where 2048 * slice >= 128 * max_vq will fail with 2472ENOENT. max_vq is the vcpu's maximum supported vector length in 128-bit 2473quadwords: see [2]_ below. 2474 2475These registers are only accessible on vcpus for which SVE is enabled. 2476See KVM_ARM_VCPU_INIT for details. 2477 2478In addition, except for KVM_REG_ARM64_SVE_VLS, these registers are not 2479accessible until the vcpu's SVE configuration has been finalized 2480using KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE). See KVM_ARM_VCPU_INIT 2481and KVM_ARM_VCPU_FINALIZE for more information about this procedure. 2482 2483KVM_REG_ARM64_SVE_VLS is a pseudo-register that allows the set of vector 2484lengths supported by the vcpu to be discovered and configured by 2485userspace. When transferred to or from user memory via KVM_GET_ONE_REG 2486or KVM_SET_ONE_REG, the value of this register is of type 2487__u64[KVM_ARM64_SVE_VLS_WORDS], and encodes the set of vector lengths as 2488follows:: 2489 2490 __u64 vector_lengths[KVM_ARM64_SVE_VLS_WORDS]; 2491 2492 if (vq >= SVE_VQ_MIN && vq <= SVE_VQ_MAX && 2493 ((vector_lengths[(vq - KVM_ARM64_SVE_VQ_MIN) / 64] >> 2494 ((vq - KVM_ARM64_SVE_VQ_MIN) % 64)) & 1)) 2495 /* Vector length vq * 16 bytes supported */ 2496 else 2497 /* Vector length vq * 16 bytes not supported */ 2498 2499.. [2] The maximum value vq for which the above condition is true is 2500 max_vq. This is the maximum vector length available to the guest on 2501 this vcpu, and determines which register slices are visible through 2502 this ioctl interface. 2503 2504(See Documentation/arm64/sve.rst for an explanation of the "vq" 2505nomenclature.) 2506 2507KVM_REG_ARM64_SVE_VLS is only accessible after KVM_ARM_VCPU_INIT. 2508KVM_ARM_VCPU_INIT initialises it to the best set of vector lengths that 2509the host supports. 2510 2511Userspace may subsequently modify it if desired until the vcpu's SVE 2512configuration is finalized using KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE). 2513 2514Apart from simply removing all vector lengths from the host set that 2515exceed some value, support for arbitrarily chosen sets of vector lengths 2516is hardware-dependent and may not be available. Attempting to configure 2517an invalid set of vector lengths via KVM_SET_ONE_REG will fail with 2518EINVAL. 2519 2520After the vcpu's SVE configuration is finalized, further attempts to 2521write this register will fail with EPERM. 2522 2523 2524MIPS registers are mapped using the lower 32 bits. The upper 16 of that is 2525the register group type: 2526 2527MIPS core registers (see above) have the following id bit patterns:: 2528 2529 0x7030 0000 0000 <reg:16> 2530 2531MIPS CP0 registers (see KVM_REG_MIPS_CP0_* above) have the following id bit 2532patterns depending on whether they're 32-bit or 64-bit registers:: 2533 2534 0x7020 0000 0001 00 <reg:5> <sel:3> (32-bit) 2535 0x7030 0000 0001 00 <reg:5> <sel:3> (64-bit) 2536 2537Note: KVM_REG_MIPS_CP0_ENTRYLO0 and KVM_REG_MIPS_CP0_ENTRYLO1 are the MIPS64 2538versions of the EntryLo registers regardless of the word size of the host 2539hardware, host kernel, guest, and whether XPA is present in the guest, i.e. 2540with the RI and XI bits (if they exist) in bits 63 and 62 respectively, and 2541the PFNX field starting at bit 30. 2542 2543MIPS MAARs (see KVM_REG_MIPS_CP0_MAAR(*) above) have the following id bit 2544patterns:: 2545 2546 0x7030 0000 0001 01 <reg:8> 2547 2548MIPS KVM control registers (see above) have the following id bit patterns:: 2549 2550 0x7030 0000 0002 <reg:16> 2551 2552MIPS FPU registers (see KVM_REG_MIPS_FPR_{32,64}() above) have the following 2553id bit patterns depending on the size of the register being accessed. They are 2554always accessed according to the current guest FPU mode (Status.FR and 2555Config5.FRE), i.e. as the guest would see them, and they become unpredictable 2556if the guest FPU mode is changed. MIPS SIMD Architecture (MSA) vector 2557registers (see KVM_REG_MIPS_VEC_128() above) have similar patterns as they 2558overlap the FPU registers:: 2559 2560 0x7020 0000 0003 00 <0:3> <reg:5> (32-bit FPU registers) 2561 0x7030 0000 0003 00 <0:3> <reg:5> (64-bit FPU registers) 2562 0x7040 0000 0003 00 <0:3> <reg:5> (128-bit MSA vector registers) 2563 2564MIPS FPU control registers (see KVM_REG_MIPS_FCR_{IR,CSR} above) have the 2565following id bit patterns:: 2566 2567 0x7020 0000 0003 01 <0:3> <reg:5> 2568 2569MIPS MSA control registers (see KVM_REG_MIPS_MSA_{IR,CSR} above) have the 2570following id bit patterns:: 2571 2572 0x7020 0000 0003 02 <0:3> <reg:5> 2573 2574 25754.69 KVM_GET_ONE_REG 2576-------------------- 2577 2578:Capability: KVM_CAP_ONE_REG 2579:Architectures: all 2580:Type: vcpu ioctl 2581:Parameters: struct kvm_one_reg (in and out) 2582:Returns: 0 on success, negative value on failure 2583 2584Errors include: 2585 2586 ======== ============================================================ 2587 ENOENT no such register 2588 EINVAL invalid register ID, or no such register or used with VMs in 2589 protected virtualization mode on s390 2590 EPERM (arm64) register access not allowed before vcpu finalization 2591 ======== ============================================================ 2592 2593(These error codes are indicative only: do not rely on a specific error 2594code being returned in a specific situation.) 2595 2596This ioctl allows to receive the value of a single register implemented 2597in a vcpu. The register to read is indicated by the "id" field of the 2598kvm_one_reg struct passed in. On success, the register value can be found 2599at the memory location pointed to by "addr". 2600 2601The list of registers accessible using this interface is identical to the 2602list in 4.68. 2603 2604 26054.70 KVM_KVMCLOCK_CTRL 2606---------------------- 2607 2608:Capability: KVM_CAP_KVMCLOCK_CTRL 2609:Architectures: Any that implement pvclocks (currently x86 only) 2610:Type: vcpu ioctl 2611:Parameters: None 2612:Returns: 0 on success, -1 on error 2613 2614This ioctl sets a flag accessible to the guest indicating that the specified 2615vCPU has been paused by the host userspace. 2616 2617The host will set a flag in the pvclock structure that is checked from the 2618soft lockup watchdog. The flag is part of the pvclock structure that is 2619shared between guest and host, specifically the second bit of the flags 2620field of the pvclock_vcpu_time_info structure. It will be set exclusively by 2621the host and read/cleared exclusively by the guest. The guest operation of 2622checking and clearing the flag must be an atomic operation so 2623load-link/store-conditional, or equivalent must be used. There are two cases 2624where the guest will clear the flag: when the soft lockup watchdog timer resets 2625itself or when a soft lockup is detected. This ioctl can be called any time 2626after pausing the vcpu, but before it is resumed. 2627 2628 26294.71 KVM_SIGNAL_MSI 2630------------------- 2631 2632:Capability: KVM_CAP_SIGNAL_MSI 2633:Architectures: x86 arm arm64 2634:Type: vm ioctl 2635:Parameters: struct kvm_msi (in) 2636:Returns: >0 on delivery, 0 if guest blocked the MSI, and -1 on error 2637 2638Directly inject a MSI message. Only valid with in-kernel irqchip that handles 2639MSI messages. 2640 2641:: 2642 2643 struct kvm_msi { 2644 __u32 address_lo; 2645 __u32 address_hi; 2646 __u32 data; 2647 __u32 flags; 2648 __u32 devid; 2649 __u8 pad[12]; 2650 }; 2651 2652flags: 2653 KVM_MSI_VALID_DEVID: devid contains a valid value. The per-VM 2654 KVM_CAP_MSI_DEVID capability advertises the requirement to provide 2655 the device ID. If this capability is not available, userspace 2656 should never set the KVM_MSI_VALID_DEVID flag as the ioctl might fail. 2657 2658If KVM_MSI_VALID_DEVID is set, devid contains a unique device identifier 2659for the device that wrote the MSI message. For PCI, this is usually a 2660BFD identifier in the lower 16 bits. 2661 2662On x86, address_hi is ignored unless the KVM_X2APIC_API_USE_32BIT_IDS 2663feature of KVM_CAP_X2APIC_API capability is enabled. If it is enabled, 2664address_hi bits 31-8 provide bits 31-8 of the destination id. Bits 7-0 of 2665address_hi must be zero. 2666 2667 26684.71 KVM_CREATE_PIT2 2669-------------------- 2670 2671:Capability: KVM_CAP_PIT2 2672:Architectures: x86 2673:Type: vm ioctl 2674:Parameters: struct kvm_pit_config (in) 2675:Returns: 0 on success, -1 on error 2676 2677Creates an in-kernel device model for the i8254 PIT. This call is only valid 2678after enabling in-kernel irqchip support via KVM_CREATE_IRQCHIP. The following 2679parameters have to be passed:: 2680 2681 struct kvm_pit_config { 2682 __u32 flags; 2683 __u32 pad[15]; 2684 }; 2685 2686Valid flags are:: 2687 2688 #define KVM_PIT_SPEAKER_DUMMY 1 /* emulate speaker port stub */ 2689 2690PIT timer interrupts may use a per-VM kernel thread for injection. If it 2691exists, this thread will have a name of the following pattern:: 2692 2693 kvm-pit/<owner-process-pid> 2694 2695When running a guest with elevated priorities, the scheduling parameters of 2696this thread may have to be adjusted accordingly. 2697 2698This IOCTL replaces the obsolete KVM_CREATE_PIT. 2699 2700 27014.72 KVM_GET_PIT2 2702----------------- 2703 2704:Capability: KVM_CAP_PIT_STATE2 2705:Architectures: x86 2706:Type: vm ioctl 2707:Parameters: struct kvm_pit_state2 (out) 2708:Returns: 0 on success, -1 on error 2709 2710Retrieves the state of the in-kernel PIT model. Only valid after 2711KVM_CREATE_PIT2. The state is returned in the following structure:: 2712 2713 struct kvm_pit_state2 { 2714 struct kvm_pit_channel_state channels[3]; 2715 __u32 flags; 2716 __u32 reserved[9]; 2717 }; 2718 2719Valid flags are:: 2720 2721 /* disable PIT in HPET legacy mode */ 2722 #define KVM_PIT_FLAGS_HPET_LEGACY 0x00000001 2723 2724This IOCTL replaces the obsolete KVM_GET_PIT. 2725 2726 27274.73 KVM_SET_PIT2 2728----------------- 2729 2730:Capability: KVM_CAP_PIT_STATE2 2731:Architectures: x86 2732:Type: vm ioctl 2733:Parameters: struct kvm_pit_state2 (in) 2734:Returns: 0 on success, -1 on error 2735 2736Sets the state of the in-kernel PIT model. Only valid after KVM_CREATE_PIT2. 2737See KVM_GET_PIT2 for details on struct kvm_pit_state2. 2738 2739This IOCTL replaces the obsolete KVM_SET_PIT. 2740 2741 27424.74 KVM_PPC_GET_SMMU_INFO 2743-------------------------- 2744 2745:Capability: KVM_CAP_PPC_GET_SMMU_INFO 2746:Architectures: powerpc 2747:Type: vm ioctl 2748:Parameters: None 2749:Returns: 0 on success, -1 on error 2750 2751This populates and returns a structure describing the features of 2752the "Server" class MMU emulation supported by KVM. 2753This can in turn be used by userspace to generate the appropriate 2754device-tree properties for the guest operating system. 2755 2756The structure contains some global information, followed by an 2757array of supported segment page sizes:: 2758 2759 struct kvm_ppc_smmu_info { 2760 __u64 flags; 2761 __u32 slb_size; 2762 __u32 pad; 2763 struct kvm_ppc_one_seg_page_size sps[KVM_PPC_PAGE_SIZES_MAX_SZ]; 2764 }; 2765 2766The supported flags are: 2767 2768 - KVM_PPC_PAGE_SIZES_REAL: 2769 When that flag is set, guest page sizes must "fit" the backing 2770 store page sizes. When not set, any page size in the list can 2771 be used regardless of how they are backed by userspace. 2772 2773 - KVM_PPC_1T_SEGMENTS 2774 The emulated MMU supports 1T segments in addition to the 2775 standard 256M ones. 2776 2777 - KVM_PPC_NO_HASH 2778 This flag indicates that HPT guests are not supported by KVM, 2779 thus all guests must use radix MMU mode. 2780 2781The "slb_size" field indicates how many SLB entries are supported 2782 2783The "sps" array contains 8 entries indicating the supported base 2784page sizes for a segment in increasing order. Each entry is defined 2785as follow:: 2786 2787 struct kvm_ppc_one_seg_page_size { 2788 __u32 page_shift; /* Base page shift of segment (or 0) */ 2789 __u32 slb_enc; /* SLB encoding for BookS */ 2790 struct kvm_ppc_one_page_size enc[KVM_PPC_PAGE_SIZES_MAX_SZ]; 2791 }; 2792 2793An entry with a "page_shift" of 0 is unused. Because the array is 2794organized in increasing order, a lookup can stop when encoutering 2795such an entry. 2796 2797The "slb_enc" field provides the encoding to use in the SLB for the 2798page size. The bits are in positions such as the value can directly 2799be OR'ed into the "vsid" argument of the slbmte instruction. 2800 2801The "enc" array is a list which for each of those segment base page 2802size provides the list of supported actual page sizes (which can be 2803only larger or equal to the base page size), along with the 2804corresponding encoding in the hash PTE. Similarly, the array is 28058 entries sorted by increasing sizes and an entry with a "0" shift 2806is an empty entry and a terminator:: 2807 2808 struct kvm_ppc_one_page_size { 2809 __u32 page_shift; /* Page shift (or 0) */ 2810 __u32 pte_enc; /* Encoding in the HPTE (>>12) */ 2811 }; 2812 2813The "pte_enc" field provides a value that can OR'ed into the hash 2814PTE's RPN field (ie, it needs to be shifted left by 12 to OR it 2815into the hash PTE second double word). 2816 28174.75 KVM_IRQFD 2818-------------- 2819 2820:Capability: KVM_CAP_IRQFD 2821:Architectures: x86 s390 arm arm64 2822:Type: vm ioctl 2823:Parameters: struct kvm_irqfd (in) 2824:Returns: 0 on success, -1 on error 2825 2826Allows setting an eventfd to directly trigger a guest interrupt. 2827kvm_irqfd.fd specifies the file descriptor to use as the eventfd and 2828kvm_irqfd.gsi specifies the irqchip pin toggled by this event. When 2829an event is triggered on the eventfd, an interrupt is injected into 2830the guest using the specified gsi pin. The irqfd is removed using 2831the KVM_IRQFD_FLAG_DEASSIGN flag, specifying both kvm_irqfd.fd 2832and kvm_irqfd.gsi. 2833 2834With KVM_CAP_IRQFD_RESAMPLE, KVM_IRQFD supports a de-assert and notify 2835mechanism allowing emulation of level-triggered, irqfd-based 2836interrupts. When KVM_IRQFD_FLAG_RESAMPLE is set the user must pass an 2837additional eventfd in the kvm_irqfd.resamplefd field. When operating 2838in resample mode, posting of an interrupt through kvm_irq.fd asserts 2839the specified gsi in the irqchip. When the irqchip is resampled, such 2840as from an EOI, the gsi is de-asserted and the user is notified via 2841kvm_irqfd.resamplefd. It is the user's responsibility to re-queue 2842the interrupt if the device making use of it still requires service. 2843Note that closing the resamplefd is not sufficient to disable the 2844irqfd. The KVM_IRQFD_FLAG_RESAMPLE is only necessary on assignment 2845and need not be specified with KVM_IRQFD_FLAG_DEASSIGN. 2846 2847On arm/arm64, gsi routing being supported, the following can happen: 2848 2849- in case no routing entry is associated to this gsi, injection fails 2850- in case the gsi is associated to an irqchip routing entry, 2851 irqchip.pin + 32 corresponds to the injected SPI ID. 2852- in case the gsi is associated to an MSI routing entry, the MSI 2853 message and device ID are translated into an LPI (support restricted 2854 to GICv3 ITS in-kernel emulation). 2855 28564.76 KVM_PPC_ALLOCATE_HTAB 2857-------------------------- 2858 2859:Capability: KVM_CAP_PPC_ALLOC_HTAB 2860:Architectures: powerpc 2861:Type: vm ioctl 2862:Parameters: Pointer to u32 containing hash table order (in/out) 2863:Returns: 0 on success, -1 on error 2864 2865This requests the host kernel to allocate an MMU hash table for a 2866guest using the PAPR paravirtualization interface. This only does 2867anything if the kernel is configured to use the Book 3S HV style of 2868virtualization. Otherwise the capability doesn't exist and the ioctl 2869returns an ENOTTY error. The rest of this description assumes Book 3S 2870HV. 2871 2872There must be no vcpus running when this ioctl is called; if there 2873are, it will do nothing and return an EBUSY error. 2874 2875The parameter is a pointer to a 32-bit unsigned integer variable 2876containing the order (log base 2) of the desired size of the hash 2877table, which must be between 18 and 46. On successful return from the 2878ioctl, the value will not be changed by the kernel. 2879 2880If no hash table has been allocated when any vcpu is asked to run 2881(with the KVM_RUN ioctl), the host kernel will allocate a 2882default-sized hash table (16 MB). 2883 2884If this ioctl is called when a hash table has already been allocated, 2885with a different order from the existing hash table, the existing hash 2886table will be freed and a new one allocated. If this is ioctl is 2887called when a hash table has already been allocated of the same order 2888as specified, the kernel will clear out the existing hash table (zero 2889all HPTEs). In either case, if the guest is using the virtualized 2890real-mode area (VRMA) facility, the kernel will re-create the VMRA 2891HPTEs on the next KVM_RUN of any vcpu. 2892 28934.77 KVM_S390_INTERRUPT 2894----------------------- 2895 2896:Capability: basic 2897:Architectures: s390 2898:Type: vm ioctl, vcpu ioctl 2899:Parameters: struct kvm_s390_interrupt (in) 2900:Returns: 0 on success, -1 on error 2901 2902Allows to inject an interrupt to the guest. Interrupts can be floating 2903(vm ioctl) or per cpu (vcpu ioctl), depending on the interrupt type. 2904 2905Interrupt parameters are passed via kvm_s390_interrupt:: 2906 2907 struct kvm_s390_interrupt { 2908 __u32 type; 2909 __u32 parm; 2910 __u64 parm64; 2911 }; 2912 2913type can be one of the following: 2914 2915KVM_S390_SIGP_STOP (vcpu) 2916 - sigp stop; optional flags in parm 2917KVM_S390_PROGRAM_INT (vcpu) 2918 - program check; code in parm 2919KVM_S390_SIGP_SET_PREFIX (vcpu) 2920 - sigp set prefix; prefix address in parm 2921KVM_S390_RESTART (vcpu) 2922 - restart 2923KVM_S390_INT_CLOCK_COMP (vcpu) 2924 - clock comparator interrupt 2925KVM_S390_INT_CPU_TIMER (vcpu) 2926 - CPU timer interrupt 2927KVM_S390_INT_VIRTIO (vm) 2928 - virtio external interrupt; external interrupt 2929 parameters in parm and parm64 2930KVM_S390_INT_SERVICE (vm) 2931 - sclp external interrupt; sclp parameter in parm 2932KVM_S390_INT_EMERGENCY (vcpu) 2933 - sigp emergency; source cpu in parm 2934KVM_S390_INT_EXTERNAL_CALL (vcpu) 2935 - sigp external call; source cpu in parm 2936KVM_S390_INT_IO(ai,cssid,ssid,schid) (vm) 2937 - compound value to indicate an 2938 I/O interrupt (ai - adapter interrupt; cssid,ssid,schid - subchannel); 2939 I/O interruption parameters in parm (subchannel) and parm64 (intparm, 2940 interruption subclass) 2941KVM_S390_MCHK (vm, vcpu) 2942 - machine check interrupt; cr 14 bits in parm, machine check interrupt 2943 code in parm64 (note that machine checks needing further payload are not 2944 supported by this ioctl) 2945 2946This is an asynchronous vcpu ioctl and can be invoked from any thread. 2947 29484.78 KVM_PPC_GET_HTAB_FD 2949------------------------ 2950 2951:Capability: KVM_CAP_PPC_HTAB_FD 2952:Architectures: powerpc 2953:Type: vm ioctl 2954:Parameters: Pointer to struct kvm_get_htab_fd (in) 2955:Returns: file descriptor number (>= 0) on success, -1 on error 2956 2957This returns a file descriptor that can be used either to read out the 2958entries in the guest's hashed page table (HPT), or to write entries to 2959initialize the HPT. The returned fd can only be written to if the 2960KVM_GET_HTAB_WRITE bit is set in the flags field of the argument, and 2961can only be read if that bit is clear. The argument struct looks like 2962this:: 2963 2964 /* For KVM_PPC_GET_HTAB_FD */ 2965 struct kvm_get_htab_fd { 2966 __u64 flags; 2967 __u64 start_index; 2968 __u64 reserved[2]; 2969 }; 2970 2971 /* Values for kvm_get_htab_fd.flags */ 2972 #define KVM_GET_HTAB_BOLTED_ONLY ((__u64)0x1) 2973 #define KVM_GET_HTAB_WRITE ((__u64)0x2) 2974 2975The 'start_index' field gives the index in the HPT of the entry at 2976which to start reading. It is ignored when writing. 2977 2978Reads on the fd will initially supply information about all 2979"interesting" HPT entries. Interesting entries are those with the 2980bolted bit set, if the KVM_GET_HTAB_BOLTED_ONLY bit is set, otherwise 2981all entries. When the end of the HPT is reached, the read() will 2982return. If read() is called again on the fd, it will start again from 2983the beginning of the HPT, but will only return HPT entries that have 2984changed since they were last read. 2985 2986Data read or written is structured as a header (8 bytes) followed by a 2987series of valid HPT entries (16 bytes) each. The header indicates how 2988many valid HPT entries there are and how many invalid entries follow 2989the valid entries. The invalid entries are not represented explicitly 2990in the stream. The header format is:: 2991 2992 struct kvm_get_htab_header { 2993 __u32 index; 2994 __u16 n_valid; 2995 __u16 n_invalid; 2996 }; 2997 2998Writes to the fd create HPT entries starting at the index given in the 2999header; first 'n_valid' valid entries with contents from the data 3000written, then 'n_invalid' invalid entries, invalidating any previously 3001valid entries found. 3002 30034.79 KVM_CREATE_DEVICE 3004---------------------- 3005 3006:Capability: KVM_CAP_DEVICE_CTRL 3007:Type: vm ioctl 3008:Parameters: struct kvm_create_device (in/out) 3009:Returns: 0 on success, -1 on error 3010 3011Errors: 3012 3013 ====== ======================================================= 3014 ENODEV The device type is unknown or unsupported 3015 EEXIST Device already created, and this type of device may not 3016 be instantiated multiple times 3017 ====== ======================================================= 3018 3019 Other error conditions may be defined by individual device types or 3020 have their standard meanings. 3021 3022Creates an emulated device in the kernel. The file descriptor returned 3023in fd can be used with KVM_SET/GET/HAS_DEVICE_ATTR. 3024 3025If the KVM_CREATE_DEVICE_TEST flag is set, only test whether the 3026device type is supported (not necessarily whether it can be created 3027in the current vm). 3028 3029Individual devices should not define flags. Attributes should be used 3030for specifying any behavior that is not implied by the device type 3031number. 3032 3033:: 3034 3035 struct kvm_create_device { 3036 __u32 type; /* in: KVM_DEV_TYPE_xxx */ 3037 __u32 fd; /* out: device handle */ 3038 __u32 flags; /* in: KVM_CREATE_DEVICE_xxx */ 3039 }; 3040 30414.80 KVM_SET_DEVICE_ATTR/KVM_GET_DEVICE_ATTR 3042-------------------------------------------- 3043 3044:Capability: KVM_CAP_DEVICE_CTRL, KVM_CAP_VM_ATTRIBUTES for vm device, 3045 KVM_CAP_VCPU_ATTRIBUTES for vcpu device 3046:Type: device ioctl, vm ioctl, vcpu ioctl 3047:Parameters: struct kvm_device_attr 3048:Returns: 0 on success, -1 on error 3049 3050Errors: 3051 3052 ===== ============================================================= 3053 ENXIO The group or attribute is unknown/unsupported for this device 3054 or hardware support is missing. 3055 EPERM The attribute cannot (currently) be accessed this way 3056 (e.g. read-only attribute, or attribute that only makes 3057 sense when the device is in a different state) 3058 ===== ============================================================= 3059 3060 Other error conditions may be defined by individual device types. 3061 3062Gets/sets a specified piece of device configuration and/or state. The 3063semantics are device-specific. See individual device documentation in 3064the "devices" directory. As with ONE_REG, the size of the data 3065transferred is defined by the particular attribute. 3066 3067:: 3068 3069 struct kvm_device_attr { 3070 __u32 flags; /* no flags currently defined */ 3071 __u32 group; /* device-defined */ 3072 __u64 attr; /* group-defined */ 3073 __u64 addr; /* userspace address of attr data */ 3074 }; 3075 30764.81 KVM_HAS_DEVICE_ATTR 3077------------------------ 3078 3079:Capability: KVM_CAP_DEVICE_CTRL, KVM_CAP_VM_ATTRIBUTES for vm device, 3080 KVM_CAP_VCPU_ATTRIBUTES for vcpu device 3081:Type: device ioctl, vm ioctl, vcpu ioctl 3082:Parameters: struct kvm_device_attr 3083:Returns: 0 on success, -1 on error 3084 3085Errors: 3086 3087 ===== ============================================================= 3088 ENXIO The group or attribute is unknown/unsupported for this device 3089 or hardware support is missing. 3090 ===== ============================================================= 3091 3092Tests whether a device supports a particular attribute. A successful 3093return indicates the attribute is implemented. It does not necessarily 3094indicate that the attribute can be read or written in the device's 3095current state. "addr" is ignored. 3096 30974.82 KVM_ARM_VCPU_INIT 3098---------------------- 3099 3100:Capability: basic 3101:Architectures: arm, arm64 3102:Type: vcpu ioctl 3103:Parameters: struct kvm_vcpu_init (in) 3104:Returns: 0 on success; -1 on error 3105 3106Errors: 3107 3108 ====== ================================================================= 3109 EINVAL the target is unknown, or the combination of features is invalid. 3110 ENOENT a features bit specified is unknown. 3111 ====== ================================================================= 3112 3113This tells KVM what type of CPU to present to the guest, and what 3114optional features it should have. This will cause a reset of the cpu 3115registers to their initial values. If this is not called, KVM_RUN will 3116return ENOEXEC for that vcpu. 3117 3118Note that because some registers reflect machine topology, all vcpus 3119should be created before this ioctl is invoked. 3120 3121Userspace can call this function multiple times for a given vcpu, including 3122after the vcpu has been run. This will reset the vcpu to its initial 3123state. All calls to this function after the initial call must use the same 3124target and same set of feature flags, otherwise EINVAL will be returned. 3125 3126Possible features: 3127 3128 - KVM_ARM_VCPU_POWER_OFF: Starts the CPU in a power-off state. 3129 Depends on KVM_CAP_ARM_PSCI. If not set, the CPU will be powered on 3130 and execute guest code when KVM_RUN is called. 3131 - KVM_ARM_VCPU_EL1_32BIT: Starts the CPU in a 32bit mode. 3132 Depends on KVM_CAP_ARM_EL1_32BIT (arm64 only). 3133 - KVM_ARM_VCPU_PSCI_0_2: Emulate PSCI v0.2 (or a future revision 3134 backward compatible with v0.2) for the CPU. 3135 Depends on KVM_CAP_ARM_PSCI_0_2. 3136 - KVM_ARM_VCPU_PMU_V3: Emulate PMUv3 for the CPU. 3137 Depends on KVM_CAP_ARM_PMU_V3. 3138 3139 - KVM_ARM_VCPU_PTRAUTH_ADDRESS: Enables Address Pointer authentication 3140 for arm64 only. 3141 Depends on KVM_CAP_ARM_PTRAUTH_ADDRESS. 3142 If KVM_CAP_ARM_PTRAUTH_ADDRESS and KVM_CAP_ARM_PTRAUTH_GENERIC are 3143 both present, then both KVM_ARM_VCPU_PTRAUTH_ADDRESS and 3144 KVM_ARM_VCPU_PTRAUTH_GENERIC must be requested or neither must be 3145 requested. 3146 3147 - KVM_ARM_VCPU_PTRAUTH_GENERIC: Enables Generic Pointer authentication 3148 for arm64 only. 3149 Depends on KVM_CAP_ARM_PTRAUTH_GENERIC. 3150 If KVM_CAP_ARM_PTRAUTH_ADDRESS and KVM_CAP_ARM_PTRAUTH_GENERIC are 3151 both present, then both KVM_ARM_VCPU_PTRAUTH_ADDRESS and 3152 KVM_ARM_VCPU_PTRAUTH_GENERIC must be requested or neither must be 3153 requested. 3154 3155 - KVM_ARM_VCPU_SVE: Enables SVE for the CPU (arm64 only). 3156 Depends on KVM_CAP_ARM_SVE. 3157 Requires KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE): 3158 3159 * After KVM_ARM_VCPU_INIT: 3160 3161 - KVM_REG_ARM64_SVE_VLS may be read using KVM_GET_ONE_REG: the 3162 initial value of this pseudo-register indicates the best set of 3163 vector lengths possible for a vcpu on this host. 3164 3165 * Before KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE): 3166 3167 - KVM_RUN and KVM_GET_REG_LIST are not available; 3168 3169 - KVM_GET_ONE_REG and KVM_SET_ONE_REG cannot be used to access 3170 the scalable archietctural SVE registers 3171 KVM_REG_ARM64_SVE_ZREG(), KVM_REG_ARM64_SVE_PREG() or 3172 KVM_REG_ARM64_SVE_FFR; 3173 3174 - KVM_REG_ARM64_SVE_VLS may optionally be written using 3175 KVM_SET_ONE_REG, to modify the set of vector lengths available 3176 for the vcpu. 3177 3178 * After KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE): 3179 3180 - the KVM_REG_ARM64_SVE_VLS pseudo-register is immutable, and can 3181 no longer be written using KVM_SET_ONE_REG. 3182 31834.83 KVM_ARM_PREFERRED_TARGET 3184----------------------------- 3185 3186:Capability: basic 3187:Architectures: arm, arm64 3188:Type: vm ioctl 3189:Parameters: struct kvm_vcpu_init (out) 3190:Returns: 0 on success; -1 on error 3191 3192Errors: 3193 3194 ====== ========================================== 3195 ENODEV no preferred target available for the host 3196 ====== ========================================== 3197 3198This queries KVM for preferred CPU target type which can be emulated 3199by KVM on underlying host. 3200 3201The ioctl returns struct kvm_vcpu_init instance containing information 3202about preferred CPU target type and recommended features for it. The 3203kvm_vcpu_init->features bitmap returned will have feature bits set if 3204the preferred target recommends setting these features, but this is 3205not mandatory. 3206 3207The information returned by this ioctl can be used to prepare an instance 3208of struct kvm_vcpu_init for KVM_ARM_VCPU_INIT ioctl which will result in 3209VCPU matching underlying host. 3210 3211 32124.84 KVM_GET_REG_LIST 3213--------------------- 3214 3215:Capability: basic 3216:Architectures: arm, arm64, mips 3217:Type: vcpu ioctl 3218:Parameters: struct kvm_reg_list (in/out) 3219:Returns: 0 on success; -1 on error 3220 3221Errors: 3222 3223 ===== ============================================================== 3224 E2BIG the reg index list is too big to fit in the array specified by 3225 the user (the number required will be written into n). 3226 ===== ============================================================== 3227 3228:: 3229 3230 struct kvm_reg_list { 3231 __u64 n; /* number of registers in reg[] */ 3232 __u64 reg[0]; 3233 }; 3234 3235This ioctl returns the guest registers that are supported for the 3236KVM_GET_ONE_REG/KVM_SET_ONE_REG calls. 3237 3238 32394.85 KVM_ARM_SET_DEVICE_ADDR (deprecated) 3240----------------------------------------- 3241 3242:Capability: KVM_CAP_ARM_SET_DEVICE_ADDR 3243:Architectures: arm, arm64 3244:Type: vm ioctl 3245:Parameters: struct kvm_arm_device_address (in) 3246:Returns: 0 on success, -1 on error 3247 3248Errors: 3249 3250 ====== ============================================ 3251 ENODEV The device id is unknown 3252 ENXIO Device not supported on current system 3253 EEXIST Address already set 3254 E2BIG Address outside guest physical address space 3255 EBUSY Address overlaps with other device range 3256 ====== ============================================ 3257 3258:: 3259 3260 struct kvm_arm_device_addr { 3261 __u64 id; 3262 __u64 addr; 3263 }; 3264 3265Specify a device address in the guest's physical address space where guests 3266can access emulated or directly exposed devices, which the host kernel needs 3267to know about. The id field is an architecture specific identifier for a 3268specific device. 3269 3270ARM/arm64 divides the id field into two parts, a device id and an 3271address type id specific to the individual device:: 3272 3273 bits: | 63 ... 32 | 31 ... 16 | 15 ... 0 | 3274 field: | 0x00000000 | device id | addr type id | 3275 3276ARM/arm64 currently only require this when using the in-kernel GIC 3277support for the hardware VGIC features, using KVM_ARM_DEVICE_VGIC_V2 3278as the device id. When setting the base address for the guest's 3279mapping of the VGIC virtual CPU and distributor interface, the ioctl 3280must be called after calling KVM_CREATE_IRQCHIP, but before calling 3281KVM_RUN on any of the VCPUs. Calling this ioctl twice for any of the 3282base addresses will return -EEXIST. 3283 3284Note, this IOCTL is deprecated and the more flexible SET/GET_DEVICE_ATTR API 3285should be used instead. 3286 3287 32884.86 KVM_PPC_RTAS_DEFINE_TOKEN 3289------------------------------ 3290 3291:Capability: KVM_CAP_PPC_RTAS 3292:Architectures: ppc 3293:Type: vm ioctl 3294:Parameters: struct kvm_rtas_token_args 3295:Returns: 0 on success, -1 on error 3296 3297Defines a token value for a RTAS (Run Time Abstraction Services) 3298service in order to allow it to be handled in the kernel. The 3299argument struct gives the name of the service, which must be the name 3300of a service that has a kernel-side implementation. If the token 3301value is non-zero, it will be associated with that service, and 3302subsequent RTAS calls by the guest specifying that token will be 3303handled by the kernel. If the token value is 0, then any token 3304associated with the service will be forgotten, and subsequent RTAS 3305calls by the guest for that service will be passed to userspace to be 3306handled. 3307 33084.87 KVM_SET_GUEST_DEBUG 3309------------------------ 3310 3311:Capability: KVM_CAP_SET_GUEST_DEBUG 3312:Architectures: x86, s390, ppc, arm64 3313:Type: vcpu ioctl 3314:Parameters: struct kvm_guest_debug (in) 3315:Returns: 0 on success; -1 on error 3316 3317:: 3318 3319 struct kvm_guest_debug { 3320 __u32 control; 3321 __u32 pad; 3322 struct kvm_guest_debug_arch arch; 3323 }; 3324 3325Set up the processor specific debug registers and configure vcpu for 3326handling guest debug events. There are two parts to the structure, the 3327first a control bitfield indicates the type of debug events to handle 3328when running. Common control bits are: 3329 3330 - KVM_GUESTDBG_ENABLE: guest debugging is enabled 3331 - KVM_GUESTDBG_SINGLESTEP: the next run should single-step 3332 3333The top 16 bits of the control field are architecture specific control 3334flags which can include the following: 3335 3336 - KVM_GUESTDBG_USE_SW_BP: using software breakpoints [x86, arm64] 3337 - KVM_GUESTDBG_USE_HW_BP: using hardware breakpoints [x86, s390, arm64] 3338 - KVM_GUESTDBG_INJECT_DB: inject DB type exception [x86] 3339 - KVM_GUESTDBG_INJECT_BP: inject BP type exception [x86] 3340 - KVM_GUESTDBG_EXIT_PENDING: trigger an immediate guest exit [s390] 3341 3342For example KVM_GUESTDBG_USE_SW_BP indicates that software breakpoints 3343are enabled in memory so we need to ensure breakpoint exceptions are 3344correctly trapped and the KVM run loop exits at the breakpoint and not 3345running off into the normal guest vector. For KVM_GUESTDBG_USE_HW_BP 3346we need to ensure the guest vCPUs architecture specific registers are 3347updated to the correct (supplied) values. 3348 3349The second part of the structure is architecture specific and 3350typically contains a set of debug registers. 3351 3352For arm64 the number of debug registers is implementation defined and 3353can be determined by querying the KVM_CAP_GUEST_DEBUG_HW_BPS and 3354KVM_CAP_GUEST_DEBUG_HW_WPS capabilities which return a positive number 3355indicating the number of supported registers. 3356 3357For ppc, the KVM_CAP_PPC_GUEST_DEBUG_SSTEP capability indicates whether 3358the single-step debug event (KVM_GUESTDBG_SINGLESTEP) is supported. 3359 3360When debug events exit the main run loop with the reason 3361KVM_EXIT_DEBUG with the kvm_debug_exit_arch part of the kvm_run 3362structure containing architecture specific debug information. 3363 33644.88 KVM_GET_EMULATED_CPUID 3365--------------------------- 3366 3367:Capability: KVM_CAP_EXT_EMUL_CPUID 3368:Architectures: x86 3369:Type: system ioctl 3370:Parameters: struct kvm_cpuid2 (in/out) 3371:Returns: 0 on success, -1 on error 3372 3373:: 3374 3375 struct kvm_cpuid2 { 3376 __u32 nent; 3377 __u32 flags; 3378 struct kvm_cpuid_entry2 entries[0]; 3379 }; 3380 3381The member 'flags' is used for passing flags from userspace. 3382 3383:: 3384 3385 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX BIT(0) 3386 #define KVM_CPUID_FLAG_STATEFUL_FUNC BIT(1) /* deprecated */ 3387 #define KVM_CPUID_FLAG_STATE_READ_NEXT BIT(2) /* deprecated */ 3388 3389 struct kvm_cpuid_entry2 { 3390 __u32 function; 3391 __u32 index; 3392 __u32 flags; 3393 __u32 eax; 3394 __u32 ebx; 3395 __u32 ecx; 3396 __u32 edx; 3397 __u32 padding[3]; 3398 }; 3399 3400This ioctl returns x86 cpuid features which are emulated by 3401kvm.Userspace can use the information returned by this ioctl to query 3402which features are emulated by kvm instead of being present natively. 3403 3404Userspace invokes KVM_GET_EMULATED_CPUID by passing a kvm_cpuid2 3405structure with the 'nent' field indicating the number of entries in 3406the variable-size array 'entries'. If the number of entries is too low 3407to describe the cpu capabilities, an error (E2BIG) is returned. If the 3408number is too high, the 'nent' field is adjusted and an error (ENOMEM) 3409is returned. If the number is just right, the 'nent' field is adjusted 3410to the number of valid entries in the 'entries' array, which is then 3411filled. 3412 3413The entries returned are the set CPUID bits of the respective features 3414which kvm emulates, as returned by the CPUID instruction, with unknown 3415or unsupported feature bits cleared. 3416 3417Features like x2apic, for example, may not be present in the host cpu 3418but are exposed by kvm in KVM_GET_SUPPORTED_CPUID because they can be 3419emulated efficiently and thus not included here. 3420 3421The fields in each entry are defined as follows: 3422 3423 function: 3424 the eax value used to obtain the entry 3425 index: 3426 the ecx value used to obtain the entry (for entries that are 3427 affected by ecx) 3428 flags: 3429 an OR of zero or more of the following: 3430 3431 KVM_CPUID_FLAG_SIGNIFCANT_INDEX: 3432 if the index field is valid 3433 3434 eax, ebx, ecx, edx: 3435 3436 the values returned by the cpuid instruction for 3437 this function/index combination 3438 34394.89 KVM_S390_MEM_OP 3440-------------------- 3441 3442:Capability: KVM_CAP_S390_MEM_OP 3443:Architectures: s390 3444:Type: vcpu ioctl 3445:Parameters: struct kvm_s390_mem_op (in) 3446:Returns: = 0 on success, 3447 < 0 on generic error (e.g. -EFAULT or -ENOMEM), 3448 > 0 if an exception occurred while walking the page tables 3449 3450Read or write data from/to the logical (virtual) memory of a VCPU. 3451 3452Parameters are specified via the following structure:: 3453 3454 struct kvm_s390_mem_op { 3455 __u64 gaddr; /* the guest address */ 3456 __u64 flags; /* flags */ 3457 __u32 size; /* amount of bytes */ 3458 __u32 op; /* type of operation */ 3459 __u64 buf; /* buffer in userspace */ 3460 __u8 ar; /* the access register number */ 3461 __u8 reserved[31]; /* should be set to 0 */ 3462 }; 3463 3464The type of operation is specified in the "op" field. It is either 3465KVM_S390_MEMOP_LOGICAL_READ for reading from logical memory space or 3466KVM_S390_MEMOP_LOGICAL_WRITE for writing to logical memory space. The 3467KVM_S390_MEMOP_F_CHECK_ONLY flag can be set in the "flags" field to check 3468whether the corresponding memory access would create an access exception 3469(without touching the data in the memory at the destination). In case an 3470access exception occurred while walking the MMU tables of the guest, the 3471ioctl returns a positive error number to indicate the type of exception. 3472This exception is also raised directly at the corresponding VCPU if the 3473flag KVM_S390_MEMOP_F_INJECT_EXCEPTION is set in the "flags" field. 3474 3475The start address of the memory region has to be specified in the "gaddr" 3476field, and the length of the region in the "size" field (which must not 3477be 0). The maximum value for "size" can be obtained by checking the 3478KVM_CAP_S390_MEM_OP capability. "buf" is the buffer supplied by the 3479userspace application where the read data should be written to for 3480KVM_S390_MEMOP_LOGICAL_READ, or where the data that should be written is 3481stored for a KVM_S390_MEMOP_LOGICAL_WRITE. When KVM_S390_MEMOP_F_CHECK_ONLY 3482is specified, "buf" is unused and can be NULL. "ar" designates the access 3483register number to be used; the valid range is 0..15. 3484 3485The "reserved" field is meant for future extensions. It is not used by 3486KVM with the currently defined set of flags. 3487 34884.90 KVM_S390_GET_SKEYS 3489----------------------- 3490 3491:Capability: KVM_CAP_S390_SKEYS 3492:Architectures: s390 3493:Type: vm ioctl 3494:Parameters: struct kvm_s390_skeys 3495:Returns: 0 on success, KVM_S390_GET_KEYS_NONE if guest is not using storage 3496 keys, negative value on error 3497 3498This ioctl is used to get guest storage key values on the s390 3499architecture. The ioctl takes parameters via the kvm_s390_skeys struct:: 3500 3501 struct kvm_s390_skeys { 3502 __u64 start_gfn; 3503 __u64 count; 3504 __u64 skeydata_addr; 3505 __u32 flags; 3506 __u32 reserved[9]; 3507 }; 3508 3509The start_gfn field is the number of the first guest frame whose storage keys 3510you want to get. 3511 3512The count field is the number of consecutive frames (starting from start_gfn) 3513whose storage keys to get. The count field must be at least 1 and the maximum 3514allowed value is defined as KVM_S390_SKEYS_ALLOC_MAX. Values outside this range 3515will cause the ioctl to return -EINVAL. 3516 3517The skeydata_addr field is the address to a buffer large enough to hold count 3518bytes. This buffer will be filled with storage key data by the ioctl. 3519 35204.91 KVM_S390_SET_SKEYS 3521----------------------- 3522 3523:Capability: KVM_CAP_S390_SKEYS 3524:Architectures: s390 3525:Type: vm ioctl 3526:Parameters: struct kvm_s390_skeys 3527:Returns: 0 on success, negative value on error 3528 3529This ioctl is used to set guest storage key values on the s390 3530architecture. The ioctl takes parameters via the kvm_s390_skeys struct. 3531See section on KVM_S390_GET_SKEYS for struct definition. 3532 3533The start_gfn field is the number of the first guest frame whose storage keys 3534you want to set. 3535 3536The count field is the number of consecutive frames (starting from start_gfn) 3537whose storage keys to get. The count field must be at least 1 and the maximum 3538allowed value is defined as KVM_S390_SKEYS_ALLOC_MAX. Values outside this range 3539will cause the ioctl to return -EINVAL. 3540 3541The skeydata_addr field is the address to a buffer containing count bytes of 3542storage keys. Each byte in the buffer will be set as the storage key for a 3543single frame starting at start_gfn for count frames. 3544 3545Note: If any architecturally invalid key value is found in the given data then 3546the ioctl will return -EINVAL. 3547 35484.92 KVM_S390_IRQ 3549----------------- 3550 3551:Capability: KVM_CAP_S390_INJECT_IRQ 3552:Architectures: s390 3553:Type: vcpu ioctl 3554:Parameters: struct kvm_s390_irq (in) 3555:Returns: 0 on success, -1 on error 3556 3557Errors: 3558 3559 3560 ====== ================================================================= 3561 EINVAL interrupt type is invalid 3562 type is KVM_S390_SIGP_STOP and flag parameter is invalid value, 3563 type is KVM_S390_INT_EXTERNAL_CALL and code is bigger 3564 than the maximum of VCPUs 3565 EBUSY type is KVM_S390_SIGP_SET_PREFIX and vcpu is not stopped, 3566 type is KVM_S390_SIGP_STOP and a stop irq is already pending, 3567 type is KVM_S390_INT_EXTERNAL_CALL and an external call interrupt 3568 is already pending 3569 ====== ================================================================= 3570 3571Allows to inject an interrupt to the guest. 3572 3573Using struct kvm_s390_irq as a parameter allows 3574to inject additional payload which is not 3575possible via KVM_S390_INTERRUPT. 3576 3577Interrupt parameters are passed via kvm_s390_irq:: 3578 3579 struct kvm_s390_irq { 3580 __u64 type; 3581 union { 3582 struct kvm_s390_io_info io; 3583 struct kvm_s390_ext_info ext; 3584 struct kvm_s390_pgm_info pgm; 3585 struct kvm_s390_emerg_info emerg; 3586 struct kvm_s390_extcall_info extcall; 3587 struct kvm_s390_prefix_info prefix; 3588 struct kvm_s390_stop_info stop; 3589 struct kvm_s390_mchk_info mchk; 3590 char reserved[64]; 3591 } u; 3592 }; 3593 3594type can be one of the following: 3595 3596- KVM_S390_SIGP_STOP - sigp stop; parameter in .stop 3597- KVM_S390_PROGRAM_INT - program check; parameters in .pgm 3598- KVM_S390_SIGP_SET_PREFIX - sigp set prefix; parameters in .prefix 3599- KVM_S390_RESTART - restart; no parameters 3600- KVM_S390_INT_CLOCK_COMP - clock comparator interrupt; no parameters 3601- KVM_S390_INT_CPU_TIMER - CPU timer interrupt; no parameters 3602- KVM_S390_INT_EMERGENCY - sigp emergency; parameters in .emerg 3603- KVM_S390_INT_EXTERNAL_CALL - sigp external call; parameters in .extcall 3604- KVM_S390_MCHK - machine check interrupt; parameters in .mchk 3605 3606This is an asynchronous vcpu ioctl and can be invoked from any thread. 3607 36084.94 KVM_S390_GET_IRQ_STATE 3609--------------------------- 3610 3611:Capability: KVM_CAP_S390_IRQ_STATE 3612:Architectures: s390 3613:Type: vcpu ioctl 3614:Parameters: struct kvm_s390_irq_state (out) 3615:Returns: >= number of bytes copied into buffer, 3616 -EINVAL if buffer size is 0, 3617 -ENOBUFS if buffer size is too small to fit all pending interrupts, 3618 -EFAULT if the buffer address was invalid 3619 3620This ioctl allows userspace to retrieve the complete state of all currently 3621pending interrupts in a single buffer. Use cases include migration 3622and introspection. The parameter structure contains the address of a 3623userspace buffer and its length:: 3624 3625 struct kvm_s390_irq_state { 3626 __u64 buf; 3627 __u32 flags; /* will stay unused for compatibility reasons */ 3628 __u32 len; 3629 __u32 reserved[4]; /* will stay unused for compatibility reasons */ 3630 }; 3631 3632Userspace passes in the above struct and for each pending interrupt a 3633struct kvm_s390_irq is copied to the provided buffer. 3634 3635The structure contains a flags and a reserved field for future extensions. As 3636the kernel never checked for flags == 0 and QEMU never pre-zeroed flags and 3637reserved, these fields can not be used in the future without breaking 3638compatibility. 3639 3640If -ENOBUFS is returned the buffer provided was too small and userspace 3641may retry with a bigger buffer. 3642 36434.95 KVM_S390_SET_IRQ_STATE 3644--------------------------- 3645 3646:Capability: KVM_CAP_S390_IRQ_STATE 3647:Architectures: s390 3648:Type: vcpu ioctl 3649:Parameters: struct kvm_s390_irq_state (in) 3650:Returns: 0 on success, 3651 -EFAULT if the buffer address was invalid, 3652 -EINVAL for an invalid buffer length (see below), 3653 -EBUSY if there were already interrupts pending, 3654 errors occurring when actually injecting the 3655 interrupt. See KVM_S390_IRQ. 3656 3657This ioctl allows userspace to set the complete state of all cpu-local 3658interrupts currently pending for the vcpu. It is intended for restoring 3659interrupt state after a migration. The input parameter is a userspace buffer 3660containing a struct kvm_s390_irq_state:: 3661 3662 struct kvm_s390_irq_state { 3663 __u64 buf; 3664 __u32 flags; /* will stay unused for compatibility reasons */ 3665 __u32 len; 3666 __u32 reserved[4]; /* will stay unused for compatibility reasons */ 3667 }; 3668 3669The restrictions for flags and reserved apply as well. 3670(see KVM_S390_GET_IRQ_STATE) 3671 3672The userspace memory referenced by buf contains a struct kvm_s390_irq 3673for each interrupt to be injected into the guest. 3674If one of the interrupts could not be injected for some reason the 3675ioctl aborts. 3676 3677len must be a multiple of sizeof(struct kvm_s390_irq). It must be > 0 3678and it must not exceed (max_vcpus + 32) * sizeof(struct kvm_s390_irq), 3679which is the maximum number of possibly pending cpu-local interrupts. 3680 36814.96 KVM_SMI 3682------------ 3683 3684:Capability: KVM_CAP_X86_SMM 3685:Architectures: x86 3686:Type: vcpu ioctl 3687:Parameters: none 3688:Returns: 0 on success, -1 on error 3689 3690Queues an SMI on the thread's vcpu. 3691 36924.97 KVM_CAP_PPC_MULTITCE 3693------------------------- 3694 3695:Capability: KVM_CAP_PPC_MULTITCE 3696:Architectures: ppc 3697:Type: vm 3698 3699This capability means the kernel is capable of handling hypercalls 3700H_PUT_TCE_INDIRECT and H_STUFF_TCE without passing those into the user 3701space. This significantly accelerates DMA operations for PPC KVM guests. 3702User space should expect that its handlers for these hypercalls 3703are not going to be called if user space previously registered LIOBN 3704in KVM (via KVM_CREATE_SPAPR_TCE or similar calls). 3705 3706In order to enable H_PUT_TCE_INDIRECT and H_STUFF_TCE use in the guest, 3707user space might have to advertise it for the guest. For example, 3708IBM pSeries (sPAPR) guest starts using them if "hcall-multi-tce" is 3709present in the "ibm,hypertas-functions" device-tree property. 3710 3711The hypercalls mentioned above may or may not be processed successfully 3712in the kernel based fast path. If they can not be handled by the kernel, 3713they will get passed on to user space. So user space still has to have 3714an implementation for these despite the in kernel acceleration. 3715 3716This capability is always enabled. 3717 37184.98 KVM_CREATE_SPAPR_TCE_64 3719---------------------------- 3720 3721:Capability: KVM_CAP_SPAPR_TCE_64 3722:Architectures: powerpc 3723:Type: vm ioctl 3724:Parameters: struct kvm_create_spapr_tce_64 (in) 3725:Returns: file descriptor for manipulating the created TCE table 3726 3727This is an extension for KVM_CAP_SPAPR_TCE which only supports 32bit 3728windows, described in 4.62 KVM_CREATE_SPAPR_TCE 3729 3730This capability uses extended struct in ioctl interface:: 3731 3732 /* for KVM_CAP_SPAPR_TCE_64 */ 3733 struct kvm_create_spapr_tce_64 { 3734 __u64 liobn; 3735 __u32 page_shift; 3736 __u32 flags; 3737 __u64 offset; /* in pages */ 3738 __u64 size; /* in pages */ 3739 }; 3740 3741The aim of extension is to support an additional bigger DMA window with 3742a variable page size. 3743KVM_CREATE_SPAPR_TCE_64 receives a 64bit window size, an IOMMU page shift and 3744a bus offset of the corresponding DMA window, @size and @offset are numbers 3745of IOMMU pages. 3746 3747@flags are not used at the moment. 3748 3749The rest of functionality is identical to KVM_CREATE_SPAPR_TCE. 3750 37514.99 KVM_REINJECT_CONTROL 3752------------------------- 3753 3754:Capability: KVM_CAP_REINJECT_CONTROL 3755:Architectures: x86 3756:Type: vm ioctl 3757:Parameters: struct kvm_reinject_control (in) 3758:Returns: 0 on success, 3759 -EFAULT if struct kvm_reinject_control cannot be read, 3760 -ENXIO if KVM_CREATE_PIT or KVM_CREATE_PIT2 didn't succeed earlier. 3761 3762i8254 (PIT) has two modes, reinject and !reinject. The default is reinject, 3763where KVM queues elapsed i8254 ticks and monitors completion of interrupt from 3764vector(s) that i8254 injects. Reinject mode dequeues a tick and injects its 3765interrupt whenever there isn't a pending interrupt from i8254. 3766!reinject mode injects an interrupt as soon as a tick arrives. 3767 3768:: 3769 3770 struct kvm_reinject_control { 3771 __u8 pit_reinject; 3772 __u8 reserved[31]; 3773 }; 3774 3775pit_reinject = 0 (!reinject mode) is recommended, unless running an old 3776operating system that uses the PIT for timing (e.g. Linux 2.4.x). 3777 37784.100 KVM_PPC_CONFIGURE_V3_MMU 3779------------------------------ 3780 3781:Capability: KVM_CAP_PPC_RADIX_MMU or KVM_CAP_PPC_HASH_MMU_V3 3782:Architectures: ppc 3783:Type: vm ioctl 3784:Parameters: struct kvm_ppc_mmuv3_cfg (in) 3785:Returns: 0 on success, 3786 -EFAULT if struct kvm_ppc_mmuv3_cfg cannot be read, 3787 -EINVAL if the configuration is invalid 3788 3789This ioctl controls whether the guest will use radix or HPT (hashed 3790page table) translation, and sets the pointer to the process table for 3791the guest. 3792 3793:: 3794 3795 struct kvm_ppc_mmuv3_cfg { 3796 __u64 flags; 3797 __u64 process_table; 3798 }; 3799 3800There are two bits that can be set in flags; KVM_PPC_MMUV3_RADIX and 3801KVM_PPC_MMUV3_GTSE. KVM_PPC_MMUV3_RADIX, if set, configures the guest 3802to use radix tree translation, and if clear, to use HPT translation. 3803KVM_PPC_MMUV3_GTSE, if set and if KVM permits it, configures the guest 3804to be able to use the global TLB and SLB invalidation instructions; 3805if clear, the guest may not use these instructions. 3806 3807The process_table field specifies the address and size of the guest 3808process table, which is in the guest's space. This field is formatted 3809as the second doubleword of the partition table entry, as defined in 3810the Power ISA V3.00, Book III section 5.7.6.1. 3811 38124.101 KVM_PPC_GET_RMMU_INFO 3813--------------------------- 3814 3815:Capability: KVM_CAP_PPC_RADIX_MMU 3816:Architectures: ppc 3817:Type: vm ioctl 3818:Parameters: struct kvm_ppc_rmmu_info (out) 3819:Returns: 0 on success, 3820 -EFAULT if struct kvm_ppc_rmmu_info cannot be written, 3821 -EINVAL if no useful information can be returned 3822 3823This ioctl returns a structure containing two things: (a) a list 3824containing supported radix tree geometries, and (b) a list that maps 3825page sizes to put in the "AP" (actual page size) field for the tlbie 3826(TLB invalidate entry) instruction. 3827 3828:: 3829 3830 struct kvm_ppc_rmmu_info { 3831 struct kvm_ppc_radix_geom { 3832 __u8 page_shift; 3833 __u8 level_bits[4]; 3834 __u8 pad[3]; 3835 } geometries[8]; 3836 __u32 ap_encodings[8]; 3837 }; 3838 3839The geometries[] field gives up to 8 supported geometries for the 3840radix page table, in terms of the log base 2 of the smallest page 3841size, and the number of bits indexed at each level of the tree, from 3842the PTE level up to the PGD level in that order. Any unused entries 3843will have 0 in the page_shift field. 3844 3845The ap_encodings gives the supported page sizes and their AP field 3846encodings, encoded with the AP value in the top 3 bits and the log 3847base 2 of the page size in the bottom 6 bits. 3848 38494.102 KVM_PPC_RESIZE_HPT_PREPARE 3850-------------------------------- 3851 3852:Capability: KVM_CAP_SPAPR_RESIZE_HPT 3853:Architectures: powerpc 3854:Type: vm ioctl 3855:Parameters: struct kvm_ppc_resize_hpt (in) 3856:Returns: 0 on successful completion, 3857 >0 if a new HPT is being prepared, the value is an estimated 3858 number of milliseconds until preparation is complete, 3859 -EFAULT if struct kvm_reinject_control cannot be read, 3860 -EINVAL if the supplied shift or flags are invalid, 3861 -ENOMEM if unable to allocate the new HPT, 3862 3863Used to implement the PAPR extension for runtime resizing of a guest's 3864Hashed Page Table (HPT). Specifically this starts, stops or monitors 3865the preparation of a new potential HPT for the guest, essentially 3866implementing the H_RESIZE_HPT_PREPARE hypercall. 3867 3868:: 3869 3870 struct kvm_ppc_resize_hpt { 3871 __u64 flags; 3872 __u32 shift; 3873 __u32 pad; 3874 }; 3875 3876If called with shift > 0 when there is no pending HPT for the guest, 3877this begins preparation of a new pending HPT of size 2^(shift) bytes. 3878It then returns a positive integer with the estimated number of 3879milliseconds until preparation is complete. 3880 3881If called when there is a pending HPT whose size does not match that 3882requested in the parameters, discards the existing pending HPT and 3883creates a new one as above. 3884 3885If called when there is a pending HPT of the size requested, will: 3886 3887 * If preparation of the pending HPT is already complete, return 0 3888 * If preparation of the pending HPT has failed, return an error 3889 code, then discard the pending HPT. 3890 * If preparation of the pending HPT is still in progress, return an 3891 estimated number of milliseconds until preparation is complete. 3892 3893If called with shift == 0, discards any currently pending HPT and 3894returns 0 (i.e. cancels any in-progress preparation). 3895 3896flags is reserved for future expansion, currently setting any bits in 3897flags will result in an -EINVAL. 3898 3899Normally this will be called repeatedly with the same parameters until 3900it returns <= 0. The first call will initiate preparation, subsequent 3901ones will monitor preparation until it completes or fails. 3902 39034.103 KVM_PPC_RESIZE_HPT_COMMIT 3904------------------------------- 3905 3906:Capability: KVM_CAP_SPAPR_RESIZE_HPT 3907:Architectures: powerpc 3908:Type: vm ioctl 3909:Parameters: struct kvm_ppc_resize_hpt (in) 3910:Returns: 0 on successful completion, 3911 -EFAULT if struct kvm_reinject_control cannot be read, 3912 -EINVAL if the supplied shift or flags are invalid, 3913 -ENXIO is there is no pending HPT, or the pending HPT doesn't 3914 have the requested size, 3915 -EBUSY if the pending HPT is not fully prepared, 3916 -ENOSPC if there was a hash collision when moving existing 3917 HPT entries to the new HPT, 3918 -EIO on other error conditions 3919 3920Used to implement the PAPR extension for runtime resizing of a guest's 3921Hashed Page Table (HPT). Specifically this requests that the guest be 3922transferred to working with the new HPT, essentially implementing the 3923H_RESIZE_HPT_COMMIT hypercall. 3924 3925:: 3926 3927 struct kvm_ppc_resize_hpt { 3928 __u64 flags; 3929 __u32 shift; 3930 __u32 pad; 3931 }; 3932 3933This should only be called after KVM_PPC_RESIZE_HPT_PREPARE has 3934returned 0 with the same parameters. In other cases 3935KVM_PPC_RESIZE_HPT_COMMIT will return an error (usually -ENXIO or 3936-EBUSY, though others may be possible if the preparation was started, 3937but failed). 3938 3939This will have undefined effects on the guest if it has not already 3940placed itself in a quiescent state where no vcpu will make MMU enabled 3941memory accesses. 3942 3943On succsful completion, the pending HPT will become the guest's active 3944HPT and the previous HPT will be discarded. 3945 3946On failure, the guest will still be operating on its previous HPT. 3947 39484.104 KVM_X86_GET_MCE_CAP_SUPPORTED 3949----------------------------------- 3950 3951:Capability: KVM_CAP_MCE 3952:Architectures: x86 3953:Type: system ioctl 3954:Parameters: u64 mce_cap (out) 3955:Returns: 0 on success, -1 on error 3956 3957Returns supported MCE capabilities. The u64 mce_cap parameter 3958has the same format as the MSR_IA32_MCG_CAP register. Supported 3959capabilities will have the corresponding bits set. 3960 39614.105 KVM_X86_SETUP_MCE 3962----------------------- 3963 3964:Capability: KVM_CAP_MCE 3965:Architectures: x86 3966:Type: vcpu ioctl 3967:Parameters: u64 mcg_cap (in) 3968:Returns: 0 on success, 3969 -EFAULT if u64 mcg_cap cannot be read, 3970 -EINVAL if the requested number of banks is invalid, 3971 -EINVAL if requested MCE capability is not supported. 3972 3973Initializes MCE support for use. The u64 mcg_cap parameter 3974has the same format as the MSR_IA32_MCG_CAP register and 3975specifies which capabilities should be enabled. The maximum 3976supported number of error-reporting banks can be retrieved when 3977checking for KVM_CAP_MCE. The supported capabilities can be 3978retrieved with KVM_X86_GET_MCE_CAP_SUPPORTED. 3979 39804.106 KVM_X86_SET_MCE 3981--------------------- 3982 3983:Capability: KVM_CAP_MCE 3984:Architectures: x86 3985:Type: vcpu ioctl 3986:Parameters: struct kvm_x86_mce (in) 3987:Returns: 0 on success, 3988 -EFAULT if struct kvm_x86_mce cannot be read, 3989 -EINVAL if the bank number is invalid, 3990 -EINVAL if VAL bit is not set in status field. 3991 3992Inject a machine check error (MCE) into the guest. The input 3993parameter is:: 3994 3995 struct kvm_x86_mce { 3996 __u64 status; 3997 __u64 addr; 3998 __u64 misc; 3999 __u64 mcg_status; 4000 __u8 bank; 4001 __u8 pad1[7]; 4002 __u64 pad2[3]; 4003 }; 4004 4005If the MCE being reported is an uncorrected error, KVM will 4006inject it as an MCE exception into the guest. If the guest 4007MCG_STATUS register reports that an MCE is in progress, KVM 4008causes an KVM_EXIT_SHUTDOWN vmexit. 4009 4010Otherwise, if the MCE is a corrected error, KVM will just 4011store it in the corresponding bank (provided this bank is 4012not holding a previously reported uncorrected error). 4013 40144.107 KVM_S390_GET_CMMA_BITS 4015---------------------------- 4016 4017:Capability: KVM_CAP_S390_CMMA_MIGRATION 4018:Architectures: s390 4019:Type: vm ioctl 4020:Parameters: struct kvm_s390_cmma_log (in, out) 4021:Returns: 0 on success, a negative value on error 4022 4023This ioctl is used to get the values of the CMMA bits on the s390 4024architecture. It is meant to be used in two scenarios: 4025 4026- During live migration to save the CMMA values. Live migration needs 4027 to be enabled via the KVM_REQ_START_MIGRATION VM property. 4028- To non-destructively peek at the CMMA values, with the flag 4029 KVM_S390_CMMA_PEEK set. 4030 4031The ioctl takes parameters via the kvm_s390_cmma_log struct. The desired 4032values are written to a buffer whose location is indicated via the "values" 4033member in the kvm_s390_cmma_log struct. The values in the input struct are 4034also updated as needed. 4035 4036Each CMMA value takes up one byte. 4037 4038:: 4039 4040 struct kvm_s390_cmma_log { 4041 __u64 start_gfn; 4042 __u32 count; 4043 __u32 flags; 4044 union { 4045 __u64 remaining; 4046 __u64 mask; 4047 }; 4048 __u64 values; 4049 }; 4050 4051start_gfn is the number of the first guest frame whose CMMA values are 4052to be retrieved, 4053 4054count is the length of the buffer in bytes, 4055 4056values points to the buffer where the result will be written to. 4057 4058If count is greater than KVM_S390_SKEYS_MAX, then it is considered to be 4059KVM_S390_SKEYS_MAX. KVM_S390_SKEYS_MAX is re-used for consistency with 4060other ioctls. 4061 4062The result is written in the buffer pointed to by the field values, and 4063the values of the input parameter are updated as follows. 4064 4065Depending on the flags, different actions are performed. The only 4066supported flag so far is KVM_S390_CMMA_PEEK. 4067 4068The default behaviour if KVM_S390_CMMA_PEEK is not set is: 4069start_gfn will indicate the first page frame whose CMMA bits were dirty. 4070It is not necessarily the same as the one passed as input, as clean pages 4071are skipped. 4072 4073count will indicate the number of bytes actually written in the buffer. 4074It can (and very often will) be smaller than the input value, since the 4075buffer is only filled until 16 bytes of clean values are found (which 4076are then not copied in the buffer). Since a CMMA migration block needs 4077the base address and the length, for a total of 16 bytes, we will send 4078back some clean data if there is some dirty data afterwards, as long as 4079the size of the clean data does not exceed the size of the header. This 4080allows to minimize the amount of data to be saved or transferred over 4081the network at the expense of more roundtrips to userspace. The next 4082invocation of the ioctl will skip over all the clean values, saving 4083potentially more than just the 16 bytes we found. 4084 4085If KVM_S390_CMMA_PEEK is set: 4086the existing storage attributes are read even when not in migration 4087mode, and no other action is performed; 4088 4089the output start_gfn will be equal to the input start_gfn, 4090 4091the output count will be equal to the input count, except if the end of 4092memory has been reached. 4093 4094In both cases: 4095the field "remaining" will indicate the total number of dirty CMMA values 4096still remaining, or 0 if KVM_S390_CMMA_PEEK is set and migration mode is 4097not enabled. 4098 4099mask is unused. 4100 4101values points to the userspace buffer where the result will be stored. 4102 4103This ioctl can fail with -ENOMEM if not enough memory can be allocated to 4104complete the task, with -ENXIO if CMMA is not enabled, with -EINVAL if 4105KVM_S390_CMMA_PEEK is not set but migration mode was not enabled, with 4106-EFAULT if the userspace address is invalid or if no page table is 4107present for the addresses (e.g. when using hugepages). 4108 41094.108 KVM_S390_SET_CMMA_BITS 4110---------------------------- 4111 4112:Capability: KVM_CAP_S390_CMMA_MIGRATION 4113:Architectures: s390 4114:Type: vm ioctl 4115:Parameters: struct kvm_s390_cmma_log (in) 4116:Returns: 0 on success, a negative value on error 4117 4118This ioctl is used to set the values of the CMMA bits on the s390 4119architecture. It is meant to be used during live migration to restore 4120the CMMA values, but there are no restrictions on its use. 4121The ioctl takes parameters via the kvm_s390_cmma_values struct. 4122Each CMMA value takes up one byte. 4123 4124:: 4125 4126 struct kvm_s390_cmma_log { 4127 __u64 start_gfn; 4128 __u32 count; 4129 __u32 flags; 4130 union { 4131 __u64 remaining; 4132 __u64 mask; 4133 }; 4134 __u64 values; 4135 }; 4136 4137start_gfn indicates the starting guest frame number, 4138 4139count indicates how many values are to be considered in the buffer, 4140 4141flags is not used and must be 0. 4142 4143mask indicates which PGSTE bits are to be considered. 4144 4145remaining is not used. 4146 4147values points to the buffer in userspace where to store the values. 4148 4149This ioctl can fail with -ENOMEM if not enough memory can be allocated to 4150complete the task, with -ENXIO if CMMA is not enabled, with -EINVAL if 4151the count field is too large (e.g. more than KVM_S390_CMMA_SIZE_MAX) or 4152if the flags field was not 0, with -EFAULT if the userspace address is 4153invalid, if invalid pages are written to (e.g. after the end of memory) 4154or if no page table is present for the addresses (e.g. when using 4155hugepages). 4156 41574.109 KVM_PPC_GET_CPU_CHAR 4158-------------------------- 4159 4160:Capability: KVM_CAP_PPC_GET_CPU_CHAR 4161:Architectures: powerpc 4162:Type: vm ioctl 4163:Parameters: struct kvm_ppc_cpu_char (out) 4164:Returns: 0 on successful completion, 4165 -EFAULT if struct kvm_ppc_cpu_char cannot be written 4166 4167This ioctl gives userspace information about certain characteristics 4168of the CPU relating to speculative execution of instructions and 4169possible information leakage resulting from speculative execution (see 4170CVE-2017-5715, CVE-2017-5753 and CVE-2017-5754). The information is 4171returned in struct kvm_ppc_cpu_char, which looks like this:: 4172 4173 struct kvm_ppc_cpu_char { 4174 __u64 character; /* characteristics of the CPU */ 4175 __u64 behaviour; /* recommended software behaviour */ 4176 __u64 character_mask; /* valid bits in character */ 4177 __u64 behaviour_mask; /* valid bits in behaviour */ 4178 }; 4179 4180For extensibility, the character_mask and behaviour_mask fields 4181indicate which bits of character and behaviour have been filled in by 4182the kernel. If the set of defined bits is extended in future then 4183userspace will be able to tell whether it is running on a kernel that 4184knows about the new bits. 4185 4186The character field describes attributes of the CPU which can help 4187with preventing inadvertent information disclosure - specifically, 4188whether there is an instruction to flash-invalidate the L1 data cache 4189(ori 30,30,0 or mtspr SPRN_TRIG2,rN), whether the L1 data cache is set 4190to a mode where entries can only be used by the thread that created 4191them, whether the bcctr[l] instruction prevents speculation, and 4192whether a speculation barrier instruction (ori 31,31,0) is provided. 4193 4194The behaviour field describes actions that software should take to 4195prevent inadvertent information disclosure, and thus describes which 4196vulnerabilities the hardware is subject to; specifically whether the 4197L1 data cache should be flushed when returning to user mode from the 4198kernel, and whether a speculation barrier should be placed between an 4199array bounds check and the array access. 4200 4201These fields use the same bit definitions as the new 4202H_GET_CPU_CHARACTERISTICS hypercall. 4203 42044.110 KVM_MEMORY_ENCRYPT_OP 4205--------------------------- 4206 4207:Capability: basic 4208:Architectures: x86 4209:Type: vm 4210:Parameters: an opaque platform specific structure (in/out) 4211:Returns: 0 on success; -1 on error 4212 4213If the platform supports creating encrypted VMs then this ioctl can be used 4214for issuing platform-specific memory encryption commands to manage those 4215encrypted VMs. 4216 4217Currently, this ioctl is used for issuing Secure Encrypted Virtualization 4218(SEV) commands on AMD Processors. The SEV commands are defined in 4219Documentation/virt/kvm/amd-memory-encryption.rst. 4220 42214.111 KVM_MEMORY_ENCRYPT_REG_REGION 4222----------------------------------- 4223 4224:Capability: basic 4225:Architectures: x86 4226:Type: system 4227:Parameters: struct kvm_enc_region (in) 4228:Returns: 0 on success; -1 on error 4229 4230This ioctl can be used to register a guest memory region which may 4231contain encrypted data (e.g. guest RAM, SMRAM etc). 4232 4233It is used in the SEV-enabled guest. When encryption is enabled, a guest 4234memory region may contain encrypted data. The SEV memory encryption 4235engine uses a tweak such that two identical plaintext pages, each at 4236different locations will have differing ciphertexts. So swapping or 4237moving ciphertext of those pages will not result in plaintext being 4238swapped. So relocating (or migrating) physical backing pages for the SEV 4239guest will require some additional steps. 4240 4241Note: The current SEV key management spec does not provide commands to 4242swap or migrate (move) ciphertext pages. Hence, for now we pin the guest 4243memory region registered with the ioctl. 4244 42454.112 KVM_MEMORY_ENCRYPT_UNREG_REGION 4246------------------------------------- 4247 4248:Capability: basic 4249:Architectures: x86 4250:Type: system 4251:Parameters: struct kvm_enc_region (in) 4252:Returns: 0 on success; -1 on error 4253 4254This ioctl can be used to unregister the guest memory region registered 4255with KVM_MEMORY_ENCRYPT_REG_REGION ioctl above. 4256 42574.113 KVM_HYPERV_EVENTFD 4258------------------------ 4259 4260:Capability: KVM_CAP_HYPERV_EVENTFD 4261:Architectures: x86 4262:Type: vm ioctl 4263:Parameters: struct kvm_hyperv_eventfd (in) 4264 4265This ioctl (un)registers an eventfd to receive notifications from the guest on 4266the specified Hyper-V connection id through the SIGNAL_EVENT hypercall, without 4267causing a user exit. SIGNAL_EVENT hypercall with non-zero event flag number 4268(bits 24-31) still triggers a KVM_EXIT_HYPERV_HCALL user exit. 4269 4270:: 4271 4272 struct kvm_hyperv_eventfd { 4273 __u32 conn_id; 4274 __s32 fd; 4275 __u32 flags; 4276 __u32 padding[3]; 4277 }; 4278 4279The conn_id field should fit within 24 bits:: 4280 4281 #define KVM_HYPERV_CONN_ID_MASK 0x00ffffff 4282 4283The acceptable values for the flags field are:: 4284 4285 #define KVM_HYPERV_EVENTFD_DEASSIGN (1 << 0) 4286 4287:Returns: 0 on success, 4288 -EINVAL if conn_id or flags is outside the allowed range, 4289 -ENOENT on deassign if the conn_id isn't registered, 4290 -EEXIST on assign if the conn_id is already registered 4291 42924.114 KVM_GET_NESTED_STATE 4293-------------------------- 4294 4295:Capability: KVM_CAP_NESTED_STATE 4296:Architectures: x86 4297:Type: vcpu ioctl 4298:Parameters: struct kvm_nested_state (in/out) 4299:Returns: 0 on success, -1 on error 4300 4301Errors: 4302 4303 ===== ============================================================= 4304 E2BIG the total state size exceeds the value of 'size' specified by 4305 the user; the size required will be written into size. 4306 ===== ============================================================= 4307 4308:: 4309 4310 struct kvm_nested_state { 4311 __u16 flags; 4312 __u16 format; 4313 __u32 size; 4314 4315 union { 4316 struct kvm_vmx_nested_state_hdr vmx; 4317 struct kvm_svm_nested_state_hdr svm; 4318 4319 /* Pad the header to 128 bytes. */ 4320 __u8 pad[120]; 4321 } hdr; 4322 4323 union { 4324 struct kvm_vmx_nested_state_data vmx[0]; 4325 struct kvm_svm_nested_state_data svm[0]; 4326 } data; 4327 }; 4328 4329 #define KVM_STATE_NESTED_GUEST_MODE 0x00000001 4330 #define KVM_STATE_NESTED_RUN_PENDING 0x00000002 4331 #define KVM_STATE_NESTED_EVMCS 0x00000004 4332 4333 #define KVM_STATE_NESTED_FORMAT_VMX 0 4334 #define KVM_STATE_NESTED_FORMAT_SVM 1 4335 4336 #define KVM_STATE_NESTED_VMX_VMCS_SIZE 0x1000 4337 4338 #define KVM_STATE_NESTED_VMX_SMM_GUEST_MODE 0x00000001 4339 #define KVM_STATE_NESTED_VMX_SMM_VMXON 0x00000002 4340 4341 #define KVM_STATE_VMX_PREEMPTION_TIMER_DEADLINE 0x00000001 4342 4343 struct kvm_vmx_nested_state_hdr { 4344 __u64 vmxon_pa; 4345 __u64 vmcs12_pa; 4346 4347 struct { 4348 __u16 flags; 4349 } smm; 4350 4351 __u32 flags; 4352 __u64 preemption_timer_deadline; 4353 }; 4354 4355 struct kvm_vmx_nested_state_data { 4356 __u8 vmcs12[KVM_STATE_NESTED_VMX_VMCS_SIZE]; 4357 __u8 shadow_vmcs12[KVM_STATE_NESTED_VMX_VMCS_SIZE]; 4358 }; 4359 4360This ioctl copies the vcpu's nested virtualization state from the kernel to 4361userspace. 4362 4363The maximum size of the state can be retrieved by passing KVM_CAP_NESTED_STATE 4364to the KVM_CHECK_EXTENSION ioctl(). 4365 43664.115 KVM_SET_NESTED_STATE 4367-------------------------- 4368 4369:Capability: KVM_CAP_NESTED_STATE 4370:Architectures: x86 4371:Type: vcpu ioctl 4372:Parameters: struct kvm_nested_state (in) 4373:Returns: 0 on success, -1 on error 4374 4375This copies the vcpu's kvm_nested_state struct from userspace to the kernel. 4376For the definition of struct kvm_nested_state, see KVM_GET_NESTED_STATE. 4377 43784.116 KVM_(UN)REGISTER_COALESCED_MMIO 4379------------------------------------- 4380 4381:Capability: KVM_CAP_COALESCED_MMIO (for coalesced mmio) 4382 KVM_CAP_COALESCED_PIO (for coalesced pio) 4383:Architectures: all 4384:Type: vm ioctl 4385:Parameters: struct kvm_coalesced_mmio_zone 4386:Returns: 0 on success, < 0 on error 4387 4388Coalesced I/O is a performance optimization that defers hardware 4389register write emulation so that userspace exits are avoided. It is 4390typically used to reduce the overhead of emulating frequently accessed 4391hardware registers. 4392 4393When a hardware register is configured for coalesced I/O, write accesses 4394do not exit to userspace and their value is recorded in a ring buffer 4395that is shared between kernel and userspace. 4396 4397Coalesced I/O is used if one or more write accesses to a hardware 4398register can be deferred until a read or a write to another hardware 4399register on the same device. This last access will cause a vmexit and 4400userspace will process accesses from the ring buffer before emulating 4401it. That will avoid exiting to userspace on repeated writes. 4402 4403Coalesced pio is based on coalesced mmio. There is little difference 4404between coalesced mmio and pio except that coalesced pio records accesses 4405to I/O ports. 4406 44074.117 KVM_CLEAR_DIRTY_LOG (vm ioctl) 4408------------------------------------ 4409 4410:Capability: KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 4411:Architectures: x86, arm, arm64, mips 4412:Type: vm ioctl 4413:Parameters: struct kvm_clear_dirty_log (in) 4414:Returns: 0 on success, -1 on error 4415 4416:: 4417 4418 /* for KVM_CLEAR_DIRTY_LOG */ 4419 struct kvm_clear_dirty_log { 4420 __u32 slot; 4421 __u32 num_pages; 4422 __u64 first_page; 4423 union { 4424 void __user *dirty_bitmap; /* one bit per page */ 4425 __u64 padding; 4426 }; 4427 }; 4428 4429The ioctl clears the dirty status of pages in a memory slot, according to 4430the bitmap that is passed in struct kvm_clear_dirty_log's dirty_bitmap 4431field. Bit 0 of the bitmap corresponds to page "first_page" in the 4432memory slot, and num_pages is the size in bits of the input bitmap. 4433first_page must be a multiple of 64; num_pages must also be a multiple of 443464 unless first_page + num_pages is the size of the memory slot. For each 4435bit that is set in the input bitmap, the corresponding page is marked "clean" 4436in KVM's dirty bitmap, and dirty tracking is re-enabled for that page 4437(for example via write-protection, or by clearing the dirty bit in 4438a page table entry). 4439 4440If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of slot field specifies 4441the address space for which you want to clear the dirty status. See 4442KVM_SET_USER_MEMORY_REGION for details on the usage of slot field. 4443 4444This ioctl is mostly useful when KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 4445is enabled; for more information, see the description of the capability. 4446However, it can always be used as long as KVM_CHECK_EXTENSION confirms 4447that KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 is present. 4448 44494.118 KVM_GET_SUPPORTED_HV_CPUID 4450-------------------------------- 4451 4452:Capability: KVM_CAP_HYPERV_CPUID (vcpu), KVM_CAP_SYS_HYPERV_CPUID (system) 4453:Architectures: x86 4454:Type: system ioctl, vcpu ioctl 4455:Parameters: struct kvm_cpuid2 (in/out) 4456:Returns: 0 on success, -1 on error 4457 4458:: 4459 4460 struct kvm_cpuid2 { 4461 __u32 nent; 4462 __u32 padding; 4463 struct kvm_cpuid_entry2 entries[0]; 4464 }; 4465 4466 struct kvm_cpuid_entry2 { 4467 __u32 function; 4468 __u32 index; 4469 __u32 flags; 4470 __u32 eax; 4471 __u32 ebx; 4472 __u32 ecx; 4473 __u32 edx; 4474 __u32 padding[3]; 4475 }; 4476 4477This ioctl returns x86 cpuid features leaves related to Hyper-V emulation in 4478KVM. Userspace can use the information returned by this ioctl to construct 4479cpuid information presented to guests consuming Hyper-V enlightenments (e.g. 4480Windows or Hyper-V guests). 4481 4482CPUID feature leaves returned by this ioctl are defined by Hyper-V Top Level 4483Functional Specification (TLFS). These leaves can't be obtained with 4484KVM_GET_SUPPORTED_CPUID ioctl because some of them intersect with KVM feature 4485leaves (0x40000000, 0x40000001). 4486 4487Currently, the following list of CPUID leaves are returned: 4488 4489 - HYPERV_CPUID_VENDOR_AND_MAX_FUNCTIONS 4490 - HYPERV_CPUID_INTERFACE 4491 - HYPERV_CPUID_VERSION 4492 - HYPERV_CPUID_FEATURES 4493 - HYPERV_CPUID_ENLIGHTMENT_INFO 4494 - HYPERV_CPUID_IMPLEMENT_LIMITS 4495 - HYPERV_CPUID_NESTED_FEATURES 4496 - HYPERV_CPUID_SYNDBG_VENDOR_AND_MAX_FUNCTIONS 4497 - HYPERV_CPUID_SYNDBG_INTERFACE 4498 - HYPERV_CPUID_SYNDBG_PLATFORM_CAPABILITIES 4499 4500Userspace invokes KVM_GET_SUPPORTED_HV_CPUID by passing a kvm_cpuid2 structure 4501with the 'nent' field indicating the number of entries in the variable-size 4502array 'entries'. If the number of entries is too low to describe all Hyper-V 4503feature leaves, an error (E2BIG) is returned. If the number is more or equal 4504to the number of Hyper-V feature leaves, the 'nent' field is adjusted to the 4505number of valid entries in the 'entries' array, which is then filled. 4506 4507'index' and 'flags' fields in 'struct kvm_cpuid_entry2' are currently reserved, 4508userspace should not expect to get any particular value there. 4509 4510Note, vcpu version of KVM_GET_SUPPORTED_HV_CPUID is currently deprecated. Unlike 4511system ioctl which exposes all supported feature bits unconditionally, vcpu 4512version has the following quirks: 4513 4514- HYPERV_CPUID_NESTED_FEATURES leaf and HV_X64_ENLIGHTENED_VMCS_RECOMMENDED 4515 feature bit are only exposed when Enlightened VMCS was previously enabled 4516 on the corresponding vCPU (KVM_CAP_HYPERV_ENLIGHTENED_VMCS). 4517- HV_STIMER_DIRECT_MODE_AVAILABLE bit is only exposed with in-kernel LAPIC. 4518 (presumes KVM_CREATE_IRQCHIP has already been called). 4519 45204.119 KVM_ARM_VCPU_FINALIZE 4521--------------------------- 4522 4523:Architectures: arm, arm64 4524:Type: vcpu ioctl 4525:Parameters: int feature (in) 4526:Returns: 0 on success, -1 on error 4527 4528Errors: 4529 4530 ====== ============================================================== 4531 EPERM feature not enabled, needs configuration, or already finalized 4532 EINVAL feature unknown or not present 4533 ====== ============================================================== 4534 4535Recognised values for feature: 4536 4537 ===== =========================================== 4538 arm64 KVM_ARM_VCPU_SVE (requires KVM_CAP_ARM_SVE) 4539 ===== =========================================== 4540 4541Finalizes the configuration of the specified vcpu feature. 4542 4543The vcpu must already have been initialised, enabling the affected feature, by 4544means of a successful KVM_ARM_VCPU_INIT call with the appropriate flag set in 4545features[]. 4546 4547For affected vcpu features, this is a mandatory step that must be performed 4548before the vcpu is fully usable. 4549 4550Between KVM_ARM_VCPU_INIT and KVM_ARM_VCPU_FINALIZE, the feature may be 4551configured by use of ioctls such as KVM_SET_ONE_REG. The exact configuration 4552that should be performaned and how to do it are feature-dependent. 4553 4554Other calls that depend on a particular feature being finalized, such as 4555KVM_RUN, KVM_GET_REG_LIST, KVM_GET_ONE_REG and KVM_SET_ONE_REG, will fail with 4556-EPERM unless the feature has already been finalized by means of a 4557KVM_ARM_VCPU_FINALIZE call. 4558 4559See KVM_ARM_VCPU_INIT for details of vcpu features that require finalization 4560using this ioctl. 4561 45624.120 KVM_SET_PMU_EVENT_FILTER 4563------------------------------ 4564 4565:Capability: KVM_CAP_PMU_EVENT_FILTER 4566:Architectures: x86 4567:Type: vm ioctl 4568:Parameters: struct kvm_pmu_event_filter (in) 4569:Returns: 0 on success, -1 on error 4570 4571:: 4572 4573 struct kvm_pmu_event_filter { 4574 __u32 action; 4575 __u32 nevents; 4576 __u32 fixed_counter_bitmap; 4577 __u32 flags; 4578 __u32 pad[4]; 4579 __u64 events[0]; 4580 }; 4581 4582This ioctl restricts the set of PMU events that the guest can program. 4583The argument holds a list of events which will be allowed or denied. 4584The eventsel+umask of each event the guest attempts to program is compared 4585against the events field to determine whether the guest should have access. 4586The events field only controls general purpose counters; fixed purpose 4587counters are controlled by the fixed_counter_bitmap. 4588 4589No flags are defined yet, the field must be zero. 4590 4591Valid values for 'action':: 4592 4593 #define KVM_PMU_EVENT_ALLOW 0 4594 #define KVM_PMU_EVENT_DENY 1 4595 45964.121 KVM_PPC_SVM_OFF 4597--------------------- 4598 4599:Capability: basic 4600:Architectures: powerpc 4601:Type: vm ioctl 4602:Parameters: none 4603:Returns: 0 on successful completion, 4604 4605Errors: 4606 4607 ====== ================================================================ 4608 EINVAL if ultravisor failed to terminate the secure guest 4609 ENOMEM if hypervisor failed to allocate new radix page tables for guest 4610 ====== ================================================================ 4611 4612This ioctl is used to turn off the secure mode of the guest or transition 4613the guest from secure mode to normal mode. This is invoked when the guest 4614is reset. This has no effect if called for a normal guest. 4615 4616This ioctl issues an ultravisor call to terminate the secure guest, 4617unpins the VPA pages and releases all the device pages that are used to 4618track the secure pages by hypervisor. 4619 46204.122 KVM_S390_NORMAL_RESET 4621--------------------------- 4622 4623:Capability: KVM_CAP_S390_VCPU_RESETS 4624:Architectures: s390 4625:Type: vcpu ioctl 4626:Parameters: none 4627:Returns: 0 4628 4629This ioctl resets VCPU registers and control structures according to 4630the cpu reset definition in the POP (Principles Of Operation). 4631 46324.123 KVM_S390_INITIAL_RESET 4633---------------------------- 4634 4635:Capability: none 4636:Architectures: s390 4637:Type: vcpu ioctl 4638:Parameters: none 4639:Returns: 0 4640 4641This ioctl resets VCPU registers and control structures according to 4642the initial cpu reset definition in the POP. However, the cpu is not 4643put into ESA mode. This reset is a superset of the normal reset. 4644 46454.124 KVM_S390_CLEAR_RESET 4646-------------------------- 4647 4648:Capability: KVM_CAP_S390_VCPU_RESETS 4649:Architectures: s390 4650:Type: vcpu ioctl 4651:Parameters: none 4652:Returns: 0 4653 4654This ioctl resets VCPU registers and control structures according to 4655the clear cpu reset definition in the POP. However, the cpu is not put 4656into ESA mode. This reset is a superset of the initial reset. 4657 4658 46594.125 KVM_S390_PV_COMMAND 4660------------------------- 4661 4662:Capability: KVM_CAP_S390_PROTECTED 4663:Architectures: s390 4664:Type: vm ioctl 4665:Parameters: struct kvm_pv_cmd 4666:Returns: 0 on success, < 0 on error 4667 4668:: 4669 4670 struct kvm_pv_cmd { 4671 __u32 cmd; /* Command to be executed */ 4672 __u16 rc; /* Ultravisor return code */ 4673 __u16 rrc; /* Ultravisor return reason code */ 4674 __u64 data; /* Data or address */ 4675 __u32 flags; /* flags for future extensions. Must be 0 for now */ 4676 __u32 reserved[3]; 4677 }; 4678 4679cmd values: 4680 4681KVM_PV_ENABLE 4682 Allocate memory and register the VM with the Ultravisor, thereby 4683 donating memory to the Ultravisor that will become inaccessible to 4684 KVM. All existing CPUs are converted to protected ones. After this 4685 command has succeeded, any CPU added via hotplug will become 4686 protected during its creation as well. 4687 4688 Errors: 4689 4690 ===== ============================= 4691 EINTR an unmasked signal is pending 4692 ===== ============================= 4693 4694KVM_PV_DISABLE 4695 4696 Deregister the VM from the Ultravisor and reclaim the memory that 4697 had been donated to the Ultravisor, making it usable by the kernel 4698 again. All registered VCPUs are converted back to non-protected 4699 ones. 4700 4701KVM_PV_VM_SET_SEC_PARMS 4702 Pass the image header from VM memory to the Ultravisor in 4703 preparation of image unpacking and verification. 4704 4705KVM_PV_VM_UNPACK 4706 Unpack (protect and decrypt) a page of the encrypted boot image. 4707 4708KVM_PV_VM_VERIFY 4709 Verify the integrity of the unpacked image. Only if this succeeds, 4710 KVM is allowed to start protected VCPUs. 4711 47124.126 KVM_X86_SET_MSR_FILTER 4713---------------------------- 4714 4715:Capability: KVM_X86_SET_MSR_FILTER 4716:Architectures: x86 4717:Type: vm ioctl 4718:Parameters: struct kvm_msr_filter 4719:Returns: 0 on success, < 0 on error 4720 4721:: 4722 4723 struct kvm_msr_filter_range { 4724 #define KVM_MSR_FILTER_READ (1 << 0) 4725 #define KVM_MSR_FILTER_WRITE (1 << 1) 4726 __u32 flags; 4727 __u32 nmsrs; /* number of msrs in bitmap */ 4728 __u32 base; /* MSR index the bitmap starts at */ 4729 __u8 *bitmap; /* a 1 bit allows the operations in flags, 0 denies */ 4730 }; 4731 4732 #define KVM_MSR_FILTER_MAX_RANGES 16 4733 struct kvm_msr_filter { 4734 #define KVM_MSR_FILTER_DEFAULT_ALLOW (0 << 0) 4735 #define KVM_MSR_FILTER_DEFAULT_DENY (1 << 0) 4736 __u32 flags; 4737 struct kvm_msr_filter_range ranges[KVM_MSR_FILTER_MAX_RANGES]; 4738 }; 4739 4740flags values for ``struct kvm_msr_filter_range``: 4741 4742``KVM_MSR_FILTER_READ`` 4743 4744 Filter read accesses to MSRs using the given bitmap. A 0 in the bitmap 4745 indicates that a read should immediately fail, while a 1 indicates that 4746 a read for a particular MSR should be handled regardless of the default 4747 filter action. 4748 4749``KVM_MSR_FILTER_WRITE`` 4750 4751 Filter write accesses to MSRs using the given bitmap. A 0 in the bitmap 4752 indicates that a write should immediately fail, while a 1 indicates that 4753 a write for a particular MSR should be handled regardless of the default 4754 filter action. 4755 4756``KVM_MSR_FILTER_READ | KVM_MSR_FILTER_WRITE`` 4757 4758 Filter both read and write accesses to MSRs using the given bitmap. A 0 4759 in the bitmap indicates that both reads and writes should immediately fail, 4760 while a 1 indicates that reads and writes for a particular MSR are not 4761 filtered by this range. 4762 4763flags values for ``struct kvm_msr_filter``: 4764 4765``KVM_MSR_FILTER_DEFAULT_ALLOW`` 4766 4767 If no filter range matches an MSR index that is getting accessed, KVM will 4768 fall back to allowing access to the MSR. 4769 4770``KVM_MSR_FILTER_DEFAULT_DENY`` 4771 4772 If no filter range matches an MSR index that is getting accessed, KVM will 4773 fall back to rejecting access to the MSR. In this mode, all MSRs that should 4774 be processed by KVM need to explicitly be marked as allowed in the bitmaps. 4775 4776This ioctl allows user space to define up to 16 bitmaps of MSR ranges to 4777specify whether a certain MSR access should be explicitly filtered for or not. 4778 4779If this ioctl has never been invoked, MSR accesses are not guarded and the 4780default KVM in-kernel emulation behavior is fully preserved. 4781 4782Calling this ioctl with an empty set of ranges (all nmsrs == 0) disables MSR 4783filtering. In that mode, ``KVM_MSR_FILTER_DEFAULT_DENY`` is invalid and causes 4784an error. 4785 4786As soon as the filtering is in place, every MSR access is processed through 4787the filtering except for accesses to the x2APIC MSRs (from 0x800 to 0x8ff); 4788x2APIC MSRs are always allowed, independent of the ``default_allow`` setting, 4789and their behavior depends on the ``X2APIC_ENABLE`` bit of the APIC base 4790register. 4791 4792If a bit is within one of the defined ranges, read and write accesses are 4793guarded by the bitmap's value for the MSR index if the kind of access 4794is included in the ``struct kvm_msr_filter_range`` flags. If no range 4795cover this particular access, the behavior is determined by the flags 4796field in the kvm_msr_filter struct: ``KVM_MSR_FILTER_DEFAULT_ALLOW`` 4797and ``KVM_MSR_FILTER_DEFAULT_DENY``. 4798 4799Each bitmap range specifies a range of MSRs to potentially allow access on. 4800The range goes from MSR index [base .. base+nmsrs]. The flags field 4801indicates whether reads, writes or both reads and writes are filtered 4802by setting a 1 bit in the bitmap for the corresponding MSR index. 4803 4804If an MSR access is not permitted through the filtering, it generates a 4805#GP inside the guest. When combined with KVM_CAP_X86_USER_SPACE_MSR, that 4806allows user space to deflect and potentially handle various MSR accesses 4807into user space. 4808 4809If a vCPU is in running state while this ioctl is invoked, the vCPU may 4810experience inconsistent filtering behavior on MSR accesses. 4811 48124.127 KVM_XEN_HVM_SET_ATTR 4813-------------------------- 4814 4815:Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO 4816:Architectures: x86 4817:Type: vm ioctl 4818:Parameters: struct kvm_xen_hvm_attr 4819:Returns: 0 on success, < 0 on error 4820 4821:: 4822 4823 struct kvm_xen_hvm_attr { 4824 __u16 type; 4825 __u16 pad[3]; 4826 union { 4827 __u8 long_mode; 4828 __u8 vector; 4829 struct { 4830 __u64 gfn; 4831 } shared_info; 4832 __u64 pad[4]; 4833 } u; 4834 }; 4835 4836type values: 4837 4838KVM_XEN_ATTR_TYPE_LONG_MODE 4839 Sets the ABI mode of the VM to 32-bit or 64-bit (long mode). This 4840 determines the layout of the shared info pages exposed to the VM. 4841 4842KVM_XEN_ATTR_TYPE_SHARED_INFO 4843 Sets the guest physical frame number at which the Xen "shared info" 4844 page resides. Note that although Xen places vcpu_info for the first 4845 32 vCPUs in the shared_info page, KVM does not automatically do so 4846 and instead requires that KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO be used 4847 explicitly even when the vcpu_info for a given vCPU resides at the 4848 "default" location in the shared_info page. This is because KVM is 4849 not aware of the Xen CPU id which is used as the index into the 4850 vcpu_info[] array, so cannot know the correct default location. 4851 4852KVM_XEN_ATTR_TYPE_UPCALL_VECTOR 4853 Sets the exception vector used to deliver Xen event channel upcalls. 4854 48554.128 KVM_XEN_HVM_GET_ATTR 4856-------------------------- 4857 4858:Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO 4859:Architectures: x86 4860:Type: vm ioctl 4861:Parameters: struct kvm_xen_hvm_attr 4862:Returns: 0 on success, < 0 on error 4863 4864Allows Xen VM attributes to be read. For the structure and types, 4865see KVM_XEN_HVM_SET_ATTR above. 4866 48674.129 KVM_XEN_VCPU_SET_ATTR 4868--------------------------- 4869 4870:Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO 4871:Architectures: x86 4872:Type: vcpu ioctl 4873:Parameters: struct kvm_xen_vcpu_attr 4874:Returns: 0 on success, < 0 on error 4875 4876:: 4877 4878 struct kvm_xen_vcpu_attr { 4879 __u16 type; 4880 __u16 pad[3]; 4881 union { 4882 __u64 gpa; 4883 __u64 pad[4]; 4884 struct { 4885 __u64 state; 4886 __u64 state_entry_time; 4887 __u64 time_running; 4888 __u64 time_runnable; 4889 __u64 time_blocked; 4890 __u64 time_offline; 4891 } runstate; 4892 } u; 4893 }; 4894 4895type values: 4896 4897KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO 4898 Sets the guest physical address of the vcpu_info for a given vCPU. 4899 4900KVM_XEN_VCPU_ATTR_TYPE_VCPU_TIME_INFO 4901 Sets the guest physical address of an additional pvclock structure 4902 for a given vCPU. This is typically used for guest vsyscall support. 4903 4904KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADDR 4905 Sets the guest physical address of the vcpu_runstate_info for a given 4906 vCPU. This is how a Xen guest tracks CPU state such as steal time. 4907 4908KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_CURRENT 4909 Sets the runstate (RUNSTATE_running/_runnable/_blocked/_offline) of 4910 the given vCPU from the .u.runstate.state member of the structure. 4911 KVM automatically accounts running and runnable time but blocked 4912 and offline states are only entered explicitly. 4913 4914KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_DATA 4915 Sets all fields of the vCPU runstate data from the .u.runstate member 4916 of the structure, including the current runstate. The state_entry_time 4917 must equal the sum of the other four times. 4918 4919KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADJUST 4920 This *adds* the contents of the .u.runstate members of the structure 4921 to the corresponding members of the given vCPU's runstate data, thus 4922 permitting atomic adjustments to the runstate times. The adjustment 4923 to the state_entry_time must equal the sum of the adjustments to the 4924 other four times. The state field must be set to -1, or to a valid 4925 runstate value (RUNSTATE_running, RUNSTATE_runnable, RUNSTATE_blocked 4926 or RUNSTATE_offline) to set the current accounted state as of the 4927 adjusted state_entry_time. 4928 49294.130 KVM_XEN_VCPU_GET_ATTR 4930--------------------------- 4931 4932:Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO 4933:Architectures: x86 4934:Type: vcpu ioctl 4935:Parameters: struct kvm_xen_vcpu_attr 4936:Returns: 0 on success, < 0 on error 4937 4938Allows Xen vCPU attributes to be read. For the structure and types, 4939see KVM_XEN_VCPU_SET_ATTR above. 4940 4941The KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADJUST type may not be used 4942with the KVM_XEN_VCPU_GET_ATTR ioctl. 4943 49445. The kvm_run structure 4945======================== 4946 4947Application code obtains a pointer to the kvm_run structure by 4948mmap()ing a vcpu fd. From that point, application code can control 4949execution by changing fields in kvm_run prior to calling the KVM_RUN 4950ioctl, and obtain information about the reason KVM_RUN returned by 4951looking up structure members. 4952 4953:: 4954 4955 struct kvm_run { 4956 /* in */ 4957 __u8 request_interrupt_window; 4958 4959Request that KVM_RUN return when it becomes possible to inject external 4960interrupts into the guest. Useful in conjunction with KVM_INTERRUPT. 4961 4962:: 4963 4964 __u8 immediate_exit; 4965 4966This field is polled once when KVM_RUN starts; if non-zero, KVM_RUN 4967exits immediately, returning -EINTR. In the common scenario where a 4968signal is used to "kick" a VCPU out of KVM_RUN, this field can be used 4969to avoid usage of KVM_SET_SIGNAL_MASK, which has worse scalability. 4970Rather than blocking the signal outside KVM_RUN, userspace can set up 4971a signal handler that sets run->immediate_exit to a non-zero value. 4972 4973This field is ignored if KVM_CAP_IMMEDIATE_EXIT is not available. 4974 4975:: 4976 4977 __u8 padding1[6]; 4978 4979 /* out */ 4980 __u32 exit_reason; 4981 4982When KVM_RUN has returned successfully (return value 0), this informs 4983application code why KVM_RUN has returned. Allowable values for this 4984field are detailed below. 4985 4986:: 4987 4988 __u8 ready_for_interrupt_injection; 4989 4990If request_interrupt_window has been specified, this field indicates 4991an interrupt can be injected now with KVM_INTERRUPT. 4992 4993:: 4994 4995 __u8 if_flag; 4996 4997The value of the current interrupt flag. Only valid if in-kernel 4998local APIC is not used. 4999 5000:: 5001 5002 __u16 flags; 5003 5004More architecture-specific flags detailing state of the VCPU that may 5005affect the device's behavior. Current defined flags:: 5006 5007 /* x86, set if the VCPU is in system management mode */ 5008 #define KVM_RUN_X86_SMM (1 << 0) 5009 /* x86, set if bus lock detected in VM */ 5010 #define KVM_RUN_BUS_LOCK (1 << 1) 5011 5012:: 5013 5014 /* in (pre_kvm_run), out (post_kvm_run) */ 5015 __u64 cr8; 5016 5017The value of the cr8 register. Only valid if in-kernel local APIC is 5018not used. Both input and output. 5019 5020:: 5021 5022 __u64 apic_base; 5023 5024The value of the APIC BASE msr. Only valid if in-kernel local 5025APIC is not used. Both input and output. 5026 5027:: 5028 5029 union { 5030 /* KVM_EXIT_UNKNOWN */ 5031 struct { 5032 __u64 hardware_exit_reason; 5033 } hw; 5034 5035If exit_reason is KVM_EXIT_UNKNOWN, the vcpu has exited due to unknown 5036reasons. Further architecture-specific information is available in 5037hardware_exit_reason. 5038 5039:: 5040 5041 /* KVM_EXIT_FAIL_ENTRY */ 5042 struct { 5043 __u64 hardware_entry_failure_reason; 5044 __u32 cpu; /* if KVM_LAST_CPU */ 5045 } fail_entry; 5046 5047If exit_reason is KVM_EXIT_FAIL_ENTRY, the vcpu could not be run due 5048to unknown reasons. Further architecture-specific information is 5049available in hardware_entry_failure_reason. 5050 5051:: 5052 5053 /* KVM_EXIT_EXCEPTION */ 5054 struct { 5055 __u32 exception; 5056 __u32 error_code; 5057 } ex; 5058 5059Unused. 5060 5061:: 5062 5063 /* KVM_EXIT_IO */ 5064 struct { 5065 #define KVM_EXIT_IO_IN 0 5066 #define KVM_EXIT_IO_OUT 1 5067 __u8 direction; 5068 __u8 size; /* bytes */ 5069 __u16 port; 5070 __u32 count; 5071 __u64 data_offset; /* relative to kvm_run start */ 5072 } io; 5073 5074If exit_reason is KVM_EXIT_IO, then the vcpu has 5075executed a port I/O instruction which could not be satisfied by kvm. 5076data_offset describes where the data is located (KVM_EXIT_IO_OUT) or 5077where kvm expects application code to place the data for the next 5078KVM_RUN invocation (KVM_EXIT_IO_IN). Data format is a packed array. 5079 5080:: 5081 5082 /* KVM_EXIT_DEBUG */ 5083 struct { 5084 struct kvm_debug_exit_arch arch; 5085 } debug; 5086 5087If the exit_reason is KVM_EXIT_DEBUG, then a vcpu is processing a debug event 5088for which architecture specific information is returned. 5089 5090:: 5091 5092 /* KVM_EXIT_MMIO */ 5093 struct { 5094 __u64 phys_addr; 5095 __u8 data[8]; 5096 __u32 len; 5097 __u8 is_write; 5098 } mmio; 5099 5100If exit_reason is KVM_EXIT_MMIO, then the vcpu has 5101executed a memory-mapped I/O instruction which could not be satisfied 5102by kvm. The 'data' member contains the written data if 'is_write' is 5103true, and should be filled by application code otherwise. 5104 5105The 'data' member contains, in its first 'len' bytes, the value as it would 5106appear if the VCPU performed a load or store of the appropriate width directly 5107to the byte array. 5108 5109.. note:: 5110 5111 For KVM_EXIT_IO, KVM_EXIT_MMIO, KVM_EXIT_OSI, KVM_EXIT_PAPR, KVM_EXIT_XEN, 5112 KVM_EXIT_EPR, KVM_EXIT_X86_RDMSR and KVM_EXIT_X86_WRMSR the corresponding 5113 operations are complete (and guest state is consistent) only after userspace 5114 has re-entered the kernel with KVM_RUN. The kernel side will first finish 5115 incomplete operations and then check for pending signals. 5116 5117 The pending state of the operation is not preserved in state which is 5118 visible to userspace, thus userspace should ensure that the operation is 5119 completed before performing a live migration. Userspace can re-enter the 5120 guest with an unmasked signal pending or with the immediate_exit field set 5121 to complete pending operations without allowing any further instructions 5122 to be executed. 5123 5124:: 5125 5126 /* KVM_EXIT_HYPERCALL */ 5127 struct { 5128 __u64 nr; 5129 __u64 args[6]; 5130 __u64 ret; 5131 __u32 longmode; 5132 __u32 pad; 5133 } hypercall; 5134 5135Unused. This was once used for 'hypercall to userspace'. To implement 5136such functionality, use KVM_EXIT_IO (x86) or KVM_EXIT_MMIO (all except s390). 5137 5138.. note:: KVM_EXIT_IO is significantly faster than KVM_EXIT_MMIO. 5139 5140:: 5141 5142 /* KVM_EXIT_TPR_ACCESS */ 5143 struct { 5144 __u64 rip; 5145 __u32 is_write; 5146 __u32 pad; 5147 } tpr_access; 5148 5149To be documented (KVM_TPR_ACCESS_REPORTING). 5150 5151:: 5152 5153 /* KVM_EXIT_S390_SIEIC */ 5154 struct { 5155 __u8 icptcode; 5156 __u64 mask; /* psw upper half */ 5157 __u64 addr; /* psw lower half */ 5158 __u16 ipa; 5159 __u32 ipb; 5160 } s390_sieic; 5161 5162s390 specific. 5163 5164:: 5165 5166 /* KVM_EXIT_S390_RESET */ 5167 #define KVM_S390_RESET_POR 1 5168 #define KVM_S390_RESET_CLEAR 2 5169 #define KVM_S390_RESET_SUBSYSTEM 4 5170 #define KVM_S390_RESET_CPU_INIT 8 5171 #define KVM_S390_RESET_IPL 16 5172 __u64 s390_reset_flags; 5173 5174s390 specific. 5175 5176:: 5177 5178 /* KVM_EXIT_S390_UCONTROL */ 5179 struct { 5180 __u64 trans_exc_code; 5181 __u32 pgm_code; 5182 } s390_ucontrol; 5183 5184s390 specific. A page fault has occurred for a user controlled virtual 5185machine (KVM_VM_S390_UNCONTROL) on it's host page table that cannot be 5186resolved by the kernel. 5187The program code and the translation exception code that were placed 5188in the cpu's lowcore are presented here as defined by the z Architecture 5189Principles of Operation Book in the Chapter for Dynamic Address Translation 5190(DAT) 5191 5192:: 5193 5194 /* KVM_EXIT_DCR */ 5195 struct { 5196 __u32 dcrn; 5197 __u32 data; 5198 __u8 is_write; 5199 } dcr; 5200 5201Deprecated - was used for 440 KVM. 5202 5203:: 5204 5205 /* KVM_EXIT_OSI */ 5206 struct { 5207 __u64 gprs[32]; 5208 } osi; 5209 5210MOL uses a special hypercall interface it calls 'OSI'. To enable it, we catch 5211hypercalls and exit with this exit struct that contains all the guest gprs. 5212 5213If exit_reason is KVM_EXIT_OSI, then the vcpu has triggered such a hypercall. 5214Userspace can now handle the hypercall and when it's done modify the gprs as 5215necessary. Upon guest entry all guest GPRs will then be replaced by the values 5216in this struct. 5217 5218:: 5219 5220 /* KVM_EXIT_PAPR_HCALL */ 5221 struct { 5222 __u64 nr; 5223 __u64 ret; 5224 __u64 args[9]; 5225 } papr_hcall; 5226 5227This is used on 64-bit PowerPC when emulating a pSeries partition, 5228e.g. with the 'pseries' machine type in qemu. It occurs when the 5229guest does a hypercall using the 'sc 1' instruction. The 'nr' field 5230contains the hypercall number (from the guest R3), and 'args' contains 5231the arguments (from the guest R4 - R12). Userspace should put the 5232return code in 'ret' and any extra returned values in args[]. 5233The possible hypercalls are defined in the Power Architecture Platform 5234Requirements (PAPR) document available from www.power.org (free 5235developer registration required to access it). 5236 5237:: 5238 5239 /* KVM_EXIT_S390_TSCH */ 5240 struct { 5241 __u16 subchannel_id; 5242 __u16 subchannel_nr; 5243 __u32 io_int_parm; 5244 __u32 io_int_word; 5245 __u32 ipb; 5246 __u8 dequeued; 5247 } s390_tsch; 5248 5249s390 specific. This exit occurs when KVM_CAP_S390_CSS_SUPPORT has been enabled 5250and TEST SUBCHANNEL was intercepted. If dequeued is set, a pending I/O 5251interrupt for the target subchannel has been dequeued and subchannel_id, 5252subchannel_nr, io_int_parm and io_int_word contain the parameters for that 5253interrupt. ipb is needed for instruction parameter decoding. 5254 5255:: 5256 5257 /* KVM_EXIT_EPR */ 5258 struct { 5259 __u32 epr; 5260 } epr; 5261 5262On FSL BookE PowerPC chips, the interrupt controller has a fast patch 5263interrupt acknowledge path to the core. When the core successfully 5264delivers an interrupt, it automatically populates the EPR register with 5265the interrupt vector number and acknowledges the interrupt inside 5266the interrupt controller. 5267 5268In case the interrupt controller lives in user space, we need to do 5269the interrupt acknowledge cycle through it to fetch the next to be 5270delivered interrupt vector using this exit. 5271 5272It gets triggered whenever both KVM_CAP_PPC_EPR are enabled and an 5273external interrupt has just been delivered into the guest. User space 5274should put the acknowledged interrupt vector into the 'epr' field. 5275 5276:: 5277 5278 /* KVM_EXIT_SYSTEM_EVENT */ 5279 struct { 5280 #define KVM_SYSTEM_EVENT_SHUTDOWN 1 5281 #define KVM_SYSTEM_EVENT_RESET 2 5282 #define KVM_SYSTEM_EVENT_CRASH 3 5283 __u32 type; 5284 __u64 flags; 5285 } system_event; 5286 5287If exit_reason is KVM_EXIT_SYSTEM_EVENT then the vcpu has triggered 5288a system-level event using some architecture specific mechanism (hypercall 5289or some special instruction). In case of ARM/ARM64, this is triggered using 5290HVC instruction based PSCI call from the vcpu. The 'type' field describes 5291the system-level event type. The 'flags' field describes architecture 5292specific flags for the system-level event. 5293 5294Valid values for 'type' are: 5295 5296 - KVM_SYSTEM_EVENT_SHUTDOWN -- the guest has requested a shutdown of the 5297 VM. Userspace is not obliged to honour this, and if it does honour 5298 this does not need to destroy the VM synchronously (ie it may call 5299 KVM_RUN again before shutdown finally occurs). 5300 - KVM_SYSTEM_EVENT_RESET -- the guest has requested a reset of the VM. 5301 As with SHUTDOWN, userspace can choose to ignore the request, or 5302 to schedule the reset to occur in the future and may call KVM_RUN again. 5303 - KVM_SYSTEM_EVENT_CRASH -- the guest crash occurred and the guest 5304 has requested a crash condition maintenance. Userspace can choose 5305 to ignore the request, or to gather VM memory core dump and/or 5306 reset/shutdown of the VM. 5307 5308:: 5309 5310 /* KVM_EXIT_IOAPIC_EOI */ 5311 struct { 5312 __u8 vector; 5313 } eoi; 5314 5315Indicates that the VCPU's in-kernel local APIC received an EOI for a 5316level-triggered IOAPIC interrupt. This exit only triggers when the 5317IOAPIC is implemented in userspace (i.e. KVM_CAP_SPLIT_IRQCHIP is enabled); 5318the userspace IOAPIC should process the EOI and retrigger the interrupt if 5319it is still asserted. Vector is the LAPIC interrupt vector for which the 5320EOI was received. 5321 5322:: 5323 5324 struct kvm_hyperv_exit { 5325 #define KVM_EXIT_HYPERV_SYNIC 1 5326 #define KVM_EXIT_HYPERV_HCALL 2 5327 #define KVM_EXIT_HYPERV_SYNDBG 3 5328 __u32 type; 5329 __u32 pad1; 5330 union { 5331 struct { 5332 __u32 msr; 5333 __u32 pad2; 5334 __u64 control; 5335 __u64 evt_page; 5336 __u64 msg_page; 5337 } synic; 5338 struct { 5339 __u64 input; 5340 __u64 result; 5341 __u64 params[2]; 5342 } hcall; 5343 struct { 5344 __u32 msr; 5345 __u32 pad2; 5346 __u64 control; 5347 __u64 status; 5348 __u64 send_page; 5349 __u64 recv_page; 5350 __u64 pending_page; 5351 } syndbg; 5352 } u; 5353 }; 5354 /* KVM_EXIT_HYPERV */ 5355 struct kvm_hyperv_exit hyperv; 5356 5357Indicates that the VCPU exits into userspace to process some tasks 5358related to Hyper-V emulation. 5359 5360Valid values for 'type' are: 5361 5362 - KVM_EXIT_HYPERV_SYNIC -- synchronously notify user-space about 5363 5364Hyper-V SynIC state change. Notification is used to remap SynIC 5365event/message pages and to enable/disable SynIC messages/events processing 5366in userspace. 5367 5368 - KVM_EXIT_HYPERV_SYNDBG -- synchronously notify user-space about 5369 5370Hyper-V Synthetic debugger state change. Notification is used to either update 5371the pending_page location or to send a control command (send the buffer located 5372in send_page or recv a buffer to recv_page). 5373 5374:: 5375 5376 /* KVM_EXIT_ARM_NISV */ 5377 struct { 5378 __u64 esr_iss; 5379 __u64 fault_ipa; 5380 } arm_nisv; 5381 5382Used on arm and arm64 systems. If a guest accesses memory not in a memslot, 5383KVM will typically return to userspace and ask it to do MMIO emulation on its 5384behalf. However, for certain classes of instructions, no instruction decode 5385(direction, length of memory access) is provided, and fetching and decoding 5386the instruction from the VM is overly complicated to live in the kernel. 5387 5388Historically, when this situation occurred, KVM would print a warning and kill 5389the VM. KVM assumed that if the guest accessed non-memslot memory, it was 5390trying to do I/O, which just couldn't be emulated, and the warning message was 5391phrased accordingly. However, what happened more often was that a guest bug 5392caused access outside the guest memory areas which should lead to a more 5393meaningful warning message and an external abort in the guest, if the access 5394did not fall within an I/O window. 5395 5396Userspace implementations can query for KVM_CAP_ARM_NISV_TO_USER, and enable 5397this capability at VM creation. Once this is done, these types of errors will 5398instead return to userspace with KVM_EXIT_ARM_NISV, with the valid bits from 5399the HSR (arm) and ESR_EL2 (arm64) in the esr_iss field, and the faulting IPA 5400in the fault_ipa field. Userspace can either fix up the access if it's 5401actually an I/O access by decoding the instruction from guest memory (if it's 5402very brave) and continue executing the guest, or it can decide to suspend, 5403dump, or restart the guest. 5404 5405Note that KVM does not skip the faulting instruction as it does for 5406KVM_EXIT_MMIO, but userspace has to emulate any change to the processing state 5407if it decides to decode and emulate the instruction. 5408 5409:: 5410 5411 /* KVM_EXIT_X86_RDMSR / KVM_EXIT_X86_WRMSR */ 5412 struct { 5413 __u8 error; /* user -> kernel */ 5414 __u8 pad[7]; 5415 __u32 reason; /* kernel -> user */ 5416 __u32 index; /* kernel -> user */ 5417 __u64 data; /* kernel <-> user */ 5418 } msr; 5419 5420Used on x86 systems. When the VM capability KVM_CAP_X86_USER_SPACE_MSR is 5421enabled, MSR accesses to registers that would invoke a #GP by KVM kernel code 5422will instead trigger a KVM_EXIT_X86_RDMSR exit for reads and KVM_EXIT_X86_WRMSR 5423exit for writes. 5424 5425The "reason" field specifies why the MSR trap occurred. User space will only 5426receive MSR exit traps when a particular reason was requested during through 5427ENABLE_CAP. Currently valid exit reasons are: 5428 5429 KVM_MSR_EXIT_REASON_UNKNOWN - access to MSR that is unknown to KVM 5430 KVM_MSR_EXIT_REASON_INVAL - access to invalid MSRs or reserved bits 5431 KVM_MSR_EXIT_REASON_FILTER - access blocked by KVM_X86_SET_MSR_FILTER 5432 5433For KVM_EXIT_X86_RDMSR, the "index" field tells user space which MSR the guest 5434wants to read. To respond to this request with a successful read, user space 5435writes the respective data into the "data" field and must continue guest 5436execution to ensure the read data is transferred into guest register state. 5437 5438If the RDMSR request was unsuccessful, user space indicates that with a "1" in 5439the "error" field. This will inject a #GP into the guest when the VCPU is 5440executed again. 5441 5442For KVM_EXIT_X86_WRMSR, the "index" field tells user space which MSR the guest 5443wants to write. Once finished processing the event, user space must continue 5444vCPU execution. If the MSR write was unsuccessful, user space also sets the 5445"error" field to "1". 5446 5447:: 5448 5449 5450 struct kvm_xen_exit { 5451 #define KVM_EXIT_XEN_HCALL 1 5452 __u32 type; 5453 union { 5454 struct { 5455 __u32 longmode; 5456 __u32 cpl; 5457 __u64 input; 5458 __u64 result; 5459 __u64 params[6]; 5460 } hcall; 5461 } u; 5462 }; 5463 /* KVM_EXIT_XEN */ 5464 struct kvm_hyperv_exit xen; 5465 5466Indicates that the VCPU exits into userspace to process some tasks 5467related to Xen emulation. 5468 5469Valid values for 'type' are: 5470 5471 - KVM_EXIT_XEN_HCALL -- synchronously notify user-space about Xen hypercall. 5472 Userspace is expected to place the hypercall result into the appropriate 5473 field before invoking KVM_RUN again. 5474 5475:: 5476 5477 /* Fix the size of the union. */ 5478 char padding[256]; 5479 }; 5480 5481 /* 5482 * shared registers between kvm and userspace. 5483 * kvm_valid_regs specifies the register classes set by the host 5484 * kvm_dirty_regs specified the register classes dirtied by userspace 5485 * struct kvm_sync_regs is architecture specific, as well as the 5486 * bits for kvm_valid_regs and kvm_dirty_regs 5487 */ 5488 __u64 kvm_valid_regs; 5489 __u64 kvm_dirty_regs; 5490 union { 5491 struct kvm_sync_regs regs; 5492 char padding[SYNC_REGS_SIZE_BYTES]; 5493 } s; 5494 5495If KVM_CAP_SYNC_REGS is defined, these fields allow userspace to access 5496certain guest registers without having to call SET/GET_*REGS. Thus we can 5497avoid some system call overhead if userspace has to handle the exit. 5498Userspace can query the validity of the structure by checking 5499kvm_valid_regs for specific bits. These bits are architecture specific 5500and usually define the validity of a groups of registers. (e.g. one bit 5501for general purpose registers) 5502 5503Please note that the kernel is allowed to use the kvm_run structure as the 5504primary storage for certain register types. Therefore, the kernel may use the 5505values in kvm_run even if the corresponding bit in kvm_dirty_regs is not set. 5506 5507:: 5508 5509 }; 5510 5511 5512 55136. Capabilities that can be enabled on vCPUs 5514============================================ 5515 5516There are certain capabilities that change the behavior of the virtual CPU or 5517the virtual machine when enabled. To enable them, please see section 4.37. 5518Below you can find a list of capabilities and what their effect on the vCPU or 5519the virtual machine is when enabling them. 5520 5521The following information is provided along with the description: 5522 5523 Architectures: 5524 which instruction set architectures provide this ioctl. 5525 x86 includes both i386 and x86_64. 5526 5527 Target: 5528 whether this is a per-vcpu or per-vm capability. 5529 5530 Parameters: 5531 what parameters are accepted by the capability. 5532 5533 Returns: 5534 the return value. General error numbers (EBADF, ENOMEM, EINVAL) 5535 are not detailed, but errors with specific meanings are. 5536 5537 55386.1 KVM_CAP_PPC_OSI 5539------------------- 5540 5541:Architectures: ppc 5542:Target: vcpu 5543:Parameters: none 5544:Returns: 0 on success; -1 on error 5545 5546This capability enables interception of OSI hypercalls that otherwise would 5547be treated as normal system calls to be injected into the guest. OSI hypercalls 5548were invented by Mac-on-Linux to have a standardized communication mechanism 5549between the guest and the host. 5550 5551When this capability is enabled, KVM_EXIT_OSI can occur. 5552 5553 55546.2 KVM_CAP_PPC_PAPR 5555-------------------- 5556 5557:Architectures: ppc 5558:Target: vcpu 5559:Parameters: none 5560:Returns: 0 on success; -1 on error 5561 5562This capability enables interception of PAPR hypercalls. PAPR hypercalls are 5563done using the hypercall instruction "sc 1". 5564 5565It also sets the guest privilege level to "supervisor" mode. Usually the guest 5566runs in "hypervisor" privilege mode with a few missing features. 5567 5568In addition to the above, it changes the semantics of SDR1. In this mode, the 5569HTAB address part of SDR1 contains an HVA instead of a GPA, as PAPR keeps the 5570HTAB invisible to the guest. 5571 5572When this capability is enabled, KVM_EXIT_PAPR_HCALL can occur. 5573 5574 55756.3 KVM_CAP_SW_TLB 5576------------------ 5577 5578:Architectures: ppc 5579:Target: vcpu 5580:Parameters: args[0] is the address of a struct kvm_config_tlb 5581:Returns: 0 on success; -1 on error 5582 5583:: 5584 5585 struct kvm_config_tlb { 5586 __u64 params; 5587 __u64 array; 5588 __u32 mmu_type; 5589 __u32 array_len; 5590 }; 5591 5592Configures the virtual CPU's TLB array, establishing a shared memory area 5593between userspace and KVM. The "params" and "array" fields are userspace 5594addresses of mmu-type-specific data structures. The "array_len" field is an 5595safety mechanism, and should be set to the size in bytes of the memory that 5596userspace has reserved for the array. It must be at least the size dictated 5597by "mmu_type" and "params". 5598 5599While KVM_RUN is active, the shared region is under control of KVM. Its 5600contents are undefined, and any modification by userspace results in 5601boundedly undefined behavior. 5602 5603On return from KVM_RUN, the shared region will reflect the current state of 5604the guest's TLB. If userspace makes any changes, it must call KVM_DIRTY_TLB 5605to tell KVM which entries have been changed, prior to calling KVM_RUN again 5606on this vcpu. 5607 5608For mmu types KVM_MMU_FSL_BOOKE_NOHV and KVM_MMU_FSL_BOOKE_HV: 5609 5610 - The "params" field is of type "struct kvm_book3e_206_tlb_params". 5611 - The "array" field points to an array of type "struct 5612 kvm_book3e_206_tlb_entry". 5613 - The array consists of all entries in the first TLB, followed by all 5614 entries in the second TLB. 5615 - Within a TLB, entries are ordered first by increasing set number. Within a 5616 set, entries are ordered by way (increasing ESEL). 5617 - The hash for determining set number in TLB0 is: (MAS2 >> 12) & (num_sets - 1) 5618 where "num_sets" is the tlb_sizes[] value divided by the tlb_ways[] value. 5619 - The tsize field of mas1 shall be set to 4K on TLB0, even though the 5620 hardware ignores this value for TLB0. 5621 56226.4 KVM_CAP_S390_CSS_SUPPORT 5623---------------------------- 5624 5625:Architectures: s390 5626:Target: vcpu 5627:Parameters: none 5628:Returns: 0 on success; -1 on error 5629 5630This capability enables support for handling of channel I/O instructions. 5631 5632TEST PENDING INTERRUPTION and the interrupt portion of TEST SUBCHANNEL are 5633handled in-kernel, while the other I/O instructions are passed to userspace. 5634 5635When this capability is enabled, KVM_EXIT_S390_TSCH will occur on TEST 5636SUBCHANNEL intercepts. 5637 5638Note that even though this capability is enabled per-vcpu, the complete 5639virtual machine is affected. 5640 56416.5 KVM_CAP_PPC_EPR 5642------------------- 5643 5644:Architectures: ppc 5645:Target: vcpu 5646:Parameters: args[0] defines whether the proxy facility is active 5647:Returns: 0 on success; -1 on error 5648 5649This capability enables or disables the delivery of interrupts through the 5650external proxy facility. 5651 5652When enabled (args[0] != 0), every time the guest gets an external interrupt 5653delivered, it automatically exits into user space with a KVM_EXIT_EPR exit 5654to receive the topmost interrupt vector. 5655 5656When disabled (args[0] == 0), behavior is as if this facility is unsupported. 5657 5658When this capability is enabled, KVM_EXIT_EPR can occur. 5659 56606.6 KVM_CAP_IRQ_MPIC 5661-------------------- 5662 5663:Architectures: ppc 5664:Parameters: args[0] is the MPIC device fd; 5665 args[1] is the MPIC CPU number for this vcpu 5666 5667This capability connects the vcpu to an in-kernel MPIC device. 5668 56696.7 KVM_CAP_IRQ_XICS 5670-------------------- 5671 5672:Architectures: ppc 5673:Target: vcpu 5674:Parameters: args[0] is the XICS device fd; 5675 args[1] is the XICS CPU number (server ID) for this vcpu 5676 5677This capability connects the vcpu to an in-kernel XICS device. 5678 56796.8 KVM_CAP_S390_IRQCHIP 5680------------------------ 5681 5682:Architectures: s390 5683:Target: vm 5684:Parameters: none 5685 5686This capability enables the in-kernel irqchip for s390. Please refer to 5687"4.24 KVM_CREATE_IRQCHIP" for details. 5688 56896.9 KVM_CAP_MIPS_FPU 5690-------------------- 5691 5692:Architectures: mips 5693:Target: vcpu 5694:Parameters: args[0] is reserved for future use (should be 0). 5695 5696This capability allows the use of the host Floating Point Unit by the guest. It 5697allows the Config1.FP bit to be set to enable the FPU in the guest. Once this is 5698done the ``KVM_REG_MIPS_FPR_*`` and ``KVM_REG_MIPS_FCR_*`` registers can be 5699accessed (depending on the current guest FPU register mode), and the Status.FR, 5700Config5.FRE bits are accessible via the KVM API and also from the guest, 5701depending on them being supported by the FPU. 5702 57036.10 KVM_CAP_MIPS_MSA 5704--------------------- 5705 5706:Architectures: mips 5707:Target: vcpu 5708:Parameters: args[0] is reserved for future use (should be 0). 5709 5710This capability allows the use of the MIPS SIMD Architecture (MSA) by the guest. 5711It allows the Config3.MSAP bit to be set to enable the use of MSA by the guest. 5712Once this is done the ``KVM_REG_MIPS_VEC_*`` and ``KVM_REG_MIPS_MSA_*`` 5713registers can be accessed, and the Config5.MSAEn bit is accessible via the 5714KVM API and also from the guest. 5715 57166.74 KVM_CAP_SYNC_REGS 5717---------------------- 5718 5719:Architectures: s390, x86 5720:Target: s390: always enabled, x86: vcpu 5721:Parameters: none 5722:Returns: x86: KVM_CHECK_EXTENSION returns a bit-array indicating which register 5723 sets are supported 5724 (bitfields defined in arch/x86/include/uapi/asm/kvm.h). 5725 5726As described above in the kvm_sync_regs struct info in section 5 (kvm_run): 5727KVM_CAP_SYNC_REGS "allow[s] userspace to access certain guest registers 5728without having to call SET/GET_*REGS". This reduces overhead by eliminating 5729repeated ioctl calls for setting and/or getting register values. This is 5730particularly important when userspace is making synchronous guest state 5731modifications, e.g. when emulating and/or intercepting instructions in 5732userspace. 5733 5734For s390 specifics, please refer to the source code. 5735 5736For x86: 5737 5738- the register sets to be copied out to kvm_run are selectable 5739 by userspace (rather that all sets being copied out for every exit). 5740- vcpu_events are available in addition to regs and sregs. 5741 5742For x86, the 'kvm_valid_regs' field of struct kvm_run is overloaded to 5743function as an input bit-array field set by userspace to indicate the 5744specific register sets to be copied out on the next exit. 5745 5746To indicate when userspace has modified values that should be copied into 5747the vCPU, the all architecture bitarray field, 'kvm_dirty_regs' must be set. 5748This is done using the same bitflags as for the 'kvm_valid_regs' field. 5749If the dirty bit is not set, then the register set values will not be copied 5750into the vCPU even if they've been modified. 5751 5752Unused bitfields in the bitarrays must be set to zero. 5753 5754:: 5755 5756 struct kvm_sync_regs { 5757 struct kvm_regs regs; 5758 struct kvm_sregs sregs; 5759 struct kvm_vcpu_events events; 5760 }; 5761 57626.75 KVM_CAP_PPC_IRQ_XIVE 5763------------------------- 5764 5765:Architectures: ppc 5766:Target: vcpu 5767:Parameters: args[0] is the XIVE device fd; 5768 args[1] is the XIVE CPU number (server ID) for this vcpu 5769 5770This capability connects the vcpu to an in-kernel XIVE device. 5771 57727. Capabilities that can be enabled on VMs 5773========================================== 5774 5775There are certain capabilities that change the behavior of the virtual 5776machine when enabled. To enable them, please see section 4.37. Below 5777you can find a list of capabilities and what their effect on the VM 5778is when enabling them. 5779 5780The following information is provided along with the description: 5781 5782 Architectures: 5783 which instruction set architectures provide this ioctl. 5784 x86 includes both i386 and x86_64. 5785 5786 Parameters: 5787 what parameters are accepted by the capability. 5788 5789 Returns: 5790 the return value. General error numbers (EBADF, ENOMEM, EINVAL) 5791 are not detailed, but errors with specific meanings are. 5792 5793 57947.1 KVM_CAP_PPC_ENABLE_HCALL 5795---------------------------- 5796 5797:Architectures: ppc 5798:Parameters: args[0] is the sPAPR hcall number; 5799 args[1] is 0 to disable, 1 to enable in-kernel handling 5800 5801This capability controls whether individual sPAPR hypercalls (hcalls) 5802get handled by the kernel or not. Enabling or disabling in-kernel 5803handling of an hcall is effective across the VM. On creation, an 5804initial set of hcalls are enabled for in-kernel handling, which 5805consists of those hcalls for which in-kernel handlers were implemented 5806before this capability was implemented. If disabled, the kernel will 5807not to attempt to handle the hcall, but will always exit to userspace 5808to handle it. Note that it may not make sense to enable some and 5809disable others of a group of related hcalls, but KVM does not prevent 5810userspace from doing that. 5811 5812If the hcall number specified is not one that has an in-kernel 5813implementation, the KVM_ENABLE_CAP ioctl will fail with an EINVAL 5814error. 5815 58167.2 KVM_CAP_S390_USER_SIGP 5817-------------------------- 5818 5819:Architectures: s390 5820:Parameters: none 5821 5822This capability controls which SIGP orders will be handled completely in user 5823space. With this capability enabled, all fast orders will be handled completely 5824in the kernel: 5825 5826- SENSE 5827- SENSE RUNNING 5828- EXTERNAL CALL 5829- EMERGENCY SIGNAL 5830- CONDITIONAL EMERGENCY SIGNAL 5831 5832All other orders will be handled completely in user space. 5833 5834Only privileged operation exceptions will be checked for in the kernel (or even 5835in the hardware prior to interception). If this capability is not enabled, the 5836old way of handling SIGP orders is used (partially in kernel and user space). 5837 58387.3 KVM_CAP_S390_VECTOR_REGISTERS 5839--------------------------------- 5840 5841:Architectures: s390 5842:Parameters: none 5843:Returns: 0 on success, negative value on error 5844 5845Allows use of the vector registers introduced with z13 processor, and 5846provides for the synchronization between host and user space. Will 5847return -EINVAL if the machine does not support vectors. 5848 58497.4 KVM_CAP_S390_USER_STSI 5850-------------------------- 5851 5852:Architectures: s390 5853:Parameters: none 5854 5855This capability allows post-handlers for the STSI instruction. After 5856initial handling in the kernel, KVM exits to user space with 5857KVM_EXIT_S390_STSI to allow user space to insert further data. 5858 5859Before exiting to userspace, kvm handlers should fill in s390_stsi field of 5860vcpu->run:: 5861 5862 struct { 5863 __u64 addr; 5864 __u8 ar; 5865 __u8 reserved; 5866 __u8 fc; 5867 __u8 sel1; 5868 __u16 sel2; 5869 } s390_stsi; 5870 5871 @addr - guest address of STSI SYSIB 5872 @fc - function code 5873 @sel1 - selector 1 5874 @sel2 - selector 2 5875 @ar - access register number 5876 5877KVM handlers should exit to userspace with rc = -EREMOTE. 5878 58797.5 KVM_CAP_SPLIT_IRQCHIP 5880------------------------- 5881 5882:Architectures: x86 5883:Parameters: args[0] - number of routes reserved for userspace IOAPICs 5884:Returns: 0 on success, -1 on error 5885 5886Create a local apic for each processor in the kernel. This can be used 5887instead of KVM_CREATE_IRQCHIP if the userspace VMM wishes to emulate the 5888IOAPIC and PIC (and also the PIT, even though this has to be enabled 5889separately). 5890 5891This capability also enables in kernel routing of interrupt requests; 5892when KVM_CAP_SPLIT_IRQCHIP only routes of KVM_IRQ_ROUTING_MSI type are 5893used in the IRQ routing table. The first args[0] MSI routes are reserved 5894for the IOAPIC pins. Whenever the LAPIC receives an EOI for these routes, 5895a KVM_EXIT_IOAPIC_EOI vmexit will be reported to userspace. 5896 5897Fails if VCPU has already been created, or if the irqchip is already in the 5898kernel (i.e. KVM_CREATE_IRQCHIP has already been called). 5899 59007.6 KVM_CAP_S390_RI 5901------------------- 5902 5903:Architectures: s390 5904:Parameters: none 5905 5906Allows use of runtime-instrumentation introduced with zEC12 processor. 5907Will return -EINVAL if the machine does not support runtime-instrumentation. 5908Will return -EBUSY if a VCPU has already been created. 5909 59107.7 KVM_CAP_X2APIC_API 5911---------------------- 5912 5913:Architectures: x86 5914:Parameters: args[0] - features that should be enabled 5915:Returns: 0 on success, -EINVAL when args[0] contains invalid features 5916 5917Valid feature flags in args[0] are:: 5918 5919 #define KVM_X2APIC_API_USE_32BIT_IDS (1ULL << 0) 5920 #define KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK (1ULL << 1) 5921 5922Enabling KVM_X2APIC_API_USE_32BIT_IDS changes the behavior of 5923KVM_SET_GSI_ROUTING, KVM_SIGNAL_MSI, KVM_SET_LAPIC, and KVM_GET_LAPIC, 5924allowing the use of 32-bit APIC IDs. See KVM_CAP_X2APIC_API in their 5925respective sections. 5926 5927KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK must be enabled for x2APIC to work 5928in logical mode or with more than 255 VCPUs. Otherwise, KVM treats 0xff 5929as a broadcast even in x2APIC mode in order to support physical x2APIC 5930without interrupt remapping. This is undesirable in logical mode, 5931where 0xff represents CPUs 0-7 in cluster 0. 5932 59337.8 KVM_CAP_S390_USER_INSTR0 5934---------------------------- 5935 5936:Architectures: s390 5937:Parameters: none 5938 5939With this capability enabled, all illegal instructions 0x0000 (2 bytes) will 5940be intercepted and forwarded to user space. User space can use this 5941mechanism e.g. to realize 2-byte software breakpoints. The kernel will 5942not inject an operating exception for these instructions, user space has 5943to take care of that. 5944 5945This capability can be enabled dynamically even if VCPUs were already 5946created and are running. 5947 59487.9 KVM_CAP_S390_GS 5949------------------- 5950 5951:Architectures: s390 5952:Parameters: none 5953:Returns: 0 on success; -EINVAL if the machine does not support 5954 guarded storage; -EBUSY if a VCPU has already been created. 5955 5956Allows use of guarded storage for the KVM guest. 5957 59587.10 KVM_CAP_S390_AIS 5959--------------------- 5960 5961:Architectures: s390 5962:Parameters: none 5963 5964Allow use of adapter-interruption suppression. 5965:Returns: 0 on success; -EBUSY if a VCPU has already been created. 5966 59677.11 KVM_CAP_PPC_SMT 5968-------------------- 5969 5970:Architectures: ppc 5971:Parameters: vsmt_mode, flags 5972 5973Enabling this capability on a VM provides userspace with a way to set 5974the desired virtual SMT mode (i.e. the number of virtual CPUs per 5975virtual core). The virtual SMT mode, vsmt_mode, must be a power of 2 5976between 1 and 8. On POWER8, vsmt_mode must also be no greater than 5977the number of threads per subcore for the host. Currently flags must 5978be 0. A successful call to enable this capability will result in 5979vsmt_mode being returned when the KVM_CAP_PPC_SMT capability is 5980subsequently queried for the VM. This capability is only supported by 5981HV KVM, and can only be set before any VCPUs have been created. 5982The KVM_CAP_PPC_SMT_POSSIBLE capability indicates which virtual SMT 5983modes are available. 5984 59857.12 KVM_CAP_PPC_FWNMI 5986---------------------- 5987 5988:Architectures: ppc 5989:Parameters: none 5990 5991With this capability a machine check exception in the guest address 5992space will cause KVM to exit the guest with NMI exit reason. This 5993enables QEMU to build error log and branch to guest kernel registered 5994machine check handling routine. Without this capability KVM will 5995branch to guests' 0x200 interrupt vector. 5996 59977.13 KVM_CAP_X86_DISABLE_EXITS 5998------------------------------ 5999 6000:Architectures: x86 6001:Parameters: args[0] defines which exits are disabled 6002:Returns: 0 on success, -EINVAL when args[0] contains invalid exits 6003 6004Valid bits in args[0] are:: 6005 6006 #define KVM_X86_DISABLE_EXITS_MWAIT (1 << 0) 6007 #define KVM_X86_DISABLE_EXITS_HLT (1 << 1) 6008 #define KVM_X86_DISABLE_EXITS_PAUSE (1 << 2) 6009 #define KVM_X86_DISABLE_EXITS_CSTATE (1 << 3) 6010 6011Enabling this capability on a VM provides userspace with a way to no 6012longer intercept some instructions for improved latency in some 6013workloads, and is suggested when vCPUs are associated to dedicated 6014physical CPUs. More bits can be added in the future; userspace can 6015just pass the KVM_CHECK_EXTENSION result to KVM_ENABLE_CAP to disable 6016all such vmexits. 6017 6018Do not enable KVM_FEATURE_PV_UNHALT if you disable HLT exits. 6019 60207.14 KVM_CAP_S390_HPAGE_1M 6021-------------------------- 6022 6023:Architectures: s390 6024:Parameters: none 6025:Returns: 0 on success, -EINVAL if hpage module parameter was not set 6026 or cmma is enabled, or the VM has the KVM_VM_S390_UCONTROL 6027 flag set 6028 6029With this capability the KVM support for memory backing with 1m pages 6030through hugetlbfs can be enabled for a VM. After the capability is 6031enabled, cmma can't be enabled anymore and pfmfi and the storage key 6032interpretation are disabled. If cmma has already been enabled or the 6033hpage module parameter is not set to 1, -EINVAL is returned. 6034 6035While it is generally possible to create a huge page backed VM without 6036this capability, the VM will not be able to run. 6037 60387.15 KVM_CAP_MSR_PLATFORM_INFO 6039------------------------------ 6040 6041:Architectures: x86 6042:Parameters: args[0] whether feature should be enabled or not 6043 6044With this capability, a guest may read the MSR_PLATFORM_INFO MSR. Otherwise, 6045a #GP would be raised when the guest tries to access. Currently, this 6046capability does not enable write permissions of this MSR for the guest. 6047 60487.16 KVM_CAP_PPC_NESTED_HV 6049-------------------------- 6050 6051:Architectures: ppc 6052:Parameters: none 6053:Returns: 0 on success, -EINVAL when the implementation doesn't support 6054 nested-HV virtualization. 6055 6056HV-KVM on POWER9 and later systems allows for "nested-HV" 6057virtualization, which provides a way for a guest VM to run guests that 6058can run using the CPU's supervisor mode (privileged non-hypervisor 6059state). Enabling this capability on a VM depends on the CPU having 6060the necessary functionality and on the facility being enabled with a 6061kvm-hv module parameter. 6062 60637.17 KVM_CAP_EXCEPTION_PAYLOAD 6064------------------------------ 6065 6066:Architectures: x86 6067:Parameters: args[0] whether feature should be enabled or not 6068 6069With this capability enabled, CR2 will not be modified prior to the 6070emulated VM-exit when L1 intercepts a #PF exception that occurs in 6071L2. Similarly, for kvm-intel only, DR6 will not be modified prior to 6072the emulated VM-exit when L1 intercepts a #DB exception that occurs in 6073L2. As a result, when KVM_GET_VCPU_EVENTS reports a pending #PF (or 6074#DB) exception for L2, exception.has_payload will be set and the 6075faulting address (or the new DR6 bits*) will be reported in the 6076exception_payload field. Similarly, when userspace injects a #PF (or 6077#DB) into L2 using KVM_SET_VCPU_EVENTS, it is expected to set 6078exception.has_payload and to put the faulting address - or the new DR6 6079bits\ [#]_ - in the exception_payload field. 6080 6081This capability also enables exception.pending in struct 6082kvm_vcpu_events, which allows userspace to distinguish between pending 6083and injected exceptions. 6084 6085 6086.. [#] For the new DR6 bits, note that bit 16 is set iff the #DB exception 6087 will clear DR6.RTM. 6088 60897.18 KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 6090 6091:Architectures: x86, arm, arm64, mips 6092:Parameters: args[0] whether feature should be enabled or not 6093 6094Valid flags are:: 6095 6096 #define KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE (1 << 0) 6097 #define KVM_DIRTY_LOG_INITIALLY_SET (1 << 1) 6098 6099With KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE is set, KVM_GET_DIRTY_LOG will not 6100automatically clear and write-protect all pages that are returned as dirty. 6101Rather, userspace will have to do this operation separately using 6102KVM_CLEAR_DIRTY_LOG. 6103 6104At the cost of a slightly more complicated operation, this provides better 6105scalability and responsiveness for two reasons. First, 6106KVM_CLEAR_DIRTY_LOG ioctl can operate on a 64-page granularity rather 6107than requiring to sync a full memslot; this ensures that KVM does not 6108take spinlocks for an extended period of time. Second, in some cases a 6109large amount of time can pass between a call to KVM_GET_DIRTY_LOG and 6110userspace actually using the data in the page. Pages can be modified 6111during this time, which is inefficient for both the guest and userspace: 6112the guest will incur a higher penalty due to write protection faults, 6113while userspace can see false reports of dirty pages. Manual reprotection 6114helps reducing this time, improving guest performance and reducing the 6115number of dirty log false positives. 6116 6117With KVM_DIRTY_LOG_INITIALLY_SET set, all the bits of the dirty bitmap 6118will be initialized to 1 when created. This also improves performance because 6119dirty logging can be enabled gradually in small chunks on the first call 6120to KVM_CLEAR_DIRTY_LOG. KVM_DIRTY_LOG_INITIALLY_SET depends on 6121KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE (it is also only available on 6122x86 and arm64 for now). 6123 6124KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 was previously available under the name 6125KVM_CAP_MANUAL_DIRTY_LOG_PROTECT, but the implementation had bugs that make 6126it hard or impossible to use it correctly. The availability of 6127KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 signals that those bugs are fixed. 6128Userspace should not try to use KVM_CAP_MANUAL_DIRTY_LOG_PROTECT. 6129 61307.19 KVM_CAP_PPC_SECURE_GUEST 6131------------------------------ 6132 6133:Architectures: ppc 6134 6135This capability indicates that KVM is running on a host that has 6136ultravisor firmware and thus can support a secure guest. On such a 6137system, a guest can ask the ultravisor to make it a secure guest, 6138one whose memory is inaccessible to the host except for pages which 6139are explicitly requested to be shared with the host. The ultravisor 6140notifies KVM when a guest requests to become a secure guest, and KVM 6141has the opportunity to veto the transition. 6142 6143If present, this capability can be enabled for a VM, meaning that KVM 6144will allow the transition to secure guest mode. Otherwise KVM will 6145veto the transition. 6146 61477.20 KVM_CAP_HALT_POLL 6148---------------------- 6149 6150:Architectures: all 6151:Target: VM 6152:Parameters: args[0] is the maximum poll time in nanoseconds 6153:Returns: 0 on success; -1 on error 6154 6155This capability overrides the kvm module parameter halt_poll_ns for the 6156target VM. 6157 6158VCPU polling allows a VCPU to poll for wakeup events instead of immediately 6159scheduling during guest halts. The maximum time a VCPU can spend polling is 6160controlled by the kvm module parameter halt_poll_ns. This capability allows 6161the maximum halt time to specified on a per-VM basis, effectively overriding 6162the module parameter for the target VM. 6163 61647.21 KVM_CAP_X86_USER_SPACE_MSR 6165------------------------------- 6166 6167:Architectures: x86 6168:Target: VM 6169:Parameters: args[0] contains the mask of KVM_MSR_EXIT_REASON_* events to report 6170:Returns: 0 on success; -1 on error 6171 6172This capability enables trapping of #GP invoking RDMSR and WRMSR instructions 6173into user space. 6174 6175When a guest requests to read or write an MSR, KVM may not implement all MSRs 6176that are relevant to a respective system. It also does not differentiate by 6177CPU type. 6178 6179To allow more fine grained control over MSR handling, user space may enable 6180this capability. With it enabled, MSR accesses that match the mask specified in 6181args[0] and trigger a #GP event inside the guest by KVM will instead trigger 6182KVM_EXIT_X86_RDMSR and KVM_EXIT_X86_WRMSR exit notifications which user space 6183can then handle to implement model specific MSR handling and/or user notifications 6184to inform a user that an MSR was not handled. 6185 61867.22 KVM_CAP_X86_BUS_LOCK_EXIT 6187------------------------------- 6188 6189:Architectures: x86 6190:Target: VM 6191:Parameters: args[0] defines the policy used when bus locks detected in guest 6192:Returns: 0 on success, -EINVAL when args[0] contains invalid bits 6193 6194Valid bits in args[0] are:: 6195 6196 #define KVM_BUS_LOCK_DETECTION_OFF (1 << 0) 6197 #define KVM_BUS_LOCK_DETECTION_EXIT (1 << 1) 6198 6199Enabling this capability on a VM provides userspace with a way to select 6200a policy to handle the bus locks detected in guest. Userspace can obtain 6201the supported modes from the result of KVM_CHECK_EXTENSION and define it 6202through the KVM_ENABLE_CAP. 6203 6204KVM_BUS_LOCK_DETECTION_OFF and KVM_BUS_LOCK_DETECTION_EXIT are supported 6205currently and mutually exclusive with each other. More bits can be added in 6206the future. 6207 6208With KVM_BUS_LOCK_DETECTION_OFF set, bus locks in guest will not cause vm exits 6209so that no additional actions are needed. This is the default mode. 6210 6211With KVM_BUS_LOCK_DETECTION_EXIT set, vm exits happen when bus lock detected 6212in VM. KVM just exits to userspace when handling them. Userspace can enforce 6213its own throttling or other policy based mitigations. 6214 6215This capability is aimed to address the thread that VM can exploit bus locks to 6216degree the performance of the whole system. Once the userspace enable this 6217capability and select the KVM_BUS_LOCK_DETECTION_EXIT mode, KVM will set the 6218KVM_RUN_BUS_LOCK flag in vcpu-run->flags field and exit to userspace. Concerning 6219the bus lock vm exit can be preempted by a higher priority VM exit, the exit 6220notifications to userspace can be KVM_EXIT_BUS_LOCK or other reasons. 6221KVM_RUN_BUS_LOCK flag is used to distinguish between them. 6222 62237.23 KVM_CAP_PPC_DAWR1 6224---------------------- 6225 6226:Architectures: ppc 6227:Parameters: none 6228:Returns: 0 on success, -EINVAL when CPU doesn't support 2nd DAWR 6229 6230This capability can be used to check / enable 2nd DAWR feature provided 6231by POWER10 processor. 6232 62338. Other capabilities. 6234====================== 6235 6236This section lists capabilities that give information about other 6237features of the KVM implementation. 6238 62398.1 KVM_CAP_PPC_HWRNG 6240--------------------- 6241 6242:Architectures: ppc 6243 6244This capability, if KVM_CHECK_EXTENSION indicates that it is 6245available, means that the kernel has an implementation of the 6246H_RANDOM hypercall backed by a hardware random-number generator. 6247If present, the kernel H_RANDOM handler can be enabled for guest use 6248with the KVM_CAP_PPC_ENABLE_HCALL capability. 6249 62508.2 KVM_CAP_HYPERV_SYNIC 6251------------------------ 6252 6253:Architectures: x86 6254 6255This capability, if KVM_CHECK_EXTENSION indicates that it is 6256available, means that the kernel has an implementation of the 6257Hyper-V Synthetic interrupt controller(SynIC). Hyper-V SynIC is 6258used to support Windows Hyper-V based guest paravirt drivers(VMBus). 6259 6260In order to use SynIC, it has to be activated by setting this 6261capability via KVM_ENABLE_CAP ioctl on the vcpu fd. Note that this 6262will disable the use of APIC hardware virtualization even if supported 6263by the CPU, as it's incompatible with SynIC auto-EOI behavior. 6264 62658.3 KVM_CAP_PPC_RADIX_MMU 6266------------------------- 6267 6268:Architectures: ppc 6269 6270This capability, if KVM_CHECK_EXTENSION indicates that it is 6271available, means that the kernel can support guests using the 6272radix MMU defined in Power ISA V3.00 (as implemented in the POWER9 6273processor). 6274 62758.4 KVM_CAP_PPC_HASH_MMU_V3 6276--------------------------- 6277 6278:Architectures: ppc 6279 6280This capability, if KVM_CHECK_EXTENSION indicates that it is 6281available, means that the kernel can support guests using the 6282hashed page table MMU defined in Power ISA V3.00 (as implemented in 6283the POWER9 processor), including in-memory segment tables. 6284 62858.5 KVM_CAP_MIPS_VZ 6286------------------- 6287 6288:Architectures: mips 6289 6290This capability, if KVM_CHECK_EXTENSION on the main kvm handle indicates that 6291it is available, means that full hardware assisted virtualization capabilities 6292of the hardware are available for use through KVM. An appropriate 6293KVM_VM_MIPS_* type must be passed to KVM_CREATE_VM to create a VM which 6294utilises it. 6295 6296If KVM_CHECK_EXTENSION on a kvm VM handle indicates that this capability is 6297available, it means that the VM is using full hardware assisted virtualization 6298capabilities of the hardware. This is useful to check after creating a VM with 6299KVM_VM_MIPS_DEFAULT. 6300 6301The value returned by KVM_CHECK_EXTENSION should be compared against known 6302values (see below). All other values are reserved. This is to allow for the 6303possibility of other hardware assisted virtualization implementations which 6304may be incompatible with the MIPS VZ ASE. 6305 6306== ========================================================================== 6307 0 The trap & emulate implementation is in use to run guest code in user 6308 mode. Guest virtual memory segments are rearranged to fit the guest in the 6309 user mode address space. 6310 6311 1 The MIPS VZ ASE is in use, providing full hardware assisted 6312 virtualization, including standard guest virtual memory segments. 6313== ========================================================================== 6314 63158.6 KVM_CAP_MIPS_TE 6316------------------- 6317 6318:Architectures: mips 6319 6320This capability, if KVM_CHECK_EXTENSION on the main kvm handle indicates that 6321it is available, means that the trap & emulate implementation is available to 6322run guest code in user mode, even if KVM_CAP_MIPS_VZ indicates that hardware 6323assisted virtualisation is also available. KVM_VM_MIPS_TE (0) must be passed 6324to KVM_CREATE_VM to create a VM which utilises it. 6325 6326If KVM_CHECK_EXTENSION on a kvm VM handle indicates that this capability is 6327available, it means that the VM is using trap & emulate. 6328 63298.7 KVM_CAP_MIPS_64BIT 6330---------------------- 6331 6332:Architectures: mips 6333 6334This capability indicates the supported architecture type of the guest, i.e. the 6335supported register and address width. 6336 6337The values returned when this capability is checked by KVM_CHECK_EXTENSION on a 6338kvm VM handle correspond roughly to the CP0_Config.AT register field, and should 6339be checked specifically against known values (see below). All other values are 6340reserved. 6341 6342== ======================================================================== 6343 0 MIPS32 or microMIPS32. 6344 Both registers and addresses are 32-bits wide. 6345 It will only be possible to run 32-bit guest code. 6346 6347 1 MIPS64 or microMIPS64 with access only to 32-bit compatibility segments. 6348 Registers are 64-bits wide, but addresses are 32-bits wide. 6349 64-bit guest code may run but cannot access MIPS64 memory segments. 6350 It will also be possible to run 32-bit guest code. 6351 6352 2 MIPS64 or microMIPS64 with access to all address segments. 6353 Both registers and addresses are 64-bits wide. 6354 It will be possible to run 64-bit or 32-bit guest code. 6355== ======================================================================== 6356 63578.9 KVM_CAP_ARM_USER_IRQ 6358------------------------ 6359 6360:Architectures: arm, arm64 6361 6362This capability, if KVM_CHECK_EXTENSION indicates that it is available, means 6363that if userspace creates a VM without an in-kernel interrupt controller, it 6364will be notified of changes to the output level of in-kernel emulated devices, 6365which can generate virtual interrupts, presented to the VM. 6366For such VMs, on every return to userspace, the kernel 6367updates the vcpu's run->s.regs.device_irq_level field to represent the actual 6368output level of the device. 6369 6370Whenever kvm detects a change in the device output level, kvm guarantees at 6371least one return to userspace before running the VM. This exit could either 6372be a KVM_EXIT_INTR or any other exit event, like KVM_EXIT_MMIO. This way, 6373userspace can always sample the device output level and re-compute the state of 6374the userspace interrupt controller. Userspace should always check the state 6375of run->s.regs.device_irq_level on every kvm exit. 6376The value in run->s.regs.device_irq_level can represent both level and edge 6377triggered interrupt signals, depending on the device. Edge triggered interrupt 6378signals will exit to userspace with the bit in run->s.regs.device_irq_level 6379set exactly once per edge signal. 6380 6381The field run->s.regs.device_irq_level is available independent of 6382run->kvm_valid_regs or run->kvm_dirty_regs bits. 6383 6384If KVM_CAP_ARM_USER_IRQ is supported, the KVM_CHECK_EXTENSION ioctl returns a 6385number larger than 0 indicating the version of this capability is implemented 6386and thereby which bits in run->s.regs.device_irq_level can signal values. 6387 6388Currently the following bits are defined for the device_irq_level bitmap:: 6389 6390 KVM_CAP_ARM_USER_IRQ >= 1: 6391 6392 KVM_ARM_DEV_EL1_VTIMER - EL1 virtual timer 6393 KVM_ARM_DEV_EL1_PTIMER - EL1 physical timer 6394 KVM_ARM_DEV_PMU - ARM PMU overflow interrupt signal 6395 6396Future versions of kvm may implement additional events. These will get 6397indicated by returning a higher number from KVM_CHECK_EXTENSION and will be 6398listed above. 6399 64008.10 KVM_CAP_PPC_SMT_POSSIBLE 6401----------------------------- 6402 6403:Architectures: ppc 6404 6405Querying this capability returns a bitmap indicating the possible 6406virtual SMT modes that can be set using KVM_CAP_PPC_SMT. If bit N 6407(counting from the right) is set, then a virtual SMT mode of 2^N is 6408available. 6409 64108.11 KVM_CAP_HYPERV_SYNIC2 6411-------------------------- 6412 6413:Architectures: x86 6414 6415This capability enables a newer version of Hyper-V Synthetic interrupt 6416controller (SynIC). The only difference with KVM_CAP_HYPERV_SYNIC is that KVM 6417doesn't clear SynIC message and event flags pages when they are enabled by 6418writing to the respective MSRs. 6419 64208.12 KVM_CAP_HYPERV_VP_INDEX 6421---------------------------- 6422 6423:Architectures: x86 6424 6425This capability indicates that userspace can load HV_X64_MSR_VP_INDEX msr. Its 6426value is used to denote the target vcpu for a SynIC interrupt. For 6427compatibilty, KVM initializes this msr to KVM's internal vcpu index. When this 6428capability is absent, userspace can still query this msr's value. 6429 64308.13 KVM_CAP_S390_AIS_MIGRATION 6431------------------------------- 6432 6433:Architectures: s390 6434:Parameters: none 6435 6436This capability indicates if the flic device will be able to get/set the 6437AIS states for migration via the KVM_DEV_FLIC_AISM_ALL attribute and allows 6438to discover this without having to create a flic device. 6439 64408.14 KVM_CAP_S390_PSW 6441--------------------- 6442 6443:Architectures: s390 6444 6445This capability indicates that the PSW is exposed via the kvm_run structure. 6446 64478.15 KVM_CAP_S390_GMAP 6448---------------------- 6449 6450:Architectures: s390 6451 6452This capability indicates that the user space memory used as guest mapping can 6453be anywhere in the user memory address space, as long as the memory slots are 6454aligned and sized to a segment (1MB) boundary. 6455 64568.16 KVM_CAP_S390_COW 6457--------------------- 6458 6459:Architectures: s390 6460 6461This capability indicates that the user space memory used as guest mapping can 6462use copy-on-write semantics as well as dirty pages tracking via read-only page 6463tables. 6464 64658.17 KVM_CAP_S390_BPB 6466--------------------- 6467 6468:Architectures: s390 6469 6470This capability indicates that kvm will implement the interfaces to handle 6471reset, migration and nested KVM for branch prediction blocking. The stfle 6472facility 82 should not be provided to the guest without this capability. 6473 64748.18 KVM_CAP_HYPERV_TLBFLUSH 6475---------------------------- 6476 6477:Architectures: x86 6478 6479This capability indicates that KVM supports paravirtualized Hyper-V TLB Flush 6480hypercalls: 6481HvFlushVirtualAddressSpace, HvFlushVirtualAddressSpaceEx, 6482HvFlushVirtualAddressList, HvFlushVirtualAddressListEx. 6483 64848.19 KVM_CAP_ARM_INJECT_SERROR_ESR 6485---------------------------------- 6486 6487:Architectures: arm, arm64 6488 6489This capability indicates that userspace can specify (via the 6490KVM_SET_VCPU_EVENTS ioctl) the syndrome value reported to the guest when it 6491takes a virtual SError interrupt exception. 6492If KVM advertises this capability, userspace can only specify the ISS field for 6493the ESR syndrome. Other parts of the ESR, such as the EC are generated by the 6494CPU when the exception is taken. If this virtual SError is taken to EL1 using 6495AArch64, this value will be reported in the ISS field of ESR_ELx. 6496 6497See KVM_CAP_VCPU_EVENTS for more details. 6498 64998.20 KVM_CAP_HYPERV_SEND_IPI 6500---------------------------- 6501 6502:Architectures: x86 6503 6504This capability indicates that KVM supports paravirtualized Hyper-V IPI send 6505hypercalls: 6506HvCallSendSyntheticClusterIpi, HvCallSendSyntheticClusterIpiEx. 6507 65088.21 KVM_CAP_HYPERV_DIRECT_TLBFLUSH 6509----------------------------------- 6510 6511:Architectures: x86 6512 6513This capability indicates that KVM running on top of Hyper-V hypervisor 6514enables Direct TLB flush for its guests meaning that TLB flush 6515hypercalls are handled by Level 0 hypervisor (Hyper-V) bypassing KVM. 6516Due to the different ABI for hypercall parameters between Hyper-V and 6517KVM, enabling this capability effectively disables all hypercall 6518handling by KVM (as some KVM hypercall may be mistakenly treated as TLB 6519flush hypercalls by Hyper-V) so userspace should disable KVM identification 6520in CPUID and only exposes Hyper-V identification. In this case, guest 6521thinks it's running on Hyper-V and only use Hyper-V hypercalls. 6522 65238.22 KVM_CAP_S390_VCPU_RESETS 6524----------------------------- 6525 6526:Architectures: s390 6527 6528This capability indicates that the KVM_S390_NORMAL_RESET and 6529KVM_S390_CLEAR_RESET ioctls are available. 6530 65318.23 KVM_CAP_S390_PROTECTED 6532--------------------------- 6533 6534:Architectures: s390 6535 6536This capability indicates that the Ultravisor has been initialized and 6537KVM can therefore start protected VMs. 6538This capability governs the KVM_S390_PV_COMMAND ioctl and the 6539KVM_MP_STATE_LOAD MP_STATE. KVM_SET_MP_STATE can fail for protected 6540guests when the state change is invalid. 6541 65428.24 KVM_CAP_STEAL_TIME 6543----------------------- 6544 6545:Architectures: arm64, x86 6546 6547This capability indicates that KVM supports steal time accounting. 6548When steal time accounting is supported it may be enabled with 6549architecture-specific interfaces. This capability and the architecture- 6550specific interfaces must be consistent, i.e. if one says the feature 6551is supported, than the other should as well and vice versa. For arm64 6552see Documentation/virt/kvm/devices/vcpu.rst "KVM_ARM_VCPU_PVTIME_CTRL". 6553For x86 see Documentation/virt/kvm/msr.rst "MSR_KVM_STEAL_TIME". 6554 65558.25 KVM_CAP_S390_DIAG318 6556------------------------- 6557 6558:Architectures: s390 6559 6560This capability enables a guest to set information about its control program 6561(i.e. guest kernel type and version). The information is helpful during 6562system/firmware service events, providing additional data about the guest 6563environments running on the machine. 6564 6565The information is associated with the DIAGNOSE 0x318 instruction, which sets 6566an 8-byte value consisting of a one-byte Control Program Name Code (CPNC) and 6567a 7-byte Control Program Version Code (CPVC). The CPNC determines what 6568environment the control program is running in (e.g. Linux, z/VM...), and the 6569CPVC is used for information specific to OS (e.g. Linux version, Linux 6570distribution...) 6571 6572If this capability is available, then the CPNC and CPVC can be synchronized 6573between KVM and userspace via the sync regs mechanism (KVM_SYNC_DIAG318). 6574 65758.26 KVM_CAP_X86_USER_SPACE_MSR 6576------------------------------- 6577 6578:Architectures: x86 6579 6580This capability indicates that KVM supports deflection of MSR reads and 6581writes to user space. It can be enabled on a VM level. If enabled, MSR 6582accesses that would usually trigger a #GP by KVM into the guest will 6583instead get bounced to user space through the KVM_EXIT_X86_RDMSR and 6584KVM_EXIT_X86_WRMSR exit notifications. 6585 65868.27 KVM_X86_SET_MSR_FILTER 6587--------------------------- 6588 6589:Architectures: x86 6590 6591This capability indicates that KVM supports that accesses to user defined MSRs 6592may be rejected. With this capability exposed, KVM exports new VM ioctl 6593KVM_X86_SET_MSR_FILTER which user space can call to specify bitmaps of MSR 6594ranges that KVM should reject access to. 6595 6596In combination with KVM_CAP_X86_USER_SPACE_MSR, this allows user space to 6597trap and emulate MSRs that are outside of the scope of KVM as well as 6598limit the attack surface on KVM's MSR emulation code. 6599 66008.28 KVM_CAP_ENFORCE_PV_CPUID 6601----------------------------- 6602 6603Architectures: x86 6604 6605When enabled, KVM will disable paravirtual features provided to the 6606guest according to the bits in the KVM_CPUID_FEATURES CPUID leaf 6607(0x40000001). Otherwise, a guest may use the paravirtual features 6608regardless of what has actually been exposed through the CPUID leaf. 6609 66108.29 KVM_CAP_DIRTY_LOG_RING 6611--------------------------- 6612 6613:Architectures: x86 6614:Parameters: args[0] - size of the dirty log ring 6615 6616KVM is capable of tracking dirty memory using ring buffers that are 6617mmaped into userspace; there is one dirty ring per vcpu. 6618 6619The dirty ring is available to userspace as an array of 6620``struct kvm_dirty_gfn``. Each dirty entry it's defined as:: 6621 6622 struct kvm_dirty_gfn { 6623 __u32 flags; 6624 __u32 slot; /* as_id | slot_id */ 6625 __u64 offset; 6626 }; 6627 6628The following values are defined for the flags field to define the 6629current state of the entry:: 6630 6631 #define KVM_DIRTY_GFN_F_DIRTY BIT(0) 6632 #define KVM_DIRTY_GFN_F_RESET BIT(1) 6633 #define KVM_DIRTY_GFN_F_MASK 0x3 6634 6635Userspace should call KVM_ENABLE_CAP ioctl right after KVM_CREATE_VM 6636ioctl to enable this capability for the new guest and set the size of 6637the rings. Enabling the capability is only allowed before creating any 6638vCPU, and the size of the ring must be a power of two. The larger the 6639ring buffer, the less likely the ring is full and the VM is forced to 6640exit to userspace. The optimal size depends on the workload, but it is 6641recommended that it be at least 64 KiB (4096 entries). 6642 6643Just like for dirty page bitmaps, the buffer tracks writes to 6644all user memory regions for which the KVM_MEM_LOG_DIRTY_PAGES flag was 6645set in KVM_SET_USER_MEMORY_REGION. Once a memory region is registered 6646with the flag set, userspace can start harvesting dirty pages from the 6647ring buffer. 6648 6649An entry in the ring buffer can be unused (flag bits ``00``), 6650dirty (flag bits ``01``) or harvested (flag bits ``1X``). The 6651state machine for the entry is as follows:: 6652 6653 dirtied harvested reset 6654 00 -----------> 01 -------------> 1X -------+ 6655 ^ | 6656 | | 6657 +------------------------------------------+ 6658 6659To harvest the dirty pages, userspace accesses the mmaped ring buffer 6660to read the dirty GFNs. If the flags has the DIRTY bit set (at this stage 6661the RESET bit must be cleared), then it means this GFN is a dirty GFN. 6662The userspace should harvest this GFN and mark the flags from state 6663``01b`` to ``1Xb`` (bit 0 will be ignored by KVM, but bit 1 must be set 6664to show that this GFN is harvested and waiting for a reset), and move 6665on to the next GFN. The userspace should continue to do this until the 6666flags of a GFN have the DIRTY bit cleared, meaning that it has harvested 6667all the dirty GFNs that were available. 6668 6669It's not necessary for userspace to harvest the all dirty GFNs at once. 6670However it must collect the dirty GFNs in sequence, i.e., the userspace 6671program cannot skip one dirty GFN to collect the one next to it. 6672 6673After processing one or more entries in the ring buffer, userspace 6674calls the VM ioctl KVM_RESET_DIRTY_RINGS to notify the kernel about 6675it, so that the kernel will reprotect those collected GFNs. 6676Therefore, the ioctl must be called *before* reading the content of 6677the dirty pages. 6678 6679The dirty ring can get full. When it happens, the KVM_RUN of the 6680vcpu will return with exit reason KVM_EXIT_DIRTY_LOG_FULL. 6681 6682The dirty ring interface has a major difference comparing to the 6683KVM_GET_DIRTY_LOG interface in that, when reading the dirty ring from 6684userspace, it's still possible that the kernel has not yet flushed the 6685processor's dirty page buffers into the kernel buffer (with dirty bitmaps, the 6686flushing is done by the KVM_GET_DIRTY_LOG ioctl). To achieve that, one 6687needs to kick the vcpu out of KVM_RUN using a signal. The resulting 6688vmexit ensures that all dirty GFNs are flushed to the dirty rings. 6689 6690NOTE: the capability KVM_CAP_DIRTY_LOG_RING and the corresponding 6691ioctl KVM_RESET_DIRTY_RINGS are mutual exclusive to the existing ioctls 6692KVM_GET_DIRTY_LOG and KVM_CLEAR_DIRTY_LOG. After enabling 6693KVM_CAP_DIRTY_LOG_RING with an acceptable dirty ring size, the virtual 6694machine will switch to ring-buffer dirty page tracking and further 6695KVM_GET_DIRTY_LOG or KVM_CLEAR_DIRTY_LOG ioctls will fail. 6696 66978.30 KVM_CAP_XEN_HVM 6698-------------------- 6699 6700:Architectures: x86 6701 6702This capability indicates the features that Xen supports for hosting Xen 6703PVHVM guests. Valid flags are:: 6704 6705 #define KVM_XEN_HVM_CONFIG_HYPERCALL_MSR (1 << 0) 6706 #define KVM_XEN_HVM_CONFIG_INTERCEPT_HCALL (1 << 1) 6707 #define KVM_XEN_HVM_CONFIG_SHARED_INFO (1 << 2) 6708 #define KVM_XEN_HVM_CONFIG_RUNSTATE (1 << 2) 6709 6710The KVM_XEN_HVM_CONFIG_HYPERCALL_MSR flag indicates that the KVM_XEN_HVM_CONFIG 6711ioctl is available, for the guest to set its hypercall page. 6712 6713If KVM_XEN_HVM_CONFIG_INTERCEPT_HCALL is also set, the same flag may also be 6714provided in the flags to KVM_XEN_HVM_CONFIG, without providing hypercall page 6715contents, to request that KVM generate hypercall page content automatically 6716and also enable interception of guest hypercalls with KVM_EXIT_XEN. 6717 6718The KVM_XEN_HVM_CONFIG_SHARED_INFO flag indicates the availability of the 6719KVM_XEN_HVM_SET_ATTR, KVM_XEN_HVM_GET_ATTR, KVM_XEN_VCPU_SET_ATTR and 6720KVM_XEN_VCPU_GET_ATTR ioctls, as well as the delivery of exception vectors 6721for event channel upcalls when the evtchn_upcall_pending field of a vcpu's 6722vcpu_info is set. 6723 6724The KVM_XEN_HVM_CONFIG_RUNSTATE flag indicates that the runstate-related 6725features KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADDR/_CURRENT/_DATA/_ADJUST are 6726supported by the KVM_XEN_VCPU_SET_ATTR/KVM_XEN_VCPU_GET_ATTR ioctls. 6727