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 although VM ioctls may only be issued from 59the process that created the VM, a VM's lifecycle is associated with its 60file descriptor, not its creator (process). In other words, the VM and 61its resources, *including the associated address space*, are not freed 62until the last reference to the VM's file descriptor has been released. 63For example, if fork() is issued after ioctl(KVM_CREATE_VM), the VM will 64not be freed until both the parent (original) process and its child have 65put their references to the VM's file descriptor. 66 67Because a VM's resources are not freed until the last reference to its 68file descriptor is released, creating additional references to a VM 69via fork(), dup(), etc... without careful consideration is strongly 70discouraged and may have unwanted side effects, e.g. memory allocated 71by and on behalf of the VM's process may not be freed/unaccounted when 72the VM is shut down. 73 74 753. Extensions 76============= 77 78As of Linux 2.6.22, the KVM ABI has been stabilized: no backward 79incompatible change are allowed. However, there is an extension 80facility that allows backward-compatible extensions to the API to be 81queried and used. 82 83The extension mechanism is not based on the Linux version number. 84Instead, kvm defines extension identifiers and a facility to query 85whether a particular extension identifier is available. If it is, a 86set of ioctls is available for application use. 87 88 894. API description 90================== 91 92This section describes ioctls that can be used to control kvm guests. 93For each ioctl, the following information is provided along with a 94description: 95 96 Capability: 97 which KVM extension provides this ioctl. Can be 'basic', 98 which means that is will be provided by any kernel that supports 99 API version 12 (see section 4.1), a KVM_CAP_xyz constant, which 100 means availability needs to be checked with KVM_CHECK_EXTENSION 101 (see section 4.4), or 'none' which means that while not all kernels 102 support this ioctl, there's no capability bit to check its 103 availability: for kernels that don't support the ioctl, 104 the ioctl returns -ENOTTY. 105 106 Architectures: 107 which instruction set architectures provide this ioctl. 108 x86 includes both i386 and x86_64. 109 110 Type: 111 system, vm, or vcpu. 112 113 Parameters: 114 what parameters are accepted by the ioctl. 115 116 Returns: 117 the return value. General error numbers (EBADF, ENOMEM, EINVAL) 118 are not detailed, but errors with specific meanings are. 119 120 1214.1 KVM_GET_API_VERSION 122----------------------- 123 124:Capability: basic 125:Architectures: all 126:Type: system ioctl 127:Parameters: none 128:Returns: the constant KVM_API_VERSION (=12) 129 130This identifies the API version as the stable kvm API. It is not 131expected that this number will change. However, Linux 2.6.20 and 1322.6.21 report earlier versions; these are not documented and not 133supported. Applications should refuse to run if KVM_GET_API_VERSION 134returns a value other than 12. If this check passes, all ioctls 135described as 'basic' will be available. 136 137 1384.2 KVM_CREATE_VM 139----------------- 140 141:Capability: basic 142:Architectures: all 143:Type: system ioctl 144:Parameters: machine type identifier (KVM_VM_*) 145:Returns: a VM fd that can be used to control the new virtual machine. 146 147The new VM has no virtual cpus and no memory. 148You probably want to use 0 as machine type. 149 150X86: 151^^^^ 152 153Supported X86 VM types can be queried via KVM_CAP_VM_TYPES. 154 155S390: 156^^^^^ 157 158In order to create user controlled virtual machines on S390, check 159KVM_CAP_S390_UCONTROL and use the flag KVM_VM_S390_UCONTROL as 160privileged user (CAP_SYS_ADMIN). 161 162MIPS: 163^^^^^ 164 165To use hardware assisted virtualization on MIPS (VZ ASE) rather than 166the default trap & emulate implementation (which changes the virtual 167memory layout to fit in user mode), check KVM_CAP_MIPS_VZ and use the 168flag KVM_VM_MIPS_VZ. 169 170ARM64: 171^^^^^^ 172 173On arm64, the physical address size for a VM (IPA Size limit) is limited 174to 40bits by default. The limit can be configured if the host supports the 175extension KVM_CAP_ARM_VM_IPA_SIZE. When supported, use 176KVM_VM_TYPE_ARM_IPA_SIZE(IPA_Bits) to set the size in the machine type 177identifier, where IPA_Bits is the maximum width of any physical 178address used by the VM. The IPA_Bits is encoded in bits[7-0] of the 179machine type identifier. 180 181e.g, to configure a guest to use 48bit physical address size:: 182 183 vm_fd = ioctl(dev_fd, KVM_CREATE_VM, KVM_VM_TYPE_ARM_IPA_SIZE(48)); 184 185The requested size (IPA_Bits) must be: 186 187 == ========================================================= 188 0 Implies default size, 40bits (for backward compatibility) 189 N Implies N bits, where N is a positive integer such that, 190 32 <= N <= Host_IPA_Limit 191 == ========================================================= 192 193Host_IPA_Limit is the maximum possible value for IPA_Bits on the host and 194is dependent on the CPU capability and the kernel configuration. The limit can 195be retrieved using KVM_CAP_ARM_VM_IPA_SIZE of the KVM_CHECK_EXTENSION 196ioctl() at run-time. 197 198Creation of the VM will fail if the requested IPA size (whether it is 199implicit or explicit) is unsupported on the host. 200 201Please note that configuring the IPA size does not affect the capability 202exposed by the guest CPUs in ID_AA64MMFR0_EL1[PARange]. It only affects 203size of the address translated by the stage2 level (guest physical to 204host physical address translations). 205 206 2074.3 KVM_GET_MSR_INDEX_LIST, KVM_GET_MSR_FEATURE_INDEX_LIST 208---------------------------------------------------------- 209 210:Capability: basic, KVM_CAP_GET_MSR_FEATURES for KVM_GET_MSR_FEATURE_INDEX_LIST 211:Architectures: x86 212:Type: system ioctl 213:Parameters: struct kvm_msr_list (in/out) 214:Returns: 0 on success; -1 on error 215 216Errors: 217 218 ====== ============================================================ 219 EFAULT the msr index list cannot be read from or written to 220 E2BIG the msr index list is too big to fit in the array specified by 221 the user. 222 ====== ============================================================ 223 224:: 225 226 struct kvm_msr_list { 227 __u32 nmsrs; /* number of msrs in entries */ 228 __u32 indices[0]; 229 }; 230 231The user fills in the size of the indices array in nmsrs, and in return 232kvm adjusts nmsrs to reflect the actual number of msrs and fills in the 233indices array with their numbers. 234 235KVM_GET_MSR_INDEX_LIST returns the guest msrs that are supported. The list 236varies by kvm version and host processor, but does not change otherwise. 237 238Note: if kvm indicates supports MCE (KVM_CAP_MCE), then the MCE bank MSRs are 239not returned in the MSR list, as different vcpus can have a different number 240of banks, as set via the KVM_X86_SETUP_MCE ioctl. 241 242KVM_GET_MSR_FEATURE_INDEX_LIST returns the list of MSRs that can be passed 243to the KVM_GET_MSRS system ioctl. This lets userspace probe host capabilities 244and processor features that are exposed via MSRs (e.g., VMX capabilities). 245This list also varies by kvm version and host processor, but does not change 246otherwise. 247 248 2494.4 KVM_CHECK_EXTENSION 250----------------------- 251 252:Capability: basic, KVM_CAP_CHECK_EXTENSION_VM for vm ioctl 253:Architectures: all 254:Type: system ioctl, vm ioctl 255:Parameters: extension identifier (KVM_CAP_*) 256:Returns: 0 if unsupported; 1 (or some other positive integer) if supported 257 258The API allows the application to query about extensions to the core 259kvm API. Userspace passes an extension identifier (an integer) and 260receives an integer that describes the extension availability. 261Generally 0 means no and 1 means yes, but some extensions may report 262additional information in the integer return value. 263 264Based on their initialization different VMs may have different capabilities. 265It is thus encouraged to use the vm ioctl to query for capabilities (available 266with KVM_CAP_CHECK_EXTENSION_VM on the vm fd) 267 2684.5 KVM_GET_VCPU_MMAP_SIZE 269-------------------------- 270 271:Capability: basic 272:Architectures: all 273:Type: system ioctl 274:Parameters: none 275:Returns: size of vcpu mmap area, in bytes 276 277The KVM_RUN ioctl (cf.) communicates with userspace via a shared 278memory region. This ioctl returns the size of that region. See the 279KVM_RUN documentation for details. 280 281Besides the size of the KVM_RUN communication region, other areas of 282the VCPU file descriptor can be mmap-ed, including: 283 284- if KVM_CAP_COALESCED_MMIO is available, a page at 285 KVM_COALESCED_MMIO_PAGE_OFFSET * PAGE_SIZE; for historical reasons, 286 this page is included in the result of KVM_GET_VCPU_MMAP_SIZE. 287 KVM_CAP_COALESCED_MMIO is not documented yet. 288 289- if KVM_CAP_DIRTY_LOG_RING is available, a number of pages at 290 KVM_DIRTY_LOG_PAGE_OFFSET * PAGE_SIZE. For more information on 291 KVM_CAP_DIRTY_LOG_RING, see section 8.3. 292 293 2944.7 KVM_CREATE_VCPU 295------------------- 296 297:Capability: basic 298:Architectures: all 299:Type: vm ioctl 300:Parameters: vcpu id (apic id on x86) 301:Returns: vcpu fd on success, -1 on error 302 303This API adds a vcpu to a virtual machine. No more than max_vcpus may be added. 304The vcpu id is an integer in the range [0, max_vcpu_id). 305 306The recommended max_vcpus value can be retrieved using the KVM_CAP_NR_VCPUS of 307the KVM_CHECK_EXTENSION ioctl() at run-time. 308The maximum possible value for max_vcpus can be retrieved using the 309KVM_CAP_MAX_VCPUS of the KVM_CHECK_EXTENSION ioctl() at run-time. 310 311If the KVM_CAP_NR_VCPUS does not exist, you should assume that max_vcpus is 4 312cpus max. 313If the KVM_CAP_MAX_VCPUS does not exist, you should assume that max_vcpus is 314same as the value returned from KVM_CAP_NR_VCPUS. 315 316The maximum possible value for max_vcpu_id can be retrieved using the 317KVM_CAP_MAX_VCPU_ID of the KVM_CHECK_EXTENSION ioctl() at run-time. 318 319If the KVM_CAP_MAX_VCPU_ID does not exist, you should assume that max_vcpu_id 320is the same as the value returned from KVM_CAP_MAX_VCPUS. 321 322On powerpc using book3s_hv mode, the vcpus are mapped onto virtual 323threads in one or more virtual CPU cores. (This is because the 324hardware requires all the hardware threads in a CPU core to be in the 325same partition.) The KVM_CAP_PPC_SMT capability indicates the number 326of vcpus per virtual core (vcore). The vcore id is obtained by 327dividing the vcpu id by the number of vcpus per vcore. The vcpus in a 328given vcore will always be in the same physical core as each other 329(though that might be a different physical core from time to time). 330Userspace can control the threading (SMT) mode of the guest by its 331allocation of vcpu ids. For example, if userspace wants 332single-threaded guest vcpus, it should make all vcpu ids be a multiple 333of the number of vcpus per vcore. 334 335For virtual cpus that have been created with S390 user controlled virtual 336machines, the resulting vcpu fd can be memory mapped at page offset 337KVM_S390_SIE_PAGE_OFFSET in order to obtain a memory map of the virtual 338cpu's hardware control block. 339 340 3414.8 KVM_GET_DIRTY_LOG (vm ioctl) 342-------------------------------- 343 344:Capability: basic 345:Architectures: all 346:Type: vm ioctl 347:Parameters: struct kvm_dirty_log (in/out) 348:Returns: 0 on success, -1 on error 349 350:: 351 352 /* for KVM_GET_DIRTY_LOG */ 353 struct kvm_dirty_log { 354 __u32 slot; 355 __u32 padding; 356 union { 357 void __user *dirty_bitmap; /* one bit per page */ 358 __u64 padding; 359 }; 360 }; 361 362Given a memory slot, return a bitmap containing any pages dirtied 363since the last call to this ioctl. Bit 0 is the first page in the 364memory slot. Ensure the entire structure is cleared to avoid padding 365issues. 366 367If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of slot field specifies 368the address space for which you want to return the dirty bitmap. See 369KVM_SET_USER_MEMORY_REGION for details on the usage of slot field. 370 371The bits in the dirty bitmap are cleared before the ioctl returns, unless 372KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 is enabled. For more information, 373see the description of the capability. 374 375Note that the Xen shared_info page, if configured, shall always be assumed 376to be dirty. KVM will not explicitly mark it such. 377 378 3794.10 KVM_RUN 380------------ 381 382:Capability: basic 383:Architectures: all 384:Type: vcpu ioctl 385:Parameters: none 386:Returns: 0 on success, -1 on error 387 388Errors: 389 390 ======= ============================================================== 391 EINTR an unmasked signal is pending 392 ENOEXEC the vcpu hasn't been initialized or the guest tried to execute 393 instructions from device memory (arm64) 394 ENOSYS data abort outside memslots with no syndrome info and 395 KVM_CAP_ARM_NISV_TO_USER not enabled (arm64) 396 EPERM SVE feature set but not finalized (arm64) 397 ======= ============================================================== 398 399This ioctl is used to run a guest virtual cpu. While there are no 400explicit parameters, there is an implicit parameter block that can be 401obtained by mmap()ing the vcpu fd at offset 0, with the size given by 402KVM_GET_VCPU_MMAP_SIZE. The parameter block is formatted as a 'struct 403kvm_run' (see below). 404 405 4064.11 KVM_GET_REGS 407----------------- 408 409:Capability: basic 410:Architectures: all except arm64 411:Type: vcpu ioctl 412:Parameters: struct kvm_regs (out) 413:Returns: 0 on success, -1 on error 414 415Reads the general purpose registers from the vcpu. 416 417:: 418 419 /* x86 */ 420 struct kvm_regs { 421 /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */ 422 __u64 rax, rbx, rcx, rdx; 423 __u64 rsi, rdi, rsp, rbp; 424 __u64 r8, r9, r10, r11; 425 __u64 r12, r13, r14, r15; 426 __u64 rip, rflags; 427 }; 428 429 /* mips */ 430 struct kvm_regs { 431 /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */ 432 __u64 gpr[32]; 433 __u64 hi; 434 __u64 lo; 435 __u64 pc; 436 }; 437 438 /* LoongArch */ 439 struct kvm_regs { 440 /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */ 441 unsigned long gpr[32]; 442 unsigned long pc; 443 }; 444 445 4464.12 KVM_SET_REGS 447----------------- 448 449:Capability: basic 450:Architectures: all except arm64 451:Type: vcpu ioctl 452:Parameters: struct kvm_regs (in) 453:Returns: 0 on success, -1 on error 454 455Writes the general purpose registers into the vcpu. 456 457See KVM_GET_REGS for the data structure. 458 459 4604.13 KVM_GET_SREGS 461------------------ 462 463:Capability: basic 464:Architectures: x86, ppc 465:Type: vcpu ioctl 466:Parameters: struct kvm_sregs (out) 467:Returns: 0 on success, -1 on error 468 469Reads special registers from the vcpu. 470 471:: 472 473 /* x86 */ 474 struct kvm_sregs { 475 struct kvm_segment cs, ds, es, fs, gs, ss; 476 struct kvm_segment tr, ldt; 477 struct kvm_dtable gdt, idt; 478 __u64 cr0, cr2, cr3, cr4, cr8; 479 __u64 efer; 480 __u64 apic_base; 481 __u64 interrupt_bitmap[(KVM_NR_INTERRUPTS + 63) / 64]; 482 }; 483 484 /* ppc -- see arch/powerpc/include/uapi/asm/kvm.h */ 485 486interrupt_bitmap is a bitmap of pending external interrupts. At most 487one bit may be set. This interrupt has been acknowledged by the APIC 488but not yet injected into the cpu core. 489 490 4914.14 KVM_SET_SREGS 492------------------ 493 494:Capability: basic 495:Architectures: x86, ppc 496:Type: vcpu ioctl 497:Parameters: struct kvm_sregs (in) 498:Returns: 0 on success, -1 on error 499 500Writes special registers into the vcpu. See KVM_GET_SREGS for the 501data structures. 502 503 5044.15 KVM_TRANSLATE 505------------------ 506 507:Capability: basic 508:Architectures: x86 509:Type: vcpu ioctl 510:Parameters: struct kvm_translation (in/out) 511:Returns: 0 on success, -1 on error 512 513Translates a virtual address according to the vcpu's current address 514translation mode. 515 516:: 517 518 struct kvm_translation { 519 /* in */ 520 __u64 linear_address; 521 522 /* out */ 523 __u64 physical_address; 524 __u8 valid; 525 __u8 writeable; 526 __u8 usermode; 527 __u8 pad[5]; 528 }; 529 530 5314.16 KVM_INTERRUPT 532------------------ 533 534:Capability: basic 535:Architectures: x86, ppc, mips, riscv, loongarch 536:Type: vcpu ioctl 537:Parameters: struct kvm_interrupt (in) 538:Returns: 0 on success, negative on failure. 539 540Queues a hardware interrupt vector to be injected. 541 542:: 543 544 /* for KVM_INTERRUPT */ 545 struct kvm_interrupt { 546 /* in */ 547 __u32 irq; 548 }; 549 550X86: 551^^^^ 552 553:Returns: 554 555 ========= =================================== 556 0 on success, 557 -EEXIST if an interrupt is already enqueued 558 -EINVAL the irq number is invalid 559 -ENXIO if the PIC is in the kernel 560 -EFAULT if the pointer is invalid 561 ========= =================================== 562 563Note 'irq' is an interrupt vector, not an interrupt pin or line. This 564ioctl is useful if the in-kernel PIC is not used. 565 566PPC: 567^^^^ 568 569Queues an external interrupt to be injected. This ioctl is overloaded 570with 3 different irq values: 571 572a) KVM_INTERRUPT_SET 573 574 This injects an edge type external interrupt into the guest once it's ready 575 to receive interrupts. When injected, the interrupt is done. 576 577b) KVM_INTERRUPT_UNSET 578 579 This unsets any pending interrupt. 580 581 Only available with KVM_CAP_PPC_UNSET_IRQ. 582 583c) KVM_INTERRUPT_SET_LEVEL 584 585 This injects a level type external interrupt into the guest context. The 586 interrupt stays pending until a specific ioctl with KVM_INTERRUPT_UNSET 587 is triggered. 588 589 Only available with KVM_CAP_PPC_IRQ_LEVEL. 590 591Note that any value for 'irq' other than the ones stated above is invalid 592and incurs unexpected behavior. 593 594This is an asynchronous vcpu ioctl and can be invoked from any thread. 595 596MIPS: 597^^^^^ 598 599Queues an external interrupt to be injected into the virtual CPU. A negative 600interrupt number dequeues the interrupt. 601 602This is an asynchronous vcpu ioctl and can be invoked from any thread. 603 604RISC-V: 605^^^^^^^ 606 607Queues an external interrupt to be injected into the virtual CPU. This ioctl 608is overloaded with 2 different irq values: 609 610a) KVM_INTERRUPT_SET 611 612 This sets external interrupt for a virtual CPU and it will receive 613 once it is ready. 614 615b) KVM_INTERRUPT_UNSET 616 617 This clears pending external interrupt for a virtual CPU. 618 619This is an asynchronous vcpu ioctl and can be invoked from any thread. 620 621LOONGARCH: 622^^^^^^^^^^ 623 624Queues an external interrupt to be injected into the virtual CPU. A negative 625interrupt number dequeues the interrupt. 626 627This is an asynchronous vcpu ioctl and can be invoked from any thread. 628 629 6304.18 KVM_GET_MSRS 631----------------- 632 633:Capability: basic (vcpu), KVM_CAP_GET_MSR_FEATURES (system) 634:Architectures: x86 635:Type: system ioctl, vcpu ioctl 636:Parameters: struct kvm_msrs (in/out) 637:Returns: number of msrs successfully returned; 638 -1 on error 639 640When used as a system ioctl: 641Reads the values of MSR-based features that are available for the VM. This 642is similar to KVM_GET_SUPPORTED_CPUID, but it returns MSR indices and values. 643The list of msr-based features can be obtained using KVM_GET_MSR_FEATURE_INDEX_LIST 644in a system ioctl. 645 646When used as a vcpu ioctl: 647Reads model-specific registers from the vcpu. Supported msr indices can 648be obtained using KVM_GET_MSR_INDEX_LIST in a system ioctl. 649 650:: 651 652 struct kvm_msrs { 653 __u32 nmsrs; /* number of msrs in entries */ 654 __u32 pad; 655 656 struct kvm_msr_entry entries[0]; 657 }; 658 659 struct kvm_msr_entry { 660 __u32 index; 661 __u32 reserved; 662 __u64 data; 663 }; 664 665Application code should set the 'nmsrs' member (which indicates the 666size of the entries array) and the 'index' member of each array entry. 667kvm will fill in the 'data' member. 668 669 6704.19 KVM_SET_MSRS 671----------------- 672 673:Capability: basic 674:Architectures: x86 675:Type: vcpu ioctl 676:Parameters: struct kvm_msrs (in) 677:Returns: number of msrs successfully set (see below), -1 on error 678 679Writes model-specific registers to the vcpu. See KVM_GET_MSRS for the 680data structures. 681 682Application code should set the 'nmsrs' member (which indicates the 683size of the entries array), and the 'index' and 'data' members of each 684array entry. 685 686It tries to set the MSRs in array entries[] one by one. If setting an MSR 687fails, e.g., due to setting reserved bits, the MSR isn't supported/emulated 688by KVM, etc..., it stops processing the MSR list and returns the number of 689MSRs that have been set successfully. 690 691 6924.20 KVM_SET_CPUID 693------------------ 694 695:Capability: basic 696:Architectures: x86 697:Type: vcpu ioctl 698:Parameters: struct kvm_cpuid (in) 699:Returns: 0 on success, -1 on error 700 701Defines the vcpu responses to the cpuid instruction. Applications 702should use the KVM_SET_CPUID2 ioctl if available. 703 704Caveat emptor: 705 - If this IOCTL fails, KVM gives no guarantees that previous valid CPUID 706 configuration (if there is) is not corrupted. Userspace can get a copy 707 of the resulting CPUID configuration through KVM_GET_CPUID2 in case. 708 - Using KVM_SET_CPUID{,2} after KVM_RUN, i.e. changing the guest vCPU model 709 after running the guest, may cause guest instability. 710 - Using heterogeneous CPUID configurations, modulo APIC IDs, topology, etc... 711 may cause guest instability. 712 713:: 714 715 struct kvm_cpuid_entry { 716 __u32 function; 717 __u32 eax; 718 __u32 ebx; 719 __u32 ecx; 720 __u32 edx; 721 __u32 padding; 722 }; 723 724 /* for KVM_SET_CPUID */ 725 struct kvm_cpuid { 726 __u32 nent; 727 __u32 padding; 728 struct kvm_cpuid_entry entries[0]; 729 }; 730 731 7324.21 KVM_SET_SIGNAL_MASK 733------------------------ 734 735:Capability: basic 736:Architectures: all 737:Type: vcpu ioctl 738:Parameters: struct kvm_signal_mask (in) 739:Returns: 0 on success, -1 on error 740 741Defines which signals are blocked during execution of KVM_RUN. This 742signal mask temporarily overrides the threads signal mask. Any 743unblocked signal received (except SIGKILL and SIGSTOP, which retain 744their traditional behaviour) will cause KVM_RUN to return with -EINTR. 745 746Note the signal will only be delivered if not blocked by the original 747signal mask. 748 749:: 750 751 /* for KVM_SET_SIGNAL_MASK */ 752 struct kvm_signal_mask { 753 __u32 len; 754 __u8 sigset[0]; 755 }; 756 757 7584.22 KVM_GET_FPU 759---------------- 760 761:Capability: basic 762:Architectures: x86, loongarch 763:Type: vcpu ioctl 764:Parameters: struct kvm_fpu (out) 765:Returns: 0 on success, -1 on error 766 767Reads the floating point state from the vcpu. 768 769:: 770 771 /* x86: for KVM_GET_FPU and KVM_SET_FPU */ 772 struct kvm_fpu { 773 __u8 fpr[8][16]; 774 __u16 fcw; 775 __u16 fsw; 776 __u8 ftwx; /* in fxsave format */ 777 __u8 pad1; 778 __u16 last_opcode; 779 __u64 last_ip; 780 __u64 last_dp; 781 __u8 xmm[16][16]; 782 __u32 mxcsr; 783 __u32 pad2; 784 }; 785 786 /* LoongArch: for KVM_GET_FPU and KVM_SET_FPU */ 787 struct kvm_fpu { 788 __u32 fcsr; 789 __u64 fcc; 790 struct kvm_fpureg { 791 __u64 val64[4]; 792 }fpr[32]; 793 }; 794 795 7964.23 KVM_SET_FPU 797---------------- 798 799:Capability: basic 800:Architectures: x86, loongarch 801:Type: vcpu ioctl 802:Parameters: struct kvm_fpu (in) 803:Returns: 0 on success, -1 on error 804 805Writes the floating point state to the vcpu. 806 807:: 808 809 /* x86: for KVM_GET_FPU and KVM_SET_FPU */ 810 struct kvm_fpu { 811 __u8 fpr[8][16]; 812 __u16 fcw; 813 __u16 fsw; 814 __u8 ftwx; /* in fxsave format */ 815 __u8 pad1; 816 __u16 last_opcode; 817 __u64 last_ip; 818 __u64 last_dp; 819 __u8 xmm[16][16]; 820 __u32 mxcsr; 821 __u32 pad2; 822 }; 823 824 /* LoongArch: for KVM_GET_FPU and KVM_SET_FPU */ 825 struct kvm_fpu { 826 __u32 fcsr; 827 __u64 fcc; 828 struct kvm_fpureg { 829 __u64 val64[4]; 830 }fpr[32]; 831 }; 832 833 8344.24 KVM_CREATE_IRQCHIP 835----------------------- 836 837:Capability: KVM_CAP_IRQCHIP, KVM_CAP_S390_IRQCHIP (s390) 838:Architectures: x86, arm64, s390 839:Type: vm ioctl 840:Parameters: none 841:Returns: 0 on success, -1 on error 842 843Creates an interrupt controller model in the kernel. 844On x86, creates a virtual ioapic, a virtual PIC (two PICs, nested), and sets up 845future vcpus to have a local APIC. IRQ routing for GSIs 0-15 is set to both 846PIC and IOAPIC; GSI 16-23 only go to the IOAPIC. 847On arm64, a GICv2 is created. Any other GIC versions require the usage of 848KVM_CREATE_DEVICE, which also supports creating a GICv2. Using 849KVM_CREATE_DEVICE is preferred over KVM_CREATE_IRQCHIP for GICv2. 850On s390, a dummy irq routing table is created. 851 852Note that on s390 the KVM_CAP_S390_IRQCHIP vm capability needs to be enabled 853before KVM_CREATE_IRQCHIP can be used. 854 855 8564.25 KVM_IRQ_LINE 857----------------- 858 859:Capability: KVM_CAP_IRQCHIP 860:Architectures: x86, arm64 861:Type: vm ioctl 862:Parameters: struct kvm_irq_level 863:Returns: 0 on success, -1 on error 864 865Sets the level of a GSI input to the interrupt controller model in the kernel. 866On some architectures it is required that an interrupt controller model has 867been previously created with KVM_CREATE_IRQCHIP. Note that edge-triggered 868interrupts require the level to be set to 1 and then back to 0. 869 870On real hardware, interrupt pins can be active-low or active-high. This 871does not matter for the level field of struct kvm_irq_level: 1 always 872means active (asserted), 0 means inactive (deasserted). 873 874x86 allows the operating system to program the interrupt polarity 875(active-low/active-high) for level-triggered interrupts, and KVM used 876to consider the polarity. However, due to bitrot in the handling of 877active-low interrupts, the above convention is now valid on x86 too. 878This is signaled by KVM_CAP_X86_IOAPIC_POLARITY_IGNORED. Userspace 879should not present interrupts to the guest as active-low unless this 880capability is present (or unless it is not using the in-kernel irqchip, 881of course). 882 883 884arm64 can signal an interrupt either at the CPU level, or at the 885in-kernel irqchip (GIC), and for in-kernel irqchip can tell the GIC to 886use PPIs designated for specific cpus. The irq field is interpreted 887like this:: 888 889 bits: | 31 ... 28 | 27 ... 24 | 23 ... 16 | 15 ... 0 | 890 field: | vcpu2_index | irq_type | vcpu_index | irq_id | 891 892The irq_type field has the following values: 893 894- KVM_ARM_IRQ_TYPE_CPU: 895 out-of-kernel GIC: irq_id 0 is IRQ, irq_id 1 is FIQ 896- KVM_ARM_IRQ_TYPE_SPI: 897 in-kernel GIC: SPI, irq_id between 32 and 1019 (incl.) 898 (the vcpu_index field is ignored) 899- KVM_ARM_IRQ_TYPE_PPI: 900 in-kernel GIC: PPI, irq_id between 16 and 31 (incl.) 901 902(The irq_id field thus corresponds nicely to the IRQ ID in the ARM GIC specs) 903 904In both cases, level is used to assert/deassert the line. 905 906When KVM_CAP_ARM_IRQ_LINE_LAYOUT_2 is supported, the target vcpu is 907identified as (256 * vcpu2_index + vcpu_index). Otherwise, vcpu2_index 908must be zero. 909 910Note that on arm64, the KVM_CAP_IRQCHIP capability only conditions 911injection of interrupts for the in-kernel irqchip. KVM_IRQ_LINE can always 912be used for a userspace interrupt controller. 913 914:: 915 916 struct kvm_irq_level { 917 union { 918 __u32 irq; /* GSI */ 919 __s32 status; /* not used for KVM_IRQ_LEVEL */ 920 }; 921 __u32 level; /* 0 or 1 */ 922 }; 923 924 9254.26 KVM_GET_IRQCHIP 926-------------------- 927 928:Capability: KVM_CAP_IRQCHIP 929:Architectures: x86 930:Type: vm ioctl 931:Parameters: struct kvm_irqchip (in/out) 932:Returns: 0 on success, -1 on error 933 934Reads the state of a kernel interrupt controller created with 935KVM_CREATE_IRQCHIP into a buffer provided by the caller. 936 937:: 938 939 struct kvm_irqchip { 940 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */ 941 __u32 pad; 942 union { 943 char dummy[512]; /* reserving space */ 944 struct kvm_pic_state pic; 945 struct kvm_ioapic_state ioapic; 946 } chip; 947 }; 948 949 9504.27 KVM_SET_IRQCHIP 951-------------------- 952 953:Capability: KVM_CAP_IRQCHIP 954:Architectures: x86 955:Type: vm ioctl 956:Parameters: struct kvm_irqchip (in) 957:Returns: 0 on success, -1 on error 958 959Sets the state of a kernel interrupt controller created with 960KVM_CREATE_IRQCHIP from a buffer provided by the caller. 961 962:: 963 964 struct kvm_irqchip { 965 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */ 966 __u32 pad; 967 union { 968 char dummy[512]; /* reserving space */ 969 struct kvm_pic_state pic; 970 struct kvm_ioapic_state ioapic; 971 } chip; 972 }; 973 974 9754.28 KVM_XEN_HVM_CONFIG 976----------------------- 977 978:Capability: KVM_CAP_XEN_HVM 979:Architectures: x86 980:Type: vm ioctl 981:Parameters: struct kvm_xen_hvm_config (in) 982:Returns: 0 on success, -1 on error 983 984Sets the MSR that the Xen HVM guest uses to initialize its hypercall 985page, and provides the starting address and size of the hypercall 986blobs in userspace. When the guest writes the MSR, kvm copies one 987page of a blob (32- or 64-bit, depending on the vcpu mode) to guest 988memory. 989 990:: 991 992 struct kvm_xen_hvm_config { 993 __u32 flags; 994 __u32 msr; 995 __u64 blob_addr_32; 996 __u64 blob_addr_64; 997 __u8 blob_size_32; 998 __u8 blob_size_64; 999 __u8 pad2[30]; 1000 }; 1001 1002If certain flags are returned from the KVM_CAP_XEN_HVM check, they may 1003be set in the flags field of this ioctl: 1004 1005The KVM_XEN_HVM_CONFIG_INTERCEPT_HCALL flag requests KVM to generate 1006the contents of the hypercall page automatically; hypercalls will be 1007intercepted and passed to userspace through KVM_EXIT_XEN. In this 1008case, all of the blob size and address fields must be zero. 1009 1010The KVM_XEN_HVM_CONFIG_EVTCHN_SEND flag indicates to KVM that userspace 1011will always use the KVM_XEN_HVM_EVTCHN_SEND ioctl to deliver event 1012channel interrupts rather than manipulating the guest's shared_info 1013structures directly. This, in turn, may allow KVM to enable features 1014such as intercepting the SCHEDOP_poll hypercall to accelerate PV 1015spinlock operation for the guest. Userspace may still use the ioctl 1016to deliver events if it was advertised, even if userspace does not 1017send this indication that it will always do so 1018 1019No other flags are currently valid in the struct kvm_xen_hvm_config. 1020 10214.29 KVM_GET_CLOCK 1022------------------ 1023 1024:Capability: KVM_CAP_ADJUST_CLOCK 1025:Architectures: x86 1026:Type: vm ioctl 1027:Parameters: struct kvm_clock_data (out) 1028:Returns: 0 on success, -1 on error 1029 1030Gets the current timestamp of kvmclock as seen by the current guest. In 1031conjunction with KVM_SET_CLOCK, it is used to ensure monotonicity on scenarios 1032such as migration. 1033 1034When KVM_CAP_ADJUST_CLOCK is passed to KVM_CHECK_EXTENSION, it returns the 1035set of bits that KVM can return in struct kvm_clock_data's flag member. 1036 1037The following flags are defined: 1038 1039KVM_CLOCK_TSC_STABLE 1040 If set, the returned value is the exact kvmclock 1041 value seen by all VCPUs at the instant when KVM_GET_CLOCK was called. 1042 If clear, the returned value is simply CLOCK_MONOTONIC plus a constant 1043 offset; the offset can be modified with KVM_SET_CLOCK. KVM will try 1044 to make all VCPUs follow this clock, but the exact value read by each 1045 VCPU could differ, because the host TSC is not stable. 1046 1047KVM_CLOCK_REALTIME 1048 If set, the `realtime` field in the kvm_clock_data 1049 structure is populated with the value of the host's real time 1050 clocksource at the instant when KVM_GET_CLOCK was called. If clear, 1051 the `realtime` field does not contain a value. 1052 1053KVM_CLOCK_HOST_TSC 1054 If set, the `host_tsc` field in the kvm_clock_data 1055 structure is populated with the value of the host's timestamp counter (TSC) 1056 at the instant when KVM_GET_CLOCK was called. If clear, the `host_tsc` field 1057 does not contain a value. 1058 1059:: 1060 1061 struct kvm_clock_data { 1062 __u64 clock; /* kvmclock current value */ 1063 __u32 flags; 1064 __u32 pad0; 1065 __u64 realtime; 1066 __u64 host_tsc; 1067 __u32 pad[4]; 1068 }; 1069 1070 10714.30 KVM_SET_CLOCK 1072------------------ 1073 1074:Capability: KVM_CAP_ADJUST_CLOCK 1075:Architectures: x86 1076:Type: vm ioctl 1077:Parameters: struct kvm_clock_data (in) 1078:Returns: 0 on success, -1 on error 1079 1080Sets the current timestamp of kvmclock to the value specified in its parameter. 1081In conjunction with KVM_GET_CLOCK, it is used to ensure monotonicity on scenarios 1082such as migration. 1083 1084The following flags can be passed: 1085 1086KVM_CLOCK_REALTIME 1087 If set, KVM will compare the value of the `realtime` field 1088 with the value of the host's real time clocksource at the instant when 1089 KVM_SET_CLOCK was called. The difference in elapsed time is added to the final 1090 kvmclock value that will be provided to guests. 1091 1092Other flags returned by ``KVM_GET_CLOCK`` are accepted but ignored. 1093 1094:: 1095 1096 struct kvm_clock_data { 1097 __u64 clock; /* kvmclock current value */ 1098 __u32 flags; 1099 __u32 pad0; 1100 __u64 realtime; 1101 __u64 host_tsc; 1102 __u32 pad[4]; 1103 }; 1104 1105 11064.31 KVM_GET_VCPU_EVENTS 1107------------------------ 1108 1109:Capability: KVM_CAP_VCPU_EVENTS 1110:Extended by: KVM_CAP_INTR_SHADOW 1111:Architectures: x86, arm64 1112:Type: vcpu ioctl 1113:Parameters: struct kvm_vcpu_events (out) 1114:Returns: 0 on success, -1 on error 1115 1116X86: 1117^^^^ 1118 1119Gets currently pending exceptions, interrupts, and NMIs as well as related 1120states of the vcpu. 1121 1122:: 1123 1124 struct kvm_vcpu_events { 1125 struct { 1126 __u8 injected; 1127 __u8 nr; 1128 __u8 has_error_code; 1129 __u8 pending; 1130 __u32 error_code; 1131 } exception; 1132 struct { 1133 __u8 injected; 1134 __u8 nr; 1135 __u8 soft; 1136 __u8 shadow; 1137 } interrupt; 1138 struct { 1139 __u8 injected; 1140 __u8 pending; 1141 __u8 masked; 1142 __u8 pad; 1143 } nmi; 1144 __u32 sipi_vector; 1145 __u32 flags; 1146 struct { 1147 __u8 smm; 1148 __u8 pending; 1149 __u8 smm_inside_nmi; 1150 __u8 latched_init; 1151 } smi; 1152 __u8 reserved[27]; 1153 __u8 exception_has_payload; 1154 __u64 exception_payload; 1155 }; 1156 1157The following bits are defined in the flags field: 1158 1159- KVM_VCPUEVENT_VALID_SHADOW may be set to signal that 1160 interrupt.shadow contains a valid state. 1161 1162- KVM_VCPUEVENT_VALID_SMM may be set to signal that smi contains a 1163 valid state. 1164 1165- KVM_VCPUEVENT_VALID_PAYLOAD may be set to signal that the 1166 exception_has_payload, exception_payload, and exception.pending 1167 fields contain a valid state. This bit will be set whenever 1168 KVM_CAP_EXCEPTION_PAYLOAD is enabled. 1169 1170- KVM_VCPUEVENT_VALID_TRIPLE_FAULT may be set to signal that the 1171 triple_fault_pending field contains a valid state. This bit will 1172 be set whenever KVM_CAP_X86_TRIPLE_FAULT_EVENT is enabled. 1173 1174ARM64: 1175^^^^^^ 1176 1177If the guest accesses a device that is being emulated by the host kernel in 1178such a way that a real device would generate a physical SError, KVM may make 1179a virtual SError pending for that VCPU. This system error interrupt remains 1180pending until the guest takes the exception by unmasking PSTATE.A. 1181 1182Running the VCPU may cause it to take a pending SError, or make an access that 1183causes an SError to become pending. The event's description is only valid while 1184the VPCU is not running. 1185 1186This API provides a way to read and write the pending 'event' state that is not 1187visible to the guest. To save, restore or migrate a VCPU the struct representing 1188the state can be read then written using this GET/SET API, along with the other 1189guest-visible registers. It is not possible to 'cancel' an SError that has been 1190made pending. 1191 1192A device being emulated in user-space may also wish to generate an SError. To do 1193this the events structure can be populated by user-space. The current state 1194should be read first, to ensure no existing SError is pending. If an existing 1195SError is pending, the architecture's 'Multiple SError interrupts' rules should 1196be followed. (2.5.3 of DDI0587.a "ARM Reliability, Availability, and 1197Serviceability (RAS) Specification"). 1198 1199SError exceptions always have an ESR value. Some CPUs have the ability to 1200specify what the virtual SError's ESR value should be. These systems will 1201advertise KVM_CAP_ARM_INJECT_SERROR_ESR. In this case exception.has_esr will 1202always have a non-zero value when read, and the agent making an SError pending 1203should specify the ISS field in the lower 24 bits of exception.serror_esr. If 1204the system supports KVM_CAP_ARM_INJECT_SERROR_ESR, but user-space sets the events 1205with exception.has_esr as zero, KVM will choose an ESR. 1206 1207Specifying exception.has_esr on a system that does not support it will return 1208-EINVAL. Setting anything other than the lower 24bits of exception.serror_esr 1209will return -EINVAL. 1210 1211It is not possible to read back a pending external abort (injected via 1212KVM_SET_VCPU_EVENTS or otherwise) because such an exception is always delivered 1213directly to the virtual CPU). 1214 1215:: 1216 1217 struct kvm_vcpu_events { 1218 struct { 1219 __u8 serror_pending; 1220 __u8 serror_has_esr; 1221 __u8 ext_dabt_pending; 1222 /* Align it to 8 bytes */ 1223 __u8 pad[5]; 1224 __u64 serror_esr; 1225 } exception; 1226 __u32 reserved[12]; 1227 }; 1228 12294.32 KVM_SET_VCPU_EVENTS 1230------------------------ 1231 1232:Capability: KVM_CAP_VCPU_EVENTS 1233:Extended by: KVM_CAP_INTR_SHADOW 1234:Architectures: x86, arm64 1235:Type: vcpu ioctl 1236:Parameters: struct kvm_vcpu_events (in) 1237:Returns: 0 on success, -1 on error 1238 1239X86: 1240^^^^ 1241 1242Set pending exceptions, interrupts, and NMIs as well as related states of the 1243vcpu. 1244 1245See KVM_GET_VCPU_EVENTS for the data structure. 1246 1247Fields that may be modified asynchronously by running VCPUs can be excluded 1248from the update. These fields are nmi.pending, sipi_vector, smi.smm, 1249smi.pending. Keep the corresponding bits in the flags field cleared to 1250suppress overwriting the current in-kernel state. The bits are: 1251 1252=============================== ================================== 1253KVM_VCPUEVENT_VALID_NMI_PENDING transfer nmi.pending to the kernel 1254KVM_VCPUEVENT_VALID_SIPI_VECTOR transfer sipi_vector 1255KVM_VCPUEVENT_VALID_SMM transfer the smi sub-struct. 1256=============================== ================================== 1257 1258If KVM_CAP_INTR_SHADOW is available, KVM_VCPUEVENT_VALID_SHADOW can be set in 1259the flags field to signal that interrupt.shadow contains a valid state and 1260shall be written into the VCPU. 1261 1262KVM_VCPUEVENT_VALID_SMM can only be set if KVM_CAP_X86_SMM is available. 1263 1264If KVM_CAP_EXCEPTION_PAYLOAD is enabled, KVM_VCPUEVENT_VALID_PAYLOAD 1265can be set in the flags field to signal that the 1266exception_has_payload, exception_payload, and exception.pending fields 1267contain a valid state and shall be written into the VCPU. 1268 1269If KVM_CAP_X86_TRIPLE_FAULT_EVENT is enabled, KVM_VCPUEVENT_VALID_TRIPLE_FAULT 1270can be set in flags field to signal that the triple_fault field contains 1271a valid state and shall be written into the VCPU. 1272 1273ARM64: 1274^^^^^^ 1275 1276User space may need to inject several types of events to the guest. 1277 1278Set the pending SError exception state for this VCPU. It is not possible to 1279'cancel' an Serror that has been made pending. 1280 1281If the guest performed an access to I/O memory which could not be handled by 1282userspace, for example because of missing instruction syndrome decode 1283information or because there is no device mapped at the accessed IPA, then 1284userspace can ask the kernel to inject an external abort using the address 1285from the exiting fault on the VCPU. It is a programming error to set 1286ext_dabt_pending after an exit which was not either KVM_EXIT_MMIO or 1287KVM_EXIT_ARM_NISV. This feature is only available if the system supports 1288KVM_CAP_ARM_INJECT_EXT_DABT. This is a helper which provides commonality in 1289how userspace reports accesses for the above cases to guests, across different 1290userspace implementations. Nevertheless, userspace can still emulate all Arm 1291exceptions by manipulating individual registers using the KVM_SET_ONE_REG API. 1292 1293See KVM_GET_VCPU_EVENTS for the data structure. 1294 1295 12964.33 KVM_GET_DEBUGREGS 1297---------------------- 1298 1299:Capability: KVM_CAP_DEBUGREGS 1300:Architectures: x86 1301:Type: vm ioctl 1302:Parameters: struct kvm_debugregs (out) 1303:Returns: 0 on success, -1 on error 1304 1305Reads debug registers from the vcpu. 1306 1307:: 1308 1309 struct kvm_debugregs { 1310 __u64 db[4]; 1311 __u64 dr6; 1312 __u64 dr7; 1313 __u64 flags; 1314 __u64 reserved[9]; 1315 }; 1316 1317 13184.34 KVM_SET_DEBUGREGS 1319---------------------- 1320 1321:Capability: KVM_CAP_DEBUGREGS 1322:Architectures: x86 1323:Type: vm ioctl 1324:Parameters: struct kvm_debugregs (in) 1325:Returns: 0 on success, -1 on error 1326 1327Writes debug registers into the vcpu. 1328 1329See KVM_GET_DEBUGREGS for the data structure. The flags field is unused 1330yet and must be cleared on entry. 1331 1332 13334.35 KVM_SET_USER_MEMORY_REGION 1334------------------------------- 1335 1336:Capability: KVM_CAP_USER_MEMORY 1337:Architectures: all 1338:Type: vm ioctl 1339:Parameters: struct kvm_userspace_memory_region (in) 1340:Returns: 0 on success, -1 on error 1341 1342:: 1343 1344 struct kvm_userspace_memory_region { 1345 __u32 slot; 1346 __u32 flags; 1347 __u64 guest_phys_addr; 1348 __u64 memory_size; /* bytes */ 1349 __u64 userspace_addr; /* start of the userspace allocated memory */ 1350 }; 1351 1352 /* for kvm_userspace_memory_region::flags */ 1353 #define KVM_MEM_LOG_DIRTY_PAGES (1UL << 0) 1354 #define KVM_MEM_READONLY (1UL << 1) 1355 1356This ioctl allows the user to create, modify or delete a guest physical 1357memory slot. Bits 0-15 of "slot" specify the slot id and this value 1358should be less than the maximum number of user memory slots supported per 1359VM. The maximum allowed slots can be queried using KVM_CAP_NR_MEMSLOTS. 1360Slots may not overlap in guest physical address space. 1361 1362If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of "slot" 1363specifies the address space which is being modified. They must be 1364less than the value that KVM_CHECK_EXTENSION returns for the 1365KVM_CAP_MULTI_ADDRESS_SPACE capability. Slots in separate address spaces 1366are unrelated; the restriction on overlapping slots only applies within 1367each address space. 1368 1369Deleting a slot is done by passing zero for memory_size. When changing 1370an existing slot, it may be moved in the guest physical memory space, 1371or its flags may be modified, but it may not be resized. 1372 1373Memory for the region is taken starting at the address denoted by the 1374field userspace_addr, which must point at user addressable memory for 1375the entire memory slot size. Any object may back this memory, including 1376anonymous memory, ordinary files, and hugetlbfs. 1377 1378On architectures that support a form of address tagging, userspace_addr must 1379be an untagged address. 1380 1381It is recommended that the lower 21 bits of guest_phys_addr and userspace_addr 1382be identical. This allows large pages in the guest to be backed by large 1383pages in the host. 1384 1385The flags field supports two flags: KVM_MEM_LOG_DIRTY_PAGES and 1386KVM_MEM_READONLY. The former can be set to instruct KVM to keep track of 1387writes to memory within the slot. See KVM_GET_DIRTY_LOG ioctl to know how to 1388use it. The latter can be set, if KVM_CAP_READONLY_MEM capability allows it, 1389to make a new slot read-only. In this case, writes to this memory will be 1390posted to userspace as KVM_EXIT_MMIO exits. 1391 1392When the KVM_CAP_SYNC_MMU capability is available, changes in the backing of 1393the memory region are automatically reflected into the guest. For example, an 1394mmap() that affects the region will be made visible immediately. Another 1395example is madvise(MADV_DROP). 1396 1397Note: On arm64, a write generated by the page-table walker (to update 1398the Access and Dirty flags, for example) never results in a 1399KVM_EXIT_MMIO exit when the slot has the KVM_MEM_READONLY flag. This 1400is because KVM cannot provide the data that would be written by the 1401page-table walker, making it impossible to emulate the access. 1402Instead, an abort (data abort if the cause of the page-table update 1403was a load or a store, instruction abort if it was an instruction 1404fetch) is injected in the guest. 1405 1406S390: 1407^^^^^ 1408 1409Returns -EINVAL if the VM has the KVM_VM_S390_UCONTROL flag set. 1410Returns -EINVAL if called on a protected VM. 1411 14124.36 KVM_SET_TSS_ADDR 1413--------------------- 1414 1415:Capability: KVM_CAP_SET_TSS_ADDR 1416:Architectures: x86 1417:Type: vm ioctl 1418:Parameters: unsigned long tss_address (in) 1419:Returns: 0 on success, -1 on error 1420 1421This ioctl defines the physical address of a three-page region in the guest 1422physical address space. The region must be within the first 4GB of the 1423guest physical address space and must not conflict with any memory slot 1424or any mmio address. The guest may malfunction if it accesses this memory 1425region. 1426 1427This ioctl is required on Intel-based hosts. This is needed on Intel hardware 1428because of a quirk in the virtualization implementation (see the internals 1429documentation when it pops into existence). 1430 1431 14324.37 KVM_ENABLE_CAP 1433------------------- 1434 1435:Capability: KVM_CAP_ENABLE_CAP 1436:Architectures: mips, ppc, s390, x86, loongarch 1437:Type: vcpu ioctl 1438:Parameters: struct kvm_enable_cap (in) 1439:Returns: 0 on success; -1 on error 1440 1441:Capability: KVM_CAP_ENABLE_CAP_VM 1442:Architectures: all 1443:Type: vm ioctl 1444:Parameters: struct kvm_enable_cap (in) 1445:Returns: 0 on success; -1 on error 1446 1447.. note:: 1448 1449 Not all extensions are enabled by default. Using this ioctl the application 1450 can enable an extension, making it available to the guest. 1451 1452On systems that do not support this ioctl, it always fails. On systems that 1453do support it, it only works for extensions that are supported for enablement. 1454 1455To check if a capability can be enabled, the KVM_CHECK_EXTENSION ioctl should 1456be used. 1457 1458:: 1459 1460 struct kvm_enable_cap { 1461 /* in */ 1462 __u32 cap; 1463 1464The capability that is supposed to get enabled. 1465 1466:: 1467 1468 __u32 flags; 1469 1470A bitfield indicating future enhancements. Has to be 0 for now. 1471 1472:: 1473 1474 __u64 args[4]; 1475 1476Arguments for enabling a feature. If a feature needs initial values to 1477function properly, this is the place to put them. 1478 1479:: 1480 1481 __u8 pad[64]; 1482 }; 1483 1484The vcpu ioctl should be used for vcpu-specific capabilities, the vm ioctl 1485for vm-wide capabilities. 1486 14874.38 KVM_GET_MP_STATE 1488--------------------- 1489 1490:Capability: KVM_CAP_MP_STATE 1491:Architectures: x86, s390, arm64, riscv, loongarch 1492:Type: vcpu ioctl 1493:Parameters: struct kvm_mp_state (out) 1494:Returns: 0 on success; -1 on error 1495 1496:: 1497 1498 struct kvm_mp_state { 1499 __u32 mp_state; 1500 }; 1501 1502Returns the vcpu's current "multiprocessing state" (though also valid on 1503uniprocessor guests). 1504 1505Possible values are: 1506 1507 ========================== =============================================== 1508 KVM_MP_STATE_RUNNABLE the vcpu is currently running 1509 [x86,arm64,riscv,loongarch] 1510 KVM_MP_STATE_UNINITIALIZED the vcpu is an application processor (AP) 1511 which has not yet received an INIT signal [x86] 1512 KVM_MP_STATE_INIT_RECEIVED the vcpu has received an INIT signal, and is 1513 now ready for a SIPI [x86] 1514 KVM_MP_STATE_HALTED the vcpu has executed a HLT instruction and 1515 is waiting for an interrupt [x86] 1516 KVM_MP_STATE_SIPI_RECEIVED the vcpu has just received a SIPI (vector 1517 accessible via KVM_GET_VCPU_EVENTS) [x86] 1518 KVM_MP_STATE_STOPPED the vcpu is stopped [s390,arm64,riscv] 1519 KVM_MP_STATE_CHECK_STOP the vcpu is in a special error state [s390] 1520 KVM_MP_STATE_OPERATING the vcpu is operating (running or halted) 1521 [s390] 1522 KVM_MP_STATE_LOAD the vcpu is in a special load/startup state 1523 [s390] 1524 KVM_MP_STATE_SUSPENDED the vcpu is in a suspend state and is waiting 1525 for a wakeup event [arm64] 1526 ========================== =============================================== 1527 1528On x86, this ioctl is only useful after KVM_CREATE_IRQCHIP. Without an 1529in-kernel irqchip, the multiprocessing state must be maintained by userspace on 1530these architectures. 1531 1532For arm64: 1533^^^^^^^^^^ 1534 1535If a vCPU is in the KVM_MP_STATE_SUSPENDED state, KVM will emulate the 1536architectural execution of a WFI instruction. 1537 1538If a wakeup event is recognized, KVM will exit to userspace with a 1539KVM_SYSTEM_EVENT exit, where the event type is KVM_SYSTEM_EVENT_WAKEUP. If 1540userspace wants to honor the wakeup, it must set the vCPU's MP state to 1541KVM_MP_STATE_RUNNABLE. If it does not, KVM will continue to await a wakeup 1542event in subsequent calls to KVM_RUN. 1543 1544.. warning:: 1545 1546 If userspace intends to keep the vCPU in a SUSPENDED state, it is 1547 strongly recommended that userspace take action to suppress the 1548 wakeup event (such as masking an interrupt). Otherwise, subsequent 1549 calls to KVM_RUN will immediately exit with a KVM_SYSTEM_EVENT_WAKEUP 1550 event and inadvertently waste CPU cycles. 1551 1552 Additionally, if userspace takes action to suppress a wakeup event, 1553 it is strongly recommended that it also restores the vCPU to its 1554 original state when the vCPU is made RUNNABLE again. For example, 1555 if userspace masked a pending interrupt to suppress the wakeup, 1556 the interrupt should be unmasked before returning control to the 1557 guest. 1558 1559For riscv: 1560^^^^^^^^^^ 1561 1562The only states that are valid are KVM_MP_STATE_STOPPED and 1563KVM_MP_STATE_RUNNABLE which reflect if the vcpu is paused or not. 1564 1565On LoongArch, only the KVM_MP_STATE_RUNNABLE state is used to reflect 1566whether the vcpu is runnable. 1567 15684.39 KVM_SET_MP_STATE 1569--------------------- 1570 1571:Capability: KVM_CAP_MP_STATE 1572:Architectures: x86, s390, arm64, riscv, loongarch 1573:Type: vcpu ioctl 1574:Parameters: struct kvm_mp_state (in) 1575:Returns: 0 on success; -1 on error 1576 1577Sets the vcpu's current "multiprocessing state"; see KVM_GET_MP_STATE for 1578arguments. 1579 1580On x86, this ioctl is only useful after KVM_CREATE_IRQCHIP. Without an 1581in-kernel irqchip, the multiprocessing state must be maintained by userspace on 1582these architectures. 1583 1584For arm64/riscv: 1585^^^^^^^^^^^^^^^^ 1586 1587The only states that are valid are KVM_MP_STATE_STOPPED and 1588KVM_MP_STATE_RUNNABLE which reflect if the vcpu should be paused or not. 1589 1590On LoongArch, only the KVM_MP_STATE_RUNNABLE state is used to reflect 1591whether the vcpu is runnable. 1592 15934.40 KVM_SET_IDENTITY_MAP_ADDR 1594------------------------------ 1595 1596:Capability: KVM_CAP_SET_IDENTITY_MAP_ADDR 1597:Architectures: x86 1598:Type: vm ioctl 1599:Parameters: unsigned long identity (in) 1600:Returns: 0 on success, -1 on error 1601 1602This ioctl defines the physical address of a one-page region in the guest 1603physical address space. The region must be within the first 4GB of the 1604guest physical address space and must not conflict with any memory slot 1605or any mmio address. The guest may malfunction if it accesses this memory 1606region. 1607 1608Setting the address to 0 will result in resetting the address to its default 1609(0xfffbc000). 1610 1611This ioctl is required on Intel-based hosts. This is needed on Intel hardware 1612because of a quirk in the virtualization implementation (see the internals 1613documentation when it pops into existence). 1614 1615Fails if any VCPU has already been created. 1616 16174.41 KVM_SET_BOOT_CPU_ID 1618------------------------ 1619 1620:Capability: KVM_CAP_SET_BOOT_CPU_ID 1621:Architectures: x86 1622:Type: vm ioctl 1623:Parameters: unsigned long vcpu_id 1624:Returns: 0 on success, -1 on error 1625 1626Define which vcpu is the Bootstrap Processor (BSP). Values are the same 1627as the vcpu id in KVM_CREATE_VCPU. If this ioctl is not called, the default 1628is vcpu 0. This ioctl has to be called before vcpu creation, 1629otherwise it will return EBUSY error. 1630 1631 16324.42 KVM_GET_XSAVE 1633------------------ 1634 1635:Capability: KVM_CAP_XSAVE 1636:Architectures: x86 1637:Type: vcpu ioctl 1638:Parameters: struct kvm_xsave (out) 1639:Returns: 0 on success, -1 on error 1640 1641 1642:: 1643 1644 struct kvm_xsave { 1645 __u32 region[1024]; 1646 __u32 extra[0]; 1647 }; 1648 1649This ioctl would copy current vcpu's xsave struct to the userspace. 1650 1651 16524.43 KVM_SET_XSAVE 1653------------------ 1654 1655:Capability: KVM_CAP_XSAVE and KVM_CAP_XSAVE2 1656:Architectures: x86 1657:Type: vcpu ioctl 1658:Parameters: struct kvm_xsave (in) 1659:Returns: 0 on success, -1 on error 1660 1661:: 1662 1663 1664 struct kvm_xsave { 1665 __u32 region[1024]; 1666 __u32 extra[0]; 1667 }; 1668 1669This ioctl would copy userspace's xsave struct to the kernel. It copies 1670as many bytes as are returned by KVM_CHECK_EXTENSION(KVM_CAP_XSAVE2), 1671when invoked on the vm file descriptor. The size value returned by 1672KVM_CHECK_EXTENSION(KVM_CAP_XSAVE2) will always be at least 4096. 1673Currently, it is only greater than 4096 if a dynamic feature has been 1674enabled with ``arch_prctl()``, but this may change in the future. 1675 1676The offsets of the state save areas in struct kvm_xsave follow the 1677contents of CPUID leaf 0xD on the host. 1678 1679 16804.44 KVM_GET_XCRS 1681----------------- 1682 1683:Capability: KVM_CAP_XCRS 1684:Architectures: x86 1685:Type: vcpu ioctl 1686:Parameters: struct kvm_xcrs (out) 1687:Returns: 0 on success, -1 on error 1688 1689:: 1690 1691 struct kvm_xcr { 1692 __u32 xcr; 1693 __u32 reserved; 1694 __u64 value; 1695 }; 1696 1697 struct kvm_xcrs { 1698 __u32 nr_xcrs; 1699 __u32 flags; 1700 struct kvm_xcr xcrs[KVM_MAX_XCRS]; 1701 __u64 padding[16]; 1702 }; 1703 1704This ioctl would copy current vcpu's xcrs to the userspace. 1705 1706 17074.45 KVM_SET_XCRS 1708----------------- 1709 1710:Capability: KVM_CAP_XCRS 1711:Architectures: x86 1712:Type: vcpu ioctl 1713:Parameters: struct kvm_xcrs (in) 1714:Returns: 0 on success, -1 on error 1715 1716:: 1717 1718 struct kvm_xcr { 1719 __u32 xcr; 1720 __u32 reserved; 1721 __u64 value; 1722 }; 1723 1724 struct kvm_xcrs { 1725 __u32 nr_xcrs; 1726 __u32 flags; 1727 struct kvm_xcr xcrs[KVM_MAX_XCRS]; 1728 __u64 padding[16]; 1729 }; 1730 1731This ioctl would set vcpu's xcr to the value userspace specified. 1732 1733 17344.46 KVM_GET_SUPPORTED_CPUID 1735---------------------------- 1736 1737:Capability: KVM_CAP_EXT_CPUID 1738:Architectures: x86 1739:Type: system ioctl 1740:Parameters: struct kvm_cpuid2 (in/out) 1741:Returns: 0 on success, -1 on error 1742 1743:: 1744 1745 struct kvm_cpuid2 { 1746 __u32 nent; 1747 __u32 padding; 1748 struct kvm_cpuid_entry2 entries[0]; 1749 }; 1750 1751 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX BIT(0) 1752 #define KVM_CPUID_FLAG_STATEFUL_FUNC BIT(1) /* deprecated */ 1753 #define KVM_CPUID_FLAG_STATE_READ_NEXT BIT(2) /* deprecated */ 1754 1755 struct kvm_cpuid_entry2 { 1756 __u32 function; 1757 __u32 index; 1758 __u32 flags; 1759 __u32 eax; 1760 __u32 ebx; 1761 __u32 ecx; 1762 __u32 edx; 1763 __u32 padding[3]; 1764 }; 1765 1766This ioctl returns x86 cpuid features which are supported by both the 1767hardware and kvm in its default configuration. Userspace can use the 1768information returned by this ioctl to construct cpuid information (for 1769KVM_SET_CPUID2) that is consistent with hardware, kernel, and 1770userspace capabilities, and with user requirements (for example, the 1771user may wish to constrain cpuid to emulate older hardware, or for 1772feature consistency across a cluster). 1773 1774Dynamically-enabled feature bits need to be requested with 1775``arch_prctl()`` before calling this ioctl. Feature bits that have not 1776been requested are excluded from the result. 1777 1778Note that certain capabilities, such as KVM_CAP_X86_DISABLE_EXITS, may 1779expose cpuid features (e.g. MONITOR) which are not supported by kvm in 1780its default configuration. If userspace enables such capabilities, it 1781is responsible for modifying the results of this ioctl appropriately. 1782 1783Userspace invokes KVM_GET_SUPPORTED_CPUID by passing a kvm_cpuid2 structure 1784with the 'nent' field indicating the number of entries in the variable-size 1785array 'entries'. If the number of entries is too low to describe the cpu 1786capabilities, an error (E2BIG) is returned. If the number is too high, 1787the 'nent' field is adjusted and an error (ENOMEM) is returned. If the 1788number is just right, the 'nent' field is adjusted to the number of valid 1789entries in the 'entries' array, which is then filled. 1790 1791The entries returned are the host cpuid as returned by the cpuid instruction, 1792with unknown or unsupported features masked out. Some features (for example, 1793x2apic), may not be present in the host cpu, but are exposed by kvm if it can 1794emulate them efficiently. The fields in each entry are defined as follows: 1795 1796 function: 1797 the eax value used to obtain the entry 1798 1799 index: 1800 the ecx value used to obtain the entry (for entries that are 1801 affected by ecx) 1802 1803 flags: 1804 an OR of zero or more of the following: 1805 1806 KVM_CPUID_FLAG_SIGNIFCANT_INDEX: 1807 if the index field is valid 1808 1809 eax, ebx, ecx, edx: 1810 the values returned by the cpuid instruction for 1811 this function/index combination 1812 1813The TSC deadline timer feature (CPUID leaf 1, ecx[24]) is always returned 1814as false, since the feature depends on KVM_CREATE_IRQCHIP for local APIC 1815support. Instead it is reported via:: 1816 1817 ioctl(KVM_CHECK_EXTENSION, KVM_CAP_TSC_DEADLINE_TIMER) 1818 1819if that returns true and you use KVM_CREATE_IRQCHIP, or if you emulate the 1820feature in userspace, then you can enable the feature for KVM_SET_CPUID2. 1821 1822 18234.47 KVM_PPC_GET_PVINFO 1824----------------------- 1825 1826:Capability: KVM_CAP_PPC_GET_PVINFO 1827:Architectures: ppc 1828:Type: vm ioctl 1829:Parameters: struct kvm_ppc_pvinfo (out) 1830:Returns: 0 on success, !0 on error 1831 1832:: 1833 1834 struct kvm_ppc_pvinfo { 1835 __u32 flags; 1836 __u32 hcall[4]; 1837 __u8 pad[108]; 1838 }; 1839 1840This ioctl fetches PV specific information that need to be passed to the guest 1841using the device tree or other means from vm context. 1842 1843The hcall array defines 4 instructions that make up a hypercall. 1844 1845If any additional field gets added to this structure later on, a bit for that 1846additional piece of information will be set in the flags bitmap. 1847 1848The flags bitmap is defined as:: 1849 1850 /* the host supports the ePAPR idle hcall 1851 #define KVM_PPC_PVINFO_FLAGS_EV_IDLE (1<<0) 1852 18534.52 KVM_SET_GSI_ROUTING 1854------------------------ 1855 1856:Capability: KVM_CAP_IRQ_ROUTING 1857:Architectures: x86 s390 arm64 1858:Type: vm ioctl 1859:Parameters: struct kvm_irq_routing (in) 1860:Returns: 0 on success, -1 on error 1861 1862Sets the GSI routing table entries, overwriting any previously set entries. 1863 1864On arm64, GSI routing has the following limitation: 1865 1866- GSI routing does not apply to KVM_IRQ_LINE but only to KVM_IRQFD. 1867 1868:: 1869 1870 struct kvm_irq_routing { 1871 __u32 nr; 1872 __u32 flags; 1873 struct kvm_irq_routing_entry entries[0]; 1874 }; 1875 1876No flags are specified so far, the corresponding field must be set to zero. 1877 1878:: 1879 1880 struct kvm_irq_routing_entry { 1881 __u32 gsi; 1882 __u32 type; 1883 __u32 flags; 1884 __u32 pad; 1885 union { 1886 struct kvm_irq_routing_irqchip irqchip; 1887 struct kvm_irq_routing_msi msi; 1888 struct kvm_irq_routing_s390_adapter adapter; 1889 struct kvm_irq_routing_hv_sint hv_sint; 1890 struct kvm_irq_routing_xen_evtchn xen_evtchn; 1891 __u32 pad[8]; 1892 } u; 1893 }; 1894 1895 /* gsi routing entry types */ 1896 #define KVM_IRQ_ROUTING_IRQCHIP 1 1897 #define KVM_IRQ_ROUTING_MSI 2 1898 #define KVM_IRQ_ROUTING_S390_ADAPTER 3 1899 #define KVM_IRQ_ROUTING_HV_SINT 4 1900 #define KVM_IRQ_ROUTING_XEN_EVTCHN 5 1901 1902flags: 1903 1904- KVM_MSI_VALID_DEVID: used along with KVM_IRQ_ROUTING_MSI routing entry 1905 type, specifies that the devid field contains a valid value. The per-VM 1906 KVM_CAP_MSI_DEVID capability advertises the requirement to provide 1907 the device ID. If this capability is not available, userspace should 1908 never set the KVM_MSI_VALID_DEVID flag as the ioctl might fail. 1909- zero otherwise 1910 1911:: 1912 1913 struct kvm_irq_routing_irqchip { 1914 __u32 irqchip; 1915 __u32 pin; 1916 }; 1917 1918 struct kvm_irq_routing_msi { 1919 __u32 address_lo; 1920 __u32 address_hi; 1921 __u32 data; 1922 union { 1923 __u32 pad; 1924 __u32 devid; 1925 }; 1926 }; 1927 1928If KVM_MSI_VALID_DEVID is set, devid contains a unique device identifier 1929for the device that wrote the MSI message. For PCI, this is usually a 1930BDF identifier in the lower 16 bits. 1931 1932On x86, address_hi is ignored unless the KVM_X2APIC_API_USE_32BIT_IDS 1933feature of KVM_CAP_X2APIC_API capability is enabled. If it is enabled, 1934address_hi bits 31-8 provide bits 31-8 of the destination id. Bits 7-0 of 1935address_hi must be zero. 1936 1937:: 1938 1939 struct kvm_irq_routing_s390_adapter { 1940 __u64 ind_addr; 1941 __u64 summary_addr; 1942 __u64 ind_offset; 1943 __u32 summary_offset; 1944 __u32 adapter_id; 1945 }; 1946 1947 struct kvm_irq_routing_hv_sint { 1948 __u32 vcpu; 1949 __u32 sint; 1950 }; 1951 1952 struct kvm_irq_routing_xen_evtchn { 1953 __u32 port; 1954 __u32 vcpu; 1955 __u32 priority; 1956 }; 1957 1958 1959When KVM_CAP_XEN_HVM includes the KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL bit 1960in its indication of supported features, routing to Xen event channels 1961is supported. Although the priority field is present, only the value 1962KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL is supported, which means delivery by 19632 level event channels. FIFO event channel support may be added in 1964the future. 1965 1966 19674.55 KVM_SET_TSC_KHZ 1968-------------------- 1969 1970:Capability: KVM_CAP_TSC_CONTROL / KVM_CAP_VM_TSC_CONTROL 1971:Architectures: x86 1972:Type: vcpu ioctl / vm ioctl 1973:Parameters: virtual tsc_khz 1974:Returns: 0 on success, -1 on error 1975 1976Specifies the tsc frequency for the virtual machine. The unit of the 1977frequency is KHz. 1978 1979If the KVM_CAP_VM_TSC_CONTROL capability is advertised, this can also 1980be used as a vm ioctl to set the initial tsc frequency of subsequently 1981created vCPUs. 1982 19834.56 KVM_GET_TSC_KHZ 1984-------------------- 1985 1986:Capability: KVM_CAP_GET_TSC_KHZ / KVM_CAP_VM_TSC_CONTROL 1987:Architectures: x86 1988:Type: vcpu ioctl / vm ioctl 1989:Parameters: none 1990:Returns: virtual tsc-khz on success, negative value on error 1991 1992Returns the tsc frequency of the guest. The unit of the return value is 1993KHz. If the host has unstable tsc this ioctl returns -EIO instead as an 1994error. 1995 1996 19974.57 KVM_GET_LAPIC 1998------------------ 1999 2000:Capability: KVM_CAP_IRQCHIP 2001:Architectures: x86 2002:Type: vcpu ioctl 2003:Parameters: struct kvm_lapic_state (out) 2004:Returns: 0 on success, -1 on error 2005 2006:: 2007 2008 #define KVM_APIC_REG_SIZE 0x400 2009 struct kvm_lapic_state { 2010 char regs[KVM_APIC_REG_SIZE]; 2011 }; 2012 2013Reads the Local APIC registers and copies them into the input argument. The 2014data format and layout are the same as documented in the architecture manual. 2015 2016If KVM_X2APIC_API_USE_32BIT_IDS feature of KVM_CAP_X2APIC_API is 2017enabled, then the format of APIC_ID register depends on the APIC mode 2018(reported by MSR_IA32_APICBASE) of its VCPU. x2APIC stores APIC ID in 2019the APIC_ID register (bytes 32-35). xAPIC only allows an 8-bit APIC ID 2020which is stored in bits 31-24 of the APIC register, or equivalently in 2021byte 35 of struct kvm_lapic_state's regs field. KVM_GET_LAPIC must then 2022be called after MSR_IA32_APICBASE has been set with KVM_SET_MSR. 2023 2024If KVM_X2APIC_API_USE_32BIT_IDS feature is disabled, struct kvm_lapic_state 2025always uses xAPIC format. 2026 2027 20284.58 KVM_SET_LAPIC 2029------------------ 2030 2031:Capability: KVM_CAP_IRQCHIP 2032:Architectures: x86 2033:Type: vcpu ioctl 2034:Parameters: struct kvm_lapic_state (in) 2035:Returns: 0 on success, -1 on error 2036 2037:: 2038 2039 #define KVM_APIC_REG_SIZE 0x400 2040 struct kvm_lapic_state { 2041 char regs[KVM_APIC_REG_SIZE]; 2042 }; 2043 2044Copies the input argument into the Local APIC registers. The data format 2045and layout are the same as documented in the architecture manual. 2046 2047The format of the APIC ID register (bytes 32-35 of struct kvm_lapic_state's 2048regs field) depends on the state of the KVM_CAP_X2APIC_API capability. 2049See the note in KVM_GET_LAPIC. 2050 2051 20524.59 KVM_IOEVENTFD 2053------------------ 2054 2055:Capability: KVM_CAP_IOEVENTFD 2056:Architectures: all 2057:Type: vm ioctl 2058:Parameters: struct kvm_ioeventfd (in) 2059:Returns: 0 on success, !0 on error 2060 2061This ioctl attaches or detaches an ioeventfd to a legal pio/mmio address 2062within the guest. A guest write in the registered address will signal the 2063provided event instead of triggering an exit. 2064 2065:: 2066 2067 struct kvm_ioeventfd { 2068 __u64 datamatch; 2069 __u64 addr; /* legal pio/mmio address */ 2070 __u32 len; /* 0, 1, 2, 4, or 8 bytes */ 2071 __s32 fd; 2072 __u32 flags; 2073 __u8 pad[36]; 2074 }; 2075 2076For the special case of virtio-ccw devices on s390, the ioevent is matched 2077to a subchannel/virtqueue tuple instead. 2078 2079The following flags are defined:: 2080 2081 #define KVM_IOEVENTFD_FLAG_DATAMATCH (1 << kvm_ioeventfd_flag_nr_datamatch) 2082 #define KVM_IOEVENTFD_FLAG_PIO (1 << kvm_ioeventfd_flag_nr_pio) 2083 #define KVM_IOEVENTFD_FLAG_DEASSIGN (1 << kvm_ioeventfd_flag_nr_deassign) 2084 #define KVM_IOEVENTFD_FLAG_VIRTIO_CCW_NOTIFY \ 2085 (1 << kvm_ioeventfd_flag_nr_virtio_ccw_notify) 2086 2087If datamatch flag is set, the event will be signaled only if the written value 2088to the registered address is equal to datamatch in struct kvm_ioeventfd. 2089 2090For virtio-ccw devices, addr contains the subchannel id and datamatch the 2091virtqueue index. 2092 2093With KVM_CAP_IOEVENTFD_ANY_LENGTH, a zero length ioeventfd is allowed, and 2094the kernel will ignore the length of guest write and may get a faster vmexit. 2095The speedup may only apply to specific architectures, but the ioeventfd will 2096work anyway. 2097 20984.60 KVM_DIRTY_TLB 2099------------------ 2100 2101:Capability: KVM_CAP_SW_TLB 2102:Architectures: ppc 2103:Type: vcpu ioctl 2104:Parameters: struct kvm_dirty_tlb (in) 2105:Returns: 0 on success, -1 on error 2106 2107:: 2108 2109 struct kvm_dirty_tlb { 2110 __u64 bitmap; 2111 __u32 num_dirty; 2112 }; 2113 2114This must be called whenever userspace has changed an entry in the shared 2115TLB, prior to calling KVM_RUN on the associated vcpu. 2116 2117The "bitmap" field is the userspace address of an array. This array 2118consists of a number of bits, equal to the total number of TLB entries as 2119determined by the last successful call to KVM_CONFIG_TLB, rounded up to the 2120nearest multiple of 64. 2121 2122Each bit corresponds to one TLB entry, ordered the same as in the shared TLB 2123array. 2124 2125The array is little-endian: the bit 0 is the least significant bit of the 2126first byte, bit 8 is the least significant bit of the second byte, etc. 2127This avoids any complications with differing word sizes. 2128 2129The "num_dirty" field is a performance hint for KVM to determine whether it 2130should skip processing the bitmap and just invalidate everything. It must 2131be set to the number of set bits in the bitmap. 2132 2133 21344.62 KVM_CREATE_SPAPR_TCE 2135------------------------- 2136 2137:Capability: KVM_CAP_SPAPR_TCE 2138:Architectures: powerpc 2139:Type: vm ioctl 2140:Parameters: struct kvm_create_spapr_tce (in) 2141:Returns: file descriptor for manipulating the created TCE table 2142 2143This creates a virtual TCE (translation control entry) table, which 2144is an IOMMU for PAPR-style virtual I/O. It is used to translate 2145logical addresses used in virtual I/O into guest physical addresses, 2146and provides a scatter/gather capability for PAPR virtual I/O. 2147 2148:: 2149 2150 /* for KVM_CAP_SPAPR_TCE */ 2151 struct kvm_create_spapr_tce { 2152 __u64 liobn; 2153 __u32 window_size; 2154 }; 2155 2156The liobn field gives the logical IO bus number for which to create a 2157TCE table. The window_size field specifies the size of the DMA window 2158which this TCE table will translate - the table will contain one 64 2159bit TCE entry for every 4kiB of the DMA window. 2160 2161When the guest issues an H_PUT_TCE hcall on a liobn for which a TCE 2162table has been created using this ioctl(), the kernel will handle it 2163in real mode, updating the TCE table. H_PUT_TCE calls for other 2164liobns will cause a vm exit and must be handled by userspace. 2165 2166The return value is a file descriptor which can be passed to mmap(2) 2167to map the created TCE table into userspace. This lets userspace read 2168the entries written by kernel-handled H_PUT_TCE calls, and also lets 2169userspace update the TCE table directly which is useful in some 2170circumstances. 2171 2172 21734.63 KVM_ALLOCATE_RMA 2174--------------------- 2175 2176:Capability: KVM_CAP_PPC_RMA 2177:Architectures: powerpc 2178:Type: vm ioctl 2179:Parameters: struct kvm_allocate_rma (out) 2180:Returns: file descriptor for mapping the allocated RMA 2181 2182This allocates a Real Mode Area (RMA) from the pool allocated at boot 2183time by the kernel. An RMA is a physically-contiguous, aligned region 2184of memory used on older POWER processors to provide the memory which 2185will be accessed by real-mode (MMU off) accesses in a KVM guest. 2186POWER processors support a set of sizes for the RMA that usually 2187includes 64MB, 128MB, 256MB and some larger powers of two. 2188 2189:: 2190 2191 /* for KVM_ALLOCATE_RMA */ 2192 struct kvm_allocate_rma { 2193 __u64 rma_size; 2194 }; 2195 2196The return value is a file descriptor which can be passed to mmap(2) 2197to map the allocated RMA into userspace. The mapped area can then be 2198passed to the KVM_SET_USER_MEMORY_REGION ioctl to establish it as the 2199RMA for a virtual machine. The size of the RMA in bytes (which is 2200fixed at host kernel boot time) is returned in the rma_size field of 2201the argument structure. 2202 2203The KVM_CAP_PPC_RMA capability is 1 or 2 if the KVM_ALLOCATE_RMA ioctl 2204is supported; 2 if the processor requires all virtual machines to have 2205an RMA, or 1 if the processor can use an RMA but doesn't require it, 2206because it supports the Virtual RMA (VRMA) facility. 2207 2208 22094.64 KVM_NMI 2210------------ 2211 2212:Capability: KVM_CAP_USER_NMI 2213:Architectures: x86 2214:Type: vcpu ioctl 2215:Parameters: none 2216:Returns: 0 on success, -1 on error 2217 2218Queues an NMI on the thread's vcpu. Note this is well defined only 2219when KVM_CREATE_IRQCHIP has not been called, since this is an interface 2220between the virtual cpu core and virtual local APIC. After KVM_CREATE_IRQCHIP 2221has been called, this interface is completely emulated within the kernel. 2222 2223To use this to emulate the LINT1 input with KVM_CREATE_IRQCHIP, use the 2224following algorithm: 2225 2226 - pause the vcpu 2227 - read the local APIC's state (KVM_GET_LAPIC) 2228 - check whether changing LINT1 will queue an NMI (see the LVT entry for LINT1) 2229 - if so, issue KVM_NMI 2230 - resume the vcpu 2231 2232Some guests configure the LINT1 NMI input to cause a panic, aiding in 2233debugging. 2234 2235 22364.65 KVM_S390_UCAS_MAP 2237---------------------- 2238 2239:Capability: KVM_CAP_S390_UCONTROL 2240:Architectures: s390 2241:Type: vcpu ioctl 2242:Parameters: struct kvm_s390_ucas_mapping (in) 2243:Returns: 0 in case of success 2244 2245The parameter is defined like this:: 2246 2247 struct kvm_s390_ucas_mapping { 2248 __u64 user_addr; 2249 __u64 vcpu_addr; 2250 __u64 length; 2251 }; 2252 2253This ioctl maps the memory at "user_addr" with the length "length" to 2254the vcpu's address space starting at "vcpu_addr". All parameters need to 2255be aligned by 1 megabyte. 2256 2257 22584.66 KVM_S390_UCAS_UNMAP 2259------------------------ 2260 2261:Capability: KVM_CAP_S390_UCONTROL 2262:Architectures: s390 2263:Type: vcpu ioctl 2264:Parameters: struct kvm_s390_ucas_mapping (in) 2265:Returns: 0 in case of success 2266 2267The parameter is defined like this:: 2268 2269 struct kvm_s390_ucas_mapping { 2270 __u64 user_addr; 2271 __u64 vcpu_addr; 2272 __u64 length; 2273 }; 2274 2275This ioctl unmaps the memory in the vcpu's address space starting at 2276"vcpu_addr" with the length "length". The field "user_addr" is ignored. 2277All parameters need to be aligned by 1 megabyte. 2278 2279 22804.67 KVM_S390_VCPU_FAULT 2281------------------------ 2282 2283:Capability: KVM_CAP_S390_UCONTROL 2284:Architectures: s390 2285:Type: vcpu ioctl 2286:Parameters: vcpu absolute address (in) 2287:Returns: 0 in case of success 2288 2289This call creates a page table entry on the virtual cpu's address space 2290(for user controlled virtual machines) or the virtual machine's address 2291space (for regular virtual machines). This only works for minor faults, 2292thus it's recommended to access subject memory page via the user page 2293table upfront. This is useful to handle validity intercepts for user 2294controlled virtual machines to fault in the virtual cpu's lowcore pages 2295prior to calling the KVM_RUN ioctl. 2296 2297 22984.68 KVM_SET_ONE_REG 2299-------------------- 2300 2301:Capability: KVM_CAP_ONE_REG 2302:Architectures: all 2303:Type: vcpu ioctl 2304:Parameters: struct kvm_one_reg (in) 2305:Returns: 0 on success, negative value on failure 2306 2307Errors: 2308 2309 ====== ============================================================ 2310 ENOENT no such register 2311 EINVAL invalid register ID, or no such register or used with VMs in 2312 protected virtualization mode on s390 2313 EPERM (arm64) register access not allowed before vcpu finalization 2314 EBUSY (riscv) changing register value not allowed after the vcpu 2315 has run at least once 2316 ====== ============================================================ 2317 2318(These error codes are indicative only: do not rely on a specific error 2319code being returned in a specific situation.) 2320 2321:: 2322 2323 struct kvm_one_reg { 2324 __u64 id; 2325 __u64 addr; 2326 }; 2327 2328Using this ioctl, a single vcpu register can be set to a specific value 2329defined by user space with the passed in struct kvm_one_reg, where id 2330refers to the register identifier as described below and addr is a pointer 2331to a variable with the respective size. There can be architecture agnostic 2332and architecture specific registers. Each have their own range of operation 2333and their own constants and width. To keep track of the implemented 2334registers, find a list below: 2335 2336 ======= =============================== ============ 2337 Arch Register Width (bits) 2338 ======= =============================== ============ 2339 PPC KVM_REG_PPC_HIOR 64 2340 PPC KVM_REG_PPC_IAC1 64 2341 PPC KVM_REG_PPC_IAC2 64 2342 PPC KVM_REG_PPC_IAC3 64 2343 PPC KVM_REG_PPC_IAC4 64 2344 PPC KVM_REG_PPC_DAC1 64 2345 PPC KVM_REG_PPC_DAC2 64 2346 PPC KVM_REG_PPC_DABR 64 2347 PPC KVM_REG_PPC_DSCR 64 2348 PPC KVM_REG_PPC_PURR 64 2349 PPC KVM_REG_PPC_SPURR 64 2350 PPC KVM_REG_PPC_DAR 64 2351 PPC KVM_REG_PPC_DSISR 32 2352 PPC KVM_REG_PPC_AMR 64 2353 PPC KVM_REG_PPC_UAMOR 64 2354 PPC KVM_REG_PPC_MMCR0 64 2355 PPC KVM_REG_PPC_MMCR1 64 2356 PPC KVM_REG_PPC_MMCRA 64 2357 PPC KVM_REG_PPC_MMCR2 64 2358 PPC KVM_REG_PPC_MMCRS 64 2359 PPC KVM_REG_PPC_MMCR3 64 2360 PPC KVM_REG_PPC_SIAR 64 2361 PPC KVM_REG_PPC_SDAR 64 2362 PPC KVM_REG_PPC_SIER 64 2363 PPC KVM_REG_PPC_SIER2 64 2364 PPC KVM_REG_PPC_SIER3 64 2365 PPC KVM_REG_PPC_PMC1 32 2366 PPC KVM_REG_PPC_PMC2 32 2367 PPC KVM_REG_PPC_PMC3 32 2368 PPC KVM_REG_PPC_PMC4 32 2369 PPC KVM_REG_PPC_PMC5 32 2370 PPC KVM_REG_PPC_PMC6 32 2371 PPC KVM_REG_PPC_PMC7 32 2372 PPC KVM_REG_PPC_PMC8 32 2373 PPC KVM_REG_PPC_FPR0 64 2374 ... 2375 PPC KVM_REG_PPC_FPR31 64 2376 PPC KVM_REG_PPC_VR0 128 2377 ... 2378 PPC KVM_REG_PPC_VR31 128 2379 PPC KVM_REG_PPC_VSR0 128 2380 ... 2381 PPC KVM_REG_PPC_VSR31 128 2382 PPC KVM_REG_PPC_FPSCR 64 2383 PPC KVM_REG_PPC_VSCR 32 2384 PPC KVM_REG_PPC_VPA_ADDR 64 2385 PPC KVM_REG_PPC_VPA_SLB 128 2386 PPC KVM_REG_PPC_VPA_DTL 128 2387 PPC KVM_REG_PPC_EPCR 32 2388 PPC KVM_REG_PPC_EPR 32 2389 PPC KVM_REG_PPC_TCR 32 2390 PPC KVM_REG_PPC_TSR 32 2391 PPC KVM_REG_PPC_OR_TSR 32 2392 PPC KVM_REG_PPC_CLEAR_TSR 32 2393 PPC KVM_REG_PPC_MAS0 32 2394 PPC KVM_REG_PPC_MAS1 32 2395 PPC KVM_REG_PPC_MAS2 64 2396 PPC KVM_REG_PPC_MAS7_3 64 2397 PPC KVM_REG_PPC_MAS4 32 2398 PPC KVM_REG_PPC_MAS6 32 2399 PPC KVM_REG_PPC_MMUCFG 32 2400 PPC KVM_REG_PPC_TLB0CFG 32 2401 PPC KVM_REG_PPC_TLB1CFG 32 2402 PPC KVM_REG_PPC_TLB2CFG 32 2403 PPC KVM_REG_PPC_TLB3CFG 32 2404 PPC KVM_REG_PPC_TLB0PS 32 2405 PPC KVM_REG_PPC_TLB1PS 32 2406 PPC KVM_REG_PPC_TLB2PS 32 2407 PPC KVM_REG_PPC_TLB3PS 32 2408 PPC KVM_REG_PPC_EPTCFG 32 2409 PPC KVM_REG_PPC_ICP_STATE 64 2410 PPC KVM_REG_PPC_VP_STATE 128 2411 PPC KVM_REG_PPC_TB_OFFSET 64 2412 PPC KVM_REG_PPC_SPMC1 32 2413 PPC KVM_REG_PPC_SPMC2 32 2414 PPC KVM_REG_PPC_IAMR 64 2415 PPC KVM_REG_PPC_TFHAR 64 2416 PPC KVM_REG_PPC_TFIAR 64 2417 PPC KVM_REG_PPC_TEXASR 64 2418 PPC KVM_REG_PPC_FSCR 64 2419 PPC KVM_REG_PPC_PSPB 32 2420 PPC KVM_REG_PPC_EBBHR 64 2421 PPC KVM_REG_PPC_EBBRR 64 2422 PPC KVM_REG_PPC_BESCR 64 2423 PPC KVM_REG_PPC_TAR 64 2424 PPC KVM_REG_PPC_DPDES 64 2425 PPC KVM_REG_PPC_DAWR 64 2426 PPC KVM_REG_PPC_DAWRX 64 2427 PPC KVM_REG_PPC_CIABR 64 2428 PPC KVM_REG_PPC_IC 64 2429 PPC KVM_REG_PPC_VTB 64 2430 PPC KVM_REG_PPC_CSIGR 64 2431 PPC KVM_REG_PPC_TACR 64 2432 PPC KVM_REG_PPC_TCSCR 64 2433 PPC KVM_REG_PPC_PID 64 2434 PPC KVM_REG_PPC_ACOP 64 2435 PPC KVM_REG_PPC_VRSAVE 32 2436 PPC KVM_REG_PPC_LPCR 32 2437 PPC KVM_REG_PPC_LPCR_64 64 2438 PPC KVM_REG_PPC_PPR 64 2439 PPC KVM_REG_PPC_ARCH_COMPAT 32 2440 PPC KVM_REG_PPC_DABRX 32 2441 PPC KVM_REG_PPC_WORT 64 2442 PPC KVM_REG_PPC_SPRG9 64 2443 PPC KVM_REG_PPC_DBSR 32 2444 PPC KVM_REG_PPC_TIDR 64 2445 PPC KVM_REG_PPC_PSSCR 64 2446 PPC KVM_REG_PPC_DEC_EXPIRY 64 2447 PPC KVM_REG_PPC_PTCR 64 2448 PPC KVM_REG_PPC_DAWR1 64 2449 PPC KVM_REG_PPC_DAWRX1 64 2450 PPC KVM_REG_PPC_TM_GPR0 64 2451 ... 2452 PPC KVM_REG_PPC_TM_GPR31 64 2453 PPC KVM_REG_PPC_TM_VSR0 128 2454 ... 2455 PPC KVM_REG_PPC_TM_VSR63 128 2456 PPC KVM_REG_PPC_TM_CR 64 2457 PPC KVM_REG_PPC_TM_LR 64 2458 PPC KVM_REG_PPC_TM_CTR 64 2459 PPC KVM_REG_PPC_TM_FPSCR 64 2460 PPC KVM_REG_PPC_TM_AMR 64 2461 PPC KVM_REG_PPC_TM_PPR 64 2462 PPC KVM_REG_PPC_TM_VRSAVE 64 2463 PPC KVM_REG_PPC_TM_VSCR 32 2464 PPC KVM_REG_PPC_TM_DSCR 64 2465 PPC KVM_REG_PPC_TM_TAR 64 2466 PPC KVM_REG_PPC_TM_XER 64 2467 2468 MIPS KVM_REG_MIPS_R0 64 2469 ... 2470 MIPS KVM_REG_MIPS_R31 64 2471 MIPS KVM_REG_MIPS_HI 64 2472 MIPS KVM_REG_MIPS_LO 64 2473 MIPS KVM_REG_MIPS_PC 64 2474 MIPS KVM_REG_MIPS_CP0_INDEX 32 2475 MIPS KVM_REG_MIPS_CP0_ENTRYLO0 64 2476 MIPS KVM_REG_MIPS_CP0_ENTRYLO1 64 2477 MIPS KVM_REG_MIPS_CP0_CONTEXT 64 2478 MIPS KVM_REG_MIPS_CP0_CONTEXTCONFIG 32 2479 MIPS KVM_REG_MIPS_CP0_USERLOCAL 64 2480 MIPS KVM_REG_MIPS_CP0_XCONTEXTCONFIG 64 2481 MIPS KVM_REG_MIPS_CP0_PAGEMASK 32 2482 MIPS KVM_REG_MIPS_CP0_PAGEGRAIN 32 2483 MIPS KVM_REG_MIPS_CP0_SEGCTL0 64 2484 MIPS KVM_REG_MIPS_CP0_SEGCTL1 64 2485 MIPS KVM_REG_MIPS_CP0_SEGCTL2 64 2486 MIPS KVM_REG_MIPS_CP0_PWBASE 64 2487 MIPS KVM_REG_MIPS_CP0_PWFIELD 64 2488 MIPS KVM_REG_MIPS_CP0_PWSIZE 64 2489 MIPS KVM_REG_MIPS_CP0_WIRED 32 2490 MIPS KVM_REG_MIPS_CP0_PWCTL 32 2491 MIPS KVM_REG_MIPS_CP0_HWRENA 32 2492 MIPS KVM_REG_MIPS_CP0_BADVADDR 64 2493 MIPS KVM_REG_MIPS_CP0_BADINSTR 32 2494 MIPS KVM_REG_MIPS_CP0_BADINSTRP 32 2495 MIPS KVM_REG_MIPS_CP0_COUNT 32 2496 MIPS KVM_REG_MIPS_CP0_ENTRYHI 64 2497 MIPS KVM_REG_MIPS_CP0_COMPARE 32 2498 MIPS KVM_REG_MIPS_CP0_STATUS 32 2499 MIPS KVM_REG_MIPS_CP0_INTCTL 32 2500 MIPS KVM_REG_MIPS_CP0_CAUSE 32 2501 MIPS KVM_REG_MIPS_CP0_EPC 64 2502 MIPS KVM_REG_MIPS_CP0_PRID 32 2503 MIPS KVM_REG_MIPS_CP0_EBASE 64 2504 MIPS KVM_REG_MIPS_CP0_CONFIG 32 2505 MIPS KVM_REG_MIPS_CP0_CONFIG1 32 2506 MIPS KVM_REG_MIPS_CP0_CONFIG2 32 2507 MIPS KVM_REG_MIPS_CP0_CONFIG3 32 2508 MIPS KVM_REG_MIPS_CP0_CONFIG4 32 2509 MIPS KVM_REG_MIPS_CP0_CONFIG5 32 2510 MIPS KVM_REG_MIPS_CP0_CONFIG7 32 2511 MIPS KVM_REG_MIPS_CP0_XCONTEXT 64 2512 MIPS KVM_REG_MIPS_CP0_ERROREPC 64 2513 MIPS KVM_REG_MIPS_CP0_KSCRATCH1 64 2514 MIPS KVM_REG_MIPS_CP0_KSCRATCH2 64 2515 MIPS KVM_REG_MIPS_CP0_KSCRATCH3 64 2516 MIPS KVM_REG_MIPS_CP0_KSCRATCH4 64 2517 MIPS KVM_REG_MIPS_CP0_KSCRATCH5 64 2518 MIPS KVM_REG_MIPS_CP0_KSCRATCH6 64 2519 MIPS KVM_REG_MIPS_CP0_MAAR(0..63) 64 2520 MIPS KVM_REG_MIPS_COUNT_CTL 64 2521 MIPS KVM_REG_MIPS_COUNT_RESUME 64 2522 MIPS KVM_REG_MIPS_COUNT_HZ 64 2523 MIPS KVM_REG_MIPS_FPR_32(0..31) 32 2524 MIPS KVM_REG_MIPS_FPR_64(0..31) 64 2525 MIPS KVM_REG_MIPS_VEC_128(0..31) 128 2526 MIPS KVM_REG_MIPS_FCR_IR 32 2527 MIPS KVM_REG_MIPS_FCR_CSR 32 2528 MIPS KVM_REG_MIPS_MSA_IR 32 2529 MIPS KVM_REG_MIPS_MSA_CSR 32 2530 ======= =============================== ============ 2531 2532ARM registers are mapped using the lower 32 bits. The upper 16 of that 2533is the register group type, or coprocessor number: 2534 2535ARM core registers have the following id bit patterns:: 2536 2537 0x4020 0000 0010 <index into the kvm_regs struct:16> 2538 2539ARM 32-bit CP15 registers have the following id bit patterns:: 2540 2541 0x4020 0000 000F <zero:1> <crn:4> <crm:4> <opc1:4> <opc2:3> 2542 2543ARM 64-bit CP15 registers have the following id bit patterns:: 2544 2545 0x4030 0000 000F <zero:1> <zero:4> <crm:4> <opc1:4> <zero:3> 2546 2547ARM CCSIDR registers are demultiplexed by CSSELR value:: 2548 2549 0x4020 0000 0011 00 <csselr:8> 2550 2551ARM 32-bit VFP control registers have the following id bit patterns:: 2552 2553 0x4020 0000 0012 1 <regno:12> 2554 2555ARM 64-bit FP registers have the following id bit patterns:: 2556 2557 0x4030 0000 0012 0 <regno:12> 2558 2559ARM firmware pseudo-registers have the following bit pattern:: 2560 2561 0x4030 0000 0014 <regno:16> 2562 2563 2564arm64 registers are mapped using the lower 32 bits. The upper 16 of 2565that is the register group type, or coprocessor number: 2566 2567arm64 core/FP-SIMD registers have the following id bit patterns. Note 2568that the size of the access is variable, as the kvm_regs structure 2569contains elements ranging from 32 to 128 bits. The index is a 32bit 2570value in the kvm_regs structure seen as a 32bit array:: 2571 2572 0x60x0 0000 0010 <index into the kvm_regs struct:16> 2573 2574Specifically: 2575 2576======================= ========= ===== ======================================= 2577 Encoding Register Bits kvm_regs member 2578======================= ========= ===== ======================================= 2579 0x6030 0000 0010 0000 X0 64 regs.regs[0] 2580 0x6030 0000 0010 0002 X1 64 regs.regs[1] 2581 ... 2582 0x6030 0000 0010 003c X30 64 regs.regs[30] 2583 0x6030 0000 0010 003e SP 64 regs.sp 2584 0x6030 0000 0010 0040 PC 64 regs.pc 2585 0x6030 0000 0010 0042 PSTATE 64 regs.pstate 2586 0x6030 0000 0010 0044 SP_EL1 64 sp_el1 2587 0x6030 0000 0010 0046 ELR_EL1 64 elr_el1 2588 0x6030 0000 0010 0048 SPSR_EL1 64 spsr[KVM_SPSR_EL1] (alias SPSR_SVC) 2589 0x6030 0000 0010 004a SPSR_ABT 64 spsr[KVM_SPSR_ABT] 2590 0x6030 0000 0010 004c SPSR_UND 64 spsr[KVM_SPSR_UND] 2591 0x6030 0000 0010 004e SPSR_IRQ 64 spsr[KVM_SPSR_IRQ] 2592 0x6060 0000 0010 0050 SPSR_FIQ 64 spsr[KVM_SPSR_FIQ] 2593 0x6040 0000 0010 0054 V0 128 fp_regs.vregs[0] [1]_ 2594 0x6040 0000 0010 0058 V1 128 fp_regs.vregs[1] [1]_ 2595 ... 2596 0x6040 0000 0010 00d0 V31 128 fp_regs.vregs[31] [1]_ 2597 0x6020 0000 0010 00d4 FPSR 32 fp_regs.fpsr 2598 0x6020 0000 0010 00d5 FPCR 32 fp_regs.fpcr 2599======================= ========= ===== ======================================= 2600 2601.. [1] These encodings are not accepted for SVE-enabled vcpus. See 2602 KVM_ARM_VCPU_INIT. 2603 2604 The equivalent register content can be accessed via bits [127:0] of 2605 the corresponding SVE Zn registers instead for vcpus that have SVE 2606 enabled (see below). 2607 2608arm64 CCSIDR registers are demultiplexed by CSSELR value:: 2609 2610 0x6020 0000 0011 00 <csselr:8> 2611 2612arm64 system registers have the following id bit patterns:: 2613 2614 0x6030 0000 0013 <op0:2> <op1:3> <crn:4> <crm:4> <op2:3> 2615 2616.. warning:: 2617 2618 Two system register IDs do not follow the specified pattern. These 2619 are KVM_REG_ARM_TIMER_CVAL and KVM_REG_ARM_TIMER_CNT, which map to 2620 system registers CNTV_CVAL_EL0 and CNTVCT_EL0 respectively. These 2621 two had their values accidentally swapped, which means TIMER_CVAL is 2622 derived from the register encoding for CNTVCT_EL0 and TIMER_CNT is 2623 derived from the register encoding for CNTV_CVAL_EL0. As this is 2624 API, it must remain this way. 2625 2626arm64 firmware pseudo-registers have the following bit pattern:: 2627 2628 0x6030 0000 0014 <regno:16> 2629 2630arm64 SVE registers have the following bit patterns:: 2631 2632 0x6080 0000 0015 00 <n:5> <slice:5> Zn bits[2048*slice + 2047 : 2048*slice] 2633 0x6050 0000 0015 04 <n:4> <slice:5> Pn bits[256*slice + 255 : 256*slice] 2634 0x6050 0000 0015 060 <slice:5> FFR bits[256*slice + 255 : 256*slice] 2635 0x6060 0000 0015 ffff KVM_REG_ARM64_SVE_VLS pseudo-register 2636 2637Access to register IDs where 2048 * slice >= 128 * max_vq will fail with 2638ENOENT. max_vq is the vcpu's maximum supported vector length in 128-bit 2639quadwords: see [2]_ below. 2640 2641These registers are only accessible on vcpus for which SVE is enabled. 2642See KVM_ARM_VCPU_INIT for details. 2643 2644In addition, except for KVM_REG_ARM64_SVE_VLS, these registers are not 2645accessible until the vcpu's SVE configuration has been finalized 2646using KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE). See KVM_ARM_VCPU_INIT 2647and KVM_ARM_VCPU_FINALIZE for more information about this procedure. 2648 2649KVM_REG_ARM64_SVE_VLS is a pseudo-register that allows the set of vector 2650lengths supported by the vcpu to be discovered and configured by 2651userspace. When transferred to or from user memory via KVM_GET_ONE_REG 2652or KVM_SET_ONE_REG, the value of this register is of type 2653__u64[KVM_ARM64_SVE_VLS_WORDS], and encodes the set of vector lengths as 2654follows:: 2655 2656 __u64 vector_lengths[KVM_ARM64_SVE_VLS_WORDS]; 2657 2658 if (vq >= SVE_VQ_MIN && vq <= SVE_VQ_MAX && 2659 ((vector_lengths[(vq - KVM_ARM64_SVE_VQ_MIN) / 64] >> 2660 ((vq - KVM_ARM64_SVE_VQ_MIN) % 64)) & 1)) 2661 /* Vector length vq * 16 bytes supported */ 2662 else 2663 /* Vector length vq * 16 bytes not supported */ 2664 2665.. [2] The maximum value vq for which the above condition is true is 2666 max_vq. This is the maximum vector length available to the guest on 2667 this vcpu, and determines which register slices are visible through 2668 this ioctl interface. 2669 2670(See Documentation/arch/arm64/sve.rst for an explanation of the "vq" 2671nomenclature.) 2672 2673KVM_REG_ARM64_SVE_VLS is only accessible after KVM_ARM_VCPU_INIT. 2674KVM_ARM_VCPU_INIT initialises it to the best set of vector lengths that 2675the host supports. 2676 2677Userspace may subsequently modify it if desired until the vcpu's SVE 2678configuration is finalized using KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE). 2679 2680Apart from simply removing all vector lengths from the host set that 2681exceed some value, support for arbitrarily chosen sets of vector lengths 2682is hardware-dependent and may not be available. Attempting to configure 2683an invalid set of vector lengths via KVM_SET_ONE_REG will fail with 2684EINVAL. 2685 2686After the vcpu's SVE configuration is finalized, further attempts to 2687write this register will fail with EPERM. 2688 2689arm64 bitmap feature firmware pseudo-registers have the following bit pattern:: 2690 2691 0x6030 0000 0016 <regno:16> 2692 2693The bitmap feature firmware registers exposes the hypercall services that 2694are available for userspace to configure. The set bits corresponds to the 2695services that are available for the guests to access. By default, KVM 2696sets all the supported bits during VM initialization. The userspace can 2697discover the available services via KVM_GET_ONE_REG, and write back the 2698bitmap corresponding to the features that it wishes guests to see via 2699KVM_SET_ONE_REG. 2700 2701Note: These registers are immutable once any of the vCPUs of the VM has 2702run at least once. A KVM_SET_ONE_REG in such a scenario will return 2703a -EBUSY to userspace. 2704 2705(See Documentation/virt/kvm/arm/hypercalls.rst for more details.) 2706 2707 2708MIPS registers are mapped using the lower 32 bits. The upper 16 of that is 2709the register group type: 2710 2711MIPS core registers (see above) have the following id bit patterns:: 2712 2713 0x7030 0000 0000 <reg:16> 2714 2715MIPS CP0 registers (see KVM_REG_MIPS_CP0_* above) have the following id bit 2716patterns depending on whether they're 32-bit or 64-bit registers:: 2717 2718 0x7020 0000 0001 00 <reg:5> <sel:3> (32-bit) 2719 0x7030 0000 0001 00 <reg:5> <sel:3> (64-bit) 2720 2721Note: KVM_REG_MIPS_CP0_ENTRYLO0 and KVM_REG_MIPS_CP0_ENTRYLO1 are the MIPS64 2722versions of the EntryLo registers regardless of the word size of the host 2723hardware, host kernel, guest, and whether XPA is present in the guest, i.e. 2724with the RI and XI bits (if they exist) in bits 63 and 62 respectively, and 2725the PFNX field starting at bit 30. 2726 2727MIPS MAARs (see KVM_REG_MIPS_CP0_MAAR(*) above) have the following id bit 2728patterns:: 2729 2730 0x7030 0000 0001 01 <reg:8> 2731 2732MIPS KVM control registers (see above) have the following id bit patterns:: 2733 2734 0x7030 0000 0002 <reg:16> 2735 2736MIPS FPU registers (see KVM_REG_MIPS_FPR_{32,64}() above) have the following 2737id bit patterns depending on the size of the register being accessed. They are 2738always accessed according to the current guest FPU mode (Status.FR and 2739Config5.FRE), i.e. as the guest would see them, and they become unpredictable 2740if the guest FPU mode is changed. MIPS SIMD Architecture (MSA) vector 2741registers (see KVM_REG_MIPS_VEC_128() above) have similar patterns as they 2742overlap the FPU registers:: 2743 2744 0x7020 0000 0003 00 <0:3> <reg:5> (32-bit FPU registers) 2745 0x7030 0000 0003 00 <0:3> <reg:5> (64-bit FPU registers) 2746 0x7040 0000 0003 00 <0:3> <reg:5> (128-bit MSA vector registers) 2747 2748MIPS FPU control registers (see KVM_REG_MIPS_FCR_{IR,CSR} above) have the 2749following id bit patterns:: 2750 2751 0x7020 0000 0003 01 <0:3> <reg:5> 2752 2753MIPS MSA control registers (see KVM_REG_MIPS_MSA_{IR,CSR} above) have the 2754following id bit patterns:: 2755 2756 0x7020 0000 0003 02 <0:3> <reg:5> 2757 2758RISC-V registers are mapped using the lower 32 bits. The upper 8 bits of 2759that is the register group type. 2760 2761RISC-V config registers are meant for configuring a Guest VCPU and it has 2762the following id bit patterns:: 2763 2764 0x8020 0000 01 <index into the kvm_riscv_config struct:24> (32bit Host) 2765 0x8030 0000 01 <index into the kvm_riscv_config struct:24> (64bit Host) 2766 2767Following are the RISC-V config registers: 2768 2769======================= ========= ============================================= 2770 Encoding Register Description 2771======================= ========= ============================================= 2772 0x80x0 0000 0100 0000 isa ISA feature bitmap of Guest VCPU 2773======================= ========= ============================================= 2774 2775The isa config register can be read anytime but can only be written before 2776a Guest VCPU runs. It will have ISA feature bits matching underlying host 2777set by default. 2778 2779RISC-V core registers represent the general execution state of a Guest VCPU 2780and it has the following id bit patterns:: 2781 2782 0x8020 0000 02 <index into the kvm_riscv_core struct:24> (32bit Host) 2783 0x8030 0000 02 <index into the kvm_riscv_core struct:24> (64bit Host) 2784 2785Following are the RISC-V core registers: 2786 2787======================= ========= ============================================= 2788 Encoding Register Description 2789======================= ========= ============================================= 2790 0x80x0 0000 0200 0000 regs.pc Program counter 2791 0x80x0 0000 0200 0001 regs.ra Return address 2792 0x80x0 0000 0200 0002 regs.sp Stack pointer 2793 0x80x0 0000 0200 0003 regs.gp Global pointer 2794 0x80x0 0000 0200 0004 regs.tp Task pointer 2795 0x80x0 0000 0200 0005 regs.t0 Caller saved register 0 2796 0x80x0 0000 0200 0006 regs.t1 Caller saved register 1 2797 0x80x0 0000 0200 0007 regs.t2 Caller saved register 2 2798 0x80x0 0000 0200 0008 regs.s0 Callee saved register 0 2799 0x80x0 0000 0200 0009 regs.s1 Callee saved register 1 2800 0x80x0 0000 0200 000a regs.a0 Function argument (or return value) 0 2801 0x80x0 0000 0200 000b regs.a1 Function argument (or return value) 1 2802 0x80x0 0000 0200 000c regs.a2 Function argument 2 2803 0x80x0 0000 0200 000d regs.a3 Function argument 3 2804 0x80x0 0000 0200 000e regs.a4 Function argument 4 2805 0x80x0 0000 0200 000f regs.a5 Function argument 5 2806 0x80x0 0000 0200 0010 regs.a6 Function argument 6 2807 0x80x0 0000 0200 0011 regs.a7 Function argument 7 2808 0x80x0 0000 0200 0012 regs.s2 Callee saved register 2 2809 0x80x0 0000 0200 0013 regs.s3 Callee saved register 3 2810 0x80x0 0000 0200 0014 regs.s4 Callee saved register 4 2811 0x80x0 0000 0200 0015 regs.s5 Callee saved register 5 2812 0x80x0 0000 0200 0016 regs.s6 Callee saved register 6 2813 0x80x0 0000 0200 0017 regs.s7 Callee saved register 7 2814 0x80x0 0000 0200 0018 regs.s8 Callee saved register 8 2815 0x80x0 0000 0200 0019 regs.s9 Callee saved register 9 2816 0x80x0 0000 0200 001a regs.s10 Callee saved register 10 2817 0x80x0 0000 0200 001b regs.s11 Callee saved register 11 2818 0x80x0 0000 0200 001c regs.t3 Caller saved register 3 2819 0x80x0 0000 0200 001d regs.t4 Caller saved register 4 2820 0x80x0 0000 0200 001e regs.t5 Caller saved register 5 2821 0x80x0 0000 0200 001f regs.t6 Caller saved register 6 2822 0x80x0 0000 0200 0020 mode Privilege mode (1 = S-mode or 0 = U-mode) 2823======================= ========= ============================================= 2824 2825RISC-V csr registers represent the supervisor mode control/status registers 2826of a Guest VCPU and it has the following id bit patterns:: 2827 2828 0x8020 0000 03 <index into the kvm_riscv_csr struct:24> (32bit Host) 2829 0x8030 0000 03 <index into the kvm_riscv_csr struct:24> (64bit Host) 2830 2831Following are the RISC-V csr registers: 2832 2833======================= ========= ============================================= 2834 Encoding Register Description 2835======================= ========= ============================================= 2836 0x80x0 0000 0300 0000 sstatus Supervisor status 2837 0x80x0 0000 0300 0001 sie Supervisor interrupt enable 2838 0x80x0 0000 0300 0002 stvec Supervisor trap vector base 2839 0x80x0 0000 0300 0003 sscratch Supervisor scratch register 2840 0x80x0 0000 0300 0004 sepc Supervisor exception program counter 2841 0x80x0 0000 0300 0005 scause Supervisor trap cause 2842 0x80x0 0000 0300 0006 stval Supervisor bad address or instruction 2843 0x80x0 0000 0300 0007 sip Supervisor interrupt pending 2844 0x80x0 0000 0300 0008 satp Supervisor address translation and protection 2845======================= ========= ============================================= 2846 2847RISC-V timer registers represent the timer state of a Guest VCPU and it has 2848the following id bit patterns:: 2849 2850 0x8030 0000 04 <index into the kvm_riscv_timer struct:24> 2851 2852Following are the RISC-V timer registers: 2853 2854======================= ========= ============================================= 2855 Encoding Register Description 2856======================= ========= ============================================= 2857 0x8030 0000 0400 0000 frequency Time base frequency (read-only) 2858 0x8030 0000 0400 0001 time Time value visible to Guest 2859 0x8030 0000 0400 0002 compare Time compare programmed by Guest 2860 0x8030 0000 0400 0003 state Time compare state (1 = ON or 0 = OFF) 2861======================= ========= ============================================= 2862 2863RISC-V F-extension registers represent the single precision floating point 2864state of a Guest VCPU and it has the following id bit patterns:: 2865 2866 0x8020 0000 05 <index into the __riscv_f_ext_state struct:24> 2867 2868Following are the RISC-V F-extension registers: 2869 2870======================= ========= ============================================= 2871 Encoding Register Description 2872======================= ========= ============================================= 2873 0x8020 0000 0500 0000 f[0] Floating point register 0 2874 ... 2875 0x8020 0000 0500 001f f[31] Floating point register 31 2876 0x8020 0000 0500 0020 fcsr Floating point control and status register 2877======================= ========= ============================================= 2878 2879RISC-V D-extension registers represent the double precision floating point 2880state of a Guest VCPU and it has the following id bit patterns:: 2881 2882 0x8020 0000 06 <index into the __riscv_d_ext_state struct:24> (fcsr) 2883 0x8030 0000 06 <index into the __riscv_d_ext_state struct:24> (non-fcsr) 2884 2885Following are the RISC-V D-extension registers: 2886 2887======================= ========= ============================================= 2888 Encoding Register Description 2889======================= ========= ============================================= 2890 0x8030 0000 0600 0000 f[0] Floating point register 0 2891 ... 2892 0x8030 0000 0600 001f f[31] Floating point register 31 2893 0x8020 0000 0600 0020 fcsr Floating point control and status register 2894======================= ========= ============================================= 2895 2896LoongArch registers are mapped using the lower 32 bits. The upper 16 bits of 2897that is the register group type. 2898 2899LoongArch csr registers are used to control guest cpu or get status of guest 2900cpu, and they have the following id bit patterns:: 2901 2902 0x9030 0000 0001 00 <reg:5> <sel:3> (64-bit) 2903 2904LoongArch KVM control registers are used to implement some new defined functions 2905such as set vcpu counter or reset vcpu, and they have the following id bit patterns:: 2906 2907 0x9030 0000 0002 <reg:16> 2908 2909 29104.69 KVM_GET_ONE_REG 2911-------------------- 2912 2913:Capability: KVM_CAP_ONE_REG 2914:Architectures: all 2915:Type: vcpu ioctl 2916:Parameters: struct kvm_one_reg (in and out) 2917:Returns: 0 on success, negative value on failure 2918 2919Errors include: 2920 2921 ======== ============================================================ 2922 ENOENT no such register 2923 EINVAL invalid register ID, or no such register or used with VMs in 2924 protected virtualization mode on s390 2925 EPERM (arm64) register access not allowed before vcpu finalization 2926 ======== ============================================================ 2927 2928(These error codes are indicative only: do not rely on a specific error 2929code being returned in a specific situation.) 2930 2931This ioctl allows to receive the value of a single register implemented 2932in a vcpu. The register to read is indicated by the "id" field of the 2933kvm_one_reg struct passed in. On success, the register value can be found 2934at the memory location pointed to by "addr". 2935 2936The list of registers accessible using this interface is identical to the 2937list in 4.68. 2938 2939 29404.70 KVM_KVMCLOCK_CTRL 2941---------------------- 2942 2943:Capability: KVM_CAP_KVMCLOCK_CTRL 2944:Architectures: Any that implement pvclocks (currently x86 only) 2945:Type: vcpu ioctl 2946:Parameters: None 2947:Returns: 0 on success, -1 on error 2948 2949This ioctl sets a flag accessible to the guest indicating that the specified 2950vCPU has been paused by the host userspace. 2951 2952The host will set a flag in the pvclock structure that is checked from the 2953soft lockup watchdog. The flag is part of the pvclock structure that is 2954shared between guest and host, specifically the second bit of the flags 2955field of the pvclock_vcpu_time_info structure. It will be set exclusively by 2956the host and read/cleared exclusively by the guest. The guest operation of 2957checking and clearing the flag must be an atomic operation so 2958load-link/store-conditional, or equivalent must be used. There are two cases 2959where the guest will clear the flag: when the soft lockup watchdog timer resets 2960itself or when a soft lockup is detected. This ioctl can be called any time 2961after pausing the vcpu, but before it is resumed. 2962 2963 29644.71 KVM_SIGNAL_MSI 2965------------------- 2966 2967:Capability: KVM_CAP_SIGNAL_MSI 2968:Architectures: x86 arm64 2969:Type: vm ioctl 2970:Parameters: struct kvm_msi (in) 2971:Returns: >0 on delivery, 0 if guest blocked the MSI, and -1 on error 2972 2973Directly inject a MSI message. Only valid with in-kernel irqchip that handles 2974MSI messages. 2975 2976:: 2977 2978 struct kvm_msi { 2979 __u32 address_lo; 2980 __u32 address_hi; 2981 __u32 data; 2982 __u32 flags; 2983 __u32 devid; 2984 __u8 pad[12]; 2985 }; 2986 2987flags: 2988 KVM_MSI_VALID_DEVID: devid contains a valid value. The per-VM 2989 KVM_CAP_MSI_DEVID capability advertises the requirement to provide 2990 the device ID. If this capability is not available, userspace 2991 should never set the KVM_MSI_VALID_DEVID flag as the ioctl might fail. 2992 2993If KVM_MSI_VALID_DEVID is set, devid contains a unique device identifier 2994for the device that wrote the MSI message. For PCI, this is usually a 2995BDF identifier in the lower 16 bits. 2996 2997On x86, address_hi is ignored unless the KVM_X2APIC_API_USE_32BIT_IDS 2998feature of KVM_CAP_X2APIC_API capability is enabled. If it is enabled, 2999address_hi bits 31-8 provide bits 31-8 of the destination id. Bits 7-0 of 3000address_hi must be zero. 3001 3002 30034.71 KVM_CREATE_PIT2 3004-------------------- 3005 3006:Capability: KVM_CAP_PIT2 3007:Architectures: x86 3008:Type: vm ioctl 3009:Parameters: struct kvm_pit_config (in) 3010:Returns: 0 on success, -1 on error 3011 3012Creates an in-kernel device model for the i8254 PIT. This call is only valid 3013after enabling in-kernel irqchip support via KVM_CREATE_IRQCHIP. The following 3014parameters have to be passed:: 3015 3016 struct kvm_pit_config { 3017 __u32 flags; 3018 __u32 pad[15]; 3019 }; 3020 3021Valid flags are:: 3022 3023 #define KVM_PIT_SPEAKER_DUMMY 1 /* emulate speaker port stub */ 3024 3025PIT timer interrupts may use a per-VM kernel thread for injection. If it 3026exists, this thread will have a name of the following pattern:: 3027 3028 kvm-pit/<owner-process-pid> 3029 3030When running a guest with elevated priorities, the scheduling parameters of 3031this thread may have to be adjusted accordingly. 3032 3033This IOCTL replaces the obsolete KVM_CREATE_PIT. 3034 3035 30364.72 KVM_GET_PIT2 3037----------------- 3038 3039:Capability: KVM_CAP_PIT_STATE2 3040:Architectures: x86 3041:Type: vm ioctl 3042:Parameters: struct kvm_pit_state2 (out) 3043:Returns: 0 on success, -1 on error 3044 3045Retrieves the state of the in-kernel PIT model. Only valid after 3046KVM_CREATE_PIT2. The state is returned in the following structure:: 3047 3048 struct kvm_pit_state2 { 3049 struct kvm_pit_channel_state channels[3]; 3050 __u32 flags; 3051 __u32 reserved[9]; 3052 }; 3053 3054Valid flags are:: 3055 3056 /* disable PIT in HPET legacy mode */ 3057 #define KVM_PIT_FLAGS_HPET_LEGACY 0x00000001 3058 /* speaker port data bit enabled */ 3059 #define KVM_PIT_FLAGS_SPEAKER_DATA_ON 0x00000002 3060 3061This IOCTL replaces the obsolete KVM_GET_PIT. 3062 3063 30644.73 KVM_SET_PIT2 3065----------------- 3066 3067:Capability: KVM_CAP_PIT_STATE2 3068:Architectures: x86 3069:Type: vm ioctl 3070:Parameters: struct kvm_pit_state2 (in) 3071:Returns: 0 on success, -1 on error 3072 3073Sets the state of the in-kernel PIT model. Only valid after KVM_CREATE_PIT2. 3074See KVM_GET_PIT2 for details on struct kvm_pit_state2. 3075 3076This IOCTL replaces the obsolete KVM_SET_PIT. 3077 3078 30794.74 KVM_PPC_GET_SMMU_INFO 3080-------------------------- 3081 3082:Capability: KVM_CAP_PPC_GET_SMMU_INFO 3083:Architectures: powerpc 3084:Type: vm ioctl 3085:Parameters: None 3086:Returns: 0 on success, -1 on error 3087 3088This populates and returns a structure describing the features of 3089the "Server" class MMU emulation supported by KVM. 3090This can in turn be used by userspace to generate the appropriate 3091device-tree properties for the guest operating system. 3092 3093The structure contains some global information, followed by an 3094array of supported segment page sizes:: 3095 3096 struct kvm_ppc_smmu_info { 3097 __u64 flags; 3098 __u32 slb_size; 3099 __u32 pad; 3100 struct kvm_ppc_one_seg_page_size sps[KVM_PPC_PAGE_SIZES_MAX_SZ]; 3101 }; 3102 3103The supported flags are: 3104 3105 - KVM_PPC_PAGE_SIZES_REAL: 3106 When that flag is set, guest page sizes must "fit" the backing 3107 store page sizes. When not set, any page size in the list can 3108 be used regardless of how they are backed by userspace. 3109 3110 - KVM_PPC_1T_SEGMENTS 3111 The emulated MMU supports 1T segments in addition to the 3112 standard 256M ones. 3113 3114 - KVM_PPC_NO_HASH 3115 This flag indicates that HPT guests are not supported by KVM, 3116 thus all guests must use radix MMU mode. 3117 3118The "slb_size" field indicates how many SLB entries are supported 3119 3120The "sps" array contains 8 entries indicating the supported base 3121page sizes for a segment in increasing order. Each entry is defined 3122as follow:: 3123 3124 struct kvm_ppc_one_seg_page_size { 3125 __u32 page_shift; /* Base page shift of segment (or 0) */ 3126 __u32 slb_enc; /* SLB encoding for BookS */ 3127 struct kvm_ppc_one_page_size enc[KVM_PPC_PAGE_SIZES_MAX_SZ]; 3128 }; 3129 3130An entry with a "page_shift" of 0 is unused. Because the array is 3131organized in increasing order, a lookup can stop when encountering 3132such an entry. 3133 3134The "slb_enc" field provides the encoding to use in the SLB for the 3135page size. The bits are in positions such as the value can directly 3136be OR'ed into the "vsid" argument of the slbmte instruction. 3137 3138The "enc" array is a list which for each of those segment base page 3139size provides the list of supported actual page sizes (which can be 3140only larger or equal to the base page size), along with the 3141corresponding encoding in the hash PTE. Similarly, the array is 31428 entries sorted by increasing sizes and an entry with a "0" shift 3143is an empty entry and a terminator:: 3144 3145 struct kvm_ppc_one_page_size { 3146 __u32 page_shift; /* Page shift (or 0) */ 3147 __u32 pte_enc; /* Encoding in the HPTE (>>12) */ 3148 }; 3149 3150The "pte_enc" field provides a value that can OR'ed into the hash 3151PTE's RPN field (ie, it needs to be shifted left by 12 to OR it 3152into the hash PTE second double word). 3153 31544.75 KVM_IRQFD 3155-------------- 3156 3157:Capability: KVM_CAP_IRQFD 3158:Architectures: x86 s390 arm64 3159:Type: vm ioctl 3160:Parameters: struct kvm_irqfd (in) 3161:Returns: 0 on success, -1 on error 3162 3163Allows setting an eventfd to directly trigger a guest interrupt. 3164kvm_irqfd.fd specifies the file descriptor to use as the eventfd and 3165kvm_irqfd.gsi specifies the irqchip pin toggled by this event. When 3166an event is triggered on the eventfd, an interrupt is injected into 3167the guest using the specified gsi pin. The irqfd is removed using 3168the KVM_IRQFD_FLAG_DEASSIGN flag, specifying both kvm_irqfd.fd 3169and kvm_irqfd.gsi. 3170 3171With KVM_CAP_IRQFD_RESAMPLE, KVM_IRQFD supports a de-assert and notify 3172mechanism allowing emulation of level-triggered, irqfd-based 3173interrupts. When KVM_IRQFD_FLAG_RESAMPLE is set the user must pass an 3174additional eventfd in the kvm_irqfd.resamplefd field. When operating 3175in resample mode, posting of an interrupt through kvm_irq.fd asserts 3176the specified gsi in the irqchip. When the irqchip is resampled, such 3177as from an EOI, the gsi is de-asserted and the user is notified via 3178kvm_irqfd.resamplefd. It is the user's responsibility to re-queue 3179the interrupt if the device making use of it still requires service. 3180Note that closing the resamplefd is not sufficient to disable the 3181irqfd. The KVM_IRQFD_FLAG_RESAMPLE is only necessary on assignment 3182and need not be specified with KVM_IRQFD_FLAG_DEASSIGN. 3183 3184On arm64, gsi routing being supported, the following can happen: 3185 3186- in case no routing entry is associated to this gsi, injection fails 3187- in case the gsi is associated to an irqchip routing entry, 3188 irqchip.pin + 32 corresponds to the injected SPI ID. 3189- in case the gsi is associated to an MSI routing entry, the MSI 3190 message and device ID are translated into an LPI (support restricted 3191 to GICv3 ITS in-kernel emulation). 3192 31934.76 KVM_PPC_ALLOCATE_HTAB 3194-------------------------- 3195 3196:Capability: KVM_CAP_PPC_ALLOC_HTAB 3197:Architectures: powerpc 3198:Type: vm ioctl 3199:Parameters: Pointer to u32 containing hash table order (in/out) 3200:Returns: 0 on success, -1 on error 3201 3202This requests the host kernel to allocate an MMU hash table for a 3203guest using the PAPR paravirtualization interface. This only does 3204anything if the kernel is configured to use the Book 3S HV style of 3205virtualization. Otherwise the capability doesn't exist and the ioctl 3206returns an ENOTTY error. The rest of this description assumes Book 3S 3207HV. 3208 3209There must be no vcpus running when this ioctl is called; if there 3210are, it will do nothing and return an EBUSY error. 3211 3212The parameter is a pointer to a 32-bit unsigned integer variable 3213containing the order (log base 2) of the desired size of the hash 3214table, which must be between 18 and 46. On successful return from the 3215ioctl, the value will not be changed by the kernel. 3216 3217If no hash table has been allocated when any vcpu is asked to run 3218(with the KVM_RUN ioctl), the host kernel will allocate a 3219default-sized hash table (16 MB). 3220 3221If this ioctl is called when a hash table has already been allocated, 3222with a different order from the existing hash table, the existing hash 3223table will be freed and a new one allocated. If this is ioctl is 3224called when a hash table has already been allocated of the same order 3225as specified, the kernel will clear out the existing hash table (zero 3226all HPTEs). In either case, if the guest is using the virtualized 3227real-mode area (VRMA) facility, the kernel will re-create the VMRA 3228HPTEs on the next KVM_RUN of any vcpu. 3229 32304.77 KVM_S390_INTERRUPT 3231----------------------- 3232 3233:Capability: basic 3234:Architectures: s390 3235:Type: vm ioctl, vcpu ioctl 3236:Parameters: struct kvm_s390_interrupt (in) 3237:Returns: 0 on success, -1 on error 3238 3239Allows to inject an interrupt to the guest. Interrupts can be floating 3240(vm ioctl) or per cpu (vcpu ioctl), depending on the interrupt type. 3241 3242Interrupt parameters are passed via kvm_s390_interrupt:: 3243 3244 struct kvm_s390_interrupt { 3245 __u32 type; 3246 __u32 parm; 3247 __u64 parm64; 3248 }; 3249 3250type can be one of the following: 3251 3252KVM_S390_SIGP_STOP (vcpu) 3253 - sigp stop; optional flags in parm 3254KVM_S390_PROGRAM_INT (vcpu) 3255 - program check; code in parm 3256KVM_S390_SIGP_SET_PREFIX (vcpu) 3257 - sigp set prefix; prefix address in parm 3258KVM_S390_RESTART (vcpu) 3259 - restart 3260KVM_S390_INT_CLOCK_COMP (vcpu) 3261 - clock comparator interrupt 3262KVM_S390_INT_CPU_TIMER (vcpu) 3263 - CPU timer interrupt 3264KVM_S390_INT_VIRTIO (vm) 3265 - virtio external interrupt; external interrupt 3266 parameters in parm and parm64 3267KVM_S390_INT_SERVICE (vm) 3268 - sclp external interrupt; sclp parameter in parm 3269KVM_S390_INT_EMERGENCY (vcpu) 3270 - sigp emergency; source cpu in parm 3271KVM_S390_INT_EXTERNAL_CALL (vcpu) 3272 - sigp external call; source cpu in parm 3273KVM_S390_INT_IO(ai,cssid,ssid,schid) (vm) 3274 - compound value to indicate an 3275 I/O interrupt (ai - adapter interrupt; cssid,ssid,schid - subchannel); 3276 I/O interruption parameters in parm (subchannel) and parm64 (intparm, 3277 interruption subclass) 3278KVM_S390_MCHK (vm, vcpu) 3279 - machine check interrupt; cr 14 bits in parm, machine check interrupt 3280 code in parm64 (note that machine checks needing further payload are not 3281 supported by this ioctl) 3282 3283This is an asynchronous vcpu ioctl and can be invoked from any thread. 3284 32854.78 KVM_PPC_GET_HTAB_FD 3286------------------------ 3287 3288:Capability: KVM_CAP_PPC_HTAB_FD 3289:Architectures: powerpc 3290:Type: vm ioctl 3291:Parameters: Pointer to struct kvm_get_htab_fd (in) 3292:Returns: file descriptor number (>= 0) on success, -1 on error 3293 3294This returns a file descriptor that can be used either to read out the 3295entries in the guest's hashed page table (HPT), or to write entries to 3296initialize the HPT. The returned fd can only be written to if the 3297KVM_GET_HTAB_WRITE bit is set in the flags field of the argument, and 3298can only be read if that bit is clear. The argument struct looks like 3299this:: 3300 3301 /* For KVM_PPC_GET_HTAB_FD */ 3302 struct kvm_get_htab_fd { 3303 __u64 flags; 3304 __u64 start_index; 3305 __u64 reserved[2]; 3306 }; 3307 3308 /* Values for kvm_get_htab_fd.flags */ 3309 #define KVM_GET_HTAB_BOLTED_ONLY ((__u64)0x1) 3310 #define KVM_GET_HTAB_WRITE ((__u64)0x2) 3311 3312The 'start_index' field gives the index in the HPT of the entry at 3313which to start reading. It is ignored when writing. 3314 3315Reads on the fd will initially supply information about all 3316"interesting" HPT entries. Interesting entries are those with the 3317bolted bit set, if the KVM_GET_HTAB_BOLTED_ONLY bit is set, otherwise 3318all entries. When the end of the HPT is reached, the read() will 3319return. If read() is called again on the fd, it will start again from 3320the beginning of the HPT, but will only return HPT entries that have 3321changed since they were last read. 3322 3323Data read or written is structured as a header (8 bytes) followed by a 3324series of valid HPT entries (16 bytes) each. The header indicates how 3325many valid HPT entries there are and how many invalid entries follow 3326the valid entries. The invalid entries are not represented explicitly 3327in the stream. The header format is:: 3328 3329 struct kvm_get_htab_header { 3330 __u32 index; 3331 __u16 n_valid; 3332 __u16 n_invalid; 3333 }; 3334 3335Writes to the fd create HPT entries starting at the index given in the 3336header; first 'n_valid' valid entries with contents from the data 3337written, then 'n_invalid' invalid entries, invalidating any previously 3338valid entries found. 3339 33404.79 KVM_CREATE_DEVICE 3341---------------------- 3342 3343:Capability: KVM_CAP_DEVICE_CTRL 3344:Architectures: all 3345:Type: vm ioctl 3346:Parameters: struct kvm_create_device (in/out) 3347:Returns: 0 on success, -1 on error 3348 3349Errors: 3350 3351 ====== ======================================================= 3352 ENODEV The device type is unknown or unsupported 3353 EEXIST Device already created, and this type of device may not 3354 be instantiated multiple times 3355 ====== ======================================================= 3356 3357 Other error conditions may be defined by individual device types or 3358 have their standard meanings. 3359 3360Creates an emulated device in the kernel. The file descriptor returned 3361in fd can be used with KVM_SET/GET/HAS_DEVICE_ATTR. 3362 3363If the KVM_CREATE_DEVICE_TEST flag is set, only test whether the 3364device type is supported (not necessarily whether it can be created 3365in the current vm). 3366 3367Individual devices should not define flags. Attributes should be used 3368for specifying any behavior that is not implied by the device type 3369number. 3370 3371:: 3372 3373 struct kvm_create_device { 3374 __u32 type; /* in: KVM_DEV_TYPE_xxx */ 3375 __u32 fd; /* out: device handle */ 3376 __u32 flags; /* in: KVM_CREATE_DEVICE_xxx */ 3377 }; 3378 33794.80 KVM_SET_DEVICE_ATTR/KVM_GET_DEVICE_ATTR 3380-------------------------------------------- 3381 3382:Capability: KVM_CAP_DEVICE_CTRL, KVM_CAP_VM_ATTRIBUTES for vm device, 3383 KVM_CAP_VCPU_ATTRIBUTES for vcpu device 3384 KVM_CAP_SYS_ATTRIBUTES for system (/dev/kvm) device (no set) 3385:Architectures: x86, arm64, s390 3386:Type: device ioctl, vm ioctl, vcpu ioctl 3387:Parameters: struct kvm_device_attr 3388:Returns: 0 on success, -1 on error 3389 3390Errors: 3391 3392 ===== ============================================================= 3393 ENXIO The group or attribute is unknown/unsupported for this device 3394 or hardware support is missing. 3395 EPERM The attribute cannot (currently) be accessed this way 3396 (e.g. read-only attribute, or attribute that only makes 3397 sense when the device is in a different state) 3398 ===== ============================================================= 3399 3400 Other error conditions may be defined by individual device types. 3401 3402Gets/sets a specified piece of device configuration and/or state. The 3403semantics are device-specific. See individual device documentation in 3404the "devices" directory. As with ONE_REG, the size of the data 3405transferred is defined by the particular attribute. 3406 3407:: 3408 3409 struct kvm_device_attr { 3410 __u32 flags; /* no flags currently defined */ 3411 __u32 group; /* device-defined */ 3412 __u64 attr; /* group-defined */ 3413 __u64 addr; /* userspace address of attr data */ 3414 }; 3415 34164.81 KVM_HAS_DEVICE_ATTR 3417------------------------ 3418 3419:Capability: KVM_CAP_DEVICE_CTRL, KVM_CAP_VM_ATTRIBUTES for vm device, 3420 KVM_CAP_VCPU_ATTRIBUTES for vcpu device 3421 KVM_CAP_SYS_ATTRIBUTES for system (/dev/kvm) device 3422:Type: device ioctl, vm ioctl, vcpu ioctl 3423:Parameters: struct kvm_device_attr 3424:Returns: 0 on success, -1 on error 3425 3426Errors: 3427 3428 ===== ============================================================= 3429 ENXIO The group or attribute is unknown/unsupported for this device 3430 or hardware support is missing. 3431 ===== ============================================================= 3432 3433Tests whether a device supports a particular attribute. A successful 3434return indicates the attribute is implemented. It does not necessarily 3435indicate that the attribute can be read or written in the device's 3436current state. "addr" is ignored. 3437 3438.. _KVM_ARM_VCPU_INIT: 3439 34404.82 KVM_ARM_VCPU_INIT 3441---------------------- 3442 3443:Capability: basic 3444:Architectures: arm64 3445:Type: vcpu ioctl 3446:Parameters: struct kvm_vcpu_init (in) 3447:Returns: 0 on success; -1 on error 3448 3449Errors: 3450 3451 ====== ================================================================= 3452 EINVAL the target is unknown, or the combination of features is invalid. 3453 ENOENT a features bit specified is unknown. 3454 ====== ================================================================= 3455 3456This tells KVM what type of CPU to present to the guest, and what 3457optional features it should have. This will cause a reset of the cpu 3458registers to their initial values. If this is not called, KVM_RUN will 3459return ENOEXEC for that vcpu. 3460 3461The initial values are defined as: 3462 - Processor state: 3463 * AArch64: EL1h, D, A, I and F bits set. All other bits 3464 are cleared. 3465 * AArch32: SVC, A, I and F bits set. All other bits are 3466 cleared. 3467 - General Purpose registers, including PC and SP: set to 0 3468 - FPSIMD/NEON registers: set to 0 3469 - SVE registers: set to 0 3470 - System registers: Reset to their architecturally defined 3471 values as for a warm reset to EL1 (resp. SVC) 3472 3473Note that because some registers reflect machine topology, all vcpus 3474should be created before this ioctl is invoked. 3475 3476Userspace can call this function multiple times for a given vcpu, including 3477after the vcpu has been run. This will reset the vcpu to its initial 3478state. All calls to this function after the initial call must use the same 3479target and same set of feature flags, otherwise EINVAL will be returned. 3480 3481Possible features: 3482 3483 - KVM_ARM_VCPU_POWER_OFF: Starts the CPU in a power-off state. 3484 Depends on KVM_CAP_ARM_PSCI. If not set, the CPU will be powered on 3485 and execute guest code when KVM_RUN is called. 3486 - KVM_ARM_VCPU_EL1_32BIT: Starts the CPU in a 32bit mode. 3487 Depends on KVM_CAP_ARM_EL1_32BIT (arm64 only). 3488 - KVM_ARM_VCPU_PSCI_0_2: Emulate PSCI v0.2 (or a future revision 3489 backward compatible with v0.2) for the CPU. 3490 Depends on KVM_CAP_ARM_PSCI_0_2. 3491 - KVM_ARM_VCPU_PMU_V3: Emulate PMUv3 for the CPU. 3492 Depends on KVM_CAP_ARM_PMU_V3. 3493 3494 - KVM_ARM_VCPU_PTRAUTH_ADDRESS: Enables Address Pointer authentication 3495 for arm64 only. 3496 Depends on KVM_CAP_ARM_PTRAUTH_ADDRESS. 3497 If KVM_CAP_ARM_PTRAUTH_ADDRESS and KVM_CAP_ARM_PTRAUTH_GENERIC are 3498 both present, then both KVM_ARM_VCPU_PTRAUTH_ADDRESS and 3499 KVM_ARM_VCPU_PTRAUTH_GENERIC must be requested or neither must be 3500 requested. 3501 3502 - KVM_ARM_VCPU_PTRAUTH_GENERIC: Enables Generic Pointer authentication 3503 for arm64 only. 3504 Depends on KVM_CAP_ARM_PTRAUTH_GENERIC. 3505 If KVM_CAP_ARM_PTRAUTH_ADDRESS and KVM_CAP_ARM_PTRAUTH_GENERIC are 3506 both present, then both KVM_ARM_VCPU_PTRAUTH_ADDRESS and 3507 KVM_ARM_VCPU_PTRAUTH_GENERIC must be requested or neither must be 3508 requested. 3509 3510 - KVM_ARM_VCPU_SVE: Enables SVE for the CPU (arm64 only). 3511 Depends on KVM_CAP_ARM_SVE. 3512 Requires KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE): 3513 3514 * After KVM_ARM_VCPU_INIT: 3515 3516 - KVM_REG_ARM64_SVE_VLS may be read using KVM_GET_ONE_REG: the 3517 initial value of this pseudo-register indicates the best set of 3518 vector lengths possible for a vcpu on this host. 3519 3520 * Before KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE): 3521 3522 - KVM_RUN and KVM_GET_REG_LIST are not available; 3523 3524 - KVM_GET_ONE_REG and KVM_SET_ONE_REG cannot be used to access 3525 the scalable architectural SVE registers 3526 KVM_REG_ARM64_SVE_ZREG(), KVM_REG_ARM64_SVE_PREG() or 3527 KVM_REG_ARM64_SVE_FFR; 3528 3529 - KVM_REG_ARM64_SVE_VLS may optionally be written using 3530 KVM_SET_ONE_REG, to modify the set of vector lengths available 3531 for the vcpu. 3532 3533 * After KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE): 3534 3535 - the KVM_REG_ARM64_SVE_VLS pseudo-register is immutable, and can 3536 no longer be written using KVM_SET_ONE_REG. 3537 35384.83 KVM_ARM_PREFERRED_TARGET 3539----------------------------- 3540 3541:Capability: basic 3542:Architectures: arm64 3543:Type: vm ioctl 3544:Parameters: struct kvm_vcpu_init (out) 3545:Returns: 0 on success; -1 on error 3546 3547Errors: 3548 3549 ====== ========================================== 3550 ENODEV no preferred target available for the host 3551 ====== ========================================== 3552 3553This queries KVM for preferred CPU target type which can be emulated 3554by KVM on underlying host. 3555 3556The ioctl returns struct kvm_vcpu_init instance containing information 3557about preferred CPU target type and recommended features for it. The 3558kvm_vcpu_init->features bitmap returned will have feature bits set if 3559the preferred target recommends setting these features, but this is 3560not mandatory. 3561 3562The information returned by this ioctl can be used to prepare an instance 3563of struct kvm_vcpu_init for KVM_ARM_VCPU_INIT ioctl which will result in 3564VCPU matching underlying host. 3565 3566 35674.84 KVM_GET_REG_LIST 3568--------------------- 3569 3570:Capability: basic 3571:Architectures: arm64, mips, riscv 3572:Type: vcpu ioctl 3573:Parameters: struct kvm_reg_list (in/out) 3574:Returns: 0 on success; -1 on error 3575 3576Errors: 3577 3578 ===== ============================================================== 3579 E2BIG the reg index list is too big to fit in the array specified by 3580 the user (the number required will be written into n). 3581 ===== ============================================================== 3582 3583:: 3584 3585 struct kvm_reg_list { 3586 __u64 n; /* number of registers in reg[] */ 3587 __u64 reg[0]; 3588 }; 3589 3590This ioctl returns the guest registers that are supported for the 3591KVM_GET_ONE_REG/KVM_SET_ONE_REG calls. 3592 3593 35944.85 KVM_ARM_SET_DEVICE_ADDR (deprecated) 3595----------------------------------------- 3596 3597:Capability: KVM_CAP_ARM_SET_DEVICE_ADDR 3598:Architectures: arm64 3599:Type: vm ioctl 3600:Parameters: struct kvm_arm_device_address (in) 3601:Returns: 0 on success, -1 on error 3602 3603Errors: 3604 3605 ====== ============================================ 3606 ENODEV The device id is unknown 3607 ENXIO Device not supported on current system 3608 EEXIST Address already set 3609 E2BIG Address outside guest physical address space 3610 EBUSY Address overlaps with other device range 3611 ====== ============================================ 3612 3613:: 3614 3615 struct kvm_arm_device_addr { 3616 __u64 id; 3617 __u64 addr; 3618 }; 3619 3620Specify a device address in the guest's physical address space where guests 3621can access emulated or directly exposed devices, which the host kernel needs 3622to know about. The id field is an architecture specific identifier for a 3623specific device. 3624 3625arm64 divides the id field into two parts, a device id and an 3626address type id specific to the individual device:: 3627 3628 bits: | 63 ... 32 | 31 ... 16 | 15 ... 0 | 3629 field: | 0x00000000 | device id | addr type id | 3630 3631arm64 currently only require this when using the in-kernel GIC 3632support for the hardware VGIC features, using KVM_ARM_DEVICE_VGIC_V2 3633as the device id. When setting the base address for the guest's 3634mapping of the VGIC virtual CPU and distributor interface, the ioctl 3635must be called after calling KVM_CREATE_IRQCHIP, but before calling 3636KVM_RUN on any of the VCPUs. Calling this ioctl twice for any of the 3637base addresses will return -EEXIST. 3638 3639Note, this IOCTL is deprecated and the more flexible SET/GET_DEVICE_ATTR API 3640should be used instead. 3641 3642 36434.86 KVM_PPC_RTAS_DEFINE_TOKEN 3644------------------------------ 3645 3646:Capability: KVM_CAP_PPC_RTAS 3647:Architectures: ppc 3648:Type: vm ioctl 3649:Parameters: struct kvm_rtas_token_args 3650:Returns: 0 on success, -1 on error 3651 3652Defines a token value for a RTAS (Run Time Abstraction Services) 3653service in order to allow it to be handled in the kernel. The 3654argument struct gives the name of the service, which must be the name 3655of a service that has a kernel-side implementation. If the token 3656value is non-zero, it will be associated with that service, and 3657subsequent RTAS calls by the guest specifying that token will be 3658handled by the kernel. If the token value is 0, then any token 3659associated with the service will be forgotten, and subsequent RTAS 3660calls by the guest for that service will be passed to userspace to be 3661handled. 3662 36634.87 KVM_SET_GUEST_DEBUG 3664------------------------ 3665 3666:Capability: KVM_CAP_SET_GUEST_DEBUG 3667:Architectures: x86, s390, ppc, arm64 3668:Type: vcpu ioctl 3669:Parameters: struct kvm_guest_debug (in) 3670:Returns: 0 on success; -1 on error 3671 3672:: 3673 3674 struct kvm_guest_debug { 3675 __u32 control; 3676 __u32 pad; 3677 struct kvm_guest_debug_arch arch; 3678 }; 3679 3680Set up the processor specific debug registers and configure vcpu for 3681handling guest debug events. There are two parts to the structure, the 3682first a control bitfield indicates the type of debug events to handle 3683when running. Common control bits are: 3684 3685 - KVM_GUESTDBG_ENABLE: guest debugging is enabled 3686 - KVM_GUESTDBG_SINGLESTEP: the next run should single-step 3687 3688The top 16 bits of the control field are architecture specific control 3689flags which can include the following: 3690 3691 - KVM_GUESTDBG_USE_SW_BP: using software breakpoints [x86, arm64] 3692 - KVM_GUESTDBG_USE_HW_BP: using hardware breakpoints [x86, s390] 3693 - KVM_GUESTDBG_USE_HW: using hardware debug events [arm64] 3694 - KVM_GUESTDBG_INJECT_DB: inject DB type exception [x86] 3695 - KVM_GUESTDBG_INJECT_BP: inject BP type exception [x86] 3696 - KVM_GUESTDBG_EXIT_PENDING: trigger an immediate guest exit [s390] 3697 - KVM_GUESTDBG_BLOCKIRQ: avoid injecting interrupts/NMI/SMI [x86] 3698 3699For example KVM_GUESTDBG_USE_SW_BP indicates that software breakpoints 3700are enabled in memory so we need to ensure breakpoint exceptions are 3701correctly trapped and the KVM run loop exits at the breakpoint and not 3702running off into the normal guest vector. For KVM_GUESTDBG_USE_HW_BP 3703we need to ensure the guest vCPUs architecture specific registers are 3704updated to the correct (supplied) values. 3705 3706The second part of the structure is architecture specific and 3707typically contains a set of debug registers. 3708 3709For arm64 the number of debug registers is implementation defined and 3710can be determined by querying the KVM_CAP_GUEST_DEBUG_HW_BPS and 3711KVM_CAP_GUEST_DEBUG_HW_WPS capabilities which return a positive number 3712indicating the number of supported registers. 3713 3714For ppc, the KVM_CAP_PPC_GUEST_DEBUG_SSTEP capability indicates whether 3715the single-step debug event (KVM_GUESTDBG_SINGLESTEP) is supported. 3716 3717Also when supported, KVM_CAP_SET_GUEST_DEBUG2 capability indicates the 3718supported KVM_GUESTDBG_* bits in the control field. 3719 3720When debug events exit the main run loop with the reason 3721KVM_EXIT_DEBUG with the kvm_debug_exit_arch part of the kvm_run 3722structure containing architecture specific debug information. 3723 37244.88 KVM_GET_EMULATED_CPUID 3725--------------------------- 3726 3727:Capability: KVM_CAP_EXT_EMUL_CPUID 3728:Architectures: x86 3729:Type: system ioctl 3730:Parameters: struct kvm_cpuid2 (in/out) 3731:Returns: 0 on success, -1 on error 3732 3733:: 3734 3735 struct kvm_cpuid2 { 3736 __u32 nent; 3737 __u32 flags; 3738 struct kvm_cpuid_entry2 entries[0]; 3739 }; 3740 3741The member 'flags' is used for passing flags from userspace. 3742 3743:: 3744 3745 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX BIT(0) 3746 #define KVM_CPUID_FLAG_STATEFUL_FUNC BIT(1) /* deprecated */ 3747 #define KVM_CPUID_FLAG_STATE_READ_NEXT BIT(2) /* deprecated */ 3748 3749 struct kvm_cpuid_entry2 { 3750 __u32 function; 3751 __u32 index; 3752 __u32 flags; 3753 __u32 eax; 3754 __u32 ebx; 3755 __u32 ecx; 3756 __u32 edx; 3757 __u32 padding[3]; 3758 }; 3759 3760This ioctl returns x86 cpuid features which are emulated by 3761kvm.Userspace can use the information returned by this ioctl to query 3762which features are emulated by kvm instead of being present natively. 3763 3764Userspace invokes KVM_GET_EMULATED_CPUID by passing a kvm_cpuid2 3765structure with the 'nent' field indicating the number of entries in 3766the variable-size array 'entries'. If the number of entries is too low 3767to describe the cpu capabilities, an error (E2BIG) is returned. If the 3768number is too high, the 'nent' field is adjusted and an error (ENOMEM) 3769is returned. If the number is just right, the 'nent' field is adjusted 3770to the number of valid entries in the 'entries' array, which is then 3771filled. 3772 3773The entries returned are the set CPUID bits of the respective features 3774which kvm emulates, as returned by the CPUID instruction, with unknown 3775or unsupported feature bits cleared. 3776 3777Features like x2apic, for example, may not be present in the host cpu 3778but are exposed by kvm in KVM_GET_SUPPORTED_CPUID because they can be 3779emulated efficiently and thus not included here. 3780 3781The fields in each entry are defined as follows: 3782 3783 function: 3784 the eax value used to obtain the entry 3785 index: 3786 the ecx value used to obtain the entry (for entries that are 3787 affected by ecx) 3788 flags: 3789 an OR of zero or more of the following: 3790 3791 KVM_CPUID_FLAG_SIGNIFCANT_INDEX: 3792 if the index field is valid 3793 3794 eax, ebx, ecx, edx: 3795 3796 the values returned by the cpuid instruction for 3797 this function/index combination 3798 37994.89 KVM_S390_MEM_OP 3800-------------------- 3801 3802:Capability: KVM_CAP_S390_MEM_OP, KVM_CAP_S390_PROTECTED, KVM_CAP_S390_MEM_OP_EXTENSION 3803:Architectures: s390 3804:Type: vm ioctl, vcpu ioctl 3805:Parameters: struct kvm_s390_mem_op (in) 3806:Returns: = 0 on success, 3807 < 0 on generic error (e.g. -EFAULT or -ENOMEM), 3808 16 bit program exception code if the access causes such an exception 3809 3810Read or write data from/to the VM's memory. 3811The KVM_CAP_S390_MEM_OP_EXTENSION capability specifies what functionality is 3812supported. 3813 3814Parameters are specified via the following structure:: 3815 3816 struct kvm_s390_mem_op { 3817 __u64 gaddr; /* the guest address */ 3818 __u64 flags; /* flags */ 3819 __u32 size; /* amount of bytes */ 3820 __u32 op; /* type of operation */ 3821 __u64 buf; /* buffer in userspace */ 3822 union { 3823 struct { 3824 __u8 ar; /* the access register number */ 3825 __u8 key; /* access key, ignored if flag unset */ 3826 __u8 pad1[6]; /* ignored */ 3827 __u64 old_addr; /* ignored if flag unset */ 3828 }; 3829 __u32 sida_offset; /* offset into the sida */ 3830 __u8 reserved[32]; /* ignored */ 3831 }; 3832 }; 3833 3834The start address of the memory region has to be specified in the "gaddr" 3835field, and the length of the region in the "size" field (which must not 3836be 0). The maximum value for "size" can be obtained by checking the 3837KVM_CAP_S390_MEM_OP capability. "buf" is the buffer supplied by the 3838userspace application where the read data should be written to for 3839a read access, or where the data that should be written is stored for 3840a write access. The "reserved" field is meant for future extensions. 3841Reserved and unused values are ignored. Future extension that add members must 3842introduce new flags. 3843 3844The type of operation is specified in the "op" field. Flags modifying 3845their behavior can be set in the "flags" field. Undefined flag bits must 3846be set to 0. 3847 3848Possible operations are: 3849 * ``KVM_S390_MEMOP_LOGICAL_READ`` 3850 * ``KVM_S390_MEMOP_LOGICAL_WRITE`` 3851 * ``KVM_S390_MEMOP_ABSOLUTE_READ`` 3852 * ``KVM_S390_MEMOP_ABSOLUTE_WRITE`` 3853 * ``KVM_S390_MEMOP_SIDA_READ`` 3854 * ``KVM_S390_MEMOP_SIDA_WRITE`` 3855 * ``KVM_S390_MEMOP_ABSOLUTE_CMPXCHG`` 3856 3857Logical read/write: 3858^^^^^^^^^^^^^^^^^^^ 3859 3860Access logical memory, i.e. translate the given guest address to an absolute 3861address given the state of the VCPU and use the absolute address as target of 3862the access. "ar" designates the access register number to be used; the valid 3863range is 0..15. 3864Logical accesses are permitted for the VCPU ioctl only. 3865Logical accesses are permitted for non-protected guests only. 3866 3867Supported flags: 3868 * ``KVM_S390_MEMOP_F_CHECK_ONLY`` 3869 * ``KVM_S390_MEMOP_F_INJECT_EXCEPTION`` 3870 * ``KVM_S390_MEMOP_F_SKEY_PROTECTION`` 3871 3872The KVM_S390_MEMOP_F_CHECK_ONLY flag can be set to check whether the 3873corresponding memory access would cause an access exception; however, 3874no actual access to the data in memory at the destination is performed. 3875In this case, "buf" is unused and can be NULL. 3876 3877In case an access exception occurred during the access (or would occur 3878in case of KVM_S390_MEMOP_F_CHECK_ONLY), the ioctl returns a positive 3879error number indicating the type of exception. This exception is also 3880raised directly at the corresponding VCPU if the flag 3881KVM_S390_MEMOP_F_INJECT_EXCEPTION is set. 3882On protection exceptions, unless specified otherwise, the injected 3883translation-exception identifier (TEID) indicates suppression. 3884 3885If the KVM_S390_MEMOP_F_SKEY_PROTECTION flag is set, storage key 3886protection is also in effect and may cause exceptions if accesses are 3887prohibited given the access key designated by "key"; the valid range is 0..15. 3888KVM_S390_MEMOP_F_SKEY_PROTECTION is available if KVM_CAP_S390_MEM_OP_EXTENSION 3889is > 0. 3890Since the accessed memory may span multiple pages and those pages might have 3891different storage keys, it is possible that a protection exception occurs 3892after memory has been modified. In this case, if the exception is injected, 3893the TEID does not indicate suppression. 3894 3895Absolute read/write: 3896^^^^^^^^^^^^^^^^^^^^ 3897 3898Access absolute memory. This operation is intended to be used with the 3899KVM_S390_MEMOP_F_SKEY_PROTECTION flag, to allow accessing memory and performing 3900the checks required for storage key protection as one operation (as opposed to 3901user space getting the storage keys, performing the checks, and accessing 3902memory thereafter, which could lead to a delay between check and access). 3903Absolute accesses are permitted for the VM ioctl if KVM_CAP_S390_MEM_OP_EXTENSION 3904has the KVM_S390_MEMOP_EXTENSION_CAP_BASE bit set. 3905Currently absolute accesses are not permitted for VCPU ioctls. 3906Absolute accesses are permitted for non-protected guests only. 3907 3908Supported flags: 3909 * ``KVM_S390_MEMOP_F_CHECK_ONLY`` 3910 * ``KVM_S390_MEMOP_F_SKEY_PROTECTION`` 3911 3912The semantics of the flags common with logical accesses are as for logical 3913accesses. 3914 3915Absolute cmpxchg: 3916^^^^^^^^^^^^^^^^^ 3917 3918Perform cmpxchg on absolute guest memory. Intended for use with the 3919KVM_S390_MEMOP_F_SKEY_PROTECTION flag. 3920Instead of doing an unconditional write, the access occurs only if the target 3921location contains the value pointed to by "old_addr". 3922This is performed as an atomic cmpxchg with the length specified by the "size" 3923parameter. "size" must be a power of two up to and including 16. 3924If the exchange did not take place because the target value doesn't match the 3925old value, the value "old_addr" points to is replaced by the target value. 3926User space can tell if an exchange took place by checking if this replacement 3927occurred. The cmpxchg op is permitted for the VM ioctl if 3928KVM_CAP_S390_MEM_OP_EXTENSION has flag KVM_S390_MEMOP_EXTENSION_CAP_CMPXCHG set. 3929 3930Supported flags: 3931 * ``KVM_S390_MEMOP_F_SKEY_PROTECTION`` 3932 3933SIDA read/write: 3934^^^^^^^^^^^^^^^^ 3935 3936Access the secure instruction data area which contains memory operands necessary 3937for instruction emulation for protected guests. 3938SIDA accesses are available if the KVM_CAP_S390_PROTECTED capability is available. 3939SIDA accesses are permitted for the VCPU ioctl only. 3940SIDA accesses are permitted for protected guests only. 3941 3942No flags are supported. 3943 39444.90 KVM_S390_GET_SKEYS 3945----------------------- 3946 3947:Capability: KVM_CAP_S390_SKEYS 3948:Architectures: s390 3949:Type: vm ioctl 3950:Parameters: struct kvm_s390_skeys 3951:Returns: 0 on success, KVM_S390_GET_SKEYS_NONE if guest is not using storage 3952 keys, negative value on error 3953 3954This ioctl is used to get guest storage key values on the s390 3955architecture. The ioctl takes parameters via the kvm_s390_skeys struct:: 3956 3957 struct kvm_s390_skeys { 3958 __u64 start_gfn; 3959 __u64 count; 3960 __u64 skeydata_addr; 3961 __u32 flags; 3962 __u32 reserved[9]; 3963 }; 3964 3965The start_gfn field is the number of the first guest frame whose storage keys 3966you want to get. 3967 3968The count field is the number of consecutive frames (starting from start_gfn) 3969whose storage keys to get. The count field must be at least 1 and the maximum 3970allowed value is defined as KVM_S390_SKEYS_MAX. Values outside this range 3971will cause the ioctl to return -EINVAL. 3972 3973The skeydata_addr field is the address to a buffer large enough to hold count 3974bytes. This buffer will be filled with storage key data by the ioctl. 3975 39764.91 KVM_S390_SET_SKEYS 3977----------------------- 3978 3979:Capability: KVM_CAP_S390_SKEYS 3980:Architectures: s390 3981:Type: vm ioctl 3982:Parameters: struct kvm_s390_skeys 3983:Returns: 0 on success, negative value on error 3984 3985This ioctl is used to set guest storage key values on the s390 3986architecture. The ioctl takes parameters via the kvm_s390_skeys struct. 3987See section on KVM_S390_GET_SKEYS for struct definition. 3988 3989The start_gfn field is the number of the first guest frame whose storage keys 3990you want to set. 3991 3992The count field is the number of consecutive frames (starting from start_gfn) 3993whose storage keys to get. The count field must be at least 1 and the maximum 3994allowed value is defined as KVM_S390_SKEYS_MAX. Values outside this range 3995will cause the ioctl to return -EINVAL. 3996 3997The skeydata_addr field is the address to a buffer containing count bytes of 3998storage keys. Each byte in the buffer will be set as the storage key for a 3999single frame starting at start_gfn for count frames. 4000 4001Note: If any architecturally invalid key value is found in the given data then 4002the ioctl will return -EINVAL. 4003 40044.92 KVM_S390_IRQ 4005----------------- 4006 4007:Capability: KVM_CAP_S390_INJECT_IRQ 4008:Architectures: s390 4009:Type: vcpu ioctl 4010:Parameters: struct kvm_s390_irq (in) 4011:Returns: 0 on success, -1 on error 4012 4013Errors: 4014 4015 4016 ====== ================================================================= 4017 EINVAL interrupt type is invalid 4018 type is KVM_S390_SIGP_STOP and flag parameter is invalid value, 4019 type is KVM_S390_INT_EXTERNAL_CALL and code is bigger 4020 than the maximum of VCPUs 4021 EBUSY type is KVM_S390_SIGP_SET_PREFIX and vcpu is not stopped, 4022 type is KVM_S390_SIGP_STOP and a stop irq is already pending, 4023 type is KVM_S390_INT_EXTERNAL_CALL and an external call interrupt 4024 is already pending 4025 ====== ================================================================= 4026 4027Allows to inject an interrupt to the guest. 4028 4029Using struct kvm_s390_irq as a parameter allows 4030to inject additional payload which is not 4031possible via KVM_S390_INTERRUPT. 4032 4033Interrupt parameters are passed via kvm_s390_irq:: 4034 4035 struct kvm_s390_irq { 4036 __u64 type; 4037 union { 4038 struct kvm_s390_io_info io; 4039 struct kvm_s390_ext_info ext; 4040 struct kvm_s390_pgm_info pgm; 4041 struct kvm_s390_emerg_info emerg; 4042 struct kvm_s390_extcall_info extcall; 4043 struct kvm_s390_prefix_info prefix; 4044 struct kvm_s390_stop_info stop; 4045 struct kvm_s390_mchk_info mchk; 4046 char reserved[64]; 4047 } u; 4048 }; 4049 4050type can be one of the following: 4051 4052- KVM_S390_SIGP_STOP - sigp stop; parameter in .stop 4053- KVM_S390_PROGRAM_INT - program check; parameters in .pgm 4054- KVM_S390_SIGP_SET_PREFIX - sigp set prefix; parameters in .prefix 4055- KVM_S390_RESTART - restart; no parameters 4056- KVM_S390_INT_CLOCK_COMP - clock comparator interrupt; no parameters 4057- KVM_S390_INT_CPU_TIMER - CPU timer interrupt; no parameters 4058- KVM_S390_INT_EMERGENCY - sigp emergency; parameters in .emerg 4059- KVM_S390_INT_EXTERNAL_CALL - sigp external call; parameters in .extcall 4060- KVM_S390_MCHK - machine check interrupt; parameters in .mchk 4061 4062This is an asynchronous vcpu ioctl and can be invoked from any thread. 4063 40644.94 KVM_S390_GET_IRQ_STATE 4065--------------------------- 4066 4067:Capability: KVM_CAP_S390_IRQ_STATE 4068:Architectures: s390 4069:Type: vcpu ioctl 4070:Parameters: struct kvm_s390_irq_state (out) 4071:Returns: >= number of bytes copied into buffer, 4072 -EINVAL if buffer size is 0, 4073 -ENOBUFS if buffer size is too small to fit all pending interrupts, 4074 -EFAULT if the buffer address was invalid 4075 4076This ioctl allows userspace to retrieve the complete state of all currently 4077pending interrupts in a single buffer. Use cases include migration 4078and introspection. The parameter structure contains the address of a 4079userspace buffer and its length:: 4080 4081 struct kvm_s390_irq_state { 4082 __u64 buf; 4083 __u32 flags; /* will stay unused for compatibility reasons */ 4084 __u32 len; 4085 __u32 reserved[4]; /* will stay unused for compatibility reasons */ 4086 }; 4087 4088Userspace passes in the above struct and for each pending interrupt a 4089struct kvm_s390_irq is copied to the provided buffer. 4090 4091The structure contains a flags and a reserved field for future extensions. As 4092the kernel never checked for flags == 0 and QEMU never pre-zeroed flags and 4093reserved, these fields can not be used in the future without breaking 4094compatibility. 4095 4096If -ENOBUFS is returned the buffer provided was too small and userspace 4097may retry with a bigger buffer. 4098 40994.95 KVM_S390_SET_IRQ_STATE 4100--------------------------- 4101 4102:Capability: KVM_CAP_S390_IRQ_STATE 4103:Architectures: s390 4104:Type: vcpu ioctl 4105:Parameters: struct kvm_s390_irq_state (in) 4106:Returns: 0 on success, 4107 -EFAULT if the buffer address was invalid, 4108 -EINVAL for an invalid buffer length (see below), 4109 -EBUSY if there were already interrupts pending, 4110 errors occurring when actually injecting the 4111 interrupt. See KVM_S390_IRQ. 4112 4113This ioctl allows userspace to set the complete state of all cpu-local 4114interrupts currently pending for the vcpu. It is intended for restoring 4115interrupt state after a migration. The input parameter is a userspace buffer 4116containing a struct kvm_s390_irq_state:: 4117 4118 struct kvm_s390_irq_state { 4119 __u64 buf; 4120 __u32 flags; /* will stay unused for compatibility reasons */ 4121 __u32 len; 4122 __u32 reserved[4]; /* will stay unused for compatibility reasons */ 4123 }; 4124 4125The restrictions for flags and reserved apply as well. 4126(see KVM_S390_GET_IRQ_STATE) 4127 4128The userspace memory referenced by buf contains a struct kvm_s390_irq 4129for each interrupt to be injected into the guest. 4130If one of the interrupts could not be injected for some reason the 4131ioctl aborts. 4132 4133len must be a multiple of sizeof(struct kvm_s390_irq). It must be > 0 4134and it must not exceed (max_vcpus + 32) * sizeof(struct kvm_s390_irq), 4135which is the maximum number of possibly pending cpu-local interrupts. 4136 41374.96 KVM_SMI 4138------------ 4139 4140:Capability: KVM_CAP_X86_SMM 4141:Architectures: x86 4142:Type: vcpu ioctl 4143:Parameters: none 4144:Returns: 0 on success, -1 on error 4145 4146Queues an SMI on the thread's vcpu. 4147 41484.97 KVM_X86_SET_MSR_FILTER 4149---------------------------- 4150 4151:Capability: KVM_CAP_X86_MSR_FILTER 4152:Architectures: x86 4153:Type: vm ioctl 4154:Parameters: struct kvm_msr_filter 4155:Returns: 0 on success, < 0 on error 4156 4157:: 4158 4159 struct kvm_msr_filter_range { 4160 #define KVM_MSR_FILTER_READ (1 << 0) 4161 #define KVM_MSR_FILTER_WRITE (1 << 1) 4162 __u32 flags; 4163 __u32 nmsrs; /* number of msrs in bitmap */ 4164 __u32 base; /* MSR index the bitmap starts at */ 4165 __u8 *bitmap; /* a 1 bit allows the operations in flags, 0 denies */ 4166 }; 4167 4168 #define KVM_MSR_FILTER_MAX_RANGES 16 4169 struct kvm_msr_filter { 4170 #define KVM_MSR_FILTER_DEFAULT_ALLOW (0 << 0) 4171 #define KVM_MSR_FILTER_DEFAULT_DENY (1 << 0) 4172 __u32 flags; 4173 struct kvm_msr_filter_range ranges[KVM_MSR_FILTER_MAX_RANGES]; 4174 }; 4175 4176flags values for ``struct kvm_msr_filter_range``: 4177 4178``KVM_MSR_FILTER_READ`` 4179 4180 Filter read accesses to MSRs using the given bitmap. A 0 in the bitmap 4181 indicates that read accesses should be denied, while a 1 indicates that 4182 a read for a particular MSR should be allowed regardless of the default 4183 filter action. 4184 4185``KVM_MSR_FILTER_WRITE`` 4186 4187 Filter write accesses to MSRs using the given bitmap. A 0 in the bitmap 4188 indicates that write accesses should be denied, while a 1 indicates that 4189 a write for a particular MSR should be allowed regardless of the default 4190 filter action. 4191 4192flags values for ``struct kvm_msr_filter``: 4193 4194``KVM_MSR_FILTER_DEFAULT_ALLOW`` 4195 4196 If no filter range matches an MSR index that is getting accessed, KVM will 4197 allow accesses to all MSRs by default. 4198 4199``KVM_MSR_FILTER_DEFAULT_DENY`` 4200 4201 If no filter range matches an MSR index that is getting accessed, KVM will 4202 deny accesses to all MSRs by default. 4203 4204This ioctl allows userspace to define up to 16 bitmaps of MSR ranges to deny 4205guest MSR accesses that would normally be allowed by KVM. If an MSR is not 4206covered by a specific range, the "default" filtering behavior applies. Each 4207bitmap range covers MSRs from [base .. base+nmsrs). 4208 4209If an MSR access is denied by userspace, the resulting KVM behavior depends on 4210whether or not KVM_CAP_X86_USER_SPACE_MSR's KVM_MSR_EXIT_REASON_FILTER is 4211enabled. If KVM_MSR_EXIT_REASON_FILTER is enabled, KVM will exit to userspace 4212on denied accesses, i.e. userspace effectively intercepts the MSR access. If 4213KVM_MSR_EXIT_REASON_FILTER is not enabled, KVM will inject a #GP into the guest 4214on denied accesses. 4215 4216If an MSR access is allowed by userspace, KVM will emulate and/or virtualize 4217the access in accordance with the vCPU model. Note, KVM may still ultimately 4218inject a #GP if an access is allowed by userspace, e.g. if KVM doesn't support 4219the MSR, or to follow architectural behavior for the MSR. 4220 4221By default, KVM operates in KVM_MSR_FILTER_DEFAULT_ALLOW mode with no MSR range 4222filters. 4223 4224Calling this ioctl with an empty set of ranges (all nmsrs == 0) disables MSR 4225filtering. In that mode, ``KVM_MSR_FILTER_DEFAULT_DENY`` is invalid and causes 4226an error. 4227 4228.. warning:: 4229 MSR accesses as part of nested VM-Enter/VM-Exit are not filtered. 4230 This includes both writes to individual VMCS fields and reads/writes 4231 through the MSR lists pointed to by the VMCS. 4232 4233 x2APIC MSR accesses cannot be filtered (KVM silently ignores filters that 4234 cover any x2APIC MSRs). 4235 4236Note, invoking this ioctl while a vCPU is running is inherently racy. However, 4237KVM does guarantee that vCPUs will see either the previous filter or the new 4238filter, e.g. MSRs with identical settings in both the old and new filter will 4239have deterministic behavior. 4240 4241Similarly, if userspace wishes to intercept on denied accesses, 4242KVM_MSR_EXIT_REASON_FILTER must be enabled before activating any filters, and 4243left enabled until after all filters are deactivated. Failure to do so may 4244result in KVM injecting a #GP instead of exiting to userspace. 4245 42464.98 KVM_CREATE_SPAPR_TCE_64 4247---------------------------- 4248 4249:Capability: KVM_CAP_SPAPR_TCE_64 4250:Architectures: powerpc 4251:Type: vm ioctl 4252:Parameters: struct kvm_create_spapr_tce_64 (in) 4253:Returns: file descriptor for manipulating the created TCE table 4254 4255This is an extension for KVM_CAP_SPAPR_TCE which only supports 32bit 4256windows, described in 4.62 KVM_CREATE_SPAPR_TCE 4257 4258This capability uses extended struct in ioctl interface:: 4259 4260 /* for KVM_CAP_SPAPR_TCE_64 */ 4261 struct kvm_create_spapr_tce_64 { 4262 __u64 liobn; 4263 __u32 page_shift; 4264 __u32 flags; 4265 __u64 offset; /* in pages */ 4266 __u64 size; /* in pages */ 4267 }; 4268 4269The aim of extension is to support an additional bigger DMA window with 4270a variable page size. 4271KVM_CREATE_SPAPR_TCE_64 receives a 64bit window size, an IOMMU page shift and 4272a bus offset of the corresponding DMA window, @size and @offset are numbers 4273of IOMMU pages. 4274 4275@flags are not used at the moment. 4276 4277The rest of functionality is identical to KVM_CREATE_SPAPR_TCE. 4278 42794.99 KVM_REINJECT_CONTROL 4280------------------------- 4281 4282:Capability: KVM_CAP_REINJECT_CONTROL 4283:Architectures: x86 4284:Type: vm ioctl 4285:Parameters: struct kvm_reinject_control (in) 4286:Returns: 0 on success, 4287 -EFAULT if struct kvm_reinject_control cannot be read, 4288 -ENXIO if KVM_CREATE_PIT or KVM_CREATE_PIT2 didn't succeed earlier. 4289 4290i8254 (PIT) has two modes, reinject and !reinject. The default is reinject, 4291where KVM queues elapsed i8254 ticks and monitors completion of interrupt from 4292vector(s) that i8254 injects. Reinject mode dequeues a tick and injects its 4293interrupt whenever there isn't a pending interrupt from i8254. 4294!reinject mode injects an interrupt as soon as a tick arrives. 4295 4296:: 4297 4298 struct kvm_reinject_control { 4299 __u8 pit_reinject; 4300 __u8 reserved[31]; 4301 }; 4302 4303pit_reinject = 0 (!reinject mode) is recommended, unless running an old 4304operating system that uses the PIT for timing (e.g. Linux 2.4.x). 4305 43064.100 KVM_PPC_CONFIGURE_V3_MMU 4307------------------------------ 4308 4309:Capability: KVM_CAP_PPC_MMU_RADIX or KVM_CAP_PPC_MMU_HASH_V3 4310:Architectures: ppc 4311:Type: vm ioctl 4312:Parameters: struct kvm_ppc_mmuv3_cfg (in) 4313:Returns: 0 on success, 4314 -EFAULT if struct kvm_ppc_mmuv3_cfg cannot be read, 4315 -EINVAL if the configuration is invalid 4316 4317This ioctl controls whether the guest will use radix or HPT (hashed 4318page table) translation, and sets the pointer to the process table for 4319the guest. 4320 4321:: 4322 4323 struct kvm_ppc_mmuv3_cfg { 4324 __u64 flags; 4325 __u64 process_table; 4326 }; 4327 4328There are two bits that can be set in flags; KVM_PPC_MMUV3_RADIX and 4329KVM_PPC_MMUV3_GTSE. KVM_PPC_MMUV3_RADIX, if set, configures the guest 4330to use radix tree translation, and if clear, to use HPT translation. 4331KVM_PPC_MMUV3_GTSE, if set and if KVM permits it, configures the guest 4332to be able to use the global TLB and SLB invalidation instructions; 4333if clear, the guest may not use these instructions. 4334 4335The process_table field specifies the address and size of the guest 4336process table, which is in the guest's space. This field is formatted 4337as the second doubleword of the partition table entry, as defined in 4338the Power ISA V3.00, Book III section 5.7.6.1. 4339 43404.101 KVM_PPC_GET_RMMU_INFO 4341--------------------------- 4342 4343:Capability: KVM_CAP_PPC_MMU_RADIX 4344:Architectures: ppc 4345:Type: vm ioctl 4346:Parameters: struct kvm_ppc_rmmu_info (out) 4347:Returns: 0 on success, 4348 -EFAULT if struct kvm_ppc_rmmu_info cannot be written, 4349 -EINVAL if no useful information can be returned 4350 4351This ioctl returns a structure containing two things: (a) a list 4352containing supported radix tree geometries, and (b) a list that maps 4353page sizes to put in the "AP" (actual page size) field for the tlbie 4354(TLB invalidate entry) instruction. 4355 4356:: 4357 4358 struct kvm_ppc_rmmu_info { 4359 struct kvm_ppc_radix_geom { 4360 __u8 page_shift; 4361 __u8 level_bits[4]; 4362 __u8 pad[3]; 4363 } geometries[8]; 4364 __u32 ap_encodings[8]; 4365 }; 4366 4367The geometries[] field gives up to 8 supported geometries for the 4368radix page table, in terms of the log base 2 of the smallest page 4369size, and the number of bits indexed at each level of the tree, from 4370the PTE level up to the PGD level in that order. Any unused entries 4371will have 0 in the page_shift field. 4372 4373The ap_encodings gives the supported page sizes and their AP field 4374encodings, encoded with the AP value in the top 3 bits and the log 4375base 2 of the page size in the bottom 6 bits. 4376 43774.102 KVM_PPC_RESIZE_HPT_PREPARE 4378-------------------------------- 4379 4380:Capability: KVM_CAP_SPAPR_RESIZE_HPT 4381:Architectures: powerpc 4382:Type: vm ioctl 4383:Parameters: struct kvm_ppc_resize_hpt (in) 4384:Returns: 0 on successful completion, 4385 >0 if a new HPT is being prepared, the value is an estimated 4386 number of milliseconds until preparation is complete, 4387 -EFAULT if struct kvm_reinject_control cannot be read, 4388 -EINVAL if the supplied shift or flags are invalid, 4389 -ENOMEM if unable to allocate the new HPT, 4390 4391Used to implement the PAPR extension for runtime resizing of a guest's 4392Hashed Page Table (HPT). Specifically this starts, stops or monitors 4393the preparation of a new potential HPT for the guest, essentially 4394implementing the H_RESIZE_HPT_PREPARE hypercall. 4395 4396:: 4397 4398 struct kvm_ppc_resize_hpt { 4399 __u64 flags; 4400 __u32 shift; 4401 __u32 pad; 4402 }; 4403 4404If called with shift > 0 when there is no pending HPT for the guest, 4405this begins preparation of a new pending HPT of size 2^(shift) bytes. 4406It then returns a positive integer with the estimated number of 4407milliseconds until preparation is complete. 4408 4409If called when there is a pending HPT whose size does not match that 4410requested in the parameters, discards the existing pending HPT and 4411creates a new one as above. 4412 4413If called when there is a pending HPT of the size requested, will: 4414 4415 * If preparation of the pending HPT is already complete, return 0 4416 * If preparation of the pending HPT has failed, return an error 4417 code, then discard the pending HPT. 4418 * If preparation of the pending HPT is still in progress, return an 4419 estimated number of milliseconds until preparation is complete. 4420 4421If called with shift == 0, discards any currently pending HPT and 4422returns 0 (i.e. cancels any in-progress preparation). 4423 4424flags is reserved for future expansion, currently setting any bits in 4425flags will result in an -EINVAL. 4426 4427Normally this will be called repeatedly with the same parameters until 4428it returns <= 0. The first call will initiate preparation, subsequent 4429ones will monitor preparation until it completes or fails. 4430 44314.103 KVM_PPC_RESIZE_HPT_COMMIT 4432------------------------------- 4433 4434:Capability: KVM_CAP_SPAPR_RESIZE_HPT 4435:Architectures: powerpc 4436:Type: vm ioctl 4437:Parameters: struct kvm_ppc_resize_hpt (in) 4438:Returns: 0 on successful completion, 4439 -EFAULT if struct kvm_reinject_control cannot be read, 4440 -EINVAL if the supplied shift or flags are invalid, 4441 -ENXIO is there is no pending HPT, or the pending HPT doesn't 4442 have the requested size, 4443 -EBUSY if the pending HPT is not fully prepared, 4444 -ENOSPC if there was a hash collision when moving existing 4445 HPT entries to the new HPT, 4446 -EIO on other error conditions 4447 4448Used to implement the PAPR extension for runtime resizing of a guest's 4449Hashed Page Table (HPT). Specifically this requests that the guest be 4450transferred to working with the new HPT, essentially implementing the 4451H_RESIZE_HPT_COMMIT hypercall. 4452 4453:: 4454 4455 struct kvm_ppc_resize_hpt { 4456 __u64 flags; 4457 __u32 shift; 4458 __u32 pad; 4459 }; 4460 4461This should only be called after KVM_PPC_RESIZE_HPT_PREPARE has 4462returned 0 with the same parameters. In other cases 4463KVM_PPC_RESIZE_HPT_COMMIT will return an error (usually -ENXIO or 4464-EBUSY, though others may be possible if the preparation was started, 4465but failed). 4466 4467This will have undefined effects on the guest if it has not already 4468placed itself in a quiescent state where no vcpu will make MMU enabled 4469memory accesses. 4470 4471On successful completion, the pending HPT will become the guest's active 4472HPT and the previous HPT will be discarded. 4473 4474On failure, the guest will still be operating on its previous HPT. 4475 44764.104 KVM_X86_GET_MCE_CAP_SUPPORTED 4477----------------------------------- 4478 4479:Capability: KVM_CAP_MCE 4480:Architectures: x86 4481:Type: system ioctl 4482:Parameters: u64 mce_cap (out) 4483:Returns: 0 on success, -1 on error 4484 4485Returns supported MCE capabilities. The u64 mce_cap parameter 4486has the same format as the MSR_IA32_MCG_CAP register. Supported 4487capabilities will have the corresponding bits set. 4488 44894.105 KVM_X86_SETUP_MCE 4490----------------------- 4491 4492:Capability: KVM_CAP_MCE 4493:Architectures: x86 4494:Type: vcpu ioctl 4495:Parameters: u64 mcg_cap (in) 4496:Returns: 0 on success, 4497 -EFAULT if u64 mcg_cap cannot be read, 4498 -EINVAL if the requested number of banks is invalid, 4499 -EINVAL if requested MCE capability is not supported. 4500 4501Initializes MCE support for use. The u64 mcg_cap parameter 4502has the same format as the MSR_IA32_MCG_CAP register and 4503specifies which capabilities should be enabled. The maximum 4504supported number of error-reporting banks can be retrieved when 4505checking for KVM_CAP_MCE. The supported capabilities can be 4506retrieved with KVM_X86_GET_MCE_CAP_SUPPORTED. 4507 45084.106 KVM_X86_SET_MCE 4509--------------------- 4510 4511:Capability: KVM_CAP_MCE 4512:Architectures: x86 4513:Type: vcpu ioctl 4514:Parameters: struct kvm_x86_mce (in) 4515:Returns: 0 on success, 4516 -EFAULT if struct kvm_x86_mce cannot be read, 4517 -EINVAL if the bank number is invalid, 4518 -EINVAL if VAL bit is not set in status field. 4519 4520Inject a machine check error (MCE) into the guest. The input 4521parameter is:: 4522 4523 struct kvm_x86_mce { 4524 __u64 status; 4525 __u64 addr; 4526 __u64 misc; 4527 __u64 mcg_status; 4528 __u8 bank; 4529 __u8 pad1[7]; 4530 __u64 pad2[3]; 4531 }; 4532 4533If the MCE being reported is an uncorrected error, KVM will 4534inject it as an MCE exception into the guest. If the guest 4535MCG_STATUS register reports that an MCE is in progress, KVM 4536causes an KVM_EXIT_SHUTDOWN vmexit. 4537 4538Otherwise, if the MCE is a corrected error, KVM will just 4539store it in the corresponding bank (provided this bank is 4540not holding a previously reported uncorrected error). 4541 45424.107 KVM_S390_GET_CMMA_BITS 4543---------------------------- 4544 4545:Capability: KVM_CAP_S390_CMMA_MIGRATION 4546:Architectures: s390 4547:Type: vm ioctl 4548:Parameters: struct kvm_s390_cmma_log (in, out) 4549:Returns: 0 on success, a negative value on error 4550 4551Errors: 4552 4553 ====== ============================================================= 4554 ENOMEM not enough memory can be allocated to complete the task 4555 ENXIO if CMMA is not enabled 4556 EINVAL if KVM_S390_CMMA_PEEK is not set but migration mode was not enabled 4557 EINVAL if KVM_S390_CMMA_PEEK is not set but dirty tracking has been 4558 disabled (and thus migration mode was automatically disabled) 4559 EFAULT if the userspace address is invalid or if no page table is 4560 present for the addresses (e.g. when using hugepages). 4561 ====== ============================================================= 4562 4563This ioctl is used to get the values of the CMMA bits on the s390 4564architecture. It is meant to be used in two scenarios: 4565 4566- During live migration to save the CMMA values. Live migration needs 4567 to be enabled via the KVM_REQ_START_MIGRATION VM property. 4568- To non-destructively peek at the CMMA values, with the flag 4569 KVM_S390_CMMA_PEEK set. 4570 4571The ioctl takes parameters via the kvm_s390_cmma_log struct. The desired 4572values are written to a buffer whose location is indicated via the "values" 4573member in the kvm_s390_cmma_log struct. The values in the input struct are 4574also updated as needed. 4575 4576Each CMMA value takes up one byte. 4577 4578:: 4579 4580 struct kvm_s390_cmma_log { 4581 __u64 start_gfn; 4582 __u32 count; 4583 __u32 flags; 4584 union { 4585 __u64 remaining; 4586 __u64 mask; 4587 }; 4588 __u64 values; 4589 }; 4590 4591start_gfn is the number of the first guest frame whose CMMA values are 4592to be retrieved, 4593 4594count is the length of the buffer in bytes, 4595 4596values points to the buffer where the result will be written to. 4597 4598If count is greater than KVM_S390_SKEYS_MAX, then it is considered to be 4599KVM_S390_SKEYS_MAX. KVM_S390_SKEYS_MAX is re-used for consistency with 4600other ioctls. 4601 4602The result is written in the buffer pointed to by the field values, and 4603the values of the input parameter are updated as follows. 4604 4605Depending on the flags, different actions are performed. The only 4606supported flag so far is KVM_S390_CMMA_PEEK. 4607 4608The default behaviour if KVM_S390_CMMA_PEEK is not set is: 4609start_gfn will indicate the first page frame whose CMMA bits were dirty. 4610It is not necessarily the same as the one passed as input, as clean pages 4611are skipped. 4612 4613count will indicate the number of bytes actually written in the buffer. 4614It can (and very often will) be smaller than the input value, since the 4615buffer is only filled until 16 bytes of clean values are found (which 4616are then not copied in the buffer). Since a CMMA migration block needs 4617the base address and the length, for a total of 16 bytes, we will send 4618back some clean data if there is some dirty data afterwards, as long as 4619the size of the clean data does not exceed the size of the header. This 4620allows to minimize the amount of data to be saved or transferred over 4621the network at the expense of more roundtrips to userspace. The next 4622invocation of the ioctl will skip over all the clean values, saving 4623potentially more than just the 16 bytes we found. 4624 4625If KVM_S390_CMMA_PEEK is set: 4626the existing storage attributes are read even when not in migration 4627mode, and no other action is performed; 4628 4629the output start_gfn will be equal to the input start_gfn, 4630 4631the output count will be equal to the input count, except if the end of 4632memory has been reached. 4633 4634In both cases: 4635the field "remaining" will indicate the total number of dirty CMMA values 4636still remaining, or 0 if KVM_S390_CMMA_PEEK is set and migration mode is 4637not enabled. 4638 4639mask is unused. 4640 4641values points to the userspace buffer where the result will be stored. 4642 46434.108 KVM_S390_SET_CMMA_BITS 4644---------------------------- 4645 4646:Capability: KVM_CAP_S390_CMMA_MIGRATION 4647:Architectures: s390 4648:Type: vm ioctl 4649:Parameters: struct kvm_s390_cmma_log (in) 4650:Returns: 0 on success, a negative value on error 4651 4652This ioctl is used to set the values of the CMMA bits on the s390 4653architecture. It is meant to be used during live migration to restore 4654the CMMA values, but there are no restrictions on its use. 4655The ioctl takes parameters via the kvm_s390_cmma_values struct. 4656Each CMMA value takes up one byte. 4657 4658:: 4659 4660 struct kvm_s390_cmma_log { 4661 __u64 start_gfn; 4662 __u32 count; 4663 __u32 flags; 4664 union { 4665 __u64 remaining; 4666 __u64 mask; 4667 }; 4668 __u64 values; 4669 }; 4670 4671start_gfn indicates the starting guest frame number, 4672 4673count indicates how many values are to be considered in the buffer, 4674 4675flags is not used and must be 0. 4676 4677mask indicates which PGSTE bits are to be considered. 4678 4679remaining is not used. 4680 4681values points to the buffer in userspace where to store the values. 4682 4683This ioctl can fail with -ENOMEM if not enough memory can be allocated to 4684complete the task, with -ENXIO if CMMA is not enabled, with -EINVAL if 4685the count field is too large (e.g. more than KVM_S390_CMMA_SIZE_MAX) or 4686if the flags field was not 0, with -EFAULT if the userspace address is 4687invalid, if invalid pages are written to (e.g. after the end of memory) 4688or if no page table is present for the addresses (e.g. when using 4689hugepages). 4690 46914.109 KVM_PPC_GET_CPU_CHAR 4692-------------------------- 4693 4694:Capability: KVM_CAP_PPC_GET_CPU_CHAR 4695:Architectures: powerpc 4696:Type: vm ioctl 4697:Parameters: struct kvm_ppc_cpu_char (out) 4698:Returns: 0 on successful completion, 4699 -EFAULT if struct kvm_ppc_cpu_char cannot be written 4700 4701This ioctl gives userspace information about certain characteristics 4702of the CPU relating to speculative execution of instructions and 4703possible information leakage resulting from speculative execution (see 4704CVE-2017-5715, CVE-2017-5753 and CVE-2017-5754). The information is 4705returned in struct kvm_ppc_cpu_char, which looks like this:: 4706 4707 struct kvm_ppc_cpu_char { 4708 __u64 character; /* characteristics of the CPU */ 4709 __u64 behaviour; /* recommended software behaviour */ 4710 __u64 character_mask; /* valid bits in character */ 4711 __u64 behaviour_mask; /* valid bits in behaviour */ 4712 }; 4713 4714For extensibility, the character_mask and behaviour_mask fields 4715indicate which bits of character and behaviour have been filled in by 4716the kernel. If the set of defined bits is extended in future then 4717userspace will be able to tell whether it is running on a kernel that 4718knows about the new bits. 4719 4720The character field describes attributes of the CPU which can help 4721with preventing inadvertent information disclosure - specifically, 4722whether there is an instruction to flash-invalidate the L1 data cache 4723(ori 30,30,0 or mtspr SPRN_TRIG2,rN), whether the L1 data cache is set 4724to a mode where entries can only be used by the thread that created 4725them, whether the bcctr[l] instruction prevents speculation, and 4726whether a speculation barrier instruction (ori 31,31,0) is provided. 4727 4728The behaviour field describes actions that software should take to 4729prevent inadvertent information disclosure, and thus describes which 4730vulnerabilities the hardware is subject to; specifically whether the 4731L1 data cache should be flushed when returning to user mode from the 4732kernel, and whether a speculation barrier should be placed between an 4733array bounds check and the array access. 4734 4735These fields use the same bit definitions as the new 4736H_GET_CPU_CHARACTERISTICS hypercall. 4737 47384.110 KVM_MEMORY_ENCRYPT_OP 4739--------------------------- 4740 4741:Capability: basic 4742:Architectures: x86 4743:Type: vm 4744:Parameters: an opaque platform specific structure (in/out) 4745:Returns: 0 on success; -1 on error 4746 4747If the platform supports creating encrypted VMs then this ioctl can be used 4748for issuing platform-specific memory encryption commands to manage those 4749encrypted VMs. 4750 4751Currently, this ioctl is used for issuing Secure Encrypted Virtualization 4752(SEV) commands on AMD Processors. The SEV commands are defined in 4753Documentation/virt/kvm/x86/amd-memory-encryption.rst. 4754 47554.111 KVM_MEMORY_ENCRYPT_REG_REGION 4756----------------------------------- 4757 4758:Capability: basic 4759:Architectures: x86 4760:Type: system 4761:Parameters: struct kvm_enc_region (in) 4762:Returns: 0 on success; -1 on error 4763 4764This ioctl can be used to register a guest memory region which may 4765contain encrypted data (e.g. guest RAM, SMRAM etc). 4766 4767It is used in the SEV-enabled guest. When encryption is enabled, a guest 4768memory region may contain encrypted data. The SEV memory encryption 4769engine uses a tweak such that two identical plaintext pages, each at 4770different locations will have differing ciphertexts. So swapping or 4771moving ciphertext of those pages will not result in plaintext being 4772swapped. So relocating (or migrating) physical backing pages for the SEV 4773guest will require some additional steps. 4774 4775Note: The current SEV key management spec does not provide commands to 4776swap or migrate (move) ciphertext pages. Hence, for now we pin the guest 4777memory region registered with the ioctl. 4778 47794.112 KVM_MEMORY_ENCRYPT_UNREG_REGION 4780------------------------------------- 4781 4782:Capability: basic 4783:Architectures: x86 4784:Type: system 4785:Parameters: struct kvm_enc_region (in) 4786:Returns: 0 on success; -1 on error 4787 4788This ioctl can be used to unregister the guest memory region registered 4789with KVM_MEMORY_ENCRYPT_REG_REGION ioctl above. 4790 47914.113 KVM_HYPERV_EVENTFD 4792------------------------ 4793 4794:Capability: KVM_CAP_HYPERV_EVENTFD 4795:Architectures: x86 4796:Type: vm ioctl 4797:Parameters: struct kvm_hyperv_eventfd (in) 4798 4799This ioctl (un)registers an eventfd to receive notifications from the guest on 4800the specified Hyper-V connection id through the SIGNAL_EVENT hypercall, without 4801causing a user exit. SIGNAL_EVENT hypercall with non-zero event flag number 4802(bits 24-31) still triggers a KVM_EXIT_HYPERV_HCALL user exit. 4803 4804:: 4805 4806 struct kvm_hyperv_eventfd { 4807 __u32 conn_id; 4808 __s32 fd; 4809 __u32 flags; 4810 __u32 padding[3]; 4811 }; 4812 4813The conn_id field should fit within 24 bits:: 4814 4815 #define KVM_HYPERV_CONN_ID_MASK 0x00ffffff 4816 4817The acceptable values for the flags field are:: 4818 4819 #define KVM_HYPERV_EVENTFD_DEASSIGN (1 << 0) 4820 4821:Returns: 0 on success, 4822 -EINVAL if conn_id or flags is outside the allowed range, 4823 -ENOENT on deassign if the conn_id isn't registered, 4824 -EEXIST on assign if the conn_id is already registered 4825 48264.114 KVM_GET_NESTED_STATE 4827-------------------------- 4828 4829:Capability: KVM_CAP_NESTED_STATE 4830:Architectures: x86 4831:Type: vcpu ioctl 4832:Parameters: struct kvm_nested_state (in/out) 4833:Returns: 0 on success, -1 on error 4834 4835Errors: 4836 4837 ===== ============================================================= 4838 E2BIG the total state size exceeds the value of 'size' specified by 4839 the user; the size required will be written into size. 4840 ===== ============================================================= 4841 4842:: 4843 4844 struct kvm_nested_state { 4845 __u16 flags; 4846 __u16 format; 4847 __u32 size; 4848 4849 union { 4850 struct kvm_vmx_nested_state_hdr vmx; 4851 struct kvm_svm_nested_state_hdr svm; 4852 4853 /* Pad the header to 128 bytes. */ 4854 __u8 pad[120]; 4855 } hdr; 4856 4857 union { 4858 struct kvm_vmx_nested_state_data vmx[0]; 4859 struct kvm_svm_nested_state_data svm[0]; 4860 } data; 4861 }; 4862 4863 #define KVM_STATE_NESTED_GUEST_MODE 0x00000001 4864 #define KVM_STATE_NESTED_RUN_PENDING 0x00000002 4865 #define KVM_STATE_NESTED_EVMCS 0x00000004 4866 4867 #define KVM_STATE_NESTED_FORMAT_VMX 0 4868 #define KVM_STATE_NESTED_FORMAT_SVM 1 4869 4870 #define KVM_STATE_NESTED_VMX_VMCS_SIZE 0x1000 4871 4872 #define KVM_STATE_NESTED_VMX_SMM_GUEST_MODE 0x00000001 4873 #define KVM_STATE_NESTED_VMX_SMM_VMXON 0x00000002 4874 4875 #define KVM_STATE_VMX_PREEMPTION_TIMER_DEADLINE 0x00000001 4876 4877 struct kvm_vmx_nested_state_hdr { 4878 __u64 vmxon_pa; 4879 __u64 vmcs12_pa; 4880 4881 struct { 4882 __u16 flags; 4883 } smm; 4884 4885 __u32 flags; 4886 __u64 preemption_timer_deadline; 4887 }; 4888 4889 struct kvm_vmx_nested_state_data { 4890 __u8 vmcs12[KVM_STATE_NESTED_VMX_VMCS_SIZE]; 4891 __u8 shadow_vmcs12[KVM_STATE_NESTED_VMX_VMCS_SIZE]; 4892 }; 4893 4894This ioctl copies the vcpu's nested virtualization state from the kernel to 4895userspace. 4896 4897The maximum size of the state can be retrieved by passing KVM_CAP_NESTED_STATE 4898to the KVM_CHECK_EXTENSION ioctl(). 4899 49004.115 KVM_SET_NESTED_STATE 4901-------------------------- 4902 4903:Capability: KVM_CAP_NESTED_STATE 4904:Architectures: x86 4905:Type: vcpu ioctl 4906:Parameters: struct kvm_nested_state (in) 4907:Returns: 0 on success, -1 on error 4908 4909This copies the vcpu's kvm_nested_state struct from userspace to the kernel. 4910For the definition of struct kvm_nested_state, see KVM_GET_NESTED_STATE. 4911 49124.116 KVM_(UN)REGISTER_COALESCED_MMIO 4913------------------------------------- 4914 4915:Capability: KVM_CAP_COALESCED_MMIO (for coalesced mmio) 4916 KVM_CAP_COALESCED_PIO (for coalesced pio) 4917:Architectures: all 4918:Type: vm ioctl 4919:Parameters: struct kvm_coalesced_mmio_zone 4920:Returns: 0 on success, < 0 on error 4921 4922Coalesced I/O is a performance optimization that defers hardware 4923register write emulation so that userspace exits are avoided. It is 4924typically used to reduce the overhead of emulating frequently accessed 4925hardware registers. 4926 4927When a hardware register is configured for coalesced I/O, write accesses 4928do not exit to userspace and their value is recorded in a ring buffer 4929that is shared between kernel and userspace. 4930 4931Coalesced I/O is used if one or more write accesses to a hardware 4932register can be deferred until a read or a write to another hardware 4933register on the same device. This last access will cause a vmexit and 4934userspace will process accesses from the ring buffer before emulating 4935it. That will avoid exiting to userspace on repeated writes. 4936 4937Coalesced pio is based on coalesced mmio. There is little difference 4938between coalesced mmio and pio except that coalesced pio records accesses 4939to I/O ports. 4940 49414.117 KVM_CLEAR_DIRTY_LOG (vm ioctl) 4942------------------------------------ 4943 4944:Capability: KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 4945:Architectures: x86, arm64, mips 4946:Type: vm ioctl 4947:Parameters: struct kvm_clear_dirty_log (in) 4948:Returns: 0 on success, -1 on error 4949 4950:: 4951 4952 /* for KVM_CLEAR_DIRTY_LOG */ 4953 struct kvm_clear_dirty_log { 4954 __u32 slot; 4955 __u32 num_pages; 4956 __u64 first_page; 4957 union { 4958 void __user *dirty_bitmap; /* one bit per page */ 4959 __u64 padding; 4960 }; 4961 }; 4962 4963The ioctl clears the dirty status of pages in a memory slot, according to 4964the bitmap that is passed in struct kvm_clear_dirty_log's dirty_bitmap 4965field. Bit 0 of the bitmap corresponds to page "first_page" in the 4966memory slot, and num_pages is the size in bits of the input bitmap. 4967first_page must be a multiple of 64; num_pages must also be a multiple of 496864 unless first_page + num_pages is the size of the memory slot. For each 4969bit that is set in the input bitmap, the corresponding page is marked "clean" 4970in KVM's dirty bitmap, and dirty tracking is re-enabled for that page 4971(for example via write-protection, or by clearing the dirty bit in 4972a page table entry). 4973 4974If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of slot field specifies 4975the address space for which you want to clear the dirty status. See 4976KVM_SET_USER_MEMORY_REGION for details on the usage of slot field. 4977 4978This ioctl is mostly useful when KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 4979is enabled; for more information, see the description of the capability. 4980However, it can always be used as long as KVM_CHECK_EXTENSION confirms 4981that KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 is present. 4982 49834.118 KVM_GET_SUPPORTED_HV_CPUID 4984-------------------------------- 4985 4986:Capability: KVM_CAP_HYPERV_CPUID (vcpu), KVM_CAP_SYS_HYPERV_CPUID (system) 4987:Architectures: x86 4988:Type: system ioctl, vcpu ioctl 4989:Parameters: struct kvm_cpuid2 (in/out) 4990:Returns: 0 on success, -1 on error 4991 4992:: 4993 4994 struct kvm_cpuid2 { 4995 __u32 nent; 4996 __u32 padding; 4997 struct kvm_cpuid_entry2 entries[0]; 4998 }; 4999 5000 struct kvm_cpuid_entry2 { 5001 __u32 function; 5002 __u32 index; 5003 __u32 flags; 5004 __u32 eax; 5005 __u32 ebx; 5006 __u32 ecx; 5007 __u32 edx; 5008 __u32 padding[3]; 5009 }; 5010 5011This ioctl returns x86 cpuid features leaves related to Hyper-V emulation in 5012KVM. Userspace can use the information returned by this ioctl to construct 5013cpuid information presented to guests consuming Hyper-V enlightenments (e.g. 5014Windows or Hyper-V guests). 5015 5016CPUID feature leaves returned by this ioctl are defined by Hyper-V Top Level 5017Functional Specification (TLFS). These leaves can't be obtained with 5018KVM_GET_SUPPORTED_CPUID ioctl because some of them intersect with KVM feature 5019leaves (0x40000000, 0x40000001). 5020 5021Currently, the following list of CPUID leaves are returned: 5022 5023 - HYPERV_CPUID_VENDOR_AND_MAX_FUNCTIONS 5024 - HYPERV_CPUID_INTERFACE 5025 - HYPERV_CPUID_VERSION 5026 - HYPERV_CPUID_FEATURES 5027 - HYPERV_CPUID_ENLIGHTMENT_INFO 5028 - HYPERV_CPUID_IMPLEMENT_LIMITS 5029 - HYPERV_CPUID_NESTED_FEATURES 5030 - HYPERV_CPUID_SYNDBG_VENDOR_AND_MAX_FUNCTIONS 5031 - HYPERV_CPUID_SYNDBG_INTERFACE 5032 - HYPERV_CPUID_SYNDBG_PLATFORM_CAPABILITIES 5033 5034Userspace invokes KVM_GET_SUPPORTED_HV_CPUID by passing a kvm_cpuid2 structure 5035with the 'nent' field indicating the number of entries in the variable-size 5036array 'entries'. If the number of entries is too low to describe all Hyper-V 5037feature leaves, an error (E2BIG) is returned. If the number is more or equal 5038to the number of Hyper-V feature leaves, the 'nent' field is adjusted to the 5039number of valid entries in the 'entries' array, which is then filled. 5040 5041'index' and 'flags' fields in 'struct kvm_cpuid_entry2' are currently reserved, 5042userspace should not expect to get any particular value there. 5043 5044Note, vcpu version of KVM_GET_SUPPORTED_HV_CPUID is currently deprecated. Unlike 5045system ioctl which exposes all supported feature bits unconditionally, vcpu 5046version has the following quirks: 5047 5048- HYPERV_CPUID_NESTED_FEATURES leaf and HV_X64_ENLIGHTENED_VMCS_RECOMMENDED 5049 feature bit are only exposed when Enlightened VMCS was previously enabled 5050 on the corresponding vCPU (KVM_CAP_HYPERV_ENLIGHTENED_VMCS). 5051- HV_STIMER_DIRECT_MODE_AVAILABLE bit is only exposed with in-kernel LAPIC. 5052 (presumes KVM_CREATE_IRQCHIP has already been called). 5053 50544.119 KVM_ARM_VCPU_FINALIZE 5055--------------------------- 5056 5057:Architectures: arm64 5058:Type: vcpu ioctl 5059:Parameters: int feature (in) 5060:Returns: 0 on success, -1 on error 5061 5062Errors: 5063 5064 ====== ============================================================== 5065 EPERM feature not enabled, needs configuration, or already finalized 5066 EINVAL feature unknown or not present 5067 ====== ============================================================== 5068 5069Recognised values for feature: 5070 5071 ===== =========================================== 5072 arm64 KVM_ARM_VCPU_SVE (requires KVM_CAP_ARM_SVE) 5073 ===== =========================================== 5074 5075Finalizes the configuration of the specified vcpu feature. 5076 5077The vcpu must already have been initialised, enabling the affected feature, by 5078means of a successful KVM_ARM_VCPU_INIT call with the appropriate flag set in 5079features[]. 5080 5081For affected vcpu features, this is a mandatory step that must be performed 5082before the vcpu is fully usable. 5083 5084Between KVM_ARM_VCPU_INIT and KVM_ARM_VCPU_FINALIZE, the feature may be 5085configured by use of ioctls such as KVM_SET_ONE_REG. The exact configuration 5086that should be performed and how to do it are feature-dependent. 5087 5088Other calls that depend on a particular feature being finalized, such as 5089KVM_RUN, KVM_GET_REG_LIST, KVM_GET_ONE_REG and KVM_SET_ONE_REG, will fail with 5090-EPERM unless the feature has already been finalized by means of a 5091KVM_ARM_VCPU_FINALIZE call. 5092 5093See KVM_ARM_VCPU_INIT for details of vcpu features that require finalization 5094using this ioctl. 5095 50964.120 KVM_SET_PMU_EVENT_FILTER 5097------------------------------ 5098 5099:Capability: KVM_CAP_PMU_EVENT_FILTER 5100:Architectures: x86 5101:Type: vm ioctl 5102:Parameters: struct kvm_pmu_event_filter (in) 5103:Returns: 0 on success, -1 on error 5104 5105Errors: 5106 5107 ====== ============================================================ 5108 EFAULT args[0] cannot be accessed 5109 EINVAL args[0] contains invalid data in the filter or filter events 5110 E2BIG nevents is too large 5111 EBUSY not enough memory to allocate the filter 5112 ====== ============================================================ 5113 5114:: 5115 5116 struct kvm_pmu_event_filter { 5117 __u32 action; 5118 __u32 nevents; 5119 __u32 fixed_counter_bitmap; 5120 __u32 flags; 5121 __u32 pad[4]; 5122 __u64 events[0]; 5123 }; 5124 5125This ioctl restricts the set of PMU events the guest can program by limiting 5126which event select and unit mask combinations are permitted. 5127 5128The argument holds a list of filter events which will be allowed or denied. 5129 5130Filter events only control general purpose counters; fixed purpose counters 5131are controlled by the fixed_counter_bitmap. 5132 5133Valid values for 'flags':: 5134 5135``0`` 5136 5137To use this mode, clear the 'flags' field. 5138 5139In this mode each event will contain an event select + unit mask. 5140 5141When the guest attempts to program the PMU the guest's event select + 5142unit mask is compared against the filter events to determine whether the 5143guest should have access. 5144 5145``KVM_PMU_EVENT_FLAG_MASKED_EVENTS`` 5146:Capability: KVM_CAP_PMU_EVENT_MASKED_EVENTS 5147 5148In this mode each filter event will contain an event select, mask, match, and 5149exclude value. To encode a masked event use:: 5150 5151 KVM_PMU_ENCODE_MASKED_ENTRY() 5152 5153An encoded event will follow this layout:: 5154 5155 Bits Description 5156 ---- ----------- 5157 7:0 event select (low bits) 5158 15:8 umask match 5159 31:16 unused 5160 35:32 event select (high bits) 5161 36:54 unused 5162 55 exclude bit 5163 63:56 umask mask 5164 5165When the guest attempts to program the PMU, these steps are followed in 5166determining if the guest should have access: 5167 5168 1. Match the event select from the guest against the filter events. 5169 2. If a match is found, match the guest's unit mask to the mask and match 5170 values of the included filter events. 5171 I.e. (unit mask & mask) == match && !exclude. 5172 3. If a match is found, match the guest's unit mask to the mask and match 5173 values of the excluded filter events. 5174 I.e. (unit mask & mask) == match && exclude. 5175 4. 5176 a. If an included match is found and an excluded match is not found, filter 5177 the event. 5178 b. For everything else, do not filter the event. 5179 5. 5180 a. If the event is filtered and it's an allow list, allow the guest to 5181 program the event. 5182 b. If the event is filtered and it's a deny list, do not allow the guest to 5183 program the event. 5184 5185When setting a new pmu event filter, -EINVAL will be returned if any of the 5186unused fields are set or if any of the high bits (35:32) in the event 5187select are set when called on Intel. 5188 5189Valid values for 'action':: 5190 5191 #define KVM_PMU_EVENT_ALLOW 0 5192 #define KVM_PMU_EVENT_DENY 1 5193 5194Via this API, KVM userspace can also control the behavior of the VM's fixed 5195counters (if any) by configuring the "action" and "fixed_counter_bitmap" fields. 5196 5197Specifically, KVM follows the following pseudo-code when determining whether to 5198allow the guest FixCtr[i] to count its pre-defined fixed event:: 5199 5200 FixCtr[i]_is_allowed = (action == ALLOW) && (bitmap & BIT(i)) || 5201 (action == DENY) && !(bitmap & BIT(i)); 5202 FixCtr[i]_is_denied = !FixCtr[i]_is_allowed; 5203 5204KVM always consumes fixed_counter_bitmap, it's userspace's responsibility to 5205ensure fixed_counter_bitmap is set correctly, e.g. if userspace wants to define 5206a filter that only affects general purpose counters. 5207 5208Note, the "events" field also applies to fixed counters' hardcoded event_select 5209and unit_mask values. "fixed_counter_bitmap" has higher priority than "events" 5210if there is a contradiction between the two. 5211 52124.121 KVM_PPC_SVM_OFF 5213--------------------- 5214 5215:Capability: basic 5216:Architectures: powerpc 5217:Type: vm ioctl 5218:Parameters: none 5219:Returns: 0 on successful completion, 5220 5221Errors: 5222 5223 ====== ================================================================ 5224 EINVAL if ultravisor failed to terminate the secure guest 5225 ENOMEM if hypervisor failed to allocate new radix page tables for guest 5226 ====== ================================================================ 5227 5228This ioctl is used to turn off the secure mode of the guest or transition 5229the guest from secure mode to normal mode. This is invoked when the guest 5230is reset. This has no effect if called for a normal guest. 5231 5232This ioctl issues an ultravisor call to terminate the secure guest, 5233unpins the VPA pages and releases all the device pages that are used to 5234track the secure pages by hypervisor. 5235 52364.122 KVM_S390_NORMAL_RESET 5237--------------------------- 5238 5239:Capability: KVM_CAP_S390_VCPU_RESETS 5240:Architectures: s390 5241:Type: vcpu ioctl 5242:Parameters: none 5243:Returns: 0 5244 5245This ioctl resets VCPU registers and control structures according to 5246the cpu reset definition in the POP (Principles Of Operation). 5247 52484.123 KVM_S390_INITIAL_RESET 5249---------------------------- 5250 5251:Capability: none 5252:Architectures: s390 5253:Type: vcpu ioctl 5254:Parameters: none 5255:Returns: 0 5256 5257This ioctl resets VCPU registers and control structures according to 5258the initial cpu reset definition in the POP. However, the cpu is not 5259put into ESA mode. This reset is a superset of the normal reset. 5260 52614.124 KVM_S390_CLEAR_RESET 5262-------------------------- 5263 5264:Capability: KVM_CAP_S390_VCPU_RESETS 5265:Architectures: s390 5266:Type: vcpu ioctl 5267:Parameters: none 5268:Returns: 0 5269 5270This ioctl resets VCPU registers and control structures according to 5271the clear cpu reset definition in the POP. However, the cpu is not put 5272into ESA mode. This reset is a superset of the initial reset. 5273 5274 52754.125 KVM_S390_PV_COMMAND 5276------------------------- 5277 5278:Capability: KVM_CAP_S390_PROTECTED 5279:Architectures: s390 5280:Type: vm ioctl 5281:Parameters: struct kvm_pv_cmd 5282:Returns: 0 on success, < 0 on error 5283 5284:: 5285 5286 struct kvm_pv_cmd { 5287 __u32 cmd; /* Command to be executed */ 5288 __u16 rc; /* Ultravisor return code */ 5289 __u16 rrc; /* Ultravisor return reason code */ 5290 __u64 data; /* Data or address */ 5291 __u32 flags; /* flags for future extensions. Must be 0 for now */ 5292 __u32 reserved[3]; 5293 }; 5294 5295**Ultravisor return codes** 5296The Ultravisor return (reason) codes are provided by the kernel if a 5297Ultravisor call has been executed to achieve the results expected by 5298the command. Therefore they are independent of the IOCTL return 5299code. If KVM changes `rc`, its value will always be greater than 0 5300hence setting it to 0 before issuing a PV command is advised to be 5301able to detect a change of `rc`. 5302 5303**cmd values:** 5304 5305KVM_PV_ENABLE 5306 Allocate memory and register the VM with the Ultravisor, thereby 5307 donating memory to the Ultravisor that will become inaccessible to 5308 KVM. All existing CPUs are converted to protected ones. After this 5309 command has succeeded, any CPU added via hotplug will become 5310 protected during its creation as well. 5311 5312 Errors: 5313 5314 ===== ============================= 5315 EINTR an unmasked signal is pending 5316 ===== ============================= 5317 5318KVM_PV_DISABLE 5319 Deregister the VM from the Ultravisor and reclaim the memory that had 5320 been donated to the Ultravisor, making it usable by the kernel again. 5321 All registered VCPUs are converted back to non-protected ones. If a 5322 previous protected VM had been prepared for asynchronous teardown with 5323 KVM_PV_ASYNC_CLEANUP_PREPARE and not subsequently torn down with 5324 KVM_PV_ASYNC_CLEANUP_PERFORM, it will be torn down in this call 5325 together with the current protected VM. 5326 5327KVM_PV_VM_SET_SEC_PARMS 5328 Pass the image header from VM memory to the Ultravisor in 5329 preparation of image unpacking and verification. 5330 5331KVM_PV_VM_UNPACK 5332 Unpack (protect and decrypt) a page of the encrypted boot image. 5333 5334KVM_PV_VM_VERIFY 5335 Verify the integrity of the unpacked image. Only if this succeeds, 5336 KVM is allowed to start protected VCPUs. 5337 5338KVM_PV_INFO 5339 :Capability: KVM_CAP_S390_PROTECTED_DUMP 5340 5341 Presents an API that provides Ultravisor related data to userspace 5342 via subcommands. len_max is the size of the user space buffer, 5343 len_written is KVM's indication of how much bytes of that buffer 5344 were actually written to. len_written can be used to determine the 5345 valid fields if more response fields are added in the future. 5346 5347 :: 5348 5349 enum pv_cmd_info_id { 5350 KVM_PV_INFO_VM, 5351 KVM_PV_INFO_DUMP, 5352 }; 5353 5354 struct kvm_s390_pv_info_header { 5355 __u32 id; 5356 __u32 len_max; 5357 __u32 len_written; 5358 __u32 reserved; 5359 }; 5360 5361 struct kvm_s390_pv_info { 5362 struct kvm_s390_pv_info_header header; 5363 struct kvm_s390_pv_info_dump dump; 5364 struct kvm_s390_pv_info_vm vm; 5365 }; 5366 5367**subcommands:** 5368 5369 KVM_PV_INFO_VM 5370 This subcommand provides basic Ultravisor information for PV 5371 hosts. These values are likely also exported as files in the sysfs 5372 firmware UV query interface but they are more easily available to 5373 programs in this API. 5374 5375 The installed calls and feature_indication members provide the 5376 installed UV calls and the UV's other feature indications. 5377 5378 The max_* members provide information about the maximum number of PV 5379 vcpus, PV guests and PV guest memory size. 5380 5381 :: 5382 5383 struct kvm_s390_pv_info_vm { 5384 __u64 inst_calls_list[4]; 5385 __u64 max_cpus; 5386 __u64 max_guests; 5387 __u64 max_guest_addr; 5388 __u64 feature_indication; 5389 }; 5390 5391 5392 KVM_PV_INFO_DUMP 5393 This subcommand provides information related to dumping PV guests. 5394 5395 :: 5396 5397 struct kvm_s390_pv_info_dump { 5398 __u64 dump_cpu_buffer_len; 5399 __u64 dump_config_mem_buffer_per_1m; 5400 __u64 dump_config_finalize_len; 5401 }; 5402 5403KVM_PV_DUMP 5404 :Capability: KVM_CAP_S390_PROTECTED_DUMP 5405 5406 Presents an API that provides calls which facilitate dumping a 5407 protected VM. 5408 5409 :: 5410 5411 struct kvm_s390_pv_dmp { 5412 __u64 subcmd; 5413 __u64 buff_addr; 5414 __u64 buff_len; 5415 __u64 gaddr; /* For dump storage state */ 5416 }; 5417 5418 **subcommands:** 5419 5420 KVM_PV_DUMP_INIT 5421 Initializes the dump process of a protected VM. If this call does 5422 not succeed all other subcommands will fail with -EINVAL. This 5423 subcommand will return -EINVAL if a dump process has not yet been 5424 completed. 5425 5426 Not all PV vms can be dumped, the owner needs to set `dump 5427 allowed` PCF bit 34 in the SE header to allow dumping. 5428 5429 KVM_PV_DUMP_CONFIG_STOR_STATE 5430 Stores `buff_len` bytes of tweak component values starting with 5431 the 1MB block specified by the absolute guest address 5432 (`gaddr`). `buff_len` needs to be `conf_dump_storage_state_len` 5433 aligned and at least >= the `conf_dump_storage_state_len` value 5434 provided by the dump uv_info data. buff_user might be written to 5435 even if an error rc is returned. For instance if we encounter a 5436 fault after writing the first page of data. 5437 5438 KVM_PV_DUMP_COMPLETE 5439 If the subcommand succeeds it completes the dump process and lets 5440 KVM_PV_DUMP_INIT be called again. 5441 5442 On success `conf_dump_finalize_len` bytes of completion data will be 5443 stored to the `buff_addr`. The completion data contains a key 5444 derivation seed, IV, tweak nonce and encryption keys as well as an 5445 authentication tag all of which are needed to decrypt the dump at a 5446 later time. 5447 5448KVM_PV_ASYNC_CLEANUP_PREPARE 5449 :Capability: KVM_CAP_S390_PROTECTED_ASYNC_DISABLE 5450 5451 Prepare the current protected VM for asynchronous teardown. Most 5452 resources used by the current protected VM will be set aside for a 5453 subsequent asynchronous teardown. The current protected VM will then 5454 resume execution immediately as non-protected. There can be at most 5455 one protected VM prepared for asynchronous teardown at any time. If 5456 a protected VM had already been prepared for teardown without 5457 subsequently calling KVM_PV_ASYNC_CLEANUP_PERFORM, this call will 5458 fail. In that case, the userspace process should issue a normal 5459 KVM_PV_DISABLE. The resources set aside with this call will need to 5460 be cleaned up with a subsequent call to KVM_PV_ASYNC_CLEANUP_PERFORM 5461 or KVM_PV_DISABLE, otherwise they will be cleaned up when KVM 5462 terminates. KVM_PV_ASYNC_CLEANUP_PREPARE can be called again as soon 5463 as cleanup starts, i.e. before KVM_PV_ASYNC_CLEANUP_PERFORM finishes. 5464 5465KVM_PV_ASYNC_CLEANUP_PERFORM 5466 :Capability: KVM_CAP_S390_PROTECTED_ASYNC_DISABLE 5467 5468 Tear down the protected VM previously prepared for teardown with 5469 KVM_PV_ASYNC_CLEANUP_PREPARE. The resources that had been set aside 5470 will be freed during the execution of this command. This PV command 5471 should ideally be issued by userspace from a separate thread. If a 5472 fatal signal is received (or the process terminates naturally), the 5473 command will terminate immediately without completing, and the normal 5474 KVM shutdown procedure will take care of cleaning up all remaining 5475 protected VMs, including the ones whose teardown was interrupted by 5476 process termination. 5477 54784.126 KVM_XEN_HVM_SET_ATTR 5479-------------------------- 5480 5481:Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO 5482:Architectures: x86 5483:Type: vm ioctl 5484:Parameters: struct kvm_xen_hvm_attr 5485:Returns: 0 on success, < 0 on error 5486 5487:: 5488 5489 struct kvm_xen_hvm_attr { 5490 __u16 type; 5491 __u16 pad[3]; 5492 union { 5493 __u8 long_mode; 5494 __u8 vector; 5495 __u8 runstate_update_flag; 5496 union { 5497 __u64 gfn; 5498 __u64 hva; 5499 } shared_info; 5500 struct { 5501 __u32 send_port; 5502 __u32 type; /* EVTCHNSTAT_ipi / EVTCHNSTAT_interdomain */ 5503 __u32 flags; 5504 union { 5505 struct { 5506 __u32 port; 5507 __u32 vcpu; 5508 __u32 priority; 5509 } port; 5510 struct { 5511 __u32 port; /* Zero for eventfd */ 5512 __s32 fd; 5513 } eventfd; 5514 __u32 padding[4]; 5515 } deliver; 5516 } evtchn; 5517 __u32 xen_version; 5518 __u64 pad[8]; 5519 } u; 5520 }; 5521 5522type values: 5523 5524KVM_XEN_ATTR_TYPE_LONG_MODE 5525 Sets the ABI mode of the VM to 32-bit or 64-bit (long mode). This 5526 determines the layout of the shared_info page exposed to the VM. 5527 5528KVM_XEN_ATTR_TYPE_SHARED_INFO 5529 Sets the guest physical frame number at which the Xen shared_info 5530 page resides. Note that although Xen places vcpu_info for the first 5531 32 vCPUs in the shared_info page, KVM does not automatically do so 5532 and instead requires that KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO or 5533 KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO_HVA be used explicitly even when 5534 the vcpu_info for a given vCPU resides at the "default" location 5535 in the shared_info page. This is because KVM may not be aware of 5536 the Xen CPU id which is used as the index into the vcpu_info[] 5537 array, so may know the correct default location. 5538 5539 Note that the shared_info page may be constantly written to by KVM; 5540 it contains the event channel bitmap used to deliver interrupts to 5541 a Xen guest, amongst other things. It is exempt from dirty tracking 5542 mechanisms — KVM will not explicitly mark the page as dirty each 5543 time an event channel interrupt is delivered to the guest! Thus, 5544 userspace should always assume that the designated GFN is dirty if 5545 any vCPU has been running or any event channel interrupts can be 5546 routed to the guest. 5547 5548 Setting the gfn to KVM_XEN_INVALID_GFN will disable the shared_info 5549 page. 5550 5551KVM_XEN_ATTR_TYPE_SHARED_INFO_HVA 5552 If the KVM_XEN_HVM_CONFIG_SHARED_INFO_HVA flag is also set in the 5553 Xen capabilities, then this attribute may be used to set the 5554 userspace address at which the shared_info page resides, which 5555 will always be fixed in the VMM regardless of where it is mapped 5556 in guest physical address space. This attribute should be used in 5557 preference to KVM_XEN_ATTR_TYPE_SHARED_INFO as it avoids 5558 unnecessary invalidation of an internal cache when the page is 5559 re-mapped in guest physcial address space. 5560 5561 Setting the hva to zero will disable the shared_info page. 5562 5563KVM_XEN_ATTR_TYPE_UPCALL_VECTOR 5564 Sets the exception vector used to deliver Xen event channel upcalls. 5565 This is the HVM-wide vector injected directly by the hypervisor 5566 (not through the local APIC), typically configured by a guest via 5567 HVM_PARAM_CALLBACK_IRQ. This can be disabled again (e.g. for guest 5568 SHUTDOWN_soft_reset) by setting it to zero. 5569 5570KVM_XEN_ATTR_TYPE_EVTCHN 5571 This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates 5572 support for KVM_XEN_HVM_CONFIG_EVTCHN_SEND features. It configures 5573 an outbound port number for interception of EVTCHNOP_send requests 5574 from the guest. A given sending port number may be directed back to 5575 a specified vCPU (by APIC ID) / port / priority on the guest, or to 5576 trigger events on an eventfd. The vCPU and priority can be changed 5577 by setting KVM_XEN_EVTCHN_UPDATE in a subsequent call, but other 5578 fields cannot change for a given sending port. A port mapping is 5579 removed by using KVM_XEN_EVTCHN_DEASSIGN in the flags field. Passing 5580 KVM_XEN_EVTCHN_RESET in the flags field removes all interception of 5581 outbound event channels. The values of the flags field are mutually 5582 exclusive and cannot be combined as a bitmask. 5583 5584KVM_XEN_ATTR_TYPE_XEN_VERSION 5585 This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates 5586 support for KVM_XEN_HVM_CONFIG_EVTCHN_SEND features. It configures 5587 the 32-bit version code returned to the guest when it invokes the 5588 XENVER_version call; typically (XEN_MAJOR << 16 | XEN_MINOR). PV 5589 Xen guests will often use this to as a dummy hypercall to trigger 5590 event channel delivery, so responding within the kernel without 5591 exiting to userspace is beneficial. 5592 5593KVM_XEN_ATTR_TYPE_RUNSTATE_UPDATE_FLAG 5594 This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates 5595 support for KVM_XEN_HVM_CONFIG_RUNSTATE_UPDATE_FLAG. It enables the 5596 XEN_RUNSTATE_UPDATE flag which allows guest vCPUs to safely read 5597 other vCPUs' vcpu_runstate_info. Xen guests enable this feature via 5598 the VMASST_TYPE_runstate_update_flag of the HYPERVISOR_vm_assist 5599 hypercall. 5600 56014.127 KVM_XEN_HVM_GET_ATTR 5602-------------------------- 5603 5604:Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO 5605:Architectures: x86 5606:Type: vm ioctl 5607:Parameters: struct kvm_xen_hvm_attr 5608:Returns: 0 on success, < 0 on error 5609 5610Allows Xen VM attributes to be read. For the structure and types, 5611see KVM_XEN_HVM_SET_ATTR above. The KVM_XEN_ATTR_TYPE_EVTCHN 5612attribute cannot be read. 5613 56144.128 KVM_XEN_VCPU_SET_ATTR 5615--------------------------- 5616 5617:Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO 5618:Architectures: x86 5619:Type: vcpu ioctl 5620:Parameters: struct kvm_xen_vcpu_attr 5621:Returns: 0 on success, < 0 on error 5622 5623:: 5624 5625 struct kvm_xen_vcpu_attr { 5626 __u16 type; 5627 __u16 pad[3]; 5628 union { 5629 __u64 gpa; 5630 __u64 pad[4]; 5631 struct { 5632 __u64 state; 5633 __u64 state_entry_time; 5634 __u64 time_running; 5635 __u64 time_runnable; 5636 __u64 time_blocked; 5637 __u64 time_offline; 5638 } runstate; 5639 __u32 vcpu_id; 5640 struct { 5641 __u32 port; 5642 __u32 priority; 5643 __u64 expires_ns; 5644 } timer; 5645 __u8 vector; 5646 } u; 5647 }; 5648 5649type values: 5650 5651KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO 5652 Sets the guest physical address of the vcpu_info for a given vCPU. 5653 As with the shared_info page for the VM, the corresponding page may be 5654 dirtied at any time if event channel interrupt delivery is enabled, so 5655 userspace should always assume that the page is dirty without relying 5656 on dirty logging. Setting the gpa to KVM_XEN_INVALID_GPA will disable 5657 the vcpu_info. 5658 5659KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO_HVA 5660 If the KVM_XEN_HVM_CONFIG_SHARED_INFO_HVA flag is also set in the 5661 Xen capabilities, then this attribute may be used to set the 5662 userspace address of the vcpu_info for a given vCPU. It should 5663 only be used when the vcpu_info resides at the "default" location 5664 in the shared_info page. In this case it is safe to assume the 5665 userspace address will not change, because the shared_info page is 5666 an overlay on guest memory and remains at a fixed host address 5667 regardless of where it is mapped in guest physical address space 5668 and hence unnecessary invalidation of an internal cache may be 5669 avoided if the guest memory layout is modified. 5670 If the vcpu_info does not reside at the "default" location then 5671 it is not guaranteed to remain at the same host address and 5672 hence the aforementioned cache invalidation is required. 5673 5674KVM_XEN_VCPU_ATTR_TYPE_VCPU_TIME_INFO 5675 Sets the guest physical address of an additional pvclock structure 5676 for a given vCPU. This is typically used for guest vsyscall support. 5677 Setting the gpa to KVM_XEN_INVALID_GPA will disable the structure. 5678 5679KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADDR 5680 Sets the guest physical address of the vcpu_runstate_info for a given 5681 vCPU. This is how a Xen guest tracks CPU state such as steal time. 5682 Setting the gpa to KVM_XEN_INVALID_GPA will disable the runstate area. 5683 5684KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_CURRENT 5685 Sets the runstate (RUNSTATE_running/_runnable/_blocked/_offline) of 5686 the given vCPU from the .u.runstate.state member of the structure. 5687 KVM automatically accounts running and runnable time but blocked 5688 and offline states are only entered explicitly. 5689 5690KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_DATA 5691 Sets all fields of the vCPU runstate data from the .u.runstate member 5692 of the structure, including the current runstate. The state_entry_time 5693 must equal the sum of the other four times. 5694 5695KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADJUST 5696 This *adds* the contents of the .u.runstate members of the structure 5697 to the corresponding members of the given vCPU's runstate data, thus 5698 permitting atomic adjustments to the runstate times. The adjustment 5699 to the state_entry_time must equal the sum of the adjustments to the 5700 other four times. The state field must be set to -1, or to a valid 5701 runstate value (RUNSTATE_running, RUNSTATE_runnable, RUNSTATE_blocked 5702 or RUNSTATE_offline) to set the current accounted state as of the 5703 adjusted state_entry_time. 5704 5705KVM_XEN_VCPU_ATTR_TYPE_VCPU_ID 5706 This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates 5707 support for KVM_XEN_HVM_CONFIG_EVTCHN_SEND features. It sets the Xen 5708 vCPU ID of the given vCPU, to allow timer-related VCPU operations to 5709 be intercepted by KVM. 5710 5711KVM_XEN_VCPU_ATTR_TYPE_TIMER 5712 This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates 5713 support for KVM_XEN_HVM_CONFIG_EVTCHN_SEND features. It sets the 5714 event channel port/priority for the VIRQ_TIMER of the vCPU, as well 5715 as allowing a pending timer to be saved/restored. Setting the timer 5716 port to zero disables kernel handling of the singleshot timer. 5717 5718KVM_XEN_VCPU_ATTR_TYPE_UPCALL_VECTOR 5719 This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates 5720 support for KVM_XEN_HVM_CONFIG_EVTCHN_SEND features. It sets the 5721 per-vCPU local APIC upcall vector, configured by a Xen guest with 5722 the HVMOP_set_evtchn_upcall_vector hypercall. This is typically 5723 used by Windows guests, and is distinct from the HVM-wide upcall 5724 vector configured with HVM_PARAM_CALLBACK_IRQ. It is disabled by 5725 setting the vector to zero. 5726 5727 57284.129 KVM_XEN_VCPU_GET_ATTR 5729--------------------------- 5730 5731:Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO 5732:Architectures: x86 5733:Type: vcpu ioctl 5734:Parameters: struct kvm_xen_vcpu_attr 5735:Returns: 0 on success, < 0 on error 5736 5737Allows Xen vCPU attributes to be read. For the structure and types, 5738see KVM_XEN_VCPU_SET_ATTR above. 5739 5740The KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADJUST type may not be used 5741with the KVM_XEN_VCPU_GET_ATTR ioctl. 5742 57434.130 KVM_ARM_MTE_COPY_TAGS 5744--------------------------- 5745 5746:Capability: KVM_CAP_ARM_MTE 5747:Architectures: arm64 5748:Type: vm ioctl 5749:Parameters: struct kvm_arm_copy_mte_tags 5750:Returns: number of bytes copied, < 0 on error (-EINVAL for incorrect 5751 arguments, -EFAULT if memory cannot be accessed). 5752 5753:: 5754 5755 struct kvm_arm_copy_mte_tags { 5756 __u64 guest_ipa; 5757 __u64 length; 5758 void __user *addr; 5759 __u64 flags; 5760 __u64 reserved[2]; 5761 }; 5762 5763Copies Memory Tagging Extension (MTE) tags to/from guest tag memory. The 5764``guest_ipa`` and ``length`` fields must be ``PAGE_SIZE`` aligned. 5765``length`` must not be bigger than 2^31 - PAGE_SIZE bytes. The ``addr`` 5766field must point to a buffer which the tags will be copied to or from. 5767 5768``flags`` specifies the direction of copy, either ``KVM_ARM_TAGS_TO_GUEST`` or 5769``KVM_ARM_TAGS_FROM_GUEST``. 5770 5771The size of the buffer to store the tags is ``(length / 16)`` bytes 5772(granules in MTE are 16 bytes long). Each byte contains a single tag 5773value. This matches the format of ``PTRACE_PEEKMTETAGS`` and 5774``PTRACE_POKEMTETAGS``. 5775 5776If an error occurs before any data is copied then a negative error code is 5777returned. If some tags have been copied before an error occurs then the number 5778of bytes successfully copied is returned. If the call completes successfully 5779then ``length`` is returned. 5780 57814.131 KVM_GET_SREGS2 5782-------------------- 5783 5784:Capability: KVM_CAP_SREGS2 5785:Architectures: x86 5786:Type: vcpu ioctl 5787:Parameters: struct kvm_sregs2 (out) 5788:Returns: 0 on success, -1 on error 5789 5790Reads special registers from the vcpu. 5791This ioctl (when supported) replaces the KVM_GET_SREGS. 5792 5793:: 5794 5795 struct kvm_sregs2 { 5796 /* out (KVM_GET_SREGS2) / in (KVM_SET_SREGS2) */ 5797 struct kvm_segment cs, ds, es, fs, gs, ss; 5798 struct kvm_segment tr, ldt; 5799 struct kvm_dtable gdt, idt; 5800 __u64 cr0, cr2, cr3, cr4, cr8; 5801 __u64 efer; 5802 __u64 apic_base; 5803 __u64 flags; 5804 __u64 pdptrs[4]; 5805 }; 5806 5807flags values for ``kvm_sregs2``: 5808 5809``KVM_SREGS2_FLAGS_PDPTRS_VALID`` 5810 5811 Indicates that the struct contains valid PDPTR values. 5812 5813 58144.132 KVM_SET_SREGS2 5815-------------------- 5816 5817:Capability: KVM_CAP_SREGS2 5818:Architectures: x86 5819:Type: vcpu ioctl 5820:Parameters: struct kvm_sregs2 (in) 5821:Returns: 0 on success, -1 on error 5822 5823Writes special registers into the vcpu. 5824See KVM_GET_SREGS2 for the data structures. 5825This ioctl (when supported) replaces the KVM_SET_SREGS. 5826 58274.133 KVM_GET_STATS_FD 5828---------------------- 5829 5830:Capability: KVM_CAP_STATS_BINARY_FD 5831:Architectures: all 5832:Type: vm ioctl, vcpu ioctl 5833:Parameters: none 5834:Returns: statistics file descriptor on success, < 0 on error 5835 5836Errors: 5837 5838 ====== ====================================================== 5839 ENOMEM if the fd could not be created due to lack of memory 5840 EMFILE if the number of opened files exceeds the limit 5841 ====== ====================================================== 5842 5843The returned file descriptor can be used to read VM/vCPU statistics data in 5844binary format. The data in the file descriptor consists of four blocks 5845organized as follows: 5846 5847+-------------+ 5848| Header | 5849+-------------+ 5850| id string | 5851+-------------+ 5852| Descriptors | 5853+-------------+ 5854| Stats Data | 5855+-------------+ 5856 5857Apart from the header starting at offset 0, please be aware that it is 5858not guaranteed that the four blocks are adjacent or in the above order; 5859the offsets of the id, descriptors and data blocks are found in the 5860header. However, all four blocks are aligned to 64 bit offsets in the 5861file and they do not overlap. 5862 5863All blocks except the data block are immutable. Userspace can read them 5864only one time after retrieving the file descriptor, and then use ``pread`` or 5865``lseek`` to read the statistics repeatedly. 5866 5867All data is in system endianness. 5868 5869The format of the header is as follows:: 5870 5871 struct kvm_stats_header { 5872 __u32 flags; 5873 __u32 name_size; 5874 __u32 num_desc; 5875 __u32 id_offset; 5876 __u32 desc_offset; 5877 __u32 data_offset; 5878 }; 5879 5880The ``flags`` field is not used at the moment. It is always read as 0. 5881 5882The ``name_size`` field is the size (in byte) of the statistics name string 5883(including trailing '\0') which is contained in the "id string" block and 5884appended at the end of every descriptor. 5885 5886The ``num_desc`` field is the number of descriptors that are included in the 5887descriptor block. (The actual number of values in the data block may be 5888larger, since each descriptor may comprise more than one value). 5889 5890The ``id_offset`` field is the offset of the id string from the start of the 5891file indicated by the file descriptor. It is a multiple of 8. 5892 5893The ``desc_offset`` field is the offset of the Descriptors block from the start 5894of the file indicated by the file descriptor. It is a multiple of 8. 5895 5896The ``data_offset`` field is the offset of the Stats Data block from the start 5897of the file indicated by the file descriptor. It is a multiple of 8. 5898 5899The id string block contains a string which identifies the file descriptor on 5900which KVM_GET_STATS_FD was invoked. The size of the block, including the 5901trailing ``'\0'``, is indicated by the ``name_size`` field in the header. 5902 5903The descriptors block is only needed to be read once for the lifetime of the 5904file descriptor contains a sequence of ``struct kvm_stats_desc``, each followed 5905by a string of size ``name_size``. 5906:: 5907 5908 #define KVM_STATS_TYPE_SHIFT 0 5909 #define KVM_STATS_TYPE_MASK (0xF << KVM_STATS_TYPE_SHIFT) 5910 #define KVM_STATS_TYPE_CUMULATIVE (0x0 << KVM_STATS_TYPE_SHIFT) 5911 #define KVM_STATS_TYPE_INSTANT (0x1 << KVM_STATS_TYPE_SHIFT) 5912 #define KVM_STATS_TYPE_PEAK (0x2 << KVM_STATS_TYPE_SHIFT) 5913 #define KVM_STATS_TYPE_LINEAR_HIST (0x3 << KVM_STATS_TYPE_SHIFT) 5914 #define KVM_STATS_TYPE_LOG_HIST (0x4 << KVM_STATS_TYPE_SHIFT) 5915 #define KVM_STATS_TYPE_MAX KVM_STATS_TYPE_LOG_HIST 5916 5917 #define KVM_STATS_UNIT_SHIFT 4 5918 #define KVM_STATS_UNIT_MASK (0xF << KVM_STATS_UNIT_SHIFT) 5919 #define KVM_STATS_UNIT_NONE (0x0 << KVM_STATS_UNIT_SHIFT) 5920 #define KVM_STATS_UNIT_BYTES (0x1 << KVM_STATS_UNIT_SHIFT) 5921 #define KVM_STATS_UNIT_SECONDS (0x2 << KVM_STATS_UNIT_SHIFT) 5922 #define KVM_STATS_UNIT_CYCLES (0x3 << KVM_STATS_UNIT_SHIFT) 5923 #define KVM_STATS_UNIT_BOOLEAN (0x4 << KVM_STATS_UNIT_SHIFT) 5924 #define KVM_STATS_UNIT_MAX KVM_STATS_UNIT_BOOLEAN 5925 5926 #define KVM_STATS_BASE_SHIFT 8 5927 #define KVM_STATS_BASE_MASK (0xF << KVM_STATS_BASE_SHIFT) 5928 #define KVM_STATS_BASE_POW10 (0x0 << KVM_STATS_BASE_SHIFT) 5929 #define KVM_STATS_BASE_POW2 (0x1 << KVM_STATS_BASE_SHIFT) 5930 #define KVM_STATS_BASE_MAX KVM_STATS_BASE_POW2 5931 5932 struct kvm_stats_desc { 5933 __u32 flags; 5934 __s16 exponent; 5935 __u16 size; 5936 __u32 offset; 5937 __u32 bucket_size; 5938 char name[]; 5939 }; 5940 5941The ``flags`` field contains the type and unit of the statistics data described 5942by this descriptor. Its endianness is CPU native. 5943The following flags are supported: 5944 5945Bits 0-3 of ``flags`` encode the type: 5946 5947 * ``KVM_STATS_TYPE_CUMULATIVE`` 5948 The statistics reports a cumulative count. The value of data can only be increased. 5949 Most of the counters used in KVM are of this type. 5950 The corresponding ``size`` field for this type is always 1. 5951 All cumulative statistics data are read/write. 5952 * ``KVM_STATS_TYPE_INSTANT`` 5953 The statistics reports an instantaneous value. Its value can be increased or 5954 decreased. This type is usually used as a measurement of some resources, 5955 like the number of dirty pages, the number of large pages, etc. 5956 All instant statistics are read only. 5957 The corresponding ``size`` field for this type is always 1. 5958 * ``KVM_STATS_TYPE_PEAK`` 5959 The statistics data reports a peak value, for example the maximum number 5960 of items in a hash table bucket, the longest time waited and so on. 5961 The value of data can only be increased. 5962 The corresponding ``size`` field for this type is always 1. 5963 * ``KVM_STATS_TYPE_LINEAR_HIST`` 5964 The statistic is reported as a linear histogram. The number of 5965 buckets is specified by the ``size`` field. The size of buckets is specified 5966 by the ``hist_param`` field. The range of the Nth bucket (1 <= N < ``size``) 5967 is [``hist_param``*(N-1), ``hist_param``*N), while the range of the last 5968 bucket is [``hist_param``*(``size``-1), +INF). (+INF means positive infinity 5969 value.) 5970 * ``KVM_STATS_TYPE_LOG_HIST`` 5971 The statistic is reported as a logarithmic histogram. The number of 5972 buckets is specified by the ``size`` field. The range of the first bucket is 5973 [0, 1), while the range of the last bucket is [pow(2, ``size``-2), +INF). 5974 Otherwise, The Nth bucket (1 < N < ``size``) covers 5975 [pow(2, N-2), pow(2, N-1)). 5976 5977Bits 4-7 of ``flags`` encode the unit: 5978 5979 * ``KVM_STATS_UNIT_NONE`` 5980 There is no unit for the value of statistics data. This usually means that 5981 the value is a simple counter of an event. 5982 * ``KVM_STATS_UNIT_BYTES`` 5983 It indicates that the statistics data is used to measure memory size, in the 5984 unit of Byte, KiByte, MiByte, GiByte, etc. The unit of the data is 5985 determined by the ``exponent`` field in the descriptor. 5986 * ``KVM_STATS_UNIT_SECONDS`` 5987 It indicates that the statistics data is used to measure time or latency. 5988 * ``KVM_STATS_UNIT_CYCLES`` 5989 It indicates that the statistics data is used to measure CPU clock cycles. 5990 * ``KVM_STATS_UNIT_BOOLEAN`` 5991 It indicates that the statistic will always be either 0 or 1. Boolean 5992 statistics of "peak" type will never go back from 1 to 0. Boolean 5993 statistics can be linear histograms (with two buckets) but not logarithmic 5994 histograms. 5995 5996Note that, in the case of histograms, the unit applies to the bucket 5997ranges, while the bucket value indicates how many samples fell in the 5998bucket's range. 5999 6000Bits 8-11 of ``flags``, together with ``exponent``, encode the scale of the 6001unit: 6002 6003 * ``KVM_STATS_BASE_POW10`` 6004 The scale is based on power of 10. It is used for measurement of time and 6005 CPU clock cycles. For example, an exponent of -9 can be used with 6006 ``KVM_STATS_UNIT_SECONDS`` to express that the unit is nanoseconds. 6007 * ``KVM_STATS_BASE_POW2`` 6008 The scale is based on power of 2. It is used for measurement of memory size. 6009 For example, an exponent of 20 can be used with ``KVM_STATS_UNIT_BYTES`` to 6010 express that the unit is MiB. 6011 6012The ``size`` field is the number of values of this statistics data. Its 6013value is usually 1 for most of simple statistics. 1 means it contains an 6014unsigned 64bit data. 6015 6016The ``offset`` field is the offset from the start of Data Block to the start of 6017the corresponding statistics data. 6018 6019The ``bucket_size`` field is used as a parameter for histogram statistics data. 6020It is only used by linear histogram statistics data, specifying the size of a 6021bucket in the unit expressed by bits 4-11 of ``flags`` together with ``exponent``. 6022 6023The ``name`` field is the name string of the statistics data. The name string 6024starts at the end of ``struct kvm_stats_desc``. The maximum length including 6025the trailing ``'\0'``, is indicated by ``name_size`` in the header. 6026 6027The Stats Data block contains an array of 64-bit values in the same order 6028as the descriptors in Descriptors block. 6029 60304.134 KVM_GET_XSAVE2 6031-------------------- 6032 6033:Capability: KVM_CAP_XSAVE2 6034:Architectures: x86 6035:Type: vcpu ioctl 6036:Parameters: struct kvm_xsave (out) 6037:Returns: 0 on success, -1 on error 6038 6039 6040:: 6041 6042 struct kvm_xsave { 6043 __u32 region[1024]; 6044 __u32 extra[0]; 6045 }; 6046 6047This ioctl would copy current vcpu's xsave struct to the userspace. It 6048copies as many bytes as are returned by KVM_CHECK_EXTENSION(KVM_CAP_XSAVE2) 6049when invoked on the vm file descriptor. The size value returned by 6050KVM_CHECK_EXTENSION(KVM_CAP_XSAVE2) will always be at least 4096. 6051Currently, it is only greater than 4096 if a dynamic feature has been 6052enabled with ``arch_prctl()``, but this may change in the future. 6053 6054The offsets of the state save areas in struct kvm_xsave follow the contents 6055of CPUID leaf 0xD on the host. 6056 60574.135 KVM_XEN_HVM_EVTCHN_SEND 6058----------------------------- 6059 6060:Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_EVTCHN_SEND 6061:Architectures: x86 6062:Type: vm ioctl 6063:Parameters: struct kvm_irq_routing_xen_evtchn 6064:Returns: 0 on success, < 0 on error 6065 6066 6067:: 6068 6069 struct kvm_irq_routing_xen_evtchn { 6070 __u32 port; 6071 __u32 vcpu; 6072 __u32 priority; 6073 }; 6074 6075This ioctl injects an event channel interrupt directly to the guest vCPU. 6076 60774.136 KVM_S390_PV_CPU_COMMAND 6078----------------------------- 6079 6080:Capability: KVM_CAP_S390_PROTECTED_DUMP 6081:Architectures: s390 6082:Type: vcpu ioctl 6083:Parameters: none 6084:Returns: 0 on success, < 0 on error 6085 6086This ioctl closely mirrors `KVM_S390_PV_COMMAND` but handles requests 6087for vcpus. It re-uses the kvm_s390_pv_dmp struct and hence also shares 6088the command ids. 6089 6090**command:** 6091 6092KVM_PV_DUMP 6093 Presents an API that provides calls which facilitate dumping a vcpu 6094 of a protected VM. 6095 6096**subcommand:** 6097 6098KVM_PV_DUMP_CPU 6099 Provides encrypted dump data like register values. 6100 The length of the returned data is provided by uv_info.guest_cpu_stor_len. 6101 61024.137 KVM_S390_ZPCI_OP 6103---------------------- 6104 6105:Capability: KVM_CAP_S390_ZPCI_OP 6106:Architectures: s390 6107:Type: vm ioctl 6108:Parameters: struct kvm_s390_zpci_op (in) 6109:Returns: 0 on success, <0 on error 6110 6111Used to manage hardware-assisted virtualization features for zPCI devices. 6112 6113Parameters are specified via the following structure:: 6114 6115 struct kvm_s390_zpci_op { 6116 /* in */ 6117 __u32 fh; /* target device */ 6118 __u8 op; /* operation to perform */ 6119 __u8 pad[3]; 6120 union { 6121 /* for KVM_S390_ZPCIOP_REG_AEN */ 6122 struct { 6123 __u64 ibv; /* Guest addr of interrupt bit vector */ 6124 __u64 sb; /* Guest addr of summary bit */ 6125 __u32 flags; 6126 __u32 noi; /* Number of interrupts */ 6127 __u8 isc; /* Guest interrupt subclass */ 6128 __u8 sbo; /* Offset of guest summary bit vector */ 6129 __u16 pad; 6130 } reg_aen; 6131 __u64 reserved[8]; 6132 } u; 6133 }; 6134 6135The type of operation is specified in the "op" field. 6136KVM_S390_ZPCIOP_REG_AEN is used to register the VM for adapter event 6137notification interpretation, which will allow firmware delivery of adapter 6138events directly to the vm, with KVM providing a backup delivery mechanism; 6139KVM_S390_ZPCIOP_DEREG_AEN is used to subsequently disable interpretation of 6140adapter event notifications. 6141 6142The target zPCI function must also be specified via the "fh" field. For the 6143KVM_S390_ZPCIOP_REG_AEN operation, additional information to establish firmware 6144delivery must be provided via the "reg_aen" struct. 6145 6146The "pad" and "reserved" fields may be used for future extensions and should be 6147set to 0s by userspace. 6148 61494.138 KVM_ARM_SET_COUNTER_OFFSET 6150-------------------------------- 6151 6152:Capability: KVM_CAP_COUNTER_OFFSET 6153:Architectures: arm64 6154:Type: vm ioctl 6155:Parameters: struct kvm_arm_counter_offset (in) 6156:Returns: 0 on success, < 0 on error 6157 6158This capability indicates that userspace is able to apply a single VM-wide 6159offset to both the virtual and physical counters as viewed by the guest 6160using the KVM_ARM_SET_CNT_OFFSET ioctl and the following data structure: 6161 6162:: 6163 6164 struct kvm_arm_counter_offset { 6165 __u64 counter_offset; 6166 __u64 reserved; 6167 }; 6168 6169The offset describes a number of counter cycles that are subtracted from 6170both virtual and physical counter views (similar to the effects of the 6171CNTVOFF_EL2 and CNTPOFF_EL2 system registers, but only global). The offset 6172always applies to all vcpus (already created or created after this ioctl) 6173for this VM. 6174 6175It is userspace's responsibility to compute the offset based, for example, 6176on previous values of the guest counters. 6177 6178Any value other than 0 for the "reserved" field may result in an error 6179(-EINVAL) being returned. This ioctl can also return -EBUSY if any vcpu 6180ioctl is issued concurrently. 6181 6182Note that using this ioctl results in KVM ignoring subsequent userspace 6183writes to the CNTVCT_EL0 and CNTPCT_EL0 registers using the SET_ONE_REG 6184interface. No error will be returned, but the resulting offset will not be 6185applied. 6186 6187.. _KVM_ARM_GET_REG_WRITABLE_MASKS: 6188 61894.139 KVM_ARM_GET_REG_WRITABLE_MASKS 6190------------------------------------------- 6191 6192:Capability: KVM_CAP_ARM_SUPPORTED_REG_MASK_RANGES 6193:Architectures: arm64 6194:Type: vm ioctl 6195:Parameters: struct reg_mask_range (in/out) 6196:Returns: 0 on success, < 0 on error 6197 6198 6199:: 6200 6201 #define KVM_ARM_FEATURE_ID_RANGE 0 6202 #define KVM_ARM_FEATURE_ID_RANGE_SIZE (3 * 8 * 8) 6203 6204 struct reg_mask_range { 6205 __u64 addr; /* Pointer to mask array */ 6206 __u32 range; /* Requested range */ 6207 __u32 reserved[13]; 6208 }; 6209 6210This ioctl copies the writable masks for a selected range of registers to 6211userspace. 6212 6213The ``addr`` field is a pointer to the destination array where KVM copies 6214the writable masks. 6215 6216The ``range`` field indicates the requested range of registers. 6217``KVM_CHECK_EXTENSION`` for the ``KVM_CAP_ARM_SUPPORTED_REG_MASK_RANGES`` 6218capability returns the supported ranges, expressed as a set of flags. Each 6219flag's bit index represents a possible value for the ``range`` field. 6220All other values are reserved for future use and KVM may return an error. 6221 6222The ``reserved[13]`` array is reserved for future use and should be 0, or 6223KVM may return an error. 6224 6225KVM_ARM_FEATURE_ID_RANGE (0) 6226^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 6227 6228The Feature ID range is defined as the AArch64 System register space with 6229op0==3, op1=={0, 1, 3}, CRn==0, CRm=={0-7}, op2=={0-7}. 6230 6231The mask returned array pointed to by ``addr`` is indexed by the macro 6232``ARM64_FEATURE_ID_RANGE_IDX(op0, op1, crn, crm, op2)``, allowing userspace 6233to know what fields can be changed for the system register described by 6234``op0, op1, crn, crm, op2``. KVM rejects ID register values that describe a 6235superset of the features supported by the system. 6236 62374.140 KVM_SET_USER_MEMORY_REGION2 6238--------------------------------- 6239 6240:Capability: KVM_CAP_USER_MEMORY2 6241:Architectures: all 6242:Type: vm ioctl 6243:Parameters: struct kvm_userspace_memory_region2 (in) 6244:Returns: 0 on success, -1 on error 6245 6246KVM_SET_USER_MEMORY_REGION2 is an extension to KVM_SET_USER_MEMORY_REGION that 6247allows mapping guest_memfd memory into a guest. All fields shared with 6248KVM_SET_USER_MEMORY_REGION identically. Userspace can set KVM_MEM_GUEST_MEMFD 6249in flags to have KVM bind the memory region to a given guest_memfd range of 6250[guest_memfd_offset, guest_memfd_offset + memory_size]. The target guest_memfd 6251must point at a file created via KVM_CREATE_GUEST_MEMFD on the current VM, and 6252the target range must not be bound to any other memory region. All standard 6253bounds checks apply (use common sense). 6254 6255:: 6256 6257 struct kvm_userspace_memory_region2 { 6258 __u32 slot; 6259 __u32 flags; 6260 __u64 guest_phys_addr; 6261 __u64 memory_size; /* bytes */ 6262 __u64 userspace_addr; /* start of the userspace allocated memory */ 6263 __u64 guest_memfd_offset; 6264 __u32 guest_memfd; 6265 __u32 pad1; 6266 __u64 pad2[14]; 6267 }; 6268 6269A KVM_MEM_GUEST_MEMFD region _must_ have a valid guest_memfd (private memory) and 6270userspace_addr (shared memory). However, "valid" for userspace_addr simply 6271means that the address itself must be a legal userspace address. The backing 6272mapping for userspace_addr is not required to be valid/populated at the time of 6273KVM_SET_USER_MEMORY_REGION2, e.g. shared memory can be lazily mapped/allocated 6274on-demand. 6275 6276When mapping a gfn into the guest, KVM selects shared vs. private, i.e consumes 6277userspace_addr vs. guest_memfd, based on the gfn's KVM_MEMORY_ATTRIBUTE_PRIVATE 6278state. At VM creation time, all memory is shared, i.e. the PRIVATE attribute 6279is '0' for all gfns. Userspace can control whether memory is shared/private by 6280toggling KVM_MEMORY_ATTRIBUTE_PRIVATE via KVM_SET_MEMORY_ATTRIBUTES as needed. 6281 6282S390: 6283^^^^^ 6284 6285Returns -EINVAL if the VM has the KVM_VM_S390_UCONTROL flag set. 6286Returns -EINVAL if called on a protected VM. 6287 62884.141 KVM_SET_MEMORY_ATTRIBUTES 6289------------------------------- 6290 6291:Capability: KVM_CAP_MEMORY_ATTRIBUTES 6292:Architectures: x86 6293:Type: vm ioctl 6294:Parameters: struct kvm_memory_attributes (in) 6295:Returns: 0 on success, <0 on error 6296 6297KVM_SET_MEMORY_ATTRIBUTES allows userspace to set memory attributes for a range 6298of guest physical memory. 6299 6300:: 6301 6302 struct kvm_memory_attributes { 6303 __u64 address; 6304 __u64 size; 6305 __u64 attributes; 6306 __u64 flags; 6307 }; 6308 6309 #define KVM_MEMORY_ATTRIBUTE_PRIVATE (1ULL << 3) 6310 6311The address and size must be page aligned. The supported attributes can be 6312retrieved via ioctl(KVM_CHECK_EXTENSION) on KVM_CAP_MEMORY_ATTRIBUTES. If 6313executed on a VM, KVM_CAP_MEMORY_ATTRIBUTES precisely returns the attributes 6314supported by that VM. If executed at system scope, KVM_CAP_MEMORY_ATTRIBUTES 6315returns all attributes supported by KVM. The only attribute defined at this 6316time is KVM_MEMORY_ATTRIBUTE_PRIVATE, which marks the associated gfn as being 6317guest private memory. 6318 6319Note, there is no "get" API. Userspace is responsible for explicitly tracking 6320the state of a gfn/page as needed. 6321 6322The "flags" field is reserved for future extensions and must be '0'. 6323 63244.142 KVM_CREATE_GUEST_MEMFD 6325---------------------------- 6326 6327:Capability: KVM_CAP_GUEST_MEMFD 6328:Architectures: none 6329:Type: vm ioctl 6330:Parameters: struct kvm_create_guest_memfd(in) 6331:Returns: A file descriptor on success, <0 on error 6332 6333KVM_CREATE_GUEST_MEMFD creates an anonymous file and returns a file descriptor 6334that refers to it. guest_memfd files are roughly analogous to files created 6335via memfd_create(), e.g. guest_memfd files live in RAM, have volatile storage, 6336and are automatically released when the last reference is dropped. Unlike 6337"regular" memfd_create() files, guest_memfd files are bound to their owning 6338virtual machine (see below), cannot be mapped, read, or written by userspace, 6339and cannot be resized (guest_memfd files do however support PUNCH_HOLE). 6340 6341:: 6342 6343 struct kvm_create_guest_memfd { 6344 __u64 size; 6345 __u64 flags; 6346 __u64 reserved[6]; 6347 }; 6348 6349Conceptually, the inode backing a guest_memfd file represents physical memory, 6350i.e. is coupled to the virtual machine as a thing, not to a "struct kvm". The 6351file itself, which is bound to a "struct kvm", is that instance's view of the 6352underlying memory, e.g. effectively provides the translation of guest addresses 6353to host memory. This allows for use cases where multiple KVM structures are 6354used to manage a single virtual machine, e.g. when performing intrahost 6355migration of a virtual machine. 6356 6357KVM currently only supports mapping guest_memfd via KVM_SET_USER_MEMORY_REGION2, 6358and more specifically via the guest_memfd and guest_memfd_offset fields in 6359"struct kvm_userspace_memory_region2", where guest_memfd_offset is the offset 6360into the guest_memfd instance. For a given guest_memfd file, there can be at 6361most one mapping per page, i.e. binding multiple memory regions to a single 6362guest_memfd range is not allowed (any number of memory regions can be bound to 6363a single guest_memfd file, but the bound ranges must not overlap). 6364 6365See KVM_SET_USER_MEMORY_REGION2 for additional details. 6366 63674.143 KVM_PRE_FAULT_MEMORY 6368------------------------ 6369 6370:Capability: KVM_CAP_PRE_FAULT_MEMORY 6371:Architectures: none 6372:Type: vcpu ioctl 6373:Parameters: struct kvm_pre_fault_memory (in/out) 6374:Returns: 0 if at least one page is processed, < 0 on error 6375 6376Errors: 6377 6378 ========== =============================================================== 6379 EINVAL The specified `gpa` and `size` were invalid (e.g. not 6380 page aligned, causes an overflow, or size is zero). 6381 ENOENT The specified `gpa` is outside defined memslots. 6382 EINTR An unmasked signal is pending and no page was processed. 6383 EFAULT The parameter address was invalid. 6384 EOPNOTSUPP Mapping memory for a GPA is unsupported by the 6385 hypervisor, and/or for the current vCPU state/mode. 6386 EIO unexpected error conditions (also causes a WARN) 6387 ========== =============================================================== 6388 6389:: 6390 6391 struct kvm_pre_fault_memory { 6392 /* in/out */ 6393 __u64 gpa; 6394 __u64 size; 6395 /* in */ 6396 __u64 flags; 6397 __u64 padding[5]; 6398 }; 6399 6400KVM_PRE_FAULT_MEMORY populates KVM's stage-2 page tables used to map memory 6401for the current vCPU state. KVM maps memory as if the vCPU generated a 6402stage-2 read page fault, e.g. faults in memory as needed, but doesn't break 6403CoW. However, KVM does not mark any newly created stage-2 PTE as Accessed. 6404 6405In some cases, multiple vCPUs might share the page tables. In this 6406case, the ioctl can be called in parallel. 6407 6408When the ioctl returns, the input values are updated to point to the 6409remaining range. If `size` > 0 on return, the caller can just issue 6410the ioctl again with the same `struct kvm_map_memory` argument. 6411 6412Shadow page tables cannot support this ioctl because they 6413are indexed by virtual address or nested guest physical address. 6414Calling this ioctl when the guest is using shadow page tables (for 6415example because it is running a nested guest with nested page tables) 6416will fail with `EOPNOTSUPP` even if `KVM_CHECK_EXTENSION` reports 6417the capability to be present. 6418 6419`flags` must currently be zero. 6420 6421 64225. The kvm_run structure 6423======================== 6424 6425Application code obtains a pointer to the kvm_run structure by 6426mmap()ing a vcpu fd. From that point, application code can control 6427execution by changing fields in kvm_run prior to calling the KVM_RUN 6428ioctl, and obtain information about the reason KVM_RUN returned by 6429looking up structure members. 6430 6431:: 6432 6433 struct kvm_run { 6434 /* in */ 6435 __u8 request_interrupt_window; 6436 6437Request that KVM_RUN return when it becomes possible to inject external 6438interrupts into the guest. Useful in conjunction with KVM_INTERRUPT. 6439 6440:: 6441 6442 __u8 immediate_exit; 6443 6444This field is polled once when KVM_RUN starts; if non-zero, KVM_RUN 6445exits immediately, returning -EINTR. In the common scenario where a 6446signal is used to "kick" a VCPU out of KVM_RUN, this field can be used 6447to avoid usage of KVM_SET_SIGNAL_MASK, which has worse scalability. 6448Rather than blocking the signal outside KVM_RUN, userspace can set up 6449a signal handler that sets run->immediate_exit to a non-zero value. 6450 6451This field is ignored if KVM_CAP_IMMEDIATE_EXIT is not available. 6452 6453:: 6454 6455 __u8 padding1[6]; 6456 6457 /* out */ 6458 __u32 exit_reason; 6459 6460When KVM_RUN has returned successfully (return value 0), this informs 6461application code why KVM_RUN has returned. Allowable values for this 6462field are detailed below. 6463 6464:: 6465 6466 __u8 ready_for_interrupt_injection; 6467 6468If request_interrupt_window has been specified, this field indicates 6469an interrupt can be injected now with KVM_INTERRUPT. 6470 6471:: 6472 6473 __u8 if_flag; 6474 6475The value of the current interrupt flag. Only valid if in-kernel 6476local APIC is not used. 6477 6478:: 6479 6480 __u16 flags; 6481 6482More architecture-specific flags detailing state of the VCPU that may 6483affect the device's behavior. Current defined flags:: 6484 6485 /* x86, set if the VCPU is in system management mode */ 6486 #define KVM_RUN_X86_SMM (1 << 0) 6487 /* x86, set if bus lock detected in VM */ 6488 #define KVM_RUN_X86_BUS_LOCK (1 << 1) 6489 /* x86, set if the VCPU is executing a nested (L2) guest */ 6490 #define KVM_RUN_X86_GUEST_MODE (1 << 2) 6491 6492 /* arm64, set for KVM_EXIT_DEBUG */ 6493 #define KVM_DEBUG_ARCH_HSR_HIGH_VALID (1 << 0) 6494 6495:: 6496 6497 /* in (pre_kvm_run), out (post_kvm_run) */ 6498 __u64 cr8; 6499 6500The value of the cr8 register. Only valid if in-kernel local APIC is 6501not used. Both input and output. 6502 6503:: 6504 6505 __u64 apic_base; 6506 6507The value of the APIC BASE msr. Only valid if in-kernel local 6508APIC is not used. Both input and output. 6509 6510:: 6511 6512 union { 6513 /* KVM_EXIT_UNKNOWN */ 6514 struct { 6515 __u64 hardware_exit_reason; 6516 } hw; 6517 6518If exit_reason is KVM_EXIT_UNKNOWN, the vcpu has exited due to unknown 6519reasons. Further architecture-specific information is available in 6520hardware_exit_reason. 6521 6522:: 6523 6524 /* KVM_EXIT_FAIL_ENTRY */ 6525 struct { 6526 __u64 hardware_entry_failure_reason; 6527 __u32 cpu; /* if KVM_LAST_CPU */ 6528 } fail_entry; 6529 6530If exit_reason is KVM_EXIT_FAIL_ENTRY, the vcpu could not be run due 6531to unknown reasons. Further architecture-specific information is 6532available in hardware_entry_failure_reason. 6533 6534:: 6535 6536 /* KVM_EXIT_EXCEPTION */ 6537 struct { 6538 __u32 exception; 6539 __u32 error_code; 6540 } ex; 6541 6542Unused. 6543 6544:: 6545 6546 /* KVM_EXIT_IO */ 6547 struct { 6548 #define KVM_EXIT_IO_IN 0 6549 #define KVM_EXIT_IO_OUT 1 6550 __u8 direction; 6551 __u8 size; /* bytes */ 6552 __u16 port; 6553 __u32 count; 6554 __u64 data_offset; /* relative to kvm_run start */ 6555 } io; 6556 6557If exit_reason is KVM_EXIT_IO, then the vcpu has 6558executed a port I/O instruction which could not be satisfied by kvm. 6559data_offset describes where the data is located (KVM_EXIT_IO_OUT) or 6560where kvm expects application code to place the data for the next 6561KVM_RUN invocation (KVM_EXIT_IO_IN). Data format is a packed array. 6562 6563:: 6564 6565 /* KVM_EXIT_DEBUG */ 6566 struct { 6567 struct kvm_debug_exit_arch arch; 6568 } debug; 6569 6570If the exit_reason is KVM_EXIT_DEBUG, then a vcpu is processing a debug event 6571for which architecture specific information is returned. 6572 6573:: 6574 6575 /* KVM_EXIT_MMIO */ 6576 struct { 6577 __u64 phys_addr; 6578 __u8 data[8]; 6579 __u32 len; 6580 __u8 is_write; 6581 } mmio; 6582 6583If exit_reason is KVM_EXIT_MMIO, then the vcpu has 6584executed a memory-mapped I/O instruction which could not be satisfied 6585by kvm. The 'data' member contains the written data if 'is_write' is 6586true, and should be filled by application code otherwise. 6587 6588The 'data' member contains, in its first 'len' bytes, the value as it would 6589appear if the VCPU performed a load or store of the appropriate width directly 6590to the byte array. 6591 6592.. note:: 6593 6594 For KVM_EXIT_IO, KVM_EXIT_MMIO, KVM_EXIT_OSI, KVM_EXIT_PAPR, KVM_EXIT_XEN, 6595 KVM_EXIT_EPR, KVM_EXIT_X86_RDMSR and KVM_EXIT_X86_WRMSR the corresponding 6596 operations are complete (and guest state is consistent) only after userspace 6597 has re-entered the kernel with KVM_RUN. The kernel side will first finish 6598 incomplete operations and then check for pending signals. 6599 6600 The pending state of the operation is not preserved in state which is 6601 visible to userspace, thus userspace should ensure that the operation is 6602 completed before performing a live migration. Userspace can re-enter the 6603 guest with an unmasked signal pending or with the immediate_exit field set 6604 to complete pending operations without allowing any further instructions 6605 to be executed. 6606 6607:: 6608 6609 /* KVM_EXIT_HYPERCALL */ 6610 struct { 6611 __u64 nr; 6612 __u64 args[6]; 6613 __u64 ret; 6614 __u64 flags; 6615 } hypercall; 6616 6617 6618It is strongly recommended that userspace use ``KVM_EXIT_IO`` (x86) or 6619``KVM_EXIT_MMIO`` (all except s390) to implement functionality that 6620requires a guest to interact with host userspace. 6621 6622.. note:: KVM_EXIT_IO is significantly faster than KVM_EXIT_MMIO. 6623 6624For arm64: 6625---------- 6626 6627SMCCC exits can be enabled depending on the configuration of the SMCCC 6628filter. See the Documentation/virt/kvm/devices/vm.rst 6629``KVM_ARM_SMCCC_FILTER`` for more details. 6630 6631``nr`` contains the function ID of the guest's SMCCC call. Userspace is 6632expected to use the ``KVM_GET_ONE_REG`` ioctl to retrieve the call 6633parameters from the vCPU's GPRs. 6634 6635Definition of ``flags``: 6636 - ``KVM_HYPERCALL_EXIT_SMC``: Indicates that the guest used the SMC 6637 conduit to initiate the SMCCC call. If this bit is 0 then the guest 6638 used the HVC conduit for the SMCCC call. 6639 6640 - ``KVM_HYPERCALL_EXIT_16BIT``: Indicates that the guest used a 16bit 6641 instruction to initiate the SMCCC call. If this bit is 0 then the 6642 guest used a 32bit instruction. An AArch64 guest always has this 6643 bit set to 0. 6644 6645At the point of exit, PC points to the instruction immediately following 6646the trapping instruction. 6647 6648:: 6649 6650 /* KVM_EXIT_TPR_ACCESS */ 6651 struct { 6652 __u64 rip; 6653 __u32 is_write; 6654 __u32 pad; 6655 } tpr_access; 6656 6657To be documented (KVM_TPR_ACCESS_REPORTING). 6658 6659:: 6660 6661 /* KVM_EXIT_S390_SIEIC */ 6662 struct { 6663 __u8 icptcode; 6664 __u64 mask; /* psw upper half */ 6665 __u64 addr; /* psw lower half */ 6666 __u16 ipa; 6667 __u32 ipb; 6668 } s390_sieic; 6669 6670s390 specific. 6671 6672:: 6673 6674 /* KVM_EXIT_S390_RESET */ 6675 #define KVM_S390_RESET_POR 1 6676 #define KVM_S390_RESET_CLEAR 2 6677 #define KVM_S390_RESET_SUBSYSTEM 4 6678 #define KVM_S390_RESET_CPU_INIT 8 6679 #define KVM_S390_RESET_IPL 16 6680 __u64 s390_reset_flags; 6681 6682s390 specific. 6683 6684:: 6685 6686 /* KVM_EXIT_S390_UCONTROL */ 6687 struct { 6688 __u64 trans_exc_code; 6689 __u32 pgm_code; 6690 } s390_ucontrol; 6691 6692s390 specific. A page fault has occurred for a user controlled virtual 6693machine (KVM_VM_S390_UNCONTROL) on its host page table that cannot be 6694resolved by the kernel. 6695The program code and the translation exception code that were placed 6696in the cpu's lowcore are presented here as defined by the z Architecture 6697Principles of Operation Book in the Chapter for Dynamic Address Translation 6698(DAT) 6699 6700:: 6701 6702 /* KVM_EXIT_DCR */ 6703 struct { 6704 __u32 dcrn; 6705 __u32 data; 6706 __u8 is_write; 6707 } dcr; 6708 6709Deprecated - was used for 440 KVM. 6710 6711:: 6712 6713 /* KVM_EXIT_OSI */ 6714 struct { 6715 __u64 gprs[32]; 6716 } osi; 6717 6718MOL uses a special hypercall interface it calls 'OSI'. To enable it, we catch 6719hypercalls and exit with this exit struct that contains all the guest gprs. 6720 6721If exit_reason is KVM_EXIT_OSI, then the vcpu has triggered such a hypercall. 6722Userspace can now handle the hypercall and when it's done modify the gprs as 6723necessary. Upon guest entry all guest GPRs will then be replaced by the values 6724in this struct. 6725 6726:: 6727 6728 /* KVM_EXIT_PAPR_HCALL */ 6729 struct { 6730 __u64 nr; 6731 __u64 ret; 6732 __u64 args[9]; 6733 } papr_hcall; 6734 6735This is used on 64-bit PowerPC when emulating a pSeries partition, 6736e.g. with the 'pseries' machine type in qemu. It occurs when the 6737guest does a hypercall using the 'sc 1' instruction. The 'nr' field 6738contains the hypercall number (from the guest R3), and 'args' contains 6739the arguments (from the guest R4 - R12). Userspace should put the 6740return code in 'ret' and any extra returned values in args[]. 6741The possible hypercalls are defined in the Power Architecture Platform 6742Requirements (PAPR) document available from www.power.org (free 6743developer registration required to access it). 6744 6745:: 6746 6747 /* KVM_EXIT_S390_TSCH */ 6748 struct { 6749 __u16 subchannel_id; 6750 __u16 subchannel_nr; 6751 __u32 io_int_parm; 6752 __u32 io_int_word; 6753 __u32 ipb; 6754 __u8 dequeued; 6755 } s390_tsch; 6756 6757s390 specific. This exit occurs when KVM_CAP_S390_CSS_SUPPORT has been enabled 6758and TEST SUBCHANNEL was intercepted. If dequeued is set, a pending I/O 6759interrupt for the target subchannel has been dequeued and subchannel_id, 6760subchannel_nr, io_int_parm and io_int_word contain the parameters for that 6761interrupt. ipb is needed for instruction parameter decoding. 6762 6763:: 6764 6765 /* KVM_EXIT_EPR */ 6766 struct { 6767 __u32 epr; 6768 } epr; 6769 6770On FSL BookE PowerPC chips, the interrupt controller has a fast patch 6771interrupt acknowledge path to the core. When the core successfully 6772delivers an interrupt, it automatically populates the EPR register with 6773the interrupt vector number and acknowledges the interrupt inside 6774the interrupt controller. 6775 6776In case the interrupt controller lives in user space, we need to do 6777the interrupt acknowledge cycle through it to fetch the next to be 6778delivered interrupt vector using this exit. 6779 6780It gets triggered whenever both KVM_CAP_PPC_EPR are enabled and an 6781external interrupt has just been delivered into the guest. User space 6782should put the acknowledged interrupt vector into the 'epr' field. 6783 6784:: 6785 6786 /* KVM_EXIT_SYSTEM_EVENT */ 6787 struct { 6788 #define KVM_SYSTEM_EVENT_SHUTDOWN 1 6789 #define KVM_SYSTEM_EVENT_RESET 2 6790 #define KVM_SYSTEM_EVENT_CRASH 3 6791 #define KVM_SYSTEM_EVENT_WAKEUP 4 6792 #define KVM_SYSTEM_EVENT_SUSPEND 5 6793 #define KVM_SYSTEM_EVENT_SEV_TERM 6 6794 __u32 type; 6795 __u32 ndata; 6796 __u64 data[16]; 6797 } system_event; 6798 6799If exit_reason is KVM_EXIT_SYSTEM_EVENT then the vcpu has triggered 6800a system-level event using some architecture specific mechanism (hypercall 6801or some special instruction). In case of ARM64, this is triggered using 6802HVC instruction based PSCI call from the vcpu. 6803 6804The 'type' field describes the system-level event type. 6805Valid values for 'type' are: 6806 6807 - KVM_SYSTEM_EVENT_SHUTDOWN -- the guest has requested a shutdown of the 6808 VM. Userspace is not obliged to honour this, and if it does honour 6809 this does not need to destroy the VM synchronously (ie it may call 6810 KVM_RUN again before shutdown finally occurs). 6811 - KVM_SYSTEM_EVENT_RESET -- the guest has requested a reset of the VM. 6812 As with SHUTDOWN, userspace can choose to ignore the request, or 6813 to schedule the reset to occur in the future and may call KVM_RUN again. 6814 - KVM_SYSTEM_EVENT_CRASH -- the guest crash occurred and the guest 6815 has requested a crash condition maintenance. Userspace can choose 6816 to ignore the request, or to gather VM memory core dump and/or 6817 reset/shutdown of the VM. 6818 - KVM_SYSTEM_EVENT_SEV_TERM -- an AMD SEV guest requested termination. 6819 The guest physical address of the guest's GHCB is stored in `data[0]`. 6820 - KVM_SYSTEM_EVENT_WAKEUP -- the exiting vCPU is in a suspended state and 6821 KVM has recognized a wakeup event. Userspace may honor this event by 6822 marking the exiting vCPU as runnable, or deny it and call KVM_RUN again. 6823 - KVM_SYSTEM_EVENT_SUSPEND -- the guest has requested a suspension of 6824 the VM. 6825 6826If KVM_CAP_SYSTEM_EVENT_DATA is present, the 'data' field can contain 6827architecture specific information for the system-level event. Only 6828the first `ndata` items (possibly zero) of the data array are valid. 6829 6830 - for arm64, data[0] is set to KVM_SYSTEM_EVENT_RESET_FLAG_PSCI_RESET2 if 6831 the guest issued a SYSTEM_RESET2 call according to v1.1 of the PSCI 6832 specification. 6833 6834 - for RISC-V, data[0] is set to the value of the second argument of the 6835 ``sbi_system_reset`` call. 6836 6837Previous versions of Linux defined a `flags` member in this struct. The 6838field is now aliased to `data[0]`. Userspace can assume that it is only 6839written if ndata is greater than 0. 6840 6841For arm/arm64: 6842-------------- 6843 6844KVM_SYSTEM_EVENT_SUSPEND exits are enabled with the 6845KVM_CAP_ARM_SYSTEM_SUSPEND VM capability. If a guest invokes the PSCI 6846SYSTEM_SUSPEND function, KVM will exit to userspace with this event 6847type. 6848 6849It is the sole responsibility of userspace to implement the PSCI 6850SYSTEM_SUSPEND call according to ARM DEN0022D.b 5.19 "SYSTEM_SUSPEND". 6851KVM does not change the vCPU's state before exiting to userspace, so 6852the call parameters are left in-place in the vCPU registers. 6853 6854Userspace is _required_ to take action for such an exit. It must 6855either: 6856 6857 - Honor the guest request to suspend the VM. Userspace can request 6858 in-kernel emulation of suspension by setting the calling vCPU's 6859 state to KVM_MP_STATE_SUSPENDED. Userspace must configure the vCPU's 6860 state according to the parameters passed to the PSCI function when 6861 the calling vCPU is resumed. See ARM DEN0022D.b 5.19.1 "Intended use" 6862 for details on the function parameters. 6863 6864 - Deny the guest request to suspend the VM. See ARM DEN0022D.b 5.19.2 6865 "Caller responsibilities" for possible return values. 6866 6867:: 6868 6869 /* KVM_EXIT_IOAPIC_EOI */ 6870 struct { 6871 __u8 vector; 6872 } eoi; 6873 6874Indicates that the VCPU's in-kernel local APIC received an EOI for a 6875level-triggered IOAPIC interrupt. This exit only triggers when the 6876IOAPIC is implemented in userspace (i.e. KVM_CAP_SPLIT_IRQCHIP is enabled); 6877the userspace IOAPIC should process the EOI and retrigger the interrupt if 6878it is still asserted. Vector is the LAPIC interrupt vector for which the 6879EOI was received. 6880 6881:: 6882 6883 struct kvm_hyperv_exit { 6884 #define KVM_EXIT_HYPERV_SYNIC 1 6885 #define KVM_EXIT_HYPERV_HCALL 2 6886 #define KVM_EXIT_HYPERV_SYNDBG 3 6887 __u32 type; 6888 __u32 pad1; 6889 union { 6890 struct { 6891 __u32 msr; 6892 __u32 pad2; 6893 __u64 control; 6894 __u64 evt_page; 6895 __u64 msg_page; 6896 } synic; 6897 struct { 6898 __u64 input; 6899 __u64 result; 6900 __u64 params[2]; 6901 } hcall; 6902 struct { 6903 __u32 msr; 6904 __u32 pad2; 6905 __u64 control; 6906 __u64 status; 6907 __u64 send_page; 6908 __u64 recv_page; 6909 __u64 pending_page; 6910 } syndbg; 6911 } u; 6912 }; 6913 /* KVM_EXIT_HYPERV */ 6914 struct kvm_hyperv_exit hyperv; 6915 6916Indicates that the VCPU exits into userspace to process some tasks 6917related to Hyper-V emulation. 6918 6919Valid values for 'type' are: 6920 6921 - KVM_EXIT_HYPERV_SYNIC -- synchronously notify user-space about 6922 6923Hyper-V SynIC state change. Notification is used to remap SynIC 6924event/message pages and to enable/disable SynIC messages/events processing 6925in userspace. 6926 6927 - KVM_EXIT_HYPERV_SYNDBG -- synchronously notify user-space about 6928 6929Hyper-V Synthetic debugger state change. Notification is used to either update 6930the pending_page location or to send a control command (send the buffer located 6931in send_page or recv a buffer to recv_page). 6932 6933:: 6934 6935 /* KVM_EXIT_ARM_NISV */ 6936 struct { 6937 __u64 esr_iss; 6938 __u64 fault_ipa; 6939 } arm_nisv; 6940 6941Used on arm64 systems. If a guest accesses memory not in a memslot, 6942KVM will typically return to userspace and ask it to do MMIO emulation on its 6943behalf. However, for certain classes of instructions, no instruction decode 6944(direction, length of memory access) is provided, and fetching and decoding 6945the instruction from the VM is overly complicated to live in the kernel. 6946 6947Historically, when this situation occurred, KVM would print a warning and kill 6948the VM. KVM assumed that if the guest accessed non-memslot memory, it was 6949trying to do I/O, which just couldn't be emulated, and the warning message was 6950phrased accordingly. However, what happened more often was that a guest bug 6951caused access outside the guest memory areas which should lead to a more 6952meaningful warning message and an external abort in the guest, if the access 6953did not fall within an I/O window. 6954 6955Userspace implementations can query for KVM_CAP_ARM_NISV_TO_USER, and enable 6956this capability at VM creation. Once this is done, these types of errors will 6957instead return to userspace with KVM_EXIT_ARM_NISV, with the valid bits from 6958the ESR_EL2 in the esr_iss field, and the faulting IPA in the fault_ipa field. 6959Userspace can either fix up the access if it's actually an I/O access by 6960decoding the instruction from guest memory (if it's very brave) and continue 6961executing the guest, or it can decide to suspend, dump, or restart the guest. 6962 6963Note that KVM does not skip the faulting instruction as it does for 6964KVM_EXIT_MMIO, but userspace has to emulate any change to the processing state 6965if it decides to decode and emulate the instruction. 6966 6967This feature isn't available to protected VMs, as userspace does not 6968have access to the state that is required to perform the emulation. 6969Instead, a data abort exception is directly injected in the guest. 6970Note that although KVM_CAP_ARM_NISV_TO_USER will be reported if 6971queried outside of a protected VM context, the feature will not be 6972exposed if queried on a protected VM file descriptor. 6973 6974:: 6975 6976 /* KVM_EXIT_X86_RDMSR / KVM_EXIT_X86_WRMSR */ 6977 struct { 6978 __u8 error; /* user -> kernel */ 6979 __u8 pad[7]; 6980 __u32 reason; /* kernel -> user */ 6981 __u32 index; /* kernel -> user */ 6982 __u64 data; /* kernel <-> user */ 6983 } msr; 6984 6985Used on x86 systems. When the VM capability KVM_CAP_X86_USER_SPACE_MSR is 6986enabled, MSR accesses to registers that would invoke a #GP by KVM kernel code 6987may instead trigger a KVM_EXIT_X86_RDMSR exit for reads and KVM_EXIT_X86_WRMSR 6988exit for writes. 6989 6990The "reason" field specifies why the MSR interception occurred. Userspace will 6991only receive MSR exits when a particular reason was requested during through 6992ENABLE_CAP. Currently valid exit reasons are: 6993 6994============================ ======================================== 6995 KVM_MSR_EXIT_REASON_UNKNOWN access to MSR that is unknown to KVM 6996 KVM_MSR_EXIT_REASON_INVAL access to invalid MSRs or reserved bits 6997 KVM_MSR_EXIT_REASON_FILTER access blocked by KVM_X86_SET_MSR_FILTER 6998============================ ======================================== 6999 7000For KVM_EXIT_X86_RDMSR, the "index" field tells userspace which MSR the guest 7001wants to read. To respond to this request with a successful read, userspace 7002writes the respective data into the "data" field and must continue guest 7003execution to ensure the read data is transferred into guest register state. 7004 7005If the RDMSR request was unsuccessful, userspace indicates that with a "1" in 7006the "error" field. This will inject a #GP into the guest when the VCPU is 7007executed again. 7008 7009For KVM_EXIT_X86_WRMSR, the "index" field tells userspace which MSR the guest 7010wants to write. Once finished processing the event, userspace must continue 7011vCPU execution. If the MSR write was unsuccessful, userspace also sets the 7012"error" field to "1". 7013 7014See KVM_X86_SET_MSR_FILTER for details on the interaction with MSR filtering. 7015 7016:: 7017 7018 7019 struct kvm_xen_exit { 7020 #define KVM_EXIT_XEN_HCALL 1 7021 __u32 type; 7022 union { 7023 struct { 7024 __u32 longmode; 7025 __u32 cpl; 7026 __u64 input; 7027 __u64 result; 7028 __u64 params[6]; 7029 } hcall; 7030 } u; 7031 }; 7032 /* KVM_EXIT_XEN */ 7033 struct kvm_hyperv_exit xen; 7034 7035Indicates that the VCPU exits into userspace to process some tasks 7036related to Xen emulation. 7037 7038Valid values for 'type' are: 7039 7040 - KVM_EXIT_XEN_HCALL -- synchronously notify user-space about Xen hypercall. 7041 Userspace is expected to place the hypercall result into the appropriate 7042 field before invoking KVM_RUN again. 7043 7044:: 7045 7046 /* KVM_EXIT_RISCV_SBI */ 7047 struct { 7048 unsigned long extension_id; 7049 unsigned long function_id; 7050 unsigned long args[6]; 7051 unsigned long ret[2]; 7052 } riscv_sbi; 7053 7054If exit reason is KVM_EXIT_RISCV_SBI then it indicates that the VCPU has 7055done a SBI call which is not handled by KVM RISC-V kernel module. The details 7056of the SBI call are available in 'riscv_sbi' member of kvm_run structure. The 7057'extension_id' field of 'riscv_sbi' represents SBI extension ID whereas the 7058'function_id' field represents function ID of given SBI extension. The 'args' 7059array field of 'riscv_sbi' represents parameters for the SBI call and 'ret' 7060array field represents return values. The userspace should update the return 7061values of SBI call before resuming the VCPU. For more details on RISC-V SBI 7062spec refer, https://github.com/riscv/riscv-sbi-doc. 7063 7064:: 7065 7066 /* KVM_EXIT_MEMORY_FAULT */ 7067 struct { 7068 #define KVM_MEMORY_EXIT_FLAG_PRIVATE (1ULL << 3) 7069 __u64 flags; 7070 __u64 gpa; 7071 __u64 size; 7072 } memory_fault; 7073 7074KVM_EXIT_MEMORY_FAULT indicates the vCPU has encountered a memory fault that 7075could not be resolved by KVM. The 'gpa' and 'size' (in bytes) describe the 7076guest physical address range [gpa, gpa + size) of the fault. The 'flags' field 7077describes properties of the faulting access that are likely pertinent: 7078 7079 - KVM_MEMORY_EXIT_FLAG_PRIVATE - When set, indicates the memory fault occurred 7080 on a private memory access. When clear, indicates the fault occurred on a 7081 shared access. 7082 7083Note! KVM_EXIT_MEMORY_FAULT is unique among all KVM exit reasons in that it 7084accompanies a return code of '-1', not '0'! errno will always be set to EFAULT 7085or EHWPOISON when KVM exits with KVM_EXIT_MEMORY_FAULT, userspace should assume 7086kvm_run.exit_reason is stale/undefined for all other error numbers. 7087 7088:: 7089 7090 /* KVM_EXIT_NOTIFY */ 7091 struct { 7092 #define KVM_NOTIFY_CONTEXT_INVALID (1 << 0) 7093 __u32 flags; 7094 } notify; 7095 7096Used on x86 systems. When the VM capability KVM_CAP_X86_NOTIFY_VMEXIT is 7097enabled, a VM exit generated if no event window occurs in VM non-root mode 7098for a specified amount of time. Once KVM_X86_NOTIFY_VMEXIT_USER is set when 7099enabling the cap, it would exit to userspace with the exit reason 7100KVM_EXIT_NOTIFY for further handling. The "flags" field contains more 7101detailed info. 7102 7103The valid value for 'flags' is: 7104 7105 - KVM_NOTIFY_CONTEXT_INVALID -- the VM context is corrupted and not valid 7106 in VMCS. It would run into unknown result if resume the target VM. 7107 7108:: 7109 7110 /* Fix the size of the union. */ 7111 char padding[256]; 7112 }; 7113 7114 /* 7115 * shared registers between kvm and userspace. 7116 * kvm_valid_regs specifies the register classes set by the host 7117 * kvm_dirty_regs specified the register classes dirtied by userspace 7118 * struct kvm_sync_regs is architecture specific, as well as the 7119 * bits for kvm_valid_regs and kvm_dirty_regs 7120 */ 7121 __u64 kvm_valid_regs; 7122 __u64 kvm_dirty_regs; 7123 union { 7124 struct kvm_sync_regs regs; 7125 char padding[SYNC_REGS_SIZE_BYTES]; 7126 } s; 7127 7128If KVM_CAP_SYNC_REGS is defined, these fields allow userspace to access 7129certain guest registers without having to call SET/GET_*REGS. Thus we can 7130avoid some system call overhead if userspace has to handle the exit. 7131Userspace can query the validity of the structure by checking 7132kvm_valid_regs for specific bits. These bits are architecture specific 7133and usually define the validity of a groups of registers. (e.g. one bit 7134for general purpose registers) 7135 7136Please note that the kernel is allowed to use the kvm_run structure as the 7137primary storage for certain register types. Therefore, the kernel may use the 7138values in kvm_run even if the corresponding bit in kvm_dirty_regs is not set. 7139 7140 71416. Capabilities that can be enabled on vCPUs 7142============================================ 7143 7144There are certain capabilities that change the behavior of the virtual CPU or 7145the virtual machine when enabled. To enable them, please see section 4.37. 7146Below you can find a list of capabilities and what their effect on the vCPU or 7147the virtual machine is when enabling them. 7148 7149The following information is provided along with the description: 7150 7151 Architectures: 7152 which instruction set architectures provide this ioctl. 7153 x86 includes both i386 and x86_64. 7154 7155 Target: 7156 whether this is a per-vcpu or per-vm capability. 7157 7158 Parameters: 7159 what parameters are accepted by the capability. 7160 7161 Returns: 7162 the return value. General error numbers (EBADF, ENOMEM, EINVAL) 7163 are not detailed, but errors with specific meanings are. 7164 7165 71666.1 KVM_CAP_PPC_OSI 7167------------------- 7168 7169:Architectures: ppc 7170:Target: vcpu 7171:Parameters: none 7172:Returns: 0 on success; -1 on error 7173 7174This capability enables interception of OSI hypercalls that otherwise would 7175be treated as normal system calls to be injected into the guest. OSI hypercalls 7176were invented by Mac-on-Linux to have a standardized communication mechanism 7177between the guest and the host. 7178 7179When this capability is enabled, KVM_EXIT_OSI can occur. 7180 7181 71826.2 KVM_CAP_PPC_PAPR 7183-------------------- 7184 7185:Architectures: ppc 7186:Target: vcpu 7187:Parameters: none 7188:Returns: 0 on success; -1 on error 7189 7190This capability enables interception of PAPR hypercalls. PAPR hypercalls are 7191done using the hypercall instruction "sc 1". 7192 7193It also sets the guest privilege level to "supervisor" mode. Usually the guest 7194runs in "hypervisor" privilege mode with a few missing features. 7195 7196In addition to the above, it changes the semantics of SDR1. In this mode, the 7197HTAB address part of SDR1 contains an HVA instead of a GPA, as PAPR keeps the 7198HTAB invisible to the guest. 7199 7200When this capability is enabled, KVM_EXIT_PAPR_HCALL can occur. 7201 7202 72036.3 KVM_CAP_SW_TLB 7204------------------ 7205 7206:Architectures: ppc 7207:Target: vcpu 7208:Parameters: args[0] is the address of a struct kvm_config_tlb 7209:Returns: 0 on success; -1 on error 7210 7211:: 7212 7213 struct kvm_config_tlb { 7214 __u64 params; 7215 __u64 array; 7216 __u32 mmu_type; 7217 __u32 array_len; 7218 }; 7219 7220Configures the virtual CPU's TLB array, establishing a shared memory area 7221between userspace and KVM. The "params" and "array" fields are userspace 7222addresses of mmu-type-specific data structures. The "array_len" field is an 7223safety mechanism, and should be set to the size in bytes of the memory that 7224userspace has reserved for the array. It must be at least the size dictated 7225by "mmu_type" and "params". 7226 7227While KVM_RUN is active, the shared region is under control of KVM. Its 7228contents are undefined, and any modification by userspace results in 7229boundedly undefined behavior. 7230 7231On return from KVM_RUN, the shared region will reflect the current state of 7232the guest's TLB. If userspace makes any changes, it must call KVM_DIRTY_TLB 7233to tell KVM which entries have been changed, prior to calling KVM_RUN again 7234on this vcpu. 7235 7236For mmu types KVM_MMU_FSL_BOOKE_NOHV and KVM_MMU_FSL_BOOKE_HV: 7237 7238 - The "params" field is of type "struct kvm_book3e_206_tlb_params". 7239 - The "array" field points to an array of type "struct 7240 kvm_book3e_206_tlb_entry". 7241 - The array consists of all entries in the first TLB, followed by all 7242 entries in the second TLB. 7243 - Within a TLB, entries are ordered first by increasing set number. Within a 7244 set, entries are ordered by way (increasing ESEL). 7245 - The hash for determining set number in TLB0 is: (MAS2 >> 12) & (num_sets - 1) 7246 where "num_sets" is the tlb_sizes[] value divided by the tlb_ways[] value. 7247 - The tsize field of mas1 shall be set to 4K on TLB0, even though the 7248 hardware ignores this value for TLB0. 7249 72506.4 KVM_CAP_S390_CSS_SUPPORT 7251---------------------------- 7252 7253:Architectures: s390 7254:Target: vcpu 7255:Parameters: none 7256:Returns: 0 on success; -1 on error 7257 7258This capability enables support for handling of channel I/O instructions. 7259 7260TEST PENDING INTERRUPTION and the interrupt portion of TEST SUBCHANNEL are 7261handled in-kernel, while the other I/O instructions are passed to userspace. 7262 7263When this capability is enabled, KVM_EXIT_S390_TSCH will occur on TEST 7264SUBCHANNEL intercepts. 7265 7266Note that even though this capability is enabled per-vcpu, the complete 7267virtual machine is affected. 7268 72696.5 KVM_CAP_PPC_EPR 7270------------------- 7271 7272:Architectures: ppc 7273:Target: vcpu 7274:Parameters: args[0] defines whether the proxy facility is active 7275:Returns: 0 on success; -1 on error 7276 7277This capability enables or disables the delivery of interrupts through the 7278external proxy facility. 7279 7280When enabled (args[0] != 0), every time the guest gets an external interrupt 7281delivered, it automatically exits into user space with a KVM_EXIT_EPR exit 7282to receive the topmost interrupt vector. 7283 7284When disabled (args[0] == 0), behavior is as if this facility is unsupported. 7285 7286When this capability is enabled, KVM_EXIT_EPR can occur. 7287 72886.6 KVM_CAP_IRQ_MPIC 7289-------------------- 7290 7291:Architectures: ppc 7292:Parameters: args[0] is the MPIC device fd; 7293 args[1] is the MPIC CPU number for this vcpu 7294 7295This capability connects the vcpu to an in-kernel MPIC device. 7296 72976.7 KVM_CAP_IRQ_XICS 7298-------------------- 7299 7300:Architectures: ppc 7301:Target: vcpu 7302:Parameters: args[0] is the XICS device fd; 7303 args[1] is the XICS CPU number (server ID) for this vcpu 7304 7305This capability connects the vcpu to an in-kernel XICS device. 7306 73076.8 KVM_CAP_S390_IRQCHIP 7308------------------------ 7309 7310:Architectures: s390 7311:Target: vm 7312:Parameters: none 7313 7314This capability enables the in-kernel irqchip for s390. Please refer to 7315"4.24 KVM_CREATE_IRQCHIP" for details. 7316 73176.9 KVM_CAP_MIPS_FPU 7318-------------------- 7319 7320:Architectures: mips 7321:Target: vcpu 7322:Parameters: args[0] is reserved for future use (should be 0). 7323 7324This capability allows the use of the host Floating Point Unit by the guest. It 7325allows the Config1.FP bit to be set to enable the FPU in the guest. Once this is 7326done the ``KVM_REG_MIPS_FPR_*`` and ``KVM_REG_MIPS_FCR_*`` registers can be 7327accessed (depending on the current guest FPU register mode), and the Status.FR, 7328Config5.FRE bits are accessible via the KVM API and also from the guest, 7329depending on them being supported by the FPU. 7330 73316.10 KVM_CAP_MIPS_MSA 7332--------------------- 7333 7334:Architectures: mips 7335:Target: vcpu 7336:Parameters: args[0] is reserved for future use (should be 0). 7337 7338This capability allows the use of the MIPS SIMD Architecture (MSA) by the guest. 7339It allows the Config3.MSAP bit to be set to enable the use of MSA by the guest. 7340Once this is done the ``KVM_REG_MIPS_VEC_*`` and ``KVM_REG_MIPS_MSA_*`` 7341registers can be accessed, and the Config5.MSAEn bit is accessible via the 7342KVM API and also from the guest. 7343 73446.74 KVM_CAP_SYNC_REGS 7345---------------------- 7346 7347:Architectures: s390, x86 7348:Target: s390: always enabled, x86: vcpu 7349:Parameters: none 7350:Returns: x86: KVM_CHECK_EXTENSION returns a bit-array indicating which register 7351 sets are supported 7352 (bitfields defined in arch/x86/include/uapi/asm/kvm.h). 7353 7354As described above in the kvm_sync_regs struct info in section 5 (kvm_run): 7355KVM_CAP_SYNC_REGS "allow[s] userspace to access certain guest registers 7356without having to call SET/GET_*REGS". This reduces overhead by eliminating 7357repeated ioctl calls for setting and/or getting register values. This is 7358particularly important when userspace is making synchronous guest state 7359modifications, e.g. when emulating and/or intercepting instructions in 7360userspace. 7361 7362For s390 specifics, please refer to the source code. 7363 7364For x86: 7365 7366- the register sets to be copied out to kvm_run are selectable 7367 by userspace (rather that all sets being copied out for every exit). 7368- vcpu_events are available in addition to regs and sregs. 7369 7370For x86, the 'kvm_valid_regs' field of struct kvm_run is overloaded to 7371function as an input bit-array field set by userspace to indicate the 7372specific register sets to be copied out on the next exit. 7373 7374To indicate when userspace has modified values that should be copied into 7375the vCPU, the all architecture bitarray field, 'kvm_dirty_regs' must be set. 7376This is done using the same bitflags as for the 'kvm_valid_regs' field. 7377If the dirty bit is not set, then the register set values will not be copied 7378into the vCPU even if they've been modified. 7379 7380Unused bitfields in the bitarrays must be set to zero. 7381 7382:: 7383 7384 struct kvm_sync_regs { 7385 struct kvm_regs regs; 7386 struct kvm_sregs sregs; 7387 struct kvm_vcpu_events events; 7388 }; 7389 73906.75 KVM_CAP_PPC_IRQ_XIVE 7391------------------------- 7392 7393:Architectures: ppc 7394:Target: vcpu 7395:Parameters: args[0] is the XIVE device fd; 7396 args[1] is the XIVE CPU number (server ID) for this vcpu 7397 7398This capability connects the vcpu to an in-kernel XIVE device. 7399 74007. Capabilities that can be enabled on VMs 7401========================================== 7402 7403There are certain capabilities that change the behavior of the virtual 7404machine when enabled. To enable them, please see section 4.37. Below 7405you can find a list of capabilities and what their effect on the VM 7406is when enabling them. 7407 7408The following information is provided along with the description: 7409 7410 Architectures: 7411 which instruction set architectures provide this ioctl. 7412 x86 includes both i386 and x86_64. 7413 7414 Parameters: 7415 what parameters are accepted by the capability. 7416 7417 Returns: 7418 the return value. General error numbers (EBADF, ENOMEM, EINVAL) 7419 are not detailed, but errors with specific meanings are. 7420 7421 74227.1 KVM_CAP_PPC_ENABLE_HCALL 7423---------------------------- 7424 7425:Architectures: ppc 7426:Parameters: args[0] is the sPAPR hcall number; 7427 args[1] is 0 to disable, 1 to enable in-kernel handling 7428 7429This capability controls whether individual sPAPR hypercalls (hcalls) 7430get handled by the kernel or not. Enabling or disabling in-kernel 7431handling of an hcall is effective across the VM. On creation, an 7432initial set of hcalls are enabled for in-kernel handling, which 7433consists of those hcalls for which in-kernel handlers were implemented 7434before this capability was implemented. If disabled, the kernel will 7435not to attempt to handle the hcall, but will always exit to userspace 7436to handle it. Note that it may not make sense to enable some and 7437disable others of a group of related hcalls, but KVM does not prevent 7438userspace from doing that. 7439 7440If the hcall number specified is not one that has an in-kernel 7441implementation, the KVM_ENABLE_CAP ioctl will fail with an EINVAL 7442error. 7443 74447.2 KVM_CAP_S390_USER_SIGP 7445-------------------------- 7446 7447:Architectures: s390 7448:Parameters: none 7449 7450This capability controls which SIGP orders will be handled completely in user 7451space. With this capability enabled, all fast orders will be handled completely 7452in the kernel: 7453 7454- SENSE 7455- SENSE RUNNING 7456- EXTERNAL CALL 7457- EMERGENCY SIGNAL 7458- CONDITIONAL EMERGENCY SIGNAL 7459 7460All other orders will be handled completely in user space. 7461 7462Only privileged operation exceptions will be checked for in the kernel (or even 7463in the hardware prior to interception). If this capability is not enabled, the 7464old way of handling SIGP orders is used (partially in kernel and user space). 7465 74667.3 KVM_CAP_S390_VECTOR_REGISTERS 7467--------------------------------- 7468 7469:Architectures: s390 7470:Parameters: none 7471:Returns: 0 on success, negative value on error 7472 7473Allows use of the vector registers introduced with z13 processor, and 7474provides for the synchronization between host and user space. Will 7475return -EINVAL if the machine does not support vectors. 7476 74777.4 KVM_CAP_S390_USER_STSI 7478-------------------------- 7479 7480:Architectures: s390 7481:Parameters: none 7482 7483This capability allows post-handlers for the STSI instruction. After 7484initial handling in the kernel, KVM exits to user space with 7485KVM_EXIT_S390_STSI to allow user space to insert further data. 7486 7487Before exiting to userspace, kvm handlers should fill in s390_stsi field of 7488vcpu->run:: 7489 7490 struct { 7491 __u64 addr; 7492 __u8 ar; 7493 __u8 reserved; 7494 __u8 fc; 7495 __u8 sel1; 7496 __u16 sel2; 7497 } s390_stsi; 7498 7499 @addr - guest address of STSI SYSIB 7500 @fc - function code 7501 @sel1 - selector 1 7502 @sel2 - selector 2 7503 @ar - access register number 7504 7505KVM handlers should exit to userspace with rc = -EREMOTE. 7506 75077.5 KVM_CAP_SPLIT_IRQCHIP 7508------------------------- 7509 7510:Architectures: x86 7511:Parameters: args[0] - number of routes reserved for userspace IOAPICs 7512:Returns: 0 on success, -1 on error 7513 7514Create a local apic for each processor in the kernel. This can be used 7515instead of KVM_CREATE_IRQCHIP if the userspace VMM wishes to emulate the 7516IOAPIC and PIC (and also the PIT, even though this has to be enabled 7517separately). 7518 7519This capability also enables in kernel routing of interrupt requests; 7520when KVM_CAP_SPLIT_IRQCHIP only routes of KVM_IRQ_ROUTING_MSI type are 7521used in the IRQ routing table. The first args[0] MSI routes are reserved 7522for the IOAPIC pins. Whenever the LAPIC receives an EOI for these routes, 7523a KVM_EXIT_IOAPIC_EOI vmexit will be reported to userspace. 7524 7525Fails if VCPU has already been created, or if the irqchip is already in the 7526kernel (i.e. KVM_CREATE_IRQCHIP has already been called). 7527 75287.6 KVM_CAP_S390_RI 7529------------------- 7530 7531:Architectures: s390 7532:Parameters: none 7533 7534Allows use of runtime-instrumentation introduced with zEC12 processor. 7535Will return -EINVAL if the machine does not support runtime-instrumentation. 7536Will return -EBUSY if a VCPU has already been created. 7537 75387.7 KVM_CAP_X2APIC_API 7539---------------------- 7540 7541:Architectures: x86 7542:Parameters: args[0] - features that should be enabled 7543:Returns: 0 on success, -EINVAL when args[0] contains invalid features 7544 7545Valid feature flags in args[0] are:: 7546 7547 #define KVM_X2APIC_API_USE_32BIT_IDS (1ULL << 0) 7548 #define KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK (1ULL << 1) 7549 7550Enabling KVM_X2APIC_API_USE_32BIT_IDS changes the behavior of 7551KVM_SET_GSI_ROUTING, KVM_SIGNAL_MSI, KVM_SET_LAPIC, and KVM_GET_LAPIC, 7552allowing the use of 32-bit APIC IDs. See KVM_CAP_X2APIC_API in their 7553respective sections. 7554 7555KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK must be enabled for x2APIC to work 7556in logical mode or with more than 255 VCPUs. Otherwise, KVM treats 0xff 7557as a broadcast even in x2APIC mode in order to support physical x2APIC 7558without interrupt remapping. This is undesirable in logical mode, 7559where 0xff represents CPUs 0-7 in cluster 0. 7560 75617.8 KVM_CAP_S390_USER_INSTR0 7562---------------------------- 7563 7564:Architectures: s390 7565:Parameters: none 7566 7567With this capability enabled, all illegal instructions 0x0000 (2 bytes) will 7568be intercepted and forwarded to user space. User space can use this 7569mechanism e.g. to realize 2-byte software breakpoints. The kernel will 7570not inject an operating exception for these instructions, user space has 7571to take care of that. 7572 7573This capability can be enabled dynamically even if VCPUs were already 7574created and are running. 7575 75767.9 KVM_CAP_S390_GS 7577------------------- 7578 7579:Architectures: s390 7580:Parameters: none 7581:Returns: 0 on success; -EINVAL if the machine does not support 7582 guarded storage; -EBUSY if a VCPU has already been created. 7583 7584Allows use of guarded storage for the KVM guest. 7585 75867.10 KVM_CAP_S390_AIS 7587--------------------- 7588 7589:Architectures: s390 7590:Parameters: none 7591 7592Allow use of adapter-interruption suppression. 7593:Returns: 0 on success; -EBUSY if a VCPU has already been created. 7594 75957.11 KVM_CAP_PPC_SMT 7596-------------------- 7597 7598:Architectures: ppc 7599:Parameters: vsmt_mode, flags 7600 7601Enabling this capability on a VM provides userspace with a way to set 7602the desired virtual SMT mode (i.e. the number of virtual CPUs per 7603virtual core). The virtual SMT mode, vsmt_mode, must be a power of 2 7604between 1 and 8. On POWER8, vsmt_mode must also be no greater than 7605the number of threads per subcore for the host. Currently flags must 7606be 0. A successful call to enable this capability will result in 7607vsmt_mode being returned when the KVM_CAP_PPC_SMT capability is 7608subsequently queried for the VM. This capability is only supported by 7609HV KVM, and can only be set before any VCPUs have been created. 7610The KVM_CAP_PPC_SMT_POSSIBLE capability indicates which virtual SMT 7611modes are available. 7612 76137.12 KVM_CAP_PPC_FWNMI 7614---------------------- 7615 7616:Architectures: ppc 7617:Parameters: none 7618 7619With this capability a machine check exception in the guest address 7620space will cause KVM to exit the guest with NMI exit reason. This 7621enables QEMU to build error log and branch to guest kernel registered 7622machine check handling routine. Without this capability KVM will 7623branch to guests' 0x200 interrupt vector. 7624 76257.13 KVM_CAP_X86_DISABLE_EXITS 7626------------------------------ 7627 7628:Architectures: x86 7629:Parameters: args[0] defines which exits are disabled 7630:Returns: 0 on success, -EINVAL when args[0] contains invalid exits 7631 7632Valid bits in args[0] are:: 7633 7634 #define KVM_X86_DISABLE_EXITS_MWAIT (1 << 0) 7635 #define KVM_X86_DISABLE_EXITS_HLT (1 << 1) 7636 #define KVM_X86_DISABLE_EXITS_PAUSE (1 << 2) 7637 #define KVM_X86_DISABLE_EXITS_CSTATE (1 << 3) 7638 7639Enabling this capability on a VM provides userspace with a way to no 7640longer intercept some instructions for improved latency in some 7641workloads, and is suggested when vCPUs are associated to dedicated 7642physical CPUs. More bits can be added in the future; userspace can 7643just pass the KVM_CHECK_EXTENSION result to KVM_ENABLE_CAP to disable 7644all such vmexits. 7645 7646Do not enable KVM_FEATURE_PV_UNHALT if you disable HLT exits. 7647 76487.14 KVM_CAP_S390_HPAGE_1M 7649-------------------------- 7650 7651:Architectures: s390 7652:Parameters: none 7653:Returns: 0 on success, -EINVAL if hpage module parameter was not set 7654 or cmma is enabled, or the VM has the KVM_VM_S390_UCONTROL 7655 flag set 7656 7657With this capability the KVM support for memory backing with 1m pages 7658through hugetlbfs can be enabled for a VM. After the capability is 7659enabled, cmma can't be enabled anymore and pfmfi and the storage key 7660interpretation are disabled. If cmma has already been enabled or the 7661hpage module parameter is not set to 1, -EINVAL is returned. 7662 7663While it is generally possible to create a huge page backed VM without 7664this capability, the VM will not be able to run. 7665 76667.15 KVM_CAP_MSR_PLATFORM_INFO 7667------------------------------ 7668 7669:Architectures: x86 7670:Parameters: args[0] whether feature should be enabled or not 7671 7672With this capability, a guest may read the MSR_PLATFORM_INFO MSR. Otherwise, 7673a #GP would be raised when the guest tries to access. Currently, this 7674capability does not enable write permissions of this MSR for the guest. 7675 76767.16 KVM_CAP_PPC_NESTED_HV 7677-------------------------- 7678 7679:Architectures: ppc 7680:Parameters: none 7681:Returns: 0 on success, -EINVAL when the implementation doesn't support 7682 nested-HV virtualization. 7683 7684HV-KVM on POWER9 and later systems allows for "nested-HV" 7685virtualization, which provides a way for a guest VM to run guests that 7686can run using the CPU's supervisor mode (privileged non-hypervisor 7687state). Enabling this capability on a VM depends on the CPU having 7688the necessary functionality and on the facility being enabled with a 7689kvm-hv module parameter. 7690 76917.17 KVM_CAP_EXCEPTION_PAYLOAD 7692------------------------------ 7693 7694:Architectures: x86 7695:Parameters: args[0] whether feature should be enabled or not 7696 7697With this capability enabled, CR2 will not be modified prior to the 7698emulated VM-exit when L1 intercepts a #PF exception that occurs in 7699L2. Similarly, for kvm-intel only, DR6 will not be modified prior to 7700the emulated VM-exit when L1 intercepts a #DB exception that occurs in 7701L2. As a result, when KVM_GET_VCPU_EVENTS reports a pending #PF (or 7702#DB) exception for L2, exception.has_payload will be set and the 7703faulting address (or the new DR6 bits*) will be reported in the 7704exception_payload field. Similarly, when userspace injects a #PF (or 7705#DB) into L2 using KVM_SET_VCPU_EVENTS, it is expected to set 7706exception.has_payload and to put the faulting address - or the new DR6 7707bits\ [#]_ - in the exception_payload field. 7708 7709This capability also enables exception.pending in struct 7710kvm_vcpu_events, which allows userspace to distinguish between pending 7711and injected exceptions. 7712 7713 7714.. [#] For the new DR6 bits, note that bit 16 is set iff the #DB exception 7715 will clear DR6.RTM. 7716 77177.18 KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 7718-------------------------------------- 7719 7720:Architectures: x86, arm64, mips 7721:Parameters: args[0] whether feature should be enabled or not 7722 7723Valid flags are:: 7724 7725 #define KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE (1 << 0) 7726 #define KVM_DIRTY_LOG_INITIALLY_SET (1 << 1) 7727 7728With KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE is set, KVM_GET_DIRTY_LOG will not 7729automatically clear and write-protect all pages that are returned as dirty. 7730Rather, userspace will have to do this operation separately using 7731KVM_CLEAR_DIRTY_LOG. 7732 7733At the cost of a slightly more complicated operation, this provides better 7734scalability and responsiveness for two reasons. First, 7735KVM_CLEAR_DIRTY_LOG ioctl can operate on a 64-page granularity rather 7736than requiring to sync a full memslot; this ensures that KVM does not 7737take spinlocks for an extended period of time. Second, in some cases a 7738large amount of time can pass between a call to KVM_GET_DIRTY_LOG and 7739userspace actually using the data in the page. Pages can be modified 7740during this time, which is inefficient for both the guest and userspace: 7741the guest will incur a higher penalty due to write protection faults, 7742while userspace can see false reports of dirty pages. Manual reprotection 7743helps reducing this time, improving guest performance and reducing the 7744number of dirty log false positives. 7745 7746With KVM_DIRTY_LOG_INITIALLY_SET set, all the bits of the dirty bitmap 7747will be initialized to 1 when created. This also improves performance because 7748dirty logging can be enabled gradually in small chunks on the first call 7749to KVM_CLEAR_DIRTY_LOG. KVM_DIRTY_LOG_INITIALLY_SET depends on 7750KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE (it is also only available on 7751x86 and arm64 for now). 7752 7753KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 was previously available under the name 7754KVM_CAP_MANUAL_DIRTY_LOG_PROTECT, but the implementation had bugs that make 7755it hard or impossible to use it correctly. The availability of 7756KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 signals that those bugs are fixed. 7757Userspace should not try to use KVM_CAP_MANUAL_DIRTY_LOG_PROTECT. 7758 77597.19 KVM_CAP_PPC_SECURE_GUEST 7760------------------------------ 7761 7762:Architectures: ppc 7763 7764This capability indicates that KVM is running on a host that has 7765ultravisor firmware and thus can support a secure guest. On such a 7766system, a guest can ask the ultravisor to make it a secure guest, 7767one whose memory is inaccessible to the host except for pages which 7768are explicitly requested to be shared with the host. The ultravisor 7769notifies KVM when a guest requests to become a secure guest, and KVM 7770has the opportunity to veto the transition. 7771 7772If present, this capability can be enabled for a VM, meaning that KVM 7773will allow the transition to secure guest mode. Otherwise KVM will 7774veto the transition. 7775 77767.20 KVM_CAP_HALT_POLL 7777---------------------- 7778 7779:Architectures: all 7780:Target: VM 7781:Parameters: args[0] is the maximum poll time in nanoseconds 7782:Returns: 0 on success; -1 on error 7783 7784KVM_CAP_HALT_POLL overrides the kvm.halt_poll_ns module parameter to set the 7785maximum halt-polling time for all vCPUs in the target VM. This capability can 7786be invoked at any time and any number of times to dynamically change the 7787maximum halt-polling time. 7788 7789See Documentation/virt/kvm/halt-polling.rst for more information on halt 7790polling. 7791 77927.21 KVM_CAP_X86_USER_SPACE_MSR 7793------------------------------- 7794 7795:Architectures: x86 7796:Target: VM 7797:Parameters: args[0] contains the mask of KVM_MSR_EXIT_REASON_* events to report 7798:Returns: 0 on success; -1 on error 7799 7800This capability allows userspace to intercept RDMSR and WRMSR instructions if 7801access to an MSR is denied. By default, KVM injects #GP on denied accesses. 7802 7803When a guest requests to read or write an MSR, KVM may not implement all MSRs 7804that are relevant to a respective system. It also does not differentiate by 7805CPU type. 7806 7807To allow more fine grained control over MSR handling, userspace may enable 7808this capability. With it enabled, MSR accesses that match the mask specified in 7809args[0] and would trigger a #GP inside the guest will instead trigger 7810KVM_EXIT_X86_RDMSR and KVM_EXIT_X86_WRMSR exit notifications. Userspace 7811can then implement model specific MSR handling and/or user notifications 7812to inform a user that an MSR was not emulated/virtualized by KVM. 7813 7814The valid mask flags are: 7815 7816============================ =============================================== 7817 KVM_MSR_EXIT_REASON_UNKNOWN intercept accesses to unknown (to KVM) MSRs 7818 KVM_MSR_EXIT_REASON_INVAL intercept accesses that are architecturally 7819 invalid according to the vCPU model and/or mode 7820 KVM_MSR_EXIT_REASON_FILTER intercept accesses that are denied by userspace 7821 via KVM_X86_SET_MSR_FILTER 7822============================ =============================================== 7823 78247.22 KVM_CAP_X86_BUS_LOCK_EXIT 7825------------------------------- 7826 7827:Architectures: x86 7828:Target: VM 7829:Parameters: args[0] defines the policy used when bus locks detected in guest 7830:Returns: 0 on success, -EINVAL when args[0] contains invalid bits 7831 7832Valid bits in args[0] are:: 7833 7834 #define KVM_BUS_LOCK_DETECTION_OFF (1 << 0) 7835 #define KVM_BUS_LOCK_DETECTION_EXIT (1 << 1) 7836 7837Enabling this capability on a VM provides userspace with a way to select a 7838policy to handle the bus locks detected in guest. Userspace can obtain the 7839supported modes from the result of KVM_CHECK_EXTENSION and define it through 7840the KVM_ENABLE_CAP. The supported modes are mutually-exclusive. 7841 7842This capability allows userspace to force VM exits on bus locks detected in the 7843guest, irrespective whether or not the host has enabled split-lock detection 7844(which triggers an #AC exception that KVM intercepts). This capability is 7845intended to mitigate attacks where a malicious/buggy guest can exploit bus 7846locks to degrade the performance of the whole system. 7847 7848If KVM_BUS_LOCK_DETECTION_OFF is set, KVM doesn't force guest bus locks to VM 7849exit, although the host kernel's split-lock #AC detection still applies, if 7850enabled. 7851 7852If KVM_BUS_LOCK_DETECTION_EXIT is set, KVM enables a CPU feature that ensures 7853bus locks in the guest trigger a VM exit, and KVM exits to userspace for all 7854such VM exits, e.g. to allow userspace to throttle the offending guest and/or 7855apply some other policy-based mitigation. When exiting to userspace, KVM sets 7856KVM_RUN_X86_BUS_LOCK in vcpu-run->flags, and conditionally sets the exit_reason 7857to KVM_EXIT_X86_BUS_LOCK. 7858 7859Note! Detected bus locks may be coincident with other exits to userspace, i.e. 7860KVM_RUN_X86_BUS_LOCK should be checked regardless of the primary exit reason if 7861userspace wants to take action on all detected bus locks. 7862 78637.23 KVM_CAP_PPC_DAWR1 7864---------------------- 7865 7866:Architectures: ppc 7867:Parameters: none 7868:Returns: 0 on success, -EINVAL when CPU doesn't support 2nd DAWR 7869 7870This capability can be used to check / enable 2nd DAWR feature provided 7871by POWER10 processor. 7872 7873 78747.24 KVM_CAP_VM_COPY_ENC_CONTEXT_FROM 7875------------------------------------- 7876 7877Architectures: x86 SEV enabled 7878Type: vm 7879Parameters: args[0] is the fd of the source vm 7880Returns: 0 on success; ENOTTY on error 7881 7882This capability enables userspace to copy encryption context from the vm 7883indicated by the fd to the vm this is called on. 7884 7885This is intended to support in-guest workloads scheduled by the host. This 7886allows the in-guest workload to maintain its own NPTs and keeps the two vms 7887from accidentally clobbering each other with interrupts and the like (separate 7888APIC/MSRs/etc). 7889 78907.25 KVM_CAP_SGX_ATTRIBUTE 7891-------------------------- 7892 7893:Architectures: x86 7894:Target: VM 7895:Parameters: args[0] is a file handle of a SGX attribute file in securityfs 7896:Returns: 0 on success, -EINVAL if the file handle is invalid or if a requested 7897 attribute is not supported by KVM. 7898 7899KVM_CAP_SGX_ATTRIBUTE enables a userspace VMM to grant a VM access to one or 7900more privileged enclave attributes. args[0] must hold a file handle to a valid 7901SGX attribute file corresponding to an attribute that is supported/restricted 7902by KVM (currently only PROVISIONKEY). 7903 7904The SGX subsystem restricts access to a subset of enclave attributes to provide 7905additional security for an uncompromised kernel, e.g. use of the PROVISIONKEY 7906is restricted to deter malware from using the PROVISIONKEY to obtain a stable 7907system fingerprint. To prevent userspace from circumventing such restrictions 7908by running an enclave in a VM, KVM prevents access to privileged attributes by 7909default. 7910 7911See Documentation/arch/x86/sgx.rst for more details. 7912 79137.26 KVM_CAP_PPC_RPT_INVALIDATE 7914------------------------------- 7915 7916:Capability: KVM_CAP_PPC_RPT_INVALIDATE 7917:Architectures: ppc 7918:Type: vm 7919 7920This capability indicates that the kernel is capable of handling 7921H_RPT_INVALIDATE hcall. 7922 7923In order to enable the use of H_RPT_INVALIDATE in the guest, 7924user space might have to advertise it for the guest. For example, 7925IBM pSeries (sPAPR) guest starts using it if "hcall-rpt-invalidate" is 7926present in the "ibm,hypertas-functions" device-tree property. 7927 7928This capability is enabled for hypervisors on platforms like POWER9 7929that support radix MMU. 7930 79317.27 KVM_CAP_EXIT_ON_EMULATION_FAILURE 7932-------------------------------------- 7933 7934:Architectures: x86 7935:Parameters: args[0] whether the feature should be enabled or not 7936 7937When this capability is enabled, an emulation failure will result in an exit 7938to userspace with KVM_INTERNAL_ERROR (except when the emulator was invoked 7939to handle a VMware backdoor instruction). Furthermore, KVM will now provide up 7940to 15 instruction bytes for any exit to userspace resulting from an emulation 7941failure. When these exits to userspace occur use the emulation_failure struct 7942instead of the internal struct. They both have the same layout, but the 7943emulation_failure struct matches the content better. It also explicitly 7944defines the 'flags' field which is used to describe the fields in the struct 7945that are valid (ie: if KVM_INTERNAL_ERROR_EMULATION_FLAG_INSTRUCTION_BYTES is 7946set in the 'flags' field then both 'insn_size' and 'insn_bytes' have valid data 7947in them.) 7948 79497.28 KVM_CAP_ARM_MTE 7950-------------------- 7951 7952:Architectures: arm64 7953:Parameters: none 7954 7955This capability indicates that KVM (and the hardware) supports exposing the 7956Memory Tagging Extensions (MTE) to the guest. It must also be enabled by the 7957VMM before creating any VCPUs to allow the guest access. Note that MTE is only 7958available to a guest running in AArch64 mode and enabling this capability will 7959cause attempts to create AArch32 VCPUs to fail. 7960 7961When enabled the guest is able to access tags associated with any memory given 7962to the guest. KVM will ensure that the tags are maintained during swap or 7963hibernation of the host; however the VMM needs to manually save/restore the 7964tags as appropriate if the VM is migrated. 7965 7966When this capability is enabled all memory in memslots must be mapped as 7967``MAP_ANONYMOUS`` or with a RAM-based file mapping (``tmpfs``, ``memfd``), 7968attempts to create a memslot with an invalid mmap will result in an 7969-EINVAL return. 7970 7971When enabled the VMM may make use of the ``KVM_ARM_MTE_COPY_TAGS`` ioctl to 7972perform a bulk copy of tags to/from the guest. 7973 79747.29 KVM_CAP_VM_MOVE_ENC_CONTEXT_FROM 7975------------------------------------- 7976 7977:Architectures: x86 SEV enabled 7978:Type: vm 7979:Parameters: args[0] is the fd of the source vm 7980:Returns: 0 on success 7981 7982This capability enables userspace to migrate the encryption context from the VM 7983indicated by the fd to the VM this is called on. 7984 7985This is intended to support intra-host migration of VMs between userspace VMMs, 7986upgrading the VMM process without interrupting the guest. 7987 79887.30 KVM_CAP_PPC_AIL_MODE_3 7989------------------------------- 7990 7991:Capability: KVM_CAP_PPC_AIL_MODE_3 7992:Architectures: ppc 7993:Type: vm 7994 7995This capability indicates that the kernel supports the mode 3 setting for the 7996"Address Translation Mode on Interrupt" aka "Alternate Interrupt Location" 7997resource that is controlled with the H_SET_MODE hypercall. 7998 7999This capability allows a guest kernel to use a better-performance mode for 8000handling interrupts and system calls. 8001 80027.31 KVM_CAP_DISABLE_QUIRKS2 8003---------------------------- 8004 8005:Capability: KVM_CAP_DISABLE_QUIRKS2 8006:Parameters: args[0] - set of KVM quirks to disable 8007:Architectures: x86 8008:Type: vm 8009 8010This capability, if enabled, will cause KVM to disable some behavior 8011quirks. 8012 8013Calling KVM_CHECK_EXTENSION for this capability returns a bitmask of 8014quirks that can be disabled in KVM. 8015 8016The argument to KVM_ENABLE_CAP for this capability is a bitmask of 8017quirks to disable, and must be a subset of the bitmask returned by 8018KVM_CHECK_EXTENSION. 8019 8020The valid bits in cap.args[0] are: 8021 8022=================================== ============================================ 8023 KVM_X86_QUIRK_LINT0_REENABLED By default, the reset value for the LVT 8024 LINT0 register is 0x700 (APIC_MODE_EXTINT). 8025 When this quirk is disabled, the reset value 8026 is 0x10000 (APIC_LVT_MASKED). 8027 8028 KVM_X86_QUIRK_CD_NW_CLEARED By default, KVM clears CR0.CD and CR0.NW on 8029 AMD CPUs to workaround buggy guest firmware 8030 that runs in perpetuity with CR0.CD, i.e. 8031 with caches in "no fill" mode. 8032 8033 When this quirk is disabled, KVM does not 8034 change the value of CR0.CD and CR0.NW. 8035 8036 KVM_X86_QUIRK_LAPIC_MMIO_HOLE By default, the MMIO LAPIC interface is 8037 available even when configured for x2APIC 8038 mode. When this quirk is disabled, KVM 8039 disables the MMIO LAPIC interface if the 8040 LAPIC is in x2APIC mode. 8041 8042 KVM_X86_QUIRK_OUT_7E_INC_RIP By default, KVM pre-increments %rip before 8043 exiting to userspace for an OUT instruction 8044 to port 0x7e. When this quirk is disabled, 8045 KVM does not pre-increment %rip before 8046 exiting to userspace. 8047 8048 KVM_X86_QUIRK_MISC_ENABLE_NO_MWAIT When this quirk is disabled, KVM sets 8049 CPUID.01H:ECX[bit 3] (MONITOR/MWAIT) if 8050 IA32_MISC_ENABLE[bit 18] (MWAIT) is set. 8051 Additionally, when this quirk is disabled, 8052 KVM clears CPUID.01H:ECX[bit 3] if 8053 IA32_MISC_ENABLE[bit 18] is cleared. 8054 8055 KVM_X86_QUIRK_FIX_HYPERCALL_INSN By default, KVM rewrites guest 8056 VMMCALL/VMCALL instructions to match the 8057 vendor's hypercall instruction for the 8058 system. When this quirk is disabled, KVM 8059 will no longer rewrite invalid guest 8060 hypercall instructions. Executing the 8061 incorrect hypercall instruction will 8062 generate a #UD within the guest. 8063 8064KVM_X86_QUIRK_MWAIT_NEVER_UD_FAULTS By default, KVM emulates MONITOR/MWAIT (if 8065 they are intercepted) as NOPs regardless of 8066 whether or not MONITOR/MWAIT are supported 8067 according to guest CPUID. When this quirk 8068 is disabled and KVM_X86_DISABLE_EXITS_MWAIT 8069 is not set (MONITOR/MWAIT are intercepted), 8070 KVM will inject a #UD on MONITOR/MWAIT if 8071 they're unsupported per guest CPUID. Note, 8072 KVM will modify MONITOR/MWAIT support in 8073 guest CPUID on writes to MISC_ENABLE if 8074 KVM_X86_QUIRK_MISC_ENABLE_NO_MWAIT is 8075 disabled. 8076=================================== ============================================ 8077 80787.32 KVM_CAP_MAX_VCPU_ID 8079------------------------ 8080 8081:Architectures: x86 8082:Target: VM 8083:Parameters: args[0] - maximum APIC ID value set for current VM 8084:Returns: 0 on success, -EINVAL if args[0] is beyond KVM_MAX_VCPU_IDS 8085 supported in KVM or if it has been set. 8086 8087This capability allows userspace to specify maximum possible APIC ID 8088assigned for current VM session prior to the creation of vCPUs, saving 8089memory for data structures indexed by the APIC ID. Userspace is able 8090to calculate the limit to APIC ID values from designated 8091CPU topology. 8092 8093The value can be changed only until KVM_ENABLE_CAP is set to a nonzero 8094value or until a vCPU is created. Upon creation of the first vCPU, 8095if the value was set to zero or KVM_ENABLE_CAP was not invoked, KVM 8096uses the return value of KVM_CHECK_EXTENSION(KVM_CAP_MAX_VCPU_ID) as 8097the maximum APIC ID. 8098 80997.33 KVM_CAP_X86_NOTIFY_VMEXIT 8100------------------------------ 8101 8102:Architectures: x86 8103:Target: VM 8104:Parameters: args[0] is the value of notify window as well as some flags 8105:Returns: 0 on success, -EINVAL if args[0] contains invalid flags or notify 8106 VM exit is unsupported. 8107 8108Bits 63:32 of args[0] are used for notify window. 8109Bits 31:0 of args[0] are for some flags. Valid bits are:: 8110 8111 #define KVM_X86_NOTIFY_VMEXIT_ENABLED (1 << 0) 8112 #define KVM_X86_NOTIFY_VMEXIT_USER (1 << 1) 8113 8114This capability allows userspace to configure the notify VM exit on/off 8115in per-VM scope during VM creation. Notify VM exit is disabled by default. 8116When userspace sets KVM_X86_NOTIFY_VMEXIT_ENABLED bit in args[0], VMM will 8117enable this feature with the notify window provided, which will generate 8118a VM exit if no event window occurs in VM non-root mode for a specified of 8119time (notify window). 8120 8121If KVM_X86_NOTIFY_VMEXIT_USER is set in args[0], upon notify VM exits happen, 8122KVM would exit to userspace for handling. 8123 8124This capability is aimed to mitigate the threat that malicious VMs can 8125cause CPU stuck (due to event windows don't open up) and make the CPU 8126unavailable to host or other VMs. 8127 81287.34 KVM_CAP_MEMORY_FAULT_INFO 8129------------------------------ 8130 8131:Architectures: x86 8132:Returns: Informational only, -EINVAL on direct KVM_ENABLE_CAP. 8133 8134The presence of this capability indicates that KVM_RUN will fill 8135kvm_run.memory_fault if KVM cannot resolve a guest page fault VM-Exit, e.g. if 8136there is a valid memslot but no backing VMA for the corresponding host virtual 8137address. 8138 8139The information in kvm_run.memory_fault is valid if and only if KVM_RUN returns 8140an error with errno=EFAULT or errno=EHWPOISON *and* kvm_run.exit_reason is set 8141to KVM_EXIT_MEMORY_FAULT. 8142 8143Note: Userspaces which attempt to resolve memory faults so that they can retry 8144KVM_RUN are encouraged to guard against repeatedly receiving the same 8145error/annotated fault. 8146 8147See KVM_EXIT_MEMORY_FAULT for more information. 8148 81497.35 KVM_CAP_X86_APIC_BUS_CYCLES_NS 8150----------------------------------- 8151 8152:Architectures: x86 8153:Target: VM 8154:Parameters: args[0] is the desired APIC bus clock rate, in nanoseconds 8155:Returns: 0 on success, -EINVAL if args[0] contains an invalid value for the 8156 frequency or if any vCPUs have been created, -ENXIO if a virtual 8157 local APIC has not been created using KVM_CREATE_IRQCHIP. 8158 8159This capability sets the VM's APIC bus clock frequency, used by KVM's in-kernel 8160virtual APIC when emulating APIC timers. KVM's default value can be retrieved 8161by KVM_CHECK_EXTENSION. 8162 8163Note: Userspace is responsible for correctly configuring CPUID 0x15, a.k.a. the 8164core crystal clock frequency, if a non-zero CPUID 0x15 is exposed to the guest. 8165 81667.36 KVM_CAP_X86_GUEST_MODE 8167------------------------------ 8168 8169:Architectures: x86 8170:Returns: Informational only, -EINVAL on direct KVM_ENABLE_CAP. 8171 8172The presence of this capability indicates that KVM_RUN will update the 8173KVM_RUN_X86_GUEST_MODE bit in kvm_run.flags to indicate whether the 8174vCPU was executing nested guest code when it exited. 8175 8176KVM exits with the register state of either the L1 or L2 guest 8177depending on which executed at the time of an exit. Userspace must 8178take care to differentiate between these cases. 8179 81808. Other capabilities. 8181====================== 8182 8183This section lists capabilities that give information about other 8184features of the KVM implementation. 8185 81868.1 KVM_CAP_PPC_HWRNG 8187--------------------- 8188 8189:Architectures: ppc 8190 8191This capability, if KVM_CHECK_EXTENSION indicates that it is 8192available, means that the kernel has an implementation of the 8193H_RANDOM hypercall backed by a hardware random-number generator. 8194If present, the kernel H_RANDOM handler can be enabled for guest use 8195with the KVM_CAP_PPC_ENABLE_HCALL capability. 8196 81978.2 KVM_CAP_HYPERV_SYNIC 8198------------------------ 8199 8200:Architectures: x86 8201 8202This capability, if KVM_CHECK_EXTENSION indicates that it is 8203available, means that the kernel has an implementation of the 8204Hyper-V Synthetic interrupt controller(SynIC). Hyper-V SynIC is 8205used to support Windows Hyper-V based guest paravirt drivers(VMBus). 8206 8207In order to use SynIC, it has to be activated by setting this 8208capability via KVM_ENABLE_CAP ioctl on the vcpu fd. Note that this 8209will disable the use of APIC hardware virtualization even if supported 8210by the CPU, as it's incompatible with SynIC auto-EOI behavior. 8211 82128.3 KVM_CAP_PPC_MMU_RADIX 8213------------------------- 8214 8215:Architectures: ppc 8216 8217This capability, if KVM_CHECK_EXTENSION indicates that it is 8218available, means that the kernel can support guests using the 8219radix MMU defined in Power ISA V3.00 (as implemented in the POWER9 8220processor). 8221 82228.4 KVM_CAP_PPC_MMU_HASH_V3 8223--------------------------- 8224 8225:Architectures: ppc 8226 8227This capability, if KVM_CHECK_EXTENSION indicates that it is 8228available, means that the kernel can support guests using the 8229hashed page table MMU defined in Power ISA V3.00 (as implemented in 8230the POWER9 processor), including in-memory segment tables. 8231 82328.5 KVM_CAP_MIPS_VZ 8233------------------- 8234 8235:Architectures: mips 8236 8237This capability, if KVM_CHECK_EXTENSION on the main kvm handle indicates that 8238it is available, means that full hardware assisted virtualization capabilities 8239of the hardware are available for use through KVM. An appropriate 8240KVM_VM_MIPS_* type must be passed to KVM_CREATE_VM to create a VM which 8241utilises it. 8242 8243If KVM_CHECK_EXTENSION on a kvm VM handle indicates that this capability is 8244available, it means that the VM is using full hardware assisted virtualization 8245capabilities of the hardware. This is useful to check after creating a VM with 8246KVM_VM_MIPS_DEFAULT. 8247 8248The value returned by KVM_CHECK_EXTENSION should be compared against known 8249values (see below). All other values are reserved. This is to allow for the 8250possibility of other hardware assisted virtualization implementations which 8251may be incompatible with the MIPS VZ ASE. 8252 8253== ========================================================================== 8254 0 The trap & emulate implementation is in use to run guest code in user 8255 mode. Guest virtual memory segments are rearranged to fit the guest in the 8256 user mode address space. 8257 8258 1 The MIPS VZ ASE is in use, providing full hardware assisted 8259 virtualization, including standard guest virtual memory segments. 8260== ========================================================================== 8261 82628.6 KVM_CAP_MIPS_TE 8263------------------- 8264 8265:Architectures: mips 8266 8267This capability, if KVM_CHECK_EXTENSION on the main kvm handle indicates that 8268it is available, means that the trap & emulate implementation is available to 8269run guest code in user mode, even if KVM_CAP_MIPS_VZ indicates that hardware 8270assisted virtualisation is also available. KVM_VM_MIPS_TE (0) must be passed 8271to KVM_CREATE_VM to create a VM which utilises it. 8272 8273If KVM_CHECK_EXTENSION on a kvm VM handle indicates that this capability is 8274available, it means that the VM is using trap & emulate. 8275 82768.7 KVM_CAP_MIPS_64BIT 8277---------------------- 8278 8279:Architectures: mips 8280 8281This capability indicates the supported architecture type of the guest, i.e. the 8282supported register and address width. 8283 8284The values returned when this capability is checked by KVM_CHECK_EXTENSION on a 8285kvm VM handle correspond roughly to the CP0_Config.AT register field, and should 8286be checked specifically against known values (see below). All other values are 8287reserved. 8288 8289== ======================================================================== 8290 0 MIPS32 or microMIPS32. 8291 Both registers and addresses are 32-bits wide. 8292 It will only be possible to run 32-bit guest code. 8293 8294 1 MIPS64 or microMIPS64 with access only to 32-bit compatibility segments. 8295 Registers are 64-bits wide, but addresses are 32-bits wide. 8296 64-bit guest code may run but cannot access MIPS64 memory segments. 8297 It will also be possible to run 32-bit guest code. 8298 8299 2 MIPS64 or microMIPS64 with access to all address segments. 8300 Both registers and addresses are 64-bits wide. 8301 It will be possible to run 64-bit or 32-bit guest code. 8302== ======================================================================== 8303 83048.9 KVM_CAP_ARM_USER_IRQ 8305------------------------ 8306 8307:Architectures: arm64 8308 8309This capability, if KVM_CHECK_EXTENSION indicates that it is available, means 8310that if userspace creates a VM without an in-kernel interrupt controller, it 8311will be notified of changes to the output level of in-kernel emulated devices, 8312which can generate virtual interrupts, presented to the VM. 8313For such VMs, on every return to userspace, the kernel 8314updates the vcpu's run->s.regs.device_irq_level field to represent the actual 8315output level of the device. 8316 8317Whenever kvm detects a change in the device output level, kvm guarantees at 8318least one return to userspace before running the VM. This exit could either 8319be a KVM_EXIT_INTR or any other exit event, like KVM_EXIT_MMIO. This way, 8320userspace can always sample the device output level and re-compute the state of 8321the userspace interrupt controller. Userspace should always check the state 8322of run->s.regs.device_irq_level on every kvm exit. 8323The value in run->s.regs.device_irq_level can represent both level and edge 8324triggered interrupt signals, depending on the device. Edge triggered interrupt 8325signals will exit to userspace with the bit in run->s.regs.device_irq_level 8326set exactly once per edge signal. 8327 8328The field run->s.regs.device_irq_level is available independent of 8329run->kvm_valid_regs or run->kvm_dirty_regs bits. 8330 8331If KVM_CAP_ARM_USER_IRQ is supported, the KVM_CHECK_EXTENSION ioctl returns a 8332number larger than 0 indicating the version of this capability is implemented 8333and thereby which bits in run->s.regs.device_irq_level can signal values. 8334 8335Currently the following bits are defined for the device_irq_level bitmap:: 8336 8337 KVM_CAP_ARM_USER_IRQ >= 1: 8338 8339 KVM_ARM_DEV_EL1_VTIMER - EL1 virtual timer 8340 KVM_ARM_DEV_EL1_PTIMER - EL1 physical timer 8341 KVM_ARM_DEV_PMU - ARM PMU overflow interrupt signal 8342 8343Future versions of kvm may implement additional events. These will get 8344indicated by returning a higher number from KVM_CHECK_EXTENSION and will be 8345listed above. 8346 83478.10 KVM_CAP_PPC_SMT_POSSIBLE 8348----------------------------- 8349 8350:Architectures: ppc 8351 8352Querying this capability returns a bitmap indicating the possible 8353virtual SMT modes that can be set using KVM_CAP_PPC_SMT. If bit N 8354(counting from the right) is set, then a virtual SMT mode of 2^N is 8355available. 8356 83578.11 KVM_CAP_HYPERV_SYNIC2 8358-------------------------- 8359 8360:Architectures: x86 8361 8362This capability enables a newer version of Hyper-V Synthetic interrupt 8363controller (SynIC). The only difference with KVM_CAP_HYPERV_SYNIC is that KVM 8364doesn't clear SynIC message and event flags pages when they are enabled by 8365writing to the respective MSRs. 8366 83678.12 KVM_CAP_HYPERV_VP_INDEX 8368---------------------------- 8369 8370:Architectures: x86 8371 8372This capability indicates that userspace can load HV_X64_MSR_VP_INDEX msr. Its 8373value is used to denote the target vcpu for a SynIC interrupt. For 8374compatibility, KVM initializes this msr to KVM's internal vcpu index. When this 8375capability is absent, userspace can still query this msr's value. 8376 83778.13 KVM_CAP_S390_AIS_MIGRATION 8378------------------------------- 8379 8380:Architectures: s390 8381:Parameters: none 8382 8383This capability indicates if the flic device will be able to get/set the 8384AIS states for migration via the KVM_DEV_FLIC_AISM_ALL attribute and allows 8385to discover this without having to create a flic device. 8386 83878.14 KVM_CAP_S390_PSW 8388--------------------- 8389 8390:Architectures: s390 8391 8392This capability indicates that the PSW is exposed via the kvm_run structure. 8393 83948.15 KVM_CAP_S390_GMAP 8395---------------------- 8396 8397:Architectures: s390 8398 8399This capability indicates that the user space memory used as guest mapping can 8400be anywhere in the user memory address space, as long as the memory slots are 8401aligned and sized to a segment (1MB) boundary. 8402 84038.16 KVM_CAP_S390_COW 8404--------------------- 8405 8406:Architectures: s390 8407 8408This capability indicates that the user space memory used as guest mapping can 8409use copy-on-write semantics as well as dirty pages tracking via read-only page 8410tables. 8411 84128.17 KVM_CAP_S390_BPB 8413--------------------- 8414 8415:Architectures: s390 8416 8417This capability indicates that kvm will implement the interfaces to handle 8418reset, migration and nested KVM for branch prediction blocking. The stfle 8419facility 82 should not be provided to the guest without this capability. 8420 84218.18 KVM_CAP_HYPERV_TLBFLUSH 8422---------------------------- 8423 8424:Architectures: x86 8425 8426This capability indicates that KVM supports paravirtualized Hyper-V TLB Flush 8427hypercalls: 8428HvFlushVirtualAddressSpace, HvFlushVirtualAddressSpaceEx, 8429HvFlushVirtualAddressList, HvFlushVirtualAddressListEx. 8430 84318.19 KVM_CAP_ARM_INJECT_SERROR_ESR 8432---------------------------------- 8433 8434:Architectures: arm64 8435 8436This capability indicates that userspace can specify (via the 8437KVM_SET_VCPU_EVENTS ioctl) the syndrome value reported to the guest when it 8438takes a virtual SError interrupt exception. 8439If KVM advertises this capability, userspace can only specify the ISS field for 8440the ESR syndrome. Other parts of the ESR, such as the EC are generated by the 8441CPU when the exception is taken. If this virtual SError is taken to EL1 using 8442AArch64, this value will be reported in the ISS field of ESR_ELx. 8443 8444See KVM_CAP_VCPU_EVENTS for more details. 8445 84468.20 KVM_CAP_HYPERV_SEND_IPI 8447---------------------------- 8448 8449:Architectures: x86 8450 8451This capability indicates that KVM supports paravirtualized Hyper-V IPI send 8452hypercalls: 8453HvCallSendSyntheticClusterIpi, HvCallSendSyntheticClusterIpiEx. 8454 84558.21 KVM_CAP_HYPERV_DIRECT_TLBFLUSH 8456----------------------------------- 8457 8458:Architectures: x86 8459 8460This capability indicates that KVM running on top of Hyper-V hypervisor 8461enables Direct TLB flush for its guests meaning that TLB flush 8462hypercalls are handled by Level 0 hypervisor (Hyper-V) bypassing KVM. 8463Due to the different ABI for hypercall parameters between Hyper-V and 8464KVM, enabling this capability effectively disables all hypercall 8465handling by KVM (as some KVM hypercall may be mistakenly treated as TLB 8466flush hypercalls by Hyper-V) so userspace should disable KVM identification 8467in CPUID and only exposes Hyper-V identification. In this case, guest 8468thinks it's running on Hyper-V and only use Hyper-V hypercalls. 8469 84708.22 KVM_CAP_S390_VCPU_RESETS 8471----------------------------- 8472 8473:Architectures: s390 8474 8475This capability indicates that the KVM_S390_NORMAL_RESET and 8476KVM_S390_CLEAR_RESET ioctls are available. 8477 84788.23 KVM_CAP_S390_PROTECTED 8479--------------------------- 8480 8481:Architectures: s390 8482 8483This capability indicates that the Ultravisor has been initialized and 8484KVM can therefore start protected VMs. 8485This capability governs the KVM_S390_PV_COMMAND ioctl and the 8486KVM_MP_STATE_LOAD MP_STATE. KVM_SET_MP_STATE can fail for protected 8487guests when the state change is invalid. 8488 84898.24 KVM_CAP_STEAL_TIME 8490----------------------- 8491 8492:Architectures: arm64, x86 8493 8494This capability indicates that KVM supports steal time accounting. 8495When steal time accounting is supported it may be enabled with 8496architecture-specific interfaces. This capability and the architecture- 8497specific interfaces must be consistent, i.e. if one says the feature 8498is supported, than the other should as well and vice versa. For arm64 8499see Documentation/virt/kvm/devices/vcpu.rst "KVM_ARM_VCPU_PVTIME_CTRL". 8500For x86 see Documentation/virt/kvm/x86/msr.rst "MSR_KVM_STEAL_TIME". 8501 85028.25 KVM_CAP_S390_DIAG318 8503------------------------- 8504 8505:Architectures: s390 8506 8507This capability enables a guest to set information about its control program 8508(i.e. guest kernel type and version). The information is helpful during 8509system/firmware service events, providing additional data about the guest 8510environments running on the machine. 8511 8512The information is associated with the DIAGNOSE 0x318 instruction, which sets 8513an 8-byte value consisting of a one-byte Control Program Name Code (CPNC) and 8514a 7-byte Control Program Version Code (CPVC). The CPNC determines what 8515environment the control program is running in (e.g. Linux, z/VM...), and the 8516CPVC is used for information specific to OS (e.g. Linux version, Linux 8517distribution...) 8518 8519If this capability is available, then the CPNC and CPVC can be synchronized 8520between KVM and userspace via the sync regs mechanism (KVM_SYNC_DIAG318). 8521 85228.26 KVM_CAP_X86_USER_SPACE_MSR 8523------------------------------- 8524 8525:Architectures: x86 8526 8527This capability indicates that KVM supports deflection of MSR reads and 8528writes to user space. It can be enabled on a VM level. If enabled, MSR 8529accesses that would usually trigger a #GP by KVM into the guest will 8530instead get bounced to user space through the KVM_EXIT_X86_RDMSR and 8531KVM_EXIT_X86_WRMSR exit notifications. 8532 85338.27 KVM_CAP_X86_MSR_FILTER 8534--------------------------- 8535 8536:Architectures: x86 8537 8538This capability indicates that KVM supports that accesses to user defined MSRs 8539may be rejected. With this capability exposed, KVM exports new VM ioctl 8540KVM_X86_SET_MSR_FILTER which user space can call to specify bitmaps of MSR 8541ranges that KVM should deny access to. 8542 8543In combination with KVM_CAP_X86_USER_SPACE_MSR, this allows user space to 8544trap and emulate MSRs that are outside of the scope of KVM as well as 8545limit the attack surface on KVM's MSR emulation code. 8546 85478.28 KVM_CAP_ENFORCE_PV_FEATURE_CPUID 8548------------------------------------- 8549 8550Architectures: x86 8551 8552When enabled, KVM will disable paravirtual features provided to the 8553guest according to the bits in the KVM_CPUID_FEATURES CPUID leaf 8554(0x40000001). Otherwise, a guest may use the paravirtual features 8555regardless of what has actually been exposed through the CPUID leaf. 8556 85578.29 KVM_CAP_DIRTY_LOG_RING/KVM_CAP_DIRTY_LOG_RING_ACQ_REL 8558---------------------------------------------------------- 8559 8560:Architectures: x86, arm64 8561:Parameters: args[0] - size of the dirty log ring 8562 8563KVM is capable of tracking dirty memory using ring buffers that are 8564mmapped into userspace; there is one dirty ring per vcpu. 8565 8566The dirty ring is available to userspace as an array of 8567``struct kvm_dirty_gfn``. Each dirty entry is defined as:: 8568 8569 struct kvm_dirty_gfn { 8570 __u32 flags; 8571 __u32 slot; /* as_id | slot_id */ 8572 __u64 offset; 8573 }; 8574 8575The following values are defined for the flags field to define the 8576current state of the entry:: 8577 8578 #define KVM_DIRTY_GFN_F_DIRTY BIT(0) 8579 #define KVM_DIRTY_GFN_F_RESET BIT(1) 8580 #define KVM_DIRTY_GFN_F_MASK 0x3 8581 8582Userspace should call KVM_ENABLE_CAP ioctl right after KVM_CREATE_VM 8583ioctl to enable this capability for the new guest and set the size of 8584the rings. Enabling the capability is only allowed before creating any 8585vCPU, and the size of the ring must be a power of two. The larger the 8586ring buffer, the less likely the ring is full and the VM is forced to 8587exit to userspace. The optimal size depends on the workload, but it is 8588recommended that it be at least 64 KiB (4096 entries). 8589 8590Just like for dirty page bitmaps, the buffer tracks writes to 8591all user memory regions for which the KVM_MEM_LOG_DIRTY_PAGES flag was 8592set in KVM_SET_USER_MEMORY_REGION. Once a memory region is registered 8593with the flag set, userspace can start harvesting dirty pages from the 8594ring buffer. 8595 8596An entry in the ring buffer can be unused (flag bits ``00``), 8597dirty (flag bits ``01``) or harvested (flag bits ``1X``). The 8598state machine for the entry is as follows:: 8599 8600 dirtied harvested reset 8601 00 -----------> 01 -------------> 1X -------+ 8602 ^ | 8603 | | 8604 +------------------------------------------+ 8605 8606To harvest the dirty pages, userspace accesses the mmapped ring buffer 8607to read the dirty GFNs. If the flags has the DIRTY bit set (at this stage 8608the RESET bit must be cleared), then it means this GFN is a dirty GFN. 8609The userspace should harvest this GFN and mark the flags from state 8610``01b`` to ``1Xb`` (bit 0 will be ignored by KVM, but bit 1 must be set 8611to show that this GFN is harvested and waiting for a reset), and move 8612on to the next GFN. The userspace should continue to do this until the 8613flags of a GFN have the DIRTY bit cleared, meaning that it has harvested 8614all the dirty GFNs that were available. 8615 8616Note that on weakly ordered architectures, userspace accesses to the 8617ring buffer (and more specifically the 'flags' field) must be ordered, 8618using load-acquire/store-release accessors when available, or any 8619other memory barrier that will ensure this ordering. 8620 8621It's not necessary for userspace to harvest the all dirty GFNs at once. 8622However it must collect the dirty GFNs in sequence, i.e., the userspace 8623program cannot skip one dirty GFN to collect the one next to it. 8624 8625After processing one or more entries in the ring buffer, userspace 8626calls the VM ioctl KVM_RESET_DIRTY_RINGS to notify the kernel about 8627it, so that the kernel will reprotect those collected GFNs. 8628Therefore, the ioctl must be called *before* reading the content of 8629the dirty pages. 8630 8631The dirty ring can get full. When it happens, the KVM_RUN of the 8632vcpu will return with exit reason KVM_EXIT_DIRTY_LOG_FULL. 8633 8634The dirty ring interface has a major difference comparing to the 8635KVM_GET_DIRTY_LOG interface in that, when reading the dirty ring from 8636userspace, it's still possible that the kernel has not yet flushed the 8637processor's dirty page buffers into the kernel buffer (with dirty bitmaps, the 8638flushing is done by the KVM_GET_DIRTY_LOG ioctl). To achieve that, one 8639needs to kick the vcpu out of KVM_RUN using a signal. The resulting 8640vmexit ensures that all dirty GFNs are flushed to the dirty rings. 8641 8642NOTE: KVM_CAP_DIRTY_LOG_RING_ACQ_REL is the only capability that 8643should be exposed by weakly ordered architecture, in order to indicate 8644the additional memory ordering requirements imposed on userspace when 8645reading the state of an entry and mutating it from DIRTY to HARVESTED. 8646Architecture with TSO-like ordering (such as x86) are allowed to 8647expose both KVM_CAP_DIRTY_LOG_RING and KVM_CAP_DIRTY_LOG_RING_ACQ_REL 8648to userspace. 8649 8650After enabling the dirty rings, the userspace needs to detect the 8651capability of KVM_CAP_DIRTY_LOG_RING_WITH_BITMAP to see whether the 8652ring structures can be backed by per-slot bitmaps. With this capability 8653advertised, it means the architecture can dirty guest pages without 8654vcpu/ring context, so that some of the dirty information will still be 8655maintained in the bitmap structure. KVM_CAP_DIRTY_LOG_RING_WITH_BITMAP 8656can't be enabled if the capability of KVM_CAP_DIRTY_LOG_RING_ACQ_REL 8657hasn't been enabled, or any memslot has been existing. 8658 8659Note that the bitmap here is only a backup of the ring structure. The 8660use of the ring and bitmap combination is only beneficial if there is 8661only a very small amount of memory that is dirtied out of vcpu/ring 8662context. Otherwise, the stand-alone per-slot bitmap mechanism needs to 8663be considered. 8664 8665To collect dirty bits in the backup bitmap, userspace can use the same 8666KVM_GET_DIRTY_LOG ioctl. KVM_CLEAR_DIRTY_LOG isn't needed as long as all 8667the generation of the dirty bits is done in a single pass. Collecting 8668the dirty bitmap should be the very last thing that the VMM does before 8669considering the state as complete. VMM needs to ensure that the dirty 8670state is final and avoid missing dirty pages from another ioctl ordered 8671after the bitmap collection. 8672 8673NOTE: Multiple examples of using the backup bitmap: (1) save vgic/its 8674tables through command KVM_DEV_ARM_{VGIC_GRP_CTRL, ITS_SAVE_TABLES} on 8675KVM device "kvm-arm-vgic-its". (2) restore vgic/its tables through 8676command KVM_DEV_ARM_{VGIC_GRP_CTRL, ITS_RESTORE_TABLES} on KVM device 8677"kvm-arm-vgic-its". VGICv3 LPI pending status is restored. (3) save 8678vgic3 pending table through KVM_DEV_ARM_VGIC_{GRP_CTRL, SAVE_PENDING_TABLES} 8679command on KVM device "kvm-arm-vgic-v3". 8680 86818.30 KVM_CAP_XEN_HVM 8682-------------------- 8683 8684:Architectures: x86 8685 8686This capability indicates the features that Xen supports for hosting Xen 8687PVHVM guests. Valid flags are:: 8688 8689 #define KVM_XEN_HVM_CONFIG_HYPERCALL_MSR (1 << 0) 8690 #define KVM_XEN_HVM_CONFIG_INTERCEPT_HCALL (1 << 1) 8691 #define KVM_XEN_HVM_CONFIG_SHARED_INFO (1 << 2) 8692 #define KVM_XEN_HVM_CONFIG_RUNSTATE (1 << 3) 8693 #define KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL (1 << 4) 8694 #define KVM_XEN_HVM_CONFIG_EVTCHN_SEND (1 << 5) 8695 #define KVM_XEN_HVM_CONFIG_RUNSTATE_UPDATE_FLAG (1 << 6) 8696 #define KVM_XEN_HVM_CONFIG_PVCLOCK_TSC_UNSTABLE (1 << 7) 8697 8698The KVM_XEN_HVM_CONFIG_HYPERCALL_MSR flag indicates that the KVM_XEN_HVM_CONFIG 8699ioctl is available, for the guest to set its hypercall page. 8700 8701If KVM_XEN_HVM_CONFIG_INTERCEPT_HCALL is also set, the same flag may also be 8702provided in the flags to KVM_XEN_HVM_CONFIG, without providing hypercall page 8703contents, to request that KVM generate hypercall page content automatically 8704and also enable interception of guest hypercalls with KVM_EXIT_XEN. 8705 8706The KVM_XEN_HVM_CONFIG_SHARED_INFO flag indicates the availability of the 8707KVM_XEN_HVM_SET_ATTR, KVM_XEN_HVM_GET_ATTR, KVM_XEN_VCPU_SET_ATTR and 8708KVM_XEN_VCPU_GET_ATTR ioctls, as well as the delivery of exception vectors 8709for event channel upcalls when the evtchn_upcall_pending field of a vcpu's 8710vcpu_info is set. 8711 8712The KVM_XEN_HVM_CONFIG_RUNSTATE flag indicates that the runstate-related 8713features KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADDR/_CURRENT/_DATA/_ADJUST are 8714supported by the KVM_XEN_VCPU_SET_ATTR/KVM_XEN_VCPU_GET_ATTR ioctls. 8715 8716The KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL flag indicates that IRQ routing entries 8717of the type KVM_IRQ_ROUTING_XEN_EVTCHN are supported, with the priority 8718field set to indicate 2 level event channel delivery. 8719 8720The KVM_XEN_HVM_CONFIG_EVTCHN_SEND flag indicates that KVM supports 8721injecting event channel events directly into the guest with the 8722KVM_XEN_HVM_EVTCHN_SEND ioctl. It also indicates support for the 8723KVM_XEN_ATTR_TYPE_EVTCHN/XEN_VERSION HVM attributes and the 8724KVM_XEN_VCPU_ATTR_TYPE_VCPU_ID/TIMER/UPCALL_VECTOR vCPU attributes. 8725related to event channel delivery, timers, and the XENVER_version 8726interception. 8727 8728The KVM_XEN_HVM_CONFIG_RUNSTATE_UPDATE_FLAG flag indicates that KVM supports 8729the KVM_XEN_ATTR_TYPE_RUNSTATE_UPDATE_FLAG attribute in the KVM_XEN_SET_ATTR 8730and KVM_XEN_GET_ATTR ioctls. This controls whether KVM will set the 8731XEN_RUNSTATE_UPDATE flag in guest memory mapped vcpu_runstate_info during 8732updates of the runstate information. Note that versions of KVM which support 8733the RUNSTATE feature above, but not the RUNSTATE_UPDATE_FLAG feature, will 8734always set the XEN_RUNSTATE_UPDATE flag when updating the guest structure, 8735which is perhaps counterintuitive. When this flag is advertised, KVM will 8736behave more correctly, not using the XEN_RUNSTATE_UPDATE flag until/unless 8737specifically enabled (by the guest making the hypercall, causing the VMM 8738to enable the KVM_XEN_ATTR_TYPE_RUNSTATE_UPDATE_FLAG attribute). 8739 8740The KVM_XEN_HVM_CONFIG_PVCLOCK_TSC_UNSTABLE flag indicates that KVM supports 8741clearing the PVCLOCK_TSC_STABLE_BIT flag in Xen pvclock sources. This will be 8742done when the KVM_CAP_XEN_HVM ioctl sets the 8743KVM_XEN_HVM_CONFIG_PVCLOCK_TSC_UNSTABLE flag. 8744 87458.31 KVM_CAP_PPC_MULTITCE 8746------------------------- 8747 8748:Capability: KVM_CAP_PPC_MULTITCE 8749:Architectures: ppc 8750:Type: vm 8751 8752This capability means the kernel is capable of handling hypercalls 8753H_PUT_TCE_INDIRECT and H_STUFF_TCE without passing those into the user 8754space. This significantly accelerates DMA operations for PPC KVM guests. 8755User space should expect that its handlers for these hypercalls 8756are not going to be called if user space previously registered LIOBN 8757in KVM (via KVM_CREATE_SPAPR_TCE or similar calls). 8758 8759In order to enable H_PUT_TCE_INDIRECT and H_STUFF_TCE use in the guest, 8760user space might have to advertise it for the guest. For example, 8761IBM pSeries (sPAPR) guest starts using them if "hcall-multi-tce" is 8762present in the "ibm,hypertas-functions" device-tree property. 8763 8764The hypercalls mentioned above may or may not be processed successfully 8765in the kernel based fast path. If they can not be handled by the kernel, 8766they will get passed on to user space. So user space still has to have 8767an implementation for these despite the in kernel acceleration. 8768 8769This capability is always enabled. 8770 87718.32 KVM_CAP_PTP_KVM 8772-------------------- 8773 8774:Architectures: arm64 8775 8776This capability indicates that the KVM virtual PTP service is 8777supported in the host. A VMM can check whether the service is 8778available to the guest on migration. 8779 87808.33 KVM_CAP_HYPERV_ENFORCE_CPUID 8781--------------------------------- 8782 8783Architectures: x86 8784 8785When enabled, KVM will disable emulated Hyper-V features provided to the 8786guest according to the bits Hyper-V CPUID feature leaves. Otherwise, all 8787currently implemented Hyper-V features are provided unconditionally when 8788Hyper-V identification is set in the HYPERV_CPUID_INTERFACE (0x40000001) 8789leaf. 8790 87918.34 KVM_CAP_EXIT_HYPERCALL 8792--------------------------- 8793 8794:Capability: KVM_CAP_EXIT_HYPERCALL 8795:Architectures: x86 8796:Type: vm 8797 8798This capability, if enabled, will cause KVM to exit to userspace 8799with KVM_EXIT_HYPERCALL exit reason to process some hypercalls. 8800 8801Calling KVM_CHECK_EXTENSION for this capability will return a bitmask 8802of hypercalls that can be configured to exit to userspace. 8803Right now, the only such hypercall is KVM_HC_MAP_GPA_RANGE. 8804 8805The argument to KVM_ENABLE_CAP is also a bitmask, and must be a subset 8806of the result of KVM_CHECK_EXTENSION. KVM will forward to userspace 8807the hypercalls whose corresponding bit is in the argument, and return 8808ENOSYS for the others. 8809 88108.35 KVM_CAP_PMU_CAPABILITY 8811--------------------------- 8812 8813:Capability: KVM_CAP_PMU_CAPABILITY 8814:Architectures: x86 8815:Type: vm 8816:Parameters: arg[0] is bitmask of PMU virtualization capabilities. 8817:Returns: 0 on success, -EINVAL when arg[0] contains invalid bits 8818 8819This capability alters PMU virtualization in KVM. 8820 8821Calling KVM_CHECK_EXTENSION for this capability returns a bitmask of 8822PMU virtualization capabilities that can be adjusted on a VM. 8823 8824The argument to KVM_ENABLE_CAP is also a bitmask and selects specific 8825PMU virtualization capabilities to be applied to the VM. This can 8826only be invoked on a VM prior to the creation of VCPUs. 8827 8828At this time, KVM_PMU_CAP_DISABLE is the only capability. Setting 8829this capability will disable PMU virtualization for that VM. Usermode 8830should adjust CPUID leaf 0xA to reflect that the PMU is disabled. 8831 88328.36 KVM_CAP_ARM_SYSTEM_SUSPEND 8833------------------------------- 8834 8835:Capability: KVM_CAP_ARM_SYSTEM_SUSPEND 8836:Architectures: arm64 8837:Type: vm 8838 8839When enabled, KVM will exit to userspace with KVM_EXIT_SYSTEM_EVENT of 8840type KVM_SYSTEM_EVENT_SUSPEND to process the guest suspend request. 8841 88428.37 KVM_CAP_S390_PROTECTED_DUMP 8843-------------------------------- 8844 8845:Capability: KVM_CAP_S390_PROTECTED_DUMP 8846:Architectures: s390 8847:Type: vm 8848 8849This capability indicates that KVM and the Ultravisor support dumping 8850PV guests. The `KVM_PV_DUMP` command is available for the 8851`KVM_S390_PV_COMMAND` ioctl and the `KVM_PV_INFO` command provides 8852dump related UV data. Also the vcpu ioctl `KVM_S390_PV_CPU_COMMAND` is 8853available and supports the `KVM_PV_DUMP_CPU` subcommand. 8854 88558.38 KVM_CAP_VM_DISABLE_NX_HUGE_PAGES 8856------------------------------------- 8857 8858:Capability: KVM_CAP_VM_DISABLE_NX_HUGE_PAGES 8859:Architectures: x86 8860:Type: vm 8861:Parameters: arg[0] must be 0. 8862:Returns: 0 on success, -EPERM if the userspace process does not 8863 have CAP_SYS_BOOT, -EINVAL if args[0] is not 0 or any vCPUs have been 8864 created. 8865 8866This capability disables the NX huge pages mitigation for iTLB MULTIHIT. 8867 8868The capability has no effect if the nx_huge_pages module parameter is not set. 8869 8870This capability may only be set before any vCPUs are created. 8871 88728.39 KVM_CAP_S390_CPU_TOPOLOGY 8873------------------------------ 8874 8875:Capability: KVM_CAP_S390_CPU_TOPOLOGY 8876:Architectures: s390 8877:Type: vm 8878 8879This capability indicates that KVM will provide the S390 CPU Topology 8880facility which consist of the interpretation of the PTF instruction for 8881the function code 2 along with interception and forwarding of both the 8882PTF instruction with function codes 0 or 1 and the STSI(15,1,x) 8883instruction to the userland hypervisor. 8884 8885The stfle facility 11, CPU Topology facility, should not be indicated 8886to the guest without this capability. 8887 8888When this capability is present, KVM provides a new attribute group 8889on vm fd, KVM_S390_VM_CPU_TOPOLOGY. 8890This new attribute allows to get, set or clear the Modified Change 8891Topology Report (MTCR) bit of the SCA through the kvm_device_attr 8892structure. 8893 8894When getting the Modified Change Topology Report value, the attr->addr 8895must point to a byte where the value will be stored or retrieved from. 8896 88978.40 KVM_CAP_ARM_EAGER_SPLIT_CHUNK_SIZE 8898--------------------------------------- 8899 8900:Capability: KVM_CAP_ARM_EAGER_SPLIT_CHUNK_SIZE 8901:Architectures: arm64 8902:Type: vm 8903:Parameters: arg[0] is the new split chunk size. 8904:Returns: 0 on success, -EINVAL if any memslot was already created. 8905 8906This capability sets the chunk size used in Eager Page Splitting. 8907 8908Eager Page Splitting improves the performance of dirty-logging (used 8909in live migrations) when guest memory is backed by huge-pages. It 8910avoids splitting huge-pages (into PAGE_SIZE pages) on fault, by doing 8911it eagerly when enabling dirty logging (with the 8912KVM_MEM_LOG_DIRTY_PAGES flag for a memory region), or when using 8913KVM_CLEAR_DIRTY_LOG. 8914 8915The chunk size specifies how many pages to break at a time, using a 8916single allocation for each chunk. Bigger the chunk size, more pages 8917need to be allocated ahead of time. 8918 8919The chunk size needs to be a valid block size. The list of acceptable 8920block sizes is exposed in KVM_CAP_ARM_SUPPORTED_BLOCK_SIZES as a 892164-bit bitmap (each bit describing a block size). The default value is 89220, to disable the eager page splitting. 8923 89248.41 KVM_CAP_VM_TYPES 8925--------------------- 8926 8927:Capability: KVM_CAP_MEMORY_ATTRIBUTES 8928:Architectures: x86 8929:Type: system ioctl 8930 8931This capability returns a bitmap of support VM types. The 1-setting of bit @n 8932means the VM type with value @n is supported. Possible values of @n are:: 8933 8934 #define KVM_X86_DEFAULT_VM 0 8935 #define KVM_X86_SW_PROTECTED_VM 1 8936 #define KVM_X86_SEV_VM 2 8937 #define KVM_X86_SEV_ES_VM 3 8938 8939Note, KVM_X86_SW_PROTECTED_VM is currently only for development and testing. 8940Do not use KVM_X86_SW_PROTECTED_VM for "real" VMs, and especially not in 8941production. The behavior and effective ABI for software-protected VMs is 8942unstable. 8943 89449. Known KVM API problems 8945========================= 8946 8947In some cases, KVM's API has some inconsistencies or common pitfalls 8948that userspace need to be aware of. This section details some of 8949these issues. 8950 8951Most of them are architecture specific, so the section is split by 8952architecture. 8953 89549.1. x86 8955-------- 8956 8957``KVM_GET_SUPPORTED_CPUID`` issues 8958^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 8959 8960In general, ``KVM_GET_SUPPORTED_CPUID`` is designed so that it is possible 8961to take its result and pass it directly to ``KVM_SET_CPUID2``. This section 8962documents some cases in which that requires some care. 8963 8964Local APIC features 8965~~~~~~~~~~~~~~~~~~~ 8966 8967CPU[EAX=1]:ECX[21] (X2APIC) is reported by ``KVM_GET_SUPPORTED_CPUID``, 8968but it can only be enabled if ``KVM_CREATE_IRQCHIP`` or 8969``KVM_ENABLE_CAP(KVM_CAP_IRQCHIP_SPLIT)`` are used to enable in-kernel emulation of 8970the local APIC. 8971 8972The same is true for the ``KVM_FEATURE_PV_UNHALT`` paravirtualized feature. 8973 8974CPU[EAX=1]:ECX[24] (TSC_DEADLINE) is not reported by ``KVM_GET_SUPPORTED_CPUID``. 8975It can be enabled if ``KVM_CAP_TSC_DEADLINE_TIMER`` is present and the kernel 8976has enabled in-kernel emulation of the local APIC. 8977 8978CPU topology 8979~~~~~~~~~~~~ 8980 8981Several CPUID values include topology information for the host CPU: 89820x0b and 0x1f for Intel systems, 0x8000001e for AMD systems. Different 8983versions of KVM return different values for this information and userspace 8984should not rely on it. Currently they return all zeroes. 8985 8986If userspace wishes to set up a guest topology, it should be careful that 8987the values of these three leaves differ for each CPU. In particular, 8988the APIC ID is found in EDX for all subleaves of 0x0b and 0x1f, and in EAX 8989for 0x8000001e; the latter also encodes the core id and node id in bits 89907:0 of EBX and ECX respectively. 8991 8992Obsolete ioctls and capabilities 8993^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 8994 8995KVM_CAP_DISABLE_QUIRKS does not let userspace know which quirks are actually 8996available. Use ``KVM_CHECK_EXTENSION(KVM_CAP_DISABLE_QUIRKS2)`` instead if 8997available. 8998 8999Ordering of KVM_GET_*/KVM_SET_* ioctls 9000^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 9001 9002TBD 9003