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 via 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 185Please note that configuring the IPA size does not affect the capability 186exposed by the guest CPUs in ID_AA64MMFR0_EL1[PARange]. It only affects 187size of the address translated by the stage2 level (guest physical to 188host physical address translations). 189 190 1914.3 KVM_GET_MSR_INDEX_LIST, KVM_GET_MSR_FEATURE_INDEX_LIST 192---------------------------------------------------------- 193 194:Capability: basic, KVM_CAP_GET_MSR_FEATURES for KVM_GET_MSR_FEATURE_INDEX_LIST 195:Architectures: x86 196:Type: system ioctl 197:Parameters: struct kvm_msr_list (in/out) 198:Returns: 0 on success; -1 on error 199 200Errors: 201 202 ====== ============================================================ 203 EFAULT the msr index list cannot be read from or written to 204 E2BIG the msr index list is to be to fit in the array specified by 205 the user. 206 ====== ============================================================ 207 208:: 209 210 struct kvm_msr_list { 211 __u32 nmsrs; /* number of msrs in entries */ 212 __u32 indices[0]; 213 }; 214 215The user fills in the size of the indices array in nmsrs, and in return 216kvm adjusts nmsrs to reflect the actual number of msrs and fills in the 217indices array with their numbers. 218 219KVM_GET_MSR_INDEX_LIST returns the guest msrs that are supported. The list 220varies by kvm version and host processor, but does not change otherwise. 221 222Note: if kvm indicates supports MCE (KVM_CAP_MCE), then the MCE bank MSRs are 223not returned in the MSR list, as different vcpus can have a different number 224of banks, as set via the KVM_X86_SETUP_MCE ioctl. 225 226KVM_GET_MSR_FEATURE_INDEX_LIST returns the list of MSRs that can be passed 227to the KVM_GET_MSRS system ioctl. This lets userspace probe host capabilities 228and processor features that are exposed via MSRs (e.g., VMX capabilities). 229This list also varies by kvm version and host processor, but does not change 230otherwise. 231 232 2334.4 KVM_CHECK_EXTENSION 234----------------------- 235 236:Capability: basic, KVM_CAP_CHECK_EXTENSION_VM for vm ioctl 237:Architectures: all 238:Type: system ioctl, vm ioctl 239:Parameters: extension identifier (KVM_CAP_*) 240:Returns: 0 if unsupported; 1 (or some other positive integer) if supported 241 242The API allows the application to query about extensions to the core 243kvm API. Userspace passes an extension identifier (an integer) and 244receives an integer that describes the extension availability. 245Generally 0 means no and 1 means yes, but some extensions may report 246additional information in the integer return value. 247 248Based on their initialization different VMs may have different capabilities. 249It is thus encouraged to use the vm ioctl to query for capabilities (available 250with KVM_CAP_CHECK_EXTENSION_VM on the vm fd) 251 2524.5 KVM_GET_VCPU_MMAP_SIZE 253-------------------------- 254 255:Capability: basic 256:Architectures: all 257:Type: system ioctl 258:Parameters: none 259:Returns: size of vcpu mmap area, in bytes 260 261The KVM_RUN ioctl (cf.) communicates with userspace via a shared 262memory region. This ioctl returns the size of that region. See the 263KVM_RUN documentation for details. 264 265 2664.6 KVM_SET_MEMORY_REGION 267------------------------- 268 269:Capability: basic 270:Architectures: all 271:Type: vm ioctl 272:Parameters: struct kvm_memory_region (in) 273:Returns: 0 on success, -1 on error 274 275This ioctl is obsolete and has been removed. 276 277 2784.7 KVM_CREATE_VCPU 279------------------- 280 281:Capability: basic 282:Architectures: all 283:Type: vm ioctl 284:Parameters: vcpu id (apic id on x86) 285:Returns: vcpu fd on success, -1 on error 286 287This API adds a vcpu to a virtual machine. No more than max_vcpus may be added. 288The vcpu id is an integer in the range [0, max_vcpu_id). 289 290The recommended max_vcpus value can be retrieved using the KVM_CAP_NR_VCPUS of 291the KVM_CHECK_EXTENSION ioctl() at run-time. 292The maximum possible value for max_vcpus can be retrieved using the 293KVM_CAP_MAX_VCPUS of the KVM_CHECK_EXTENSION ioctl() at run-time. 294 295If the KVM_CAP_NR_VCPUS does not exist, you should assume that max_vcpus is 4 296cpus max. 297If the KVM_CAP_MAX_VCPUS does not exist, you should assume that max_vcpus is 298same as the value returned from KVM_CAP_NR_VCPUS. 299 300The maximum possible value for max_vcpu_id can be retrieved using the 301KVM_CAP_MAX_VCPU_ID of the KVM_CHECK_EXTENSION ioctl() at run-time. 302 303If the KVM_CAP_MAX_VCPU_ID does not exist, you should assume that max_vcpu_id 304is the same as the value returned from KVM_CAP_MAX_VCPUS. 305 306On powerpc using book3s_hv mode, the vcpus are mapped onto virtual 307threads in one or more virtual CPU cores. (This is because the 308hardware requires all the hardware threads in a CPU core to be in the 309same partition.) The KVM_CAP_PPC_SMT capability indicates the number 310of vcpus per virtual core (vcore). The vcore id is obtained by 311dividing the vcpu id by the number of vcpus per vcore. The vcpus in a 312given vcore will always be in the same physical core as each other 313(though that might be a different physical core from time to time). 314Userspace can control the threading (SMT) mode of the guest by its 315allocation of vcpu ids. For example, if userspace wants 316single-threaded guest vcpus, it should make all vcpu ids be a multiple 317of the number of vcpus per vcore. 318 319For virtual cpus that have been created with S390 user controlled virtual 320machines, the resulting vcpu fd can be memory mapped at page offset 321KVM_S390_SIE_PAGE_OFFSET in order to obtain a memory map of the virtual 322cpu's hardware control block. 323 324 3254.8 KVM_GET_DIRTY_LOG (vm ioctl) 326-------------------------------- 327 328:Capability: basic 329:Architectures: all 330:Type: vm ioctl 331:Parameters: struct kvm_dirty_log (in/out) 332:Returns: 0 on success, -1 on error 333 334:: 335 336 /* for KVM_GET_DIRTY_LOG */ 337 struct kvm_dirty_log { 338 __u32 slot; 339 __u32 padding; 340 union { 341 void __user *dirty_bitmap; /* one bit per page */ 342 __u64 padding; 343 }; 344 }; 345 346Given a memory slot, return a bitmap containing any pages dirtied 347since the last call to this ioctl. Bit 0 is the first page in the 348memory slot. Ensure the entire structure is cleared to avoid padding 349issues. 350 351If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 specifies 352the address space for which you want to return the dirty bitmap. 353They must be less than the value that KVM_CHECK_EXTENSION returns for 354the KVM_CAP_MULTI_ADDRESS_SPACE capability. 355 356The bits in the dirty bitmap are cleared before the ioctl returns, unless 357KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 is enabled. For more information, 358see the description of the capability. 359 3604.9 KVM_SET_MEMORY_ALIAS 361------------------------ 362 363:Capability: basic 364:Architectures: x86 365:Type: vm ioctl 366:Parameters: struct kvm_memory_alias (in) 367:Returns: 0 (success), -1 (error) 368 369This ioctl is obsolete and has been removed. 370 371 3724.10 KVM_RUN 373------------ 374 375:Capability: basic 376:Architectures: all 377:Type: vcpu ioctl 378:Parameters: none 379:Returns: 0 on success, -1 on error 380 381Errors: 382 383 ===== ============================= 384 EINTR an unmasked signal is pending 385 ===== ============================= 386 387This ioctl is used to run a guest virtual cpu. While there are no 388explicit parameters, there is an implicit parameter block that can be 389obtained by mmap()ing the vcpu fd at offset 0, with the size given by 390KVM_GET_VCPU_MMAP_SIZE. The parameter block is formatted as a 'struct 391kvm_run' (see below). 392 393 3944.11 KVM_GET_REGS 395----------------- 396 397:Capability: basic 398:Architectures: all except ARM, arm64 399:Type: vcpu ioctl 400:Parameters: struct kvm_regs (out) 401:Returns: 0 on success, -1 on error 402 403Reads the general purpose registers from the vcpu. 404 405:: 406 407 /* x86 */ 408 struct kvm_regs { 409 /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */ 410 __u64 rax, rbx, rcx, rdx; 411 __u64 rsi, rdi, rsp, rbp; 412 __u64 r8, r9, r10, r11; 413 __u64 r12, r13, r14, r15; 414 __u64 rip, rflags; 415 }; 416 417 /* mips */ 418 struct kvm_regs { 419 /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */ 420 __u64 gpr[32]; 421 __u64 hi; 422 __u64 lo; 423 __u64 pc; 424 }; 425 426 4274.12 KVM_SET_REGS 428----------------- 429 430:Capability: basic 431:Architectures: all except ARM, arm64 432:Type: vcpu ioctl 433:Parameters: struct kvm_regs (in) 434:Returns: 0 on success, -1 on error 435 436Writes the general purpose registers into the vcpu. 437 438See KVM_GET_REGS for the data structure. 439 440 4414.13 KVM_GET_SREGS 442------------------ 443 444:Capability: basic 445:Architectures: x86, ppc 446:Type: vcpu ioctl 447:Parameters: struct kvm_sregs (out) 448:Returns: 0 on success, -1 on error 449 450Reads special registers from the vcpu. 451 452:: 453 454 /* x86 */ 455 struct kvm_sregs { 456 struct kvm_segment cs, ds, es, fs, gs, ss; 457 struct kvm_segment tr, ldt; 458 struct kvm_dtable gdt, idt; 459 __u64 cr0, cr2, cr3, cr4, cr8; 460 __u64 efer; 461 __u64 apic_base; 462 __u64 interrupt_bitmap[(KVM_NR_INTERRUPTS + 63) / 64]; 463 }; 464 465 /* ppc -- see arch/powerpc/include/uapi/asm/kvm.h */ 466 467interrupt_bitmap is a bitmap of pending external interrupts. At most 468one bit may be set. This interrupt has been acknowledged by the APIC 469but not yet injected into the cpu core. 470 471 4724.14 KVM_SET_SREGS 473------------------ 474 475:Capability: basic 476:Architectures: x86, ppc 477:Type: vcpu ioctl 478:Parameters: struct kvm_sregs (in) 479:Returns: 0 on success, -1 on error 480 481Writes special registers into the vcpu. See KVM_GET_SREGS for the 482data structures. 483 484 4854.15 KVM_TRANSLATE 486------------------ 487 488:Capability: basic 489:Architectures: x86 490:Type: vcpu ioctl 491:Parameters: struct kvm_translation (in/out) 492:Returns: 0 on success, -1 on error 493 494Translates a virtual address according to the vcpu's current address 495translation mode. 496 497:: 498 499 struct kvm_translation { 500 /* in */ 501 __u64 linear_address; 502 503 /* out */ 504 __u64 physical_address; 505 __u8 valid; 506 __u8 writeable; 507 __u8 usermode; 508 __u8 pad[5]; 509 }; 510 511 5124.16 KVM_INTERRUPT 513------------------ 514 515:Capability: basic 516:Architectures: x86, ppc, mips 517:Type: vcpu ioctl 518:Parameters: struct kvm_interrupt (in) 519:Returns: 0 on success, negative on failure. 520 521Queues a hardware interrupt vector to be injected. 522 523:: 524 525 /* for KVM_INTERRUPT */ 526 struct kvm_interrupt { 527 /* in */ 528 __u32 irq; 529 }; 530 531X86: 532^^^^ 533 534:Returns: 535 536 ========= =================================== 537 0 on success, 538 -EEXIST if an interrupt is already enqueued 539 -EINVAL the the irq number is invalid 540 -ENXIO if the PIC is in the kernel 541 -EFAULT if the pointer is invalid 542 ========= =================================== 543 544Note 'irq' is an interrupt vector, not an interrupt pin or line. This 545ioctl is useful if the in-kernel PIC is not used. 546 547PPC: 548^^^^ 549 550Queues an external interrupt to be injected. This ioctl is overleaded 551with 3 different irq values: 552 553a) KVM_INTERRUPT_SET 554 555 This injects an edge type external interrupt into the guest once it's ready 556 to receive interrupts. When injected, the interrupt is done. 557 558b) KVM_INTERRUPT_UNSET 559 560 This unsets any pending interrupt. 561 562 Only available with KVM_CAP_PPC_UNSET_IRQ. 563 564c) KVM_INTERRUPT_SET_LEVEL 565 566 This injects a level type external interrupt into the guest context. The 567 interrupt stays pending until a specific ioctl with KVM_INTERRUPT_UNSET 568 is triggered. 569 570 Only available with KVM_CAP_PPC_IRQ_LEVEL. 571 572Note that any value for 'irq' other than the ones stated above is invalid 573and incurs unexpected behavior. 574 575This is an asynchronous vcpu ioctl and can be invoked from any thread. 576 577MIPS: 578^^^^^ 579 580Queues an external interrupt to be injected into the virtual CPU. A negative 581interrupt number dequeues the interrupt. 582 583This is an asynchronous vcpu ioctl and can be invoked from any thread. 584 585 5864.17 KVM_DEBUG_GUEST 587-------------------- 588 589:Capability: basic 590:Architectures: none 591:Type: vcpu ioctl 592:Parameters: none) 593:Returns: -1 on error 594 595Support for this has been removed. Use KVM_SET_GUEST_DEBUG instead. 596 597 5984.18 KVM_GET_MSRS 599----------------- 600 601:Capability: basic (vcpu), KVM_CAP_GET_MSR_FEATURES (system) 602:Architectures: x86 603:Type: system ioctl, vcpu ioctl 604:Parameters: struct kvm_msrs (in/out) 605:Returns: number of msrs successfully returned; 606 -1 on error 607 608When used as a system ioctl: 609Reads the values of MSR-based features that are available for the VM. This 610is similar to KVM_GET_SUPPORTED_CPUID, but it returns MSR indices and values. 611The list of msr-based features can be obtained using KVM_GET_MSR_FEATURE_INDEX_LIST 612in a system ioctl. 613 614When used as a vcpu ioctl: 615Reads model-specific registers from the vcpu. Supported msr indices can 616be obtained using KVM_GET_MSR_INDEX_LIST in a system ioctl. 617 618:: 619 620 struct kvm_msrs { 621 __u32 nmsrs; /* number of msrs in entries */ 622 __u32 pad; 623 624 struct kvm_msr_entry entries[0]; 625 }; 626 627 struct kvm_msr_entry { 628 __u32 index; 629 __u32 reserved; 630 __u64 data; 631 }; 632 633Application code should set the 'nmsrs' member (which indicates the 634size of the entries array) and the 'index' member of each array entry. 635kvm will fill in the 'data' member. 636 637 6384.19 KVM_SET_MSRS 639----------------- 640 641:Capability: basic 642:Architectures: x86 643:Type: vcpu ioctl 644:Parameters: struct kvm_msrs (in) 645:Returns: number of msrs successfully set (see below), -1 on error 646 647Writes model-specific registers to the vcpu. See KVM_GET_MSRS for the 648data structures. 649 650Application code should set the 'nmsrs' member (which indicates the 651size of the entries array), and the 'index' and 'data' members of each 652array entry. 653 654It tries to set the MSRs in array entries[] one by one. If setting an MSR 655fails, e.g., due to setting reserved bits, the MSR isn't supported/emulated 656by KVM, etc..., it stops processing the MSR list and returns the number of 657MSRs that have been set successfully. 658 659 6604.20 KVM_SET_CPUID 661------------------ 662 663:Capability: basic 664:Architectures: x86 665:Type: vcpu ioctl 666:Parameters: struct kvm_cpuid (in) 667:Returns: 0 on success, -1 on error 668 669Defines the vcpu responses to the cpuid instruction. Applications 670should use the KVM_SET_CPUID2 ioctl if available. 671 672:: 673 674 struct kvm_cpuid_entry { 675 __u32 function; 676 __u32 eax; 677 __u32 ebx; 678 __u32 ecx; 679 __u32 edx; 680 __u32 padding; 681 }; 682 683 /* for KVM_SET_CPUID */ 684 struct kvm_cpuid { 685 __u32 nent; 686 __u32 padding; 687 struct kvm_cpuid_entry entries[0]; 688 }; 689 690 6914.21 KVM_SET_SIGNAL_MASK 692------------------------ 693 694:Capability: basic 695:Architectures: all 696:Type: vcpu ioctl 697:Parameters: struct kvm_signal_mask (in) 698:Returns: 0 on success, -1 on error 699 700Defines which signals are blocked during execution of KVM_RUN. This 701signal mask temporarily overrides the threads signal mask. Any 702unblocked signal received (except SIGKILL and SIGSTOP, which retain 703their traditional behaviour) will cause KVM_RUN to return with -EINTR. 704 705Note the signal will only be delivered if not blocked by the original 706signal mask. 707 708:: 709 710 /* for KVM_SET_SIGNAL_MASK */ 711 struct kvm_signal_mask { 712 __u32 len; 713 __u8 sigset[0]; 714 }; 715 716 7174.22 KVM_GET_FPU 718---------------- 719 720:Capability: basic 721:Architectures: x86 722:Type: vcpu ioctl 723:Parameters: struct kvm_fpu (out) 724:Returns: 0 on success, -1 on error 725 726Reads the floating point state from the vcpu. 727 728:: 729 730 /* for KVM_GET_FPU and KVM_SET_FPU */ 731 struct kvm_fpu { 732 __u8 fpr[8][16]; 733 __u16 fcw; 734 __u16 fsw; 735 __u8 ftwx; /* in fxsave format */ 736 __u8 pad1; 737 __u16 last_opcode; 738 __u64 last_ip; 739 __u64 last_dp; 740 __u8 xmm[16][16]; 741 __u32 mxcsr; 742 __u32 pad2; 743 }; 744 745 7464.23 KVM_SET_FPU 747---------------- 748 749:Capability: basic 750:Architectures: x86 751:Type: vcpu ioctl 752:Parameters: struct kvm_fpu (in) 753:Returns: 0 on success, -1 on error 754 755Writes the floating point state to the vcpu. 756 757:: 758 759 /* for KVM_GET_FPU and KVM_SET_FPU */ 760 struct kvm_fpu { 761 __u8 fpr[8][16]; 762 __u16 fcw; 763 __u16 fsw; 764 __u8 ftwx; /* in fxsave format */ 765 __u8 pad1; 766 __u16 last_opcode; 767 __u64 last_ip; 768 __u64 last_dp; 769 __u8 xmm[16][16]; 770 __u32 mxcsr; 771 __u32 pad2; 772 }; 773 774 7754.24 KVM_CREATE_IRQCHIP 776----------------------- 777 778:Capability: KVM_CAP_IRQCHIP, KVM_CAP_S390_IRQCHIP (s390) 779:Architectures: x86, ARM, arm64, s390 780:Type: vm ioctl 781:Parameters: none 782:Returns: 0 on success, -1 on error 783 784Creates an interrupt controller model in the kernel. 785On x86, creates a virtual ioapic, a virtual PIC (two PICs, nested), and sets up 786future vcpus to have a local APIC. IRQ routing for GSIs 0-15 is set to both 787PIC and IOAPIC; GSI 16-23 only go to the IOAPIC. 788On ARM/arm64, a GICv2 is created. Any other GIC versions require the usage of 789KVM_CREATE_DEVICE, which also supports creating a GICv2. Using 790KVM_CREATE_DEVICE is preferred over KVM_CREATE_IRQCHIP for GICv2. 791On s390, a dummy irq routing table is created. 792 793Note that on s390 the KVM_CAP_S390_IRQCHIP vm capability needs to be enabled 794before KVM_CREATE_IRQCHIP can be used. 795 796 7974.25 KVM_IRQ_LINE 798----------------- 799 800:Capability: KVM_CAP_IRQCHIP 801:Architectures: x86, arm, arm64 802:Type: vm ioctl 803:Parameters: struct kvm_irq_level 804:Returns: 0 on success, -1 on error 805 806Sets the level of a GSI input to the interrupt controller model in the kernel. 807On some architectures it is required that an interrupt controller model has 808been previously created with KVM_CREATE_IRQCHIP. Note that edge-triggered 809interrupts require the level to be set to 1 and then back to 0. 810 811On real hardware, interrupt pins can be active-low or active-high. This 812does not matter for the level field of struct kvm_irq_level: 1 always 813means active (asserted), 0 means inactive (deasserted). 814 815x86 allows the operating system to program the interrupt polarity 816(active-low/active-high) for level-triggered interrupts, and KVM used 817to consider the polarity. However, due to bitrot in the handling of 818active-low interrupts, the above convention is now valid on x86 too. 819This is signaled by KVM_CAP_X86_IOAPIC_POLARITY_IGNORED. Userspace 820should not present interrupts to the guest as active-low unless this 821capability is present (or unless it is not using the in-kernel irqchip, 822of course). 823 824 825ARM/arm64 can signal an interrupt either at the CPU level, or at the 826in-kernel irqchip (GIC), and for in-kernel irqchip can tell the GIC to 827use PPIs designated for specific cpus. The irq field is interpreted 828like this:: 829 830 bits: | 31 ... 28 | 27 ... 24 | 23 ... 16 | 15 ... 0 | 831 field: | vcpu2_index | irq_type | vcpu_index | irq_id | 832 833The irq_type field has the following values: 834 835- irq_type[0]: 836 out-of-kernel GIC: irq_id 0 is IRQ, irq_id 1 is FIQ 837- irq_type[1]: 838 in-kernel GIC: SPI, irq_id between 32 and 1019 (incl.) 839 (the vcpu_index field is ignored) 840- irq_type[2]: 841 in-kernel GIC: PPI, irq_id between 16 and 31 (incl.) 842 843(The irq_id field thus corresponds nicely to the IRQ ID in the ARM GIC specs) 844 845In both cases, level is used to assert/deassert the line. 846 847When KVM_CAP_ARM_IRQ_LINE_LAYOUT_2 is supported, the target vcpu is 848identified as (256 * vcpu2_index + vcpu_index). Otherwise, vcpu2_index 849must be zero. 850 851Note that on arm/arm64, the KVM_CAP_IRQCHIP capability only conditions 852injection of interrupts for the in-kernel irqchip. KVM_IRQ_LINE can always 853be used for a userspace interrupt controller. 854 855:: 856 857 struct kvm_irq_level { 858 union { 859 __u32 irq; /* GSI */ 860 __s32 status; /* not used for KVM_IRQ_LEVEL */ 861 }; 862 __u32 level; /* 0 or 1 */ 863 }; 864 865 8664.26 KVM_GET_IRQCHIP 867-------------------- 868 869:Capability: KVM_CAP_IRQCHIP 870:Architectures: x86 871:Type: vm ioctl 872:Parameters: struct kvm_irqchip (in/out) 873:Returns: 0 on success, -1 on error 874 875Reads the state of a kernel interrupt controller created with 876KVM_CREATE_IRQCHIP into a buffer provided by the caller. 877 878:: 879 880 struct kvm_irqchip { 881 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */ 882 __u32 pad; 883 union { 884 char dummy[512]; /* reserving space */ 885 struct kvm_pic_state pic; 886 struct kvm_ioapic_state ioapic; 887 } chip; 888 }; 889 890 8914.27 KVM_SET_IRQCHIP 892-------------------- 893 894:Capability: KVM_CAP_IRQCHIP 895:Architectures: x86 896:Type: vm ioctl 897:Parameters: struct kvm_irqchip (in) 898:Returns: 0 on success, -1 on error 899 900Sets the state of a kernel interrupt controller created with 901KVM_CREATE_IRQCHIP from a buffer provided by the caller. 902 903:: 904 905 struct kvm_irqchip { 906 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */ 907 __u32 pad; 908 union { 909 char dummy[512]; /* reserving space */ 910 struct kvm_pic_state pic; 911 struct kvm_ioapic_state ioapic; 912 } chip; 913 }; 914 915 9164.28 KVM_XEN_HVM_CONFIG 917----------------------- 918 919:Capability: KVM_CAP_XEN_HVM 920:Architectures: x86 921:Type: vm ioctl 922:Parameters: struct kvm_xen_hvm_config (in) 923:Returns: 0 on success, -1 on error 924 925Sets the MSR that the Xen HVM guest uses to initialize its hypercall 926page, and provides the starting address and size of the hypercall 927blobs in userspace. When the guest writes the MSR, kvm copies one 928page of a blob (32- or 64-bit, depending on the vcpu mode) to guest 929memory. 930 931:: 932 933 struct kvm_xen_hvm_config { 934 __u32 flags; 935 __u32 msr; 936 __u64 blob_addr_32; 937 __u64 blob_addr_64; 938 __u8 blob_size_32; 939 __u8 blob_size_64; 940 __u8 pad2[30]; 941 }; 942 943 9444.29 KVM_GET_CLOCK 945------------------ 946 947:Capability: KVM_CAP_ADJUST_CLOCK 948:Architectures: x86 949:Type: vm ioctl 950:Parameters: struct kvm_clock_data (out) 951:Returns: 0 on success, -1 on error 952 953Gets the current timestamp of kvmclock as seen by the current guest. In 954conjunction with KVM_SET_CLOCK, it is used to ensure monotonicity on scenarios 955such as migration. 956 957When KVM_CAP_ADJUST_CLOCK is passed to KVM_CHECK_EXTENSION, it returns the 958set of bits that KVM can return in struct kvm_clock_data's flag member. 959 960The only flag defined now is KVM_CLOCK_TSC_STABLE. If set, the returned 961value is the exact kvmclock value seen by all VCPUs at the instant 962when KVM_GET_CLOCK was called. If clear, the returned value is simply 963CLOCK_MONOTONIC plus a constant offset; the offset can be modified 964with KVM_SET_CLOCK. KVM will try to make all VCPUs follow this clock, 965but the exact value read by each VCPU could differ, because the host 966TSC is not stable. 967 968:: 969 970 struct kvm_clock_data { 971 __u64 clock; /* kvmclock current value */ 972 __u32 flags; 973 __u32 pad[9]; 974 }; 975 976 9774.30 KVM_SET_CLOCK 978------------------ 979 980:Capability: KVM_CAP_ADJUST_CLOCK 981:Architectures: x86 982:Type: vm ioctl 983:Parameters: struct kvm_clock_data (in) 984:Returns: 0 on success, -1 on error 985 986Sets the current timestamp of kvmclock to the value specified in its parameter. 987In conjunction with KVM_GET_CLOCK, it is used to ensure monotonicity on scenarios 988such as migration. 989 990:: 991 992 struct kvm_clock_data { 993 __u64 clock; /* kvmclock current value */ 994 __u32 flags; 995 __u32 pad[9]; 996 }; 997 998 9994.31 KVM_GET_VCPU_EVENTS 1000------------------------ 1001 1002:Capability: KVM_CAP_VCPU_EVENTS 1003:Extended by: KVM_CAP_INTR_SHADOW 1004:Architectures: x86, arm, arm64 1005:Type: vcpu ioctl 1006:Parameters: struct kvm_vcpu_event (out) 1007:Returns: 0 on success, -1 on error 1008 1009X86: 1010^^^^ 1011 1012Gets currently pending exceptions, interrupts, and NMIs as well as related 1013states of the vcpu. 1014 1015:: 1016 1017 struct kvm_vcpu_events { 1018 struct { 1019 __u8 injected; 1020 __u8 nr; 1021 __u8 has_error_code; 1022 __u8 pending; 1023 __u32 error_code; 1024 } exception; 1025 struct { 1026 __u8 injected; 1027 __u8 nr; 1028 __u8 soft; 1029 __u8 shadow; 1030 } interrupt; 1031 struct { 1032 __u8 injected; 1033 __u8 pending; 1034 __u8 masked; 1035 __u8 pad; 1036 } nmi; 1037 __u32 sipi_vector; 1038 __u32 flags; 1039 struct { 1040 __u8 smm; 1041 __u8 pending; 1042 __u8 smm_inside_nmi; 1043 __u8 latched_init; 1044 } smi; 1045 __u8 reserved[27]; 1046 __u8 exception_has_payload; 1047 __u64 exception_payload; 1048 }; 1049 1050The following bits are defined in the flags field: 1051 1052- KVM_VCPUEVENT_VALID_SHADOW may be set to signal that 1053 interrupt.shadow contains a valid state. 1054 1055- KVM_VCPUEVENT_VALID_SMM may be set to signal that smi contains a 1056 valid state. 1057 1058- KVM_VCPUEVENT_VALID_PAYLOAD may be set to signal that the 1059 exception_has_payload, exception_payload, and exception.pending 1060 fields contain a valid state. This bit will be set whenever 1061 KVM_CAP_EXCEPTION_PAYLOAD is enabled. 1062 1063ARM/ARM64: 1064^^^^^^^^^^ 1065 1066If the guest accesses a device that is being emulated by the host kernel in 1067such a way that a real device would generate a physical SError, KVM may make 1068a virtual SError pending for that VCPU. This system error interrupt remains 1069pending until the guest takes the exception by unmasking PSTATE.A. 1070 1071Running the VCPU may cause it to take a pending SError, or make an access that 1072causes an SError to become pending. The event's description is only valid while 1073the VPCU is not running. 1074 1075This API provides a way to read and write the pending 'event' state that is not 1076visible to the guest. To save, restore or migrate a VCPU the struct representing 1077the state can be read then written using this GET/SET API, along with the other 1078guest-visible registers. It is not possible to 'cancel' an SError that has been 1079made pending. 1080 1081A device being emulated in user-space may also wish to generate an SError. To do 1082this the events structure can be populated by user-space. The current state 1083should be read first, to ensure no existing SError is pending. If an existing 1084SError is pending, the architecture's 'Multiple SError interrupts' rules should 1085be followed. (2.5.3 of DDI0587.a "ARM Reliability, Availability, and 1086Serviceability (RAS) Specification"). 1087 1088SError exceptions always have an ESR value. Some CPUs have the ability to 1089specify what the virtual SError's ESR value should be. These systems will 1090advertise KVM_CAP_ARM_INJECT_SERROR_ESR. In this case exception.has_esr will 1091always have a non-zero value when read, and the agent making an SError pending 1092should specify the ISS field in the lower 24 bits of exception.serror_esr. If 1093the system supports KVM_CAP_ARM_INJECT_SERROR_ESR, but user-space sets the events 1094with exception.has_esr as zero, KVM will choose an ESR. 1095 1096Specifying exception.has_esr on a system that does not support it will return 1097-EINVAL. Setting anything other than the lower 24bits of exception.serror_esr 1098will return -EINVAL. 1099 1100It is not possible to read back a pending external abort (injected via 1101KVM_SET_VCPU_EVENTS or otherwise) because such an exception is always delivered 1102directly to the virtual CPU). 1103 1104:: 1105 1106 struct kvm_vcpu_events { 1107 struct { 1108 __u8 serror_pending; 1109 __u8 serror_has_esr; 1110 __u8 ext_dabt_pending; 1111 /* Align it to 8 bytes */ 1112 __u8 pad[5]; 1113 __u64 serror_esr; 1114 } exception; 1115 __u32 reserved[12]; 1116 }; 1117 11184.32 KVM_SET_VCPU_EVENTS 1119------------------------ 1120 1121:Capability: KVM_CAP_VCPU_EVENTS 1122:Extended by: KVM_CAP_INTR_SHADOW 1123:Architectures: x86, arm, arm64 1124:Type: vcpu ioctl 1125:Parameters: struct kvm_vcpu_event (in) 1126:Returns: 0 on success, -1 on error 1127 1128X86: 1129^^^^ 1130 1131Set pending exceptions, interrupts, and NMIs as well as related states of the 1132vcpu. 1133 1134See KVM_GET_VCPU_EVENTS for the data structure. 1135 1136Fields that may be modified asynchronously by running VCPUs can be excluded 1137from the update. These fields are nmi.pending, sipi_vector, smi.smm, 1138smi.pending. Keep the corresponding bits in the flags field cleared to 1139suppress overwriting the current in-kernel state. The bits are: 1140 1141=============================== ================================== 1142KVM_VCPUEVENT_VALID_NMI_PENDING transfer nmi.pending to the kernel 1143KVM_VCPUEVENT_VALID_SIPI_VECTOR transfer sipi_vector 1144KVM_VCPUEVENT_VALID_SMM transfer the smi sub-struct. 1145=============================== ================================== 1146 1147If KVM_CAP_INTR_SHADOW is available, KVM_VCPUEVENT_VALID_SHADOW can be set in 1148the flags field to signal that interrupt.shadow contains a valid state and 1149shall be written into the VCPU. 1150 1151KVM_VCPUEVENT_VALID_SMM can only be set if KVM_CAP_X86_SMM is available. 1152 1153If KVM_CAP_EXCEPTION_PAYLOAD is enabled, KVM_VCPUEVENT_VALID_PAYLOAD 1154can be set in the flags field to signal that the 1155exception_has_payload, exception_payload, and exception.pending fields 1156contain a valid state and shall be written into the VCPU. 1157 1158ARM/ARM64: 1159^^^^^^^^^^ 1160 1161User space may need to inject several types of events to the guest. 1162 1163Set the pending SError exception state for this VCPU. It is not possible to 1164'cancel' an Serror that has been made pending. 1165 1166If the guest performed an access to I/O memory which could not be handled by 1167userspace, for example because of missing instruction syndrome decode 1168information or because there is no device mapped at the accessed IPA, then 1169userspace can ask the kernel to inject an external abort using the address 1170from the exiting fault on the VCPU. It is a programming error to set 1171ext_dabt_pending after an exit which was not either KVM_EXIT_MMIO or 1172KVM_EXIT_ARM_NISV. This feature is only available if the system supports 1173KVM_CAP_ARM_INJECT_EXT_DABT. This is a helper which provides commonality in 1174how userspace reports accesses for the above cases to guests, across different 1175userspace implementations. Nevertheless, userspace can still emulate all Arm 1176exceptions by manipulating individual registers using the KVM_SET_ONE_REG API. 1177 1178See KVM_GET_VCPU_EVENTS for the data structure. 1179 1180 11814.33 KVM_GET_DEBUGREGS 1182---------------------- 1183 1184:Capability: KVM_CAP_DEBUGREGS 1185:Architectures: x86 1186:Type: vm ioctl 1187:Parameters: struct kvm_debugregs (out) 1188:Returns: 0 on success, -1 on error 1189 1190Reads debug registers from the vcpu. 1191 1192:: 1193 1194 struct kvm_debugregs { 1195 __u64 db[4]; 1196 __u64 dr6; 1197 __u64 dr7; 1198 __u64 flags; 1199 __u64 reserved[9]; 1200 }; 1201 1202 12034.34 KVM_SET_DEBUGREGS 1204---------------------- 1205 1206:Capability: KVM_CAP_DEBUGREGS 1207:Architectures: x86 1208:Type: vm ioctl 1209:Parameters: struct kvm_debugregs (in) 1210:Returns: 0 on success, -1 on error 1211 1212Writes debug registers into the vcpu. 1213 1214See KVM_GET_DEBUGREGS for the data structure. The flags field is unused 1215yet and must be cleared on entry. 1216 1217 12184.35 KVM_SET_USER_MEMORY_REGION 1219------------------------------- 1220 1221:Capability: KVM_CAP_USER_MEMORY 1222:Architectures: all 1223:Type: vm ioctl 1224:Parameters: struct kvm_userspace_memory_region (in) 1225:Returns: 0 on success, -1 on error 1226 1227:: 1228 1229 struct kvm_userspace_memory_region { 1230 __u32 slot; 1231 __u32 flags; 1232 __u64 guest_phys_addr; 1233 __u64 memory_size; /* bytes */ 1234 __u64 userspace_addr; /* start of the userspace allocated memory */ 1235 }; 1236 1237 /* for kvm_memory_region::flags */ 1238 #define KVM_MEM_LOG_DIRTY_PAGES (1UL << 0) 1239 #define KVM_MEM_READONLY (1UL << 1) 1240 1241This ioctl allows the user to create, modify or delete a guest physical 1242memory slot. Bits 0-15 of "slot" specify the slot id and this value 1243should be less than the maximum number of user memory slots supported per 1244VM. The maximum allowed slots can be queried using KVM_CAP_NR_MEMSLOTS. 1245Slots may not overlap in guest physical address space. 1246 1247If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of "slot" 1248specifies the address space which is being modified. They must be 1249less than the value that KVM_CHECK_EXTENSION returns for the 1250KVM_CAP_MULTI_ADDRESS_SPACE capability. Slots in separate address spaces 1251are unrelated; the restriction on overlapping slots only applies within 1252each address space. 1253 1254Deleting a slot is done by passing zero for memory_size. When changing 1255an existing slot, it may be moved in the guest physical memory space, 1256or its flags may be modified, but it may not be resized. 1257 1258Memory for the region is taken starting at the address denoted by the 1259field userspace_addr, which must point at user addressable memory for 1260the entire memory slot size. Any object may back this memory, including 1261anonymous memory, ordinary files, and hugetlbfs. 1262 1263It is recommended that the lower 21 bits of guest_phys_addr and userspace_addr 1264be identical. This allows large pages in the guest to be backed by large 1265pages in the host. 1266 1267The flags field supports two flags: KVM_MEM_LOG_DIRTY_PAGES and 1268KVM_MEM_READONLY. The former can be set to instruct KVM to keep track of 1269writes to memory within the slot. See KVM_GET_DIRTY_LOG ioctl to know how to 1270use it. The latter can be set, if KVM_CAP_READONLY_MEM capability allows it, 1271to make a new slot read-only. In this case, writes to this memory will be 1272posted to userspace as KVM_EXIT_MMIO exits. 1273 1274When the KVM_CAP_SYNC_MMU capability is available, changes in the backing of 1275the memory region are automatically reflected into the guest. For example, an 1276mmap() that affects the region will be made visible immediately. Another 1277example is madvise(MADV_DROP). 1278 1279It is recommended to use this API instead of the KVM_SET_MEMORY_REGION ioctl. 1280The KVM_SET_MEMORY_REGION does not allow fine grained control over memory 1281allocation and is deprecated. 1282 1283 12844.36 KVM_SET_TSS_ADDR 1285--------------------- 1286 1287:Capability: KVM_CAP_SET_TSS_ADDR 1288:Architectures: x86 1289:Type: vm ioctl 1290:Parameters: unsigned long tss_address (in) 1291:Returns: 0 on success, -1 on error 1292 1293This ioctl defines the physical address of a three-page region in the guest 1294physical address space. The region must be within the first 4GB of the 1295guest physical address space and must not conflict with any memory slot 1296or any mmio address. The guest may malfunction if it accesses this memory 1297region. 1298 1299This ioctl is required on Intel-based hosts. This is needed on Intel hardware 1300because of a quirk in the virtualization implementation (see the internals 1301documentation when it pops into existence). 1302 1303 13044.37 KVM_ENABLE_CAP 1305------------------- 1306 1307:Capability: KVM_CAP_ENABLE_CAP 1308:Architectures: mips, ppc, s390 1309:Type: vcpu ioctl 1310:Parameters: struct kvm_enable_cap (in) 1311:Returns: 0 on success; -1 on error 1312 1313:Capability: KVM_CAP_ENABLE_CAP_VM 1314:Architectures: all 1315:Type: vcpu ioctl 1316:Parameters: struct kvm_enable_cap (in) 1317:Returns: 0 on success; -1 on error 1318 1319.. note:: 1320 1321 Not all extensions are enabled by default. Using this ioctl the application 1322 can enable an extension, making it available to the guest. 1323 1324On systems that do not support this ioctl, it always fails. On systems that 1325do support it, it only works for extensions that are supported for enablement. 1326 1327To check if a capability can be enabled, the KVM_CHECK_EXTENSION ioctl should 1328be used. 1329 1330:: 1331 1332 struct kvm_enable_cap { 1333 /* in */ 1334 __u32 cap; 1335 1336The capability that is supposed to get enabled. 1337 1338:: 1339 1340 __u32 flags; 1341 1342A bitfield indicating future enhancements. Has to be 0 for now. 1343 1344:: 1345 1346 __u64 args[4]; 1347 1348Arguments for enabling a feature. If a feature needs initial values to 1349function properly, this is the place to put them. 1350 1351:: 1352 1353 __u8 pad[64]; 1354 }; 1355 1356The vcpu ioctl should be used for vcpu-specific capabilities, the vm ioctl 1357for vm-wide capabilities. 1358 13594.38 KVM_GET_MP_STATE 1360--------------------- 1361 1362:Capability: KVM_CAP_MP_STATE 1363:Architectures: x86, s390, arm, arm64 1364:Type: vcpu ioctl 1365:Parameters: struct kvm_mp_state (out) 1366:Returns: 0 on success; -1 on error 1367 1368:: 1369 1370 struct kvm_mp_state { 1371 __u32 mp_state; 1372 }; 1373 1374Returns the vcpu's current "multiprocessing state" (though also valid on 1375uniprocessor guests). 1376 1377Possible values are: 1378 1379 ========================== =============================================== 1380 KVM_MP_STATE_RUNNABLE the vcpu is currently running [x86,arm/arm64] 1381 KVM_MP_STATE_UNINITIALIZED the vcpu is an application processor (AP) 1382 which has not yet received an INIT signal [x86] 1383 KVM_MP_STATE_INIT_RECEIVED the vcpu has received an INIT signal, and is 1384 now ready for a SIPI [x86] 1385 KVM_MP_STATE_HALTED the vcpu has executed a HLT instruction and 1386 is waiting for an interrupt [x86] 1387 KVM_MP_STATE_SIPI_RECEIVED the vcpu has just received a SIPI (vector 1388 accessible via KVM_GET_VCPU_EVENTS) [x86] 1389 KVM_MP_STATE_STOPPED the vcpu is stopped [s390,arm/arm64] 1390 KVM_MP_STATE_CHECK_STOP the vcpu is in a special error state [s390] 1391 KVM_MP_STATE_OPERATING the vcpu is operating (running or halted) 1392 [s390] 1393 KVM_MP_STATE_LOAD the vcpu is in a special load/startup state 1394 [s390] 1395 ========================== =============================================== 1396 1397On x86, this ioctl is only useful after KVM_CREATE_IRQCHIP. Without an 1398in-kernel irqchip, the multiprocessing state must be maintained by userspace on 1399these architectures. 1400 1401For arm/arm64: 1402^^^^^^^^^^^^^^ 1403 1404The only states that are valid are KVM_MP_STATE_STOPPED and 1405KVM_MP_STATE_RUNNABLE which reflect if the vcpu is paused or not. 1406 14074.39 KVM_SET_MP_STATE 1408--------------------- 1409 1410:Capability: KVM_CAP_MP_STATE 1411:Architectures: x86, s390, arm, arm64 1412:Type: vcpu ioctl 1413:Parameters: struct kvm_mp_state (in) 1414:Returns: 0 on success; -1 on error 1415 1416Sets the vcpu's current "multiprocessing state"; see KVM_GET_MP_STATE for 1417arguments. 1418 1419On x86, this ioctl is only useful after KVM_CREATE_IRQCHIP. Without an 1420in-kernel irqchip, the multiprocessing state must be maintained by userspace on 1421these architectures. 1422 1423For arm/arm64: 1424^^^^^^^^^^^^^^ 1425 1426The only states that are valid are KVM_MP_STATE_STOPPED and 1427KVM_MP_STATE_RUNNABLE which reflect if the vcpu should be paused or not. 1428 14294.40 KVM_SET_IDENTITY_MAP_ADDR 1430------------------------------ 1431 1432:Capability: KVM_CAP_SET_IDENTITY_MAP_ADDR 1433:Architectures: x86 1434:Type: vm ioctl 1435:Parameters: unsigned long identity (in) 1436:Returns: 0 on success, -1 on error 1437 1438This ioctl defines the physical address of a one-page region in the guest 1439physical address space. The region must be within the first 4GB of the 1440guest physical address space and must not conflict with any memory slot 1441or any mmio address. The guest may malfunction if it accesses this memory 1442region. 1443 1444Setting the address to 0 will result in resetting the address to its default 1445(0xfffbc000). 1446 1447This ioctl is required on Intel-based hosts. This is needed on Intel hardware 1448because of a quirk in the virtualization implementation (see the internals 1449documentation when it pops into existence). 1450 1451Fails if any VCPU has already been created. 1452 14534.41 KVM_SET_BOOT_CPU_ID 1454------------------------ 1455 1456:Capability: KVM_CAP_SET_BOOT_CPU_ID 1457:Architectures: x86 1458:Type: vm ioctl 1459:Parameters: unsigned long vcpu_id 1460:Returns: 0 on success, -1 on error 1461 1462Define which vcpu is the Bootstrap Processor (BSP). Values are the same 1463as the vcpu id in KVM_CREATE_VCPU. If this ioctl is not called, the default 1464is vcpu 0. 1465 1466 14674.42 KVM_GET_XSAVE 1468------------------ 1469 1470:Capability: KVM_CAP_XSAVE 1471:Architectures: x86 1472:Type: vcpu ioctl 1473:Parameters: struct kvm_xsave (out) 1474:Returns: 0 on success, -1 on error 1475 1476 1477:: 1478 1479 struct kvm_xsave { 1480 __u32 region[1024]; 1481 }; 1482 1483This ioctl would copy current vcpu's xsave struct to the userspace. 1484 1485 14864.43 KVM_SET_XSAVE 1487------------------ 1488 1489:Capability: KVM_CAP_XSAVE 1490:Architectures: x86 1491:Type: vcpu ioctl 1492:Parameters: struct kvm_xsave (in) 1493:Returns: 0 on success, -1 on error 1494 1495:: 1496 1497 1498 struct kvm_xsave { 1499 __u32 region[1024]; 1500 }; 1501 1502This ioctl would copy userspace's xsave struct to the kernel. 1503 1504 15054.44 KVM_GET_XCRS 1506----------------- 1507 1508:Capability: KVM_CAP_XCRS 1509:Architectures: x86 1510:Type: vcpu ioctl 1511:Parameters: struct kvm_xcrs (out) 1512:Returns: 0 on success, -1 on error 1513 1514:: 1515 1516 struct kvm_xcr { 1517 __u32 xcr; 1518 __u32 reserved; 1519 __u64 value; 1520 }; 1521 1522 struct kvm_xcrs { 1523 __u32 nr_xcrs; 1524 __u32 flags; 1525 struct kvm_xcr xcrs[KVM_MAX_XCRS]; 1526 __u64 padding[16]; 1527 }; 1528 1529This ioctl would copy current vcpu's xcrs to the userspace. 1530 1531 15324.45 KVM_SET_XCRS 1533----------------- 1534 1535:Capability: KVM_CAP_XCRS 1536:Architectures: x86 1537:Type: vcpu ioctl 1538:Parameters: struct kvm_xcrs (in) 1539:Returns: 0 on success, -1 on error 1540 1541:: 1542 1543 struct kvm_xcr { 1544 __u32 xcr; 1545 __u32 reserved; 1546 __u64 value; 1547 }; 1548 1549 struct kvm_xcrs { 1550 __u32 nr_xcrs; 1551 __u32 flags; 1552 struct kvm_xcr xcrs[KVM_MAX_XCRS]; 1553 __u64 padding[16]; 1554 }; 1555 1556This ioctl would set vcpu's xcr to the value userspace specified. 1557 1558 15594.46 KVM_GET_SUPPORTED_CPUID 1560---------------------------- 1561 1562:Capability: KVM_CAP_EXT_CPUID 1563:Architectures: x86 1564:Type: system ioctl 1565:Parameters: struct kvm_cpuid2 (in/out) 1566:Returns: 0 on success, -1 on error 1567 1568:: 1569 1570 struct kvm_cpuid2 { 1571 __u32 nent; 1572 __u32 padding; 1573 struct kvm_cpuid_entry2 entries[0]; 1574 }; 1575 1576 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX BIT(0) 1577 #define KVM_CPUID_FLAG_STATEFUL_FUNC BIT(1) 1578 #define KVM_CPUID_FLAG_STATE_READ_NEXT BIT(2) 1579 1580 struct kvm_cpuid_entry2 { 1581 __u32 function; 1582 __u32 index; 1583 __u32 flags; 1584 __u32 eax; 1585 __u32 ebx; 1586 __u32 ecx; 1587 __u32 edx; 1588 __u32 padding[3]; 1589 }; 1590 1591This ioctl returns x86 cpuid features which are supported by both the 1592hardware and kvm in its default configuration. Userspace can use the 1593information returned by this ioctl to construct cpuid information (for 1594KVM_SET_CPUID2) that is consistent with hardware, kernel, and 1595userspace capabilities, and with user requirements (for example, the 1596user may wish to constrain cpuid to emulate older hardware, or for 1597feature consistency across a cluster). 1598 1599Note that certain capabilities, such as KVM_CAP_X86_DISABLE_EXITS, may 1600expose cpuid features (e.g. MONITOR) which are not supported by kvm in 1601its default configuration. If userspace enables such capabilities, it 1602is responsible for modifying the results of this ioctl appropriately. 1603 1604Userspace invokes KVM_GET_SUPPORTED_CPUID by passing a kvm_cpuid2 structure 1605with the 'nent' field indicating the number of entries in the variable-size 1606array 'entries'. If the number of entries is too low to describe the cpu 1607capabilities, an error (E2BIG) is returned. If the number is too high, 1608the 'nent' field is adjusted and an error (ENOMEM) is returned. If the 1609number is just right, the 'nent' field is adjusted to the number of valid 1610entries in the 'entries' array, which is then filled. 1611 1612The entries returned are the host cpuid as returned by the cpuid instruction, 1613with unknown or unsupported features masked out. Some features (for example, 1614x2apic), may not be present in the host cpu, but are exposed by kvm if it can 1615emulate them efficiently. The fields in each entry are defined as follows: 1616 1617 function: 1618 the eax value used to obtain the entry 1619 1620 index: 1621 the ecx value used to obtain the entry (for entries that are 1622 affected by ecx) 1623 1624 flags: 1625 an OR of zero or more of the following: 1626 1627 KVM_CPUID_FLAG_SIGNIFCANT_INDEX: 1628 if the index field is valid 1629 KVM_CPUID_FLAG_STATEFUL_FUNC: 1630 if cpuid for this function returns different values for successive 1631 invocations; there will be several entries with the same function, 1632 all with this flag set 1633 KVM_CPUID_FLAG_STATE_READ_NEXT: 1634 for KVM_CPUID_FLAG_STATEFUL_FUNC entries, set if this entry is 1635 the first entry to be read by a cpu 1636 1637 eax, ebx, ecx, edx: 1638 the values returned by the cpuid instruction for 1639 this function/index combination 1640 1641The TSC deadline timer feature (CPUID leaf 1, ecx[24]) is always returned 1642as false, since the feature depends on KVM_CREATE_IRQCHIP for local APIC 1643support. Instead it is reported via:: 1644 1645 ioctl(KVM_CHECK_EXTENSION, KVM_CAP_TSC_DEADLINE_TIMER) 1646 1647if that returns true and you use KVM_CREATE_IRQCHIP, or if you emulate the 1648feature in userspace, then you can enable the feature for KVM_SET_CPUID2. 1649 1650 16514.47 KVM_PPC_GET_PVINFO 1652----------------------- 1653 1654:Capability: KVM_CAP_PPC_GET_PVINFO 1655:Architectures: ppc 1656:Type: vm ioctl 1657:Parameters: struct kvm_ppc_pvinfo (out) 1658:Returns: 0 on success, !0 on error 1659 1660:: 1661 1662 struct kvm_ppc_pvinfo { 1663 __u32 flags; 1664 __u32 hcall[4]; 1665 __u8 pad[108]; 1666 }; 1667 1668This ioctl fetches PV specific information that need to be passed to the guest 1669using the device tree or other means from vm context. 1670 1671The hcall array defines 4 instructions that make up a hypercall. 1672 1673If any additional field gets added to this structure later on, a bit for that 1674additional piece of information will be set in the flags bitmap. 1675 1676The flags bitmap is defined as:: 1677 1678 /* the host supports the ePAPR idle hcall 1679 #define KVM_PPC_PVINFO_FLAGS_EV_IDLE (1<<0) 1680 16814.52 KVM_SET_GSI_ROUTING 1682------------------------ 1683 1684:Capability: KVM_CAP_IRQ_ROUTING 1685:Architectures: x86 s390 arm arm64 1686:Type: vm ioctl 1687:Parameters: struct kvm_irq_routing (in) 1688:Returns: 0 on success, -1 on error 1689 1690Sets the GSI routing table entries, overwriting any previously set entries. 1691 1692On arm/arm64, GSI routing has the following limitation: 1693 1694- GSI routing does not apply to KVM_IRQ_LINE but only to KVM_IRQFD. 1695 1696:: 1697 1698 struct kvm_irq_routing { 1699 __u32 nr; 1700 __u32 flags; 1701 struct kvm_irq_routing_entry entries[0]; 1702 }; 1703 1704No flags are specified so far, the corresponding field must be set to zero. 1705 1706:: 1707 1708 struct kvm_irq_routing_entry { 1709 __u32 gsi; 1710 __u32 type; 1711 __u32 flags; 1712 __u32 pad; 1713 union { 1714 struct kvm_irq_routing_irqchip irqchip; 1715 struct kvm_irq_routing_msi msi; 1716 struct kvm_irq_routing_s390_adapter adapter; 1717 struct kvm_irq_routing_hv_sint hv_sint; 1718 __u32 pad[8]; 1719 } u; 1720 }; 1721 1722 /* gsi routing entry types */ 1723 #define KVM_IRQ_ROUTING_IRQCHIP 1 1724 #define KVM_IRQ_ROUTING_MSI 2 1725 #define KVM_IRQ_ROUTING_S390_ADAPTER 3 1726 #define KVM_IRQ_ROUTING_HV_SINT 4 1727 1728flags: 1729 1730- KVM_MSI_VALID_DEVID: used along with KVM_IRQ_ROUTING_MSI routing entry 1731 type, specifies that the devid field contains a valid value. The per-VM 1732 KVM_CAP_MSI_DEVID capability advertises the requirement to provide 1733 the device ID. If this capability is not available, userspace should 1734 never set the KVM_MSI_VALID_DEVID flag as the ioctl might fail. 1735- zero otherwise 1736 1737:: 1738 1739 struct kvm_irq_routing_irqchip { 1740 __u32 irqchip; 1741 __u32 pin; 1742 }; 1743 1744 struct kvm_irq_routing_msi { 1745 __u32 address_lo; 1746 __u32 address_hi; 1747 __u32 data; 1748 union { 1749 __u32 pad; 1750 __u32 devid; 1751 }; 1752 }; 1753 1754If KVM_MSI_VALID_DEVID is set, devid contains a unique device identifier 1755for the device that wrote the MSI message. For PCI, this is usually a 1756BFD identifier in the lower 16 bits. 1757 1758On x86, address_hi is ignored unless the KVM_X2APIC_API_USE_32BIT_IDS 1759feature of KVM_CAP_X2APIC_API capability is enabled. If it is enabled, 1760address_hi bits 31-8 provide bits 31-8 of the destination id. Bits 7-0 of 1761address_hi must be zero. 1762 1763:: 1764 1765 struct kvm_irq_routing_s390_adapter { 1766 __u64 ind_addr; 1767 __u64 summary_addr; 1768 __u64 ind_offset; 1769 __u32 summary_offset; 1770 __u32 adapter_id; 1771 }; 1772 1773 struct kvm_irq_routing_hv_sint { 1774 __u32 vcpu; 1775 __u32 sint; 1776 }; 1777 1778 17794.55 KVM_SET_TSC_KHZ 1780-------------------- 1781 1782:Capability: KVM_CAP_TSC_CONTROL 1783:Architectures: x86 1784:Type: vcpu ioctl 1785:Parameters: virtual tsc_khz 1786:Returns: 0 on success, -1 on error 1787 1788Specifies the tsc frequency for the virtual machine. The unit of the 1789frequency is KHz. 1790 1791 17924.56 KVM_GET_TSC_KHZ 1793-------------------- 1794 1795:Capability: KVM_CAP_GET_TSC_KHZ 1796:Architectures: x86 1797:Type: vcpu ioctl 1798:Parameters: none 1799:Returns: virtual tsc-khz on success, negative value on error 1800 1801Returns the tsc frequency of the guest. The unit of the return value is 1802KHz. If the host has unstable tsc this ioctl returns -EIO instead as an 1803error. 1804 1805 18064.57 KVM_GET_LAPIC 1807------------------ 1808 1809:Capability: KVM_CAP_IRQCHIP 1810:Architectures: x86 1811:Type: vcpu ioctl 1812:Parameters: struct kvm_lapic_state (out) 1813:Returns: 0 on success, -1 on error 1814 1815:: 1816 1817 #define KVM_APIC_REG_SIZE 0x400 1818 struct kvm_lapic_state { 1819 char regs[KVM_APIC_REG_SIZE]; 1820 }; 1821 1822Reads the Local APIC registers and copies them into the input argument. The 1823data format and layout are the same as documented in the architecture manual. 1824 1825If KVM_X2APIC_API_USE_32BIT_IDS feature of KVM_CAP_X2APIC_API is 1826enabled, then the format of APIC_ID register depends on the APIC mode 1827(reported by MSR_IA32_APICBASE) of its VCPU. x2APIC stores APIC ID in 1828the APIC_ID register (bytes 32-35). xAPIC only allows an 8-bit APIC ID 1829which is stored in bits 31-24 of the APIC register, or equivalently in 1830byte 35 of struct kvm_lapic_state's regs field. KVM_GET_LAPIC must then 1831be called after MSR_IA32_APICBASE has been set with KVM_SET_MSR. 1832 1833If KVM_X2APIC_API_USE_32BIT_IDS feature is disabled, struct kvm_lapic_state 1834always uses xAPIC format. 1835 1836 18374.58 KVM_SET_LAPIC 1838------------------ 1839 1840:Capability: KVM_CAP_IRQCHIP 1841:Architectures: x86 1842:Type: vcpu ioctl 1843:Parameters: struct kvm_lapic_state (in) 1844:Returns: 0 on success, -1 on error 1845 1846:: 1847 1848 #define KVM_APIC_REG_SIZE 0x400 1849 struct kvm_lapic_state { 1850 char regs[KVM_APIC_REG_SIZE]; 1851 }; 1852 1853Copies the input argument into the Local APIC registers. The data format 1854and layout are the same as documented in the architecture manual. 1855 1856The format of the APIC ID register (bytes 32-35 of struct kvm_lapic_state's 1857regs field) depends on the state of the KVM_CAP_X2APIC_API capability. 1858See the note in KVM_GET_LAPIC. 1859 1860 18614.59 KVM_IOEVENTFD 1862------------------ 1863 1864:Capability: KVM_CAP_IOEVENTFD 1865:Architectures: all 1866:Type: vm ioctl 1867:Parameters: struct kvm_ioeventfd (in) 1868:Returns: 0 on success, !0 on error 1869 1870This ioctl attaches or detaches an ioeventfd to a legal pio/mmio address 1871within the guest. A guest write in the registered address will signal the 1872provided event instead of triggering an exit. 1873 1874:: 1875 1876 struct kvm_ioeventfd { 1877 __u64 datamatch; 1878 __u64 addr; /* legal pio/mmio address */ 1879 __u32 len; /* 0, 1, 2, 4, or 8 bytes */ 1880 __s32 fd; 1881 __u32 flags; 1882 __u8 pad[36]; 1883 }; 1884 1885For the special case of virtio-ccw devices on s390, the ioevent is matched 1886to a subchannel/virtqueue tuple instead. 1887 1888The following flags are defined:: 1889 1890 #define KVM_IOEVENTFD_FLAG_DATAMATCH (1 << kvm_ioeventfd_flag_nr_datamatch) 1891 #define KVM_IOEVENTFD_FLAG_PIO (1 << kvm_ioeventfd_flag_nr_pio) 1892 #define KVM_IOEVENTFD_FLAG_DEASSIGN (1 << kvm_ioeventfd_flag_nr_deassign) 1893 #define KVM_IOEVENTFD_FLAG_VIRTIO_CCW_NOTIFY \ 1894 (1 << kvm_ioeventfd_flag_nr_virtio_ccw_notify) 1895 1896If datamatch flag is set, the event will be signaled only if the written value 1897to the registered address is equal to datamatch in struct kvm_ioeventfd. 1898 1899For virtio-ccw devices, addr contains the subchannel id and datamatch the 1900virtqueue index. 1901 1902With KVM_CAP_IOEVENTFD_ANY_LENGTH, a zero length ioeventfd is allowed, and 1903the kernel will ignore the length of guest write and may get a faster vmexit. 1904The speedup may only apply to specific architectures, but the ioeventfd will 1905work anyway. 1906 19074.60 KVM_DIRTY_TLB 1908------------------ 1909 1910:Capability: KVM_CAP_SW_TLB 1911:Architectures: ppc 1912:Type: vcpu ioctl 1913:Parameters: struct kvm_dirty_tlb (in) 1914:Returns: 0 on success, -1 on error 1915 1916:: 1917 1918 struct kvm_dirty_tlb { 1919 __u64 bitmap; 1920 __u32 num_dirty; 1921 }; 1922 1923This must be called whenever userspace has changed an entry in the shared 1924TLB, prior to calling KVM_RUN on the associated vcpu. 1925 1926The "bitmap" field is the userspace address of an array. This array 1927consists of a number of bits, equal to the total number of TLB entries as 1928determined by the last successful call to KVM_CONFIG_TLB, rounded up to the 1929nearest multiple of 64. 1930 1931Each bit corresponds to one TLB entry, ordered the same as in the shared TLB 1932array. 1933 1934The array is little-endian: the bit 0 is the least significant bit of the 1935first byte, bit 8 is the least significant bit of the second byte, etc. 1936This avoids any complications with differing word sizes. 1937 1938The "num_dirty" field is a performance hint for KVM to determine whether it 1939should skip processing the bitmap and just invalidate everything. It must 1940be set to the number of set bits in the bitmap. 1941 1942 19434.62 KVM_CREATE_SPAPR_TCE 1944------------------------- 1945 1946:Capability: KVM_CAP_SPAPR_TCE 1947:Architectures: powerpc 1948:Type: vm ioctl 1949:Parameters: struct kvm_create_spapr_tce (in) 1950:Returns: file descriptor for manipulating the created TCE table 1951 1952This creates a virtual TCE (translation control entry) table, which 1953is an IOMMU for PAPR-style virtual I/O. It is used to translate 1954logical addresses used in virtual I/O into guest physical addresses, 1955and provides a scatter/gather capability for PAPR virtual I/O. 1956 1957:: 1958 1959 /* for KVM_CAP_SPAPR_TCE */ 1960 struct kvm_create_spapr_tce { 1961 __u64 liobn; 1962 __u32 window_size; 1963 }; 1964 1965The liobn field gives the logical IO bus number for which to create a 1966TCE table. The window_size field specifies the size of the DMA window 1967which this TCE table will translate - the table will contain one 64 1968bit TCE entry for every 4kiB of the DMA window. 1969 1970When the guest issues an H_PUT_TCE hcall on a liobn for which a TCE 1971table has been created using this ioctl(), the kernel will handle it 1972in real mode, updating the TCE table. H_PUT_TCE calls for other 1973liobns will cause a vm exit and must be handled by userspace. 1974 1975The return value is a file descriptor which can be passed to mmap(2) 1976to map the created TCE table into userspace. This lets userspace read 1977the entries written by kernel-handled H_PUT_TCE calls, and also lets 1978userspace update the TCE table directly which is useful in some 1979circumstances. 1980 1981 19824.63 KVM_ALLOCATE_RMA 1983--------------------- 1984 1985:Capability: KVM_CAP_PPC_RMA 1986:Architectures: powerpc 1987:Type: vm ioctl 1988:Parameters: struct kvm_allocate_rma (out) 1989:Returns: file descriptor for mapping the allocated RMA 1990 1991This allocates a Real Mode Area (RMA) from the pool allocated at boot 1992time by the kernel. An RMA is a physically-contiguous, aligned region 1993of memory used on older POWER processors to provide the memory which 1994will be accessed by real-mode (MMU off) accesses in a KVM guest. 1995POWER processors support a set of sizes for the RMA that usually 1996includes 64MB, 128MB, 256MB and some larger powers of two. 1997 1998:: 1999 2000 /* for KVM_ALLOCATE_RMA */ 2001 struct kvm_allocate_rma { 2002 __u64 rma_size; 2003 }; 2004 2005The return value is a file descriptor which can be passed to mmap(2) 2006to map the allocated RMA into userspace. The mapped area can then be 2007passed to the KVM_SET_USER_MEMORY_REGION ioctl to establish it as the 2008RMA for a virtual machine. The size of the RMA in bytes (which is 2009fixed at host kernel boot time) is returned in the rma_size field of 2010the argument structure. 2011 2012The KVM_CAP_PPC_RMA capability is 1 or 2 if the KVM_ALLOCATE_RMA ioctl 2013is supported; 2 if the processor requires all virtual machines to have 2014an RMA, or 1 if the processor can use an RMA but doesn't require it, 2015because it supports the Virtual RMA (VRMA) facility. 2016 2017 20184.64 KVM_NMI 2019------------ 2020 2021:Capability: KVM_CAP_USER_NMI 2022:Architectures: x86 2023:Type: vcpu ioctl 2024:Parameters: none 2025:Returns: 0 on success, -1 on error 2026 2027Queues an NMI on the thread's vcpu. Note this is well defined only 2028when KVM_CREATE_IRQCHIP has not been called, since this is an interface 2029between the virtual cpu core and virtual local APIC. After KVM_CREATE_IRQCHIP 2030has been called, this interface is completely emulated within the kernel. 2031 2032To use this to emulate the LINT1 input with KVM_CREATE_IRQCHIP, use the 2033following algorithm: 2034 2035 - pause the vcpu 2036 - read the local APIC's state (KVM_GET_LAPIC) 2037 - check whether changing LINT1 will queue an NMI (see the LVT entry for LINT1) 2038 - if so, issue KVM_NMI 2039 - resume the vcpu 2040 2041Some guests configure the LINT1 NMI input to cause a panic, aiding in 2042debugging. 2043 2044 20454.65 KVM_S390_UCAS_MAP 2046---------------------- 2047 2048:Capability: KVM_CAP_S390_UCONTROL 2049:Architectures: s390 2050:Type: vcpu ioctl 2051:Parameters: struct kvm_s390_ucas_mapping (in) 2052:Returns: 0 in case of success 2053 2054The parameter is defined like this:: 2055 2056 struct kvm_s390_ucas_mapping { 2057 __u64 user_addr; 2058 __u64 vcpu_addr; 2059 __u64 length; 2060 }; 2061 2062This ioctl maps the memory at "user_addr" with the length "length" to 2063the vcpu's address space starting at "vcpu_addr". All parameters need to 2064be aligned by 1 megabyte. 2065 2066 20674.66 KVM_S390_UCAS_UNMAP 2068------------------------ 2069 2070:Capability: KVM_CAP_S390_UCONTROL 2071:Architectures: s390 2072:Type: vcpu ioctl 2073:Parameters: struct kvm_s390_ucas_mapping (in) 2074:Returns: 0 in case of success 2075 2076The parameter is defined like this:: 2077 2078 struct kvm_s390_ucas_mapping { 2079 __u64 user_addr; 2080 __u64 vcpu_addr; 2081 __u64 length; 2082 }; 2083 2084This ioctl unmaps the memory in the vcpu's address space starting at 2085"vcpu_addr" with the length "length". The field "user_addr" is ignored. 2086All parameters need to be aligned by 1 megabyte. 2087 2088 20894.67 KVM_S390_VCPU_FAULT 2090------------------------ 2091 2092:Capability: KVM_CAP_S390_UCONTROL 2093:Architectures: s390 2094:Type: vcpu ioctl 2095:Parameters: vcpu absolute address (in) 2096:Returns: 0 in case of success 2097 2098This call creates a page table entry on the virtual cpu's address space 2099(for user controlled virtual machines) or the virtual machine's address 2100space (for regular virtual machines). This only works for minor faults, 2101thus it's recommended to access subject memory page via the user page 2102table upfront. This is useful to handle validity intercepts for user 2103controlled virtual machines to fault in the virtual cpu's lowcore pages 2104prior to calling the KVM_RUN ioctl. 2105 2106 21074.68 KVM_SET_ONE_REG 2108-------------------- 2109 2110:Capability: KVM_CAP_ONE_REG 2111:Architectures: all 2112:Type: vcpu ioctl 2113:Parameters: struct kvm_one_reg (in) 2114:Returns: 0 on success, negative value on failure 2115 2116Errors: 2117 2118 ====== ============================================================ 2119 ENOENT no such register 2120 EINVAL invalid register ID, or no such register 2121 EPERM (arm64) register access not allowed before vcpu finalization 2122 ====== ============================================================ 2123 2124(These error codes are indicative only: do not rely on a specific error 2125code being returned in a specific situation.) 2126 2127:: 2128 2129 struct kvm_one_reg { 2130 __u64 id; 2131 __u64 addr; 2132 }; 2133 2134Using this ioctl, a single vcpu register can be set to a specific value 2135defined by user space with the passed in struct kvm_one_reg, where id 2136refers to the register identifier as described below and addr is a pointer 2137to a variable with the respective size. There can be architecture agnostic 2138and architecture specific registers. Each have their own range of operation 2139and their own constants and width. To keep track of the implemented 2140registers, find a list below: 2141 2142 ======= =============================== ============ 2143 Arch Register Width (bits) 2144 ======= =============================== ============ 2145 PPC KVM_REG_PPC_HIOR 64 2146 PPC KVM_REG_PPC_IAC1 64 2147 PPC KVM_REG_PPC_IAC2 64 2148 PPC KVM_REG_PPC_IAC3 64 2149 PPC KVM_REG_PPC_IAC4 64 2150 PPC KVM_REG_PPC_DAC1 64 2151 PPC KVM_REG_PPC_DAC2 64 2152 PPC KVM_REG_PPC_DABR 64 2153 PPC KVM_REG_PPC_DSCR 64 2154 PPC KVM_REG_PPC_PURR 64 2155 PPC KVM_REG_PPC_SPURR 64 2156 PPC KVM_REG_PPC_DAR 64 2157 PPC KVM_REG_PPC_DSISR 32 2158 PPC KVM_REG_PPC_AMR 64 2159 PPC KVM_REG_PPC_UAMOR 64 2160 PPC KVM_REG_PPC_MMCR0 64 2161 PPC KVM_REG_PPC_MMCR1 64 2162 PPC KVM_REG_PPC_MMCRA 64 2163 PPC KVM_REG_PPC_MMCR2 64 2164 PPC KVM_REG_PPC_MMCRS 64 2165 PPC KVM_REG_PPC_SIAR 64 2166 PPC KVM_REG_PPC_SDAR 64 2167 PPC KVM_REG_PPC_SIER 64 2168 PPC KVM_REG_PPC_PMC1 32 2169 PPC KVM_REG_PPC_PMC2 32 2170 PPC KVM_REG_PPC_PMC3 32 2171 PPC KVM_REG_PPC_PMC4 32 2172 PPC KVM_REG_PPC_PMC5 32 2173 PPC KVM_REG_PPC_PMC6 32 2174 PPC KVM_REG_PPC_PMC7 32 2175 PPC KVM_REG_PPC_PMC8 32 2176 PPC KVM_REG_PPC_FPR0 64 2177 ... 2178 PPC KVM_REG_PPC_FPR31 64 2179 PPC KVM_REG_PPC_VR0 128 2180 ... 2181 PPC KVM_REG_PPC_VR31 128 2182 PPC KVM_REG_PPC_VSR0 128 2183 ... 2184 PPC KVM_REG_PPC_VSR31 128 2185 PPC KVM_REG_PPC_FPSCR 64 2186 PPC KVM_REG_PPC_VSCR 32 2187 PPC KVM_REG_PPC_VPA_ADDR 64 2188 PPC KVM_REG_PPC_VPA_SLB 128 2189 PPC KVM_REG_PPC_VPA_DTL 128 2190 PPC KVM_REG_PPC_EPCR 32 2191 PPC KVM_REG_PPC_EPR 32 2192 PPC KVM_REG_PPC_TCR 32 2193 PPC KVM_REG_PPC_TSR 32 2194 PPC KVM_REG_PPC_OR_TSR 32 2195 PPC KVM_REG_PPC_CLEAR_TSR 32 2196 PPC KVM_REG_PPC_MAS0 32 2197 PPC KVM_REG_PPC_MAS1 32 2198 PPC KVM_REG_PPC_MAS2 64 2199 PPC KVM_REG_PPC_MAS7_3 64 2200 PPC KVM_REG_PPC_MAS4 32 2201 PPC KVM_REG_PPC_MAS6 32 2202 PPC KVM_REG_PPC_MMUCFG 32 2203 PPC KVM_REG_PPC_TLB0CFG 32 2204 PPC KVM_REG_PPC_TLB1CFG 32 2205 PPC KVM_REG_PPC_TLB2CFG 32 2206 PPC KVM_REG_PPC_TLB3CFG 32 2207 PPC KVM_REG_PPC_TLB0PS 32 2208 PPC KVM_REG_PPC_TLB1PS 32 2209 PPC KVM_REG_PPC_TLB2PS 32 2210 PPC KVM_REG_PPC_TLB3PS 32 2211 PPC KVM_REG_PPC_EPTCFG 32 2212 PPC KVM_REG_PPC_ICP_STATE 64 2213 PPC KVM_REG_PPC_VP_STATE 128 2214 PPC KVM_REG_PPC_TB_OFFSET 64 2215 PPC KVM_REG_PPC_SPMC1 32 2216 PPC KVM_REG_PPC_SPMC2 32 2217 PPC KVM_REG_PPC_IAMR 64 2218 PPC KVM_REG_PPC_TFHAR 64 2219 PPC KVM_REG_PPC_TFIAR 64 2220 PPC KVM_REG_PPC_TEXASR 64 2221 PPC KVM_REG_PPC_FSCR 64 2222 PPC KVM_REG_PPC_PSPB 32 2223 PPC KVM_REG_PPC_EBBHR 64 2224 PPC KVM_REG_PPC_EBBRR 64 2225 PPC KVM_REG_PPC_BESCR 64 2226 PPC KVM_REG_PPC_TAR 64 2227 PPC KVM_REG_PPC_DPDES 64 2228 PPC KVM_REG_PPC_DAWR 64 2229 PPC KVM_REG_PPC_DAWRX 64 2230 PPC KVM_REG_PPC_CIABR 64 2231 PPC KVM_REG_PPC_IC 64 2232 PPC KVM_REG_PPC_VTB 64 2233 PPC KVM_REG_PPC_CSIGR 64 2234 PPC KVM_REG_PPC_TACR 64 2235 PPC KVM_REG_PPC_TCSCR 64 2236 PPC KVM_REG_PPC_PID 64 2237 PPC KVM_REG_PPC_ACOP 64 2238 PPC KVM_REG_PPC_VRSAVE 32 2239 PPC KVM_REG_PPC_LPCR 32 2240 PPC KVM_REG_PPC_LPCR_64 64 2241 PPC KVM_REG_PPC_PPR 64 2242 PPC KVM_REG_PPC_ARCH_COMPAT 32 2243 PPC KVM_REG_PPC_DABRX 32 2244 PPC KVM_REG_PPC_WORT 64 2245 PPC KVM_REG_PPC_SPRG9 64 2246 PPC KVM_REG_PPC_DBSR 32 2247 PPC KVM_REG_PPC_TIDR 64 2248 PPC KVM_REG_PPC_PSSCR 64 2249 PPC KVM_REG_PPC_DEC_EXPIRY 64 2250 PPC KVM_REG_PPC_PTCR 64 2251 PPC KVM_REG_PPC_TM_GPR0 64 2252 ... 2253 PPC KVM_REG_PPC_TM_GPR31 64 2254 PPC KVM_REG_PPC_TM_VSR0 128 2255 ... 2256 PPC KVM_REG_PPC_TM_VSR63 128 2257 PPC KVM_REG_PPC_TM_CR 64 2258 PPC KVM_REG_PPC_TM_LR 64 2259 PPC KVM_REG_PPC_TM_CTR 64 2260 PPC KVM_REG_PPC_TM_FPSCR 64 2261 PPC KVM_REG_PPC_TM_AMR 64 2262 PPC KVM_REG_PPC_TM_PPR 64 2263 PPC KVM_REG_PPC_TM_VRSAVE 64 2264 PPC KVM_REG_PPC_TM_VSCR 32 2265 PPC KVM_REG_PPC_TM_DSCR 64 2266 PPC KVM_REG_PPC_TM_TAR 64 2267 PPC KVM_REG_PPC_TM_XER 64 2268 2269 MIPS KVM_REG_MIPS_R0 64 2270 ... 2271 MIPS KVM_REG_MIPS_R31 64 2272 MIPS KVM_REG_MIPS_HI 64 2273 MIPS KVM_REG_MIPS_LO 64 2274 MIPS KVM_REG_MIPS_PC 64 2275 MIPS KVM_REG_MIPS_CP0_INDEX 32 2276 MIPS KVM_REG_MIPS_CP0_ENTRYLO0 64 2277 MIPS KVM_REG_MIPS_CP0_ENTRYLO1 64 2278 MIPS KVM_REG_MIPS_CP0_CONTEXT 64 2279 MIPS KVM_REG_MIPS_CP0_CONTEXTCONFIG 32 2280 MIPS KVM_REG_MIPS_CP0_USERLOCAL 64 2281 MIPS KVM_REG_MIPS_CP0_XCONTEXTCONFIG 64 2282 MIPS KVM_REG_MIPS_CP0_PAGEMASK 32 2283 MIPS KVM_REG_MIPS_CP0_PAGEGRAIN 32 2284 MIPS KVM_REG_MIPS_CP0_SEGCTL0 64 2285 MIPS KVM_REG_MIPS_CP0_SEGCTL1 64 2286 MIPS KVM_REG_MIPS_CP0_SEGCTL2 64 2287 MIPS KVM_REG_MIPS_CP0_PWBASE 64 2288 MIPS KVM_REG_MIPS_CP0_PWFIELD 64 2289 MIPS KVM_REG_MIPS_CP0_PWSIZE 64 2290 MIPS KVM_REG_MIPS_CP0_WIRED 32 2291 MIPS KVM_REG_MIPS_CP0_PWCTL 32 2292 MIPS KVM_REG_MIPS_CP0_HWRENA 32 2293 MIPS KVM_REG_MIPS_CP0_BADVADDR 64 2294 MIPS KVM_REG_MIPS_CP0_BADINSTR 32 2295 MIPS KVM_REG_MIPS_CP0_BADINSTRP 32 2296 MIPS KVM_REG_MIPS_CP0_COUNT 32 2297 MIPS KVM_REG_MIPS_CP0_ENTRYHI 64 2298 MIPS KVM_REG_MIPS_CP0_COMPARE 32 2299 MIPS KVM_REG_MIPS_CP0_STATUS 32 2300 MIPS KVM_REG_MIPS_CP0_INTCTL 32 2301 MIPS KVM_REG_MIPS_CP0_CAUSE 32 2302 MIPS KVM_REG_MIPS_CP0_EPC 64 2303 MIPS KVM_REG_MIPS_CP0_PRID 32 2304 MIPS KVM_REG_MIPS_CP0_EBASE 64 2305 MIPS KVM_REG_MIPS_CP0_CONFIG 32 2306 MIPS KVM_REG_MIPS_CP0_CONFIG1 32 2307 MIPS KVM_REG_MIPS_CP0_CONFIG2 32 2308 MIPS KVM_REG_MIPS_CP0_CONFIG3 32 2309 MIPS KVM_REG_MIPS_CP0_CONFIG4 32 2310 MIPS KVM_REG_MIPS_CP0_CONFIG5 32 2311 MIPS KVM_REG_MIPS_CP0_CONFIG7 32 2312 MIPS KVM_REG_MIPS_CP0_XCONTEXT 64 2313 MIPS KVM_REG_MIPS_CP0_ERROREPC 64 2314 MIPS KVM_REG_MIPS_CP0_KSCRATCH1 64 2315 MIPS KVM_REG_MIPS_CP0_KSCRATCH2 64 2316 MIPS KVM_REG_MIPS_CP0_KSCRATCH3 64 2317 MIPS KVM_REG_MIPS_CP0_KSCRATCH4 64 2318 MIPS KVM_REG_MIPS_CP0_KSCRATCH5 64 2319 MIPS KVM_REG_MIPS_CP0_KSCRATCH6 64 2320 MIPS KVM_REG_MIPS_CP0_MAAR(0..63) 64 2321 MIPS KVM_REG_MIPS_COUNT_CTL 64 2322 MIPS KVM_REG_MIPS_COUNT_RESUME 64 2323 MIPS KVM_REG_MIPS_COUNT_HZ 64 2324 MIPS KVM_REG_MIPS_FPR_32(0..31) 32 2325 MIPS KVM_REG_MIPS_FPR_64(0..31) 64 2326 MIPS KVM_REG_MIPS_VEC_128(0..31) 128 2327 MIPS KVM_REG_MIPS_FCR_IR 32 2328 MIPS KVM_REG_MIPS_FCR_CSR 32 2329 MIPS KVM_REG_MIPS_MSA_IR 32 2330 MIPS KVM_REG_MIPS_MSA_CSR 32 2331 ======= =============================== ============ 2332 2333ARM registers are mapped using the lower 32 bits. The upper 16 of that 2334is the register group type, or coprocessor number: 2335 2336ARM core registers have the following id bit patterns:: 2337 2338 0x4020 0000 0010 <index into the kvm_regs struct:16> 2339 2340ARM 32-bit CP15 registers have the following id bit patterns:: 2341 2342 0x4020 0000 000F <zero:1> <crn:4> <crm:4> <opc1:4> <opc2:3> 2343 2344ARM 64-bit CP15 registers have the following id bit patterns:: 2345 2346 0x4030 0000 000F <zero:1> <zero:4> <crm:4> <opc1:4> <zero:3> 2347 2348ARM CCSIDR registers are demultiplexed by CSSELR value:: 2349 2350 0x4020 0000 0011 00 <csselr:8> 2351 2352ARM 32-bit VFP control registers have the following id bit patterns:: 2353 2354 0x4020 0000 0012 1 <regno:12> 2355 2356ARM 64-bit FP registers have the following id bit patterns:: 2357 2358 0x4030 0000 0012 0 <regno:12> 2359 2360ARM firmware pseudo-registers have the following bit pattern:: 2361 2362 0x4030 0000 0014 <regno:16> 2363 2364 2365arm64 registers are mapped using the lower 32 bits. The upper 16 of 2366that is the register group type, or coprocessor number: 2367 2368arm64 core/FP-SIMD registers have the following id bit patterns. Note 2369that the size of the access is variable, as the kvm_regs structure 2370contains elements ranging from 32 to 128 bits. The index is a 32bit 2371value in the kvm_regs structure seen as a 32bit array:: 2372 2373 0x60x0 0000 0010 <index into the kvm_regs struct:16> 2374 2375Specifically: 2376 2377======================= ========= ===== ======================================= 2378 Encoding Register Bits kvm_regs member 2379======================= ========= ===== ======================================= 2380 0x6030 0000 0010 0000 X0 64 regs.regs[0] 2381 0x6030 0000 0010 0002 X1 64 regs.regs[1] 2382 ... 2383 0x6030 0000 0010 003c X30 64 regs.regs[30] 2384 0x6030 0000 0010 003e SP 64 regs.sp 2385 0x6030 0000 0010 0040 PC 64 regs.pc 2386 0x6030 0000 0010 0042 PSTATE 64 regs.pstate 2387 0x6030 0000 0010 0044 SP_EL1 64 sp_el1 2388 0x6030 0000 0010 0046 ELR_EL1 64 elr_el1 2389 0x6030 0000 0010 0048 SPSR_EL1 64 spsr[KVM_SPSR_EL1] (alias SPSR_SVC) 2390 0x6030 0000 0010 004a SPSR_ABT 64 spsr[KVM_SPSR_ABT] 2391 0x6030 0000 0010 004c SPSR_UND 64 spsr[KVM_SPSR_UND] 2392 0x6030 0000 0010 004e SPSR_IRQ 64 spsr[KVM_SPSR_IRQ] 2393 0x6060 0000 0010 0050 SPSR_FIQ 64 spsr[KVM_SPSR_FIQ] 2394 0x6040 0000 0010 0054 V0 128 fp_regs.vregs[0] [1]_ 2395 0x6040 0000 0010 0058 V1 128 fp_regs.vregs[1] [1]_ 2396 ... 2397 0x6040 0000 0010 00d0 V31 128 fp_regs.vregs[31] [1]_ 2398 0x6020 0000 0010 00d4 FPSR 32 fp_regs.fpsr 2399 0x6020 0000 0010 00d5 FPCR 32 fp_regs.fpcr 2400======================= ========= ===== ======================================= 2401 2402.. [1] These encodings are not accepted for SVE-enabled vcpus. See 2403 KVM_ARM_VCPU_INIT. 2404 2405 The equivalent register content can be accessed via bits [127:0] of 2406 the corresponding SVE Zn registers instead for vcpus that have SVE 2407 enabled (see below). 2408 2409arm64 CCSIDR registers are demultiplexed by CSSELR value:: 2410 2411 0x6020 0000 0011 00 <csselr:8> 2412 2413arm64 system registers have the following id bit patterns:: 2414 2415 0x6030 0000 0013 <op0:2> <op1:3> <crn:4> <crm:4> <op2:3> 2416 2417.. warning:: 2418 2419 Two system register IDs do not follow the specified pattern. These 2420 are KVM_REG_ARM_TIMER_CVAL and KVM_REG_ARM_TIMER_CNT, which map to 2421 system registers CNTV_CVAL_EL0 and CNTVCT_EL0 respectively. These 2422 two had their values accidentally swapped, which means TIMER_CVAL is 2423 derived from the register encoding for CNTVCT_EL0 and TIMER_CNT is 2424 derived from the register encoding for CNTV_CVAL_EL0. As this is 2425 API, it must remain this way. 2426 2427arm64 firmware pseudo-registers have the following bit pattern:: 2428 2429 0x6030 0000 0014 <regno:16> 2430 2431arm64 SVE registers have the following bit patterns:: 2432 2433 0x6080 0000 0015 00 <n:5> <slice:5> Zn bits[2048*slice + 2047 : 2048*slice] 2434 0x6050 0000 0015 04 <n:4> <slice:5> Pn bits[256*slice + 255 : 256*slice] 2435 0x6050 0000 0015 060 <slice:5> FFR bits[256*slice + 255 : 256*slice] 2436 0x6060 0000 0015 ffff KVM_REG_ARM64_SVE_VLS pseudo-register 2437 2438Access to register IDs where 2048 * slice >= 128 * max_vq will fail with 2439ENOENT. max_vq is the vcpu's maximum supported vector length in 128-bit 2440quadwords: see [2]_ below. 2441 2442These registers are only accessible on vcpus for which SVE is enabled. 2443See KVM_ARM_VCPU_INIT for details. 2444 2445In addition, except for KVM_REG_ARM64_SVE_VLS, these registers are not 2446accessible until the vcpu's SVE configuration has been finalized 2447using KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE). See KVM_ARM_VCPU_INIT 2448and KVM_ARM_VCPU_FINALIZE for more information about this procedure. 2449 2450KVM_REG_ARM64_SVE_VLS is a pseudo-register that allows the set of vector 2451lengths supported by the vcpu to be discovered and configured by 2452userspace. When transferred to or from user memory via KVM_GET_ONE_REG 2453or KVM_SET_ONE_REG, the value of this register is of type 2454__u64[KVM_ARM64_SVE_VLS_WORDS], and encodes the set of vector lengths as 2455follows:: 2456 2457 __u64 vector_lengths[KVM_ARM64_SVE_VLS_WORDS]; 2458 2459 if (vq >= SVE_VQ_MIN && vq <= SVE_VQ_MAX && 2460 ((vector_lengths[(vq - KVM_ARM64_SVE_VQ_MIN) / 64] >> 2461 ((vq - KVM_ARM64_SVE_VQ_MIN) % 64)) & 1)) 2462 /* Vector length vq * 16 bytes supported */ 2463 else 2464 /* Vector length vq * 16 bytes not supported */ 2465 2466.. [2] The maximum value vq for which the above condition is true is 2467 max_vq. This is the maximum vector length available to the guest on 2468 this vcpu, and determines which register slices are visible through 2469 this ioctl interface. 2470 2471(See Documentation/arm64/sve.rst for an explanation of the "vq" 2472nomenclature.) 2473 2474KVM_REG_ARM64_SVE_VLS is only accessible after KVM_ARM_VCPU_INIT. 2475KVM_ARM_VCPU_INIT initialises it to the best set of vector lengths that 2476the host supports. 2477 2478Userspace may subsequently modify it if desired until the vcpu's SVE 2479configuration is finalized using KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE). 2480 2481Apart from simply removing all vector lengths from the host set that 2482exceed some value, support for arbitrarily chosen sets of vector lengths 2483is hardware-dependent and may not be available. Attempting to configure 2484an invalid set of vector lengths via KVM_SET_ONE_REG will fail with 2485EINVAL. 2486 2487After the vcpu's SVE configuration is finalized, further attempts to 2488write this register will fail with EPERM. 2489 2490 2491MIPS registers are mapped using the lower 32 bits. The upper 16 of that is 2492the register group type: 2493 2494MIPS core registers (see above) have the following id bit patterns:: 2495 2496 0x7030 0000 0000 <reg:16> 2497 2498MIPS CP0 registers (see KVM_REG_MIPS_CP0_* above) have the following id bit 2499patterns depending on whether they're 32-bit or 64-bit registers:: 2500 2501 0x7020 0000 0001 00 <reg:5> <sel:3> (32-bit) 2502 0x7030 0000 0001 00 <reg:5> <sel:3> (64-bit) 2503 2504Note: KVM_REG_MIPS_CP0_ENTRYLO0 and KVM_REG_MIPS_CP0_ENTRYLO1 are the MIPS64 2505versions of the EntryLo registers regardless of the word size of the host 2506hardware, host kernel, guest, and whether XPA is present in the guest, i.e. 2507with the RI and XI bits (if they exist) in bits 63 and 62 respectively, and 2508the PFNX field starting at bit 30. 2509 2510MIPS MAARs (see KVM_REG_MIPS_CP0_MAAR(*) above) have the following id bit 2511patterns:: 2512 2513 0x7030 0000 0001 01 <reg:8> 2514 2515MIPS KVM control registers (see above) have the following id bit patterns:: 2516 2517 0x7030 0000 0002 <reg:16> 2518 2519MIPS FPU registers (see KVM_REG_MIPS_FPR_{32,64}() above) have the following 2520id bit patterns depending on the size of the register being accessed. They are 2521always accessed according to the current guest FPU mode (Status.FR and 2522Config5.FRE), i.e. as the guest would see them, and they become unpredictable 2523if the guest FPU mode is changed. MIPS SIMD Architecture (MSA) vector 2524registers (see KVM_REG_MIPS_VEC_128() above) have similar patterns as they 2525overlap the FPU registers:: 2526 2527 0x7020 0000 0003 00 <0:3> <reg:5> (32-bit FPU registers) 2528 0x7030 0000 0003 00 <0:3> <reg:5> (64-bit FPU registers) 2529 0x7040 0000 0003 00 <0:3> <reg:5> (128-bit MSA vector registers) 2530 2531MIPS FPU control registers (see KVM_REG_MIPS_FCR_{IR,CSR} above) have the 2532following id bit patterns:: 2533 2534 0x7020 0000 0003 01 <0:3> <reg:5> 2535 2536MIPS MSA control registers (see KVM_REG_MIPS_MSA_{IR,CSR} above) have the 2537following id bit patterns:: 2538 2539 0x7020 0000 0003 02 <0:3> <reg:5> 2540 2541 25424.69 KVM_GET_ONE_REG 2543-------------------- 2544 2545:Capability: KVM_CAP_ONE_REG 2546:Architectures: all 2547:Type: vcpu ioctl 2548:Parameters: struct kvm_one_reg (in and out) 2549:Returns: 0 on success, negative value on failure 2550 2551Errors include: 2552 2553 ======== ============================================================ 2554 ENOENT no such register 2555 EINVAL invalid register ID, or no such register 2556 EPERM (arm64) register access not allowed before vcpu finalization 2557 ======== ============================================================ 2558 2559(These error codes are indicative only: do not rely on a specific error 2560code being returned in a specific situation.) 2561 2562This ioctl allows to receive the value of a single register implemented 2563in a vcpu. The register to read is indicated by the "id" field of the 2564kvm_one_reg struct passed in. On success, the register value can be found 2565at the memory location pointed to by "addr". 2566 2567The list of registers accessible using this interface is identical to the 2568list in 4.68. 2569 2570 25714.70 KVM_KVMCLOCK_CTRL 2572---------------------- 2573 2574:Capability: KVM_CAP_KVMCLOCK_CTRL 2575:Architectures: Any that implement pvclocks (currently x86 only) 2576:Type: vcpu ioctl 2577:Parameters: None 2578:Returns: 0 on success, -1 on error 2579 2580This signals to the host kernel that the specified guest is being paused by 2581userspace. The host will set a flag in the pvclock structure that is checked 2582from the soft lockup watchdog. The flag is part of the pvclock structure that 2583is shared between guest and host, specifically the second bit of the flags 2584field of the pvclock_vcpu_time_info structure. It will be set exclusively by 2585the host and read/cleared exclusively by the guest. The guest operation of 2586checking and clearing the flag must an atomic operation so 2587load-link/store-conditional, or equivalent must be used. There are two cases 2588where the guest will clear the flag: when the soft lockup watchdog timer resets 2589itself or when a soft lockup is detected. This ioctl can be called any time 2590after pausing the vcpu, but before it is resumed. 2591 2592 25934.71 KVM_SIGNAL_MSI 2594------------------- 2595 2596:Capability: KVM_CAP_SIGNAL_MSI 2597:Architectures: x86 arm arm64 2598:Type: vm ioctl 2599:Parameters: struct kvm_msi (in) 2600:Returns: >0 on delivery, 0 if guest blocked the MSI, and -1 on error 2601 2602Directly inject a MSI message. Only valid with in-kernel irqchip that handles 2603MSI messages. 2604 2605:: 2606 2607 struct kvm_msi { 2608 __u32 address_lo; 2609 __u32 address_hi; 2610 __u32 data; 2611 __u32 flags; 2612 __u32 devid; 2613 __u8 pad[12]; 2614 }; 2615 2616flags: 2617 KVM_MSI_VALID_DEVID: devid contains a valid value. The per-VM 2618 KVM_CAP_MSI_DEVID capability advertises the requirement to provide 2619 the device ID. If this capability is not available, userspace 2620 should never set the KVM_MSI_VALID_DEVID flag as the ioctl might fail. 2621 2622If KVM_MSI_VALID_DEVID is set, devid contains a unique device identifier 2623for the device that wrote the MSI message. For PCI, this is usually a 2624BFD identifier in the lower 16 bits. 2625 2626On x86, address_hi is ignored unless the KVM_X2APIC_API_USE_32BIT_IDS 2627feature of KVM_CAP_X2APIC_API capability is enabled. If it is enabled, 2628address_hi bits 31-8 provide bits 31-8 of the destination id. Bits 7-0 of 2629address_hi must be zero. 2630 2631 26324.71 KVM_CREATE_PIT2 2633-------------------- 2634 2635:Capability: KVM_CAP_PIT2 2636:Architectures: x86 2637:Type: vm ioctl 2638:Parameters: struct kvm_pit_config (in) 2639:Returns: 0 on success, -1 on error 2640 2641Creates an in-kernel device model for the i8254 PIT. This call is only valid 2642after enabling in-kernel irqchip support via KVM_CREATE_IRQCHIP. The following 2643parameters have to be passed:: 2644 2645 struct kvm_pit_config { 2646 __u32 flags; 2647 __u32 pad[15]; 2648 }; 2649 2650Valid flags are:: 2651 2652 #define KVM_PIT_SPEAKER_DUMMY 1 /* emulate speaker port stub */ 2653 2654PIT timer interrupts may use a per-VM kernel thread for injection. If it 2655exists, this thread will have a name of the following pattern:: 2656 2657 kvm-pit/<owner-process-pid> 2658 2659When running a guest with elevated priorities, the scheduling parameters of 2660this thread may have to be adjusted accordingly. 2661 2662This IOCTL replaces the obsolete KVM_CREATE_PIT. 2663 2664 26654.72 KVM_GET_PIT2 2666----------------- 2667 2668:Capability: KVM_CAP_PIT_STATE2 2669:Architectures: x86 2670:Type: vm ioctl 2671:Parameters: struct kvm_pit_state2 (out) 2672:Returns: 0 on success, -1 on error 2673 2674Retrieves the state of the in-kernel PIT model. Only valid after 2675KVM_CREATE_PIT2. The state is returned in the following structure:: 2676 2677 struct kvm_pit_state2 { 2678 struct kvm_pit_channel_state channels[3]; 2679 __u32 flags; 2680 __u32 reserved[9]; 2681 }; 2682 2683Valid flags are:: 2684 2685 /* disable PIT in HPET legacy mode */ 2686 #define KVM_PIT_FLAGS_HPET_LEGACY 0x00000001 2687 2688This IOCTL replaces the obsolete KVM_GET_PIT. 2689 2690 26914.73 KVM_SET_PIT2 2692----------------- 2693 2694:Capability: KVM_CAP_PIT_STATE2 2695:Architectures: x86 2696:Type: vm ioctl 2697:Parameters: struct kvm_pit_state2 (in) 2698:Returns: 0 on success, -1 on error 2699 2700Sets the state of the in-kernel PIT model. Only valid after KVM_CREATE_PIT2. 2701See KVM_GET_PIT2 for details on struct kvm_pit_state2. 2702 2703This IOCTL replaces the obsolete KVM_SET_PIT. 2704 2705 27064.74 KVM_PPC_GET_SMMU_INFO 2707-------------------------- 2708 2709:Capability: KVM_CAP_PPC_GET_SMMU_INFO 2710:Architectures: powerpc 2711:Type: vm ioctl 2712:Parameters: None 2713:Returns: 0 on success, -1 on error 2714 2715This populates and returns a structure describing the features of 2716the "Server" class MMU emulation supported by KVM. 2717This can in turn be used by userspace to generate the appropriate 2718device-tree properties for the guest operating system. 2719 2720The structure contains some global information, followed by an 2721array of supported segment page sizes:: 2722 2723 struct kvm_ppc_smmu_info { 2724 __u64 flags; 2725 __u32 slb_size; 2726 __u32 pad; 2727 struct kvm_ppc_one_seg_page_size sps[KVM_PPC_PAGE_SIZES_MAX_SZ]; 2728 }; 2729 2730The supported flags are: 2731 2732 - KVM_PPC_PAGE_SIZES_REAL: 2733 When that flag is set, guest page sizes must "fit" the backing 2734 store page sizes. When not set, any page size in the list can 2735 be used regardless of how they are backed by userspace. 2736 2737 - KVM_PPC_1T_SEGMENTS 2738 The emulated MMU supports 1T segments in addition to the 2739 standard 256M ones. 2740 2741 - KVM_PPC_NO_HASH 2742 This flag indicates that HPT guests are not supported by KVM, 2743 thus all guests must use radix MMU mode. 2744 2745The "slb_size" field indicates how many SLB entries are supported 2746 2747The "sps" array contains 8 entries indicating the supported base 2748page sizes for a segment in increasing order. Each entry is defined 2749as follow:: 2750 2751 struct kvm_ppc_one_seg_page_size { 2752 __u32 page_shift; /* Base page shift of segment (or 0) */ 2753 __u32 slb_enc; /* SLB encoding for BookS */ 2754 struct kvm_ppc_one_page_size enc[KVM_PPC_PAGE_SIZES_MAX_SZ]; 2755 }; 2756 2757An entry with a "page_shift" of 0 is unused. Because the array is 2758organized in increasing order, a lookup can stop when encoutering 2759such an entry. 2760 2761The "slb_enc" field provides the encoding to use in the SLB for the 2762page size. The bits are in positions such as the value can directly 2763be OR'ed into the "vsid" argument of the slbmte instruction. 2764 2765The "enc" array is a list which for each of those segment base page 2766size provides the list of supported actual page sizes (which can be 2767only larger or equal to the base page size), along with the 2768corresponding encoding in the hash PTE. Similarly, the array is 27698 entries sorted by increasing sizes and an entry with a "0" shift 2770is an empty entry and a terminator:: 2771 2772 struct kvm_ppc_one_page_size { 2773 __u32 page_shift; /* Page shift (or 0) */ 2774 __u32 pte_enc; /* Encoding in the HPTE (>>12) */ 2775 }; 2776 2777The "pte_enc" field provides a value that can OR'ed into the hash 2778PTE's RPN field (ie, it needs to be shifted left by 12 to OR it 2779into the hash PTE second double word). 2780 27814.75 KVM_IRQFD 2782-------------- 2783 2784:Capability: KVM_CAP_IRQFD 2785:Architectures: x86 s390 arm arm64 2786:Type: vm ioctl 2787:Parameters: struct kvm_irqfd (in) 2788:Returns: 0 on success, -1 on error 2789 2790Allows setting an eventfd to directly trigger a guest interrupt. 2791kvm_irqfd.fd specifies the file descriptor to use as the eventfd and 2792kvm_irqfd.gsi specifies the irqchip pin toggled by this event. When 2793an event is triggered on the eventfd, an interrupt is injected into 2794the guest using the specified gsi pin. The irqfd is removed using 2795the KVM_IRQFD_FLAG_DEASSIGN flag, specifying both kvm_irqfd.fd 2796and kvm_irqfd.gsi. 2797 2798With KVM_CAP_IRQFD_RESAMPLE, KVM_IRQFD supports a de-assert and notify 2799mechanism allowing emulation of level-triggered, irqfd-based 2800interrupts. When KVM_IRQFD_FLAG_RESAMPLE is set the user must pass an 2801additional eventfd in the kvm_irqfd.resamplefd field. When operating 2802in resample mode, posting of an interrupt through kvm_irq.fd asserts 2803the specified gsi in the irqchip. When the irqchip is resampled, such 2804as from an EOI, the gsi is de-asserted and the user is notified via 2805kvm_irqfd.resamplefd. It is the user's responsibility to re-queue 2806the interrupt if the device making use of it still requires service. 2807Note that closing the resamplefd is not sufficient to disable the 2808irqfd. The KVM_IRQFD_FLAG_RESAMPLE is only necessary on assignment 2809and need not be specified with KVM_IRQFD_FLAG_DEASSIGN. 2810 2811On arm/arm64, gsi routing being supported, the following can happen: 2812 2813- in case no routing entry is associated to this gsi, injection fails 2814- in case the gsi is associated to an irqchip routing entry, 2815 irqchip.pin + 32 corresponds to the injected SPI ID. 2816- in case the gsi is associated to an MSI routing entry, the MSI 2817 message and device ID are translated into an LPI (support restricted 2818 to GICv3 ITS in-kernel emulation). 2819 28204.76 KVM_PPC_ALLOCATE_HTAB 2821-------------------------- 2822 2823:Capability: KVM_CAP_PPC_ALLOC_HTAB 2824:Architectures: powerpc 2825:Type: vm ioctl 2826:Parameters: Pointer to u32 containing hash table order (in/out) 2827:Returns: 0 on success, -1 on error 2828 2829This requests the host kernel to allocate an MMU hash table for a 2830guest using the PAPR paravirtualization interface. This only does 2831anything if the kernel is configured to use the Book 3S HV style of 2832virtualization. Otherwise the capability doesn't exist and the ioctl 2833returns an ENOTTY error. The rest of this description assumes Book 3S 2834HV. 2835 2836There must be no vcpus running when this ioctl is called; if there 2837are, it will do nothing and return an EBUSY error. 2838 2839The parameter is a pointer to a 32-bit unsigned integer variable 2840containing the order (log base 2) of the desired size of the hash 2841table, which must be between 18 and 46. On successful return from the 2842ioctl, the value will not be changed by the kernel. 2843 2844If no hash table has been allocated when any vcpu is asked to run 2845(with the KVM_RUN ioctl), the host kernel will allocate a 2846default-sized hash table (16 MB). 2847 2848If this ioctl is called when a hash table has already been allocated, 2849with a different order from the existing hash table, the existing hash 2850table will be freed and a new one allocated. If this is ioctl is 2851called when a hash table has already been allocated of the same order 2852as specified, the kernel will clear out the existing hash table (zero 2853all HPTEs). In either case, if the guest is using the virtualized 2854real-mode area (VRMA) facility, the kernel will re-create the VMRA 2855HPTEs on the next KVM_RUN of any vcpu. 2856 28574.77 KVM_S390_INTERRUPT 2858----------------------- 2859 2860:Capability: basic 2861:Architectures: s390 2862:Type: vm ioctl, vcpu ioctl 2863:Parameters: struct kvm_s390_interrupt (in) 2864:Returns: 0 on success, -1 on error 2865 2866Allows to inject an interrupt to the guest. Interrupts can be floating 2867(vm ioctl) or per cpu (vcpu ioctl), depending on the interrupt type. 2868 2869Interrupt parameters are passed via kvm_s390_interrupt:: 2870 2871 struct kvm_s390_interrupt { 2872 __u32 type; 2873 __u32 parm; 2874 __u64 parm64; 2875 }; 2876 2877type can be one of the following: 2878 2879KVM_S390_SIGP_STOP (vcpu) 2880 - sigp stop; optional flags in parm 2881KVM_S390_PROGRAM_INT (vcpu) 2882 - program check; code in parm 2883KVM_S390_SIGP_SET_PREFIX (vcpu) 2884 - sigp set prefix; prefix address in parm 2885KVM_S390_RESTART (vcpu) 2886 - restart 2887KVM_S390_INT_CLOCK_COMP (vcpu) 2888 - clock comparator interrupt 2889KVM_S390_INT_CPU_TIMER (vcpu) 2890 - CPU timer interrupt 2891KVM_S390_INT_VIRTIO (vm) 2892 - virtio external interrupt; external interrupt 2893 parameters in parm and parm64 2894KVM_S390_INT_SERVICE (vm) 2895 - sclp external interrupt; sclp parameter in parm 2896KVM_S390_INT_EMERGENCY (vcpu) 2897 - sigp emergency; source cpu in parm 2898KVM_S390_INT_EXTERNAL_CALL (vcpu) 2899 - sigp external call; source cpu in parm 2900KVM_S390_INT_IO(ai,cssid,ssid,schid) (vm) 2901 - compound value to indicate an 2902 I/O interrupt (ai - adapter interrupt; cssid,ssid,schid - subchannel); 2903 I/O interruption parameters in parm (subchannel) and parm64 (intparm, 2904 interruption subclass) 2905KVM_S390_MCHK (vm, vcpu) 2906 - machine check interrupt; cr 14 bits in parm, machine check interrupt 2907 code in parm64 (note that machine checks needing further payload are not 2908 supported by this ioctl) 2909 2910This is an asynchronous vcpu ioctl and can be invoked from any thread. 2911 29124.78 KVM_PPC_GET_HTAB_FD 2913------------------------ 2914 2915:Capability: KVM_CAP_PPC_HTAB_FD 2916:Architectures: powerpc 2917:Type: vm ioctl 2918:Parameters: Pointer to struct kvm_get_htab_fd (in) 2919:Returns: file descriptor number (>= 0) on success, -1 on error 2920 2921This returns a file descriptor that can be used either to read out the 2922entries in the guest's hashed page table (HPT), or to write entries to 2923initialize the HPT. The returned fd can only be written to if the 2924KVM_GET_HTAB_WRITE bit is set in the flags field of the argument, and 2925can only be read if that bit is clear. The argument struct looks like 2926this:: 2927 2928 /* For KVM_PPC_GET_HTAB_FD */ 2929 struct kvm_get_htab_fd { 2930 __u64 flags; 2931 __u64 start_index; 2932 __u64 reserved[2]; 2933 }; 2934 2935 /* Values for kvm_get_htab_fd.flags */ 2936 #define KVM_GET_HTAB_BOLTED_ONLY ((__u64)0x1) 2937 #define KVM_GET_HTAB_WRITE ((__u64)0x2) 2938 2939The 'start_index' field gives the index in the HPT of the entry at 2940which to start reading. It is ignored when writing. 2941 2942Reads on the fd will initially supply information about all 2943"interesting" HPT entries. Interesting entries are those with the 2944bolted bit set, if the KVM_GET_HTAB_BOLTED_ONLY bit is set, otherwise 2945all entries. When the end of the HPT is reached, the read() will 2946return. If read() is called again on the fd, it will start again from 2947the beginning of the HPT, but will only return HPT entries that have 2948changed since they were last read. 2949 2950Data read or written is structured as a header (8 bytes) followed by a 2951series of valid HPT entries (16 bytes) each. The header indicates how 2952many valid HPT entries there are and how many invalid entries follow 2953the valid entries. The invalid entries are not represented explicitly 2954in the stream. The header format is:: 2955 2956 struct kvm_get_htab_header { 2957 __u32 index; 2958 __u16 n_valid; 2959 __u16 n_invalid; 2960 }; 2961 2962Writes to the fd create HPT entries starting at the index given in the 2963header; first 'n_valid' valid entries with contents from the data 2964written, then 'n_invalid' invalid entries, invalidating any previously 2965valid entries found. 2966 29674.79 KVM_CREATE_DEVICE 2968---------------------- 2969 2970:Capability: KVM_CAP_DEVICE_CTRL 2971:Type: vm ioctl 2972:Parameters: struct kvm_create_device (in/out) 2973:Returns: 0 on success, -1 on error 2974 2975Errors: 2976 2977 ====== ======================================================= 2978 ENODEV The device type is unknown or unsupported 2979 EEXIST Device already created, and this type of device may not 2980 be instantiated multiple times 2981 ====== ======================================================= 2982 2983 Other error conditions may be defined by individual device types or 2984 have their standard meanings. 2985 2986Creates an emulated device in the kernel. The file descriptor returned 2987in fd can be used with KVM_SET/GET/HAS_DEVICE_ATTR. 2988 2989If the KVM_CREATE_DEVICE_TEST flag is set, only test whether the 2990device type is supported (not necessarily whether it can be created 2991in the current vm). 2992 2993Individual devices should not define flags. Attributes should be used 2994for specifying any behavior that is not implied by the device type 2995number. 2996 2997:: 2998 2999 struct kvm_create_device { 3000 __u32 type; /* in: KVM_DEV_TYPE_xxx */ 3001 __u32 fd; /* out: device handle */ 3002 __u32 flags; /* in: KVM_CREATE_DEVICE_xxx */ 3003 }; 3004 30054.80 KVM_SET_DEVICE_ATTR/KVM_GET_DEVICE_ATTR 3006-------------------------------------------- 3007 3008:Capability: KVM_CAP_DEVICE_CTRL, KVM_CAP_VM_ATTRIBUTES for vm device, 3009 KVM_CAP_VCPU_ATTRIBUTES for vcpu device 3010:Type: device ioctl, vm ioctl, vcpu ioctl 3011:Parameters: struct kvm_device_attr 3012:Returns: 0 on success, -1 on error 3013 3014Errors: 3015 3016 ===== ============================================================= 3017 ENXIO The group or attribute is unknown/unsupported for this device 3018 or hardware support is missing. 3019 EPERM The attribute cannot (currently) be accessed this way 3020 (e.g. read-only attribute, or attribute that only makes 3021 sense when the device is in a different state) 3022 ===== ============================================================= 3023 3024 Other error conditions may be defined by individual device types. 3025 3026Gets/sets a specified piece of device configuration and/or state. The 3027semantics are device-specific. See individual device documentation in 3028the "devices" directory. As with ONE_REG, the size of the data 3029transferred is defined by the particular attribute. 3030 3031:: 3032 3033 struct kvm_device_attr { 3034 __u32 flags; /* no flags currently defined */ 3035 __u32 group; /* device-defined */ 3036 __u64 attr; /* group-defined */ 3037 __u64 addr; /* userspace address of attr data */ 3038 }; 3039 30404.81 KVM_HAS_DEVICE_ATTR 3041------------------------ 3042 3043:Capability: KVM_CAP_DEVICE_CTRL, KVM_CAP_VM_ATTRIBUTES for vm device, 3044 KVM_CAP_VCPU_ATTRIBUTES for vcpu device 3045:Type: device ioctl, vm ioctl, vcpu ioctl 3046:Parameters: struct kvm_device_attr 3047:Returns: 0 on success, -1 on error 3048 3049Errors: 3050 3051 ===== ============================================================= 3052 ENXIO The group or attribute is unknown/unsupported for this device 3053 or hardware support is missing. 3054 ===== ============================================================= 3055 3056Tests whether a device supports a particular attribute. A successful 3057return indicates the attribute is implemented. It does not necessarily 3058indicate that the attribute can be read or written in the device's 3059current state. "addr" is ignored. 3060 30614.82 KVM_ARM_VCPU_INIT 3062---------------------- 3063 3064:Capability: basic 3065:Architectures: arm, arm64 3066:Type: vcpu ioctl 3067:Parameters: struct kvm_vcpu_init (in) 3068:Returns: 0 on success; -1 on error 3069 3070Errors: 3071 3072 ====== ================================================================= 3073 EINVAL the target is unknown, or the combination of features is invalid. 3074 ENOENT a features bit specified is unknown. 3075 ====== ================================================================= 3076 3077This tells KVM what type of CPU to present to the guest, and what 3078optional features it should have. This will cause a reset of the cpu 3079registers to their initial values. If this is not called, KVM_RUN will 3080return ENOEXEC for that vcpu. 3081 3082Note that because some registers reflect machine topology, all vcpus 3083should be created before this ioctl is invoked. 3084 3085Userspace can call this function multiple times for a given vcpu, including 3086after the vcpu has been run. This will reset the vcpu to its initial 3087state. All calls to this function after the initial call must use the same 3088target and same set of feature flags, otherwise EINVAL will be returned. 3089 3090Possible features: 3091 3092 - KVM_ARM_VCPU_POWER_OFF: Starts the CPU in a power-off state. 3093 Depends on KVM_CAP_ARM_PSCI. If not set, the CPU will be powered on 3094 and execute guest code when KVM_RUN is called. 3095 - KVM_ARM_VCPU_EL1_32BIT: Starts the CPU in a 32bit mode. 3096 Depends on KVM_CAP_ARM_EL1_32BIT (arm64 only). 3097 - KVM_ARM_VCPU_PSCI_0_2: Emulate PSCI v0.2 (or a future revision 3098 backward compatible with v0.2) for the CPU. 3099 Depends on KVM_CAP_ARM_PSCI_0_2. 3100 - KVM_ARM_VCPU_PMU_V3: Emulate PMUv3 for the CPU. 3101 Depends on KVM_CAP_ARM_PMU_V3. 3102 3103 - KVM_ARM_VCPU_PTRAUTH_ADDRESS: Enables Address Pointer authentication 3104 for arm64 only. 3105 Depends on KVM_CAP_ARM_PTRAUTH_ADDRESS. 3106 If KVM_CAP_ARM_PTRAUTH_ADDRESS and KVM_CAP_ARM_PTRAUTH_GENERIC are 3107 both present, then both KVM_ARM_VCPU_PTRAUTH_ADDRESS and 3108 KVM_ARM_VCPU_PTRAUTH_GENERIC must be requested or neither must be 3109 requested. 3110 3111 - KVM_ARM_VCPU_PTRAUTH_GENERIC: Enables Generic Pointer authentication 3112 for arm64 only. 3113 Depends on KVM_CAP_ARM_PTRAUTH_GENERIC. 3114 If KVM_CAP_ARM_PTRAUTH_ADDRESS and KVM_CAP_ARM_PTRAUTH_GENERIC are 3115 both present, then both KVM_ARM_VCPU_PTRAUTH_ADDRESS and 3116 KVM_ARM_VCPU_PTRAUTH_GENERIC must be requested or neither must be 3117 requested. 3118 3119 - KVM_ARM_VCPU_SVE: Enables SVE for the CPU (arm64 only). 3120 Depends on KVM_CAP_ARM_SVE. 3121 Requires KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE): 3122 3123 * After KVM_ARM_VCPU_INIT: 3124 3125 - KVM_REG_ARM64_SVE_VLS may be read using KVM_GET_ONE_REG: the 3126 initial value of this pseudo-register indicates the best set of 3127 vector lengths possible for a vcpu on this host. 3128 3129 * Before KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE): 3130 3131 - KVM_RUN and KVM_GET_REG_LIST are not available; 3132 3133 - KVM_GET_ONE_REG and KVM_SET_ONE_REG cannot be used to access 3134 the scalable archietctural SVE registers 3135 KVM_REG_ARM64_SVE_ZREG(), KVM_REG_ARM64_SVE_PREG() or 3136 KVM_REG_ARM64_SVE_FFR; 3137 3138 - KVM_REG_ARM64_SVE_VLS may optionally be written using 3139 KVM_SET_ONE_REG, to modify the set of vector lengths available 3140 for the vcpu. 3141 3142 * After KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE): 3143 3144 - the KVM_REG_ARM64_SVE_VLS pseudo-register is immutable, and can 3145 no longer be written using KVM_SET_ONE_REG. 3146 31474.83 KVM_ARM_PREFERRED_TARGET 3148----------------------------- 3149 3150:Capability: basic 3151:Architectures: arm, arm64 3152:Type: vm ioctl 3153:Parameters: struct struct kvm_vcpu_init (out) 3154:Returns: 0 on success; -1 on error 3155 3156Errors: 3157 3158 ====== ========================================== 3159 ENODEV no preferred target available for the host 3160 ====== ========================================== 3161 3162This queries KVM for preferred CPU target type which can be emulated 3163by KVM on underlying host. 3164 3165The ioctl returns struct kvm_vcpu_init instance containing information 3166about preferred CPU target type and recommended features for it. The 3167kvm_vcpu_init->features bitmap returned will have feature bits set if 3168the preferred target recommends setting these features, but this is 3169not mandatory. 3170 3171The information returned by this ioctl can be used to prepare an instance 3172of struct kvm_vcpu_init for KVM_ARM_VCPU_INIT ioctl which will result in 3173in VCPU matching underlying host. 3174 3175 31764.84 KVM_GET_REG_LIST 3177--------------------- 3178 3179:Capability: basic 3180:Architectures: arm, arm64, mips 3181:Type: vcpu ioctl 3182:Parameters: struct kvm_reg_list (in/out) 3183:Returns: 0 on success; -1 on error 3184 3185Errors: 3186 3187 ===== ============================================================== 3188 E2BIG the reg index list is too big to fit in the array specified by 3189 the user (the number required will be written into n). 3190 ===== ============================================================== 3191 3192:: 3193 3194 struct kvm_reg_list { 3195 __u64 n; /* number of registers in reg[] */ 3196 __u64 reg[0]; 3197 }; 3198 3199This ioctl returns the guest registers that are supported for the 3200KVM_GET_ONE_REG/KVM_SET_ONE_REG calls. 3201 3202 32034.85 KVM_ARM_SET_DEVICE_ADDR (deprecated) 3204----------------------------------------- 3205 3206:Capability: KVM_CAP_ARM_SET_DEVICE_ADDR 3207:Architectures: arm, arm64 3208:Type: vm ioctl 3209:Parameters: struct kvm_arm_device_address (in) 3210:Returns: 0 on success, -1 on error 3211 3212Errors: 3213 3214 ====== ============================================ 3215 ENODEV The device id is unknown 3216 ENXIO Device not supported on current system 3217 EEXIST Address already set 3218 E2BIG Address outside guest physical address space 3219 EBUSY Address overlaps with other device range 3220 ====== ============================================ 3221 3222:: 3223 3224 struct kvm_arm_device_addr { 3225 __u64 id; 3226 __u64 addr; 3227 }; 3228 3229Specify a device address in the guest's physical address space where guests 3230can access emulated or directly exposed devices, which the host kernel needs 3231to know about. The id field is an architecture specific identifier for a 3232specific device. 3233 3234ARM/arm64 divides the id field into two parts, a device id and an 3235address type id specific to the individual device:: 3236 3237 bits: | 63 ... 32 | 31 ... 16 | 15 ... 0 | 3238 field: | 0x00000000 | device id | addr type id | 3239 3240ARM/arm64 currently only require this when using the in-kernel GIC 3241support for the hardware VGIC features, using KVM_ARM_DEVICE_VGIC_V2 3242as the device id. When setting the base address for the guest's 3243mapping of the VGIC virtual CPU and distributor interface, the ioctl 3244must be called after calling KVM_CREATE_IRQCHIP, but before calling 3245KVM_RUN on any of the VCPUs. Calling this ioctl twice for any of the 3246base addresses will return -EEXIST. 3247 3248Note, this IOCTL is deprecated and the more flexible SET/GET_DEVICE_ATTR API 3249should be used instead. 3250 3251 32524.86 KVM_PPC_RTAS_DEFINE_TOKEN 3253------------------------------ 3254 3255:Capability: KVM_CAP_PPC_RTAS 3256:Architectures: ppc 3257:Type: vm ioctl 3258:Parameters: struct kvm_rtas_token_args 3259:Returns: 0 on success, -1 on error 3260 3261Defines a token value for a RTAS (Run Time Abstraction Services) 3262service in order to allow it to be handled in the kernel. The 3263argument struct gives the name of the service, which must be the name 3264of a service that has a kernel-side implementation. If the token 3265value is non-zero, it will be associated with that service, and 3266subsequent RTAS calls by the guest specifying that token will be 3267handled by the kernel. If the token value is 0, then any token 3268associated with the service will be forgotten, and subsequent RTAS 3269calls by the guest for that service will be passed to userspace to be 3270handled. 3271 32724.87 KVM_SET_GUEST_DEBUG 3273------------------------ 3274 3275:Capability: KVM_CAP_SET_GUEST_DEBUG 3276:Architectures: x86, s390, ppc, arm64 3277:Type: vcpu ioctl 3278:Parameters: struct kvm_guest_debug (in) 3279:Returns: 0 on success; -1 on error 3280 3281:: 3282 3283 struct kvm_guest_debug { 3284 __u32 control; 3285 __u32 pad; 3286 struct kvm_guest_debug_arch arch; 3287 }; 3288 3289Set up the processor specific debug registers and configure vcpu for 3290handling guest debug events. There are two parts to the structure, the 3291first a control bitfield indicates the type of debug events to handle 3292when running. Common control bits are: 3293 3294 - KVM_GUESTDBG_ENABLE: guest debugging is enabled 3295 - KVM_GUESTDBG_SINGLESTEP: the next run should single-step 3296 3297The top 16 bits of the control field are architecture specific control 3298flags which can include the following: 3299 3300 - KVM_GUESTDBG_USE_SW_BP: using software breakpoints [x86, arm64] 3301 - KVM_GUESTDBG_USE_HW_BP: using hardware breakpoints [x86, s390, arm64] 3302 - KVM_GUESTDBG_INJECT_DB: inject DB type exception [x86] 3303 - KVM_GUESTDBG_INJECT_BP: inject BP type exception [x86] 3304 - KVM_GUESTDBG_EXIT_PENDING: trigger an immediate guest exit [s390] 3305 3306For example KVM_GUESTDBG_USE_SW_BP indicates that software breakpoints 3307are enabled in memory so we need to ensure breakpoint exceptions are 3308correctly trapped and the KVM run loop exits at the breakpoint and not 3309running off into the normal guest vector. For KVM_GUESTDBG_USE_HW_BP 3310we need to ensure the guest vCPUs architecture specific registers are 3311updated to the correct (supplied) values. 3312 3313The second part of the structure is architecture specific and 3314typically contains a set of debug registers. 3315 3316For arm64 the number of debug registers is implementation defined and 3317can be determined by querying the KVM_CAP_GUEST_DEBUG_HW_BPS and 3318KVM_CAP_GUEST_DEBUG_HW_WPS capabilities which return a positive number 3319indicating the number of supported registers. 3320 3321For ppc, the KVM_CAP_PPC_GUEST_DEBUG_SSTEP capability indicates whether 3322the single-step debug event (KVM_GUESTDBG_SINGLESTEP) is supported. 3323 3324When debug events exit the main run loop with the reason 3325KVM_EXIT_DEBUG with the kvm_debug_exit_arch part of the kvm_run 3326structure containing architecture specific debug information. 3327 33284.88 KVM_GET_EMULATED_CPUID 3329--------------------------- 3330 3331:Capability: KVM_CAP_EXT_EMUL_CPUID 3332:Architectures: x86 3333:Type: system ioctl 3334:Parameters: struct kvm_cpuid2 (in/out) 3335:Returns: 0 on success, -1 on error 3336 3337:: 3338 3339 struct kvm_cpuid2 { 3340 __u32 nent; 3341 __u32 flags; 3342 struct kvm_cpuid_entry2 entries[0]; 3343 }; 3344 3345The member 'flags' is used for passing flags from userspace. 3346 3347:: 3348 3349 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX BIT(0) 3350 #define KVM_CPUID_FLAG_STATEFUL_FUNC BIT(1) 3351 #define KVM_CPUID_FLAG_STATE_READ_NEXT BIT(2) 3352 3353 struct kvm_cpuid_entry2 { 3354 __u32 function; 3355 __u32 index; 3356 __u32 flags; 3357 __u32 eax; 3358 __u32 ebx; 3359 __u32 ecx; 3360 __u32 edx; 3361 __u32 padding[3]; 3362 }; 3363 3364This ioctl returns x86 cpuid features which are emulated by 3365kvm.Userspace can use the information returned by this ioctl to query 3366which features are emulated by kvm instead of being present natively. 3367 3368Userspace invokes KVM_GET_EMULATED_CPUID by passing a kvm_cpuid2 3369structure with the 'nent' field indicating the number of entries in 3370the variable-size array 'entries'. If the number of entries is too low 3371to describe the cpu capabilities, an error (E2BIG) is returned. If the 3372number is too high, the 'nent' field is adjusted and an error (ENOMEM) 3373is returned. If the number is just right, the 'nent' field is adjusted 3374to the number of valid entries in the 'entries' array, which is then 3375filled. 3376 3377The entries returned are the set CPUID bits of the respective features 3378which kvm emulates, as returned by the CPUID instruction, with unknown 3379or unsupported feature bits cleared. 3380 3381Features like x2apic, for example, may not be present in the host cpu 3382but are exposed by kvm in KVM_GET_SUPPORTED_CPUID because they can be 3383emulated efficiently and thus not included here. 3384 3385The fields in each entry are defined as follows: 3386 3387 function: 3388 the eax value used to obtain the entry 3389 index: 3390 the ecx value used to obtain the entry (for entries that are 3391 affected by ecx) 3392 flags: 3393 an OR of zero or more of the following: 3394 3395 KVM_CPUID_FLAG_SIGNIFCANT_INDEX: 3396 if the index field is valid 3397 KVM_CPUID_FLAG_STATEFUL_FUNC: 3398 if cpuid for this function returns different values for successive 3399 invocations; there will be several entries with the same function, 3400 all with this flag set 3401 KVM_CPUID_FLAG_STATE_READ_NEXT: 3402 for KVM_CPUID_FLAG_STATEFUL_FUNC entries, set if this entry is 3403 the first entry to be read by a cpu 3404 3405 eax, ebx, ecx, edx: 3406 3407 the values returned by the cpuid instruction for 3408 this function/index combination 3409 34104.89 KVM_S390_MEM_OP 3411-------------------- 3412 3413:Capability: KVM_CAP_S390_MEM_OP 3414:Architectures: s390 3415:Type: vcpu ioctl 3416:Parameters: struct kvm_s390_mem_op (in) 3417:Returns: = 0 on success, 3418 < 0 on generic error (e.g. -EFAULT or -ENOMEM), 3419 > 0 if an exception occurred while walking the page tables 3420 3421Read or write data from/to the logical (virtual) memory of a VCPU. 3422 3423Parameters are specified via the following structure:: 3424 3425 struct kvm_s390_mem_op { 3426 __u64 gaddr; /* the guest address */ 3427 __u64 flags; /* flags */ 3428 __u32 size; /* amount of bytes */ 3429 __u32 op; /* type of operation */ 3430 __u64 buf; /* buffer in userspace */ 3431 __u8 ar; /* the access register number */ 3432 __u8 reserved[31]; /* should be set to 0 */ 3433 }; 3434 3435The type of operation is specified in the "op" field. It is either 3436KVM_S390_MEMOP_LOGICAL_READ for reading from logical memory space or 3437KVM_S390_MEMOP_LOGICAL_WRITE for writing to logical memory space. The 3438KVM_S390_MEMOP_F_CHECK_ONLY flag can be set in the "flags" field to check 3439whether the corresponding memory access would create an access exception 3440(without touching the data in the memory at the destination). In case an 3441access exception occurred while walking the MMU tables of the guest, the 3442ioctl returns a positive error number to indicate the type of exception. 3443This exception is also raised directly at the corresponding VCPU if the 3444flag KVM_S390_MEMOP_F_INJECT_EXCEPTION is set in the "flags" field. 3445 3446The start address of the memory region has to be specified in the "gaddr" 3447field, and the length of the region in the "size" field (which must not 3448be 0). The maximum value for "size" can be obtained by checking the 3449KVM_CAP_S390_MEM_OP capability. "buf" is the buffer supplied by the 3450userspace application where the read data should be written to for 3451KVM_S390_MEMOP_LOGICAL_READ, or where the data that should be written is 3452stored for a KVM_S390_MEMOP_LOGICAL_WRITE. When KVM_S390_MEMOP_F_CHECK_ONLY 3453is specified, "buf" is unused and can be NULL. "ar" designates the access 3454register number to be used; the valid range is 0..15. 3455 3456The "reserved" field is meant for future extensions. It is not used by 3457KVM with the currently defined set of flags. 3458 34594.90 KVM_S390_GET_SKEYS 3460----------------------- 3461 3462:Capability: KVM_CAP_S390_SKEYS 3463:Architectures: s390 3464:Type: vm ioctl 3465:Parameters: struct kvm_s390_skeys 3466:Returns: 0 on success, KVM_S390_GET_KEYS_NONE if guest is not using storage 3467 keys, negative value on error 3468 3469This ioctl is used to get guest storage key values on the s390 3470architecture. The ioctl takes parameters via the kvm_s390_skeys struct:: 3471 3472 struct kvm_s390_skeys { 3473 __u64 start_gfn; 3474 __u64 count; 3475 __u64 skeydata_addr; 3476 __u32 flags; 3477 __u32 reserved[9]; 3478 }; 3479 3480The start_gfn field is the number of the first guest frame whose storage keys 3481you want to get. 3482 3483The count field is the number of consecutive frames (starting from start_gfn) 3484whose storage keys to get. The count field must be at least 1 and the maximum 3485allowed value is defined as KVM_S390_SKEYS_ALLOC_MAX. Values outside this range 3486will cause the ioctl to return -EINVAL. 3487 3488The skeydata_addr field is the address to a buffer large enough to hold count 3489bytes. This buffer will be filled with storage key data by the ioctl. 3490 34914.91 KVM_S390_SET_SKEYS 3492----------------------- 3493 3494:Capability: KVM_CAP_S390_SKEYS 3495:Architectures: s390 3496:Type: vm ioctl 3497:Parameters: struct kvm_s390_skeys 3498:Returns: 0 on success, negative value on error 3499 3500This ioctl is used to set guest storage key values on the s390 3501architecture. The ioctl takes parameters via the kvm_s390_skeys struct. 3502See section on KVM_S390_GET_SKEYS for struct definition. 3503 3504The start_gfn field is the number of the first guest frame whose storage keys 3505you want to set. 3506 3507The count field is the number of consecutive frames (starting from start_gfn) 3508whose storage keys to get. The count field must be at least 1 and the maximum 3509allowed value is defined as KVM_S390_SKEYS_ALLOC_MAX. Values outside this range 3510will cause the ioctl to return -EINVAL. 3511 3512The skeydata_addr field is the address to a buffer containing count bytes of 3513storage keys. Each byte in the buffer will be set as the storage key for a 3514single frame starting at start_gfn for count frames. 3515 3516Note: If any architecturally invalid key value is found in the given data then 3517the ioctl will return -EINVAL. 3518 35194.92 KVM_S390_IRQ 3520----------------- 3521 3522:Capability: KVM_CAP_S390_INJECT_IRQ 3523:Architectures: s390 3524:Type: vcpu ioctl 3525:Parameters: struct kvm_s390_irq (in) 3526:Returns: 0 on success, -1 on error 3527 3528Errors: 3529 3530 3531 ====== ================================================================= 3532 EINVAL interrupt type is invalid 3533 type is KVM_S390_SIGP_STOP and flag parameter is invalid value, 3534 type is KVM_S390_INT_EXTERNAL_CALL and code is bigger 3535 than the maximum of VCPUs 3536 EBUSY type is KVM_S390_SIGP_SET_PREFIX and vcpu is not stopped, 3537 type is KVM_S390_SIGP_STOP and a stop irq is already pending, 3538 type is KVM_S390_INT_EXTERNAL_CALL and an external call interrupt 3539 is already pending 3540 ====== ================================================================= 3541 3542Allows to inject an interrupt to the guest. 3543 3544Using struct kvm_s390_irq as a parameter allows 3545to inject additional payload which is not 3546possible via KVM_S390_INTERRUPT. 3547 3548Interrupt parameters are passed via kvm_s390_irq:: 3549 3550 struct kvm_s390_irq { 3551 __u64 type; 3552 union { 3553 struct kvm_s390_io_info io; 3554 struct kvm_s390_ext_info ext; 3555 struct kvm_s390_pgm_info pgm; 3556 struct kvm_s390_emerg_info emerg; 3557 struct kvm_s390_extcall_info extcall; 3558 struct kvm_s390_prefix_info prefix; 3559 struct kvm_s390_stop_info stop; 3560 struct kvm_s390_mchk_info mchk; 3561 char reserved[64]; 3562 } u; 3563 }; 3564 3565type can be one of the following: 3566 3567- KVM_S390_SIGP_STOP - sigp stop; parameter in .stop 3568- KVM_S390_PROGRAM_INT - program check; parameters in .pgm 3569- KVM_S390_SIGP_SET_PREFIX - sigp set prefix; parameters in .prefix 3570- KVM_S390_RESTART - restart; no parameters 3571- KVM_S390_INT_CLOCK_COMP - clock comparator interrupt; no parameters 3572- KVM_S390_INT_CPU_TIMER - CPU timer interrupt; no parameters 3573- KVM_S390_INT_EMERGENCY - sigp emergency; parameters in .emerg 3574- KVM_S390_INT_EXTERNAL_CALL - sigp external call; parameters in .extcall 3575- KVM_S390_MCHK - machine check interrupt; parameters in .mchk 3576 3577This is an asynchronous vcpu ioctl and can be invoked from any thread. 3578 35794.94 KVM_S390_GET_IRQ_STATE 3580--------------------------- 3581 3582:Capability: KVM_CAP_S390_IRQ_STATE 3583:Architectures: s390 3584:Type: vcpu ioctl 3585:Parameters: struct kvm_s390_irq_state (out) 3586:Returns: >= number of bytes copied into buffer, 3587 -EINVAL if buffer size is 0, 3588 -ENOBUFS if buffer size is too small to fit all pending interrupts, 3589 -EFAULT if the buffer address was invalid 3590 3591This ioctl allows userspace to retrieve the complete state of all currently 3592pending interrupts in a single buffer. Use cases include migration 3593and introspection. The parameter structure contains the address of a 3594userspace buffer and its length:: 3595 3596 struct kvm_s390_irq_state { 3597 __u64 buf; 3598 __u32 flags; /* will stay unused for compatibility reasons */ 3599 __u32 len; 3600 __u32 reserved[4]; /* will stay unused for compatibility reasons */ 3601 }; 3602 3603Userspace passes in the above struct and for each pending interrupt a 3604struct kvm_s390_irq is copied to the provided buffer. 3605 3606The structure contains a flags and a reserved field for future extensions. As 3607the kernel never checked for flags == 0 and QEMU never pre-zeroed flags and 3608reserved, these fields can not be used in the future without breaking 3609compatibility. 3610 3611If -ENOBUFS is returned the buffer provided was too small and userspace 3612may retry with a bigger buffer. 3613 36144.95 KVM_S390_SET_IRQ_STATE 3615--------------------------- 3616 3617:Capability: KVM_CAP_S390_IRQ_STATE 3618:Architectures: s390 3619:Type: vcpu ioctl 3620:Parameters: struct kvm_s390_irq_state (in) 3621:Returns: 0 on success, 3622 -EFAULT if the buffer address was invalid, 3623 -EINVAL for an invalid buffer length (see below), 3624 -EBUSY if there were already interrupts pending, 3625 errors occurring when actually injecting the 3626 interrupt. See KVM_S390_IRQ. 3627 3628This ioctl allows userspace to set the complete state of all cpu-local 3629interrupts currently pending for the vcpu. It is intended for restoring 3630interrupt state after a migration. The input parameter is a userspace buffer 3631containing a struct kvm_s390_irq_state:: 3632 3633 struct kvm_s390_irq_state { 3634 __u64 buf; 3635 __u32 flags; /* will stay unused for compatibility reasons */ 3636 __u32 len; 3637 __u32 reserved[4]; /* will stay unused for compatibility reasons */ 3638 }; 3639 3640The restrictions for flags and reserved apply as well. 3641(see KVM_S390_GET_IRQ_STATE) 3642 3643The userspace memory referenced by buf contains a struct kvm_s390_irq 3644for each interrupt to be injected into the guest. 3645If one of the interrupts could not be injected for some reason the 3646ioctl aborts. 3647 3648len must be a multiple of sizeof(struct kvm_s390_irq). It must be > 0 3649and it must not exceed (max_vcpus + 32) * sizeof(struct kvm_s390_irq), 3650which is the maximum number of possibly pending cpu-local interrupts. 3651 36524.96 KVM_SMI 3653------------ 3654 3655:Capability: KVM_CAP_X86_SMM 3656:Architectures: x86 3657:Type: vcpu ioctl 3658:Parameters: none 3659:Returns: 0 on success, -1 on error 3660 3661Queues an SMI on the thread's vcpu. 3662 36634.97 KVM_CAP_PPC_MULTITCE 3664------------------------- 3665 3666:Capability: KVM_CAP_PPC_MULTITCE 3667:Architectures: ppc 3668:Type: vm 3669 3670This capability means the kernel is capable of handling hypercalls 3671H_PUT_TCE_INDIRECT and H_STUFF_TCE without passing those into the user 3672space. This significantly accelerates DMA operations for PPC KVM guests. 3673User space should expect that its handlers for these hypercalls 3674are not going to be called if user space previously registered LIOBN 3675in KVM (via KVM_CREATE_SPAPR_TCE or similar calls). 3676 3677In order to enable H_PUT_TCE_INDIRECT and H_STUFF_TCE use in the guest, 3678user space might have to advertise it for the guest. For example, 3679IBM pSeries (sPAPR) guest starts using them if "hcall-multi-tce" is 3680present in the "ibm,hypertas-functions" device-tree property. 3681 3682The hypercalls mentioned above may or may not be processed successfully 3683in the kernel based fast path. If they can not be handled by the kernel, 3684they will get passed on to user space. So user space still has to have 3685an implementation for these despite the in kernel acceleration. 3686 3687This capability is always enabled. 3688 36894.98 KVM_CREATE_SPAPR_TCE_64 3690---------------------------- 3691 3692:Capability: KVM_CAP_SPAPR_TCE_64 3693:Architectures: powerpc 3694:Type: vm ioctl 3695:Parameters: struct kvm_create_spapr_tce_64 (in) 3696:Returns: file descriptor for manipulating the created TCE table 3697 3698This is an extension for KVM_CAP_SPAPR_TCE which only supports 32bit 3699windows, described in 4.62 KVM_CREATE_SPAPR_TCE 3700 3701This capability uses extended struct in ioctl interface:: 3702 3703 /* for KVM_CAP_SPAPR_TCE_64 */ 3704 struct kvm_create_spapr_tce_64 { 3705 __u64 liobn; 3706 __u32 page_shift; 3707 __u32 flags; 3708 __u64 offset; /* in pages */ 3709 __u64 size; /* in pages */ 3710 }; 3711 3712The aim of extension is to support an additional bigger DMA window with 3713a variable page size. 3714KVM_CREATE_SPAPR_TCE_64 receives a 64bit window size, an IOMMU page shift and 3715a bus offset of the corresponding DMA window, @size and @offset are numbers 3716of IOMMU pages. 3717 3718@flags are not used at the moment. 3719 3720The rest of functionality is identical to KVM_CREATE_SPAPR_TCE. 3721 37224.99 KVM_REINJECT_CONTROL 3723------------------------- 3724 3725:Capability: KVM_CAP_REINJECT_CONTROL 3726:Architectures: x86 3727:Type: vm ioctl 3728:Parameters: struct kvm_reinject_control (in) 3729:Returns: 0 on success, 3730 -EFAULT if struct kvm_reinject_control cannot be read, 3731 -ENXIO if KVM_CREATE_PIT or KVM_CREATE_PIT2 didn't succeed earlier. 3732 3733i8254 (PIT) has two modes, reinject and !reinject. The default is reinject, 3734where KVM queues elapsed i8254 ticks and monitors completion of interrupt from 3735vector(s) that i8254 injects. Reinject mode dequeues a tick and injects its 3736interrupt whenever there isn't a pending interrupt from i8254. 3737!reinject mode injects an interrupt as soon as a tick arrives. 3738 3739:: 3740 3741 struct kvm_reinject_control { 3742 __u8 pit_reinject; 3743 __u8 reserved[31]; 3744 }; 3745 3746pit_reinject = 0 (!reinject mode) is recommended, unless running an old 3747operating system that uses the PIT for timing (e.g. Linux 2.4.x). 3748 37494.100 KVM_PPC_CONFIGURE_V3_MMU 3750------------------------------ 3751 3752:Capability: KVM_CAP_PPC_RADIX_MMU or KVM_CAP_PPC_HASH_MMU_V3 3753:Architectures: ppc 3754:Type: vm ioctl 3755:Parameters: struct kvm_ppc_mmuv3_cfg (in) 3756:Returns: 0 on success, 3757 -EFAULT if struct kvm_ppc_mmuv3_cfg cannot be read, 3758 -EINVAL if the configuration is invalid 3759 3760This ioctl controls whether the guest will use radix or HPT (hashed 3761page table) translation, and sets the pointer to the process table for 3762the guest. 3763 3764:: 3765 3766 struct kvm_ppc_mmuv3_cfg { 3767 __u64 flags; 3768 __u64 process_table; 3769 }; 3770 3771There are two bits that can be set in flags; KVM_PPC_MMUV3_RADIX and 3772KVM_PPC_MMUV3_GTSE. KVM_PPC_MMUV3_RADIX, if set, configures the guest 3773to use radix tree translation, and if clear, to use HPT translation. 3774KVM_PPC_MMUV3_GTSE, if set and if KVM permits it, configures the guest 3775to be able to use the global TLB and SLB invalidation instructions; 3776if clear, the guest may not use these instructions. 3777 3778The process_table field specifies the address and size of the guest 3779process table, which is in the guest's space. This field is formatted 3780as the second doubleword of the partition table entry, as defined in 3781the Power ISA V3.00, Book III section 5.7.6.1. 3782 37834.101 KVM_PPC_GET_RMMU_INFO 3784--------------------------- 3785 3786:Capability: KVM_CAP_PPC_RADIX_MMU 3787:Architectures: ppc 3788:Type: vm ioctl 3789:Parameters: struct kvm_ppc_rmmu_info (out) 3790:Returns: 0 on success, 3791 -EFAULT if struct kvm_ppc_rmmu_info cannot be written, 3792 -EINVAL if no useful information can be returned 3793 3794This ioctl returns a structure containing two things: (a) a list 3795containing supported radix tree geometries, and (b) a list that maps 3796page sizes to put in the "AP" (actual page size) field for the tlbie 3797(TLB invalidate entry) instruction. 3798 3799:: 3800 3801 struct kvm_ppc_rmmu_info { 3802 struct kvm_ppc_radix_geom { 3803 __u8 page_shift; 3804 __u8 level_bits[4]; 3805 __u8 pad[3]; 3806 } geometries[8]; 3807 __u32 ap_encodings[8]; 3808 }; 3809 3810The geometries[] field gives up to 8 supported geometries for the 3811radix page table, in terms of the log base 2 of the smallest page 3812size, and the number of bits indexed at each level of the tree, from 3813the PTE level up to the PGD level in that order. Any unused entries 3814will have 0 in the page_shift field. 3815 3816The ap_encodings gives the supported page sizes and their AP field 3817encodings, encoded with the AP value in the top 3 bits and the log 3818base 2 of the page size in the bottom 6 bits. 3819 38204.102 KVM_PPC_RESIZE_HPT_PREPARE 3821-------------------------------- 3822 3823:Capability: KVM_CAP_SPAPR_RESIZE_HPT 3824:Architectures: powerpc 3825:Type: vm ioctl 3826:Parameters: struct kvm_ppc_resize_hpt (in) 3827:Returns: 0 on successful completion, 3828 >0 if a new HPT is being prepared, the value is an estimated 3829 number of milliseconds until preparation is complete, 3830 -EFAULT if struct kvm_reinject_control cannot be read, 3831 -EINVAL if the supplied shift or flags are invalid, 3832 -ENOMEM if unable to allocate the new HPT, 3833 -ENOSPC if there was a hash collision 3834 3835:: 3836 3837 struct kvm_ppc_rmmu_info { 3838 struct kvm_ppc_radix_geom { 3839 __u8 page_shift; 3840 __u8 level_bits[4]; 3841 __u8 pad[3]; 3842 } geometries[8]; 3843 __u32 ap_encodings[8]; 3844 }; 3845 3846The geometries[] field gives up to 8 supported geometries for the 3847radix page table, in terms of the log base 2 of the smallest page 3848size, and the number of bits indexed at each level of the tree, from 3849the PTE level up to the PGD level in that order. Any unused entries 3850will have 0 in the page_shift field. 3851 3852The ap_encodings gives the supported page sizes and their AP field 3853encodings, encoded with the AP value in the top 3 bits and the log 3854base 2 of the page size in the bottom 6 bits. 3855 38564.102 KVM_PPC_RESIZE_HPT_PREPARE 3857-------------------------------- 3858 3859:Capability: KVM_CAP_SPAPR_RESIZE_HPT 3860:Architectures: powerpc 3861:Type: vm ioctl 3862:Parameters: struct kvm_ppc_resize_hpt (in) 3863:Returns: 0 on successful completion, 3864 >0 if a new HPT is being prepared, the value is an estimated 3865 number of milliseconds until preparation is complete, 3866 -EFAULT if struct kvm_reinject_control cannot be read, 3867 -EINVAL if the supplied shift or flags are invalid,when moving existing 3868 HPT entries to the new HPT, 3869 -EIO on other error conditions 3870 3871Used to implement the PAPR extension for runtime resizing of a guest's 3872Hashed Page Table (HPT). Specifically this starts, stops or monitors 3873the preparation of a new potential HPT for the guest, essentially 3874implementing the H_RESIZE_HPT_PREPARE hypercall. 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 3903:: 3904 3905 struct kvm_ppc_resize_hpt { 3906 __u64 flags; 3907 __u32 shift; 3908 __u32 pad; 3909 }; 3910 39114.103 KVM_PPC_RESIZE_HPT_COMMIT 3912------------------------------- 3913 3914:Capability: KVM_CAP_SPAPR_RESIZE_HPT 3915:Architectures: powerpc 3916:Type: vm ioctl 3917:Parameters: struct kvm_ppc_resize_hpt (in) 3918:Returns: 0 on successful completion, 3919 -EFAULT if struct kvm_reinject_control cannot be read, 3920 -EINVAL if the supplied shift or flags are invalid, 3921 -ENXIO is there is no pending HPT, or the pending HPT doesn't 3922 have the requested size, 3923 -EBUSY if the pending HPT is not fully prepared, 3924 -ENOSPC if there was a hash collision when moving existing 3925 HPT entries to the new HPT, 3926 -EIO on other error conditions 3927 3928Used to implement the PAPR extension for runtime resizing of a guest's 3929Hashed Page Table (HPT). Specifically this requests that the guest be 3930transferred to working with the new HPT, essentially implementing the 3931H_RESIZE_HPT_COMMIT hypercall. 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 3948:: 3949 3950 struct kvm_ppc_resize_hpt { 3951 __u64 flags; 3952 __u32 shift; 3953 __u32 pad; 3954 }; 3955 39564.104 KVM_X86_GET_MCE_CAP_SUPPORTED 3957----------------------------------- 3958 3959:Capability: KVM_CAP_MCE 3960:Architectures: x86 3961:Type: system ioctl 3962:Parameters: u64 mce_cap (out) 3963:Returns: 0 on success, -1 on error 3964 3965Returns supported MCE capabilities. The u64 mce_cap parameter 3966has the same format as the MSR_IA32_MCG_CAP register. Supported 3967capabilities will have the corresponding bits set. 3968 39694.105 KVM_X86_SETUP_MCE 3970----------------------- 3971 3972:Capability: KVM_CAP_MCE 3973:Architectures: x86 3974:Type: vcpu ioctl 3975:Parameters: u64 mcg_cap (in) 3976:Returns: 0 on success, 3977 -EFAULT if u64 mcg_cap cannot be read, 3978 -EINVAL if the requested number of banks is invalid, 3979 -EINVAL if requested MCE capability is not supported. 3980 3981Initializes MCE support for use. The u64 mcg_cap parameter 3982has the same format as the MSR_IA32_MCG_CAP register and 3983specifies which capabilities should be enabled. The maximum 3984supported number of error-reporting banks can be retrieved when 3985checking for KVM_CAP_MCE. The supported capabilities can be 3986retrieved with KVM_X86_GET_MCE_CAP_SUPPORTED. 3987 39884.106 KVM_X86_SET_MCE 3989--------------------- 3990 3991:Capability: KVM_CAP_MCE 3992:Architectures: x86 3993:Type: vcpu ioctl 3994:Parameters: struct kvm_x86_mce (in) 3995:Returns: 0 on success, 3996 -EFAULT if struct kvm_x86_mce cannot be read, 3997 -EINVAL if the bank number is invalid, 3998 -EINVAL if VAL bit is not set in status field. 3999 4000Inject a machine check error (MCE) into the guest. The input 4001parameter is:: 4002 4003 struct kvm_x86_mce { 4004 __u64 status; 4005 __u64 addr; 4006 __u64 misc; 4007 __u64 mcg_status; 4008 __u8 bank; 4009 __u8 pad1[7]; 4010 __u64 pad2[3]; 4011 }; 4012 4013If the MCE being reported is an uncorrected error, KVM will 4014inject it as an MCE exception into the guest. If the guest 4015MCG_STATUS register reports that an MCE is in progress, KVM 4016causes an KVM_EXIT_SHUTDOWN vmexit. 4017 4018Otherwise, if the MCE is a corrected error, KVM will just 4019store it in the corresponding bank (provided this bank is 4020not holding a previously reported uncorrected error). 4021 40224.107 KVM_S390_GET_CMMA_BITS 4023---------------------------- 4024 4025:Capability: KVM_CAP_S390_CMMA_MIGRATION 4026:Architectures: s390 4027:Type: vm ioctl 4028:Parameters: struct kvm_s390_cmma_log (in, out) 4029:Returns: 0 on success, a negative value on error 4030 4031This ioctl is used to get the values of the CMMA bits on the s390 4032architecture. It is meant to be used in two scenarios: 4033 4034- During live migration to save the CMMA values. Live migration needs 4035 to be enabled via the KVM_REQ_START_MIGRATION VM property. 4036- To non-destructively peek at the CMMA values, with the flag 4037 KVM_S390_CMMA_PEEK set. 4038 4039The ioctl takes parameters via the kvm_s390_cmma_log struct. The desired 4040values are written to a buffer whose location is indicated via the "values" 4041member in the kvm_s390_cmma_log struct. The values in the input struct are 4042also updated as needed. 4043 4044Each CMMA value takes up one byte. 4045 4046:: 4047 4048 struct kvm_s390_cmma_log { 4049 __u64 start_gfn; 4050 __u32 count; 4051 __u32 flags; 4052 union { 4053 __u64 remaining; 4054 __u64 mask; 4055 }; 4056 __u64 values; 4057 }; 4058 4059start_gfn is the number of the first guest frame whose CMMA values are 4060to be retrieved, 4061 4062count is the length of the buffer in bytes, 4063 4064values points to the buffer where the result will be written to. 4065 4066If count is greater than KVM_S390_SKEYS_MAX, then it is considered to be 4067KVM_S390_SKEYS_MAX. KVM_S390_SKEYS_MAX is re-used for consistency with 4068other ioctls. 4069 4070The result is written in the buffer pointed to by the field values, and 4071the values of the input parameter are updated as follows. 4072 4073Depending on the flags, different actions are performed. The only 4074supported flag so far is KVM_S390_CMMA_PEEK. 4075 4076The default behaviour if KVM_S390_CMMA_PEEK is not set is: 4077start_gfn will indicate the first page frame whose CMMA bits were dirty. 4078It is not necessarily the same as the one passed as input, as clean pages 4079are skipped. 4080 4081count will indicate the number of bytes actually written in the buffer. 4082It can (and very often will) be smaller than the input value, since the 4083buffer is only filled until 16 bytes of clean values are found (which 4084are then not copied in the buffer). Since a CMMA migration block needs 4085the base address and the length, for a total of 16 bytes, we will send 4086back some clean data if there is some dirty data afterwards, as long as 4087the size of the clean data does not exceed the size of the header. This 4088allows to minimize the amount of data to be saved or transferred over 4089the network at the expense of more roundtrips to userspace. The next 4090invocation of the ioctl will skip over all the clean values, saving 4091potentially more than just the 16 bytes we found. 4092 4093If KVM_S390_CMMA_PEEK is set: 4094the existing storage attributes are read even when not in migration 4095mode, and no other action is performed; 4096 4097the output start_gfn will be equal to the input start_gfn, 4098 4099the output count will be equal to the input count, except if the end of 4100memory has been reached. 4101 4102In both cases: 4103the field "remaining" will indicate the total number of dirty CMMA values 4104still remaining, or 0 if KVM_S390_CMMA_PEEK is set and migration mode is 4105not enabled. 4106 4107mask is unused. 4108 4109values points to the userspace buffer where the result will be stored. 4110 4111This ioctl can fail with -ENOMEM if not enough memory can be allocated to 4112complete the task, with -ENXIO if CMMA is not enabled, with -EINVAL if 4113KVM_S390_CMMA_PEEK is not set but migration mode was not enabled, with 4114-EFAULT if the userspace address is invalid or if no page table is 4115present for the addresses (e.g. when using hugepages). 4116 41174.108 KVM_S390_SET_CMMA_BITS 4118---------------------------- 4119 4120:Capability: KVM_CAP_S390_CMMA_MIGRATION 4121:Architectures: s390 4122:Type: vm ioctl 4123:Parameters: struct kvm_s390_cmma_log (in) 4124:Returns: 0 on success, a negative value on error 4125 4126This ioctl is used to set the values of the CMMA bits on the s390 4127architecture. It is meant to be used during live migration to restore 4128the CMMA values, but there are no restrictions on its use. 4129The ioctl takes parameters via the kvm_s390_cmma_values struct. 4130Each CMMA value takes up one byte. 4131 4132:: 4133 4134 struct kvm_s390_cmma_log { 4135 __u64 start_gfn; 4136 __u32 count; 4137 __u32 flags; 4138 union { 4139 __u64 remaining; 4140 __u64 mask; 4141 }; 4142 __u64 values; 4143 }; 4144 4145start_gfn indicates the starting guest frame number, 4146 4147count indicates how many values are to be considered in the buffer, 4148 4149flags is not used and must be 0. 4150 4151mask indicates which PGSTE bits are to be considered. 4152 4153remaining is not used. 4154 4155values points to the buffer in userspace where to store the values. 4156 4157This ioctl can fail with -ENOMEM if not enough memory can be allocated to 4158complete the task, with -ENXIO if CMMA is not enabled, with -EINVAL if 4159the count field is too large (e.g. more than KVM_S390_CMMA_SIZE_MAX) or 4160if the flags field was not 0, with -EFAULT if the userspace address is 4161invalid, if invalid pages are written to (e.g. after the end of memory) 4162or if no page table is present for the addresses (e.g. when using 4163hugepages). 4164 41654.109 KVM_PPC_GET_CPU_CHAR 4166-------------------------- 4167 4168:Capability: KVM_CAP_PPC_GET_CPU_CHAR 4169:Architectures: powerpc 4170:Type: vm ioctl 4171:Parameters: struct kvm_ppc_cpu_char (out) 4172:Returns: 0 on successful completion, 4173 -EFAULT if struct kvm_ppc_cpu_char cannot be written 4174 4175This ioctl gives userspace information about certain characteristics 4176of the CPU relating to speculative execution of instructions and 4177possible information leakage resulting from speculative execution (see 4178CVE-2017-5715, CVE-2017-5753 and CVE-2017-5754). The information is 4179returned in struct kvm_ppc_cpu_char, which looks like this:: 4180 4181 struct kvm_ppc_cpu_char { 4182 __u64 character; /* characteristics of the CPU */ 4183 __u64 behaviour; /* recommended software behaviour */ 4184 __u64 character_mask; /* valid bits in character */ 4185 __u64 behaviour_mask; /* valid bits in behaviour */ 4186 }; 4187 4188For extensibility, the character_mask and behaviour_mask fields 4189indicate which bits of character and behaviour have been filled in by 4190the kernel. If the set of defined bits is extended in future then 4191userspace will be able to tell whether it is running on a kernel that 4192knows about the new bits. 4193 4194The character field describes attributes of the CPU which can help 4195with preventing inadvertent information disclosure - specifically, 4196whether there is an instruction to flash-invalidate the L1 data cache 4197(ori 30,30,0 or mtspr SPRN_TRIG2,rN), whether the L1 data cache is set 4198to a mode where entries can only be used by the thread that created 4199them, whether the bcctr[l] instruction prevents speculation, and 4200whether a speculation barrier instruction (ori 31,31,0) is provided. 4201 4202The behaviour field describes actions that software should take to 4203prevent inadvertent information disclosure, and thus describes which 4204vulnerabilities the hardware is subject to; specifically whether the 4205L1 data cache should be flushed when returning to user mode from the 4206kernel, and whether a speculation barrier should be placed between an 4207array bounds check and the array access. 4208 4209These fields use the same bit definitions as the new 4210H_GET_CPU_CHARACTERISTICS hypercall. 4211 42124.110 KVM_MEMORY_ENCRYPT_OP 4213--------------------------- 4214 4215:Capability: basic 4216:Architectures: x86 4217:Type: system 4218:Parameters: an opaque platform specific structure (in/out) 4219:Returns: 0 on success; -1 on error 4220 4221If the platform supports creating encrypted VMs then this ioctl can be used 4222for issuing platform-specific memory encryption commands to manage those 4223encrypted VMs. 4224 4225Currently, this ioctl is used for issuing Secure Encrypted Virtualization 4226(SEV) commands on AMD Processors. The SEV commands are defined in 4227Documentation/virt/kvm/amd-memory-encryption.rst. 4228 42294.111 KVM_MEMORY_ENCRYPT_REG_REGION 4230----------------------------------- 4231 4232:Capability: basic 4233:Architectures: x86 4234:Type: system 4235:Parameters: struct kvm_enc_region (in) 4236:Returns: 0 on success; -1 on error 4237 4238This ioctl can be used to register a guest memory region which may 4239contain encrypted data (e.g. guest RAM, SMRAM etc). 4240 4241It is used in the SEV-enabled guest. When encryption is enabled, a guest 4242memory region may contain encrypted data. The SEV memory encryption 4243engine uses a tweak such that two identical plaintext pages, each at 4244different locations will have differing ciphertexts. So swapping or 4245moving ciphertext of those pages will not result in plaintext being 4246swapped. So relocating (or migrating) physical backing pages for the SEV 4247guest will require some additional steps. 4248 4249Note: The current SEV key management spec does not provide commands to 4250swap or migrate (move) ciphertext pages. Hence, for now we pin the guest 4251memory region registered with the ioctl. 4252 42534.112 KVM_MEMORY_ENCRYPT_UNREG_REGION 4254------------------------------------- 4255 4256:Capability: basic 4257:Architectures: x86 4258:Type: system 4259:Parameters: struct kvm_enc_region (in) 4260:Returns: 0 on success; -1 on error 4261 4262This ioctl can be used to unregister the guest memory region registered 4263with KVM_MEMORY_ENCRYPT_REG_REGION ioctl above. 4264 42654.113 KVM_HYPERV_EVENTFD 4266------------------------ 4267 4268:Capability: KVM_CAP_HYPERV_EVENTFD 4269:Architectures: x86 4270:Type: vm ioctl 4271:Parameters: struct kvm_hyperv_eventfd (in) 4272 4273This ioctl (un)registers an eventfd to receive notifications from the guest on 4274the specified Hyper-V connection id through the SIGNAL_EVENT hypercall, without 4275causing a user exit. SIGNAL_EVENT hypercall with non-zero event flag number 4276(bits 24-31) still triggers a KVM_EXIT_HYPERV_HCALL user exit. 4277 4278:: 4279 4280 struct kvm_hyperv_eventfd { 4281 __u32 conn_id; 4282 __s32 fd; 4283 __u32 flags; 4284 __u32 padding[3]; 4285 }; 4286 4287The conn_id field should fit within 24 bits:: 4288 4289 #define KVM_HYPERV_CONN_ID_MASK 0x00ffffff 4290 4291The acceptable values for the flags field are:: 4292 4293 #define KVM_HYPERV_EVENTFD_DEASSIGN (1 << 0) 4294 4295:Returns: 0 on success, 4296 -EINVAL if conn_id or flags is outside the allowed range, 4297 -ENOENT on deassign if the conn_id isn't registered, 4298 -EEXIST on assign if the conn_id is already registered 4299 43004.114 KVM_GET_NESTED_STATE 4301-------------------------- 4302 4303:Capability: KVM_CAP_NESTED_STATE 4304:Architectures: x86 4305:Type: vcpu ioctl 4306:Parameters: struct kvm_nested_state (in/out) 4307:Returns: 0 on success, -1 on error 4308 4309Errors: 4310 4311 ===== ============================================================= 4312 E2BIG the total state size exceeds the value of 'size' specified by 4313 the user; the size required will be written into size. 4314 ===== ============================================================= 4315 4316:: 4317 4318 struct kvm_nested_state { 4319 __u16 flags; 4320 __u16 format; 4321 __u32 size; 4322 4323 union { 4324 struct kvm_vmx_nested_state_hdr vmx; 4325 struct kvm_svm_nested_state_hdr svm; 4326 4327 /* Pad the header to 128 bytes. */ 4328 __u8 pad[120]; 4329 } hdr; 4330 4331 union { 4332 struct kvm_vmx_nested_state_data vmx[0]; 4333 struct kvm_svm_nested_state_data svm[0]; 4334 } data; 4335 }; 4336 4337 #define KVM_STATE_NESTED_GUEST_MODE 0x00000001 4338 #define KVM_STATE_NESTED_RUN_PENDING 0x00000002 4339 #define KVM_STATE_NESTED_EVMCS 0x00000004 4340 4341 #define KVM_STATE_NESTED_FORMAT_VMX 0 4342 #define KVM_STATE_NESTED_FORMAT_SVM 1 4343 4344 #define KVM_STATE_NESTED_VMX_VMCS_SIZE 0x1000 4345 4346 #define KVM_STATE_NESTED_VMX_SMM_GUEST_MODE 0x00000001 4347 #define KVM_STATE_NESTED_VMX_SMM_VMXON 0x00000002 4348 4349 struct kvm_vmx_nested_state_hdr { 4350 __u64 vmxon_pa; 4351 __u64 vmcs12_pa; 4352 4353 struct { 4354 __u16 flags; 4355 } smm; 4356 }; 4357 4358 struct kvm_vmx_nested_state_data { 4359 __u8 vmcs12[KVM_STATE_NESTED_VMX_VMCS_SIZE]; 4360 __u8 shadow_vmcs12[KVM_STATE_NESTED_VMX_VMCS_SIZE]; 4361 }; 4362 4363This ioctl copies the vcpu's nested virtualization state from the kernel to 4364userspace. 4365 4366The maximum size of the state can be retrieved by passing KVM_CAP_NESTED_STATE 4367to the KVM_CHECK_EXTENSION ioctl(). 4368 43694.115 KVM_SET_NESTED_STATE 4370-------------------------- 4371 4372:Capability: KVM_CAP_NESTED_STATE 4373:Architectures: x86 4374:Type: vcpu ioctl 4375:Parameters: struct kvm_nested_state (in) 4376:Returns: 0 on success, -1 on error 4377 4378This copies the vcpu's kvm_nested_state struct from userspace to the kernel. 4379For the definition of struct kvm_nested_state, see KVM_GET_NESTED_STATE. 4380 43814.116 KVM_(UN)REGISTER_COALESCED_MMIO 4382------------------------------------- 4383 4384:Capability: KVM_CAP_COALESCED_MMIO (for coalesced mmio) 4385 KVM_CAP_COALESCED_PIO (for coalesced pio) 4386:Architectures: all 4387:Type: vm ioctl 4388:Parameters: struct kvm_coalesced_mmio_zone 4389:Returns: 0 on success, < 0 on error 4390 4391Coalesced I/O is a performance optimization that defers hardware 4392register write emulation so that userspace exits are avoided. It is 4393typically used to reduce the overhead of emulating frequently accessed 4394hardware registers. 4395 4396When a hardware register is configured for coalesced I/O, write accesses 4397do not exit to userspace and their value is recorded in a ring buffer 4398that is shared between kernel and userspace. 4399 4400Coalesced I/O is used if one or more write accesses to a hardware 4401register can be deferred until a read or a write to another hardware 4402register on the same device. This last access will cause a vmexit and 4403userspace will process accesses from the ring buffer before emulating 4404it. That will avoid exiting to userspace on repeated writes. 4405 4406Coalesced pio is based on coalesced mmio. There is little difference 4407between coalesced mmio and pio except that coalesced pio records accesses 4408to I/O ports. 4409 44104.117 KVM_CLEAR_DIRTY_LOG (vm ioctl) 4411------------------------------------ 4412 4413:Capability: KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 4414:Architectures: x86, arm, arm64, mips 4415:Type: vm ioctl 4416:Parameters: struct kvm_dirty_log (in) 4417:Returns: 0 on success, -1 on error 4418 4419:: 4420 4421 /* for KVM_CLEAR_DIRTY_LOG */ 4422 struct kvm_clear_dirty_log { 4423 __u32 slot; 4424 __u32 num_pages; 4425 __u64 first_page; 4426 union { 4427 void __user *dirty_bitmap; /* one bit per page */ 4428 __u64 padding; 4429 }; 4430 }; 4431 4432The ioctl clears the dirty status of pages in a memory slot, according to 4433the bitmap that is passed in struct kvm_clear_dirty_log's dirty_bitmap 4434field. Bit 0 of the bitmap corresponds to page "first_page" in the 4435memory slot, and num_pages is the size in bits of the input bitmap. 4436first_page must be a multiple of 64; num_pages must also be a multiple of 443764 unless first_page + num_pages is the size of the memory slot. For each 4438bit that is set in the input bitmap, the corresponding page is marked "clean" 4439in KVM's dirty bitmap, and dirty tracking is re-enabled for that page 4440(for example via write-protection, or by clearing the dirty bit in 4441a page table entry). 4442 4443If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 specifies 4444the address space for which you want to return the dirty bitmap. 4445They must be less than the value that KVM_CHECK_EXTENSION returns for 4446the KVM_CAP_MULTI_ADDRESS_SPACE capability. 4447 4448This ioctl is mostly useful when KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 4449is enabled; for more information, see the description of the capability. 4450However, it can always be used as long as KVM_CHECK_EXTENSION confirms 4451that KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 is present. 4452 44534.118 KVM_GET_SUPPORTED_HV_CPUID 4454-------------------------------- 4455 4456:Capability: KVM_CAP_HYPERV_CPUID 4457:Architectures: x86 4458:Type: vcpu ioctl 4459:Parameters: struct kvm_cpuid2 (in/out) 4460:Returns: 0 on success, -1 on error 4461 4462:: 4463 4464 struct kvm_cpuid2 { 4465 __u32 nent; 4466 __u32 padding; 4467 struct kvm_cpuid_entry2 entries[0]; 4468 }; 4469 4470 struct kvm_cpuid_entry2 { 4471 __u32 function; 4472 __u32 index; 4473 __u32 flags; 4474 __u32 eax; 4475 __u32 ebx; 4476 __u32 ecx; 4477 __u32 edx; 4478 __u32 padding[3]; 4479 }; 4480 4481This ioctl returns x86 cpuid features leaves related to Hyper-V emulation in 4482KVM. Userspace can use the information returned by this ioctl to construct 4483cpuid information presented to guests consuming Hyper-V enlightenments (e.g. 4484Windows or Hyper-V guests). 4485 4486CPUID feature leaves returned by this ioctl are defined by Hyper-V Top Level 4487Functional Specification (TLFS). These leaves can't be obtained with 4488KVM_GET_SUPPORTED_CPUID ioctl because some of them intersect with KVM feature 4489leaves (0x40000000, 0x40000001). 4490 4491Currently, the following list of CPUID leaves are returned: 4492 - HYPERV_CPUID_VENDOR_AND_MAX_FUNCTIONS 4493 - HYPERV_CPUID_INTERFACE 4494 - HYPERV_CPUID_VERSION 4495 - HYPERV_CPUID_FEATURES 4496 - HYPERV_CPUID_ENLIGHTMENT_INFO 4497 - HYPERV_CPUID_IMPLEMENT_LIMITS 4498 - HYPERV_CPUID_NESTED_FEATURES 4499 4500HYPERV_CPUID_NESTED_FEATURES leaf is only exposed when Enlightened VMCS was 4501enabled on the corresponding vCPU (KVM_CAP_HYPERV_ENLIGHTENED_VMCS). 4502 4503Userspace invokes KVM_GET_SUPPORTED_CPUID by passing a kvm_cpuid2 structure 4504with the 'nent' field indicating the number of entries in the variable-size 4505array 'entries'. If the number of entries is too low to describe all Hyper-V 4506feature leaves, an error (E2BIG) is returned. If the number is more or equal 4507to the number of Hyper-V feature leaves, the 'nent' field is adjusted to the 4508number of valid entries in the 'entries' array, which is then filled. 4509 4510'index' and 'flags' fields in 'struct kvm_cpuid_entry2' are currently reserved, 4511userspace should not expect to get any particular value there. 4512 45134.119 KVM_ARM_VCPU_FINALIZE 4514--------------------------- 4515 4516:Architectures: arm, arm64 4517:Type: vcpu ioctl 4518:Parameters: int feature (in) 4519:Returns: 0 on success, -1 on error 4520 4521Errors: 4522 4523 ====== ============================================================== 4524 EPERM feature not enabled, needs configuration, or already finalized 4525 EINVAL feature unknown or not present 4526 ====== ============================================================== 4527 4528Recognised values for feature: 4529 4530 ===== =========================================== 4531 arm64 KVM_ARM_VCPU_SVE (requires KVM_CAP_ARM_SVE) 4532 ===== =========================================== 4533 4534Finalizes the configuration of the specified vcpu feature. 4535 4536The vcpu must already have been initialised, enabling the affected feature, by 4537means of a successful KVM_ARM_VCPU_INIT call with the appropriate flag set in 4538features[]. 4539 4540For affected vcpu features, this is a mandatory step that must be performed 4541before the vcpu is fully usable. 4542 4543Between KVM_ARM_VCPU_INIT and KVM_ARM_VCPU_FINALIZE, the feature may be 4544configured by use of ioctls such as KVM_SET_ONE_REG. The exact configuration 4545that should be performaned and how to do it are feature-dependent. 4546 4547Other calls that depend on a particular feature being finalized, such as 4548KVM_RUN, KVM_GET_REG_LIST, KVM_GET_ONE_REG and KVM_SET_ONE_REG, will fail with 4549-EPERM unless the feature has already been finalized by means of a 4550KVM_ARM_VCPU_FINALIZE call. 4551 4552See KVM_ARM_VCPU_INIT for details of vcpu features that require finalization 4553using this ioctl. 4554 45554.120 KVM_SET_PMU_EVENT_FILTER 4556------------------------------ 4557 4558:Capability: KVM_CAP_PMU_EVENT_FILTER 4559:Architectures: x86 4560:Type: vm ioctl 4561:Parameters: struct kvm_pmu_event_filter (in) 4562:Returns: 0 on success, -1 on error 4563 4564:: 4565 4566 struct kvm_pmu_event_filter { 4567 __u32 action; 4568 __u32 nevents; 4569 __u32 fixed_counter_bitmap; 4570 __u32 flags; 4571 __u32 pad[4]; 4572 __u64 events[0]; 4573 }; 4574 4575This ioctl restricts the set of PMU events that the guest can program. 4576The argument holds a list of events which will be allowed or denied. 4577The eventsel+umask of each event the guest attempts to program is compared 4578against the events field to determine whether the guest should have access. 4579The events field only controls general purpose counters; fixed purpose 4580counters are controlled by the fixed_counter_bitmap. 4581 4582No flags are defined yet, the field must be zero. 4583 4584Valid values for 'action':: 4585 4586 #define KVM_PMU_EVENT_ALLOW 0 4587 #define KVM_PMU_EVENT_DENY 1 4588 45894.121 KVM_PPC_SVM_OFF 4590--------------------- 4591 4592:Capability: basic 4593:Architectures: powerpc 4594:Type: vm ioctl 4595:Parameters: none 4596:Returns: 0 on successful completion, 4597 4598Errors: 4599 4600 ====== ================================================================ 4601 EINVAL if ultravisor failed to terminate the secure guest 4602 ENOMEM if hypervisor failed to allocate new radix page tables for guest 4603 ====== ================================================================ 4604 4605This ioctl is used to turn off the secure mode of the guest or transition 4606the guest from secure mode to normal mode. This is invoked when the guest 4607is reset. This has no effect if called for a normal guest. 4608 4609This ioctl issues an ultravisor call to terminate the secure guest, 4610unpins the VPA pages and releases all the device pages that are used to 4611track the secure pages by hypervisor. 4612 46134.122 KVM_S390_NORMAL_RESET 4614 4615Capability: KVM_CAP_S390_VCPU_RESETS 4616Architectures: s390 4617Type: vcpu ioctl 4618Parameters: none 4619Returns: 0 4620 4621This ioctl resets VCPU registers and control structures according to 4622the cpu reset definition in the POP (Principles Of Operation). 4623 46244.123 KVM_S390_INITIAL_RESET 4625 4626Capability: none 4627Architectures: s390 4628Type: vcpu ioctl 4629Parameters: none 4630Returns: 0 4631 4632This ioctl resets VCPU registers and control structures according to 4633the initial cpu reset definition in the POP. However, the cpu is not 4634put into ESA mode. This reset is a superset of the normal reset. 4635 46364.124 KVM_S390_CLEAR_RESET 4637 4638Capability: KVM_CAP_S390_VCPU_RESETS 4639Architectures: s390 4640Type: vcpu ioctl 4641Parameters: none 4642Returns: 0 4643 4644This ioctl resets VCPU registers and control structures according to 4645the clear cpu reset definition in the POP. However, the cpu is not put 4646into ESA mode. This reset is a superset of the initial reset. 4647 4648 46495. The kvm_run structure 4650======================== 4651 4652Application code obtains a pointer to the kvm_run structure by 4653mmap()ing a vcpu fd. From that point, application code can control 4654execution by changing fields in kvm_run prior to calling the KVM_RUN 4655ioctl, and obtain information about the reason KVM_RUN returned by 4656looking up structure members. 4657 4658:: 4659 4660 struct kvm_run { 4661 /* in */ 4662 __u8 request_interrupt_window; 4663 4664Request that KVM_RUN return when it becomes possible to inject external 4665interrupts into the guest. Useful in conjunction with KVM_INTERRUPT. 4666 4667:: 4668 4669 __u8 immediate_exit; 4670 4671This field is polled once when KVM_RUN starts; if non-zero, KVM_RUN 4672exits immediately, returning -EINTR. In the common scenario where a 4673signal is used to "kick" a VCPU out of KVM_RUN, this field can be used 4674to avoid usage of KVM_SET_SIGNAL_MASK, which has worse scalability. 4675Rather than blocking the signal outside KVM_RUN, userspace can set up 4676a signal handler that sets run->immediate_exit to a non-zero value. 4677 4678This field is ignored if KVM_CAP_IMMEDIATE_EXIT is not available. 4679 4680:: 4681 4682 __u8 padding1[6]; 4683 4684 /* out */ 4685 __u32 exit_reason; 4686 4687When KVM_RUN has returned successfully (return value 0), this informs 4688application code why KVM_RUN has returned. Allowable values for this 4689field are detailed below. 4690 4691:: 4692 4693 __u8 ready_for_interrupt_injection; 4694 4695If request_interrupt_window has been specified, this field indicates 4696an interrupt can be injected now with KVM_INTERRUPT. 4697 4698:: 4699 4700 __u8 if_flag; 4701 4702The value of the current interrupt flag. Only valid if in-kernel 4703local APIC is not used. 4704 4705:: 4706 4707 __u16 flags; 4708 4709More architecture-specific flags detailing state of the VCPU that may 4710affect the device's behavior. The only currently defined flag is 4711KVM_RUN_X86_SMM, which is valid on x86 machines and is set if the 4712VCPU is in system management mode. 4713 4714:: 4715 4716 /* in (pre_kvm_run), out (post_kvm_run) */ 4717 __u64 cr8; 4718 4719The value of the cr8 register. Only valid if in-kernel local APIC is 4720not used. Both input and output. 4721 4722:: 4723 4724 __u64 apic_base; 4725 4726The value of the APIC BASE msr. Only valid if in-kernel local 4727APIC is not used. Both input and output. 4728 4729:: 4730 4731 union { 4732 /* KVM_EXIT_UNKNOWN */ 4733 struct { 4734 __u64 hardware_exit_reason; 4735 } hw; 4736 4737If exit_reason is KVM_EXIT_UNKNOWN, the vcpu has exited due to unknown 4738reasons. Further architecture-specific information is available in 4739hardware_exit_reason. 4740 4741:: 4742 4743 /* KVM_EXIT_FAIL_ENTRY */ 4744 struct { 4745 __u64 hardware_entry_failure_reason; 4746 } fail_entry; 4747 4748If exit_reason is KVM_EXIT_FAIL_ENTRY, the vcpu could not be run due 4749to unknown reasons. Further architecture-specific information is 4750available in hardware_entry_failure_reason. 4751 4752:: 4753 4754 /* KVM_EXIT_EXCEPTION */ 4755 struct { 4756 __u32 exception; 4757 __u32 error_code; 4758 } ex; 4759 4760Unused. 4761 4762:: 4763 4764 /* KVM_EXIT_IO */ 4765 struct { 4766 #define KVM_EXIT_IO_IN 0 4767 #define KVM_EXIT_IO_OUT 1 4768 __u8 direction; 4769 __u8 size; /* bytes */ 4770 __u16 port; 4771 __u32 count; 4772 __u64 data_offset; /* relative to kvm_run start */ 4773 } io; 4774 4775If exit_reason is KVM_EXIT_IO, then the vcpu has 4776executed a port I/O instruction which could not be satisfied by kvm. 4777data_offset describes where the data is located (KVM_EXIT_IO_OUT) or 4778where kvm expects application code to place the data for the next 4779KVM_RUN invocation (KVM_EXIT_IO_IN). Data format is a packed array. 4780 4781:: 4782 4783 /* KVM_EXIT_DEBUG */ 4784 struct { 4785 struct kvm_debug_exit_arch arch; 4786 } debug; 4787 4788If the exit_reason is KVM_EXIT_DEBUG, then a vcpu is processing a debug event 4789for which architecture specific information is returned. 4790 4791:: 4792 4793 /* KVM_EXIT_MMIO */ 4794 struct { 4795 __u64 phys_addr; 4796 __u8 data[8]; 4797 __u32 len; 4798 __u8 is_write; 4799 } mmio; 4800 4801If exit_reason is KVM_EXIT_MMIO, then the vcpu has 4802executed a memory-mapped I/O instruction which could not be satisfied 4803by kvm. The 'data' member contains the written data if 'is_write' is 4804true, and should be filled by application code otherwise. 4805 4806The 'data' member contains, in its first 'len' bytes, the value as it would 4807appear if the VCPU performed a load or store of the appropriate width directly 4808to the byte array. 4809 4810.. note:: 4811 4812 For KVM_EXIT_IO, KVM_EXIT_MMIO, KVM_EXIT_OSI, KVM_EXIT_PAPR and 4813 KVM_EXIT_EPR the corresponding 4814 4815operations are complete (and guest state is consistent) only after userspace 4816has re-entered the kernel with KVM_RUN. The kernel side will first finish 4817incomplete operations and then check for pending signals. Userspace 4818can re-enter the guest with an unmasked signal pending to complete 4819pending operations. 4820 4821:: 4822 4823 /* KVM_EXIT_HYPERCALL */ 4824 struct { 4825 __u64 nr; 4826 __u64 args[6]; 4827 __u64 ret; 4828 __u32 longmode; 4829 __u32 pad; 4830 } hypercall; 4831 4832Unused. This was once used for 'hypercall to userspace'. To implement 4833such functionality, use KVM_EXIT_IO (x86) or KVM_EXIT_MMIO (all except s390). 4834 4835.. note:: KVM_EXIT_IO is significantly faster than KVM_EXIT_MMIO. 4836 4837:: 4838 4839 /* KVM_EXIT_TPR_ACCESS */ 4840 struct { 4841 __u64 rip; 4842 __u32 is_write; 4843 __u32 pad; 4844 } tpr_access; 4845 4846To be documented (KVM_TPR_ACCESS_REPORTING). 4847 4848:: 4849 4850 /* KVM_EXIT_S390_SIEIC */ 4851 struct { 4852 __u8 icptcode; 4853 __u64 mask; /* psw upper half */ 4854 __u64 addr; /* psw lower half */ 4855 __u16 ipa; 4856 __u32 ipb; 4857 } s390_sieic; 4858 4859s390 specific. 4860 4861:: 4862 4863 /* KVM_EXIT_S390_RESET */ 4864 #define KVM_S390_RESET_POR 1 4865 #define KVM_S390_RESET_CLEAR 2 4866 #define KVM_S390_RESET_SUBSYSTEM 4 4867 #define KVM_S390_RESET_CPU_INIT 8 4868 #define KVM_S390_RESET_IPL 16 4869 __u64 s390_reset_flags; 4870 4871s390 specific. 4872 4873:: 4874 4875 /* KVM_EXIT_S390_UCONTROL */ 4876 struct { 4877 __u64 trans_exc_code; 4878 __u32 pgm_code; 4879 } s390_ucontrol; 4880 4881s390 specific. A page fault has occurred for a user controlled virtual 4882machine (KVM_VM_S390_UNCONTROL) on it's host page table that cannot be 4883resolved by the kernel. 4884The program code and the translation exception code that were placed 4885in the cpu's lowcore are presented here as defined by the z Architecture 4886Principles of Operation Book in the Chapter for Dynamic Address Translation 4887(DAT) 4888 4889:: 4890 4891 /* KVM_EXIT_DCR */ 4892 struct { 4893 __u32 dcrn; 4894 __u32 data; 4895 __u8 is_write; 4896 } dcr; 4897 4898Deprecated - was used for 440 KVM. 4899 4900:: 4901 4902 /* KVM_EXIT_OSI */ 4903 struct { 4904 __u64 gprs[32]; 4905 } osi; 4906 4907MOL uses a special hypercall interface it calls 'OSI'. To enable it, we catch 4908hypercalls and exit with this exit struct that contains all the guest gprs. 4909 4910If exit_reason is KVM_EXIT_OSI, then the vcpu has triggered such a hypercall. 4911Userspace can now handle the hypercall and when it's done modify the gprs as 4912necessary. Upon guest entry all guest GPRs will then be replaced by the values 4913in this struct. 4914 4915:: 4916 4917 /* KVM_EXIT_PAPR_HCALL */ 4918 struct { 4919 __u64 nr; 4920 __u64 ret; 4921 __u64 args[9]; 4922 } papr_hcall; 4923 4924This is used on 64-bit PowerPC when emulating a pSeries partition, 4925e.g. with the 'pseries' machine type in qemu. It occurs when the 4926guest does a hypercall using the 'sc 1' instruction. The 'nr' field 4927contains the hypercall number (from the guest R3), and 'args' contains 4928the arguments (from the guest R4 - R12). Userspace should put the 4929return code in 'ret' and any extra returned values in args[]. 4930The possible hypercalls are defined in the Power Architecture Platform 4931Requirements (PAPR) document available from www.power.org (free 4932developer registration required to access it). 4933 4934:: 4935 4936 /* KVM_EXIT_S390_TSCH */ 4937 struct { 4938 __u16 subchannel_id; 4939 __u16 subchannel_nr; 4940 __u32 io_int_parm; 4941 __u32 io_int_word; 4942 __u32 ipb; 4943 __u8 dequeued; 4944 } s390_tsch; 4945 4946s390 specific. This exit occurs when KVM_CAP_S390_CSS_SUPPORT has been enabled 4947and TEST SUBCHANNEL was intercepted. If dequeued is set, a pending I/O 4948interrupt for the target subchannel has been dequeued and subchannel_id, 4949subchannel_nr, io_int_parm and io_int_word contain the parameters for that 4950interrupt. ipb is needed for instruction parameter decoding. 4951 4952:: 4953 4954 /* KVM_EXIT_EPR */ 4955 struct { 4956 __u32 epr; 4957 } epr; 4958 4959On FSL BookE PowerPC chips, the interrupt controller has a fast patch 4960interrupt acknowledge path to the core. When the core successfully 4961delivers an interrupt, it automatically populates the EPR register with 4962the interrupt vector number and acknowledges the interrupt inside 4963the interrupt controller. 4964 4965In case the interrupt controller lives in user space, we need to do 4966the interrupt acknowledge cycle through it to fetch the next to be 4967delivered interrupt vector using this exit. 4968 4969It gets triggered whenever both KVM_CAP_PPC_EPR are enabled and an 4970external interrupt has just been delivered into the guest. User space 4971should put the acknowledged interrupt vector into the 'epr' field. 4972 4973:: 4974 4975 /* KVM_EXIT_SYSTEM_EVENT */ 4976 struct { 4977 #define KVM_SYSTEM_EVENT_SHUTDOWN 1 4978 #define KVM_SYSTEM_EVENT_RESET 2 4979 #define KVM_SYSTEM_EVENT_CRASH 3 4980 __u32 type; 4981 __u64 flags; 4982 } system_event; 4983 4984If exit_reason is KVM_EXIT_SYSTEM_EVENT then the vcpu has triggered 4985a system-level event using some architecture specific mechanism (hypercall 4986or some special instruction). In case of ARM/ARM64, this is triggered using 4987HVC instruction based PSCI call from the vcpu. The 'type' field describes 4988the system-level event type. The 'flags' field describes architecture 4989specific flags for the system-level event. 4990 4991Valid values for 'type' are: 4992 4993 - KVM_SYSTEM_EVENT_SHUTDOWN -- the guest has requested a shutdown of the 4994 VM. Userspace is not obliged to honour this, and if it does honour 4995 this does not need to destroy the VM synchronously (ie it may call 4996 KVM_RUN again before shutdown finally occurs). 4997 - KVM_SYSTEM_EVENT_RESET -- the guest has requested a reset of the VM. 4998 As with SHUTDOWN, userspace can choose to ignore the request, or 4999 to schedule the reset to occur in the future and may call KVM_RUN again. 5000 - KVM_SYSTEM_EVENT_CRASH -- the guest crash occurred and the guest 5001 has requested a crash condition maintenance. Userspace can choose 5002 to ignore the request, or to gather VM memory core dump and/or 5003 reset/shutdown of the VM. 5004 5005:: 5006 5007 /* KVM_EXIT_IOAPIC_EOI */ 5008 struct { 5009 __u8 vector; 5010 } eoi; 5011 5012Indicates that the VCPU's in-kernel local APIC received an EOI for a 5013level-triggered IOAPIC interrupt. This exit only triggers when the 5014IOAPIC is implemented in userspace (i.e. KVM_CAP_SPLIT_IRQCHIP is enabled); 5015the userspace IOAPIC should process the EOI and retrigger the interrupt if 5016it is still asserted. Vector is the LAPIC interrupt vector for which the 5017EOI was received. 5018 5019:: 5020 5021 struct kvm_hyperv_exit { 5022 #define KVM_EXIT_HYPERV_SYNIC 1 5023 #define KVM_EXIT_HYPERV_HCALL 2 5024 __u32 type; 5025 union { 5026 struct { 5027 __u32 msr; 5028 __u64 control; 5029 __u64 evt_page; 5030 __u64 msg_page; 5031 } synic; 5032 struct { 5033 __u64 input; 5034 __u64 result; 5035 __u64 params[2]; 5036 } hcall; 5037 } u; 5038 }; 5039 /* KVM_EXIT_HYPERV */ 5040 struct kvm_hyperv_exit hyperv; 5041 5042Indicates that the VCPU exits into userspace to process some tasks 5043related to Hyper-V emulation. 5044 5045Valid values for 'type' are: 5046 5047 - KVM_EXIT_HYPERV_SYNIC -- synchronously notify user-space about 5048 5049Hyper-V SynIC state change. Notification is used to remap SynIC 5050event/message pages and to enable/disable SynIC messages/events processing 5051in userspace. 5052 5053:: 5054 5055 /* KVM_EXIT_ARM_NISV */ 5056 struct { 5057 __u64 esr_iss; 5058 __u64 fault_ipa; 5059 } arm_nisv; 5060 5061Used on arm and arm64 systems. If a guest accesses memory not in a memslot, 5062KVM will typically return to userspace and ask it to do MMIO emulation on its 5063behalf. However, for certain classes of instructions, no instruction decode 5064(direction, length of memory access) is provided, and fetching and decoding 5065the instruction from the VM is overly complicated to live in the kernel. 5066 5067Historically, when this situation occurred, KVM would print a warning and kill 5068the VM. KVM assumed that if the guest accessed non-memslot memory, it was 5069trying to do I/O, which just couldn't be emulated, and the warning message was 5070phrased accordingly. However, what happened more often was that a guest bug 5071caused access outside the guest memory areas which should lead to a more 5072meaningful warning message and an external abort in the guest, if the access 5073did not fall within an I/O window. 5074 5075Userspace implementations can query for KVM_CAP_ARM_NISV_TO_USER, and enable 5076this capability at VM creation. Once this is done, these types of errors will 5077instead return to userspace with KVM_EXIT_ARM_NISV, with the valid bits from 5078the HSR (arm) and ESR_EL2 (arm64) in the esr_iss field, and the faulting IPA 5079in the fault_ipa field. Userspace can either fix up the access if it's 5080actually an I/O access by decoding the instruction from guest memory (if it's 5081very brave) and continue executing the guest, or it can decide to suspend, 5082dump, or restart the guest. 5083 5084Note that KVM does not skip the faulting instruction as it does for 5085KVM_EXIT_MMIO, but userspace has to emulate any change to the processing state 5086if it decides to decode and emulate the instruction. 5087 5088:: 5089 5090 /* Fix the size of the union. */ 5091 char padding[256]; 5092 }; 5093 5094 /* 5095 * shared registers between kvm and userspace. 5096 * kvm_valid_regs specifies the register classes set by the host 5097 * kvm_dirty_regs specified the register classes dirtied by userspace 5098 * struct kvm_sync_regs is architecture specific, as well as the 5099 * bits for kvm_valid_regs and kvm_dirty_regs 5100 */ 5101 __u64 kvm_valid_regs; 5102 __u64 kvm_dirty_regs; 5103 union { 5104 struct kvm_sync_regs regs; 5105 char padding[SYNC_REGS_SIZE_BYTES]; 5106 } s; 5107 5108If KVM_CAP_SYNC_REGS is defined, these fields allow userspace to access 5109certain guest registers without having to call SET/GET_*REGS. Thus we can 5110avoid some system call overhead if userspace has to handle the exit. 5111Userspace can query the validity of the structure by checking 5112kvm_valid_regs for specific bits. These bits are architecture specific 5113and usually define the validity of a groups of registers. (e.g. one bit 5114for general purpose registers) 5115 5116Please note that the kernel is allowed to use the kvm_run structure as the 5117primary storage for certain register types. Therefore, the kernel may use the 5118values in kvm_run even if the corresponding bit in kvm_dirty_regs is not set. 5119 5120:: 5121 5122 }; 5123 5124 5125 51266. Capabilities that can be enabled on vCPUs 5127============================================ 5128 5129There are certain capabilities that change the behavior of the virtual CPU or 5130the virtual machine when enabled. To enable them, please see section 4.37. 5131Below you can find a list of capabilities and what their effect on the vCPU or 5132the virtual machine is when enabling them. 5133 5134The following information is provided along with the description: 5135 5136 Architectures: 5137 which instruction set architectures provide this ioctl. 5138 x86 includes both i386 and x86_64. 5139 5140 Target: 5141 whether this is a per-vcpu or per-vm capability. 5142 5143 Parameters: 5144 what parameters are accepted by the capability. 5145 5146 Returns: 5147 the return value. General error numbers (EBADF, ENOMEM, EINVAL) 5148 are not detailed, but errors with specific meanings are. 5149 5150 51516.1 KVM_CAP_PPC_OSI 5152------------------- 5153 5154:Architectures: ppc 5155:Target: vcpu 5156:Parameters: none 5157:Returns: 0 on success; -1 on error 5158 5159This capability enables interception of OSI hypercalls that otherwise would 5160be treated as normal system calls to be injected into the guest. OSI hypercalls 5161were invented by Mac-on-Linux to have a standardized communication mechanism 5162between the guest and the host. 5163 5164When this capability is enabled, KVM_EXIT_OSI can occur. 5165 5166 51676.2 KVM_CAP_PPC_PAPR 5168-------------------- 5169 5170:Architectures: ppc 5171:Target: vcpu 5172:Parameters: none 5173:Returns: 0 on success; -1 on error 5174 5175This capability enables interception of PAPR hypercalls. PAPR hypercalls are 5176done using the hypercall instruction "sc 1". 5177 5178It also sets the guest privilege level to "supervisor" mode. Usually the guest 5179runs in "hypervisor" privilege mode with a few missing features. 5180 5181In addition to the above, it changes the semantics of SDR1. In this mode, the 5182HTAB address part of SDR1 contains an HVA instead of a GPA, as PAPR keeps the 5183HTAB invisible to the guest. 5184 5185When this capability is enabled, KVM_EXIT_PAPR_HCALL can occur. 5186 5187 51886.3 KVM_CAP_SW_TLB 5189------------------ 5190 5191:Architectures: ppc 5192:Target: vcpu 5193:Parameters: args[0] is the address of a struct kvm_config_tlb 5194:Returns: 0 on success; -1 on error 5195 5196:: 5197 5198 struct kvm_config_tlb { 5199 __u64 params; 5200 __u64 array; 5201 __u32 mmu_type; 5202 __u32 array_len; 5203 }; 5204 5205Configures the virtual CPU's TLB array, establishing a shared memory area 5206between userspace and KVM. The "params" and "array" fields are userspace 5207addresses of mmu-type-specific data structures. The "array_len" field is an 5208safety mechanism, and should be set to the size in bytes of the memory that 5209userspace has reserved for the array. It must be at least the size dictated 5210by "mmu_type" and "params". 5211 5212While KVM_RUN is active, the shared region is under control of KVM. Its 5213contents are undefined, and any modification by userspace results in 5214boundedly undefined behavior. 5215 5216On return from KVM_RUN, the shared region will reflect the current state of 5217the guest's TLB. If userspace makes any changes, it must call KVM_DIRTY_TLB 5218to tell KVM which entries have been changed, prior to calling KVM_RUN again 5219on this vcpu. 5220 5221For mmu types KVM_MMU_FSL_BOOKE_NOHV and KVM_MMU_FSL_BOOKE_HV: 5222 5223 - The "params" field is of type "struct kvm_book3e_206_tlb_params". 5224 - The "array" field points to an array of type "struct 5225 kvm_book3e_206_tlb_entry". 5226 - The array consists of all entries in the first TLB, followed by all 5227 entries in the second TLB. 5228 - Within a TLB, entries are ordered first by increasing set number. Within a 5229 set, entries are ordered by way (increasing ESEL). 5230 - The hash for determining set number in TLB0 is: (MAS2 >> 12) & (num_sets - 1) 5231 where "num_sets" is the tlb_sizes[] value divided by the tlb_ways[] value. 5232 - The tsize field of mas1 shall be set to 4K on TLB0, even though the 5233 hardware ignores this value for TLB0. 5234 52356.4 KVM_CAP_S390_CSS_SUPPORT 5236---------------------------- 5237 5238:Architectures: s390 5239:Target: vcpu 5240:Parameters: none 5241:Returns: 0 on success; -1 on error 5242 5243This capability enables support for handling of channel I/O instructions. 5244 5245TEST PENDING INTERRUPTION and the interrupt portion of TEST SUBCHANNEL are 5246handled in-kernel, while the other I/O instructions are passed to userspace. 5247 5248When this capability is enabled, KVM_EXIT_S390_TSCH will occur on TEST 5249SUBCHANNEL intercepts. 5250 5251Note that even though this capability is enabled per-vcpu, the complete 5252virtual machine is affected. 5253 52546.5 KVM_CAP_PPC_EPR 5255------------------- 5256 5257:Architectures: ppc 5258:Target: vcpu 5259:Parameters: args[0] defines whether the proxy facility is active 5260:Returns: 0 on success; -1 on error 5261 5262This capability enables or disables the delivery of interrupts through the 5263external proxy facility. 5264 5265When enabled (args[0] != 0), every time the guest gets an external interrupt 5266delivered, it automatically exits into user space with a KVM_EXIT_EPR exit 5267to receive the topmost interrupt vector. 5268 5269When disabled (args[0] == 0), behavior is as if this facility is unsupported. 5270 5271When this capability is enabled, KVM_EXIT_EPR can occur. 5272 52736.6 KVM_CAP_IRQ_MPIC 5274-------------------- 5275 5276:Architectures: ppc 5277:Parameters: args[0] is the MPIC device fd; 5278 args[1] is the MPIC CPU number for this vcpu 5279 5280This capability connects the vcpu to an in-kernel MPIC device. 5281 52826.7 KVM_CAP_IRQ_XICS 5283-------------------- 5284 5285:Architectures: ppc 5286:Target: vcpu 5287:Parameters: args[0] is the XICS device fd; 5288 args[1] is the XICS CPU number (server ID) for this vcpu 5289 5290This capability connects the vcpu to an in-kernel XICS device. 5291 52926.8 KVM_CAP_S390_IRQCHIP 5293------------------------ 5294 5295:Architectures: s390 5296:Target: vm 5297:Parameters: none 5298 5299This capability enables the in-kernel irqchip for s390. Please refer to 5300"4.24 KVM_CREATE_IRQCHIP" for details. 5301 53026.9 KVM_CAP_MIPS_FPU 5303-------------------- 5304 5305:Architectures: mips 5306:Target: vcpu 5307:Parameters: args[0] is reserved for future use (should be 0). 5308 5309This capability allows the use of the host Floating Point Unit by the guest. It 5310allows the Config1.FP bit to be set to enable the FPU in the guest. Once this is 5311done the ``KVM_REG_MIPS_FPR_*`` and ``KVM_REG_MIPS_FCR_*`` registers can be 5312accessed (depending on the current guest FPU register mode), and the Status.FR, 5313Config5.FRE bits are accessible via the KVM API and also from the guest, 5314depending on them being supported by the FPU. 5315 53166.10 KVM_CAP_MIPS_MSA 5317--------------------- 5318 5319:Architectures: mips 5320:Target: vcpu 5321:Parameters: args[0] is reserved for future use (should be 0). 5322 5323This capability allows the use of the MIPS SIMD Architecture (MSA) by the guest. 5324It allows the Config3.MSAP bit to be set to enable the use of MSA by the guest. 5325Once this is done the ``KVM_REG_MIPS_VEC_*`` and ``KVM_REG_MIPS_MSA_*`` 5326registers can be accessed, and the Config5.MSAEn bit is accessible via the 5327KVM API and also from the guest. 5328 53296.74 KVM_CAP_SYNC_REGS 5330---------------------- 5331 5332:Architectures: s390, x86 5333:Target: s390: always enabled, x86: vcpu 5334:Parameters: none 5335:Returns: x86: KVM_CHECK_EXTENSION returns a bit-array indicating which register 5336 sets are supported 5337 (bitfields defined in arch/x86/include/uapi/asm/kvm.h). 5338 5339As described above in the kvm_sync_regs struct info in section 5 (kvm_run): 5340KVM_CAP_SYNC_REGS "allow[s] userspace to access certain guest registers 5341without having to call SET/GET_*REGS". This reduces overhead by eliminating 5342repeated ioctl calls for setting and/or getting register values. This is 5343particularly important when userspace is making synchronous guest state 5344modifications, e.g. when emulating and/or intercepting instructions in 5345userspace. 5346 5347For s390 specifics, please refer to the source code. 5348 5349For x86: 5350 5351- the register sets to be copied out to kvm_run are selectable 5352 by userspace (rather that all sets being copied out for every exit). 5353- vcpu_events are available in addition to regs and sregs. 5354 5355For x86, the 'kvm_valid_regs' field of struct kvm_run is overloaded to 5356function as an input bit-array field set by userspace to indicate the 5357specific register sets to be copied out on the next exit. 5358 5359To indicate when userspace has modified values that should be copied into 5360the vCPU, the all architecture bitarray field, 'kvm_dirty_regs' must be set. 5361This is done using the same bitflags as for the 'kvm_valid_regs' field. 5362If the dirty bit is not set, then the register set values will not be copied 5363into the vCPU even if they've been modified. 5364 5365Unused bitfields in the bitarrays must be set to zero. 5366 5367:: 5368 5369 struct kvm_sync_regs { 5370 struct kvm_regs regs; 5371 struct kvm_sregs sregs; 5372 struct kvm_vcpu_events events; 5373 }; 5374 53756.75 KVM_CAP_PPC_IRQ_XIVE 5376------------------------- 5377 5378:Architectures: ppc 5379:Target: vcpu 5380:Parameters: args[0] is the XIVE device fd; 5381 args[1] is the XIVE CPU number (server ID) for this vcpu 5382 5383This capability connects the vcpu to an in-kernel XIVE device. 5384 53857. Capabilities that can be enabled on VMs 5386========================================== 5387 5388There are certain capabilities that change the behavior of the virtual 5389machine when enabled. To enable them, please see section 4.37. Below 5390you can find a list of capabilities and what their effect on the VM 5391is when enabling them. 5392 5393The following information is provided along with the description: 5394 5395 Architectures: 5396 which instruction set architectures provide this ioctl. 5397 x86 includes both i386 and x86_64. 5398 5399 Parameters: 5400 what parameters are accepted by the capability. 5401 5402 Returns: 5403 the return value. General error numbers (EBADF, ENOMEM, EINVAL) 5404 are not detailed, but errors with specific meanings are. 5405 5406 54077.1 KVM_CAP_PPC_ENABLE_HCALL 5408---------------------------- 5409 5410:Architectures: ppc 5411:Parameters: args[0] is the sPAPR hcall number; 5412 args[1] is 0 to disable, 1 to enable in-kernel handling 5413 5414This capability controls whether individual sPAPR hypercalls (hcalls) 5415get handled by the kernel or not. Enabling or disabling in-kernel 5416handling of an hcall is effective across the VM. On creation, an 5417initial set of hcalls are enabled for in-kernel handling, which 5418consists of those hcalls for which in-kernel handlers were implemented 5419before this capability was implemented. If disabled, the kernel will 5420not to attempt to handle the hcall, but will always exit to userspace 5421to handle it. Note that it may not make sense to enable some and 5422disable others of a group of related hcalls, but KVM does not prevent 5423userspace from doing that. 5424 5425If the hcall number specified is not one that has an in-kernel 5426implementation, the KVM_ENABLE_CAP ioctl will fail with an EINVAL 5427error. 5428 54297.2 KVM_CAP_S390_USER_SIGP 5430-------------------------- 5431 5432:Architectures: s390 5433:Parameters: none 5434 5435This capability controls which SIGP orders will be handled completely in user 5436space. With this capability enabled, all fast orders will be handled completely 5437in the kernel: 5438 5439- SENSE 5440- SENSE RUNNING 5441- EXTERNAL CALL 5442- EMERGENCY SIGNAL 5443- CONDITIONAL EMERGENCY SIGNAL 5444 5445All other orders will be handled completely in user space. 5446 5447Only privileged operation exceptions will be checked for in the kernel (or even 5448in the hardware prior to interception). If this capability is not enabled, the 5449old way of handling SIGP orders is used (partially in kernel and user space). 5450 54517.3 KVM_CAP_S390_VECTOR_REGISTERS 5452--------------------------------- 5453 5454:Architectures: s390 5455:Parameters: none 5456:Returns: 0 on success, negative value on error 5457 5458Allows use of the vector registers introduced with z13 processor, and 5459provides for the synchronization between host and user space. Will 5460return -EINVAL if the machine does not support vectors. 5461 54627.4 KVM_CAP_S390_USER_STSI 5463-------------------------- 5464 5465:Architectures: s390 5466:Parameters: none 5467 5468This capability allows post-handlers for the STSI instruction. After 5469initial handling in the kernel, KVM exits to user space with 5470KVM_EXIT_S390_STSI to allow user space to insert further data. 5471 5472Before exiting to userspace, kvm handlers should fill in s390_stsi field of 5473vcpu->run:: 5474 5475 struct { 5476 __u64 addr; 5477 __u8 ar; 5478 __u8 reserved; 5479 __u8 fc; 5480 __u8 sel1; 5481 __u16 sel2; 5482 } s390_stsi; 5483 5484 @addr - guest address of STSI SYSIB 5485 @fc - function code 5486 @sel1 - selector 1 5487 @sel2 - selector 2 5488 @ar - access register number 5489 5490KVM handlers should exit to userspace with rc = -EREMOTE. 5491 54927.5 KVM_CAP_SPLIT_IRQCHIP 5493------------------------- 5494 5495:Architectures: x86 5496:Parameters: args[0] - number of routes reserved for userspace IOAPICs 5497:Returns: 0 on success, -1 on error 5498 5499Create a local apic for each processor in the kernel. This can be used 5500instead of KVM_CREATE_IRQCHIP if the userspace VMM wishes to emulate the 5501IOAPIC and PIC (and also the PIT, even though this has to be enabled 5502separately). 5503 5504This capability also enables in kernel routing of interrupt requests; 5505when KVM_CAP_SPLIT_IRQCHIP only routes of KVM_IRQ_ROUTING_MSI type are 5506used in the IRQ routing table. The first args[0] MSI routes are reserved 5507for the IOAPIC pins. Whenever the LAPIC receives an EOI for these routes, 5508a KVM_EXIT_IOAPIC_EOI vmexit will be reported to userspace. 5509 5510Fails if VCPU has already been created, or if the irqchip is already in the 5511kernel (i.e. KVM_CREATE_IRQCHIP has already been called). 5512 55137.6 KVM_CAP_S390_RI 5514------------------- 5515 5516:Architectures: s390 5517:Parameters: none 5518 5519Allows use of runtime-instrumentation introduced with zEC12 processor. 5520Will return -EINVAL if the machine does not support runtime-instrumentation. 5521Will return -EBUSY if a VCPU has already been created. 5522 55237.7 KVM_CAP_X2APIC_API 5524---------------------- 5525 5526:Architectures: x86 5527:Parameters: args[0] - features that should be enabled 5528:Returns: 0 on success, -EINVAL when args[0] contains invalid features 5529 5530Valid feature flags in args[0] are:: 5531 5532 #define KVM_X2APIC_API_USE_32BIT_IDS (1ULL << 0) 5533 #define KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK (1ULL << 1) 5534 5535Enabling KVM_X2APIC_API_USE_32BIT_IDS changes the behavior of 5536KVM_SET_GSI_ROUTING, KVM_SIGNAL_MSI, KVM_SET_LAPIC, and KVM_GET_LAPIC, 5537allowing the use of 32-bit APIC IDs. See KVM_CAP_X2APIC_API in their 5538respective sections. 5539 5540KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK must be enabled for x2APIC to work 5541in logical mode or with more than 255 VCPUs. Otherwise, KVM treats 0xff 5542as a broadcast even in x2APIC mode in order to support physical x2APIC 5543without interrupt remapping. This is undesirable in logical mode, 5544where 0xff represents CPUs 0-7 in cluster 0. 5545 55467.8 KVM_CAP_S390_USER_INSTR0 5547---------------------------- 5548 5549:Architectures: s390 5550:Parameters: none 5551 5552With this capability enabled, all illegal instructions 0x0000 (2 bytes) will 5553be intercepted and forwarded to user space. User space can use this 5554mechanism e.g. to realize 2-byte software breakpoints. The kernel will 5555not inject an operating exception for these instructions, user space has 5556to take care of that. 5557 5558This capability can be enabled dynamically even if VCPUs were already 5559created and are running. 5560 55617.9 KVM_CAP_S390_GS 5562------------------- 5563 5564:Architectures: s390 5565:Parameters: none 5566:Returns: 0 on success; -EINVAL if the machine does not support 5567 guarded storage; -EBUSY if a VCPU has already been created. 5568 5569Allows use of guarded storage for the KVM guest. 5570 55717.10 KVM_CAP_S390_AIS 5572--------------------- 5573 5574:Architectures: s390 5575:Parameters: none 5576 5577Allow use of adapter-interruption suppression. 5578:Returns: 0 on success; -EBUSY if a VCPU has already been created. 5579 55807.11 KVM_CAP_PPC_SMT 5581-------------------- 5582 5583:Architectures: ppc 5584:Parameters: vsmt_mode, flags 5585 5586Enabling this capability on a VM provides userspace with a way to set 5587the desired virtual SMT mode (i.e. the number of virtual CPUs per 5588virtual core). The virtual SMT mode, vsmt_mode, must be a power of 2 5589between 1 and 8. On POWER8, vsmt_mode must also be no greater than 5590the number of threads per subcore for the host. Currently flags must 5591be 0. A successful call to enable this capability will result in 5592vsmt_mode being returned when the KVM_CAP_PPC_SMT capability is 5593subsequently queried for the VM. This capability is only supported by 5594HV KVM, and can only be set before any VCPUs have been created. 5595The KVM_CAP_PPC_SMT_POSSIBLE capability indicates which virtual SMT 5596modes are available. 5597 55987.12 KVM_CAP_PPC_FWNMI 5599---------------------- 5600 5601:Architectures: ppc 5602:Parameters: none 5603 5604With this capability a machine check exception in the guest address 5605space will cause KVM to exit the guest with NMI exit reason. This 5606enables QEMU to build error log and branch to guest kernel registered 5607machine check handling routine. Without this capability KVM will 5608branch to guests' 0x200 interrupt vector. 5609 56107.13 KVM_CAP_X86_DISABLE_EXITS 5611------------------------------ 5612 5613:Architectures: x86 5614:Parameters: args[0] defines which exits are disabled 5615:Returns: 0 on success, -EINVAL when args[0] contains invalid exits 5616 5617Valid bits in args[0] are:: 5618 5619 #define KVM_X86_DISABLE_EXITS_MWAIT (1 << 0) 5620 #define KVM_X86_DISABLE_EXITS_HLT (1 << 1) 5621 #define KVM_X86_DISABLE_EXITS_PAUSE (1 << 2) 5622 #define KVM_X86_DISABLE_EXITS_CSTATE (1 << 3) 5623 5624Enabling this capability on a VM provides userspace with a way to no 5625longer intercept some instructions for improved latency in some 5626workloads, and is suggested when vCPUs are associated to dedicated 5627physical CPUs. More bits can be added in the future; userspace can 5628just pass the KVM_CHECK_EXTENSION result to KVM_ENABLE_CAP to disable 5629all such vmexits. 5630 5631Do not enable KVM_FEATURE_PV_UNHALT if you disable HLT exits. 5632 56337.14 KVM_CAP_S390_HPAGE_1M 5634-------------------------- 5635 5636:Architectures: s390 5637:Parameters: none 5638:Returns: 0 on success, -EINVAL if hpage module parameter was not set 5639 or cmma is enabled, or the VM has the KVM_VM_S390_UCONTROL 5640 flag set 5641 5642With this capability the KVM support for memory backing with 1m pages 5643through hugetlbfs can be enabled for a VM. After the capability is 5644enabled, cmma can't be enabled anymore and pfmfi and the storage key 5645interpretation are disabled. If cmma has already been enabled or the 5646hpage module parameter is not set to 1, -EINVAL is returned. 5647 5648While it is generally possible to create a huge page backed VM without 5649this capability, the VM will not be able to run. 5650 56517.15 KVM_CAP_MSR_PLATFORM_INFO 5652------------------------------ 5653 5654:Architectures: x86 5655:Parameters: args[0] whether feature should be enabled or not 5656 5657With this capability, a guest may read the MSR_PLATFORM_INFO MSR. Otherwise, 5658a #GP would be raised when the guest tries to access. Currently, this 5659capability does not enable write permissions of this MSR for the guest. 5660 56617.16 KVM_CAP_PPC_NESTED_HV 5662-------------------------- 5663 5664:Architectures: ppc 5665:Parameters: none 5666:Returns: 0 on success, -EINVAL when the implementation doesn't support 5667 nested-HV virtualization. 5668 5669HV-KVM on POWER9 and later systems allows for "nested-HV" 5670virtualization, which provides a way for a guest VM to run guests that 5671can run using the CPU's supervisor mode (privileged non-hypervisor 5672state). Enabling this capability on a VM depends on the CPU having 5673the necessary functionality and on the facility being enabled with a 5674kvm-hv module parameter. 5675 56767.17 KVM_CAP_EXCEPTION_PAYLOAD 5677------------------------------ 5678 5679:Architectures: x86 5680:Parameters: args[0] whether feature should be enabled or not 5681 5682With this capability enabled, CR2 will not be modified prior to the 5683emulated VM-exit when L1 intercepts a #PF exception that occurs in 5684L2. Similarly, for kvm-intel only, DR6 will not be modified prior to 5685the emulated VM-exit when L1 intercepts a #DB exception that occurs in 5686L2. As a result, when KVM_GET_VCPU_EVENTS reports a pending #PF (or 5687#DB) exception for L2, exception.has_payload will be set and the 5688faulting address (or the new DR6 bits*) will be reported in the 5689exception_payload field. Similarly, when userspace injects a #PF (or 5690#DB) into L2 using KVM_SET_VCPU_EVENTS, it is expected to set 5691exception.has_payload and to put the faulting address - or the new DR6 5692bits\ [#]_ - in the exception_payload field. 5693 5694This capability also enables exception.pending in struct 5695kvm_vcpu_events, which allows userspace to distinguish between pending 5696and injected exceptions. 5697 5698 5699.. [#] For the new DR6 bits, note that bit 16 is set iff the #DB exception 5700 will clear DR6.RTM. 5701 57027.18 KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 5703 5704:Architectures: x86, arm, arm64, mips 5705:Parameters: args[0] whether feature should be enabled or not 5706 5707With this capability enabled, KVM_GET_DIRTY_LOG will not automatically 5708clear and write-protect all pages that are returned as dirty. 5709Rather, userspace will have to do this operation separately using 5710KVM_CLEAR_DIRTY_LOG. 5711 5712At the cost of a slightly more complicated operation, this provides better 5713scalability and responsiveness for two reasons. First, 5714KVM_CLEAR_DIRTY_LOG ioctl can operate on a 64-page granularity rather 5715than requiring to sync a full memslot; this ensures that KVM does not 5716take spinlocks for an extended period of time. Second, in some cases a 5717large amount of time can pass between a call to KVM_GET_DIRTY_LOG and 5718userspace actually using the data in the page. Pages can be modified 5719during this time, which is inefficint for both the guest and userspace: 5720the guest will incur a higher penalty due to write protection faults, 5721while userspace can see false reports of dirty pages. Manual reprotection 5722helps reducing this time, improving guest performance and reducing the 5723number of dirty log false positives. 5724 5725KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 was previously available under the name 5726KVM_CAP_MANUAL_DIRTY_LOG_PROTECT, but the implementation had bugs that make 5727it hard or impossible to use it correctly. The availability of 5728KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 signals that those bugs are fixed. 5729Userspace should not try to use KVM_CAP_MANUAL_DIRTY_LOG_PROTECT. 5730 57318. Other capabilities. 5732====================== 5733 5734This section lists capabilities that give information about other 5735features of the KVM implementation. 5736 57378.1 KVM_CAP_PPC_HWRNG 5738--------------------- 5739 5740:Architectures: ppc 5741 5742This capability, if KVM_CHECK_EXTENSION indicates that it is 5743available, means that that the kernel has an implementation of the 5744H_RANDOM hypercall backed by a hardware random-number generator. 5745If present, the kernel H_RANDOM handler can be enabled for guest use 5746with the KVM_CAP_PPC_ENABLE_HCALL capability. 5747 57488.2 KVM_CAP_HYPERV_SYNIC 5749------------------------ 5750 5751:Architectures: x86 5752 5753This capability, if KVM_CHECK_EXTENSION indicates that it is 5754available, means that that the kernel has an implementation of the 5755Hyper-V Synthetic interrupt controller(SynIC). Hyper-V SynIC is 5756used to support Windows Hyper-V based guest paravirt drivers(VMBus). 5757 5758In order to use SynIC, it has to be activated by setting this 5759capability via KVM_ENABLE_CAP ioctl on the vcpu fd. Note that this 5760will disable the use of APIC hardware virtualization even if supported 5761by the CPU, as it's incompatible with SynIC auto-EOI behavior. 5762 57638.3 KVM_CAP_PPC_RADIX_MMU 5764------------------------- 5765 5766:Architectures: ppc 5767 5768This capability, if KVM_CHECK_EXTENSION indicates that it is 5769available, means that that the kernel can support guests using the 5770radix MMU defined in Power ISA V3.00 (as implemented in the POWER9 5771processor). 5772 57738.4 KVM_CAP_PPC_HASH_MMU_V3 5774--------------------------- 5775 5776:Architectures: ppc 5777 5778This capability, if KVM_CHECK_EXTENSION indicates that it is 5779available, means that that the kernel can support guests using the 5780hashed page table MMU defined in Power ISA V3.00 (as implemented in 5781the POWER9 processor), including in-memory segment tables. 5782 57838.5 KVM_CAP_MIPS_VZ 5784------------------- 5785 5786:Architectures: mips 5787 5788This capability, if KVM_CHECK_EXTENSION on the main kvm handle indicates that 5789it is available, means that full hardware assisted virtualization capabilities 5790of the hardware are available for use through KVM. An appropriate 5791KVM_VM_MIPS_* type must be passed to KVM_CREATE_VM to create a VM which 5792utilises it. 5793 5794If KVM_CHECK_EXTENSION on a kvm VM handle indicates that this capability is 5795available, it means that the VM is using full hardware assisted virtualization 5796capabilities of the hardware. This is useful to check after creating a VM with 5797KVM_VM_MIPS_DEFAULT. 5798 5799The value returned by KVM_CHECK_EXTENSION should be compared against known 5800values (see below). All other values are reserved. This is to allow for the 5801possibility of other hardware assisted virtualization implementations which 5802may be incompatible with the MIPS VZ ASE. 5803 5804== ========================================================================== 5805 0 The trap & emulate implementation is in use to run guest code in user 5806 mode. Guest virtual memory segments are rearranged to fit the guest in the 5807 user mode address space. 5808 5809 1 The MIPS VZ ASE is in use, providing full hardware assisted 5810 virtualization, including standard guest virtual memory segments. 5811== ========================================================================== 5812 58138.6 KVM_CAP_MIPS_TE 5814------------------- 5815 5816:Architectures: mips 5817 5818This capability, if KVM_CHECK_EXTENSION on the main kvm handle indicates that 5819it is available, means that the trap & emulate implementation is available to 5820run guest code in user mode, even if KVM_CAP_MIPS_VZ indicates that hardware 5821assisted virtualisation is also available. KVM_VM_MIPS_TE (0) must be passed 5822to KVM_CREATE_VM to create a VM which utilises it. 5823 5824If KVM_CHECK_EXTENSION on a kvm VM handle indicates that this capability is 5825available, it means that the VM is using trap & emulate. 5826 58278.7 KVM_CAP_MIPS_64BIT 5828---------------------- 5829 5830:Architectures: mips 5831 5832This capability indicates the supported architecture type of the guest, i.e. the 5833supported register and address width. 5834 5835The values returned when this capability is checked by KVM_CHECK_EXTENSION on a 5836kvm VM handle correspond roughly to the CP0_Config.AT register field, and should 5837be checked specifically against known values (see below). All other values are 5838reserved. 5839 5840== ======================================================================== 5841 0 MIPS32 or microMIPS32. 5842 Both registers and addresses are 32-bits wide. 5843 It will only be possible to run 32-bit guest code. 5844 5845 1 MIPS64 or microMIPS64 with access only to 32-bit compatibility segments. 5846 Registers are 64-bits wide, but addresses are 32-bits wide. 5847 64-bit guest code may run but cannot access MIPS64 memory segments. 5848 It will also be possible to run 32-bit guest code. 5849 5850 2 MIPS64 or microMIPS64 with access to all address segments. 5851 Both registers and addresses are 64-bits wide. 5852 It will be possible to run 64-bit or 32-bit guest code. 5853== ======================================================================== 5854 58558.9 KVM_CAP_ARM_USER_IRQ 5856------------------------ 5857 5858:Architectures: arm, arm64 5859 5860This capability, if KVM_CHECK_EXTENSION indicates that it is available, means 5861that if userspace creates a VM without an in-kernel interrupt controller, it 5862will be notified of changes to the output level of in-kernel emulated devices, 5863which can generate virtual interrupts, presented to the VM. 5864For such VMs, on every return to userspace, the kernel 5865updates the vcpu's run->s.regs.device_irq_level field to represent the actual 5866output level of the device. 5867 5868Whenever kvm detects a change in the device output level, kvm guarantees at 5869least one return to userspace before running the VM. This exit could either 5870be a KVM_EXIT_INTR or any other exit event, like KVM_EXIT_MMIO. This way, 5871userspace can always sample the device output level and re-compute the state of 5872the userspace interrupt controller. Userspace should always check the state 5873of run->s.regs.device_irq_level on every kvm exit. 5874The value in run->s.regs.device_irq_level can represent both level and edge 5875triggered interrupt signals, depending on the device. Edge triggered interrupt 5876signals will exit to userspace with the bit in run->s.regs.device_irq_level 5877set exactly once per edge signal. 5878 5879The field run->s.regs.device_irq_level is available independent of 5880run->kvm_valid_regs or run->kvm_dirty_regs bits. 5881 5882If KVM_CAP_ARM_USER_IRQ is supported, the KVM_CHECK_EXTENSION ioctl returns a 5883number larger than 0 indicating the version of this capability is implemented 5884and thereby which bits in in run->s.regs.device_irq_level can signal values. 5885 5886Currently the following bits are defined for the device_irq_level bitmap:: 5887 5888 KVM_CAP_ARM_USER_IRQ >= 1: 5889 5890 KVM_ARM_DEV_EL1_VTIMER - EL1 virtual timer 5891 KVM_ARM_DEV_EL1_PTIMER - EL1 physical timer 5892 KVM_ARM_DEV_PMU - ARM PMU overflow interrupt signal 5893 5894Future versions of kvm may implement additional events. These will get 5895indicated by returning a higher number from KVM_CHECK_EXTENSION and will be 5896listed above. 5897 58988.10 KVM_CAP_PPC_SMT_POSSIBLE 5899----------------------------- 5900 5901:Architectures: ppc 5902 5903Querying this capability returns a bitmap indicating the possible 5904virtual SMT modes that can be set using KVM_CAP_PPC_SMT. If bit N 5905(counting from the right) is set, then a virtual SMT mode of 2^N is 5906available. 5907 59088.11 KVM_CAP_HYPERV_SYNIC2 5909-------------------------- 5910 5911:Architectures: x86 5912 5913This capability enables a newer version of Hyper-V Synthetic interrupt 5914controller (SynIC). The only difference with KVM_CAP_HYPERV_SYNIC is that KVM 5915doesn't clear SynIC message and event flags pages when they are enabled by 5916writing to the respective MSRs. 5917 59188.12 KVM_CAP_HYPERV_VP_INDEX 5919---------------------------- 5920 5921:Architectures: x86 5922 5923This capability indicates that userspace can load HV_X64_MSR_VP_INDEX msr. Its 5924value is used to denote the target vcpu for a SynIC interrupt. For 5925compatibilty, KVM initializes this msr to KVM's internal vcpu index. When this 5926capability is absent, userspace can still query this msr's value. 5927 59288.13 KVM_CAP_S390_AIS_MIGRATION 5929------------------------------- 5930 5931:Architectures: s390 5932:Parameters: none 5933 5934This capability indicates if the flic device will be able to get/set the 5935AIS states for migration via the KVM_DEV_FLIC_AISM_ALL attribute and allows 5936to discover this without having to create a flic device. 5937 59388.14 KVM_CAP_S390_PSW 5939--------------------- 5940 5941:Architectures: s390 5942 5943This capability indicates that the PSW is exposed via the kvm_run structure. 5944 59458.15 KVM_CAP_S390_GMAP 5946---------------------- 5947 5948:Architectures: s390 5949 5950This capability indicates that the user space memory used as guest mapping can 5951be anywhere in the user memory address space, as long as the memory slots are 5952aligned and sized to a segment (1MB) boundary. 5953 59548.16 KVM_CAP_S390_COW 5955--------------------- 5956 5957:Architectures: s390 5958 5959This capability indicates that the user space memory used as guest mapping can 5960use copy-on-write semantics as well as dirty pages tracking via read-only page 5961tables. 5962 59638.17 KVM_CAP_S390_BPB 5964--------------------- 5965 5966:Architectures: s390 5967 5968This capability indicates that kvm will implement the interfaces to handle 5969reset, migration and nested KVM for branch prediction blocking. The stfle 5970facility 82 should not be provided to the guest without this capability. 5971 59728.18 KVM_CAP_HYPERV_TLBFLUSH 5973---------------------------- 5974 5975:Architectures: x86 5976 5977This capability indicates that KVM supports paravirtualized Hyper-V TLB Flush 5978hypercalls: 5979HvFlushVirtualAddressSpace, HvFlushVirtualAddressSpaceEx, 5980HvFlushVirtualAddressList, HvFlushVirtualAddressListEx. 5981 59828.19 KVM_CAP_ARM_INJECT_SERROR_ESR 5983---------------------------------- 5984 5985:Architectures: arm, arm64 5986 5987This capability indicates that userspace can specify (via the 5988KVM_SET_VCPU_EVENTS ioctl) the syndrome value reported to the guest when it 5989takes a virtual SError interrupt exception. 5990If KVM advertises this capability, userspace can only specify the ISS field for 5991the ESR syndrome. Other parts of the ESR, such as the EC are generated by the 5992CPU when the exception is taken. If this virtual SError is taken to EL1 using 5993AArch64, this value will be reported in the ISS field of ESR_ELx. 5994 5995See KVM_CAP_VCPU_EVENTS for more details. 5996 59978.20 KVM_CAP_HYPERV_SEND_IPI 5998---------------------------- 5999 6000:Architectures: x86 6001 6002This capability indicates that KVM supports paravirtualized Hyper-V IPI send 6003hypercalls: 6004HvCallSendSyntheticClusterIpi, HvCallSendSyntheticClusterIpiEx. 6005 60068.21 KVM_CAP_HYPERV_DIRECT_TLBFLUSH 6007----------------------------------- 6008 6009:Architecture: x86 6010 6011This capability indicates that KVM running on top of Hyper-V hypervisor 6012enables Direct TLB flush for its guests meaning that TLB flush 6013hypercalls are handled by Level 0 hypervisor (Hyper-V) bypassing KVM. 6014Due to the different ABI for hypercall parameters between Hyper-V and 6015KVM, enabling this capability effectively disables all hypercall 6016handling by KVM (as some KVM hypercall may be mistakenly treated as TLB 6017flush hypercalls by Hyper-V) so userspace should disable KVM identification 6018in CPUID and only exposes Hyper-V identification. In this case, guest 6019thinks it's running on Hyper-V and only use Hyper-V hypercalls. 6020 60218.22 KVM_CAP_S390_VCPU_RESETS 6022 6023Architectures: s390 6024 6025This capability indicates that the KVM_S390_NORMAL_RESET and 6026KVM_S390_CLEAR_RESET ioctls are available. 6027