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- irq_type[0]: 895 out-of-kernel GIC: irq_id 0 is IRQ, irq_id 1 is FIQ 896- irq_type[1]: 897 in-kernel GIC: SPI, irq_id between 32 and 1019 (incl.) 898 (the vcpu_index field is ignored) 899- irq_type[2]: 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 14064.36 KVM_SET_TSS_ADDR 1407--------------------- 1408 1409:Capability: KVM_CAP_SET_TSS_ADDR 1410:Architectures: x86 1411:Type: vm ioctl 1412:Parameters: unsigned long tss_address (in) 1413:Returns: 0 on success, -1 on error 1414 1415This ioctl defines the physical address of a three-page region in the guest 1416physical address space. The region must be within the first 4GB of the 1417guest physical address space and must not conflict with any memory slot 1418or any mmio address. The guest may malfunction if it accesses this memory 1419region. 1420 1421This ioctl is required on Intel-based hosts. This is needed on Intel hardware 1422because of a quirk in the virtualization implementation (see the internals 1423documentation when it pops into existence). 1424 1425 14264.37 KVM_ENABLE_CAP 1427------------------- 1428 1429:Capability: KVM_CAP_ENABLE_CAP 1430:Architectures: mips, ppc, s390, x86, loongarch 1431:Type: vcpu ioctl 1432:Parameters: struct kvm_enable_cap (in) 1433:Returns: 0 on success; -1 on error 1434 1435:Capability: KVM_CAP_ENABLE_CAP_VM 1436:Architectures: all 1437:Type: vm ioctl 1438:Parameters: struct kvm_enable_cap (in) 1439:Returns: 0 on success; -1 on error 1440 1441.. note:: 1442 1443 Not all extensions are enabled by default. Using this ioctl the application 1444 can enable an extension, making it available to the guest. 1445 1446On systems that do not support this ioctl, it always fails. On systems that 1447do support it, it only works for extensions that are supported for enablement. 1448 1449To check if a capability can be enabled, the KVM_CHECK_EXTENSION ioctl should 1450be used. 1451 1452:: 1453 1454 struct kvm_enable_cap { 1455 /* in */ 1456 __u32 cap; 1457 1458The capability that is supposed to get enabled. 1459 1460:: 1461 1462 __u32 flags; 1463 1464A bitfield indicating future enhancements. Has to be 0 for now. 1465 1466:: 1467 1468 __u64 args[4]; 1469 1470Arguments for enabling a feature. If a feature needs initial values to 1471function properly, this is the place to put them. 1472 1473:: 1474 1475 __u8 pad[64]; 1476 }; 1477 1478The vcpu ioctl should be used for vcpu-specific capabilities, the vm ioctl 1479for vm-wide capabilities. 1480 14814.38 KVM_GET_MP_STATE 1482--------------------- 1483 1484:Capability: KVM_CAP_MP_STATE 1485:Architectures: x86, s390, arm64, riscv, loongarch 1486:Type: vcpu ioctl 1487:Parameters: struct kvm_mp_state (out) 1488:Returns: 0 on success; -1 on error 1489 1490:: 1491 1492 struct kvm_mp_state { 1493 __u32 mp_state; 1494 }; 1495 1496Returns the vcpu's current "multiprocessing state" (though also valid on 1497uniprocessor guests). 1498 1499Possible values are: 1500 1501 ========================== =============================================== 1502 KVM_MP_STATE_RUNNABLE the vcpu is currently running 1503 [x86,arm64,riscv,loongarch] 1504 KVM_MP_STATE_UNINITIALIZED the vcpu is an application processor (AP) 1505 which has not yet received an INIT signal [x86] 1506 KVM_MP_STATE_INIT_RECEIVED the vcpu has received an INIT signal, and is 1507 now ready for a SIPI [x86] 1508 KVM_MP_STATE_HALTED the vcpu has executed a HLT instruction and 1509 is waiting for an interrupt [x86] 1510 KVM_MP_STATE_SIPI_RECEIVED the vcpu has just received a SIPI (vector 1511 accessible via KVM_GET_VCPU_EVENTS) [x86] 1512 KVM_MP_STATE_STOPPED the vcpu is stopped [s390,arm64,riscv] 1513 KVM_MP_STATE_CHECK_STOP the vcpu is in a special error state [s390] 1514 KVM_MP_STATE_OPERATING the vcpu is operating (running or halted) 1515 [s390] 1516 KVM_MP_STATE_LOAD the vcpu is in a special load/startup state 1517 [s390] 1518 KVM_MP_STATE_SUSPENDED the vcpu is in a suspend state and is waiting 1519 for a wakeup event [arm64] 1520 ========================== =============================================== 1521 1522On x86, this ioctl is only useful after KVM_CREATE_IRQCHIP. Without an 1523in-kernel irqchip, the multiprocessing state must be maintained by userspace on 1524these architectures. 1525 1526For arm64: 1527^^^^^^^^^^ 1528 1529If a vCPU is in the KVM_MP_STATE_SUSPENDED state, KVM will emulate the 1530architectural execution of a WFI instruction. 1531 1532If a wakeup event is recognized, KVM will exit to userspace with a 1533KVM_SYSTEM_EVENT exit, where the event type is KVM_SYSTEM_EVENT_WAKEUP. If 1534userspace wants to honor the wakeup, it must set the vCPU's MP state to 1535KVM_MP_STATE_RUNNABLE. If it does not, KVM will continue to await a wakeup 1536event in subsequent calls to KVM_RUN. 1537 1538.. warning:: 1539 1540 If userspace intends to keep the vCPU in a SUSPENDED state, it is 1541 strongly recommended that userspace take action to suppress the 1542 wakeup event (such as masking an interrupt). Otherwise, subsequent 1543 calls to KVM_RUN will immediately exit with a KVM_SYSTEM_EVENT_WAKEUP 1544 event and inadvertently waste CPU cycles. 1545 1546 Additionally, if userspace takes action to suppress a wakeup event, 1547 it is strongly recommended that it also restores the vCPU to its 1548 original state when the vCPU is made RUNNABLE again. For example, 1549 if userspace masked a pending interrupt to suppress the wakeup, 1550 the interrupt should be unmasked before returning control to the 1551 guest. 1552 1553For riscv: 1554^^^^^^^^^^ 1555 1556The only states that are valid are KVM_MP_STATE_STOPPED and 1557KVM_MP_STATE_RUNNABLE which reflect if the vcpu is paused or not. 1558 1559On LoongArch, only the KVM_MP_STATE_RUNNABLE state is used to reflect 1560whether the vcpu is runnable. 1561 15624.39 KVM_SET_MP_STATE 1563--------------------- 1564 1565:Capability: KVM_CAP_MP_STATE 1566:Architectures: x86, s390, arm64, riscv, loongarch 1567:Type: vcpu ioctl 1568:Parameters: struct kvm_mp_state (in) 1569:Returns: 0 on success; -1 on error 1570 1571Sets the vcpu's current "multiprocessing state"; see KVM_GET_MP_STATE for 1572arguments. 1573 1574On x86, this ioctl is only useful after KVM_CREATE_IRQCHIP. Without an 1575in-kernel irqchip, the multiprocessing state must be maintained by userspace on 1576these architectures. 1577 1578For arm64/riscv: 1579^^^^^^^^^^^^^^^^ 1580 1581The only states that are valid are KVM_MP_STATE_STOPPED and 1582KVM_MP_STATE_RUNNABLE which reflect if the vcpu should be paused or not. 1583 1584On LoongArch, only the KVM_MP_STATE_RUNNABLE state is used to reflect 1585whether the vcpu is runnable. 1586 15874.40 KVM_SET_IDENTITY_MAP_ADDR 1588------------------------------ 1589 1590:Capability: KVM_CAP_SET_IDENTITY_MAP_ADDR 1591:Architectures: x86 1592:Type: vm ioctl 1593:Parameters: unsigned long identity (in) 1594:Returns: 0 on success, -1 on error 1595 1596This ioctl defines the physical address of a one-page region in the guest 1597physical address space. The region must be within the first 4GB of the 1598guest physical address space and must not conflict with any memory slot 1599or any mmio address. The guest may malfunction if it accesses this memory 1600region. 1601 1602Setting the address to 0 will result in resetting the address to its default 1603(0xfffbc000). 1604 1605This ioctl is required on Intel-based hosts. This is needed on Intel hardware 1606because of a quirk in the virtualization implementation (see the internals 1607documentation when it pops into existence). 1608 1609Fails if any VCPU has already been created. 1610 16114.41 KVM_SET_BOOT_CPU_ID 1612------------------------ 1613 1614:Capability: KVM_CAP_SET_BOOT_CPU_ID 1615:Architectures: x86 1616:Type: vm ioctl 1617:Parameters: unsigned long vcpu_id 1618:Returns: 0 on success, -1 on error 1619 1620Define which vcpu is the Bootstrap Processor (BSP). Values are the same 1621as the vcpu id in KVM_CREATE_VCPU. If this ioctl is not called, the default 1622is vcpu 0. This ioctl has to be called before vcpu creation, 1623otherwise it will return EBUSY error. 1624 1625 16264.42 KVM_GET_XSAVE 1627------------------ 1628 1629:Capability: KVM_CAP_XSAVE 1630:Architectures: x86 1631:Type: vcpu ioctl 1632:Parameters: struct kvm_xsave (out) 1633:Returns: 0 on success, -1 on error 1634 1635 1636:: 1637 1638 struct kvm_xsave { 1639 __u32 region[1024]; 1640 __u32 extra[0]; 1641 }; 1642 1643This ioctl would copy current vcpu's xsave struct to the userspace. 1644 1645 16464.43 KVM_SET_XSAVE 1647------------------ 1648 1649:Capability: KVM_CAP_XSAVE and KVM_CAP_XSAVE2 1650:Architectures: x86 1651:Type: vcpu ioctl 1652:Parameters: struct kvm_xsave (in) 1653:Returns: 0 on success, -1 on error 1654 1655:: 1656 1657 1658 struct kvm_xsave { 1659 __u32 region[1024]; 1660 __u32 extra[0]; 1661 }; 1662 1663This ioctl would copy userspace's xsave struct to the kernel. It copies 1664as many bytes as are returned by KVM_CHECK_EXTENSION(KVM_CAP_XSAVE2), 1665when invoked on the vm file descriptor. The size value returned by 1666KVM_CHECK_EXTENSION(KVM_CAP_XSAVE2) will always be at least 4096. 1667Currently, it is only greater than 4096 if a dynamic feature has been 1668enabled with ``arch_prctl()``, but this may change in the future. 1669 1670The offsets of the state save areas in struct kvm_xsave follow the 1671contents of CPUID leaf 0xD on the host. 1672 1673 16744.44 KVM_GET_XCRS 1675----------------- 1676 1677:Capability: KVM_CAP_XCRS 1678:Architectures: x86 1679:Type: vcpu ioctl 1680:Parameters: struct kvm_xcrs (out) 1681:Returns: 0 on success, -1 on error 1682 1683:: 1684 1685 struct kvm_xcr { 1686 __u32 xcr; 1687 __u32 reserved; 1688 __u64 value; 1689 }; 1690 1691 struct kvm_xcrs { 1692 __u32 nr_xcrs; 1693 __u32 flags; 1694 struct kvm_xcr xcrs[KVM_MAX_XCRS]; 1695 __u64 padding[16]; 1696 }; 1697 1698This ioctl would copy current vcpu's xcrs to the userspace. 1699 1700 17014.45 KVM_SET_XCRS 1702----------------- 1703 1704:Capability: KVM_CAP_XCRS 1705:Architectures: x86 1706:Type: vcpu ioctl 1707:Parameters: struct kvm_xcrs (in) 1708:Returns: 0 on success, -1 on error 1709 1710:: 1711 1712 struct kvm_xcr { 1713 __u32 xcr; 1714 __u32 reserved; 1715 __u64 value; 1716 }; 1717 1718 struct kvm_xcrs { 1719 __u32 nr_xcrs; 1720 __u32 flags; 1721 struct kvm_xcr xcrs[KVM_MAX_XCRS]; 1722 __u64 padding[16]; 1723 }; 1724 1725This ioctl would set vcpu's xcr to the value userspace specified. 1726 1727 17284.46 KVM_GET_SUPPORTED_CPUID 1729---------------------------- 1730 1731:Capability: KVM_CAP_EXT_CPUID 1732:Architectures: x86 1733:Type: system ioctl 1734:Parameters: struct kvm_cpuid2 (in/out) 1735:Returns: 0 on success, -1 on error 1736 1737:: 1738 1739 struct kvm_cpuid2 { 1740 __u32 nent; 1741 __u32 padding; 1742 struct kvm_cpuid_entry2 entries[0]; 1743 }; 1744 1745 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX BIT(0) 1746 #define KVM_CPUID_FLAG_STATEFUL_FUNC BIT(1) /* deprecated */ 1747 #define KVM_CPUID_FLAG_STATE_READ_NEXT BIT(2) /* deprecated */ 1748 1749 struct kvm_cpuid_entry2 { 1750 __u32 function; 1751 __u32 index; 1752 __u32 flags; 1753 __u32 eax; 1754 __u32 ebx; 1755 __u32 ecx; 1756 __u32 edx; 1757 __u32 padding[3]; 1758 }; 1759 1760This ioctl returns x86 cpuid features which are supported by both the 1761hardware and kvm in its default configuration. Userspace can use the 1762information returned by this ioctl to construct cpuid information (for 1763KVM_SET_CPUID2) that is consistent with hardware, kernel, and 1764userspace capabilities, and with user requirements (for example, the 1765user may wish to constrain cpuid to emulate older hardware, or for 1766feature consistency across a cluster). 1767 1768Dynamically-enabled feature bits need to be requested with 1769``arch_prctl()`` before calling this ioctl. Feature bits that have not 1770been requested are excluded from the result. 1771 1772Note that certain capabilities, such as KVM_CAP_X86_DISABLE_EXITS, may 1773expose cpuid features (e.g. MONITOR) which are not supported by kvm in 1774its default configuration. If userspace enables such capabilities, it 1775is responsible for modifying the results of this ioctl appropriately. 1776 1777Userspace invokes KVM_GET_SUPPORTED_CPUID by passing a kvm_cpuid2 structure 1778with the 'nent' field indicating the number of entries in the variable-size 1779array 'entries'. If the number of entries is too low to describe the cpu 1780capabilities, an error (E2BIG) is returned. If the number is too high, 1781the 'nent' field is adjusted and an error (ENOMEM) is returned. If the 1782number is just right, the 'nent' field is adjusted to the number of valid 1783entries in the 'entries' array, which is then filled. 1784 1785The entries returned are the host cpuid as returned by the cpuid instruction, 1786with unknown or unsupported features masked out. Some features (for example, 1787x2apic), may not be present in the host cpu, but are exposed by kvm if it can 1788emulate them efficiently. The fields in each entry are defined as follows: 1789 1790 function: 1791 the eax value used to obtain the entry 1792 1793 index: 1794 the ecx value used to obtain the entry (for entries that are 1795 affected by ecx) 1796 1797 flags: 1798 an OR of zero or more of the following: 1799 1800 KVM_CPUID_FLAG_SIGNIFCANT_INDEX: 1801 if the index field is valid 1802 1803 eax, ebx, ecx, edx: 1804 the values returned by the cpuid instruction for 1805 this function/index combination 1806 1807The TSC deadline timer feature (CPUID leaf 1, ecx[24]) is always returned 1808as false, since the feature depends on KVM_CREATE_IRQCHIP for local APIC 1809support. Instead it is reported via:: 1810 1811 ioctl(KVM_CHECK_EXTENSION, KVM_CAP_TSC_DEADLINE_TIMER) 1812 1813if that returns true and you use KVM_CREATE_IRQCHIP, or if you emulate the 1814feature in userspace, then you can enable the feature for KVM_SET_CPUID2. 1815 1816 18174.47 KVM_PPC_GET_PVINFO 1818----------------------- 1819 1820:Capability: KVM_CAP_PPC_GET_PVINFO 1821:Architectures: ppc 1822:Type: vm ioctl 1823:Parameters: struct kvm_ppc_pvinfo (out) 1824:Returns: 0 on success, !0 on error 1825 1826:: 1827 1828 struct kvm_ppc_pvinfo { 1829 __u32 flags; 1830 __u32 hcall[4]; 1831 __u8 pad[108]; 1832 }; 1833 1834This ioctl fetches PV specific information that need to be passed to the guest 1835using the device tree or other means from vm context. 1836 1837The hcall array defines 4 instructions that make up a hypercall. 1838 1839If any additional field gets added to this structure later on, a bit for that 1840additional piece of information will be set in the flags bitmap. 1841 1842The flags bitmap is defined as:: 1843 1844 /* the host supports the ePAPR idle hcall 1845 #define KVM_PPC_PVINFO_FLAGS_EV_IDLE (1<<0) 1846 18474.52 KVM_SET_GSI_ROUTING 1848------------------------ 1849 1850:Capability: KVM_CAP_IRQ_ROUTING 1851:Architectures: x86 s390 arm64 1852:Type: vm ioctl 1853:Parameters: struct kvm_irq_routing (in) 1854:Returns: 0 on success, -1 on error 1855 1856Sets the GSI routing table entries, overwriting any previously set entries. 1857 1858On arm64, GSI routing has the following limitation: 1859 1860- GSI routing does not apply to KVM_IRQ_LINE but only to KVM_IRQFD. 1861 1862:: 1863 1864 struct kvm_irq_routing { 1865 __u32 nr; 1866 __u32 flags; 1867 struct kvm_irq_routing_entry entries[0]; 1868 }; 1869 1870No flags are specified so far, the corresponding field must be set to zero. 1871 1872:: 1873 1874 struct kvm_irq_routing_entry { 1875 __u32 gsi; 1876 __u32 type; 1877 __u32 flags; 1878 __u32 pad; 1879 union { 1880 struct kvm_irq_routing_irqchip irqchip; 1881 struct kvm_irq_routing_msi msi; 1882 struct kvm_irq_routing_s390_adapter adapter; 1883 struct kvm_irq_routing_hv_sint hv_sint; 1884 struct kvm_irq_routing_xen_evtchn xen_evtchn; 1885 __u32 pad[8]; 1886 } u; 1887 }; 1888 1889 /* gsi routing entry types */ 1890 #define KVM_IRQ_ROUTING_IRQCHIP 1 1891 #define KVM_IRQ_ROUTING_MSI 2 1892 #define KVM_IRQ_ROUTING_S390_ADAPTER 3 1893 #define KVM_IRQ_ROUTING_HV_SINT 4 1894 #define KVM_IRQ_ROUTING_XEN_EVTCHN 5 1895 1896flags: 1897 1898- KVM_MSI_VALID_DEVID: used along with KVM_IRQ_ROUTING_MSI routing entry 1899 type, specifies that the devid field contains a valid value. The per-VM 1900 KVM_CAP_MSI_DEVID capability advertises the requirement to provide 1901 the device ID. If this capability is not available, userspace should 1902 never set the KVM_MSI_VALID_DEVID flag as the ioctl might fail. 1903- zero otherwise 1904 1905:: 1906 1907 struct kvm_irq_routing_irqchip { 1908 __u32 irqchip; 1909 __u32 pin; 1910 }; 1911 1912 struct kvm_irq_routing_msi { 1913 __u32 address_lo; 1914 __u32 address_hi; 1915 __u32 data; 1916 union { 1917 __u32 pad; 1918 __u32 devid; 1919 }; 1920 }; 1921 1922If KVM_MSI_VALID_DEVID is set, devid contains a unique device identifier 1923for the device that wrote the MSI message. For PCI, this is usually a 1924BFD identifier in the lower 16 bits. 1925 1926On x86, address_hi is ignored unless the KVM_X2APIC_API_USE_32BIT_IDS 1927feature of KVM_CAP_X2APIC_API capability is enabled. If it is enabled, 1928address_hi bits 31-8 provide bits 31-8 of the destination id. Bits 7-0 of 1929address_hi must be zero. 1930 1931:: 1932 1933 struct kvm_irq_routing_s390_adapter { 1934 __u64 ind_addr; 1935 __u64 summary_addr; 1936 __u64 ind_offset; 1937 __u32 summary_offset; 1938 __u32 adapter_id; 1939 }; 1940 1941 struct kvm_irq_routing_hv_sint { 1942 __u32 vcpu; 1943 __u32 sint; 1944 }; 1945 1946 struct kvm_irq_routing_xen_evtchn { 1947 __u32 port; 1948 __u32 vcpu; 1949 __u32 priority; 1950 }; 1951 1952 1953When KVM_CAP_XEN_HVM includes the KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL bit 1954in its indication of supported features, routing to Xen event channels 1955is supported. Although the priority field is present, only the value 1956KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL is supported, which means delivery by 19572 level event channels. FIFO event channel support may be added in 1958the future. 1959 1960 19614.55 KVM_SET_TSC_KHZ 1962-------------------- 1963 1964:Capability: KVM_CAP_TSC_CONTROL / KVM_CAP_VM_TSC_CONTROL 1965:Architectures: x86 1966:Type: vcpu ioctl / vm ioctl 1967:Parameters: virtual tsc_khz 1968:Returns: 0 on success, -1 on error 1969 1970Specifies the tsc frequency for the virtual machine. The unit of the 1971frequency is KHz. 1972 1973If the KVM_CAP_VM_TSC_CONTROL capability is advertised, this can also 1974be used as a vm ioctl to set the initial tsc frequency of subsequently 1975created vCPUs. 1976 19774.56 KVM_GET_TSC_KHZ 1978-------------------- 1979 1980:Capability: KVM_CAP_GET_TSC_KHZ / KVM_CAP_VM_TSC_CONTROL 1981:Architectures: x86 1982:Type: vcpu ioctl / vm ioctl 1983:Parameters: none 1984:Returns: virtual tsc-khz on success, negative value on error 1985 1986Returns the tsc frequency of the guest. The unit of the return value is 1987KHz. If the host has unstable tsc this ioctl returns -EIO instead as an 1988error. 1989 1990 19914.57 KVM_GET_LAPIC 1992------------------ 1993 1994:Capability: KVM_CAP_IRQCHIP 1995:Architectures: x86 1996:Type: vcpu ioctl 1997:Parameters: struct kvm_lapic_state (out) 1998:Returns: 0 on success, -1 on error 1999 2000:: 2001 2002 #define KVM_APIC_REG_SIZE 0x400 2003 struct kvm_lapic_state { 2004 char regs[KVM_APIC_REG_SIZE]; 2005 }; 2006 2007Reads the Local APIC registers and copies them into the input argument. The 2008data format and layout are the same as documented in the architecture manual. 2009 2010If KVM_X2APIC_API_USE_32BIT_IDS feature of KVM_CAP_X2APIC_API is 2011enabled, then the format of APIC_ID register depends on the APIC mode 2012(reported by MSR_IA32_APICBASE) of its VCPU. x2APIC stores APIC ID in 2013the APIC_ID register (bytes 32-35). xAPIC only allows an 8-bit APIC ID 2014which is stored in bits 31-24 of the APIC register, or equivalently in 2015byte 35 of struct kvm_lapic_state's regs field. KVM_GET_LAPIC must then 2016be called after MSR_IA32_APICBASE has been set with KVM_SET_MSR. 2017 2018If KVM_X2APIC_API_USE_32BIT_IDS feature is disabled, struct kvm_lapic_state 2019always uses xAPIC format. 2020 2021 20224.58 KVM_SET_LAPIC 2023------------------ 2024 2025:Capability: KVM_CAP_IRQCHIP 2026:Architectures: x86 2027:Type: vcpu ioctl 2028:Parameters: struct kvm_lapic_state (in) 2029:Returns: 0 on success, -1 on error 2030 2031:: 2032 2033 #define KVM_APIC_REG_SIZE 0x400 2034 struct kvm_lapic_state { 2035 char regs[KVM_APIC_REG_SIZE]; 2036 }; 2037 2038Copies the input argument into the Local APIC registers. The data format 2039and layout are the same as documented in the architecture manual. 2040 2041The format of the APIC ID register (bytes 32-35 of struct kvm_lapic_state's 2042regs field) depends on the state of the KVM_CAP_X2APIC_API capability. 2043See the note in KVM_GET_LAPIC. 2044 2045 20464.59 KVM_IOEVENTFD 2047------------------ 2048 2049:Capability: KVM_CAP_IOEVENTFD 2050:Architectures: all 2051:Type: vm ioctl 2052:Parameters: struct kvm_ioeventfd (in) 2053:Returns: 0 on success, !0 on error 2054 2055This ioctl attaches or detaches an ioeventfd to a legal pio/mmio address 2056within the guest. A guest write in the registered address will signal the 2057provided event instead of triggering an exit. 2058 2059:: 2060 2061 struct kvm_ioeventfd { 2062 __u64 datamatch; 2063 __u64 addr; /* legal pio/mmio address */ 2064 __u32 len; /* 0, 1, 2, 4, or 8 bytes */ 2065 __s32 fd; 2066 __u32 flags; 2067 __u8 pad[36]; 2068 }; 2069 2070For the special case of virtio-ccw devices on s390, the ioevent is matched 2071to a subchannel/virtqueue tuple instead. 2072 2073The following flags are defined:: 2074 2075 #define KVM_IOEVENTFD_FLAG_DATAMATCH (1 << kvm_ioeventfd_flag_nr_datamatch) 2076 #define KVM_IOEVENTFD_FLAG_PIO (1 << kvm_ioeventfd_flag_nr_pio) 2077 #define KVM_IOEVENTFD_FLAG_DEASSIGN (1 << kvm_ioeventfd_flag_nr_deassign) 2078 #define KVM_IOEVENTFD_FLAG_VIRTIO_CCW_NOTIFY \ 2079 (1 << kvm_ioeventfd_flag_nr_virtio_ccw_notify) 2080 2081If datamatch flag is set, the event will be signaled only if the written value 2082to the registered address is equal to datamatch in struct kvm_ioeventfd. 2083 2084For virtio-ccw devices, addr contains the subchannel id and datamatch the 2085virtqueue index. 2086 2087With KVM_CAP_IOEVENTFD_ANY_LENGTH, a zero length ioeventfd is allowed, and 2088the kernel will ignore the length of guest write and may get a faster vmexit. 2089The speedup may only apply to specific architectures, but the ioeventfd will 2090work anyway. 2091 20924.60 KVM_DIRTY_TLB 2093------------------ 2094 2095:Capability: KVM_CAP_SW_TLB 2096:Architectures: ppc 2097:Type: vcpu ioctl 2098:Parameters: struct kvm_dirty_tlb (in) 2099:Returns: 0 on success, -1 on error 2100 2101:: 2102 2103 struct kvm_dirty_tlb { 2104 __u64 bitmap; 2105 __u32 num_dirty; 2106 }; 2107 2108This must be called whenever userspace has changed an entry in the shared 2109TLB, prior to calling KVM_RUN on the associated vcpu. 2110 2111The "bitmap" field is the userspace address of an array. This array 2112consists of a number of bits, equal to the total number of TLB entries as 2113determined by the last successful call to KVM_CONFIG_TLB, rounded up to the 2114nearest multiple of 64. 2115 2116Each bit corresponds to one TLB entry, ordered the same as in the shared TLB 2117array. 2118 2119The array is little-endian: the bit 0 is the least significant bit of the 2120first byte, bit 8 is the least significant bit of the second byte, etc. 2121This avoids any complications with differing word sizes. 2122 2123The "num_dirty" field is a performance hint for KVM to determine whether it 2124should skip processing the bitmap and just invalidate everything. It must 2125be set to the number of set bits in the bitmap. 2126 2127 21284.62 KVM_CREATE_SPAPR_TCE 2129------------------------- 2130 2131:Capability: KVM_CAP_SPAPR_TCE 2132:Architectures: powerpc 2133:Type: vm ioctl 2134:Parameters: struct kvm_create_spapr_tce (in) 2135:Returns: file descriptor for manipulating the created TCE table 2136 2137This creates a virtual TCE (translation control entry) table, which 2138is an IOMMU for PAPR-style virtual I/O. It is used to translate 2139logical addresses used in virtual I/O into guest physical addresses, 2140and provides a scatter/gather capability for PAPR virtual I/O. 2141 2142:: 2143 2144 /* for KVM_CAP_SPAPR_TCE */ 2145 struct kvm_create_spapr_tce { 2146 __u64 liobn; 2147 __u32 window_size; 2148 }; 2149 2150The liobn field gives the logical IO bus number for which to create a 2151TCE table. The window_size field specifies the size of the DMA window 2152which this TCE table will translate - the table will contain one 64 2153bit TCE entry for every 4kiB of the DMA window. 2154 2155When the guest issues an H_PUT_TCE hcall on a liobn for which a TCE 2156table has been created using this ioctl(), the kernel will handle it 2157in real mode, updating the TCE table. H_PUT_TCE calls for other 2158liobns will cause a vm exit and must be handled by userspace. 2159 2160The return value is a file descriptor which can be passed to mmap(2) 2161to map the created TCE table into userspace. This lets userspace read 2162the entries written by kernel-handled H_PUT_TCE calls, and also lets 2163userspace update the TCE table directly which is useful in some 2164circumstances. 2165 2166 21674.63 KVM_ALLOCATE_RMA 2168--------------------- 2169 2170:Capability: KVM_CAP_PPC_RMA 2171:Architectures: powerpc 2172:Type: vm ioctl 2173:Parameters: struct kvm_allocate_rma (out) 2174:Returns: file descriptor for mapping the allocated RMA 2175 2176This allocates a Real Mode Area (RMA) from the pool allocated at boot 2177time by the kernel. An RMA is a physically-contiguous, aligned region 2178of memory used on older POWER processors to provide the memory which 2179will be accessed by real-mode (MMU off) accesses in a KVM guest. 2180POWER processors support a set of sizes for the RMA that usually 2181includes 64MB, 128MB, 256MB and some larger powers of two. 2182 2183:: 2184 2185 /* for KVM_ALLOCATE_RMA */ 2186 struct kvm_allocate_rma { 2187 __u64 rma_size; 2188 }; 2189 2190The return value is a file descriptor which can be passed to mmap(2) 2191to map the allocated RMA into userspace. The mapped area can then be 2192passed to the KVM_SET_USER_MEMORY_REGION ioctl to establish it as the 2193RMA for a virtual machine. The size of the RMA in bytes (which is 2194fixed at host kernel boot time) is returned in the rma_size field of 2195the argument structure. 2196 2197The KVM_CAP_PPC_RMA capability is 1 or 2 if the KVM_ALLOCATE_RMA ioctl 2198is supported; 2 if the processor requires all virtual machines to have 2199an RMA, or 1 if the processor can use an RMA but doesn't require it, 2200because it supports the Virtual RMA (VRMA) facility. 2201 2202 22034.64 KVM_NMI 2204------------ 2205 2206:Capability: KVM_CAP_USER_NMI 2207:Architectures: x86 2208:Type: vcpu ioctl 2209:Parameters: none 2210:Returns: 0 on success, -1 on error 2211 2212Queues an NMI on the thread's vcpu. Note this is well defined only 2213when KVM_CREATE_IRQCHIP has not been called, since this is an interface 2214between the virtual cpu core and virtual local APIC. After KVM_CREATE_IRQCHIP 2215has been called, this interface is completely emulated within the kernel. 2216 2217To use this to emulate the LINT1 input with KVM_CREATE_IRQCHIP, use the 2218following algorithm: 2219 2220 - pause the vcpu 2221 - read the local APIC's state (KVM_GET_LAPIC) 2222 - check whether changing LINT1 will queue an NMI (see the LVT entry for LINT1) 2223 - if so, issue KVM_NMI 2224 - resume the vcpu 2225 2226Some guests configure the LINT1 NMI input to cause a panic, aiding in 2227debugging. 2228 2229 22304.65 KVM_S390_UCAS_MAP 2231---------------------- 2232 2233:Capability: KVM_CAP_S390_UCONTROL 2234:Architectures: s390 2235:Type: vcpu ioctl 2236:Parameters: struct kvm_s390_ucas_mapping (in) 2237:Returns: 0 in case of success 2238 2239The parameter is defined like this:: 2240 2241 struct kvm_s390_ucas_mapping { 2242 __u64 user_addr; 2243 __u64 vcpu_addr; 2244 __u64 length; 2245 }; 2246 2247This ioctl maps the memory at "user_addr" with the length "length" to 2248the vcpu's address space starting at "vcpu_addr". All parameters need to 2249be aligned by 1 megabyte. 2250 2251 22524.66 KVM_S390_UCAS_UNMAP 2253------------------------ 2254 2255:Capability: KVM_CAP_S390_UCONTROL 2256:Architectures: s390 2257:Type: vcpu ioctl 2258:Parameters: struct kvm_s390_ucas_mapping (in) 2259:Returns: 0 in case of success 2260 2261The parameter is defined like this:: 2262 2263 struct kvm_s390_ucas_mapping { 2264 __u64 user_addr; 2265 __u64 vcpu_addr; 2266 __u64 length; 2267 }; 2268 2269This ioctl unmaps the memory in the vcpu's address space starting at 2270"vcpu_addr" with the length "length". The field "user_addr" is ignored. 2271All parameters need to be aligned by 1 megabyte. 2272 2273 22744.67 KVM_S390_VCPU_FAULT 2275------------------------ 2276 2277:Capability: KVM_CAP_S390_UCONTROL 2278:Architectures: s390 2279:Type: vcpu ioctl 2280:Parameters: vcpu absolute address (in) 2281:Returns: 0 in case of success 2282 2283This call creates a page table entry on the virtual cpu's address space 2284(for user controlled virtual machines) or the virtual machine's address 2285space (for regular virtual machines). This only works for minor faults, 2286thus it's recommended to access subject memory page via the user page 2287table upfront. This is useful to handle validity intercepts for user 2288controlled virtual machines to fault in the virtual cpu's lowcore pages 2289prior to calling the KVM_RUN ioctl. 2290 2291 22924.68 KVM_SET_ONE_REG 2293-------------------- 2294 2295:Capability: KVM_CAP_ONE_REG 2296:Architectures: all 2297:Type: vcpu ioctl 2298:Parameters: struct kvm_one_reg (in) 2299:Returns: 0 on success, negative value on failure 2300 2301Errors: 2302 2303 ====== ============================================================ 2304 ENOENT no such register 2305 EINVAL invalid register ID, or no such register or used with VMs in 2306 protected virtualization mode on s390 2307 EPERM (arm64) register access not allowed before vcpu finalization 2308 EBUSY (riscv) changing register value not allowed after the vcpu 2309 has run at least once 2310 ====== ============================================================ 2311 2312(These error codes are indicative only: do not rely on a specific error 2313code being returned in a specific situation.) 2314 2315:: 2316 2317 struct kvm_one_reg { 2318 __u64 id; 2319 __u64 addr; 2320 }; 2321 2322Using this ioctl, a single vcpu register can be set to a specific value 2323defined by user space with the passed in struct kvm_one_reg, where id 2324refers to the register identifier as described below and addr is a pointer 2325to a variable with the respective size. There can be architecture agnostic 2326and architecture specific registers. Each have their own range of operation 2327and their own constants and width. To keep track of the implemented 2328registers, find a list below: 2329 2330 ======= =============================== ============ 2331 Arch Register Width (bits) 2332 ======= =============================== ============ 2333 PPC KVM_REG_PPC_HIOR 64 2334 PPC KVM_REG_PPC_IAC1 64 2335 PPC KVM_REG_PPC_IAC2 64 2336 PPC KVM_REG_PPC_IAC3 64 2337 PPC KVM_REG_PPC_IAC4 64 2338 PPC KVM_REG_PPC_DAC1 64 2339 PPC KVM_REG_PPC_DAC2 64 2340 PPC KVM_REG_PPC_DABR 64 2341 PPC KVM_REG_PPC_DSCR 64 2342 PPC KVM_REG_PPC_PURR 64 2343 PPC KVM_REG_PPC_SPURR 64 2344 PPC KVM_REG_PPC_DAR 64 2345 PPC KVM_REG_PPC_DSISR 32 2346 PPC KVM_REG_PPC_AMR 64 2347 PPC KVM_REG_PPC_UAMOR 64 2348 PPC KVM_REG_PPC_MMCR0 64 2349 PPC KVM_REG_PPC_MMCR1 64 2350 PPC KVM_REG_PPC_MMCRA 64 2351 PPC KVM_REG_PPC_MMCR2 64 2352 PPC KVM_REG_PPC_MMCRS 64 2353 PPC KVM_REG_PPC_MMCR3 64 2354 PPC KVM_REG_PPC_SIAR 64 2355 PPC KVM_REG_PPC_SDAR 64 2356 PPC KVM_REG_PPC_SIER 64 2357 PPC KVM_REG_PPC_SIER2 64 2358 PPC KVM_REG_PPC_SIER3 64 2359 PPC KVM_REG_PPC_PMC1 32 2360 PPC KVM_REG_PPC_PMC2 32 2361 PPC KVM_REG_PPC_PMC3 32 2362 PPC KVM_REG_PPC_PMC4 32 2363 PPC KVM_REG_PPC_PMC5 32 2364 PPC KVM_REG_PPC_PMC6 32 2365 PPC KVM_REG_PPC_PMC7 32 2366 PPC KVM_REG_PPC_PMC8 32 2367 PPC KVM_REG_PPC_FPR0 64 2368 ... 2369 PPC KVM_REG_PPC_FPR31 64 2370 PPC KVM_REG_PPC_VR0 128 2371 ... 2372 PPC KVM_REG_PPC_VR31 128 2373 PPC KVM_REG_PPC_VSR0 128 2374 ... 2375 PPC KVM_REG_PPC_VSR31 128 2376 PPC KVM_REG_PPC_FPSCR 64 2377 PPC KVM_REG_PPC_VSCR 32 2378 PPC KVM_REG_PPC_VPA_ADDR 64 2379 PPC KVM_REG_PPC_VPA_SLB 128 2380 PPC KVM_REG_PPC_VPA_DTL 128 2381 PPC KVM_REG_PPC_EPCR 32 2382 PPC KVM_REG_PPC_EPR 32 2383 PPC KVM_REG_PPC_TCR 32 2384 PPC KVM_REG_PPC_TSR 32 2385 PPC KVM_REG_PPC_OR_TSR 32 2386 PPC KVM_REG_PPC_CLEAR_TSR 32 2387 PPC KVM_REG_PPC_MAS0 32 2388 PPC KVM_REG_PPC_MAS1 32 2389 PPC KVM_REG_PPC_MAS2 64 2390 PPC KVM_REG_PPC_MAS7_3 64 2391 PPC KVM_REG_PPC_MAS4 32 2392 PPC KVM_REG_PPC_MAS6 32 2393 PPC KVM_REG_PPC_MMUCFG 32 2394 PPC KVM_REG_PPC_TLB0CFG 32 2395 PPC KVM_REG_PPC_TLB1CFG 32 2396 PPC KVM_REG_PPC_TLB2CFG 32 2397 PPC KVM_REG_PPC_TLB3CFG 32 2398 PPC KVM_REG_PPC_TLB0PS 32 2399 PPC KVM_REG_PPC_TLB1PS 32 2400 PPC KVM_REG_PPC_TLB2PS 32 2401 PPC KVM_REG_PPC_TLB3PS 32 2402 PPC KVM_REG_PPC_EPTCFG 32 2403 PPC KVM_REG_PPC_ICP_STATE 64 2404 PPC KVM_REG_PPC_VP_STATE 128 2405 PPC KVM_REG_PPC_TB_OFFSET 64 2406 PPC KVM_REG_PPC_SPMC1 32 2407 PPC KVM_REG_PPC_SPMC2 32 2408 PPC KVM_REG_PPC_IAMR 64 2409 PPC KVM_REG_PPC_TFHAR 64 2410 PPC KVM_REG_PPC_TFIAR 64 2411 PPC KVM_REG_PPC_TEXASR 64 2412 PPC KVM_REG_PPC_FSCR 64 2413 PPC KVM_REG_PPC_PSPB 32 2414 PPC KVM_REG_PPC_EBBHR 64 2415 PPC KVM_REG_PPC_EBBRR 64 2416 PPC KVM_REG_PPC_BESCR 64 2417 PPC KVM_REG_PPC_TAR 64 2418 PPC KVM_REG_PPC_DPDES 64 2419 PPC KVM_REG_PPC_DAWR 64 2420 PPC KVM_REG_PPC_DAWRX 64 2421 PPC KVM_REG_PPC_CIABR 64 2422 PPC KVM_REG_PPC_IC 64 2423 PPC KVM_REG_PPC_VTB 64 2424 PPC KVM_REG_PPC_CSIGR 64 2425 PPC KVM_REG_PPC_TACR 64 2426 PPC KVM_REG_PPC_TCSCR 64 2427 PPC KVM_REG_PPC_PID 64 2428 PPC KVM_REG_PPC_ACOP 64 2429 PPC KVM_REG_PPC_VRSAVE 32 2430 PPC KVM_REG_PPC_LPCR 32 2431 PPC KVM_REG_PPC_LPCR_64 64 2432 PPC KVM_REG_PPC_PPR 64 2433 PPC KVM_REG_PPC_ARCH_COMPAT 32 2434 PPC KVM_REG_PPC_DABRX 32 2435 PPC KVM_REG_PPC_WORT 64 2436 PPC KVM_REG_PPC_SPRG9 64 2437 PPC KVM_REG_PPC_DBSR 32 2438 PPC KVM_REG_PPC_TIDR 64 2439 PPC KVM_REG_PPC_PSSCR 64 2440 PPC KVM_REG_PPC_DEC_EXPIRY 64 2441 PPC KVM_REG_PPC_PTCR 64 2442 PPC KVM_REG_PPC_HASHKEYR 64 2443 PPC KVM_REG_PPC_HASHPKEYR 64 2444 PPC KVM_REG_PPC_DAWR1 64 2445 PPC KVM_REG_PPC_DAWRX1 64 2446 PPC KVM_REG_PPC_DEXCR 64 2447 PPC KVM_REG_PPC_TM_GPR0 64 2448 ... 2449 PPC KVM_REG_PPC_TM_GPR31 64 2450 PPC KVM_REG_PPC_TM_VSR0 128 2451 ... 2452 PPC KVM_REG_PPC_TM_VSR63 128 2453 PPC KVM_REG_PPC_TM_CR 64 2454 PPC KVM_REG_PPC_TM_LR 64 2455 PPC KVM_REG_PPC_TM_CTR 64 2456 PPC KVM_REG_PPC_TM_FPSCR 64 2457 PPC KVM_REG_PPC_TM_AMR 64 2458 PPC KVM_REG_PPC_TM_PPR 64 2459 PPC KVM_REG_PPC_TM_VRSAVE 64 2460 PPC KVM_REG_PPC_TM_VSCR 32 2461 PPC KVM_REG_PPC_TM_DSCR 64 2462 PPC KVM_REG_PPC_TM_TAR 64 2463 PPC KVM_REG_PPC_TM_XER 64 2464 2465 MIPS KVM_REG_MIPS_R0 64 2466 ... 2467 MIPS KVM_REG_MIPS_R31 64 2468 MIPS KVM_REG_MIPS_HI 64 2469 MIPS KVM_REG_MIPS_LO 64 2470 MIPS KVM_REG_MIPS_PC 64 2471 MIPS KVM_REG_MIPS_CP0_INDEX 32 2472 MIPS KVM_REG_MIPS_CP0_ENTRYLO0 64 2473 MIPS KVM_REG_MIPS_CP0_ENTRYLO1 64 2474 MIPS KVM_REG_MIPS_CP0_CONTEXT 64 2475 MIPS KVM_REG_MIPS_CP0_CONTEXTCONFIG 32 2476 MIPS KVM_REG_MIPS_CP0_USERLOCAL 64 2477 MIPS KVM_REG_MIPS_CP0_XCONTEXTCONFIG 64 2478 MIPS KVM_REG_MIPS_CP0_PAGEMASK 32 2479 MIPS KVM_REG_MIPS_CP0_PAGEGRAIN 32 2480 MIPS KVM_REG_MIPS_CP0_SEGCTL0 64 2481 MIPS KVM_REG_MIPS_CP0_SEGCTL1 64 2482 MIPS KVM_REG_MIPS_CP0_SEGCTL2 64 2483 MIPS KVM_REG_MIPS_CP0_PWBASE 64 2484 MIPS KVM_REG_MIPS_CP0_PWFIELD 64 2485 MIPS KVM_REG_MIPS_CP0_PWSIZE 64 2486 MIPS KVM_REG_MIPS_CP0_WIRED 32 2487 MIPS KVM_REG_MIPS_CP0_PWCTL 32 2488 MIPS KVM_REG_MIPS_CP0_HWRENA 32 2489 MIPS KVM_REG_MIPS_CP0_BADVADDR 64 2490 MIPS KVM_REG_MIPS_CP0_BADINSTR 32 2491 MIPS KVM_REG_MIPS_CP0_BADINSTRP 32 2492 MIPS KVM_REG_MIPS_CP0_COUNT 32 2493 MIPS KVM_REG_MIPS_CP0_ENTRYHI 64 2494 MIPS KVM_REG_MIPS_CP0_COMPARE 32 2495 MIPS KVM_REG_MIPS_CP0_STATUS 32 2496 MIPS KVM_REG_MIPS_CP0_INTCTL 32 2497 MIPS KVM_REG_MIPS_CP0_CAUSE 32 2498 MIPS KVM_REG_MIPS_CP0_EPC 64 2499 MIPS KVM_REG_MIPS_CP0_PRID 32 2500 MIPS KVM_REG_MIPS_CP0_EBASE 64 2501 MIPS KVM_REG_MIPS_CP0_CONFIG 32 2502 MIPS KVM_REG_MIPS_CP0_CONFIG1 32 2503 MIPS KVM_REG_MIPS_CP0_CONFIG2 32 2504 MIPS KVM_REG_MIPS_CP0_CONFIG3 32 2505 MIPS KVM_REG_MIPS_CP0_CONFIG4 32 2506 MIPS KVM_REG_MIPS_CP0_CONFIG5 32 2507 MIPS KVM_REG_MIPS_CP0_CONFIG7 32 2508 MIPS KVM_REG_MIPS_CP0_XCONTEXT 64 2509 MIPS KVM_REG_MIPS_CP0_ERROREPC 64 2510 MIPS KVM_REG_MIPS_CP0_KSCRATCH1 64 2511 MIPS KVM_REG_MIPS_CP0_KSCRATCH2 64 2512 MIPS KVM_REG_MIPS_CP0_KSCRATCH3 64 2513 MIPS KVM_REG_MIPS_CP0_KSCRATCH4 64 2514 MIPS KVM_REG_MIPS_CP0_KSCRATCH5 64 2515 MIPS KVM_REG_MIPS_CP0_KSCRATCH6 64 2516 MIPS KVM_REG_MIPS_CP0_MAAR(0..63) 64 2517 MIPS KVM_REG_MIPS_COUNT_CTL 64 2518 MIPS KVM_REG_MIPS_COUNT_RESUME 64 2519 MIPS KVM_REG_MIPS_COUNT_HZ 64 2520 MIPS KVM_REG_MIPS_FPR_32(0..31) 32 2521 MIPS KVM_REG_MIPS_FPR_64(0..31) 64 2522 MIPS KVM_REG_MIPS_VEC_128(0..31) 128 2523 MIPS KVM_REG_MIPS_FCR_IR 32 2524 MIPS KVM_REG_MIPS_FCR_CSR 32 2525 MIPS KVM_REG_MIPS_MSA_IR 32 2526 MIPS KVM_REG_MIPS_MSA_CSR 32 2527 ======= =============================== ============ 2528 2529ARM registers are mapped using the lower 32 bits. The upper 16 of that 2530is the register group type, or coprocessor number: 2531 2532ARM core registers have the following id bit patterns:: 2533 2534 0x4020 0000 0010 <index into the kvm_regs struct:16> 2535 2536ARM 32-bit CP15 registers have the following id bit patterns:: 2537 2538 0x4020 0000 000F <zero:1> <crn:4> <crm:4> <opc1:4> <opc2:3> 2539 2540ARM 64-bit CP15 registers have the following id bit patterns:: 2541 2542 0x4030 0000 000F <zero:1> <zero:4> <crm:4> <opc1:4> <zero:3> 2543 2544ARM CCSIDR registers are demultiplexed by CSSELR value:: 2545 2546 0x4020 0000 0011 00 <csselr:8> 2547 2548ARM 32-bit VFP control registers have the following id bit patterns:: 2549 2550 0x4020 0000 0012 1 <regno:12> 2551 2552ARM 64-bit FP registers have the following id bit patterns:: 2553 2554 0x4030 0000 0012 0 <regno:12> 2555 2556ARM firmware pseudo-registers have the following bit pattern:: 2557 2558 0x4030 0000 0014 <regno:16> 2559 2560 2561arm64 registers are mapped using the lower 32 bits. The upper 16 of 2562that is the register group type, or coprocessor number: 2563 2564arm64 core/FP-SIMD registers have the following id bit patterns. Note 2565that the size of the access is variable, as the kvm_regs structure 2566contains elements ranging from 32 to 128 bits. The index is a 32bit 2567value in the kvm_regs structure seen as a 32bit array:: 2568 2569 0x60x0 0000 0010 <index into the kvm_regs struct:16> 2570 2571Specifically: 2572 2573======================= ========= ===== ======================================= 2574 Encoding Register Bits kvm_regs member 2575======================= ========= ===== ======================================= 2576 0x6030 0000 0010 0000 X0 64 regs.regs[0] 2577 0x6030 0000 0010 0002 X1 64 regs.regs[1] 2578 ... 2579 0x6030 0000 0010 003c X30 64 regs.regs[30] 2580 0x6030 0000 0010 003e SP 64 regs.sp 2581 0x6030 0000 0010 0040 PC 64 regs.pc 2582 0x6030 0000 0010 0042 PSTATE 64 regs.pstate 2583 0x6030 0000 0010 0044 SP_EL1 64 sp_el1 2584 0x6030 0000 0010 0046 ELR_EL1 64 elr_el1 2585 0x6030 0000 0010 0048 SPSR_EL1 64 spsr[KVM_SPSR_EL1] (alias SPSR_SVC) 2586 0x6030 0000 0010 004a SPSR_ABT 64 spsr[KVM_SPSR_ABT] 2587 0x6030 0000 0010 004c SPSR_UND 64 spsr[KVM_SPSR_UND] 2588 0x6030 0000 0010 004e SPSR_IRQ 64 spsr[KVM_SPSR_IRQ] 2589 0x6060 0000 0010 0050 SPSR_FIQ 64 spsr[KVM_SPSR_FIQ] 2590 0x6040 0000 0010 0054 V0 128 fp_regs.vregs[0] [1]_ 2591 0x6040 0000 0010 0058 V1 128 fp_regs.vregs[1] [1]_ 2592 ... 2593 0x6040 0000 0010 00d0 V31 128 fp_regs.vregs[31] [1]_ 2594 0x6020 0000 0010 00d4 FPSR 32 fp_regs.fpsr 2595 0x6020 0000 0010 00d5 FPCR 32 fp_regs.fpcr 2596======================= ========= ===== ======================================= 2597 2598.. [1] These encodings are not accepted for SVE-enabled vcpus. See 2599 KVM_ARM_VCPU_INIT. 2600 2601 The equivalent register content can be accessed via bits [127:0] of 2602 the corresponding SVE Zn registers instead for vcpus that have SVE 2603 enabled (see below). 2604 2605arm64 CCSIDR registers are demultiplexed by CSSELR value:: 2606 2607 0x6020 0000 0011 00 <csselr:8> 2608 2609arm64 system registers have the following id bit patterns:: 2610 2611 0x6030 0000 0013 <op0:2> <op1:3> <crn:4> <crm:4> <op2:3> 2612 2613.. warning:: 2614 2615 Two system register IDs do not follow the specified pattern. These 2616 are KVM_REG_ARM_TIMER_CVAL and KVM_REG_ARM_TIMER_CNT, which map to 2617 system registers CNTV_CVAL_EL0 and CNTVCT_EL0 respectively. These 2618 two had their values accidentally swapped, which means TIMER_CVAL is 2619 derived from the register encoding for CNTVCT_EL0 and TIMER_CNT is 2620 derived from the register encoding for CNTV_CVAL_EL0. As this is 2621 API, it must remain this way. 2622 2623arm64 firmware pseudo-registers have the following bit pattern:: 2624 2625 0x6030 0000 0014 <regno:16> 2626 2627arm64 SVE registers have the following bit patterns:: 2628 2629 0x6080 0000 0015 00 <n:5> <slice:5> Zn bits[2048*slice + 2047 : 2048*slice] 2630 0x6050 0000 0015 04 <n:4> <slice:5> Pn bits[256*slice + 255 : 256*slice] 2631 0x6050 0000 0015 060 <slice:5> FFR bits[256*slice + 255 : 256*slice] 2632 0x6060 0000 0015 ffff KVM_REG_ARM64_SVE_VLS pseudo-register 2633 2634Access to register IDs where 2048 * slice >= 128 * max_vq will fail with 2635ENOENT. max_vq is the vcpu's maximum supported vector length in 128-bit 2636quadwords: see [2]_ below. 2637 2638These registers are only accessible on vcpus for which SVE is enabled. 2639See KVM_ARM_VCPU_INIT for details. 2640 2641In addition, except for KVM_REG_ARM64_SVE_VLS, these registers are not 2642accessible until the vcpu's SVE configuration has been finalized 2643using KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE). See KVM_ARM_VCPU_INIT 2644and KVM_ARM_VCPU_FINALIZE for more information about this procedure. 2645 2646KVM_REG_ARM64_SVE_VLS is a pseudo-register that allows the set of vector 2647lengths supported by the vcpu to be discovered and configured by 2648userspace. When transferred to or from user memory via KVM_GET_ONE_REG 2649or KVM_SET_ONE_REG, the value of this register is of type 2650__u64[KVM_ARM64_SVE_VLS_WORDS], and encodes the set of vector lengths as 2651follows:: 2652 2653 __u64 vector_lengths[KVM_ARM64_SVE_VLS_WORDS]; 2654 2655 if (vq >= SVE_VQ_MIN && vq <= SVE_VQ_MAX && 2656 ((vector_lengths[(vq - KVM_ARM64_SVE_VQ_MIN) / 64] >> 2657 ((vq - KVM_ARM64_SVE_VQ_MIN) % 64)) & 1)) 2658 /* Vector length vq * 16 bytes supported */ 2659 else 2660 /* Vector length vq * 16 bytes not supported */ 2661 2662.. [2] The maximum value vq for which the above condition is true is 2663 max_vq. This is the maximum vector length available to the guest on 2664 this vcpu, and determines which register slices are visible through 2665 this ioctl interface. 2666 2667(See Documentation/arch/arm64/sve.rst for an explanation of the "vq" 2668nomenclature.) 2669 2670KVM_REG_ARM64_SVE_VLS is only accessible after KVM_ARM_VCPU_INIT. 2671KVM_ARM_VCPU_INIT initialises it to the best set of vector lengths that 2672the host supports. 2673 2674Userspace may subsequently modify it if desired until the vcpu's SVE 2675configuration is finalized using KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE). 2676 2677Apart from simply removing all vector lengths from the host set that 2678exceed some value, support for arbitrarily chosen sets of vector lengths 2679is hardware-dependent and may not be available. Attempting to configure 2680an invalid set of vector lengths via KVM_SET_ONE_REG will fail with 2681EINVAL. 2682 2683After the vcpu's SVE configuration is finalized, further attempts to 2684write this register will fail with EPERM. 2685 2686arm64 bitmap feature firmware pseudo-registers have the following bit pattern:: 2687 2688 0x6030 0000 0016 <regno:16> 2689 2690The bitmap feature firmware registers exposes the hypercall services that 2691are available for userspace to configure. The set bits corresponds to the 2692services that are available for the guests to access. By default, KVM 2693sets all the supported bits during VM initialization. The userspace can 2694discover the available services via KVM_GET_ONE_REG, and write back the 2695bitmap corresponding to the features that it wishes guests to see via 2696KVM_SET_ONE_REG. 2697 2698Note: These registers are immutable once any of the vCPUs of the VM has 2699run at least once. A KVM_SET_ONE_REG in such a scenario will return 2700a -EBUSY to userspace. 2701 2702(See Documentation/virt/kvm/arm/hypercalls.rst for more details.) 2703 2704 2705MIPS registers are mapped using the lower 32 bits. The upper 16 of that is 2706the register group type: 2707 2708MIPS core registers (see above) have the following id bit patterns:: 2709 2710 0x7030 0000 0000 <reg:16> 2711 2712MIPS CP0 registers (see KVM_REG_MIPS_CP0_* above) have the following id bit 2713patterns depending on whether they're 32-bit or 64-bit registers:: 2714 2715 0x7020 0000 0001 00 <reg:5> <sel:3> (32-bit) 2716 0x7030 0000 0001 00 <reg:5> <sel:3> (64-bit) 2717 2718Note: KVM_REG_MIPS_CP0_ENTRYLO0 and KVM_REG_MIPS_CP0_ENTRYLO1 are the MIPS64 2719versions of the EntryLo registers regardless of the word size of the host 2720hardware, host kernel, guest, and whether XPA is present in the guest, i.e. 2721with the RI and XI bits (if they exist) in bits 63 and 62 respectively, and 2722the PFNX field starting at bit 30. 2723 2724MIPS MAARs (see KVM_REG_MIPS_CP0_MAAR(*) above) have the following id bit 2725patterns:: 2726 2727 0x7030 0000 0001 01 <reg:8> 2728 2729MIPS KVM control registers (see above) have the following id bit patterns:: 2730 2731 0x7030 0000 0002 <reg:16> 2732 2733MIPS FPU registers (see KVM_REG_MIPS_FPR_{32,64}() above) have the following 2734id bit patterns depending on the size of the register being accessed. They are 2735always accessed according to the current guest FPU mode (Status.FR and 2736Config5.FRE), i.e. as the guest would see them, and they become unpredictable 2737if the guest FPU mode is changed. MIPS SIMD Architecture (MSA) vector 2738registers (see KVM_REG_MIPS_VEC_128() above) have similar patterns as they 2739overlap the FPU registers:: 2740 2741 0x7020 0000 0003 00 <0:3> <reg:5> (32-bit FPU registers) 2742 0x7030 0000 0003 00 <0:3> <reg:5> (64-bit FPU registers) 2743 0x7040 0000 0003 00 <0:3> <reg:5> (128-bit MSA vector registers) 2744 2745MIPS FPU control registers (see KVM_REG_MIPS_FCR_{IR,CSR} above) have the 2746following id bit patterns:: 2747 2748 0x7020 0000 0003 01 <0:3> <reg:5> 2749 2750MIPS MSA control registers (see KVM_REG_MIPS_MSA_{IR,CSR} above) have the 2751following id bit patterns:: 2752 2753 0x7020 0000 0003 02 <0:3> <reg:5> 2754 2755RISC-V registers are mapped using the lower 32 bits. The upper 8 bits of 2756that is the register group type. 2757 2758RISC-V config registers are meant for configuring a Guest VCPU and it has 2759the following id bit patterns:: 2760 2761 0x8020 0000 01 <index into the kvm_riscv_config struct:24> (32bit Host) 2762 0x8030 0000 01 <index into the kvm_riscv_config struct:24> (64bit Host) 2763 2764Following are the RISC-V config registers: 2765 2766======================= ========= ============================================= 2767 Encoding Register Description 2768======================= ========= ============================================= 2769 0x80x0 0000 0100 0000 isa ISA feature bitmap of Guest VCPU 2770======================= ========= ============================================= 2771 2772The isa config register can be read anytime but can only be written before 2773a Guest VCPU runs. It will have ISA feature bits matching underlying host 2774set by default. 2775 2776RISC-V core registers represent the general execution state of a Guest VCPU 2777and it has the following id bit patterns:: 2778 2779 0x8020 0000 02 <index into the kvm_riscv_core struct:24> (32bit Host) 2780 0x8030 0000 02 <index into the kvm_riscv_core struct:24> (64bit Host) 2781 2782Following are the RISC-V core registers: 2783 2784======================= ========= ============================================= 2785 Encoding Register Description 2786======================= ========= ============================================= 2787 0x80x0 0000 0200 0000 regs.pc Program counter 2788 0x80x0 0000 0200 0001 regs.ra Return address 2789 0x80x0 0000 0200 0002 regs.sp Stack pointer 2790 0x80x0 0000 0200 0003 regs.gp Global pointer 2791 0x80x0 0000 0200 0004 regs.tp Task pointer 2792 0x80x0 0000 0200 0005 regs.t0 Caller saved register 0 2793 0x80x0 0000 0200 0006 regs.t1 Caller saved register 1 2794 0x80x0 0000 0200 0007 regs.t2 Caller saved register 2 2795 0x80x0 0000 0200 0008 regs.s0 Callee saved register 0 2796 0x80x0 0000 0200 0009 regs.s1 Callee saved register 1 2797 0x80x0 0000 0200 000a regs.a0 Function argument (or return value) 0 2798 0x80x0 0000 0200 000b regs.a1 Function argument (or return value) 1 2799 0x80x0 0000 0200 000c regs.a2 Function argument 2 2800 0x80x0 0000 0200 000d regs.a3 Function argument 3 2801 0x80x0 0000 0200 000e regs.a4 Function argument 4 2802 0x80x0 0000 0200 000f regs.a5 Function argument 5 2803 0x80x0 0000 0200 0010 regs.a6 Function argument 6 2804 0x80x0 0000 0200 0011 regs.a7 Function argument 7 2805 0x80x0 0000 0200 0012 regs.s2 Callee saved register 2 2806 0x80x0 0000 0200 0013 regs.s3 Callee saved register 3 2807 0x80x0 0000 0200 0014 regs.s4 Callee saved register 4 2808 0x80x0 0000 0200 0015 regs.s5 Callee saved register 5 2809 0x80x0 0000 0200 0016 regs.s6 Callee saved register 6 2810 0x80x0 0000 0200 0017 regs.s7 Callee saved register 7 2811 0x80x0 0000 0200 0018 regs.s8 Callee saved register 8 2812 0x80x0 0000 0200 0019 regs.s9 Callee saved register 9 2813 0x80x0 0000 0200 001a regs.s10 Callee saved register 10 2814 0x80x0 0000 0200 001b regs.s11 Callee saved register 11 2815 0x80x0 0000 0200 001c regs.t3 Caller saved register 3 2816 0x80x0 0000 0200 001d regs.t4 Caller saved register 4 2817 0x80x0 0000 0200 001e regs.t5 Caller saved register 5 2818 0x80x0 0000 0200 001f regs.t6 Caller saved register 6 2819 0x80x0 0000 0200 0020 mode Privilege mode (1 = S-mode or 0 = U-mode) 2820======================= ========= ============================================= 2821 2822RISC-V csr registers represent the supervisor mode control/status registers 2823of a Guest VCPU and it has the following id bit patterns:: 2824 2825 0x8020 0000 03 <index into the kvm_riscv_csr struct:24> (32bit Host) 2826 0x8030 0000 03 <index into the kvm_riscv_csr struct:24> (64bit Host) 2827 2828Following are the RISC-V csr registers: 2829 2830======================= ========= ============================================= 2831 Encoding Register Description 2832======================= ========= ============================================= 2833 0x80x0 0000 0300 0000 sstatus Supervisor status 2834 0x80x0 0000 0300 0001 sie Supervisor interrupt enable 2835 0x80x0 0000 0300 0002 stvec Supervisor trap vector base 2836 0x80x0 0000 0300 0003 sscratch Supervisor scratch register 2837 0x80x0 0000 0300 0004 sepc Supervisor exception program counter 2838 0x80x0 0000 0300 0005 scause Supervisor trap cause 2839 0x80x0 0000 0300 0006 stval Supervisor bad address or instruction 2840 0x80x0 0000 0300 0007 sip Supervisor interrupt pending 2841 0x80x0 0000 0300 0008 satp Supervisor address translation and protection 2842======================= ========= ============================================= 2843 2844RISC-V timer registers represent the timer state of a Guest VCPU and it has 2845the following id bit patterns:: 2846 2847 0x8030 0000 04 <index into the kvm_riscv_timer struct:24> 2848 2849Following are the RISC-V timer registers: 2850 2851======================= ========= ============================================= 2852 Encoding Register Description 2853======================= ========= ============================================= 2854 0x8030 0000 0400 0000 frequency Time base frequency (read-only) 2855 0x8030 0000 0400 0001 time Time value visible to Guest 2856 0x8030 0000 0400 0002 compare Time compare programmed by Guest 2857 0x8030 0000 0400 0003 state Time compare state (1 = ON or 0 = OFF) 2858======================= ========= ============================================= 2859 2860RISC-V F-extension registers represent the single precision floating point 2861state of a Guest VCPU and it has the following id bit patterns:: 2862 2863 0x8020 0000 05 <index into the __riscv_f_ext_state struct:24> 2864 2865Following are the RISC-V F-extension registers: 2866 2867======================= ========= ============================================= 2868 Encoding Register Description 2869======================= ========= ============================================= 2870 0x8020 0000 0500 0000 f[0] Floating point register 0 2871 ... 2872 0x8020 0000 0500 001f f[31] Floating point register 31 2873 0x8020 0000 0500 0020 fcsr Floating point control and status register 2874======================= ========= ============================================= 2875 2876RISC-V D-extension registers represent the double precision floating point 2877state of a Guest VCPU and it has the following id bit patterns:: 2878 2879 0x8020 0000 06 <index into the __riscv_d_ext_state struct:24> (fcsr) 2880 0x8030 0000 06 <index into the __riscv_d_ext_state struct:24> (non-fcsr) 2881 2882Following are the RISC-V D-extension registers: 2883 2884======================= ========= ============================================= 2885 Encoding Register Description 2886======================= ========= ============================================= 2887 0x8030 0000 0600 0000 f[0] Floating point register 0 2888 ... 2889 0x8030 0000 0600 001f f[31] Floating point register 31 2890 0x8020 0000 0600 0020 fcsr Floating point control and status register 2891======================= ========= ============================================= 2892 2893LoongArch registers are mapped using the lower 32 bits. The upper 16 bits of 2894that is the register group type. 2895 2896LoongArch csr registers are used to control guest cpu or get status of guest 2897cpu, and they have the following id bit patterns:: 2898 2899 0x9030 0000 0001 00 <reg:5> <sel:3> (64-bit) 2900 2901LoongArch KVM control registers are used to implement some new defined functions 2902such as set vcpu counter or reset vcpu, and they have the following id bit patterns:: 2903 2904 0x9030 0000 0002 <reg:16> 2905 2906 29074.69 KVM_GET_ONE_REG 2908-------------------- 2909 2910:Capability: KVM_CAP_ONE_REG 2911:Architectures: all 2912:Type: vcpu ioctl 2913:Parameters: struct kvm_one_reg (in and out) 2914:Returns: 0 on success, negative value on failure 2915 2916Errors include: 2917 2918 ======== ============================================================ 2919 ENOENT no such register 2920 EINVAL invalid register ID, or no such register or used with VMs in 2921 protected virtualization mode on s390 2922 EPERM (arm64) register access not allowed before vcpu finalization 2923 ======== ============================================================ 2924 2925(These error codes are indicative only: do not rely on a specific error 2926code being returned in a specific situation.) 2927 2928This ioctl allows to receive the value of a single register implemented 2929in a vcpu. The register to read is indicated by the "id" field of the 2930kvm_one_reg struct passed in. On success, the register value can be found 2931at the memory location pointed to by "addr". 2932 2933The list of registers accessible using this interface is identical to the 2934list in 4.68. 2935 2936 29374.70 KVM_KVMCLOCK_CTRL 2938---------------------- 2939 2940:Capability: KVM_CAP_KVMCLOCK_CTRL 2941:Architectures: Any that implement pvclocks (currently x86 only) 2942:Type: vcpu ioctl 2943:Parameters: None 2944:Returns: 0 on success, -1 on error 2945 2946This ioctl sets a flag accessible to the guest indicating that the specified 2947vCPU has been paused by the host userspace. 2948 2949The host will set a flag in the pvclock structure that is checked from the 2950soft lockup watchdog. The flag is part of the pvclock structure that is 2951shared between guest and host, specifically the second bit of the flags 2952field of the pvclock_vcpu_time_info structure. It will be set exclusively by 2953the host and read/cleared exclusively by the guest. The guest operation of 2954checking and clearing the flag must be an atomic operation so 2955load-link/store-conditional, or equivalent must be used. There are two cases 2956where the guest will clear the flag: when the soft lockup watchdog timer resets 2957itself or when a soft lockup is detected. This ioctl can be called any time 2958after pausing the vcpu, but before it is resumed. 2959 2960 29614.71 KVM_SIGNAL_MSI 2962------------------- 2963 2964:Capability: KVM_CAP_SIGNAL_MSI 2965:Architectures: x86 arm64 2966:Type: vm ioctl 2967:Parameters: struct kvm_msi (in) 2968:Returns: >0 on delivery, 0 if guest blocked the MSI, and -1 on error 2969 2970Directly inject a MSI message. Only valid with in-kernel irqchip that handles 2971MSI messages. 2972 2973:: 2974 2975 struct kvm_msi { 2976 __u32 address_lo; 2977 __u32 address_hi; 2978 __u32 data; 2979 __u32 flags; 2980 __u32 devid; 2981 __u8 pad[12]; 2982 }; 2983 2984flags: 2985 KVM_MSI_VALID_DEVID: devid contains a valid value. The per-VM 2986 KVM_CAP_MSI_DEVID capability advertises the requirement to provide 2987 the device ID. If this capability is not available, userspace 2988 should never set the KVM_MSI_VALID_DEVID flag as the ioctl might fail. 2989 2990If KVM_MSI_VALID_DEVID is set, devid contains a unique device identifier 2991for the device that wrote the MSI message. For PCI, this is usually a 2992BFD identifier in the lower 16 bits. 2993 2994On x86, address_hi is ignored unless the KVM_X2APIC_API_USE_32BIT_IDS 2995feature of KVM_CAP_X2APIC_API capability is enabled. If it is enabled, 2996address_hi bits 31-8 provide bits 31-8 of the destination id. Bits 7-0 of 2997address_hi must be zero. 2998 2999 30004.71 KVM_CREATE_PIT2 3001-------------------- 3002 3003:Capability: KVM_CAP_PIT2 3004:Architectures: x86 3005:Type: vm ioctl 3006:Parameters: struct kvm_pit_config (in) 3007:Returns: 0 on success, -1 on error 3008 3009Creates an in-kernel device model for the i8254 PIT. This call is only valid 3010after enabling in-kernel irqchip support via KVM_CREATE_IRQCHIP. The following 3011parameters have to be passed:: 3012 3013 struct kvm_pit_config { 3014 __u32 flags; 3015 __u32 pad[15]; 3016 }; 3017 3018Valid flags are:: 3019 3020 #define KVM_PIT_SPEAKER_DUMMY 1 /* emulate speaker port stub */ 3021 3022PIT timer interrupts may use a per-VM kernel thread for injection. If it 3023exists, this thread will have a name of the following pattern:: 3024 3025 kvm-pit/<owner-process-pid> 3026 3027When running a guest with elevated priorities, the scheduling parameters of 3028this thread may have to be adjusted accordingly. 3029 3030This IOCTL replaces the obsolete KVM_CREATE_PIT. 3031 3032 30334.72 KVM_GET_PIT2 3034----------------- 3035 3036:Capability: KVM_CAP_PIT_STATE2 3037:Architectures: x86 3038:Type: vm ioctl 3039:Parameters: struct kvm_pit_state2 (out) 3040:Returns: 0 on success, -1 on error 3041 3042Retrieves the state of the in-kernel PIT model. Only valid after 3043KVM_CREATE_PIT2. The state is returned in the following structure:: 3044 3045 struct kvm_pit_state2 { 3046 struct kvm_pit_channel_state channels[3]; 3047 __u32 flags; 3048 __u32 reserved[9]; 3049 }; 3050 3051Valid flags are:: 3052 3053 /* disable PIT in HPET legacy mode */ 3054 #define KVM_PIT_FLAGS_HPET_LEGACY 0x00000001 3055 /* speaker port data bit enabled */ 3056 #define KVM_PIT_FLAGS_SPEAKER_DATA_ON 0x00000002 3057 3058This IOCTL replaces the obsolete KVM_GET_PIT. 3059 3060 30614.73 KVM_SET_PIT2 3062----------------- 3063 3064:Capability: KVM_CAP_PIT_STATE2 3065:Architectures: x86 3066:Type: vm ioctl 3067:Parameters: struct kvm_pit_state2 (in) 3068:Returns: 0 on success, -1 on error 3069 3070Sets the state of the in-kernel PIT model. Only valid after KVM_CREATE_PIT2. 3071See KVM_GET_PIT2 for details on struct kvm_pit_state2. 3072 3073This IOCTL replaces the obsolete KVM_SET_PIT. 3074 3075 30764.74 KVM_PPC_GET_SMMU_INFO 3077-------------------------- 3078 3079:Capability: KVM_CAP_PPC_GET_SMMU_INFO 3080:Architectures: powerpc 3081:Type: vm ioctl 3082:Parameters: None 3083:Returns: 0 on success, -1 on error 3084 3085This populates and returns a structure describing the features of 3086the "Server" class MMU emulation supported by KVM. 3087This can in turn be used by userspace to generate the appropriate 3088device-tree properties for the guest operating system. 3089 3090The structure contains some global information, followed by an 3091array of supported segment page sizes:: 3092 3093 struct kvm_ppc_smmu_info { 3094 __u64 flags; 3095 __u32 slb_size; 3096 __u32 pad; 3097 struct kvm_ppc_one_seg_page_size sps[KVM_PPC_PAGE_SIZES_MAX_SZ]; 3098 }; 3099 3100The supported flags are: 3101 3102 - KVM_PPC_PAGE_SIZES_REAL: 3103 When that flag is set, guest page sizes must "fit" the backing 3104 store page sizes. When not set, any page size in the list can 3105 be used regardless of how they are backed by userspace. 3106 3107 - KVM_PPC_1T_SEGMENTS 3108 The emulated MMU supports 1T segments in addition to the 3109 standard 256M ones. 3110 3111 - KVM_PPC_NO_HASH 3112 This flag indicates that HPT guests are not supported by KVM, 3113 thus all guests must use radix MMU mode. 3114 3115The "slb_size" field indicates how many SLB entries are supported 3116 3117The "sps" array contains 8 entries indicating the supported base 3118page sizes for a segment in increasing order. Each entry is defined 3119as follow:: 3120 3121 struct kvm_ppc_one_seg_page_size { 3122 __u32 page_shift; /* Base page shift of segment (or 0) */ 3123 __u32 slb_enc; /* SLB encoding for BookS */ 3124 struct kvm_ppc_one_page_size enc[KVM_PPC_PAGE_SIZES_MAX_SZ]; 3125 }; 3126 3127An entry with a "page_shift" of 0 is unused. Because the array is 3128organized in increasing order, a lookup can stop when encountering 3129such an entry. 3130 3131The "slb_enc" field provides the encoding to use in the SLB for the 3132page size. The bits are in positions such as the value can directly 3133be OR'ed into the "vsid" argument of the slbmte instruction. 3134 3135The "enc" array is a list which for each of those segment base page 3136size provides the list of supported actual page sizes (which can be 3137only larger or equal to the base page size), along with the 3138corresponding encoding in the hash PTE. Similarly, the array is 31398 entries sorted by increasing sizes and an entry with a "0" shift 3140is an empty entry and a terminator:: 3141 3142 struct kvm_ppc_one_page_size { 3143 __u32 page_shift; /* Page shift (or 0) */ 3144 __u32 pte_enc; /* Encoding in the HPTE (>>12) */ 3145 }; 3146 3147The "pte_enc" field provides a value that can OR'ed into the hash 3148PTE's RPN field (ie, it needs to be shifted left by 12 to OR it 3149into the hash PTE second double word). 3150 31514.75 KVM_IRQFD 3152-------------- 3153 3154:Capability: KVM_CAP_IRQFD 3155:Architectures: x86 s390 arm64 3156:Type: vm ioctl 3157:Parameters: struct kvm_irqfd (in) 3158:Returns: 0 on success, -1 on error 3159 3160Allows setting an eventfd to directly trigger a guest interrupt. 3161kvm_irqfd.fd specifies the file descriptor to use as the eventfd and 3162kvm_irqfd.gsi specifies the irqchip pin toggled by this event. When 3163an event is triggered on the eventfd, an interrupt is injected into 3164the guest using the specified gsi pin. The irqfd is removed using 3165the KVM_IRQFD_FLAG_DEASSIGN flag, specifying both kvm_irqfd.fd 3166and kvm_irqfd.gsi. 3167 3168With KVM_CAP_IRQFD_RESAMPLE, KVM_IRQFD supports a de-assert and notify 3169mechanism allowing emulation of level-triggered, irqfd-based 3170interrupts. When KVM_IRQFD_FLAG_RESAMPLE is set the user must pass an 3171additional eventfd in the kvm_irqfd.resamplefd field. When operating 3172in resample mode, posting of an interrupt through kvm_irq.fd asserts 3173the specified gsi in the irqchip. When the irqchip is resampled, such 3174as from an EOI, the gsi is de-asserted and the user is notified via 3175kvm_irqfd.resamplefd. It is the user's responsibility to re-queue 3176the interrupt if the device making use of it still requires service. 3177Note that closing the resamplefd is not sufficient to disable the 3178irqfd. The KVM_IRQFD_FLAG_RESAMPLE is only necessary on assignment 3179and need not be specified with KVM_IRQFD_FLAG_DEASSIGN. 3180 3181On arm64, gsi routing being supported, the following can happen: 3182 3183- in case no routing entry is associated to this gsi, injection fails 3184- in case the gsi is associated to an irqchip routing entry, 3185 irqchip.pin + 32 corresponds to the injected SPI ID. 3186- in case the gsi is associated to an MSI routing entry, the MSI 3187 message and device ID are translated into an LPI (support restricted 3188 to GICv3 ITS in-kernel emulation). 3189 31904.76 KVM_PPC_ALLOCATE_HTAB 3191-------------------------- 3192 3193:Capability: KVM_CAP_PPC_ALLOC_HTAB 3194:Architectures: powerpc 3195:Type: vm ioctl 3196:Parameters: Pointer to u32 containing hash table order (in/out) 3197:Returns: 0 on success, -1 on error 3198 3199This requests the host kernel to allocate an MMU hash table for a 3200guest using the PAPR paravirtualization interface. This only does 3201anything if the kernel is configured to use the Book 3S HV style of 3202virtualization. Otherwise the capability doesn't exist and the ioctl 3203returns an ENOTTY error. The rest of this description assumes Book 3S 3204HV. 3205 3206There must be no vcpus running when this ioctl is called; if there 3207are, it will do nothing and return an EBUSY error. 3208 3209The parameter is a pointer to a 32-bit unsigned integer variable 3210containing the order (log base 2) of the desired size of the hash 3211table, which must be between 18 and 46. On successful return from the 3212ioctl, the value will not be changed by the kernel. 3213 3214If no hash table has been allocated when any vcpu is asked to run 3215(with the KVM_RUN ioctl), the host kernel will allocate a 3216default-sized hash table (16 MB). 3217 3218If this ioctl is called when a hash table has already been allocated, 3219with a different order from the existing hash table, the existing hash 3220table will be freed and a new one allocated. If this is ioctl is 3221called when a hash table has already been allocated of the same order 3222as specified, the kernel will clear out the existing hash table (zero 3223all HPTEs). In either case, if the guest is using the virtualized 3224real-mode area (VRMA) facility, the kernel will re-create the VMRA 3225HPTEs on the next KVM_RUN of any vcpu. 3226 32274.77 KVM_S390_INTERRUPT 3228----------------------- 3229 3230:Capability: basic 3231:Architectures: s390 3232:Type: vm ioctl, vcpu ioctl 3233:Parameters: struct kvm_s390_interrupt (in) 3234:Returns: 0 on success, -1 on error 3235 3236Allows to inject an interrupt to the guest. Interrupts can be floating 3237(vm ioctl) or per cpu (vcpu ioctl), depending on the interrupt type. 3238 3239Interrupt parameters are passed via kvm_s390_interrupt:: 3240 3241 struct kvm_s390_interrupt { 3242 __u32 type; 3243 __u32 parm; 3244 __u64 parm64; 3245 }; 3246 3247type can be one of the following: 3248 3249KVM_S390_SIGP_STOP (vcpu) 3250 - sigp stop; optional flags in parm 3251KVM_S390_PROGRAM_INT (vcpu) 3252 - program check; code in parm 3253KVM_S390_SIGP_SET_PREFIX (vcpu) 3254 - sigp set prefix; prefix address in parm 3255KVM_S390_RESTART (vcpu) 3256 - restart 3257KVM_S390_INT_CLOCK_COMP (vcpu) 3258 - clock comparator interrupt 3259KVM_S390_INT_CPU_TIMER (vcpu) 3260 - CPU timer interrupt 3261KVM_S390_INT_VIRTIO (vm) 3262 - virtio external interrupt; external interrupt 3263 parameters in parm and parm64 3264KVM_S390_INT_SERVICE (vm) 3265 - sclp external interrupt; sclp parameter in parm 3266KVM_S390_INT_EMERGENCY (vcpu) 3267 - sigp emergency; source cpu in parm 3268KVM_S390_INT_EXTERNAL_CALL (vcpu) 3269 - sigp external call; source cpu in parm 3270KVM_S390_INT_IO(ai,cssid,ssid,schid) (vm) 3271 - compound value to indicate an 3272 I/O interrupt (ai - adapter interrupt; cssid,ssid,schid - subchannel); 3273 I/O interruption parameters in parm (subchannel) and parm64 (intparm, 3274 interruption subclass) 3275KVM_S390_MCHK (vm, vcpu) 3276 - machine check interrupt; cr 14 bits in parm, machine check interrupt 3277 code in parm64 (note that machine checks needing further payload are not 3278 supported by this ioctl) 3279 3280This is an asynchronous vcpu ioctl and can be invoked from any thread. 3281 32824.78 KVM_PPC_GET_HTAB_FD 3283------------------------ 3284 3285:Capability: KVM_CAP_PPC_HTAB_FD 3286:Architectures: powerpc 3287:Type: vm ioctl 3288:Parameters: Pointer to struct kvm_get_htab_fd (in) 3289:Returns: file descriptor number (>= 0) on success, -1 on error 3290 3291This returns a file descriptor that can be used either to read out the 3292entries in the guest's hashed page table (HPT), or to write entries to 3293initialize the HPT. The returned fd can only be written to if the 3294KVM_GET_HTAB_WRITE bit is set in the flags field of the argument, and 3295can only be read if that bit is clear. The argument struct looks like 3296this:: 3297 3298 /* For KVM_PPC_GET_HTAB_FD */ 3299 struct kvm_get_htab_fd { 3300 __u64 flags; 3301 __u64 start_index; 3302 __u64 reserved[2]; 3303 }; 3304 3305 /* Values for kvm_get_htab_fd.flags */ 3306 #define KVM_GET_HTAB_BOLTED_ONLY ((__u64)0x1) 3307 #define KVM_GET_HTAB_WRITE ((__u64)0x2) 3308 3309The 'start_index' field gives the index in the HPT of the entry at 3310which to start reading. It is ignored when writing. 3311 3312Reads on the fd will initially supply information about all 3313"interesting" HPT entries. Interesting entries are those with the 3314bolted bit set, if the KVM_GET_HTAB_BOLTED_ONLY bit is set, otherwise 3315all entries. When the end of the HPT is reached, the read() will 3316return. If read() is called again on the fd, it will start again from 3317the beginning of the HPT, but will only return HPT entries that have 3318changed since they were last read. 3319 3320Data read or written is structured as a header (8 bytes) followed by a 3321series of valid HPT entries (16 bytes) each. The header indicates how 3322many valid HPT entries there are and how many invalid entries follow 3323the valid entries. The invalid entries are not represented explicitly 3324in the stream. The header format is:: 3325 3326 struct kvm_get_htab_header { 3327 __u32 index; 3328 __u16 n_valid; 3329 __u16 n_invalid; 3330 }; 3331 3332Writes to the fd create HPT entries starting at the index given in the 3333header; first 'n_valid' valid entries with contents from the data 3334written, then 'n_invalid' invalid entries, invalidating any previously 3335valid entries found. 3336 33374.79 KVM_CREATE_DEVICE 3338---------------------- 3339 3340:Capability: KVM_CAP_DEVICE_CTRL 3341:Architectures: all 3342:Type: vm ioctl 3343:Parameters: struct kvm_create_device (in/out) 3344:Returns: 0 on success, -1 on error 3345 3346Errors: 3347 3348 ====== ======================================================= 3349 ENODEV The device type is unknown or unsupported 3350 EEXIST Device already created, and this type of device may not 3351 be instantiated multiple times 3352 ====== ======================================================= 3353 3354 Other error conditions may be defined by individual device types or 3355 have their standard meanings. 3356 3357Creates an emulated device in the kernel. The file descriptor returned 3358in fd can be used with KVM_SET/GET/HAS_DEVICE_ATTR. 3359 3360If the KVM_CREATE_DEVICE_TEST flag is set, only test whether the 3361device type is supported (not necessarily whether it can be created 3362in the current vm). 3363 3364Individual devices should not define flags. Attributes should be used 3365for specifying any behavior that is not implied by the device type 3366number. 3367 3368:: 3369 3370 struct kvm_create_device { 3371 __u32 type; /* in: KVM_DEV_TYPE_xxx */ 3372 __u32 fd; /* out: device handle */ 3373 __u32 flags; /* in: KVM_CREATE_DEVICE_xxx */ 3374 }; 3375 33764.80 KVM_SET_DEVICE_ATTR/KVM_GET_DEVICE_ATTR 3377-------------------------------------------- 3378 3379:Capability: KVM_CAP_DEVICE_CTRL, KVM_CAP_VM_ATTRIBUTES for vm device, 3380 KVM_CAP_VCPU_ATTRIBUTES for vcpu device 3381 KVM_CAP_SYS_ATTRIBUTES for system (/dev/kvm) device (no set) 3382:Architectures: x86, arm64, s390 3383:Type: device ioctl, vm ioctl, vcpu ioctl 3384:Parameters: struct kvm_device_attr 3385:Returns: 0 on success, -1 on error 3386 3387Errors: 3388 3389 ===== ============================================================= 3390 ENXIO The group or attribute is unknown/unsupported for this device 3391 or hardware support is missing. 3392 EPERM The attribute cannot (currently) be accessed this way 3393 (e.g. read-only attribute, or attribute that only makes 3394 sense when the device is in a different state) 3395 ===== ============================================================= 3396 3397 Other error conditions may be defined by individual device types. 3398 3399Gets/sets a specified piece of device configuration and/or state. The 3400semantics are device-specific. See individual device documentation in 3401the "devices" directory. As with ONE_REG, the size of the data 3402transferred is defined by the particular attribute. 3403 3404:: 3405 3406 struct kvm_device_attr { 3407 __u32 flags; /* no flags currently defined */ 3408 __u32 group; /* device-defined */ 3409 __u64 attr; /* group-defined */ 3410 __u64 addr; /* userspace address of attr data */ 3411 }; 3412 34134.81 KVM_HAS_DEVICE_ATTR 3414------------------------ 3415 3416:Capability: KVM_CAP_DEVICE_CTRL, KVM_CAP_VM_ATTRIBUTES for vm device, 3417 KVM_CAP_VCPU_ATTRIBUTES for vcpu device 3418 KVM_CAP_SYS_ATTRIBUTES for system (/dev/kvm) device 3419:Type: device ioctl, vm ioctl, vcpu ioctl 3420:Parameters: struct kvm_device_attr 3421:Returns: 0 on success, -1 on error 3422 3423Errors: 3424 3425 ===== ============================================================= 3426 ENXIO The group or attribute is unknown/unsupported for this device 3427 or hardware support is missing. 3428 ===== ============================================================= 3429 3430Tests whether a device supports a particular attribute. A successful 3431return indicates the attribute is implemented. It does not necessarily 3432indicate that the attribute can be read or written in the device's 3433current state. "addr" is ignored. 3434 3435.. _KVM_ARM_VCPU_INIT: 3436 34374.82 KVM_ARM_VCPU_INIT 3438---------------------- 3439 3440:Capability: basic 3441:Architectures: arm64 3442:Type: vcpu ioctl 3443:Parameters: struct kvm_vcpu_init (in) 3444:Returns: 0 on success; -1 on error 3445 3446Errors: 3447 3448 ====== ================================================================= 3449 EINVAL the target is unknown, or the combination of features is invalid. 3450 ENOENT a features bit specified is unknown. 3451 ====== ================================================================= 3452 3453This tells KVM what type of CPU to present to the guest, and what 3454optional features it should have. This will cause a reset of the cpu 3455registers to their initial values. If this is not called, KVM_RUN will 3456return ENOEXEC for that vcpu. 3457 3458The initial values are defined as: 3459 - Processor state: 3460 * AArch64: EL1h, D, A, I and F bits set. All other bits 3461 are cleared. 3462 * AArch32: SVC, A, I and F bits set. All other bits are 3463 cleared. 3464 - General Purpose registers, including PC and SP: set to 0 3465 - FPSIMD/NEON registers: set to 0 3466 - SVE registers: set to 0 3467 - System registers: Reset to their architecturally defined 3468 values as for a warm reset to EL1 (resp. SVC) 3469 3470Note that because some registers reflect machine topology, all vcpus 3471should be created before this ioctl is invoked. 3472 3473Userspace can call this function multiple times for a given vcpu, including 3474after the vcpu has been run. This will reset the vcpu to its initial 3475state. All calls to this function after the initial call must use the same 3476target and same set of feature flags, otherwise EINVAL will be returned. 3477 3478Possible features: 3479 3480 - KVM_ARM_VCPU_POWER_OFF: Starts the CPU in a power-off state. 3481 Depends on KVM_CAP_ARM_PSCI. If not set, the CPU will be powered on 3482 and execute guest code when KVM_RUN is called. 3483 - KVM_ARM_VCPU_EL1_32BIT: Starts the CPU in a 32bit mode. 3484 Depends on KVM_CAP_ARM_EL1_32BIT (arm64 only). 3485 - KVM_ARM_VCPU_PSCI_0_2: Emulate PSCI v0.2 (or a future revision 3486 backward compatible with v0.2) for the CPU. 3487 Depends on KVM_CAP_ARM_PSCI_0_2. 3488 - KVM_ARM_VCPU_PMU_V3: Emulate PMUv3 for the CPU. 3489 Depends on KVM_CAP_ARM_PMU_V3. 3490 3491 - KVM_ARM_VCPU_PTRAUTH_ADDRESS: Enables Address Pointer authentication 3492 for arm64 only. 3493 Depends on KVM_CAP_ARM_PTRAUTH_ADDRESS. 3494 If KVM_CAP_ARM_PTRAUTH_ADDRESS and KVM_CAP_ARM_PTRAUTH_GENERIC are 3495 both present, then both KVM_ARM_VCPU_PTRAUTH_ADDRESS and 3496 KVM_ARM_VCPU_PTRAUTH_GENERIC must be requested or neither must be 3497 requested. 3498 3499 - KVM_ARM_VCPU_PTRAUTH_GENERIC: Enables Generic Pointer authentication 3500 for arm64 only. 3501 Depends on KVM_CAP_ARM_PTRAUTH_GENERIC. 3502 If KVM_CAP_ARM_PTRAUTH_ADDRESS and KVM_CAP_ARM_PTRAUTH_GENERIC are 3503 both present, then both KVM_ARM_VCPU_PTRAUTH_ADDRESS and 3504 KVM_ARM_VCPU_PTRAUTH_GENERIC must be requested or neither must be 3505 requested. 3506 3507 - KVM_ARM_VCPU_SVE: Enables SVE for the CPU (arm64 only). 3508 Depends on KVM_CAP_ARM_SVE. 3509 Requires KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE): 3510 3511 * After KVM_ARM_VCPU_INIT: 3512 3513 - KVM_REG_ARM64_SVE_VLS may be read using KVM_GET_ONE_REG: the 3514 initial value of this pseudo-register indicates the best set of 3515 vector lengths possible for a vcpu on this host. 3516 3517 * Before KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE): 3518 3519 - KVM_RUN and KVM_GET_REG_LIST are not available; 3520 3521 - KVM_GET_ONE_REG and KVM_SET_ONE_REG cannot be used to access 3522 the scalable architectural SVE registers 3523 KVM_REG_ARM64_SVE_ZREG(), KVM_REG_ARM64_SVE_PREG() or 3524 KVM_REG_ARM64_SVE_FFR; 3525 3526 - KVM_REG_ARM64_SVE_VLS may optionally be written using 3527 KVM_SET_ONE_REG, to modify the set of vector lengths available 3528 for the vcpu. 3529 3530 * After KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE): 3531 3532 - the KVM_REG_ARM64_SVE_VLS pseudo-register is immutable, and can 3533 no longer be written using KVM_SET_ONE_REG. 3534 35354.83 KVM_ARM_PREFERRED_TARGET 3536----------------------------- 3537 3538:Capability: basic 3539:Architectures: arm64 3540:Type: vm ioctl 3541:Parameters: struct kvm_vcpu_init (out) 3542:Returns: 0 on success; -1 on error 3543 3544Errors: 3545 3546 ====== ========================================== 3547 ENODEV no preferred target available for the host 3548 ====== ========================================== 3549 3550This queries KVM for preferred CPU target type which can be emulated 3551by KVM on underlying host. 3552 3553The ioctl returns struct kvm_vcpu_init instance containing information 3554about preferred CPU target type and recommended features for it. The 3555kvm_vcpu_init->features bitmap returned will have feature bits set if 3556the preferred target recommends setting these features, but this is 3557not mandatory. 3558 3559The information returned by this ioctl can be used to prepare an instance 3560of struct kvm_vcpu_init for KVM_ARM_VCPU_INIT ioctl which will result in 3561VCPU matching underlying host. 3562 3563 35644.84 KVM_GET_REG_LIST 3565--------------------- 3566 3567:Capability: basic 3568:Architectures: arm64, mips, riscv 3569:Type: vcpu ioctl 3570:Parameters: struct kvm_reg_list (in/out) 3571:Returns: 0 on success; -1 on error 3572 3573Errors: 3574 3575 ===== ============================================================== 3576 E2BIG the reg index list is too big to fit in the array specified by 3577 the user (the number required will be written into n). 3578 ===== ============================================================== 3579 3580:: 3581 3582 struct kvm_reg_list { 3583 __u64 n; /* number of registers in reg[] */ 3584 __u64 reg[0]; 3585 }; 3586 3587This ioctl returns the guest registers that are supported for the 3588KVM_GET_ONE_REG/KVM_SET_ONE_REG calls. 3589 3590 35914.85 KVM_ARM_SET_DEVICE_ADDR (deprecated) 3592----------------------------------------- 3593 3594:Capability: KVM_CAP_ARM_SET_DEVICE_ADDR 3595:Architectures: arm64 3596:Type: vm ioctl 3597:Parameters: struct kvm_arm_device_address (in) 3598:Returns: 0 on success, -1 on error 3599 3600Errors: 3601 3602 ====== ============================================ 3603 ENODEV The device id is unknown 3604 ENXIO Device not supported on current system 3605 EEXIST Address already set 3606 E2BIG Address outside guest physical address space 3607 EBUSY Address overlaps with other device range 3608 ====== ============================================ 3609 3610:: 3611 3612 struct kvm_arm_device_addr { 3613 __u64 id; 3614 __u64 addr; 3615 }; 3616 3617Specify a device address in the guest's physical address space where guests 3618can access emulated or directly exposed devices, which the host kernel needs 3619to know about. The id field is an architecture specific identifier for a 3620specific device. 3621 3622arm64 divides the id field into two parts, a device id and an 3623address type id specific to the individual device:: 3624 3625 bits: | 63 ... 32 | 31 ... 16 | 15 ... 0 | 3626 field: | 0x00000000 | device id | addr type id | 3627 3628arm64 currently only require this when using the in-kernel GIC 3629support for the hardware VGIC features, using KVM_ARM_DEVICE_VGIC_V2 3630as the device id. When setting the base address for the guest's 3631mapping of the VGIC virtual CPU and distributor interface, the ioctl 3632must be called after calling KVM_CREATE_IRQCHIP, but before calling 3633KVM_RUN on any of the VCPUs. Calling this ioctl twice for any of the 3634base addresses will return -EEXIST. 3635 3636Note, this IOCTL is deprecated and the more flexible SET/GET_DEVICE_ATTR API 3637should be used instead. 3638 3639 36404.86 KVM_PPC_RTAS_DEFINE_TOKEN 3641------------------------------ 3642 3643:Capability: KVM_CAP_PPC_RTAS 3644:Architectures: ppc 3645:Type: vm ioctl 3646:Parameters: struct kvm_rtas_token_args 3647:Returns: 0 on success, -1 on error 3648 3649Defines a token value for a RTAS (Run Time Abstraction Services) 3650service in order to allow it to be handled in the kernel. The 3651argument struct gives the name of the service, which must be the name 3652of a service that has a kernel-side implementation. If the token 3653value is non-zero, it will be associated with that service, and 3654subsequent RTAS calls by the guest specifying that token will be 3655handled by the kernel. If the token value is 0, then any token 3656associated with the service will be forgotten, and subsequent RTAS 3657calls by the guest for that service will be passed to userspace to be 3658handled. 3659 36604.87 KVM_SET_GUEST_DEBUG 3661------------------------ 3662 3663:Capability: KVM_CAP_SET_GUEST_DEBUG 3664:Architectures: x86, s390, ppc, arm64 3665:Type: vcpu ioctl 3666:Parameters: struct kvm_guest_debug (in) 3667:Returns: 0 on success; -1 on error 3668 3669:: 3670 3671 struct kvm_guest_debug { 3672 __u32 control; 3673 __u32 pad; 3674 struct kvm_guest_debug_arch arch; 3675 }; 3676 3677Set up the processor specific debug registers and configure vcpu for 3678handling guest debug events. There are two parts to the structure, the 3679first a control bitfield indicates the type of debug events to handle 3680when running. Common control bits are: 3681 3682 - KVM_GUESTDBG_ENABLE: guest debugging is enabled 3683 - KVM_GUESTDBG_SINGLESTEP: the next run should single-step 3684 3685The top 16 bits of the control field are architecture specific control 3686flags which can include the following: 3687 3688 - KVM_GUESTDBG_USE_SW_BP: using software breakpoints [x86, arm64] 3689 - KVM_GUESTDBG_USE_HW_BP: using hardware breakpoints [x86, s390] 3690 - KVM_GUESTDBG_USE_HW: using hardware debug events [arm64] 3691 - KVM_GUESTDBG_INJECT_DB: inject DB type exception [x86] 3692 - KVM_GUESTDBG_INJECT_BP: inject BP type exception [x86] 3693 - KVM_GUESTDBG_EXIT_PENDING: trigger an immediate guest exit [s390] 3694 - KVM_GUESTDBG_BLOCKIRQ: avoid injecting interrupts/NMI/SMI [x86] 3695 3696For example KVM_GUESTDBG_USE_SW_BP indicates that software breakpoints 3697are enabled in memory so we need to ensure breakpoint exceptions are 3698correctly trapped and the KVM run loop exits at the breakpoint and not 3699running off into the normal guest vector. For KVM_GUESTDBG_USE_HW_BP 3700we need to ensure the guest vCPUs architecture specific registers are 3701updated to the correct (supplied) values. 3702 3703The second part of the structure is architecture specific and 3704typically contains a set of debug registers. 3705 3706For arm64 the number of debug registers is implementation defined and 3707can be determined by querying the KVM_CAP_GUEST_DEBUG_HW_BPS and 3708KVM_CAP_GUEST_DEBUG_HW_WPS capabilities which return a positive number 3709indicating the number of supported registers. 3710 3711For ppc, the KVM_CAP_PPC_GUEST_DEBUG_SSTEP capability indicates whether 3712the single-step debug event (KVM_GUESTDBG_SINGLESTEP) is supported. 3713 3714Also when supported, KVM_CAP_SET_GUEST_DEBUG2 capability indicates the 3715supported KVM_GUESTDBG_* bits in the control field. 3716 3717When debug events exit the main run loop with the reason 3718KVM_EXIT_DEBUG with the kvm_debug_exit_arch part of the kvm_run 3719structure containing architecture specific debug information. 3720 37214.88 KVM_GET_EMULATED_CPUID 3722--------------------------- 3723 3724:Capability: KVM_CAP_EXT_EMUL_CPUID 3725:Architectures: x86 3726:Type: system ioctl 3727:Parameters: struct kvm_cpuid2 (in/out) 3728:Returns: 0 on success, -1 on error 3729 3730:: 3731 3732 struct kvm_cpuid2 { 3733 __u32 nent; 3734 __u32 flags; 3735 struct kvm_cpuid_entry2 entries[0]; 3736 }; 3737 3738The member 'flags' is used for passing flags from userspace. 3739 3740:: 3741 3742 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX BIT(0) 3743 #define KVM_CPUID_FLAG_STATEFUL_FUNC BIT(1) /* deprecated */ 3744 #define KVM_CPUID_FLAG_STATE_READ_NEXT BIT(2) /* deprecated */ 3745 3746 struct kvm_cpuid_entry2 { 3747 __u32 function; 3748 __u32 index; 3749 __u32 flags; 3750 __u32 eax; 3751 __u32 ebx; 3752 __u32 ecx; 3753 __u32 edx; 3754 __u32 padding[3]; 3755 }; 3756 3757This ioctl returns x86 cpuid features which are emulated by 3758kvm.Userspace can use the information returned by this ioctl to query 3759which features are emulated by kvm instead of being present natively. 3760 3761Userspace invokes KVM_GET_EMULATED_CPUID by passing a kvm_cpuid2 3762structure with the 'nent' field indicating the number of entries in 3763the variable-size array 'entries'. If the number of entries is too low 3764to describe the cpu capabilities, an error (E2BIG) is returned. If the 3765number is too high, the 'nent' field is adjusted and an error (ENOMEM) 3766is returned. If the number is just right, the 'nent' field is adjusted 3767to the number of valid entries in the 'entries' array, which is then 3768filled. 3769 3770The entries returned are the set CPUID bits of the respective features 3771which kvm emulates, as returned by the CPUID instruction, with unknown 3772or unsupported feature bits cleared. 3773 3774Features like x2apic, for example, may not be present in the host cpu 3775but are exposed by kvm in KVM_GET_SUPPORTED_CPUID because they can be 3776emulated efficiently and thus not included here. 3777 3778The fields in each entry are defined as follows: 3779 3780 function: 3781 the eax value used to obtain the entry 3782 index: 3783 the ecx value used to obtain the entry (for entries that are 3784 affected by ecx) 3785 flags: 3786 an OR of zero or more of the following: 3787 3788 KVM_CPUID_FLAG_SIGNIFCANT_INDEX: 3789 if the index field is valid 3790 3791 eax, ebx, ecx, edx: 3792 3793 the values returned by the cpuid instruction for 3794 this function/index combination 3795 37964.89 KVM_S390_MEM_OP 3797-------------------- 3798 3799:Capability: KVM_CAP_S390_MEM_OP, KVM_CAP_S390_PROTECTED, KVM_CAP_S390_MEM_OP_EXTENSION 3800:Architectures: s390 3801:Type: vm ioctl, vcpu ioctl 3802:Parameters: struct kvm_s390_mem_op (in) 3803:Returns: = 0 on success, 3804 < 0 on generic error (e.g. -EFAULT or -ENOMEM), 3805 16 bit program exception code if the access causes such an exception 3806 3807Read or write data from/to the VM's memory. 3808The KVM_CAP_S390_MEM_OP_EXTENSION capability specifies what functionality is 3809supported. 3810 3811Parameters are specified via the following structure:: 3812 3813 struct kvm_s390_mem_op { 3814 __u64 gaddr; /* the guest address */ 3815 __u64 flags; /* flags */ 3816 __u32 size; /* amount of bytes */ 3817 __u32 op; /* type of operation */ 3818 __u64 buf; /* buffer in userspace */ 3819 union { 3820 struct { 3821 __u8 ar; /* the access register number */ 3822 __u8 key; /* access key, ignored if flag unset */ 3823 __u8 pad1[6]; /* ignored */ 3824 __u64 old_addr; /* ignored if flag unset */ 3825 }; 3826 __u32 sida_offset; /* offset into the sida */ 3827 __u8 reserved[32]; /* ignored */ 3828 }; 3829 }; 3830 3831The start address of the memory region has to be specified in the "gaddr" 3832field, and the length of the region in the "size" field (which must not 3833be 0). The maximum value for "size" can be obtained by checking the 3834KVM_CAP_S390_MEM_OP capability. "buf" is the buffer supplied by the 3835userspace application where the read data should be written to for 3836a read access, or where the data that should be written is stored for 3837a write access. The "reserved" field is meant for future extensions. 3838Reserved and unused values are ignored. Future extension that add members must 3839introduce new flags. 3840 3841The type of operation is specified in the "op" field. Flags modifying 3842their behavior can be set in the "flags" field. Undefined flag bits must 3843be set to 0. 3844 3845Possible operations are: 3846 * ``KVM_S390_MEMOP_LOGICAL_READ`` 3847 * ``KVM_S390_MEMOP_LOGICAL_WRITE`` 3848 * ``KVM_S390_MEMOP_ABSOLUTE_READ`` 3849 * ``KVM_S390_MEMOP_ABSOLUTE_WRITE`` 3850 * ``KVM_S390_MEMOP_SIDA_READ`` 3851 * ``KVM_S390_MEMOP_SIDA_WRITE`` 3852 * ``KVM_S390_MEMOP_ABSOLUTE_CMPXCHG`` 3853 3854Logical read/write: 3855^^^^^^^^^^^^^^^^^^^ 3856 3857Access logical memory, i.e. translate the given guest address to an absolute 3858address given the state of the VCPU and use the absolute address as target of 3859the access. "ar" designates the access register number to be used; the valid 3860range is 0..15. 3861Logical accesses are permitted for the VCPU ioctl only. 3862Logical accesses are permitted for non-protected guests only. 3863 3864Supported flags: 3865 * ``KVM_S390_MEMOP_F_CHECK_ONLY`` 3866 * ``KVM_S390_MEMOP_F_INJECT_EXCEPTION`` 3867 * ``KVM_S390_MEMOP_F_SKEY_PROTECTION`` 3868 3869The KVM_S390_MEMOP_F_CHECK_ONLY flag can be set to check whether the 3870corresponding memory access would cause an access exception; however, 3871no actual access to the data in memory at the destination is performed. 3872In this case, "buf" is unused and can be NULL. 3873 3874In case an access exception occurred during the access (or would occur 3875in case of KVM_S390_MEMOP_F_CHECK_ONLY), the ioctl returns a positive 3876error number indicating the type of exception. This exception is also 3877raised directly at the corresponding VCPU if the flag 3878KVM_S390_MEMOP_F_INJECT_EXCEPTION is set. 3879On protection exceptions, unless specified otherwise, the injected 3880translation-exception identifier (TEID) indicates suppression. 3881 3882If the KVM_S390_MEMOP_F_SKEY_PROTECTION flag is set, storage key 3883protection is also in effect and may cause exceptions if accesses are 3884prohibited given the access key designated by "key"; the valid range is 0..15. 3885KVM_S390_MEMOP_F_SKEY_PROTECTION is available if KVM_CAP_S390_MEM_OP_EXTENSION 3886is > 0. 3887Since the accessed memory may span multiple pages and those pages might have 3888different storage keys, it is possible that a protection exception occurs 3889after memory has been modified. In this case, if the exception is injected, 3890the TEID does not indicate suppression. 3891 3892Absolute read/write: 3893^^^^^^^^^^^^^^^^^^^^ 3894 3895Access absolute memory. This operation is intended to be used with the 3896KVM_S390_MEMOP_F_SKEY_PROTECTION flag, to allow accessing memory and performing 3897the checks required for storage key protection as one operation (as opposed to 3898user space getting the storage keys, performing the checks, and accessing 3899memory thereafter, which could lead to a delay between check and access). 3900Absolute accesses are permitted for the VM ioctl if KVM_CAP_S390_MEM_OP_EXTENSION 3901has the KVM_S390_MEMOP_EXTENSION_CAP_BASE bit set. 3902Currently absolute accesses are not permitted for VCPU ioctls. 3903Absolute accesses are permitted for non-protected guests only. 3904 3905Supported flags: 3906 * ``KVM_S390_MEMOP_F_CHECK_ONLY`` 3907 * ``KVM_S390_MEMOP_F_SKEY_PROTECTION`` 3908 3909The semantics of the flags common with logical accesses are as for logical 3910accesses. 3911 3912Absolute cmpxchg: 3913^^^^^^^^^^^^^^^^^ 3914 3915Perform cmpxchg on absolute guest memory. Intended for use with the 3916KVM_S390_MEMOP_F_SKEY_PROTECTION flag. 3917Instead of doing an unconditional write, the access occurs only if the target 3918location contains the value pointed to by "old_addr". 3919This is performed as an atomic cmpxchg with the length specified by the "size" 3920parameter. "size" must be a power of two up to and including 16. 3921If the exchange did not take place because the target value doesn't match the 3922old value, the value "old_addr" points to is replaced by the target value. 3923User space can tell if an exchange took place by checking if this replacement 3924occurred. The cmpxchg op is permitted for the VM ioctl if 3925KVM_CAP_S390_MEM_OP_EXTENSION has flag KVM_S390_MEMOP_EXTENSION_CAP_CMPXCHG set. 3926 3927Supported flags: 3928 * ``KVM_S390_MEMOP_F_SKEY_PROTECTION`` 3929 3930SIDA read/write: 3931^^^^^^^^^^^^^^^^ 3932 3933Access the secure instruction data area which contains memory operands necessary 3934for instruction emulation for protected guests. 3935SIDA accesses are available if the KVM_CAP_S390_PROTECTED capability is available. 3936SIDA accesses are permitted for the VCPU ioctl only. 3937SIDA accesses are permitted for protected guests only. 3938 3939No flags are supported. 3940 39414.90 KVM_S390_GET_SKEYS 3942----------------------- 3943 3944:Capability: KVM_CAP_S390_SKEYS 3945:Architectures: s390 3946:Type: vm ioctl 3947:Parameters: struct kvm_s390_skeys 3948:Returns: 0 on success, KVM_S390_GET_SKEYS_NONE if guest is not using storage 3949 keys, negative value on error 3950 3951This ioctl is used to get guest storage key values on the s390 3952architecture. The ioctl takes parameters via the kvm_s390_skeys struct:: 3953 3954 struct kvm_s390_skeys { 3955 __u64 start_gfn; 3956 __u64 count; 3957 __u64 skeydata_addr; 3958 __u32 flags; 3959 __u32 reserved[9]; 3960 }; 3961 3962The start_gfn field is the number of the first guest frame whose storage keys 3963you want to get. 3964 3965The count field is the number of consecutive frames (starting from start_gfn) 3966whose storage keys to get. The count field must be at least 1 and the maximum 3967allowed value is defined as KVM_S390_SKEYS_MAX. Values outside this range 3968will cause the ioctl to return -EINVAL. 3969 3970The skeydata_addr field is the address to a buffer large enough to hold count 3971bytes. This buffer will be filled with storage key data by the ioctl. 3972 39734.91 KVM_S390_SET_SKEYS 3974----------------------- 3975 3976:Capability: KVM_CAP_S390_SKEYS 3977:Architectures: s390 3978:Type: vm ioctl 3979:Parameters: struct kvm_s390_skeys 3980:Returns: 0 on success, negative value on error 3981 3982This ioctl is used to set guest storage key values on the s390 3983architecture. The ioctl takes parameters via the kvm_s390_skeys struct. 3984See section on KVM_S390_GET_SKEYS for struct definition. 3985 3986The start_gfn field is the number of the first guest frame whose storage keys 3987you want to set. 3988 3989The count field is the number of consecutive frames (starting from start_gfn) 3990whose storage keys to get. The count field must be at least 1 and the maximum 3991allowed value is defined as KVM_S390_SKEYS_MAX. Values outside this range 3992will cause the ioctl to return -EINVAL. 3993 3994The skeydata_addr field is the address to a buffer containing count bytes of 3995storage keys. Each byte in the buffer will be set as the storage key for a 3996single frame starting at start_gfn for count frames. 3997 3998Note: If any architecturally invalid key value is found in the given data then 3999the ioctl will return -EINVAL. 4000 40014.92 KVM_S390_IRQ 4002----------------- 4003 4004:Capability: KVM_CAP_S390_INJECT_IRQ 4005:Architectures: s390 4006:Type: vcpu ioctl 4007:Parameters: struct kvm_s390_irq (in) 4008:Returns: 0 on success, -1 on error 4009 4010Errors: 4011 4012 4013 ====== ================================================================= 4014 EINVAL interrupt type is invalid 4015 type is KVM_S390_SIGP_STOP and flag parameter is invalid value, 4016 type is KVM_S390_INT_EXTERNAL_CALL and code is bigger 4017 than the maximum of VCPUs 4018 EBUSY type is KVM_S390_SIGP_SET_PREFIX and vcpu is not stopped, 4019 type is KVM_S390_SIGP_STOP and a stop irq is already pending, 4020 type is KVM_S390_INT_EXTERNAL_CALL and an external call interrupt 4021 is already pending 4022 ====== ================================================================= 4023 4024Allows to inject an interrupt to the guest. 4025 4026Using struct kvm_s390_irq as a parameter allows 4027to inject additional payload which is not 4028possible via KVM_S390_INTERRUPT. 4029 4030Interrupt parameters are passed via kvm_s390_irq:: 4031 4032 struct kvm_s390_irq { 4033 __u64 type; 4034 union { 4035 struct kvm_s390_io_info io; 4036 struct kvm_s390_ext_info ext; 4037 struct kvm_s390_pgm_info pgm; 4038 struct kvm_s390_emerg_info emerg; 4039 struct kvm_s390_extcall_info extcall; 4040 struct kvm_s390_prefix_info prefix; 4041 struct kvm_s390_stop_info stop; 4042 struct kvm_s390_mchk_info mchk; 4043 char reserved[64]; 4044 } u; 4045 }; 4046 4047type can be one of the following: 4048 4049- KVM_S390_SIGP_STOP - sigp stop; parameter in .stop 4050- KVM_S390_PROGRAM_INT - program check; parameters in .pgm 4051- KVM_S390_SIGP_SET_PREFIX - sigp set prefix; parameters in .prefix 4052- KVM_S390_RESTART - restart; no parameters 4053- KVM_S390_INT_CLOCK_COMP - clock comparator interrupt; no parameters 4054- KVM_S390_INT_CPU_TIMER - CPU timer interrupt; no parameters 4055- KVM_S390_INT_EMERGENCY - sigp emergency; parameters in .emerg 4056- KVM_S390_INT_EXTERNAL_CALL - sigp external call; parameters in .extcall 4057- KVM_S390_MCHK - machine check interrupt; parameters in .mchk 4058 4059This is an asynchronous vcpu ioctl and can be invoked from any thread. 4060 40614.94 KVM_S390_GET_IRQ_STATE 4062--------------------------- 4063 4064:Capability: KVM_CAP_S390_IRQ_STATE 4065:Architectures: s390 4066:Type: vcpu ioctl 4067:Parameters: struct kvm_s390_irq_state (out) 4068:Returns: >= number of bytes copied into buffer, 4069 -EINVAL if buffer size is 0, 4070 -ENOBUFS if buffer size is too small to fit all pending interrupts, 4071 -EFAULT if the buffer address was invalid 4072 4073This ioctl allows userspace to retrieve the complete state of all currently 4074pending interrupts in a single buffer. Use cases include migration 4075and introspection. The parameter structure contains the address of a 4076userspace buffer and its length:: 4077 4078 struct kvm_s390_irq_state { 4079 __u64 buf; 4080 __u32 flags; /* will stay unused for compatibility reasons */ 4081 __u32 len; 4082 __u32 reserved[4]; /* will stay unused for compatibility reasons */ 4083 }; 4084 4085Userspace passes in the above struct and for each pending interrupt a 4086struct kvm_s390_irq is copied to the provided buffer. 4087 4088The structure contains a flags and a reserved field for future extensions. As 4089the kernel never checked for flags == 0 and QEMU never pre-zeroed flags and 4090reserved, these fields can not be used in the future without breaking 4091compatibility. 4092 4093If -ENOBUFS is returned the buffer provided was too small and userspace 4094may retry with a bigger buffer. 4095 40964.95 KVM_S390_SET_IRQ_STATE 4097--------------------------- 4098 4099:Capability: KVM_CAP_S390_IRQ_STATE 4100:Architectures: s390 4101:Type: vcpu ioctl 4102:Parameters: struct kvm_s390_irq_state (in) 4103:Returns: 0 on success, 4104 -EFAULT if the buffer address was invalid, 4105 -EINVAL for an invalid buffer length (see below), 4106 -EBUSY if there were already interrupts pending, 4107 errors occurring when actually injecting the 4108 interrupt. See KVM_S390_IRQ. 4109 4110This ioctl allows userspace to set the complete state of all cpu-local 4111interrupts currently pending for the vcpu. It is intended for restoring 4112interrupt state after a migration. The input parameter is a userspace buffer 4113containing a struct kvm_s390_irq_state:: 4114 4115 struct kvm_s390_irq_state { 4116 __u64 buf; 4117 __u32 flags; /* will stay unused for compatibility reasons */ 4118 __u32 len; 4119 __u32 reserved[4]; /* will stay unused for compatibility reasons */ 4120 }; 4121 4122The restrictions for flags and reserved apply as well. 4123(see KVM_S390_GET_IRQ_STATE) 4124 4125The userspace memory referenced by buf contains a struct kvm_s390_irq 4126for each interrupt to be injected into the guest. 4127If one of the interrupts could not be injected for some reason the 4128ioctl aborts. 4129 4130len must be a multiple of sizeof(struct kvm_s390_irq). It must be > 0 4131and it must not exceed (max_vcpus + 32) * sizeof(struct kvm_s390_irq), 4132which is the maximum number of possibly pending cpu-local interrupts. 4133 41344.96 KVM_SMI 4135------------ 4136 4137:Capability: KVM_CAP_X86_SMM 4138:Architectures: x86 4139:Type: vcpu ioctl 4140:Parameters: none 4141:Returns: 0 on success, -1 on error 4142 4143Queues an SMI on the thread's vcpu. 4144 41454.97 KVM_X86_SET_MSR_FILTER 4146---------------------------- 4147 4148:Capability: KVM_CAP_X86_MSR_FILTER 4149:Architectures: x86 4150:Type: vm ioctl 4151:Parameters: struct kvm_msr_filter 4152:Returns: 0 on success, < 0 on error 4153 4154:: 4155 4156 struct kvm_msr_filter_range { 4157 #define KVM_MSR_FILTER_READ (1 << 0) 4158 #define KVM_MSR_FILTER_WRITE (1 << 1) 4159 __u32 flags; 4160 __u32 nmsrs; /* number of msrs in bitmap */ 4161 __u32 base; /* MSR index the bitmap starts at */ 4162 __u8 *bitmap; /* a 1 bit allows the operations in flags, 0 denies */ 4163 }; 4164 4165 #define KVM_MSR_FILTER_MAX_RANGES 16 4166 struct kvm_msr_filter { 4167 #define KVM_MSR_FILTER_DEFAULT_ALLOW (0 << 0) 4168 #define KVM_MSR_FILTER_DEFAULT_DENY (1 << 0) 4169 __u32 flags; 4170 struct kvm_msr_filter_range ranges[KVM_MSR_FILTER_MAX_RANGES]; 4171 }; 4172 4173flags values for ``struct kvm_msr_filter_range``: 4174 4175``KVM_MSR_FILTER_READ`` 4176 4177 Filter read accesses to MSRs using the given bitmap. A 0 in the bitmap 4178 indicates that read accesses should be denied, while a 1 indicates that 4179 a read for a particular MSR should be allowed regardless of the default 4180 filter action. 4181 4182``KVM_MSR_FILTER_WRITE`` 4183 4184 Filter write accesses to MSRs using the given bitmap. A 0 in the bitmap 4185 indicates that write accesses should be denied, while a 1 indicates that 4186 a write for a particular MSR should be allowed regardless of the default 4187 filter action. 4188 4189flags values for ``struct kvm_msr_filter``: 4190 4191``KVM_MSR_FILTER_DEFAULT_ALLOW`` 4192 4193 If no filter range matches an MSR index that is getting accessed, KVM will 4194 allow accesses to all MSRs by default. 4195 4196``KVM_MSR_FILTER_DEFAULT_DENY`` 4197 4198 If no filter range matches an MSR index that is getting accessed, KVM will 4199 deny accesses to all MSRs by default. 4200 4201This ioctl allows userspace to define up to 16 bitmaps of MSR ranges to deny 4202guest MSR accesses that would normally be allowed by KVM. If an MSR is not 4203covered by a specific range, the "default" filtering behavior applies. Each 4204bitmap range covers MSRs from [base .. base+nmsrs). 4205 4206If an MSR access is denied by userspace, the resulting KVM behavior depends on 4207whether or not KVM_CAP_X86_USER_SPACE_MSR's KVM_MSR_EXIT_REASON_FILTER is 4208enabled. If KVM_MSR_EXIT_REASON_FILTER is enabled, KVM will exit to userspace 4209on denied accesses, i.e. userspace effectively intercepts the MSR access. If 4210KVM_MSR_EXIT_REASON_FILTER is not enabled, KVM will inject a #GP into the guest 4211on denied accesses. 4212 4213If an MSR access is allowed by userspace, KVM will emulate and/or virtualize 4214the access in accordance with the vCPU model. Note, KVM may still ultimately 4215inject a #GP if an access is allowed by userspace, e.g. if KVM doesn't support 4216the MSR, or to follow architectural behavior for the MSR. 4217 4218By default, KVM operates in KVM_MSR_FILTER_DEFAULT_ALLOW mode with no MSR range 4219filters. 4220 4221Calling this ioctl with an empty set of ranges (all nmsrs == 0) disables MSR 4222filtering. In that mode, ``KVM_MSR_FILTER_DEFAULT_DENY`` is invalid and causes 4223an error. 4224 4225.. warning:: 4226 MSR accesses as part of nested VM-Enter/VM-Exit are not filtered. 4227 This includes both writes to individual VMCS fields and reads/writes 4228 through the MSR lists pointed to by the VMCS. 4229 4230 x2APIC MSR accesses cannot be filtered (KVM silently ignores filters that 4231 cover any x2APIC MSRs). 4232 4233Note, invoking this ioctl while a vCPU is running is inherently racy. However, 4234KVM does guarantee that vCPUs will see either the previous filter or the new 4235filter, e.g. MSRs with identical settings in both the old and new filter will 4236have deterministic behavior. 4237 4238Similarly, if userspace wishes to intercept on denied accesses, 4239KVM_MSR_EXIT_REASON_FILTER must be enabled before activating any filters, and 4240left enabled until after all filters are deactivated. Failure to do so may 4241result in KVM injecting a #GP instead of exiting to userspace. 4242 42434.98 KVM_CREATE_SPAPR_TCE_64 4244---------------------------- 4245 4246:Capability: KVM_CAP_SPAPR_TCE_64 4247:Architectures: powerpc 4248:Type: vm ioctl 4249:Parameters: struct kvm_create_spapr_tce_64 (in) 4250:Returns: file descriptor for manipulating the created TCE table 4251 4252This is an extension for KVM_CAP_SPAPR_TCE which only supports 32bit 4253windows, described in 4.62 KVM_CREATE_SPAPR_TCE 4254 4255This capability uses extended struct in ioctl interface:: 4256 4257 /* for KVM_CAP_SPAPR_TCE_64 */ 4258 struct kvm_create_spapr_tce_64 { 4259 __u64 liobn; 4260 __u32 page_shift; 4261 __u32 flags; 4262 __u64 offset; /* in pages */ 4263 __u64 size; /* in pages */ 4264 }; 4265 4266The aim of extension is to support an additional bigger DMA window with 4267a variable page size. 4268KVM_CREATE_SPAPR_TCE_64 receives a 64bit window size, an IOMMU page shift and 4269a bus offset of the corresponding DMA window, @size and @offset are numbers 4270of IOMMU pages. 4271 4272@flags are not used at the moment. 4273 4274The rest of functionality is identical to KVM_CREATE_SPAPR_TCE. 4275 42764.99 KVM_REINJECT_CONTROL 4277------------------------- 4278 4279:Capability: KVM_CAP_REINJECT_CONTROL 4280:Architectures: x86 4281:Type: vm ioctl 4282:Parameters: struct kvm_reinject_control (in) 4283:Returns: 0 on success, 4284 -EFAULT if struct kvm_reinject_control cannot be read, 4285 -ENXIO if KVM_CREATE_PIT or KVM_CREATE_PIT2 didn't succeed earlier. 4286 4287i8254 (PIT) has two modes, reinject and !reinject. The default is reinject, 4288where KVM queues elapsed i8254 ticks and monitors completion of interrupt from 4289vector(s) that i8254 injects. Reinject mode dequeues a tick and injects its 4290interrupt whenever there isn't a pending interrupt from i8254. 4291!reinject mode injects an interrupt as soon as a tick arrives. 4292 4293:: 4294 4295 struct kvm_reinject_control { 4296 __u8 pit_reinject; 4297 __u8 reserved[31]; 4298 }; 4299 4300pit_reinject = 0 (!reinject mode) is recommended, unless running an old 4301operating system that uses the PIT for timing (e.g. Linux 2.4.x). 4302 43034.100 KVM_PPC_CONFIGURE_V3_MMU 4304------------------------------ 4305 4306:Capability: KVM_CAP_PPC_MMU_RADIX or KVM_CAP_PPC_MMU_HASH_V3 4307:Architectures: ppc 4308:Type: vm ioctl 4309:Parameters: struct kvm_ppc_mmuv3_cfg (in) 4310:Returns: 0 on success, 4311 -EFAULT if struct kvm_ppc_mmuv3_cfg cannot be read, 4312 -EINVAL if the configuration is invalid 4313 4314This ioctl controls whether the guest will use radix or HPT (hashed 4315page table) translation, and sets the pointer to the process table for 4316the guest. 4317 4318:: 4319 4320 struct kvm_ppc_mmuv3_cfg { 4321 __u64 flags; 4322 __u64 process_table; 4323 }; 4324 4325There are two bits that can be set in flags; KVM_PPC_MMUV3_RADIX and 4326KVM_PPC_MMUV3_GTSE. KVM_PPC_MMUV3_RADIX, if set, configures the guest 4327to use radix tree translation, and if clear, to use HPT translation. 4328KVM_PPC_MMUV3_GTSE, if set and if KVM permits it, configures the guest 4329to be able to use the global TLB and SLB invalidation instructions; 4330if clear, the guest may not use these instructions. 4331 4332The process_table field specifies the address and size of the guest 4333process table, which is in the guest's space. This field is formatted 4334as the second doubleword of the partition table entry, as defined in 4335the Power ISA V3.00, Book III section 5.7.6.1. 4336 43374.101 KVM_PPC_GET_RMMU_INFO 4338--------------------------- 4339 4340:Capability: KVM_CAP_PPC_MMU_RADIX 4341:Architectures: ppc 4342:Type: vm ioctl 4343:Parameters: struct kvm_ppc_rmmu_info (out) 4344:Returns: 0 on success, 4345 -EFAULT if struct kvm_ppc_rmmu_info cannot be written, 4346 -EINVAL if no useful information can be returned 4347 4348This ioctl returns a structure containing two things: (a) a list 4349containing supported radix tree geometries, and (b) a list that maps 4350page sizes to put in the "AP" (actual page size) field for the tlbie 4351(TLB invalidate entry) instruction. 4352 4353:: 4354 4355 struct kvm_ppc_rmmu_info { 4356 struct kvm_ppc_radix_geom { 4357 __u8 page_shift; 4358 __u8 level_bits[4]; 4359 __u8 pad[3]; 4360 } geometries[8]; 4361 __u32 ap_encodings[8]; 4362 }; 4363 4364The geometries[] field gives up to 8 supported geometries for the 4365radix page table, in terms of the log base 2 of the smallest page 4366size, and the number of bits indexed at each level of the tree, from 4367the PTE level up to the PGD level in that order. Any unused entries 4368will have 0 in the page_shift field. 4369 4370The ap_encodings gives the supported page sizes and their AP field 4371encodings, encoded with the AP value in the top 3 bits and the log 4372base 2 of the page size in the bottom 6 bits. 4373 43744.102 KVM_PPC_RESIZE_HPT_PREPARE 4375-------------------------------- 4376 4377:Capability: KVM_CAP_SPAPR_RESIZE_HPT 4378:Architectures: powerpc 4379:Type: vm ioctl 4380:Parameters: struct kvm_ppc_resize_hpt (in) 4381:Returns: 0 on successful completion, 4382 >0 if a new HPT is being prepared, the value is an estimated 4383 number of milliseconds until preparation is complete, 4384 -EFAULT if struct kvm_reinject_control cannot be read, 4385 -EINVAL if the supplied shift or flags are invalid, 4386 -ENOMEM if unable to allocate the new HPT, 4387 4388Used to implement the PAPR extension for runtime resizing of a guest's 4389Hashed Page Table (HPT). Specifically this starts, stops or monitors 4390the preparation of a new potential HPT for the guest, essentially 4391implementing the H_RESIZE_HPT_PREPARE hypercall. 4392 4393:: 4394 4395 struct kvm_ppc_resize_hpt { 4396 __u64 flags; 4397 __u32 shift; 4398 __u32 pad; 4399 }; 4400 4401If called with shift > 0 when there is no pending HPT for the guest, 4402this begins preparation of a new pending HPT of size 2^(shift) bytes. 4403It then returns a positive integer with the estimated number of 4404milliseconds until preparation is complete. 4405 4406If called when there is a pending HPT whose size does not match that 4407requested in the parameters, discards the existing pending HPT and 4408creates a new one as above. 4409 4410If called when there is a pending HPT of the size requested, will: 4411 4412 * If preparation of the pending HPT is already complete, return 0 4413 * If preparation of the pending HPT has failed, return an error 4414 code, then discard the pending HPT. 4415 * If preparation of the pending HPT is still in progress, return an 4416 estimated number of milliseconds until preparation is complete. 4417 4418If called with shift == 0, discards any currently pending HPT and 4419returns 0 (i.e. cancels any in-progress preparation). 4420 4421flags is reserved for future expansion, currently setting any bits in 4422flags will result in an -EINVAL. 4423 4424Normally this will be called repeatedly with the same parameters until 4425it returns <= 0. The first call will initiate preparation, subsequent 4426ones will monitor preparation until it completes or fails. 4427 44284.103 KVM_PPC_RESIZE_HPT_COMMIT 4429------------------------------- 4430 4431:Capability: KVM_CAP_SPAPR_RESIZE_HPT 4432:Architectures: powerpc 4433:Type: vm ioctl 4434:Parameters: struct kvm_ppc_resize_hpt (in) 4435:Returns: 0 on successful completion, 4436 -EFAULT if struct kvm_reinject_control cannot be read, 4437 -EINVAL if the supplied shift or flags are invalid, 4438 -ENXIO is there is no pending HPT, or the pending HPT doesn't 4439 have the requested size, 4440 -EBUSY if the pending HPT is not fully prepared, 4441 -ENOSPC if there was a hash collision when moving existing 4442 HPT entries to the new HPT, 4443 -EIO on other error conditions 4444 4445Used to implement the PAPR extension for runtime resizing of a guest's 4446Hashed Page Table (HPT). Specifically this requests that the guest be 4447transferred to working with the new HPT, essentially implementing the 4448H_RESIZE_HPT_COMMIT hypercall. 4449 4450:: 4451 4452 struct kvm_ppc_resize_hpt { 4453 __u64 flags; 4454 __u32 shift; 4455 __u32 pad; 4456 }; 4457 4458This should only be called after KVM_PPC_RESIZE_HPT_PREPARE has 4459returned 0 with the same parameters. In other cases 4460KVM_PPC_RESIZE_HPT_COMMIT will return an error (usually -ENXIO or 4461-EBUSY, though others may be possible if the preparation was started, 4462but failed). 4463 4464This will have undefined effects on the guest if it has not already 4465placed itself in a quiescent state where no vcpu will make MMU enabled 4466memory accesses. 4467 4468On successful completion, the pending HPT will become the guest's active 4469HPT and the previous HPT will be discarded. 4470 4471On failure, the guest will still be operating on its previous HPT. 4472 44734.104 KVM_X86_GET_MCE_CAP_SUPPORTED 4474----------------------------------- 4475 4476:Capability: KVM_CAP_MCE 4477:Architectures: x86 4478:Type: system ioctl 4479:Parameters: u64 mce_cap (out) 4480:Returns: 0 on success, -1 on error 4481 4482Returns supported MCE capabilities. The u64 mce_cap parameter 4483has the same format as the MSR_IA32_MCG_CAP register. Supported 4484capabilities will have the corresponding bits set. 4485 44864.105 KVM_X86_SETUP_MCE 4487----------------------- 4488 4489:Capability: KVM_CAP_MCE 4490:Architectures: x86 4491:Type: vcpu ioctl 4492:Parameters: u64 mcg_cap (in) 4493:Returns: 0 on success, 4494 -EFAULT if u64 mcg_cap cannot be read, 4495 -EINVAL if the requested number of banks is invalid, 4496 -EINVAL if requested MCE capability is not supported. 4497 4498Initializes MCE support for use. The u64 mcg_cap parameter 4499has the same format as the MSR_IA32_MCG_CAP register and 4500specifies which capabilities should be enabled. The maximum 4501supported number of error-reporting banks can be retrieved when 4502checking for KVM_CAP_MCE. The supported capabilities can be 4503retrieved with KVM_X86_GET_MCE_CAP_SUPPORTED. 4504 45054.106 KVM_X86_SET_MCE 4506--------------------- 4507 4508:Capability: KVM_CAP_MCE 4509:Architectures: x86 4510:Type: vcpu ioctl 4511:Parameters: struct kvm_x86_mce (in) 4512:Returns: 0 on success, 4513 -EFAULT if struct kvm_x86_mce cannot be read, 4514 -EINVAL if the bank number is invalid, 4515 -EINVAL if VAL bit is not set in status field. 4516 4517Inject a machine check error (MCE) into the guest. The input 4518parameter is:: 4519 4520 struct kvm_x86_mce { 4521 __u64 status; 4522 __u64 addr; 4523 __u64 misc; 4524 __u64 mcg_status; 4525 __u8 bank; 4526 __u8 pad1[7]; 4527 __u64 pad2[3]; 4528 }; 4529 4530If the MCE being reported is an uncorrected error, KVM will 4531inject it as an MCE exception into the guest. If the guest 4532MCG_STATUS register reports that an MCE is in progress, KVM 4533causes an KVM_EXIT_SHUTDOWN vmexit. 4534 4535Otherwise, if the MCE is a corrected error, KVM will just 4536store it in the corresponding bank (provided this bank is 4537not holding a previously reported uncorrected error). 4538 45394.107 KVM_S390_GET_CMMA_BITS 4540---------------------------- 4541 4542:Capability: KVM_CAP_S390_CMMA_MIGRATION 4543:Architectures: s390 4544:Type: vm ioctl 4545:Parameters: struct kvm_s390_cmma_log (in, out) 4546:Returns: 0 on success, a negative value on error 4547 4548Errors: 4549 4550 ====== ============================================================= 4551 ENOMEM not enough memory can be allocated to complete the task 4552 ENXIO if CMMA is not enabled 4553 EINVAL if KVM_S390_CMMA_PEEK is not set but migration mode was not enabled 4554 EINVAL if KVM_S390_CMMA_PEEK is not set but dirty tracking has been 4555 disabled (and thus migration mode was automatically disabled) 4556 EFAULT if the userspace address is invalid or if no page table is 4557 present for the addresses (e.g. when using hugepages). 4558 ====== ============================================================= 4559 4560This ioctl is used to get the values of the CMMA bits on the s390 4561architecture. It is meant to be used in two scenarios: 4562 4563- During live migration to save the CMMA values. Live migration needs 4564 to be enabled via the KVM_REQ_START_MIGRATION VM property. 4565- To non-destructively peek at the CMMA values, with the flag 4566 KVM_S390_CMMA_PEEK set. 4567 4568The ioctl takes parameters via the kvm_s390_cmma_log struct. The desired 4569values are written to a buffer whose location is indicated via the "values" 4570member in the kvm_s390_cmma_log struct. The values in the input struct are 4571also updated as needed. 4572 4573Each CMMA value takes up one byte. 4574 4575:: 4576 4577 struct kvm_s390_cmma_log { 4578 __u64 start_gfn; 4579 __u32 count; 4580 __u32 flags; 4581 union { 4582 __u64 remaining; 4583 __u64 mask; 4584 }; 4585 __u64 values; 4586 }; 4587 4588start_gfn is the number of the first guest frame whose CMMA values are 4589to be retrieved, 4590 4591count is the length of the buffer in bytes, 4592 4593values points to the buffer where the result will be written to. 4594 4595If count is greater than KVM_S390_SKEYS_MAX, then it is considered to be 4596KVM_S390_SKEYS_MAX. KVM_S390_SKEYS_MAX is re-used for consistency with 4597other ioctls. 4598 4599The result is written in the buffer pointed to by the field values, and 4600the values of the input parameter are updated as follows. 4601 4602Depending on the flags, different actions are performed. The only 4603supported flag so far is KVM_S390_CMMA_PEEK. 4604 4605The default behaviour if KVM_S390_CMMA_PEEK is not set is: 4606start_gfn will indicate the first page frame whose CMMA bits were dirty. 4607It is not necessarily the same as the one passed as input, as clean pages 4608are skipped. 4609 4610count will indicate the number of bytes actually written in the buffer. 4611It can (and very often will) be smaller than the input value, since the 4612buffer is only filled until 16 bytes of clean values are found (which 4613are then not copied in the buffer). Since a CMMA migration block needs 4614the base address and the length, for a total of 16 bytes, we will send 4615back some clean data if there is some dirty data afterwards, as long as 4616the size of the clean data does not exceed the size of the header. This 4617allows to minimize the amount of data to be saved or transferred over 4618the network at the expense of more roundtrips to userspace. The next 4619invocation of the ioctl will skip over all the clean values, saving 4620potentially more than just the 16 bytes we found. 4621 4622If KVM_S390_CMMA_PEEK is set: 4623the existing storage attributes are read even when not in migration 4624mode, and no other action is performed; 4625 4626the output start_gfn will be equal to the input start_gfn, 4627 4628the output count will be equal to the input count, except if the end of 4629memory has been reached. 4630 4631In both cases: 4632the field "remaining" will indicate the total number of dirty CMMA values 4633still remaining, or 0 if KVM_S390_CMMA_PEEK is set and migration mode is 4634not enabled. 4635 4636mask is unused. 4637 4638values points to the userspace buffer where the result will be stored. 4639 46404.108 KVM_S390_SET_CMMA_BITS 4641---------------------------- 4642 4643:Capability: KVM_CAP_S390_CMMA_MIGRATION 4644:Architectures: s390 4645:Type: vm ioctl 4646:Parameters: struct kvm_s390_cmma_log (in) 4647:Returns: 0 on success, a negative value on error 4648 4649This ioctl is used to set the values of the CMMA bits on the s390 4650architecture. It is meant to be used during live migration to restore 4651the CMMA values, but there are no restrictions on its use. 4652The ioctl takes parameters via the kvm_s390_cmma_values struct. 4653Each CMMA value takes up one byte. 4654 4655:: 4656 4657 struct kvm_s390_cmma_log { 4658 __u64 start_gfn; 4659 __u32 count; 4660 __u32 flags; 4661 union { 4662 __u64 remaining; 4663 __u64 mask; 4664 }; 4665 __u64 values; 4666 }; 4667 4668start_gfn indicates the starting guest frame number, 4669 4670count indicates how many values are to be considered in the buffer, 4671 4672flags is not used and must be 0. 4673 4674mask indicates which PGSTE bits are to be considered. 4675 4676remaining is not used. 4677 4678values points to the buffer in userspace where to store the values. 4679 4680This ioctl can fail with -ENOMEM if not enough memory can be allocated to 4681complete the task, with -ENXIO if CMMA is not enabled, with -EINVAL if 4682the count field is too large (e.g. more than KVM_S390_CMMA_SIZE_MAX) or 4683if the flags field was not 0, with -EFAULT if the userspace address is 4684invalid, if invalid pages are written to (e.g. after the end of memory) 4685or if no page table is present for the addresses (e.g. when using 4686hugepages). 4687 46884.109 KVM_PPC_GET_CPU_CHAR 4689-------------------------- 4690 4691:Capability: KVM_CAP_PPC_GET_CPU_CHAR 4692:Architectures: powerpc 4693:Type: vm ioctl 4694:Parameters: struct kvm_ppc_cpu_char (out) 4695:Returns: 0 on successful completion, 4696 -EFAULT if struct kvm_ppc_cpu_char cannot be written 4697 4698This ioctl gives userspace information about certain characteristics 4699of the CPU relating to speculative execution of instructions and 4700possible information leakage resulting from speculative execution (see 4701CVE-2017-5715, CVE-2017-5753 and CVE-2017-5754). The information is 4702returned in struct kvm_ppc_cpu_char, which looks like this:: 4703 4704 struct kvm_ppc_cpu_char { 4705 __u64 character; /* characteristics of the CPU */ 4706 __u64 behaviour; /* recommended software behaviour */ 4707 __u64 character_mask; /* valid bits in character */ 4708 __u64 behaviour_mask; /* valid bits in behaviour */ 4709 }; 4710 4711For extensibility, the character_mask and behaviour_mask fields 4712indicate which bits of character and behaviour have been filled in by 4713the kernel. If the set of defined bits is extended in future then 4714userspace will be able to tell whether it is running on a kernel that 4715knows about the new bits. 4716 4717The character field describes attributes of the CPU which can help 4718with preventing inadvertent information disclosure - specifically, 4719whether there is an instruction to flash-invalidate the L1 data cache 4720(ori 30,30,0 or mtspr SPRN_TRIG2,rN), whether the L1 data cache is set 4721to a mode where entries can only be used by the thread that created 4722them, whether the bcctr[l] instruction prevents speculation, and 4723whether a speculation barrier instruction (ori 31,31,0) is provided. 4724 4725The behaviour field describes actions that software should take to 4726prevent inadvertent information disclosure, and thus describes which 4727vulnerabilities the hardware is subject to; specifically whether the 4728L1 data cache should be flushed when returning to user mode from the 4729kernel, and whether a speculation barrier should be placed between an 4730array bounds check and the array access. 4731 4732These fields use the same bit definitions as the new 4733H_GET_CPU_CHARACTERISTICS hypercall. 4734 47354.110 KVM_MEMORY_ENCRYPT_OP 4736--------------------------- 4737 4738:Capability: basic 4739:Architectures: x86 4740:Type: vm 4741:Parameters: an opaque platform specific structure (in/out) 4742:Returns: 0 on success; -1 on error 4743 4744If the platform supports creating encrypted VMs then this ioctl can be used 4745for issuing platform-specific memory encryption commands to manage those 4746encrypted VMs. 4747 4748Currently, this ioctl is used for issuing Secure Encrypted Virtualization 4749(SEV) commands on AMD Processors. The SEV commands are defined in 4750Documentation/virt/kvm/x86/amd-memory-encryption.rst. 4751 47524.111 KVM_MEMORY_ENCRYPT_REG_REGION 4753----------------------------------- 4754 4755:Capability: basic 4756:Architectures: x86 4757:Type: system 4758:Parameters: struct kvm_enc_region (in) 4759:Returns: 0 on success; -1 on error 4760 4761This ioctl can be used to register a guest memory region which may 4762contain encrypted data (e.g. guest RAM, SMRAM etc). 4763 4764It is used in the SEV-enabled guest. When encryption is enabled, a guest 4765memory region may contain encrypted data. The SEV memory encryption 4766engine uses a tweak such that two identical plaintext pages, each at 4767different locations will have differing ciphertexts. So swapping or 4768moving ciphertext of those pages will not result in plaintext being 4769swapped. So relocating (or migrating) physical backing pages for the SEV 4770guest will require some additional steps. 4771 4772Note: The current SEV key management spec does not provide commands to 4773swap or migrate (move) ciphertext pages. Hence, for now we pin the guest 4774memory region registered with the ioctl. 4775 47764.112 KVM_MEMORY_ENCRYPT_UNREG_REGION 4777------------------------------------- 4778 4779:Capability: basic 4780:Architectures: x86 4781:Type: system 4782:Parameters: struct kvm_enc_region (in) 4783:Returns: 0 on success; -1 on error 4784 4785This ioctl can be used to unregister the guest memory region registered 4786with KVM_MEMORY_ENCRYPT_REG_REGION ioctl above. 4787 47884.113 KVM_HYPERV_EVENTFD 4789------------------------ 4790 4791:Capability: KVM_CAP_HYPERV_EVENTFD 4792:Architectures: x86 4793:Type: vm ioctl 4794:Parameters: struct kvm_hyperv_eventfd (in) 4795 4796This ioctl (un)registers an eventfd to receive notifications from the guest on 4797the specified Hyper-V connection id through the SIGNAL_EVENT hypercall, without 4798causing a user exit. SIGNAL_EVENT hypercall with non-zero event flag number 4799(bits 24-31) still triggers a KVM_EXIT_HYPERV_HCALL user exit. 4800 4801:: 4802 4803 struct kvm_hyperv_eventfd { 4804 __u32 conn_id; 4805 __s32 fd; 4806 __u32 flags; 4807 __u32 padding[3]; 4808 }; 4809 4810The conn_id field should fit within 24 bits:: 4811 4812 #define KVM_HYPERV_CONN_ID_MASK 0x00ffffff 4813 4814The acceptable values for the flags field are:: 4815 4816 #define KVM_HYPERV_EVENTFD_DEASSIGN (1 << 0) 4817 4818:Returns: 0 on success, 4819 -EINVAL if conn_id or flags is outside the allowed range, 4820 -ENOENT on deassign if the conn_id isn't registered, 4821 -EEXIST on assign if the conn_id is already registered 4822 48234.114 KVM_GET_NESTED_STATE 4824-------------------------- 4825 4826:Capability: KVM_CAP_NESTED_STATE 4827:Architectures: x86 4828:Type: vcpu ioctl 4829:Parameters: struct kvm_nested_state (in/out) 4830:Returns: 0 on success, -1 on error 4831 4832Errors: 4833 4834 ===== ============================================================= 4835 E2BIG the total state size exceeds the value of 'size' specified by 4836 the user; the size required will be written into size. 4837 ===== ============================================================= 4838 4839:: 4840 4841 struct kvm_nested_state { 4842 __u16 flags; 4843 __u16 format; 4844 __u32 size; 4845 4846 union { 4847 struct kvm_vmx_nested_state_hdr vmx; 4848 struct kvm_svm_nested_state_hdr svm; 4849 4850 /* Pad the header to 128 bytes. */ 4851 __u8 pad[120]; 4852 } hdr; 4853 4854 union { 4855 struct kvm_vmx_nested_state_data vmx[0]; 4856 struct kvm_svm_nested_state_data svm[0]; 4857 } data; 4858 }; 4859 4860 #define KVM_STATE_NESTED_GUEST_MODE 0x00000001 4861 #define KVM_STATE_NESTED_RUN_PENDING 0x00000002 4862 #define KVM_STATE_NESTED_EVMCS 0x00000004 4863 4864 #define KVM_STATE_NESTED_FORMAT_VMX 0 4865 #define KVM_STATE_NESTED_FORMAT_SVM 1 4866 4867 #define KVM_STATE_NESTED_VMX_VMCS_SIZE 0x1000 4868 4869 #define KVM_STATE_NESTED_VMX_SMM_GUEST_MODE 0x00000001 4870 #define KVM_STATE_NESTED_VMX_SMM_VMXON 0x00000002 4871 4872 #define KVM_STATE_VMX_PREEMPTION_TIMER_DEADLINE 0x00000001 4873 4874 struct kvm_vmx_nested_state_hdr { 4875 __u64 vmxon_pa; 4876 __u64 vmcs12_pa; 4877 4878 struct { 4879 __u16 flags; 4880 } smm; 4881 4882 __u32 flags; 4883 __u64 preemption_timer_deadline; 4884 }; 4885 4886 struct kvm_vmx_nested_state_data { 4887 __u8 vmcs12[KVM_STATE_NESTED_VMX_VMCS_SIZE]; 4888 __u8 shadow_vmcs12[KVM_STATE_NESTED_VMX_VMCS_SIZE]; 4889 }; 4890 4891This ioctl copies the vcpu's nested virtualization state from the kernel to 4892userspace. 4893 4894The maximum size of the state can be retrieved by passing KVM_CAP_NESTED_STATE 4895to the KVM_CHECK_EXTENSION ioctl(). 4896 48974.115 KVM_SET_NESTED_STATE 4898-------------------------- 4899 4900:Capability: KVM_CAP_NESTED_STATE 4901:Architectures: x86 4902:Type: vcpu ioctl 4903:Parameters: struct kvm_nested_state (in) 4904:Returns: 0 on success, -1 on error 4905 4906This copies the vcpu's kvm_nested_state struct from userspace to the kernel. 4907For the definition of struct kvm_nested_state, see KVM_GET_NESTED_STATE. 4908 49094.116 KVM_(UN)REGISTER_COALESCED_MMIO 4910------------------------------------- 4911 4912:Capability: KVM_CAP_COALESCED_MMIO (for coalesced mmio) 4913 KVM_CAP_COALESCED_PIO (for coalesced pio) 4914:Architectures: all 4915:Type: vm ioctl 4916:Parameters: struct kvm_coalesced_mmio_zone 4917:Returns: 0 on success, < 0 on error 4918 4919Coalesced I/O is a performance optimization that defers hardware 4920register write emulation so that userspace exits are avoided. It is 4921typically used to reduce the overhead of emulating frequently accessed 4922hardware registers. 4923 4924When a hardware register is configured for coalesced I/O, write accesses 4925do not exit to userspace and their value is recorded in a ring buffer 4926that is shared between kernel and userspace. 4927 4928Coalesced I/O is used if one or more write accesses to a hardware 4929register can be deferred until a read or a write to another hardware 4930register on the same device. This last access will cause a vmexit and 4931userspace will process accesses from the ring buffer before emulating 4932it. That will avoid exiting to userspace on repeated writes. 4933 4934Coalesced pio is based on coalesced mmio. There is little difference 4935between coalesced mmio and pio except that coalesced pio records accesses 4936to I/O ports. 4937 49384.117 KVM_CLEAR_DIRTY_LOG (vm ioctl) 4939------------------------------------ 4940 4941:Capability: KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 4942:Architectures: x86, arm64, mips 4943:Type: vm ioctl 4944:Parameters: struct kvm_clear_dirty_log (in) 4945:Returns: 0 on success, -1 on error 4946 4947:: 4948 4949 /* for KVM_CLEAR_DIRTY_LOG */ 4950 struct kvm_clear_dirty_log { 4951 __u32 slot; 4952 __u32 num_pages; 4953 __u64 first_page; 4954 union { 4955 void __user *dirty_bitmap; /* one bit per page */ 4956 __u64 padding; 4957 }; 4958 }; 4959 4960The ioctl clears the dirty status of pages in a memory slot, according to 4961the bitmap that is passed in struct kvm_clear_dirty_log's dirty_bitmap 4962field. Bit 0 of the bitmap corresponds to page "first_page" in the 4963memory slot, and num_pages is the size in bits of the input bitmap. 4964first_page must be a multiple of 64; num_pages must also be a multiple of 496564 unless first_page + num_pages is the size of the memory slot. For each 4966bit that is set in the input bitmap, the corresponding page is marked "clean" 4967in KVM's dirty bitmap, and dirty tracking is re-enabled for that page 4968(for example via write-protection, or by clearing the dirty bit in 4969a page table entry). 4970 4971If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of slot field specifies 4972the address space for which you want to clear the dirty status. See 4973KVM_SET_USER_MEMORY_REGION for details on the usage of slot field. 4974 4975This ioctl is mostly useful when KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 4976is enabled; for more information, see the description of the capability. 4977However, it can always be used as long as KVM_CHECK_EXTENSION confirms 4978that KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 is present. 4979 49804.118 KVM_GET_SUPPORTED_HV_CPUID 4981-------------------------------- 4982 4983:Capability: KVM_CAP_HYPERV_CPUID (vcpu), KVM_CAP_SYS_HYPERV_CPUID (system) 4984:Architectures: x86 4985:Type: system ioctl, vcpu ioctl 4986:Parameters: struct kvm_cpuid2 (in/out) 4987:Returns: 0 on success, -1 on error 4988 4989:: 4990 4991 struct kvm_cpuid2 { 4992 __u32 nent; 4993 __u32 padding; 4994 struct kvm_cpuid_entry2 entries[0]; 4995 }; 4996 4997 struct kvm_cpuid_entry2 { 4998 __u32 function; 4999 __u32 index; 5000 __u32 flags; 5001 __u32 eax; 5002 __u32 ebx; 5003 __u32 ecx; 5004 __u32 edx; 5005 __u32 padding[3]; 5006 }; 5007 5008This ioctl returns x86 cpuid features leaves related to Hyper-V emulation in 5009KVM. Userspace can use the information returned by this ioctl to construct 5010cpuid information presented to guests consuming Hyper-V enlightenments (e.g. 5011Windows or Hyper-V guests). 5012 5013CPUID feature leaves returned by this ioctl are defined by Hyper-V Top Level 5014Functional Specification (TLFS). These leaves can't be obtained with 5015KVM_GET_SUPPORTED_CPUID ioctl because some of them intersect with KVM feature 5016leaves (0x40000000, 0x40000001). 5017 5018Currently, the following list of CPUID leaves are returned: 5019 5020 - HYPERV_CPUID_VENDOR_AND_MAX_FUNCTIONS 5021 - HYPERV_CPUID_INTERFACE 5022 - HYPERV_CPUID_VERSION 5023 - HYPERV_CPUID_FEATURES 5024 - HYPERV_CPUID_ENLIGHTMENT_INFO 5025 - HYPERV_CPUID_IMPLEMENT_LIMITS 5026 - HYPERV_CPUID_NESTED_FEATURES 5027 - HYPERV_CPUID_SYNDBG_VENDOR_AND_MAX_FUNCTIONS 5028 - HYPERV_CPUID_SYNDBG_INTERFACE 5029 - HYPERV_CPUID_SYNDBG_PLATFORM_CAPABILITIES 5030 5031Userspace invokes KVM_GET_SUPPORTED_HV_CPUID by passing a kvm_cpuid2 structure 5032with the 'nent' field indicating the number of entries in the variable-size 5033array 'entries'. If the number of entries is too low to describe all Hyper-V 5034feature leaves, an error (E2BIG) is returned. If the number is more or equal 5035to the number of Hyper-V feature leaves, the 'nent' field is adjusted to the 5036number of valid entries in the 'entries' array, which is then filled. 5037 5038'index' and 'flags' fields in 'struct kvm_cpuid_entry2' are currently reserved, 5039userspace should not expect to get any particular value there. 5040 5041Note, vcpu version of KVM_GET_SUPPORTED_HV_CPUID is currently deprecated. Unlike 5042system ioctl which exposes all supported feature bits unconditionally, vcpu 5043version has the following quirks: 5044 5045- HYPERV_CPUID_NESTED_FEATURES leaf and HV_X64_ENLIGHTENED_VMCS_RECOMMENDED 5046 feature bit are only exposed when Enlightened VMCS was previously enabled 5047 on the corresponding vCPU (KVM_CAP_HYPERV_ENLIGHTENED_VMCS). 5048- HV_STIMER_DIRECT_MODE_AVAILABLE bit is only exposed with in-kernel LAPIC. 5049 (presumes KVM_CREATE_IRQCHIP has already been called). 5050 50514.119 KVM_ARM_VCPU_FINALIZE 5052--------------------------- 5053 5054:Architectures: arm64 5055:Type: vcpu ioctl 5056:Parameters: int feature (in) 5057:Returns: 0 on success, -1 on error 5058 5059Errors: 5060 5061 ====== ============================================================== 5062 EPERM feature not enabled, needs configuration, or already finalized 5063 EINVAL feature unknown or not present 5064 ====== ============================================================== 5065 5066Recognised values for feature: 5067 5068 ===== =========================================== 5069 arm64 KVM_ARM_VCPU_SVE (requires KVM_CAP_ARM_SVE) 5070 ===== =========================================== 5071 5072Finalizes the configuration of the specified vcpu feature. 5073 5074The vcpu must already have been initialised, enabling the affected feature, by 5075means of a successful KVM_ARM_VCPU_INIT call with the appropriate flag set in 5076features[]. 5077 5078For affected vcpu features, this is a mandatory step that must be performed 5079before the vcpu is fully usable. 5080 5081Between KVM_ARM_VCPU_INIT and KVM_ARM_VCPU_FINALIZE, the feature may be 5082configured by use of ioctls such as KVM_SET_ONE_REG. The exact configuration 5083that should be performed and how to do it are feature-dependent. 5084 5085Other calls that depend on a particular feature being finalized, such as 5086KVM_RUN, KVM_GET_REG_LIST, KVM_GET_ONE_REG and KVM_SET_ONE_REG, will fail with 5087-EPERM unless the feature has already been finalized by means of a 5088KVM_ARM_VCPU_FINALIZE call. 5089 5090See KVM_ARM_VCPU_INIT for details of vcpu features that require finalization 5091using this ioctl. 5092 50934.120 KVM_SET_PMU_EVENT_FILTER 5094------------------------------ 5095 5096:Capability: KVM_CAP_PMU_EVENT_FILTER 5097:Architectures: x86 5098:Type: vm ioctl 5099:Parameters: struct kvm_pmu_event_filter (in) 5100:Returns: 0 on success, -1 on error 5101 5102Errors: 5103 5104 ====== ============================================================ 5105 EFAULT args[0] cannot be accessed 5106 EINVAL args[0] contains invalid data in the filter or filter events 5107 E2BIG nevents is too large 5108 EBUSY not enough memory to allocate the filter 5109 ====== ============================================================ 5110 5111:: 5112 5113 struct kvm_pmu_event_filter { 5114 __u32 action; 5115 __u32 nevents; 5116 __u32 fixed_counter_bitmap; 5117 __u32 flags; 5118 __u32 pad[4]; 5119 __u64 events[0]; 5120 }; 5121 5122This ioctl restricts the set of PMU events the guest can program by limiting 5123which event select and unit mask combinations are permitted. 5124 5125The argument holds a list of filter events which will be allowed or denied. 5126 5127Filter events only control general purpose counters; fixed purpose counters 5128are controlled by the fixed_counter_bitmap. 5129 5130Valid values for 'flags':: 5131 5132``0`` 5133 5134To use this mode, clear the 'flags' field. 5135 5136In this mode each event will contain an event select + unit mask. 5137 5138When the guest attempts to program the PMU the guest's event select + 5139unit mask is compared against the filter events to determine whether the 5140guest should have access. 5141 5142``KVM_PMU_EVENT_FLAG_MASKED_EVENTS`` 5143:Capability: KVM_CAP_PMU_EVENT_MASKED_EVENTS 5144 5145In this mode each filter event will contain an event select, mask, match, and 5146exclude value. To encode a masked event use:: 5147 5148 KVM_PMU_ENCODE_MASKED_ENTRY() 5149 5150An encoded event will follow this layout:: 5151 5152 Bits Description 5153 ---- ----------- 5154 7:0 event select (low bits) 5155 15:8 umask match 5156 31:16 unused 5157 35:32 event select (high bits) 5158 36:54 unused 5159 55 exclude bit 5160 63:56 umask mask 5161 5162When the guest attempts to program the PMU, these steps are followed in 5163determining if the guest should have access: 5164 5165 1. Match the event select from the guest against the filter events. 5166 2. If a match is found, match the guest's unit mask to the mask and match 5167 values of the included filter events. 5168 I.e. (unit mask & mask) == match && !exclude. 5169 3. If a match is found, match the guest's unit mask to the mask and match 5170 values of the excluded filter events. 5171 I.e. (unit mask & mask) == match && exclude. 5172 4. 5173 a. If an included match is found and an excluded match is not found, filter 5174 the event. 5175 b. For everything else, do not filter the event. 5176 5. 5177 a. If the event is filtered and it's an allow list, allow the guest to 5178 program the event. 5179 b. If the event is filtered and it's a deny list, do not allow the guest to 5180 program the event. 5181 5182When setting a new pmu event filter, -EINVAL will be returned if any of the 5183unused fields are set or if any of the high bits (35:32) in the event 5184select are set when called on Intel. 5185 5186Valid values for 'action':: 5187 5188 #define KVM_PMU_EVENT_ALLOW 0 5189 #define KVM_PMU_EVENT_DENY 1 5190 5191Via this API, KVM userspace can also control the behavior of the VM's fixed 5192counters (if any) by configuring the "action" and "fixed_counter_bitmap" fields. 5193 5194Specifically, KVM follows the following pseudo-code when determining whether to 5195allow the guest FixCtr[i] to count its pre-defined fixed event:: 5196 5197 FixCtr[i]_is_allowed = (action == ALLOW) && (bitmap & BIT(i)) || 5198 (action == DENY) && !(bitmap & BIT(i)); 5199 FixCtr[i]_is_denied = !FixCtr[i]_is_allowed; 5200 5201KVM always consumes fixed_counter_bitmap, it's userspace's responsibility to 5202ensure fixed_counter_bitmap is set correctly, e.g. if userspace wants to define 5203a filter that only affects general purpose counters. 5204 5205Note, the "events" field also applies to fixed counters' hardcoded event_select 5206and unit_mask values. "fixed_counter_bitmap" has higher priority than "events" 5207if there is a contradiction between the two. 5208 52094.121 KVM_PPC_SVM_OFF 5210--------------------- 5211 5212:Capability: basic 5213:Architectures: powerpc 5214:Type: vm ioctl 5215:Parameters: none 5216:Returns: 0 on successful completion, 5217 5218Errors: 5219 5220 ====== ================================================================ 5221 EINVAL if ultravisor failed to terminate the secure guest 5222 ENOMEM if hypervisor failed to allocate new radix page tables for guest 5223 ====== ================================================================ 5224 5225This ioctl is used to turn off the secure mode of the guest or transition 5226the guest from secure mode to normal mode. This is invoked when the guest 5227is reset. This has no effect if called for a normal guest. 5228 5229This ioctl issues an ultravisor call to terminate the secure guest, 5230unpins the VPA pages and releases all the device pages that are used to 5231track the secure pages by hypervisor. 5232 52334.122 KVM_S390_NORMAL_RESET 5234--------------------------- 5235 5236:Capability: KVM_CAP_S390_VCPU_RESETS 5237:Architectures: s390 5238:Type: vcpu ioctl 5239:Parameters: none 5240:Returns: 0 5241 5242This ioctl resets VCPU registers and control structures according to 5243the cpu reset definition in the POP (Principles Of Operation). 5244 52454.123 KVM_S390_INITIAL_RESET 5246---------------------------- 5247 5248:Capability: none 5249:Architectures: s390 5250:Type: vcpu ioctl 5251:Parameters: none 5252:Returns: 0 5253 5254This ioctl resets VCPU registers and control structures according to 5255the initial cpu reset definition in the POP. However, the cpu is not 5256put into ESA mode. This reset is a superset of the normal reset. 5257 52584.124 KVM_S390_CLEAR_RESET 5259-------------------------- 5260 5261:Capability: KVM_CAP_S390_VCPU_RESETS 5262:Architectures: s390 5263:Type: vcpu ioctl 5264:Parameters: none 5265:Returns: 0 5266 5267This ioctl resets VCPU registers and control structures according to 5268the clear cpu reset definition in the POP. However, the cpu is not put 5269into ESA mode. This reset is a superset of the initial reset. 5270 5271 52724.125 KVM_S390_PV_COMMAND 5273------------------------- 5274 5275:Capability: KVM_CAP_S390_PROTECTED 5276:Architectures: s390 5277:Type: vm ioctl 5278:Parameters: struct kvm_pv_cmd 5279:Returns: 0 on success, < 0 on error 5280 5281:: 5282 5283 struct kvm_pv_cmd { 5284 __u32 cmd; /* Command to be executed */ 5285 __u16 rc; /* Ultravisor return code */ 5286 __u16 rrc; /* Ultravisor return reason code */ 5287 __u64 data; /* Data or address */ 5288 __u32 flags; /* flags for future extensions. Must be 0 for now */ 5289 __u32 reserved[3]; 5290 }; 5291 5292**Ultravisor return codes** 5293The Ultravisor return (reason) codes are provided by the kernel if a 5294Ultravisor call has been executed to achieve the results expected by 5295the command. Therefore they are independent of the IOCTL return 5296code. If KVM changes `rc`, its value will always be greater than 0 5297hence setting it to 0 before issuing a PV command is advised to be 5298able to detect a change of `rc`. 5299 5300**cmd values:** 5301 5302KVM_PV_ENABLE 5303 Allocate memory and register the VM with the Ultravisor, thereby 5304 donating memory to the Ultravisor that will become inaccessible to 5305 KVM. All existing CPUs are converted to protected ones. After this 5306 command has succeeded, any CPU added via hotplug will become 5307 protected during its creation as well. 5308 5309 Errors: 5310 5311 ===== ============================= 5312 EINTR an unmasked signal is pending 5313 ===== ============================= 5314 5315KVM_PV_DISABLE 5316 Deregister the VM from the Ultravisor and reclaim the memory that had 5317 been donated to the Ultravisor, making it usable by the kernel again. 5318 All registered VCPUs are converted back to non-protected ones. If a 5319 previous protected VM had been prepared for asynchronous teardown with 5320 KVM_PV_ASYNC_CLEANUP_PREPARE and not subsequently torn down with 5321 KVM_PV_ASYNC_CLEANUP_PERFORM, it will be torn down in this call 5322 together with the current protected VM. 5323 5324KVM_PV_VM_SET_SEC_PARMS 5325 Pass the image header from VM memory to the Ultravisor in 5326 preparation of image unpacking and verification. 5327 5328KVM_PV_VM_UNPACK 5329 Unpack (protect and decrypt) a page of the encrypted boot image. 5330 5331KVM_PV_VM_VERIFY 5332 Verify the integrity of the unpacked image. Only if this succeeds, 5333 KVM is allowed to start protected VCPUs. 5334 5335KVM_PV_INFO 5336 :Capability: KVM_CAP_S390_PROTECTED_DUMP 5337 5338 Presents an API that provides Ultravisor related data to userspace 5339 via subcommands. len_max is the size of the user space buffer, 5340 len_written is KVM's indication of how much bytes of that buffer 5341 were actually written to. len_written can be used to determine the 5342 valid fields if more response fields are added in the future. 5343 5344 :: 5345 5346 enum pv_cmd_info_id { 5347 KVM_PV_INFO_VM, 5348 KVM_PV_INFO_DUMP, 5349 }; 5350 5351 struct kvm_s390_pv_info_header { 5352 __u32 id; 5353 __u32 len_max; 5354 __u32 len_written; 5355 __u32 reserved; 5356 }; 5357 5358 struct kvm_s390_pv_info { 5359 struct kvm_s390_pv_info_header header; 5360 struct kvm_s390_pv_info_dump dump; 5361 struct kvm_s390_pv_info_vm vm; 5362 }; 5363 5364**subcommands:** 5365 5366 KVM_PV_INFO_VM 5367 This subcommand provides basic Ultravisor information for PV 5368 hosts. These values are likely also exported as files in the sysfs 5369 firmware UV query interface but they are more easily available to 5370 programs in this API. 5371 5372 The installed calls and feature_indication members provide the 5373 installed UV calls and the UV's other feature indications. 5374 5375 The max_* members provide information about the maximum number of PV 5376 vcpus, PV guests and PV guest memory size. 5377 5378 :: 5379 5380 struct kvm_s390_pv_info_vm { 5381 __u64 inst_calls_list[4]; 5382 __u64 max_cpus; 5383 __u64 max_guests; 5384 __u64 max_guest_addr; 5385 __u64 feature_indication; 5386 }; 5387 5388 5389 KVM_PV_INFO_DUMP 5390 This subcommand provides information related to dumping PV guests. 5391 5392 :: 5393 5394 struct kvm_s390_pv_info_dump { 5395 __u64 dump_cpu_buffer_len; 5396 __u64 dump_config_mem_buffer_per_1m; 5397 __u64 dump_config_finalize_len; 5398 }; 5399 5400KVM_PV_DUMP 5401 :Capability: KVM_CAP_S390_PROTECTED_DUMP 5402 5403 Presents an API that provides calls which facilitate dumping a 5404 protected VM. 5405 5406 :: 5407 5408 struct kvm_s390_pv_dmp { 5409 __u64 subcmd; 5410 __u64 buff_addr; 5411 __u64 buff_len; 5412 __u64 gaddr; /* For dump storage state */ 5413 }; 5414 5415 **subcommands:** 5416 5417 KVM_PV_DUMP_INIT 5418 Initializes the dump process of a protected VM. If this call does 5419 not succeed all other subcommands will fail with -EINVAL. This 5420 subcommand will return -EINVAL if a dump process has not yet been 5421 completed. 5422 5423 Not all PV vms can be dumped, the owner needs to set `dump 5424 allowed` PCF bit 34 in the SE header to allow dumping. 5425 5426 KVM_PV_DUMP_CONFIG_STOR_STATE 5427 Stores `buff_len` bytes of tweak component values starting with 5428 the 1MB block specified by the absolute guest address 5429 (`gaddr`). `buff_len` needs to be `conf_dump_storage_state_len` 5430 aligned and at least >= the `conf_dump_storage_state_len` value 5431 provided by the dump uv_info data. buff_user might be written to 5432 even if an error rc is returned. For instance if we encounter a 5433 fault after writing the first page of data. 5434 5435 KVM_PV_DUMP_COMPLETE 5436 If the subcommand succeeds it completes the dump process and lets 5437 KVM_PV_DUMP_INIT be called again. 5438 5439 On success `conf_dump_finalize_len` bytes of completion data will be 5440 stored to the `buff_addr`. The completion data contains a key 5441 derivation seed, IV, tweak nonce and encryption keys as well as an 5442 authentication tag all of which are needed to decrypt the dump at a 5443 later time. 5444 5445KVM_PV_ASYNC_CLEANUP_PREPARE 5446 :Capability: KVM_CAP_S390_PROTECTED_ASYNC_DISABLE 5447 5448 Prepare the current protected VM for asynchronous teardown. Most 5449 resources used by the current protected VM will be set aside for a 5450 subsequent asynchronous teardown. The current protected VM will then 5451 resume execution immediately as non-protected. There can be at most 5452 one protected VM prepared for asynchronous teardown at any time. If 5453 a protected VM had already been prepared for teardown without 5454 subsequently calling KVM_PV_ASYNC_CLEANUP_PERFORM, this call will 5455 fail. In that case, the userspace process should issue a normal 5456 KVM_PV_DISABLE. The resources set aside with this call will need to 5457 be cleaned up with a subsequent call to KVM_PV_ASYNC_CLEANUP_PERFORM 5458 or KVM_PV_DISABLE, otherwise they will be cleaned up when KVM 5459 terminates. KVM_PV_ASYNC_CLEANUP_PREPARE can be called again as soon 5460 as cleanup starts, i.e. before KVM_PV_ASYNC_CLEANUP_PERFORM finishes. 5461 5462KVM_PV_ASYNC_CLEANUP_PERFORM 5463 :Capability: KVM_CAP_S390_PROTECTED_ASYNC_DISABLE 5464 5465 Tear down the protected VM previously prepared for teardown with 5466 KVM_PV_ASYNC_CLEANUP_PREPARE. The resources that had been set aside 5467 will be freed during the execution of this command. This PV command 5468 should ideally be issued by userspace from a separate thread. If a 5469 fatal signal is received (or the process terminates naturally), the 5470 command will terminate immediately without completing, and the normal 5471 KVM shutdown procedure will take care of cleaning up all remaining 5472 protected VMs, including the ones whose teardown was interrupted by 5473 process termination. 5474 54754.126 KVM_XEN_HVM_SET_ATTR 5476-------------------------- 5477 5478:Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO 5479:Architectures: x86 5480:Type: vm ioctl 5481:Parameters: struct kvm_xen_hvm_attr 5482:Returns: 0 on success, < 0 on error 5483 5484:: 5485 5486 struct kvm_xen_hvm_attr { 5487 __u16 type; 5488 __u16 pad[3]; 5489 union { 5490 __u8 long_mode; 5491 __u8 vector; 5492 __u8 runstate_update_flag; 5493 union { 5494 __u64 gfn; 5495 __u64 hva; 5496 } shared_info; 5497 struct { 5498 __u32 send_port; 5499 __u32 type; /* EVTCHNSTAT_ipi / EVTCHNSTAT_interdomain */ 5500 __u32 flags; 5501 union { 5502 struct { 5503 __u32 port; 5504 __u32 vcpu; 5505 __u32 priority; 5506 } port; 5507 struct { 5508 __u32 port; /* Zero for eventfd */ 5509 __s32 fd; 5510 } eventfd; 5511 __u32 padding[4]; 5512 } deliver; 5513 } evtchn; 5514 __u32 xen_version; 5515 __u64 pad[8]; 5516 } u; 5517 }; 5518 5519type values: 5520 5521KVM_XEN_ATTR_TYPE_LONG_MODE 5522 Sets the ABI mode of the VM to 32-bit or 64-bit (long mode). This 5523 determines the layout of the shared_info page exposed to the VM. 5524 5525KVM_XEN_ATTR_TYPE_SHARED_INFO 5526 Sets the guest physical frame number at which the Xen shared_info 5527 page resides. Note that although Xen places vcpu_info for the first 5528 32 vCPUs in the shared_info page, KVM does not automatically do so 5529 and instead requires that KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO or 5530 KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO_HVA be used explicitly even when 5531 the vcpu_info for a given vCPU resides at the "default" location 5532 in the shared_info page. This is because KVM may not be aware of 5533 the Xen CPU id which is used as the index into the vcpu_info[] 5534 array, so may know the correct default location. 5535 5536 Note that the shared_info page may be constantly written to by KVM; 5537 it contains the event channel bitmap used to deliver interrupts to 5538 a Xen guest, amongst other things. It is exempt from dirty tracking 5539 mechanisms — KVM will not explicitly mark the page as dirty each 5540 time an event channel interrupt is delivered to the guest! Thus, 5541 userspace should always assume that the designated GFN is dirty if 5542 any vCPU has been running or any event channel interrupts can be 5543 routed to the guest. 5544 5545 Setting the gfn to KVM_XEN_INVALID_GFN will disable the shared_info 5546 page. 5547 5548KVM_XEN_ATTR_TYPE_SHARED_INFO_HVA 5549 If the KVM_XEN_HVM_CONFIG_SHARED_INFO_HVA flag is also set in the 5550 Xen capabilities, then this attribute may be used to set the 5551 userspace address at which the shared_info page resides, which 5552 will always be fixed in the VMM regardless of where it is mapped 5553 in guest physical address space. This attribute should be used in 5554 preference to KVM_XEN_ATTR_TYPE_SHARED_INFO as it avoids 5555 unnecessary invalidation of an internal cache when the page is 5556 re-mapped in guest physcial address space. 5557 5558 Setting the hva to zero will disable the shared_info page. 5559 5560KVM_XEN_ATTR_TYPE_UPCALL_VECTOR 5561 Sets the exception vector used to deliver Xen event channel upcalls. 5562 This is the HVM-wide vector injected directly by the hypervisor 5563 (not through the local APIC), typically configured by a guest via 5564 HVM_PARAM_CALLBACK_IRQ. This can be disabled again (e.g. for guest 5565 SHUTDOWN_soft_reset) by setting it to zero. 5566 5567KVM_XEN_ATTR_TYPE_EVTCHN 5568 This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates 5569 support for KVM_XEN_HVM_CONFIG_EVTCHN_SEND features. It configures 5570 an outbound port number for interception of EVTCHNOP_send requests 5571 from the guest. A given sending port number may be directed back to 5572 a specified vCPU (by APIC ID) / port / priority on the guest, or to 5573 trigger events on an eventfd. The vCPU and priority can be changed 5574 by setting KVM_XEN_EVTCHN_UPDATE in a subsequent call, but other 5575 fields cannot change for a given sending port. A port mapping is 5576 removed by using KVM_XEN_EVTCHN_DEASSIGN in the flags field. Passing 5577 KVM_XEN_EVTCHN_RESET in the flags field removes all interception of 5578 outbound event channels. The values of the flags field are mutually 5579 exclusive and cannot be combined as a bitmask. 5580 5581KVM_XEN_ATTR_TYPE_XEN_VERSION 5582 This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates 5583 support for KVM_XEN_HVM_CONFIG_EVTCHN_SEND features. It configures 5584 the 32-bit version code returned to the guest when it invokes the 5585 XENVER_version call; typically (XEN_MAJOR << 16 | XEN_MINOR). PV 5586 Xen guests will often use this to as a dummy hypercall to trigger 5587 event channel delivery, so responding within the kernel without 5588 exiting to userspace is beneficial. 5589 5590KVM_XEN_ATTR_TYPE_RUNSTATE_UPDATE_FLAG 5591 This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates 5592 support for KVM_XEN_HVM_CONFIG_RUNSTATE_UPDATE_FLAG. It enables the 5593 XEN_RUNSTATE_UPDATE flag which allows guest vCPUs to safely read 5594 other vCPUs' vcpu_runstate_info. Xen guests enable this feature via 5595 the VMASST_TYPE_runstate_update_flag of the HYPERVISOR_vm_assist 5596 hypercall. 5597 55984.127 KVM_XEN_HVM_GET_ATTR 5599-------------------------- 5600 5601:Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO 5602:Architectures: x86 5603:Type: vm ioctl 5604:Parameters: struct kvm_xen_hvm_attr 5605:Returns: 0 on success, < 0 on error 5606 5607Allows Xen VM attributes to be read. For the structure and types, 5608see KVM_XEN_HVM_SET_ATTR above. The KVM_XEN_ATTR_TYPE_EVTCHN 5609attribute cannot be read. 5610 56114.128 KVM_XEN_VCPU_SET_ATTR 5612--------------------------- 5613 5614:Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO 5615:Architectures: x86 5616:Type: vcpu ioctl 5617:Parameters: struct kvm_xen_vcpu_attr 5618:Returns: 0 on success, < 0 on error 5619 5620:: 5621 5622 struct kvm_xen_vcpu_attr { 5623 __u16 type; 5624 __u16 pad[3]; 5625 union { 5626 __u64 gpa; 5627 __u64 pad[4]; 5628 struct { 5629 __u64 state; 5630 __u64 state_entry_time; 5631 __u64 time_running; 5632 __u64 time_runnable; 5633 __u64 time_blocked; 5634 __u64 time_offline; 5635 } runstate; 5636 __u32 vcpu_id; 5637 struct { 5638 __u32 port; 5639 __u32 priority; 5640 __u64 expires_ns; 5641 } timer; 5642 __u8 vector; 5643 } u; 5644 }; 5645 5646type values: 5647 5648KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO 5649 Sets the guest physical address of the vcpu_info for a given vCPU. 5650 As with the shared_info page for the VM, the corresponding page may be 5651 dirtied at any time if event channel interrupt delivery is enabled, so 5652 userspace should always assume that the page is dirty without relying 5653 on dirty logging. Setting the gpa to KVM_XEN_INVALID_GPA will disable 5654 the vcpu_info. 5655 5656KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO_HVA 5657 If the KVM_XEN_HVM_CONFIG_SHARED_INFO_HVA flag is also set in the 5658 Xen capabilities, then this attribute may be used to set the 5659 userspace address of the vcpu_info for a given vCPU. It should 5660 only be used when the vcpu_info resides at the "default" location 5661 in the shared_info page. In this case it is safe to assume the 5662 userspace address will not change, because the shared_info page is 5663 an overlay on guest memory and remains at a fixed host address 5664 regardless of where it is mapped in guest physical address space 5665 and hence unnecessary invalidation of an internal cache may be 5666 avoided if the guest memory layout is modified. 5667 If the vcpu_info does not reside at the "default" location then 5668 it is not guaranteed to remain at the same host address and 5669 hence the aforementioned cache invalidation is required. 5670 5671KVM_XEN_VCPU_ATTR_TYPE_VCPU_TIME_INFO 5672 Sets the guest physical address of an additional pvclock structure 5673 for a given vCPU. This is typically used for guest vsyscall support. 5674 Setting the gpa to KVM_XEN_INVALID_GPA will disable the structure. 5675 5676KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADDR 5677 Sets the guest physical address of the vcpu_runstate_info for a given 5678 vCPU. This is how a Xen guest tracks CPU state such as steal time. 5679 Setting the gpa to KVM_XEN_INVALID_GPA will disable the runstate area. 5680 5681KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_CURRENT 5682 Sets the runstate (RUNSTATE_running/_runnable/_blocked/_offline) of 5683 the given vCPU from the .u.runstate.state member of the structure. 5684 KVM automatically accounts running and runnable time but blocked 5685 and offline states are only entered explicitly. 5686 5687KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_DATA 5688 Sets all fields of the vCPU runstate data from the .u.runstate member 5689 of the structure, including the current runstate. The state_entry_time 5690 must equal the sum of the other four times. 5691 5692KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADJUST 5693 This *adds* the contents of the .u.runstate members of the structure 5694 to the corresponding members of the given vCPU's runstate data, thus 5695 permitting atomic adjustments to the runstate times. The adjustment 5696 to the state_entry_time must equal the sum of the adjustments to the 5697 other four times. The state field must be set to -1, or to a valid 5698 runstate value (RUNSTATE_running, RUNSTATE_runnable, RUNSTATE_blocked 5699 or RUNSTATE_offline) to set the current accounted state as of the 5700 adjusted state_entry_time. 5701 5702KVM_XEN_VCPU_ATTR_TYPE_VCPU_ID 5703 This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates 5704 support for KVM_XEN_HVM_CONFIG_EVTCHN_SEND features. It sets the Xen 5705 vCPU ID of the given vCPU, to allow timer-related VCPU operations to 5706 be intercepted by KVM. 5707 5708KVM_XEN_VCPU_ATTR_TYPE_TIMER 5709 This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates 5710 support for KVM_XEN_HVM_CONFIG_EVTCHN_SEND features. It sets the 5711 event channel port/priority for the VIRQ_TIMER of the vCPU, as well 5712 as allowing a pending timer to be saved/restored. Setting the timer 5713 port to zero disables kernel handling of the singleshot timer. 5714 5715KVM_XEN_VCPU_ATTR_TYPE_UPCALL_VECTOR 5716 This attribute is available when the KVM_CAP_XEN_HVM ioctl indicates 5717 support for KVM_XEN_HVM_CONFIG_EVTCHN_SEND features. It sets the 5718 per-vCPU local APIC upcall vector, configured by a Xen guest with 5719 the HVMOP_set_evtchn_upcall_vector hypercall. This is typically 5720 used by Windows guests, and is distinct from the HVM-wide upcall 5721 vector configured with HVM_PARAM_CALLBACK_IRQ. It is disabled by 5722 setting the vector to zero. 5723 5724 57254.129 KVM_XEN_VCPU_GET_ATTR 5726--------------------------- 5727 5728:Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO 5729:Architectures: x86 5730:Type: vcpu ioctl 5731:Parameters: struct kvm_xen_vcpu_attr 5732:Returns: 0 on success, < 0 on error 5733 5734Allows Xen vCPU attributes to be read. For the structure and types, 5735see KVM_XEN_VCPU_SET_ATTR above. 5736 5737The KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADJUST type may not be used 5738with the KVM_XEN_VCPU_GET_ATTR ioctl. 5739 57404.130 KVM_ARM_MTE_COPY_TAGS 5741--------------------------- 5742 5743:Capability: KVM_CAP_ARM_MTE 5744:Architectures: arm64 5745:Type: vm ioctl 5746:Parameters: struct kvm_arm_copy_mte_tags 5747:Returns: number of bytes copied, < 0 on error (-EINVAL for incorrect 5748 arguments, -EFAULT if memory cannot be accessed). 5749 5750:: 5751 5752 struct kvm_arm_copy_mte_tags { 5753 __u64 guest_ipa; 5754 __u64 length; 5755 void __user *addr; 5756 __u64 flags; 5757 __u64 reserved[2]; 5758 }; 5759 5760Copies Memory Tagging Extension (MTE) tags to/from guest tag memory. The 5761``guest_ipa`` and ``length`` fields must be ``PAGE_SIZE`` aligned. 5762``length`` must not be bigger than 2^31 - PAGE_SIZE bytes. The ``addr`` 5763field must point to a buffer which the tags will be copied to or from. 5764 5765``flags`` specifies the direction of copy, either ``KVM_ARM_TAGS_TO_GUEST`` or 5766``KVM_ARM_TAGS_FROM_GUEST``. 5767 5768The size of the buffer to store the tags is ``(length / 16)`` bytes 5769(granules in MTE are 16 bytes long). Each byte contains a single tag 5770value. This matches the format of ``PTRACE_PEEKMTETAGS`` and 5771``PTRACE_POKEMTETAGS``. 5772 5773If an error occurs before any data is copied then a negative error code is 5774returned. If some tags have been copied before an error occurs then the number 5775of bytes successfully copied is returned. If the call completes successfully 5776then ``length`` is returned. 5777 57784.131 KVM_GET_SREGS2 5779-------------------- 5780 5781:Capability: KVM_CAP_SREGS2 5782:Architectures: x86 5783:Type: vcpu ioctl 5784:Parameters: struct kvm_sregs2 (out) 5785:Returns: 0 on success, -1 on error 5786 5787Reads special registers from the vcpu. 5788This ioctl (when supported) replaces the KVM_GET_SREGS. 5789 5790:: 5791 5792 struct kvm_sregs2 { 5793 /* out (KVM_GET_SREGS2) / in (KVM_SET_SREGS2) */ 5794 struct kvm_segment cs, ds, es, fs, gs, ss; 5795 struct kvm_segment tr, ldt; 5796 struct kvm_dtable gdt, idt; 5797 __u64 cr0, cr2, cr3, cr4, cr8; 5798 __u64 efer; 5799 __u64 apic_base; 5800 __u64 flags; 5801 __u64 pdptrs[4]; 5802 }; 5803 5804flags values for ``kvm_sregs2``: 5805 5806``KVM_SREGS2_FLAGS_PDPTRS_VALID`` 5807 5808 Indicates that the struct contains valid PDPTR values. 5809 5810 58114.132 KVM_SET_SREGS2 5812-------------------- 5813 5814:Capability: KVM_CAP_SREGS2 5815:Architectures: x86 5816:Type: vcpu ioctl 5817:Parameters: struct kvm_sregs2 (in) 5818:Returns: 0 on success, -1 on error 5819 5820Writes special registers into the vcpu. 5821See KVM_GET_SREGS2 for the data structures. 5822This ioctl (when supported) replaces the KVM_SET_SREGS. 5823 58244.133 KVM_GET_STATS_FD 5825---------------------- 5826 5827:Capability: KVM_CAP_STATS_BINARY_FD 5828:Architectures: all 5829:Type: vm ioctl, vcpu ioctl 5830:Parameters: none 5831:Returns: statistics file descriptor on success, < 0 on error 5832 5833Errors: 5834 5835 ====== ====================================================== 5836 ENOMEM if the fd could not be created due to lack of memory 5837 EMFILE if the number of opened files exceeds the limit 5838 ====== ====================================================== 5839 5840The returned file descriptor can be used to read VM/vCPU statistics data in 5841binary format. The data in the file descriptor consists of four blocks 5842organized as follows: 5843 5844+-------------+ 5845| Header | 5846+-------------+ 5847| id string | 5848+-------------+ 5849| Descriptors | 5850+-------------+ 5851| Stats Data | 5852+-------------+ 5853 5854Apart from the header starting at offset 0, please be aware that it is 5855not guaranteed that the four blocks are adjacent or in the above order; 5856the offsets of the id, descriptors and data blocks are found in the 5857header. However, all four blocks are aligned to 64 bit offsets in the 5858file and they do not overlap. 5859 5860All blocks except the data block are immutable. Userspace can read them 5861only one time after retrieving the file descriptor, and then use ``pread`` or 5862``lseek`` to read the statistics repeatedly. 5863 5864All data is in system endianness. 5865 5866The format of the header is as follows:: 5867 5868 struct kvm_stats_header { 5869 __u32 flags; 5870 __u32 name_size; 5871 __u32 num_desc; 5872 __u32 id_offset; 5873 __u32 desc_offset; 5874 __u32 data_offset; 5875 }; 5876 5877The ``flags`` field is not used at the moment. It is always read as 0. 5878 5879The ``name_size`` field is the size (in byte) of the statistics name string 5880(including trailing '\0') which is contained in the "id string" block and 5881appended at the end of every descriptor. 5882 5883The ``num_desc`` field is the number of descriptors that are included in the 5884descriptor block. (The actual number of values in the data block may be 5885larger, since each descriptor may comprise more than one value). 5886 5887The ``id_offset`` field is the offset of the id string from the start of the 5888file indicated by the file descriptor. It is a multiple of 8. 5889 5890The ``desc_offset`` field is the offset of the Descriptors block from the start 5891of the file indicated by the file descriptor. It is a multiple of 8. 5892 5893The ``data_offset`` field is the offset of the Stats Data block from the start 5894of the file indicated by the file descriptor. It is a multiple of 8. 5895 5896The id string block contains a string which identifies the file descriptor on 5897which KVM_GET_STATS_FD was invoked. The size of the block, including the 5898trailing ``'\0'``, is indicated by the ``name_size`` field in the header. 5899 5900The descriptors block is only needed to be read once for the lifetime of the 5901file descriptor contains a sequence of ``struct kvm_stats_desc``, each followed 5902by a string of size ``name_size``. 5903:: 5904 5905 #define KVM_STATS_TYPE_SHIFT 0 5906 #define KVM_STATS_TYPE_MASK (0xF << KVM_STATS_TYPE_SHIFT) 5907 #define KVM_STATS_TYPE_CUMULATIVE (0x0 << KVM_STATS_TYPE_SHIFT) 5908 #define KVM_STATS_TYPE_INSTANT (0x1 << KVM_STATS_TYPE_SHIFT) 5909 #define KVM_STATS_TYPE_PEAK (0x2 << KVM_STATS_TYPE_SHIFT) 5910 #define KVM_STATS_TYPE_LINEAR_HIST (0x3 << KVM_STATS_TYPE_SHIFT) 5911 #define KVM_STATS_TYPE_LOG_HIST (0x4 << KVM_STATS_TYPE_SHIFT) 5912 #define KVM_STATS_TYPE_MAX KVM_STATS_TYPE_LOG_HIST 5913 5914 #define KVM_STATS_UNIT_SHIFT 4 5915 #define KVM_STATS_UNIT_MASK (0xF << KVM_STATS_UNIT_SHIFT) 5916 #define KVM_STATS_UNIT_NONE (0x0 << KVM_STATS_UNIT_SHIFT) 5917 #define KVM_STATS_UNIT_BYTES (0x1 << KVM_STATS_UNIT_SHIFT) 5918 #define KVM_STATS_UNIT_SECONDS (0x2 << KVM_STATS_UNIT_SHIFT) 5919 #define KVM_STATS_UNIT_CYCLES (0x3 << KVM_STATS_UNIT_SHIFT) 5920 #define KVM_STATS_UNIT_BOOLEAN (0x4 << KVM_STATS_UNIT_SHIFT) 5921 #define KVM_STATS_UNIT_MAX KVM_STATS_UNIT_BOOLEAN 5922 5923 #define KVM_STATS_BASE_SHIFT 8 5924 #define KVM_STATS_BASE_MASK (0xF << KVM_STATS_BASE_SHIFT) 5925 #define KVM_STATS_BASE_POW10 (0x0 << KVM_STATS_BASE_SHIFT) 5926 #define KVM_STATS_BASE_POW2 (0x1 << KVM_STATS_BASE_SHIFT) 5927 #define KVM_STATS_BASE_MAX KVM_STATS_BASE_POW2 5928 5929 struct kvm_stats_desc { 5930 __u32 flags; 5931 __s16 exponent; 5932 __u16 size; 5933 __u32 offset; 5934 __u32 bucket_size; 5935 char name[]; 5936 }; 5937 5938The ``flags`` field contains the type and unit of the statistics data described 5939by this descriptor. Its endianness is CPU native. 5940The following flags are supported: 5941 5942Bits 0-3 of ``flags`` encode the type: 5943 5944 * ``KVM_STATS_TYPE_CUMULATIVE`` 5945 The statistics reports a cumulative count. The value of data can only be increased. 5946 Most of the counters used in KVM are of this type. 5947 The corresponding ``size`` field for this type is always 1. 5948 All cumulative statistics data are read/write. 5949 * ``KVM_STATS_TYPE_INSTANT`` 5950 The statistics reports an instantaneous value. Its value can be increased or 5951 decreased. This type is usually used as a measurement of some resources, 5952 like the number of dirty pages, the number of large pages, etc. 5953 All instant statistics are read only. 5954 The corresponding ``size`` field for this type is always 1. 5955 * ``KVM_STATS_TYPE_PEAK`` 5956 The statistics data reports a peak value, for example the maximum number 5957 of items in a hash table bucket, the longest time waited and so on. 5958 The value of data can only be increased. 5959 The corresponding ``size`` field for this type is always 1. 5960 * ``KVM_STATS_TYPE_LINEAR_HIST`` 5961 The statistic is reported as a linear histogram. The number of 5962 buckets is specified by the ``size`` field. The size of buckets is specified 5963 by the ``hist_param`` field. The range of the Nth bucket (1 <= N < ``size``) 5964 is [``hist_param``*(N-1), ``hist_param``*N), while the range of the last 5965 bucket is [``hist_param``*(``size``-1), +INF). (+INF means positive infinity 5966 value.) 5967 * ``KVM_STATS_TYPE_LOG_HIST`` 5968 The statistic is reported as a logarithmic histogram. The number of 5969 buckets is specified by the ``size`` field. The range of the first bucket is 5970 [0, 1), while the range of the last bucket is [pow(2, ``size``-2), +INF). 5971 Otherwise, The Nth bucket (1 < N < ``size``) covers 5972 [pow(2, N-2), pow(2, N-1)). 5973 5974Bits 4-7 of ``flags`` encode the unit: 5975 5976 * ``KVM_STATS_UNIT_NONE`` 5977 There is no unit for the value of statistics data. This usually means that 5978 the value is a simple counter of an event. 5979 * ``KVM_STATS_UNIT_BYTES`` 5980 It indicates that the statistics data is used to measure memory size, in the 5981 unit of Byte, KiByte, MiByte, GiByte, etc. The unit of the data is 5982 determined by the ``exponent`` field in the descriptor. 5983 * ``KVM_STATS_UNIT_SECONDS`` 5984 It indicates that the statistics data is used to measure time or latency. 5985 * ``KVM_STATS_UNIT_CYCLES`` 5986 It indicates that the statistics data is used to measure CPU clock cycles. 5987 * ``KVM_STATS_UNIT_BOOLEAN`` 5988 It indicates that the statistic will always be either 0 or 1. Boolean 5989 statistics of "peak" type will never go back from 1 to 0. Boolean 5990 statistics can be linear histograms (with two buckets) but not logarithmic 5991 histograms. 5992 5993Note that, in the case of histograms, the unit applies to the bucket 5994ranges, while the bucket value indicates how many samples fell in the 5995bucket's range. 5996 5997Bits 8-11 of ``flags``, together with ``exponent``, encode the scale of the 5998unit: 5999 6000 * ``KVM_STATS_BASE_POW10`` 6001 The scale is based on power of 10. It is used for measurement of time and 6002 CPU clock cycles. For example, an exponent of -9 can be used with 6003 ``KVM_STATS_UNIT_SECONDS`` to express that the unit is nanoseconds. 6004 * ``KVM_STATS_BASE_POW2`` 6005 The scale is based on power of 2. It is used for measurement of memory size. 6006 For example, an exponent of 20 can be used with ``KVM_STATS_UNIT_BYTES`` to 6007 express that the unit is MiB. 6008 6009The ``size`` field is the number of values of this statistics data. Its 6010value is usually 1 for most of simple statistics. 1 means it contains an 6011unsigned 64bit data. 6012 6013The ``offset`` field is the offset from the start of Data Block to the start of 6014the corresponding statistics data. 6015 6016The ``bucket_size`` field is used as a parameter for histogram statistics data. 6017It is only used by linear histogram statistics data, specifying the size of a 6018bucket in the unit expressed by bits 4-11 of ``flags`` together with ``exponent``. 6019 6020The ``name`` field is the name string of the statistics data. The name string 6021starts at the end of ``struct kvm_stats_desc``. The maximum length including 6022the trailing ``'\0'``, is indicated by ``name_size`` in the header. 6023 6024The Stats Data block contains an array of 64-bit values in the same order 6025as the descriptors in Descriptors block. 6026 60274.134 KVM_GET_XSAVE2 6028-------------------- 6029 6030:Capability: KVM_CAP_XSAVE2 6031:Architectures: x86 6032:Type: vcpu ioctl 6033:Parameters: struct kvm_xsave (out) 6034:Returns: 0 on success, -1 on error 6035 6036 6037:: 6038 6039 struct kvm_xsave { 6040 __u32 region[1024]; 6041 __u32 extra[0]; 6042 }; 6043 6044This ioctl would copy current vcpu's xsave struct to the userspace. It 6045copies as many bytes as are returned by KVM_CHECK_EXTENSION(KVM_CAP_XSAVE2) 6046when invoked on the vm file descriptor. The size value returned by 6047KVM_CHECK_EXTENSION(KVM_CAP_XSAVE2) will always be at least 4096. 6048Currently, it is only greater than 4096 if a dynamic feature has been 6049enabled with ``arch_prctl()``, but this may change in the future. 6050 6051The offsets of the state save areas in struct kvm_xsave follow the contents 6052of CPUID leaf 0xD on the host. 6053 60544.135 KVM_XEN_HVM_EVTCHN_SEND 6055----------------------------- 6056 6057:Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_EVTCHN_SEND 6058:Architectures: x86 6059:Type: vm ioctl 6060:Parameters: struct kvm_irq_routing_xen_evtchn 6061:Returns: 0 on success, < 0 on error 6062 6063 6064:: 6065 6066 struct kvm_irq_routing_xen_evtchn { 6067 __u32 port; 6068 __u32 vcpu; 6069 __u32 priority; 6070 }; 6071 6072This ioctl injects an event channel interrupt directly to the guest vCPU. 6073 60744.136 KVM_S390_PV_CPU_COMMAND 6075----------------------------- 6076 6077:Capability: KVM_CAP_S390_PROTECTED_DUMP 6078:Architectures: s390 6079:Type: vcpu ioctl 6080:Parameters: none 6081:Returns: 0 on success, < 0 on error 6082 6083This ioctl closely mirrors `KVM_S390_PV_COMMAND` but handles requests 6084for vcpus. It re-uses the kvm_s390_pv_dmp struct and hence also shares 6085the command ids. 6086 6087**command:** 6088 6089KVM_PV_DUMP 6090 Presents an API that provides calls which facilitate dumping a vcpu 6091 of a protected VM. 6092 6093**subcommand:** 6094 6095KVM_PV_DUMP_CPU 6096 Provides encrypted dump data like register values. 6097 The length of the returned data is provided by uv_info.guest_cpu_stor_len. 6098 60994.137 KVM_S390_ZPCI_OP 6100---------------------- 6101 6102:Capability: KVM_CAP_S390_ZPCI_OP 6103:Architectures: s390 6104:Type: vm ioctl 6105:Parameters: struct kvm_s390_zpci_op (in) 6106:Returns: 0 on success, <0 on error 6107 6108Used to manage hardware-assisted virtualization features for zPCI devices. 6109 6110Parameters are specified via the following structure:: 6111 6112 struct kvm_s390_zpci_op { 6113 /* in */ 6114 __u32 fh; /* target device */ 6115 __u8 op; /* operation to perform */ 6116 __u8 pad[3]; 6117 union { 6118 /* for KVM_S390_ZPCIOP_REG_AEN */ 6119 struct { 6120 __u64 ibv; /* Guest addr of interrupt bit vector */ 6121 __u64 sb; /* Guest addr of summary bit */ 6122 __u32 flags; 6123 __u32 noi; /* Number of interrupts */ 6124 __u8 isc; /* Guest interrupt subclass */ 6125 __u8 sbo; /* Offset of guest summary bit vector */ 6126 __u16 pad; 6127 } reg_aen; 6128 __u64 reserved[8]; 6129 } u; 6130 }; 6131 6132The type of operation is specified in the "op" field. 6133KVM_S390_ZPCIOP_REG_AEN is used to register the VM for adapter event 6134notification interpretation, which will allow firmware delivery of adapter 6135events directly to the vm, with KVM providing a backup delivery mechanism; 6136KVM_S390_ZPCIOP_DEREG_AEN is used to subsequently disable interpretation of 6137adapter event notifications. 6138 6139The target zPCI function must also be specified via the "fh" field. For the 6140KVM_S390_ZPCIOP_REG_AEN operation, additional information to establish firmware 6141delivery must be provided via the "reg_aen" struct. 6142 6143The "pad" and "reserved" fields may be used for future extensions and should be 6144set to 0s by userspace. 6145 61464.138 KVM_ARM_SET_COUNTER_OFFSET 6147-------------------------------- 6148 6149:Capability: KVM_CAP_COUNTER_OFFSET 6150:Architectures: arm64 6151:Type: vm ioctl 6152:Parameters: struct kvm_arm_counter_offset (in) 6153:Returns: 0 on success, < 0 on error 6154 6155This capability indicates that userspace is able to apply a single VM-wide 6156offset to both the virtual and physical counters as viewed by the guest 6157using the KVM_ARM_SET_CNT_OFFSET ioctl and the following data structure: 6158 6159:: 6160 6161 struct kvm_arm_counter_offset { 6162 __u64 counter_offset; 6163 __u64 reserved; 6164 }; 6165 6166The offset describes a number of counter cycles that are subtracted from 6167both virtual and physical counter views (similar to the effects of the 6168CNTVOFF_EL2 and CNTPOFF_EL2 system registers, but only global). The offset 6169always applies to all vcpus (already created or created after this ioctl) 6170for this VM. 6171 6172It is userspace's responsibility to compute the offset based, for example, 6173on previous values of the guest counters. 6174 6175Any value other than 0 for the "reserved" field may result in an error 6176(-EINVAL) being returned. This ioctl can also return -EBUSY if any vcpu 6177ioctl is issued concurrently. 6178 6179Note that using this ioctl results in KVM ignoring subsequent userspace 6180writes to the CNTVCT_EL0 and CNTPCT_EL0 registers using the SET_ONE_REG 6181interface. No error will be returned, but the resulting offset will not be 6182applied. 6183 6184.. _KVM_ARM_GET_REG_WRITABLE_MASKS: 6185 61864.139 KVM_ARM_GET_REG_WRITABLE_MASKS 6187------------------------------------------- 6188 6189:Capability: KVM_CAP_ARM_SUPPORTED_REG_MASK_RANGES 6190:Architectures: arm64 6191:Type: vm ioctl 6192:Parameters: struct reg_mask_range (in/out) 6193:Returns: 0 on success, < 0 on error 6194 6195 6196:: 6197 6198 #define KVM_ARM_FEATURE_ID_RANGE 0 6199 #define KVM_ARM_FEATURE_ID_RANGE_SIZE (3 * 8 * 8) 6200 6201 struct reg_mask_range { 6202 __u64 addr; /* Pointer to mask array */ 6203 __u32 range; /* Requested range */ 6204 __u32 reserved[13]; 6205 }; 6206 6207This ioctl copies the writable masks for a selected range of registers to 6208userspace. 6209 6210The ``addr`` field is a pointer to the destination array where KVM copies 6211the writable masks. 6212 6213The ``range`` field indicates the requested range of registers. 6214``KVM_CHECK_EXTENSION`` for the ``KVM_CAP_ARM_SUPPORTED_REG_MASK_RANGES`` 6215capability returns the supported ranges, expressed as a set of flags. Each 6216flag's bit index represents a possible value for the ``range`` field. 6217All other values are reserved for future use and KVM may return an error. 6218 6219The ``reserved[13]`` array is reserved for future use and should be 0, or 6220KVM may return an error. 6221 6222KVM_ARM_FEATURE_ID_RANGE (0) 6223^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 6224 6225The Feature ID range is defined as the AArch64 System register space with 6226op0==3, op1=={0, 1, 3}, CRn==0, CRm=={0-7}, op2=={0-7}. 6227 6228The mask returned array pointed to by ``addr`` is indexed by the macro 6229``ARM64_FEATURE_ID_RANGE_IDX(op0, op1, crn, crm, op2)``, allowing userspace 6230to know what fields can be changed for the system register described by 6231``op0, op1, crn, crm, op2``. KVM rejects ID register values that describe a 6232superset of the features supported by the system. 6233 62344.140 KVM_SET_USER_MEMORY_REGION2 6235--------------------------------- 6236 6237:Capability: KVM_CAP_USER_MEMORY2 6238:Architectures: all 6239:Type: vm ioctl 6240:Parameters: struct kvm_userspace_memory_region2 (in) 6241:Returns: 0 on success, -1 on error 6242 6243KVM_SET_USER_MEMORY_REGION2 is an extension to KVM_SET_USER_MEMORY_REGION that 6244allows mapping guest_memfd memory into a guest. All fields shared with 6245KVM_SET_USER_MEMORY_REGION identically. Userspace can set KVM_MEM_GUEST_MEMFD 6246in flags to have KVM bind the memory region to a given guest_memfd range of 6247[guest_memfd_offset, guest_memfd_offset + memory_size]. The target guest_memfd 6248must point at a file created via KVM_CREATE_GUEST_MEMFD on the current VM, and 6249the target range must not be bound to any other memory region. All standard 6250bounds checks apply (use common sense). 6251 6252:: 6253 6254 struct kvm_userspace_memory_region2 { 6255 __u32 slot; 6256 __u32 flags; 6257 __u64 guest_phys_addr; 6258 __u64 memory_size; /* bytes */ 6259 __u64 userspace_addr; /* start of the userspace allocated memory */ 6260 __u64 guest_memfd_offset; 6261 __u32 guest_memfd; 6262 __u32 pad1; 6263 __u64 pad2[14]; 6264 }; 6265 6266A KVM_MEM_GUEST_MEMFD region _must_ have a valid guest_memfd (private memory) and 6267userspace_addr (shared memory). However, "valid" for userspace_addr simply 6268means that the address itself must be a legal userspace address. The backing 6269mapping for userspace_addr is not required to be valid/populated at the time of 6270KVM_SET_USER_MEMORY_REGION2, e.g. shared memory can be lazily mapped/allocated 6271on-demand. 6272 6273When mapping a gfn into the guest, KVM selects shared vs. private, i.e consumes 6274userspace_addr vs. guest_memfd, based on the gfn's KVM_MEMORY_ATTRIBUTE_PRIVATE 6275state. At VM creation time, all memory is shared, i.e. the PRIVATE attribute 6276is '0' for all gfns. Userspace can control whether memory is shared/private by 6277toggling KVM_MEMORY_ATTRIBUTE_PRIVATE via KVM_SET_MEMORY_ATTRIBUTES as needed. 6278 62794.141 KVM_SET_MEMORY_ATTRIBUTES 6280------------------------------- 6281 6282:Capability: KVM_CAP_MEMORY_ATTRIBUTES 6283:Architectures: x86 6284:Type: vm ioctl 6285:Parameters: struct kvm_memory_attributes (in) 6286:Returns: 0 on success, <0 on error 6287 6288KVM_SET_MEMORY_ATTRIBUTES allows userspace to set memory attributes for a range 6289of guest physical memory. 6290 6291:: 6292 6293 struct kvm_memory_attributes { 6294 __u64 address; 6295 __u64 size; 6296 __u64 attributes; 6297 __u64 flags; 6298 }; 6299 6300 #define KVM_MEMORY_ATTRIBUTE_PRIVATE (1ULL << 3) 6301 6302The address and size must be page aligned. The supported attributes can be 6303retrieved via ioctl(KVM_CHECK_EXTENSION) on KVM_CAP_MEMORY_ATTRIBUTES. If 6304executed on a VM, KVM_CAP_MEMORY_ATTRIBUTES precisely returns the attributes 6305supported by that VM. If executed at system scope, KVM_CAP_MEMORY_ATTRIBUTES 6306returns all attributes supported by KVM. The only attribute defined at this 6307time is KVM_MEMORY_ATTRIBUTE_PRIVATE, which marks the associated gfn as being 6308guest private memory. 6309 6310Note, there is no "get" API. Userspace is responsible for explicitly tracking 6311the state of a gfn/page as needed. 6312 6313The "flags" field is reserved for future extensions and must be '0'. 6314 63154.142 KVM_CREATE_GUEST_MEMFD 6316---------------------------- 6317 6318:Capability: KVM_CAP_GUEST_MEMFD 6319:Architectures: none 6320:Type: vm ioctl 6321:Parameters: struct kvm_create_guest_memfd(in) 6322:Returns: A file descriptor on success, <0 on error 6323 6324KVM_CREATE_GUEST_MEMFD creates an anonymous file and returns a file descriptor 6325that refers to it. guest_memfd files are roughly analogous to files created 6326via memfd_create(), e.g. guest_memfd files live in RAM, have volatile storage, 6327and are automatically released when the last reference is dropped. Unlike 6328"regular" memfd_create() files, guest_memfd files are bound to their owning 6329virtual machine (see below), cannot be mapped, read, or written by userspace, 6330and cannot be resized (guest_memfd files do however support PUNCH_HOLE). 6331 6332:: 6333 6334 struct kvm_create_guest_memfd { 6335 __u64 size; 6336 __u64 flags; 6337 __u64 reserved[6]; 6338 }; 6339 6340Conceptually, the inode backing a guest_memfd file represents physical memory, 6341i.e. is coupled to the virtual machine as a thing, not to a "struct kvm". The 6342file itself, which is bound to a "struct kvm", is that instance's view of the 6343underlying memory, e.g. effectively provides the translation of guest addresses 6344to host memory. This allows for use cases where multiple KVM structures are 6345used to manage a single virtual machine, e.g. when performing intrahost 6346migration of a virtual machine. 6347 6348KVM currently only supports mapping guest_memfd via KVM_SET_USER_MEMORY_REGION2, 6349and more specifically via the guest_memfd and guest_memfd_offset fields in 6350"struct kvm_userspace_memory_region2", where guest_memfd_offset is the offset 6351into the guest_memfd instance. For a given guest_memfd file, there can be at 6352most one mapping per page, i.e. binding multiple memory regions to a single 6353guest_memfd range is not allowed (any number of memory regions can be bound to 6354a single guest_memfd file, but the bound ranges must not overlap). 6355 6356See KVM_SET_USER_MEMORY_REGION2 for additional details. 6357 63585. The kvm_run structure 6359======================== 6360 6361Application code obtains a pointer to the kvm_run structure by 6362mmap()ing a vcpu fd. From that point, application code can control 6363execution by changing fields in kvm_run prior to calling the KVM_RUN 6364ioctl, and obtain information about the reason KVM_RUN returned by 6365looking up structure members. 6366 6367:: 6368 6369 struct kvm_run { 6370 /* in */ 6371 __u8 request_interrupt_window; 6372 6373Request that KVM_RUN return when it becomes possible to inject external 6374interrupts into the guest. Useful in conjunction with KVM_INTERRUPT. 6375 6376:: 6377 6378 __u8 immediate_exit; 6379 6380This field is polled once when KVM_RUN starts; if non-zero, KVM_RUN 6381exits immediately, returning -EINTR. In the common scenario where a 6382signal is used to "kick" a VCPU out of KVM_RUN, this field can be used 6383to avoid usage of KVM_SET_SIGNAL_MASK, which has worse scalability. 6384Rather than blocking the signal outside KVM_RUN, userspace can set up 6385a signal handler that sets run->immediate_exit to a non-zero value. 6386 6387This field is ignored if KVM_CAP_IMMEDIATE_EXIT is not available. 6388 6389:: 6390 6391 __u8 padding1[6]; 6392 6393 /* out */ 6394 __u32 exit_reason; 6395 6396When KVM_RUN has returned successfully (return value 0), this informs 6397application code why KVM_RUN has returned. Allowable values for this 6398field are detailed below. 6399 6400:: 6401 6402 __u8 ready_for_interrupt_injection; 6403 6404If request_interrupt_window has been specified, this field indicates 6405an interrupt can be injected now with KVM_INTERRUPT. 6406 6407:: 6408 6409 __u8 if_flag; 6410 6411The value of the current interrupt flag. Only valid if in-kernel 6412local APIC is not used. 6413 6414:: 6415 6416 __u16 flags; 6417 6418More architecture-specific flags detailing state of the VCPU that may 6419affect the device's behavior. Current defined flags:: 6420 6421 /* x86, set if the VCPU is in system management mode */ 6422 #define KVM_RUN_X86_SMM (1 << 0) 6423 /* x86, set if bus lock detected in VM */ 6424 #define KVM_RUN_BUS_LOCK (1 << 1) 6425 /* arm64, set for KVM_EXIT_DEBUG */ 6426 #define KVM_DEBUG_ARCH_HSR_HIGH_VALID (1 << 0) 6427 6428:: 6429 6430 /* in (pre_kvm_run), out (post_kvm_run) */ 6431 __u64 cr8; 6432 6433The value of the cr8 register. Only valid if in-kernel local APIC is 6434not used. Both input and output. 6435 6436:: 6437 6438 __u64 apic_base; 6439 6440The value of the APIC BASE msr. Only valid if in-kernel local 6441APIC is not used. Both input and output. 6442 6443:: 6444 6445 union { 6446 /* KVM_EXIT_UNKNOWN */ 6447 struct { 6448 __u64 hardware_exit_reason; 6449 } hw; 6450 6451If exit_reason is KVM_EXIT_UNKNOWN, the vcpu has exited due to unknown 6452reasons. Further architecture-specific information is available in 6453hardware_exit_reason. 6454 6455:: 6456 6457 /* KVM_EXIT_FAIL_ENTRY */ 6458 struct { 6459 __u64 hardware_entry_failure_reason; 6460 __u32 cpu; /* if KVM_LAST_CPU */ 6461 } fail_entry; 6462 6463If exit_reason is KVM_EXIT_FAIL_ENTRY, the vcpu could not be run due 6464to unknown reasons. Further architecture-specific information is 6465available in hardware_entry_failure_reason. 6466 6467:: 6468 6469 /* KVM_EXIT_EXCEPTION */ 6470 struct { 6471 __u32 exception; 6472 __u32 error_code; 6473 } ex; 6474 6475Unused. 6476 6477:: 6478 6479 /* KVM_EXIT_IO */ 6480 struct { 6481 #define KVM_EXIT_IO_IN 0 6482 #define KVM_EXIT_IO_OUT 1 6483 __u8 direction; 6484 __u8 size; /* bytes */ 6485 __u16 port; 6486 __u32 count; 6487 __u64 data_offset; /* relative to kvm_run start */ 6488 } io; 6489 6490If exit_reason is KVM_EXIT_IO, then the vcpu has 6491executed a port I/O instruction which could not be satisfied by kvm. 6492data_offset describes where the data is located (KVM_EXIT_IO_OUT) or 6493where kvm expects application code to place the data for the next 6494KVM_RUN invocation (KVM_EXIT_IO_IN). Data format is a packed array. 6495 6496:: 6497 6498 /* KVM_EXIT_DEBUG */ 6499 struct { 6500 struct kvm_debug_exit_arch arch; 6501 } debug; 6502 6503If the exit_reason is KVM_EXIT_DEBUG, then a vcpu is processing a debug event 6504for which architecture specific information is returned. 6505 6506:: 6507 6508 /* KVM_EXIT_MMIO */ 6509 struct { 6510 __u64 phys_addr; 6511 __u8 data[8]; 6512 __u32 len; 6513 __u8 is_write; 6514 } mmio; 6515 6516If exit_reason is KVM_EXIT_MMIO, then the vcpu has 6517executed a memory-mapped I/O instruction which could not be satisfied 6518by kvm. The 'data' member contains the written data if 'is_write' is 6519true, and should be filled by application code otherwise. 6520 6521The 'data' member contains, in its first 'len' bytes, the value as it would 6522appear if the VCPU performed a load or store of the appropriate width directly 6523to the byte array. 6524 6525.. note:: 6526 6527 For KVM_EXIT_IO, KVM_EXIT_MMIO, KVM_EXIT_OSI, KVM_EXIT_PAPR, KVM_EXIT_XEN, 6528 KVM_EXIT_EPR, KVM_EXIT_X86_RDMSR and KVM_EXIT_X86_WRMSR the corresponding 6529 operations are complete (and guest state is consistent) only after userspace 6530 has re-entered the kernel with KVM_RUN. The kernel side will first finish 6531 incomplete operations and then check for pending signals. 6532 6533 The pending state of the operation is not preserved in state which is 6534 visible to userspace, thus userspace should ensure that the operation is 6535 completed before performing a live migration. Userspace can re-enter the 6536 guest with an unmasked signal pending or with the immediate_exit field set 6537 to complete pending operations without allowing any further instructions 6538 to be executed. 6539 6540:: 6541 6542 /* KVM_EXIT_HYPERCALL */ 6543 struct { 6544 __u64 nr; 6545 __u64 args[6]; 6546 __u64 ret; 6547 __u64 flags; 6548 } hypercall; 6549 6550 6551It is strongly recommended that userspace use ``KVM_EXIT_IO`` (x86) or 6552``KVM_EXIT_MMIO`` (all except s390) to implement functionality that 6553requires a guest to interact with host userspace. 6554 6555.. note:: KVM_EXIT_IO is significantly faster than KVM_EXIT_MMIO. 6556 6557For arm64: 6558---------- 6559 6560SMCCC exits can be enabled depending on the configuration of the SMCCC 6561filter. See the Documentation/virt/kvm/devices/vm.rst 6562``KVM_ARM_SMCCC_FILTER`` for more details. 6563 6564``nr`` contains the function ID of the guest's SMCCC call. Userspace is 6565expected to use the ``KVM_GET_ONE_REG`` ioctl to retrieve the call 6566parameters from the vCPU's GPRs. 6567 6568Definition of ``flags``: 6569 - ``KVM_HYPERCALL_EXIT_SMC``: Indicates that the guest used the SMC 6570 conduit to initiate the SMCCC call. If this bit is 0 then the guest 6571 used the HVC conduit for the SMCCC call. 6572 6573 - ``KVM_HYPERCALL_EXIT_16BIT``: Indicates that the guest used a 16bit 6574 instruction to initiate the SMCCC call. If this bit is 0 then the 6575 guest used a 32bit instruction. An AArch64 guest always has this 6576 bit set to 0. 6577 6578At the point of exit, PC points to the instruction immediately following 6579the trapping instruction. 6580 6581:: 6582 6583 /* KVM_EXIT_TPR_ACCESS */ 6584 struct { 6585 __u64 rip; 6586 __u32 is_write; 6587 __u32 pad; 6588 } tpr_access; 6589 6590To be documented (KVM_TPR_ACCESS_REPORTING). 6591 6592:: 6593 6594 /* KVM_EXIT_S390_SIEIC */ 6595 struct { 6596 __u8 icptcode; 6597 __u64 mask; /* psw upper half */ 6598 __u64 addr; /* psw lower half */ 6599 __u16 ipa; 6600 __u32 ipb; 6601 } s390_sieic; 6602 6603s390 specific. 6604 6605:: 6606 6607 /* KVM_EXIT_S390_RESET */ 6608 #define KVM_S390_RESET_POR 1 6609 #define KVM_S390_RESET_CLEAR 2 6610 #define KVM_S390_RESET_SUBSYSTEM 4 6611 #define KVM_S390_RESET_CPU_INIT 8 6612 #define KVM_S390_RESET_IPL 16 6613 __u64 s390_reset_flags; 6614 6615s390 specific. 6616 6617:: 6618 6619 /* KVM_EXIT_S390_UCONTROL */ 6620 struct { 6621 __u64 trans_exc_code; 6622 __u32 pgm_code; 6623 } s390_ucontrol; 6624 6625s390 specific. A page fault has occurred for a user controlled virtual 6626machine (KVM_VM_S390_UNCONTROL) on its host page table that cannot be 6627resolved by the kernel. 6628The program code and the translation exception code that were placed 6629in the cpu's lowcore are presented here as defined by the z Architecture 6630Principles of Operation Book in the Chapter for Dynamic Address Translation 6631(DAT) 6632 6633:: 6634 6635 /* KVM_EXIT_DCR */ 6636 struct { 6637 __u32 dcrn; 6638 __u32 data; 6639 __u8 is_write; 6640 } dcr; 6641 6642Deprecated - was used for 440 KVM. 6643 6644:: 6645 6646 /* KVM_EXIT_OSI */ 6647 struct { 6648 __u64 gprs[32]; 6649 } osi; 6650 6651MOL uses a special hypercall interface it calls 'OSI'. To enable it, we catch 6652hypercalls and exit with this exit struct that contains all the guest gprs. 6653 6654If exit_reason is KVM_EXIT_OSI, then the vcpu has triggered such a hypercall. 6655Userspace can now handle the hypercall and when it's done modify the gprs as 6656necessary. Upon guest entry all guest GPRs will then be replaced by the values 6657in this struct. 6658 6659:: 6660 6661 /* KVM_EXIT_PAPR_HCALL */ 6662 struct { 6663 __u64 nr; 6664 __u64 ret; 6665 __u64 args[9]; 6666 } papr_hcall; 6667 6668This is used on 64-bit PowerPC when emulating a pSeries partition, 6669e.g. with the 'pseries' machine type in qemu. It occurs when the 6670guest does a hypercall using the 'sc 1' instruction. The 'nr' field 6671contains the hypercall number (from the guest R3), and 'args' contains 6672the arguments (from the guest R4 - R12). Userspace should put the 6673return code in 'ret' and any extra returned values in args[]. 6674The possible hypercalls are defined in the Power Architecture Platform 6675Requirements (PAPR) document available from www.power.org (free 6676developer registration required to access it). 6677 6678:: 6679 6680 /* KVM_EXIT_S390_TSCH */ 6681 struct { 6682 __u16 subchannel_id; 6683 __u16 subchannel_nr; 6684 __u32 io_int_parm; 6685 __u32 io_int_word; 6686 __u32 ipb; 6687 __u8 dequeued; 6688 } s390_tsch; 6689 6690s390 specific. This exit occurs when KVM_CAP_S390_CSS_SUPPORT has been enabled 6691and TEST SUBCHANNEL was intercepted. If dequeued is set, a pending I/O 6692interrupt for the target subchannel has been dequeued and subchannel_id, 6693subchannel_nr, io_int_parm and io_int_word contain the parameters for that 6694interrupt. ipb is needed for instruction parameter decoding. 6695 6696:: 6697 6698 /* KVM_EXIT_EPR */ 6699 struct { 6700 __u32 epr; 6701 } epr; 6702 6703On FSL BookE PowerPC chips, the interrupt controller has a fast patch 6704interrupt acknowledge path to the core. When the core successfully 6705delivers an interrupt, it automatically populates the EPR register with 6706the interrupt vector number and acknowledges the interrupt inside 6707the interrupt controller. 6708 6709In case the interrupt controller lives in user space, we need to do 6710the interrupt acknowledge cycle through it to fetch the next to be 6711delivered interrupt vector using this exit. 6712 6713It gets triggered whenever both KVM_CAP_PPC_EPR are enabled and an 6714external interrupt has just been delivered into the guest. User space 6715should put the acknowledged interrupt vector into the 'epr' field. 6716 6717:: 6718 6719 /* KVM_EXIT_SYSTEM_EVENT */ 6720 struct { 6721 #define KVM_SYSTEM_EVENT_SHUTDOWN 1 6722 #define KVM_SYSTEM_EVENT_RESET 2 6723 #define KVM_SYSTEM_EVENT_CRASH 3 6724 #define KVM_SYSTEM_EVENT_WAKEUP 4 6725 #define KVM_SYSTEM_EVENT_SUSPEND 5 6726 #define KVM_SYSTEM_EVENT_SEV_TERM 6 6727 __u32 type; 6728 __u32 ndata; 6729 __u64 data[16]; 6730 } system_event; 6731 6732If exit_reason is KVM_EXIT_SYSTEM_EVENT then the vcpu has triggered 6733a system-level event using some architecture specific mechanism (hypercall 6734or some special instruction). In case of ARM64, this is triggered using 6735HVC instruction based PSCI call from the vcpu. 6736 6737The 'type' field describes the system-level event type. 6738Valid values for 'type' are: 6739 6740 - KVM_SYSTEM_EVENT_SHUTDOWN -- the guest has requested a shutdown of the 6741 VM. Userspace is not obliged to honour this, and if it does honour 6742 this does not need to destroy the VM synchronously (ie it may call 6743 KVM_RUN again before shutdown finally occurs). 6744 - KVM_SYSTEM_EVENT_RESET -- the guest has requested a reset of the VM. 6745 As with SHUTDOWN, userspace can choose to ignore the request, or 6746 to schedule the reset to occur in the future and may call KVM_RUN again. 6747 - KVM_SYSTEM_EVENT_CRASH -- the guest crash occurred and the guest 6748 has requested a crash condition maintenance. Userspace can choose 6749 to ignore the request, or to gather VM memory core dump and/or 6750 reset/shutdown of the VM. 6751 - KVM_SYSTEM_EVENT_SEV_TERM -- an AMD SEV guest requested termination. 6752 The guest physical address of the guest's GHCB is stored in `data[0]`. 6753 - KVM_SYSTEM_EVENT_WAKEUP -- the exiting vCPU is in a suspended state and 6754 KVM has recognized a wakeup event. Userspace may honor this event by 6755 marking the exiting vCPU as runnable, or deny it and call KVM_RUN again. 6756 - KVM_SYSTEM_EVENT_SUSPEND -- the guest has requested a suspension of 6757 the VM. 6758 6759If KVM_CAP_SYSTEM_EVENT_DATA is present, the 'data' field can contain 6760architecture specific information for the system-level event. Only 6761the first `ndata` items (possibly zero) of the data array are valid. 6762 6763 - for arm64, data[0] is set to KVM_SYSTEM_EVENT_RESET_FLAG_PSCI_RESET2 if 6764 the guest issued a SYSTEM_RESET2 call according to v1.1 of the PSCI 6765 specification. 6766 6767 - for RISC-V, data[0] is set to the value of the second argument of the 6768 ``sbi_system_reset`` call. 6769 6770Previous versions of Linux defined a `flags` member in this struct. The 6771field is now aliased to `data[0]`. Userspace can assume that it is only 6772written if ndata is greater than 0. 6773 6774For arm/arm64: 6775-------------- 6776 6777KVM_SYSTEM_EVENT_SUSPEND exits are enabled with the 6778KVM_CAP_ARM_SYSTEM_SUSPEND VM capability. If a guest invokes the PSCI 6779SYSTEM_SUSPEND function, KVM will exit to userspace with this event 6780type. 6781 6782It is the sole responsibility of userspace to implement the PSCI 6783SYSTEM_SUSPEND call according to ARM DEN0022D.b 5.19 "SYSTEM_SUSPEND". 6784KVM does not change the vCPU's state before exiting to userspace, so 6785the call parameters are left in-place in the vCPU registers. 6786 6787Userspace is _required_ to take action for such an exit. It must 6788either: 6789 6790 - Honor the guest request to suspend the VM. Userspace can request 6791 in-kernel emulation of suspension by setting the calling vCPU's 6792 state to KVM_MP_STATE_SUSPENDED. Userspace must configure the vCPU's 6793 state according to the parameters passed to the PSCI function when 6794 the calling vCPU is resumed. See ARM DEN0022D.b 5.19.1 "Intended use" 6795 for details on the function parameters. 6796 6797 - Deny the guest request to suspend the VM. See ARM DEN0022D.b 5.19.2 6798 "Caller responsibilities" for possible return values. 6799 6800:: 6801 6802 /* KVM_EXIT_IOAPIC_EOI */ 6803 struct { 6804 __u8 vector; 6805 } eoi; 6806 6807Indicates that the VCPU's in-kernel local APIC received an EOI for a 6808level-triggered IOAPIC interrupt. This exit only triggers when the 6809IOAPIC is implemented in userspace (i.e. KVM_CAP_SPLIT_IRQCHIP is enabled); 6810the userspace IOAPIC should process the EOI and retrigger the interrupt if 6811it is still asserted. Vector is the LAPIC interrupt vector for which the 6812EOI was received. 6813 6814:: 6815 6816 struct kvm_hyperv_exit { 6817 #define KVM_EXIT_HYPERV_SYNIC 1 6818 #define KVM_EXIT_HYPERV_HCALL 2 6819 #define KVM_EXIT_HYPERV_SYNDBG 3 6820 __u32 type; 6821 __u32 pad1; 6822 union { 6823 struct { 6824 __u32 msr; 6825 __u32 pad2; 6826 __u64 control; 6827 __u64 evt_page; 6828 __u64 msg_page; 6829 } synic; 6830 struct { 6831 __u64 input; 6832 __u64 result; 6833 __u64 params[2]; 6834 } hcall; 6835 struct { 6836 __u32 msr; 6837 __u32 pad2; 6838 __u64 control; 6839 __u64 status; 6840 __u64 send_page; 6841 __u64 recv_page; 6842 __u64 pending_page; 6843 } syndbg; 6844 } u; 6845 }; 6846 /* KVM_EXIT_HYPERV */ 6847 struct kvm_hyperv_exit hyperv; 6848 6849Indicates that the VCPU exits into userspace to process some tasks 6850related to Hyper-V emulation. 6851 6852Valid values for 'type' are: 6853 6854 - KVM_EXIT_HYPERV_SYNIC -- synchronously notify user-space about 6855 6856Hyper-V SynIC state change. Notification is used to remap SynIC 6857event/message pages and to enable/disable SynIC messages/events processing 6858in userspace. 6859 6860 - KVM_EXIT_HYPERV_SYNDBG -- synchronously notify user-space about 6861 6862Hyper-V Synthetic debugger state change. Notification is used to either update 6863the pending_page location or to send a control command (send the buffer located 6864in send_page or recv a buffer to recv_page). 6865 6866:: 6867 6868 /* KVM_EXIT_ARM_NISV */ 6869 struct { 6870 __u64 esr_iss; 6871 __u64 fault_ipa; 6872 } arm_nisv; 6873 6874Used on arm64 systems. If a guest accesses memory not in a memslot, 6875KVM will typically return to userspace and ask it to do MMIO emulation on its 6876behalf. However, for certain classes of instructions, no instruction decode 6877(direction, length of memory access) is provided, and fetching and decoding 6878the instruction from the VM is overly complicated to live in the kernel. 6879 6880Historically, when this situation occurred, KVM would print a warning and kill 6881the VM. KVM assumed that if the guest accessed non-memslot memory, it was 6882trying to do I/O, which just couldn't be emulated, and the warning message was 6883phrased accordingly. However, what happened more often was that a guest bug 6884caused access outside the guest memory areas which should lead to a more 6885meaningful warning message and an external abort in the guest, if the access 6886did not fall within an I/O window. 6887 6888Userspace implementations can query for KVM_CAP_ARM_NISV_TO_USER, and enable 6889this capability at VM creation. Once this is done, these types of errors will 6890instead return to userspace with KVM_EXIT_ARM_NISV, with the valid bits from 6891the ESR_EL2 in the esr_iss field, and the faulting IPA in the fault_ipa field. 6892Userspace can either fix up the access if it's actually an I/O access by 6893decoding the instruction from guest memory (if it's very brave) and continue 6894executing the guest, or it can decide to suspend, dump, or restart the guest. 6895 6896Note that KVM does not skip the faulting instruction as it does for 6897KVM_EXIT_MMIO, but userspace has to emulate any change to the processing state 6898if it decides to decode and emulate the instruction. 6899 6900This feature isn't available to protected VMs, as userspace does not 6901have access to the state that is required to perform the emulation. 6902Instead, a data abort exception is directly injected in the guest. 6903Note that although KVM_CAP_ARM_NISV_TO_USER will be reported if 6904queried outside of a protected VM context, the feature will not be 6905exposed if queried on a protected VM file descriptor. 6906 6907:: 6908 6909 /* KVM_EXIT_X86_RDMSR / KVM_EXIT_X86_WRMSR */ 6910 struct { 6911 __u8 error; /* user -> kernel */ 6912 __u8 pad[7]; 6913 __u32 reason; /* kernel -> user */ 6914 __u32 index; /* kernel -> user */ 6915 __u64 data; /* kernel <-> user */ 6916 } msr; 6917 6918Used on x86 systems. When the VM capability KVM_CAP_X86_USER_SPACE_MSR is 6919enabled, MSR accesses to registers that would invoke a #GP by KVM kernel code 6920may instead trigger a KVM_EXIT_X86_RDMSR exit for reads and KVM_EXIT_X86_WRMSR 6921exit for writes. 6922 6923The "reason" field specifies why the MSR interception occurred. Userspace will 6924only receive MSR exits when a particular reason was requested during through 6925ENABLE_CAP. Currently valid exit reasons are: 6926 6927============================ ======================================== 6928 KVM_MSR_EXIT_REASON_UNKNOWN access to MSR that is unknown to KVM 6929 KVM_MSR_EXIT_REASON_INVAL access to invalid MSRs or reserved bits 6930 KVM_MSR_EXIT_REASON_FILTER access blocked by KVM_X86_SET_MSR_FILTER 6931============================ ======================================== 6932 6933For KVM_EXIT_X86_RDMSR, the "index" field tells userspace which MSR the guest 6934wants to read. To respond to this request with a successful read, userspace 6935writes the respective data into the "data" field and must continue guest 6936execution to ensure the read data is transferred into guest register state. 6937 6938If the RDMSR request was unsuccessful, userspace indicates that with a "1" in 6939the "error" field. This will inject a #GP into the guest when the VCPU is 6940executed again. 6941 6942For KVM_EXIT_X86_WRMSR, the "index" field tells userspace which MSR the guest 6943wants to write. Once finished processing the event, userspace must continue 6944vCPU execution. If the MSR write was unsuccessful, userspace also sets the 6945"error" field to "1". 6946 6947See KVM_X86_SET_MSR_FILTER for details on the interaction with MSR filtering. 6948 6949:: 6950 6951 6952 struct kvm_xen_exit { 6953 #define KVM_EXIT_XEN_HCALL 1 6954 __u32 type; 6955 union { 6956 struct { 6957 __u32 longmode; 6958 __u32 cpl; 6959 __u64 input; 6960 __u64 result; 6961 __u64 params[6]; 6962 } hcall; 6963 } u; 6964 }; 6965 /* KVM_EXIT_XEN */ 6966 struct kvm_hyperv_exit xen; 6967 6968Indicates that the VCPU exits into userspace to process some tasks 6969related to Xen emulation. 6970 6971Valid values for 'type' are: 6972 6973 - KVM_EXIT_XEN_HCALL -- synchronously notify user-space about Xen hypercall. 6974 Userspace is expected to place the hypercall result into the appropriate 6975 field before invoking KVM_RUN again. 6976 6977:: 6978 6979 /* KVM_EXIT_RISCV_SBI */ 6980 struct { 6981 unsigned long extension_id; 6982 unsigned long function_id; 6983 unsigned long args[6]; 6984 unsigned long ret[2]; 6985 } riscv_sbi; 6986 6987If exit reason is KVM_EXIT_RISCV_SBI then it indicates that the VCPU has 6988done a SBI call which is not handled by KVM RISC-V kernel module. The details 6989of the SBI call are available in 'riscv_sbi' member of kvm_run structure. The 6990'extension_id' field of 'riscv_sbi' represents SBI extension ID whereas the 6991'function_id' field represents function ID of given SBI extension. The 'args' 6992array field of 'riscv_sbi' represents parameters for the SBI call and 'ret' 6993array field represents return values. The userspace should update the return 6994values of SBI call before resuming the VCPU. For more details on RISC-V SBI 6995spec refer, https://github.com/riscv/riscv-sbi-doc. 6996 6997:: 6998 6999 /* KVM_EXIT_MEMORY_FAULT */ 7000 struct { 7001 #define KVM_MEMORY_EXIT_FLAG_PRIVATE (1ULL << 3) 7002 __u64 flags; 7003 __u64 gpa; 7004 __u64 size; 7005 } memory_fault; 7006 7007KVM_EXIT_MEMORY_FAULT indicates the vCPU has encountered a memory fault that 7008could not be resolved by KVM. The 'gpa' and 'size' (in bytes) describe the 7009guest physical address range [gpa, gpa + size) of the fault. The 'flags' field 7010describes properties of the faulting access that are likely pertinent: 7011 7012 - KVM_MEMORY_EXIT_FLAG_PRIVATE - When set, indicates the memory fault occurred 7013 on a private memory access. When clear, indicates the fault occurred on a 7014 shared access. 7015 7016Note! KVM_EXIT_MEMORY_FAULT is unique among all KVM exit reasons in that it 7017accompanies a return code of '-1', not '0'! errno will always be set to EFAULT 7018or EHWPOISON when KVM exits with KVM_EXIT_MEMORY_FAULT, userspace should assume 7019kvm_run.exit_reason is stale/undefined for all other error numbers. 7020 7021:: 7022 7023 /* KVM_EXIT_NOTIFY */ 7024 struct { 7025 #define KVM_NOTIFY_CONTEXT_INVALID (1 << 0) 7026 __u32 flags; 7027 } notify; 7028 7029Used on x86 systems. When the VM capability KVM_CAP_X86_NOTIFY_VMEXIT is 7030enabled, a VM exit generated if no event window occurs in VM non-root mode 7031for a specified amount of time. Once KVM_X86_NOTIFY_VMEXIT_USER is set when 7032enabling the cap, it would exit to userspace with the exit reason 7033KVM_EXIT_NOTIFY for further handling. The "flags" field contains more 7034detailed info. 7035 7036The valid value for 'flags' is: 7037 7038 - KVM_NOTIFY_CONTEXT_INVALID -- the VM context is corrupted and not valid 7039 in VMCS. It would run into unknown result if resume the target VM. 7040 7041:: 7042 7043 /* Fix the size of the union. */ 7044 char padding[256]; 7045 }; 7046 7047 /* 7048 * shared registers between kvm and userspace. 7049 * kvm_valid_regs specifies the register classes set by the host 7050 * kvm_dirty_regs specified the register classes dirtied by userspace 7051 * struct kvm_sync_regs is architecture specific, as well as the 7052 * bits for kvm_valid_regs and kvm_dirty_regs 7053 */ 7054 __u64 kvm_valid_regs; 7055 __u64 kvm_dirty_regs; 7056 union { 7057 struct kvm_sync_regs regs; 7058 char padding[SYNC_REGS_SIZE_BYTES]; 7059 } s; 7060 7061If KVM_CAP_SYNC_REGS is defined, these fields allow userspace to access 7062certain guest registers without having to call SET/GET_*REGS. Thus we can 7063avoid some system call overhead if userspace has to handle the exit. 7064Userspace can query the validity of the structure by checking 7065kvm_valid_regs for specific bits. These bits are architecture specific 7066and usually define the validity of a groups of registers. (e.g. one bit 7067for general purpose registers) 7068 7069Please note that the kernel is allowed to use the kvm_run structure as the 7070primary storage for certain register types. Therefore, the kernel may use the 7071values in kvm_run even if the corresponding bit in kvm_dirty_regs is not set. 7072 7073 70746. Capabilities that can be enabled on vCPUs 7075============================================ 7076 7077There are certain capabilities that change the behavior of the virtual CPU or 7078the virtual machine when enabled. To enable them, please see section 4.37. 7079Below you can find a list of capabilities and what their effect on the vCPU or 7080the virtual machine is when enabling them. 7081 7082The following information is provided along with the description: 7083 7084 Architectures: 7085 which instruction set architectures provide this ioctl. 7086 x86 includes both i386 and x86_64. 7087 7088 Target: 7089 whether this is a per-vcpu or per-vm capability. 7090 7091 Parameters: 7092 what parameters are accepted by the capability. 7093 7094 Returns: 7095 the return value. General error numbers (EBADF, ENOMEM, EINVAL) 7096 are not detailed, but errors with specific meanings are. 7097 7098 70996.1 KVM_CAP_PPC_OSI 7100------------------- 7101 7102:Architectures: ppc 7103:Target: vcpu 7104:Parameters: none 7105:Returns: 0 on success; -1 on error 7106 7107This capability enables interception of OSI hypercalls that otherwise would 7108be treated as normal system calls to be injected into the guest. OSI hypercalls 7109were invented by Mac-on-Linux to have a standardized communication mechanism 7110between the guest and the host. 7111 7112When this capability is enabled, KVM_EXIT_OSI can occur. 7113 7114 71156.2 KVM_CAP_PPC_PAPR 7116-------------------- 7117 7118:Architectures: ppc 7119:Target: vcpu 7120:Parameters: none 7121:Returns: 0 on success; -1 on error 7122 7123This capability enables interception of PAPR hypercalls. PAPR hypercalls are 7124done using the hypercall instruction "sc 1". 7125 7126It also sets the guest privilege level to "supervisor" mode. Usually the guest 7127runs in "hypervisor" privilege mode with a few missing features. 7128 7129In addition to the above, it changes the semantics of SDR1. In this mode, the 7130HTAB address part of SDR1 contains an HVA instead of a GPA, as PAPR keeps the 7131HTAB invisible to the guest. 7132 7133When this capability is enabled, KVM_EXIT_PAPR_HCALL can occur. 7134 7135 71366.3 KVM_CAP_SW_TLB 7137------------------ 7138 7139:Architectures: ppc 7140:Target: vcpu 7141:Parameters: args[0] is the address of a struct kvm_config_tlb 7142:Returns: 0 on success; -1 on error 7143 7144:: 7145 7146 struct kvm_config_tlb { 7147 __u64 params; 7148 __u64 array; 7149 __u32 mmu_type; 7150 __u32 array_len; 7151 }; 7152 7153Configures the virtual CPU's TLB array, establishing a shared memory area 7154between userspace and KVM. The "params" and "array" fields are userspace 7155addresses of mmu-type-specific data structures. The "array_len" field is an 7156safety mechanism, and should be set to the size in bytes of the memory that 7157userspace has reserved for the array. It must be at least the size dictated 7158by "mmu_type" and "params". 7159 7160While KVM_RUN is active, the shared region is under control of KVM. Its 7161contents are undefined, and any modification by userspace results in 7162boundedly undefined behavior. 7163 7164On return from KVM_RUN, the shared region will reflect the current state of 7165the guest's TLB. If userspace makes any changes, it must call KVM_DIRTY_TLB 7166to tell KVM which entries have been changed, prior to calling KVM_RUN again 7167on this vcpu. 7168 7169For mmu types KVM_MMU_FSL_BOOKE_NOHV and KVM_MMU_FSL_BOOKE_HV: 7170 7171 - The "params" field is of type "struct kvm_book3e_206_tlb_params". 7172 - The "array" field points to an array of type "struct 7173 kvm_book3e_206_tlb_entry". 7174 - The array consists of all entries in the first TLB, followed by all 7175 entries in the second TLB. 7176 - Within a TLB, entries are ordered first by increasing set number. Within a 7177 set, entries are ordered by way (increasing ESEL). 7178 - The hash for determining set number in TLB0 is: (MAS2 >> 12) & (num_sets - 1) 7179 where "num_sets" is the tlb_sizes[] value divided by the tlb_ways[] value. 7180 - The tsize field of mas1 shall be set to 4K on TLB0, even though the 7181 hardware ignores this value for TLB0. 7182 71836.4 KVM_CAP_S390_CSS_SUPPORT 7184---------------------------- 7185 7186:Architectures: s390 7187:Target: vcpu 7188:Parameters: none 7189:Returns: 0 on success; -1 on error 7190 7191This capability enables support for handling of channel I/O instructions. 7192 7193TEST PENDING INTERRUPTION and the interrupt portion of TEST SUBCHANNEL are 7194handled in-kernel, while the other I/O instructions are passed to userspace. 7195 7196When this capability is enabled, KVM_EXIT_S390_TSCH will occur on TEST 7197SUBCHANNEL intercepts. 7198 7199Note that even though this capability is enabled per-vcpu, the complete 7200virtual machine is affected. 7201 72026.5 KVM_CAP_PPC_EPR 7203------------------- 7204 7205:Architectures: ppc 7206:Target: vcpu 7207:Parameters: args[0] defines whether the proxy facility is active 7208:Returns: 0 on success; -1 on error 7209 7210This capability enables or disables the delivery of interrupts through the 7211external proxy facility. 7212 7213When enabled (args[0] != 0), every time the guest gets an external interrupt 7214delivered, it automatically exits into user space with a KVM_EXIT_EPR exit 7215to receive the topmost interrupt vector. 7216 7217When disabled (args[0] == 0), behavior is as if this facility is unsupported. 7218 7219When this capability is enabled, KVM_EXIT_EPR can occur. 7220 72216.6 KVM_CAP_IRQ_MPIC 7222-------------------- 7223 7224:Architectures: ppc 7225:Parameters: args[0] is the MPIC device fd; 7226 args[1] is the MPIC CPU number for this vcpu 7227 7228This capability connects the vcpu to an in-kernel MPIC device. 7229 72306.7 KVM_CAP_IRQ_XICS 7231-------------------- 7232 7233:Architectures: ppc 7234:Target: vcpu 7235:Parameters: args[0] is the XICS device fd; 7236 args[1] is the XICS CPU number (server ID) for this vcpu 7237 7238This capability connects the vcpu to an in-kernel XICS device. 7239 72406.8 KVM_CAP_S390_IRQCHIP 7241------------------------ 7242 7243:Architectures: s390 7244:Target: vm 7245:Parameters: none 7246 7247This capability enables the in-kernel irqchip for s390. Please refer to 7248"4.24 KVM_CREATE_IRQCHIP" for details. 7249 72506.9 KVM_CAP_MIPS_FPU 7251-------------------- 7252 7253:Architectures: mips 7254:Target: vcpu 7255:Parameters: args[0] is reserved for future use (should be 0). 7256 7257This capability allows the use of the host Floating Point Unit by the guest. It 7258allows the Config1.FP bit to be set to enable the FPU in the guest. Once this is 7259done the ``KVM_REG_MIPS_FPR_*`` and ``KVM_REG_MIPS_FCR_*`` registers can be 7260accessed (depending on the current guest FPU register mode), and the Status.FR, 7261Config5.FRE bits are accessible via the KVM API and also from the guest, 7262depending on them being supported by the FPU. 7263 72646.10 KVM_CAP_MIPS_MSA 7265--------------------- 7266 7267:Architectures: mips 7268:Target: vcpu 7269:Parameters: args[0] is reserved for future use (should be 0). 7270 7271This capability allows the use of the MIPS SIMD Architecture (MSA) by the guest. 7272It allows the Config3.MSAP bit to be set to enable the use of MSA by the guest. 7273Once this is done the ``KVM_REG_MIPS_VEC_*`` and ``KVM_REG_MIPS_MSA_*`` 7274registers can be accessed, and the Config5.MSAEn bit is accessible via the 7275KVM API and also from the guest. 7276 72776.74 KVM_CAP_SYNC_REGS 7278---------------------- 7279 7280:Architectures: s390, x86 7281:Target: s390: always enabled, x86: vcpu 7282:Parameters: none 7283:Returns: x86: KVM_CHECK_EXTENSION returns a bit-array indicating which register 7284 sets are supported 7285 (bitfields defined in arch/x86/include/uapi/asm/kvm.h). 7286 7287As described above in the kvm_sync_regs struct info in section 5 (kvm_run): 7288KVM_CAP_SYNC_REGS "allow[s] userspace to access certain guest registers 7289without having to call SET/GET_*REGS". This reduces overhead by eliminating 7290repeated ioctl calls for setting and/or getting register values. This is 7291particularly important when userspace is making synchronous guest state 7292modifications, e.g. when emulating and/or intercepting instructions in 7293userspace. 7294 7295For s390 specifics, please refer to the source code. 7296 7297For x86: 7298 7299- the register sets to be copied out to kvm_run are selectable 7300 by userspace (rather that all sets being copied out for every exit). 7301- vcpu_events are available in addition to regs and sregs. 7302 7303For x86, the 'kvm_valid_regs' field of struct kvm_run is overloaded to 7304function as an input bit-array field set by userspace to indicate the 7305specific register sets to be copied out on the next exit. 7306 7307To indicate when userspace has modified values that should be copied into 7308the vCPU, the all architecture bitarray field, 'kvm_dirty_regs' must be set. 7309This is done using the same bitflags as for the 'kvm_valid_regs' field. 7310If the dirty bit is not set, then the register set values will not be copied 7311into the vCPU even if they've been modified. 7312 7313Unused bitfields in the bitarrays must be set to zero. 7314 7315:: 7316 7317 struct kvm_sync_regs { 7318 struct kvm_regs regs; 7319 struct kvm_sregs sregs; 7320 struct kvm_vcpu_events events; 7321 }; 7322 73236.75 KVM_CAP_PPC_IRQ_XIVE 7324------------------------- 7325 7326:Architectures: ppc 7327:Target: vcpu 7328:Parameters: args[0] is the XIVE device fd; 7329 args[1] is the XIVE CPU number (server ID) for this vcpu 7330 7331This capability connects the vcpu to an in-kernel XIVE device. 7332 73337. Capabilities that can be enabled on VMs 7334========================================== 7335 7336There are certain capabilities that change the behavior of the virtual 7337machine when enabled. To enable them, please see section 4.37. Below 7338you can find a list of capabilities and what their effect on the VM 7339is when enabling them. 7340 7341The following information is provided along with the description: 7342 7343 Architectures: 7344 which instruction set architectures provide this ioctl. 7345 x86 includes both i386 and x86_64. 7346 7347 Parameters: 7348 what parameters are accepted by the capability. 7349 7350 Returns: 7351 the return value. General error numbers (EBADF, ENOMEM, EINVAL) 7352 are not detailed, but errors with specific meanings are. 7353 7354 73557.1 KVM_CAP_PPC_ENABLE_HCALL 7356---------------------------- 7357 7358:Architectures: ppc 7359:Parameters: args[0] is the sPAPR hcall number; 7360 args[1] is 0 to disable, 1 to enable in-kernel handling 7361 7362This capability controls whether individual sPAPR hypercalls (hcalls) 7363get handled by the kernel or not. Enabling or disabling in-kernel 7364handling of an hcall is effective across the VM. On creation, an 7365initial set of hcalls are enabled for in-kernel handling, which 7366consists of those hcalls for which in-kernel handlers were implemented 7367before this capability was implemented. If disabled, the kernel will 7368not to attempt to handle the hcall, but will always exit to userspace 7369to handle it. Note that it may not make sense to enable some and 7370disable others of a group of related hcalls, but KVM does not prevent 7371userspace from doing that. 7372 7373If the hcall number specified is not one that has an in-kernel 7374implementation, the KVM_ENABLE_CAP ioctl will fail with an EINVAL 7375error. 7376 73777.2 KVM_CAP_S390_USER_SIGP 7378-------------------------- 7379 7380:Architectures: s390 7381:Parameters: none 7382 7383This capability controls which SIGP orders will be handled completely in user 7384space. With this capability enabled, all fast orders will be handled completely 7385in the kernel: 7386 7387- SENSE 7388- SENSE RUNNING 7389- EXTERNAL CALL 7390- EMERGENCY SIGNAL 7391- CONDITIONAL EMERGENCY SIGNAL 7392 7393All other orders will be handled completely in user space. 7394 7395Only privileged operation exceptions will be checked for in the kernel (or even 7396in the hardware prior to interception). If this capability is not enabled, the 7397old way of handling SIGP orders is used (partially in kernel and user space). 7398 73997.3 KVM_CAP_S390_VECTOR_REGISTERS 7400--------------------------------- 7401 7402:Architectures: s390 7403:Parameters: none 7404:Returns: 0 on success, negative value on error 7405 7406Allows use of the vector registers introduced with z13 processor, and 7407provides for the synchronization between host and user space. Will 7408return -EINVAL if the machine does not support vectors. 7409 74107.4 KVM_CAP_S390_USER_STSI 7411-------------------------- 7412 7413:Architectures: s390 7414:Parameters: none 7415 7416This capability allows post-handlers for the STSI instruction. After 7417initial handling in the kernel, KVM exits to user space with 7418KVM_EXIT_S390_STSI to allow user space to insert further data. 7419 7420Before exiting to userspace, kvm handlers should fill in s390_stsi field of 7421vcpu->run:: 7422 7423 struct { 7424 __u64 addr; 7425 __u8 ar; 7426 __u8 reserved; 7427 __u8 fc; 7428 __u8 sel1; 7429 __u16 sel2; 7430 } s390_stsi; 7431 7432 @addr - guest address of STSI SYSIB 7433 @fc - function code 7434 @sel1 - selector 1 7435 @sel2 - selector 2 7436 @ar - access register number 7437 7438KVM handlers should exit to userspace with rc = -EREMOTE. 7439 74407.5 KVM_CAP_SPLIT_IRQCHIP 7441------------------------- 7442 7443:Architectures: x86 7444:Parameters: args[0] - number of routes reserved for userspace IOAPICs 7445:Returns: 0 on success, -1 on error 7446 7447Create a local apic for each processor in the kernel. This can be used 7448instead of KVM_CREATE_IRQCHIP if the userspace VMM wishes to emulate the 7449IOAPIC and PIC (and also the PIT, even though this has to be enabled 7450separately). 7451 7452This capability also enables in kernel routing of interrupt requests; 7453when KVM_CAP_SPLIT_IRQCHIP only routes of KVM_IRQ_ROUTING_MSI type are 7454used in the IRQ routing table. The first args[0] MSI routes are reserved 7455for the IOAPIC pins. Whenever the LAPIC receives an EOI for these routes, 7456a KVM_EXIT_IOAPIC_EOI vmexit will be reported to userspace. 7457 7458Fails if VCPU has already been created, or if the irqchip is already in the 7459kernel (i.e. KVM_CREATE_IRQCHIP has already been called). 7460 74617.6 KVM_CAP_S390_RI 7462------------------- 7463 7464:Architectures: s390 7465:Parameters: none 7466 7467Allows use of runtime-instrumentation introduced with zEC12 processor. 7468Will return -EINVAL if the machine does not support runtime-instrumentation. 7469Will return -EBUSY if a VCPU has already been created. 7470 74717.7 KVM_CAP_X2APIC_API 7472---------------------- 7473 7474:Architectures: x86 7475:Parameters: args[0] - features that should be enabled 7476:Returns: 0 on success, -EINVAL when args[0] contains invalid features 7477 7478Valid feature flags in args[0] are:: 7479 7480 #define KVM_X2APIC_API_USE_32BIT_IDS (1ULL << 0) 7481 #define KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK (1ULL << 1) 7482 7483Enabling KVM_X2APIC_API_USE_32BIT_IDS changes the behavior of 7484KVM_SET_GSI_ROUTING, KVM_SIGNAL_MSI, KVM_SET_LAPIC, and KVM_GET_LAPIC, 7485allowing the use of 32-bit APIC IDs. See KVM_CAP_X2APIC_API in their 7486respective sections. 7487 7488KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK must be enabled for x2APIC to work 7489in logical mode or with more than 255 VCPUs. Otherwise, KVM treats 0xff 7490as a broadcast even in x2APIC mode in order to support physical x2APIC 7491without interrupt remapping. This is undesirable in logical mode, 7492where 0xff represents CPUs 0-7 in cluster 0. 7493 74947.8 KVM_CAP_S390_USER_INSTR0 7495---------------------------- 7496 7497:Architectures: s390 7498:Parameters: none 7499 7500With this capability enabled, all illegal instructions 0x0000 (2 bytes) will 7501be intercepted and forwarded to user space. User space can use this 7502mechanism e.g. to realize 2-byte software breakpoints. The kernel will 7503not inject an operating exception for these instructions, user space has 7504to take care of that. 7505 7506This capability can be enabled dynamically even if VCPUs were already 7507created and are running. 7508 75097.9 KVM_CAP_S390_GS 7510------------------- 7511 7512:Architectures: s390 7513:Parameters: none 7514:Returns: 0 on success; -EINVAL if the machine does not support 7515 guarded storage; -EBUSY if a VCPU has already been created. 7516 7517Allows use of guarded storage for the KVM guest. 7518 75197.10 KVM_CAP_S390_AIS 7520--------------------- 7521 7522:Architectures: s390 7523:Parameters: none 7524 7525Allow use of adapter-interruption suppression. 7526:Returns: 0 on success; -EBUSY if a VCPU has already been created. 7527 75287.11 KVM_CAP_PPC_SMT 7529-------------------- 7530 7531:Architectures: ppc 7532:Parameters: vsmt_mode, flags 7533 7534Enabling this capability on a VM provides userspace with a way to set 7535the desired virtual SMT mode (i.e. the number of virtual CPUs per 7536virtual core). The virtual SMT mode, vsmt_mode, must be a power of 2 7537between 1 and 8. On POWER8, vsmt_mode must also be no greater than 7538the number of threads per subcore for the host. Currently flags must 7539be 0. A successful call to enable this capability will result in 7540vsmt_mode being returned when the KVM_CAP_PPC_SMT capability is 7541subsequently queried for the VM. This capability is only supported by 7542HV KVM, and can only be set before any VCPUs have been created. 7543The KVM_CAP_PPC_SMT_POSSIBLE capability indicates which virtual SMT 7544modes are available. 7545 75467.12 KVM_CAP_PPC_FWNMI 7547---------------------- 7548 7549:Architectures: ppc 7550:Parameters: none 7551 7552With this capability a machine check exception in the guest address 7553space will cause KVM to exit the guest with NMI exit reason. This 7554enables QEMU to build error log and branch to guest kernel registered 7555machine check handling routine. Without this capability KVM will 7556branch to guests' 0x200 interrupt vector. 7557 75587.13 KVM_CAP_X86_DISABLE_EXITS 7559------------------------------ 7560 7561:Architectures: x86 7562:Parameters: args[0] defines which exits are disabled 7563:Returns: 0 on success, -EINVAL when args[0] contains invalid exits 7564 7565Valid bits in args[0] are:: 7566 7567 #define KVM_X86_DISABLE_EXITS_MWAIT (1 << 0) 7568 #define KVM_X86_DISABLE_EXITS_HLT (1 << 1) 7569 #define KVM_X86_DISABLE_EXITS_PAUSE (1 << 2) 7570 #define KVM_X86_DISABLE_EXITS_CSTATE (1 << 3) 7571 7572Enabling this capability on a VM provides userspace with a way to no 7573longer intercept some instructions for improved latency in some 7574workloads, and is suggested when vCPUs are associated to dedicated 7575physical CPUs. More bits can be added in the future; userspace can 7576just pass the KVM_CHECK_EXTENSION result to KVM_ENABLE_CAP to disable 7577all such vmexits. 7578 7579Do not enable KVM_FEATURE_PV_UNHALT if you disable HLT exits. 7580 75817.14 KVM_CAP_S390_HPAGE_1M 7582-------------------------- 7583 7584:Architectures: s390 7585:Parameters: none 7586:Returns: 0 on success, -EINVAL if hpage module parameter was not set 7587 or cmma is enabled, or the VM has the KVM_VM_S390_UCONTROL 7588 flag set 7589 7590With this capability the KVM support for memory backing with 1m pages 7591through hugetlbfs can be enabled for a VM. After the capability is 7592enabled, cmma can't be enabled anymore and pfmfi and the storage key 7593interpretation are disabled. If cmma has already been enabled or the 7594hpage module parameter is not set to 1, -EINVAL is returned. 7595 7596While it is generally possible to create a huge page backed VM without 7597this capability, the VM will not be able to run. 7598 75997.15 KVM_CAP_MSR_PLATFORM_INFO 7600------------------------------ 7601 7602:Architectures: x86 7603:Parameters: args[0] whether feature should be enabled or not 7604 7605With this capability, a guest may read the MSR_PLATFORM_INFO MSR. Otherwise, 7606a #GP would be raised when the guest tries to access. Currently, this 7607capability does not enable write permissions of this MSR for the guest. 7608 76097.16 KVM_CAP_PPC_NESTED_HV 7610-------------------------- 7611 7612:Architectures: ppc 7613:Parameters: none 7614:Returns: 0 on success, -EINVAL when the implementation doesn't support 7615 nested-HV virtualization. 7616 7617HV-KVM on POWER9 and later systems allows for "nested-HV" 7618virtualization, which provides a way for a guest VM to run guests that 7619can run using the CPU's supervisor mode (privileged non-hypervisor 7620state). Enabling this capability on a VM depends on the CPU having 7621the necessary functionality and on the facility being enabled with a 7622kvm-hv module parameter. 7623 76247.17 KVM_CAP_EXCEPTION_PAYLOAD 7625------------------------------ 7626 7627:Architectures: x86 7628:Parameters: args[0] whether feature should be enabled or not 7629 7630With this capability enabled, CR2 will not be modified prior to the 7631emulated VM-exit when L1 intercepts a #PF exception that occurs in 7632L2. Similarly, for kvm-intel only, DR6 will not be modified prior to 7633the emulated VM-exit when L1 intercepts a #DB exception that occurs in 7634L2. As a result, when KVM_GET_VCPU_EVENTS reports a pending #PF (or 7635#DB) exception for L2, exception.has_payload will be set and the 7636faulting address (or the new DR6 bits*) will be reported in the 7637exception_payload field. Similarly, when userspace injects a #PF (or 7638#DB) into L2 using KVM_SET_VCPU_EVENTS, it is expected to set 7639exception.has_payload and to put the faulting address - or the new DR6 7640bits\ [#]_ - in the exception_payload field. 7641 7642This capability also enables exception.pending in struct 7643kvm_vcpu_events, which allows userspace to distinguish between pending 7644and injected exceptions. 7645 7646 7647.. [#] For the new DR6 bits, note that bit 16 is set iff the #DB exception 7648 will clear DR6.RTM. 7649 76507.18 KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 7651-------------------------------------- 7652 7653:Architectures: x86, arm64, mips 7654:Parameters: args[0] whether feature should be enabled or not 7655 7656Valid flags are:: 7657 7658 #define KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE (1 << 0) 7659 #define KVM_DIRTY_LOG_INITIALLY_SET (1 << 1) 7660 7661With KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE is set, KVM_GET_DIRTY_LOG will not 7662automatically clear and write-protect all pages that are returned as dirty. 7663Rather, userspace will have to do this operation separately using 7664KVM_CLEAR_DIRTY_LOG. 7665 7666At the cost of a slightly more complicated operation, this provides better 7667scalability and responsiveness for two reasons. First, 7668KVM_CLEAR_DIRTY_LOG ioctl can operate on a 64-page granularity rather 7669than requiring to sync a full memslot; this ensures that KVM does not 7670take spinlocks for an extended period of time. Second, in some cases a 7671large amount of time can pass between a call to KVM_GET_DIRTY_LOG and 7672userspace actually using the data in the page. Pages can be modified 7673during this time, which is inefficient for both the guest and userspace: 7674the guest will incur a higher penalty due to write protection faults, 7675while userspace can see false reports of dirty pages. Manual reprotection 7676helps reducing this time, improving guest performance and reducing the 7677number of dirty log false positives. 7678 7679With KVM_DIRTY_LOG_INITIALLY_SET set, all the bits of the dirty bitmap 7680will be initialized to 1 when created. This also improves performance because 7681dirty logging can be enabled gradually in small chunks on the first call 7682to KVM_CLEAR_DIRTY_LOG. KVM_DIRTY_LOG_INITIALLY_SET depends on 7683KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE (it is also only available on 7684x86 and arm64 for now). 7685 7686KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 was previously available under the name 7687KVM_CAP_MANUAL_DIRTY_LOG_PROTECT, but the implementation had bugs that make 7688it hard or impossible to use it correctly. The availability of 7689KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 signals that those bugs are fixed. 7690Userspace should not try to use KVM_CAP_MANUAL_DIRTY_LOG_PROTECT. 7691 76927.19 KVM_CAP_PPC_SECURE_GUEST 7693------------------------------ 7694 7695:Architectures: ppc 7696 7697This capability indicates that KVM is running on a host that has 7698ultravisor firmware and thus can support a secure guest. On such a 7699system, a guest can ask the ultravisor to make it a secure guest, 7700one whose memory is inaccessible to the host except for pages which 7701are explicitly requested to be shared with the host. The ultravisor 7702notifies KVM when a guest requests to become a secure guest, and KVM 7703has the opportunity to veto the transition. 7704 7705If present, this capability can be enabled for a VM, meaning that KVM 7706will allow the transition to secure guest mode. Otherwise KVM will 7707veto the transition. 7708 77097.20 KVM_CAP_HALT_POLL 7710---------------------- 7711 7712:Architectures: all 7713:Target: VM 7714:Parameters: args[0] is the maximum poll time in nanoseconds 7715:Returns: 0 on success; -1 on error 7716 7717KVM_CAP_HALT_POLL overrides the kvm.halt_poll_ns module parameter to set the 7718maximum halt-polling time for all vCPUs in the target VM. This capability can 7719be invoked at any time and any number of times to dynamically change the 7720maximum halt-polling time. 7721 7722See Documentation/virt/kvm/halt-polling.rst for more information on halt 7723polling. 7724 77257.21 KVM_CAP_X86_USER_SPACE_MSR 7726------------------------------- 7727 7728:Architectures: x86 7729:Target: VM 7730:Parameters: args[0] contains the mask of KVM_MSR_EXIT_REASON_* events to report 7731:Returns: 0 on success; -1 on error 7732 7733This capability allows userspace to intercept RDMSR and WRMSR instructions if 7734access to an MSR is denied. By default, KVM injects #GP on denied accesses. 7735 7736When a guest requests to read or write an MSR, KVM may not implement all MSRs 7737that are relevant to a respective system. It also does not differentiate by 7738CPU type. 7739 7740To allow more fine grained control over MSR handling, userspace may enable 7741this capability. With it enabled, MSR accesses that match the mask specified in 7742args[0] and would trigger a #GP inside the guest will instead trigger 7743KVM_EXIT_X86_RDMSR and KVM_EXIT_X86_WRMSR exit notifications. Userspace 7744can then implement model specific MSR handling and/or user notifications 7745to inform a user that an MSR was not emulated/virtualized by KVM. 7746 7747The valid mask flags are: 7748 7749============================ =============================================== 7750 KVM_MSR_EXIT_REASON_UNKNOWN intercept accesses to unknown (to KVM) MSRs 7751 KVM_MSR_EXIT_REASON_INVAL intercept accesses that are architecturally 7752 invalid according to the vCPU model and/or mode 7753 KVM_MSR_EXIT_REASON_FILTER intercept accesses that are denied by userspace 7754 via KVM_X86_SET_MSR_FILTER 7755============================ =============================================== 7756 77577.22 KVM_CAP_X86_BUS_LOCK_EXIT 7758------------------------------- 7759 7760:Architectures: x86 7761:Target: VM 7762:Parameters: args[0] defines the policy used when bus locks detected in guest 7763:Returns: 0 on success, -EINVAL when args[0] contains invalid bits 7764 7765Valid bits in args[0] are:: 7766 7767 #define KVM_BUS_LOCK_DETECTION_OFF (1 << 0) 7768 #define KVM_BUS_LOCK_DETECTION_EXIT (1 << 1) 7769 7770Enabling this capability on a VM provides userspace with a way to select 7771a policy to handle the bus locks detected in guest. Userspace can obtain 7772the supported modes from the result of KVM_CHECK_EXTENSION and define it 7773through the KVM_ENABLE_CAP. 7774 7775KVM_BUS_LOCK_DETECTION_OFF and KVM_BUS_LOCK_DETECTION_EXIT are supported 7776currently and mutually exclusive with each other. More bits can be added in 7777the future. 7778 7779With KVM_BUS_LOCK_DETECTION_OFF set, bus locks in guest will not cause vm exits 7780so that no additional actions are needed. This is the default mode. 7781 7782With KVM_BUS_LOCK_DETECTION_EXIT set, vm exits happen when bus lock detected 7783in VM. KVM just exits to userspace when handling them. Userspace can enforce 7784its own throttling or other policy based mitigations. 7785 7786This capability is aimed to address the thread that VM can exploit bus locks to 7787degree the performance of the whole system. Once the userspace enable this 7788capability and select the KVM_BUS_LOCK_DETECTION_EXIT mode, KVM will set the 7789KVM_RUN_BUS_LOCK flag in vcpu-run->flags field and exit to userspace. Concerning 7790the bus lock vm exit can be preempted by a higher priority VM exit, the exit 7791notifications to userspace can be KVM_EXIT_BUS_LOCK or other reasons. 7792KVM_RUN_BUS_LOCK flag is used to distinguish between them. 7793 77947.23 KVM_CAP_PPC_DAWR1 7795---------------------- 7796 7797:Architectures: ppc 7798:Parameters: none 7799:Returns: 0 on success, -EINVAL when CPU doesn't support 2nd DAWR 7800 7801This capability can be used to check / enable 2nd DAWR feature provided 7802by POWER10 processor. 7803 7804 78057.24 KVM_CAP_VM_COPY_ENC_CONTEXT_FROM 7806------------------------------------- 7807 7808Architectures: x86 SEV enabled 7809Type: vm 7810Parameters: args[0] is the fd of the source vm 7811Returns: 0 on success; ENOTTY on error 7812 7813This capability enables userspace to copy encryption context from the vm 7814indicated by the fd to the vm this is called on. 7815 7816This is intended to support in-guest workloads scheduled by the host. This 7817allows the in-guest workload to maintain its own NPTs and keeps the two vms 7818from accidentally clobbering each other with interrupts and the like (separate 7819APIC/MSRs/etc). 7820 78217.25 KVM_CAP_SGX_ATTRIBUTE 7822-------------------------- 7823 7824:Architectures: x86 7825:Target: VM 7826:Parameters: args[0] is a file handle of a SGX attribute file in securityfs 7827:Returns: 0 on success, -EINVAL if the file handle is invalid or if a requested 7828 attribute is not supported by KVM. 7829 7830KVM_CAP_SGX_ATTRIBUTE enables a userspace VMM to grant a VM access to one or 7831more privileged enclave attributes. args[0] must hold a file handle to a valid 7832SGX attribute file corresponding to an attribute that is supported/restricted 7833by KVM (currently only PROVISIONKEY). 7834 7835The SGX subsystem restricts access to a subset of enclave attributes to provide 7836additional security for an uncompromised kernel, e.g. use of the PROVISIONKEY 7837is restricted to deter malware from using the PROVISIONKEY to obtain a stable 7838system fingerprint. To prevent userspace from circumventing such restrictions 7839by running an enclave in a VM, KVM prevents access to privileged attributes by 7840default. 7841 7842See Documentation/arch/x86/sgx.rst for more details. 7843 78447.26 KVM_CAP_PPC_RPT_INVALIDATE 7845------------------------------- 7846 7847:Capability: KVM_CAP_PPC_RPT_INVALIDATE 7848:Architectures: ppc 7849:Type: vm 7850 7851This capability indicates that the kernel is capable of handling 7852H_RPT_INVALIDATE hcall. 7853 7854In order to enable the use of H_RPT_INVALIDATE in the guest, 7855user space might have to advertise it for the guest. For example, 7856IBM pSeries (sPAPR) guest starts using it if "hcall-rpt-invalidate" is 7857present in the "ibm,hypertas-functions" device-tree property. 7858 7859This capability is enabled for hypervisors on platforms like POWER9 7860that support radix MMU. 7861 78627.27 KVM_CAP_EXIT_ON_EMULATION_FAILURE 7863-------------------------------------- 7864 7865:Architectures: x86 7866:Parameters: args[0] whether the feature should be enabled or not 7867 7868When this capability is enabled, an emulation failure will result in an exit 7869to userspace with KVM_INTERNAL_ERROR (except when the emulator was invoked 7870to handle a VMware backdoor instruction). Furthermore, KVM will now provide up 7871to 15 instruction bytes for any exit to userspace resulting from an emulation 7872failure. When these exits to userspace occur use the emulation_failure struct 7873instead of the internal struct. They both have the same layout, but the 7874emulation_failure struct matches the content better. It also explicitly 7875defines the 'flags' field which is used to describe the fields in the struct 7876that are valid (ie: if KVM_INTERNAL_ERROR_EMULATION_FLAG_INSTRUCTION_BYTES is 7877set in the 'flags' field then both 'insn_size' and 'insn_bytes' have valid data 7878in them.) 7879 78807.28 KVM_CAP_ARM_MTE 7881-------------------- 7882 7883:Architectures: arm64 7884:Parameters: none 7885 7886This capability indicates that KVM (and the hardware) supports exposing the 7887Memory Tagging Extensions (MTE) to the guest. It must also be enabled by the 7888VMM before creating any VCPUs to allow the guest access. Note that MTE is only 7889available to a guest running in AArch64 mode and enabling this capability will 7890cause attempts to create AArch32 VCPUs to fail. 7891 7892When enabled the guest is able to access tags associated with any memory given 7893to the guest. KVM will ensure that the tags are maintained during swap or 7894hibernation of the host; however the VMM needs to manually save/restore the 7895tags as appropriate if the VM is migrated. 7896 7897When this capability is enabled all memory in memslots must be mapped as 7898``MAP_ANONYMOUS`` or with a RAM-based file mapping (``tmpfs``, ``memfd``), 7899attempts to create a memslot with an invalid mmap will result in an 7900-EINVAL return. 7901 7902When enabled the VMM may make use of the ``KVM_ARM_MTE_COPY_TAGS`` ioctl to 7903perform a bulk copy of tags to/from the guest. 7904 79057.29 KVM_CAP_VM_MOVE_ENC_CONTEXT_FROM 7906------------------------------------- 7907 7908Architectures: x86 SEV enabled 7909Type: vm 7910Parameters: args[0] is the fd of the source vm 7911Returns: 0 on success 7912 7913This capability enables userspace to migrate the encryption context from the VM 7914indicated by the fd to the VM this is called on. 7915 7916This is intended to support intra-host migration of VMs between userspace VMMs, 7917upgrading the VMM process without interrupting the guest. 7918 79197.30 KVM_CAP_PPC_AIL_MODE_3 7920------------------------------- 7921 7922:Capability: KVM_CAP_PPC_AIL_MODE_3 7923:Architectures: ppc 7924:Type: vm 7925 7926This capability indicates that the kernel supports the mode 3 setting for the 7927"Address Translation Mode on Interrupt" aka "Alternate Interrupt Location" 7928resource that is controlled with the H_SET_MODE hypercall. 7929 7930This capability allows a guest kernel to use a better-performance mode for 7931handling interrupts and system calls. 7932 79337.31 KVM_CAP_DISABLE_QUIRKS2 7934---------------------------- 7935 7936:Capability: KVM_CAP_DISABLE_QUIRKS2 7937:Parameters: args[0] - set of KVM quirks to disable 7938:Architectures: x86 7939:Type: vm 7940 7941This capability, if enabled, will cause KVM to disable some behavior 7942quirks. 7943 7944Calling KVM_CHECK_EXTENSION for this capability returns a bitmask of 7945quirks that can be disabled in KVM. 7946 7947The argument to KVM_ENABLE_CAP for this capability is a bitmask of 7948quirks to disable, and must be a subset of the bitmask returned by 7949KVM_CHECK_EXTENSION. 7950 7951The valid bits in cap.args[0] are: 7952 7953=================================== ============================================ 7954 KVM_X86_QUIRK_LINT0_REENABLED By default, the reset value for the LVT 7955 LINT0 register is 0x700 (APIC_MODE_EXTINT). 7956 When this quirk is disabled, the reset value 7957 is 0x10000 (APIC_LVT_MASKED). 7958 7959 KVM_X86_QUIRK_CD_NW_CLEARED By default, KVM clears CR0.CD and CR0.NW. 7960 When this quirk is disabled, KVM does not 7961 change the value of CR0.CD and CR0.NW. 7962 7963 KVM_X86_QUIRK_LAPIC_MMIO_HOLE By default, the MMIO LAPIC interface is 7964 available even when configured for x2APIC 7965 mode. When this quirk is disabled, KVM 7966 disables the MMIO LAPIC interface if the 7967 LAPIC is in x2APIC mode. 7968 7969 KVM_X86_QUIRK_OUT_7E_INC_RIP By default, KVM pre-increments %rip before 7970 exiting to userspace for an OUT instruction 7971 to port 0x7e. When this quirk is disabled, 7972 KVM does not pre-increment %rip before 7973 exiting to userspace. 7974 7975 KVM_X86_QUIRK_MISC_ENABLE_NO_MWAIT When this quirk is disabled, KVM sets 7976 CPUID.01H:ECX[bit 3] (MONITOR/MWAIT) if 7977 IA32_MISC_ENABLE[bit 18] (MWAIT) is set. 7978 Additionally, when this quirk is disabled, 7979 KVM clears CPUID.01H:ECX[bit 3] if 7980 IA32_MISC_ENABLE[bit 18] is cleared. 7981 7982 KVM_X86_QUIRK_FIX_HYPERCALL_INSN By default, KVM rewrites guest 7983 VMMCALL/VMCALL instructions to match the 7984 vendor's hypercall instruction for the 7985 system. When this quirk is disabled, KVM 7986 will no longer rewrite invalid guest 7987 hypercall instructions. Executing the 7988 incorrect hypercall instruction will 7989 generate a #UD within the guest. 7990 7991KVM_X86_QUIRK_MWAIT_NEVER_UD_FAULTS By default, KVM emulates MONITOR/MWAIT (if 7992 they are intercepted) as NOPs regardless of 7993 whether or not MONITOR/MWAIT are supported 7994 according to guest CPUID. When this quirk 7995 is disabled and KVM_X86_DISABLE_EXITS_MWAIT 7996 is not set (MONITOR/MWAIT are intercepted), 7997 KVM will inject a #UD on MONITOR/MWAIT if 7998 they're unsupported per guest CPUID. Note, 7999 KVM will modify MONITOR/MWAIT support in 8000 guest CPUID on writes to MISC_ENABLE if 8001 KVM_X86_QUIRK_MISC_ENABLE_NO_MWAIT is 8002 disabled. 8003=================================== ============================================ 8004 80057.32 KVM_CAP_MAX_VCPU_ID 8006------------------------ 8007 8008:Architectures: x86 8009:Target: VM 8010:Parameters: args[0] - maximum APIC ID value set for current VM 8011:Returns: 0 on success, -EINVAL if args[0] is beyond KVM_MAX_VCPU_IDS 8012 supported in KVM or if it has been set. 8013 8014This capability allows userspace to specify maximum possible APIC ID 8015assigned for current VM session prior to the creation of vCPUs, saving 8016memory for data structures indexed by the APIC ID. Userspace is able 8017to calculate the limit to APIC ID values from designated 8018CPU topology. 8019 8020The value can be changed only until KVM_ENABLE_CAP is set to a nonzero 8021value or until a vCPU is created. Upon creation of the first vCPU, 8022if the value was set to zero or KVM_ENABLE_CAP was not invoked, KVM 8023uses the return value of KVM_CHECK_EXTENSION(KVM_CAP_MAX_VCPU_ID) as 8024the maximum APIC ID. 8025 80267.33 KVM_CAP_X86_NOTIFY_VMEXIT 8027------------------------------ 8028 8029:Architectures: x86 8030:Target: VM 8031:Parameters: args[0] is the value of notify window as well as some flags 8032:Returns: 0 on success, -EINVAL if args[0] contains invalid flags or notify 8033 VM exit is unsupported. 8034 8035Bits 63:32 of args[0] are used for notify window. 8036Bits 31:0 of args[0] are for some flags. Valid bits are:: 8037 8038 #define KVM_X86_NOTIFY_VMEXIT_ENABLED (1 << 0) 8039 #define KVM_X86_NOTIFY_VMEXIT_USER (1 << 1) 8040 8041This capability allows userspace to configure the notify VM exit on/off 8042in per-VM scope during VM creation. Notify VM exit is disabled by default. 8043When userspace sets KVM_X86_NOTIFY_VMEXIT_ENABLED bit in args[0], VMM will 8044enable this feature with the notify window provided, which will generate 8045a VM exit if no event window occurs in VM non-root mode for a specified of 8046time (notify window). 8047 8048If KVM_X86_NOTIFY_VMEXIT_USER is set in args[0], upon notify VM exits happen, 8049KVM would exit to userspace for handling. 8050 8051This capability is aimed to mitigate the threat that malicious VMs can 8052cause CPU stuck (due to event windows don't open up) and make the CPU 8053unavailable to host or other VMs. 8054 80557.34 KVM_CAP_MEMORY_FAULT_INFO 8056------------------------------ 8057 8058:Architectures: x86 8059:Returns: Informational only, -EINVAL on direct KVM_ENABLE_CAP. 8060 8061The presence of this capability indicates that KVM_RUN will fill 8062kvm_run.memory_fault if KVM cannot resolve a guest page fault VM-Exit, e.g. if 8063there is a valid memslot but no backing VMA for the corresponding host virtual 8064address. 8065 8066The information in kvm_run.memory_fault is valid if and only if KVM_RUN returns 8067an error with errno=EFAULT or errno=EHWPOISON *and* kvm_run.exit_reason is set 8068to KVM_EXIT_MEMORY_FAULT. 8069 8070Note: Userspaces which attempt to resolve memory faults so that they can retry 8071KVM_RUN are encouraged to guard against repeatedly receiving the same 8072error/annotated fault. 8073 8074See KVM_EXIT_MEMORY_FAULT for more information. 8075 80768. Other capabilities. 8077====================== 8078 8079This section lists capabilities that give information about other 8080features of the KVM implementation. 8081 80828.1 KVM_CAP_PPC_HWRNG 8083--------------------- 8084 8085:Architectures: ppc 8086 8087This capability, if KVM_CHECK_EXTENSION indicates that it is 8088available, means that the kernel has an implementation of the 8089H_RANDOM hypercall backed by a hardware random-number generator. 8090If present, the kernel H_RANDOM handler can be enabled for guest use 8091with the KVM_CAP_PPC_ENABLE_HCALL capability. 8092 80938.2 KVM_CAP_HYPERV_SYNIC 8094------------------------ 8095 8096:Architectures: x86 8097 8098This capability, if KVM_CHECK_EXTENSION indicates that it is 8099available, means that the kernel has an implementation of the 8100Hyper-V Synthetic interrupt controller(SynIC). Hyper-V SynIC is 8101used to support Windows Hyper-V based guest paravirt drivers(VMBus). 8102 8103In order to use SynIC, it has to be activated by setting this 8104capability via KVM_ENABLE_CAP ioctl on the vcpu fd. Note that this 8105will disable the use of APIC hardware virtualization even if supported 8106by the CPU, as it's incompatible with SynIC auto-EOI behavior. 8107 81088.3 KVM_CAP_PPC_MMU_RADIX 8109------------------------- 8110 8111:Architectures: ppc 8112 8113This capability, if KVM_CHECK_EXTENSION indicates that it is 8114available, means that the kernel can support guests using the 8115radix MMU defined in Power ISA V3.00 (as implemented in the POWER9 8116processor). 8117 81188.4 KVM_CAP_PPC_MMU_HASH_V3 8119--------------------------- 8120 8121:Architectures: ppc 8122 8123This capability, if KVM_CHECK_EXTENSION indicates that it is 8124available, means that the kernel can support guests using the 8125hashed page table MMU defined in Power ISA V3.00 (as implemented in 8126the POWER9 processor), including in-memory segment tables. 8127 81288.5 KVM_CAP_MIPS_VZ 8129------------------- 8130 8131:Architectures: mips 8132 8133This capability, if KVM_CHECK_EXTENSION on the main kvm handle indicates that 8134it is available, means that full hardware assisted virtualization capabilities 8135of the hardware are available for use through KVM. An appropriate 8136KVM_VM_MIPS_* type must be passed to KVM_CREATE_VM to create a VM which 8137utilises it. 8138 8139If KVM_CHECK_EXTENSION on a kvm VM handle indicates that this capability is 8140available, it means that the VM is using full hardware assisted virtualization 8141capabilities of the hardware. This is useful to check after creating a VM with 8142KVM_VM_MIPS_DEFAULT. 8143 8144The value returned by KVM_CHECK_EXTENSION should be compared against known 8145values (see below). All other values are reserved. This is to allow for the 8146possibility of other hardware assisted virtualization implementations which 8147may be incompatible with the MIPS VZ ASE. 8148 8149== ========================================================================== 8150 0 The trap & emulate implementation is in use to run guest code in user 8151 mode. Guest virtual memory segments are rearranged to fit the guest in the 8152 user mode address space. 8153 8154 1 The MIPS VZ ASE is in use, providing full hardware assisted 8155 virtualization, including standard guest virtual memory segments. 8156== ========================================================================== 8157 81588.6 KVM_CAP_MIPS_TE 8159------------------- 8160 8161:Architectures: mips 8162 8163This capability, if KVM_CHECK_EXTENSION on the main kvm handle indicates that 8164it is available, means that the trap & emulate implementation is available to 8165run guest code in user mode, even if KVM_CAP_MIPS_VZ indicates that hardware 8166assisted virtualisation is also available. KVM_VM_MIPS_TE (0) must be passed 8167to KVM_CREATE_VM to create a VM which utilises it. 8168 8169If KVM_CHECK_EXTENSION on a kvm VM handle indicates that this capability is 8170available, it means that the VM is using trap & emulate. 8171 81728.7 KVM_CAP_MIPS_64BIT 8173---------------------- 8174 8175:Architectures: mips 8176 8177This capability indicates the supported architecture type of the guest, i.e. the 8178supported register and address width. 8179 8180The values returned when this capability is checked by KVM_CHECK_EXTENSION on a 8181kvm VM handle correspond roughly to the CP0_Config.AT register field, and should 8182be checked specifically against known values (see below). All other values are 8183reserved. 8184 8185== ======================================================================== 8186 0 MIPS32 or microMIPS32. 8187 Both registers and addresses are 32-bits wide. 8188 It will only be possible to run 32-bit guest code. 8189 8190 1 MIPS64 or microMIPS64 with access only to 32-bit compatibility segments. 8191 Registers are 64-bits wide, but addresses are 32-bits wide. 8192 64-bit guest code may run but cannot access MIPS64 memory segments. 8193 It will also be possible to run 32-bit guest code. 8194 8195 2 MIPS64 or microMIPS64 with access to all address segments. 8196 Both registers and addresses are 64-bits wide. 8197 It will be possible to run 64-bit or 32-bit guest code. 8198== ======================================================================== 8199 82008.9 KVM_CAP_ARM_USER_IRQ 8201------------------------ 8202 8203:Architectures: arm64 8204 8205This capability, if KVM_CHECK_EXTENSION indicates that it is available, means 8206that if userspace creates a VM without an in-kernel interrupt controller, it 8207will be notified of changes to the output level of in-kernel emulated devices, 8208which can generate virtual interrupts, presented to the VM. 8209For such VMs, on every return to userspace, the kernel 8210updates the vcpu's run->s.regs.device_irq_level field to represent the actual 8211output level of the device. 8212 8213Whenever kvm detects a change in the device output level, kvm guarantees at 8214least one return to userspace before running the VM. This exit could either 8215be a KVM_EXIT_INTR or any other exit event, like KVM_EXIT_MMIO. This way, 8216userspace can always sample the device output level and re-compute the state of 8217the userspace interrupt controller. Userspace should always check the state 8218of run->s.regs.device_irq_level on every kvm exit. 8219The value in run->s.regs.device_irq_level can represent both level and edge 8220triggered interrupt signals, depending on the device. Edge triggered interrupt 8221signals will exit to userspace with the bit in run->s.regs.device_irq_level 8222set exactly once per edge signal. 8223 8224The field run->s.regs.device_irq_level is available independent of 8225run->kvm_valid_regs or run->kvm_dirty_regs bits. 8226 8227If KVM_CAP_ARM_USER_IRQ is supported, the KVM_CHECK_EXTENSION ioctl returns a 8228number larger than 0 indicating the version of this capability is implemented 8229and thereby which bits in run->s.regs.device_irq_level can signal values. 8230 8231Currently the following bits are defined for the device_irq_level bitmap:: 8232 8233 KVM_CAP_ARM_USER_IRQ >= 1: 8234 8235 KVM_ARM_DEV_EL1_VTIMER - EL1 virtual timer 8236 KVM_ARM_DEV_EL1_PTIMER - EL1 physical timer 8237 KVM_ARM_DEV_PMU - ARM PMU overflow interrupt signal 8238 8239Future versions of kvm may implement additional events. These will get 8240indicated by returning a higher number from KVM_CHECK_EXTENSION and will be 8241listed above. 8242 82438.10 KVM_CAP_PPC_SMT_POSSIBLE 8244----------------------------- 8245 8246:Architectures: ppc 8247 8248Querying this capability returns a bitmap indicating the possible 8249virtual SMT modes that can be set using KVM_CAP_PPC_SMT. If bit N 8250(counting from the right) is set, then a virtual SMT mode of 2^N is 8251available. 8252 82538.11 KVM_CAP_HYPERV_SYNIC2 8254-------------------------- 8255 8256:Architectures: x86 8257 8258This capability enables a newer version of Hyper-V Synthetic interrupt 8259controller (SynIC). The only difference with KVM_CAP_HYPERV_SYNIC is that KVM 8260doesn't clear SynIC message and event flags pages when they are enabled by 8261writing to the respective MSRs. 8262 82638.12 KVM_CAP_HYPERV_VP_INDEX 8264---------------------------- 8265 8266:Architectures: x86 8267 8268This capability indicates that userspace can load HV_X64_MSR_VP_INDEX msr. Its 8269value is used to denote the target vcpu for a SynIC interrupt. For 8270compatibility, KVM initializes this msr to KVM's internal vcpu index. When this 8271capability is absent, userspace can still query this msr's value. 8272 82738.13 KVM_CAP_S390_AIS_MIGRATION 8274------------------------------- 8275 8276:Architectures: s390 8277:Parameters: none 8278 8279This capability indicates if the flic device will be able to get/set the 8280AIS states for migration via the KVM_DEV_FLIC_AISM_ALL attribute and allows 8281to discover this without having to create a flic device. 8282 82838.14 KVM_CAP_S390_PSW 8284--------------------- 8285 8286:Architectures: s390 8287 8288This capability indicates that the PSW is exposed via the kvm_run structure. 8289 82908.15 KVM_CAP_S390_GMAP 8291---------------------- 8292 8293:Architectures: s390 8294 8295This capability indicates that the user space memory used as guest mapping can 8296be anywhere in the user memory address space, as long as the memory slots are 8297aligned and sized to a segment (1MB) boundary. 8298 82998.16 KVM_CAP_S390_COW 8300--------------------- 8301 8302:Architectures: s390 8303 8304This capability indicates that the user space memory used as guest mapping can 8305use copy-on-write semantics as well as dirty pages tracking via read-only page 8306tables. 8307 83088.17 KVM_CAP_S390_BPB 8309--------------------- 8310 8311:Architectures: s390 8312 8313This capability indicates that kvm will implement the interfaces to handle 8314reset, migration and nested KVM for branch prediction blocking. The stfle 8315facility 82 should not be provided to the guest without this capability. 8316 83178.18 KVM_CAP_HYPERV_TLBFLUSH 8318---------------------------- 8319 8320:Architectures: x86 8321 8322This capability indicates that KVM supports paravirtualized Hyper-V TLB Flush 8323hypercalls: 8324HvFlushVirtualAddressSpace, HvFlushVirtualAddressSpaceEx, 8325HvFlushVirtualAddressList, HvFlushVirtualAddressListEx. 8326 83278.19 KVM_CAP_ARM_INJECT_SERROR_ESR 8328---------------------------------- 8329 8330:Architectures: arm64 8331 8332This capability indicates that userspace can specify (via the 8333KVM_SET_VCPU_EVENTS ioctl) the syndrome value reported to the guest when it 8334takes a virtual SError interrupt exception. 8335If KVM advertises this capability, userspace can only specify the ISS field for 8336the ESR syndrome. Other parts of the ESR, such as the EC are generated by the 8337CPU when the exception is taken. If this virtual SError is taken to EL1 using 8338AArch64, this value will be reported in the ISS field of ESR_ELx. 8339 8340See KVM_CAP_VCPU_EVENTS for more details. 8341 83428.20 KVM_CAP_HYPERV_SEND_IPI 8343---------------------------- 8344 8345:Architectures: x86 8346 8347This capability indicates that KVM supports paravirtualized Hyper-V IPI send 8348hypercalls: 8349HvCallSendSyntheticClusterIpi, HvCallSendSyntheticClusterIpiEx. 8350 83518.21 KVM_CAP_HYPERV_DIRECT_TLBFLUSH 8352----------------------------------- 8353 8354:Architectures: x86 8355 8356This capability indicates that KVM running on top of Hyper-V hypervisor 8357enables Direct TLB flush for its guests meaning that TLB flush 8358hypercalls are handled by Level 0 hypervisor (Hyper-V) bypassing KVM. 8359Due to the different ABI for hypercall parameters between Hyper-V and 8360KVM, enabling this capability effectively disables all hypercall 8361handling by KVM (as some KVM hypercall may be mistakenly treated as TLB 8362flush hypercalls by Hyper-V) so userspace should disable KVM identification 8363in CPUID and only exposes Hyper-V identification. In this case, guest 8364thinks it's running on Hyper-V and only use Hyper-V hypercalls. 8365 83668.22 KVM_CAP_S390_VCPU_RESETS 8367----------------------------- 8368 8369:Architectures: s390 8370 8371This capability indicates that the KVM_S390_NORMAL_RESET and 8372KVM_S390_CLEAR_RESET ioctls are available. 8373 83748.23 KVM_CAP_S390_PROTECTED 8375--------------------------- 8376 8377:Architectures: s390 8378 8379This capability indicates that the Ultravisor has been initialized and 8380KVM can therefore start protected VMs. 8381This capability governs the KVM_S390_PV_COMMAND ioctl and the 8382KVM_MP_STATE_LOAD MP_STATE. KVM_SET_MP_STATE can fail for protected 8383guests when the state change is invalid. 8384 83858.24 KVM_CAP_STEAL_TIME 8386----------------------- 8387 8388:Architectures: arm64, x86 8389 8390This capability indicates that KVM supports steal time accounting. 8391When steal time accounting is supported it may be enabled with 8392architecture-specific interfaces. This capability and the architecture- 8393specific interfaces must be consistent, i.e. if one says the feature 8394is supported, than the other should as well and vice versa. For arm64 8395see Documentation/virt/kvm/devices/vcpu.rst "KVM_ARM_VCPU_PVTIME_CTRL". 8396For x86 see Documentation/virt/kvm/x86/msr.rst "MSR_KVM_STEAL_TIME". 8397 83988.25 KVM_CAP_S390_DIAG318 8399------------------------- 8400 8401:Architectures: s390 8402 8403This capability enables a guest to set information about its control program 8404(i.e. guest kernel type and version). The information is helpful during 8405system/firmware service events, providing additional data about the guest 8406environments running on the machine. 8407 8408The information is associated with the DIAGNOSE 0x318 instruction, which sets 8409an 8-byte value consisting of a one-byte Control Program Name Code (CPNC) and 8410a 7-byte Control Program Version Code (CPVC). The CPNC determines what 8411environment the control program is running in (e.g. Linux, z/VM...), and the 8412CPVC is used for information specific to OS (e.g. Linux version, Linux 8413distribution...) 8414 8415If this capability is available, then the CPNC and CPVC can be synchronized 8416between KVM and userspace via the sync regs mechanism (KVM_SYNC_DIAG318). 8417 84188.26 KVM_CAP_X86_USER_SPACE_MSR 8419------------------------------- 8420 8421:Architectures: x86 8422 8423This capability indicates that KVM supports deflection of MSR reads and 8424writes to user space. It can be enabled on a VM level. If enabled, MSR 8425accesses that would usually trigger a #GP by KVM into the guest will 8426instead get bounced to user space through the KVM_EXIT_X86_RDMSR and 8427KVM_EXIT_X86_WRMSR exit notifications. 8428 84298.27 KVM_CAP_X86_MSR_FILTER 8430--------------------------- 8431 8432:Architectures: x86 8433 8434This capability indicates that KVM supports that accesses to user defined MSRs 8435may be rejected. With this capability exposed, KVM exports new VM ioctl 8436KVM_X86_SET_MSR_FILTER which user space can call to specify bitmaps of MSR 8437ranges that KVM should deny access to. 8438 8439In combination with KVM_CAP_X86_USER_SPACE_MSR, this allows user space to 8440trap and emulate MSRs that are outside of the scope of KVM as well as 8441limit the attack surface on KVM's MSR emulation code. 8442 84438.28 KVM_CAP_ENFORCE_PV_FEATURE_CPUID 8444------------------------------------- 8445 8446Architectures: x86 8447 8448When enabled, KVM will disable paravirtual features provided to the 8449guest according to the bits in the KVM_CPUID_FEATURES CPUID leaf 8450(0x40000001). Otherwise, a guest may use the paravirtual features 8451regardless of what has actually been exposed through the CPUID leaf. 8452 84538.29 KVM_CAP_DIRTY_LOG_RING/KVM_CAP_DIRTY_LOG_RING_ACQ_REL 8454---------------------------------------------------------- 8455 8456:Architectures: x86, arm64 8457:Parameters: args[0] - size of the dirty log ring 8458 8459KVM is capable of tracking dirty memory using ring buffers that are 8460mmapped into userspace; there is one dirty ring per vcpu. 8461 8462The dirty ring is available to userspace as an array of 8463``struct kvm_dirty_gfn``. Each dirty entry is defined as:: 8464 8465 struct kvm_dirty_gfn { 8466 __u32 flags; 8467 __u32 slot; /* as_id | slot_id */ 8468 __u64 offset; 8469 }; 8470 8471The following values are defined for the flags field to define the 8472current state of the entry:: 8473 8474 #define KVM_DIRTY_GFN_F_DIRTY BIT(0) 8475 #define KVM_DIRTY_GFN_F_RESET BIT(1) 8476 #define KVM_DIRTY_GFN_F_MASK 0x3 8477 8478Userspace should call KVM_ENABLE_CAP ioctl right after KVM_CREATE_VM 8479ioctl to enable this capability for the new guest and set the size of 8480the rings. Enabling the capability is only allowed before creating any 8481vCPU, and the size of the ring must be a power of two. The larger the 8482ring buffer, the less likely the ring is full and the VM is forced to 8483exit to userspace. The optimal size depends on the workload, but it is 8484recommended that it be at least 64 KiB (4096 entries). 8485 8486Just like for dirty page bitmaps, the buffer tracks writes to 8487all user memory regions for which the KVM_MEM_LOG_DIRTY_PAGES flag was 8488set in KVM_SET_USER_MEMORY_REGION. Once a memory region is registered 8489with the flag set, userspace can start harvesting dirty pages from the 8490ring buffer. 8491 8492An entry in the ring buffer can be unused (flag bits ``00``), 8493dirty (flag bits ``01``) or harvested (flag bits ``1X``). The 8494state machine for the entry is as follows:: 8495 8496 dirtied harvested reset 8497 00 -----------> 01 -------------> 1X -------+ 8498 ^ | 8499 | | 8500 +------------------------------------------+ 8501 8502To harvest the dirty pages, userspace accesses the mmapped ring buffer 8503to read the dirty GFNs. If the flags has the DIRTY bit set (at this stage 8504the RESET bit must be cleared), then it means this GFN is a dirty GFN. 8505The userspace should harvest this GFN and mark the flags from state 8506``01b`` to ``1Xb`` (bit 0 will be ignored by KVM, but bit 1 must be set 8507to show that this GFN is harvested and waiting for a reset), and move 8508on to the next GFN. The userspace should continue to do this until the 8509flags of a GFN have the DIRTY bit cleared, meaning that it has harvested 8510all the dirty GFNs that were available. 8511 8512Note that on weakly ordered architectures, userspace accesses to the 8513ring buffer (and more specifically the 'flags' field) must be ordered, 8514using load-acquire/store-release accessors when available, or any 8515other memory barrier that will ensure this ordering. 8516 8517It's not necessary for userspace to harvest the all dirty GFNs at once. 8518However it must collect the dirty GFNs in sequence, i.e., the userspace 8519program cannot skip one dirty GFN to collect the one next to it. 8520 8521After processing one or more entries in the ring buffer, userspace 8522calls the VM ioctl KVM_RESET_DIRTY_RINGS to notify the kernel about 8523it, so that the kernel will reprotect those collected GFNs. 8524Therefore, the ioctl must be called *before* reading the content of 8525the dirty pages. 8526 8527The dirty ring can get full. When it happens, the KVM_RUN of the 8528vcpu will return with exit reason KVM_EXIT_DIRTY_LOG_FULL. 8529 8530The dirty ring interface has a major difference comparing to the 8531KVM_GET_DIRTY_LOG interface in that, when reading the dirty ring from 8532userspace, it's still possible that the kernel has not yet flushed the 8533processor's dirty page buffers into the kernel buffer (with dirty bitmaps, the 8534flushing is done by the KVM_GET_DIRTY_LOG ioctl). To achieve that, one 8535needs to kick the vcpu out of KVM_RUN using a signal. The resulting 8536vmexit ensures that all dirty GFNs are flushed to the dirty rings. 8537 8538NOTE: KVM_CAP_DIRTY_LOG_RING_ACQ_REL is the only capability that 8539should be exposed by weakly ordered architecture, in order to indicate 8540the additional memory ordering requirements imposed on userspace when 8541reading the state of an entry and mutating it from DIRTY to HARVESTED. 8542Architecture with TSO-like ordering (such as x86) are allowed to 8543expose both KVM_CAP_DIRTY_LOG_RING and KVM_CAP_DIRTY_LOG_RING_ACQ_REL 8544to userspace. 8545 8546After enabling the dirty rings, the userspace needs to detect the 8547capability of KVM_CAP_DIRTY_LOG_RING_WITH_BITMAP to see whether the 8548ring structures can be backed by per-slot bitmaps. With this capability 8549advertised, it means the architecture can dirty guest pages without 8550vcpu/ring context, so that some of the dirty information will still be 8551maintained in the bitmap structure. KVM_CAP_DIRTY_LOG_RING_WITH_BITMAP 8552can't be enabled if the capability of KVM_CAP_DIRTY_LOG_RING_ACQ_REL 8553hasn't been enabled, or any memslot has been existing. 8554 8555Note that the bitmap here is only a backup of the ring structure. The 8556use of the ring and bitmap combination is only beneficial if there is 8557only a very small amount of memory that is dirtied out of vcpu/ring 8558context. Otherwise, the stand-alone per-slot bitmap mechanism needs to 8559be considered. 8560 8561To collect dirty bits in the backup bitmap, userspace can use the same 8562KVM_GET_DIRTY_LOG ioctl. KVM_CLEAR_DIRTY_LOG isn't needed as long as all 8563the generation of the dirty bits is done in a single pass. Collecting 8564the dirty bitmap should be the very last thing that the VMM does before 8565considering the state as complete. VMM needs to ensure that the dirty 8566state is final and avoid missing dirty pages from another ioctl ordered 8567after the bitmap collection. 8568 8569NOTE: Multiple examples of using the backup bitmap: (1) save vgic/its 8570tables through command KVM_DEV_ARM_{VGIC_GRP_CTRL, ITS_SAVE_TABLES} on 8571KVM device "kvm-arm-vgic-its". (2) restore vgic/its tables through 8572command KVM_DEV_ARM_{VGIC_GRP_CTRL, ITS_RESTORE_TABLES} on KVM device 8573"kvm-arm-vgic-its". VGICv3 LPI pending status is restored. (3) save 8574vgic3 pending table through KVM_DEV_ARM_VGIC_{GRP_CTRL, SAVE_PENDING_TABLES} 8575command on KVM device "kvm-arm-vgic-v3". 8576 85778.30 KVM_CAP_XEN_HVM 8578-------------------- 8579 8580:Architectures: x86 8581 8582This capability indicates the features that Xen supports for hosting Xen 8583PVHVM guests. Valid flags are:: 8584 8585 #define KVM_XEN_HVM_CONFIG_HYPERCALL_MSR (1 << 0) 8586 #define KVM_XEN_HVM_CONFIG_INTERCEPT_HCALL (1 << 1) 8587 #define KVM_XEN_HVM_CONFIG_SHARED_INFO (1 << 2) 8588 #define KVM_XEN_HVM_CONFIG_RUNSTATE (1 << 3) 8589 #define KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL (1 << 4) 8590 #define KVM_XEN_HVM_CONFIG_EVTCHN_SEND (1 << 5) 8591 #define KVM_XEN_HVM_CONFIG_RUNSTATE_UPDATE_FLAG (1 << 6) 8592 #define KVM_XEN_HVM_CONFIG_PVCLOCK_TSC_UNSTABLE (1 << 7) 8593 8594The KVM_XEN_HVM_CONFIG_HYPERCALL_MSR flag indicates that the KVM_XEN_HVM_CONFIG 8595ioctl is available, for the guest to set its hypercall page. 8596 8597If KVM_XEN_HVM_CONFIG_INTERCEPT_HCALL is also set, the same flag may also be 8598provided in the flags to KVM_XEN_HVM_CONFIG, without providing hypercall page 8599contents, to request that KVM generate hypercall page content automatically 8600and also enable interception of guest hypercalls with KVM_EXIT_XEN. 8601 8602The KVM_XEN_HVM_CONFIG_SHARED_INFO flag indicates the availability of the 8603KVM_XEN_HVM_SET_ATTR, KVM_XEN_HVM_GET_ATTR, KVM_XEN_VCPU_SET_ATTR and 8604KVM_XEN_VCPU_GET_ATTR ioctls, as well as the delivery of exception vectors 8605for event channel upcalls when the evtchn_upcall_pending field of a vcpu's 8606vcpu_info is set. 8607 8608The KVM_XEN_HVM_CONFIG_RUNSTATE flag indicates that the runstate-related 8609features KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADDR/_CURRENT/_DATA/_ADJUST are 8610supported by the KVM_XEN_VCPU_SET_ATTR/KVM_XEN_VCPU_GET_ATTR ioctls. 8611 8612The KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL flag indicates that IRQ routing entries 8613of the type KVM_IRQ_ROUTING_XEN_EVTCHN are supported, with the priority 8614field set to indicate 2 level event channel delivery. 8615 8616The KVM_XEN_HVM_CONFIG_EVTCHN_SEND flag indicates that KVM supports 8617injecting event channel events directly into the guest with the 8618KVM_XEN_HVM_EVTCHN_SEND ioctl. It also indicates support for the 8619KVM_XEN_ATTR_TYPE_EVTCHN/XEN_VERSION HVM attributes and the 8620KVM_XEN_VCPU_ATTR_TYPE_VCPU_ID/TIMER/UPCALL_VECTOR vCPU attributes. 8621related to event channel delivery, timers, and the XENVER_version 8622interception. 8623 8624The KVM_XEN_HVM_CONFIG_RUNSTATE_UPDATE_FLAG flag indicates that KVM supports 8625the KVM_XEN_ATTR_TYPE_RUNSTATE_UPDATE_FLAG attribute in the KVM_XEN_SET_ATTR 8626and KVM_XEN_GET_ATTR ioctls. This controls whether KVM will set the 8627XEN_RUNSTATE_UPDATE flag in guest memory mapped vcpu_runstate_info during 8628updates of the runstate information. Note that versions of KVM which support 8629the RUNSTATE feature above, but not the RUNSTATE_UPDATE_FLAG feature, will 8630always set the XEN_RUNSTATE_UPDATE flag when updating the guest structure, 8631which is perhaps counterintuitive. When this flag is advertised, KVM will 8632behave more correctly, not using the XEN_RUNSTATE_UPDATE flag until/unless 8633specifically enabled (by the guest making the hypercall, causing the VMM 8634to enable the KVM_XEN_ATTR_TYPE_RUNSTATE_UPDATE_FLAG attribute). 8635 8636The KVM_XEN_HVM_CONFIG_PVCLOCK_TSC_UNSTABLE flag indicates that KVM supports 8637clearing the PVCLOCK_TSC_STABLE_BIT flag in Xen pvclock sources. This will be 8638done when the KVM_CAP_XEN_HVM ioctl sets the 8639KVM_XEN_HVM_CONFIG_PVCLOCK_TSC_UNSTABLE flag. 8640 86418.31 KVM_CAP_PPC_MULTITCE 8642------------------------- 8643 8644:Capability: KVM_CAP_PPC_MULTITCE 8645:Architectures: ppc 8646:Type: vm 8647 8648This capability means the kernel is capable of handling hypercalls 8649H_PUT_TCE_INDIRECT and H_STUFF_TCE without passing those into the user 8650space. This significantly accelerates DMA operations for PPC KVM guests. 8651User space should expect that its handlers for these hypercalls 8652are not going to be called if user space previously registered LIOBN 8653in KVM (via KVM_CREATE_SPAPR_TCE or similar calls). 8654 8655In order to enable H_PUT_TCE_INDIRECT and H_STUFF_TCE use in the guest, 8656user space might have to advertise it for the guest. For example, 8657IBM pSeries (sPAPR) guest starts using them if "hcall-multi-tce" is 8658present in the "ibm,hypertas-functions" device-tree property. 8659 8660The hypercalls mentioned above may or may not be processed successfully 8661in the kernel based fast path. If they can not be handled by the kernel, 8662they will get passed on to user space. So user space still has to have 8663an implementation for these despite the in kernel acceleration. 8664 8665This capability is always enabled. 8666 86678.32 KVM_CAP_PTP_KVM 8668-------------------- 8669 8670:Architectures: arm64 8671 8672This capability indicates that the KVM virtual PTP service is 8673supported in the host. A VMM can check whether the service is 8674available to the guest on migration. 8675 86768.33 KVM_CAP_HYPERV_ENFORCE_CPUID 8677--------------------------------- 8678 8679Architectures: x86 8680 8681When enabled, KVM will disable emulated Hyper-V features provided to the 8682guest according to the bits Hyper-V CPUID feature leaves. Otherwise, all 8683currently implemented Hyper-V features are provided unconditionally when 8684Hyper-V identification is set in the HYPERV_CPUID_INTERFACE (0x40000001) 8685leaf. 8686 86878.34 KVM_CAP_EXIT_HYPERCALL 8688--------------------------- 8689 8690:Capability: KVM_CAP_EXIT_HYPERCALL 8691:Architectures: x86 8692:Type: vm 8693 8694This capability, if enabled, will cause KVM to exit to userspace 8695with KVM_EXIT_HYPERCALL exit reason to process some hypercalls. 8696 8697Calling KVM_CHECK_EXTENSION for this capability will return a bitmask 8698of hypercalls that can be configured to exit to userspace. 8699Right now, the only such hypercall is KVM_HC_MAP_GPA_RANGE. 8700 8701The argument to KVM_ENABLE_CAP is also a bitmask, and must be a subset 8702of the result of KVM_CHECK_EXTENSION. KVM will forward to userspace 8703the hypercalls whose corresponding bit is in the argument, and return 8704ENOSYS for the others. 8705 87068.35 KVM_CAP_PMU_CAPABILITY 8707--------------------------- 8708 8709:Capability: KVM_CAP_PMU_CAPABILITY 8710:Architectures: x86 8711:Type: vm 8712:Parameters: arg[0] is bitmask of PMU virtualization capabilities. 8713:Returns: 0 on success, -EINVAL when arg[0] contains invalid bits 8714 8715This capability alters PMU virtualization in KVM. 8716 8717Calling KVM_CHECK_EXTENSION for this capability returns a bitmask of 8718PMU virtualization capabilities that can be adjusted on a VM. 8719 8720The argument to KVM_ENABLE_CAP is also a bitmask and selects specific 8721PMU virtualization capabilities to be applied to the VM. This can 8722only be invoked on a VM prior to the creation of VCPUs. 8723 8724At this time, KVM_PMU_CAP_DISABLE is the only capability. Setting 8725this capability will disable PMU virtualization for that VM. Usermode 8726should adjust CPUID leaf 0xA to reflect that the PMU is disabled. 8727 87288.36 KVM_CAP_ARM_SYSTEM_SUSPEND 8729------------------------------- 8730 8731:Capability: KVM_CAP_ARM_SYSTEM_SUSPEND 8732:Architectures: arm64 8733:Type: vm 8734 8735When enabled, KVM will exit to userspace with KVM_EXIT_SYSTEM_EVENT of 8736type KVM_SYSTEM_EVENT_SUSPEND to process the guest suspend request. 8737 87388.37 KVM_CAP_S390_PROTECTED_DUMP 8739-------------------------------- 8740 8741:Capability: KVM_CAP_S390_PROTECTED_DUMP 8742:Architectures: s390 8743:Type: vm 8744 8745This capability indicates that KVM and the Ultravisor support dumping 8746PV guests. The `KVM_PV_DUMP` command is available for the 8747`KVM_S390_PV_COMMAND` ioctl and the `KVM_PV_INFO` command provides 8748dump related UV data. Also the vcpu ioctl `KVM_S390_PV_CPU_COMMAND` is 8749available and supports the `KVM_PV_DUMP_CPU` subcommand. 8750 87518.38 KVM_CAP_VM_DISABLE_NX_HUGE_PAGES 8752------------------------------------- 8753 8754:Capability: KVM_CAP_VM_DISABLE_NX_HUGE_PAGES 8755:Architectures: x86 8756:Type: vm 8757:Parameters: arg[0] must be 0. 8758:Returns: 0 on success, -EPERM if the userspace process does not 8759 have CAP_SYS_BOOT, -EINVAL if args[0] is not 0 or any vCPUs have been 8760 created. 8761 8762This capability disables the NX huge pages mitigation for iTLB MULTIHIT. 8763 8764The capability has no effect if the nx_huge_pages module parameter is not set. 8765 8766This capability may only be set before any vCPUs are created. 8767 87688.39 KVM_CAP_S390_CPU_TOPOLOGY 8769------------------------------ 8770 8771:Capability: KVM_CAP_S390_CPU_TOPOLOGY 8772:Architectures: s390 8773:Type: vm 8774 8775This capability indicates that KVM will provide the S390 CPU Topology 8776facility which consist of the interpretation of the PTF instruction for 8777the function code 2 along with interception and forwarding of both the 8778PTF instruction with function codes 0 or 1 and the STSI(15,1,x) 8779instruction to the userland hypervisor. 8780 8781The stfle facility 11, CPU Topology facility, should not be indicated 8782to the guest without this capability. 8783 8784When this capability is present, KVM provides a new attribute group 8785on vm fd, KVM_S390_VM_CPU_TOPOLOGY. 8786This new attribute allows to get, set or clear the Modified Change 8787Topology Report (MTCR) bit of the SCA through the kvm_device_attr 8788structure. 8789 8790When getting the Modified Change Topology Report value, the attr->addr 8791must point to a byte where the value will be stored or retrieved from. 8792 87938.40 KVM_CAP_ARM_EAGER_SPLIT_CHUNK_SIZE 8794--------------------------------------- 8795 8796:Capability: KVM_CAP_ARM_EAGER_SPLIT_CHUNK_SIZE 8797:Architectures: arm64 8798:Type: vm 8799:Parameters: arg[0] is the new split chunk size. 8800:Returns: 0 on success, -EINVAL if any memslot was already created. 8801 8802This capability sets the chunk size used in Eager Page Splitting. 8803 8804Eager Page Splitting improves the performance of dirty-logging (used 8805in live migrations) when guest memory is backed by huge-pages. It 8806avoids splitting huge-pages (into PAGE_SIZE pages) on fault, by doing 8807it eagerly when enabling dirty logging (with the 8808KVM_MEM_LOG_DIRTY_PAGES flag for a memory region), or when using 8809KVM_CLEAR_DIRTY_LOG. 8810 8811The chunk size specifies how many pages to break at a time, using a 8812single allocation for each chunk. Bigger the chunk size, more pages 8813need to be allocated ahead of time. 8814 8815The chunk size needs to be a valid block size. The list of acceptable 8816block sizes is exposed in KVM_CAP_ARM_SUPPORTED_BLOCK_SIZES as a 881764-bit bitmap (each bit describing a block size). The default value is 88180, to disable the eager page splitting. 8819 88208.41 KVM_CAP_VM_TYPES 8821--------------------- 8822 8823:Capability: KVM_CAP_MEMORY_ATTRIBUTES 8824:Architectures: x86 8825:Type: system ioctl 8826 8827This capability returns a bitmap of support VM types. The 1-setting of bit @n 8828means the VM type with value @n is supported. Possible values of @n are:: 8829 8830 #define KVM_X86_DEFAULT_VM 0 8831 #define KVM_X86_SW_PROTECTED_VM 1 8832 #define KVM_X86_SEV_VM 2 8833 #define KVM_X86_SEV_ES_VM 3 8834 8835Note, KVM_X86_SW_PROTECTED_VM is currently only for development and testing. 8836Do not use KVM_X86_SW_PROTECTED_VM for "real" VMs, and especially not in 8837production. The behavior and effective ABI for software-protected VMs is 8838unstable. 8839 88409. Known KVM API problems 8841========================= 8842 8843In some cases, KVM's API has some inconsistencies or common pitfalls 8844that userspace need to be aware of. This section details some of 8845these issues. 8846 8847Most of them are architecture specific, so the section is split by 8848architecture. 8849 88509.1. x86 8851-------- 8852 8853``KVM_GET_SUPPORTED_CPUID`` issues 8854^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 8855 8856In general, ``KVM_GET_SUPPORTED_CPUID`` is designed so that it is possible 8857to take its result and pass it directly to ``KVM_SET_CPUID2``. This section 8858documents some cases in which that requires some care. 8859 8860Local APIC features 8861~~~~~~~~~~~~~~~~~~~ 8862 8863CPU[EAX=1]:ECX[21] (X2APIC) is reported by ``KVM_GET_SUPPORTED_CPUID``, 8864but it can only be enabled if ``KVM_CREATE_IRQCHIP`` or 8865``KVM_ENABLE_CAP(KVM_CAP_IRQCHIP_SPLIT)`` are used to enable in-kernel emulation of 8866the local APIC. 8867 8868The same is true for the ``KVM_FEATURE_PV_UNHALT`` paravirtualized feature. 8869 8870CPU[EAX=1]:ECX[24] (TSC_DEADLINE) is not reported by ``KVM_GET_SUPPORTED_CPUID``. 8871It can be enabled if ``KVM_CAP_TSC_DEADLINE_TIMER`` is present and the kernel 8872has enabled in-kernel emulation of the local APIC. 8873 8874CPU topology 8875~~~~~~~~~~~~ 8876 8877Several CPUID values include topology information for the host CPU: 88780x0b and 0x1f for Intel systems, 0x8000001e for AMD systems. Different 8879versions of KVM return different values for this information and userspace 8880should not rely on it. Currently they return all zeroes. 8881 8882If userspace wishes to set up a guest topology, it should be careful that 8883the values of these three leaves differ for each CPU. In particular, 8884the APIC ID is found in EDX for all subleaves of 0x0b and 0x1f, and in EAX 8885for 0x8000001e; the latter also encodes the core id and node id in bits 88867:0 of EBX and ECX respectively. 8887 8888Obsolete ioctls and capabilities 8889^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 8890 8891KVM_CAP_DISABLE_QUIRKS does not let userspace know which quirks are actually 8892available. Use ``KVM_CHECK_EXTENSION(KVM_CAP_DISABLE_QUIRKS2)`` instead if 8893available. 8894 8895Ordering of KVM_GET_*/KVM_SET_* ioctls 8896^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 8897 8898TBD 8899