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 150In order to create user controlled virtual machines on S390, check 151KVM_CAP_S390_UCONTROL and use the flag KVM_VM_S390_UCONTROL as 152privileged user (CAP_SYS_ADMIN). 153 154On arm64, the physical address size for a VM (IPA Size limit) is limited 155to 40bits by default. The limit can be configured if the host supports the 156extension KVM_CAP_ARM_VM_IPA_SIZE. When supported, use 157KVM_VM_TYPE_ARM_IPA_SIZE(IPA_Bits) to set the size in the machine type 158identifier, where IPA_Bits is the maximum width of any physical 159address used by the VM. The IPA_Bits is encoded in bits[7-0] of the 160machine type identifier. 161 162e.g, to configure a guest to use 48bit physical address size:: 163 164 vm_fd = ioctl(dev_fd, KVM_CREATE_VM, KVM_VM_TYPE_ARM_IPA_SIZE(48)); 165 166The requested size (IPA_Bits) must be: 167 168 == ========================================================= 169 0 Implies default size, 40bits (for backward compatibility) 170 N Implies N bits, where N is a positive integer such that, 171 32 <= N <= Host_IPA_Limit 172 == ========================================================= 173 174Host_IPA_Limit is the maximum possible value for IPA_Bits on the host and 175is dependent on the CPU capability and the kernel configuration. The limit can 176be retrieved using KVM_CAP_ARM_VM_IPA_SIZE of the KVM_CHECK_EXTENSION 177ioctl() at run-time. 178 179Creation of the VM will fail if the requested IPA size (whether it is 180implicit or explicit) is unsupported on the host. 181 182Please note that configuring the IPA size does not affect the capability 183exposed by the guest CPUs in ID_AA64MMFR0_EL1[PARange]. It only affects 184size of the address translated by the stage2 level (guest physical to 185host physical address translations). 186 187 1884.3 KVM_GET_MSR_INDEX_LIST, KVM_GET_MSR_FEATURE_INDEX_LIST 189---------------------------------------------------------- 190 191:Capability: basic, KVM_CAP_GET_MSR_FEATURES for KVM_GET_MSR_FEATURE_INDEX_LIST 192:Architectures: x86 193:Type: system ioctl 194:Parameters: struct kvm_msr_list (in/out) 195:Returns: 0 on success; -1 on error 196 197Errors: 198 199 ====== ============================================================ 200 EFAULT the msr index list cannot be read from or written to 201 E2BIG the msr index list is too big to fit in the array specified by 202 the user. 203 ====== ============================================================ 204 205:: 206 207 struct kvm_msr_list { 208 __u32 nmsrs; /* number of msrs in entries */ 209 __u32 indices[0]; 210 }; 211 212The user fills in the size of the indices array in nmsrs, and in return 213kvm adjusts nmsrs to reflect the actual number of msrs and fills in the 214indices array with their numbers. 215 216KVM_GET_MSR_INDEX_LIST returns the guest msrs that are supported. The list 217varies by kvm version and host processor, but does not change otherwise. 218 219Note: if kvm indicates supports MCE (KVM_CAP_MCE), then the MCE bank MSRs are 220not returned in the MSR list, as different vcpus can have a different number 221of banks, as set via the KVM_X86_SETUP_MCE ioctl. 222 223KVM_GET_MSR_FEATURE_INDEX_LIST returns the list of MSRs that can be passed 224to the KVM_GET_MSRS system ioctl. This lets userspace probe host capabilities 225and processor features that are exposed via MSRs (e.g., VMX capabilities). 226This list also varies by kvm version and host processor, but does not change 227otherwise. 228 229 2304.4 KVM_CHECK_EXTENSION 231----------------------- 232 233:Capability: basic, KVM_CAP_CHECK_EXTENSION_VM for vm ioctl 234:Architectures: all 235:Type: system ioctl, vm ioctl 236:Parameters: extension identifier (KVM_CAP_*) 237:Returns: 0 if unsupported; 1 (or some other positive integer) if supported 238 239The API allows the application to query about extensions to the core 240kvm API. Userspace passes an extension identifier (an integer) and 241receives an integer that describes the extension availability. 242Generally 0 means no and 1 means yes, but some extensions may report 243additional information in the integer return value. 244 245Based on their initialization different VMs may have different capabilities. 246It is thus encouraged to use the vm ioctl to query for capabilities (available 247with KVM_CAP_CHECK_EXTENSION_VM on the vm fd) 248 2494.5 KVM_GET_VCPU_MMAP_SIZE 250-------------------------- 251 252:Capability: basic 253:Architectures: all 254:Type: system ioctl 255:Parameters: none 256:Returns: size of vcpu mmap area, in bytes 257 258The KVM_RUN ioctl (cf.) communicates with userspace via a shared 259memory region. This ioctl returns the size of that region. See the 260KVM_RUN documentation for details. 261 262Besides the size of the KVM_RUN communication region, other areas of 263the VCPU file descriptor can be mmap-ed, including: 264 265- if KVM_CAP_COALESCED_MMIO is available, a page at 266 KVM_COALESCED_MMIO_PAGE_OFFSET * PAGE_SIZE; for historical reasons, 267 this page is included in the result of KVM_GET_VCPU_MMAP_SIZE. 268 KVM_CAP_COALESCED_MMIO is not documented yet. 269 270- if KVM_CAP_DIRTY_LOG_RING is available, a number of pages at 271 KVM_DIRTY_LOG_PAGE_OFFSET * PAGE_SIZE. For more information on 272 KVM_CAP_DIRTY_LOG_RING, see section 8.3. 273 274 2754.6 KVM_SET_MEMORY_REGION 276------------------------- 277 278:Capability: basic 279:Architectures: all 280:Type: vm ioctl 281:Parameters: struct kvm_memory_region (in) 282:Returns: 0 on success, -1 on error 283 284This ioctl is obsolete and has been removed. 285 286 2874.7 KVM_CREATE_VCPU 288------------------- 289 290:Capability: basic 291:Architectures: all 292:Type: vm ioctl 293:Parameters: vcpu id (apic id on x86) 294:Returns: vcpu fd on success, -1 on error 295 296This API adds a vcpu to a virtual machine. No more than max_vcpus may be added. 297The vcpu id is an integer in the range [0, max_vcpu_id). 298 299The recommended max_vcpus value can be retrieved using the KVM_CAP_NR_VCPUS of 300the KVM_CHECK_EXTENSION ioctl() at run-time. 301The maximum possible value for max_vcpus can be retrieved using the 302KVM_CAP_MAX_VCPUS of the KVM_CHECK_EXTENSION ioctl() at run-time. 303 304If the KVM_CAP_NR_VCPUS does not exist, you should assume that max_vcpus is 4 305cpus max. 306If the KVM_CAP_MAX_VCPUS does not exist, you should assume that max_vcpus is 307same as the value returned from KVM_CAP_NR_VCPUS. 308 309The maximum possible value for max_vcpu_id can be retrieved using the 310KVM_CAP_MAX_VCPU_ID of the KVM_CHECK_EXTENSION ioctl() at run-time. 311 312If the KVM_CAP_MAX_VCPU_ID does not exist, you should assume that max_vcpu_id 313is the same as the value returned from KVM_CAP_MAX_VCPUS. 314 315On powerpc using book3s_hv mode, the vcpus are mapped onto virtual 316threads in one or more virtual CPU cores. (This is because the 317hardware requires all the hardware threads in a CPU core to be in the 318same partition.) The KVM_CAP_PPC_SMT capability indicates the number 319of vcpus per virtual core (vcore). The vcore id is obtained by 320dividing the vcpu id by the number of vcpus per vcore. The vcpus in a 321given vcore will always be in the same physical core as each other 322(though that might be a different physical core from time to time). 323Userspace can control the threading (SMT) mode of the guest by its 324allocation of vcpu ids. For example, if userspace wants 325single-threaded guest vcpus, it should make all vcpu ids be a multiple 326of the number of vcpus per vcore. 327 328For virtual cpus that have been created with S390 user controlled virtual 329machines, the resulting vcpu fd can be memory mapped at page offset 330KVM_S390_SIE_PAGE_OFFSET in order to obtain a memory map of the virtual 331cpu's hardware control block. 332 333 3344.8 KVM_GET_DIRTY_LOG (vm ioctl) 335-------------------------------- 336 337:Capability: basic 338:Architectures: all 339:Type: vm ioctl 340:Parameters: struct kvm_dirty_log (in/out) 341:Returns: 0 on success, -1 on error 342 343:: 344 345 /* for KVM_GET_DIRTY_LOG */ 346 struct kvm_dirty_log { 347 __u32 slot; 348 __u32 padding; 349 union { 350 void __user *dirty_bitmap; /* one bit per page */ 351 __u64 padding; 352 }; 353 }; 354 355Given a memory slot, return a bitmap containing any pages dirtied 356since the last call to this ioctl. Bit 0 is the first page in the 357memory slot. Ensure the entire structure is cleared to avoid padding 358issues. 359 360If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of slot field specifies 361the address space for which you want to return the dirty bitmap. See 362KVM_SET_USER_MEMORY_REGION for details on the usage of slot field. 363 364The bits in the dirty bitmap are cleared before the ioctl returns, unless 365KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 is enabled. For more information, 366see the description of the capability. 367 368Note that the Xen shared info page, if configured, shall always be assumed 369to be dirty. KVM will not explicitly mark it such. 370 3714.9 KVM_SET_MEMORY_ALIAS 372------------------------ 373 374:Capability: basic 375:Architectures: x86 376:Type: vm ioctl 377:Parameters: struct kvm_memory_alias (in) 378:Returns: 0 (success), -1 (error) 379 380This ioctl is obsolete and has been removed. 381 382 3834.10 KVM_RUN 384------------ 385 386:Capability: basic 387:Architectures: all 388:Type: vcpu ioctl 389:Parameters: none 390:Returns: 0 on success, -1 on error 391 392Errors: 393 394 ======= ============================================================== 395 EINTR an unmasked signal is pending 396 ENOEXEC the vcpu hasn't been initialized or the guest tried to execute 397 instructions from device memory (arm64) 398 ENOSYS data abort outside memslots with no syndrome info and 399 KVM_CAP_ARM_NISV_TO_USER not enabled (arm64) 400 EPERM SVE feature set but not finalized (arm64) 401 ======= ============================================================== 402 403This ioctl is used to run a guest virtual cpu. While there are no 404explicit parameters, there is an implicit parameter block that can be 405obtained by mmap()ing the vcpu fd at offset 0, with the size given by 406KVM_GET_VCPU_MMAP_SIZE. The parameter block is formatted as a 'struct 407kvm_run' (see below). 408 409 4104.11 KVM_GET_REGS 411----------------- 412 413:Capability: basic 414:Architectures: all except arm64 415:Type: vcpu ioctl 416:Parameters: struct kvm_regs (out) 417:Returns: 0 on success, -1 on error 418 419Reads the general purpose registers from the vcpu. 420 421:: 422 423 /* x86 */ 424 struct kvm_regs { 425 /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */ 426 __u64 rax, rbx, rcx, rdx; 427 __u64 rsi, rdi, rsp, rbp; 428 __u64 r8, r9, r10, r11; 429 __u64 r12, r13, r14, r15; 430 __u64 rip, rflags; 431 }; 432 433 /* mips */ 434 struct kvm_regs { 435 /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */ 436 __u64 gpr[32]; 437 __u64 hi; 438 __u64 lo; 439 __u64 pc; 440 }; 441 442 4434.12 KVM_SET_REGS 444----------------- 445 446:Capability: basic 447:Architectures: all except arm64 448:Type: vcpu ioctl 449:Parameters: struct kvm_regs (in) 450:Returns: 0 on success, -1 on error 451 452Writes the general purpose registers into the vcpu. 453 454See KVM_GET_REGS for the data structure. 455 456 4574.13 KVM_GET_SREGS 458------------------ 459 460:Capability: basic 461:Architectures: x86, ppc 462:Type: vcpu ioctl 463:Parameters: struct kvm_sregs (out) 464:Returns: 0 on success, -1 on error 465 466Reads special registers from the vcpu. 467 468:: 469 470 /* x86 */ 471 struct kvm_sregs { 472 struct kvm_segment cs, ds, es, fs, gs, ss; 473 struct kvm_segment tr, ldt; 474 struct kvm_dtable gdt, idt; 475 __u64 cr0, cr2, cr3, cr4, cr8; 476 __u64 efer; 477 __u64 apic_base; 478 __u64 interrupt_bitmap[(KVM_NR_INTERRUPTS + 63) / 64]; 479 }; 480 481 /* ppc -- see arch/powerpc/include/uapi/asm/kvm.h */ 482 483interrupt_bitmap is a bitmap of pending external interrupts. At most 484one bit may be set. This interrupt has been acknowledged by the APIC 485but not yet injected into the cpu core. 486 487 4884.14 KVM_SET_SREGS 489------------------ 490 491:Capability: basic 492:Architectures: x86, ppc 493:Type: vcpu ioctl 494:Parameters: struct kvm_sregs (in) 495:Returns: 0 on success, -1 on error 496 497Writes special registers into the vcpu. See KVM_GET_SREGS for the 498data structures. 499 500 5014.15 KVM_TRANSLATE 502------------------ 503 504:Capability: basic 505:Architectures: x86 506:Type: vcpu ioctl 507:Parameters: struct kvm_translation (in/out) 508:Returns: 0 on success, -1 on error 509 510Translates a virtual address according to the vcpu's current address 511translation mode. 512 513:: 514 515 struct kvm_translation { 516 /* in */ 517 __u64 linear_address; 518 519 /* out */ 520 __u64 physical_address; 521 __u8 valid; 522 __u8 writeable; 523 __u8 usermode; 524 __u8 pad[5]; 525 }; 526 527 5284.16 KVM_INTERRUPT 529------------------ 530 531:Capability: basic 532:Architectures: x86, ppc, mips, riscv 533:Type: vcpu ioctl 534:Parameters: struct kvm_interrupt (in) 535:Returns: 0 on success, negative on failure. 536 537Queues a hardware interrupt vector to be injected. 538 539:: 540 541 /* for KVM_INTERRUPT */ 542 struct kvm_interrupt { 543 /* in */ 544 __u32 irq; 545 }; 546 547X86: 548^^^^ 549 550:Returns: 551 552 ========= =================================== 553 0 on success, 554 -EEXIST if an interrupt is already enqueued 555 -EINVAL the irq number is invalid 556 -ENXIO if the PIC is in the kernel 557 -EFAULT if the pointer is invalid 558 ========= =================================== 559 560Note 'irq' is an interrupt vector, not an interrupt pin or line. This 561ioctl is useful if the in-kernel PIC is not used. 562 563PPC: 564^^^^ 565 566Queues an external interrupt to be injected. This ioctl is overleaded 567with 3 different irq values: 568 569a) KVM_INTERRUPT_SET 570 571 This injects an edge type external interrupt into the guest once it's ready 572 to receive interrupts. When injected, the interrupt is done. 573 574b) KVM_INTERRUPT_UNSET 575 576 This unsets any pending interrupt. 577 578 Only available with KVM_CAP_PPC_UNSET_IRQ. 579 580c) KVM_INTERRUPT_SET_LEVEL 581 582 This injects a level type external interrupt into the guest context. The 583 interrupt stays pending until a specific ioctl with KVM_INTERRUPT_UNSET 584 is triggered. 585 586 Only available with KVM_CAP_PPC_IRQ_LEVEL. 587 588Note that any value for 'irq' other than the ones stated above is invalid 589and incurs unexpected behavior. 590 591This is an asynchronous vcpu ioctl and can be invoked from any thread. 592 593MIPS: 594^^^^^ 595 596Queues an external interrupt to be injected into the virtual CPU. A negative 597interrupt number dequeues the interrupt. 598 599This is an asynchronous vcpu ioctl and can be invoked from any thread. 600 601RISC-V: 602^^^^^^^ 603 604Queues an external interrupt to be injected into the virutal CPU. This ioctl 605is overloaded with 2 different irq values: 606 607a) KVM_INTERRUPT_SET 608 609 This sets external interrupt for a virtual CPU and it will receive 610 once it is ready. 611 612b) KVM_INTERRUPT_UNSET 613 614 This clears pending external interrupt for a virtual CPU. 615 616This is an asynchronous vcpu ioctl and can be invoked from any thread. 617 618 6194.17 KVM_DEBUG_GUEST 620-------------------- 621 622:Capability: basic 623:Architectures: none 624:Type: vcpu ioctl 625:Parameters: none) 626:Returns: -1 on error 627 628Support for this has been removed. Use KVM_SET_GUEST_DEBUG instead. 629 630 6314.18 KVM_GET_MSRS 632----------------- 633 634:Capability: basic (vcpu), KVM_CAP_GET_MSR_FEATURES (system) 635:Architectures: x86 636:Type: system ioctl, vcpu ioctl 637:Parameters: struct kvm_msrs (in/out) 638:Returns: number of msrs successfully returned; 639 -1 on error 640 641When used as a system ioctl: 642Reads the values of MSR-based features that are available for the VM. This 643is similar to KVM_GET_SUPPORTED_CPUID, but it returns MSR indices and values. 644The list of msr-based features can be obtained using KVM_GET_MSR_FEATURE_INDEX_LIST 645in a system ioctl. 646 647When used as a vcpu ioctl: 648Reads model-specific registers from the vcpu. Supported msr indices can 649be obtained using KVM_GET_MSR_INDEX_LIST in a system ioctl. 650 651:: 652 653 struct kvm_msrs { 654 __u32 nmsrs; /* number of msrs in entries */ 655 __u32 pad; 656 657 struct kvm_msr_entry entries[0]; 658 }; 659 660 struct kvm_msr_entry { 661 __u32 index; 662 __u32 reserved; 663 __u64 data; 664 }; 665 666Application code should set the 'nmsrs' member (which indicates the 667size of the entries array) and the 'index' member of each array entry. 668kvm will fill in the 'data' member. 669 670 6714.19 KVM_SET_MSRS 672----------------- 673 674:Capability: basic 675:Architectures: x86 676:Type: vcpu ioctl 677:Parameters: struct kvm_msrs (in) 678:Returns: number of msrs successfully set (see below), -1 on error 679 680Writes model-specific registers to the vcpu. See KVM_GET_MSRS for the 681data structures. 682 683Application code should set the 'nmsrs' member (which indicates the 684size of the entries array), and the 'index' and 'data' members of each 685array entry. 686 687It tries to set the MSRs in array entries[] one by one. If setting an MSR 688fails, e.g., due to setting reserved bits, the MSR isn't supported/emulated 689by KVM, etc..., it stops processing the MSR list and returns the number of 690MSRs that have been set successfully. 691 692 6934.20 KVM_SET_CPUID 694------------------ 695 696:Capability: basic 697:Architectures: x86 698:Type: vcpu ioctl 699:Parameters: struct kvm_cpuid (in) 700:Returns: 0 on success, -1 on error 701 702Defines the vcpu responses to the cpuid instruction. Applications 703should use the KVM_SET_CPUID2 ioctl if available. 704 705Caveat emptor: 706 - If this IOCTL fails, KVM gives no guarantees that previous valid CPUID 707 configuration (if there is) is not corrupted. Userspace can get a copy 708 of the resulting CPUID configuration through KVM_GET_CPUID2 in case. 709 - Using KVM_SET_CPUID{,2} after KVM_RUN, i.e. changing the guest vCPU model 710 after running the guest, may cause guest instability. 711 - Using heterogeneous CPUID configurations, modulo APIC IDs, topology, etc... 712 may cause guest instability. 713 714:: 715 716 struct kvm_cpuid_entry { 717 __u32 function; 718 __u32 eax; 719 __u32 ebx; 720 __u32 ecx; 721 __u32 edx; 722 __u32 padding; 723 }; 724 725 /* for KVM_SET_CPUID */ 726 struct kvm_cpuid { 727 __u32 nent; 728 __u32 padding; 729 struct kvm_cpuid_entry entries[0]; 730 }; 731 732 7334.21 KVM_SET_SIGNAL_MASK 734------------------------ 735 736:Capability: basic 737:Architectures: all 738:Type: vcpu ioctl 739:Parameters: struct kvm_signal_mask (in) 740:Returns: 0 on success, -1 on error 741 742Defines which signals are blocked during execution of KVM_RUN. This 743signal mask temporarily overrides the threads signal mask. Any 744unblocked signal received (except SIGKILL and SIGSTOP, which retain 745their traditional behaviour) will cause KVM_RUN to return with -EINTR. 746 747Note the signal will only be delivered if not blocked by the original 748signal mask. 749 750:: 751 752 /* for KVM_SET_SIGNAL_MASK */ 753 struct kvm_signal_mask { 754 __u32 len; 755 __u8 sigset[0]; 756 }; 757 758 7594.22 KVM_GET_FPU 760---------------- 761 762:Capability: basic 763:Architectures: x86 764:Type: vcpu ioctl 765:Parameters: struct kvm_fpu (out) 766:Returns: 0 on success, -1 on error 767 768Reads the floating point state from the vcpu. 769 770:: 771 772 /* for KVM_GET_FPU and KVM_SET_FPU */ 773 struct kvm_fpu { 774 __u8 fpr[8][16]; 775 __u16 fcw; 776 __u16 fsw; 777 __u8 ftwx; /* in fxsave format */ 778 __u8 pad1; 779 __u16 last_opcode; 780 __u64 last_ip; 781 __u64 last_dp; 782 __u8 xmm[16][16]; 783 __u32 mxcsr; 784 __u32 pad2; 785 }; 786 787 7884.23 KVM_SET_FPU 789---------------- 790 791:Capability: basic 792:Architectures: x86 793:Type: vcpu ioctl 794:Parameters: struct kvm_fpu (in) 795:Returns: 0 on success, -1 on error 796 797Writes the floating point state to the vcpu. 798 799:: 800 801 /* for KVM_GET_FPU and KVM_SET_FPU */ 802 struct kvm_fpu { 803 __u8 fpr[8][16]; 804 __u16 fcw; 805 __u16 fsw; 806 __u8 ftwx; /* in fxsave format */ 807 __u8 pad1; 808 __u16 last_opcode; 809 __u64 last_ip; 810 __u64 last_dp; 811 __u8 xmm[16][16]; 812 __u32 mxcsr; 813 __u32 pad2; 814 }; 815 816 8174.24 KVM_CREATE_IRQCHIP 818----------------------- 819 820:Capability: KVM_CAP_IRQCHIP, KVM_CAP_S390_IRQCHIP (s390) 821:Architectures: x86, arm64, s390 822:Type: vm ioctl 823:Parameters: none 824:Returns: 0 on success, -1 on error 825 826Creates an interrupt controller model in the kernel. 827On x86, creates a virtual ioapic, a virtual PIC (two PICs, nested), and sets up 828future vcpus to have a local APIC. IRQ routing for GSIs 0-15 is set to both 829PIC and IOAPIC; GSI 16-23 only go to the IOAPIC. 830On arm64, a GICv2 is created. Any other GIC versions require the usage of 831KVM_CREATE_DEVICE, which also supports creating a GICv2. Using 832KVM_CREATE_DEVICE is preferred over KVM_CREATE_IRQCHIP for GICv2. 833On s390, a dummy irq routing table is created. 834 835Note that on s390 the KVM_CAP_S390_IRQCHIP vm capability needs to be enabled 836before KVM_CREATE_IRQCHIP can be used. 837 838 8394.25 KVM_IRQ_LINE 840----------------- 841 842:Capability: KVM_CAP_IRQCHIP 843:Architectures: x86, arm64 844:Type: vm ioctl 845:Parameters: struct kvm_irq_level 846:Returns: 0 on success, -1 on error 847 848Sets the level of a GSI input to the interrupt controller model in the kernel. 849On some architectures it is required that an interrupt controller model has 850been previously created with KVM_CREATE_IRQCHIP. Note that edge-triggered 851interrupts require the level to be set to 1 and then back to 0. 852 853On real hardware, interrupt pins can be active-low or active-high. This 854does not matter for the level field of struct kvm_irq_level: 1 always 855means active (asserted), 0 means inactive (deasserted). 856 857x86 allows the operating system to program the interrupt polarity 858(active-low/active-high) for level-triggered interrupts, and KVM used 859to consider the polarity. However, due to bitrot in the handling of 860active-low interrupts, the above convention is now valid on x86 too. 861This is signaled by KVM_CAP_X86_IOAPIC_POLARITY_IGNORED. Userspace 862should not present interrupts to the guest as active-low unless this 863capability is present (or unless it is not using the in-kernel irqchip, 864of course). 865 866 867arm64 can signal an interrupt either at the CPU level, or at the 868in-kernel irqchip (GIC), and for in-kernel irqchip can tell the GIC to 869use PPIs designated for specific cpus. The irq field is interpreted 870like this:: 871 872 bits: | 31 ... 28 | 27 ... 24 | 23 ... 16 | 15 ... 0 | 873 field: | vcpu2_index | irq_type | vcpu_index | irq_id | 874 875The irq_type field has the following values: 876 877- irq_type[0]: 878 out-of-kernel GIC: irq_id 0 is IRQ, irq_id 1 is FIQ 879- irq_type[1]: 880 in-kernel GIC: SPI, irq_id between 32 and 1019 (incl.) 881 (the vcpu_index field is ignored) 882- irq_type[2]: 883 in-kernel GIC: PPI, irq_id between 16 and 31 (incl.) 884 885(The irq_id field thus corresponds nicely to the IRQ ID in the ARM GIC specs) 886 887In both cases, level is used to assert/deassert the line. 888 889When KVM_CAP_ARM_IRQ_LINE_LAYOUT_2 is supported, the target vcpu is 890identified as (256 * vcpu2_index + vcpu_index). Otherwise, vcpu2_index 891must be zero. 892 893Note that on arm64, the KVM_CAP_IRQCHIP capability only conditions 894injection of interrupts for the in-kernel irqchip. KVM_IRQ_LINE can always 895be used for a userspace interrupt controller. 896 897:: 898 899 struct kvm_irq_level { 900 union { 901 __u32 irq; /* GSI */ 902 __s32 status; /* not used for KVM_IRQ_LEVEL */ 903 }; 904 __u32 level; /* 0 or 1 */ 905 }; 906 907 9084.26 KVM_GET_IRQCHIP 909-------------------- 910 911:Capability: KVM_CAP_IRQCHIP 912:Architectures: x86 913:Type: vm ioctl 914:Parameters: struct kvm_irqchip (in/out) 915:Returns: 0 on success, -1 on error 916 917Reads the state of a kernel interrupt controller created with 918KVM_CREATE_IRQCHIP into a buffer provided by the caller. 919 920:: 921 922 struct kvm_irqchip { 923 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */ 924 __u32 pad; 925 union { 926 char dummy[512]; /* reserving space */ 927 struct kvm_pic_state pic; 928 struct kvm_ioapic_state ioapic; 929 } chip; 930 }; 931 932 9334.27 KVM_SET_IRQCHIP 934-------------------- 935 936:Capability: KVM_CAP_IRQCHIP 937:Architectures: x86 938:Type: vm ioctl 939:Parameters: struct kvm_irqchip (in) 940:Returns: 0 on success, -1 on error 941 942Sets the state of a kernel interrupt controller created with 943KVM_CREATE_IRQCHIP from a buffer provided by the caller. 944 945:: 946 947 struct kvm_irqchip { 948 __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */ 949 __u32 pad; 950 union { 951 char dummy[512]; /* reserving space */ 952 struct kvm_pic_state pic; 953 struct kvm_ioapic_state ioapic; 954 } chip; 955 }; 956 957 9584.28 KVM_XEN_HVM_CONFIG 959----------------------- 960 961:Capability: KVM_CAP_XEN_HVM 962:Architectures: x86 963:Type: vm ioctl 964:Parameters: struct kvm_xen_hvm_config (in) 965:Returns: 0 on success, -1 on error 966 967Sets the MSR that the Xen HVM guest uses to initialize its hypercall 968page, and provides the starting address and size of the hypercall 969blobs in userspace. When the guest writes the MSR, kvm copies one 970page of a blob (32- or 64-bit, depending on the vcpu mode) to guest 971memory. 972 973:: 974 975 struct kvm_xen_hvm_config { 976 __u32 flags; 977 __u32 msr; 978 __u64 blob_addr_32; 979 __u64 blob_addr_64; 980 __u8 blob_size_32; 981 __u8 blob_size_64; 982 __u8 pad2[30]; 983 }; 984 985If the KVM_XEN_HVM_CONFIG_INTERCEPT_HCALL flag is returned from the 986KVM_CAP_XEN_HVM check, it may be set in the flags field of this ioctl. 987This requests KVM to generate the contents of the hypercall page 988automatically; hypercalls will be intercepted and passed to userspace 989through KVM_EXIT_XEN. In this case, all of the blob size and address 990fields must be zero. 991 992No other flags are currently valid in the struct kvm_xen_hvm_config. 993 9944.29 KVM_GET_CLOCK 995------------------ 996 997:Capability: KVM_CAP_ADJUST_CLOCK 998:Architectures: x86 999:Type: vm ioctl 1000:Parameters: struct kvm_clock_data (out) 1001:Returns: 0 on success, -1 on error 1002 1003Gets the current timestamp of kvmclock as seen by the current guest. In 1004conjunction with KVM_SET_CLOCK, it is used to ensure monotonicity on scenarios 1005such as migration. 1006 1007When KVM_CAP_ADJUST_CLOCK is passed to KVM_CHECK_EXTENSION, it returns the 1008set of bits that KVM can return in struct kvm_clock_data's flag member. 1009 1010The following flags are defined: 1011 1012KVM_CLOCK_TSC_STABLE 1013 If set, the returned value is the exact kvmclock 1014 value seen by all VCPUs at the instant when KVM_GET_CLOCK was called. 1015 If clear, the returned value is simply CLOCK_MONOTONIC plus a constant 1016 offset; the offset can be modified with KVM_SET_CLOCK. KVM will try 1017 to make all VCPUs follow this clock, but the exact value read by each 1018 VCPU could differ, because the host TSC is not stable. 1019 1020KVM_CLOCK_REALTIME 1021 If set, the `realtime` field in the kvm_clock_data 1022 structure is populated with the value of the host's real time 1023 clocksource at the instant when KVM_GET_CLOCK was called. If clear, 1024 the `realtime` field does not contain a value. 1025 1026KVM_CLOCK_HOST_TSC 1027 If set, the `host_tsc` field in the kvm_clock_data 1028 structure is populated with the value of the host's timestamp counter (TSC) 1029 at the instant when KVM_GET_CLOCK was called. If clear, the `host_tsc` field 1030 does not contain a value. 1031 1032:: 1033 1034 struct kvm_clock_data { 1035 __u64 clock; /* kvmclock current value */ 1036 __u32 flags; 1037 __u32 pad0; 1038 __u64 realtime; 1039 __u64 host_tsc; 1040 __u32 pad[4]; 1041 }; 1042 1043 10444.30 KVM_SET_CLOCK 1045------------------ 1046 1047:Capability: KVM_CAP_ADJUST_CLOCK 1048:Architectures: x86 1049:Type: vm ioctl 1050:Parameters: struct kvm_clock_data (in) 1051:Returns: 0 on success, -1 on error 1052 1053Sets the current timestamp of kvmclock to the value specified in its parameter. 1054In conjunction with KVM_GET_CLOCK, it is used to ensure monotonicity on scenarios 1055such as migration. 1056 1057The following flags can be passed: 1058 1059KVM_CLOCK_REALTIME 1060 If set, KVM will compare the value of the `realtime` field 1061 with the value of the host's real time clocksource at the instant when 1062 KVM_SET_CLOCK was called. The difference in elapsed time is added to the final 1063 kvmclock value that will be provided to guests. 1064 1065Other flags returned by ``KVM_GET_CLOCK`` are accepted but ignored. 1066 1067:: 1068 1069 struct kvm_clock_data { 1070 __u64 clock; /* kvmclock current value */ 1071 __u32 flags; 1072 __u32 pad0; 1073 __u64 realtime; 1074 __u64 host_tsc; 1075 __u32 pad[4]; 1076 }; 1077 1078 10794.31 KVM_GET_VCPU_EVENTS 1080------------------------ 1081 1082:Capability: KVM_CAP_VCPU_EVENTS 1083:Extended by: KVM_CAP_INTR_SHADOW 1084:Architectures: x86, arm64 1085:Type: vcpu ioctl 1086:Parameters: struct kvm_vcpu_event (out) 1087:Returns: 0 on success, -1 on error 1088 1089X86: 1090^^^^ 1091 1092Gets currently pending exceptions, interrupts, and NMIs as well as related 1093states of the vcpu. 1094 1095:: 1096 1097 struct kvm_vcpu_events { 1098 struct { 1099 __u8 injected; 1100 __u8 nr; 1101 __u8 has_error_code; 1102 __u8 pending; 1103 __u32 error_code; 1104 } exception; 1105 struct { 1106 __u8 injected; 1107 __u8 nr; 1108 __u8 soft; 1109 __u8 shadow; 1110 } interrupt; 1111 struct { 1112 __u8 injected; 1113 __u8 pending; 1114 __u8 masked; 1115 __u8 pad; 1116 } nmi; 1117 __u32 sipi_vector; 1118 __u32 flags; 1119 struct { 1120 __u8 smm; 1121 __u8 pending; 1122 __u8 smm_inside_nmi; 1123 __u8 latched_init; 1124 } smi; 1125 __u8 reserved[27]; 1126 __u8 exception_has_payload; 1127 __u64 exception_payload; 1128 }; 1129 1130The following bits are defined in the flags field: 1131 1132- KVM_VCPUEVENT_VALID_SHADOW may be set to signal that 1133 interrupt.shadow contains a valid state. 1134 1135- KVM_VCPUEVENT_VALID_SMM may be set to signal that smi contains a 1136 valid state. 1137 1138- KVM_VCPUEVENT_VALID_PAYLOAD may be set to signal that the 1139 exception_has_payload, exception_payload, and exception.pending 1140 fields contain a valid state. This bit will be set whenever 1141 KVM_CAP_EXCEPTION_PAYLOAD is enabled. 1142 1143ARM64: 1144^^^^^^ 1145 1146If the guest accesses a device that is being emulated by the host kernel in 1147such a way that a real device would generate a physical SError, KVM may make 1148a virtual SError pending for that VCPU. This system error interrupt remains 1149pending until the guest takes the exception by unmasking PSTATE.A. 1150 1151Running the VCPU may cause it to take a pending SError, or make an access that 1152causes an SError to become pending. The event's description is only valid while 1153the VPCU is not running. 1154 1155This API provides a way to read and write the pending 'event' state that is not 1156visible to the guest. To save, restore or migrate a VCPU the struct representing 1157the state can be read then written using this GET/SET API, along with the other 1158guest-visible registers. It is not possible to 'cancel' an SError that has been 1159made pending. 1160 1161A device being emulated in user-space may also wish to generate an SError. To do 1162this the events structure can be populated by user-space. The current state 1163should be read first, to ensure no existing SError is pending. If an existing 1164SError is pending, the architecture's 'Multiple SError interrupts' rules should 1165be followed. (2.5.3 of DDI0587.a "ARM Reliability, Availability, and 1166Serviceability (RAS) Specification"). 1167 1168SError exceptions always have an ESR value. Some CPUs have the ability to 1169specify what the virtual SError's ESR value should be. These systems will 1170advertise KVM_CAP_ARM_INJECT_SERROR_ESR. In this case exception.has_esr will 1171always have a non-zero value when read, and the agent making an SError pending 1172should specify the ISS field in the lower 24 bits of exception.serror_esr. If 1173the system supports KVM_CAP_ARM_INJECT_SERROR_ESR, but user-space sets the events 1174with exception.has_esr as zero, KVM will choose an ESR. 1175 1176Specifying exception.has_esr on a system that does not support it will return 1177-EINVAL. Setting anything other than the lower 24bits of exception.serror_esr 1178will return -EINVAL. 1179 1180It is not possible to read back a pending external abort (injected via 1181KVM_SET_VCPU_EVENTS or otherwise) because such an exception is always delivered 1182directly to the virtual CPU). 1183 1184:: 1185 1186 struct kvm_vcpu_events { 1187 struct { 1188 __u8 serror_pending; 1189 __u8 serror_has_esr; 1190 __u8 ext_dabt_pending; 1191 /* Align it to 8 bytes */ 1192 __u8 pad[5]; 1193 __u64 serror_esr; 1194 } exception; 1195 __u32 reserved[12]; 1196 }; 1197 11984.32 KVM_SET_VCPU_EVENTS 1199------------------------ 1200 1201:Capability: KVM_CAP_VCPU_EVENTS 1202:Extended by: KVM_CAP_INTR_SHADOW 1203:Architectures: x86, arm64 1204:Type: vcpu ioctl 1205:Parameters: struct kvm_vcpu_event (in) 1206:Returns: 0 on success, -1 on error 1207 1208X86: 1209^^^^ 1210 1211Set pending exceptions, interrupts, and NMIs as well as related states of the 1212vcpu. 1213 1214See KVM_GET_VCPU_EVENTS for the data structure. 1215 1216Fields that may be modified asynchronously by running VCPUs can be excluded 1217from the update. These fields are nmi.pending, sipi_vector, smi.smm, 1218smi.pending. Keep the corresponding bits in the flags field cleared to 1219suppress overwriting the current in-kernel state. The bits are: 1220 1221=============================== ================================== 1222KVM_VCPUEVENT_VALID_NMI_PENDING transfer nmi.pending to the kernel 1223KVM_VCPUEVENT_VALID_SIPI_VECTOR transfer sipi_vector 1224KVM_VCPUEVENT_VALID_SMM transfer the smi sub-struct. 1225=============================== ================================== 1226 1227If KVM_CAP_INTR_SHADOW is available, KVM_VCPUEVENT_VALID_SHADOW can be set in 1228the flags field to signal that interrupt.shadow contains a valid state and 1229shall be written into the VCPU. 1230 1231KVM_VCPUEVENT_VALID_SMM can only be set if KVM_CAP_X86_SMM is available. 1232 1233If KVM_CAP_EXCEPTION_PAYLOAD is enabled, KVM_VCPUEVENT_VALID_PAYLOAD 1234can be set in the flags field to signal that the 1235exception_has_payload, exception_payload, and exception.pending fields 1236contain a valid state and shall be written into the VCPU. 1237 1238ARM64: 1239^^^^^^ 1240 1241User space may need to inject several types of events to the guest. 1242 1243Set the pending SError exception state for this VCPU. It is not possible to 1244'cancel' an Serror that has been made pending. 1245 1246If the guest performed an access to I/O memory which could not be handled by 1247userspace, for example because of missing instruction syndrome decode 1248information or because there is no device mapped at the accessed IPA, then 1249userspace can ask the kernel to inject an external abort using the address 1250from the exiting fault on the VCPU. It is a programming error to set 1251ext_dabt_pending after an exit which was not either KVM_EXIT_MMIO or 1252KVM_EXIT_ARM_NISV. This feature is only available if the system supports 1253KVM_CAP_ARM_INJECT_EXT_DABT. This is a helper which provides commonality in 1254how userspace reports accesses for the above cases to guests, across different 1255userspace implementations. Nevertheless, userspace can still emulate all Arm 1256exceptions by manipulating individual registers using the KVM_SET_ONE_REG API. 1257 1258See KVM_GET_VCPU_EVENTS for the data structure. 1259 1260 12614.33 KVM_GET_DEBUGREGS 1262---------------------- 1263 1264:Capability: KVM_CAP_DEBUGREGS 1265:Architectures: x86 1266:Type: vm ioctl 1267:Parameters: struct kvm_debugregs (out) 1268:Returns: 0 on success, -1 on error 1269 1270Reads debug registers from the vcpu. 1271 1272:: 1273 1274 struct kvm_debugregs { 1275 __u64 db[4]; 1276 __u64 dr6; 1277 __u64 dr7; 1278 __u64 flags; 1279 __u64 reserved[9]; 1280 }; 1281 1282 12834.34 KVM_SET_DEBUGREGS 1284---------------------- 1285 1286:Capability: KVM_CAP_DEBUGREGS 1287:Architectures: x86 1288:Type: vm ioctl 1289:Parameters: struct kvm_debugregs (in) 1290:Returns: 0 on success, -1 on error 1291 1292Writes debug registers into the vcpu. 1293 1294See KVM_GET_DEBUGREGS for the data structure. The flags field is unused 1295yet and must be cleared on entry. 1296 1297 12984.35 KVM_SET_USER_MEMORY_REGION 1299------------------------------- 1300 1301:Capability: KVM_CAP_USER_MEMORY 1302:Architectures: all 1303:Type: vm ioctl 1304:Parameters: struct kvm_userspace_memory_region (in) 1305:Returns: 0 on success, -1 on error 1306 1307:: 1308 1309 struct kvm_userspace_memory_region { 1310 __u32 slot; 1311 __u32 flags; 1312 __u64 guest_phys_addr; 1313 __u64 memory_size; /* bytes */ 1314 __u64 userspace_addr; /* start of the userspace allocated memory */ 1315 }; 1316 1317 /* for kvm_memory_region::flags */ 1318 #define KVM_MEM_LOG_DIRTY_PAGES (1UL << 0) 1319 #define KVM_MEM_READONLY (1UL << 1) 1320 1321This ioctl allows the user to create, modify or delete a guest physical 1322memory slot. Bits 0-15 of "slot" specify the slot id and this value 1323should be less than the maximum number of user memory slots supported per 1324VM. The maximum allowed slots can be queried using KVM_CAP_NR_MEMSLOTS. 1325Slots may not overlap in guest physical address space. 1326 1327If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of "slot" 1328specifies the address space which is being modified. They must be 1329less than the value that KVM_CHECK_EXTENSION returns for the 1330KVM_CAP_MULTI_ADDRESS_SPACE capability. Slots in separate address spaces 1331are unrelated; the restriction on overlapping slots only applies within 1332each address space. 1333 1334Deleting a slot is done by passing zero for memory_size. When changing 1335an existing slot, it may be moved in the guest physical memory space, 1336or its flags may be modified, but it may not be resized. 1337 1338Memory for the region is taken starting at the address denoted by the 1339field userspace_addr, which must point at user addressable memory for 1340the entire memory slot size. Any object may back this memory, including 1341anonymous memory, ordinary files, and hugetlbfs. 1342 1343On architectures that support a form of address tagging, userspace_addr must 1344be an untagged address. 1345 1346It is recommended that the lower 21 bits of guest_phys_addr and userspace_addr 1347be identical. This allows large pages in the guest to be backed by large 1348pages in the host. 1349 1350The flags field supports two flags: KVM_MEM_LOG_DIRTY_PAGES and 1351KVM_MEM_READONLY. The former can be set to instruct KVM to keep track of 1352writes to memory within the slot. See KVM_GET_DIRTY_LOG ioctl to know how to 1353use it. The latter can be set, if KVM_CAP_READONLY_MEM capability allows it, 1354to make a new slot read-only. In this case, writes to this memory will be 1355posted to userspace as KVM_EXIT_MMIO exits. 1356 1357When the KVM_CAP_SYNC_MMU capability is available, changes in the backing of 1358the memory region are automatically reflected into the guest. For example, an 1359mmap() that affects the region will be made visible immediately. Another 1360example is madvise(MADV_DROP). 1361 1362It is recommended to use this API instead of the KVM_SET_MEMORY_REGION ioctl. 1363The KVM_SET_MEMORY_REGION does not allow fine grained control over memory 1364allocation and is deprecated. 1365 1366 13674.36 KVM_SET_TSS_ADDR 1368--------------------- 1369 1370:Capability: KVM_CAP_SET_TSS_ADDR 1371:Architectures: x86 1372:Type: vm ioctl 1373:Parameters: unsigned long tss_address (in) 1374:Returns: 0 on success, -1 on error 1375 1376This ioctl defines the physical address of a three-page region in the guest 1377physical address space. The region must be within the first 4GB of the 1378guest physical address space and must not conflict with any memory slot 1379or any mmio address. The guest may malfunction if it accesses this memory 1380region. 1381 1382This ioctl is required on Intel-based hosts. This is needed on Intel hardware 1383because of a quirk in the virtualization implementation (see the internals 1384documentation when it pops into existence). 1385 1386 13874.37 KVM_ENABLE_CAP 1388------------------- 1389 1390:Capability: KVM_CAP_ENABLE_CAP 1391:Architectures: mips, ppc, s390, x86 1392:Type: vcpu ioctl 1393:Parameters: struct kvm_enable_cap (in) 1394:Returns: 0 on success; -1 on error 1395 1396:Capability: KVM_CAP_ENABLE_CAP_VM 1397:Architectures: all 1398:Type: vm ioctl 1399:Parameters: struct kvm_enable_cap (in) 1400:Returns: 0 on success; -1 on error 1401 1402.. note:: 1403 1404 Not all extensions are enabled by default. Using this ioctl the application 1405 can enable an extension, making it available to the guest. 1406 1407On systems that do not support this ioctl, it always fails. On systems that 1408do support it, it only works for extensions that are supported for enablement. 1409 1410To check if a capability can be enabled, the KVM_CHECK_EXTENSION ioctl should 1411be used. 1412 1413:: 1414 1415 struct kvm_enable_cap { 1416 /* in */ 1417 __u32 cap; 1418 1419The capability that is supposed to get enabled. 1420 1421:: 1422 1423 __u32 flags; 1424 1425A bitfield indicating future enhancements. Has to be 0 for now. 1426 1427:: 1428 1429 __u64 args[4]; 1430 1431Arguments for enabling a feature. If a feature needs initial values to 1432function properly, this is the place to put them. 1433 1434:: 1435 1436 __u8 pad[64]; 1437 }; 1438 1439The vcpu ioctl should be used for vcpu-specific capabilities, the vm ioctl 1440for vm-wide capabilities. 1441 14424.38 KVM_GET_MP_STATE 1443--------------------- 1444 1445:Capability: KVM_CAP_MP_STATE 1446:Architectures: x86, s390, arm64, riscv 1447:Type: vcpu ioctl 1448:Parameters: struct kvm_mp_state (out) 1449:Returns: 0 on success; -1 on error 1450 1451:: 1452 1453 struct kvm_mp_state { 1454 __u32 mp_state; 1455 }; 1456 1457Returns the vcpu's current "multiprocessing state" (though also valid on 1458uniprocessor guests). 1459 1460Possible values are: 1461 1462 ========================== =============================================== 1463 KVM_MP_STATE_RUNNABLE the vcpu is currently running 1464 [x86,arm64,riscv] 1465 KVM_MP_STATE_UNINITIALIZED the vcpu is an application processor (AP) 1466 which has not yet received an INIT signal [x86] 1467 KVM_MP_STATE_INIT_RECEIVED the vcpu has received an INIT signal, and is 1468 now ready for a SIPI [x86] 1469 KVM_MP_STATE_HALTED the vcpu has executed a HLT instruction and 1470 is waiting for an interrupt [x86] 1471 KVM_MP_STATE_SIPI_RECEIVED the vcpu has just received a SIPI (vector 1472 accessible via KVM_GET_VCPU_EVENTS) [x86] 1473 KVM_MP_STATE_STOPPED the vcpu is stopped [s390,arm64,riscv] 1474 KVM_MP_STATE_CHECK_STOP the vcpu is in a special error state [s390] 1475 KVM_MP_STATE_OPERATING the vcpu is operating (running or halted) 1476 [s390] 1477 KVM_MP_STATE_LOAD the vcpu is in a special load/startup state 1478 [s390] 1479 ========================== =============================================== 1480 1481On x86, this ioctl is only useful after KVM_CREATE_IRQCHIP. Without an 1482in-kernel irqchip, the multiprocessing state must be maintained by userspace on 1483these architectures. 1484 1485For arm64/riscv: 1486^^^^^^^^^^^^^^^^ 1487 1488The only states that are valid are KVM_MP_STATE_STOPPED and 1489KVM_MP_STATE_RUNNABLE which reflect if the vcpu is paused or not. 1490 14914.39 KVM_SET_MP_STATE 1492--------------------- 1493 1494:Capability: KVM_CAP_MP_STATE 1495:Architectures: x86, s390, arm64, riscv 1496:Type: vcpu ioctl 1497:Parameters: struct kvm_mp_state (in) 1498:Returns: 0 on success; -1 on error 1499 1500Sets the vcpu's current "multiprocessing state"; see KVM_GET_MP_STATE for 1501arguments. 1502 1503On x86, this ioctl is only useful after KVM_CREATE_IRQCHIP. Without an 1504in-kernel irqchip, the multiprocessing state must be maintained by userspace on 1505these architectures. 1506 1507For arm64/riscv: 1508^^^^^^^^^^^^^^^^ 1509 1510The only states that are valid are KVM_MP_STATE_STOPPED and 1511KVM_MP_STATE_RUNNABLE which reflect if the vcpu should be paused or not. 1512 15134.40 KVM_SET_IDENTITY_MAP_ADDR 1514------------------------------ 1515 1516:Capability: KVM_CAP_SET_IDENTITY_MAP_ADDR 1517:Architectures: x86 1518:Type: vm ioctl 1519:Parameters: unsigned long identity (in) 1520:Returns: 0 on success, -1 on error 1521 1522This ioctl defines the physical address of a one-page region in the guest 1523physical address space. The region must be within the first 4GB of the 1524guest physical address space and must not conflict with any memory slot 1525or any mmio address. The guest may malfunction if it accesses this memory 1526region. 1527 1528Setting the address to 0 will result in resetting the address to its default 1529(0xfffbc000). 1530 1531This ioctl is required on Intel-based hosts. This is needed on Intel hardware 1532because of a quirk in the virtualization implementation (see the internals 1533documentation when it pops into existence). 1534 1535Fails if any VCPU has already been created. 1536 15374.41 KVM_SET_BOOT_CPU_ID 1538------------------------ 1539 1540:Capability: KVM_CAP_SET_BOOT_CPU_ID 1541:Architectures: x86 1542:Type: vm ioctl 1543:Parameters: unsigned long vcpu_id 1544:Returns: 0 on success, -1 on error 1545 1546Define which vcpu is the Bootstrap Processor (BSP). Values are the same 1547as the vcpu id in KVM_CREATE_VCPU. If this ioctl is not called, the default 1548is vcpu 0. This ioctl has to be called before vcpu creation, 1549otherwise it will return EBUSY error. 1550 1551 15524.42 KVM_GET_XSAVE 1553------------------ 1554 1555:Capability: KVM_CAP_XSAVE 1556:Architectures: x86 1557:Type: vcpu ioctl 1558:Parameters: struct kvm_xsave (out) 1559:Returns: 0 on success, -1 on error 1560 1561 1562:: 1563 1564 struct kvm_xsave { 1565 __u32 region[1024]; 1566 __u32 extra[0]; 1567 }; 1568 1569This ioctl would copy current vcpu's xsave struct to the userspace. 1570 1571 15724.43 KVM_SET_XSAVE 1573------------------ 1574 1575:Capability: KVM_CAP_XSAVE and KVM_CAP_XSAVE2 1576:Architectures: x86 1577:Type: vcpu ioctl 1578:Parameters: struct kvm_xsave (in) 1579:Returns: 0 on success, -1 on error 1580 1581:: 1582 1583 1584 struct kvm_xsave { 1585 __u32 region[1024]; 1586 __u32 extra[0]; 1587 }; 1588 1589This ioctl would copy userspace's xsave struct to the kernel. It copies 1590as many bytes as are returned by KVM_CHECK_EXTENSION(KVM_CAP_XSAVE2), 1591when invoked on the vm file descriptor. The size value returned by 1592KVM_CHECK_EXTENSION(KVM_CAP_XSAVE2) will always be at least 4096. 1593Currently, it is only greater than 4096 if a dynamic feature has been 1594enabled with ``arch_prctl()``, but this may change in the future. 1595 1596The offsets of the state save areas in struct kvm_xsave follow the 1597contents of CPUID leaf 0xD on the host. 1598 1599 16004.44 KVM_GET_XCRS 1601----------------- 1602 1603:Capability: KVM_CAP_XCRS 1604:Architectures: x86 1605:Type: vcpu ioctl 1606:Parameters: struct kvm_xcrs (out) 1607:Returns: 0 on success, -1 on error 1608 1609:: 1610 1611 struct kvm_xcr { 1612 __u32 xcr; 1613 __u32 reserved; 1614 __u64 value; 1615 }; 1616 1617 struct kvm_xcrs { 1618 __u32 nr_xcrs; 1619 __u32 flags; 1620 struct kvm_xcr xcrs[KVM_MAX_XCRS]; 1621 __u64 padding[16]; 1622 }; 1623 1624This ioctl would copy current vcpu's xcrs to the userspace. 1625 1626 16274.45 KVM_SET_XCRS 1628----------------- 1629 1630:Capability: KVM_CAP_XCRS 1631:Architectures: x86 1632:Type: vcpu ioctl 1633:Parameters: struct kvm_xcrs (in) 1634:Returns: 0 on success, -1 on error 1635 1636:: 1637 1638 struct kvm_xcr { 1639 __u32 xcr; 1640 __u32 reserved; 1641 __u64 value; 1642 }; 1643 1644 struct kvm_xcrs { 1645 __u32 nr_xcrs; 1646 __u32 flags; 1647 struct kvm_xcr xcrs[KVM_MAX_XCRS]; 1648 __u64 padding[16]; 1649 }; 1650 1651This ioctl would set vcpu's xcr to the value userspace specified. 1652 1653 16544.46 KVM_GET_SUPPORTED_CPUID 1655---------------------------- 1656 1657:Capability: KVM_CAP_EXT_CPUID 1658:Architectures: x86 1659:Type: system ioctl 1660:Parameters: struct kvm_cpuid2 (in/out) 1661:Returns: 0 on success, -1 on error 1662 1663:: 1664 1665 struct kvm_cpuid2 { 1666 __u32 nent; 1667 __u32 padding; 1668 struct kvm_cpuid_entry2 entries[0]; 1669 }; 1670 1671 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX BIT(0) 1672 #define KVM_CPUID_FLAG_STATEFUL_FUNC BIT(1) /* deprecated */ 1673 #define KVM_CPUID_FLAG_STATE_READ_NEXT BIT(2) /* deprecated */ 1674 1675 struct kvm_cpuid_entry2 { 1676 __u32 function; 1677 __u32 index; 1678 __u32 flags; 1679 __u32 eax; 1680 __u32 ebx; 1681 __u32 ecx; 1682 __u32 edx; 1683 __u32 padding[3]; 1684 }; 1685 1686This ioctl returns x86 cpuid features which are supported by both the 1687hardware and kvm in its default configuration. Userspace can use the 1688information returned by this ioctl to construct cpuid information (for 1689KVM_SET_CPUID2) that is consistent with hardware, kernel, and 1690userspace capabilities, and with user requirements (for example, the 1691user may wish to constrain cpuid to emulate older hardware, or for 1692feature consistency across a cluster). 1693 1694Dynamically-enabled feature bits need to be requested with 1695``arch_prctl()`` before calling this ioctl. Feature bits that have not 1696been requested are excluded from the result. 1697 1698Note that certain capabilities, such as KVM_CAP_X86_DISABLE_EXITS, may 1699expose cpuid features (e.g. MONITOR) which are not supported by kvm in 1700its default configuration. If userspace enables such capabilities, it 1701is responsible for modifying the results of this ioctl appropriately. 1702 1703Userspace invokes KVM_GET_SUPPORTED_CPUID by passing a kvm_cpuid2 structure 1704with the 'nent' field indicating the number of entries in the variable-size 1705array 'entries'. If the number of entries is too low to describe the cpu 1706capabilities, an error (E2BIG) is returned. If the number is too high, 1707the 'nent' field is adjusted and an error (ENOMEM) is returned. If the 1708number is just right, the 'nent' field is adjusted to the number of valid 1709entries in the 'entries' array, which is then filled. 1710 1711The entries returned are the host cpuid as returned by the cpuid instruction, 1712with unknown or unsupported features masked out. Some features (for example, 1713x2apic), may not be present in the host cpu, but are exposed by kvm if it can 1714emulate them efficiently. The fields in each entry are defined as follows: 1715 1716 function: 1717 the eax value used to obtain the entry 1718 1719 index: 1720 the ecx value used to obtain the entry (for entries that are 1721 affected by ecx) 1722 1723 flags: 1724 an OR of zero or more of the following: 1725 1726 KVM_CPUID_FLAG_SIGNIFCANT_INDEX: 1727 if the index field is valid 1728 1729 eax, ebx, ecx, edx: 1730 the values returned by the cpuid instruction for 1731 this function/index combination 1732 1733The TSC deadline timer feature (CPUID leaf 1, ecx[24]) is always returned 1734as false, since the feature depends on KVM_CREATE_IRQCHIP for local APIC 1735support. Instead it is reported via:: 1736 1737 ioctl(KVM_CHECK_EXTENSION, KVM_CAP_TSC_DEADLINE_TIMER) 1738 1739if that returns true and you use KVM_CREATE_IRQCHIP, or if you emulate the 1740feature in userspace, then you can enable the feature for KVM_SET_CPUID2. 1741 1742 17434.47 KVM_PPC_GET_PVINFO 1744----------------------- 1745 1746:Capability: KVM_CAP_PPC_GET_PVINFO 1747:Architectures: ppc 1748:Type: vm ioctl 1749:Parameters: struct kvm_ppc_pvinfo (out) 1750:Returns: 0 on success, !0 on error 1751 1752:: 1753 1754 struct kvm_ppc_pvinfo { 1755 __u32 flags; 1756 __u32 hcall[4]; 1757 __u8 pad[108]; 1758 }; 1759 1760This ioctl fetches PV specific information that need to be passed to the guest 1761using the device tree or other means from vm context. 1762 1763The hcall array defines 4 instructions that make up a hypercall. 1764 1765If any additional field gets added to this structure later on, a bit for that 1766additional piece of information will be set in the flags bitmap. 1767 1768The flags bitmap is defined as:: 1769 1770 /* the host supports the ePAPR idle hcall 1771 #define KVM_PPC_PVINFO_FLAGS_EV_IDLE (1<<0) 1772 17734.52 KVM_SET_GSI_ROUTING 1774------------------------ 1775 1776:Capability: KVM_CAP_IRQ_ROUTING 1777:Architectures: x86 s390 arm64 1778:Type: vm ioctl 1779:Parameters: struct kvm_irq_routing (in) 1780:Returns: 0 on success, -1 on error 1781 1782Sets the GSI routing table entries, overwriting any previously set entries. 1783 1784On arm64, GSI routing has the following limitation: 1785 1786- GSI routing does not apply to KVM_IRQ_LINE but only to KVM_IRQFD. 1787 1788:: 1789 1790 struct kvm_irq_routing { 1791 __u32 nr; 1792 __u32 flags; 1793 struct kvm_irq_routing_entry entries[0]; 1794 }; 1795 1796No flags are specified so far, the corresponding field must be set to zero. 1797 1798:: 1799 1800 struct kvm_irq_routing_entry { 1801 __u32 gsi; 1802 __u32 type; 1803 __u32 flags; 1804 __u32 pad; 1805 union { 1806 struct kvm_irq_routing_irqchip irqchip; 1807 struct kvm_irq_routing_msi msi; 1808 struct kvm_irq_routing_s390_adapter adapter; 1809 struct kvm_irq_routing_hv_sint hv_sint; 1810 struct kvm_irq_routing_xen_evtchn xen_evtchn; 1811 __u32 pad[8]; 1812 } u; 1813 }; 1814 1815 /* gsi routing entry types */ 1816 #define KVM_IRQ_ROUTING_IRQCHIP 1 1817 #define KVM_IRQ_ROUTING_MSI 2 1818 #define KVM_IRQ_ROUTING_S390_ADAPTER 3 1819 #define KVM_IRQ_ROUTING_HV_SINT 4 1820 #define KVM_IRQ_ROUTING_XEN_EVTCHN 5 1821 1822flags: 1823 1824- KVM_MSI_VALID_DEVID: used along with KVM_IRQ_ROUTING_MSI routing entry 1825 type, specifies that the devid field contains a valid value. The per-VM 1826 KVM_CAP_MSI_DEVID capability advertises the requirement to provide 1827 the device ID. If this capability is not available, userspace should 1828 never set the KVM_MSI_VALID_DEVID flag as the ioctl might fail. 1829- zero otherwise 1830 1831:: 1832 1833 struct kvm_irq_routing_irqchip { 1834 __u32 irqchip; 1835 __u32 pin; 1836 }; 1837 1838 struct kvm_irq_routing_msi { 1839 __u32 address_lo; 1840 __u32 address_hi; 1841 __u32 data; 1842 union { 1843 __u32 pad; 1844 __u32 devid; 1845 }; 1846 }; 1847 1848If KVM_MSI_VALID_DEVID is set, devid contains a unique device identifier 1849for the device that wrote the MSI message. For PCI, this is usually a 1850BFD identifier in the lower 16 bits. 1851 1852On x86, address_hi is ignored unless the KVM_X2APIC_API_USE_32BIT_IDS 1853feature of KVM_CAP_X2APIC_API capability is enabled. If it is enabled, 1854address_hi bits 31-8 provide bits 31-8 of the destination id. Bits 7-0 of 1855address_hi must be zero. 1856 1857:: 1858 1859 struct kvm_irq_routing_s390_adapter { 1860 __u64 ind_addr; 1861 __u64 summary_addr; 1862 __u64 ind_offset; 1863 __u32 summary_offset; 1864 __u32 adapter_id; 1865 }; 1866 1867 struct kvm_irq_routing_hv_sint { 1868 __u32 vcpu; 1869 __u32 sint; 1870 }; 1871 1872 struct kvm_irq_routing_xen_evtchn { 1873 __u32 port; 1874 __u32 vcpu; 1875 __u32 priority; 1876 }; 1877 1878 1879When KVM_CAP_XEN_HVM includes the KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL bit 1880in its indication of supported features, routing to Xen event channels 1881is supported. Although the priority field is present, only the value 1882KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL is supported, which means delivery by 18832 level event channels. FIFO event channel support may be added in 1884the future. 1885 1886 18874.55 KVM_SET_TSC_KHZ 1888-------------------- 1889 1890:Capability: KVM_CAP_TSC_CONTROL 1891:Architectures: x86 1892:Type: vcpu ioctl 1893:Parameters: virtual tsc_khz 1894:Returns: 0 on success, -1 on error 1895 1896Specifies the tsc frequency for the virtual machine. The unit of the 1897frequency is KHz. 1898 1899 19004.56 KVM_GET_TSC_KHZ 1901-------------------- 1902 1903:Capability: KVM_CAP_GET_TSC_KHZ 1904:Architectures: x86 1905:Type: vcpu ioctl 1906:Parameters: none 1907:Returns: virtual tsc-khz on success, negative value on error 1908 1909Returns the tsc frequency of the guest. The unit of the return value is 1910KHz. If the host has unstable tsc this ioctl returns -EIO instead as an 1911error. 1912 1913 19144.57 KVM_GET_LAPIC 1915------------------ 1916 1917:Capability: KVM_CAP_IRQCHIP 1918:Architectures: x86 1919:Type: vcpu ioctl 1920:Parameters: struct kvm_lapic_state (out) 1921:Returns: 0 on success, -1 on error 1922 1923:: 1924 1925 #define KVM_APIC_REG_SIZE 0x400 1926 struct kvm_lapic_state { 1927 char regs[KVM_APIC_REG_SIZE]; 1928 }; 1929 1930Reads the Local APIC registers and copies them into the input argument. The 1931data format and layout are the same as documented in the architecture manual. 1932 1933If KVM_X2APIC_API_USE_32BIT_IDS feature of KVM_CAP_X2APIC_API is 1934enabled, then the format of APIC_ID register depends on the APIC mode 1935(reported by MSR_IA32_APICBASE) of its VCPU. x2APIC stores APIC ID in 1936the APIC_ID register (bytes 32-35). xAPIC only allows an 8-bit APIC ID 1937which is stored in bits 31-24 of the APIC register, or equivalently in 1938byte 35 of struct kvm_lapic_state's regs field. KVM_GET_LAPIC must then 1939be called after MSR_IA32_APICBASE has been set with KVM_SET_MSR. 1940 1941If KVM_X2APIC_API_USE_32BIT_IDS feature is disabled, struct kvm_lapic_state 1942always uses xAPIC format. 1943 1944 19454.58 KVM_SET_LAPIC 1946------------------ 1947 1948:Capability: KVM_CAP_IRQCHIP 1949:Architectures: x86 1950:Type: vcpu ioctl 1951:Parameters: struct kvm_lapic_state (in) 1952:Returns: 0 on success, -1 on error 1953 1954:: 1955 1956 #define KVM_APIC_REG_SIZE 0x400 1957 struct kvm_lapic_state { 1958 char regs[KVM_APIC_REG_SIZE]; 1959 }; 1960 1961Copies the input argument into the Local APIC registers. The data format 1962and layout are the same as documented in the architecture manual. 1963 1964The format of the APIC ID register (bytes 32-35 of struct kvm_lapic_state's 1965regs field) depends on the state of the KVM_CAP_X2APIC_API capability. 1966See the note in KVM_GET_LAPIC. 1967 1968 19694.59 KVM_IOEVENTFD 1970------------------ 1971 1972:Capability: KVM_CAP_IOEVENTFD 1973:Architectures: all 1974:Type: vm ioctl 1975:Parameters: struct kvm_ioeventfd (in) 1976:Returns: 0 on success, !0 on error 1977 1978This ioctl attaches or detaches an ioeventfd to a legal pio/mmio address 1979within the guest. A guest write in the registered address will signal the 1980provided event instead of triggering an exit. 1981 1982:: 1983 1984 struct kvm_ioeventfd { 1985 __u64 datamatch; 1986 __u64 addr; /* legal pio/mmio address */ 1987 __u32 len; /* 0, 1, 2, 4, or 8 bytes */ 1988 __s32 fd; 1989 __u32 flags; 1990 __u8 pad[36]; 1991 }; 1992 1993For the special case of virtio-ccw devices on s390, the ioevent is matched 1994to a subchannel/virtqueue tuple instead. 1995 1996The following flags are defined:: 1997 1998 #define KVM_IOEVENTFD_FLAG_DATAMATCH (1 << kvm_ioeventfd_flag_nr_datamatch) 1999 #define KVM_IOEVENTFD_FLAG_PIO (1 << kvm_ioeventfd_flag_nr_pio) 2000 #define KVM_IOEVENTFD_FLAG_DEASSIGN (1 << kvm_ioeventfd_flag_nr_deassign) 2001 #define KVM_IOEVENTFD_FLAG_VIRTIO_CCW_NOTIFY \ 2002 (1 << kvm_ioeventfd_flag_nr_virtio_ccw_notify) 2003 2004If datamatch flag is set, the event will be signaled only if the written value 2005to the registered address is equal to datamatch in struct kvm_ioeventfd. 2006 2007For virtio-ccw devices, addr contains the subchannel id and datamatch the 2008virtqueue index. 2009 2010With KVM_CAP_IOEVENTFD_ANY_LENGTH, a zero length ioeventfd is allowed, and 2011the kernel will ignore the length of guest write and may get a faster vmexit. 2012The speedup may only apply to specific architectures, but the ioeventfd will 2013work anyway. 2014 20154.60 KVM_DIRTY_TLB 2016------------------ 2017 2018:Capability: KVM_CAP_SW_TLB 2019:Architectures: ppc 2020:Type: vcpu ioctl 2021:Parameters: struct kvm_dirty_tlb (in) 2022:Returns: 0 on success, -1 on error 2023 2024:: 2025 2026 struct kvm_dirty_tlb { 2027 __u64 bitmap; 2028 __u32 num_dirty; 2029 }; 2030 2031This must be called whenever userspace has changed an entry in the shared 2032TLB, prior to calling KVM_RUN on the associated vcpu. 2033 2034The "bitmap" field is the userspace address of an array. This array 2035consists of a number of bits, equal to the total number of TLB entries as 2036determined by the last successful call to KVM_CONFIG_TLB, rounded up to the 2037nearest multiple of 64. 2038 2039Each bit corresponds to one TLB entry, ordered the same as in the shared TLB 2040array. 2041 2042The array is little-endian: the bit 0 is the least significant bit of the 2043first byte, bit 8 is the least significant bit of the second byte, etc. 2044This avoids any complications with differing word sizes. 2045 2046The "num_dirty" field is a performance hint for KVM to determine whether it 2047should skip processing the bitmap and just invalidate everything. It must 2048be set to the number of set bits in the bitmap. 2049 2050 20514.62 KVM_CREATE_SPAPR_TCE 2052------------------------- 2053 2054:Capability: KVM_CAP_SPAPR_TCE 2055:Architectures: powerpc 2056:Type: vm ioctl 2057:Parameters: struct kvm_create_spapr_tce (in) 2058:Returns: file descriptor for manipulating the created TCE table 2059 2060This creates a virtual TCE (translation control entry) table, which 2061is an IOMMU for PAPR-style virtual I/O. It is used to translate 2062logical addresses used in virtual I/O into guest physical addresses, 2063and provides a scatter/gather capability for PAPR virtual I/O. 2064 2065:: 2066 2067 /* for KVM_CAP_SPAPR_TCE */ 2068 struct kvm_create_spapr_tce { 2069 __u64 liobn; 2070 __u32 window_size; 2071 }; 2072 2073The liobn field gives the logical IO bus number for which to create a 2074TCE table. The window_size field specifies the size of the DMA window 2075which this TCE table will translate - the table will contain one 64 2076bit TCE entry for every 4kiB of the DMA window. 2077 2078When the guest issues an H_PUT_TCE hcall on a liobn for which a TCE 2079table has been created using this ioctl(), the kernel will handle it 2080in real mode, updating the TCE table. H_PUT_TCE calls for other 2081liobns will cause a vm exit and must be handled by userspace. 2082 2083The return value is a file descriptor which can be passed to mmap(2) 2084to map the created TCE table into userspace. This lets userspace read 2085the entries written by kernel-handled H_PUT_TCE calls, and also lets 2086userspace update the TCE table directly which is useful in some 2087circumstances. 2088 2089 20904.63 KVM_ALLOCATE_RMA 2091--------------------- 2092 2093:Capability: KVM_CAP_PPC_RMA 2094:Architectures: powerpc 2095:Type: vm ioctl 2096:Parameters: struct kvm_allocate_rma (out) 2097:Returns: file descriptor for mapping the allocated RMA 2098 2099This allocates a Real Mode Area (RMA) from the pool allocated at boot 2100time by the kernel. An RMA is a physically-contiguous, aligned region 2101of memory used on older POWER processors to provide the memory which 2102will be accessed by real-mode (MMU off) accesses in a KVM guest. 2103POWER processors support a set of sizes for the RMA that usually 2104includes 64MB, 128MB, 256MB and some larger powers of two. 2105 2106:: 2107 2108 /* for KVM_ALLOCATE_RMA */ 2109 struct kvm_allocate_rma { 2110 __u64 rma_size; 2111 }; 2112 2113The return value is a file descriptor which can be passed to mmap(2) 2114to map the allocated RMA into userspace. The mapped area can then be 2115passed to the KVM_SET_USER_MEMORY_REGION ioctl to establish it as the 2116RMA for a virtual machine. The size of the RMA in bytes (which is 2117fixed at host kernel boot time) is returned in the rma_size field of 2118the argument structure. 2119 2120The KVM_CAP_PPC_RMA capability is 1 or 2 if the KVM_ALLOCATE_RMA ioctl 2121is supported; 2 if the processor requires all virtual machines to have 2122an RMA, or 1 if the processor can use an RMA but doesn't require it, 2123because it supports the Virtual RMA (VRMA) facility. 2124 2125 21264.64 KVM_NMI 2127------------ 2128 2129:Capability: KVM_CAP_USER_NMI 2130:Architectures: x86 2131:Type: vcpu ioctl 2132:Parameters: none 2133:Returns: 0 on success, -1 on error 2134 2135Queues an NMI on the thread's vcpu. Note this is well defined only 2136when KVM_CREATE_IRQCHIP has not been called, since this is an interface 2137between the virtual cpu core and virtual local APIC. After KVM_CREATE_IRQCHIP 2138has been called, this interface is completely emulated within the kernel. 2139 2140To use this to emulate the LINT1 input with KVM_CREATE_IRQCHIP, use the 2141following algorithm: 2142 2143 - pause the vcpu 2144 - read the local APIC's state (KVM_GET_LAPIC) 2145 - check whether changing LINT1 will queue an NMI (see the LVT entry for LINT1) 2146 - if so, issue KVM_NMI 2147 - resume the vcpu 2148 2149Some guests configure the LINT1 NMI input to cause a panic, aiding in 2150debugging. 2151 2152 21534.65 KVM_S390_UCAS_MAP 2154---------------------- 2155 2156:Capability: KVM_CAP_S390_UCONTROL 2157:Architectures: s390 2158:Type: vcpu ioctl 2159:Parameters: struct kvm_s390_ucas_mapping (in) 2160:Returns: 0 in case of success 2161 2162The parameter is defined like this:: 2163 2164 struct kvm_s390_ucas_mapping { 2165 __u64 user_addr; 2166 __u64 vcpu_addr; 2167 __u64 length; 2168 }; 2169 2170This ioctl maps the memory at "user_addr" with the length "length" to 2171the vcpu's address space starting at "vcpu_addr". All parameters need to 2172be aligned by 1 megabyte. 2173 2174 21754.66 KVM_S390_UCAS_UNMAP 2176------------------------ 2177 2178:Capability: KVM_CAP_S390_UCONTROL 2179:Architectures: s390 2180:Type: vcpu ioctl 2181:Parameters: struct kvm_s390_ucas_mapping (in) 2182:Returns: 0 in case of success 2183 2184The parameter is defined like this:: 2185 2186 struct kvm_s390_ucas_mapping { 2187 __u64 user_addr; 2188 __u64 vcpu_addr; 2189 __u64 length; 2190 }; 2191 2192This ioctl unmaps the memory in the vcpu's address space starting at 2193"vcpu_addr" with the length "length". The field "user_addr" is ignored. 2194All parameters need to be aligned by 1 megabyte. 2195 2196 21974.67 KVM_S390_VCPU_FAULT 2198------------------------ 2199 2200:Capability: KVM_CAP_S390_UCONTROL 2201:Architectures: s390 2202:Type: vcpu ioctl 2203:Parameters: vcpu absolute address (in) 2204:Returns: 0 in case of success 2205 2206This call creates a page table entry on the virtual cpu's address space 2207(for user controlled virtual machines) or the virtual machine's address 2208space (for regular virtual machines). This only works for minor faults, 2209thus it's recommended to access subject memory page via the user page 2210table upfront. This is useful to handle validity intercepts for user 2211controlled virtual machines to fault in the virtual cpu's lowcore pages 2212prior to calling the KVM_RUN ioctl. 2213 2214 22154.68 KVM_SET_ONE_REG 2216-------------------- 2217 2218:Capability: KVM_CAP_ONE_REG 2219:Architectures: all 2220:Type: vcpu ioctl 2221:Parameters: struct kvm_one_reg (in) 2222:Returns: 0 on success, negative value on failure 2223 2224Errors: 2225 2226 ====== ============================================================ 2227 ENOENT no such register 2228 EINVAL invalid register ID, or no such register or used with VMs in 2229 protected virtualization mode on s390 2230 EPERM (arm64) register access not allowed before vcpu finalization 2231 ====== ============================================================ 2232 2233(These error codes are indicative only: do not rely on a specific error 2234code being returned in a specific situation.) 2235 2236:: 2237 2238 struct kvm_one_reg { 2239 __u64 id; 2240 __u64 addr; 2241 }; 2242 2243Using this ioctl, a single vcpu register can be set to a specific value 2244defined by user space with the passed in struct kvm_one_reg, where id 2245refers to the register identifier as described below and addr is a pointer 2246to a variable with the respective size. There can be architecture agnostic 2247and architecture specific registers. Each have their own range of operation 2248and their own constants and width. To keep track of the implemented 2249registers, find a list below: 2250 2251 ======= =============================== ============ 2252 Arch Register Width (bits) 2253 ======= =============================== ============ 2254 PPC KVM_REG_PPC_HIOR 64 2255 PPC KVM_REG_PPC_IAC1 64 2256 PPC KVM_REG_PPC_IAC2 64 2257 PPC KVM_REG_PPC_IAC3 64 2258 PPC KVM_REG_PPC_IAC4 64 2259 PPC KVM_REG_PPC_DAC1 64 2260 PPC KVM_REG_PPC_DAC2 64 2261 PPC KVM_REG_PPC_DABR 64 2262 PPC KVM_REG_PPC_DSCR 64 2263 PPC KVM_REG_PPC_PURR 64 2264 PPC KVM_REG_PPC_SPURR 64 2265 PPC KVM_REG_PPC_DAR 64 2266 PPC KVM_REG_PPC_DSISR 32 2267 PPC KVM_REG_PPC_AMR 64 2268 PPC KVM_REG_PPC_UAMOR 64 2269 PPC KVM_REG_PPC_MMCR0 64 2270 PPC KVM_REG_PPC_MMCR1 64 2271 PPC KVM_REG_PPC_MMCRA 64 2272 PPC KVM_REG_PPC_MMCR2 64 2273 PPC KVM_REG_PPC_MMCRS 64 2274 PPC KVM_REG_PPC_MMCR3 64 2275 PPC KVM_REG_PPC_SIAR 64 2276 PPC KVM_REG_PPC_SDAR 64 2277 PPC KVM_REG_PPC_SIER 64 2278 PPC KVM_REG_PPC_SIER2 64 2279 PPC KVM_REG_PPC_SIER3 64 2280 PPC KVM_REG_PPC_PMC1 32 2281 PPC KVM_REG_PPC_PMC2 32 2282 PPC KVM_REG_PPC_PMC3 32 2283 PPC KVM_REG_PPC_PMC4 32 2284 PPC KVM_REG_PPC_PMC5 32 2285 PPC KVM_REG_PPC_PMC6 32 2286 PPC KVM_REG_PPC_PMC7 32 2287 PPC KVM_REG_PPC_PMC8 32 2288 PPC KVM_REG_PPC_FPR0 64 2289 ... 2290 PPC KVM_REG_PPC_FPR31 64 2291 PPC KVM_REG_PPC_VR0 128 2292 ... 2293 PPC KVM_REG_PPC_VR31 128 2294 PPC KVM_REG_PPC_VSR0 128 2295 ... 2296 PPC KVM_REG_PPC_VSR31 128 2297 PPC KVM_REG_PPC_FPSCR 64 2298 PPC KVM_REG_PPC_VSCR 32 2299 PPC KVM_REG_PPC_VPA_ADDR 64 2300 PPC KVM_REG_PPC_VPA_SLB 128 2301 PPC KVM_REG_PPC_VPA_DTL 128 2302 PPC KVM_REG_PPC_EPCR 32 2303 PPC KVM_REG_PPC_EPR 32 2304 PPC KVM_REG_PPC_TCR 32 2305 PPC KVM_REG_PPC_TSR 32 2306 PPC KVM_REG_PPC_OR_TSR 32 2307 PPC KVM_REG_PPC_CLEAR_TSR 32 2308 PPC KVM_REG_PPC_MAS0 32 2309 PPC KVM_REG_PPC_MAS1 32 2310 PPC KVM_REG_PPC_MAS2 64 2311 PPC KVM_REG_PPC_MAS7_3 64 2312 PPC KVM_REG_PPC_MAS4 32 2313 PPC KVM_REG_PPC_MAS6 32 2314 PPC KVM_REG_PPC_MMUCFG 32 2315 PPC KVM_REG_PPC_TLB0CFG 32 2316 PPC KVM_REG_PPC_TLB1CFG 32 2317 PPC KVM_REG_PPC_TLB2CFG 32 2318 PPC KVM_REG_PPC_TLB3CFG 32 2319 PPC KVM_REG_PPC_TLB0PS 32 2320 PPC KVM_REG_PPC_TLB1PS 32 2321 PPC KVM_REG_PPC_TLB2PS 32 2322 PPC KVM_REG_PPC_TLB3PS 32 2323 PPC KVM_REG_PPC_EPTCFG 32 2324 PPC KVM_REG_PPC_ICP_STATE 64 2325 PPC KVM_REG_PPC_VP_STATE 128 2326 PPC KVM_REG_PPC_TB_OFFSET 64 2327 PPC KVM_REG_PPC_SPMC1 32 2328 PPC KVM_REG_PPC_SPMC2 32 2329 PPC KVM_REG_PPC_IAMR 64 2330 PPC KVM_REG_PPC_TFHAR 64 2331 PPC KVM_REG_PPC_TFIAR 64 2332 PPC KVM_REG_PPC_TEXASR 64 2333 PPC KVM_REG_PPC_FSCR 64 2334 PPC KVM_REG_PPC_PSPB 32 2335 PPC KVM_REG_PPC_EBBHR 64 2336 PPC KVM_REG_PPC_EBBRR 64 2337 PPC KVM_REG_PPC_BESCR 64 2338 PPC KVM_REG_PPC_TAR 64 2339 PPC KVM_REG_PPC_DPDES 64 2340 PPC KVM_REG_PPC_DAWR 64 2341 PPC KVM_REG_PPC_DAWRX 64 2342 PPC KVM_REG_PPC_CIABR 64 2343 PPC KVM_REG_PPC_IC 64 2344 PPC KVM_REG_PPC_VTB 64 2345 PPC KVM_REG_PPC_CSIGR 64 2346 PPC KVM_REG_PPC_TACR 64 2347 PPC KVM_REG_PPC_TCSCR 64 2348 PPC KVM_REG_PPC_PID 64 2349 PPC KVM_REG_PPC_ACOP 64 2350 PPC KVM_REG_PPC_VRSAVE 32 2351 PPC KVM_REG_PPC_LPCR 32 2352 PPC KVM_REG_PPC_LPCR_64 64 2353 PPC KVM_REG_PPC_PPR 64 2354 PPC KVM_REG_PPC_ARCH_COMPAT 32 2355 PPC KVM_REG_PPC_DABRX 32 2356 PPC KVM_REG_PPC_WORT 64 2357 PPC KVM_REG_PPC_SPRG9 64 2358 PPC KVM_REG_PPC_DBSR 32 2359 PPC KVM_REG_PPC_TIDR 64 2360 PPC KVM_REG_PPC_PSSCR 64 2361 PPC KVM_REG_PPC_DEC_EXPIRY 64 2362 PPC KVM_REG_PPC_PTCR 64 2363 PPC KVM_REG_PPC_DAWR1 64 2364 PPC KVM_REG_PPC_DAWRX1 64 2365 PPC KVM_REG_PPC_TM_GPR0 64 2366 ... 2367 PPC KVM_REG_PPC_TM_GPR31 64 2368 PPC KVM_REG_PPC_TM_VSR0 128 2369 ... 2370 PPC KVM_REG_PPC_TM_VSR63 128 2371 PPC KVM_REG_PPC_TM_CR 64 2372 PPC KVM_REG_PPC_TM_LR 64 2373 PPC KVM_REG_PPC_TM_CTR 64 2374 PPC KVM_REG_PPC_TM_FPSCR 64 2375 PPC KVM_REG_PPC_TM_AMR 64 2376 PPC KVM_REG_PPC_TM_PPR 64 2377 PPC KVM_REG_PPC_TM_VRSAVE 64 2378 PPC KVM_REG_PPC_TM_VSCR 32 2379 PPC KVM_REG_PPC_TM_DSCR 64 2380 PPC KVM_REG_PPC_TM_TAR 64 2381 PPC KVM_REG_PPC_TM_XER 64 2382 2383 MIPS KVM_REG_MIPS_R0 64 2384 ... 2385 MIPS KVM_REG_MIPS_R31 64 2386 MIPS KVM_REG_MIPS_HI 64 2387 MIPS KVM_REG_MIPS_LO 64 2388 MIPS KVM_REG_MIPS_PC 64 2389 MIPS KVM_REG_MIPS_CP0_INDEX 32 2390 MIPS KVM_REG_MIPS_CP0_ENTRYLO0 64 2391 MIPS KVM_REG_MIPS_CP0_ENTRYLO1 64 2392 MIPS KVM_REG_MIPS_CP0_CONTEXT 64 2393 MIPS KVM_REG_MIPS_CP0_CONTEXTCONFIG 32 2394 MIPS KVM_REG_MIPS_CP0_USERLOCAL 64 2395 MIPS KVM_REG_MIPS_CP0_XCONTEXTCONFIG 64 2396 MIPS KVM_REG_MIPS_CP0_PAGEMASK 32 2397 MIPS KVM_REG_MIPS_CP0_PAGEGRAIN 32 2398 MIPS KVM_REG_MIPS_CP0_SEGCTL0 64 2399 MIPS KVM_REG_MIPS_CP0_SEGCTL1 64 2400 MIPS KVM_REG_MIPS_CP0_SEGCTL2 64 2401 MIPS KVM_REG_MIPS_CP0_PWBASE 64 2402 MIPS KVM_REG_MIPS_CP0_PWFIELD 64 2403 MIPS KVM_REG_MIPS_CP0_PWSIZE 64 2404 MIPS KVM_REG_MIPS_CP0_WIRED 32 2405 MIPS KVM_REG_MIPS_CP0_PWCTL 32 2406 MIPS KVM_REG_MIPS_CP0_HWRENA 32 2407 MIPS KVM_REG_MIPS_CP0_BADVADDR 64 2408 MIPS KVM_REG_MIPS_CP0_BADINSTR 32 2409 MIPS KVM_REG_MIPS_CP0_BADINSTRP 32 2410 MIPS KVM_REG_MIPS_CP0_COUNT 32 2411 MIPS KVM_REG_MIPS_CP0_ENTRYHI 64 2412 MIPS KVM_REG_MIPS_CP0_COMPARE 32 2413 MIPS KVM_REG_MIPS_CP0_STATUS 32 2414 MIPS KVM_REG_MIPS_CP0_INTCTL 32 2415 MIPS KVM_REG_MIPS_CP0_CAUSE 32 2416 MIPS KVM_REG_MIPS_CP0_EPC 64 2417 MIPS KVM_REG_MIPS_CP0_PRID 32 2418 MIPS KVM_REG_MIPS_CP0_EBASE 64 2419 MIPS KVM_REG_MIPS_CP0_CONFIG 32 2420 MIPS KVM_REG_MIPS_CP0_CONFIG1 32 2421 MIPS KVM_REG_MIPS_CP0_CONFIG2 32 2422 MIPS KVM_REG_MIPS_CP0_CONFIG3 32 2423 MIPS KVM_REG_MIPS_CP0_CONFIG4 32 2424 MIPS KVM_REG_MIPS_CP0_CONFIG5 32 2425 MIPS KVM_REG_MIPS_CP0_CONFIG7 32 2426 MIPS KVM_REG_MIPS_CP0_XCONTEXT 64 2427 MIPS KVM_REG_MIPS_CP0_ERROREPC 64 2428 MIPS KVM_REG_MIPS_CP0_KSCRATCH1 64 2429 MIPS KVM_REG_MIPS_CP0_KSCRATCH2 64 2430 MIPS KVM_REG_MIPS_CP0_KSCRATCH3 64 2431 MIPS KVM_REG_MIPS_CP0_KSCRATCH4 64 2432 MIPS KVM_REG_MIPS_CP0_KSCRATCH5 64 2433 MIPS KVM_REG_MIPS_CP0_KSCRATCH6 64 2434 MIPS KVM_REG_MIPS_CP0_MAAR(0..63) 64 2435 MIPS KVM_REG_MIPS_COUNT_CTL 64 2436 MIPS KVM_REG_MIPS_COUNT_RESUME 64 2437 MIPS KVM_REG_MIPS_COUNT_HZ 64 2438 MIPS KVM_REG_MIPS_FPR_32(0..31) 32 2439 MIPS KVM_REG_MIPS_FPR_64(0..31) 64 2440 MIPS KVM_REG_MIPS_VEC_128(0..31) 128 2441 MIPS KVM_REG_MIPS_FCR_IR 32 2442 MIPS KVM_REG_MIPS_FCR_CSR 32 2443 MIPS KVM_REG_MIPS_MSA_IR 32 2444 MIPS KVM_REG_MIPS_MSA_CSR 32 2445 ======= =============================== ============ 2446 2447ARM registers are mapped using the lower 32 bits. The upper 16 of that 2448is the register group type, or coprocessor number: 2449 2450ARM core registers have the following id bit patterns:: 2451 2452 0x4020 0000 0010 <index into the kvm_regs struct:16> 2453 2454ARM 32-bit CP15 registers have the following id bit patterns:: 2455 2456 0x4020 0000 000F <zero:1> <crn:4> <crm:4> <opc1:4> <opc2:3> 2457 2458ARM 64-bit CP15 registers have the following id bit patterns:: 2459 2460 0x4030 0000 000F <zero:1> <zero:4> <crm:4> <opc1:4> <zero:3> 2461 2462ARM CCSIDR registers are demultiplexed by CSSELR value:: 2463 2464 0x4020 0000 0011 00 <csselr:8> 2465 2466ARM 32-bit VFP control registers have the following id bit patterns:: 2467 2468 0x4020 0000 0012 1 <regno:12> 2469 2470ARM 64-bit FP registers have the following id bit patterns:: 2471 2472 0x4030 0000 0012 0 <regno:12> 2473 2474ARM firmware pseudo-registers have the following bit pattern:: 2475 2476 0x4030 0000 0014 <regno:16> 2477 2478 2479arm64 registers are mapped using the lower 32 bits. The upper 16 of 2480that is the register group type, or coprocessor number: 2481 2482arm64 core/FP-SIMD registers have the following id bit patterns. Note 2483that the size of the access is variable, as the kvm_regs structure 2484contains elements ranging from 32 to 128 bits. The index is a 32bit 2485value in the kvm_regs structure seen as a 32bit array:: 2486 2487 0x60x0 0000 0010 <index into the kvm_regs struct:16> 2488 2489Specifically: 2490 2491======================= ========= ===== ======================================= 2492 Encoding Register Bits kvm_regs member 2493======================= ========= ===== ======================================= 2494 0x6030 0000 0010 0000 X0 64 regs.regs[0] 2495 0x6030 0000 0010 0002 X1 64 regs.regs[1] 2496 ... 2497 0x6030 0000 0010 003c X30 64 regs.regs[30] 2498 0x6030 0000 0010 003e SP 64 regs.sp 2499 0x6030 0000 0010 0040 PC 64 regs.pc 2500 0x6030 0000 0010 0042 PSTATE 64 regs.pstate 2501 0x6030 0000 0010 0044 SP_EL1 64 sp_el1 2502 0x6030 0000 0010 0046 ELR_EL1 64 elr_el1 2503 0x6030 0000 0010 0048 SPSR_EL1 64 spsr[KVM_SPSR_EL1] (alias SPSR_SVC) 2504 0x6030 0000 0010 004a SPSR_ABT 64 spsr[KVM_SPSR_ABT] 2505 0x6030 0000 0010 004c SPSR_UND 64 spsr[KVM_SPSR_UND] 2506 0x6030 0000 0010 004e SPSR_IRQ 64 spsr[KVM_SPSR_IRQ] 2507 0x6060 0000 0010 0050 SPSR_FIQ 64 spsr[KVM_SPSR_FIQ] 2508 0x6040 0000 0010 0054 V0 128 fp_regs.vregs[0] [1]_ 2509 0x6040 0000 0010 0058 V1 128 fp_regs.vregs[1] [1]_ 2510 ... 2511 0x6040 0000 0010 00d0 V31 128 fp_regs.vregs[31] [1]_ 2512 0x6020 0000 0010 00d4 FPSR 32 fp_regs.fpsr 2513 0x6020 0000 0010 00d5 FPCR 32 fp_regs.fpcr 2514======================= ========= ===== ======================================= 2515 2516.. [1] These encodings are not accepted for SVE-enabled vcpus. See 2517 KVM_ARM_VCPU_INIT. 2518 2519 The equivalent register content can be accessed via bits [127:0] of 2520 the corresponding SVE Zn registers instead for vcpus that have SVE 2521 enabled (see below). 2522 2523arm64 CCSIDR registers are demultiplexed by CSSELR value:: 2524 2525 0x6020 0000 0011 00 <csselr:8> 2526 2527arm64 system registers have the following id bit patterns:: 2528 2529 0x6030 0000 0013 <op0:2> <op1:3> <crn:4> <crm:4> <op2:3> 2530 2531.. warning:: 2532 2533 Two system register IDs do not follow the specified pattern. These 2534 are KVM_REG_ARM_TIMER_CVAL and KVM_REG_ARM_TIMER_CNT, which map to 2535 system registers CNTV_CVAL_EL0 and CNTVCT_EL0 respectively. These 2536 two had their values accidentally swapped, which means TIMER_CVAL is 2537 derived from the register encoding for CNTVCT_EL0 and TIMER_CNT is 2538 derived from the register encoding for CNTV_CVAL_EL0. As this is 2539 API, it must remain this way. 2540 2541arm64 firmware pseudo-registers have the following bit pattern:: 2542 2543 0x6030 0000 0014 <regno:16> 2544 2545arm64 SVE registers have the following bit patterns:: 2546 2547 0x6080 0000 0015 00 <n:5> <slice:5> Zn bits[2048*slice + 2047 : 2048*slice] 2548 0x6050 0000 0015 04 <n:4> <slice:5> Pn bits[256*slice + 255 : 256*slice] 2549 0x6050 0000 0015 060 <slice:5> FFR bits[256*slice + 255 : 256*slice] 2550 0x6060 0000 0015 ffff KVM_REG_ARM64_SVE_VLS pseudo-register 2551 2552Access to register IDs where 2048 * slice >= 128 * max_vq will fail with 2553ENOENT. max_vq is the vcpu's maximum supported vector length in 128-bit 2554quadwords: see [2]_ below. 2555 2556These registers are only accessible on vcpus for which SVE is enabled. 2557See KVM_ARM_VCPU_INIT for details. 2558 2559In addition, except for KVM_REG_ARM64_SVE_VLS, these registers are not 2560accessible until the vcpu's SVE configuration has been finalized 2561using KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE). See KVM_ARM_VCPU_INIT 2562and KVM_ARM_VCPU_FINALIZE for more information about this procedure. 2563 2564KVM_REG_ARM64_SVE_VLS is a pseudo-register that allows the set of vector 2565lengths supported by the vcpu to be discovered and configured by 2566userspace. When transferred to or from user memory via KVM_GET_ONE_REG 2567or KVM_SET_ONE_REG, the value of this register is of type 2568__u64[KVM_ARM64_SVE_VLS_WORDS], and encodes the set of vector lengths as 2569follows:: 2570 2571 __u64 vector_lengths[KVM_ARM64_SVE_VLS_WORDS]; 2572 2573 if (vq >= SVE_VQ_MIN && vq <= SVE_VQ_MAX && 2574 ((vector_lengths[(vq - KVM_ARM64_SVE_VQ_MIN) / 64] >> 2575 ((vq - KVM_ARM64_SVE_VQ_MIN) % 64)) & 1)) 2576 /* Vector length vq * 16 bytes supported */ 2577 else 2578 /* Vector length vq * 16 bytes not supported */ 2579 2580.. [2] The maximum value vq for which the above condition is true is 2581 max_vq. This is the maximum vector length available to the guest on 2582 this vcpu, and determines which register slices are visible through 2583 this ioctl interface. 2584 2585(See Documentation/arm64/sve.rst for an explanation of the "vq" 2586nomenclature.) 2587 2588KVM_REG_ARM64_SVE_VLS is only accessible after KVM_ARM_VCPU_INIT. 2589KVM_ARM_VCPU_INIT initialises it to the best set of vector lengths that 2590the host supports. 2591 2592Userspace may subsequently modify it if desired until the vcpu's SVE 2593configuration is finalized using KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE). 2594 2595Apart from simply removing all vector lengths from the host set that 2596exceed some value, support for arbitrarily chosen sets of vector lengths 2597is hardware-dependent and may not be available. Attempting to configure 2598an invalid set of vector lengths via KVM_SET_ONE_REG will fail with 2599EINVAL. 2600 2601After the vcpu's SVE configuration is finalized, further attempts to 2602write this register will fail with EPERM. 2603 2604 2605MIPS registers are mapped using the lower 32 bits. The upper 16 of that is 2606the register group type: 2607 2608MIPS core registers (see above) have the following id bit patterns:: 2609 2610 0x7030 0000 0000 <reg:16> 2611 2612MIPS CP0 registers (see KVM_REG_MIPS_CP0_* above) have the following id bit 2613patterns depending on whether they're 32-bit or 64-bit registers:: 2614 2615 0x7020 0000 0001 00 <reg:5> <sel:3> (32-bit) 2616 0x7030 0000 0001 00 <reg:5> <sel:3> (64-bit) 2617 2618Note: KVM_REG_MIPS_CP0_ENTRYLO0 and KVM_REG_MIPS_CP0_ENTRYLO1 are the MIPS64 2619versions of the EntryLo registers regardless of the word size of the host 2620hardware, host kernel, guest, and whether XPA is present in the guest, i.e. 2621with the RI and XI bits (if they exist) in bits 63 and 62 respectively, and 2622the PFNX field starting at bit 30. 2623 2624MIPS MAARs (see KVM_REG_MIPS_CP0_MAAR(*) above) have the following id bit 2625patterns:: 2626 2627 0x7030 0000 0001 01 <reg:8> 2628 2629MIPS KVM control registers (see above) have the following id bit patterns:: 2630 2631 0x7030 0000 0002 <reg:16> 2632 2633MIPS FPU registers (see KVM_REG_MIPS_FPR_{32,64}() above) have the following 2634id bit patterns depending on the size of the register being accessed. They are 2635always accessed according to the current guest FPU mode (Status.FR and 2636Config5.FRE), i.e. as the guest would see them, and they become unpredictable 2637if the guest FPU mode is changed. MIPS SIMD Architecture (MSA) vector 2638registers (see KVM_REG_MIPS_VEC_128() above) have similar patterns as they 2639overlap the FPU registers:: 2640 2641 0x7020 0000 0003 00 <0:3> <reg:5> (32-bit FPU registers) 2642 0x7030 0000 0003 00 <0:3> <reg:5> (64-bit FPU registers) 2643 0x7040 0000 0003 00 <0:3> <reg:5> (128-bit MSA vector registers) 2644 2645MIPS FPU control registers (see KVM_REG_MIPS_FCR_{IR,CSR} above) have the 2646following id bit patterns:: 2647 2648 0x7020 0000 0003 01 <0:3> <reg:5> 2649 2650MIPS MSA control registers (see KVM_REG_MIPS_MSA_{IR,CSR} above) have the 2651following id bit patterns:: 2652 2653 0x7020 0000 0003 02 <0:3> <reg:5> 2654 2655RISC-V registers are mapped using the lower 32 bits. The upper 8 bits of 2656that is the register group type. 2657 2658RISC-V config registers are meant for configuring a Guest VCPU and it has 2659the following id bit patterns:: 2660 2661 0x8020 0000 01 <index into the kvm_riscv_config struct:24> (32bit Host) 2662 0x8030 0000 01 <index into the kvm_riscv_config struct:24> (64bit Host) 2663 2664Following are the RISC-V config registers: 2665 2666======================= ========= ============================================= 2667 Encoding Register Description 2668======================= ========= ============================================= 2669 0x80x0 0000 0100 0000 isa ISA feature bitmap of Guest VCPU 2670======================= ========= ============================================= 2671 2672The isa config register can be read anytime but can only be written before 2673a Guest VCPU runs. It will have ISA feature bits matching underlying host 2674set by default. 2675 2676RISC-V core registers represent the general excution state of a Guest VCPU 2677and it has the following id bit patterns:: 2678 2679 0x8020 0000 02 <index into the kvm_riscv_core struct:24> (32bit Host) 2680 0x8030 0000 02 <index into the kvm_riscv_core struct:24> (64bit Host) 2681 2682Following are the RISC-V core registers: 2683 2684======================= ========= ============================================= 2685 Encoding Register Description 2686======================= ========= ============================================= 2687 0x80x0 0000 0200 0000 regs.pc Program counter 2688 0x80x0 0000 0200 0001 regs.ra Return address 2689 0x80x0 0000 0200 0002 regs.sp Stack pointer 2690 0x80x0 0000 0200 0003 regs.gp Global pointer 2691 0x80x0 0000 0200 0004 regs.tp Task pointer 2692 0x80x0 0000 0200 0005 regs.t0 Caller saved register 0 2693 0x80x0 0000 0200 0006 regs.t1 Caller saved register 1 2694 0x80x0 0000 0200 0007 regs.t2 Caller saved register 2 2695 0x80x0 0000 0200 0008 regs.s0 Callee saved register 0 2696 0x80x0 0000 0200 0009 regs.s1 Callee saved register 1 2697 0x80x0 0000 0200 000a regs.a0 Function argument (or return value) 0 2698 0x80x0 0000 0200 000b regs.a1 Function argument (or return value) 1 2699 0x80x0 0000 0200 000c regs.a2 Function argument 2 2700 0x80x0 0000 0200 000d regs.a3 Function argument 3 2701 0x80x0 0000 0200 000e regs.a4 Function argument 4 2702 0x80x0 0000 0200 000f regs.a5 Function argument 5 2703 0x80x0 0000 0200 0010 regs.a6 Function argument 6 2704 0x80x0 0000 0200 0011 regs.a7 Function argument 7 2705 0x80x0 0000 0200 0012 regs.s2 Callee saved register 2 2706 0x80x0 0000 0200 0013 regs.s3 Callee saved register 3 2707 0x80x0 0000 0200 0014 regs.s4 Callee saved register 4 2708 0x80x0 0000 0200 0015 regs.s5 Callee saved register 5 2709 0x80x0 0000 0200 0016 regs.s6 Callee saved register 6 2710 0x80x0 0000 0200 0017 regs.s7 Callee saved register 7 2711 0x80x0 0000 0200 0018 regs.s8 Callee saved register 8 2712 0x80x0 0000 0200 0019 regs.s9 Callee saved register 9 2713 0x80x0 0000 0200 001a regs.s10 Callee saved register 10 2714 0x80x0 0000 0200 001b regs.s11 Callee saved register 11 2715 0x80x0 0000 0200 001c regs.t3 Caller saved register 3 2716 0x80x0 0000 0200 001d regs.t4 Caller saved register 4 2717 0x80x0 0000 0200 001e regs.t5 Caller saved register 5 2718 0x80x0 0000 0200 001f regs.t6 Caller saved register 6 2719 0x80x0 0000 0200 0020 mode Privilege mode (1 = S-mode or 0 = U-mode) 2720======================= ========= ============================================= 2721 2722RISC-V csr registers represent the supervisor mode control/status registers 2723of a Guest VCPU and it has the following id bit patterns:: 2724 2725 0x8020 0000 03 <index into the kvm_riscv_csr struct:24> (32bit Host) 2726 0x8030 0000 03 <index into the kvm_riscv_csr struct:24> (64bit Host) 2727 2728Following are the RISC-V csr registers: 2729 2730======================= ========= ============================================= 2731 Encoding Register Description 2732======================= ========= ============================================= 2733 0x80x0 0000 0300 0000 sstatus Supervisor status 2734 0x80x0 0000 0300 0001 sie Supervisor interrupt enable 2735 0x80x0 0000 0300 0002 stvec Supervisor trap vector base 2736 0x80x0 0000 0300 0003 sscratch Supervisor scratch register 2737 0x80x0 0000 0300 0004 sepc Supervisor exception program counter 2738 0x80x0 0000 0300 0005 scause Supervisor trap cause 2739 0x80x0 0000 0300 0006 stval Supervisor bad address or instruction 2740 0x80x0 0000 0300 0007 sip Supervisor interrupt pending 2741 0x80x0 0000 0300 0008 satp Supervisor address translation and protection 2742======================= ========= ============================================= 2743 2744RISC-V timer registers represent the timer state of a Guest VCPU and it has 2745the following id bit patterns:: 2746 2747 0x8030 0000 04 <index into the kvm_riscv_timer struct:24> 2748 2749Following are the RISC-V timer registers: 2750 2751======================= ========= ============================================= 2752 Encoding Register Description 2753======================= ========= ============================================= 2754 0x8030 0000 0400 0000 frequency Time base frequency (read-only) 2755 0x8030 0000 0400 0001 time Time value visible to Guest 2756 0x8030 0000 0400 0002 compare Time compare programmed by Guest 2757 0x8030 0000 0400 0003 state Time compare state (1 = ON or 0 = OFF) 2758======================= ========= ============================================= 2759 2760RISC-V F-extension registers represent the single precision floating point 2761state of a Guest VCPU and it has the following id bit patterns:: 2762 2763 0x8020 0000 05 <index into the __riscv_f_ext_state struct:24> 2764 2765Following are the RISC-V F-extension registers: 2766 2767======================= ========= ============================================= 2768 Encoding Register Description 2769======================= ========= ============================================= 2770 0x8020 0000 0500 0000 f[0] Floating point register 0 2771 ... 2772 0x8020 0000 0500 001f f[31] Floating point register 31 2773 0x8020 0000 0500 0020 fcsr Floating point control and status register 2774======================= ========= ============================================= 2775 2776RISC-V D-extension registers represent the double precision floating point 2777state of a Guest VCPU and it has the following id bit patterns:: 2778 2779 0x8020 0000 06 <index into the __riscv_d_ext_state struct:24> (fcsr) 2780 0x8030 0000 06 <index into the __riscv_d_ext_state struct:24> (non-fcsr) 2781 2782Following are the RISC-V D-extension registers: 2783 2784======================= ========= ============================================= 2785 Encoding Register Description 2786======================= ========= ============================================= 2787 0x8030 0000 0600 0000 f[0] Floating point register 0 2788 ... 2789 0x8030 0000 0600 001f f[31] Floating point register 31 2790 0x8020 0000 0600 0020 fcsr Floating point control and status register 2791======================= ========= ============================================= 2792 2793 27944.69 KVM_GET_ONE_REG 2795-------------------- 2796 2797:Capability: KVM_CAP_ONE_REG 2798:Architectures: all 2799:Type: vcpu ioctl 2800:Parameters: struct kvm_one_reg (in and out) 2801:Returns: 0 on success, negative value on failure 2802 2803Errors include: 2804 2805 ======== ============================================================ 2806 ENOENT no such register 2807 EINVAL invalid register ID, or no such register or used with VMs in 2808 protected virtualization mode on s390 2809 EPERM (arm64) register access not allowed before vcpu finalization 2810 ======== ============================================================ 2811 2812(These error codes are indicative only: do not rely on a specific error 2813code being returned in a specific situation.) 2814 2815This ioctl allows to receive the value of a single register implemented 2816in a vcpu. The register to read is indicated by the "id" field of the 2817kvm_one_reg struct passed in. On success, the register value can be found 2818at the memory location pointed to by "addr". 2819 2820The list of registers accessible using this interface is identical to the 2821list in 4.68. 2822 2823 28244.70 KVM_KVMCLOCK_CTRL 2825---------------------- 2826 2827:Capability: KVM_CAP_KVMCLOCK_CTRL 2828:Architectures: Any that implement pvclocks (currently x86 only) 2829:Type: vcpu ioctl 2830:Parameters: None 2831:Returns: 0 on success, -1 on error 2832 2833This ioctl sets a flag accessible to the guest indicating that the specified 2834vCPU has been paused by the host userspace. 2835 2836The host will set a flag in the pvclock structure that is checked from the 2837soft lockup watchdog. The flag is part of the pvclock structure that is 2838shared between guest and host, specifically the second bit of the flags 2839field of the pvclock_vcpu_time_info structure. It will be set exclusively by 2840the host and read/cleared exclusively by the guest. The guest operation of 2841checking and clearing the flag must be an atomic operation so 2842load-link/store-conditional, or equivalent must be used. There are two cases 2843where the guest will clear the flag: when the soft lockup watchdog timer resets 2844itself or when a soft lockup is detected. This ioctl can be called any time 2845after pausing the vcpu, but before it is resumed. 2846 2847 28484.71 KVM_SIGNAL_MSI 2849------------------- 2850 2851:Capability: KVM_CAP_SIGNAL_MSI 2852:Architectures: x86 arm64 2853:Type: vm ioctl 2854:Parameters: struct kvm_msi (in) 2855:Returns: >0 on delivery, 0 if guest blocked the MSI, and -1 on error 2856 2857Directly inject a MSI message. Only valid with in-kernel irqchip that handles 2858MSI messages. 2859 2860:: 2861 2862 struct kvm_msi { 2863 __u32 address_lo; 2864 __u32 address_hi; 2865 __u32 data; 2866 __u32 flags; 2867 __u32 devid; 2868 __u8 pad[12]; 2869 }; 2870 2871flags: 2872 KVM_MSI_VALID_DEVID: devid contains a valid value. The per-VM 2873 KVM_CAP_MSI_DEVID capability advertises the requirement to provide 2874 the device ID. If this capability is not available, userspace 2875 should never set the KVM_MSI_VALID_DEVID flag as the ioctl might fail. 2876 2877If KVM_MSI_VALID_DEVID is set, devid contains a unique device identifier 2878for the device that wrote the MSI message. For PCI, this is usually a 2879BFD identifier in the lower 16 bits. 2880 2881On x86, address_hi is ignored unless the KVM_X2APIC_API_USE_32BIT_IDS 2882feature of KVM_CAP_X2APIC_API capability is enabled. If it is enabled, 2883address_hi bits 31-8 provide bits 31-8 of the destination id. Bits 7-0 of 2884address_hi must be zero. 2885 2886 28874.71 KVM_CREATE_PIT2 2888-------------------- 2889 2890:Capability: KVM_CAP_PIT2 2891:Architectures: x86 2892:Type: vm ioctl 2893:Parameters: struct kvm_pit_config (in) 2894:Returns: 0 on success, -1 on error 2895 2896Creates an in-kernel device model for the i8254 PIT. This call is only valid 2897after enabling in-kernel irqchip support via KVM_CREATE_IRQCHIP. The following 2898parameters have to be passed:: 2899 2900 struct kvm_pit_config { 2901 __u32 flags; 2902 __u32 pad[15]; 2903 }; 2904 2905Valid flags are:: 2906 2907 #define KVM_PIT_SPEAKER_DUMMY 1 /* emulate speaker port stub */ 2908 2909PIT timer interrupts may use a per-VM kernel thread for injection. If it 2910exists, this thread will have a name of the following pattern:: 2911 2912 kvm-pit/<owner-process-pid> 2913 2914When running a guest with elevated priorities, the scheduling parameters of 2915this thread may have to be adjusted accordingly. 2916 2917This IOCTL replaces the obsolete KVM_CREATE_PIT. 2918 2919 29204.72 KVM_GET_PIT2 2921----------------- 2922 2923:Capability: KVM_CAP_PIT_STATE2 2924:Architectures: x86 2925:Type: vm ioctl 2926:Parameters: struct kvm_pit_state2 (out) 2927:Returns: 0 on success, -1 on error 2928 2929Retrieves the state of the in-kernel PIT model. Only valid after 2930KVM_CREATE_PIT2. The state is returned in the following structure:: 2931 2932 struct kvm_pit_state2 { 2933 struct kvm_pit_channel_state channels[3]; 2934 __u32 flags; 2935 __u32 reserved[9]; 2936 }; 2937 2938Valid flags are:: 2939 2940 /* disable PIT in HPET legacy mode */ 2941 #define KVM_PIT_FLAGS_HPET_LEGACY 0x00000001 2942 2943This IOCTL replaces the obsolete KVM_GET_PIT. 2944 2945 29464.73 KVM_SET_PIT2 2947----------------- 2948 2949:Capability: KVM_CAP_PIT_STATE2 2950:Architectures: x86 2951:Type: vm ioctl 2952:Parameters: struct kvm_pit_state2 (in) 2953:Returns: 0 on success, -1 on error 2954 2955Sets the state of the in-kernel PIT model. Only valid after KVM_CREATE_PIT2. 2956See KVM_GET_PIT2 for details on struct kvm_pit_state2. 2957 2958This IOCTL replaces the obsolete KVM_SET_PIT. 2959 2960 29614.74 KVM_PPC_GET_SMMU_INFO 2962-------------------------- 2963 2964:Capability: KVM_CAP_PPC_GET_SMMU_INFO 2965:Architectures: powerpc 2966:Type: vm ioctl 2967:Parameters: None 2968:Returns: 0 on success, -1 on error 2969 2970This populates and returns a structure describing the features of 2971the "Server" class MMU emulation supported by KVM. 2972This can in turn be used by userspace to generate the appropriate 2973device-tree properties for the guest operating system. 2974 2975The structure contains some global information, followed by an 2976array of supported segment page sizes:: 2977 2978 struct kvm_ppc_smmu_info { 2979 __u64 flags; 2980 __u32 slb_size; 2981 __u32 pad; 2982 struct kvm_ppc_one_seg_page_size sps[KVM_PPC_PAGE_SIZES_MAX_SZ]; 2983 }; 2984 2985The supported flags are: 2986 2987 - KVM_PPC_PAGE_SIZES_REAL: 2988 When that flag is set, guest page sizes must "fit" the backing 2989 store page sizes. When not set, any page size in the list can 2990 be used regardless of how they are backed by userspace. 2991 2992 - KVM_PPC_1T_SEGMENTS 2993 The emulated MMU supports 1T segments in addition to the 2994 standard 256M ones. 2995 2996 - KVM_PPC_NO_HASH 2997 This flag indicates that HPT guests are not supported by KVM, 2998 thus all guests must use radix MMU mode. 2999 3000The "slb_size" field indicates how many SLB entries are supported 3001 3002The "sps" array contains 8 entries indicating the supported base 3003page sizes for a segment in increasing order. Each entry is defined 3004as follow:: 3005 3006 struct kvm_ppc_one_seg_page_size { 3007 __u32 page_shift; /* Base page shift of segment (or 0) */ 3008 __u32 slb_enc; /* SLB encoding for BookS */ 3009 struct kvm_ppc_one_page_size enc[KVM_PPC_PAGE_SIZES_MAX_SZ]; 3010 }; 3011 3012An entry with a "page_shift" of 0 is unused. Because the array is 3013organized in increasing order, a lookup can stop when encoutering 3014such an entry. 3015 3016The "slb_enc" field provides the encoding to use in the SLB for the 3017page size. The bits are in positions such as the value can directly 3018be OR'ed into the "vsid" argument of the slbmte instruction. 3019 3020The "enc" array is a list which for each of those segment base page 3021size provides the list of supported actual page sizes (which can be 3022only larger or equal to the base page size), along with the 3023corresponding encoding in the hash PTE. Similarly, the array is 30248 entries sorted by increasing sizes and an entry with a "0" shift 3025is an empty entry and a terminator:: 3026 3027 struct kvm_ppc_one_page_size { 3028 __u32 page_shift; /* Page shift (or 0) */ 3029 __u32 pte_enc; /* Encoding in the HPTE (>>12) */ 3030 }; 3031 3032The "pte_enc" field provides a value that can OR'ed into the hash 3033PTE's RPN field (ie, it needs to be shifted left by 12 to OR it 3034into the hash PTE second double word). 3035 30364.75 KVM_IRQFD 3037-------------- 3038 3039:Capability: KVM_CAP_IRQFD 3040:Architectures: x86 s390 arm64 3041:Type: vm ioctl 3042:Parameters: struct kvm_irqfd (in) 3043:Returns: 0 on success, -1 on error 3044 3045Allows setting an eventfd to directly trigger a guest interrupt. 3046kvm_irqfd.fd specifies the file descriptor to use as the eventfd and 3047kvm_irqfd.gsi specifies the irqchip pin toggled by this event. When 3048an event is triggered on the eventfd, an interrupt is injected into 3049the guest using the specified gsi pin. The irqfd is removed using 3050the KVM_IRQFD_FLAG_DEASSIGN flag, specifying both kvm_irqfd.fd 3051and kvm_irqfd.gsi. 3052 3053With KVM_CAP_IRQFD_RESAMPLE, KVM_IRQFD supports a de-assert and notify 3054mechanism allowing emulation of level-triggered, irqfd-based 3055interrupts. When KVM_IRQFD_FLAG_RESAMPLE is set the user must pass an 3056additional eventfd in the kvm_irqfd.resamplefd field. When operating 3057in resample mode, posting of an interrupt through kvm_irq.fd asserts 3058the specified gsi in the irqchip. When the irqchip is resampled, such 3059as from an EOI, the gsi is de-asserted and the user is notified via 3060kvm_irqfd.resamplefd. It is the user's responsibility to re-queue 3061the interrupt if the device making use of it still requires service. 3062Note that closing the resamplefd is not sufficient to disable the 3063irqfd. The KVM_IRQFD_FLAG_RESAMPLE is only necessary on assignment 3064and need not be specified with KVM_IRQFD_FLAG_DEASSIGN. 3065 3066On arm64, gsi routing being supported, the following can happen: 3067 3068- in case no routing entry is associated to this gsi, injection fails 3069- in case the gsi is associated to an irqchip routing entry, 3070 irqchip.pin + 32 corresponds to the injected SPI ID. 3071- in case the gsi is associated to an MSI routing entry, the MSI 3072 message and device ID are translated into an LPI (support restricted 3073 to GICv3 ITS in-kernel emulation). 3074 30754.76 KVM_PPC_ALLOCATE_HTAB 3076-------------------------- 3077 3078:Capability: KVM_CAP_PPC_ALLOC_HTAB 3079:Architectures: powerpc 3080:Type: vm ioctl 3081:Parameters: Pointer to u32 containing hash table order (in/out) 3082:Returns: 0 on success, -1 on error 3083 3084This requests the host kernel to allocate an MMU hash table for a 3085guest using the PAPR paravirtualization interface. This only does 3086anything if the kernel is configured to use the Book 3S HV style of 3087virtualization. Otherwise the capability doesn't exist and the ioctl 3088returns an ENOTTY error. The rest of this description assumes Book 3S 3089HV. 3090 3091There must be no vcpus running when this ioctl is called; if there 3092are, it will do nothing and return an EBUSY error. 3093 3094The parameter is a pointer to a 32-bit unsigned integer variable 3095containing the order (log base 2) of the desired size of the hash 3096table, which must be between 18 and 46. On successful return from the 3097ioctl, the value will not be changed by the kernel. 3098 3099If no hash table has been allocated when any vcpu is asked to run 3100(with the KVM_RUN ioctl), the host kernel will allocate a 3101default-sized hash table (16 MB). 3102 3103If this ioctl is called when a hash table has already been allocated, 3104with a different order from the existing hash table, the existing hash 3105table will be freed and a new one allocated. If this is ioctl is 3106called when a hash table has already been allocated of the same order 3107as specified, the kernel will clear out the existing hash table (zero 3108all HPTEs). In either case, if the guest is using the virtualized 3109real-mode area (VRMA) facility, the kernel will re-create the VMRA 3110HPTEs on the next KVM_RUN of any vcpu. 3111 31124.77 KVM_S390_INTERRUPT 3113----------------------- 3114 3115:Capability: basic 3116:Architectures: s390 3117:Type: vm ioctl, vcpu ioctl 3118:Parameters: struct kvm_s390_interrupt (in) 3119:Returns: 0 on success, -1 on error 3120 3121Allows to inject an interrupt to the guest. Interrupts can be floating 3122(vm ioctl) or per cpu (vcpu ioctl), depending on the interrupt type. 3123 3124Interrupt parameters are passed via kvm_s390_interrupt:: 3125 3126 struct kvm_s390_interrupt { 3127 __u32 type; 3128 __u32 parm; 3129 __u64 parm64; 3130 }; 3131 3132type can be one of the following: 3133 3134KVM_S390_SIGP_STOP (vcpu) 3135 - sigp stop; optional flags in parm 3136KVM_S390_PROGRAM_INT (vcpu) 3137 - program check; code in parm 3138KVM_S390_SIGP_SET_PREFIX (vcpu) 3139 - sigp set prefix; prefix address in parm 3140KVM_S390_RESTART (vcpu) 3141 - restart 3142KVM_S390_INT_CLOCK_COMP (vcpu) 3143 - clock comparator interrupt 3144KVM_S390_INT_CPU_TIMER (vcpu) 3145 - CPU timer interrupt 3146KVM_S390_INT_VIRTIO (vm) 3147 - virtio external interrupt; external interrupt 3148 parameters in parm and parm64 3149KVM_S390_INT_SERVICE (vm) 3150 - sclp external interrupt; sclp parameter in parm 3151KVM_S390_INT_EMERGENCY (vcpu) 3152 - sigp emergency; source cpu in parm 3153KVM_S390_INT_EXTERNAL_CALL (vcpu) 3154 - sigp external call; source cpu in parm 3155KVM_S390_INT_IO(ai,cssid,ssid,schid) (vm) 3156 - compound value to indicate an 3157 I/O interrupt (ai - adapter interrupt; cssid,ssid,schid - subchannel); 3158 I/O interruption parameters in parm (subchannel) and parm64 (intparm, 3159 interruption subclass) 3160KVM_S390_MCHK (vm, vcpu) 3161 - machine check interrupt; cr 14 bits in parm, machine check interrupt 3162 code in parm64 (note that machine checks needing further payload are not 3163 supported by this ioctl) 3164 3165This is an asynchronous vcpu ioctl and can be invoked from any thread. 3166 31674.78 KVM_PPC_GET_HTAB_FD 3168------------------------ 3169 3170:Capability: KVM_CAP_PPC_HTAB_FD 3171:Architectures: powerpc 3172:Type: vm ioctl 3173:Parameters: Pointer to struct kvm_get_htab_fd (in) 3174:Returns: file descriptor number (>= 0) on success, -1 on error 3175 3176This returns a file descriptor that can be used either to read out the 3177entries in the guest's hashed page table (HPT), or to write entries to 3178initialize the HPT. The returned fd can only be written to if the 3179KVM_GET_HTAB_WRITE bit is set in the flags field of the argument, and 3180can only be read if that bit is clear. The argument struct looks like 3181this:: 3182 3183 /* For KVM_PPC_GET_HTAB_FD */ 3184 struct kvm_get_htab_fd { 3185 __u64 flags; 3186 __u64 start_index; 3187 __u64 reserved[2]; 3188 }; 3189 3190 /* Values for kvm_get_htab_fd.flags */ 3191 #define KVM_GET_HTAB_BOLTED_ONLY ((__u64)0x1) 3192 #define KVM_GET_HTAB_WRITE ((__u64)0x2) 3193 3194The 'start_index' field gives the index in the HPT of the entry at 3195which to start reading. It is ignored when writing. 3196 3197Reads on the fd will initially supply information about all 3198"interesting" HPT entries. Interesting entries are those with the 3199bolted bit set, if the KVM_GET_HTAB_BOLTED_ONLY bit is set, otherwise 3200all entries. When the end of the HPT is reached, the read() will 3201return. If read() is called again on the fd, it will start again from 3202the beginning of the HPT, but will only return HPT entries that have 3203changed since they were last read. 3204 3205Data read or written is structured as a header (8 bytes) followed by a 3206series of valid HPT entries (16 bytes) each. The header indicates how 3207many valid HPT entries there are and how many invalid entries follow 3208the valid entries. The invalid entries are not represented explicitly 3209in the stream. The header format is:: 3210 3211 struct kvm_get_htab_header { 3212 __u32 index; 3213 __u16 n_valid; 3214 __u16 n_invalid; 3215 }; 3216 3217Writes to the fd create HPT entries starting at the index given in the 3218header; first 'n_valid' valid entries with contents from the data 3219written, then 'n_invalid' invalid entries, invalidating any previously 3220valid entries found. 3221 32224.79 KVM_CREATE_DEVICE 3223---------------------- 3224 3225:Capability: KVM_CAP_DEVICE_CTRL 3226:Type: vm ioctl 3227:Parameters: struct kvm_create_device (in/out) 3228:Returns: 0 on success, -1 on error 3229 3230Errors: 3231 3232 ====== ======================================================= 3233 ENODEV The device type is unknown or unsupported 3234 EEXIST Device already created, and this type of device may not 3235 be instantiated multiple times 3236 ====== ======================================================= 3237 3238 Other error conditions may be defined by individual device types or 3239 have their standard meanings. 3240 3241Creates an emulated device in the kernel. The file descriptor returned 3242in fd can be used with KVM_SET/GET/HAS_DEVICE_ATTR. 3243 3244If the KVM_CREATE_DEVICE_TEST flag is set, only test whether the 3245device type is supported (not necessarily whether it can be created 3246in the current vm). 3247 3248Individual devices should not define flags. Attributes should be used 3249for specifying any behavior that is not implied by the device type 3250number. 3251 3252:: 3253 3254 struct kvm_create_device { 3255 __u32 type; /* in: KVM_DEV_TYPE_xxx */ 3256 __u32 fd; /* out: device handle */ 3257 __u32 flags; /* in: KVM_CREATE_DEVICE_xxx */ 3258 }; 3259 32604.80 KVM_SET_DEVICE_ATTR/KVM_GET_DEVICE_ATTR 3261-------------------------------------------- 3262 3263:Capability: KVM_CAP_DEVICE_CTRL, KVM_CAP_VM_ATTRIBUTES for vm device, 3264 KVM_CAP_VCPU_ATTRIBUTES for vcpu device 3265 KVM_CAP_SYS_ATTRIBUTES for system (/dev/kvm) device (no set) 3266:Type: device ioctl, vm ioctl, vcpu ioctl 3267:Parameters: struct kvm_device_attr 3268:Returns: 0 on success, -1 on error 3269 3270Errors: 3271 3272 ===== ============================================================= 3273 ENXIO The group or attribute is unknown/unsupported for this device 3274 or hardware support is missing. 3275 EPERM The attribute cannot (currently) be accessed this way 3276 (e.g. read-only attribute, or attribute that only makes 3277 sense when the device is in a different state) 3278 ===== ============================================================= 3279 3280 Other error conditions may be defined by individual device types. 3281 3282Gets/sets a specified piece of device configuration and/or state. The 3283semantics are device-specific. See individual device documentation in 3284the "devices" directory. As with ONE_REG, the size of the data 3285transferred is defined by the particular attribute. 3286 3287:: 3288 3289 struct kvm_device_attr { 3290 __u32 flags; /* no flags currently defined */ 3291 __u32 group; /* device-defined */ 3292 __u64 attr; /* group-defined */ 3293 __u64 addr; /* userspace address of attr data */ 3294 }; 3295 32964.81 KVM_HAS_DEVICE_ATTR 3297------------------------ 3298 3299:Capability: KVM_CAP_DEVICE_CTRL, KVM_CAP_VM_ATTRIBUTES for vm device, 3300 KVM_CAP_VCPU_ATTRIBUTES for vcpu device 3301 KVM_CAP_SYS_ATTRIBUTES for system (/dev/kvm) device 3302:Type: device ioctl, vm ioctl, vcpu ioctl 3303:Parameters: struct kvm_device_attr 3304:Returns: 0 on success, -1 on error 3305 3306Errors: 3307 3308 ===== ============================================================= 3309 ENXIO The group or attribute is unknown/unsupported for this device 3310 or hardware support is missing. 3311 ===== ============================================================= 3312 3313Tests whether a device supports a particular attribute. A successful 3314return indicates the attribute is implemented. It does not necessarily 3315indicate that the attribute can be read or written in the device's 3316current state. "addr" is ignored. 3317 33184.82 KVM_ARM_VCPU_INIT 3319---------------------- 3320 3321:Capability: basic 3322:Architectures: arm64 3323:Type: vcpu ioctl 3324:Parameters: struct kvm_vcpu_init (in) 3325:Returns: 0 on success; -1 on error 3326 3327Errors: 3328 3329 ====== ================================================================= 3330 EINVAL the target is unknown, or the combination of features is invalid. 3331 ENOENT a features bit specified is unknown. 3332 ====== ================================================================= 3333 3334This tells KVM what type of CPU to present to the guest, and what 3335optional features it should have. This will cause a reset of the cpu 3336registers to their initial values. If this is not called, KVM_RUN will 3337return ENOEXEC for that vcpu. 3338 3339The initial values are defined as: 3340 - Processor state: 3341 * AArch64: EL1h, D, A, I and F bits set. All other bits 3342 are cleared. 3343 * AArch32: SVC, A, I and F bits set. All other bits are 3344 cleared. 3345 - General Purpose registers, including PC and SP: set to 0 3346 - FPSIMD/NEON registers: set to 0 3347 - SVE registers: set to 0 3348 - System registers: Reset to their architecturally defined 3349 values as for a warm reset to EL1 (resp. SVC) 3350 3351Note that because some registers reflect machine topology, all vcpus 3352should be created before this ioctl is invoked. 3353 3354Userspace can call this function multiple times for a given vcpu, including 3355after the vcpu has been run. This will reset the vcpu to its initial 3356state. All calls to this function after the initial call must use the same 3357target and same set of feature flags, otherwise EINVAL will be returned. 3358 3359Possible features: 3360 3361 - KVM_ARM_VCPU_POWER_OFF: Starts the CPU in a power-off state. 3362 Depends on KVM_CAP_ARM_PSCI. If not set, the CPU will be powered on 3363 and execute guest code when KVM_RUN is called. 3364 - KVM_ARM_VCPU_EL1_32BIT: Starts the CPU in a 32bit mode. 3365 Depends on KVM_CAP_ARM_EL1_32BIT (arm64 only). 3366 - KVM_ARM_VCPU_PSCI_0_2: Emulate PSCI v0.2 (or a future revision 3367 backward compatible with v0.2) for the CPU. 3368 Depends on KVM_CAP_ARM_PSCI_0_2. 3369 - KVM_ARM_VCPU_PMU_V3: Emulate PMUv3 for the CPU. 3370 Depends on KVM_CAP_ARM_PMU_V3. 3371 3372 - KVM_ARM_VCPU_PTRAUTH_ADDRESS: Enables Address Pointer authentication 3373 for arm64 only. 3374 Depends on KVM_CAP_ARM_PTRAUTH_ADDRESS. 3375 If KVM_CAP_ARM_PTRAUTH_ADDRESS and KVM_CAP_ARM_PTRAUTH_GENERIC are 3376 both present, then both KVM_ARM_VCPU_PTRAUTH_ADDRESS and 3377 KVM_ARM_VCPU_PTRAUTH_GENERIC must be requested or neither must be 3378 requested. 3379 3380 - KVM_ARM_VCPU_PTRAUTH_GENERIC: Enables Generic Pointer authentication 3381 for arm64 only. 3382 Depends on KVM_CAP_ARM_PTRAUTH_GENERIC. 3383 If KVM_CAP_ARM_PTRAUTH_ADDRESS and KVM_CAP_ARM_PTRAUTH_GENERIC are 3384 both present, then both KVM_ARM_VCPU_PTRAUTH_ADDRESS and 3385 KVM_ARM_VCPU_PTRAUTH_GENERIC must be requested or neither must be 3386 requested. 3387 3388 - KVM_ARM_VCPU_SVE: Enables SVE for the CPU (arm64 only). 3389 Depends on KVM_CAP_ARM_SVE. 3390 Requires KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE): 3391 3392 * After KVM_ARM_VCPU_INIT: 3393 3394 - KVM_REG_ARM64_SVE_VLS may be read using KVM_GET_ONE_REG: the 3395 initial value of this pseudo-register indicates the best set of 3396 vector lengths possible for a vcpu on this host. 3397 3398 * Before KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE): 3399 3400 - KVM_RUN and KVM_GET_REG_LIST are not available; 3401 3402 - KVM_GET_ONE_REG and KVM_SET_ONE_REG cannot be used to access 3403 the scalable archietctural SVE registers 3404 KVM_REG_ARM64_SVE_ZREG(), KVM_REG_ARM64_SVE_PREG() or 3405 KVM_REG_ARM64_SVE_FFR; 3406 3407 - KVM_REG_ARM64_SVE_VLS may optionally be written using 3408 KVM_SET_ONE_REG, to modify the set of vector lengths available 3409 for the vcpu. 3410 3411 * After KVM_ARM_VCPU_FINALIZE(KVM_ARM_VCPU_SVE): 3412 3413 - the KVM_REG_ARM64_SVE_VLS pseudo-register is immutable, and can 3414 no longer be written using KVM_SET_ONE_REG. 3415 34164.83 KVM_ARM_PREFERRED_TARGET 3417----------------------------- 3418 3419:Capability: basic 3420:Architectures: arm64 3421:Type: vm ioctl 3422:Parameters: struct kvm_vcpu_init (out) 3423:Returns: 0 on success; -1 on error 3424 3425Errors: 3426 3427 ====== ========================================== 3428 ENODEV no preferred target available for the host 3429 ====== ========================================== 3430 3431This queries KVM for preferred CPU target type which can be emulated 3432by KVM on underlying host. 3433 3434The ioctl returns struct kvm_vcpu_init instance containing information 3435about preferred CPU target type and recommended features for it. The 3436kvm_vcpu_init->features bitmap returned will have feature bits set if 3437the preferred target recommends setting these features, but this is 3438not mandatory. 3439 3440The information returned by this ioctl can be used to prepare an instance 3441of struct kvm_vcpu_init for KVM_ARM_VCPU_INIT ioctl which will result in 3442VCPU matching underlying host. 3443 3444 34454.84 KVM_GET_REG_LIST 3446--------------------- 3447 3448:Capability: basic 3449:Architectures: arm64, mips 3450:Type: vcpu ioctl 3451:Parameters: struct kvm_reg_list (in/out) 3452:Returns: 0 on success; -1 on error 3453 3454Errors: 3455 3456 ===== ============================================================== 3457 E2BIG the reg index list is too big to fit in the array specified by 3458 the user (the number required will be written into n). 3459 ===== ============================================================== 3460 3461:: 3462 3463 struct kvm_reg_list { 3464 __u64 n; /* number of registers in reg[] */ 3465 __u64 reg[0]; 3466 }; 3467 3468This ioctl returns the guest registers that are supported for the 3469KVM_GET_ONE_REG/KVM_SET_ONE_REG calls. 3470 3471 34724.85 KVM_ARM_SET_DEVICE_ADDR (deprecated) 3473----------------------------------------- 3474 3475:Capability: KVM_CAP_ARM_SET_DEVICE_ADDR 3476:Architectures: arm64 3477:Type: vm ioctl 3478:Parameters: struct kvm_arm_device_address (in) 3479:Returns: 0 on success, -1 on error 3480 3481Errors: 3482 3483 ====== ============================================ 3484 ENODEV The device id is unknown 3485 ENXIO Device not supported on current system 3486 EEXIST Address already set 3487 E2BIG Address outside guest physical address space 3488 EBUSY Address overlaps with other device range 3489 ====== ============================================ 3490 3491:: 3492 3493 struct kvm_arm_device_addr { 3494 __u64 id; 3495 __u64 addr; 3496 }; 3497 3498Specify a device address in the guest's physical address space where guests 3499can access emulated or directly exposed devices, which the host kernel needs 3500to know about. The id field is an architecture specific identifier for a 3501specific device. 3502 3503arm64 divides the id field into two parts, a device id and an 3504address type id specific to the individual device:: 3505 3506 bits: | 63 ... 32 | 31 ... 16 | 15 ... 0 | 3507 field: | 0x00000000 | device id | addr type id | 3508 3509arm64 currently only require this when using the in-kernel GIC 3510support for the hardware VGIC features, using KVM_ARM_DEVICE_VGIC_V2 3511as the device id. When setting the base address for the guest's 3512mapping of the VGIC virtual CPU and distributor interface, the ioctl 3513must be called after calling KVM_CREATE_IRQCHIP, but before calling 3514KVM_RUN on any of the VCPUs. Calling this ioctl twice for any of the 3515base addresses will return -EEXIST. 3516 3517Note, this IOCTL is deprecated and the more flexible SET/GET_DEVICE_ATTR API 3518should be used instead. 3519 3520 35214.86 KVM_PPC_RTAS_DEFINE_TOKEN 3522------------------------------ 3523 3524:Capability: KVM_CAP_PPC_RTAS 3525:Architectures: ppc 3526:Type: vm ioctl 3527:Parameters: struct kvm_rtas_token_args 3528:Returns: 0 on success, -1 on error 3529 3530Defines a token value for a RTAS (Run Time Abstraction Services) 3531service in order to allow it to be handled in the kernel. The 3532argument struct gives the name of the service, which must be the name 3533of a service that has a kernel-side implementation. If the token 3534value is non-zero, it will be associated with that service, and 3535subsequent RTAS calls by the guest specifying that token will be 3536handled by the kernel. If the token value is 0, then any token 3537associated with the service will be forgotten, and subsequent RTAS 3538calls by the guest for that service will be passed to userspace to be 3539handled. 3540 35414.87 KVM_SET_GUEST_DEBUG 3542------------------------ 3543 3544:Capability: KVM_CAP_SET_GUEST_DEBUG 3545:Architectures: x86, s390, ppc, arm64 3546:Type: vcpu ioctl 3547:Parameters: struct kvm_guest_debug (in) 3548:Returns: 0 on success; -1 on error 3549 3550:: 3551 3552 struct kvm_guest_debug { 3553 __u32 control; 3554 __u32 pad; 3555 struct kvm_guest_debug_arch arch; 3556 }; 3557 3558Set up the processor specific debug registers and configure vcpu for 3559handling guest debug events. There are two parts to the structure, the 3560first a control bitfield indicates the type of debug events to handle 3561when running. Common control bits are: 3562 3563 - KVM_GUESTDBG_ENABLE: guest debugging is enabled 3564 - KVM_GUESTDBG_SINGLESTEP: the next run should single-step 3565 3566The top 16 bits of the control field are architecture specific control 3567flags which can include the following: 3568 3569 - KVM_GUESTDBG_USE_SW_BP: using software breakpoints [x86, arm64] 3570 - KVM_GUESTDBG_USE_HW_BP: using hardware breakpoints [x86, s390] 3571 - KVM_GUESTDBG_USE_HW: using hardware debug events [arm64] 3572 - KVM_GUESTDBG_INJECT_DB: inject DB type exception [x86] 3573 - KVM_GUESTDBG_INJECT_BP: inject BP type exception [x86] 3574 - KVM_GUESTDBG_EXIT_PENDING: trigger an immediate guest exit [s390] 3575 - KVM_GUESTDBG_BLOCKIRQ: avoid injecting interrupts/NMI/SMI [x86] 3576 3577For example KVM_GUESTDBG_USE_SW_BP indicates that software breakpoints 3578are enabled in memory so we need to ensure breakpoint exceptions are 3579correctly trapped and the KVM run loop exits at the breakpoint and not 3580running off into the normal guest vector. For KVM_GUESTDBG_USE_HW_BP 3581we need to ensure the guest vCPUs architecture specific registers are 3582updated to the correct (supplied) values. 3583 3584The second part of the structure is architecture specific and 3585typically contains a set of debug registers. 3586 3587For arm64 the number of debug registers is implementation defined and 3588can be determined by querying the KVM_CAP_GUEST_DEBUG_HW_BPS and 3589KVM_CAP_GUEST_DEBUG_HW_WPS capabilities which return a positive number 3590indicating the number of supported registers. 3591 3592For ppc, the KVM_CAP_PPC_GUEST_DEBUG_SSTEP capability indicates whether 3593the single-step debug event (KVM_GUESTDBG_SINGLESTEP) is supported. 3594 3595Also when supported, KVM_CAP_SET_GUEST_DEBUG2 capability indicates the 3596supported KVM_GUESTDBG_* bits in the control field. 3597 3598When debug events exit the main run loop with the reason 3599KVM_EXIT_DEBUG with the kvm_debug_exit_arch part of the kvm_run 3600structure containing architecture specific debug information. 3601 36024.88 KVM_GET_EMULATED_CPUID 3603--------------------------- 3604 3605:Capability: KVM_CAP_EXT_EMUL_CPUID 3606:Architectures: x86 3607:Type: system ioctl 3608:Parameters: struct kvm_cpuid2 (in/out) 3609:Returns: 0 on success, -1 on error 3610 3611:: 3612 3613 struct kvm_cpuid2 { 3614 __u32 nent; 3615 __u32 flags; 3616 struct kvm_cpuid_entry2 entries[0]; 3617 }; 3618 3619The member 'flags' is used for passing flags from userspace. 3620 3621:: 3622 3623 #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX BIT(0) 3624 #define KVM_CPUID_FLAG_STATEFUL_FUNC BIT(1) /* deprecated */ 3625 #define KVM_CPUID_FLAG_STATE_READ_NEXT BIT(2) /* deprecated */ 3626 3627 struct kvm_cpuid_entry2 { 3628 __u32 function; 3629 __u32 index; 3630 __u32 flags; 3631 __u32 eax; 3632 __u32 ebx; 3633 __u32 ecx; 3634 __u32 edx; 3635 __u32 padding[3]; 3636 }; 3637 3638This ioctl returns x86 cpuid features which are emulated by 3639kvm.Userspace can use the information returned by this ioctl to query 3640which features are emulated by kvm instead of being present natively. 3641 3642Userspace invokes KVM_GET_EMULATED_CPUID by passing a kvm_cpuid2 3643structure with the 'nent' field indicating the number of entries in 3644the variable-size array 'entries'. If the number of entries is too low 3645to describe the cpu capabilities, an error (E2BIG) is returned. If the 3646number is too high, the 'nent' field is adjusted and an error (ENOMEM) 3647is returned. If the number is just right, the 'nent' field is adjusted 3648to the number of valid entries in the 'entries' array, which is then 3649filled. 3650 3651The entries returned are the set CPUID bits of the respective features 3652which kvm emulates, as returned by the CPUID instruction, with unknown 3653or unsupported feature bits cleared. 3654 3655Features like x2apic, for example, may not be present in the host cpu 3656but are exposed by kvm in KVM_GET_SUPPORTED_CPUID because they can be 3657emulated efficiently and thus not included here. 3658 3659The fields in each entry are defined as follows: 3660 3661 function: 3662 the eax value used to obtain the entry 3663 index: 3664 the ecx value used to obtain the entry (for entries that are 3665 affected by ecx) 3666 flags: 3667 an OR of zero or more of the following: 3668 3669 KVM_CPUID_FLAG_SIGNIFCANT_INDEX: 3670 if the index field is valid 3671 3672 eax, ebx, ecx, edx: 3673 3674 the values returned by the cpuid instruction for 3675 this function/index combination 3676 36774.89 KVM_S390_MEM_OP 3678-------------------- 3679 3680:Capability: KVM_CAP_S390_MEM_OP, KVM_CAP_S390_PROTECTED, KVM_CAP_S390_MEM_OP_EXTENSION 3681:Architectures: s390 3682:Type: vm ioctl, vcpu ioctl 3683:Parameters: struct kvm_s390_mem_op (in) 3684:Returns: = 0 on success, 3685 < 0 on generic error (e.g. -EFAULT or -ENOMEM), 3686 > 0 if an exception occurred while walking the page tables 3687 3688Read or write data from/to the VM's memory. 3689The KVM_CAP_S390_MEM_OP_EXTENSION capability specifies what functionality is 3690supported. 3691 3692Parameters are specified via the following structure:: 3693 3694 struct kvm_s390_mem_op { 3695 __u64 gaddr; /* the guest address */ 3696 __u64 flags; /* flags */ 3697 __u32 size; /* amount of bytes */ 3698 __u32 op; /* type of operation */ 3699 __u64 buf; /* buffer in userspace */ 3700 union { 3701 struct { 3702 __u8 ar; /* the access register number */ 3703 __u8 key; /* access key, ignored if flag unset */ 3704 }; 3705 __u32 sida_offset; /* offset into the sida */ 3706 __u8 reserved[32]; /* ignored */ 3707 }; 3708 }; 3709 3710The start address of the memory region has to be specified in the "gaddr" 3711field, and the length of the region in the "size" field (which must not 3712be 0). The maximum value for "size" can be obtained by checking the 3713KVM_CAP_S390_MEM_OP capability. "buf" is the buffer supplied by the 3714userspace application where the read data should be written to for 3715a read access, or where the data that should be written is stored for 3716a write access. The "reserved" field is meant for future extensions. 3717Reserved and unused values are ignored. Future extension that add members must 3718introduce new flags. 3719 3720The type of operation is specified in the "op" field. Flags modifying 3721their behavior can be set in the "flags" field. Undefined flag bits must 3722be set to 0. 3723 3724Possible operations are: 3725 * ``KVM_S390_MEMOP_LOGICAL_READ`` 3726 * ``KVM_S390_MEMOP_LOGICAL_WRITE`` 3727 * ``KVM_S390_MEMOP_ABSOLUTE_READ`` 3728 * ``KVM_S390_MEMOP_ABSOLUTE_WRITE`` 3729 * ``KVM_S390_MEMOP_SIDA_READ`` 3730 * ``KVM_S390_MEMOP_SIDA_WRITE`` 3731 3732Logical read/write: 3733^^^^^^^^^^^^^^^^^^^ 3734 3735Access logical memory, i.e. translate the given guest address to an absolute 3736address given the state of the VCPU and use the absolute address as target of 3737the access. "ar" designates the access register number to be used; the valid 3738range is 0..15. 3739Logical accesses are permitted for the VCPU ioctl only. 3740Logical accesses are permitted for non-protected guests only. 3741 3742Supported flags: 3743 * ``KVM_S390_MEMOP_F_CHECK_ONLY`` 3744 * ``KVM_S390_MEMOP_F_INJECT_EXCEPTION`` 3745 * ``KVM_S390_MEMOP_F_SKEY_PROTECTION`` 3746 3747The KVM_S390_MEMOP_F_CHECK_ONLY flag can be set to check whether the 3748corresponding memory access would cause an access exception; however, 3749no actual access to the data in memory at the destination is performed. 3750In this case, "buf" is unused and can be NULL. 3751 3752In case an access exception occurred during the access (or would occur 3753in case of KVM_S390_MEMOP_F_CHECK_ONLY), the ioctl returns a positive 3754error number indicating the type of exception. This exception is also 3755raised directly at the corresponding VCPU if the flag 3756KVM_S390_MEMOP_F_INJECT_EXCEPTION is set. 3757 3758If the KVM_S390_MEMOP_F_SKEY_PROTECTION flag is set, storage key 3759protection is also in effect and may cause exceptions if accesses are 3760prohibited given the access key designated by "key"; the valid range is 0..15. 3761KVM_S390_MEMOP_F_SKEY_PROTECTION is available if KVM_CAP_S390_MEM_OP_EXTENSION 3762is > 0. 3763 3764Absolute read/write: 3765^^^^^^^^^^^^^^^^^^^^ 3766 3767Access absolute memory. This operation is intended to be used with the 3768KVM_S390_MEMOP_F_SKEY_PROTECTION flag, to allow accessing memory and performing 3769the checks required for storage key protection as one operation (as opposed to 3770user space getting the storage keys, performing the checks, and accessing 3771memory thereafter, which could lead to a delay between check and access). 3772Absolute accesses are permitted for the VM ioctl if KVM_CAP_S390_MEM_OP_EXTENSION 3773is > 0. 3774Currently absolute accesses are not permitted for VCPU ioctls. 3775Absolute accesses are permitted for non-protected guests only. 3776 3777Supported flags: 3778 * ``KVM_S390_MEMOP_F_CHECK_ONLY`` 3779 * ``KVM_S390_MEMOP_F_SKEY_PROTECTION`` 3780 3781The semantics of the flags are as for logical accesses. 3782 3783SIDA read/write: 3784^^^^^^^^^^^^^^^^ 3785 3786Access the secure instruction data area which contains memory operands necessary 3787for instruction emulation for protected guests. 3788SIDA accesses are available if the KVM_CAP_S390_PROTECTED capability is available. 3789SIDA accesses are permitted for the VCPU ioctl only. 3790SIDA accesses are permitted for protected guests only. 3791 3792No flags are supported. 3793 37944.90 KVM_S390_GET_SKEYS 3795----------------------- 3796 3797:Capability: KVM_CAP_S390_SKEYS 3798:Architectures: s390 3799:Type: vm ioctl 3800:Parameters: struct kvm_s390_skeys 3801:Returns: 0 on success, KVM_S390_GET_SKEYS_NONE if guest is not using storage 3802 keys, negative value on error 3803 3804This ioctl is used to get guest storage key values on the s390 3805architecture. The ioctl takes parameters via the kvm_s390_skeys struct:: 3806 3807 struct kvm_s390_skeys { 3808 __u64 start_gfn; 3809 __u64 count; 3810 __u64 skeydata_addr; 3811 __u32 flags; 3812 __u32 reserved[9]; 3813 }; 3814 3815The start_gfn field is the number of the first guest frame whose storage keys 3816you want to get. 3817 3818The count field is the number of consecutive frames (starting from start_gfn) 3819whose storage keys to get. The count field must be at least 1 and the maximum 3820allowed value is defined as KVM_S390_SKEYS_MAX. Values outside this range 3821will cause the ioctl to return -EINVAL. 3822 3823The skeydata_addr field is the address to a buffer large enough to hold count 3824bytes. This buffer will be filled with storage key data by the ioctl. 3825 38264.91 KVM_S390_SET_SKEYS 3827----------------------- 3828 3829:Capability: KVM_CAP_S390_SKEYS 3830:Architectures: s390 3831:Type: vm ioctl 3832:Parameters: struct kvm_s390_skeys 3833:Returns: 0 on success, negative value on error 3834 3835This ioctl is used to set guest storage key values on the s390 3836architecture. The ioctl takes parameters via the kvm_s390_skeys struct. 3837See section on KVM_S390_GET_SKEYS for struct definition. 3838 3839The start_gfn field is the number of the first guest frame whose storage keys 3840you want to set. 3841 3842The count field is the number of consecutive frames (starting from start_gfn) 3843whose storage keys to get. The count field must be at least 1 and the maximum 3844allowed value is defined as KVM_S390_SKEYS_MAX. Values outside this range 3845will cause the ioctl to return -EINVAL. 3846 3847The skeydata_addr field is the address to a buffer containing count bytes of 3848storage keys. Each byte in the buffer will be set as the storage key for a 3849single frame starting at start_gfn for count frames. 3850 3851Note: If any architecturally invalid key value is found in the given data then 3852the ioctl will return -EINVAL. 3853 38544.92 KVM_S390_IRQ 3855----------------- 3856 3857:Capability: KVM_CAP_S390_INJECT_IRQ 3858:Architectures: s390 3859:Type: vcpu ioctl 3860:Parameters: struct kvm_s390_irq (in) 3861:Returns: 0 on success, -1 on error 3862 3863Errors: 3864 3865 3866 ====== ================================================================= 3867 EINVAL interrupt type is invalid 3868 type is KVM_S390_SIGP_STOP and flag parameter is invalid value, 3869 type is KVM_S390_INT_EXTERNAL_CALL and code is bigger 3870 than the maximum of VCPUs 3871 EBUSY type is KVM_S390_SIGP_SET_PREFIX and vcpu is not stopped, 3872 type is KVM_S390_SIGP_STOP and a stop irq is already pending, 3873 type is KVM_S390_INT_EXTERNAL_CALL and an external call interrupt 3874 is already pending 3875 ====== ================================================================= 3876 3877Allows to inject an interrupt to the guest. 3878 3879Using struct kvm_s390_irq as a parameter allows 3880to inject additional payload which is not 3881possible via KVM_S390_INTERRUPT. 3882 3883Interrupt parameters are passed via kvm_s390_irq:: 3884 3885 struct kvm_s390_irq { 3886 __u64 type; 3887 union { 3888 struct kvm_s390_io_info io; 3889 struct kvm_s390_ext_info ext; 3890 struct kvm_s390_pgm_info pgm; 3891 struct kvm_s390_emerg_info emerg; 3892 struct kvm_s390_extcall_info extcall; 3893 struct kvm_s390_prefix_info prefix; 3894 struct kvm_s390_stop_info stop; 3895 struct kvm_s390_mchk_info mchk; 3896 char reserved[64]; 3897 } u; 3898 }; 3899 3900type can be one of the following: 3901 3902- KVM_S390_SIGP_STOP - sigp stop; parameter in .stop 3903- KVM_S390_PROGRAM_INT - program check; parameters in .pgm 3904- KVM_S390_SIGP_SET_PREFIX - sigp set prefix; parameters in .prefix 3905- KVM_S390_RESTART - restart; no parameters 3906- KVM_S390_INT_CLOCK_COMP - clock comparator interrupt; no parameters 3907- KVM_S390_INT_CPU_TIMER - CPU timer interrupt; no parameters 3908- KVM_S390_INT_EMERGENCY - sigp emergency; parameters in .emerg 3909- KVM_S390_INT_EXTERNAL_CALL - sigp external call; parameters in .extcall 3910- KVM_S390_MCHK - machine check interrupt; parameters in .mchk 3911 3912This is an asynchronous vcpu ioctl and can be invoked from any thread. 3913 39144.94 KVM_S390_GET_IRQ_STATE 3915--------------------------- 3916 3917:Capability: KVM_CAP_S390_IRQ_STATE 3918:Architectures: s390 3919:Type: vcpu ioctl 3920:Parameters: struct kvm_s390_irq_state (out) 3921:Returns: >= number of bytes copied into buffer, 3922 -EINVAL if buffer size is 0, 3923 -ENOBUFS if buffer size is too small to fit all pending interrupts, 3924 -EFAULT if the buffer address was invalid 3925 3926This ioctl allows userspace to retrieve the complete state of all currently 3927pending interrupts in a single buffer. Use cases include migration 3928and introspection. The parameter structure contains the address of a 3929userspace buffer and its length:: 3930 3931 struct kvm_s390_irq_state { 3932 __u64 buf; 3933 __u32 flags; /* will stay unused for compatibility reasons */ 3934 __u32 len; 3935 __u32 reserved[4]; /* will stay unused for compatibility reasons */ 3936 }; 3937 3938Userspace passes in the above struct and for each pending interrupt a 3939struct kvm_s390_irq is copied to the provided buffer. 3940 3941The structure contains a flags and a reserved field for future extensions. As 3942the kernel never checked for flags == 0 and QEMU never pre-zeroed flags and 3943reserved, these fields can not be used in the future without breaking 3944compatibility. 3945 3946If -ENOBUFS is returned the buffer provided was too small and userspace 3947may retry with a bigger buffer. 3948 39494.95 KVM_S390_SET_IRQ_STATE 3950--------------------------- 3951 3952:Capability: KVM_CAP_S390_IRQ_STATE 3953:Architectures: s390 3954:Type: vcpu ioctl 3955:Parameters: struct kvm_s390_irq_state (in) 3956:Returns: 0 on success, 3957 -EFAULT if the buffer address was invalid, 3958 -EINVAL for an invalid buffer length (see below), 3959 -EBUSY if there were already interrupts pending, 3960 errors occurring when actually injecting the 3961 interrupt. See KVM_S390_IRQ. 3962 3963This ioctl allows userspace to set the complete state of all cpu-local 3964interrupts currently pending for the vcpu. It is intended for restoring 3965interrupt state after a migration. The input parameter is a userspace buffer 3966containing a struct kvm_s390_irq_state:: 3967 3968 struct kvm_s390_irq_state { 3969 __u64 buf; 3970 __u32 flags; /* will stay unused for compatibility reasons */ 3971 __u32 len; 3972 __u32 reserved[4]; /* will stay unused for compatibility reasons */ 3973 }; 3974 3975The restrictions for flags and reserved apply as well. 3976(see KVM_S390_GET_IRQ_STATE) 3977 3978The userspace memory referenced by buf contains a struct kvm_s390_irq 3979for each interrupt to be injected into the guest. 3980If one of the interrupts could not be injected for some reason the 3981ioctl aborts. 3982 3983len must be a multiple of sizeof(struct kvm_s390_irq). It must be > 0 3984and it must not exceed (max_vcpus + 32) * sizeof(struct kvm_s390_irq), 3985which is the maximum number of possibly pending cpu-local interrupts. 3986 39874.96 KVM_SMI 3988------------ 3989 3990:Capability: KVM_CAP_X86_SMM 3991:Architectures: x86 3992:Type: vcpu ioctl 3993:Parameters: none 3994:Returns: 0 on success, -1 on error 3995 3996Queues an SMI on the thread's vcpu. 3997 39984.97 KVM_X86_SET_MSR_FILTER 3999---------------------------- 4000 4001:Capability: KVM_X86_SET_MSR_FILTER 4002:Architectures: x86 4003:Type: vm ioctl 4004:Parameters: struct kvm_msr_filter 4005:Returns: 0 on success, < 0 on error 4006 4007:: 4008 4009 struct kvm_msr_filter_range { 4010 #define KVM_MSR_FILTER_READ (1 << 0) 4011 #define KVM_MSR_FILTER_WRITE (1 << 1) 4012 __u32 flags; 4013 __u32 nmsrs; /* number of msrs in bitmap */ 4014 __u32 base; /* MSR index the bitmap starts at */ 4015 __u8 *bitmap; /* a 1 bit allows the operations in flags, 0 denies */ 4016 }; 4017 4018 #define KVM_MSR_FILTER_MAX_RANGES 16 4019 struct kvm_msr_filter { 4020 #define KVM_MSR_FILTER_DEFAULT_ALLOW (0 << 0) 4021 #define KVM_MSR_FILTER_DEFAULT_DENY (1 << 0) 4022 __u32 flags; 4023 struct kvm_msr_filter_range ranges[KVM_MSR_FILTER_MAX_RANGES]; 4024 }; 4025 4026flags values for ``struct kvm_msr_filter_range``: 4027 4028``KVM_MSR_FILTER_READ`` 4029 4030 Filter read accesses to MSRs using the given bitmap. A 0 in the bitmap 4031 indicates that a read should immediately fail, while a 1 indicates that 4032 a read for a particular MSR should be handled regardless of the default 4033 filter action. 4034 4035``KVM_MSR_FILTER_WRITE`` 4036 4037 Filter write accesses to MSRs using the given bitmap. A 0 in the bitmap 4038 indicates that a write should immediately fail, while a 1 indicates that 4039 a write for a particular MSR should be handled regardless of the default 4040 filter action. 4041 4042``KVM_MSR_FILTER_READ | KVM_MSR_FILTER_WRITE`` 4043 4044 Filter both read and write accesses to MSRs using the given bitmap. A 0 4045 in the bitmap indicates that both reads and writes should immediately fail, 4046 while a 1 indicates that reads and writes for a particular MSR are not 4047 filtered by this range. 4048 4049flags values for ``struct kvm_msr_filter``: 4050 4051``KVM_MSR_FILTER_DEFAULT_ALLOW`` 4052 4053 If no filter range matches an MSR index that is getting accessed, KVM will 4054 fall back to allowing access to the MSR. 4055 4056``KVM_MSR_FILTER_DEFAULT_DENY`` 4057 4058 If no filter range matches an MSR index that is getting accessed, KVM will 4059 fall back to rejecting access to the MSR. In this mode, all MSRs that should 4060 be processed by KVM need to explicitly be marked as allowed in the bitmaps. 4061 4062This ioctl allows user space to define up to 16 bitmaps of MSR ranges to 4063specify whether a certain MSR access should be explicitly filtered for or not. 4064 4065If this ioctl has never been invoked, MSR accesses are not guarded and the 4066default KVM in-kernel emulation behavior is fully preserved. 4067 4068Calling this ioctl with an empty set of ranges (all nmsrs == 0) disables MSR 4069filtering. In that mode, ``KVM_MSR_FILTER_DEFAULT_DENY`` is invalid and causes 4070an error. 4071 4072As soon as the filtering is in place, every MSR access is processed through 4073the filtering except for accesses to the x2APIC MSRs (from 0x800 to 0x8ff); 4074x2APIC MSRs are always allowed, independent of the ``default_allow`` setting, 4075and their behavior depends on the ``X2APIC_ENABLE`` bit of the APIC base 4076register. 4077 4078.. warning:: 4079 MSR accesses coming from nested vmentry/vmexit are not filtered. 4080 This includes both writes to individual VMCS fields and reads/writes 4081 through the MSR lists pointed to by the VMCS. 4082 4083If a bit is within one of the defined ranges, read and write accesses are 4084guarded by the bitmap's value for the MSR index if the kind of access 4085is included in the ``struct kvm_msr_filter_range`` flags. If no range 4086cover this particular access, the behavior is determined by the flags 4087field in the kvm_msr_filter struct: ``KVM_MSR_FILTER_DEFAULT_ALLOW`` 4088and ``KVM_MSR_FILTER_DEFAULT_DENY``. 4089 4090Each bitmap range specifies a range of MSRs to potentially allow access on. 4091The range goes from MSR index [base .. base+nmsrs]. The flags field 4092indicates whether reads, writes or both reads and writes are filtered 4093by setting a 1 bit in the bitmap for the corresponding MSR index. 4094 4095If an MSR access is not permitted through the filtering, it generates a 4096#GP inside the guest. When combined with KVM_CAP_X86_USER_SPACE_MSR, that 4097allows user space to deflect and potentially handle various MSR accesses 4098into user space. 4099 4100If a vCPU is in running state while this ioctl is invoked, the vCPU may 4101experience inconsistent filtering behavior on MSR accesses. 4102 41034.98 KVM_CREATE_SPAPR_TCE_64 4104---------------------------- 4105 4106:Capability: KVM_CAP_SPAPR_TCE_64 4107:Architectures: powerpc 4108:Type: vm ioctl 4109:Parameters: struct kvm_create_spapr_tce_64 (in) 4110:Returns: file descriptor for manipulating the created TCE table 4111 4112This is an extension for KVM_CAP_SPAPR_TCE which only supports 32bit 4113windows, described in 4.62 KVM_CREATE_SPAPR_TCE 4114 4115This capability uses extended struct in ioctl interface:: 4116 4117 /* for KVM_CAP_SPAPR_TCE_64 */ 4118 struct kvm_create_spapr_tce_64 { 4119 __u64 liobn; 4120 __u32 page_shift; 4121 __u32 flags; 4122 __u64 offset; /* in pages */ 4123 __u64 size; /* in pages */ 4124 }; 4125 4126The aim of extension is to support an additional bigger DMA window with 4127a variable page size. 4128KVM_CREATE_SPAPR_TCE_64 receives a 64bit window size, an IOMMU page shift and 4129a bus offset of the corresponding DMA window, @size and @offset are numbers 4130of IOMMU pages. 4131 4132@flags are not used at the moment. 4133 4134The rest of functionality is identical to KVM_CREATE_SPAPR_TCE. 4135 41364.99 KVM_REINJECT_CONTROL 4137------------------------- 4138 4139:Capability: KVM_CAP_REINJECT_CONTROL 4140:Architectures: x86 4141:Type: vm ioctl 4142:Parameters: struct kvm_reinject_control (in) 4143:Returns: 0 on success, 4144 -EFAULT if struct kvm_reinject_control cannot be read, 4145 -ENXIO if KVM_CREATE_PIT or KVM_CREATE_PIT2 didn't succeed earlier. 4146 4147i8254 (PIT) has two modes, reinject and !reinject. The default is reinject, 4148where KVM queues elapsed i8254 ticks and monitors completion of interrupt from 4149vector(s) that i8254 injects. Reinject mode dequeues a tick and injects its 4150interrupt whenever there isn't a pending interrupt from i8254. 4151!reinject mode injects an interrupt as soon as a tick arrives. 4152 4153:: 4154 4155 struct kvm_reinject_control { 4156 __u8 pit_reinject; 4157 __u8 reserved[31]; 4158 }; 4159 4160pit_reinject = 0 (!reinject mode) is recommended, unless running an old 4161operating system that uses the PIT for timing (e.g. Linux 2.4.x). 4162 41634.100 KVM_PPC_CONFIGURE_V3_MMU 4164------------------------------ 4165 4166:Capability: KVM_CAP_PPC_RADIX_MMU or KVM_CAP_PPC_HASH_MMU_V3 4167:Architectures: ppc 4168:Type: vm ioctl 4169:Parameters: struct kvm_ppc_mmuv3_cfg (in) 4170:Returns: 0 on success, 4171 -EFAULT if struct kvm_ppc_mmuv3_cfg cannot be read, 4172 -EINVAL if the configuration is invalid 4173 4174This ioctl controls whether the guest will use radix or HPT (hashed 4175page table) translation, and sets the pointer to the process table for 4176the guest. 4177 4178:: 4179 4180 struct kvm_ppc_mmuv3_cfg { 4181 __u64 flags; 4182 __u64 process_table; 4183 }; 4184 4185There are two bits that can be set in flags; KVM_PPC_MMUV3_RADIX and 4186KVM_PPC_MMUV3_GTSE. KVM_PPC_MMUV3_RADIX, if set, configures the guest 4187to use radix tree translation, and if clear, to use HPT translation. 4188KVM_PPC_MMUV3_GTSE, if set and if KVM permits it, configures the guest 4189to be able to use the global TLB and SLB invalidation instructions; 4190if clear, the guest may not use these instructions. 4191 4192The process_table field specifies the address and size of the guest 4193process table, which is in the guest's space. This field is formatted 4194as the second doubleword of the partition table entry, as defined in 4195the Power ISA V3.00, Book III section 5.7.6.1. 4196 41974.101 KVM_PPC_GET_RMMU_INFO 4198--------------------------- 4199 4200:Capability: KVM_CAP_PPC_RADIX_MMU 4201:Architectures: ppc 4202:Type: vm ioctl 4203:Parameters: struct kvm_ppc_rmmu_info (out) 4204:Returns: 0 on success, 4205 -EFAULT if struct kvm_ppc_rmmu_info cannot be written, 4206 -EINVAL if no useful information can be returned 4207 4208This ioctl returns a structure containing two things: (a) a list 4209containing supported radix tree geometries, and (b) a list that maps 4210page sizes to put in the "AP" (actual page size) field for the tlbie 4211(TLB invalidate entry) instruction. 4212 4213:: 4214 4215 struct kvm_ppc_rmmu_info { 4216 struct kvm_ppc_radix_geom { 4217 __u8 page_shift; 4218 __u8 level_bits[4]; 4219 __u8 pad[3]; 4220 } geometries[8]; 4221 __u32 ap_encodings[8]; 4222 }; 4223 4224The geometries[] field gives up to 8 supported geometries for the 4225radix page table, in terms of the log base 2 of the smallest page 4226size, and the number of bits indexed at each level of the tree, from 4227the PTE level up to the PGD level in that order. Any unused entries 4228will have 0 in the page_shift field. 4229 4230The ap_encodings gives the supported page sizes and their AP field 4231encodings, encoded with the AP value in the top 3 bits and the log 4232base 2 of the page size in the bottom 6 bits. 4233 42344.102 KVM_PPC_RESIZE_HPT_PREPARE 4235-------------------------------- 4236 4237:Capability: KVM_CAP_SPAPR_RESIZE_HPT 4238:Architectures: powerpc 4239:Type: vm ioctl 4240:Parameters: struct kvm_ppc_resize_hpt (in) 4241:Returns: 0 on successful completion, 4242 >0 if a new HPT is being prepared, the value is an estimated 4243 number of milliseconds until preparation is complete, 4244 -EFAULT if struct kvm_reinject_control cannot be read, 4245 -EINVAL if the supplied shift or flags are invalid, 4246 -ENOMEM if unable to allocate the new HPT, 4247 4248Used to implement the PAPR extension for runtime resizing of a guest's 4249Hashed Page Table (HPT). Specifically this starts, stops or monitors 4250the preparation of a new potential HPT for the guest, essentially 4251implementing the H_RESIZE_HPT_PREPARE hypercall. 4252 4253:: 4254 4255 struct kvm_ppc_resize_hpt { 4256 __u64 flags; 4257 __u32 shift; 4258 __u32 pad; 4259 }; 4260 4261If called with shift > 0 when there is no pending HPT for the guest, 4262this begins preparation of a new pending HPT of size 2^(shift) bytes. 4263It then returns a positive integer with the estimated number of 4264milliseconds until preparation is complete. 4265 4266If called when there is a pending HPT whose size does not match that 4267requested in the parameters, discards the existing pending HPT and 4268creates a new one as above. 4269 4270If called when there is a pending HPT of the size requested, will: 4271 4272 * If preparation of the pending HPT is already complete, return 0 4273 * If preparation of the pending HPT has failed, return an error 4274 code, then discard the pending HPT. 4275 * If preparation of the pending HPT is still in progress, return an 4276 estimated number of milliseconds until preparation is complete. 4277 4278If called with shift == 0, discards any currently pending HPT and 4279returns 0 (i.e. cancels any in-progress preparation). 4280 4281flags is reserved for future expansion, currently setting any bits in 4282flags will result in an -EINVAL. 4283 4284Normally this will be called repeatedly with the same parameters until 4285it returns <= 0. The first call will initiate preparation, subsequent 4286ones will monitor preparation until it completes or fails. 4287 42884.103 KVM_PPC_RESIZE_HPT_COMMIT 4289------------------------------- 4290 4291:Capability: KVM_CAP_SPAPR_RESIZE_HPT 4292:Architectures: powerpc 4293:Type: vm ioctl 4294:Parameters: struct kvm_ppc_resize_hpt (in) 4295:Returns: 0 on successful completion, 4296 -EFAULT if struct kvm_reinject_control cannot be read, 4297 -EINVAL if the supplied shift or flags are invalid, 4298 -ENXIO is there is no pending HPT, or the pending HPT doesn't 4299 have the requested size, 4300 -EBUSY if the pending HPT is not fully prepared, 4301 -ENOSPC if there was a hash collision when moving existing 4302 HPT entries to the new HPT, 4303 -EIO on other error conditions 4304 4305Used to implement the PAPR extension for runtime resizing of a guest's 4306Hashed Page Table (HPT). Specifically this requests that the guest be 4307transferred to working with the new HPT, essentially implementing the 4308H_RESIZE_HPT_COMMIT hypercall. 4309 4310:: 4311 4312 struct kvm_ppc_resize_hpt { 4313 __u64 flags; 4314 __u32 shift; 4315 __u32 pad; 4316 }; 4317 4318This should only be called after KVM_PPC_RESIZE_HPT_PREPARE has 4319returned 0 with the same parameters. In other cases 4320KVM_PPC_RESIZE_HPT_COMMIT will return an error (usually -ENXIO or 4321-EBUSY, though others may be possible if the preparation was started, 4322but failed). 4323 4324This will have undefined effects on the guest if it has not already 4325placed itself in a quiescent state where no vcpu will make MMU enabled 4326memory accesses. 4327 4328On succsful completion, the pending HPT will become the guest's active 4329HPT and the previous HPT will be discarded. 4330 4331On failure, the guest will still be operating on its previous HPT. 4332 43334.104 KVM_X86_GET_MCE_CAP_SUPPORTED 4334----------------------------------- 4335 4336:Capability: KVM_CAP_MCE 4337:Architectures: x86 4338:Type: system ioctl 4339:Parameters: u64 mce_cap (out) 4340:Returns: 0 on success, -1 on error 4341 4342Returns supported MCE capabilities. The u64 mce_cap parameter 4343has the same format as the MSR_IA32_MCG_CAP register. Supported 4344capabilities will have the corresponding bits set. 4345 43464.105 KVM_X86_SETUP_MCE 4347----------------------- 4348 4349:Capability: KVM_CAP_MCE 4350:Architectures: x86 4351:Type: vcpu ioctl 4352:Parameters: u64 mcg_cap (in) 4353:Returns: 0 on success, 4354 -EFAULT if u64 mcg_cap cannot be read, 4355 -EINVAL if the requested number of banks is invalid, 4356 -EINVAL if requested MCE capability is not supported. 4357 4358Initializes MCE support for use. The u64 mcg_cap parameter 4359has the same format as the MSR_IA32_MCG_CAP register and 4360specifies which capabilities should be enabled. The maximum 4361supported number of error-reporting banks can be retrieved when 4362checking for KVM_CAP_MCE. The supported capabilities can be 4363retrieved with KVM_X86_GET_MCE_CAP_SUPPORTED. 4364 43654.106 KVM_X86_SET_MCE 4366--------------------- 4367 4368:Capability: KVM_CAP_MCE 4369:Architectures: x86 4370:Type: vcpu ioctl 4371:Parameters: struct kvm_x86_mce (in) 4372:Returns: 0 on success, 4373 -EFAULT if struct kvm_x86_mce cannot be read, 4374 -EINVAL if the bank number is invalid, 4375 -EINVAL if VAL bit is not set in status field. 4376 4377Inject a machine check error (MCE) into the guest. The input 4378parameter is:: 4379 4380 struct kvm_x86_mce { 4381 __u64 status; 4382 __u64 addr; 4383 __u64 misc; 4384 __u64 mcg_status; 4385 __u8 bank; 4386 __u8 pad1[7]; 4387 __u64 pad2[3]; 4388 }; 4389 4390If the MCE being reported is an uncorrected error, KVM will 4391inject it as an MCE exception into the guest. If the guest 4392MCG_STATUS register reports that an MCE is in progress, KVM 4393causes an KVM_EXIT_SHUTDOWN vmexit. 4394 4395Otherwise, if the MCE is a corrected error, KVM will just 4396store it in the corresponding bank (provided this bank is 4397not holding a previously reported uncorrected error). 4398 43994.107 KVM_S390_GET_CMMA_BITS 4400---------------------------- 4401 4402:Capability: KVM_CAP_S390_CMMA_MIGRATION 4403:Architectures: s390 4404:Type: vm ioctl 4405:Parameters: struct kvm_s390_cmma_log (in, out) 4406:Returns: 0 on success, a negative value on error 4407 4408This ioctl is used to get the values of the CMMA bits on the s390 4409architecture. It is meant to be used in two scenarios: 4410 4411- During live migration to save the CMMA values. Live migration needs 4412 to be enabled via the KVM_REQ_START_MIGRATION VM property. 4413- To non-destructively peek at the CMMA values, with the flag 4414 KVM_S390_CMMA_PEEK set. 4415 4416The ioctl takes parameters via the kvm_s390_cmma_log struct. The desired 4417values are written to a buffer whose location is indicated via the "values" 4418member in the kvm_s390_cmma_log struct. The values in the input struct are 4419also updated as needed. 4420 4421Each CMMA value takes up one byte. 4422 4423:: 4424 4425 struct kvm_s390_cmma_log { 4426 __u64 start_gfn; 4427 __u32 count; 4428 __u32 flags; 4429 union { 4430 __u64 remaining; 4431 __u64 mask; 4432 }; 4433 __u64 values; 4434 }; 4435 4436start_gfn is the number of the first guest frame whose CMMA values are 4437to be retrieved, 4438 4439count is the length of the buffer in bytes, 4440 4441values points to the buffer where the result will be written to. 4442 4443If count is greater than KVM_S390_SKEYS_MAX, then it is considered to be 4444KVM_S390_SKEYS_MAX. KVM_S390_SKEYS_MAX is re-used for consistency with 4445other ioctls. 4446 4447The result is written in the buffer pointed to by the field values, and 4448the values of the input parameter are updated as follows. 4449 4450Depending on the flags, different actions are performed. The only 4451supported flag so far is KVM_S390_CMMA_PEEK. 4452 4453The default behaviour if KVM_S390_CMMA_PEEK is not set is: 4454start_gfn will indicate the first page frame whose CMMA bits were dirty. 4455It is not necessarily the same as the one passed as input, as clean pages 4456are skipped. 4457 4458count will indicate the number of bytes actually written in the buffer. 4459It can (and very often will) be smaller than the input value, since the 4460buffer is only filled until 16 bytes of clean values are found (which 4461are then not copied in the buffer). Since a CMMA migration block needs 4462the base address and the length, for a total of 16 bytes, we will send 4463back some clean data if there is some dirty data afterwards, as long as 4464the size of the clean data does not exceed the size of the header. This 4465allows to minimize the amount of data to be saved or transferred over 4466the network at the expense of more roundtrips to userspace. The next 4467invocation of the ioctl will skip over all the clean values, saving 4468potentially more than just the 16 bytes we found. 4469 4470If KVM_S390_CMMA_PEEK is set: 4471the existing storage attributes are read even when not in migration 4472mode, and no other action is performed; 4473 4474the output start_gfn will be equal to the input start_gfn, 4475 4476the output count will be equal to the input count, except if the end of 4477memory has been reached. 4478 4479In both cases: 4480the field "remaining" will indicate the total number of dirty CMMA values 4481still remaining, or 0 if KVM_S390_CMMA_PEEK is set and migration mode is 4482not enabled. 4483 4484mask is unused. 4485 4486values points to the userspace buffer where the result will be stored. 4487 4488This ioctl can fail with -ENOMEM if not enough memory can be allocated to 4489complete the task, with -ENXIO if CMMA is not enabled, with -EINVAL if 4490KVM_S390_CMMA_PEEK is not set but migration mode was not enabled, with 4491-EFAULT if the userspace address is invalid or if no page table is 4492present for the addresses (e.g. when using hugepages). 4493 44944.108 KVM_S390_SET_CMMA_BITS 4495---------------------------- 4496 4497:Capability: KVM_CAP_S390_CMMA_MIGRATION 4498:Architectures: s390 4499:Type: vm ioctl 4500:Parameters: struct kvm_s390_cmma_log (in) 4501:Returns: 0 on success, a negative value on error 4502 4503This ioctl is used to set the values of the CMMA bits on the s390 4504architecture. It is meant to be used during live migration to restore 4505the CMMA values, but there are no restrictions on its use. 4506The ioctl takes parameters via the kvm_s390_cmma_values struct. 4507Each CMMA value takes up one byte. 4508 4509:: 4510 4511 struct kvm_s390_cmma_log { 4512 __u64 start_gfn; 4513 __u32 count; 4514 __u32 flags; 4515 union { 4516 __u64 remaining; 4517 __u64 mask; 4518 }; 4519 __u64 values; 4520 }; 4521 4522start_gfn indicates the starting guest frame number, 4523 4524count indicates how many values are to be considered in the buffer, 4525 4526flags is not used and must be 0. 4527 4528mask indicates which PGSTE bits are to be considered. 4529 4530remaining is not used. 4531 4532values points to the buffer in userspace where to store the values. 4533 4534This ioctl can fail with -ENOMEM if not enough memory can be allocated to 4535complete the task, with -ENXIO if CMMA is not enabled, with -EINVAL if 4536the count field is too large (e.g. more than KVM_S390_CMMA_SIZE_MAX) or 4537if the flags field was not 0, with -EFAULT if the userspace address is 4538invalid, if invalid pages are written to (e.g. after the end of memory) 4539or if no page table is present for the addresses (e.g. when using 4540hugepages). 4541 45424.109 KVM_PPC_GET_CPU_CHAR 4543-------------------------- 4544 4545:Capability: KVM_CAP_PPC_GET_CPU_CHAR 4546:Architectures: powerpc 4547:Type: vm ioctl 4548:Parameters: struct kvm_ppc_cpu_char (out) 4549:Returns: 0 on successful completion, 4550 -EFAULT if struct kvm_ppc_cpu_char cannot be written 4551 4552This ioctl gives userspace information about certain characteristics 4553of the CPU relating to speculative execution of instructions and 4554possible information leakage resulting from speculative execution (see 4555CVE-2017-5715, CVE-2017-5753 and CVE-2017-5754). The information is 4556returned in struct kvm_ppc_cpu_char, which looks like this:: 4557 4558 struct kvm_ppc_cpu_char { 4559 __u64 character; /* characteristics of the CPU */ 4560 __u64 behaviour; /* recommended software behaviour */ 4561 __u64 character_mask; /* valid bits in character */ 4562 __u64 behaviour_mask; /* valid bits in behaviour */ 4563 }; 4564 4565For extensibility, the character_mask and behaviour_mask fields 4566indicate which bits of character and behaviour have been filled in by 4567the kernel. If the set of defined bits is extended in future then 4568userspace will be able to tell whether it is running on a kernel that 4569knows about the new bits. 4570 4571The character field describes attributes of the CPU which can help 4572with preventing inadvertent information disclosure - specifically, 4573whether there is an instruction to flash-invalidate the L1 data cache 4574(ori 30,30,0 or mtspr SPRN_TRIG2,rN), whether the L1 data cache is set 4575to a mode where entries can only be used by the thread that created 4576them, whether the bcctr[l] instruction prevents speculation, and 4577whether a speculation barrier instruction (ori 31,31,0) is provided. 4578 4579The behaviour field describes actions that software should take to 4580prevent inadvertent information disclosure, and thus describes which 4581vulnerabilities the hardware is subject to; specifically whether the 4582L1 data cache should be flushed when returning to user mode from the 4583kernel, and whether a speculation barrier should be placed between an 4584array bounds check and the array access. 4585 4586These fields use the same bit definitions as the new 4587H_GET_CPU_CHARACTERISTICS hypercall. 4588 45894.110 KVM_MEMORY_ENCRYPT_OP 4590--------------------------- 4591 4592:Capability: basic 4593:Architectures: x86 4594:Type: vm 4595:Parameters: an opaque platform specific structure (in/out) 4596:Returns: 0 on success; -1 on error 4597 4598If the platform supports creating encrypted VMs then this ioctl can be used 4599for issuing platform-specific memory encryption commands to manage those 4600encrypted VMs. 4601 4602Currently, this ioctl is used for issuing Secure Encrypted Virtualization 4603(SEV) commands on AMD Processors. The SEV commands are defined in 4604Documentation/virt/kvm/amd-memory-encryption.rst. 4605 46064.111 KVM_MEMORY_ENCRYPT_REG_REGION 4607----------------------------------- 4608 4609:Capability: basic 4610:Architectures: x86 4611:Type: system 4612:Parameters: struct kvm_enc_region (in) 4613:Returns: 0 on success; -1 on error 4614 4615This ioctl can be used to register a guest memory region which may 4616contain encrypted data (e.g. guest RAM, SMRAM etc). 4617 4618It is used in the SEV-enabled guest. When encryption is enabled, a guest 4619memory region may contain encrypted data. The SEV memory encryption 4620engine uses a tweak such that two identical plaintext pages, each at 4621different locations will have differing ciphertexts. So swapping or 4622moving ciphertext of those pages will not result in plaintext being 4623swapped. So relocating (or migrating) physical backing pages for the SEV 4624guest will require some additional steps. 4625 4626Note: The current SEV key management spec does not provide commands to 4627swap or migrate (move) ciphertext pages. Hence, for now we pin the guest 4628memory region registered with the ioctl. 4629 46304.112 KVM_MEMORY_ENCRYPT_UNREG_REGION 4631------------------------------------- 4632 4633:Capability: basic 4634:Architectures: x86 4635:Type: system 4636:Parameters: struct kvm_enc_region (in) 4637:Returns: 0 on success; -1 on error 4638 4639This ioctl can be used to unregister the guest memory region registered 4640with KVM_MEMORY_ENCRYPT_REG_REGION ioctl above. 4641 46424.113 KVM_HYPERV_EVENTFD 4643------------------------ 4644 4645:Capability: KVM_CAP_HYPERV_EVENTFD 4646:Architectures: x86 4647:Type: vm ioctl 4648:Parameters: struct kvm_hyperv_eventfd (in) 4649 4650This ioctl (un)registers an eventfd to receive notifications from the guest on 4651the specified Hyper-V connection id through the SIGNAL_EVENT hypercall, without 4652causing a user exit. SIGNAL_EVENT hypercall with non-zero event flag number 4653(bits 24-31) still triggers a KVM_EXIT_HYPERV_HCALL user exit. 4654 4655:: 4656 4657 struct kvm_hyperv_eventfd { 4658 __u32 conn_id; 4659 __s32 fd; 4660 __u32 flags; 4661 __u32 padding[3]; 4662 }; 4663 4664The conn_id field should fit within 24 bits:: 4665 4666 #define KVM_HYPERV_CONN_ID_MASK 0x00ffffff 4667 4668The acceptable values for the flags field are:: 4669 4670 #define KVM_HYPERV_EVENTFD_DEASSIGN (1 << 0) 4671 4672:Returns: 0 on success, 4673 -EINVAL if conn_id or flags is outside the allowed range, 4674 -ENOENT on deassign if the conn_id isn't registered, 4675 -EEXIST on assign if the conn_id is already registered 4676 46774.114 KVM_GET_NESTED_STATE 4678-------------------------- 4679 4680:Capability: KVM_CAP_NESTED_STATE 4681:Architectures: x86 4682:Type: vcpu ioctl 4683:Parameters: struct kvm_nested_state (in/out) 4684:Returns: 0 on success, -1 on error 4685 4686Errors: 4687 4688 ===== ============================================================= 4689 E2BIG the total state size exceeds the value of 'size' specified by 4690 the user; the size required will be written into size. 4691 ===== ============================================================= 4692 4693:: 4694 4695 struct kvm_nested_state { 4696 __u16 flags; 4697 __u16 format; 4698 __u32 size; 4699 4700 union { 4701 struct kvm_vmx_nested_state_hdr vmx; 4702 struct kvm_svm_nested_state_hdr svm; 4703 4704 /* Pad the header to 128 bytes. */ 4705 __u8 pad[120]; 4706 } hdr; 4707 4708 union { 4709 struct kvm_vmx_nested_state_data vmx[0]; 4710 struct kvm_svm_nested_state_data svm[0]; 4711 } data; 4712 }; 4713 4714 #define KVM_STATE_NESTED_GUEST_MODE 0x00000001 4715 #define KVM_STATE_NESTED_RUN_PENDING 0x00000002 4716 #define KVM_STATE_NESTED_EVMCS 0x00000004 4717 4718 #define KVM_STATE_NESTED_FORMAT_VMX 0 4719 #define KVM_STATE_NESTED_FORMAT_SVM 1 4720 4721 #define KVM_STATE_NESTED_VMX_VMCS_SIZE 0x1000 4722 4723 #define KVM_STATE_NESTED_VMX_SMM_GUEST_MODE 0x00000001 4724 #define KVM_STATE_NESTED_VMX_SMM_VMXON 0x00000002 4725 4726 #define KVM_STATE_VMX_PREEMPTION_TIMER_DEADLINE 0x00000001 4727 4728 struct kvm_vmx_nested_state_hdr { 4729 __u64 vmxon_pa; 4730 __u64 vmcs12_pa; 4731 4732 struct { 4733 __u16 flags; 4734 } smm; 4735 4736 __u32 flags; 4737 __u64 preemption_timer_deadline; 4738 }; 4739 4740 struct kvm_vmx_nested_state_data { 4741 __u8 vmcs12[KVM_STATE_NESTED_VMX_VMCS_SIZE]; 4742 __u8 shadow_vmcs12[KVM_STATE_NESTED_VMX_VMCS_SIZE]; 4743 }; 4744 4745This ioctl copies the vcpu's nested virtualization state from the kernel to 4746userspace. 4747 4748The maximum size of the state can be retrieved by passing KVM_CAP_NESTED_STATE 4749to the KVM_CHECK_EXTENSION ioctl(). 4750 47514.115 KVM_SET_NESTED_STATE 4752-------------------------- 4753 4754:Capability: KVM_CAP_NESTED_STATE 4755:Architectures: x86 4756:Type: vcpu ioctl 4757:Parameters: struct kvm_nested_state (in) 4758:Returns: 0 on success, -1 on error 4759 4760This copies the vcpu's kvm_nested_state struct from userspace to the kernel. 4761For the definition of struct kvm_nested_state, see KVM_GET_NESTED_STATE. 4762 47634.116 KVM_(UN)REGISTER_COALESCED_MMIO 4764------------------------------------- 4765 4766:Capability: KVM_CAP_COALESCED_MMIO (for coalesced mmio) 4767 KVM_CAP_COALESCED_PIO (for coalesced pio) 4768:Architectures: all 4769:Type: vm ioctl 4770:Parameters: struct kvm_coalesced_mmio_zone 4771:Returns: 0 on success, < 0 on error 4772 4773Coalesced I/O is a performance optimization that defers hardware 4774register write emulation so that userspace exits are avoided. It is 4775typically used to reduce the overhead of emulating frequently accessed 4776hardware registers. 4777 4778When a hardware register is configured for coalesced I/O, write accesses 4779do not exit to userspace and their value is recorded in a ring buffer 4780that is shared between kernel and userspace. 4781 4782Coalesced I/O is used if one or more write accesses to a hardware 4783register can be deferred until a read or a write to another hardware 4784register on the same device. This last access will cause a vmexit and 4785userspace will process accesses from the ring buffer before emulating 4786it. That will avoid exiting to userspace on repeated writes. 4787 4788Coalesced pio is based on coalesced mmio. There is little difference 4789between coalesced mmio and pio except that coalesced pio records accesses 4790to I/O ports. 4791 47924.117 KVM_CLEAR_DIRTY_LOG (vm ioctl) 4793------------------------------------ 4794 4795:Capability: KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 4796:Architectures: x86, arm64, mips 4797:Type: vm ioctl 4798:Parameters: struct kvm_clear_dirty_log (in) 4799:Returns: 0 on success, -1 on error 4800 4801:: 4802 4803 /* for KVM_CLEAR_DIRTY_LOG */ 4804 struct kvm_clear_dirty_log { 4805 __u32 slot; 4806 __u32 num_pages; 4807 __u64 first_page; 4808 union { 4809 void __user *dirty_bitmap; /* one bit per page */ 4810 __u64 padding; 4811 }; 4812 }; 4813 4814The ioctl clears the dirty status of pages in a memory slot, according to 4815the bitmap that is passed in struct kvm_clear_dirty_log's dirty_bitmap 4816field. Bit 0 of the bitmap corresponds to page "first_page" in the 4817memory slot, and num_pages is the size in bits of the input bitmap. 4818first_page must be a multiple of 64; num_pages must also be a multiple of 481964 unless first_page + num_pages is the size of the memory slot. For each 4820bit that is set in the input bitmap, the corresponding page is marked "clean" 4821in KVM's dirty bitmap, and dirty tracking is re-enabled for that page 4822(for example via write-protection, or by clearing the dirty bit in 4823a page table entry). 4824 4825If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of slot field specifies 4826the address space for which you want to clear the dirty status. See 4827KVM_SET_USER_MEMORY_REGION for details on the usage of slot field. 4828 4829This ioctl is mostly useful when KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 4830is enabled; for more information, see the description of the capability. 4831However, it can always be used as long as KVM_CHECK_EXTENSION confirms 4832that KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 is present. 4833 48344.118 KVM_GET_SUPPORTED_HV_CPUID 4835-------------------------------- 4836 4837:Capability: KVM_CAP_HYPERV_CPUID (vcpu), KVM_CAP_SYS_HYPERV_CPUID (system) 4838:Architectures: x86 4839:Type: system ioctl, vcpu ioctl 4840:Parameters: struct kvm_cpuid2 (in/out) 4841:Returns: 0 on success, -1 on error 4842 4843:: 4844 4845 struct kvm_cpuid2 { 4846 __u32 nent; 4847 __u32 padding; 4848 struct kvm_cpuid_entry2 entries[0]; 4849 }; 4850 4851 struct kvm_cpuid_entry2 { 4852 __u32 function; 4853 __u32 index; 4854 __u32 flags; 4855 __u32 eax; 4856 __u32 ebx; 4857 __u32 ecx; 4858 __u32 edx; 4859 __u32 padding[3]; 4860 }; 4861 4862This ioctl returns x86 cpuid features leaves related to Hyper-V emulation in 4863KVM. Userspace can use the information returned by this ioctl to construct 4864cpuid information presented to guests consuming Hyper-V enlightenments (e.g. 4865Windows or Hyper-V guests). 4866 4867CPUID feature leaves returned by this ioctl are defined by Hyper-V Top Level 4868Functional Specification (TLFS). These leaves can't be obtained with 4869KVM_GET_SUPPORTED_CPUID ioctl because some of them intersect with KVM feature 4870leaves (0x40000000, 0x40000001). 4871 4872Currently, the following list of CPUID leaves are returned: 4873 4874 - HYPERV_CPUID_VENDOR_AND_MAX_FUNCTIONS 4875 - HYPERV_CPUID_INTERFACE 4876 - HYPERV_CPUID_VERSION 4877 - HYPERV_CPUID_FEATURES 4878 - HYPERV_CPUID_ENLIGHTMENT_INFO 4879 - HYPERV_CPUID_IMPLEMENT_LIMITS 4880 - HYPERV_CPUID_NESTED_FEATURES 4881 - HYPERV_CPUID_SYNDBG_VENDOR_AND_MAX_FUNCTIONS 4882 - HYPERV_CPUID_SYNDBG_INTERFACE 4883 - HYPERV_CPUID_SYNDBG_PLATFORM_CAPABILITIES 4884 4885Userspace invokes KVM_GET_SUPPORTED_HV_CPUID by passing a kvm_cpuid2 structure 4886with the 'nent' field indicating the number of entries in the variable-size 4887array 'entries'. If the number of entries is too low to describe all Hyper-V 4888feature leaves, an error (E2BIG) is returned. If the number is more or equal 4889to the number of Hyper-V feature leaves, the 'nent' field is adjusted to the 4890number of valid entries in the 'entries' array, which is then filled. 4891 4892'index' and 'flags' fields in 'struct kvm_cpuid_entry2' are currently reserved, 4893userspace should not expect to get any particular value there. 4894 4895Note, vcpu version of KVM_GET_SUPPORTED_HV_CPUID is currently deprecated. Unlike 4896system ioctl which exposes all supported feature bits unconditionally, vcpu 4897version has the following quirks: 4898 4899- HYPERV_CPUID_NESTED_FEATURES leaf and HV_X64_ENLIGHTENED_VMCS_RECOMMENDED 4900 feature bit are only exposed when Enlightened VMCS was previously enabled 4901 on the corresponding vCPU (KVM_CAP_HYPERV_ENLIGHTENED_VMCS). 4902- HV_STIMER_DIRECT_MODE_AVAILABLE bit is only exposed with in-kernel LAPIC. 4903 (presumes KVM_CREATE_IRQCHIP has already been called). 4904 49054.119 KVM_ARM_VCPU_FINALIZE 4906--------------------------- 4907 4908:Architectures: arm64 4909:Type: vcpu ioctl 4910:Parameters: int feature (in) 4911:Returns: 0 on success, -1 on error 4912 4913Errors: 4914 4915 ====== ============================================================== 4916 EPERM feature not enabled, needs configuration, or already finalized 4917 EINVAL feature unknown or not present 4918 ====== ============================================================== 4919 4920Recognised values for feature: 4921 4922 ===== =========================================== 4923 arm64 KVM_ARM_VCPU_SVE (requires KVM_CAP_ARM_SVE) 4924 ===== =========================================== 4925 4926Finalizes the configuration of the specified vcpu feature. 4927 4928The vcpu must already have been initialised, enabling the affected feature, by 4929means of a successful KVM_ARM_VCPU_INIT call with the appropriate flag set in 4930features[]. 4931 4932For affected vcpu features, this is a mandatory step that must be performed 4933before the vcpu is fully usable. 4934 4935Between KVM_ARM_VCPU_INIT and KVM_ARM_VCPU_FINALIZE, the feature may be 4936configured by use of ioctls such as KVM_SET_ONE_REG. The exact configuration 4937that should be performaned and how to do it are feature-dependent. 4938 4939Other calls that depend on a particular feature being finalized, such as 4940KVM_RUN, KVM_GET_REG_LIST, KVM_GET_ONE_REG and KVM_SET_ONE_REG, will fail with 4941-EPERM unless the feature has already been finalized by means of a 4942KVM_ARM_VCPU_FINALIZE call. 4943 4944See KVM_ARM_VCPU_INIT for details of vcpu features that require finalization 4945using this ioctl. 4946 49474.120 KVM_SET_PMU_EVENT_FILTER 4948------------------------------ 4949 4950:Capability: KVM_CAP_PMU_EVENT_FILTER 4951:Architectures: x86 4952:Type: vm ioctl 4953:Parameters: struct kvm_pmu_event_filter (in) 4954:Returns: 0 on success, -1 on error 4955 4956:: 4957 4958 struct kvm_pmu_event_filter { 4959 __u32 action; 4960 __u32 nevents; 4961 __u32 fixed_counter_bitmap; 4962 __u32 flags; 4963 __u32 pad[4]; 4964 __u64 events[0]; 4965 }; 4966 4967This ioctl restricts the set of PMU events that the guest can program. 4968The argument holds a list of events which will be allowed or denied. 4969The eventsel+umask of each event the guest attempts to program is compared 4970against the events field to determine whether the guest should have access. 4971The events field only controls general purpose counters; fixed purpose 4972counters are controlled by the fixed_counter_bitmap. 4973 4974No flags are defined yet, the field must be zero. 4975 4976Valid values for 'action':: 4977 4978 #define KVM_PMU_EVENT_ALLOW 0 4979 #define KVM_PMU_EVENT_DENY 1 4980 49814.121 KVM_PPC_SVM_OFF 4982--------------------- 4983 4984:Capability: basic 4985:Architectures: powerpc 4986:Type: vm ioctl 4987:Parameters: none 4988:Returns: 0 on successful completion, 4989 4990Errors: 4991 4992 ====== ================================================================ 4993 EINVAL if ultravisor failed to terminate the secure guest 4994 ENOMEM if hypervisor failed to allocate new radix page tables for guest 4995 ====== ================================================================ 4996 4997This ioctl is used to turn off the secure mode of the guest or transition 4998the guest from secure mode to normal mode. This is invoked when the guest 4999is reset. This has no effect if called for a normal guest. 5000 5001This ioctl issues an ultravisor call to terminate the secure guest, 5002unpins the VPA pages and releases all the device pages that are used to 5003track the secure pages by hypervisor. 5004 50054.122 KVM_S390_NORMAL_RESET 5006--------------------------- 5007 5008:Capability: KVM_CAP_S390_VCPU_RESETS 5009:Architectures: s390 5010:Type: vcpu ioctl 5011:Parameters: none 5012:Returns: 0 5013 5014This ioctl resets VCPU registers and control structures according to 5015the cpu reset definition in the POP (Principles Of Operation). 5016 50174.123 KVM_S390_INITIAL_RESET 5018---------------------------- 5019 5020:Capability: none 5021:Architectures: s390 5022:Type: vcpu ioctl 5023:Parameters: none 5024:Returns: 0 5025 5026This ioctl resets VCPU registers and control structures according to 5027the initial cpu reset definition in the POP. However, the cpu is not 5028put into ESA mode. This reset is a superset of the normal reset. 5029 50304.124 KVM_S390_CLEAR_RESET 5031-------------------------- 5032 5033:Capability: KVM_CAP_S390_VCPU_RESETS 5034:Architectures: s390 5035:Type: vcpu ioctl 5036:Parameters: none 5037:Returns: 0 5038 5039This ioctl resets VCPU registers and control structures according to 5040the clear cpu reset definition in the POP. However, the cpu is not put 5041into ESA mode. This reset is a superset of the initial reset. 5042 5043 50444.125 KVM_S390_PV_COMMAND 5045------------------------- 5046 5047:Capability: KVM_CAP_S390_PROTECTED 5048:Architectures: s390 5049:Type: vm ioctl 5050:Parameters: struct kvm_pv_cmd 5051:Returns: 0 on success, < 0 on error 5052 5053:: 5054 5055 struct kvm_pv_cmd { 5056 __u32 cmd; /* Command to be executed */ 5057 __u16 rc; /* Ultravisor return code */ 5058 __u16 rrc; /* Ultravisor return reason code */ 5059 __u64 data; /* Data or address */ 5060 __u32 flags; /* flags for future extensions. Must be 0 for now */ 5061 __u32 reserved[3]; 5062 }; 5063 5064cmd values: 5065 5066KVM_PV_ENABLE 5067 Allocate memory and register the VM with the Ultravisor, thereby 5068 donating memory to the Ultravisor that will become inaccessible to 5069 KVM. All existing CPUs are converted to protected ones. After this 5070 command has succeeded, any CPU added via hotplug will become 5071 protected during its creation as well. 5072 5073 Errors: 5074 5075 ===== ============================= 5076 EINTR an unmasked signal is pending 5077 ===== ============================= 5078 5079KVM_PV_DISABLE 5080 5081 Deregister the VM from the Ultravisor and reclaim the memory that 5082 had been donated to the Ultravisor, making it usable by the kernel 5083 again. All registered VCPUs are converted back to non-protected 5084 ones. 5085 5086KVM_PV_VM_SET_SEC_PARMS 5087 Pass the image header from VM memory to the Ultravisor in 5088 preparation of image unpacking and verification. 5089 5090KVM_PV_VM_UNPACK 5091 Unpack (protect and decrypt) a page of the encrypted boot image. 5092 5093KVM_PV_VM_VERIFY 5094 Verify the integrity of the unpacked image. Only if this succeeds, 5095 KVM is allowed to start protected VCPUs. 5096 50974.126 KVM_X86_SET_MSR_FILTER 5098---------------------------- 5099 5100:Capability: KVM_CAP_X86_MSR_FILTER 5101:Architectures: x86 5102:Type: vm ioctl 5103:Parameters: struct kvm_msr_filter 5104:Returns: 0 on success, < 0 on error 5105 5106:: 5107 5108 struct kvm_msr_filter_range { 5109 #define KVM_MSR_FILTER_READ (1 << 0) 5110 #define KVM_MSR_FILTER_WRITE (1 << 1) 5111 __u32 flags; 5112 __u32 nmsrs; /* number of msrs in bitmap */ 5113 __u32 base; /* MSR index the bitmap starts at */ 5114 __u8 *bitmap; /* a 1 bit allows the operations in flags, 0 denies */ 5115 }; 5116 5117 #define KVM_MSR_FILTER_MAX_RANGES 16 5118 struct kvm_msr_filter { 5119 #define KVM_MSR_FILTER_DEFAULT_ALLOW (0 << 0) 5120 #define KVM_MSR_FILTER_DEFAULT_DENY (1 << 0) 5121 __u32 flags; 5122 struct kvm_msr_filter_range ranges[KVM_MSR_FILTER_MAX_RANGES]; 5123 }; 5124 5125flags values for ``struct kvm_msr_filter_range``: 5126 5127``KVM_MSR_FILTER_READ`` 5128 5129 Filter read accesses to MSRs using the given bitmap. A 0 in the bitmap 5130 indicates that a read should immediately fail, while a 1 indicates that 5131 a read for a particular MSR should be handled regardless of the default 5132 filter action. 5133 5134``KVM_MSR_FILTER_WRITE`` 5135 5136 Filter write accesses to MSRs using the given bitmap. A 0 in the bitmap 5137 indicates that a write should immediately fail, while a 1 indicates that 5138 a write for a particular MSR should be handled regardless of the default 5139 filter action. 5140 5141``KVM_MSR_FILTER_READ | KVM_MSR_FILTER_WRITE`` 5142 5143 Filter both read and write accesses to MSRs using the given bitmap. A 0 5144 in the bitmap indicates that both reads and writes should immediately fail, 5145 while a 1 indicates that reads and writes for a particular MSR are not 5146 filtered by this range. 5147 5148flags values for ``struct kvm_msr_filter``: 5149 5150``KVM_MSR_FILTER_DEFAULT_ALLOW`` 5151 5152 If no filter range matches an MSR index that is getting accessed, KVM will 5153 fall back to allowing access to the MSR. 5154 5155``KVM_MSR_FILTER_DEFAULT_DENY`` 5156 5157 If no filter range matches an MSR index that is getting accessed, KVM will 5158 fall back to rejecting access to the MSR. In this mode, all MSRs that should 5159 be processed by KVM need to explicitly be marked as allowed in the bitmaps. 5160 5161This ioctl allows user space to define up to 16 bitmaps of MSR ranges to 5162specify whether a certain MSR access should be explicitly filtered for or not. 5163 5164If this ioctl has never been invoked, MSR accesses are not guarded and the 5165default KVM in-kernel emulation behavior is fully preserved. 5166 5167Calling this ioctl with an empty set of ranges (all nmsrs == 0) disables MSR 5168filtering. In that mode, ``KVM_MSR_FILTER_DEFAULT_DENY`` is invalid and causes 5169an error. 5170 5171As soon as the filtering is in place, every MSR access is processed through 5172the filtering except for accesses to the x2APIC MSRs (from 0x800 to 0x8ff); 5173x2APIC MSRs are always allowed, independent of the ``default_allow`` setting, 5174and their behavior depends on the ``X2APIC_ENABLE`` bit of the APIC base 5175register. 5176 5177If a bit is within one of the defined ranges, read and write accesses are 5178guarded by the bitmap's value for the MSR index if the kind of access 5179is included in the ``struct kvm_msr_filter_range`` flags. If no range 5180cover this particular access, the behavior is determined by the flags 5181field in the kvm_msr_filter struct: ``KVM_MSR_FILTER_DEFAULT_ALLOW`` 5182and ``KVM_MSR_FILTER_DEFAULT_DENY``. 5183 5184Each bitmap range specifies a range of MSRs to potentially allow access on. 5185The range goes from MSR index [base .. base+nmsrs]. The flags field 5186indicates whether reads, writes or both reads and writes are filtered 5187by setting a 1 bit in the bitmap for the corresponding MSR index. 5188 5189If an MSR access is not permitted through the filtering, it generates a 5190#GP inside the guest. When combined with KVM_CAP_X86_USER_SPACE_MSR, that 5191allows user space to deflect and potentially handle various MSR accesses 5192into user space. 5193 5194Note, invoking this ioctl with a vCPU is running is inherently racy. However, 5195KVM does guarantee that vCPUs will see either the previous filter or the new 5196filter, e.g. MSRs with identical settings in both the old and new filter will 5197have deterministic behavior. 5198 51994.127 KVM_XEN_HVM_SET_ATTR 5200-------------------------- 5201 5202:Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO 5203:Architectures: x86 5204:Type: vm ioctl 5205:Parameters: struct kvm_xen_hvm_attr 5206:Returns: 0 on success, < 0 on error 5207 5208:: 5209 5210 struct kvm_xen_hvm_attr { 5211 __u16 type; 5212 __u16 pad[3]; 5213 union { 5214 __u8 long_mode; 5215 __u8 vector; 5216 struct { 5217 __u64 gfn; 5218 } shared_info; 5219 __u64 pad[4]; 5220 } u; 5221 }; 5222 5223type values: 5224 5225KVM_XEN_ATTR_TYPE_LONG_MODE 5226 Sets the ABI mode of the VM to 32-bit or 64-bit (long mode). This 5227 determines the layout of the shared info pages exposed to the VM. 5228 5229KVM_XEN_ATTR_TYPE_SHARED_INFO 5230 Sets the guest physical frame number at which the Xen "shared info" 5231 page resides. Note that although Xen places vcpu_info for the first 5232 32 vCPUs in the shared_info page, KVM does not automatically do so 5233 and instead requires that KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO be used 5234 explicitly even when the vcpu_info for a given vCPU resides at the 5235 "default" location in the shared_info page. This is because KVM is 5236 not aware of the Xen CPU id which is used as the index into the 5237 vcpu_info[] array, so cannot know the correct default location. 5238 5239 Note that the shared info page may be constantly written to by KVM; 5240 it contains the event channel bitmap used to deliver interrupts to 5241 a Xen guest, amongst other things. It is exempt from dirty tracking 5242 mechanisms — KVM will not explicitly mark the page as dirty each 5243 time an event channel interrupt is delivered to the guest! Thus, 5244 userspace should always assume that the designated GFN is dirty if 5245 any vCPU has been running or any event channel interrupts can be 5246 routed to the guest. 5247 5248KVM_XEN_ATTR_TYPE_UPCALL_VECTOR 5249 Sets the exception vector used to deliver Xen event channel upcalls. 5250 52514.127 KVM_XEN_HVM_GET_ATTR 5252-------------------------- 5253 5254:Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO 5255:Architectures: x86 5256:Type: vm ioctl 5257:Parameters: struct kvm_xen_hvm_attr 5258:Returns: 0 on success, < 0 on error 5259 5260Allows Xen VM attributes to be read. For the structure and types, 5261see KVM_XEN_HVM_SET_ATTR above. 5262 52634.128 KVM_XEN_VCPU_SET_ATTR 5264--------------------------- 5265 5266:Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO 5267:Architectures: x86 5268:Type: vcpu ioctl 5269:Parameters: struct kvm_xen_vcpu_attr 5270:Returns: 0 on success, < 0 on error 5271 5272:: 5273 5274 struct kvm_xen_vcpu_attr { 5275 __u16 type; 5276 __u16 pad[3]; 5277 union { 5278 __u64 gpa; 5279 __u64 pad[4]; 5280 struct { 5281 __u64 state; 5282 __u64 state_entry_time; 5283 __u64 time_running; 5284 __u64 time_runnable; 5285 __u64 time_blocked; 5286 __u64 time_offline; 5287 } runstate; 5288 } u; 5289 }; 5290 5291type values: 5292 5293KVM_XEN_VCPU_ATTR_TYPE_VCPU_INFO 5294 Sets the guest physical address of the vcpu_info for a given vCPU. 5295 As with the shared_info page for the VM, the corresponding page may be 5296 dirtied at any time if event channel interrupt delivery is enabled, so 5297 userspace should always assume that the page is dirty without relying 5298 on dirty logging. 5299 5300KVM_XEN_VCPU_ATTR_TYPE_VCPU_TIME_INFO 5301 Sets the guest physical address of an additional pvclock structure 5302 for a given vCPU. This is typically used for guest vsyscall support. 5303 5304KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADDR 5305 Sets the guest physical address of the vcpu_runstate_info for a given 5306 vCPU. This is how a Xen guest tracks CPU state such as steal time. 5307 5308KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_CURRENT 5309 Sets the runstate (RUNSTATE_running/_runnable/_blocked/_offline) of 5310 the given vCPU from the .u.runstate.state member of the structure. 5311 KVM automatically accounts running and runnable time but blocked 5312 and offline states are only entered explicitly. 5313 5314KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_DATA 5315 Sets all fields of the vCPU runstate data from the .u.runstate member 5316 of the structure, including the current runstate. The state_entry_time 5317 must equal the sum of the other four times. 5318 5319KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADJUST 5320 This *adds* the contents of the .u.runstate members of the structure 5321 to the corresponding members of the given vCPU's runstate data, thus 5322 permitting atomic adjustments to the runstate times. The adjustment 5323 to the state_entry_time must equal the sum of the adjustments to the 5324 other four times. The state field must be set to -1, or to a valid 5325 runstate value (RUNSTATE_running, RUNSTATE_runnable, RUNSTATE_blocked 5326 or RUNSTATE_offline) to set the current accounted state as of the 5327 adjusted state_entry_time. 5328 53294.129 KVM_XEN_VCPU_GET_ATTR 5330--------------------------- 5331 5332:Capability: KVM_CAP_XEN_HVM / KVM_XEN_HVM_CONFIG_SHARED_INFO 5333:Architectures: x86 5334:Type: vcpu ioctl 5335:Parameters: struct kvm_xen_vcpu_attr 5336:Returns: 0 on success, < 0 on error 5337 5338Allows Xen vCPU attributes to be read. For the structure and types, 5339see KVM_XEN_VCPU_SET_ATTR above. 5340 5341The KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADJUST type may not be used 5342with the KVM_XEN_VCPU_GET_ATTR ioctl. 5343 53444.130 KVM_ARM_MTE_COPY_TAGS 5345--------------------------- 5346 5347:Capability: KVM_CAP_ARM_MTE 5348:Architectures: arm64 5349:Type: vm ioctl 5350:Parameters: struct kvm_arm_copy_mte_tags 5351:Returns: number of bytes copied, < 0 on error (-EINVAL for incorrect 5352 arguments, -EFAULT if memory cannot be accessed). 5353 5354:: 5355 5356 struct kvm_arm_copy_mte_tags { 5357 __u64 guest_ipa; 5358 __u64 length; 5359 void __user *addr; 5360 __u64 flags; 5361 __u64 reserved[2]; 5362 }; 5363 5364Copies Memory Tagging Extension (MTE) tags to/from guest tag memory. The 5365``guest_ipa`` and ``length`` fields must be ``PAGE_SIZE`` aligned. The ``addr`` 5366field must point to a buffer which the tags will be copied to or from. 5367 5368``flags`` specifies the direction of copy, either ``KVM_ARM_TAGS_TO_GUEST`` or 5369``KVM_ARM_TAGS_FROM_GUEST``. 5370 5371The size of the buffer to store the tags is ``(length / 16)`` bytes 5372(granules in MTE are 16 bytes long). Each byte contains a single tag 5373value. This matches the format of ``PTRACE_PEEKMTETAGS`` and 5374``PTRACE_POKEMTETAGS``. 5375 5376If an error occurs before any data is copied then a negative error code is 5377returned. If some tags have been copied before an error occurs then the number 5378of bytes successfully copied is returned. If the call completes successfully 5379then ``length`` is returned. 5380 53814.131 KVM_GET_SREGS2 5382-------------------- 5383 5384:Capability: KVM_CAP_SREGS2 5385:Architectures: x86 5386:Type: vcpu ioctl 5387:Parameters: struct kvm_sregs2 (out) 5388:Returns: 0 on success, -1 on error 5389 5390Reads special registers from the vcpu. 5391This ioctl (when supported) replaces the KVM_GET_SREGS. 5392 5393:: 5394 5395 struct kvm_sregs2 { 5396 /* out (KVM_GET_SREGS2) / in (KVM_SET_SREGS2) */ 5397 struct kvm_segment cs, ds, es, fs, gs, ss; 5398 struct kvm_segment tr, ldt; 5399 struct kvm_dtable gdt, idt; 5400 __u64 cr0, cr2, cr3, cr4, cr8; 5401 __u64 efer; 5402 __u64 apic_base; 5403 __u64 flags; 5404 __u64 pdptrs[4]; 5405 }; 5406 5407flags values for ``kvm_sregs2``: 5408 5409``KVM_SREGS2_FLAGS_PDPTRS_VALID`` 5410 5411 Indicates thats the struct contain valid PDPTR values. 5412 5413 54144.132 KVM_SET_SREGS2 5415-------------------- 5416 5417:Capability: KVM_CAP_SREGS2 5418:Architectures: x86 5419:Type: vcpu ioctl 5420:Parameters: struct kvm_sregs2 (in) 5421:Returns: 0 on success, -1 on error 5422 5423Writes special registers into the vcpu. 5424See KVM_GET_SREGS2 for the data structures. 5425This ioctl (when supported) replaces the KVM_SET_SREGS. 5426 54274.133 KVM_GET_STATS_FD 5428---------------------- 5429 5430:Capability: KVM_CAP_STATS_BINARY_FD 5431:Architectures: all 5432:Type: vm ioctl, vcpu ioctl 5433:Parameters: none 5434:Returns: statistics file descriptor on success, < 0 on error 5435 5436Errors: 5437 5438 ====== ====================================================== 5439 ENOMEM if the fd could not be created due to lack of memory 5440 EMFILE if the number of opened files exceeds the limit 5441 ====== ====================================================== 5442 5443The returned file descriptor can be used to read VM/vCPU statistics data in 5444binary format. The data in the file descriptor consists of four blocks 5445organized as follows: 5446 5447+-------------+ 5448| Header | 5449+-------------+ 5450| id string | 5451+-------------+ 5452| Descriptors | 5453+-------------+ 5454| Stats Data | 5455+-------------+ 5456 5457Apart from the header starting at offset 0, please be aware that it is 5458not guaranteed that the four blocks are adjacent or in the above order; 5459the offsets of the id, descriptors and data blocks are found in the 5460header. However, all four blocks are aligned to 64 bit offsets in the 5461file and they do not overlap. 5462 5463All blocks except the data block are immutable. Userspace can read them 5464only one time after retrieving the file descriptor, and then use ``pread`` or 5465``lseek`` to read the statistics repeatedly. 5466 5467All data is in system endianness. 5468 5469The format of the header is as follows:: 5470 5471 struct kvm_stats_header { 5472 __u32 flags; 5473 __u32 name_size; 5474 __u32 num_desc; 5475 __u32 id_offset; 5476 __u32 desc_offset; 5477 __u32 data_offset; 5478 }; 5479 5480The ``flags`` field is not used at the moment. It is always read as 0. 5481 5482The ``name_size`` field is the size (in byte) of the statistics name string 5483(including trailing '\0') which is contained in the "id string" block and 5484appended at the end of every descriptor. 5485 5486The ``num_desc`` field is the number of descriptors that are included in the 5487descriptor block. (The actual number of values in the data block may be 5488larger, since each descriptor may comprise more than one value). 5489 5490The ``id_offset`` field is the offset of the id string from the start of the 5491file indicated by the file descriptor. It is a multiple of 8. 5492 5493The ``desc_offset`` field is the offset of the Descriptors block from the start 5494of the file indicated by the file descriptor. It is a multiple of 8. 5495 5496The ``data_offset`` field is the offset of the Stats Data block from the start 5497of the file indicated by the file descriptor. It is a multiple of 8. 5498 5499The id string block contains a string which identifies the file descriptor on 5500which KVM_GET_STATS_FD was invoked. The size of the block, including the 5501trailing ``'\0'``, is indicated by the ``name_size`` field in the header. 5502 5503The descriptors block is only needed to be read once for the lifetime of the 5504file descriptor contains a sequence of ``struct kvm_stats_desc``, each followed 5505by a string of size ``name_size``. 5506:: 5507 5508 #define KVM_STATS_TYPE_SHIFT 0 5509 #define KVM_STATS_TYPE_MASK (0xF << KVM_STATS_TYPE_SHIFT) 5510 #define KVM_STATS_TYPE_CUMULATIVE (0x0 << KVM_STATS_TYPE_SHIFT) 5511 #define KVM_STATS_TYPE_INSTANT (0x1 << KVM_STATS_TYPE_SHIFT) 5512 #define KVM_STATS_TYPE_PEAK (0x2 << KVM_STATS_TYPE_SHIFT) 5513 #define KVM_STATS_TYPE_LINEAR_HIST (0x3 << KVM_STATS_TYPE_SHIFT) 5514 #define KVM_STATS_TYPE_LOG_HIST (0x4 << KVM_STATS_TYPE_SHIFT) 5515 #define KVM_STATS_TYPE_MAX KVM_STATS_TYPE_LOG_HIST 5516 5517 #define KVM_STATS_UNIT_SHIFT 4 5518 #define KVM_STATS_UNIT_MASK (0xF << KVM_STATS_UNIT_SHIFT) 5519 #define KVM_STATS_UNIT_NONE (0x0 << KVM_STATS_UNIT_SHIFT) 5520 #define KVM_STATS_UNIT_BYTES (0x1 << KVM_STATS_UNIT_SHIFT) 5521 #define KVM_STATS_UNIT_SECONDS (0x2 << KVM_STATS_UNIT_SHIFT) 5522 #define KVM_STATS_UNIT_CYCLES (0x3 << KVM_STATS_UNIT_SHIFT) 5523 #define KVM_STATS_UNIT_MAX KVM_STATS_UNIT_CYCLES 5524 5525 #define KVM_STATS_BASE_SHIFT 8 5526 #define KVM_STATS_BASE_MASK (0xF << KVM_STATS_BASE_SHIFT) 5527 #define KVM_STATS_BASE_POW10 (0x0 << KVM_STATS_BASE_SHIFT) 5528 #define KVM_STATS_BASE_POW2 (0x1 << KVM_STATS_BASE_SHIFT) 5529 #define KVM_STATS_BASE_MAX KVM_STATS_BASE_POW2 5530 5531 struct kvm_stats_desc { 5532 __u32 flags; 5533 __s16 exponent; 5534 __u16 size; 5535 __u32 offset; 5536 __u32 bucket_size; 5537 char name[]; 5538 }; 5539 5540The ``flags`` field contains the type and unit of the statistics data described 5541by this descriptor. Its endianness is CPU native. 5542The following flags are supported: 5543 5544Bits 0-3 of ``flags`` encode the type: 5545 5546 * ``KVM_STATS_TYPE_CUMULATIVE`` 5547 The statistics reports a cumulative count. The value of data can only be increased. 5548 Most of the counters used in KVM are of this type. 5549 The corresponding ``size`` field for this type is always 1. 5550 All cumulative statistics data are read/write. 5551 * ``KVM_STATS_TYPE_INSTANT`` 5552 The statistics reports an instantaneous value. Its value can be increased or 5553 decreased. This type is usually used as a measurement of some resources, 5554 like the number of dirty pages, the number of large pages, etc. 5555 All instant statistics are read only. 5556 The corresponding ``size`` field for this type is always 1. 5557 * ``KVM_STATS_TYPE_PEAK`` 5558 The statistics data reports a peak value, for example the maximum number 5559 of items in a hash table bucket, the longest time waited and so on. 5560 The value of data can only be increased. 5561 The corresponding ``size`` field for this type is always 1. 5562 * ``KVM_STATS_TYPE_LINEAR_HIST`` 5563 The statistic is reported as a linear histogram. The number of 5564 buckets is specified by the ``size`` field. The size of buckets is specified 5565 by the ``hist_param`` field. The range of the Nth bucket (1 <= N < ``size``) 5566 is [``hist_param``*(N-1), ``hist_param``*N), while the range of the last 5567 bucket is [``hist_param``*(``size``-1), +INF). (+INF means positive infinity 5568 value.) The bucket value indicates how many samples fell in the bucket's range. 5569 * ``KVM_STATS_TYPE_LOG_HIST`` 5570 The statistic is reported as a logarithmic histogram. The number of 5571 buckets is specified by the ``size`` field. The range of the first bucket is 5572 [0, 1), while the range of the last bucket is [pow(2, ``size``-2), +INF). 5573 Otherwise, The Nth bucket (1 < N < ``size``) covers 5574 [pow(2, N-2), pow(2, N-1)). The bucket value indicates how many samples fell 5575 in the bucket's range. 5576 5577Bits 4-7 of ``flags`` encode the unit: 5578 5579 * ``KVM_STATS_UNIT_NONE`` 5580 There is no unit for the value of statistics data. This usually means that 5581 the value is a simple counter of an event. 5582 * ``KVM_STATS_UNIT_BYTES`` 5583 It indicates that the statistics data is used to measure memory size, in the 5584 unit of Byte, KiByte, MiByte, GiByte, etc. The unit of the data is 5585 determined by the ``exponent`` field in the descriptor. 5586 * ``KVM_STATS_UNIT_SECONDS`` 5587 It indicates that the statistics data is used to measure time or latency. 5588 * ``KVM_STATS_UNIT_CYCLES`` 5589 It indicates that the statistics data is used to measure CPU clock cycles. 5590 5591Bits 8-11 of ``flags``, together with ``exponent``, encode the scale of the 5592unit: 5593 5594 * ``KVM_STATS_BASE_POW10`` 5595 The scale is based on power of 10. It is used for measurement of time and 5596 CPU clock cycles. For example, an exponent of -9 can be used with 5597 ``KVM_STATS_UNIT_SECONDS`` to express that the unit is nanoseconds. 5598 * ``KVM_STATS_BASE_POW2`` 5599 The scale is based on power of 2. It is used for measurement of memory size. 5600 For example, an exponent of 20 can be used with ``KVM_STATS_UNIT_BYTES`` to 5601 express that the unit is MiB. 5602 5603The ``size`` field is the number of values of this statistics data. Its 5604value is usually 1 for most of simple statistics. 1 means it contains an 5605unsigned 64bit data. 5606 5607The ``offset`` field is the offset from the start of Data Block to the start of 5608the corresponding statistics data. 5609 5610The ``bucket_size`` field is used as a parameter for histogram statistics data. 5611It is only used by linear histogram statistics data, specifying the size of a 5612bucket. 5613 5614The ``name`` field is the name string of the statistics data. The name string 5615starts at the end of ``struct kvm_stats_desc``. The maximum length including 5616the trailing ``'\0'``, is indicated by ``name_size`` in the header. 5617 5618The Stats Data block contains an array of 64-bit values in the same order 5619as the descriptors in Descriptors block. 5620 56214.134 KVM_GET_XSAVE2 5622-------------------- 5623 5624:Capability: KVM_CAP_XSAVE2 5625:Architectures: x86 5626:Type: vcpu ioctl 5627:Parameters: struct kvm_xsave (out) 5628:Returns: 0 on success, -1 on error 5629 5630 5631:: 5632 5633 struct kvm_xsave { 5634 __u32 region[1024]; 5635 __u32 extra[0]; 5636 }; 5637 5638This ioctl would copy current vcpu's xsave struct to the userspace. It 5639copies as many bytes as are returned by KVM_CHECK_EXTENSION(KVM_CAP_XSAVE2) 5640when invoked on the vm file descriptor. The size value returned by 5641KVM_CHECK_EXTENSION(KVM_CAP_XSAVE2) will always be at least 4096. 5642Currently, it is only greater than 4096 if a dynamic feature has been 5643enabled with ``arch_prctl()``, but this may change in the future. 5644 5645The offsets of the state save areas in struct kvm_xsave follow the contents 5646of CPUID leaf 0xD on the host. 5647 5648 56495. The kvm_run structure 5650======================== 5651 5652Application code obtains a pointer to the kvm_run structure by 5653mmap()ing a vcpu fd. From that point, application code can control 5654execution by changing fields in kvm_run prior to calling the KVM_RUN 5655ioctl, and obtain information about the reason KVM_RUN returned by 5656looking up structure members. 5657 5658:: 5659 5660 struct kvm_run { 5661 /* in */ 5662 __u8 request_interrupt_window; 5663 5664Request that KVM_RUN return when it becomes possible to inject external 5665interrupts into the guest. Useful in conjunction with KVM_INTERRUPT. 5666 5667:: 5668 5669 __u8 immediate_exit; 5670 5671This field is polled once when KVM_RUN starts; if non-zero, KVM_RUN 5672exits immediately, returning -EINTR. In the common scenario where a 5673signal is used to "kick" a VCPU out of KVM_RUN, this field can be used 5674to avoid usage of KVM_SET_SIGNAL_MASK, which has worse scalability. 5675Rather than blocking the signal outside KVM_RUN, userspace can set up 5676a signal handler that sets run->immediate_exit to a non-zero value. 5677 5678This field is ignored if KVM_CAP_IMMEDIATE_EXIT is not available. 5679 5680:: 5681 5682 __u8 padding1[6]; 5683 5684 /* out */ 5685 __u32 exit_reason; 5686 5687When KVM_RUN has returned successfully (return value 0), this informs 5688application code why KVM_RUN has returned. Allowable values for this 5689field are detailed below. 5690 5691:: 5692 5693 __u8 ready_for_interrupt_injection; 5694 5695If request_interrupt_window has been specified, this field indicates 5696an interrupt can be injected now with KVM_INTERRUPT. 5697 5698:: 5699 5700 __u8 if_flag; 5701 5702The value of the current interrupt flag. Only valid if in-kernel 5703local APIC is not used. 5704 5705:: 5706 5707 __u16 flags; 5708 5709More architecture-specific flags detailing state of the VCPU that may 5710affect the device's behavior. Current defined flags:: 5711 5712 /* x86, set if the VCPU is in system management mode */ 5713 #define KVM_RUN_X86_SMM (1 << 0) 5714 /* x86, set if bus lock detected in VM */ 5715 #define KVM_RUN_BUS_LOCK (1 << 1) 5716 5717:: 5718 5719 /* in (pre_kvm_run), out (post_kvm_run) */ 5720 __u64 cr8; 5721 5722The value of the cr8 register. Only valid if in-kernel local APIC is 5723not used. Both input and output. 5724 5725:: 5726 5727 __u64 apic_base; 5728 5729The value of the APIC BASE msr. Only valid if in-kernel local 5730APIC is not used. Both input and output. 5731 5732:: 5733 5734 union { 5735 /* KVM_EXIT_UNKNOWN */ 5736 struct { 5737 __u64 hardware_exit_reason; 5738 } hw; 5739 5740If exit_reason is KVM_EXIT_UNKNOWN, the vcpu has exited due to unknown 5741reasons. Further architecture-specific information is available in 5742hardware_exit_reason. 5743 5744:: 5745 5746 /* KVM_EXIT_FAIL_ENTRY */ 5747 struct { 5748 __u64 hardware_entry_failure_reason; 5749 __u32 cpu; /* if KVM_LAST_CPU */ 5750 } fail_entry; 5751 5752If exit_reason is KVM_EXIT_FAIL_ENTRY, the vcpu could not be run due 5753to unknown reasons. Further architecture-specific information is 5754available in hardware_entry_failure_reason. 5755 5756:: 5757 5758 /* KVM_EXIT_EXCEPTION */ 5759 struct { 5760 __u32 exception; 5761 __u32 error_code; 5762 } ex; 5763 5764Unused. 5765 5766:: 5767 5768 /* KVM_EXIT_IO */ 5769 struct { 5770 #define KVM_EXIT_IO_IN 0 5771 #define KVM_EXIT_IO_OUT 1 5772 __u8 direction; 5773 __u8 size; /* bytes */ 5774 __u16 port; 5775 __u32 count; 5776 __u64 data_offset; /* relative to kvm_run start */ 5777 } io; 5778 5779If exit_reason is KVM_EXIT_IO, then the vcpu has 5780executed a port I/O instruction which could not be satisfied by kvm. 5781data_offset describes where the data is located (KVM_EXIT_IO_OUT) or 5782where kvm expects application code to place the data for the next 5783KVM_RUN invocation (KVM_EXIT_IO_IN). Data format is a packed array. 5784 5785:: 5786 5787 /* KVM_EXIT_DEBUG */ 5788 struct { 5789 struct kvm_debug_exit_arch arch; 5790 } debug; 5791 5792If the exit_reason is KVM_EXIT_DEBUG, then a vcpu is processing a debug event 5793for which architecture specific information is returned. 5794 5795:: 5796 5797 /* KVM_EXIT_MMIO */ 5798 struct { 5799 __u64 phys_addr; 5800 __u8 data[8]; 5801 __u32 len; 5802 __u8 is_write; 5803 } mmio; 5804 5805If exit_reason is KVM_EXIT_MMIO, then the vcpu has 5806executed a memory-mapped I/O instruction which could not be satisfied 5807by kvm. The 'data' member contains the written data if 'is_write' is 5808true, and should be filled by application code otherwise. 5809 5810The 'data' member contains, in its first 'len' bytes, the value as it would 5811appear if the VCPU performed a load or store of the appropriate width directly 5812to the byte array. 5813 5814.. note:: 5815 5816 For KVM_EXIT_IO, KVM_EXIT_MMIO, KVM_EXIT_OSI, KVM_EXIT_PAPR, KVM_EXIT_XEN, 5817 KVM_EXIT_EPR, KVM_EXIT_X86_RDMSR and KVM_EXIT_X86_WRMSR the corresponding 5818 operations are complete (and guest state is consistent) only after userspace 5819 has re-entered the kernel with KVM_RUN. The kernel side will first finish 5820 incomplete operations and then check for pending signals. 5821 5822 The pending state of the operation is not preserved in state which is 5823 visible to userspace, thus userspace should ensure that the operation is 5824 completed before performing a live migration. Userspace can re-enter the 5825 guest with an unmasked signal pending or with the immediate_exit field set 5826 to complete pending operations without allowing any further instructions 5827 to be executed. 5828 5829:: 5830 5831 /* KVM_EXIT_HYPERCALL */ 5832 struct { 5833 __u64 nr; 5834 __u64 args[6]; 5835 __u64 ret; 5836 __u32 longmode; 5837 __u32 pad; 5838 } hypercall; 5839 5840Unused. This was once used for 'hypercall to userspace'. To implement 5841such functionality, use KVM_EXIT_IO (x86) or KVM_EXIT_MMIO (all except s390). 5842 5843.. note:: KVM_EXIT_IO is significantly faster than KVM_EXIT_MMIO. 5844 5845:: 5846 5847 /* KVM_EXIT_TPR_ACCESS */ 5848 struct { 5849 __u64 rip; 5850 __u32 is_write; 5851 __u32 pad; 5852 } tpr_access; 5853 5854To be documented (KVM_TPR_ACCESS_REPORTING). 5855 5856:: 5857 5858 /* KVM_EXIT_S390_SIEIC */ 5859 struct { 5860 __u8 icptcode; 5861 __u64 mask; /* psw upper half */ 5862 __u64 addr; /* psw lower half */ 5863 __u16 ipa; 5864 __u32 ipb; 5865 } s390_sieic; 5866 5867s390 specific. 5868 5869:: 5870 5871 /* KVM_EXIT_S390_RESET */ 5872 #define KVM_S390_RESET_POR 1 5873 #define KVM_S390_RESET_CLEAR 2 5874 #define KVM_S390_RESET_SUBSYSTEM 4 5875 #define KVM_S390_RESET_CPU_INIT 8 5876 #define KVM_S390_RESET_IPL 16 5877 __u64 s390_reset_flags; 5878 5879s390 specific. 5880 5881:: 5882 5883 /* KVM_EXIT_S390_UCONTROL */ 5884 struct { 5885 __u64 trans_exc_code; 5886 __u32 pgm_code; 5887 } s390_ucontrol; 5888 5889s390 specific. A page fault has occurred for a user controlled virtual 5890machine (KVM_VM_S390_UNCONTROL) on it's host page table that cannot be 5891resolved by the kernel. 5892The program code and the translation exception code that were placed 5893in the cpu's lowcore are presented here as defined by the z Architecture 5894Principles of Operation Book in the Chapter for Dynamic Address Translation 5895(DAT) 5896 5897:: 5898 5899 /* KVM_EXIT_DCR */ 5900 struct { 5901 __u32 dcrn; 5902 __u32 data; 5903 __u8 is_write; 5904 } dcr; 5905 5906Deprecated - was used for 440 KVM. 5907 5908:: 5909 5910 /* KVM_EXIT_OSI */ 5911 struct { 5912 __u64 gprs[32]; 5913 } osi; 5914 5915MOL uses a special hypercall interface it calls 'OSI'. To enable it, we catch 5916hypercalls and exit with this exit struct that contains all the guest gprs. 5917 5918If exit_reason is KVM_EXIT_OSI, then the vcpu has triggered such a hypercall. 5919Userspace can now handle the hypercall and when it's done modify the gprs as 5920necessary. Upon guest entry all guest GPRs will then be replaced by the values 5921in this struct. 5922 5923:: 5924 5925 /* KVM_EXIT_PAPR_HCALL */ 5926 struct { 5927 __u64 nr; 5928 __u64 ret; 5929 __u64 args[9]; 5930 } papr_hcall; 5931 5932This is used on 64-bit PowerPC when emulating a pSeries partition, 5933e.g. with the 'pseries' machine type in qemu. It occurs when the 5934guest does a hypercall using the 'sc 1' instruction. The 'nr' field 5935contains the hypercall number (from the guest R3), and 'args' contains 5936the arguments (from the guest R4 - R12). Userspace should put the 5937return code in 'ret' and any extra returned values in args[]. 5938The possible hypercalls are defined in the Power Architecture Platform 5939Requirements (PAPR) document available from www.power.org (free 5940developer registration required to access it). 5941 5942:: 5943 5944 /* KVM_EXIT_S390_TSCH */ 5945 struct { 5946 __u16 subchannel_id; 5947 __u16 subchannel_nr; 5948 __u32 io_int_parm; 5949 __u32 io_int_word; 5950 __u32 ipb; 5951 __u8 dequeued; 5952 } s390_tsch; 5953 5954s390 specific. This exit occurs when KVM_CAP_S390_CSS_SUPPORT has been enabled 5955and TEST SUBCHANNEL was intercepted. If dequeued is set, a pending I/O 5956interrupt for the target subchannel has been dequeued and subchannel_id, 5957subchannel_nr, io_int_parm and io_int_word contain the parameters for that 5958interrupt. ipb is needed for instruction parameter decoding. 5959 5960:: 5961 5962 /* KVM_EXIT_EPR */ 5963 struct { 5964 __u32 epr; 5965 } epr; 5966 5967On FSL BookE PowerPC chips, the interrupt controller has a fast patch 5968interrupt acknowledge path to the core. When the core successfully 5969delivers an interrupt, it automatically populates the EPR register with 5970the interrupt vector number and acknowledges the interrupt inside 5971the interrupt controller. 5972 5973In case the interrupt controller lives in user space, we need to do 5974the interrupt acknowledge cycle through it to fetch the next to be 5975delivered interrupt vector using this exit. 5976 5977It gets triggered whenever both KVM_CAP_PPC_EPR are enabled and an 5978external interrupt has just been delivered into the guest. User space 5979should put the acknowledged interrupt vector into the 'epr' field. 5980 5981:: 5982 5983 /* KVM_EXIT_SYSTEM_EVENT */ 5984 struct { 5985 #define KVM_SYSTEM_EVENT_SHUTDOWN 1 5986 #define KVM_SYSTEM_EVENT_RESET 2 5987 #define KVM_SYSTEM_EVENT_CRASH 3 5988 __u32 type; 5989 __u64 flags; 5990 } system_event; 5991 5992If exit_reason is KVM_EXIT_SYSTEM_EVENT then the vcpu has triggered 5993a system-level event using some architecture specific mechanism (hypercall 5994or some special instruction). In case of ARM64, this is triggered using 5995HVC instruction based PSCI call from the vcpu. The 'type' field describes 5996the system-level event type. The 'flags' field describes architecture 5997specific flags for the system-level event. 5998 5999Valid values for 'type' are: 6000 6001 - KVM_SYSTEM_EVENT_SHUTDOWN -- the guest has requested a shutdown of the 6002 VM. Userspace is not obliged to honour this, and if it does honour 6003 this does not need to destroy the VM synchronously (ie it may call 6004 KVM_RUN again before shutdown finally occurs). 6005 - KVM_SYSTEM_EVENT_RESET -- the guest has requested a reset of the VM. 6006 As with SHUTDOWN, userspace can choose to ignore the request, or 6007 to schedule the reset to occur in the future and may call KVM_RUN again. 6008 - KVM_SYSTEM_EVENT_CRASH -- the guest crash occurred and the guest 6009 has requested a crash condition maintenance. Userspace can choose 6010 to ignore the request, or to gather VM memory core dump and/or 6011 reset/shutdown of the VM. 6012 6013Valid flags are: 6014 6015 - KVM_SYSTEM_EVENT_RESET_FLAG_PSCI_RESET2 (arm64 only) -- the guest issued 6016 a SYSTEM_RESET2 call according to v1.1 of the PSCI specification. 6017 6018:: 6019 6020 /* KVM_EXIT_IOAPIC_EOI */ 6021 struct { 6022 __u8 vector; 6023 } eoi; 6024 6025Indicates that the VCPU's in-kernel local APIC received an EOI for a 6026level-triggered IOAPIC interrupt. This exit only triggers when the 6027IOAPIC is implemented in userspace (i.e. KVM_CAP_SPLIT_IRQCHIP is enabled); 6028the userspace IOAPIC should process the EOI and retrigger the interrupt if 6029it is still asserted. Vector is the LAPIC interrupt vector for which the 6030EOI was received. 6031 6032:: 6033 6034 struct kvm_hyperv_exit { 6035 #define KVM_EXIT_HYPERV_SYNIC 1 6036 #define KVM_EXIT_HYPERV_HCALL 2 6037 #define KVM_EXIT_HYPERV_SYNDBG 3 6038 __u32 type; 6039 __u32 pad1; 6040 union { 6041 struct { 6042 __u32 msr; 6043 __u32 pad2; 6044 __u64 control; 6045 __u64 evt_page; 6046 __u64 msg_page; 6047 } synic; 6048 struct { 6049 __u64 input; 6050 __u64 result; 6051 __u64 params[2]; 6052 } hcall; 6053 struct { 6054 __u32 msr; 6055 __u32 pad2; 6056 __u64 control; 6057 __u64 status; 6058 __u64 send_page; 6059 __u64 recv_page; 6060 __u64 pending_page; 6061 } syndbg; 6062 } u; 6063 }; 6064 /* KVM_EXIT_HYPERV */ 6065 struct kvm_hyperv_exit hyperv; 6066 6067Indicates that the VCPU exits into userspace to process some tasks 6068related to Hyper-V emulation. 6069 6070Valid values for 'type' are: 6071 6072 - KVM_EXIT_HYPERV_SYNIC -- synchronously notify user-space about 6073 6074Hyper-V SynIC state change. Notification is used to remap SynIC 6075event/message pages and to enable/disable SynIC messages/events processing 6076in userspace. 6077 6078 - KVM_EXIT_HYPERV_SYNDBG -- synchronously notify user-space about 6079 6080Hyper-V Synthetic debugger state change. Notification is used to either update 6081the pending_page location or to send a control command (send the buffer located 6082in send_page or recv a buffer to recv_page). 6083 6084:: 6085 6086 /* KVM_EXIT_ARM_NISV */ 6087 struct { 6088 __u64 esr_iss; 6089 __u64 fault_ipa; 6090 } arm_nisv; 6091 6092Used on arm64 systems. If a guest accesses memory not in a memslot, 6093KVM will typically return to userspace and ask it to do MMIO emulation on its 6094behalf. However, for certain classes of instructions, no instruction decode 6095(direction, length of memory access) is provided, and fetching and decoding 6096the instruction from the VM is overly complicated to live in the kernel. 6097 6098Historically, when this situation occurred, KVM would print a warning and kill 6099the VM. KVM assumed that if the guest accessed non-memslot memory, it was 6100trying to do I/O, which just couldn't be emulated, and the warning message was 6101phrased accordingly. However, what happened more often was that a guest bug 6102caused access outside the guest memory areas which should lead to a more 6103meaningful warning message and an external abort in the guest, if the access 6104did not fall within an I/O window. 6105 6106Userspace implementations can query for KVM_CAP_ARM_NISV_TO_USER, and enable 6107this capability at VM creation. Once this is done, these types of errors will 6108instead return to userspace with KVM_EXIT_ARM_NISV, with the valid bits from 6109the ESR_EL2 in the esr_iss field, and the faulting IPA in the fault_ipa field. 6110Userspace can either fix up the access if it's actually an I/O access by 6111decoding the instruction from guest memory (if it's very brave) and continue 6112executing the guest, or it can decide to suspend, dump, or restart the guest. 6113 6114Note that KVM does not skip the faulting instruction as it does for 6115KVM_EXIT_MMIO, but userspace has to emulate any change to the processing state 6116if it decides to decode and emulate the instruction. 6117 6118:: 6119 6120 /* KVM_EXIT_X86_RDMSR / KVM_EXIT_X86_WRMSR */ 6121 struct { 6122 __u8 error; /* user -> kernel */ 6123 __u8 pad[7]; 6124 __u32 reason; /* kernel -> user */ 6125 __u32 index; /* kernel -> user */ 6126 __u64 data; /* kernel <-> user */ 6127 } msr; 6128 6129Used on x86 systems. When the VM capability KVM_CAP_X86_USER_SPACE_MSR is 6130enabled, MSR accesses to registers that would invoke a #GP by KVM kernel code 6131will instead trigger a KVM_EXIT_X86_RDMSR exit for reads and KVM_EXIT_X86_WRMSR 6132exit for writes. 6133 6134The "reason" field specifies why the MSR trap occurred. User space will only 6135receive MSR exit traps when a particular reason was requested during through 6136ENABLE_CAP. Currently valid exit reasons are: 6137 6138 KVM_MSR_EXIT_REASON_UNKNOWN - access to MSR that is unknown to KVM 6139 KVM_MSR_EXIT_REASON_INVAL - access to invalid MSRs or reserved bits 6140 KVM_MSR_EXIT_REASON_FILTER - access blocked by KVM_X86_SET_MSR_FILTER 6141 6142For KVM_EXIT_X86_RDMSR, the "index" field tells user space which MSR the guest 6143wants to read. To respond to this request with a successful read, user space 6144writes the respective data into the "data" field and must continue guest 6145execution to ensure the read data is transferred into guest register state. 6146 6147If the RDMSR request was unsuccessful, user space indicates that with a "1" in 6148the "error" field. This will inject a #GP into the guest when the VCPU is 6149executed again. 6150 6151For KVM_EXIT_X86_WRMSR, the "index" field tells user space which MSR the guest 6152wants to write. Once finished processing the event, user space must continue 6153vCPU execution. If the MSR write was unsuccessful, user space also sets the 6154"error" field to "1". 6155 6156:: 6157 6158 6159 struct kvm_xen_exit { 6160 #define KVM_EXIT_XEN_HCALL 1 6161 __u32 type; 6162 union { 6163 struct { 6164 __u32 longmode; 6165 __u32 cpl; 6166 __u64 input; 6167 __u64 result; 6168 __u64 params[6]; 6169 } hcall; 6170 } u; 6171 }; 6172 /* KVM_EXIT_XEN */ 6173 struct kvm_hyperv_exit xen; 6174 6175Indicates that the VCPU exits into userspace to process some tasks 6176related to Xen emulation. 6177 6178Valid values for 'type' are: 6179 6180 - KVM_EXIT_XEN_HCALL -- synchronously notify user-space about Xen hypercall. 6181 Userspace is expected to place the hypercall result into the appropriate 6182 field before invoking KVM_RUN again. 6183 6184:: 6185 6186 /* KVM_EXIT_RISCV_SBI */ 6187 struct { 6188 unsigned long extension_id; 6189 unsigned long function_id; 6190 unsigned long args[6]; 6191 unsigned long ret[2]; 6192 } riscv_sbi; 6193If exit reason is KVM_EXIT_RISCV_SBI then it indicates that the VCPU has 6194done a SBI call which is not handled by KVM RISC-V kernel module. The details 6195of the SBI call are available in 'riscv_sbi' member of kvm_run structure. The 6196'extension_id' field of 'riscv_sbi' represents SBI extension ID whereas the 6197'function_id' field represents function ID of given SBI extension. The 'args' 6198array field of 'riscv_sbi' represents parameters for the SBI call and 'ret' 6199array field represents return values. The userspace should update the return 6200values of SBI call before resuming the VCPU. For more details on RISC-V SBI 6201spec refer, https://github.com/riscv/riscv-sbi-doc. 6202 6203:: 6204 6205 /* Fix the size of the union. */ 6206 char padding[256]; 6207 }; 6208 6209 /* 6210 * shared registers between kvm and userspace. 6211 * kvm_valid_regs specifies the register classes set by the host 6212 * kvm_dirty_regs specified the register classes dirtied by userspace 6213 * struct kvm_sync_regs is architecture specific, as well as the 6214 * bits for kvm_valid_regs and kvm_dirty_regs 6215 */ 6216 __u64 kvm_valid_regs; 6217 __u64 kvm_dirty_regs; 6218 union { 6219 struct kvm_sync_regs regs; 6220 char padding[SYNC_REGS_SIZE_BYTES]; 6221 } s; 6222 6223If KVM_CAP_SYNC_REGS is defined, these fields allow userspace to access 6224certain guest registers without having to call SET/GET_*REGS. Thus we can 6225avoid some system call overhead if userspace has to handle the exit. 6226Userspace can query the validity of the structure by checking 6227kvm_valid_regs for specific bits. These bits are architecture specific 6228and usually define the validity of a groups of registers. (e.g. one bit 6229for general purpose registers) 6230 6231Please note that the kernel is allowed to use the kvm_run structure as the 6232primary storage for certain register types. Therefore, the kernel may use the 6233values in kvm_run even if the corresponding bit in kvm_dirty_regs is not set. 6234 6235:: 6236 6237 }; 6238 6239 6240 62416. Capabilities that can be enabled on vCPUs 6242============================================ 6243 6244There are certain capabilities that change the behavior of the virtual CPU or 6245the virtual machine when enabled. To enable them, please see section 4.37. 6246Below you can find a list of capabilities and what their effect on the vCPU or 6247the virtual machine is when enabling them. 6248 6249The following information is provided along with the description: 6250 6251 Architectures: 6252 which instruction set architectures provide this ioctl. 6253 x86 includes both i386 and x86_64. 6254 6255 Target: 6256 whether this is a per-vcpu or per-vm capability. 6257 6258 Parameters: 6259 what parameters are accepted by the capability. 6260 6261 Returns: 6262 the return value. General error numbers (EBADF, ENOMEM, EINVAL) 6263 are not detailed, but errors with specific meanings are. 6264 6265 62666.1 KVM_CAP_PPC_OSI 6267------------------- 6268 6269:Architectures: ppc 6270:Target: vcpu 6271:Parameters: none 6272:Returns: 0 on success; -1 on error 6273 6274This capability enables interception of OSI hypercalls that otherwise would 6275be treated as normal system calls to be injected into the guest. OSI hypercalls 6276were invented by Mac-on-Linux to have a standardized communication mechanism 6277between the guest and the host. 6278 6279When this capability is enabled, KVM_EXIT_OSI can occur. 6280 6281 62826.2 KVM_CAP_PPC_PAPR 6283-------------------- 6284 6285:Architectures: ppc 6286:Target: vcpu 6287:Parameters: none 6288:Returns: 0 on success; -1 on error 6289 6290This capability enables interception of PAPR hypercalls. PAPR hypercalls are 6291done using the hypercall instruction "sc 1". 6292 6293It also sets the guest privilege level to "supervisor" mode. Usually the guest 6294runs in "hypervisor" privilege mode with a few missing features. 6295 6296In addition to the above, it changes the semantics of SDR1. In this mode, the 6297HTAB address part of SDR1 contains an HVA instead of a GPA, as PAPR keeps the 6298HTAB invisible to the guest. 6299 6300When this capability is enabled, KVM_EXIT_PAPR_HCALL can occur. 6301 6302 63036.3 KVM_CAP_SW_TLB 6304------------------ 6305 6306:Architectures: ppc 6307:Target: vcpu 6308:Parameters: args[0] is the address of a struct kvm_config_tlb 6309:Returns: 0 on success; -1 on error 6310 6311:: 6312 6313 struct kvm_config_tlb { 6314 __u64 params; 6315 __u64 array; 6316 __u32 mmu_type; 6317 __u32 array_len; 6318 }; 6319 6320Configures the virtual CPU's TLB array, establishing a shared memory area 6321between userspace and KVM. The "params" and "array" fields are userspace 6322addresses of mmu-type-specific data structures. The "array_len" field is an 6323safety mechanism, and should be set to the size in bytes of the memory that 6324userspace has reserved for the array. It must be at least the size dictated 6325by "mmu_type" and "params". 6326 6327While KVM_RUN is active, the shared region is under control of KVM. Its 6328contents are undefined, and any modification by userspace results in 6329boundedly undefined behavior. 6330 6331On return from KVM_RUN, the shared region will reflect the current state of 6332the guest's TLB. If userspace makes any changes, it must call KVM_DIRTY_TLB 6333to tell KVM which entries have been changed, prior to calling KVM_RUN again 6334on this vcpu. 6335 6336For mmu types KVM_MMU_FSL_BOOKE_NOHV and KVM_MMU_FSL_BOOKE_HV: 6337 6338 - The "params" field is of type "struct kvm_book3e_206_tlb_params". 6339 - The "array" field points to an array of type "struct 6340 kvm_book3e_206_tlb_entry". 6341 - The array consists of all entries in the first TLB, followed by all 6342 entries in the second TLB. 6343 - Within a TLB, entries are ordered first by increasing set number. Within a 6344 set, entries are ordered by way (increasing ESEL). 6345 - The hash for determining set number in TLB0 is: (MAS2 >> 12) & (num_sets - 1) 6346 where "num_sets" is the tlb_sizes[] value divided by the tlb_ways[] value. 6347 - The tsize field of mas1 shall be set to 4K on TLB0, even though the 6348 hardware ignores this value for TLB0. 6349 63506.4 KVM_CAP_S390_CSS_SUPPORT 6351---------------------------- 6352 6353:Architectures: s390 6354:Target: vcpu 6355:Parameters: none 6356:Returns: 0 on success; -1 on error 6357 6358This capability enables support for handling of channel I/O instructions. 6359 6360TEST PENDING INTERRUPTION and the interrupt portion of TEST SUBCHANNEL are 6361handled in-kernel, while the other I/O instructions are passed to userspace. 6362 6363When this capability is enabled, KVM_EXIT_S390_TSCH will occur on TEST 6364SUBCHANNEL intercepts. 6365 6366Note that even though this capability is enabled per-vcpu, the complete 6367virtual machine is affected. 6368 63696.5 KVM_CAP_PPC_EPR 6370------------------- 6371 6372:Architectures: ppc 6373:Target: vcpu 6374:Parameters: args[0] defines whether the proxy facility is active 6375:Returns: 0 on success; -1 on error 6376 6377This capability enables or disables the delivery of interrupts through the 6378external proxy facility. 6379 6380When enabled (args[0] != 0), every time the guest gets an external interrupt 6381delivered, it automatically exits into user space with a KVM_EXIT_EPR exit 6382to receive the topmost interrupt vector. 6383 6384When disabled (args[0] == 0), behavior is as if this facility is unsupported. 6385 6386When this capability is enabled, KVM_EXIT_EPR can occur. 6387 63886.6 KVM_CAP_IRQ_MPIC 6389-------------------- 6390 6391:Architectures: ppc 6392:Parameters: args[0] is the MPIC device fd; 6393 args[1] is the MPIC CPU number for this vcpu 6394 6395This capability connects the vcpu to an in-kernel MPIC device. 6396 63976.7 KVM_CAP_IRQ_XICS 6398-------------------- 6399 6400:Architectures: ppc 6401:Target: vcpu 6402:Parameters: args[0] is the XICS device fd; 6403 args[1] is the XICS CPU number (server ID) for this vcpu 6404 6405This capability connects the vcpu to an in-kernel XICS device. 6406 64076.8 KVM_CAP_S390_IRQCHIP 6408------------------------ 6409 6410:Architectures: s390 6411:Target: vm 6412:Parameters: none 6413 6414This capability enables the in-kernel irqchip for s390. Please refer to 6415"4.24 KVM_CREATE_IRQCHIP" for details. 6416 64176.9 KVM_CAP_MIPS_FPU 6418-------------------- 6419 6420:Architectures: mips 6421:Target: vcpu 6422:Parameters: args[0] is reserved for future use (should be 0). 6423 6424This capability allows the use of the host Floating Point Unit by the guest. It 6425allows the Config1.FP bit to be set to enable the FPU in the guest. Once this is 6426done the ``KVM_REG_MIPS_FPR_*`` and ``KVM_REG_MIPS_FCR_*`` registers can be 6427accessed (depending on the current guest FPU register mode), and the Status.FR, 6428Config5.FRE bits are accessible via the KVM API and also from the guest, 6429depending on them being supported by the FPU. 6430 64316.10 KVM_CAP_MIPS_MSA 6432--------------------- 6433 6434:Architectures: mips 6435:Target: vcpu 6436:Parameters: args[0] is reserved for future use (should be 0). 6437 6438This capability allows the use of the MIPS SIMD Architecture (MSA) by the guest. 6439It allows the Config3.MSAP bit to be set to enable the use of MSA by the guest. 6440Once this is done the ``KVM_REG_MIPS_VEC_*`` and ``KVM_REG_MIPS_MSA_*`` 6441registers can be accessed, and the Config5.MSAEn bit is accessible via the 6442KVM API and also from the guest. 6443 64446.74 KVM_CAP_SYNC_REGS 6445---------------------- 6446 6447:Architectures: s390, x86 6448:Target: s390: always enabled, x86: vcpu 6449:Parameters: none 6450:Returns: x86: KVM_CHECK_EXTENSION returns a bit-array indicating which register 6451 sets are supported 6452 (bitfields defined in arch/x86/include/uapi/asm/kvm.h). 6453 6454As described above in the kvm_sync_regs struct info in section 5 (kvm_run): 6455KVM_CAP_SYNC_REGS "allow[s] userspace to access certain guest registers 6456without having to call SET/GET_*REGS". This reduces overhead by eliminating 6457repeated ioctl calls for setting and/or getting register values. This is 6458particularly important when userspace is making synchronous guest state 6459modifications, e.g. when emulating and/or intercepting instructions in 6460userspace. 6461 6462For s390 specifics, please refer to the source code. 6463 6464For x86: 6465 6466- the register sets to be copied out to kvm_run are selectable 6467 by userspace (rather that all sets being copied out for every exit). 6468- vcpu_events are available in addition to regs and sregs. 6469 6470For x86, the 'kvm_valid_regs' field of struct kvm_run is overloaded to 6471function as an input bit-array field set by userspace to indicate the 6472specific register sets to be copied out on the next exit. 6473 6474To indicate when userspace has modified values that should be copied into 6475the vCPU, the all architecture bitarray field, 'kvm_dirty_regs' must be set. 6476This is done using the same bitflags as for the 'kvm_valid_regs' field. 6477If the dirty bit is not set, then the register set values will not be copied 6478into the vCPU even if they've been modified. 6479 6480Unused bitfields in the bitarrays must be set to zero. 6481 6482:: 6483 6484 struct kvm_sync_regs { 6485 struct kvm_regs regs; 6486 struct kvm_sregs sregs; 6487 struct kvm_vcpu_events events; 6488 }; 6489 64906.75 KVM_CAP_PPC_IRQ_XIVE 6491------------------------- 6492 6493:Architectures: ppc 6494:Target: vcpu 6495:Parameters: args[0] is the XIVE device fd; 6496 args[1] is the XIVE CPU number (server ID) for this vcpu 6497 6498This capability connects the vcpu to an in-kernel XIVE device. 6499 65007. Capabilities that can be enabled on VMs 6501========================================== 6502 6503There are certain capabilities that change the behavior of the virtual 6504machine when enabled. To enable them, please see section 4.37. Below 6505you can find a list of capabilities and what their effect on the VM 6506is when enabling them. 6507 6508The following information is provided along with the description: 6509 6510 Architectures: 6511 which instruction set architectures provide this ioctl. 6512 x86 includes both i386 and x86_64. 6513 6514 Parameters: 6515 what parameters are accepted by the capability. 6516 6517 Returns: 6518 the return value. General error numbers (EBADF, ENOMEM, EINVAL) 6519 are not detailed, but errors with specific meanings are. 6520 6521 65227.1 KVM_CAP_PPC_ENABLE_HCALL 6523---------------------------- 6524 6525:Architectures: ppc 6526:Parameters: args[0] is the sPAPR hcall number; 6527 args[1] is 0 to disable, 1 to enable in-kernel handling 6528 6529This capability controls whether individual sPAPR hypercalls (hcalls) 6530get handled by the kernel or not. Enabling or disabling in-kernel 6531handling of an hcall is effective across the VM. On creation, an 6532initial set of hcalls are enabled for in-kernel handling, which 6533consists of those hcalls for which in-kernel handlers were implemented 6534before this capability was implemented. If disabled, the kernel will 6535not to attempt to handle the hcall, but will always exit to userspace 6536to handle it. Note that it may not make sense to enable some and 6537disable others of a group of related hcalls, but KVM does not prevent 6538userspace from doing that. 6539 6540If the hcall number specified is not one that has an in-kernel 6541implementation, the KVM_ENABLE_CAP ioctl will fail with an EINVAL 6542error. 6543 65447.2 KVM_CAP_S390_USER_SIGP 6545-------------------------- 6546 6547:Architectures: s390 6548:Parameters: none 6549 6550This capability controls which SIGP orders will be handled completely in user 6551space. With this capability enabled, all fast orders will be handled completely 6552in the kernel: 6553 6554- SENSE 6555- SENSE RUNNING 6556- EXTERNAL CALL 6557- EMERGENCY SIGNAL 6558- CONDITIONAL EMERGENCY SIGNAL 6559 6560All other orders will be handled completely in user space. 6561 6562Only privileged operation exceptions will be checked for in the kernel (or even 6563in the hardware prior to interception). If this capability is not enabled, the 6564old way of handling SIGP orders is used (partially in kernel and user space). 6565 65667.3 KVM_CAP_S390_VECTOR_REGISTERS 6567--------------------------------- 6568 6569:Architectures: s390 6570:Parameters: none 6571:Returns: 0 on success, negative value on error 6572 6573Allows use of the vector registers introduced with z13 processor, and 6574provides for the synchronization between host and user space. Will 6575return -EINVAL if the machine does not support vectors. 6576 65777.4 KVM_CAP_S390_USER_STSI 6578-------------------------- 6579 6580:Architectures: s390 6581:Parameters: none 6582 6583This capability allows post-handlers for the STSI instruction. After 6584initial handling in the kernel, KVM exits to user space with 6585KVM_EXIT_S390_STSI to allow user space to insert further data. 6586 6587Before exiting to userspace, kvm handlers should fill in s390_stsi field of 6588vcpu->run:: 6589 6590 struct { 6591 __u64 addr; 6592 __u8 ar; 6593 __u8 reserved; 6594 __u8 fc; 6595 __u8 sel1; 6596 __u16 sel2; 6597 } s390_stsi; 6598 6599 @addr - guest address of STSI SYSIB 6600 @fc - function code 6601 @sel1 - selector 1 6602 @sel2 - selector 2 6603 @ar - access register number 6604 6605KVM handlers should exit to userspace with rc = -EREMOTE. 6606 66077.5 KVM_CAP_SPLIT_IRQCHIP 6608------------------------- 6609 6610:Architectures: x86 6611:Parameters: args[0] - number of routes reserved for userspace IOAPICs 6612:Returns: 0 on success, -1 on error 6613 6614Create a local apic for each processor in the kernel. This can be used 6615instead of KVM_CREATE_IRQCHIP if the userspace VMM wishes to emulate the 6616IOAPIC and PIC (and also the PIT, even though this has to be enabled 6617separately). 6618 6619This capability also enables in kernel routing of interrupt requests; 6620when KVM_CAP_SPLIT_IRQCHIP only routes of KVM_IRQ_ROUTING_MSI type are 6621used in the IRQ routing table. The first args[0] MSI routes are reserved 6622for the IOAPIC pins. Whenever the LAPIC receives an EOI for these routes, 6623a KVM_EXIT_IOAPIC_EOI vmexit will be reported to userspace. 6624 6625Fails if VCPU has already been created, or if the irqchip is already in the 6626kernel (i.e. KVM_CREATE_IRQCHIP has already been called). 6627 66287.6 KVM_CAP_S390_RI 6629------------------- 6630 6631:Architectures: s390 6632:Parameters: none 6633 6634Allows use of runtime-instrumentation introduced with zEC12 processor. 6635Will return -EINVAL if the machine does not support runtime-instrumentation. 6636Will return -EBUSY if a VCPU has already been created. 6637 66387.7 KVM_CAP_X2APIC_API 6639---------------------- 6640 6641:Architectures: x86 6642:Parameters: args[0] - features that should be enabled 6643:Returns: 0 on success, -EINVAL when args[0] contains invalid features 6644 6645Valid feature flags in args[0] are:: 6646 6647 #define KVM_X2APIC_API_USE_32BIT_IDS (1ULL << 0) 6648 #define KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK (1ULL << 1) 6649 6650Enabling KVM_X2APIC_API_USE_32BIT_IDS changes the behavior of 6651KVM_SET_GSI_ROUTING, KVM_SIGNAL_MSI, KVM_SET_LAPIC, and KVM_GET_LAPIC, 6652allowing the use of 32-bit APIC IDs. See KVM_CAP_X2APIC_API in their 6653respective sections. 6654 6655KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK must be enabled for x2APIC to work 6656in logical mode or with more than 255 VCPUs. Otherwise, KVM treats 0xff 6657as a broadcast even in x2APIC mode in order to support physical x2APIC 6658without interrupt remapping. This is undesirable in logical mode, 6659where 0xff represents CPUs 0-7 in cluster 0. 6660 66617.8 KVM_CAP_S390_USER_INSTR0 6662---------------------------- 6663 6664:Architectures: s390 6665:Parameters: none 6666 6667With this capability enabled, all illegal instructions 0x0000 (2 bytes) will 6668be intercepted and forwarded to user space. User space can use this 6669mechanism e.g. to realize 2-byte software breakpoints. The kernel will 6670not inject an operating exception for these instructions, user space has 6671to take care of that. 6672 6673This capability can be enabled dynamically even if VCPUs were already 6674created and are running. 6675 66767.9 KVM_CAP_S390_GS 6677------------------- 6678 6679:Architectures: s390 6680:Parameters: none 6681:Returns: 0 on success; -EINVAL if the machine does not support 6682 guarded storage; -EBUSY if a VCPU has already been created. 6683 6684Allows use of guarded storage for the KVM guest. 6685 66867.10 KVM_CAP_S390_AIS 6687--------------------- 6688 6689:Architectures: s390 6690:Parameters: none 6691 6692Allow use of adapter-interruption suppression. 6693:Returns: 0 on success; -EBUSY if a VCPU has already been created. 6694 66957.11 KVM_CAP_PPC_SMT 6696-------------------- 6697 6698:Architectures: ppc 6699:Parameters: vsmt_mode, flags 6700 6701Enabling this capability on a VM provides userspace with a way to set 6702the desired virtual SMT mode (i.e. the number of virtual CPUs per 6703virtual core). The virtual SMT mode, vsmt_mode, must be a power of 2 6704between 1 and 8. On POWER8, vsmt_mode must also be no greater than 6705the number of threads per subcore for the host. Currently flags must 6706be 0. A successful call to enable this capability will result in 6707vsmt_mode being returned when the KVM_CAP_PPC_SMT capability is 6708subsequently queried for the VM. This capability is only supported by 6709HV KVM, and can only be set before any VCPUs have been created. 6710The KVM_CAP_PPC_SMT_POSSIBLE capability indicates which virtual SMT 6711modes are available. 6712 67137.12 KVM_CAP_PPC_FWNMI 6714---------------------- 6715 6716:Architectures: ppc 6717:Parameters: none 6718 6719With this capability a machine check exception in the guest address 6720space will cause KVM to exit the guest with NMI exit reason. This 6721enables QEMU to build error log and branch to guest kernel registered 6722machine check handling routine. Without this capability KVM will 6723branch to guests' 0x200 interrupt vector. 6724 67257.13 KVM_CAP_X86_DISABLE_EXITS 6726------------------------------ 6727 6728:Architectures: x86 6729:Parameters: args[0] defines which exits are disabled 6730:Returns: 0 on success, -EINVAL when args[0] contains invalid exits 6731 6732Valid bits in args[0] are:: 6733 6734 #define KVM_X86_DISABLE_EXITS_MWAIT (1 << 0) 6735 #define KVM_X86_DISABLE_EXITS_HLT (1 << 1) 6736 #define KVM_X86_DISABLE_EXITS_PAUSE (1 << 2) 6737 #define KVM_X86_DISABLE_EXITS_CSTATE (1 << 3) 6738 6739Enabling this capability on a VM provides userspace with a way to no 6740longer intercept some instructions for improved latency in some 6741workloads, and is suggested when vCPUs are associated to dedicated 6742physical CPUs. More bits can be added in the future; userspace can 6743just pass the KVM_CHECK_EXTENSION result to KVM_ENABLE_CAP to disable 6744all such vmexits. 6745 6746Do not enable KVM_FEATURE_PV_UNHALT if you disable HLT exits. 6747 67487.14 KVM_CAP_S390_HPAGE_1M 6749-------------------------- 6750 6751:Architectures: s390 6752:Parameters: none 6753:Returns: 0 on success, -EINVAL if hpage module parameter was not set 6754 or cmma is enabled, or the VM has the KVM_VM_S390_UCONTROL 6755 flag set 6756 6757With this capability the KVM support for memory backing with 1m pages 6758through hugetlbfs can be enabled for a VM. After the capability is 6759enabled, cmma can't be enabled anymore and pfmfi and the storage key 6760interpretation are disabled. If cmma has already been enabled or the 6761hpage module parameter is not set to 1, -EINVAL is returned. 6762 6763While it is generally possible to create a huge page backed VM without 6764this capability, the VM will not be able to run. 6765 67667.15 KVM_CAP_MSR_PLATFORM_INFO 6767------------------------------ 6768 6769:Architectures: x86 6770:Parameters: args[0] whether feature should be enabled or not 6771 6772With this capability, a guest may read the MSR_PLATFORM_INFO MSR. Otherwise, 6773a #GP would be raised when the guest tries to access. Currently, this 6774capability does not enable write permissions of this MSR for the guest. 6775 67767.16 KVM_CAP_PPC_NESTED_HV 6777-------------------------- 6778 6779:Architectures: ppc 6780:Parameters: none 6781:Returns: 0 on success, -EINVAL when the implementation doesn't support 6782 nested-HV virtualization. 6783 6784HV-KVM on POWER9 and later systems allows for "nested-HV" 6785virtualization, which provides a way for a guest VM to run guests that 6786can run using the CPU's supervisor mode (privileged non-hypervisor 6787state). Enabling this capability on a VM depends on the CPU having 6788the necessary functionality and on the facility being enabled with a 6789kvm-hv module parameter. 6790 67917.17 KVM_CAP_EXCEPTION_PAYLOAD 6792------------------------------ 6793 6794:Architectures: x86 6795:Parameters: args[0] whether feature should be enabled or not 6796 6797With this capability enabled, CR2 will not be modified prior to the 6798emulated VM-exit when L1 intercepts a #PF exception that occurs in 6799L2. Similarly, for kvm-intel only, DR6 will not be modified prior to 6800the emulated VM-exit when L1 intercepts a #DB exception that occurs in 6801L2. As a result, when KVM_GET_VCPU_EVENTS reports a pending #PF (or 6802#DB) exception for L2, exception.has_payload will be set and the 6803faulting address (or the new DR6 bits*) will be reported in the 6804exception_payload field. Similarly, when userspace injects a #PF (or 6805#DB) into L2 using KVM_SET_VCPU_EVENTS, it is expected to set 6806exception.has_payload and to put the faulting address - or the new DR6 6807bits\ [#]_ - in the exception_payload field. 6808 6809This capability also enables exception.pending in struct 6810kvm_vcpu_events, which allows userspace to distinguish between pending 6811and injected exceptions. 6812 6813 6814.. [#] For the new DR6 bits, note that bit 16 is set iff the #DB exception 6815 will clear DR6.RTM. 6816 68177.18 KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 6818 6819:Architectures: x86, arm64, mips 6820:Parameters: args[0] whether feature should be enabled or not 6821 6822Valid flags are:: 6823 6824 #define KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE (1 << 0) 6825 #define KVM_DIRTY_LOG_INITIALLY_SET (1 << 1) 6826 6827With KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE is set, KVM_GET_DIRTY_LOG will not 6828automatically clear and write-protect all pages that are returned as dirty. 6829Rather, userspace will have to do this operation separately using 6830KVM_CLEAR_DIRTY_LOG. 6831 6832At the cost of a slightly more complicated operation, this provides better 6833scalability and responsiveness for two reasons. First, 6834KVM_CLEAR_DIRTY_LOG ioctl can operate on a 64-page granularity rather 6835than requiring to sync a full memslot; this ensures that KVM does not 6836take spinlocks for an extended period of time. Second, in some cases a 6837large amount of time can pass between a call to KVM_GET_DIRTY_LOG and 6838userspace actually using the data in the page. Pages can be modified 6839during this time, which is inefficient for both the guest and userspace: 6840the guest will incur a higher penalty due to write protection faults, 6841while userspace can see false reports of dirty pages. Manual reprotection 6842helps reducing this time, improving guest performance and reducing the 6843number of dirty log false positives. 6844 6845With KVM_DIRTY_LOG_INITIALLY_SET set, all the bits of the dirty bitmap 6846will be initialized to 1 when created. This also improves performance because 6847dirty logging can be enabled gradually in small chunks on the first call 6848to KVM_CLEAR_DIRTY_LOG. KVM_DIRTY_LOG_INITIALLY_SET depends on 6849KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE (it is also only available on 6850x86 and arm64 for now). 6851 6852KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 was previously available under the name 6853KVM_CAP_MANUAL_DIRTY_LOG_PROTECT, but the implementation had bugs that make 6854it hard or impossible to use it correctly. The availability of 6855KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2 signals that those bugs are fixed. 6856Userspace should not try to use KVM_CAP_MANUAL_DIRTY_LOG_PROTECT. 6857 68587.19 KVM_CAP_PPC_SECURE_GUEST 6859------------------------------ 6860 6861:Architectures: ppc 6862 6863This capability indicates that KVM is running on a host that has 6864ultravisor firmware and thus can support a secure guest. On such a 6865system, a guest can ask the ultravisor to make it a secure guest, 6866one whose memory is inaccessible to the host except for pages which 6867are explicitly requested to be shared with the host. The ultravisor 6868notifies KVM when a guest requests to become a secure guest, and KVM 6869has the opportunity to veto the transition. 6870 6871If present, this capability can be enabled for a VM, meaning that KVM 6872will allow the transition to secure guest mode. Otherwise KVM will 6873veto the transition. 6874 68757.20 KVM_CAP_HALT_POLL 6876---------------------- 6877 6878:Architectures: all 6879:Target: VM 6880:Parameters: args[0] is the maximum poll time in nanoseconds 6881:Returns: 0 on success; -1 on error 6882 6883This capability overrides the kvm module parameter halt_poll_ns for the 6884target VM. 6885 6886VCPU polling allows a VCPU to poll for wakeup events instead of immediately 6887scheduling during guest halts. The maximum time a VCPU can spend polling is 6888controlled by the kvm module parameter halt_poll_ns. This capability allows 6889the maximum halt time to specified on a per-VM basis, effectively overriding 6890the module parameter for the target VM. 6891 68927.21 KVM_CAP_X86_USER_SPACE_MSR 6893------------------------------- 6894 6895:Architectures: x86 6896:Target: VM 6897:Parameters: args[0] contains the mask of KVM_MSR_EXIT_REASON_* events to report 6898:Returns: 0 on success; -1 on error 6899 6900This capability enables trapping of #GP invoking RDMSR and WRMSR instructions 6901into user space. 6902 6903When a guest requests to read or write an MSR, KVM may not implement all MSRs 6904that are relevant to a respective system. It also does not differentiate by 6905CPU type. 6906 6907To allow more fine grained control over MSR handling, user space may enable 6908this capability. With it enabled, MSR accesses that match the mask specified in 6909args[0] and trigger a #GP event inside the guest by KVM will instead trigger 6910KVM_EXIT_X86_RDMSR and KVM_EXIT_X86_WRMSR exit notifications which user space 6911can then handle to implement model specific MSR handling and/or user notifications 6912to inform a user that an MSR was not handled. 6913 69147.22 KVM_CAP_X86_BUS_LOCK_EXIT 6915------------------------------- 6916 6917:Architectures: x86 6918:Target: VM 6919:Parameters: args[0] defines the policy used when bus locks detected in guest 6920:Returns: 0 on success, -EINVAL when args[0] contains invalid bits 6921 6922Valid bits in args[0] are:: 6923 6924 #define KVM_BUS_LOCK_DETECTION_OFF (1 << 0) 6925 #define KVM_BUS_LOCK_DETECTION_EXIT (1 << 1) 6926 6927Enabling this capability on a VM provides userspace with a way to select 6928a policy to handle the bus locks detected in guest. Userspace can obtain 6929the supported modes from the result of KVM_CHECK_EXTENSION and define it 6930through the KVM_ENABLE_CAP. 6931 6932KVM_BUS_LOCK_DETECTION_OFF and KVM_BUS_LOCK_DETECTION_EXIT are supported 6933currently and mutually exclusive with each other. More bits can be added in 6934the future. 6935 6936With KVM_BUS_LOCK_DETECTION_OFF set, bus locks in guest will not cause vm exits 6937so that no additional actions are needed. This is the default mode. 6938 6939With KVM_BUS_LOCK_DETECTION_EXIT set, vm exits happen when bus lock detected 6940in VM. KVM just exits to userspace when handling them. Userspace can enforce 6941its own throttling or other policy based mitigations. 6942 6943This capability is aimed to address the thread that VM can exploit bus locks to 6944degree the performance of the whole system. Once the userspace enable this 6945capability and select the KVM_BUS_LOCK_DETECTION_EXIT mode, KVM will set the 6946KVM_RUN_BUS_LOCK flag in vcpu-run->flags field and exit to userspace. Concerning 6947the bus lock vm exit can be preempted by a higher priority VM exit, the exit 6948notifications to userspace can be KVM_EXIT_BUS_LOCK or other reasons. 6949KVM_RUN_BUS_LOCK flag is used to distinguish between them. 6950 69517.23 KVM_CAP_PPC_DAWR1 6952---------------------- 6953 6954:Architectures: ppc 6955:Parameters: none 6956:Returns: 0 on success, -EINVAL when CPU doesn't support 2nd DAWR 6957 6958This capability can be used to check / enable 2nd DAWR feature provided 6959by POWER10 processor. 6960 6961 69627.24 KVM_CAP_VM_COPY_ENC_CONTEXT_FROM 6963------------------------------------- 6964 6965Architectures: x86 SEV enabled 6966Type: vm 6967Parameters: args[0] is the fd of the source vm 6968Returns: 0 on success; ENOTTY on error 6969 6970This capability enables userspace to copy encryption context from the vm 6971indicated by the fd to the vm this is called on. 6972 6973This is intended to support in-guest workloads scheduled by the host. This 6974allows the in-guest workload to maintain its own NPTs and keeps the two vms 6975from accidentally clobbering each other with interrupts and the like (separate 6976APIC/MSRs/etc). 6977 69787.25 KVM_CAP_SGX_ATTRIBUTE 6979-------------------------- 6980 6981:Architectures: x86 6982:Target: VM 6983:Parameters: args[0] is a file handle of a SGX attribute file in securityfs 6984:Returns: 0 on success, -EINVAL if the file handle is invalid or if a requested 6985 attribute is not supported by KVM. 6986 6987KVM_CAP_SGX_ATTRIBUTE enables a userspace VMM to grant a VM access to one or 6988more priveleged enclave attributes. args[0] must hold a file handle to a valid 6989SGX attribute file corresponding to an attribute that is supported/restricted 6990by KVM (currently only PROVISIONKEY). 6991 6992The SGX subsystem restricts access to a subset of enclave attributes to provide 6993additional security for an uncompromised kernel, e.g. use of the PROVISIONKEY 6994is restricted to deter malware from using the PROVISIONKEY to obtain a stable 6995system fingerprint. To prevent userspace from circumventing such restrictions 6996by running an enclave in a VM, KVM prevents access to privileged attributes by 6997default. 6998 6999See Documentation/x86/sgx.rst for more details. 7000 70017.26 KVM_CAP_PPC_RPT_INVALIDATE 7002------------------------------- 7003 7004:Capability: KVM_CAP_PPC_RPT_INVALIDATE 7005:Architectures: ppc 7006:Type: vm 7007 7008This capability indicates that the kernel is capable of handling 7009H_RPT_INVALIDATE hcall. 7010 7011In order to enable the use of H_RPT_INVALIDATE in the guest, 7012user space might have to advertise it for the guest. For example, 7013IBM pSeries (sPAPR) guest starts using it if "hcall-rpt-invalidate" is 7014present in the "ibm,hypertas-functions" device-tree property. 7015 7016This capability is enabled for hypervisors on platforms like POWER9 7017that support radix MMU. 7018 70197.27 KVM_CAP_EXIT_ON_EMULATION_FAILURE 7020-------------------------------------- 7021 7022:Architectures: x86 7023:Parameters: args[0] whether the feature should be enabled or not 7024 7025When this capability is enabled, an emulation failure will result in an exit 7026to userspace with KVM_INTERNAL_ERROR (except when the emulator was invoked 7027to handle a VMware backdoor instruction). Furthermore, KVM will now provide up 7028to 15 instruction bytes for any exit to userspace resulting from an emulation 7029failure. When these exits to userspace occur use the emulation_failure struct 7030instead of the internal struct. They both have the same layout, but the 7031emulation_failure struct matches the content better. It also explicitly 7032defines the 'flags' field which is used to describe the fields in the struct 7033that are valid (ie: if KVM_INTERNAL_ERROR_EMULATION_FLAG_INSTRUCTION_BYTES is 7034set in the 'flags' field then both 'insn_size' and 'insn_bytes' have valid data 7035in them.) 7036 70377.28 KVM_CAP_ARM_MTE 7038-------------------- 7039 7040:Architectures: arm64 7041:Parameters: none 7042 7043This capability indicates that KVM (and the hardware) supports exposing the 7044Memory Tagging Extensions (MTE) to the guest. It must also be enabled by the 7045VMM before creating any VCPUs to allow the guest access. Note that MTE is only 7046available to a guest running in AArch64 mode and enabling this capability will 7047cause attempts to create AArch32 VCPUs to fail. 7048 7049When enabled the guest is able to access tags associated with any memory given 7050to the guest. KVM will ensure that the tags are maintained during swap or 7051hibernation of the host; however the VMM needs to manually save/restore the 7052tags as appropriate if the VM is migrated. 7053 7054When this capability is enabled all memory in memslots must be mapped as 7055not-shareable (no MAP_SHARED), attempts to create a memslot with a 7056MAP_SHARED mmap will result in an -EINVAL return. 7057 7058When enabled the VMM may make use of the ``KVM_ARM_MTE_COPY_TAGS`` ioctl to 7059perform a bulk copy of tags to/from the guest. 7060 70617.29 KVM_CAP_VM_MOVE_ENC_CONTEXT_FROM 7062------------------------------------- 7063 7064Architectures: x86 SEV enabled 7065Type: vm 7066Parameters: args[0] is the fd of the source vm 7067Returns: 0 on success 7068 7069This capability enables userspace to migrate the encryption context from the VM 7070indicated by the fd to the VM this is called on. 7071 7072This is intended to support intra-host migration of VMs between userspace VMMs, 7073upgrading the VMM process without interrupting the guest. 7074 70757.30 KVM_CAP_PPC_AIL_MODE_3 7076------------------------------- 7077 7078:Capability: KVM_CAP_PPC_AIL_MODE_3 7079:Architectures: ppc 7080:Type: vm 7081 7082This capability indicates that the kernel supports the mode 3 setting for the 7083"Address Translation Mode on Interrupt" aka "Alternate Interrupt Location" 7084resource that is controlled with the H_SET_MODE hypercall. 7085 7086This capability allows a guest kernel to use a better-performance mode for 7087handling interrupts and system calls. 7088 70897.31 KVM_CAP_DISABLE_QUIRKS2 7090---------------------------- 7091 7092:Capability: KVM_CAP_DISABLE_QUIRKS2 7093:Parameters: args[0] - set of KVM quirks to disable 7094:Architectures: x86 7095:Type: vm 7096 7097This capability, if enabled, will cause KVM to disable some behavior 7098quirks. 7099 7100Calling KVM_CHECK_EXTENSION for this capability returns a bitmask of 7101quirks that can be disabled in KVM. 7102 7103The argument to KVM_ENABLE_CAP for this capability is a bitmask of 7104quirks to disable, and must be a subset of the bitmask returned by 7105KVM_CHECK_EXTENSION. 7106 7107The valid bits in cap.args[0] are: 7108 7109=================================== ============================================ 7110 KVM_X86_QUIRK_LINT0_REENABLED By default, the reset value for the LVT 7111 LINT0 register is 0x700 (APIC_MODE_EXTINT). 7112 When this quirk is disabled, the reset value 7113 is 0x10000 (APIC_LVT_MASKED). 7114 7115 KVM_X86_QUIRK_CD_NW_CLEARED By default, KVM clears CR0.CD and CR0.NW. 7116 When this quirk is disabled, KVM does not 7117 change the value of CR0.CD and CR0.NW. 7118 7119 KVM_X86_QUIRK_LAPIC_MMIO_HOLE By default, the MMIO LAPIC interface is 7120 available even when configured for x2APIC 7121 mode. When this quirk is disabled, KVM 7122 disables the MMIO LAPIC interface if the 7123 LAPIC is in x2APIC mode. 7124 7125 KVM_X86_QUIRK_OUT_7E_INC_RIP By default, KVM pre-increments %rip before 7126 exiting to userspace for an OUT instruction 7127 to port 0x7e. When this quirk is disabled, 7128 KVM does not pre-increment %rip before 7129 exiting to userspace. 7130 7131 KVM_X86_QUIRK_MISC_ENABLE_NO_MWAIT When this quirk is disabled, KVM sets 7132 CPUID.01H:ECX[bit 3] (MONITOR/MWAIT) if 7133 IA32_MISC_ENABLE[bit 18] (MWAIT) is set. 7134 Additionally, when this quirk is disabled, 7135 KVM clears CPUID.01H:ECX[bit 3] if 7136 IA32_MISC_ENABLE[bit 18] is cleared. 7137=================================== ============================================ 7138 71398. Other capabilities. 7140====================== 7141 7142This section lists capabilities that give information about other 7143features of the KVM implementation. 7144 71458.1 KVM_CAP_PPC_HWRNG 7146--------------------- 7147 7148:Architectures: ppc 7149 7150This capability, if KVM_CHECK_EXTENSION indicates that it is 7151available, means that the kernel has an implementation of the 7152H_RANDOM hypercall backed by a hardware random-number generator. 7153If present, the kernel H_RANDOM handler can be enabled for guest use 7154with the KVM_CAP_PPC_ENABLE_HCALL capability. 7155 71568.2 KVM_CAP_HYPERV_SYNIC 7157------------------------ 7158 7159:Architectures: x86 7160 7161This capability, if KVM_CHECK_EXTENSION indicates that it is 7162available, means that the kernel has an implementation of the 7163Hyper-V Synthetic interrupt controller(SynIC). Hyper-V SynIC is 7164used to support Windows Hyper-V based guest paravirt drivers(VMBus). 7165 7166In order to use SynIC, it has to be activated by setting this 7167capability via KVM_ENABLE_CAP ioctl on the vcpu fd. Note that this 7168will disable the use of APIC hardware virtualization even if supported 7169by the CPU, as it's incompatible with SynIC auto-EOI behavior. 7170 71718.3 KVM_CAP_PPC_RADIX_MMU 7172------------------------- 7173 7174:Architectures: ppc 7175 7176This capability, if KVM_CHECK_EXTENSION indicates that it is 7177available, means that the kernel can support guests using the 7178radix MMU defined in Power ISA V3.00 (as implemented in the POWER9 7179processor). 7180 71818.4 KVM_CAP_PPC_HASH_MMU_V3 7182--------------------------- 7183 7184:Architectures: ppc 7185 7186This capability, if KVM_CHECK_EXTENSION indicates that it is 7187available, means that the kernel can support guests using the 7188hashed page table MMU defined in Power ISA V3.00 (as implemented in 7189the POWER9 processor), including in-memory segment tables. 7190 71918.5 KVM_CAP_MIPS_VZ 7192------------------- 7193 7194:Architectures: mips 7195 7196This capability, if KVM_CHECK_EXTENSION on the main kvm handle indicates that 7197it is available, means that full hardware assisted virtualization capabilities 7198of the hardware are available for use through KVM. An appropriate 7199KVM_VM_MIPS_* type must be passed to KVM_CREATE_VM to create a VM which 7200utilises it. 7201 7202If KVM_CHECK_EXTENSION on a kvm VM handle indicates that this capability is 7203available, it means that the VM is using full hardware assisted virtualization 7204capabilities of the hardware. This is useful to check after creating a VM with 7205KVM_VM_MIPS_DEFAULT. 7206 7207The value returned by KVM_CHECK_EXTENSION should be compared against known 7208values (see below). All other values are reserved. This is to allow for the 7209possibility of other hardware assisted virtualization implementations which 7210may be incompatible with the MIPS VZ ASE. 7211 7212== ========================================================================== 7213 0 The trap & emulate implementation is in use to run guest code in user 7214 mode. Guest virtual memory segments are rearranged to fit the guest in the 7215 user mode address space. 7216 7217 1 The MIPS VZ ASE is in use, providing full hardware assisted 7218 virtualization, including standard guest virtual memory segments. 7219== ========================================================================== 7220 72218.6 KVM_CAP_MIPS_TE 7222------------------- 7223 7224:Architectures: mips 7225 7226This capability, if KVM_CHECK_EXTENSION on the main kvm handle indicates that 7227it is available, means that the trap & emulate implementation is available to 7228run guest code in user mode, even if KVM_CAP_MIPS_VZ indicates that hardware 7229assisted virtualisation is also available. KVM_VM_MIPS_TE (0) must be passed 7230to KVM_CREATE_VM to create a VM which utilises it. 7231 7232If KVM_CHECK_EXTENSION on a kvm VM handle indicates that this capability is 7233available, it means that the VM is using trap & emulate. 7234 72358.7 KVM_CAP_MIPS_64BIT 7236---------------------- 7237 7238:Architectures: mips 7239 7240This capability indicates the supported architecture type of the guest, i.e. the 7241supported register and address width. 7242 7243The values returned when this capability is checked by KVM_CHECK_EXTENSION on a 7244kvm VM handle correspond roughly to the CP0_Config.AT register field, and should 7245be checked specifically against known values (see below). All other values are 7246reserved. 7247 7248== ======================================================================== 7249 0 MIPS32 or microMIPS32. 7250 Both registers and addresses are 32-bits wide. 7251 It will only be possible to run 32-bit guest code. 7252 7253 1 MIPS64 or microMIPS64 with access only to 32-bit compatibility segments. 7254 Registers are 64-bits wide, but addresses are 32-bits wide. 7255 64-bit guest code may run but cannot access MIPS64 memory segments. 7256 It will also be possible to run 32-bit guest code. 7257 7258 2 MIPS64 or microMIPS64 with access to all address segments. 7259 Both registers and addresses are 64-bits wide. 7260 It will be possible to run 64-bit or 32-bit guest code. 7261== ======================================================================== 7262 72638.9 KVM_CAP_ARM_USER_IRQ 7264------------------------ 7265 7266:Architectures: arm64 7267 7268This capability, if KVM_CHECK_EXTENSION indicates that it is available, means 7269that if userspace creates a VM without an in-kernel interrupt controller, it 7270will be notified of changes to the output level of in-kernel emulated devices, 7271which can generate virtual interrupts, presented to the VM. 7272For such VMs, on every return to userspace, the kernel 7273updates the vcpu's run->s.regs.device_irq_level field to represent the actual 7274output level of the device. 7275 7276Whenever kvm detects a change in the device output level, kvm guarantees at 7277least one return to userspace before running the VM. This exit could either 7278be a KVM_EXIT_INTR or any other exit event, like KVM_EXIT_MMIO. This way, 7279userspace can always sample the device output level and re-compute the state of 7280the userspace interrupt controller. Userspace should always check the state 7281of run->s.regs.device_irq_level on every kvm exit. 7282The value in run->s.regs.device_irq_level can represent both level and edge 7283triggered interrupt signals, depending on the device. Edge triggered interrupt 7284signals will exit to userspace with the bit in run->s.regs.device_irq_level 7285set exactly once per edge signal. 7286 7287The field run->s.regs.device_irq_level is available independent of 7288run->kvm_valid_regs or run->kvm_dirty_regs bits. 7289 7290If KVM_CAP_ARM_USER_IRQ is supported, the KVM_CHECK_EXTENSION ioctl returns a 7291number larger than 0 indicating the version of this capability is implemented 7292and thereby which bits in run->s.regs.device_irq_level can signal values. 7293 7294Currently the following bits are defined for the device_irq_level bitmap:: 7295 7296 KVM_CAP_ARM_USER_IRQ >= 1: 7297 7298 KVM_ARM_DEV_EL1_VTIMER - EL1 virtual timer 7299 KVM_ARM_DEV_EL1_PTIMER - EL1 physical timer 7300 KVM_ARM_DEV_PMU - ARM PMU overflow interrupt signal 7301 7302Future versions of kvm may implement additional events. These will get 7303indicated by returning a higher number from KVM_CHECK_EXTENSION and will be 7304listed above. 7305 73068.10 KVM_CAP_PPC_SMT_POSSIBLE 7307----------------------------- 7308 7309:Architectures: ppc 7310 7311Querying this capability returns a bitmap indicating the possible 7312virtual SMT modes that can be set using KVM_CAP_PPC_SMT. If bit N 7313(counting from the right) is set, then a virtual SMT mode of 2^N is 7314available. 7315 73168.11 KVM_CAP_HYPERV_SYNIC2 7317-------------------------- 7318 7319:Architectures: x86 7320 7321This capability enables a newer version of Hyper-V Synthetic interrupt 7322controller (SynIC). The only difference with KVM_CAP_HYPERV_SYNIC is that KVM 7323doesn't clear SynIC message and event flags pages when they are enabled by 7324writing to the respective MSRs. 7325 73268.12 KVM_CAP_HYPERV_VP_INDEX 7327---------------------------- 7328 7329:Architectures: x86 7330 7331This capability indicates that userspace can load HV_X64_MSR_VP_INDEX msr. Its 7332value is used to denote the target vcpu for a SynIC interrupt. For 7333compatibilty, KVM initializes this msr to KVM's internal vcpu index. When this 7334capability is absent, userspace can still query this msr's value. 7335 73368.13 KVM_CAP_S390_AIS_MIGRATION 7337------------------------------- 7338 7339:Architectures: s390 7340:Parameters: none 7341 7342This capability indicates if the flic device will be able to get/set the 7343AIS states for migration via the KVM_DEV_FLIC_AISM_ALL attribute and allows 7344to discover this without having to create a flic device. 7345 73468.14 KVM_CAP_S390_PSW 7347--------------------- 7348 7349:Architectures: s390 7350 7351This capability indicates that the PSW is exposed via the kvm_run structure. 7352 73538.15 KVM_CAP_S390_GMAP 7354---------------------- 7355 7356:Architectures: s390 7357 7358This capability indicates that the user space memory used as guest mapping can 7359be anywhere in the user memory address space, as long as the memory slots are 7360aligned and sized to a segment (1MB) boundary. 7361 73628.16 KVM_CAP_S390_COW 7363--------------------- 7364 7365:Architectures: s390 7366 7367This capability indicates that the user space memory used as guest mapping can 7368use copy-on-write semantics as well as dirty pages tracking via read-only page 7369tables. 7370 73718.17 KVM_CAP_S390_BPB 7372--------------------- 7373 7374:Architectures: s390 7375 7376This capability indicates that kvm will implement the interfaces to handle 7377reset, migration and nested KVM for branch prediction blocking. The stfle 7378facility 82 should not be provided to the guest without this capability. 7379 73808.18 KVM_CAP_HYPERV_TLBFLUSH 7381---------------------------- 7382 7383:Architectures: x86 7384 7385This capability indicates that KVM supports paravirtualized Hyper-V TLB Flush 7386hypercalls: 7387HvFlushVirtualAddressSpace, HvFlushVirtualAddressSpaceEx, 7388HvFlushVirtualAddressList, HvFlushVirtualAddressListEx. 7389 73908.19 KVM_CAP_ARM_INJECT_SERROR_ESR 7391---------------------------------- 7392 7393:Architectures: arm64 7394 7395This capability indicates that userspace can specify (via the 7396KVM_SET_VCPU_EVENTS ioctl) the syndrome value reported to the guest when it 7397takes a virtual SError interrupt exception. 7398If KVM advertises this capability, userspace can only specify the ISS field for 7399the ESR syndrome. Other parts of the ESR, such as the EC are generated by the 7400CPU when the exception is taken. If this virtual SError is taken to EL1 using 7401AArch64, this value will be reported in the ISS field of ESR_ELx. 7402 7403See KVM_CAP_VCPU_EVENTS for more details. 7404 74058.20 KVM_CAP_HYPERV_SEND_IPI 7406---------------------------- 7407 7408:Architectures: x86 7409 7410This capability indicates that KVM supports paravirtualized Hyper-V IPI send 7411hypercalls: 7412HvCallSendSyntheticClusterIpi, HvCallSendSyntheticClusterIpiEx. 7413 74148.21 KVM_CAP_HYPERV_DIRECT_TLBFLUSH 7415----------------------------------- 7416 7417:Architectures: x86 7418 7419This capability indicates that KVM running on top of Hyper-V hypervisor 7420enables Direct TLB flush for its guests meaning that TLB flush 7421hypercalls are handled by Level 0 hypervisor (Hyper-V) bypassing KVM. 7422Due to the different ABI for hypercall parameters between Hyper-V and 7423KVM, enabling this capability effectively disables all hypercall 7424handling by KVM (as some KVM hypercall may be mistakenly treated as TLB 7425flush hypercalls by Hyper-V) so userspace should disable KVM identification 7426in CPUID and only exposes Hyper-V identification. In this case, guest 7427thinks it's running on Hyper-V and only use Hyper-V hypercalls. 7428 74298.22 KVM_CAP_S390_VCPU_RESETS 7430----------------------------- 7431 7432:Architectures: s390 7433 7434This capability indicates that the KVM_S390_NORMAL_RESET and 7435KVM_S390_CLEAR_RESET ioctls are available. 7436 74378.23 KVM_CAP_S390_PROTECTED 7438--------------------------- 7439 7440:Architectures: s390 7441 7442This capability indicates that the Ultravisor has been initialized and 7443KVM can therefore start protected VMs. 7444This capability governs the KVM_S390_PV_COMMAND ioctl and the 7445KVM_MP_STATE_LOAD MP_STATE. KVM_SET_MP_STATE can fail for protected 7446guests when the state change is invalid. 7447 74488.24 KVM_CAP_STEAL_TIME 7449----------------------- 7450 7451:Architectures: arm64, x86 7452 7453This capability indicates that KVM supports steal time accounting. 7454When steal time accounting is supported it may be enabled with 7455architecture-specific interfaces. This capability and the architecture- 7456specific interfaces must be consistent, i.e. if one says the feature 7457is supported, than the other should as well and vice versa. For arm64 7458see Documentation/virt/kvm/devices/vcpu.rst "KVM_ARM_VCPU_PVTIME_CTRL". 7459For x86 see Documentation/virt/kvm/msr.rst "MSR_KVM_STEAL_TIME". 7460 74618.25 KVM_CAP_S390_DIAG318 7462------------------------- 7463 7464:Architectures: s390 7465 7466This capability enables a guest to set information about its control program 7467(i.e. guest kernel type and version). The information is helpful during 7468system/firmware service events, providing additional data about the guest 7469environments running on the machine. 7470 7471The information is associated with the DIAGNOSE 0x318 instruction, which sets 7472an 8-byte value consisting of a one-byte Control Program Name Code (CPNC) and 7473a 7-byte Control Program Version Code (CPVC). The CPNC determines what 7474environment the control program is running in (e.g. Linux, z/VM...), and the 7475CPVC is used for information specific to OS (e.g. Linux version, Linux 7476distribution...) 7477 7478If this capability is available, then the CPNC and CPVC can be synchronized 7479between KVM and userspace via the sync regs mechanism (KVM_SYNC_DIAG318). 7480 74818.26 KVM_CAP_X86_USER_SPACE_MSR 7482------------------------------- 7483 7484:Architectures: x86 7485 7486This capability indicates that KVM supports deflection of MSR reads and 7487writes to user space. It can be enabled on a VM level. If enabled, MSR 7488accesses that would usually trigger a #GP by KVM into the guest will 7489instead get bounced to user space through the KVM_EXIT_X86_RDMSR and 7490KVM_EXIT_X86_WRMSR exit notifications. 7491 74928.27 KVM_CAP_X86_MSR_FILTER 7493--------------------------- 7494 7495:Architectures: x86 7496 7497This capability indicates that KVM supports that accesses to user defined MSRs 7498may be rejected. With this capability exposed, KVM exports new VM ioctl 7499KVM_X86_SET_MSR_FILTER which user space can call to specify bitmaps of MSR 7500ranges that KVM should reject access to. 7501 7502In combination with KVM_CAP_X86_USER_SPACE_MSR, this allows user space to 7503trap and emulate MSRs that are outside of the scope of KVM as well as 7504limit the attack surface on KVM's MSR emulation code. 7505 75068.28 KVM_CAP_ENFORCE_PV_FEATURE_CPUID 7507------------------------------------- 7508 7509Architectures: x86 7510 7511When enabled, KVM will disable paravirtual features provided to the 7512guest according to the bits in the KVM_CPUID_FEATURES CPUID leaf 7513(0x40000001). Otherwise, a guest may use the paravirtual features 7514regardless of what has actually been exposed through the CPUID leaf. 7515 75168.29 KVM_CAP_DIRTY_LOG_RING 7517--------------------------- 7518 7519:Architectures: x86 7520:Parameters: args[0] - size of the dirty log ring 7521 7522KVM is capable of tracking dirty memory using ring buffers that are 7523mmaped into userspace; there is one dirty ring per vcpu. 7524 7525The dirty ring is available to userspace as an array of 7526``struct kvm_dirty_gfn``. Each dirty entry it's defined as:: 7527 7528 struct kvm_dirty_gfn { 7529 __u32 flags; 7530 __u32 slot; /* as_id | slot_id */ 7531 __u64 offset; 7532 }; 7533 7534The following values are defined for the flags field to define the 7535current state of the entry:: 7536 7537 #define KVM_DIRTY_GFN_F_DIRTY BIT(0) 7538 #define KVM_DIRTY_GFN_F_RESET BIT(1) 7539 #define KVM_DIRTY_GFN_F_MASK 0x3 7540 7541Userspace should call KVM_ENABLE_CAP ioctl right after KVM_CREATE_VM 7542ioctl to enable this capability for the new guest and set the size of 7543the rings. Enabling the capability is only allowed before creating any 7544vCPU, and the size of the ring must be a power of two. The larger the 7545ring buffer, the less likely the ring is full and the VM is forced to 7546exit to userspace. The optimal size depends on the workload, but it is 7547recommended that it be at least 64 KiB (4096 entries). 7548 7549Just like for dirty page bitmaps, the buffer tracks writes to 7550all user memory regions for which the KVM_MEM_LOG_DIRTY_PAGES flag was 7551set in KVM_SET_USER_MEMORY_REGION. Once a memory region is registered 7552with the flag set, userspace can start harvesting dirty pages from the 7553ring buffer. 7554 7555An entry in the ring buffer can be unused (flag bits ``00``), 7556dirty (flag bits ``01``) or harvested (flag bits ``1X``). The 7557state machine for the entry is as follows:: 7558 7559 dirtied harvested reset 7560 00 -----------> 01 -------------> 1X -------+ 7561 ^ | 7562 | | 7563 +------------------------------------------+ 7564 7565To harvest the dirty pages, userspace accesses the mmaped ring buffer 7566to read the dirty GFNs. If the flags has the DIRTY bit set (at this stage 7567the RESET bit must be cleared), then it means this GFN is a dirty GFN. 7568The userspace should harvest this GFN and mark the flags from state 7569``01b`` to ``1Xb`` (bit 0 will be ignored by KVM, but bit 1 must be set 7570to show that this GFN is harvested and waiting for a reset), and move 7571on to the next GFN. The userspace should continue to do this until the 7572flags of a GFN have the DIRTY bit cleared, meaning that it has harvested 7573all the dirty GFNs that were available. 7574 7575It's not necessary for userspace to harvest the all dirty GFNs at once. 7576However it must collect the dirty GFNs in sequence, i.e., the userspace 7577program cannot skip one dirty GFN to collect the one next to it. 7578 7579After processing one or more entries in the ring buffer, userspace 7580calls the VM ioctl KVM_RESET_DIRTY_RINGS to notify the kernel about 7581it, so that the kernel will reprotect those collected GFNs. 7582Therefore, the ioctl must be called *before* reading the content of 7583the dirty pages. 7584 7585The dirty ring can get full. When it happens, the KVM_RUN of the 7586vcpu will return with exit reason KVM_EXIT_DIRTY_LOG_FULL. 7587 7588The dirty ring interface has a major difference comparing to the 7589KVM_GET_DIRTY_LOG interface in that, when reading the dirty ring from 7590userspace, it's still possible that the kernel has not yet flushed the 7591processor's dirty page buffers into the kernel buffer (with dirty bitmaps, the 7592flushing is done by the KVM_GET_DIRTY_LOG ioctl). To achieve that, one 7593needs to kick the vcpu out of KVM_RUN using a signal. The resulting 7594vmexit ensures that all dirty GFNs are flushed to the dirty rings. 7595 7596NOTE: the capability KVM_CAP_DIRTY_LOG_RING and the corresponding 7597ioctl KVM_RESET_DIRTY_RINGS are mutual exclusive to the existing ioctls 7598KVM_GET_DIRTY_LOG and KVM_CLEAR_DIRTY_LOG. After enabling 7599KVM_CAP_DIRTY_LOG_RING with an acceptable dirty ring size, the virtual 7600machine will switch to ring-buffer dirty page tracking and further 7601KVM_GET_DIRTY_LOG or KVM_CLEAR_DIRTY_LOG ioctls will fail. 7602 76038.30 KVM_CAP_XEN_HVM 7604-------------------- 7605 7606:Architectures: x86 7607 7608This capability indicates the features that Xen supports for hosting Xen 7609PVHVM guests. Valid flags are:: 7610 7611 #define KVM_XEN_HVM_CONFIG_HYPERCALL_MSR (1 << 0) 7612 #define KVM_XEN_HVM_CONFIG_INTERCEPT_HCALL (1 << 1) 7613 #define KVM_XEN_HVM_CONFIG_SHARED_INFO (1 << 2) 7614 #define KVM_XEN_HVM_CONFIG_RUNSTATE (1 << 2) 7615 #define KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL (1 << 3) 7616 7617The KVM_XEN_HVM_CONFIG_HYPERCALL_MSR flag indicates that the KVM_XEN_HVM_CONFIG 7618ioctl is available, for the guest to set its hypercall page. 7619 7620If KVM_XEN_HVM_CONFIG_INTERCEPT_HCALL is also set, the same flag may also be 7621provided in the flags to KVM_XEN_HVM_CONFIG, without providing hypercall page 7622contents, to request that KVM generate hypercall page content automatically 7623and also enable interception of guest hypercalls with KVM_EXIT_XEN. 7624 7625The KVM_XEN_HVM_CONFIG_SHARED_INFO flag indicates the availability of the 7626KVM_XEN_HVM_SET_ATTR, KVM_XEN_HVM_GET_ATTR, KVM_XEN_VCPU_SET_ATTR and 7627KVM_XEN_VCPU_GET_ATTR ioctls, as well as the delivery of exception vectors 7628for event channel upcalls when the evtchn_upcall_pending field of a vcpu's 7629vcpu_info is set. 7630 7631The KVM_XEN_HVM_CONFIG_RUNSTATE flag indicates that the runstate-related 7632features KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADDR/_CURRENT/_DATA/_ADJUST are 7633supported by the KVM_XEN_VCPU_SET_ATTR/KVM_XEN_VCPU_GET_ATTR ioctls. 7634 7635The KVM_XEN_HVM_CONFIG_EVTCHN_2LEVEL flag indicates that IRQ routing entries 7636of the type KVM_IRQ_ROUTING_XEN_EVTCHN are supported, with the priority 7637field set to indicate 2 level event channel delivery. 7638 76398.31 KVM_CAP_PPC_MULTITCE 7640------------------------- 7641 7642:Capability: KVM_CAP_PPC_MULTITCE 7643:Architectures: ppc 7644:Type: vm 7645 7646This capability means the kernel is capable of handling hypercalls 7647H_PUT_TCE_INDIRECT and H_STUFF_TCE without passing those into the user 7648space. This significantly accelerates DMA operations for PPC KVM guests. 7649User space should expect that its handlers for these hypercalls 7650are not going to be called if user space previously registered LIOBN 7651in KVM (via KVM_CREATE_SPAPR_TCE or similar calls). 7652 7653In order to enable H_PUT_TCE_INDIRECT and H_STUFF_TCE use in the guest, 7654user space might have to advertise it for the guest. For example, 7655IBM pSeries (sPAPR) guest starts using them if "hcall-multi-tce" is 7656present in the "ibm,hypertas-functions" device-tree property. 7657 7658The hypercalls mentioned above may or may not be processed successfully 7659in the kernel based fast path. If they can not be handled by the kernel, 7660they will get passed on to user space. So user space still has to have 7661an implementation for these despite the in kernel acceleration. 7662 7663This capability is always enabled. 7664 76658.32 KVM_CAP_PTP_KVM 7666-------------------- 7667 7668:Architectures: arm64 7669 7670This capability indicates that the KVM virtual PTP service is 7671supported in the host. A VMM can check whether the service is 7672available to the guest on migration. 7673 76748.33 KVM_CAP_HYPERV_ENFORCE_CPUID 7675--------------------------------- 7676 7677Architectures: x86 7678 7679When enabled, KVM will disable emulated Hyper-V features provided to the 7680guest according to the bits Hyper-V CPUID feature leaves. Otherwise, all 7681currently implmented Hyper-V features are provided unconditionally when 7682Hyper-V identification is set in the HYPERV_CPUID_INTERFACE (0x40000001) 7683leaf. 7684 76858.34 KVM_CAP_EXIT_HYPERCALL 7686--------------------------- 7687 7688:Capability: KVM_CAP_EXIT_HYPERCALL 7689:Architectures: x86 7690:Type: vm 7691 7692This capability, if enabled, will cause KVM to exit to userspace 7693with KVM_EXIT_HYPERCALL exit reason to process some hypercalls. 7694 7695Calling KVM_CHECK_EXTENSION for this capability will return a bitmask 7696of hypercalls that can be configured to exit to userspace. 7697Right now, the only such hypercall is KVM_HC_MAP_GPA_RANGE. 7698 7699The argument to KVM_ENABLE_CAP is also a bitmask, and must be a subset 7700of the result of KVM_CHECK_EXTENSION. KVM will forward to userspace 7701the hypercalls whose corresponding bit is in the argument, and return 7702ENOSYS for the others. 7703 77048.35 KVM_CAP_PMU_CAPABILITY 7705--------------------------- 7706 7707:Capability KVM_CAP_PMU_CAPABILITY 7708:Architectures: x86 7709:Type: vm 7710:Parameters: arg[0] is bitmask of PMU virtualization capabilities. 7711:Returns 0 on success, -EINVAL when arg[0] contains invalid bits 7712 7713This capability alters PMU virtualization in KVM. 7714 7715Calling KVM_CHECK_EXTENSION for this capability returns a bitmask of 7716PMU virtualization capabilities that can be adjusted on a VM. 7717 7718The argument to KVM_ENABLE_CAP is also a bitmask and selects specific 7719PMU virtualization capabilities to be applied to the VM. This can 7720only be invoked on a VM prior to the creation of VCPUs. 7721 7722At this time, KVM_PMU_CAP_DISABLE is the only capability. Setting 7723this capability will disable PMU virtualization for that VM. Usermode 7724should adjust CPUID leaf 0xA to reflect that the PMU is disabled. 7725 77269. Known KVM API problems 7727========================= 7728 7729In some cases, KVM's API has some inconsistencies or common pitfalls 7730that userspace need to be aware of. This section details some of 7731these issues. 7732 7733Most of them are architecture specific, so the section is split by 7734architecture. 7735 77369.1. x86 7737-------- 7738 7739``KVM_GET_SUPPORTED_CPUID`` issues 7740^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 7741 7742In general, ``KVM_GET_SUPPORTED_CPUID`` is designed so that it is possible 7743to take its result and pass it directly to ``KVM_SET_CPUID2``. This section 7744documents some cases in which that requires some care. 7745 7746Local APIC features 7747~~~~~~~~~~~~~~~~~~~ 7748 7749CPU[EAX=1]:ECX[21] (X2APIC) is reported by ``KVM_GET_SUPPORTED_CPUID``, 7750but it can only be enabled if ``KVM_CREATE_IRQCHIP`` or 7751``KVM_ENABLE_CAP(KVM_CAP_IRQCHIP_SPLIT)`` are used to enable in-kernel emulation of 7752the local APIC. 7753 7754The same is true for the ``KVM_FEATURE_PV_UNHALT`` paravirtualized feature. 7755 7756CPU[EAX=1]:ECX[24] (TSC_DEADLINE) is not reported by ``KVM_GET_SUPPORTED_CPUID``. 7757It can be enabled if ``KVM_CAP_TSC_DEADLINE_TIMER`` is present and the kernel 7758has enabled in-kernel emulation of the local APIC. 7759 7760Obsolete ioctls and capabilities 7761^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 7762 7763KVM_CAP_DISABLE_QUIRKS does not let userspace know which quirks are actually 7764available. Use ``KVM_CHECK_EXTENSION(KVM_CAP_DISABLE_QUIRKS2)`` instead if 7765available. 7766 7767Ordering of KVM_GET_*/KVM_SET_* ioctls 7768^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 7769 7770TBD 7771