]>
Commit | Line | Data |
---|---|---|
9c1b96e3 AK |
1 | The Definitive KVM (Kernel-based Virtual Machine) API Documentation |
2 | =================================================================== | |
3 | ||
4 | 1. General description | |
414fa985 | 5 | ---------------------- |
9c1b96e3 AK |
6 | |
7 | The kvm API is a set of ioctls that are issued to control various aspects | |
8 | of a virtual machine. The ioctls belong to three classes | |
9 | ||
10 | - System ioctls: These query and set global attributes which affect the | |
11 | whole kvm subsystem. In addition a system ioctl is used to create | |
12 | virtual machines | |
13 | ||
14 | - VM ioctls: These query and set attributes that affect an entire virtual | |
15 | machine, for example memory layout. In addition a VM ioctl is used to | |
16 | create virtual cpus (vcpus). | |
17 | ||
18 | Only run VM ioctls from the same process (address space) that was used | |
19 | to create the VM. | |
20 | ||
21 | - vcpu ioctls: These query and set attributes that control the operation | |
22 | of a single virtual cpu. | |
23 | ||
24 | Only run vcpu ioctls from the same thread that was used to create the | |
25 | vcpu. | |
26 | ||
414fa985 | 27 | |
2044892d | 28 | 2. File descriptors |
414fa985 | 29 | ------------------- |
9c1b96e3 AK |
30 | |
31 | The kvm API is centered around file descriptors. An initial | |
32 | open("/dev/kvm") obtains a handle to the kvm subsystem; this handle | |
33 | can be used to issue system ioctls. A KVM_CREATE_VM ioctl on this | |
2044892d | 34 | handle will create a VM file descriptor which can be used to issue VM |
9c1b96e3 AK |
35 | ioctls. A KVM_CREATE_VCPU ioctl on a VM fd will create a virtual cpu |
36 | and return a file descriptor pointing to it. Finally, ioctls on a vcpu | |
37 | fd can be used to control the vcpu, including the important task of | |
38 | actually running guest code. | |
39 | ||
40 | In general file descriptors can be migrated among processes by means | |
41 | of fork() and the SCM_RIGHTS facility of unix domain socket. These | |
42 | kinds of tricks are explicitly not supported by kvm. While they will | |
43 | not cause harm to the host, their actual behavior is not guaranteed by | |
44 | the API. The only supported use is one virtual machine per process, | |
45 | and one vcpu per thread. | |
46 | ||
414fa985 | 47 | |
9c1b96e3 | 48 | 3. Extensions |
414fa985 | 49 | ------------- |
9c1b96e3 AK |
50 | |
51 | As of Linux 2.6.22, the KVM ABI has been stabilized: no backward | |
52 | incompatible change are allowed. However, there is an extension | |
53 | facility that allows backward-compatible extensions to the API to be | |
54 | queried and used. | |
55 | ||
56 | The extension mechanism is not based on on the Linux version number. | |
57 | Instead, kvm defines extension identifiers and a facility to query | |
58 | whether a particular extension identifier is available. If it is, a | |
59 | set of ioctls is available for application use. | |
60 | ||
414fa985 | 61 | |
9c1b96e3 | 62 | 4. API description |
414fa985 | 63 | ------------------ |
9c1b96e3 AK |
64 | |
65 | This section describes ioctls that can be used to control kvm guests. | |
66 | For each ioctl, the following information is provided along with a | |
67 | description: | |
68 | ||
69 | Capability: which KVM extension provides this ioctl. Can be 'basic', | |
70 | which means that is will be provided by any kernel that supports | |
71 | API version 12 (see section 4.1), or a KVM_CAP_xyz constant, which | |
72 | means availability needs to be checked with KVM_CHECK_EXTENSION | |
73 | (see section 4.4). | |
74 | ||
75 | Architectures: which instruction set architectures provide this ioctl. | |
76 | x86 includes both i386 and x86_64. | |
77 | ||
78 | Type: system, vm, or vcpu. | |
79 | ||
80 | Parameters: what parameters are accepted by the ioctl. | |
81 | ||
82 | Returns: the return value. General error numbers (EBADF, ENOMEM, EINVAL) | |
83 | are not detailed, but errors with specific meanings are. | |
84 | ||
414fa985 | 85 | |
9c1b96e3 AK |
86 | 4.1 KVM_GET_API_VERSION |
87 | ||
88 | Capability: basic | |
89 | Architectures: all | |
90 | Type: system ioctl | |
91 | Parameters: none | |
92 | Returns: the constant KVM_API_VERSION (=12) | |
93 | ||
94 | This identifies the API version as the stable kvm API. It is not | |
95 | expected that this number will change. However, Linux 2.6.20 and | |
96 | 2.6.21 report earlier versions; these are not documented and not | |
97 | supported. Applications should refuse to run if KVM_GET_API_VERSION | |
98 | returns a value other than 12. If this check passes, all ioctls | |
99 | described as 'basic' will be available. | |
100 | ||
414fa985 | 101 | |
9c1b96e3 AK |
102 | 4.2 KVM_CREATE_VM |
103 | ||
104 | Capability: basic | |
105 | Architectures: all | |
106 | Type: system ioctl | |
e08b9637 | 107 | Parameters: machine type identifier (KVM_VM_*) |
9c1b96e3 AK |
108 | Returns: a VM fd that can be used to control the new virtual machine. |
109 | ||
110 | The new VM has no virtual cpus and no memory. An mmap() of a VM fd | |
111 | will access the virtual machine's physical address space; offset zero | |
112 | corresponds to guest physical address zero. Use of mmap() on a VM fd | |
113 | is discouraged if userspace memory allocation (KVM_CAP_USER_MEMORY) is | |
114 | available. | |
e08b9637 CO |
115 | You most certainly want to use 0 as machine type. |
116 | ||
117 | In order to create user controlled virtual machines on S390, check | |
118 | KVM_CAP_S390_UCONTROL and use the flag KVM_VM_S390_UCONTROL as | |
119 | privileged user (CAP_SYS_ADMIN). | |
9c1b96e3 | 120 | |
414fa985 | 121 | |
9c1b96e3 AK |
122 | 4.3 KVM_GET_MSR_INDEX_LIST |
123 | ||
124 | Capability: basic | |
125 | Architectures: x86 | |
126 | Type: system | |
127 | Parameters: struct kvm_msr_list (in/out) | |
128 | Returns: 0 on success; -1 on error | |
129 | Errors: | |
130 | E2BIG: the msr index list is to be to fit in the array specified by | |
131 | the user. | |
132 | ||
133 | struct kvm_msr_list { | |
134 | __u32 nmsrs; /* number of msrs in entries */ | |
135 | __u32 indices[0]; | |
136 | }; | |
137 | ||
138 | This ioctl returns the guest msrs that are supported. The list varies | |
139 | by kvm version and host processor, but does not change otherwise. The | |
140 | user fills in the size of the indices array in nmsrs, and in return | |
141 | kvm adjusts nmsrs to reflect the actual number of msrs and fills in | |
142 | the indices array with their numbers. | |
143 | ||
2e2602ca AK |
144 | Note: if kvm indicates supports MCE (KVM_CAP_MCE), then the MCE bank MSRs are |
145 | not returned in the MSR list, as different vcpus can have a different number | |
146 | of banks, as set via the KVM_X86_SETUP_MCE ioctl. | |
147 | ||
414fa985 | 148 | |
9c1b96e3 AK |
149 | 4.4 KVM_CHECK_EXTENSION |
150 | ||
151 | Capability: basic | |
152 | Architectures: all | |
153 | Type: system ioctl | |
154 | Parameters: extension identifier (KVM_CAP_*) | |
155 | Returns: 0 if unsupported; 1 (or some other positive integer) if supported | |
156 | ||
157 | The API allows the application to query about extensions to the core | |
158 | kvm API. Userspace passes an extension identifier (an integer) and | |
159 | receives an integer that describes the extension availability. | |
160 | Generally 0 means no and 1 means yes, but some extensions may report | |
161 | additional information in the integer return value. | |
162 | ||
414fa985 | 163 | |
9c1b96e3 AK |
164 | 4.5 KVM_GET_VCPU_MMAP_SIZE |
165 | ||
166 | Capability: basic | |
167 | Architectures: all | |
168 | Type: system ioctl | |
169 | Parameters: none | |
170 | Returns: size of vcpu mmap area, in bytes | |
171 | ||
172 | The KVM_RUN ioctl (cf.) communicates with userspace via a shared | |
173 | memory region. This ioctl returns the size of that region. See the | |
174 | KVM_RUN documentation for details. | |
175 | ||
414fa985 | 176 | |
9c1b96e3 AK |
177 | 4.6 KVM_SET_MEMORY_REGION |
178 | ||
179 | Capability: basic | |
180 | Architectures: all | |
181 | Type: vm ioctl | |
182 | Parameters: struct kvm_memory_region (in) | |
183 | Returns: 0 on success, -1 on error | |
184 | ||
b74a07be | 185 | This ioctl is obsolete and has been removed. |
9c1b96e3 | 186 | |
414fa985 | 187 | |
68ba6974 | 188 | 4.7 KVM_CREATE_VCPU |
9c1b96e3 AK |
189 | |
190 | Capability: basic | |
191 | Architectures: all | |
192 | Type: vm ioctl | |
193 | Parameters: vcpu id (apic id on x86) | |
194 | Returns: vcpu fd on success, -1 on error | |
195 | ||
196 | This API adds a vcpu to a virtual machine. The vcpu id is a small integer | |
8c3ba334 SL |
197 | in the range [0, max_vcpus). |
198 | ||
199 | The recommended max_vcpus value can be retrieved using the KVM_CAP_NR_VCPUS of | |
200 | the KVM_CHECK_EXTENSION ioctl() at run-time. | |
201 | The maximum possible value for max_vcpus can be retrieved using the | |
202 | KVM_CAP_MAX_VCPUS of the KVM_CHECK_EXTENSION ioctl() at run-time. | |
203 | ||
76d25402 PE |
204 | If the KVM_CAP_NR_VCPUS does not exist, you should assume that max_vcpus is 4 |
205 | cpus max. | |
8c3ba334 SL |
206 | If the KVM_CAP_MAX_VCPUS does not exist, you should assume that max_vcpus is |
207 | same as the value returned from KVM_CAP_NR_VCPUS. | |
9c1b96e3 | 208 | |
371fefd6 PM |
209 | On powerpc using book3s_hv mode, the vcpus are mapped onto virtual |
210 | threads in one or more virtual CPU cores. (This is because the | |
211 | hardware requires all the hardware threads in a CPU core to be in the | |
212 | same partition.) The KVM_CAP_PPC_SMT capability indicates the number | |
36442687 AK |
213 | of vcpus per virtual core (vcore). The vcore id is obtained by |
214 | dividing the vcpu id by the number of vcpus per vcore. The vcpus in a | |
215 | given vcore will always be in the same physical core as each other | |
216 | (though that might be a different physical core from time to time). | |
217 | Userspace can control the threading (SMT) mode of the guest by its | |
218 | allocation of vcpu ids. For example, if userspace wants | |
219 | single-threaded guest vcpus, it should make all vcpu ids be a multiple | |
220 | of the number of vcpus per vcore. | |
221 | ||
222 | On powerpc using book3s_hv mode, the vcpus are mapped onto virtual | |
223 | threads in one or more virtual CPU cores. (This is because the | |
224 | hardware requires all the hardware threads in a CPU core to be in the | |
225 | same partition.) The KVM_CAP_PPC_SMT capability indicates the number | |
371fefd6 PM |
226 | of vcpus per virtual core (vcore). The vcore id is obtained by |
227 | dividing the vcpu id by the number of vcpus per vcore. The vcpus in a | |
228 | given vcore will always be in the same physical core as each other | |
229 | (though that might be a different physical core from time to time). | |
230 | Userspace can control the threading (SMT) mode of the guest by its | |
231 | allocation of vcpu ids. For example, if userspace wants | |
232 | single-threaded guest vcpus, it should make all vcpu ids be a multiple | |
233 | of the number of vcpus per vcore. | |
234 | ||
5b1c1493 CO |
235 | For virtual cpus that have been created with S390 user controlled virtual |
236 | machines, the resulting vcpu fd can be memory mapped at page offset | |
237 | KVM_S390_SIE_PAGE_OFFSET in order to obtain a memory map of the virtual | |
238 | cpu's hardware control block. | |
239 | ||
414fa985 | 240 | |
68ba6974 | 241 | 4.8 KVM_GET_DIRTY_LOG (vm ioctl) |
9c1b96e3 AK |
242 | |
243 | Capability: basic | |
244 | Architectures: x86 | |
245 | Type: vm ioctl | |
246 | Parameters: struct kvm_dirty_log (in/out) | |
247 | Returns: 0 on success, -1 on error | |
248 | ||
249 | /* for KVM_GET_DIRTY_LOG */ | |
250 | struct kvm_dirty_log { | |
251 | __u32 slot; | |
252 | __u32 padding; | |
253 | union { | |
254 | void __user *dirty_bitmap; /* one bit per page */ | |
255 | __u64 padding; | |
256 | }; | |
257 | }; | |
258 | ||
259 | Given a memory slot, return a bitmap containing any pages dirtied | |
260 | since the last call to this ioctl. Bit 0 is the first page in the | |
261 | memory slot. Ensure the entire structure is cleared to avoid padding | |
262 | issues. | |
263 | ||
414fa985 | 264 | |
68ba6974 | 265 | 4.9 KVM_SET_MEMORY_ALIAS |
9c1b96e3 AK |
266 | |
267 | Capability: basic | |
268 | Architectures: x86 | |
269 | Type: vm ioctl | |
270 | Parameters: struct kvm_memory_alias (in) | |
271 | Returns: 0 (success), -1 (error) | |
272 | ||
a1f4d395 | 273 | This ioctl is obsolete and has been removed. |
9c1b96e3 | 274 | |
414fa985 | 275 | |
68ba6974 | 276 | 4.10 KVM_RUN |
9c1b96e3 AK |
277 | |
278 | Capability: basic | |
279 | Architectures: all | |
280 | Type: vcpu ioctl | |
281 | Parameters: none | |
282 | Returns: 0 on success, -1 on error | |
283 | Errors: | |
284 | EINTR: an unmasked signal is pending | |
285 | ||
286 | This ioctl is used to run a guest virtual cpu. While there are no | |
287 | explicit parameters, there is an implicit parameter block that can be | |
288 | obtained by mmap()ing the vcpu fd at offset 0, with the size given by | |
289 | KVM_GET_VCPU_MMAP_SIZE. The parameter block is formatted as a 'struct | |
290 | kvm_run' (see below). | |
291 | ||
414fa985 | 292 | |
68ba6974 | 293 | 4.11 KVM_GET_REGS |
9c1b96e3 AK |
294 | |
295 | Capability: basic | |
296 | Architectures: all | |
297 | Type: vcpu ioctl | |
298 | Parameters: struct kvm_regs (out) | |
299 | Returns: 0 on success, -1 on error | |
300 | ||
301 | Reads the general purpose registers from the vcpu. | |
302 | ||
303 | /* x86 */ | |
304 | struct kvm_regs { | |
305 | /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */ | |
306 | __u64 rax, rbx, rcx, rdx; | |
307 | __u64 rsi, rdi, rsp, rbp; | |
308 | __u64 r8, r9, r10, r11; | |
309 | __u64 r12, r13, r14, r15; | |
310 | __u64 rip, rflags; | |
311 | }; | |
312 | ||
414fa985 | 313 | |
68ba6974 | 314 | 4.12 KVM_SET_REGS |
9c1b96e3 AK |
315 | |
316 | Capability: basic | |
317 | Architectures: all | |
318 | Type: vcpu ioctl | |
319 | Parameters: struct kvm_regs (in) | |
320 | Returns: 0 on success, -1 on error | |
321 | ||
322 | Writes the general purpose registers into the vcpu. | |
323 | ||
324 | See KVM_GET_REGS for the data structure. | |
325 | ||
414fa985 | 326 | |
68ba6974 | 327 | 4.13 KVM_GET_SREGS |
9c1b96e3 AK |
328 | |
329 | Capability: basic | |
5ce941ee | 330 | Architectures: x86, ppc |
9c1b96e3 AK |
331 | Type: vcpu ioctl |
332 | Parameters: struct kvm_sregs (out) | |
333 | Returns: 0 on success, -1 on error | |
334 | ||
335 | Reads special registers from the vcpu. | |
336 | ||
337 | /* x86 */ | |
338 | struct kvm_sregs { | |
339 | struct kvm_segment cs, ds, es, fs, gs, ss; | |
340 | struct kvm_segment tr, ldt; | |
341 | struct kvm_dtable gdt, idt; | |
342 | __u64 cr0, cr2, cr3, cr4, cr8; | |
343 | __u64 efer; | |
344 | __u64 apic_base; | |
345 | __u64 interrupt_bitmap[(KVM_NR_INTERRUPTS + 63) / 64]; | |
346 | }; | |
347 | ||
5ce941ee SW |
348 | /* ppc -- see arch/powerpc/include/asm/kvm.h */ |
349 | ||
9c1b96e3 AK |
350 | interrupt_bitmap is a bitmap of pending external interrupts. At most |
351 | one bit may be set. This interrupt has been acknowledged by the APIC | |
352 | but not yet injected into the cpu core. | |
353 | ||
414fa985 | 354 | |
68ba6974 | 355 | 4.14 KVM_SET_SREGS |
9c1b96e3 AK |
356 | |
357 | Capability: basic | |
5ce941ee | 358 | Architectures: x86, ppc |
9c1b96e3 AK |
359 | Type: vcpu ioctl |
360 | Parameters: struct kvm_sregs (in) | |
361 | Returns: 0 on success, -1 on error | |
362 | ||
363 | Writes special registers into the vcpu. See KVM_GET_SREGS for the | |
364 | data structures. | |
365 | ||
414fa985 | 366 | |
68ba6974 | 367 | 4.15 KVM_TRANSLATE |
9c1b96e3 AK |
368 | |
369 | Capability: basic | |
370 | Architectures: x86 | |
371 | Type: vcpu ioctl | |
372 | Parameters: struct kvm_translation (in/out) | |
373 | Returns: 0 on success, -1 on error | |
374 | ||
375 | Translates a virtual address according to the vcpu's current address | |
376 | translation mode. | |
377 | ||
378 | struct kvm_translation { | |
379 | /* in */ | |
380 | __u64 linear_address; | |
381 | ||
382 | /* out */ | |
383 | __u64 physical_address; | |
384 | __u8 valid; | |
385 | __u8 writeable; | |
386 | __u8 usermode; | |
387 | __u8 pad[5]; | |
388 | }; | |
389 | ||
414fa985 | 390 | |
68ba6974 | 391 | 4.16 KVM_INTERRUPT |
9c1b96e3 AK |
392 | |
393 | Capability: basic | |
6f7a2bd4 | 394 | Architectures: x86, ppc |
9c1b96e3 AK |
395 | Type: vcpu ioctl |
396 | Parameters: struct kvm_interrupt (in) | |
397 | Returns: 0 on success, -1 on error | |
398 | ||
399 | Queues a hardware interrupt vector to be injected. This is only | |
6f7a2bd4 | 400 | useful if in-kernel local APIC or equivalent is not used. |
9c1b96e3 AK |
401 | |
402 | /* for KVM_INTERRUPT */ | |
403 | struct kvm_interrupt { | |
404 | /* in */ | |
405 | __u32 irq; | |
406 | }; | |
407 | ||
6f7a2bd4 AG |
408 | X86: |
409 | ||
9c1b96e3 AK |
410 | Note 'irq' is an interrupt vector, not an interrupt pin or line. |
411 | ||
6f7a2bd4 AG |
412 | PPC: |
413 | ||
414 | Queues an external interrupt to be injected. This ioctl is overleaded | |
415 | with 3 different irq values: | |
416 | ||
417 | a) KVM_INTERRUPT_SET | |
418 | ||
419 | This injects an edge type external interrupt into the guest once it's ready | |
420 | to receive interrupts. When injected, the interrupt is done. | |
421 | ||
422 | b) KVM_INTERRUPT_UNSET | |
423 | ||
424 | This unsets any pending interrupt. | |
425 | ||
426 | Only available with KVM_CAP_PPC_UNSET_IRQ. | |
427 | ||
428 | c) KVM_INTERRUPT_SET_LEVEL | |
429 | ||
430 | This injects a level type external interrupt into the guest context. The | |
431 | interrupt stays pending until a specific ioctl with KVM_INTERRUPT_UNSET | |
432 | is triggered. | |
433 | ||
434 | Only available with KVM_CAP_PPC_IRQ_LEVEL. | |
435 | ||
436 | Note that any value for 'irq' other than the ones stated above is invalid | |
437 | and incurs unexpected behavior. | |
438 | ||
414fa985 | 439 | |
68ba6974 | 440 | 4.17 KVM_DEBUG_GUEST |
9c1b96e3 AK |
441 | |
442 | Capability: basic | |
443 | Architectures: none | |
444 | Type: vcpu ioctl | |
445 | Parameters: none) | |
446 | Returns: -1 on error | |
447 | ||
448 | Support for this has been removed. Use KVM_SET_GUEST_DEBUG instead. | |
449 | ||
414fa985 | 450 | |
68ba6974 | 451 | 4.18 KVM_GET_MSRS |
9c1b96e3 AK |
452 | |
453 | Capability: basic | |
454 | Architectures: x86 | |
455 | Type: vcpu ioctl | |
456 | Parameters: struct kvm_msrs (in/out) | |
457 | Returns: 0 on success, -1 on error | |
458 | ||
459 | Reads model-specific registers from the vcpu. Supported msr indices can | |
460 | be obtained using KVM_GET_MSR_INDEX_LIST. | |
461 | ||
462 | struct kvm_msrs { | |
463 | __u32 nmsrs; /* number of msrs in entries */ | |
464 | __u32 pad; | |
465 | ||
466 | struct kvm_msr_entry entries[0]; | |
467 | }; | |
468 | ||
469 | struct kvm_msr_entry { | |
470 | __u32 index; | |
471 | __u32 reserved; | |
472 | __u64 data; | |
473 | }; | |
474 | ||
475 | Application code should set the 'nmsrs' member (which indicates the | |
476 | size of the entries array) and the 'index' member of each array entry. | |
477 | kvm will fill in the 'data' member. | |
478 | ||
414fa985 | 479 | |
68ba6974 | 480 | 4.19 KVM_SET_MSRS |
9c1b96e3 AK |
481 | |
482 | Capability: basic | |
483 | Architectures: x86 | |
484 | Type: vcpu ioctl | |
485 | Parameters: struct kvm_msrs (in) | |
486 | Returns: 0 on success, -1 on error | |
487 | ||
488 | Writes model-specific registers to the vcpu. See KVM_GET_MSRS for the | |
489 | data structures. | |
490 | ||
491 | Application code should set the 'nmsrs' member (which indicates the | |
492 | size of the entries array), and the 'index' and 'data' members of each | |
493 | array entry. | |
494 | ||
414fa985 | 495 | |
68ba6974 | 496 | 4.20 KVM_SET_CPUID |
9c1b96e3 AK |
497 | |
498 | Capability: basic | |
499 | Architectures: x86 | |
500 | Type: vcpu ioctl | |
501 | Parameters: struct kvm_cpuid (in) | |
502 | Returns: 0 on success, -1 on error | |
503 | ||
504 | Defines the vcpu responses to the cpuid instruction. Applications | |
505 | should use the KVM_SET_CPUID2 ioctl if available. | |
506 | ||
507 | ||
508 | struct kvm_cpuid_entry { | |
509 | __u32 function; | |
510 | __u32 eax; | |
511 | __u32 ebx; | |
512 | __u32 ecx; | |
513 | __u32 edx; | |
514 | __u32 padding; | |
515 | }; | |
516 | ||
517 | /* for KVM_SET_CPUID */ | |
518 | struct kvm_cpuid { | |
519 | __u32 nent; | |
520 | __u32 padding; | |
521 | struct kvm_cpuid_entry entries[0]; | |
522 | }; | |
523 | ||
414fa985 | 524 | |
68ba6974 | 525 | 4.21 KVM_SET_SIGNAL_MASK |
9c1b96e3 AK |
526 | |
527 | Capability: basic | |
528 | Architectures: x86 | |
529 | Type: vcpu ioctl | |
530 | Parameters: struct kvm_signal_mask (in) | |
531 | Returns: 0 on success, -1 on error | |
532 | ||
533 | Defines which signals are blocked during execution of KVM_RUN. This | |
534 | signal mask temporarily overrides the threads signal mask. Any | |
535 | unblocked signal received (except SIGKILL and SIGSTOP, which retain | |
536 | their traditional behaviour) will cause KVM_RUN to return with -EINTR. | |
537 | ||
538 | Note the signal will only be delivered if not blocked by the original | |
539 | signal mask. | |
540 | ||
541 | /* for KVM_SET_SIGNAL_MASK */ | |
542 | struct kvm_signal_mask { | |
543 | __u32 len; | |
544 | __u8 sigset[0]; | |
545 | }; | |
546 | ||
414fa985 | 547 | |
68ba6974 | 548 | 4.22 KVM_GET_FPU |
9c1b96e3 AK |
549 | |
550 | Capability: basic | |
551 | Architectures: x86 | |
552 | Type: vcpu ioctl | |
553 | Parameters: struct kvm_fpu (out) | |
554 | Returns: 0 on success, -1 on error | |
555 | ||
556 | Reads the floating point state from the vcpu. | |
557 | ||
558 | /* for KVM_GET_FPU and KVM_SET_FPU */ | |
559 | struct kvm_fpu { | |
560 | __u8 fpr[8][16]; | |
561 | __u16 fcw; | |
562 | __u16 fsw; | |
563 | __u8 ftwx; /* in fxsave format */ | |
564 | __u8 pad1; | |
565 | __u16 last_opcode; | |
566 | __u64 last_ip; | |
567 | __u64 last_dp; | |
568 | __u8 xmm[16][16]; | |
569 | __u32 mxcsr; | |
570 | __u32 pad2; | |
571 | }; | |
572 | ||
414fa985 | 573 | |
68ba6974 | 574 | 4.23 KVM_SET_FPU |
9c1b96e3 AK |
575 | |
576 | Capability: basic | |
577 | Architectures: x86 | |
578 | Type: vcpu ioctl | |
579 | Parameters: struct kvm_fpu (in) | |
580 | Returns: 0 on success, -1 on error | |
581 | ||
582 | Writes the floating point state to the vcpu. | |
583 | ||
584 | /* for KVM_GET_FPU and KVM_SET_FPU */ | |
585 | struct kvm_fpu { | |
586 | __u8 fpr[8][16]; | |
587 | __u16 fcw; | |
588 | __u16 fsw; | |
589 | __u8 ftwx; /* in fxsave format */ | |
590 | __u8 pad1; | |
591 | __u16 last_opcode; | |
592 | __u64 last_ip; | |
593 | __u64 last_dp; | |
594 | __u8 xmm[16][16]; | |
595 | __u32 mxcsr; | |
596 | __u32 pad2; | |
597 | }; | |
598 | ||
414fa985 | 599 | |
68ba6974 | 600 | 4.24 KVM_CREATE_IRQCHIP |
5dadbfd6 AK |
601 | |
602 | Capability: KVM_CAP_IRQCHIP | |
603 | Architectures: x86, ia64 | |
604 | Type: vm ioctl | |
605 | Parameters: none | |
606 | Returns: 0 on success, -1 on error | |
607 | ||
608 | Creates an interrupt controller model in the kernel. On x86, creates a virtual | |
609 | ioapic, a virtual PIC (two PICs, nested), and sets up future vcpus to have a | |
610 | local APIC. IRQ routing for GSIs 0-15 is set to both PIC and IOAPIC; GSI 16-23 | |
611 | only go to the IOAPIC. On ia64, a IOSAPIC is created. | |
612 | ||
414fa985 | 613 | |
68ba6974 | 614 | 4.25 KVM_IRQ_LINE |
5dadbfd6 AK |
615 | |
616 | Capability: KVM_CAP_IRQCHIP | |
617 | Architectures: x86, ia64 | |
618 | Type: vm ioctl | |
619 | Parameters: struct kvm_irq_level | |
620 | Returns: 0 on success, -1 on error | |
621 | ||
622 | Sets the level of a GSI input to the interrupt controller model in the kernel. | |
623 | Requires that an interrupt controller model has been previously created with | |
624 | KVM_CREATE_IRQCHIP. Note that edge-triggered interrupts require the level | |
625 | to be set to 1 and then back to 0. | |
626 | ||
627 | struct kvm_irq_level { | |
628 | union { | |
629 | __u32 irq; /* GSI */ | |
630 | __s32 status; /* not used for KVM_IRQ_LEVEL */ | |
631 | }; | |
632 | __u32 level; /* 0 or 1 */ | |
633 | }; | |
634 | ||
414fa985 | 635 | |
68ba6974 | 636 | 4.26 KVM_GET_IRQCHIP |
5dadbfd6 AK |
637 | |
638 | Capability: KVM_CAP_IRQCHIP | |
639 | Architectures: x86, ia64 | |
640 | Type: vm ioctl | |
641 | Parameters: struct kvm_irqchip (in/out) | |
642 | Returns: 0 on success, -1 on error | |
643 | ||
644 | Reads the state of a kernel interrupt controller created with | |
645 | KVM_CREATE_IRQCHIP into a buffer provided by the caller. | |
646 | ||
647 | struct kvm_irqchip { | |
648 | __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */ | |
649 | __u32 pad; | |
650 | union { | |
651 | char dummy[512]; /* reserving space */ | |
652 | struct kvm_pic_state pic; | |
653 | struct kvm_ioapic_state ioapic; | |
654 | } chip; | |
655 | }; | |
656 | ||
414fa985 | 657 | |
68ba6974 | 658 | 4.27 KVM_SET_IRQCHIP |
5dadbfd6 AK |
659 | |
660 | Capability: KVM_CAP_IRQCHIP | |
661 | Architectures: x86, ia64 | |
662 | Type: vm ioctl | |
663 | Parameters: struct kvm_irqchip (in) | |
664 | Returns: 0 on success, -1 on error | |
665 | ||
666 | Sets the state of a kernel interrupt controller created with | |
667 | KVM_CREATE_IRQCHIP from a buffer provided by the caller. | |
668 | ||
669 | struct kvm_irqchip { | |
670 | __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */ | |
671 | __u32 pad; | |
672 | union { | |
673 | char dummy[512]; /* reserving space */ | |
674 | struct kvm_pic_state pic; | |
675 | struct kvm_ioapic_state ioapic; | |
676 | } chip; | |
677 | }; | |
678 | ||
414fa985 | 679 | |
68ba6974 | 680 | 4.28 KVM_XEN_HVM_CONFIG |
ffde22ac ES |
681 | |
682 | Capability: KVM_CAP_XEN_HVM | |
683 | Architectures: x86 | |
684 | Type: vm ioctl | |
685 | Parameters: struct kvm_xen_hvm_config (in) | |
686 | Returns: 0 on success, -1 on error | |
687 | ||
688 | Sets the MSR that the Xen HVM guest uses to initialize its hypercall | |
689 | page, and provides the starting address and size of the hypercall | |
690 | blobs in userspace. When the guest writes the MSR, kvm copies one | |
691 | page of a blob (32- or 64-bit, depending on the vcpu mode) to guest | |
692 | memory. | |
693 | ||
694 | struct kvm_xen_hvm_config { | |
695 | __u32 flags; | |
696 | __u32 msr; | |
697 | __u64 blob_addr_32; | |
698 | __u64 blob_addr_64; | |
699 | __u8 blob_size_32; | |
700 | __u8 blob_size_64; | |
701 | __u8 pad2[30]; | |
702 | }; | |
703 | ||
414fa985 | 704 | |
68ba6974 | 705 | 4.29 KVM_GET_CLOCK |
afbcf7ab GC |
706 | |
707 | Capability: KVM_CAP_ADJUST_CLOCK | |
708 | Architectures: x86 | |
709 | Type: vm ioctl | |
710 | Parameters: struct kvm_clock_data (out) | |
711 | Returns: 0 on success, -1 on error | |
712 | ||
713 | Gets the current timestamp of kvmclock as seen by the current guest. In | |
714 | conjunction with KVM_SET_CLOCK, it is used to ensure monotonicity on scenarios | |
715 | such as migration. | |
716 | ||
717 | struct kvm_clock_data { | |
718 | __u64 clock; /* kvmclock current value */ | |
719 | __u32 flags; | |
720 | __u32 pad[9]; | |
721 | }; | |
722 | ||
414fa985 | 723 | |
68ba6974 | 724 | 4.30 KVM_SET_CLOCK |
afbcf7ab GC |
725 | |
726 | Capability: KVM_CAP_ADJUST_CLOCK | |
727 | Architectures: x86 | |
728 | Type: vm ioctl | |
729 | Parameters: struct kvm_clock_data (in) | |
730 | Returns: 0 on success, -1 on error | |
731 | ||
2044892d | 732 | Sets the current timestamp of kvmclock to the value specified in its parameter. |
afbcf7ab GC |
733 | In conjunction with KVM_GET_CLOCK, it is used to ensure monotonicity on scenarios |
734 | such as migration. | |
735 | ||
736 | struct kvm_clock_data { | |
737 | __u64 clock; /* kvmclock current value */ | |
738 | __u32 flags; | |
739 | __u32 pad[9]; | |
740 | }; | |
741 | ||
414fa985 | 742 | |
68ba6974 | 743 | 4.31 KVM_GET_VCPU_EVENTS |
3cfc3092 JK |
744 | |
745 | Capability: KVM_CAP_VCPU_EVENTS | |
48005f64 | 746 | Extended by: KVM_CAP_INTR_SHADOW |
3cfc3092 JK |
747 | Architectures: x86 |
748 | Type: vm ioctl | |
749 | Parameters: struct kvm_vcpu_event (out) | |
750 | Returns: 0 on success, -1 on error | |
751 | ||
752 | Gets currently pending exceptions, interrupts, and NMIs as well as related | |
753 | states of the vcpu. | |
754 | ||
755 | struct kvm_vcpu_events { | |
756 | struct { | |
757 | __u8 injected; | |
758 | __u8 nr; | |
759 | __u8 has_error_code; | |
760 | __u8 pad; | |
761 | __u32 error_code; | |
762 | } exception; | |
763 | struct { | |
764 | __u8 injected; | |
765 | __u8 nr; | |
766 | __u8 soft; | |
48005f64 | 767 | __u8 shadow; |
3cfc3092 JK |
768 | } interrupt; |
769 | struct { | |
770 | __u8 injected; | |
771 | __u8 pending; | |
772 | __u8 masked; | |
773 | __u8 pad; | |
774 | } nmi; | |
775 | __u32 sipi_vector; | |
dab4b911 | 776 | __u32 flags; |
3cfc3092 JK |
777 | }; |
778 | ||
48005f64 JK |
779 | KVM_VCPUEVENT_VALID_SHADOW may be set in the flags field to signal that |
780 | interrupt.shadow contains a valid state. Otherwise, this field is undefined. | |
781 | ||
414fa985 | 782 | |
68ba6974 | 783 | 4.32 KVM_SET_VCPU_EVENTS |
3cfc3092 JK |
784 | |
785 | Capability: KVM_CAP_VCPU_EVENTS | |
48005f64 | 786 | Extended by: KVM_CAP_INTR_SHADOW |
3cfc3092 JK |
787 | Architectures: x86 |
788 | Type: vm ioctl | |
789 | Parameters: struct kvm_vcpu_event (in) | |
790 | Returns: 0 on success, -1 on error | |
791 | ||
792 | Set pending exceptions, interrupts, and NMIs as well as related states of the | |
793 | vcpu. | |
794 | ||
795 | See KVM_GET_VCPU_EVENTS for the data structure. | |
796 | ||
dab4b911 JK |
797 | Fields that may be modified asynchronously by running VCPUs can be excluded |
798 | from the update. These fields are nmi.pending and sipi_vector. Keep the | |
799 | corresponding bits in the flags field cleared to suppress overwriting the | |
800 | current in-kernel state. The bits are: | |
801 | ||
802 | KVM_VCPUEVENT_VALID_NMI_PENDING - transfer nmi.pending to the kernel | |
803 | KVM_VCPUEVENT_VALID_SIPI_VECTOR - transfer sipi_vector | |
804 | ||
48005f64 JK |
805 | If KVM_CAP_INTR_SHADOW is available, KVM_VCPUEVENT_VALID_SHADOW can be set in |
806 | the flags field to signal that interrupt.shadow contains a valid state and | |
807 | shall be written into the VCPU. | |
808 | ||
414fa985 | 809 | |
68ba6974 | 810 | 4.33 KVM_GET_DEBUGREGS |
a1efbe77 JK |
811 | |
812 | Capability: KVM_CAP_DEBUGREGS | |
813 | Architectures: x86 | |
814 | Type: vm ioctl | |
815 | Parameters: struct kvm_debugregs (out) | |
816 | Returns: 0 on success, -1 on error | |
817 | ||
818 | Reads debug registers from the vcpu. | |
819 | ||
820 | struct kvm_debugregs { | |
821 | __u64 db[4]; | |
822 | __u64 dr6; | |
823 | __u64 dr7; | |
824 | __u64 flags; | |
825 | __u64 reserved[9]; | |
826 | }; | |
827 | ||
414fa985 | 828 | |
68ba6974 | 829 | 4.34 KVM_SET_DEBUGREGS |
a1efbe77 JK |
830 | |
831 | Capability: KVM_CAP_DEBUGREGS | |
832 | Architectures: x86 | |
833 | Type: vm ioctl | |
834 | Parameters: struct kvm_debugregs (in) | |
835 | Returns: 0 on success, -1 on error | |
836 | ||
837 | Writes debug registers into the vcpu. | |
838 | ||
839 | See KVM_GET_DEBUGREGS for the data structure. The flags field is unused | |
840 | yet and must be cleared on entry. | |
841 | ||
414fa985 | 842 | |
68ba6974 | 843 | 4.35 KVM_SET_USER_MEMORY_REGION |
0f2d8f4d AK |
844 | |
845 | Capability: KVM_CAP_USER_MEM | |
846 | Architectures: all | |
847 | Type: vm ioctl | |
848 | Parameters: struct kvm_userspace_memory_region (in) | |
849 | Returns: 0 on success, -1 on error | |
850 | ||
851 | struct kvm_userspace_memory_region { | |
852 | __u32 slot; | |
853 | __u32 flags; | |
854 | __u64 guest_phys_addr; | |
855 | __u64 memory_size; /* bytes */ | |
856 | __u64 userspace_addr; /* start of the userspace allocated memory */ | |
857 | }; | |
858 | ||
859 | /* for kvm_memory_region::flags */ | |
4d8b81ab XG |
860 | #define KVM_MEM_LOG_DIRTY_PAGES (1UL << 0) |
861 | #define KVM_MEM_READONLY (1UL << 1) | |
0f2d8f4d AK |
862 | |
863 | This ioctl allows the user to create or modify a guest physical memory | |
864 | slot. When changing an existing slot, it may be moved in the guest | |
865 | physical memory space, or its flags may be modified. It may not be | |
866 | resized. Slots may not overlap in guest physical address space. | |
867 | ||
868 | Memory for the region is taken starting at the address denoted by the | |
869 | field userspace_addr, which must point at user addressable memory for | |
870 | the entire memory slot size. Any object may back this memory, including | |
871 | anonymous memory, ordinary files, and hugetlbfs. | |
872 | ||
873 | It is recommended that the lower 21 bits of guest_phys_addr and userspace_addr | |
874 | be identical. This allows large pages in the guest to be backed by large | |
875 | pages in the host. | |
876 | ||
7efd8fa1 JK |
877 | The flags field supports two flag, KVM_MEM_LOG_DIRTY_PAGES, which instructs |
878 | kvm to keep track of writes to memory within the slot. See KVM_GET_DIRTY_LOG | |
879 | ioctl. The KVM_CAP_READONLY_MEM capability indicates the availability of the | |
880 | KVM_MEM_READONLY flag. When this flag is set for a memory region, KVM only | |
881 | allows read accesses. Writes will be posted to userspace as KVM_EXIT_MMIO | |
882 | exits. | |
883 | ||
884 | When the KVM_CAP_SYNC_MMU capability is available, changes in the backing of | |
885 | the memory region are automatically reflected into the guest. For example, an | |
886 | mmap() that affects the region will be made visible immediately. Another | |
887 | example is madvise(MADV_DROP). | |
0f2d8f4d AK |
888 | |
889 | It is recommended to use this API instead of the KVM_SET_MEMORY_REGION ioctl. | |
890 | The KVM_SET_MEMORY_REGION does not allow fine grained control over memory | |
891 | allocation and is deprecated. | |
3cfc3092 | 892 | |
414fa985 | 893 | |
68ba6974 | 894 | 4.36 KVM_SET_TSS_ADDR |
8a5416db AK |
895 | |
896 | Capability: KVM_CAP_SET_TSS_ADDR | |
897 | Architectures: x86 | |
898 | Type: vm ioctl | |
899 | Parameters: unsigned long tss_address (in) | |
900 | Returns: 0 on success, -1 on error | |
901 | ||
902 | This ioctl defines the physical address of a three-page region in the guest | |
903 | physical address space. The region must be within the first 4GB of the | |
904 | guest physical address space and must not conflict with any memory slot | |
905 | or any mmio address. The guest may malfunction if it accesses this memory | |
906 | region. | |
907 | ||
908 | This ioctl is required on Intel-based hosts. This is needed on Intel hardware | |
909 | because of a quirk in the virtualization implementation (see the internals | |
910 | documentation when it pops into existence). | |
911 | ||
414fa985 | 912 | |
68ba6974 | 913 | 4.37 KVM_ENABLE_CAP |
71fbfd5f AG |
914 | |
915 | Capability: KVM_CAP_ENABLE_CAP | |
916 | Architectures: ppc | |
917 | Type: vcpu ioctl | |
918 | Parameters: struct kvm_enable_cap (in) | |
919 | Returns: 0 on success; -1 on error | |
920 | ||
921 | +Not all extensions are enabled by default. Using this ioctl the application | |
922 | can enable an extension, making it available to the guest. | |
923 | ||
924 | On systems that do not support this ioctl, it always fails. On systems that | |
925 | do support it, it only works for extensions that are supported for enablement. | |
926 | ||
927 | To check if a capability can be enabled, the KVM_CHECK_EXTENSION ioctl should | |
928 | be used. | |
929 | ||
930 | struct kvm_enable_cap { | |
931 | /* in */ | |
932 | __u32 cap; | |
933 | ||
934 | The capability that is supposed to get enabled. | |
935 | ||
936 | __u32 flags; | |
937 | ||
938 | A bitfield indicating future enhancements. Has to be 0 for now. | |
939 | ||
940 | __u64 args[4]; | |
941 | ||
942 | Arguments for enabling a feature. If a feature needs initial values to | |
943 | function properly, this is the place to put them. | |
944 | ||
945 | __u8 pad[64]; | |
946 | }; | |
947 | ||
414fa985 | 948 | |
68ba6974 | 949 | 4.38 KVM_GET_MP_STATE |
b843f065 AK |
950 | |
951 | Capability: KVM_CAP_MP_STATE | |
952 | Architectures: x86, ia64 | |
953 | Type: vcpu ioctl | |
954 | Parameters: struct kvm_mp_state (out) | |
955 | Returns: 0 on success; -1 on error | |
956 | ||
957 | struct kvm_mp_state { | |
958 | __u32 mp_state; | |
959 | }; | |
960 | ||
961 | Returns the vcpu's current "multiprocessing state" (though also valid on | |
962 | uniprocessor guests). | |
963 | ||
964 | Possible values are: | |
965 | ||
966 | - KVM_MP_STATE_RUNNABLE: the vcpu is currently running | |
967 | - KVM_MP_STATE_UNINITIALIZED: the vcpu is an application processor (AP) | |
968 | which has not yet received an INIT signal | |
969 | - KVM_MP_STATE_INIT_RECEIVED: the vcpu has received an INIT signal, and is | |
970 | now ready for a SIPI | |
971 | - KVM_MP_STATE_HALTED: the vcpu has executed a HLT instruction and | |
972 | is waiting for an interrupt | |
973 | - KVM_MP_STATE_SIPI_RECEIVED: the vcpu has just received a SIPI (vector | |
b595076a | 974 | accessible via KVM_GET_VCPU_EVENTS) |
b843f065 AK |
975 | |
976 | This ioctl is only useful after KVM_CREATE_IRQCHIP. Without an in-kernel | |
977 | irqchip, the multiprocessing state must be maintained by userspace. | |
978 | ||
414fa985 | 979 | |
68ba6974 | 980 | 4.39 KVM_SET_MP_STATE |
b843f065 AK |
981 | |
982 | Capability: KVM_CAP_MP_STATE | |
983 | Architectures: x86, ia64 | |
984 | Type: vcpu ioctl | |
985 | Parameters: struct kvm_mp_state (in) | |
986 | Returns: 0 on success; -1 on error | |
987 | ||
988 | Sets the vcpu's current "multiprocessing state"; see KVM_GET_MP_STATE for | |
989 | arguments. | |
990 | ||
991 | This ioctl is only useful after KVM_CREATE_IRQCHIP. Without an in-kernel | |
992 | irqchip, the multiprocessing state must be maintained by userspace. | |
993 | ||
414fa985 | 994 | |
68ba6974 | 995 | 4.40 KVM_SET_IDENTITY_MAP_ADDR |
47dbb84f AK |
996 | |
997 | Capability: KVM_CAP_SET_IDENTITY_MAP_ADDR | |
998 | Architectures: x86 | |
999 | Type: vm ioctl | |
1000 | Parameters: unsigned long identity (in) | |
1001 | Returns: 0 on success, -1 on error | |
1002 | ||
1003 | This ioctl defines the physical address of a one-page region in the guest | |
1004 | physical address space. The region must be within the first 4GB of the | |
1005 | guest physical address space and must not conflict with any memory slot | |
1006 | or any mmio address. The guest may malfunction if it accesses this memory | |
1007 | region. | |
1008 | ||
1009 | This ioctl is required on Intel-based hosts. This is needed on Intel hardware | |
1010 | because of a quirk in the virtualization implementation (see the internals | |
1011 | documentation when it pops into existence). | |
1012 | ||
414fa985 | 1013 | |
68ba6974 | 1014 | 4.41 KVM_SET_BOOT_CPU_ID |
57bc24cf AK |
1015 | |
1016 | Capability: KVM_CAP_SET_BOOT_CPU_ID | |
1017 | Architectures: x86, ia64 | |
1018 | Type: vm ioctl | |
1019 | Parameters: unsigned long vcpu_id | |
1020 | Returns: 0 on success, -1 on error | |
1021 | ||
1022 | Define which vcpu is the Bootstrap Processor (BSP). Values are the same | |
1023 | as the vcpu id in KVM_CREATE_VCPU. If this ioctl is not called, the default | |
1024 | is vcpu 0. | |
1025 | ||
414fa985 | 1026 | |
68ba6974 | 1027 | 4.42 KVM_GET_XSAVE |
2d5b5a66 SY |
1028 | |
1029 | Capability: KVM_CAP_XSAVE | |
1030 | Architectures: x86 | |
1031 | Type: vcpu ioctl | |
1032 | Parameters: struct kvm_xsave (out) | |
1033 | Returns: 0 on success, -1 on error | |
1034 | ||
1035 | struct kvm_xsave { | |
1036 | __u32 region[1024]; | |
1037 | }; | |
1038 | ||
1039 | This ioctl would copy current vcpu's xsave struct to the userspace. | |
1040 | ||
414fa985 | 1041 | |
68ba6974 | 1042 | 4.43 KVM_SET_XSAVE |
2d5b5a66 SY |
1043 | |
1044 | Capability: KVM_CAP_XSAVE | |
1045 | Architectures: x86 | |
1046 | Type: vcpu ioctl | |
1047 | Parameters: struct kvm_xsave (in) | |
1048 | Returns: 0 on success, -1 on error | |
1049 | ||
1050 | struct kvm_xsave { | |
1051 | __u32 region[1024]; | |
1052 | }; | |
1053 | ||
1054 | This ioctl would copy userspace's xsave struct to the kernel. | |
1055 | ||
414fa985 | 1056 | |
68ba6974 | 1057 | 4.44 KVM_GET_XCRS |
2d5b5a66 SY |
1058 | |
1059 | Capability: KVM_CAP_XCRS | |
1060 | Architectures: x86 | |
1061 | Type: vcpu ioctl | |
1062 | Parameters: struct kvm_xcrs (out) | |
1063 | Returns: 0 on success, -1 on error | |
1064 | ||
1065 | struct kvm_xcr { | |
1066 | __u32 xcr; | |
1067 | __u32 reserved; | |
1068 | __u64 value; | |
1069 | }; | |
1070 | ||
1071 | struct kvm_xcrs { | |
1072 | __u32 nr_xcrs; | |
1073 | __u32 flags; | |
1074 | struct kvm_xcr xcrs[KVM_MAX_XCRS]; | |
1075 | __u64 padding[16]; | |
1076 | }; | |
1077 | ||
1078 | This ioctl would copy current vcpu's xcrs to the userspace. | |
1079 | ||
414fa985 | 1080 | |
68ba6974 | 1081 | 4.45 KVM_SET_XCRS |
2d5b5a66 SY |
1082 | |
1083 | Capability: KVM_CAP_XCRS | |
1084 | Architectures: x86 | |
1085 | Type: vcpu ioctl | |
1086 | Parameters: struct kvm_xcrs (in) | |
1087 | Returns: 0 on success, -1 on error | |
1088 | ||
1089 | struct kvm_xcr { | |
1090 | __u32 xcr; | |
1091 | __u32 reserved; | |
1092 | __u64 value; | |
1093 | }; | |
1094 | ||
1095 | struct kvm_xcrs { | |
1096 | __u32 nr_xcrs; | |
1097 | __u32 flags; | |
1098 | struct kvm_xcr xcrs[KVM_MAX_XCRS]; | |
1099 | __u64 padding[16]; | |
1100 | }; | |
1101 | ||
1102 | This ioctl would set vcpu's xcr to the value userspace specified. | |
1103 | ||
414fa985 | 1104 | |
68ba6974 | 1105 | 4.46 KVM_GET_SUPPORTED_CPUID |
d153513d AK |
1106 | |
1107 | Capability: KVM_CAP_EXT_CPUID | |
1108 | Architectures: x86 | |
1109 | Type: system ioctl | |
1110 | Parameters: struct kvm_cpuid2 (in/out) | |
1111 | Returns: 0 on success, -1 on error | |
1112 | ||
1113 | struct kvm_cpuid2 { | |
1114 | __u32 nent; | |
1115 | __u32 padding; | |
1116 | struct kvm_cpuid_entry2 entries[0]; | |
1117 | }; | |
1118 | ||
1119 | #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX 1 | |
1120 | #define KVM_CPUID_FLAG_STATEFUL_FUNC 2 | |
1121 | #define KVM_CPUID_FLAG_STATE_READ_NEXT 4 | |
1122 | ||
1123 | struct kvm_cpuid_entry2 { | |
1124 | __u32 function; | |
1125 | __u32 index; | |
1126 | __u32 flags; | |
1127 | __u32 eax; | |
1128 | __u32 ebx; | |
1129 | __u32 ecx; | |
1130 | __u32 edx; | |
1131 | __u32 padding[3]; | |
1132 | }; | |
1133 | ||
1134 | This ioctl returns x86 cpuid features which are supported by both the hardware | |
1135 | and kvm. Userspace can use the information returned by this ioctl to | |
1136 | construct cpuid information (for KVM_SET_CPUID2) that is consistent with | |
1137 | hardware, kernel, and userspace capabilities, and with user requirements (for | |
1138 | example, the user may wish to constrain cpuid to emulate older hardware, | |
1139 | or for feature consistency across a cluster). | |
1140 | ||
1141 | Userspace invokes KVM_GET_SUPPORTED_CPUID by passing a kvm_cpuid2 structure | |
1142 | with the 'nent' field indicating the number of entries in the variable-size | |
1143 | array 'entries'. If the number of entries is too low to describe the cpu | |
1144 | capabilities, an error (E2BIG) is returned. If the number is too high, | |
1145 | the 'nent' field is adjusted and an error (ENOMEM) is returned. If the | |
1146 | number is just right, the 'nent' field is adjusted to the number of valid | |
1147 | entries in the 'entries' array, which is then filled. | |
1148 | ||
1149 | The entries returned are the host cpuid as returned by the cpuid instruction, | |
c39cbd2a AK |
1150 | with unknown or unsupported features masked out. Some features (for example, |
1151 | x2apic), may not be present in the host cpu, but are exposed by kvm if it can | |
1152 | emulate them efficiently. The fields in each entry are defined as follows: | |
d153513d AK |
1153 | |
1154 | function: the eax value used to obtain the entry | |
1155 | index: the ecx value used to obtain the entry (for entries that are | |
1156 | affected by ecx) | |
1157 | flags: an OR of zero or more of the following: | |
1158 | KVM_CPUID_FLAG_SIGNIFCANT_INDEX: | |
1159 | if the index field is valid | |
1160 | KVM_CPUID_FLAG_STATEFUL_FUNC: | |
1161 | if cpuid for this function returns different values for successive | |
1162 | invocations; there will be several entries with the same function, | |
1163 | all with this flag set | |
1164 | KVM_CPUID_FLAG_STATE_READ_NEXT: | |
1165 | for KVM_CPUID_FLAG_STATEFUL_FUNC entries, set if this entry is | |
1166 | the first entry to be read by a cpu | |
1167 | eax, ebx, ecx, edx: the values returned by the cpuid instruction for | |
1168 | this function/index combination | |
1169 | ||
4d25a066 JK |
1170 | The TSC deadline timer feature (CPUID leaf 1, ecx[24]) is always returned |
1171 | as false, since the feature depends on KVM_CREATE_IRQCHIP for local APIC | |
1172 | support. Instead it is reported via | |
1173 | ||
1174 | ioctl(KVM_CHECK_EXTENSION, KVM_CAP_TSC_DEADLINE_TIMER) | |
1175 | ||
1176 | if that returns true and you use KVM_CREATE_IRQCHIP, or if you emulate the | |
1177 | feature in userspace, then you can enable the feature for KVM_SET_CPUID2. | |
1178 | ||
414fa985 | 1179 | |
68ba6974 | 1180 | 4.47 KVM_PPC_GET_PVINFO |
15711e9c AG |
1181 | |
1182 | Capability: KVM_CAP_PPC_GET_PVINFO | |
1183 | Architectures: ppc | |
1184 | Type: vm ioctl | |
1185 | Parameters: struct kvm_ppc_pvinfo (out) | |
1186 | Returns: 0 on success, !0 on error | |
1187 | ||
1188 | struct kvm_ppc_pvinfo { | |
1189 | __u32 flags; | |
1190 | __u32 hcall[4]; | |
1191 | __u8 pad[108]; | |
1192 | }; | |
1193 | ||
1194 | This ioctl fetches PV specific information that need to be passed to the guest | |
1195 | using the device tree or other means from vm context. | |
1196 | ||
9202e076 | 1197 | The hcall array defines 4 instructions that make up a hypercall. |
15711e9c AG |
1198 | |
1199 | If any additional field gets added to this structure later on, a bit for that | |
1200 | additional piece of information will be set in the flags bitmap. | |
1201 | ||
9202e076 LYB |
1202 | The flags bitmap is defined as: |
1203 | ||
1204 | /* the host supports the ePAPR idle hcall | |
1205 | #define KVM_PPC_PVINFO_FLAGS_EV_IDLE (1<<0) | |
414fa985 | 1206 | |
68ba6974 | 1207 | 4.48 KVM_ASSIGN_PCI_DEVICE |
49f48172 JK |
1208 | |
1209 | Capability: KVM_CAP_DEVICE_ASSIGNMENT | |
1210 | Architectures: x86 ia64 | |
1211 | Type: vm ioctl | |
1212 | Parameters: struct kvm_assigned_pci_dev (in) | |
1213 | Returns: 0 on success, -1 on error | |
1214 | ||
1215 | Assigns a host PCI device to the VM. | |
1216 | ||
1217 | struct kvm_assigned_pci_dev { | |
1218 | __u32 assigned_dev_id; | |
1219 | __u32 busnr; | |
1220 | __u32 devfn; | |
1221 | __u32 flags; | |
1222 | __u32 segnr; | |
1223 | union { | |
1224 | __u32 reserved[11]; | |
1225 | }; | |
1226 | }; | |
1227 | ||
1228 | The PCI device is specified by the triple segnr, busnr, and devfn. | |
1229 | Identification in succeeding service requests is done via assigned_dev_id. The | |
1230 | following flags are specified: | |
1231 | ||
1232 | /* Depends on KVM_CAP_IOMMU */ | |
1233 | #define KVM_DEV_ASSIGN_ENABLE_IOMMU (1 << 0) | |
07700a94 JK |
1234 | /* The following two depend on KVM_CAP_PCI_2_3 */ |
1235 | #define KVM_DEV_ASSIGN_PCI_2_3 (1 << 1) | |
1236 | #define KVM_DEV_ASSIGN_MASK_INTX (1 << 2) | |
1237 | ||
1238 | If KVM_DEV_ASSIGN_PCI_2_3 is set, the kernel will manage legacy INTx interrupts | |
1239 | via the PCI-2.3-compliant device-level mask, thus enable IRQ sharing with other | |
1240 | assigned devices or host devices. KVM_DEV_ASSIGN_MASK_INTX specifies the | |
1241 | guest's view on the INTx mask, see KVM_ASSIGN_SET_INTX_MASK for details. | |
49f48172 | 1242 | |
42387373 AW |
1243 | The KVM_DEV_ASSIGN_ENABLE_IOMMU flag is a mandatory option to ensure |
1244 | isolation of the device. Usages not specifying this flag are deprecated. | |
1245 | ||
3d27e23b AW |
1246 | Only PCI header type 0 devices with PCI BAR resources are supported by |
1247 | device assignment. The user requesting this ioctl must have read/write | |
1248 | access to the PCI sysfs resource files associated with the device. | |
1249 | ||
414fa985 | 1250 | |
68ba6974 | 1251 | 4.49 KVM_DEASSIGN_PCI_DEVICE |
49f48172 JK |
1252 | |
1253 | Capability: KVM_CAP_DEVICE_DEASSIGNMENT | |
1254 | Architectures: x86 ia64 | |
1255 | Type: vm ioctl | |
1256 | Parameters: struct kvm_assigned_pci_dev (in) | |
1257 | Returns: 0 on success, -1 on error | |
1258 | ||
1259 | Ends PCI device assignment, releasing all associated resources. | |
1260 | ||
1261 | See KVM_CAP_DEVICE_ASSIGNMENT for the data structure. Only assigned_dev_id is | |
1262 | used in kvm_assigned_pci_dev to identify the device. | |
1263 | ||
414fa985 | 1264 | |
68ba6974 | 1265 | 4.50 KVM_ASSIGN_DEV_IRQ |
49f48172 JK |
1266 | |
1267 | Capability: KVM_CAP_ASSIGN_DEV_IRQ | |
1268 | Architectures: x86 ia64 | |
1269 | Type: vm ioctl | |
1270 | Parameters: struct kvm_assigned_irq (in) | |
1271 | Returns: 0 on success, -1 on error | |
1272 | ||
1273 | Assigns an IRQ to a passed-through device. | |
1274 | ||
1275 | struct kvm_assigned_irq { | |
1276 | __u32 assigned_dev_id; | |
91e3d71d | 1277 | __u32 host_irq; /* ignored (legacy field) */ |
49f48172 JK |
1278 | __u32 guest_irq; |
1279 | __u32 flags; | |
1280 | union { | |
49f48172 JK |
1281 | __u32 reserved[12]; |
1282 | }; | |
1283 | }; | |
1284 | ||
1285 | The following flags are defined: | |
1286 | ||
1287 | #define KVM_DEV_IRQ_HOST_INTX (1 << 0) | |
1288 | #define KVM_DEV_IRQ_HOST_MSI (1 << 1) | |
1289 | #define KVM_DEV_IRQ_HOST_MSIX (1 << 2) | |
1290 | ||
1291 | #define KVM_DEV_IRQ_GUEST_INTX (1 << 8) | |
1292 | #define KVM_DEV_IRQ_GUEST_MSI (1 << 9) | |
1293 | #define KVM_DEV_IRQ_GUEST_MSIX (1 << 10) | |
1294 | ||
1295 | It is not valid to specify multiple types per host or guest IRQ. However, the | |
1296 | IRQ type of host and guest can differ or can even be null. | |
1297 | ||
414fa985 | 1298 | |
68ba6974 | 1299 | 4.51 KVM_DEASSIGN_DEV_IRQ |
49f48172 JK |
1300 | |
1301 | Capability: KVM_CAP_ASSIGN_DEV_IRQ | |
1302 | Architectures: x86 ia64 | |
1303 | Type: vm ioctl | |
1304 | Parameters: struct kvm_assigned_irq (in) | |
1305 | Returns: 0 on success, -1 on error | |
1306 | ||
1307 | Ends an IRQ assignment to a passed-through device. | |
1308 | ||
1309 | See KVM_ASSIGN_DEV_IRQ for the data structure. The target device is specified | |
1310 | by assigned_dev_id, flags must correspond to the IRQ type specified on | |
1311 | KVM_ASSIGN_DEV_IRQ. Partial deassignment of host or guest IRQ is allowed. | |
1312 | ||
414fa985 | 1313 | |
68ba6974 | 1314 | 4.52 KVM_SET_GSI_ROUTING |
49f48172 JK |
1315 | |
1316 | Capability: KVM_CAP_IRQ_ROUTING | |
1317 | Architectures: x86 ia64 | |
1318 | Type: vm ioctl | |
1319 | Parameters: struct kvm_irq_routing (in) | |
1320 | Returns: 0 on success, -1 on error | |
1321 | ||
1322 | Sets the GSI routing table entries, overwriting any previously set entries. | |
1323 | ||
1324 | struct kvm_irq_routing { | |
1325 | __u32 nr; | |
1326 | __u32 flags; | |
1327 | struct kvm_irq_routing_entry entries[0]; | |
1328 | }; | |
1329 | ||
1330 | No flags are specified so far, the corresponding field must be set to zero. | |
1331 | ||
1332 | struct kvm_irq_routing_entry { | |
1333 | __u32 gsi; | |
1334 | __u32 type; | |
1335 | __u32 flags; | |
1336 | __u32 pad; | |
1337 | union { | |
1338 | struct kvm_irq_routing_irqchip irqchip; | |
1339 | struct kvm_irq_routing_msi msi; | |
1340 | __u32 pad[8]; | |
1341 | } u; | |
1342 | }; | |
1343 | ||
1344 | /* gsi routing entry types */ | |
1345 | #define KVM_IRQ_ROUTING_IRQCHIP 1 | |
1346 | #define KVM_IRQ_ROUTING_MSI 2 | |
1347 | ||
1348 | No flags are specified so far, the corresponding field must be set to zero. | |
1349 | ||
1350 | struct kvm_irq_routing_irqchip { | |
1351 | __u32 irqchip; | |
1352 | __u32 pin; | |
1353 | }; | |
1354 | ||
1355 | struct kvm_irq_routing_msi { | |
1356 | __u32 address_lo; | |
1357 | __u32 address_hi; | |
1358 | __u32 data; | |
1359 | __u32 pad; | |
1360 | }; | |
1361 | ||
414fa985 | 1362 | |
68ba6974 | 1363 | 4.53 KVM_ASSIGN_SET_MSIX_NR |
49f48172 JK |
1364 | |
1365 | Capability: KVM_CAP_DEVICE_MSIX | |
1366 | Architectures: x86 ia64 | |
1367 | Type: vm ioctl | |
1368 | Parameters: struct kvm_assigned_msix_nr (in) | |
1369 | Returns: 0 on success, -1 on error | |
1370 | ||
58f0964e JK |
1371 | Set the number of MSI-X interrupts for an assigned device. The number is |
1372 | reset again by terminating the MSI-X assignment of the device via | |
1373 | KVM_DEASSIGN_DEV_IRQ. Calling this service more than once at any earlier | |
1374 | point will fail. | |
49f48172 JK |
1375 | |
1376 | struct kvm_assigned_msix_nr { | |
1377 | __u32 assigned_dev_id; | |
1378 | __u16 entry_nr; | |
1379 | __u16 padding; | |
1380 | }; | |
1381 | ||
1382 | #define KVM_MAX_MSIX_PER_DEV 256 | |
1383 | ||
414fa985 | 1384 | |
68ba6974 | 1385 | 4.54 KVM_ASSIGN_SET_MSIX_ENTRY |
49f48172 JK |
1386 | |
1387 | Capability: KVM_CAP_DEVICE_MSIX | |
1388 | Architectures: x86 ia64 | |
1389 | Type: vm ioctl | |
1390 | Parameters: struct kvm_assigned_msix_entry (in) | |
1391 | Returns: 0 on success, -1 on error | |
1392 | ||
1393 | Specifies the routing of an MSI-X assigned device interrupt to a GSI. Setting | |
1394 | the GSI vector to zero means disabling the interrupt. | |
1395 | ||
1396 | struct kvm_assigned_msix_entry { | |
1397 | __u32 assigned_dev_id; | |
1398 | __u32 gsi; | |
1399 | __u16 entry; /* The index of entry in the MSI-X table */ | |
1400 | __u16 padding[3]; | |
1401 | }; | |
1402 | ||
414fa985 JK |
1403 | |
1404 | 4.55 KVM_SET_TSC_KHZ | |
92a1f12d JR |
1405 | |
1406 | Capability: KVM_CAP_TSC_CONTROL | |
1407 | Architectures: x86 | |
1408 | Type: vcpu ioctl | |
1409 | Parameters: virtual tsc_khz | |
1410 | Returns: 0 on success, -1 on error | |
1411 | ||
1412 | Specifies the tsc frequency for the virtual machine. The unit of the | |
1413 | frequency is KHz. | |
1414 | ||
414fa985 JK |
1415 | |
1416 | 4.56 KVM_GET_TSC_KHZ | |
92a1f12d JR |
1417 | |
1418 | Capability: KVM_CAP_GET_TSC_KHZ | |
1419 | Architectures: x86 | |
1420 | Type: vcpu ioctl | |
1421 | Parameters: none | |
1422 | Returns: virtual tsc-khz on success, negative value on error | |
1423 | ||
1424 | Returns the tsc frequency of the guest. The unit of the return value is | |
1425 | KHz. If the host has unstable tsc this ioctl returns -EIO instead as an | |
1426 | error. | |
1427 | ||
414fa985 JK |
1428 | |
1429 | 4.57 KVM_GET_LAPIC | |
e7677933 AK |
1430 | |
1431 | Capability: KVM_CAP_IRQCHIP | |
1432 | Architectures: x86 | |
1433 | Type: vcpu ioctl | |
1434 | Parameters: struct kvm_lapic_state (out) | |
1435 | Returns: 0 on success, -1 on error | |
1436 | ||
1437 | #define KVM_APIC_REG_SIZE 0x400 | |
1438 | struct kvm_lapic_state { | |
1439 | char regs[KVM_APIC_REG_SIZE]; | |
1440 | }; | |
1441 | ||
1442 | Reads the Local APIC registers and copies them into the input argument. The | |
1443 | data format and layout are the same as documented in the architecture manual. | |
1444 | ||
414fa985 JK |
1445 | |
1446 | 4.58 KVM_SET_LAPIC | |
e7677933 AK |
1447 | |
1448 | Capability: KVM_CAP_IRQCHIP | |
1449 | Architectures: x86 | |
1450 | Type: vcpu ioctl | |
1451 | Parameters: struct kvm_lapic_state (in) | |
1452 | Returns: 0 on success, -1 on error | |
1453 | ||
1454 | #define KVM_APIC_REG_SIZE 0x400 | |
1455 | struct kvm_lapic_state { | |
1456 | char regs[KVM_APIC_REG_SIZE]; | |
1457 | }; | |
1458 | ||
1459 | Copies the input argument into the the Local APIC registers. The data format | |
1460 | and layout are the same as documented in the architecture manual. | |
1461 | ||
414fa985 JK |
1462 | |
1463 | 4.59 KVM_IOEVENTFD | |
55399a02 SL |
1464 | |
1465 | Capability: KVM_CAP_IOEVENTFD | |
1466 | Architectures: all | |
1467 | Type: vm ioctl | |
1468 | Parameters: struct kvm_ioeventfd (in) | |
1469 | Returns: 0 on success, !0 on error | |
1470 | ||
1471 | This ioctl attaches or detaches an ioeventfd to a legal pio/mmio address | |
1472 | within the guest. A guest write in the registered address will signal the | |
1473 | provided event instead of triggering an exit. | |
1474 | ||
1475 | struct kvm_ioeventfd { | |
1476 | __u64 datamatch; | |
1477 | __u64 addr; /* legal pio/mmio address */ | |
1478 | __u32 len; /* 1, 2, 4, or 8 bytes */ | |
1479 | __s32 fd; | |
1480 | __u32 flags; | |
1481 | __u8 pad[36]; | |
1482 | }; | |
1483 | ||
1484 | The following flags are defined: | |
1485 | ||
1486 | #define KVM_IOEVENTFD_FLAG_DATAMATCH (1 << kvm_ioeventfd_flag_nr_datamatch) | |
1487 | #define KVM_IOEVENTFD_FLAG_PIO (1 << kvm_ioeventfd_flag_nr_pio) | |
1488 | #define KVM_IOEVENTFD_FLAG_DEASSIGN (1 << kvm_ioeventfd_flag_nr_deassign) | |
1489 | ||
1490 | If datamatch flag is set, the event will be signaled only if the written value | |
1491 | to the registered address is equal to datamatch in struct kvm_ioeventfd. | |
1492 | ||
414fa985 JK |
1493 | |
1494 | 4.60 KVM_DIRTY_TLB | |
dc83b8bc SW |
1495 | |
1496 | Capability: KVM_CAP_SW_TLB | |
1497 | Architectures: ppc | |
1498 | Type: vcpu ioctl | |
1499 | Parameters: struct kvm_dirty_tlb (in) | |
1500 | Returns: 0 on success, -1 on error | |
1501 | ||
1502 | struct kvm_dirty_tlb { | |
1503 | __u64 bitmap; | |
1504 | __u32 num_dirty; | |
1505 | }; | |
1506 | ||
1507 | This must be called whenever userspace has changed an entry in the shared | |
1508 | TLB, prior to calling KVM_RUN on the associated vcpu. | |
1509 | ||
1510 | The "bitmap" field is the userspace address of an array. This array | |
1511 | consists of a number of bits, equal to the total number of TLB entries as | |
1512 | determined by the last successful call to KVM_CONFIG_TLB, rounded up to the | |
1513 | nearest multiple of 64. | |
1514 | ||
1515 | Each bit corresponds to one TLB entry, ordered the same as in the shared TLB | |
1516 | array. | |
1517 | ||
1518 | The array is little-endian: the bit 0 is the least significant bit of the | |
1519 | first byte, bit 8 is the least significant bit of the second byte, etc. | |
1520 | This avoids any complications with differing word sizes. | |
1521 | ||
1522 | The "num_dirty" field is a performance hint for KVM to determine whether it | |
1523 | should skip processing the bitmap and just invalidate everything. It must | |
1524 | be set to the number of set bits in the bitmap. | |
1525 | ||
414fa985 JK |
1526 | |
1527 | 4.61 KVM_ASSIGN_SET_INTX_MASK | |
07700a94 JK |
1528 | |
1529 | Capability: KVM_CAP_PCI_2_3 | |
1530 | Architectures: x86 | |
1531 | Type: vm ioctl | |
1532 | Parameters: struct kvm_assigned_pci_dev (in) | |
1533 | Returns: 0 on success, -1 on error | |
1534 | ||
1535 | Allows userspace to mask PCI INTx interrupts from the assigned device. The | |
1536 | kernel will not deliver INTx interrupts to the guest between setting and | |
1537 | clearing of KVM_ASSIGN_SET_INTX_MASK via this interface. This enables use of | |
1538 | and emulation of PCI 2.3 INTx disable command register behavior. | |
1539 | ||
1540 | This may be used for both PCI 2.3 devices supporting INTx disable natively and | |
1541 | older devices lacking this support. Userspace is responsible for emulating the | |
1542 | read value of the INTx disable bit in the guest visible PCI command register. | |
1543 | When modifying the INTx disable state, userspace should precede updating the | |
1544 | physical device command register by calling this ioctl to inform the kernel of | |
1545 | the new intended INTx mask state. | |
1546 | ||
1547 | Note that the kernel uses the device INTx disable bit to internally manage the | |
1548 | device interrupt state for PCI 2.3 devices. Reads of this register may | |
1549 | therefore not match the expected value. Writes should always use the guest | |
1550 | intended INTx disable value rather than attempting to read-copy-update the | |
1551 | current physical device state. Races between user and kernel updates to the | |
1552 | INTx disable bit are handled lazily in the kernel. It's possible the device | |
1553 | may generate unintended interrupts, but they will not be injected into the | |
1554 | guest. | |
1555 | ||
1556 | See KVM_ASSIGN_DEV_IRQ for the data structure. The target device is specified | |
1557 | by assigned_dev_id. In the flags field, only KVM_DEV_ASSIGN_MASK_INTX is | |
1558 | evaluated. | |
1559 | ||
414fa985 | 1560 | |
54738c09 DG |
1561 | 4.62 KVM_CREATE_SPAPR_TCE |
1562 | ||
1563 | Capability: KVM_CAP_SPAPR_TCE | |
1564 | Architectures: powerpc | |
1565 | Type: vm ioctl | |
1566 | Parameters: struct kvm_create_spapr_tce (in) | |
1567 | Returns: file descriptor for manipulating the created TCE table | |
1568 | ||
1569 | This creates a virtual TCE (translation control entry) table, which | |
1570 | is an IOMMU for PAPR-style virtual I/O. It is used to translate | |
1571 | logical addresses used in virtual I/O into guest physical addresses, | |
1572 | and provides a scatter/gather capability for PAPR virtual I/O. | |
1573 | ||
1574 | /* for KVM_CAP_SPAPR_TCE */ | |
1575 | struct kvm_create_spapr_tce { | |
1576 | __u64 liobn; | |
1577 | __u32 window_size; | |
1578 | }; | |
1579 | ||
1580 | The liobn field gives the logical IO bus number for which to create a | |
1581 | TCE table. The window_size field specifies the size of the DMA window | |
1582 | which this TCE table will translate - the table will contain one 64 | |
1583 | bit TCE entry for every 4kiB of the DMA window. | |
1584 | ||
1585 | When the guest issues an H_PUT_TCE hcall on a liobn for which a TCE | |
1586 | table has been created using this ioctl(), the kernel will handle it | |
1587 | in real mode, updating the TCE table. H_PUT_TCE calls for other | |
1588 | liobns will cause a vm exit and must be handled by userspace. | |
1589 | ||
1590 | The return value is a file descriptor which can be passed to mmap(2) | |
1591 | to map the created TCE table into userspace. This lets userspace read | |
1592 | the entries written by kernel-handled H_PUT_TCE calls, and also lets | |
1593 | userspace update the TCE table directly which is useful in some | |
1594 | circumstances. | |
1595 | ||
414fa985 | 1596 | |
aa04b4cc PM |
1597 | 4.63 KVM_ALLOCATE_RMA |
1598 | ||
1599 | Capability: KVM_CAP_PPC_RMA | |
1600 | Architectures: powerpc | |
1601 | Type: vm ioctl | |
1602 | Parameters: struct kvm_allocate_rma (out) | |
1603 | Returns: file descriptor for mapping the allocated RMA | |
1604 | ||
1605 | This allocates a Real Mode Area (RMA) from the pool allocated at boot | |
1606 | time by the kernel. An RMA is a physically-contiguous, aligned region | |
1607 | of memory used on older POWER processors to provide the memory which | |
1608 | will be accessed by real-mode (MMU off) accesses in a KVM guest. | |
1609 | POWER processors support a set of sizes for the RMA that usually | |
1610 | includes 64MB, 128MB, 256MB and some larger powers of two. | |
1611 | ||
1612 | /* for KVM_ALLOCATE_RMA */ | |
1613 | struct kvm_allocate_rma { | |
1614 | __u64 rma_size; | |
1615 | }; | |
1616 | ||
1617 | The return value is a file descriptor which can be passed to mmap(2) | |
1618 | to map the allocated RMA into userspace. The mapped area can then be | |
1619 | passed to the KVM_SET_USER_MEMORY_REGION ioctl to establish it as the | |
1620 | RMA for a virtual machine. The size of the RMA in bytes (which is | |
1621 | fixed at host kernel boot time) is returned in the rma_size field of | |
1622 | the argument structure. | |
1623 | ||
1624 | The KVM_CAP_PPC_RMA capability is 1 or 2 if the KVM_ALLOCATE_RMA ioctl | |
1625 | is supported; 2 if the processor requires all virtual machines to have | |
1626 | an RMA, or 1 if the processor can use an RMA but doesn't require it, | |
1627 | because it supports the Virtual RMA (VRMA) facility. | |
1628 | ||
414fa985 | 1629 | |
3f745f1e AK |
1630 | 4.64 KVM_NMI |
1631 | ||
1632 | Capability: KVM_CAP_USER_NMI | |
1633 | Architectures: x86 | |
1634 | Type: vcpu ioctl | |
1635 | Parameters: none | |
1636 | Returns: 0 on success, -1 on error | |
1637 | ||
1638 | Queues an NMI on the thread's vcpu. Note this is well defined only | |
1639 | when KVM_CREATE_IRQCHIP has not been called, since this is an interface | |
1640 | between the virtual cpu core and virtual local APIC. After KVM_CREATE_IRQCHIP | |
1641 | has been called, this interface is completely emulated within the kernel. | |
1642 | ||
1643 | To use this to emulate the LINT1 input with KVM_CREATE_IRQCHIP, use the | |
1644 | following algorithm: | |
1645 | ||
1646 | - pause the vpcu | |
1647 | - read the local APIC's state (KVM_GET_LAPIC) | |
1648 | - check whether changing LINT1 will queue an NMI (see the LVT entry for LINT1) | |
1649 | - if so, issue KVM_NMI | |
1650 | - resume the vcpu | |
1651 | ||
1652 | Some guests configure the LINT1 NMI input to cause a panic, aiding in | |
1653 | debugging. | |
1654 | ||
414fa985 | 1655 | |
e24ed81f | 1656 | 4.65 KVM_S390_UCAS_MAP |
27e0393f CO |
1657 | |
1658 | Capability: KVM_CAP_S390_UCONTROL | |
1659 | Architectures: s390 | |
1660 | Type: vcpu ioctl | |
1661 | Parameters: struct kvm_s390_ucas_mapping (in) | |
1662 | Returns: 0 in case of success | |
1663 | ||
1664 | The parameter is defined like this: | |
1665 | struct kvm_s390_ucas_mapping { | |
1666 | __u64 user_addr; | |
1667 | __u64 vcpu_addr; | |
1668 | __u64 length; | |
1669 | }; | |
1670 | ||
1671 | This ioctl maps the memory at "user_addr" with the length "length" to | |
1672 | the vcpu's address space starting at "vcpu_addr". All parameters need to | |
1673 | be alligned by 1 megabyte. | |
1674 | ||
414fa985 | 1675 | |
e24ed81f | 1676 | 4.66 KVM_S390_UCAS_UNMAP |
27e0393f CO |
1677 | |
1678 | Capability: KVM_CAP_S390_UCONTROL | |
1679 | Architectures: s390 | |
1680 | Type: vcpu ioctl | |
1681 | Parameters: struct kvm_s390_ucas_mapping (in) | |
1682 | Returns: 0 in case of success | |
1683 | ||
1684 | The parameter is defined like this: | |
1685 | struct kvm_s390_ucas_mapping { | |
1686 | __u64 user_addr; | |
1687 | __u64 vcpu_addr; | |
1688 | __u64 length; | |
1689 | }; | |
1690 | ||
1691 | This ioctl unmaps the memory in the vcpu's address space starting at | |
1692 | "vcpu_addr" with the length "length". The field "user_addr" is ignored. | |
1693 | All parameters need to be alligned by 1 megabyte. | |
1694 | ||
414fa985 | 1695 | |
e24ed81f | 1696 | 4.67 KVM_S390_VCPU_FAULT |
ccc7910f CO |
1697 | |
1698 | Capability: KVM_CAP_S390_UCONTROL | |
1699 | Architectures: s390 | |
1700 | Type: vcpu ioctl | |
1701 | Parameters: vcpu absolute address (in) | |
1702 | Returns: 0 in case of success | |
1703 | ||
1704 | This call creates a page table entry on the virtual cpu's address space | |
1705 | (for user controlled virtual machines) or the virtual machine's address | |
1706 | space (for regular virtual machines). This only works for minor faults, | |
1707 | thus it's recommended to access subject memory page via the user page | |
1708 | table upfront. This is useful to handle validity intercepts for user | |
1709 | controlled virtual machines to fault in the virtual cpu's lowcore pages | |
1710 | prior to calling the KVM_RUN ioctl. | |
1711 | ||
414fa985 | 1712 | |
e24ed81f AG |
1713 | 4.68 KVM_SET_ONE_REG |
1714 | ||
1715 | Capability: KVM_CAP_ONE_REG | |
1716 | Architectures: all | |
1717 | Type: vcpu ioctl | |
1718 | Parameters: struct kvm_one_reg (in) | |
1719 | Returns: 0 on success, negative value on failure | |
1720 | ||
1721 | struct kvm_one_reg { | |
1722 | __u64 id; | |
1723 | __u64 addr; | |
1724 | }; | |
1725 | ||
1726 | Using this ioctl, a single vcpu register can be set to a specific value | |
1727 | defined by user space with the passed in struct kvm_one_reg, where id | |
1728 | refers to the register identifier as described below and addr is a pointer | |
1729 | to a variable with the respective size. There can be architecture agnostic | |
1730 | and architecture specific registers. Each have their own range of operation | |
1731 | and their own constants and width. To keep track of the implemented | |
1732 | registers, find a list below: | |
1733 | ||
1734 | Arch | Register | Width (bits) | |
1735 | | | | |
1022fc3d | 1736 | PPC | KVM_REG_PPC_HIOR | 64 |
2e232702 BB |
1737 | PPC | KVM_REG_PPC_IAC1 | 64 |
1738 | PPC | KVM_REG_PPC_IAC2 | 64 | |
1739 | PPC | KVM_REG_PPC_IAC3 | 64 | |
1740 | PPC | KVM_REG_PPC_IAC4 | 64 | |
1741 | PPC | KVM_REG_PPC_DAC1 | 64 | |
1742 | PPC | KVM_REG_PPC_DAC2 | 64 | |
a136a8bd PM |
1743 | PPC | KVM_REG_PPC_DABR | 64 |
1744 | PPC | KVM_REG_PPC_DSCR | 64 | |
1745 | PPC | KVM_REG_PPC_PURR | 64 | |
1746 | PPC | KVM_REG_PPC_SPURR | 64 | |
1747 | PPC | KVM_REG_PPC_DAR | 64 | |
1748 | PPC | KVM_REG_PPC_DSISR | 32 | |
1749 | PPC | KVM_REG_PPC_AMR | 64 | |
1750 | PPC | KVM_REG_PPC_UAMOR | 64 | |
1751 | PPC | KVM_REG_PPC_MMCR0 | 64 | |
1752 | PPC | KVM_REG_PPC_MMCR1 | 64 | |
1753 | PPC | KVM_REG_PPC_MMCRA | 64 | |
1754 | PPC | KVM_REG_PPC_PMC1 | 32 | |
1755 | PPC | KVM_REG_PPC_PMC2 | 32 | |
1756 | PPC | KVM_REG_PPC_PMC3 | 32 | |
1757 | PPC | KVM_REG_PPC_PMC4 | 32 | |
1758 | PPC | KVM_REG_PPC_PMC5 | 32 | |
1759 | PPC | KVM_REG_PPC_PMC6 | 32 | |
1760 | PPC | KVM_REG_PPC_PMC7 | 32 | |
1761 | PPC | KVM_REG_PPC_PMC8 | 32 | |
a8bd19ef PM |
1762 | PPC | KVM_REG_PPC_FPR0 | 64 |
1763 | ... | |
1764 | PPC | KVM_REG_PPC_FPR31 | 64 | |
1765 | PPC | KVM_REG_PPC_VR0 | 128 | |
1766 | ... | |
1767 | PPC | KVM_REG_PPC_VR31 | 128 | |
1768 | PPC | KVM_REG_PPC_VSR0 | 128 | |
1769 | ... | |
1770 | PPC | KVM_REG_PPC_VSR31 | 128 | |
1771 | PPC | KVM_REG_PPC_FPSCR | 64 | |
1772 | PPC | KVM_REG_PPC_VSCR | 32 | |
55b665b0 PM |
1773 | PPC | KVM_REG_PPC_VPA_ADDR | 64 |
1774 | PPC | KVM_REG_PPC_VPA_SLB | 128 | |
1775 | PPC | KVM_REG_PPC_VPA_DTL | 128 | |
414fa985 | 1776 | |
e24ed81f AG |
1777 | 4.69 KVM_GET_ONE_REG |
1778 | ||
1779 | Capability: KVM_CAP_ONE_REG | |
1780 | Architectures: all | |
1781 | Type: vcpu ioctl | |
1782 | Parameters: struct kvm_one_reg (in and out) | |
1783 | Returns: 0 on success, negative value on failure | |
1784 | ||
1785 | This ioctl allows to receive the value of a single register implemented | |
1786 | in a vcpu. The register to read is indicated by the "id" field of the | |
1787 | kvm_one_reg struct passed in. On success, the register value can be found | |
1788 | at the memory location pointed to by "addr". | |
1789 | ||
1790 | The list of registers accessible using this interface is identical to the | |
2e232702 | 1791 | list in 4.68. |
e24ed81f | 1792 | |
414fa985 | 1793 | |
1c0b28c2 EM |
1794 | 4.70 KVM_KVMCLOCK_CTRL |
1795 | ||
1796 | Capability: KVM_CAP_KVMCLOCK_CTRL | |
1797 | Architectures: Any that implement pvclocks (currently x86 only) | |
1798 | Type: vcpu ioctl | |
1799 | Parameters: None | |
1800 | Returns: 0 on success, -1 on error | |
1801 | ||
1802 | This signals to the host kernel that the specified guest is being paused by | |
1803 | userspace. The host will set a flag in the pvclock structure that is checked | |
1804 | from the soft lockup watchdog. The flag is part of the pvclock structure that | |
1805 | is shared between guest and host, specifically the second bit of the flags | |
1806 | field of the pvclock_vcpu_time_info structure. It will be set exclusively by | |
1807 | the host and read/cleared exclusively by the guest. The guest operation of | |
1808 | checking and clearing the flag must an atomic operation so | |
1809 | load-link/store-conditional, or equivalent must be used. There are two cases | |
1810 | where the guest will clear the flag: when the soft lockup watchdog timer resets | |
1811 | itself or when a soft lockup is detected. This ioctl can be called any time | |
1812 | after pausing the vcpu, but before it is resumed. | |
1813 | ||
414fa985 | 1814 | |
07975ad3 JK |
1815 | 4.71 KVM_SIGNAL_MSI |
1816 | ||
1817 | Capability: KVM_CAP_SIGNAL_MSI | |
1818 | Architectures: x86 | |
1819 | Type: vm ioctl | |
1820 | Parameters: struct kvm_msi (in) | |
1821 | Returns: >0 on delivery, 0 if guest blocked the MSI, and -1 on error | |
1822 | ||
1823 | Directly inject a MSI message. Only valid with in-kernel irqchip that handles | |
1824 | MSI messages. | |
1825 | ||
1826 | struct kvm_msi { | |
1827 | __u32 address_lo; | |
1828 | __u32 address_hi; | |
1829 | __u32 data; | |
1830 | __u32 flags; | |
1831 | __u8 pad[16]; | |
1832 | }; | |
1833 | ||
1834 | No flags are defined so far. The corresponding field must be 0. | |
1835 | ||
414fa985 | 1836 | |
0589ff6c JK |
1837 | 4.71 KVM_CREATE_PIT2 |
1838 | ||
1839 | Capability: KVM_CAP_PIT2 | |
1840 | Architectures: x86 | |
1841 | Type: vm ioctl | |
1842 | Parameters: struct kvm_pit_config (in) | |
1843 | Returns: 0 on success, -1 on error | |
1844 | ||
1845 | Creates an in-kernel device model for the i8254 PIT. This call is only valid | |
1846 | after enabling in-kernel irqchip support via KVM_CREATE_IRQCHIP. The following | |
1847 | parameters have to be passed: | |
1848 | ||
1849 | struct kvm_pit_config { | |
1850 | __u32 flags; | |
1851 | __u32 pad[15]; | |
1852 | }; | |
1853 | ||
1854 | Valid flags are: | |
1855 | ||
1856 | #define KVM_PIT_SPEAKER_DUMMY 1 /* emulate speaker port stub */ | |
1857 | ||
b6ddf05f JK |
1858 | PIT timer interrupts may use a per-VM kernel thread for injection. If it |
1859 | exists, this thread will have a name of the following pattern: | |
1860 | ||
1861 | kvm-pit/<owner-process-pid> | |
1862 | ||
1863 | When running a guest with elevated priorities, the scheduling parameters of | |
1864 | this thread may have to be adjusted accordingly. | |
1865 | ||
0589ff6c JK |
1866 | This IOCTL replaces the obsolete KVM_CREATE_PIT. |
1867 | ||
1868 | ||
1869 | 4.72 KVM_GET_PIT2 | |
1870 | ||
1871 | Capability: KVM_CAP_PIT_STATE2 | |
1872 | Architectures: x86 | |
1873 | Type: vm ioctl | |
1874 | Parameters: struct kvm_pit_state2 (out) | |
1875 | Returns: 0 on success, -1 on error | |
1876 | ||
1877 | Retrieves the state of the in-kernel PIT model. Only valid after | |
1878 | KVM_CREATE_PIT2. The state is returned in the following structure: | |
1879 | ||
1880 | struct kvm_pit_state2 { | |
1881 | struct kvm_pit_channel_state channels[3]; | |
1882 | __u32 flags; | |
1883 | __u32 reserved[9]; | |
1884 | }; | |
1885 | ||
1886 | Valid flags are: | |
1887 | ||
1888 | /* disable PIT in HPET legacy mode */ | |
1889 | #define KVM_PIT_FLAGS_HPET_LEGACY 0x00000001 | |
1890 | ||
1891 | This IOCTL replaces the obsolete KVM_GET_PIT. | |
1892 | ||
1893 | ||
1894 | 4.73 KVM_SET_PIT2 | |
1895 | ||
1896 | Capability: KVM_CAP_PIT_STATE2 | |
1897 | Architectures: x86 | |
1898 | Type: vm ioctl | |
1899 | Parameters: struct kvm_pit_state2 (in) | |
1900 | Returns: 0 on success, -1 on error | |
1901 | ||
1902 | Sets the state of the in-kernel PIT model. Only valid after KVM_CREATE_PIT2. | |
1903 | See KVM_GET_PIT2 for details on struct kvm_pit_state2. | |
1904 | ||
1905 | This IOCTL replaces the obsolete KVM_SET_PIT. | |
1906 | ||
1907 | ||
5b74716e BH |
1908 | 4.74 KVM_PPC_GET_SMMU_INFO |
1909 | ||
1910 | Capability: KVM_CAP_PPC_GET_SMMU_INFO | |
1911 | Architectures: powerpc | |
1912 | Type: vm ioctl | |
1913 | Parameters: None | |
1914 | Returns: 0 on success, -1 on error | |
1915 | ||
1916 | This populates and returns a structure describing the features of | |
1917 | the "Server" class MMU emulation supported by KVM. | |
1918 | This can in turn be used by userspace to generate the appropariate | |
1919 | device-tree properties for the guest operating system. | |
1920 | ||
1921 | The structure contains some global informations, followed by an | |
1922 | array of supported segment page sizes: | |
1923 | ||
1924 | struct kvm_ppc_smmu_info { | |
1925 | __u64 flags; | |
1926 | __u32 slb_size; | |
1927 | __u32 pad; | |
1928 | struct kvm_ppc_one_seg_page_size sps[KVM_PPC_PAGE_SIZES_MAX_SZ]; | |
1929 | }; | |
1930 | ||
1931 | The supported flags are: | |
1932 | ||
1933 | - KVM_PPC_PAGE_SIZES_REAL: | |
1934 | When that flag is set, guest page sizes must "fit" the backing | |
1935 | store page sizes. When not set, any page size in the list can | |
1936 | be used regardless of how they are backed by userspace. | |
1937 | ||
1938 | - KVM_PPC_1T_SEGMENTS | |
1939 | The emulated MMU supports 1T segments in addition to the | |
1940 | standard 256M ones. | |
1941 | ||
1942 | The "slb_size" field indicates how many SLB entries are supported | |
1943 | ||
1944 | The "sps" array contains 8 entries indicating the supported base | |
1945 | page sizes for a segment in increasing order. Each entry is defined | |
1946 | as follow: | |
1947 | ||
1948 | struct kvm_ppc_one_seg_page_size { | |
1949 | __u32 page_shift; /* Base page shift of segment (or 0) */ | |
1950 | __u32 slb_enc; /* SLB encoding for BookS */ | |
1951 | struct kvm_ppc_one_page_size enc[KVM_PPC_PAGE_SIZES_MAX_SZ]; | |
1952 | }; | |
1953 | ||
1954 | An entry with a "page_shift" of 0 is unused. Because the array is | |
1955 | organized in increasing order, a lookup can stop when encoutering | |
1956 | such an entry. | |
1957 | ||
1958 | The "slb_enc" field provides the encoding to use in the SLB for the | |
1959 | page size. The bits are in positions such as the value can directly | |
1960 | be OR'ed into the "vsid" argument of the slbmte instruction. | |
1961 | ||
1962 | The "enc" array is a list which for each of those segment base page | |
1963 | size provides the list of supported actual page sizes (which can be | |
1964 | only larger or equal to the base page size), along with the | |
1965 | corresponding encoding in the hash PTE. Similarily, the array is | |
1966 | 8 entries sorted by increasing sizes and an entry with a "0" shift | |
1967 | is an empty entry and a terminator: | |
1968 | ||
1969 | struct kvm_ppc_one_page_size { | |
1970 | __u32 page_shift; /* Page shift (or 0) */ | |
1971 | __u32 pte_enc; /* Encoding in the HPTE (>>12) */ | |
1972 | }; | |
1973 | ||
1974 | The "pte_enc" field provides a value that can OR'ed into the hash | |
1975 | PTE's RPN field (ie, it needs to be shifted left by 12 to OR it | |
1976 | into the hash PTE second double word). | |
1977 | ||
f36992e3 AW |
1978 | 4.75 KVM_IRQFD |
1979 | ||
1980 | Capability: KVM_CAP_IRQFD | |
1981 | Architectures: x86 | |
1982 | Type: vm ioctl | |
1983 | Parameters: struct kvm_irqfd (in) | |
1984 | Returns: 0 on success, -1 on error | |
1985 | ||
1986 | Allows setting an eventfd to directly trigger a guest interrupt. | |
1987 | kvm_irqfd.fd specifies the file descriptor to use as the eventfd and | |
1988 | kvm_irqfd.gsi specifies the irqchip pin toggled by this event. When | |
1989 | an event is tiggered on the eventfd, an interrupt is injected into | |
1990 | the guest using the specified gsi pin. The irqfd is removed using | |
1991 | the KVM_IRQFD_FLAG_DEASSIGN flag, specifying both kvm_irqfd.fd | |
1992 | and kvm_irqfd.gsi. | |
1993 | ||
7a84428a AW |
1994 | With KVM_CAP_IRQFD_RESAMPLE, KVM_IRQFD supports a de-assert and notify |
1995 | mechanism allowing emulation of level-triggered, irqfd-based | |
1996 | interrupts. When KVM_IRQFD_FLAG_RESAMPLE is set the user must pass an | |
1997 | additional eventfd in the kvm_irqfd.resamplefd field. When operating | |
1998 | in resample mode, posting of an interrupt through kvm_irq.fd asserts | |
1999 | the specified gsi in the irqchip. When the irqchip is resampled, such | |
2000 | as from an EOI, the gsi is de-asserted and the user is notifed via | |
2001 | kvm_irqfd.resamplefd. It is the user's responsibility to re-queue | |
2002 | the interrupt if the device making use of it still requires service. | |
2003 | Note that closing the resamplefd is not sufficient to disable the | |
2004 | irqfd. The KVM_IRQFD_FLAG_RESAMPLE is only necessary on assignment | |
2005 | and need not be specified with KVM_IRQFD_FLAG_DEASSIGN. | |
2006 | ||
5fecc9d8 | 2007 | 4.76 KVM_PPC_ALLOCATE_HTAB |
32fad281 PM |
2008 | |
2009 | Capability: KVM_CAP_PPC_ALLOC_HTAB | |
2010 | Architectures: powerpc | |
2011 | Type: vm ioctl | |
2012 | Parameters: Pointer to u32 containing hash table order (in/out) | |
2013 | Returns: 0 on success, -1 on error | |
2014 | ||
2015 | This requests the host kernel to allocate an MMU hash table for a | |
2016 | guest using the PAPR paravirtualization interface. This only does | |
2017 | anything if the kernel is configured to use the Book 3S HV style of | |
2018 | virtualization. Otherwise the capability doesn't exist and the ioctl | |
2019 | returns an ENOTTY error. The rest of this description assumes Book 3S | |
2020 | HV. | |
2021 | ||
2022 | There must be no vcpus running when this ioctl is called; if there | |
2023 | are, it will do nothing and return an EBUSY error. | |
2024 | ||
2025 | The parameter is a pointer to a 32-bit unsigned integer variable | |
2026 | containing the order (log base 2) of the desired size of the hash | |
2027 | table, which must be between 18 and 46. On successful return from the | |
2028 | ioctl, it will have been updated with the order of the hash table that | |
2029 | was allocated. | |
2030 | ||
2031 | If no hash table has been allocated when any vcpu is asked to run | |
2032 | (with the KVM_RUN ioctl), the host kernel will allocate a | |
2033 | default-sized hash table (16 MB). | |
2034 | ||
2035 | If this ioctl is called when a hash table has already been allocated, | |
2036 | the kernel will clear out the existing hash table (zero all HPTEs) and | |
2037 | return the hash table order in the parameter. (If the guest is using | |
2038 | the virtualized real-mode area (VRMA) facility, the kernel will | |
2039 | re-create the VMRA HPTEs on the next KVM_RUN of any vcpu.) | |
2040 | ||
416ad65f CH |
2041 | 4.77 KVM_S390_INTERRUPT |
2042 | ||
2043 | Capability: basic | |
2044 | Architectures: s390 | |
2045 | Type: vm ioctl, vcpu ioctl | |
2046 | Parameters: struct kvm_s390_interrupt (in) | |
2047 | Returns: 0 on success, -1 on error | |
2048 | ||
2049 | Allows to inject an interrupt to the guest. Interrupts can be floating | |
2050 | (vm ioctl) or per cpu (vcpu ioctl), depending on the interrupt type. | |
2051 | ||
2052 | Interrupt parameters are passed via kvm_s390_interrupt: | |
2053 | ||
2054 | struct kvm_s390_interrupt { | |
2055 | __u32 type; | |
2056 | __u32 parm; | |
2057 | __u64 parm64; | |
2058 | }; | |
2059 | ||
2060 | type can be one of the following: | |
2061 | ||
2062 | KVM_S390_SIGP_STOP (vcpu) - sigp restart | |
2063 | KVM_S390_PROGRAM_INT (vcpu) - program check; code in parm | |
2064 | KVM_S390_SIGP_SET_PREFIX (vcpu) - sigp set prefix; prefix address in parm | |
2065 | KVM_S390_RESTART (vcpu) - restart | |
2066 | KVM_S390_INT_VIRTIO (vm) - virtio external interrupt; external interrupt | |
2067 | parameters in parm and parm64 | |
2068 | KVM_S390_INT_SERVICE (vm) - sclp external interrupt; sclp parameter in parm | |
2069 | KVM_S390_INT_EMERGENCY (vcpu) - sigp emergency; source cpu in parm | |
2070 | KVM_S390_INT_EXTERNAL_CALL (vcpu) - sigp external call; source cpu in parm | |
2071 | ||
2072 | Note that the vcpu ioctl is asynchronous to vcpu execution. | |
2073 | ||
a2932923 PM |
2074 | 4.78 KVM_PPC_GET_HTAB_FD |
2075 | ||
2076 | Capability: KVM_CAP_PPC_HTAB_FD | |
2077 | Architectures: powerpc | |
2078 | Type: vm ioctl | |
2079 | Parameters: Pointer to struct kvm_get_htab_fd (in) | |
2080 | Returns: file descriptor number (>= 0) on success, -1 on error | |
2081 | ||
2082 | This returns a file descriptor that can be used either to read out the | |
2083 | entries in the guest's hashed page table (HPT), or to write entries to | |
2084 | initialize the HPT. The returned fd can only be written to if the | |
2085 | KVM_GET_HTAB_WRITE bit is set in the flags field of the argument, and | |
2086 | can only be read if that bit is clear. The argument struct looks like | |
2087 | this: | |
2088 | ||
2089 | /* For KVM_PPC_GET_HTAB_FD */ | |
2090 | struct kvm_get_htab_fd { | |
2091 | __u64 flags; | |
2092 | __u64 start_index; | |
2093 | __u64 reserved[2]; | |
2094 | }; | |
2095 | ||
2096 | /* Values for kvm_get_htab_fd.flags */ | |
2097 | #define KVM_GET_HTAB_BOLTED_ONLY ((__u64)0x1) | |
2098 | #define KVM_GET_HTAB_WRITE ((__u64)0x2) | |
2099 | ||
2100 | The `start_index' field gives the index in the HPT of the entry at | |
2101 | which to start reading. It is ignored when writing. | |
2102 | ||
2103 | Reads on the fd will initially supply information about all | |
2104 | "interesting" HPT entries. Interesting entries are those with the | |
2105 | bolted bit set, if the KVM_GET_HTAB_BOLTED_ONLY bit is set, otherwise | |
2106 | all entries. When the end of the HPT is reached, the read() will | |
2107 | return. If read() is called again on the fd, it will start again from | |
2108 | the beginning of the HPT, but will only return HPT entries that have | |
2109 | changed since they were last read. | |
2110 | ||
2111 | Data read or written is structured as a header (8 bytes) followed by a | |
2112 | series of valid HPT entries (16 bytes) each. The header indicates how | |
2113 | many valid HPT entries there are and how many invalid entries follow | |
2114 | the valid entries. The invalid entries are not represented explicitly | |
2115 | in the stream. The header format is: | |
2116 | ||
2117 | struct kvm_get_htab_header { | |
2118 | __u32 index; | |
2119 | __u16 n_valid; | |
2120 | __u16 n_invalid; | |
2121 | }; | |
2122 | ||
2123 | Writes to the fd create HPT entries starting at the index given in the | |
2124 | header; first `n_valid' valid entries with contents from the data | |
2125 | written, then `n_invalid' invalid entries, invalidating any previously | |
2126 | valid entries found. | |
2127 | ||
f36992e3 | 2128 | |
9c1b96e3 | 2129 | 5. The kvm_run structure |
414fa985 | 2130 | ------------------------ |
9c1b96e3 AK |
2131 | |
2132 | Application code obtains a pointer to the kvm_run structure by | |
2133 | mmap()ing a vcpu fd. From that point, application code can control | |
2134 | execution by changing fields in kvm_run prior to calling the KVM_RUN | |
2135 | ioctl, and obtain information about the reason KVM_RUN returned by | |
2136 | looking up structure members. | |
2137 | ||
2138 | struct kvm_run { | |
2139 | /* in */ | |
2140 | __u8 request_interrupt_window; | |
2141 | ||
2142 | Request that KVM_RUN return when it becomes possible to inject external | |
2143 | interrupts into the guest. Useful in conjunction with KVM_INTERRUPT. | |
2144 | ||
2145 | __u8 padding1[7]; | |
2146 | ||
2147 | /* out */ | |
2148 | __u32 exit_reason; | |
2149 | ||
2150 | When KVM_RUN has returned successfully (return value 0), this informs | |
2151 | application code why KVM_RUN has returned. Allowable values for this | |
2152 | field are detailed below. | |
2153 | ||
2154 | __u8 ready_for_interrupt_injection; | |
2155 | ||
2156 | If request_interrupt_window has been specified, this field indicates | |
2157 | an interrupt can be injected now with KVM_INTERRUPT. | |
2158 | ||
2159 | __u8 if_flag; | |
2160 | ||
2161 | The value of the current interrupt flag. Only valid if in-kernel | |
2162 | local APIC is not used. | |
2163 | ||
2164 | __u8 padding2[2]; | |
2165 | ||
2166 | /* in (pre_kvm_run), out (post_kvm_run) */ | |
2167 | __u64 cr8; | |
2168 | ||
2169 | The value of the cr8 register. Only valid if in-kernel local APIC is | |
2170 | not used. Both input and output. | |
2171 | ||
2172 | __u64 apic_base; | |
2173 | ||
2174 | The value of the APIC BASE msr. Only valid if in-kernel local | |
2175 | APIC is not used. Both input and output. | |
2176 | ||
2177 | union { | |
2178 | /* KVM_EXIT_UNKNOWN */ | |
2179 | struct { | |
2180 | __u64 hardware_exit_reason; | |
2181 | } hw; | |
2182 | ||
2183 | If exit_reason is KVM_EXIT_UNKNOWN, the vcpu has exited due to unknown | |
2184 | reasons. Further architecture-specific information is available in | |
2185 | hardware_exit_reason. | |
2186 | ||
2187 | /* KVM_EXIT_FAIL_ENTRY */ | |
2188 | struct { | |
2189 | __u64 hardware_entry_failure_reason; | |
2190 | } fail_entry; | |
2191 | ||
2192 | If exit_reason is KVM_EXIT_FAIL_ENTRY, the vcpu could not be run due | |
2193 | to unknown reasons. Further architecture-specific information is | |
2194 | available in hardware_entry_failure_reason. | |
2195 | ||
2196 | /* KVM_EXIT_EXCEPTION */ | |
2197 | struct { | |
2198 | __u32 exception; | |
2199 | __u32 error_code; | |
2200 | } ex; | |
2201 | ||
2202 | Unused. | |
2203 | ||
2204 | /* KVM_EXIT_IO */ | |
2205 | struct { | |
2206 | #define KVM_EXIT_IO_IN 0 | |
2207 | #define KVM_EXIT_IO_OUT 1 | |
2208 | __u8 direction; | |
2209 | __u8 size; /* bytes */ | |
2210 | __u16 port; | |
2211 | __u32 count; | |
2212 | __u64 data_offset; /* relative to kvm_run start */ | |
2213 | } io; | |
2214 | ||
2044892d | 2215 | If exit_reason is KVM_EXIT_IO, then the vcpu has |
9c1b96e3 AK |
2216 | executed a port I/O instruction which could not be satisfied by kvm. |
2217 | data_offset describes where the data is located (KVM_EXIT_IO_OUT) or | |
2218 | where kvm expects application code to place the data for the next | |
2044892d | 2219 | KVM_RUN invocation (KVM_EXIT_IO_IN). Data format is a packed array. |
9c1b96e3 AK |
2220 | |
2221 | struct { | |
2222 | struct kvm_debug_exit_arch arch; | |
2223 | } debug; | |
2224 | ||
2225 | Unused. | |
2226 | ||
2227 | /* KVM_EXIT_MMIO */ | |
2228 | struct { | |
2229 | __u64 phys_addr; | |
2230 | __u8 data[8]; | |
2231 | __u32 len; | |
2232 | __u8 is_write; | |
2233 | } mmio; | |
2234 | ||
2044892d | 2235 | If exit_reason is KVM_EXIT_MMIO, then the vcpu has |
9c1b96e3 AK |
2236 | executed a memory-mapped I/O instruction which could not be satisfied |
2237 | by kvm. The 'data' member contains the written data if 'is_write' is | |
2238 | true, and should be filled by application code otherwise. | |
2239 | ||
686de182 AG |
2240 | NOTE: For KVM_EXIT_IO, KVM_EXIT_MMIO, KVM_EXIT_OSI, KVM_EXIT_DCR |
2241 | and KVM_EXIT_PAPR the corresponding | |
ad0a048b AG |
2242 | operations are complete (and guest state is consistent) only after userspace |
2243 | has re-entered the kernel with KVM_RUN. The kernel side will first finish | |
67961344 MT |
2244 | incomplete operations and then check for pending signals. Userspace |
2245 | can re-enter the guest with an unmasked signal pending to complete | |
2246 | pending operations. | |
2247 | ||
9c1b96e3 AK |
2248 | /* KVM_EXIT_HYPERCALL */ |
2249 | struct { | |
2250 | __u64 nr; | |
2251 | __u64 args[6]; | |
2252 | __u64 ret; | |
2253 | __u32 longmode; | |
2254 | __u32 pad; | |
2255 | } hypercall; | |
2256 | ||
647dc49e AK |
2257 | Unused. This was once used for 'hypercall to userspace'. To implement |
2258 | such functionality, use KVM_EXIT_IO (x86) or KVM_EXIT_MMIO (all except s390). | |
2259 | Note KVM_EXIT_IO is significantly faster than KVM_EXIT_MMIO. | |
9c1b96e3 AK |
2260 | |
2261 | /* KVM_EXIT_TPR_ACCESS */ | |
2262 | struct { | |
2263 | __u64 rip; | |
2264 | __u32 is_write; | |
2265 | __u32 pad; | |
2266 | } tpr_access; | |
2267 | ||
2268 | To be documented (KVM_TPR_ACCESS_REPORTING). | |
2269 | ||
2270 | /* KVM_EXIT_S390_SIEIC */ | |
2271 | struct { | |
2272 | __u8 icptcode; | |
2273 | __u64 mask; /* psw upper half */ | |
2274 | __u64 addr; /* psw lower half */ | |
2275 | __u16 ipa; | |
2276 | __u32 ipb; | |
2277 | } s390_sieic; | |
2278 | ||
2279 | s390 specific. | |
2280 | ||
2281 | /* KVM_EXIT_S390_RESET */ | |
2282 | #define KVM_S390_RESET_POR 1 | |
2283 | #define KVM_S390_RESET_CLEAR 2 | |
2284 | #define KVM_S390_RESET_SUBSYSTEM 4 | |
2285 | #define KVM_S390_RESET_CPU_INIT 8 | |
2286 | #define KVM_S390_RESET_IPL 16 | |
2287 | __u64 s390_reset_flags; | |
2288 | ||
2289 | s390 specific. | |
2290 | ||
e168bf8d CO |
2291 | /* KVM_EXIT_S390_UCONTROL */ |
2292 | struct { | |
2293 | __u64 trans_exc_code; | |
2294 | __u32 pgm_code; | |
2295 | } s390_ucontrol; | |
2296 | ||
2297 | s390 specific. A page fault has occurred for a user controlled virtual | |
2298 | machine (KVM_VM_S390_UNCONTROL) on it's host page table that cannot be | |
2299 | resolved by the kernel. | |
2300 | The program code and the translation exception code that were placed | |
2301 | in the cpu's lowcore are presented here as defined by the z Architecture | |
2302 | Principles of Operation Book in the Chapter for Dynamic Address Translation | |
2303 | (DAT) | |
2304 | ||
9c1b96e3 AK |
2305 | /* KVM_EXIT_DCR */ |
2306 | struct { | |
2307 | __u32 dcrn; | |
2308 | __u32 data; | |
2309 | __u8 is_write; | |
2310 | } dcr; | |
2311 | ||
2312 | powerpc specific. | |
2313 | ||
ad0a048b AG |
2314 | /* KVM_EXIT_OSI */ |
2315 | struct { | |
2316 | __u64 gprs[32]; | |
2317 | } osi; | |
2318 | ||
2319 | MOL uses a special hypercall interface it calls 'OSI'. To enable it, we catch | |
2320 | hypercalls and exit with this exit struct that contains all the guest gprs. | |
2321 | ||
2322 | If exit_reason is KVM_EXIT_OSI, then the vcpu has triggered such a hypercall. | |
2323 | Userspace can now handle the hypercall and when it's done modify the gprs as | |
2324 | necessary. Upon guest entry all guest GPRs will then be replaced by the values | |
2325 | in this struct. | |
2326 | ||
de56a948 PM |
2327 | /* KVM_EXIT_PAPR_HCALL */ |
2328 | struct { | |
2329 | __u64 nr; | |
2330 | __u64 ret; | |
2331 | __u64 args[9]; | |
2332 | } papr_hcall; | |
2333 | ||
2334 | This is used on 64-bit PowerPC when emulating a pSeries partition, | |
2335 | e.g. with the 'pseries' machine type in qemu. It occurs when the | |
2336 | guest does a hypercall using the 'sc 1' instruction. The 'nr' field | |
2337 | contains the hypercall number (from the guest R3), and 'args' contains | |
2338 | the arguments (from the guest R4 - R12). Userspace should put the | |
2339 | return code in 'ret' and any extra returned values in args[]. | |
2340 | The possible hypercalls are defined in the Power Architecture Platform | |
2341 | Requirements (PAPR) document available from www.power.org (free | |
2342 | developer registration required to access it). | |
2343 | ||
9c1b96e3 AK |
2344 | /* Fix the size of the union. */ |
2345 | char padding[256]; | |
2346 | }; | |
b9e5dc8d CB |
2347 | |
2348 | /* | |
2349 | * shared registers between kvm and userspace. | |
2350 | * kvm_valid_regs specifies the register classes set by the host | |
2351 | * kvm_dirty_regs specified the register classes dirtied by userspace | |
2352 | * struct kvm_sync_regs is architecture specific, as well as the | |
2353 | * bits for kvm_valid_regs and kvm_dirty_regs | |
2354 | */ | |
2355 | __u64 kvm_valid_regs; | |
2356 | __u64 kvm_dirty_regs; | |
2357 | union { | |
2358 | struct kvm_sync_regs regs; | |
2359 | char padding[1024]; | |
2360 | } s; | |
2361 | ||
2362 | If KVM_CAP_SYNC_REGS is defined, these fields allow userspace to access | |
2363 | certain guest registers without having to call SET/GET_*REGS. Thus we can | |
2364 | avoid some system call overhead if userspace has to handle the exit. | |
2365 | Userspace can query the validity of the structure by checking | |
2366 | kvm_valid_regs for specific bits. These bits are architecture specific | |
2367 | and usually define the validity of a groups of registers. (e.g. one bit | |
2368 | for general purpose registers) | |
2369 | ||
9c1b96e3 | 2370 | }; |
821246a5 | 2371 | |
414fa985 | 2372 | |
821246a5 | 2373 | 6. Capabilities that can be enabled |
414fa985 | 2374 | ----------------------------------- |
821246a5 AG |
2375 | |
2376 | There are certain capabilities that change the behavior of the virtual CPU when | |
2377 | enabled. To enable them, please see section 4.37. Below you can find a list of | |
2378 | capabilities and what their effect on the vCPU is when enabling them. | |
2379 | ||
2380 | The following information is provided along with the description: | |
2381 | ||
2382 | Architectures: which instruction set architectures provide this ioctl. | |
2383 | x86 includes both i386 and x86_64. | |
2384 | ||
2385 | Parameters: what parameters are accepted by the capability. | |
2386 | ||
2387 | Returns: the return value. General error numbers (EBADF, ENOMEM, EINVAL) | |
2388 | are not detailed, but errors with specific meanings are. | |
2389 | ||
414fa985 | 2390 | |
821246a5 AG |
2391 | 6.1 KVM_CAP_PPC_OSI |
2392 | ||
2393 | Architectures: ppc | |
2394 | Parameters: none | |
2395 | Returns: 0 on success; -1 on error | |
2396 | ||
2397 | This capability enables interception of OSI hypercalls that otherwise would | |
2398 | be treated as normal system calls to be injected into the guest. OSI hypercalls | |
2399 | were invented by Mac-on-Linux to have a standardized communication mechanism | |
2400 | between the guest and the host. | |
2401 | ||
2402 | When this capability is enabled, KVM_EXIT_OSI can occur. | |
2403 | ||
414fa985 | 2404 | |
821246a5 AG |
2405 | 6.2 KVM_CAP_PPC_PAPR |
2406 | ||
2407 | Architectures: ppc | |
2408 | Parameters: none | |
2409 | Returns: 0 on success; -1 on error | |
2410 | ||
2411 | This capability enables interception of PAPR hypercalls. PAPR hypercalls are | |
2412 | done using the hypercall instruction "sc 1". | |
2413 | ||
2414 | It also sets the guest privilege level to "supervisor" mode. Usually the guest | |
2415 | runs in "hypervisor" privilege mode with a few missing features. | |
2416 | ||
2417 | In addition to the above, it changes the semantics of SDR1. In this mode, the | |
2418 | HTAB address part of SDR1 contains an HVA instead of a GPA, as PAPR keeps the | |
2419 | HTAB invisible to the guest. | |
2420 | ||
2421 | When this capability is enabled, KVM_EXIT_PAPR_HCALL can occur. | |
dc83b8bc | 2422 | |
414fa985 | 2423 | |
dc83b8bc SW |
2424 | 6.3 KVM_CAP_SW_TLB |
2425 | ||
2426 | Architectures: ppc | |
2427 | Parameters: args[0] is the address of a struct kvm_config_tlb | |
2428 | Returns: 0 on success; -1 on error | |
2429 | ||
2430 | struct kvm_config_tlb { | |
2431 | __u64 params; | |
2432 | __u64 array; | |
2433 | __u32 mmu_type; | |
2434 | __u32 array_len; | |
2435 | }; | |
2436 | ||
2437 | Configures the virtual CPU's TLB array, establishing a shared memory area | |
2438 | between userspace and KVM. The "params" and "array" fields are userspace | |
2439 | addresses of mmu-type-specific data structures. The "array_len" field is an | |
2440 | safety mechanism, and should be set to the size in bytes of the memory that | |
2441 | userspace has reserved for the array. It must be at least the size dictated | |
2442 | by "mmu_type" and "params". | |
2443 | ||
2444 | While KVM_RUN is active, the shared region is under control of KVM. Its | |
2445 | contents are undefined, and any modification by userspace results in | |
2446 | boundedly undefined behavior. | |
2447 | ||
2448 | On return from KVM_RUN, the shared region will reflect the current state of | |
2449 | the guest's TLB. If userspace makes any changes, it must call KVM_DIRTY_TLB | |
2450 | to tell KVM which entries have been changed, prior to calling KVM_RUN again | |
2451 | on this vcpu. | |
2452 | ||
2453 | For mmu types KVM_MMU_FSL_BOOKE_NOHV and KVM_MMU_FSL_BOOKE_HV: | |
2454 | - The "params" field is of type "struct kvm_book3e_206_tlb_params". | |
2455 | - The "array" field points to an array of type "struct | |
2456 | kvm_book3e_206_tlb_entry". | |
2457 | - The array consists of all entries in the first TLB, followed by all | |
2458 | entries in the second TLB. | |
2459 | - Within a TLB, entries are ordered first by increasing set number. Within a | |
2460 | set, entries are ordered by way (increasing ESEL). | |
2461 | - The hash for determining set number in TLB0 is: (MAS2 >> 12) & (num_sets - 1) | |
2462 | where "num_sets" is the tlb_sizes[] value divided by the tlb_ways[] value. | |
2463 | - The tsize field of mas1 shall be set to 4K on TLB0, even though the | |
2464 | hardware ignores this value for TLB0. |