]>
Commit | Line | Data |
---|---|---|
9c1b96e3 AK |
1 | The Definitive KVM (Kernel-based Virtual Machine) API Documentation |
2 | =================================================================== | |
3 | ||
4 | 1. General description | |
5 | ||
6 | The kvm API is a set of ioctls that are issued to control various aspects | |
7 | of a virtual machine. The ioctls belong to three classes | |
8 | ||
9 | - System ioctls: These query and set global attributes which affect the | |
10 | whole kvm subsystem. In addition a system ioctl is used to create | |
11 | virtual machines | |
12 | ||
13 | - VM ioctls: These query and set attributes that affect an entire virtual | |
14 | machine, for example memory layout. In addition a VM ioctl is used to | |
15 | create virtual cpus (vcpus). | |
16 | ||
17 | Only run VM ioctls from the same process (address space) that was used | |
18 | to create the VM. | |
19 | ||
20 | - vcpu ioctls: These query and set attributes that control the operation | |
21 | of a single virtual cpu. | |
22 | ||
23 | Only run vcpu ioctls from the same thread that was used to create the | |
24 | vcpu. | |
25 | ||
2044892d | 26 | 2. File descriptors |
9c1b96e3 AK |
27 | |
28 | The kvm API is centered around file descriptors. An initial | |
29 | open("/dev/kvm") obtains a handle to the kvm subsystem; this handle | |
30 | can be used to issue system ioctls. A KVM_CREATE_VM ioctl on this | |
2044892d | 31 | handle will create a VM file descriptor which can be used to issue VM |
9c1b96e3 AK |
32 | ioctls. A KVM_CREATE_VCPU ioctl on a VM fd will create a virtual cpu |
33 | and return a file descriptor pointing to it. Finally, ioctls on a vcpu | |
34 | fd can be used to control the vcpu, including the important task of | |
35 | actually running guest code. | |
36 | ||
37 | In general file descriptors can be migrated among processes by means | |
38 | of fork() and the SCM_RIGHTS facility of unix domain socket. These | |
39 | kinds of tricks are explicitly not supported by kvm. While they will | |
40 | not cause harm to the host, their actual behavior is not guaranteed by | |
41 | the API. The only supported use is one virtual machine per process, | |
42 | and one vcpu per thread. | |
43 | ||
44 | 3. Extensions | |
45 | ||
46 | As of Linux 2.6.22, the KVM ABI has been stabilized: no backward | |
47 | incompatible change are allowed. However, there is an extension | |
48 | facility that allows backward-compatible extensions to the API to be | |
49 | queried and used. | |
50 | ||
51 | The extension mechanism is not based on on the Linux version number. | |
52 | Instead, kvm defines extension identifiers and a facility to query | |
53 | whether a particular extension identifier is available. If it is, a | |
54 | set of ioctls is available for application use. | |
55 | ||
56 | 4. API description | |
57 | ||
58 | This section describes ioctls that can be used to control kvm guests. | |
59 | For each ioctl, the following information is provided along with a | |
60 | description: | |
61 | ||
62 | Capability: which KVM extension provides this ioctl. Can be 'basic', | |
63 | which means that is will be provided by any kernel that supports | |
64 | API version 12 (see section 4.1), or a KVM_CAP_xyz constant, which | |
65 | means availability needs to be checked with KVM_CHECK_EXTENSION | |
66 | (see section 4.4). | |
67 | ||
68 | Architectures: which instruction set architectures provide this ioctl. | |
69 | x86 includes both i386 and x86_64. | |
70 | ||
71 | Type: system, vm, or vcpu. | |
72 | ||
73 | Parameters: what parameters are accepted by the ioctl. | |
74 | ||
75 | Returns: the return value. General error numbers (EBADF, ENOMEM, EINVAL) | |
76 | are not detailed, but errors with specific meanings are. | |
77 | ||
78 | 4.1 KVM_GET_API_VERSION | |
79 | ||
80 | Capability: basic | |
81 | Architectures: all | |
82 | Type: system ioctl | |
83 | Parameters: none | |
84 | Returns: the constant KVM_API_VERSION (=12) | |
85 | ||
86 | This identifies the API version as the stable kvm API. It is not | |
87 | expected that this number will change. However, Linux 2.6.20 and | |
88 | 2.6.21 report earlier versions; these are not documented and not | |
89 | supported. Applications should refuse to run if KVM_GET_API_VERSION | |
90 | returns a value other than 12. If this check passes, all ioctls | |
91 | described as 'basic' will be available. | |
92 | ||
93 | 4.2 KVM_CREATE_VM | |
94 | ||
95 | Capability: basic | |
96 | Architectures: all | |
97 | Type: system ioctl | |
98 | Parameters: none | |
99 | Returns: a VM fd that can be used to control the new virtual machine. | |
100 | ||
101 | The new VM has no virtual cpus and no memory. An mmap() of a VM fd | |
102 | will access the virtual machine's physical address space; offset zero | |
103 | corresponds to guest physical address zero. Use of mmap() on a VM fd | |
104 | is discouraged if userspace memory allocation (KVM_CAP_USER_MEMORY) is | |
105 | available. | |
106 | ||
107 | 4.3 KVM_GET_MSR_INDEX_LIST | |
108 | ||
109 | Capability: basic | |
110 | Architectures: x86 | |
111 | Type: system | |
112 | Parameters: struct kvm_msr_list (in/out) | |
113 | Returns: 0 on success; -1 on error | |
114 | Errors: | |
115 | E2BIG: the msr index list is to be to fit in the array specified by | |
116 | the user. | |
117 | ||
118 | struct kvm_msr_list { | |
119 | __u32 nmsrs; /* number of msrs in entries */ | |
120 | __u32 indices[0]; | |
121 | }; | |
122 | ||
123 | This ioctl returns the guest msrs that are supported. The list varies | |
124 | by kvm version and host processor, but does not change otherwise. The | |
125 | user fills in the size of the indices array in nmsrs, and in return | |
126 | kvm adjusts nmsrs to reflect the actual number of msrs and fills in | |
127 | the indices array with their numbers. | |
128 | ||
2e2602ca AK |
129 | Note: if kvm indicates supports MCE (KVM_CAP_MCE), then the MCE bank MSRs are |
130 | not returned in the MSR list, as different vcpus can have a different number | |
131 | of banks, as set via the KVM_X86_SETUP_MCE ioctl. | |
132 | ||
9c1b96e3 AK |
133 | 4.4 KVM_CHECK_EXTENSION |
134 | ||
135 | Capability: basic | |
136 | Architectures: all | |
137 | Type: system ioctl | |
138 | Parameters: extension identifier (KVM_CAP_*) | |
139 | Returns: 0 if unsupported; 1 (or some other positive integer) if supported | |
140 | ||
141 | The API allows the application to query about extensions to the core | |
142 | kvm API. Userspace passes an extension identifier (an integer) and | |
143 | receives an integer that describes the extension availability. | |
144 | Generally 0 means no and 1 means yes, but some extensions may report | |
145 | additional information in the integer return value. | |
146 | ||
147 | 4.5 KVM_GET_VCPU_MMAP_SIZE | |
148 | ||
149 | Capability: basic | |
150 | Architectures: all | |
151 | Type: system ioctl | |
152 | Parameters: none | |
153 | Returns: size of vcpu mmap area, in bytes | |
154 | ||
155 | The KVM_RUN ioctl (cf.) communicates with userspace via a shared | |
156 | memory region. This ioctl returns the size of that region. See the | |
157 | KVM_RUN documentation for details. | |
158 | ||
159 | 4.6 KVM_SET_MEMORY_REGION | |
160 | ||
161 | Capability: basic | |
162 | Architectures: all | |
163 | Type: vm ioctl | |
164 | Parameters: struct kvm_memory_region (in) | |
165 | Returns: 0 on success, -1 on error | |
166 | ||
b74a07be | 167 | This ioctl is obsolete and has been removed. |
9c1b96e3 | 168 | |
68ba6974 | 169 | 4.7 KVM_CREATE_VCPU |
9c1b96e3 AK |
170 | |
171 | Capability: basic | |
172 | Architectures: all | |
173 | Type: vm ioctl | |
174 | Parameters: vcpu id (apic id on x86) | |
175 | Returns: vcpu fd on success, -1 on error | |
176 | ||
177 | This API adds a vcpu to a virtual machine. The vcpu id is a small integer | |
76d25402 PE |
178 | in the range [0, max_vcpus). You can use KVM_CAP_NR_VCPUS of the |
179 | KVM_CHECK_EXTENSION ioctl() to determine the value for max_vcpus at run-time. | |
180 | If the KVM_CAP_NR_VCPUS does not exist, you should assume that max_vcpus is 4 | |
181 | cpus max. | |
9c1b96e3 | 182 | |
371fefd6 PM |
183 | On powerpc using book3s_hv mode, the vcpus are mapped onto virtual |
184 | threads in one or more virtual CPU cores. (This is because the | |
185 | hardware requires all the hardware threads in a CPU core to be in the | |
186 | same partition.) The KVM_CAP_PPC_SMT capability indicates the number | |
187 | of vcpus per virtual core (vcore). The vcore id is obtained by | |
188 | dividing the vcpu id by the number of vcpus per vcore. The vcpus in a | |
189 | given vcore will always be in the same physical core as each other | |
190 | (though that might be a different physical core from time to time). | |
191 | Userspace can control the threading (SMT) mode of the guest by its | |
192 | allocation of vcpu ids. For example, if userspace wants | |
193 | single-threaded guest vcpus, it should make all vcpu ids be a multiple | |
194 | of the number of vcpus per vcore. | |
195 | ||
68ba6974 | 196 | 4.8 KVM_GET_DIRTY_LOG (vm ioctl) |
9c1b96e3 AK |
197 | |
198 | Capability: basic | |
199 | Architectures: x86 | |
200 | Type: vm ioctl | |
201 | Parameters: struct kvm_dirty_log (in/out) | |
202 | Returns: 0 on success, -1 on error | |
203 | ||
204 | /* for KVM_GET_DIRTY_LOG */ | |
205 | struct kvm_dirty_log { | |
206 | __u32 slot; | |
207 | __u32 padding; | |
208 | union { | |
209 | void __user *dirty_bitmap; /* one bit per page */ | |
210 | __u64 padding; | |
211 | }; | |
212 | }; | |
213 | ||
214 | Given a memory slot, return a bitmap containing any pages dirtied | |
215 | since the last call to this ioctl. Bit 0 is the first page in the | |
216 | memory slot. Ensure the entire structure is cleared to avoid padding | |
217 | issues. | |
218 | ||
68ba6974 | 219 | 4.9 KVM_SET_MEMORY_ALIAS |
9c1b96e3 AK |
220 | |
221 | Capability: basic | |
222 | Architectures: x86 | |
223 | Type: vm ioctl | |
224 | Parameters: struct kvm_memory_alias (in) | |
225 | Returns: 0 (success), -1 (error) | |
226 | ||
a1f4d395 | 227 | This ioctl is obsolete and has been removed. |
9c1b96e3 | 228 | |
68ba6974 | 229 | 4.10 KVM_RUN |
9c1b96e3 AK |
230 | |
231 | Capability: basic | |
232 | Architectures: all | |
233 | Type: vcpu ioctl | |
234 | Parameters: none | |
235 | Returns: 0 on success, -1 on error | |
236 | Errors: | |
237 | EINTR: an unmasked signal is pending | |
238 | ||
239 | This ioctl is used to run a guest virtual cpu. While there are no | |
240 | explicit parameters, there is an implicit parameter block that can be | |
241 | obtained by mmap()ing the vcpu fd at offset 0, with the size given by | |
242 | KVM_GET_VCPU_MMAP_SIZE. The parameter block is formatted as a 'struct | |
243 | kvm_run' (see below). | |
244 | ||
68ba6974 | 245 | 4.11 KVM_GET_REGS |
9c1b96e3 AK |
246 | |
247 | Capability: basic | |
248 | Architectures: all | |
249 | Type: vcpu ioctl | |
250 | Parameters: struct kvm_regs (out) | |
251 | Returns: 0 on success, -1 on error | |
252 | ||
253 | Reads the general purpose registers from the vcpu. | |
254 | ||
255 | /* x86 */ | |
256 | struct kvm_regs { | |
257 | /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */ | |
258 | __u64 rax, rbx, rcx, rdx; | |
259 | __u64 rsi, rdi, rsp, rbp; | |
260 | __u64 r8, r9, r10, r11; | |
261 | __u64 r12, r13, r14, r15; | |
262 | __u64 rip, rflags; | |
263 | }; | |
264 | ||
68ba6974 | 265 | 4.12 KVM_SET_REGS |
9c1b96e3 AK |
266 | |
267 | Capability: basic | |
268 | Architectures: all | |
269 | Type: vcpu ioctl | |
270 | Parameters: struct kvm_regs (in) | |
271 | Returns: 0 on success, -1 on error | |
272 | ||
273 | Writes the general purpose registers into the vcpu. | |
274 | ||
275 | See KVM_GET_REGS for the data structure. | |
276 | ||
68ba6974 | 277 | 4.13 KVM_GET_SREGS |
9c1b96e3 AK |
278 | |
279 | Capability: basic | |
5ce941ee | 280 | Architectures: x86, ppc |
9c1b96e3 AK |
281 | Type: vcpu ioctl |
282 | Parameters: struct kvm_sregs (out) | |
283 | Returns: 0 on success, -1 on error | |
284 | ||
285 | Reads special registers from the vcpu. | |
286 | ||
287 | /* x86 */ | |
288 | struct kvm_sregs { | |
289 | struct kvm_segment cs, ds, es, fs, gs, ss; | |
290 | struct kvm_segment tr, ldt; | |
291 | struct kvm_dtable gdt, idt; | |
292 | __u64 cr0, cr2, cr3, cr4, cr8; | |
293 | __u64 efer; | |
294 | __u64 apic_base; | |
295 | __u64 interrupt_bitmap[(KVM_NR_INTERRUPTS + 63) / 64]; | |
296 | }; | |
297 | ||
5ce941ee SW |
298 | /* ppc -- see arch/powerpc/include/asm/kvm.h */ |
299 | ||
9c1b96e3 AK |
300 | interrupt_bitmap is a bitmap of pending external interrupts. At most |
301 | one bit may be set. This interrupt has been acknowledged by the APIC | |
302 | but not yet injected into the cpu core. | |
303 | ||
68ba6974 | 304 | 4.14 KVM_SET_SREGS |
9c1b96e3 AK |
305 | |
306 | Capability: basic | |
5ce941ee | 307 | Architectures: x86, ppc |
9c1b96e3 AK |
308 | Type: vcpu ioctl |
309 | Parameters: struct kvm_sregs (in) | |
310 | Returns: 0 on success, -1 on error | |
311 | ||
312 | Writes special registers into the vcpu. See KVM_GET_SREGS for the | |
313 | data structures. | |
314 | ||
68ba6974 | 315 | 4.15 KVM_TRANSLATE |
9c1b96e3 AK |
316 | |
317 | Capability: basic | |
318 | Architectures: x86 | |
319 | Type: vcpu ioctl | |
320 | Parameters: struct kvm_translation (in/out) | |
321 | Returns: 0 on success, -1 on error | |
322 | ||
323 | Translates a virtual address according to the vcpu's current address | |
324 | translation mode. | |
325 | ||
326 | struct kvm_translation { | |
327 | /* in */ | |
328 | __u64 linear_address; | |
329 | ||
330 | /* out */ | |
331 | __u64 physical_address; | |
332 | __u8 valid; | |
333 | __u8 writeable; | |
334 | __u8 usermode; | |
335 | __u8 pad[5]; | |
336 | }; | |
337 | ||
68ba6974 | 338 | 4.16 KVM_INTERRUPT |
9c1b96e3 AK |
339 | |
340 | Capability: basic | |
6f7a2bd4 | 341 | Architectures: x86, ppc |
9c1b96e3 AK |
342 | Type: vcpu ioctl |
343 | Parameters: struct kvm_interrupt (in) | |
344 | Returns: 0 on success, -1 on error | |
345 | ||
346 | Queues a hardware interrupt vector to be injected. This is only | |
6f7a2bd4 | 347 | useful if in-kernel local APIC or equivalent is not used. |
9c1b96e3 AK |
348 | |
349 | /* for KVM_INTERRUPT */ | |
350 | struct kvm_interrupt { | |
351 | /* in */ | |
352 | __u32 irq; | |
353 | }; | |
354 | ||
6f7a2bd4 AG |
355 | X86: |
356 | ||
9c1b96e3 AK |
357 | Note 'irq' is an interrupt vector, not an interrupt pin or line. |
358 | ||
6f7a2bd4 AG |
359 | PPC: |
360 | ||
361 | Queues an external interrupt to be injected. This ioctl is overleaded | |
362 | with 3 different irq values: | |
363 | ||
364 | a) KVM_INTERRUPT_SET | |
365 | ||
366 | This injects an edge type external interrupt into the guest once it's ready | |
367 | to receive interrupts. When injected, the interrupt is done. | |
368 | ||
369 | b) KVM_INTERRUPT_UNSET | |
370 | ||
371 | This unsets any pending interrupt. | |
372 | ||
373 | Only available with KVM_CAP_PPC_UNSET_IRQ. | |
374 | ||
375 | c) KVM_INTERRUPT_SET_LEVEL | |
376 | ||
377 | This injects a level type external interrupt into the guest context. The | |
378 | interrupt stays pending until a specific ioctl with KVM_INTERRUPT_UNSET | |
379 | is triggered. | |
380 | ||
381 | Only available with KVM_CAP_PPC_IRQ_LEVEL. | |
382 | ||
383 | Note that any value for 'irq' other than the ones stated above is invalid | |
384 | and incurs unexpected behavior. | |
385 | ||
68ba6974 | 386 | 4.17 KVM_DEBUG_GUEST |
9c1b96e3 AK |
387 | |
388 | Capability: basic | |
389 | Architectures: none | |
390 | Type: vcpu ioctl | |
391 | Parameters: none) | |
392 | Returns: -1 on error | |
393 | ||
394 | Support for this has been removed. Use KVM_SET_GUEST_DEBUG instead. | |
395 | ||
68ba6974 | 396 | 4.18 KVM_GET_MSRS |
9c1b96e3 AK |
397 | |
398 | Capability: basic | |
399 | Architectures: x86 | |
400 | Type: vcpu ioctl | |
401 | Parameters: struct kvm_msrs (in/out) | |
402 | Returns: 0 on success, -1 on error | |
403 | ||
404 | Reads model-specific registers from the vcpu. Supported msr indices can | |
405 | be obtained using KVM_GET_MSR_INDEX_LIST. | |
406 | ||
407 | struct kvm_msrs { | |
408 | __u32 nmsrs; /* number of msrs in entries */ | |
409 | __u32 pad; | |
410 | ||
411 | struct kvm_msr_entry entries[0]; | |
412 | }; | |
413 | ||
414 | struct kvm_msr_entry { | |
415 | __u32 index; | |
416 | __u32 reserved; | |
417 | __u64 data; | |
418 | }; | |
419 | ||
420 | Application code should set the 'nmsrs' member (which indicates the | |
421 | size of the entries array) and the 'index' member of each array entry. | |
422 | kvm will fill in the 'data' member. | |
423 | ||
68ba6974 | 424 | 4.19 KVM_SET_MSRS |
9c1b96e3 AK |
425 | |
426 | Capability: basic | |
427 | Architectures: x86 | |
428 | Type: vcpu ioctl | |
429 | Parameters: struct kvm_msrs (in) | |
430 | Returns: 0 on success, -1 on error | |
431 | ||
432 | Writes model-specific registers to the vcpu. See KVM_GET_MSRS for the | |
433 | data structures. | |
434 | ||
435 | Application code should set the 'nmsrs' member (which indicates the | |
436 | size of the entries array), and the 'index' and 'data' members of each | |
437 | array entry. | |
438 | ||
68ba6974 | 439 | 4.20 KVM_SET_CPUID |
9c1b96e3 AK |
440 | |
441 | Capability: basic | |
442 | Architectures: x86 | |
443 | Type: vcpu ioctl | |
444 | Parameters: struct kvm_cpuid (in) | |
445 | Returns: 0 on success, -1 on error | |
446 | ||
447 | Defines the vcpu responses to the cpuid instruction. Applications | |
448 | should use the KVM_SET_CPUID2 ioctl if available. | |
449 | ||
450 | ||
451 | struct kvm_cpuid_entry { | |
452 | __u32 function; | |
453 | __u32 eax; | |
454 | __u32 ebx; | |
455 | __u32 ecx; | |
456 | __u32 edx; | |
457 | __u32 padding; | |
458 | }; | |
459 | ||
460 | /* for KVM_SET_CPUID */ | |
461 | struct kvm_cpuid { | |
462 | __u32 nent; | |
463 | __u32 padding; | |
464 | struct kvm_cpuid_entry entries[0]; | |
465 | }; | |
466 | ||
68ba6974 | 467 | 4.21 KVM_SET_SIGNAL_MASK |
9c1b96e3 AK |
468 | |
469 | Capability: basic | |
470 | Architectures: x86 | |
471 | Type: vcpu ioctl | |
472 | Parameters: struct kvm_signal_mask (in) | |
473 | Returns: 0 on success, -1 on error | |
474 | ||
475 | Defines which signals are blocked during execution of KVM_RUN. This | |
476 | signal mask temporarily overrides the threads signal mask. Any | |
477 | unblocked signal received (except SIGKILL and SIGSTOP, which retain | |
478 | their traditional behaviour) will cause KVM_RUN to return with -EINTR. | |
479 | ||
480 | Note the signal will only be delivered if not blocked by the original | |
481 | signal mask. | |
482 | ||
483 | /* for KVM_SET_SIGNAL_MASK */ | |
484 | struct kvm_signal_mask { | |
485 | __u32 len; | |
486 | __u8 sigset[0]; | |
487 | }; | |
488 | ||
68ba6974 | 489 | 4.22 KVM_GET_FPU |
9c1b96e3 AK |
490 | |
491 | Capability: basic | |
492 | Architectures: x86 | |
493 | Type: vcpu ioctl | |
494 | Parameters: struct kvm_fpu (out) | |
495 | Returns: 0 on success, -1 on error | |
496 | ||
497 | Reads the floating point state from the vcpu. | |
498 | ||
499 | /* for KVM_GET_FPU and KVM_SET_FPU */ | |
500 | struct kvm_fpu { | |
501 | __u8 fpr[8][16]; | |
502 | __u16 fcw; | |
503 | __u16 fsw; | |
504 | __u8 ftwx; /* in fxsave format */ | |
505 | __u8 pad1; | |
506 | __u16 last_opcode; | |
507 | __u64 last_ip; | |
508 | __u64 last_dp; | |
509 | __u8 xmm[16][16]; | |
510 | __u32 mxcsr; | |
511 | __u32 pad2; | |
512 | }; | |
513 | ||
68ba6974 | 514 | 4.23 KVM_SET_FPU |
9c1b96e3 AK |
515 | |
516 | Capability: basic | |
517 | Architectures: x86 | |
518 | Type: vcpu ioctl | |
519 | Parameters: struct kvm_fpu (in) | |
520 | Returns: 0 on success, -1 on error | |
521 | ||
522 | Writes the floating point state to the vcpu. | |
523 | ||
524 | /* for KVM_GET_FPU and KVM_SET_FPU */ | |
525 | struct kvm_fpu { | |
526 | __u8 fpr[8][16]; | |
527 | __u16 fcw; | |
528 | __u16 fsw; | |
529 | __u8 ftwx; /* in fxsave format */ | |
530 | __u8 pad1; | |
531 | __u16 last_opcode; | |
532 | __u64 last_ip; | |
533 | __u64 last_dp; | |
534 | __u8 xmm[16][16]; | |
535 | __u32 mxcsr; | |
536 | __u32 pad2; | |
537 | }; | |
538 | ||
68ba6974 | 539 | 4.24 KVM_CREATE_IRQCHIP |
5dadbfd6 AK |
540 | |
541 | Capability: KVM_CAP_IRQCHIP | |
542 | Architectures: x86, ia64 | |
543 | Type: vm ioctl | |
544 | Parameters: none | |
545 | Returns: 0 on success, -1 on error | |
546 | ||
547 | Creates an interrupt controller model in the kernel. On x86, creates a virtual | |
548 | ioapic, a virtual PIC (two PICs, nested), and sets up future vcpus to have a | |
549 | local APIC. IRQ routing for GSIs 0-15 is set to both PIC and IOAPIC; GSI 16-23 | |
550 | only go to the IOAPIC. On ia64, a IOSAPIC is created. | |
551 | ||
68ba6974 | 552 | 4.25 KVM_IRQ_LINE |
5dadbfd6 AK |
553 | |
554 | Capability: KVM_CAP_IRQCHIP | |
555 | Architectures: x86, ia64 | |
556 | Type: vm ioctl | |
557 | Parameters: struct kvm_irq_level | |
558 | Returns: 0 on success, -1 on error | |
559 | ||
560 | Sets the level of a GSI input to the interrupt controller model in the kernel. | |
561 | Requires that an interrupt controller model has been previously created with | |
562 | KVM_CREATE_IRQCHIP. Note that edge-triggered interrupts require the level | |
563 | to be set to 1 and then back to 0. | |
564 | ||
565 | struct kvm_irq_level { | |
566 | union { | |
567 | __u32 irq; /* GSI */ | |
568 | __s32 status; /* not used for KVM_IRQ_LEVEL */ | |
569 | }; | |
570 | __u32 level; /* 0 or 1 */ | |
571 | }; | |
572 | ||
68ba6974 | 573 | 4.26 KVM_GET_IRQCHIP |
5dadbfd6 AK |
574 | |
575 | Capability: KVM_CAP_IRQCHIP | |
576 | Architectures: x86, ia64 | |
577 | Type: vm ioctl | |
578 | Parameters: struct kvm_irqchip (in/out) | |
579 | Returns: 0 on success, -1 on error | |
580 | ||
581 | Reads the state of a kernel interrupt controller created with | |
582 | KVM_CREATE_IRQCHIP into a buffer provided by the caller. | |
583 | ||
584 | struct kvm_irqchip { | |
585 | __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */ | |
586 | __u32 pad; | |
587 | union { | |
588 | char dummy[512]; /* reserving space */ | |
589 | struct kvm_pic_state pic; | |
590 | struct kvm_ioapic_state ioapic; | |
591 | } chip; | |
592 | }; | |
593 | ||
68ba6974 | 594 | 4.27 KVM_SET_IRQCHIP |
5dadbfd6 AK |
595 | |
596 | Capability: KVM_CAP_IRQCHIP | |
597 | Architectures: x86, ia64 | |
598 | Type: vm ioctl | |
599 | Parameters: struct kvm_irqchip (in) | |
600 | Returns: 0 on success, -1 on error | |
601 | ||
602 | Sets the state of a kernel interrupt controller created with | |
603 | KVM_CREATE_IRQCHIP from a buffer provided by the caller. | |
604 | ||
605 | struct kvm_irqchip { | |
606 | __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */ | |
607 | __u32 pad; | |
608 | union { | |
609 | char dummy[512]; /* reserving space */ | |
610 | struct kvm_pic_state pic; | |
611 | struct kvm_ioapic_state ioapic; | |
612 | } chip; | |
613 | }; | |
614 | ||
68ba6974 | 615 | 4.28 KVM_XEN_HVM_CONFIG |
ffde22ac ES |
616 | |
617 | Capability: KVM_CAP_XEN_HVM | |
618 | Architectures: x86 | |
619 | Type: vm ioctl | |
620 | Parameters: struct kvm_xen_hvm_config (in) | |
621 | Returns: 0 on success, -1 on error | |
622 | ||
623 | Sets the MSR that the Xen HVM guest uses to initialize its hypercall | |
624 | page, and provides the starting address and size of the hypercall | |
625 | blobs in userspace. When the guest writes the MSR, kvm copies one | |
626 | page of a blob (32- or 64-bit, depending on the vcpu mode) to guest | |
627 | memory. | |
628 | ||
629 | struct kvm_xen_hvm_config { | |
630 | __u32 flags; | |
631 | __u32 msr; | |
632 | __u64 blob_addr_32; | |
633 | __u64 blob_addr_64; | |
634 | __u8 blob_size_32; | |
635 | __u8 blob_size_64; | |
636 | __u8 pad2[30]; | |
637 | }; | |
638 | ||
68ba6974 | 639 | 4.29 KVM_GET_CLOCK |
afbcf7ab GC |
640 | |
641 | Capability: KVM_CAP_ADJUST_CLOCK | |
642 | Architectures: x86 | |
643 | Type: vm ioctl | |
644 | Parameters: struct kvm_clock_data (out) | |
645 | Returns: 0 on success, -1 on error | |
646 | ||
647 | Gets the current timestamp of kvmclock as seen by the current guest. In | |
648 | conjunction with KVM_SET_CLOCK, it is used to ensure monotonicity on scenarios | |
649 | such as migration. | |
650 | ||
651 | struct kvm_clock_data { | |
652 | __u64 clock; /* kvmclock current value */ | |
653 | __u32 flags; | |
654 | __u32 pad[9]; | |
655 | }; | |
656 | ||
68ba6974 | 657 | 4.30 KVM_SET_CLOCK |
afbcf7ab GC |
658 | |
659 | Capability: KVM_CAP_ADJUST_CLOCK | |
660 | Architectures: x86 | |
661 | Type: vm ioctl | |
662 | Parameters: struct kvm_clock_data (in) | |
663 | Returns: 0 on success, -1 on error | |
664 | ||
2044892d | 665 | Sets the current timestamp of kvmclock to the value specified in its parameter. |
afbcf7ab GC |
666 | In conjunction with KVM_GET_CLOCK, it is used to ensure monotonicity on scenarios |
667 | such as migration. | |
668 | ||
669 | struct kvm_clock_data { | |
670 | __u64 clock; /* kvmclock current value */ | |
671 | __u32 flags; | |
672 | __u32 pad[9]; | |
673 | }; | |
674 | ||
68ba6974 | 675 | 4.31 KVM_GET_VCPU_EVENTS |
3cfc3092 JK |
676 | |
677 | Capability: KVM_CAP_VCPU_EVENTS | |
48005f64 | 678 | Extended by: KVM_CAP_INTR_SHADOW |
3cfc3092 JK |
679 | Architectures: x86 |
680 | Type: vm ioctl | |
681 | Parameters: struct kvm_vcpu_event (out) | |
682 | Returns: 0 on success, -1 on error | |
683 | ||
684 | Gets currently pending exceptions, interrupts, and NMIs as well as related | |
685 | states of the vcpu. | |
686 | ||
687 | struct kvm_vcpu_events { | |
688 | struct { | |
689 | __u8 injected; | |
690 | __u8 nr; | |
691 | __u8 has_error_code; | |
692 | __u8 pad; | |
693 | __u32 error_code; | |
694 | } exception; | |
695 | struct { | |
696 | __u8 injected; | |
697 | __u8 nr; | |
698 | __u8 soft; | |
48005f64 | 699 | __u8 shadow; |
3cfc3092 JK |
700 | } interrupt; |
701 | struct { | |
702 | __u8 injected; | |
703 | __u8 pending; | |
704 | __u8 masked; | |
705 | __u8 pad; | |
706 | } nmi; | |
707 | __u32 sipi_vector; | |
dab4b911 | 708 | __u32 flags; |
3cfc3092 JK |
709 | }; |
710 | ||
48005f64 JK |
711 | KVM_VCPUEVENT_VALID_SHADOW may be set in the flags field to signal that |
712 | interrupt.shadow contains a valid state. Otherwise, this field is undefined. | |
713 | ||
68ba6974 | 714 | 4.32 KVM_SET_VCPU_EVENTS |
3cfc3092 JK |
715 | |
716 | Capability: KVM_CAP_VCPU_EVENTS | |
48005f64 | 717 | Extended by: KVM_CAP_INTR_SHADOW |
3cfc3092 JK |
718 | Architectures: x86 |
719 | Type: vm ioctl | |
720 | Parameters: struct kvm_vcpu_event (in) | |
721 | Returns: 0 on success, -1 on error | |
722 | ||
723 | Set pending exceptions, interrupts, and NMIs as well as related states of the | |
724 | vcpu. | |
725 | ||
726 | See KVM_GET_VCPU_EVENTS for the data structure. | |
727 | ||
dab4b911 JK |
728 | Fields that may be modified asynchronously by running VCPUs can be excluded |
729 | from the update. These fields are nmi.pending and sipi_vector. Keep the | |
730 | corresponding bits in the flags field cleared to suppress overwriting the | |
731 | current in-kernel state. The bits are: | |
732 | ||
733 | KVM_VCPUEVENT_VALID_NMI_PENDING - transfer nmi.pending to the kernel | |
734 | KVM_VCPUEVENT_VALID_SIPI_VECTOR - transfer sipi_vector | |
735 | ||
48005f64 JK |
736 | If KVM_CAP_INTR_SHADOW is available, KVM_VCPUEVENT_VALID_SHADOW can be set in |
737 | the flags field to signal that interrupt.shadow contains a valid state and | |
738 | shall be written into the VCPU. | |
739 | ||
68ba6974 | 740 | 4.33 KVM_GET_DEBUGREGS |
a1efbe77 JK |
741 | |
742 | Capability: KVM_CAP_DEBUGREGS | |
743 | Architectures: x86 | |
744 | Type: vm ioctl | |
745 | Parameters: struct kvm_debugregs (out) | |
746 | Returns: 0 on success, -1 on error | |
747 | ||
748 | Reads debug registers from the vcpu. | |
749 | ||
750 | struct kvm_debugregs { | |
751 | __u64 db[4]; | |
752 | __u64 dr6; | |
753 | __u64 dr7; | |
754 | __u64 flags; | |
755 | __u64 reserved[9]; | |
756 | }; | |
757 | ||
68ba6974 | 758 | 4.34 KVM_SET_DEBUGREGS |
a1efbe77 JK |
759 | |
760 | Capability: KVM_CAP_DEBUGREGS | |
761 | Architectures: x86 | |
762 | Type: vm ioctl | |
763 | Parameters: struct kvm_debugregs (in) | |
764 | Returns: 0 on success, -1 on error | |
765 | ||
766 | Writes debug registers into the vcpu. | |
767 | ||
768 | See KVM_GET_DEBUGREGS for the data structure. The flags field is unused | |
769 | yet and must be cleared on entry. | |
770 | ||
68ba6974 | 771 | 4.35 KVM_SET_USER_MEMORY_REGION |
0f2d8f4d AK |
772 | |
773 | Capability: KVM_CAP_USER_MEM | |
774 | Architectures: all | |
775 | Type: vm ioctl | |
776 | Parameters: struct kvm_userspace_memory_region (in) | |
777 | Returns: 0 on success, -1 on error | |
778 | ||
779 | struct kvm_userspace_memory_region { | |
780 | __u32 slot; | |
781 | __u32 flags; | |
782 | __u64 guest_phys_addr; | |
783 | __u64 memory_size; /* bytes */ | |
784 | __u64 userspace_addr; /* start of the userspace allocated memory */ | |
785 | }; | |
786 | ||
787 | /* for kvm_memory_region::flags */ | |
788 | #define KVM_MEM_LOG_DIRTY_PAGES 1UL | |
789 | ||
790 | This ioctl allows the user to create or modify a guest physical memory | |
791 | slot. When changing an existing slot, it may be moved in the guest | |
792 | physical memory space, or its flags may be modified. It may not be | |
793 | resized. Slots may not overlap in guest physical address space. | |
794 | ||
795 | Memory for the region is taken starting at the address denoted by the | |
796 | field userspace_addr, which must point at user addressable memory for | |
797 | the entire memory slot size. Any object may back this memory, including | |
798 | anonymous memory, ordinary files, and hugetlbfs. | |
799 | ||
800 | It is recommended that the lower 21 bits of guest_phys_addr and userspace_addr | |
801 | be identical. This allows large pages in the guest to be backed by large | |
802 | pages in the host. | |
803 | ||
804 | The flags field supports just one flag, KVM_MEM_LOG_DIRTY_PAGES, which | |
805 | instructs kvm to keep track of writes to memory within the slot. See | |
806 | the KVM_GET_DIRTY_LOG ioctl. | |
807 | ||
808 | When the KVM_CAP_SYNC_MMU capability, changes in the backing of the memory | |
809 | region are automatically reflected into the guest. For example, an mmap() | |
810 | that affects the region will be made visible immediately. Another example | |
811 | is madvise(MADV_DROP). | |
812 | ||
813 | It is recommended to use this API instead of the KVM_SET_MEMORY_REGION ioctl. | |
814 | The KVM_SET_MEMORY_REGION does not allow fine grained control over memory | |
815 | allocation and is deprecated. | |
3cfc3092 | 816 | |
68ba6974 | 817 | 4.36 KVM_SET_TSS_ADDR |
8a5416db AK |
818 | |
819 | Capability: KVM_CAP_SET_TSS_ADDR | |
820 | Architectures: x86 | |
821 | Type: vm ioctl | |
822 | Parameters: unsigned long tss_address (in) | |
823 | Returns: 0 on success, -1 on error | |
824 | ||
825 | This ioctl defines the physical address of a three-page region in the guest | |
826 | physical address space. The region must be within the first 4GB of the | |
827 | guest physical address space and must not conflict with any memory slot | |
828 | or any mmio address. The guest may malfunction if it accesses this memory | |
829 | region. | |
830 | ||
831 | This ioctl is required on Intel-based hosts. This is needed on Intel hardware | |
832 | because of a quirk in the virtualization implementation (see the internals | |
833 | documentation when it pops into existence). | |
834 | ||
68ba6974 | 835 | 4.37 KVM_ENABLE_CAP |
71fbfd5f AG |
836 | |
837 | Capability: KVM_CAP_ENABLE_CAP | |
838 | Architectures: ppc | |
839 | Type: vcpu ioctl | |
840 | Parameters: struct kvm_enable_cap (in) | |
841 | Returns: 0 on success; -1 on error | |
842 | ||
843 | +Not all extensions are enabled by default. Using this ioctl the application | |
844 | can enable an extension, making it available to the guest. | |
845 | ||
846 | On systems that do not support this ioctl, it always fails. On systems that | |
847 | do support it, it only works for extensions that are supported for enablement. | |
848 | ||
849 | To check if a capability can be enabled, the KVM_CHECK_EXTENSION ioctl should | |
850 | be used. | |
851 | ||
852 | struct kvm_enable_cap { | |
853 | /* in */ | |
854 | __u32 cap; | |
855 | ||
856 | The capability that is supposed to get enabled. | |
857 | ||
858 | __u32 flags; | |
859 | ||
860 | A bitfield indicating future enhancements. Has to be 0 for now. | |
861 | ||
862 | __u64 args[4]; | |
863 | ||
864 | Arguments for enabling a feature. If a feature needs initial values to | |
865 | function properly, this is the place to put them. | |
866 | ||
867 | __u8 pad[64]; | |
868 | }; | |
869 | ||
68ba6974 | 870 | 4.38 KVM_GET_MP_STATE |
b843f065 AK |
871 | |
872 | Capability: KVM_CAP_MP_STATE | |
873 | Architectures: x86, ia64 | |
874 | Type: vcpu ioctl | |
875 | Parameters: struct kvm_mp_state (out) | |
876 | Returns: 0 on success; -1 on error | |
877 | ||
878 | struct kvm_mp_state { | |
879 | __u32 mp_state; | |
880 | }; | |
881 | ||
882 | Returns the vcpu's current "multiprocessing state" (though also valid on | |
883 | uniprocessor guests). | |
884 | ||
885 | Possible values are: | |
886 | ||
887 | - KVM_MP_STATE_RUNNABLE: the vcpu is currently running | |
888 | - KVM_MP_STATE_UNINITIALIZED: the vcpu is an application processor (AP) | |
889 | which has not yet received an INIT signal | |
890 | - KVM_MP_STATE_INIT_RECEIVED: the vcpu has received an INIT signal, and is | |
891 | now ready for a SIPI | |
892 | - KVM_MP_STATE_HALTED: the vcpu has executed a HLT instruction and | |
893 | is waiting for an interrupt | |
894 | - KVM_MP_STATE_SIPI_RECEIVED: the vcpu has just received a SIPI (vector | |
b595076a | 895 | accessible via KVM_GET_VCPU_EVENTS) |
b843f065 AK |
896 | |
897 | This ioctl is only useful after KVM_CREATE_IRQCHIP. Without an in-kernel | |
898 | irqchip, the multiprocessing state must be maintained by userspace. | |
899 | ||
68ba6974 | 900 | 4.39 KVM_SET_MP_STATE |
b843f065 AK |
901 | |
902 | Capability: KVM_CAP_MP_STATE | |
903 | Architectures: x86, ia64 | |
904 | Type: vcpu ioctl | |
905 | Parameters: struct kvm_mp_state (in) | |
906 | Returns: 0 on success; -1 on error | |
907 | ||
908 | Sets the vcpu's current "multiprocessing state"; see KVM_GET_MP_STATE for | |
909 | arguments. | |
910 | ||
911 | This ioctl is only useful after KVM_CREATE_IRQCHIP. Without an in-kernel | |
912 | irqchip, the multiprocessing state must be maintained by userspace. | |
913 | ||
68ba6974 | 914 | 4.40 KVM_SET_IDENTITY_MAP_ADDR |
47dbb84f AK |
915 | |
916 | Capability: KVM_CAP_SET_IDENTITY_MAP_ADDR | |
917 | Architectures: x86 | |
918 | Type: vm ioctl | |
919 | Parameters: unsigned long identity (in) | |
920 | Returns: 0 on success, -1 on error | |
921 | ||
922 | This ioctl defines the physical address of a one-page region in the guest | |
923 | physical address space. The region must be within the first 4GB of the | |
924 | guest physical address space and must not conflict with any memory slot | |
925 | or any mmio address. The guest may malfunction if it accesses this memory | |
926 | region. | |
927 | ||
928 | This ioctl is required on Intel-based hosts. This is needed on Intel hardware | |
929 | because of a quirk in the virtualization implementation (see the internals | |
930 | documentation when it pops into existence). | |
931 | ||
68ba6974 | 932 | 4.41 KVM_SET_BOOT_CPU_ID |
57bc24cf AK |
933 | |
934 | Capability: KVM_CAP_SET_BOOT_CPU_ID | |
935 | Architectures: x86, ia64 | |
936 | Type: vm ioctl | |
937 | Parameters: unsigned long vcpu_id | |
938 | Returns: 0 on success, -1 on error | |
939 | ||
940 | Define which vcpu is the Bootstrap Processor (BSP). Values are the same | |
941 | as the vcpu id in KVM_CREATE_VCPU. If this ioctl is not called, the default | |
942 | is vcpu 0. | |
943 | ||
68ba6974 | 944 | 4.42 KVM_GET_XSAVE |
2d5b5a66 SY |
945 | |
946 | Capability: KVM_CAP_XSAVE | |
947 | Architectures: x86 | |
948 | Type: vcpu ioctl | |
949 | Parameters: struct kvm_xsave (out) | |
950 | Returns: 0 on success, -1 on error | |
951 | ||
952 | struct kvm_xsave { | |
953 | __u32 region[1024]; | |
954 | }; | |
955 | ||
956 | This ioctl would copy current vcpu's xsave struct to the userspace. | |
957 | ||
68ba6974 | 958 | 4.43 KVM_SET_XSAVE |
2d5b5a66 SY |
959 | |
960 | Capability: KVM_CAP_XSAVE | |
961 | Architectures: x86 | |
962 | Type: vcpu ioctl | |
963 | Parameters: struct kvm_xsave (in) | |
964 | Returns: 0 on success, -1 on error | |
965 | ||
966 | struct kvm_xsave { | |
967 | __u32 region[1024]; | |
968 | }; | |
969 | ||
970 | This ioctl would copy userspace's xsave struct to the kernel. | |
971 | ||
68ba6974 | 972 | 4.44 KVM_GET_XCRS |
2d5b5a66 SY |
973 | |
974 | Capability: KVM_CAP_XCRS | |
975 | Architectures: x86 | |
976 | Type: vcpu ioctl | |
977 | Parameters: struct kvm_xcrs (out) | |
978 | Returns: 0 on success, -1 on error | |
979 | ||
980 | struct kvm_xcr { | |
981 | __u32 xcr; | |
982 | __u32 reserved; | |
983 | __u64 value; | |
984 | }; | |
985 | ||
986 | struct kvm_xcrs { | |
987 | __u32 nr_xcrs; | |
988 | __u32 flags; | |
989 | struct kvm_xcr xcrs[KVM_MAX_XCRS]; | |
990 | __u64 padding[16]; | |
991 | }; | |
992 | ||
993 | This ioctl would copy current vcpu's xcrs to the userspace. | |
994 | ||
68ba6974 | 995 | 4.45 KVM_SET_XCRS |
2d5b5a66 SY |
996 | |
997 | Capability: KVM_CAP_XCRS | |
998 | Architectures: x86 | |
999 | Type: vcpu ioctl | |
1000 | Parameters: struct kvm_xcrs (in) | |
1001 | Returns: 0 on success, -1 on error | |
1002 | ||
1003 | struct kvm_xcr { | |
1004 | __u32 xcr; | |
1005 | __u32 reserved; | |
1006 | __u64 value; | |
1007 | }; | |
1008 | ||
1009 | struct kvm_xcrs { | |
1010 | __u32 nr_xcrs; | |
1011 | __u32 flags; | |
1012 | struct kvm_xcr xcrs[KVM_MAX_XCRS]; | |
1013 | __u64 padding[16]; | |
1014 | }; | |
1015 | ||
1016 | This ioctl would set vcpu's xcr to the value userspace specified. | |
1017 | ||
68ba6974 | 1018 | 4.46 KVM_GET_SUPPORTED_CPUID |
d153513d AK |
1019 | |
1020 | Capability: KVM_CAP_EXT_CPUID | |
1021 | Architectures: x86 | |
1022 | Type: system ioctl | |
1023 | Parameters: struct kvm_cpuid2 (in/out) | |
1024 | Returns: 0 on success, -1 on error | |
1025 | ||
1026 | struct kvm_cpuid2 { | |
1027 | __u32 nent; | |
1028 | __u32 padding; | |
1029 | struct kvm_cpuid_entry2 entries[0]; | |
1030 | }; | |
1031 | ||
1032 | #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX 1 | |
1033 | #define KVM_CPUID_FLAG_STATEFUL_FUNC 2 | |
1034 | #define KVM_CPUID_FLAG_STATE_READ_NEXT 4 | |
1035 | ||
1036 | struct kvm_cpuid_entry2 { | |
1037 | __u32 function; | |
1038 | __u32 index; | |
1039 | __u32 flags; | |
1040 | __u32 eax; | |
1041 | __u32 ebx; | |
1042 | __u32 ecx; | |
1043 | __u32 edx; | |
1044 | __u32 padding[3]; | |
1045 | }; | |
1046 | ||
1047 | This ioctl returns x86 cpuid features which are supported by both the hardware | |
1048 | and kvm. Userspace can use the information returned by this ioctl to | |
1049 | construct cpuid information (for KVM_SET_CPUID2) that is consistent with | |
1050 | hardware, kernel, and userspace capabilities, and with user requirements (for | |
1051 | example, the user may wish to constrain cpuid to emulate older hardware, | |
1052 | or for feature consistency across a cluster). | |
1053 | ||
1054 | Userspace invokes KVM_GET_SUPPORTED_CPUID by passing a kvm_cpuid2 structure | |
1055 | with the 'nent' field indicating the number of entries in the variable-size | |
1056 | array 'entries'. If the number of entries is too low to describe the cpu | |
1057 | capabilities, an error (E2BIG) is returned. If the number is too high, | |
1058 | the 'nent' field is adjusted and an error (ENOMEM) is returned. If the | |
1059 | number is just right, the 'nent' field is adjusted to the number of valid | |
1060 | entries in the 'entries' array, which is then filled. | |
1061 | ||
1062 | The entries returned are the host cpuid as returned by the cpuid instruction, | |
c39cbd2a AK |
1063 | with unknown or unsupported features masked out. Some features (for example, |
1064 | x2apic), may not be present in the host cpu, but are exposed by kvm if it can | |
1065 | emulate them efficiently. The fields in each entry are defined as follows: | |
d153513d AK |
1066 | |
1067 | function: the eax value used to obtain the entry | |
1068 | index: the ecx value used to obtain the entry (for entries that are | |
1069 | affected by ecx) | |
1070 | flags: an OR of zero or more of the following: | |
1071 | KVM_CPUID_FLAG_SIGNIFCANT_INDEX: | |
1072 | if the index field is valid | |
1073 | KVM_CPUID_FLAG_STATEFUL_FUNC: | |
1074 | if cpuid for this function returns different values for successive | |
1075 | invocations; there will be several entries with the same function, | |
1076 | all with this flag set | |
1077 | KVM_CPUID_FLAG_STATE_READ_NEXT: | |
1078 | for KVM_CPUID_FLAG_STATEFUL_FUNC entries, set if this entry is | |
1079 | the first entry to be read by a cpu | |
1080 | eax, ebx, ecx, edx: the values returned by the cpuid instruction for | |
1081 | this function/index combination | |
1082 | ||
68ba6974 | 1083 | 4.47 KVM_PPC_GET_PVINFO |
15711e9c AG |
1084 | |
1085 | Capability: KVM_CAP_PPC_GET_PVINFO | |
1086 | Architectures: ppc | |
1087 | Type: vm ioctl | |
1088 | Parameters: struct kvm_ppc_pvinfo (out) | |
1089 | Returns: 0 on success, !0 on error | |
1090 | ||
1091 | struct kvm_ppc_pvinfo { | |
1092 | __u32 flags; | |
1093 | __u32 hcall[4]; | |
1094 | __u8 pad[108]; | |
1095 | }; | |
1096 | ||
1097 | This ioctl fetches PV specific information that need to be passed to the guest | |
1098 | using the device tree or other means from vm context. | |
1099 | ||
1100 | For now the only implemented piece of information distributed here is an array | |
1101 | of 4 instructions that make up a hypercall. | |
1102 | ||
1103 | If any additional field gets added to this structure later on, a bit for that | |
1104 | additional piece of information will be set in the flags bitmap. | |
1105 | ||
68ba6974 | 1106 | 4.48 KVM_ASSIGN_PCI_DEVICE |
49f48172 JK |
1107 | |
1108 | Capability: KVM_CAP_DEVICE_ASSIGNMENT | |
1109 | Architectures: x86 ia64 | |
1110 | Type: vm ioctl | |
1111 | Parameters: struct kvm_assigned_pci_dev (in) | |
1112 | Returns: 0 on success, -1 on error | |
1113 | ||
1114 | Assigns a host PCI device to the VM. | |
1115 | ||
1116 | struct kvm_assigned_pci_dev { | |
1117 | __u32 assigned_dev_id; | |
1118 | __u32 busnr; | |
1119 | __u32 devfn; | |
1120 | __u32 flags; | |
1121 | __u32 segnr; | |
1122 | union { | |
1123 | __u32 reserved[11]; | |
1124 | }; | |
1125 | }; | |
1126 | ||
1127 | The PCI device is specified by the triple segnr, busnr, and devfn. | |
1128 | Identification in succeeding service requests is done via assigned_dev_id. The | |
1129 | following flags are specified: | |
1130 | ||
1131 | /* Depends on KVM_CAP_IOMMU */ | |
1132 | #define KVM_DEV_ASSIGN_ENABLE_IOMMU (1 << 0) | |
1133 | ||
68ba6974 | 1134 | 4.49 KVM_DEASSIGN_PCI_DEVICE |
49f48172 JK |
1135 | |
1136 | Capability: KVM_CAP_DEVICE_DEASSIGNMENT | |
1137 | Architectures: x86 ia64 | |
1138 | Type: vm ioctl | |
1139 | Parameters: struct kvm_assigned_pci_dev (in) | |
1140 | Returns: 0 on success, -1 on error | |
1141 | ||
1142 | Ends PCI device assignment, releasing all associated resources. | |
1143 | ||
1144 | See KVM_CAP_DEVICE_ASSIGNMENT for the data structure. Only assigned_dev_id is | |
1145 | used in kvm_assigned_pci_dev to identify the device. | |
1146 | ||
68ba6974 | 1147 | 4.50 KVM_ASSIGN_DEV_IRQ |
49f48172 JK |
1148 | |
1149 | Capability: KVM_CAP_ASSIGN_DEV_IRQ | |
1150 | Architectures: x86 ia64 | |
1151 | Type: vm ioctl | |
1152 | Parameters: struct kvm_assigned_irq (in) | |
1153 | Returns: 0 on success, -1 on error | |
1154 | ||
1155 | Assigns an IRQ to a passed-through device. | |
1156 | ||
1157 | struct kvm_assigned_irq { | |
1158 | __u32 assigned_dev_id; | |
91e3d71d | 1159 | __u32 host_irq; /* ignored (legacy field) */ |
49f48172 JK |
1160 | __u32 guest_irq; |
1161 | __u32 flags; | |
1162 | union { | |
49f48172 JK |
1163 | __u32 reserved[12]; |
1164 | }; | |
1165 | }; | |
1166 | ||
1167 | The following flags are defined: | |
1168 | ||
1169 | #define KVM_DEV_IRQ_HOST_INTX (1 << 0) | |
1170 | #define KVM_DEV_IRQ_HOST_MSI (1 << 1) | |
1171 | #define KVM_DEV_IRQ_HOST_MSIX (1 << 2) | |
1172 | ||
1173 | #define KVM_DEV_IRQ_GUEST_INTX (1 << 8) | |
1174 | #define KVM_DEV_IRQ_GUEST_MSI (1 << 9) | |
1175 | #define KVM_DEV_IRQ_GUEST_MSIX (1 << 10) | |
1176 | ||
1177 | It is not valid to specify multiple types per host or guest IRQ. However, the | |
1178 | IRQ type of host and guest can differ or can even be null. | |
1179 | ||
68ba6974 | 1180 | 4.51 KVM_DEASSIGN_DEV_IRQ |
49f48172 JK |
1181 | |
1182 | Capability: KVM_CAP_ASSIGN_DEV_IRQ | |
1183 | Architectures: x86 ia64 | |
1184 | Type: vm ioctl | |
1185 | Parameters: struct kvm_assigned_irq (in) | |
1186 | Returns: 0 on success, -1 on error | |
1187 | ||
1188 | Ends an IRQ assignment to a passed-through device. | |
1189 | ||
1190 | See KVM_ASSIGN_DEV_IRQ for the data structure. The target device is specified | |
1191 | by assigned_dev_id, flags must correspond to the IRQ type specified on | |
1192 | KVM_ASSIGN_DEV_IRQ. Partial deassignment of host or guest IRQ is allowed. | |
1193 | ||
68ba6974 | 1194 | 4.52 KVM_SET_GSI_ROUTING |
49f48172 JK |
1195 | |
1196 | Capability: KVM_CAP_IRQ_ROUTING | |
1197 | Architectures: x86 ia64 | |
1198 | Type: vm ioctl | |
1199 | Parameters: struct kvm_irq_routing (in) | |
1200 | Returns: 0 on success, -1 on error | |
1201 | ||
1202 | Sets the GSI routing table entries, overwriting any previously set entries. | |
1203 | ||
1204 | struct kvm_irq_routing { | |
1205 | __u32 nr; | |
1206 | __u32 flags; | |
1207 | struct kvm_irq_routing_entry entries[0]; | |
1208 | }; | |
1209 | ||
1210 | No flags are specified so far, the corresponding field must be set to zero. | |
1211 | ||
1212 | struct kvm_irq_routing_entry { | |
1213 | __u32 gsi; | |
1214 | __u32 type; | |
1215 | __u32 flags; | |
1216 | __u32 pad; | |
1217 | union { | |
1218 | struct kvm_irq_routing_irqchip irqchip; | |
1219 | struct kvm_irq_routing_msi msi; | |
1220 | __u32 pad[8]; | |
1221 | } u; | |
1222 | }; | |
1223 | ||
1224 | /* gsi routing entry types */ | |
1225 | #define KVM_IRQ_ROUTING_IRQCHIP 1 | |
1226 | #define KVM_IRQ_ROUTING_MSI 2 | |
1227 | ||
1228 | No flags are specified so far, the corresponding field must be set to zero. | |
1229 | ||
1230 | struct kvm_irq_routing_irqchip { | |
1231 | __u32 irqchip; | |
1232 | __u32 pin; | |
1233 | }; | |
1234 | ||
1235 | struct kvm_irq_routing_msi { | |
1236 | __u32 address_lo; | |
1237 | __u32 address_hi; | |
1238 | __u32 data; | |
1239 | __u32 pad; | |
1240 | }; | |
1241 | ||
68ba6974 | 1242 | 4.53 KVM_ASSIGN_SET_MSIX_NR |
49f48172 JK |
1243 | |
1244 | Capability: KVM_CAP_DEVICE_MSIX | |
1245 | Architectures: x86 ia64 | |
1246 | Type: vm ioctl | |
1247 | Parameters: struct kvm_assigned_msix_nr (in) | |
1248 | Returns: 0 on success, -1 on error | |
1249 | ||
58f0964e JK |
1250 | Set the number of MSI-X interrupts for an assigned device. The number is |
1251 | reset again by terminating the MSI-X assignment of the device via | |
1252 | KVM_DEASSIGN_DEV_IRQ. Calling this service more than once at any earlier | |
1253 | point will fail. | |
49f48172 JK |
1254 | |
1255 | struct kvm_assigned_msix_nr { | |
1256 | __u32 assigned_dev_id; | |
1257 | __u16 entry_nr; | |
1258 | __u16 padding; | |
1259 | }; | |
1260 | ||
1261 | #define KVM_MAX_MSIX_PER_DEV 256 | |
1262 | ||
68ba6974 | 1263 | 4.54 KVM_ASSIGN_SET_MSIX_ENTRY |
49f48172 JK |
1264 | |
1265 | Capability: KVM_CAP_DEVICE_MSIX | |
1266 | Architectures: x86 ia64 | |
1267 | Type: vm ioctl | |
1268 | Parameters: struct kvm_assigned_msix_entry (in) | |
1269 | Returns: 0 on success, -1 on error | |
1270 | ||
1271 | Specifies the routing of an MSI-X assigned device interrupt to a GSI. Setting | |
1272 | the GSI vector to zero means disabling the interrupt. | |
1273 | ||
1274 | struct kvm_assigned_msix_entry { | |
1275 | __u32 assigned_dev_id; | |
1276 | __u32 gsi; | |
1277 | __u16 entry; /* The index of entry in the MSI-X table */ | |
1278 | __u16 padding[3]; | |
1279 | }; | |
1280 | ||
92a1f12d JR |
1281 | 4.54 KVM_SET_TSC_KHZ |
1282 | ||
1283 | Capability: KVM_CAP_TSC_CONTROL | |
1284 | Architectures: x86 | |
1285 | Type: vcpu ioctl | |
1286 | Parameters: virtual tsc_khz | |
1287 | Returns: 0 on success, -1 on error | |
1288 | ||
1289 | Specifies the tsc frequency for the virtual machine. The unit of the | |
1290 | frequency is KHz. | |
1291 | ||
1292 | 4.55 KVM_GET_TSC_KHZ | |
1293 | ||
1294 | Capability: KVM_CAP_GET_TSC_KHZ | |
1295 | Architectures: x86 | |
1296 | Type: vcpu ioctl | |
1297 | Parameters: none | |
1298 | Returns: virtual tsc-khz on success, negative value on error | |
1299 | ||
1300 | Returns the tsc frequency of the guest. The unit of the return value is | |
1301 | KHz. If the host has unstable tsc this ioctl returns -EIO instead as an | |
1302 | error. | |
1303 | ||
e7677933 AK |
1304 | 4.56 KVM_GET_LAPIC |
1305 | ||
1306 | Capability: KVM_CAP_IRQCHIP | |
1307 | Architectures: x86 | |
1308 | Type: vcpu ioctl | |
1309 | Parameters: struct kvm_lapic_state (out) | |
1310 | Returns: 0 on success, -1 on error | |
1311 | ||
1312 | #define KVM_APIC_REG_SIZE 0x400 | |
1313 | struct kvm_lapic_state { | |
1314 | char regs[KVM_APIC_REG_SIZE]; | |
1315 | }; | |
1316 | ||
1317 | Reads the Local APIC registers and copies them into the input argument. The | |
1318 | data format and layout are the same as documented in the architecture manual. | |
1319 | ||
1320 | 4.57 KVM_SET_LAPIC | |
1321 | ||
1322 | Capability: KVM_CAP_IRQCHIP | |
1323 | Architectures: x86 | |
1324 | Type: vcpu ioctl | |
1325 | Parameters: struct kvm_lapic_state (in) | |
1326 | Returns: 0 on success, -1 on error | |
1327 | ||
1328 | #define KVM_APIC_REG_SIZE 0x400 | |
1329 | struct kvm_lapic_state { | |
1330 | char regs[KVM_APIC_REG_SIZE]; | |
1331 | }; | |
1332 | ||
1333 | Copies the input argument into the the Local APIC registers. The data format | |
1334 | and layout are the same as documented in the architecture manual. | |
1335 | ||
7f4382e8 | 1336 | 4.58 KVM_IOEVENTFD |
55399a02 SL |
1337 | |
1338 | Capability: KVM_CAP_IOEVENTFD | |
1339 | Architectures: all | |
1340 | Type: vm ioctl | |
1341 | Parameters: struct kvm_ioeventfd (in) | |
1342 | Returns: 0 on success, !0 on error | |
1343 | ||
1344 | This ioctl attaches or detaches an ioeventfd to a legal pio/mmio address | |
1345 | within the guest. A guest write in the registered address will signal the | |
1346 | provided event instead of triggering an exit. | |
1347 | ||
1348 | struct kvm_ioeventfd { | |
1349 | __u64 datamatch; | |
1350 | __u64 addr; /* legal pio/mmio address */ | |
1351 | __u32 len; /* 1, 2, 4, or 8 bytes */ | |
1352 | __s32 fd; | |
1353 | __u32 flags; | |
1354 | __u8 pad[36]; | |
1355 | }; | |
1356 | ||
1357 | The following flags are defined: | |
1358 | ||
1359 | #define KVM_IOEVENTFD_FLAG_DATAMATCH (1 << kvm_ioeventfd_flag_nr_datamatch) | |
1360 | #define KVM_IOEVENTFD_FLAG_PIO (1 << kvm_ioeventfd_flag_nr_pio) | |
1361 | #define KVM_IOEVENTFD_FLAG_DEASSIGN (1 << kvm_ioeventfd_flag_nr_deassign) | |
1362 | ||
1363 | If datamatch flag is set, the event will be signaled only if the written value | |
1364 | to the registered address is equal to datamatch in struct kvm_ioeventfd. | |
1365 | ||
54738c09 DG |
1366 | 4.62 KVM_CREATE_SPAPR_TCE |
1367 | ||
1368 | Capability: KVM_CAP_SPAPR_TCE | |
1369 | Architectures: powerpc | |
1370 | Type: vm ioctl | |
1371 | Parameters: struct kvm_create_spapr_tce (in) | |
1372 | Returns: file descriptor for manipulating the created TCE table | |
1373 | ||
1374 | This creates a virtual TCE (translation control entry) table, which | |
1375 | is an IOMMU for PAPR-style virtual I/O. It is used to translate | |
1376 | logical addresses used in virtual I/O into guest physical addresses, | |
1377 | and provides a scatter/gather capability for PAPR virtual I/O. | |
1378 | ||
1379 | /* for KVM_CAP_SPAPR_TCE */ | |
1380 | struct kvm_create_spapr_tce { | |
1381 | __u64 liobn; | |
1382 | __u32 window_size; | |
1383 | }; | |
1384 | ||
1385 | The liobn field gives the logical IO bus number for which to create a | |
1386 | TCE table. The window_size field specifies the size of the DMA window | |
1387 | which this TCE table will translate - the table will contain one 64 | |
1388 | bit TCE entry for every 4kiB of the DMA window. | |
1389 | ||
1390 | When the guest issues an H_PUT_TCE hcall on a liobn for which a TCE | |
1391 | table has been created using this ioctl(), the kernel will handle it | |
1392 | in real mode, updating the TCE table. H_PUT_TCE calls for other | |
1393 | liobns will cause a vm exit and must be handled by userspace. | |
1394 | ||
1395 | The return value is a file descriptor which can be passed to mmap(2) | |
1396 | to map the created TCE table into userspace. This lets userspace read | |
1397 | the entries written by kernel-handled H_PUT_TCE calls, and also lets | |
1398 | userspace update the TCE table directly which is useful in some | |
1399 | circumstances. | |
1400 | ||
9c1b96e3 AK |
1401 | 5. The kvm_run structure |
1402 | ||
1403 | Application code obtains a pointer to the kvm_run structure by | |
1404 | mmap()ing a vcpu fd. From that point, application code can control | |
1405 | execution by changing fields in kvm_run prior to calling the KVM_RUN | |
1406 | ioctl, and obtain information about the reason KVM_RUN returned by | |
1407 | looking up structure members. | |
1408 | ||
1409 | struct kvm_run { | |
1410 | /* in */ | |
1411 | __u8 request_interrupt_window; | |
1412 | ||
1413 | Request that KVM_RUN return when it becomes possible to inject external | |
1414 | interrupts into the guest. Useful in conjunction with KVM_INTERRUPT. | |
1415 | ||
1416 | __u8 padding1[7]; | |
1417 | ||
1418 | /* out */ | |
1419 | __u32 exit_reason; | |
1420 | ||
1421 | When KVM_RUN has returned successfully (return value 0), this informs | |
1422 | application code why KVM_RUN has returned. Allowable values for this | |
1423 | field are detailed below. | |
1424 | ||
1425 | __u8 ready_for_interrupt_injection; | |
1426 | ||
1427 | If request_interrupt_window has been specified, this field indicates | |
1428 | an interrupt can be injected now with KVM_INTERRUPT. | |
1429 | ||
1430 | __u8 if_flag; | |
1431 | ||
1432 | The value of the current interrupt flag. Only valid if in-kernel | |
1433 | local APIC is not used. | |
1434 | ||
1435 | __u8 padding2[2]; | |
1436 | ||
1437 | /* in (pre_kvm_run), out (post_kvm_run) */ | |
1438 | __u64 cr8; | |
1439 | ||
1440 | The value of the cr8 register. Only valid if in-kernel local APIC is | |
1441 | not used. Both input and output. | |
1442 | ||
1443 | __u64 apic_base; | |
1444 | ||
1445 | The value of the APIC BASE msr. Only valid if in-kernel local | |
1446 | APIC is not used. Both input and output. | |
1447 | ||
1448 | union { | |
1449 | /* KVM_EXIT_UNKNOWN */ | |
1450 | struct { | |
1451 | __u64 hardware_exit_reason; | |
1452 | } hw; | |
1453 | ||
1454 | If exit_reason is KVM_EXIT_UNKNOWN, the vcpu has exited due to unknown | |
1455 | reasons. Further architecture-specific information is available in | |
1456 | hardware_exit_reason. | |
1457 | ||
1458 | /* KVM_EXIT_FAIL_ENTRY */ | |
1459 | struct { | |
1460 | __u64 hardware_entry_failure_reason; | |
1461 | } fail_entry; | |
1462 | ||
1463 | If exit_reason is KVM_EXIT_FAIL_ENTRY, the vcpu could not be run due | |
1464 | to unknown reasons. Further architecture-specific information is | |
1465 | available in hardware_entry_failure_reason. | |
1466 | ||
1467 | /* KVM_EXIT_EXCEPTION */ | |
1468 | struct { | |
1469 | __u32 exception; | |
1470 | __u32 error_code; | |
1471 | } ex; | |
1472 | ||
1473 | Unused. | |
1474 | ||
1475 | /* KVM_EXIT_IO */ | |
1476 | struct { | |
1477 | #define KVM_EXIT_IO_IN 0 | |
1478 | #define KVM_EXIT_IO_OUT 1 | |
1479 | __u8 direction; | |
1480 | __u8 size; /* bytes */ | |
1481 | __u16 port; | |
1482 | __u32 count; | |
1483 | __u64 data_offset; /* relative to kvm_run start */ | |
1484 | } io; | |
1485 | ||
2044892d | 1486 | If exit_reason is KVM_EXIT_IO, then the vcpu has |
9c1b96e3 AK |
1487 | executed a port I/O instruction which could not be satisfied by kvm. |
1488 | data_offset describes where the data is located (KVM_EXIT_IO_OUT) or | |
1489 | where kvm expects application code to place the data for the next | |
2044892d | 1490 | KVM_RUN invocation (KVM_EXIT_IO_IN). Data format is a packed array. |
9c1b96e3 AK |
1491 | |
1492 | struct { | |
1493 | struct kvm_debug_exit_arch arch; | |
1494 | } debug; | |
1495 | ||
1496 | Unused. | |
1497 | ||
1498 | /* KVM_EXIT_MMIO */ | |
1499 | struct { | |
1500 | __u64 phys_addr; | |
1501 | __u8 data[8]; | |
1502 | __u32 len; | |
1503 | __u8 is_write; | |
1504 | } mmio; | |
1505 | ||
2044892d | 1506 | If exit_reason is KVM_EXIT_MMIO, then the vcpu has |
9c1b96e3 AK |
1507 | executed a memory-mapped I/O instruction which could not be satisfied |
1508 | by kvm. The 'data' member contains the written data if 'is_write' is | |
1509 | true, and should be filled by application code otherwise. | |
1510 | ||
ad0a048b AG |
1511 | NOTE: For KVM_EXIT_IO, KVM_EXIT_MMIO and KVM_EXIT_OSI, the corresponding |
1512 | operations are complete (and guest state is consistent) only after userspace | |
1513 | has re-entered the kernel with KVM_RUN. The kernel side will first finish | |
67961344 MT |
1514 | incomplete operations and then check for pending signals. Userspace |
1515 | can re-enter the guest with an unmasked signal pending to complete | |
1516 | pending operations. | |
1517 | ||
9c1b96e3 AK |
1518 | /* KVM_EXIT_HYPERCALL */ |
1519 | struct { | |
1520 | __u64 nr; | |
1521 | __u64 args[6]; | |
1522 | __u64 ret; | |
1523 | __u32 longmode; | |
1524 | __u32 pad; | |
1525 | } hypercall; | |
1526 | ||
647dc49e AK |
1527 | Unused. This was once used for 'hypercall to userspace'. To implement |
1528 | such functionality, use KVM_EXIT_IO (x86) or KVM_EXIT_MMIO (all except s390). | |
1529 | Note KVM_EXIT_IO is significantly faster than KVM_EXIT_MMIO. | |
9c1b96e3 AK |
1530 | |
1531 | /* KVM_EXIT_TPR_ACCESS */ | |
1532 | struct { | |
1533 | __u64 rip; | |
1534 | __u32 is_write; | |
1535 | __u32 pad; | |
1536 | } tpr_access; | |
1537 | ||
1538 | To be documented (KVM_TPR_ACCESS_REPORTING). | |
1539 | ||
1540 | /* KVM_EXIT_S390_SIEIC */ | |
1541 | struct { | |
1542 | __u8 icptcode; | |
1543 | __u64 mask; /* psw upper half */ | |
1544 | __u64 addr; /* psw lower half */ | |
1545 | __u16 ipa; | |
1546 | __u32 ipb; | |
1547 | } s390_sieic; | |
1548 | ||
1549 | s390 specific. | |
1550 | ||
1551 | /* KVM_EXIT_S390_RESET */ | |
1552 | #define KVM_S390_RESET_POR 1 | |
1553 | #define KVM_S390_RESET_CLEAR 2 | |
1554 | #define KVM_S390_RESET_SUBSYSTEM 4 | |
1555 | #define KVM_S390_RESET_CPU_INIT 8 | |
1556 | #define KVM_S390_RESET_IPL 16 | |
1557 | __u64 s390_reset_flags; | |
1558 | ||
1559 | s390 specific. | |
1560 | ||
1561 | /* KVM_EXIT_DCR */ | |
1562 | struct { | |
1563 | __u32 dcrn; | |
1564 | __u32 data; | |
1565 | __u8 is_write; | |
1566 | } dcr; | |
1567 | ||
1568 | powerpc specific. | |
1569 | ||
ad0a048b AG |
1570 | /* KVM_EXIT_OSI */ |
1571 | struct { | |
1572 | __u64 gprs[32]; | |
1573 | } osi; | |
1574 | ||
1575 | MOL uses a special hypercall interface it calls 'OSI'. To enable it, we catch | |
1576 | hypercalls and exit with this exit struct that contains all the guest gprs. | |
1577 | ||
1578 | If exit_reason is KVM_EXIT_OSI, then the vcpu has triggered such a hypercall. | |
1579 | Userspace can now handle the hypercall and when it's done modify the gprs as | |
1580 | necessary. Upon guest entry all guest GPRs will then be replaced by the values | |
1581 | in this struct. | |
1582 | ||
de56a948 PM |
1583 | /* KVM_EXIT_PAPR_HCALL */ |
1584 | struct { | |
1585 | __u64 nr; | |
1586 | __u64 ret; | |
1587 | __u64 args[9]; | |
1588 | } papr_hcall; | |
1589 | ||
1590 | This is used on 64-bit PowerPC when emulating a pSeries partition, | |
1591 | e.g. with the 'pseries' machine type in qemu. It occurs when the | |
1592 | guest does a hypercall using the 'sc 1' instruction. The 'nr' field | |
1593 | contains the hypercall number (from the guest R3), and 'args' contains | |
1594 | the arguments (from the guest R4 - R12). Userspace should put the | |
1595 | return code in 'ret' and any extra returned values in args[]. | |
1596 | The possible hypercalls are defined in the Power Architecture Platform | |
1597 | Requirements (PAPR) document available from www.power.org (free | |
1598 | developer registration required to access it). | |
1599 | ||
9c1b96e3 AK |
1600 | /* Fix the size of the union. */ |
1601 | char padding[256]; | |
1602 | }; | |
1603 | }; |