]>
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 | ||
eca6be56 SC |
47 | |
48 | It is important to note that althought VM ioctls may only be issued from | |
49 | the process that created the VM, a VM's lifecycle is associated with its | |
50 | file descriptor, not its creator (process). In other words, the VM and | |
51 | its resources, *including the associated address space*, are not freed | |
52 | until the last reference to the VM's file descriptor has been released. | |
53 | For example, if fork() is issued after ioctl(KVM_CREATE_VM), the VM will | |
54 | not be freed until both the parent (original) process and its child have | |
55 | put their references to the VM's file descriptor. | |
56 | ||
57 | Because a VM's resources are not freed until the last reference to its | |
58 | file descriptor is released, creating additional references to a VM via | |
59 | via fork(), dup(), etc... without careful consideration is strongly | |
60 | discouraged and may have unwanted side effects, e.g. memory allocated | |
61 | by and on behalf of the VM's process may not be freed/unaccounted when | |
62 | the VM is shut down. | |
63 | ||
414fa985 | 64 | |
9c1b96e3 | 65 | 3. Extensions |
414fa985 | 66 | ------------- |
9c1b96e3 AK |
67 | |
68 | As of Linux 2.6.22, the KVM ABI has been stabilized: no backward | |
69 | incompatible change are allowed. However, there is an extension | |
70 | facility that allows backward-compatible extensions to the API to be | |
71 | queried and used. | |
72 | ||
c9f3f2d8 | 73 | The extension mechanism is not based on the Linux version number. |
9c1b96e3 AK |
74 | Instead, kvm defines extension identifiers and a facility to query |
75 | whether a particular extension identifier is available. If it is, a | |
76 | set of ioctls is available for application use. | |
77 | ||
414fa985 | 78 | |
9c1b96e3 | 79 | 4. API description |
414fa985 | 80 | ------------------ |
9c1b96e3 AK |
81 | |
82 | This section describes ioctls that can be used to control kvm guests. | |
83 | For each ioctl, the following information is provided along with a | |
84 | description: | |
85 | ||
86 | Capability: which KVM extension provides this ioctl. Can be 'basic', | |
87 | which means that is will be provided by any kernel that supports | |
7f05db6a | 88 | API version 12 (see section 4.1), a KVM_CAP_xyz constant, which |
9c1b96e3 | 89 | means availability needs to be checked with KVM_CHECK_EXTENSION |
7f05db6a MT |
90 | (see section 4.4), or 'none' which means that while not all kernels |
91 | support this ioctl, there's no capability bit to check its | |
92 | availability: for kernels that don't support the ioctl, | |
93 | the ioctl returns -ENOTTY. | |
9c1b96e3 AK |
94 | |
95 | Architectures: which instruction set architectures provide this ioctl. | |
96 | x86 includes both i386 and x86_64. | |
97 | ||
98 | Type: system, vm, or vcpu. | |
99 | ||
100 | Parameters: what parameters are accepted by the ioctl. | |
101 | ||
102 | Returns: the return value. General error numbers (EBADF, ENOMEM, EINVAL) | |
103 | are not detailed, but errors with specific meanings are. | |
104 | ||
414fa985 | 105 | |
9c1b96e3 AK |
106 | 4.1 KVM_GET_API_VERSION |
107 | ||
108 | Capability: basic | |
109 | Architectures: all | |
110 | Type: system ioctl | |
111 | Parameters: none | |
112 | Returns: the constant KVM_API_VERSION (=12) | |
113 | ||
114 | This identifies the API version as the stable kvm API. It is not | |
115 | expected that this number will change. However, Linux 2.6.20 and | |
116 | 2.6.21 report earlier versions; these are not documented and not | |
117 | supported. Applications should refuse to run if KVM_GET_API_VERSION | |
118 | returns a value other than 12. If this check passes, all ioctls | |
119 | described as 'basic' will be available. | |
120 | ||
414fa985 | 121 | |
9c1b96e3 AK |
122 | 4.2 KVM_CREATE_VM |
123 | ||
124 | Capability: basic | |
125 | Architectures: all | |
126 | Type: system ioctl | |
e08b9637 | 127 | Parameters: machine type identifier (KVM_VM_*) |
9c1b96e3 AK |
128 | Returns: a VM fd that can be used to control the new virtual machine. |
129 | ||
bcb85c88 | 130 | The new VM has no virtual cpus and no memory. |
a8a3c426 | 131 | You probably want to use 0 as machine type. |
e08b9637 CO |
132 | |
133 | In order to create user controlled virtual machines on S390, check | |
134 | KVM_CAP_S390_UCONTROL and use the flag KVM_VM_S390_UCONTROL as | |
135 | privileged user (CAP_SYS_ADMIN). | |
9c1b96e3 | 136 | |
a8a3c426 JH |
137 | To use hardware assisted virtualization on MIPS (VZ ASE) rather than |
138 | the default trap & emulate implementation (which changes the virtual | |
139 | memory layout to fit in user mode), check KVM_CAP_MIPS_VZ and use the | |
140 | flag KVM_VM_MIPS_VZ. | |
141 | ||
414fa985 | 142 | |
233a7cb2 SP |
143 | On arm64, the physical address size for a VM (IPA Size limit) is limited |
144 | to 40bits by default. The limit can be configured if the host supports the | |
145 | extension KVM_CAP_ARM_VM_IPA_SIZE. When supported, use | |
146 | KVM_VM_TYPE_ARM_IPA_SIZE(IPA_Bits) to set the size in the machine type | |
147 | identifier, where IPA_Bits is the maximum width of any physical | |
148 | address used by the VM. The IPA_Bits is encoded in bits[7-0] of the | |
149 | machine type identifier. | |
150 | ||
151 | e.g, to configure a guest to use 48bit physical address size : | |
152 | ||
153 | vm_fd = ioctl(dev_fd, KVM_CREATE_VM, KVM_VM_TYPE_ARM_IPA_SIZE(48)); | |
154 | ||
155 | The requested size (IPA_Bits) must be : | |
156 | 0 - Implies default size, 40bits (for backward compatibility) | |
157 | ||
158 | or | |
159 | ||
160 | N - Implies N bits, where N is a positive integer such that, | |
161 | 32 <= N <= Host_IPA_Limit | |
162 | ||
163 | Host_IPA_Limit is the maximum possible value for IPA_Bits on the host and | |
164 | is dependent on the CPU capability and the kernel configuration. The limit can | |
165 | be retrieved using KVM_CAP_ARM_VM_IPA_SIZE of the KVM_CHECK_EXTENSION | |
166 | ioctl() at run-time. | |
167 | ||
168 | Please note that configuring the IPA size does not affect the capability | |
169 | exposed by the guest CPUs in ID_AA64MMFR0_EL1[PARange]. It only affects | |
170 | size of the address translated by the stage2 level (guest physical to | |
171 | host physical address translations). | |
172 | ||
173 | ||
801e459a | 174 | 4.3 KVM_GET_MSR_INDEX_LIST, KVM_GET_MSR_FEATURE_INDEX_LIST |
9c1b96e3 | 175 | |
801e459a | 176 | Capability: basic, KVM_CAP_GET_MSR_FEATURES for KVM_GET_MSR_FEATURE_INDEX_LIST |
9c1b96e3 | 177 | Architectures: x86 |
801e459a | 178 | Type: system ioctl |
9c1b96e3 AK |
179 | Parameters: struct kvm_msr_list (in/out) |
180 | Returns: 0 on success; -1 on error | |
181 | Errors: | |
801e459a | 182 | EFAULT: the msr index list cannot be read from or written to |
9c1b96e3 AK |
183 | E2BIG: the msr index list is to be to fit in the array specified by |
184 | the user. | |
185 | ||
186 | struct kvm_msr_list { | |
187 | __u32 nmsrs; /* number of msrs in entries */ | |
188 | __u32 indices[0]; | |
189 | }; | |
190 | ||
801e459a TL |
191 | The user fills in the size of the indices array in nmsrs, and in return |
192 | kvm adjusts nmsrs to reflect the actual number of msrs and fills in the | |
193 | indices array with their numbers. | |
194 | ||
195 | KVM_GET_MSR_INDEX_LIST returns the guest msrs that are supported. The list | |
196 | varies by kvm version and host processor, but does not change otherwise. | |
9c1b96e3 | 197 | |
2e2602ca AK |
198 | Note: if kvm indicates supports MCE (KVM_CAP_MCE), then the MCE bank MSRs are |
199 | not returned in the MSR list, as different vcpus can have a different number | |
200 | of banks, as set via the KVM_X86_SETUP_MCE ioctl. | |
201 | ||
801e459a TL |
202 | KVM_GET_MSR_FEATURE_INDEX_LIST returns the list of MSRs that can be passed |
203 | to the KVM_GET_MSRS system ioctl. This lets userspace probe host capabilities | |
204 | and processor features that are exposed via MSRs (e.g., VMX capabilities). | |
205 | This list also varies by kvm version and host processor, but does not change | |
206 | otherwise. | |
207 | ||
414fa985 | 208 | |
9c1b96e3 AK |
209 | 4.4 KVM_CHECK_EXTENSION |
210 | ||
92b591a4 | 211 | Capability: basic, KVM_CAP_CHECK_EXTENSION_VM for vm ioctl |
9c1b96e3 | 212 | Architectures: all |
92b591a4 | 213 | Type: system ioctl, vm ioctl |
9c1b96e3 AK |
214 | Parameters: extension identifier (KVM_CAP_*) |
215 | Returns: 0 if unsupported; 1 (or some other positive integer) if supported | |
216 | ||
217 | The API allows the application to query about extensions to the core | |
218 | kvm API. Userspace passes an extension identifier (an integer) and | |
219 | receives an integer that describes the extension availability. | |
220 | Generally 0 means no and 1 means yes, but some extensions may report | |
221 | additional information in the integer return value. | |
222 | ||
92b591a4 AG |
223 | Based on their initialization different VMs may have different capabilities. |
224 | It is thus encouraged to use the vm ioctl to query for capabilities (available | |
225 | with KVM_CAP_CHECK_EXTENSION_VM on the vm fd) | |
414fa985 | 226 | |
9c1b96e3 AK |
227 | 4.5 KVM_GET_VCPU_MMAP_SIZE |
228 | ||
229 | Capability: basic | |
230 | Architectures: all | |
231 | Type: system ioctl | |
232 | Parameters: none | |
233 | Returns: size of vcpu mmap area, in bytes | |
234 | ||
235 | The KVM_RUN ioctl (cf.) communicates with userspace via a shared | |
236 | memory region. This ioctl returns the size of that region. See the | |
237 | KVM_RUN documentation for details. | |
238 | ||
414fa985 | 239 | |
9c1b96e3 AK |
240 | 4.6 KVM_SET_MEMORY_REGION |
241 | ||
242 | Capability: basic | |
243 | Architectures: all | |
244 | Type: vm ioctl | |
245 | Parameters: struct kvm_memory_region (in) | |
246 | Returns: 0 on success, -1 on error | |
247 | ||
b74a07be | 248 | This ioctl is obsolete and has been removed. |
9c1b96e3 | 249 | |
414fa985 | 250 | |
68ba6974 | 251 | 4.7 KVM_CREATE_VCPU |
9c1b96e3 AK |
252 | |
253 | Capability: basic | |
254 | Architectures: all | |
255 | Type: vm ioctl | |
256 | Parameters: vcpu id (apic id on x86) | |
257 | Returns: vcpu fd on success, -1 on error | |
258 | ||
0b1b1dfd GK |
259 | This API adds a vcpu to a virtual machine. No more than max_vcpus may be added. |
260 | The vcpu id is an integer in the range [0, max_vcpu_id). | |
8c3ba334 SL |
261 | |
262 | The recommended max_vcpus value can be retrieved using the KVM_CAP_NR_VCPUS of | |
263 | the KVM_CHECK_EXTENSION ioctl() at run-time. | |
264 | The maximum possible value for max_vcpus can be retrieved using the | |
265 | KVM_CAP_MAX_VCPUS of the KVM_CHECK_EXTENSION ioctl() at run-time. | |
266 | ||
76d25402 PE |
267 | If the KVM_CAP_NR_VCPUS does not exist, you should assume that max_vcpus is 4 |
268 | cpus max. | |
8c3ba334 SL |
269 | If the KVM_CAP_MAX_VCPUS does not exist, you should assume that max_vcpus is |
270 | same as the value returned from KVM_CAP_NR_VCPUS. | |
9c1b96e3 | 271 | |
0b1b1dfd GK |
272 | The maximum possible value for max_vcpu_id can be retrieved using the |
273 | KVM_CAP_MAX_VCPU_ID of the KVM_CHECK_EXTENSION ioctl() at run-time. | |
274 | ||
275 | If the KVM_CAP_MAX_VCPU_ID does not exist, you should assume that max_vcpu_id | |
276 | is the same as the value returned from KVM_CAP_MAX_VCPUS. | |
277 | ||
371fefd6 PM |
278 | On powerpc using book3s_hv mode, the vcpus are mapped onto virtual |
279 | threads in one or more virtual CPU cores. (This is because the | |
280 | hardware requires all the hardware threads in a CPU core to be in the | |
281 | same partition.) The KVM_CAP_PPC_SMT capability indicates the number | |
36442687 AK |
282 | of vcpus per virtual core (vcore). The vcore id is obtained by |
283 | dividing the vcpu id by the number of vcpus per vcore. The vcpus in a | |
284 | given vcore will always be in the same physical core as each other | |
285 | (though that might be a different physical core from time to time). | |
286 | Userspace can control the threading (SMT) mode of the guest by its | |
287 | allocation of vcpu ids. For example, if userspace wants | |
288 | single-threaded guest vcpus, it should make all vcpu ids be a multiple | |
289 | of the number of vcpus per vcore. | |
290 | ||
5b1c1493 CO |
291 | For virtual cpus that have been created with S390 user controlled virtual |
292 | machines, the resulting vcpu fd can be memory mapped at page offset | |
293 | KVM_S390_SIE_PAGE_OFFSET in order to obtain a memory map of the virtual | |
294 | cpu's hardware control block. | |
295 | ||
414fa985 | 296 | |
68ba6974 | 297 | 4.8 KVM_GET_DIRTY_LOG (vm ioctl) |
9c1b96e3 AK |
298 | |
299 | Capability: basic | |
300 | Architectures: x86 | |
301 | Type: vm ioctl | |
302 | Parameters: struct kvm_dirty_log (in/out) | |
303 | Returns: 0 on success, -1 on error | |
304 | ||
305 | /* for KVM_GET_DIRTY_LOG */ | |
306 | struct kvm_dirty_log { | |
307 | __u32 slot; | |
308 | __u32 padding; | |
309 | union { | |
310 | void __user *dirty_bitmap; /* one bit per page */ | |
311 | __u64 padding; | |
312 | }; | |
313 | }; | |
314 | ||
315 | Given a memory slot, return a bitmap containing any pages dirtied | |
316 | since the last call to this ioctl. Bit 0 is the first page in the | |
317 | memory slot. Ensure the entire structure is cleared to avoid padding | |
318 | issues. | |
319 | ||
f481b069 PB |
320 | If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 specifies |
321 | the address space for which you want to return the dirty bitmap. | |
322 | They must be less than the value that KVM_CHECK_EXTENSION returns for | |
323 | the KVM_CAP_MULTI_ADDRESS_SPACE capability. | |
324 | ||
2a31b9db PB |
325 | The bits in the dirty bitmap are cleared before the ioctl returns, unless |
326 | KVM_CAP_MANUAL_DIRTY_LOG_PROTECT is enabled. For more information, | |
327 | see the description of the capability. | |
414fa985 | 328 | |
68ba6974 | 329 | 4.9 KVM_SET_MEMORY_ALIAS |
9c1b96e3 AK |
330 | |
331 | Capability: basic | |
332 | Architectures: x86 | |
333 | Type: vm ioctl | |
334 | Parameters: struct kvm_memory_alias (in) | |
335 | Returns: 0 (success), -1 (error) | |
336 | ||
a1f4d395 | 337 | This ioctl is obsolete and has been removed. |
9c1b96e3 | 338 | |
414fa985 | 339 | |
68ba6974 | 340 | 4.10 KVM_RUN |
9c1b96e3 AK |
341 | |
342 | Capability: basic | |
343 | Architectures: all | |
344 | Type: vcpu ioctl | |
345 | Parameters: none | |
346 | Returns: 0 on success, -1 on error | |
347 | Errors: | |
348 | EINTR: an unmasked signal is pending | |
349 | ||
350 | This ioctl is used to run a guest virtual cpu. While there are no | |
351 | explicit parameters, there is an implicit parameter block that can be | |
352 | obtained by mmap()ing the vcpu fd at offset 0, with the size given by | |
353 | KVM_GET_VCPU_MMAP_SIZE. The parameter block is formatted as a 'struct | |
354 | kvm_run' (see below). | |
355 | ||
414fa985 | 356 | |
68ba6974 | 357 | 4.11 KVM_GET_REGS |
9c1b96e3 AK |
358 | |
359 | Capability: basic | |
379e04c7 | 360 | Architectures: all except ARM, arm64 |
9c1b96e3 AK |
361 | Type: vcpu ioctl |
362 | Parameters: struct kvm_regs (out) | |
363 | Returns: 0 on success, -1 on error | |
364 | ||
365 | Reads the general purpose registers from the vcpu. | |
366 | ||
367 | /* x86 */ | |
368 | struct kvm_regs { | |
369 | /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */ | |
370 | __u64 rax, rbx, rcx, rdx; | |
371 | __u64 rsi, rdi, rsp, rbp; | |
372 | __u64 r8, r9, r10, r11; | |
373 | __u64 r12, r13, r14, r15; | |
374 | __u64 rip, rflags; | |
375 | }; | |
376 | ||
c2d2c21b JH |
377 | /* mips */ |
378 | struct kvm_regs { | |
379 | /* out (KVM_GET_REGS) / in (KVM_SET_REGS) */ | |
380 | __u64 gpr[32]; | |
381 | __u64 hi; | |
382 | __u64 lo; | |
383 | __u64 pc; | |
384 | }; | |
385 | ||
414fa985 | 386 | |
68ba6974 | 387 | 4.12 KVM_SET_REGS |
9c1b96e3 AK |
388 | |
389 | Capability: basic | |
379e04c7 | 390 | Architectures: all except ARM, arm64 |
9c1b96e3 AK |
391 | Type: vcpu ioctl |
392 | Parameters: struct kvm_regs (in) | |
393 | Returns: 0 on success, -1 on error | |
394 | ||
395 | Writes the general purpose registers into the vcpu. | |
396 | ||
397 | See KVM_GET_REGS for the data structure. | |
398 | ||
414fa985 | 399 | |
68ba6974 | 400 | 4.13 KVM_GET_SREGS |
9c1b96e3 AK |
401 | |
402 | Capability: basic | |
5ce941ee | 403 | Architectures: x86, ppc |
9c1b96e3 AK |
404 | Type: vcpu ioctl |
405 | Parameters: struct kvm_sregs (out) | |
406 | Returns: 0 on success, -1 on error | |
407 | ||
408 | Reads special registers from the vcpu. | |
409 | ||
410 | /* x86 */ | |
411 | struct kvm_sregs { | |
412 | struct kvm_segment cs, ds, es, fs, gs, ss; | |
413 | struct kvm_segment tr, ldt; | |
414 | struct kvm_dtable gdt, idt; | |
415 | __u64 cr0, cr2, cr3, cr4, cr8; | |
416 | __u64 efer; | |
417 | __u64 apic_base; | |
418 | __u64 interrupt_bitmap[(KVM_NR_INTERRUPTS + 63) / 64]; | |
419 | }; | |
420 | ||
68e2ffed | 421 | /* ppc -- see arch/powerpc/include/uapi/asm/kvm.h */ |
5ce941ee | 422 | |
9c1b96e3 AK |
423 | interrupt_bitmap is a bitmap of pending external interrupts. At most |
424 | one bit may be set. This interrupt has been acknowledged by the APIC | |
425 | but not yet injected into the cpu core. | |
426 | ||
414fa985 | 427 | |
68ba6974 | 428 | 4.14 KVM_SET_SREGS |
9c1b96e3 AK |
429 | |
430 | Capability: basic | |
5ce941ee | 431 | Architectures: x86, ppc |
9c1b96e3 AK |
432 | Type: vcpu ioctl |
433 | Parameters: struct kvm_sregs (in) | |
434 | Returns: 0 on success, -1 on error | |
435 | ||
436 | Writes special registers into the vcpu. See KVM_GET_SREGS for the | |
437 | data structures. | |
438 | ||
414fa985 | 439 | |
68ba6974 | 440 | 4.15 KVM_TRANSLATE |
9c1b96e3 AK |
441 | |
442 | Capability: basic | |
443 | Architectures: x86 | |
444 | Type: vcpu ioctl | |
445 | Parameters: struct kvm_translation (in/out) | |
446 | Returns: 0 on success, -1 on error | |
447 | ||
448 | Translates a virtual address according to the vcpu's current address | |
449 | translation mode. | |
450 | ||
451 | struct kvm_translation { | |
452 | /* in */ | |
453 | __u64 linear_address; | |
454 | ||
455 | /* out */ | |
456 | __u64 physical_address; | |
457 | __u8 valid; | |
458 | __u8 writeable; | |
459 | __u8 usermode; | |
460 | __u8 pad[5]; | |
461 | }; | |
462 | ||
414fa985 | 463 | |
68ba6974 | 464 | 4.16 KVM_INTERRUPT |
9c1b96e3 AK |
465 | |
466 | Capability: basic | |
c2d2c21b | 467 | Architectures: x86, ppc, mips |
9c1b96e3 AK |
468 | Type: vcpu ioctl |
469 | Parameters: struct kvm_interrupt (in) | |
1c1a9ce9 | 470 | Returns: 0 on success, negative on failure. |
9c1b96e3 | 471 | |
1c1a9ce9 | 472 | Queues a hardware interrupt vector to be injected. |
9c1b96e3 AK |
473 | |
474 | /* for KVM_INTERRUPT */ | |
475 | struct kvm_interrupt { | |
476 | /* in */ | |
477 | __u32 irq; | |
478 | }; | |
479 | ||
6f7a2bd4 AG |
480 | X86: |
481 | ||
1c1a9ce9 SR |
482 | Returns: 0 on success, |
483 | -EEXIST if an interrupt is already enqueued | |
484 | -EINVAL the the irq number is invalid | |
485 | -ENXIO if the PIC is in the kernel | |
486 | -EFAULT if the pointer is invalid | |
487 | ||
488 | Note 'irq' is an interrupt vector, not an interrupt pin or line. This | |
489 | ioctl is useful if the in-kernel PIC is not used. | |
9c1b96e3 | 490 | |
6f7a2bd4 AG |
491 | PPC: |
492 | ||
493 | Queues an external interrupt to be injected. This ioctl is overleaded | |
494 | with 3 different irq values: | |
495 | ||
496 | a) KVM_INTERRUPT_SET | |
497 | ||
498 | This injects an edge type external interrupt into the guest once it's ready | |
499 | to receive interrupts. When injected, the interrupt is done. | |
500 | ||
501 | b) KVM_INTERRUPT_UNSET | |
502 | ||
503 | This unsets any pending interrupt. | |
504 | ||
505 | Only available with KVM_CAP_PPC_UNSET_IRQ. | |
506 | ||
507 | c) KVM_INTERRUPT_SET_LEVEL | |
508 | ||
509 | This injects a level type external interrupt into the guest context. The | |
510 | interrupt stays pending until a specific ioctl with KVM_INTERRUPT_UNSET | |
511 | is triggered. | |
512 | ||
513 | Only available with KVM_CAP_PPC_IRQ_LEVEL. | |
514 | ||
515 | Note that any value for 'irq' other than the ones stated above is invalid | |
516 | and incurs unexpected behavior. | |
517 | ||
c2d2c21b JH |
518 | MIPS: |
519 | ||
520 | Queues an external interrupt to be injected into the virtual CPU. A negative | |
521 | interrupt number dequeues the interrupt. | |
522 | ||
414fa985 | 523 | |
68ba6974 | 524 | 4.17 KVM_DEBUG_GUEST |
9c1b96e3 AK |
525 | |
526 | Capability: basic | |
527 | Architectures: none | |
528 | Type: vcpu ioctl | |
529 | Parameters: none) | |
530 | Returns: -1 on error | |
531 | ||
532 | Support for this has been removed. Use KVM_SET_GUEST_DEBUG instead. | |
533 | ||
414fa985 | 534 | |
68ba6974 | 535 | 4.18 KVM_GET_MSRS |
9c1b96e3 | 536 | |
801e459a | 537 | Capability: basic (vcpu), KVM_CAP_GET_MSR_FEATURES (system) |
9c1b96e3 | 538 | Architectures: x86 |
801e459a | 539 | Type: system ioctl, vcpu ioctl |
9c1b96e3 | 540 | Parameters: struct kvm_msrs (in/out) |
801e459a TL |
541 | Returns: number of msrs successfully returned; |
542 | -1 on error | |
543 | ||
544 | When used as a system ioctl: | |
545 | Reads the values of MSR-based features that are available for the VM. This | |
546 | is similar to KVM_GET_SUPPORTED_CPUID, but it returns MSR indices and values. | |
547 | The list of msr-based features can be obtained using KVM_GET_MSR_FEATURE_INDEX_LIST | |
548 | in a system ioctl. | |
9c1b96e3 | 549 | |
801e459a | 550 | When used as a vcpu ioctl: |
9c1b96e3 | 551 | Reads model-specific registers from the vcpu. Supported msr indices can |
801e459a | 552 | be obtained using KVM_GET_MSR_INDEX_LIST in a system ioctl. |
9c1b96e3 AK |
553 | |
554 | struct kvm_msrs { | |
555 | __u32 nmsrs; /* number of msrs in entries */ | |
556 | __u32 pad; | |
557 | ||
558 | struct kvm_msr_entry entries[0]; | |
559 | }; | |
560 | ||
561 | struct kvm_msr_entry { | |
562 | __u32 index; | |
563 | __u32 reserved; | |
564 | __u64 data; | |
565 | }; | |
566 | ||
567 | Application code should set the 'nmsrs' member (which indicates the | |
568 | size of the entries array) and the 'index' member of each array entry. | |
569 | kvm will fill in the 'data' member. | |
570 | ||
414fa985 | 571 | |
68ba6974 | 572 | 4.19 KVM_SET_MSRS |
9c1b96e3 AK |
573 | |
574 | Capability: basic | |
575 | Architectures: x86 | |
576 | Type: vcpu ioctl | |
577 | Parameters: struct kvm_msrs (in) | |
578 | Returns: 0 on success, -1 on error | |
579 | ||
580 | Writes model-specific registers to the vcpu. See KVM_GET_MSRS for the | |
581 | data structures. | |
582 | ||
583 | Application code should set the 'nmsrs' member (which indicates the | |
584 | size of the entries array), and the 'index' and 'data' members of each | |
585 | array entry. | |
586 | ||
414fa985 | 587 | |
68ba6974 | 588 | 4.20 KVM_SET_CPUID |
9c1b96e3 AK |
589 | |
590 | Capability: basic | |
591 | Architectures: x86 | |
592 | Type: vcpu ioctl | |
593 | Parameters: struct kvm_cpuid (in) | |
594 | Returns: 0 on success, -1 on error | |
595 | ||
596 | Defines the vcpu responses to the cpuid instruction. Applications | |
597 | should use the KVM_SET_CPUID2 ioctl if available. | |
598 | ||
599 | ||
600 | struct kvm_cpuid_entry { | |
601 | __u32 function; | |
602 | __u32 eax; | |
603 | __u32 ebx; | |
604 | __u32 ecx; | |
605 | __u32 edx; | |
606 | __u32 padding; | |
607 | }; | |
608 | ||
609 | /* for KVM_SET_CPUID */ | |
610 | struct kvm_cpuid { | |
611 | __u32 nent; | |
612 | __u32 padding; | |
613 | struct kvm_cpuid_entry entries[0]; | |
614 | }; | |
615 | ||
414fa985 | 616 | |
68ba6974 | 617 | 4.21 KVM_SET_SIGNAL_MASK |
9c1b96e3 AK |
618 | |
619 | Capability: basic | |
572e0929 | 620 | Architectures: all |
9c1b96e3 AK |
621 | Type: vcpu ioctl |
622 | Parameters: struct kvm_signal_mask (in) | |
623 | Returns: 0 on success, -1 on error | |
624 | ||
625 | Defines which signals are blocked during execution of KVM_RUN. This | |
626 | signal mask temporarily overrides the threads signal mask. Any | |
627 | unblocked signal received (except SIGKILL and SIGSTOP, which retain | |
628 | their traditional behaviour) will cause KVM_RUN to return with -EINTR. | |
629 | ||
630 | Note the signal will only be delivered if not blocked by the original | |
631 | signal mask. | |
632 | ||
633 | /* for KVM_SET_SIGNAL_MASK */ | |
634 | struct kvm_signal_mask { | |
635 | __u32 len; | |
636 | __u8 sigset[0]; | |
637 | }; | |
638 | ||
414fa985 | 639 | |
68ba6974 | 640 | 4.22 KVM_GET_FPU |
9c1b96e3 AK |
641 | |
642 | Capability: basic | |
643 | Architectures: x86 | |
644 | Type: vcpu ioctl | |
645 | Parameters: struct kvm_fpu (out) | |
646 | Returns: 0 on success, -1 on error | |
647 | ||
648 | Reads the floating point state from the vcpu. | |
649 | ||
650 | /* for KVM_GET_FPU and KVM_SET_FPU */ | |
651 | struct kvm_fpu { | |
652 | __u8 fpr[8][16]; | |
653 | __u16 fcw; | |
654 | __u16 fsw; | |
655 | __u8 ftwx; /* in fxsave format */ | |
656 | __u8 pad1; | |
657 | __u16 last_opcode; | |
658 | __u64 last_ip; | |
659 | __u64 last_dp; | |
660 | __u8 xmm[16][16]; | |
661 | __u32 mxcsr; | |
662 | __u32 pad2; | |
663 | }; | |
664 | ||
414fa985 | 665 | |
68ba6974 | 666 | 4.23 KVM_SET_FPU |
9c1b96e3 AK |
667 | |
668 | Capability: basic | |
669 | Architectures: x86 | |
670 | Type: vcpu ioctl | |
671 | Parameters: struct kvm_fpu (in) | |
672 | Returns: 0 on success, -1 on error | |
673 | ||
674 | Writes the floating point state to the vcpu. | |
675 | ||
676 | /* for KVM_GET_FPU and KVM_SET_FPU */ | |
677 | struct kvm_fpu { | |
678 | __u8 fpr[8][16]; | |
679 | __u16 fcw; | |
680 | __u16 fsw; | |
681 | __u8 ftwx; /* in fxsave format */ | |
682 | __u8 pad1; | |
683 | __u16 last_opcode; | |
684 | __u64 last_ip; | |
685 | __u64 last_dp; | |
686 | __u8 xmm[16][16]; | |
687 | __u32 mxcsr; | |
688 | __u32 pad2; | |
689 | }; | |
690 | ||
414fa985 | 691 | |
68ba6974 | 692 | 4.24 KVM_CREATE_IRQCHIP |
5dadbfd6 | 693 | |
84223598 | 694 | Capability: KVM_CAP_IRQCHIP, KVM_CAP_S390_IRQCHIP (s390) |
c32a4272 | 695 | Architectures: x86, ARM, arm64, s390 |
5dadbfd6 AK |
696 | Type: vm ioctl |
697 | Parameters: none | |
698 | Returns: 0 on success, -1 on error | |
699 | ||
ac3d3735 AP |
700 | Creates an interrupt controller model in the kernel. |
701 | On x86, creates a virtual ioapic, a virtual PIC (two PICs, nested), and sets up | |
702 | future vcpus to have a local APIC. IRQ routing for GSIs 0-15 is set to both | |
703 | PIC and IOAPIC; GSI 16-23 only go to the IOAPIC. | |
704 | On ARM/arm64, a GICv2 is created. Any other GIC versions require the usage of | |
705 | KVM_CREATE_DEVICE, which also supports creating a GICv2. Using | |
706 | KVM_CREATE_DEVICE is preferred over KVM_CREATE_IRQCHIP for GICv2. | |
707 | On s390, a dummy irq routing table is created. | |
84223598 CH |
708 | |
709 | Note that on s390 the KVM_CAP_S390_IRQCHIP vm capability needs to be enabled | |
710 | before KVM_CREATE_IRQCHIP can be used. | |
5dadbfd6 | 711 | |
414fa985 | 712 | |
68ba6974 | 713 | 4.25 KVM_IRQ_LINE |
5dadbfd6 AK |
714 | |
715 | Capability: KVM_CAP_IRQCHIP | |
c32a4272 | 716 | Architectures: x86, arm, arm64 |
5dadbfd6 AK |
717 | Type: vm ioctl |
718 | Parameters: struct kvm_irq_level | |
719 | Returns: 0 on success, -1 on error | |
720 | ||
721 | Sets the level of a GSI input to the interrupt controller model in the kernel. | |
86ce8535 CD |
722 | On some architectures it is required that an interrupt controller model has |
723 | been previously created with KVM_CREATE_IRQCHIP. Note that edge-triggered | |
724 | interrupts require the level to be set to 1 and then back to 0. | |
725 | ||
100943c5 GS |
726 | On real hardware, interrupt pins can be active-low or active-high. This |
727 | does not matter for the level field of struct kvm_irq_level: 1 always | |
728 | means active (asserted), 0 means inactive (deasserted). | |
729 | ||
730 | x86 allows the operating system to program the interrupt polarity | |
731 | (active-low/active-high) for level-triggered interrupts, and KVM used | |
732 | to consider the polarity. However, due to bitrot in the handling of | |
733 | active-low interrupts, the above convention is now valid on x86 too. | |
734 | This is signaled by KVM_CAP_X86_IOAPIC_POLARITY_IGNORED. Userspace | |
735 | should not present interrupts to the guest as active-low unless this | |
736 | capability is present (or unless it is not using the in-kernel irqchip, | |
737 | of course). | |
738 | ||
739 | ||
379e04c7 MZ |
740 | ARM/arm64 can signal an interrupt either at the CPU level, or at the |
741 | in-kernel irqchip (GIC), and for in-kernel irqchip can tell the GIC to | |
742 | use PPIs designated for specific cpus. The irq field is interpreted | |
743 | like this: | |
86ce8535 CD |
744 | |
745 | bits: | 31 ... 24 | 23 ... 16 | 15 ... 0 | | |
746 | field: | irq_type | vcpu_index | irq_id | | |
747 | ||
748 | The irq_type field has the following values: | |
749 | - irq_type[0]: out-of-kernel GIC: irq_id 0 is IRQ, irq_id 1 is FIQ | |
750 | - irq_type[1]: in-kernel GIC: SPI, irq_id between 32 and 1019 (incl.) | |
751 | (the vcpu_index field is ignored) | |
752 | - irq_type[2]: in-kernel GIC: PPI, irq_id between 16 and 31 (incl.) | |
753 | ||
754 | (The irq_id field thus corresponds nicely to the IRQ ID in the ARM GIC specs) | |
755 | ||
100943c5 | 756 | In both cases, level is used to assert/deassert the line. |
5dadbfd6 AK |
757 | |
758 | struct kvm_irq_level { | |
759 | union { | |
760 | __u32 irq; /* GSI */ | |
761 | __s32 status; /* not used for KVM_IRQ_LEVEL */ | |
762 | }; | |
763 | __u32 level; /* 0 or 1 */ | |
764 | }; | |
765 | ||
414fa985 | 766 | |
68ba6974 | 767 | 4.26 KVM_GET_IRQCHIP |
5dadbfd6 AK |
768 | |
769 | Capability: KVM_CAP_IRQCHIP | |
c32a4272 | 770 | Architectures: x86 |
5dadbfd6 AK |
771 | Type: vm ioctl |
772 | Parameters: struct kvm_irqchip (in/out) | |
773 | Returns: 0 on success, -1 on error | |
774 | ||
775 | Reads the state of a kernel interrupt controller created with | |
776 | KVM_CREATE_IRQCHIP into a buffer provided by the caller. | |
777 | ||
778 | struct kvm_irqchip { | |
779 | __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */ | |
780 | __u32 pad; | |
781 | union { | |
782 | char dummy[512]; /* reserving space */ | |
783 | struct kvm_pic_state pic; | |
784 | struct kvm_ioapic_state ioapic; | |
785 | } chip; | |
786 | }; | |
787 | ||
414fa985 | 788 | |
68ba6974 | 789 | 4.27 KVM_SET_IRQCHIP |
5dadbfd6 AK |
790 | |
791 | Capability: KVM_CAP_IRQCHIP | |
c32a4272 | 792 | Architectures: x86 |
5dadbfd6 AK |
793 | Type: vm ioctl |
794 | Parameters: struct kvm_irqchip (in) | |
795 | Returns: 0 on success, -1 on error | |
796 | ||
797 | Sets the state of a kernel interrupt controller created with | |
798 | KVM_CREATE_IRQCHIP from a buffer provided by the caller. | |
799 | ||
800 | struct kvm_irqchip { | |
801 | __u32 chip_id; /* 0 = PIC1, 1 = PIC2, 2 = IOAPIC */ | |
802 | __u32 pad; | |
803 | union { | |
804 | char dummy[512]; /* reserving space */ | |
805 | struct kvm_pic_state pic; | |
806 | struct kvm_ioapic_state ioapic; | |
807 | } chip; | |
808 | }; | |
809 | ||
414fa985 | 810 | |
68ba6974 | 811 | 4.28 KVM_XEN_HVM_CONFIG |
ffde22ac ES |
812 | |
813 | Capability: KVM_CAP_XEN_HVM | |
814 | Architectures: x86 | |
815 | Type: vm ioctl | |
816 | Parameters: struct kvm_xen_hvm_config (in) | |
817 | Returns: 0 on success, -1 on error | |
818 | ||
819 | Sets the MSR that the Xen HVM guest uses to initialize its hypercall | |
820 | page, and provides the starting address and size of the hypercall | |
821 | blobs in userspace. When the guest writes the MSR, kvm copies one | |
822 | page of a blob (32- or 64-bit, depending on the vcpu mode) to guest | |
823 | memory. | |
824 | ||
825 | struct kvm_xen_hvm_config { | |
826 | __u32 flags; | |
827 | __u32 msr; | |
828 | __u64 blob_addr_32; | |
829 | __u64 blob_addr_64; | |
830 | __u8 blob_size_32; | |
831 | __u8 blob_size_64; | |
832 | __u8 pad2[30]; | |
833 | }; | |
834 | ||
414fa985 | 835 | |
68ba6974 | 836 | 4.29 KVM_GET_CLOCK |
afbcf7ab GC |
837 | |
838 | Capability: KVM_CAP_ADJUST_CLOCK | |
839 | Architectures: x86 | |
840 | Type: vm ioctl | |
841 | Parameters: struct kvm_clock_data (out) | |
842 | Returns: 0 on success, -1 on error | |
843 | ||
844 | Gets the current timestamp of kvmclock as seen by the current guest. In | |
845 | conjunction with KVM_SET_CLOCK, it is used to ensure monotonicity on scenarios | |
846 | such as migration. | |
847 | ||
e3fd9a93 PB |
848 | When KVM_CAP_ADJUST_CLOCK is passed to KVM_CHECK_EXTENSION, it returns the |
849 | set of bits that KVM can return in struct kvm_clock_data's flag member. | |
850 | ||
851 | The only flag defined now is KVM_CLOCK_TSC_STABLE. If set, the returned | |
852 | value is the exact kvmclock value seen by all VCPUs at the instant | |
853 | when KVM_GET_CLOCK was called. If clear, the returned value is simply | |
854 | CLOCK_MONOTONIC plus a constant offset; the offset can be modified | |
855 | with KVM_SET_CLOCK. KVM will try to make all VCPUs follow this clock, | |
856 | but the exact value read by each VCPU could differ, because the host | |
857 | TSC is not stable. | |
858 | ||
afbcf7ab GC |
859 | struct kvm_clock_data { |
860 | __u64 clock; /* kvmclock current value */ | |
861 | __u32 flags; | |
862 | __u32 pad[9]; | |
863 | }; | |
864 | ||
414fa985 | 865 | |
68ba6974 | 866 | 4.30 KVM_SET_CLOCK |
afbcf7ab GC |
867 | |
868 | Capability: KVM_CAP_ADJUST_CLOCK | |
869 | Architectures: x86 | |
870 | Type: vm ioctl | |
871 | Parameters: struct kvm_clock_data (in) | |
872 | Returns: 0 on success, -1 on error | |
873 | ||
2044892d | 874 | Sets the current timestamp of kvmclock to the value specified in its parameter. |
afbcf7ab GC |
875 | In conjunction with KVM_GET_CLOCK, it is used to ensure monotonicity on scenarios |
876 | such as migration. | |
877 | ||
878 | struct kvm_clock_data { | |
879 | __u64 clock; /* kvmclock current value */ | |
880 | __u32 flags; | |
881 | __u32 pad[9]; | |
882 | }; | |
883 | ||
414fa985 | 884 | |
68ba6974 | 885 | 4.31 KVM_GET_VCPU_EVENTS |
3cfc3092 JK |
886 | |
887 | Capability: KVM_CAP_VCPU_EVENTS | |
48005f64 | 888 | Extended by: KVM_CAP_INTR_SHADOW |
b0960b95 | 889 | Architectures: x86, arm, arm64 |
b7b27fac | 890 | Type: vcpu ioctl |
3cfc3092 JK |
891 | Parameters: struct kvm_vcpu_event (out) |
892 | Returns: 0 on success, -1 on error | |
893 | ||
b7b27fac DG |
894 | X86: |
895 | ||
3cfc3092 JK |
896 | Gets currently pending exceptions, interrupts, and NMIs as well as related |
897 | states of the vcpu. | |
898 | ||
899 | struct kvm_vcpu_events { | |
900 | struct { | |
901 | __u8 injected; | |
902 | __u8 nr; | |
903 | __u8 has_error_code; | |
59073aaf | 904 | __u8 pending; |
3cfc3092 JK |
905 | __u32 error_code; |
906 | } exception; | |
907 | struct { | |
908 | __u8 injected; | |
909 | __u8 nr; | |
910 | __u8 soft; | |
48005f64 | 911 | __u8 shadow; |
3cfc3092 JK |
912 | } interrupt; |
913 | struct { | |
914 | __u8 injected; | |
915 | __u8 pending; | |
916 | __u8 masked; | |
917 | __u8 pad; | |
918 | } nmi; | |
919 | __u32 sipi_vector; | |
dab4b911 | 920 | __u32 flags; |
f077825a PB |
921 | struct { |
922 | __u8 smm; | |
923 | __u8 pending; | |
924 | __u8 smm_inside_nmi; | |
925 | __u8 latched_init; | |
926 | } smi; | |
59073aaf JM |
927 | __u8 reserved[27]; |
928 | __u8 exception_has_payload; | |
929 | __u64 exception_payload; | |
3cfc3092 JK |
930 | }; |
931 | ||
59073aaf | 932 | The following bits are defined in the flags field: |
f077825a | 933 | |
59073aaf | 934 | - KVM_VCPUEVENT_VALID_SHADOW may be set to signal that |
f077825a | 935 | interrupt.shadow contains a valid state. |
48005f64 | 936 | |
59073aaf JM |
937 | - KVM_VCPUEVENT_VALID_SMM may be set to signal that smi contains a |
938 | valid state. | |
939 | ||
940 | - KVM_VCPUEVENT_VALID_PAYLOAD may be set to signal that the | |
941 | exception_has_payload, exception_payload, and exception.pending | |
942 | fields contain a valid state. This bit will be set whenever | |
943 | KVM_CAP_EXCEPTION_PAYLOAD is enabled. | |
414fa985 | 944 | |
b0960b95 | 945 | ARM/ARM64: |
b7b27fac DG |
946 | |
947 | If the guest accesses a device that is being emulated by the host kernel in | |
948 | such a way that a real device would generate a physical SError, KVM may make | |
949 | a virtual SError pending for that VCPU. This system error interrupt remains | |
950 | pending until the guest takes the exception by unmasking PSTATE.A. | |
951 | ||
952 | Running the VCPU may cause it to take a pending SError, or make an access that | |
953 | causes an SError to become pending. The event's description is only valid while | |
954 | the VPCU is not running. | |
955 | ||
956 | This API provides a way to read and write the pending 'event' state that is not | |
957 | visible to the guest. To save, restore or migrate a VCPU the struct representing | |
958 | the state can be read then written using this GET/SET API, along with the other | |
959 | guest-visible registers. It is not possible to 'cancel' an SError that has been | |
960 | made pending. | |
961 | ||
962 | A device being emulated in user-space may also wish to generate an SError. To do | |
963 | this the events structure can be populated by user-space. The current state | |
964 | should be read first, to ensure no existing SError is pending. If an existing | |
965 | SError is pending, the architecture's 'Multiple SError interrupts' rules should | |
966 | be followed. (2.5.3 of DDI0587.a "ARM Reliability, Availability, and | |
967 | Serviceability (RAS) Specification"). | |
968 | ||
be26b3a7 DG |
969 | SError exceptions always have an ESR value. Some CPUs have the ability to |
970 | specify what the virtual SError's ESR value should be. These systems will | |
688e0581 | 971 | advertise KVM_CAP_ARM_INJECT_SERROR_ESR. In this case exception.has_esr will |
be26b3a7 DG |
972 | always have a non-zero value when read, and the agent making an SError pending |
973 | should specify the ISS field in the lower 24 bits of exception.serror_esr. If | |
688e0581 | 974 | the system supports KVM_CAP_ARM_INJECT_SERROR_ESR, but user-space sets the events |
be26b3a7 DG |
975 | with exception.has_esr as zero, KVM will choose an ESR. |
976 | ||
977 | Specifying exception.has_esr on a system that does not support it will return | |
978 | -EINVAL. Setting anything other than the lower 24bits of exception.serror_esr | |
979 | will return -EINVAL. | |
980 | ||
b7b27fac DG |
981 | struct kvm_vcpu_events { |
982 | struct { | |
983 | __u8 serror_pending; | |
984 | __u8 serror_has_esr; | |
985 | /* Align it to 8 bytes */ | |
986 | __u8 pad[6]; | |
987 | __u64 serror_esr; | |
988 | } exception; | |
989 | __u32 reserved[12]; | |
990 | }; | |
991 | ||
68ba6974 | 992 | 4.32 KVM_SET_VCPU_EVENTS |
3cfc3092 JK |
993 | |
994 | Capability: KVM_CAP_VCPU_EVENTS | |
48005f64 | 995 | Extended by: KVM_CAP_INTR_SHADOW |
b0960b95 | 996 | Architectures: x86, arm, arm64 |
b7b27fac | 997 | Type: vcpu ioctl |
3cfc3092 JK |
998 | Parameters: struct kvm_vcpu_event (in) |
999 | Returns: 0 on success, -1 on error | |
1000 | ||
b7b27fac DG |
1001 | X86: |
1002 | ||
3cfc3092 JK |
1003 | Set pending exceptions, interrupts, and NMIs as well as related states of the |
1004 | vcpu. | |
1005 | ||
1006 | See KVM_GET_VCPU_EVENTS for the data structure. | |
1007 | ||
dab4b911 | 1008 | Fields that may be modified asynchronously by running VCPUs can be excluded |
f077825a PB |
1009 | from the update. These fields are nmi.pending, sipi_vector, smi.smm, |
1010 | smi.pending. Keep the corresponding bits in the flags field cleared to | |
1011 | suppress overwriting the current in-kernel state. The bits are: | |
dab4b911 JK |
1012 | |
1013 | KVM_VCPUEVENT_VALID_NMI_PENDING - transfer nmi.pending to the kernel | |
1014 | KVM_VCPUEVENT_VALID_SIPI_VECTOR - transfer sipi_vector | |
f077825a | 1015 | KVM_VCPUEVENT_VALID_SMM - transfer the smi sub-struct. |
dab4b911 | 1016 | |
48005f64 JK |
1017 | If KVM_CAP_INTR_SHADOW is available, KVM_VCPUEVENT_VALID_SHADOW can be set in |
1018 | the flags field to signal that interrupt.shadow contains a valid state and | |
1019 | shall be written into the VCPU. | |
1020 | ||
f077825a PB |
1021 | KVM_VCPUEVENT_VALID_SMM can only be set if KVM_CAP_X86_SMM is available. |
1022 | ||
59073aaf JM |
1023 | If KVM_CAP_EXCEPTION_PAYLOAD is enabled, KVM_VCPUEVENT_VALID_PAYLOAD |
1024 | can be set in the flags field to signal that the | |
1025 | exception_has_payload, exception_payload, and exception.pending fields | |
1026 | contain a valid state and shall be written into the VCPU. | |
1027 | ||
b0960b95 | 1028 | ARM/ARM64: |
b7b27fac DG |
1029 | |
1030 | Set the pending SError exception state for this VCPU. It is not possible to | |
1031 | 'cancel' an Serror that has been made pending. | |
1032 | ||
1033 | See KVM_GET_VCPU_EVENTS for the data structure. | |
1034 | ||
414fa985 | 1035 | |
68ba6974 | 1036 | 4.33 KVM_GET_DEBUGREGS |
a1efbe77 JK |
1037 | |
1038 | Capability: KVM_CAP_DEBUGREGS | |
1039 | Architectures: x86 | |
1040 | Type: vm ioctl | |
1041 | Parameters: struct kvm_debugregs (out) | |
1042 | Returns: 0 on success, -1 on error | |
1043 | ||
1044 | Reads debug registers from the vcpu. | |
1045 | ||
1046 | struct kvm_debugregs { | |
1047 | __u64 db[4]; | |
1048 | __u64 dr6; | |
1049 | __u64 dr7; | |
1050 | __u64 flags; | |
1051 | __u64 reserved[9]; | |
1052 | }; | |
1053 | ||
414fa985 | 1054 | |
68ba6974 | 1055 | 4.34 KVM_SET_DEBUGREGS |
a1efbe77 JK |
1056 | |
1057 | Capability: KVM_CAP_DEBUGREGS | |
1058 | Architectures: x86 | |
1059 | Type: vm ioctl | |
1060 | Parameters: struct kvm_debugregs (in) | |
1061 | Returns: 0 on success, -1 on error | |
1062 | ||
1063 | Writes debug registers into the vcpu. | |
1064 | ||
1065 | See KVM_GET_DEBUGREGS for the data structure. The flags field is unused | |
1066 | yet and must be cleared on entry. | |
1067 | ||
414fa985 | 1068 | |
68ba6974 | 1069 | 4.35 KVM_SET_USER_MEMORY_REGION |
0f2d8f4d AK |
1070 | |
1071 | Capability: KVM_CAP_USER_MEM | |
1072 | Architectures: all | |
1073 | Type: vm ioctl | |
1074 | Parameters: struct kvm_userspace_memory_region (in) | |
1075 | Returns: 0 on success, -1 on error | |
1076 | ||
1077 | struct kvm_userspace_memory_region { | |
1078 | __u32 slot; | |
1079 | __u32 flags; | |
1080 | __u64 guest_phys_addr; | |
1081 | __u64 memory_size; /* bytes */ | |
1082 | __u64 userspace_addr; /* start of the userspace allocated memory */ | |
1083 | }; | |
1084 | ||
1085 | /* for kvm_memory_region::flags */ | |
4d8b81ab XG |
1086 | #define KVM_MEM_LOG_DIRTY_PAGES (1UL << 0) |
1087 | #define KVM_MEM_READONLY (1UL << 1) | |
0f2d8f4d AK |
1088 | |
1089 | This ioctl allows the user to create or modify a guest physical memory | |
1090 | slot. When changing an existing slot, it may be moved in the guest | |
1091 | physical memory space, or its flags may be modified. It may not be | |
1092 | resized. Slots may not overlap in guest physical address space. | |
a677e704 LC |
1093 | Bits 0-15 of "slot" specifies the slot id and this value should be |
1094 | less than the maximum number of user memory slots supported per VM. | |
1095 | The maximum allowed slots can be queried using KVM_CAP_NR_MEMSLOTS, | |
1096 | if this capability is supported by the architecture. | |
0f2d8f4d | 1097 | |
f481b069 PB |
1098 | If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 of "slot" |
1099 | specifies the address space which is being modified. They must be | |
1100 | less than the value that KVM_CHECK_EXTENSION returns for the | |
1101 | KVM_CAP_MULTI_ADDRESS_SPACE capability. Slots in separate address spaces | |
1102 | are unrelated; the restriction on overlapping slots only applies within | |
1103 | each address space. | |
1104 | ||
0f2d8f4d AK |
1105 | Memory for the region is taken starting at the address denoted by the |
1106 | field userspace_addr, which must point at user addressable memory for | |
1107 | the entire memory slot size. Any object may back this memory, including | |
1108 | anonymous memory, ordinary files, and hugetlbfs. | |
1109 | ||
1110 | It is recommended that the lower 21 bits of guest_phys_addr and userspace_addr | |
1111 | be identical. This allows large pages in the guest to be backed by large | |
1112 | pages in the host. | |
1113 | ||
75d61fbc TY |
1114 | The flags field supports two flags: KVM_MEM_LOG_DIRTY_PAGES and |
1115 | KVM_MEM_READONLY. The former can be set to instruct KVM to keep track of | |
1116 | writes to memory within the slot. See KVM_GET_DIRTY_LOG ioctl to know how to | |
1117 | use it. The latter can be set, if KVM_CAP_READONLY_MEM capability allows it, | |
1118 | to make a new slot read-only. In this case, writes to this memory will be | |
1119 | posted to userspace as KVM_EXIT_MMIO exits. | |
7efd8fa1 JK |
1120 | |
1121 | When the KVM_CAP_SYNC_MMU capability is available, changes in the backing of | |
1122 | the memory region are automatically reflected into the guest. For example, an | |
1123 | mmap() that affects the region will be made visible immediately. Another | |
1124 | example is madvise(MADV_DROP). | |
0f2d8f4d AK |
1125 | |
1126 | It is recommended to use this API instead of the KVM_SET_MEMORY_REGION ioctl. | |
1127 | The KVM_SET_MEMORY_REGION does not allow fine grained control over memory | |
1128 | allocation and is deprecated. | |
3cfc3092 | 1129 | |
414fa985 | 1130 | |
68ba6974 | 1131 | 4.36 KVM_SET_TSS_ADDR |
8a5416db AK |
1132 | |
1133 | Capability: KVM_CAP_SET_TSS_ADDR | |
1134 | Architectures: x86 | |
1135 | Type: vm ioctl | |
1136 | Parameters: unsigned long tss_address (in) | |
1137 | Returns: 0 on success, -1 on error | |
1138 | ||
1139 | This ioctl defines the physical address of a three-page region in the guest | |
1140 | physical address space. The region must be within the first 4GB of the | |
1141 | guest physical address space and must not conflict with any memory slot | |
1142 | or any mmio address. The guest may malfunction if it accesses this memory | |
1143 | region. | |
1144 | ||
1145 | This ioctl is required on Intel-based hosts. This is needed on Intel hardware | |
1146 | because of a quirk in the virtualization implementation (see the internals | |
1147 | documentation when it pops into existence). | |
1148 | ||
414fa985 | 1149 | |
68ba6974 | 1150 | 4.37 KVM_ENABLE_CAP |
71fbfd5f | 1151 | |
e5d83c74 PB |
1152 | Capability: KVM_CAP_ENABLE_CAP |
1153 | Architectures: mips, ppc, s390 | |
1154 | Type: vcpu ioctl | |
1155 | Parameters: struct kvm_enable_cap (in) | |
1156 | Returns: 0 on success; -1 on error | |
1157 | ||
1158 | Capability: KVM_CAP_ENABLE_CAP_VM | |
1159 | Architectures: all | |
1160 | Type: vcpu ioctl | |
71fbfd5f AG |
1161 | Parameters: struct kvm_enable_cap (in) |
1162 | Returns: 0 on success; -1 on error | |
1163 | ||
1164 | +Not all extensions are enabled by default. Using this ioctl the application | |
1165 | can enable an extension, making it available to the guest. | |
1166 | ||
1167 | On systems that do not support this ioctl, it always fails. On systems that | |
1168 | do support it, it only works for extensions that are supported for enablement. | |
1169 | ||
1170 | To check if a capability can be enabled, the KVM_CHECK_EXTENSION ioctl should | |
1171 | be used. | |
1172 | ||
1173 | struct kvm_enable_cap { | |
1174 | /* in */ | |
1175 | __u32 cap; | |
1176 | ||
1177 | The capability that is supposed to get enabled. | |
1178 | ||
1179 | __u32 flags; | |
1180 | ||
1181 | A bitfield indicating future enhancements. Has to be 0 for now. | |
1182 | ||
1183 | __u64 args[4]; | |
1184 | ||
1185 | Arguments for enabling a feature. If a feature needs initial values to | |
1186 | function properly, this is the place to put them. | |
1187 | ||
1188 | __u8 pad[64]; | |
1189 | }; | |
1190 | ||
d938dc55 CH |
1191 | The vcpu ioctl should be used for vcpu-specific capabilities, the vm ioctl |
1192 | for vm-wide capabilities. | |
414fa985 | 1193 | |
68ba6974 | 1194 | 4.38 KVM_GET_MP_STATE |
b843f065 AK |
1195 | |
1196 | Capability: KVM_CAP_MP_STATE | |
ecccf0cc | 1197 | Architectures: x86, s390, arm, arm64 |
b843f065 AK |
1198 | Type: vcpu ioctl |
1199 | Parameters: struct kvm_mp_state (out) | |
1200 | Returns: 0 on success; -1 on error | |
1201 | ||
1202 | struct kvm_mp_state { | |
1203 | __u32 mp_state; | |
1204 | }; | |
1205 | ||
1206 | Returns the vcpu's current "multiprocessing state" (though also valid on | |
1207 | uniprocessor guests). | |
1208 | ||
1209 | Possible values are: | |
1210 | ||
ecccf0cc | 1211 | - KVM_MP_STATE_RUNNABLE: the vcpu is currently running [x86,arm/arm64] |
b843f065 | 1212 | - KVM_MP_STATE_UNINITIALIZED: the vcpu is an application processor (AP) |
c32a4272 | 1213 | which has not yet received an INIT signal [x86] |
b843f065 | 1214 | - KVM_MP_STATE_INIT_RECEIVED: the vcpu has received an INIT signal, and is |
c32a4272 | 1215 | now ready for a SIPI [x86] |
b843f065 | 1216 | - KVM_MP_STATE_HALTED: the vcpu has executed a HLT instruction and |
c32a4272 | 1217 | is waiting for an interrupt [x86] |
b843f065 | 1218 | - KVM_MP_STATE_SIPI_RECEIVED: the vcpu has just received a SIPI (vector |
c32a4272 | 1219 | accessible via KVM_GET_VCPU_EVENTS) [x86] |
ecccf0cc | 1220 | - KVM_MP_STATE_STOPPED: the vcpu is stopped [s390,arm/arm64] |
6352e4d2 DH |
1221 | - KVM_MP_STATE_CHECK_STOP: the vcpu is in a special error state [s390] |
1222 | - KVM_MP_STATE_OPERATING: the vcpu is operating (running or halted) | |
1223 | [s390] | |
1224 | - KVM_MP_STATE_LOAD: the vcpu is in a special load/startup state | |
1225 | [s390] | |
b843f065 | 1226 | |
c32a4272 | 1227 | On x86, this ioctl is only useful after KVM_CREATE_IRQCHIP. Without an |
0b4820d6 DH |
1228 | in-kernel irqchip, the multiprocessing state must be maintained by userspace on |
1229 | these architectures. | |
b843f065 | 1230 | |
ecccf0cc AB |
1231 | For arm/arm64: |
1232 | ||
1233 | The only states that are valid are KVM_MP_STATE_STOPPED and | |
1234 | KVM_MP_STATE_RUNNABLE which reflect if the vcpu is paused or not. | |
414fa985 | 1235 | |
68ba6974 | 1236 | 4.39 KVM_SET_MP_STATE |
b843f065 AK |
1237 | |
1238 | Capability: KVM_CAP_MP_STATE | |
ecccf0cc | 1239 | Architectures: x86, s390, arm, arm64 |
b843f065 AK |
1240 | Type: vcpu ioctl |
1241 | Parameters: struct kvm_mp_state (in) | |
1242 | Returns: 0 on success; -1 on error | |
1243 | ||
1244 | Sets the vcpu's current "multiprocessing state"; see KVM_GET_MP_STATE for | |
1245 | arguments. | |
1246 | ||
c32a4272 | 1247 | On x86, this ioctl is only useful after KVM_CREATE_IRQCHIP. Without an |
0b4820d6 DH |
1248 | in-kernel irqchip, the multiprocessing state must be maintained by userspace on |
1249 | these architectures. | |
b843f065 | 1250 | |
ecccf0cc AB |
1251 | For arm/arm64: |
1252 | ||
1253 | The only states that are valid are KVM_MP_STATE_STOPPED and | |
1254 | KVM_MP_STATE_RUNNABLE which reflect if the vcpu should be paused or not. | |
414fa985 | 1255 | |
68ba6974 | 1256 | 4.40 KVM_SET_IDENTITY_MAP_ADDR |
47dbb84f AK |
1257 | |
1258 | Capability: KVM_CAP_SET_IDENTITY_MAP_ADDR | |
1259 | Architectures: x86 | |
1260 | Type: vm ioctl | |
1261 | Parameters: unsigned long identity (in) | |
1262 | Returns: 0 on success, -1 on error | |
1263 | ||
1264 | This ioctl defines the physical address of a one-page region in the guest | |
1265 | physical address space. The region must be within the first 4GB of the | |
1266 | guest physical address space and must not conflict with any memory slot | |
1267 | or any mmio address. The guest may malfunction if it accesses this memory | |
1268 | region. | |
1269 | ||
726b99c4 DH |
1270 | Setting the address to 0 will result in resetting the address to its default |
1271 | (0xfffbc000). | |
1272 | ||
47dbb84f AK |
1273 | This ioctl is required on Intel-based hosts. This is needed on Intel hardware |
1274 | because of a quirk in the virtualization implementation (see the internals | |
1275 | documentation when it pops into existence). | |
1276 | ||
1af1ac91 | 1277 | Fails if any VCPU has already been created. |
414fa985 | 1278 | |
68ba6974 | 1279 | 4.41 KVM_SET_BOOT_CPU_ID |
57bc24cf AK |
1280 | |
1281 | Capability: KVM_CAP_SET_BOOT_CPU_ID | |
c32a4272 | 1282 | Architectures: x86 |
57bc24cf AK |
1283 | Type: vm ioctl |
1284 | Parameters: unsigned long vcpu_id | |
1285 | Returns: 0 on success, -1 on error | |
1286 | ||
1287 | Define which vcpu is the Bootstrap Processor (BSP). Values are the same | |
1288 | as the vcpu id in KVM_CREATE_VCPU. If this ioctl is not called, the default | |
1289 | is vcpu 0. | |
1290 | ||
414fa985 | 1291 | |
68ba6974 | 1292 | 4.42 KVM_GET_XSAVE |
2d5b5a66 SY |
1293 | |
1294 | Capability: KVM_CAP_XSAVE | |
1295 | Architectures: x86 | |
1296 | Type: vcpu ioctl | |
1297 | Parameters: struct kvm_xsave (out) | |
1298 | Returns: 0 on success, -1 on error | |
1299 | ||
1300 | struct kvm_xsave { | |
1301 | __u32 region[1024]; | |
1302 | }; | |
1303 | ||
1304 | This ioctl would copy current vcpu's xsave struct to the userspace. | |
1305 | ||
414fa985 | 1306 | |
68ba6974 | 1307 | 4.43 KVM_SET_XSAVE |
2d5b5a66 SY |
1308 | |
1309 | Capability: KVM_CAP_XSAVE | |
1310 | Architectures: x86 | |
1311 | Type: vcpu ioctl | |
1312 | Parameters: struct kvm_xsave (in) | |
1313 | Returns: 0 on success, -1 on error | |
1314 | ||
1315 | struct kvm_xsave { | |
1316 | __u32 region[1024]; | |
1317 | }; | |
1318 | ||
1319 | This ioctl would copy userspace's xsave struct to the kernel. | |
1320 | ||
414fa985 | 1321 | |
68ba6974 | 1322 | 4.44 KVM_GET_XCRS |
2d5b5a66 SY |
1323 | |
1324 | Capability: KVM_CAP_XCRS | |
1325 | Architectures: x86 | |
1326 | Type: vcpu ioctl | |
1327 | Parameters: struct kvm_xcrs (out) | |
1328 | Returns: 0 on success, -1 on error | |
1329 | ||
1330 | struct kvm_xcr { | |
1331 | __u32 xcr; | |
1332 | __u32 reserved; | |
1333 | __u64 value; | |
1334 | }; | |
1335 | ||
1336 | struct kvm_xcrs { | |
1337 | __u32 nr_xcrs; | |
1338 | __u32 flags; | |
1339 | struct kvm_xcr xcrs[KVM_MAX_XCRS]; | |
1340 | __u64 padding[16]; | |
1341 | }; | |
1342 | ||
1343 | This ioctl would copy current vcpu's xcrs to the userspace. | |
1344 | ||
414fa985 | 1345 | |
68ba6974 | 1346 | 4.45 KVM_SET_XCRS |
2d5b5a66 SY |
1347 | |
1348 | Capability: KVM_CAP_XCRS | |
1349 | Architectures: x86 | |
1350 | Type: vcpu ioctl | |
1351 | Parameters: struct kvm_xcrs (in) | |
1352 | Returns: 0 on success, -1 on error | |
1353 | ||
1354 | struct kvm_xcr { | |
1355 | __u32 xcr; | |
1356 | __u32 reserved; | |
1357 | __u64 value; | |
1358 | }; | |
1359 | ||
1360 | struct kvm_xcrs { | |
1361 | __u32 nr_xcrs; | |
1362 | __u32 flags; | |
1363 | struct kvm_xcr xcrs[KVM_MAX_XCRS]; | |
1364 | __u64 padding[16]; | |
1365 | }; | |
1366 | ||
1367 | This ioctl would set vcpu's xcr to the value userspace specified. | |
1368 | ||
414fa985 | 1369 | |
68ba6974 | 1370 | 4.46 KVM_GET_SUPPORTED_CPUID |
d153513d AK |
1371 | |
1372 | Capability: KVM_CAP_EXT_CPUID | |
1373 | Architectures: x86 | |
1374 | Type: system ioctl | |
1375 | Parameters: struct kvm_cpuid2 (in/out) | |
1376 | Returns: 0 on success, -1 on error | |
1377 | ||
1378 | struct kvm_cpuid2 { | |
1379 | __u32 nent; | |
1380 | __u32 padding; | |
1381 | struct kvm_cpuid_entry2 entries[0]; | |
1382 | }; | |
1383 | ||
9c15bb1d BP |
1384 | #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX BIT(0) |
1385 | #define KVM_CPUID_FLAG_STATEFUL_FUNC BIT(1) | |
1386 | #define KVM_CPUID_FLAG_STATE_READ_NEXT BIT(2) | |
d153513d AK |
1387 | |
1388 | struct kvm_cpuid_entry2 { | |
1389 | __u32 function; | |
1390 | __u32 index; | |
1391 | __u32 flags; | |
1392 | __u32 eax; | |
1393 | __u32 ebx; | |
1394 | __u32 ecx; | |
1395 | __u32 edx; | |
1396 | __u32 padding[3]; | |
1397 | }; | |
1398 | ||
df9cb9cc JM |
1399 | This ioctl returns x86 cpuid features which are supported by both the |
1400 | hardware and kvm in its default configuration. Userspace can use the | |
1401 | information returned by this ioctl to construct cpuid information (for | |
1402 | KVM_SET_CPUID2) that is consistent with hardware, kernel, and | |
1403 | userspace capabilities, and with user requirements (for example, the | |
1404 | user may wish to constrain cpuid to emulate older hardware, or for | |
1405 | feature consistency across a cluster). | |
1406 | ||
1407 | Note that certain capabilities, such as KVM_CAP_X86_DISABLE_EXITS, may | |
1408 | expose cpuid features (e.g. MONITOR) which are not supported by kvm in | |
1409 | its default configuration. If userspace enables such capabilities, it | |
1410 | is responsible for modifying the results of this ioctl appropriately. | |
d153513d AK |
1411 | |
1412 | Userspace invokes KVM_GET_SUPPORTED_CPUID by passing a kvm_cpuid2 structure | |
1413 | with the 'nent' field indicating the number of entries in the variable-size | |
1414 | array 'entries'. If the number of entries is too low to describe the cpu | |
1415 | capabilities, an error (E2BIG) is returned. If the number is too high, | |
1416 | the 'nent' field is adjusted and an error (ENOMEM) is returned. If the | |
1417 | number is just right, the 'nent' field is adjusted to the number of valid | |
1418 | entries in the 'entries' array, which is then filled. | |
1419 | ||
1420 | The entries returned are the host cpuid as returned by the cpuid instruction, | |
c39cbd2a AK |
1421 | with unknown or unsupported features masked out. Some features (for example, |
1422 | x2apic), may not be present in the host cpu, but are exposed by kvm if it can | |
1423 | emulate them efficiently. The fields in each entry are defined as follows: | |
d153513d AK |
1424 | |
1425 | function: the eax value used to obtain the entry | |
1426 | index: the ecx value used to obtain the entry (for entries that are | |
1427 | affected by ecx) | |
1428 | flags: an OR of zero or more of the following: | |
1429 | KVM_CPUID_FLAG_SIGNIFCANT_INDEX: | |
1430 | if the index field is valid | |
1431 | KVM_CPUID_FLAG_STATEFUL_FUNC: | |
1432 | if cpuid for this function returns different values for successive | |
1433 | invocations; there will be several entries with the same function, | |
1434 | all with this flag set | |
1435 | KVM_CPUID_FLAG_STATE_READ_NEXT: | |
1436 | for KVM_CPUID_FLAG_STATEFUL_FUNC entries, set if this entry is | |
1437 | the first entry to be read by a cpu | |
1438 | eax, ebx, ecx, edx: the values returned by the cpuid instruction for | |
1439 | this function/index combination | |
1440 | ||
4d25a066 JK |
1441 | The TSC deadline timer feature (CPUID leaf 1, ecx[24]) is always returned |
1442 | as false, since the feature depends on KVM_CREATE_IRQCHIP for local APIC | |
1443 | support. Instead it is reported via | |
1444 | ||
1445 | ioctl(KVM_CHECK_EXTENSION, KVM_CAP_TSC_DEADLINE_TIMER) | |
1446 | ||
1447 | if that returns true and you use KVM_CREATE_IRQCHIP, or if you emulate the | |
1448 | feature in userspace, then you can enable the feature for KVM_SET_CPUID2. | |
1449 | ||
414fa985 | 1450 | |
68ba6974 | 1451 | 4.47 KVM_PPC_GET_PVINFO |
15711e9c AG |
1452 | |
1453 | Capability: KVM_CAP_PPC_GET_PVINFO | |
1454 | Architectures: ppc | |
1455 | Type: vm ioctl | |
1456 | Parameters: struct kvm_ppc_pvinfo (out) | |
1457 | Returns: 0 on success, !0 on error | |
1458 | ||
1459 | struct kvm_ppc_pvinfo { | |
1460 | __u32 flags; | |
1461 | __u32 hcall[4]; | |
1462 | __u8 pad[108]; | |
1463 | }; | |
1464 | ||
1465 | This ioctl fetches PV specific information that need to be passed to the guest | |
1466 | using the device tree or other means from vm context. | |
1467 | ||
9202e076 | 1468 | The hcall array defines 4 instructions that make up a hypercall. |
15711e9c AG |
1469 | |
1470 | If any additional field gets added to this structure later on, a bit for that | |
1471 | additional piece of information will be set in the flags bitmap. | |
1472 | ||
9202e076 LYB |
1473 | The flags bitmap is defined as: |
1474 | ||
1475 | /* the host supports the ePAPR idle hcall | |
1476 | #define KVM_PPC_PVINFO_FLAGS_EV_IDLE (1<<0) | |
414fa985 | 1477 | |
68ba6974 | 1478 | 4.52 KVM_SET_GSI_ROUTING |
49f48172 JK |
1479 | |
1480 | Capability: KVM_CAP_IRQ_ROUTING | |
180ae7b1 | 1481 | Architectures: x86 s390 arm arm64 |
49f48172 JK |
1482 | Type: vm ioctl |
1483 | Parameters: struct kvm_irq_routing (in) | |
1484 | Returns: 0 on success, -1 on error | |
1485 | ||
1486 | Sets the GSI routing table entries, overwriting any previously set entries. | |
1487 | ||
180ae7b1 EA |
1488 | On arm/arm64, GSI routing has the following limitation: |
1489 | - GSI routing does not apply to KVM_IRQ_LINE but only to KVM_IRQFD. | |
1490 | ||
49f48172 JK |
1491 | struct kvm_irq_routing { |
1492 | __u32 nr; | |
1493 | __u32 flags; | |
1494 | struct kvm_irq_routing_entry entries[0]; | |
1495 | }; | |
1496 | ||
1497 | No flags are specified so far, the corresponding field must be set to zero. | |
1498 | ||
1499 | struct kvm_irq_routing_entry { | |
1500 | __u32 gsi; | |
1501 | __u32 type; | |
1502 | __u32 flags; | |
1503 | __u32 pad; | |
1504 | union { | |
1505 | struct kvm_irq_routing_irqchip irqchip; | |
1506 | struct kvm_irq_routing_msi msi; | |
84223598 | 1507 | struct kvm_irq_routing_s390_adapter adapter; |
5c919412 | 1508 | struct kvm_irq_routing_hv_sint hv_sint; |
49f48172 JK |
1509 | __u32 pad[8]; |
1510 | } u; | |
1511 | }; | |
1512 | ||
1513 | /* gsi routing entry types */ | |
1514 | #define KVM_IRQ_ROUTING_IRQCHIP 1 | |
1515 | #define KVM_IRQ_ROUTING_MSI 2 | |
84223598 | 1516 | #define KVM_IRQ_ROUTING_S390_ADAPTER 3 |
5c919412 | 1517 | #define KVM_IRQ_ROUTING_HV_SINT 4 |
49f48172 | 1518 | |
76a10b86 | 1519 | flags: |
6f49b2f3 PB |
1520 | - KVM_MSI_VALID_DEVID: used along with KVM_IRQ_ROUTING_MSI routing entry |
1521 | type, specifies that the devid field contains a valid value. The per-VM | |
1522 | KVM_CAP_MSI_DEVID capability advertises the requirement to provide | |
1523 | the device ID. If this capability is not available, userspace should | |
1524 | never set the KVM_MSI_VALID_DEVID flag as the ioctl might fail. | |
76a10b86 | 1525 | - zero otherwise |
49f48172 JK |
1526 | |
1527 | struct kvm_irq_routing_irqchip { | |
1528 | __u32 irqchip; | |
1529 | __u32 pin; | |
1530 | }; | |
1531 | ||
1532 | struct kvm_irq_routing_msi { | |
1533 | __u32 address_lo; | |
1534 | __u32 address_hi; | |
1535 | __u32 data; | |
76a10b86 EA |
1536 | union { |
1537 | __u32 pad; | |
1538 | __u32 devid; | |
1539 | }; | |
49f48172 JK |
1540 | }; |
1541 | ||
6f49b2f3 PB |
1542 | If KVM_MSI_VALID_DEVID is set, devid contains a unique device identifier |
1543 | for the device that wrote the MSI message. For PCI, this is usually a | |
1544 | BFD identifier in the lower 16 bits. | |
76a10b86 | 1545 | |
37131313 RK |
1546 | On x86, address_hi is ignored unless the KVM_X2APIC_API_USE_32BIT_IDS |
1547 | feature of KVM_CAP_X2APIC_API capability is enabled. If it is enabled, | |
1548 | address_hi bits 31-8 provide bits 31-8 of the destination id. Bits 7-0 of | |
1549 | address_hi must be zero. | |
1550 | ||
84223598 CH |
1551 | struct kvm_irq_routing_s390_adapter { |
1552 | __u64 ind_addr; | |
1553 | __u64 summary_addr; | |
1554 | __u64 ind_offset; | |
1555 | __u32 summary_offset; | |
1556 | __u32 adapter_id; | |
1557 | }; | |
1558 | ||
5c919412 AS |
1559 | struct kvm_irq_routing_hv_sint { |
1560 | __u32 vcpu; | |
1561 | __u32 sint; | |
1562 | }; | |
414fa985 | 1563 | |
414fa985 JK |
1564 | |
1565 | 4.55 KVM_SET_TSC_KHZ | |
92a1f12d JR |
1566 | |
1567 | Capability: KVM_CAP_TSC_CONTROL | |
1568 | Architectures: x86 | |
1569 | Type: vcpu ioctl | |
1570 | Parameters: virtual tsc_khz | |
1571 | Returns: 0 on success, -1 on error | |
1572 | ||
1573 | Specifies the tsc frequency for the virtual machine. The unit of the | |
1574 | frequency is KHz. | |
1575 | ||
414fa985 JK |
1576 | |
1577 | 4.56 KVM_GET_TSC_KHZ | |
92a1f12d JR |
1578 | |
1579 | Capability: KVM_CAP_GET_TSC_KHZ | |
1580 | Architectures: x86 | |
1581 | Type: vcpu ioctl | |
1582 | Parameters: none | |
1583 | Returns: virtual tsc-khz on success, negative value on error | |
1584 | ||
1585 | Returns the tsc frequency of the guest. The unit of the return value is | |
1586 | KHz. If the host has unstable tsc this ioctl returns -EIO instead as an | |
1587 | error. | |
1588 | ||
414fa985 JK |
1589 | |
1590 | 4.57 KVM_GET_LAPIC | |
e7677933 AK |
1591 | |
1592 | Capability: KVM_CAP_IRQCHIP | |
1593 | Architectures: x86 | |
1594 | Type: vcpu ioctl | |
1595 | Parameters: struct kvm_lapic_state (out) | |
1596 | Returns: 0 on success, -1 on error | |
1597 | ||
1598 | #define KVM_APIC_REG_SIZE 0x400 | |
1599 | struct kvm_lapic_state { | |
1600 | char regs[KVM_APIC_REG_SIZE]; | |
1601 | }; | |
1602 | ||
1603 | Reads the Local APIC registers and copies them into the input argument. The | |
1604 | data format and layout are the same as documented in the architecture manual. | |
1605 | ||
37131313 RK |
1606 | If KVM_X2APIC_API_USE_32BIT_IDS feature of KVM_CAP_X2APIC_API is |
1607 | enabled, then the format of APIC_ID register depends on the APIC mode | |
1608 | (reported by MSR_IA32_APICBASE) of its VCPU. x2APIC stores APIC ID in | |
1609 | the APIC_ID register (bytes 32-35). xAPIC only allows an 8-bit APIC ID | |
1610 | which is stored in bits 31-24 of the APIC register, or equivalently in | |
1611 | byte 35 of struct kvm_lapic_state's regs field. KVM_GET_LAPIC must then | |
1612 | be called after MSR_IA32_APICBASE has been set with KVM_SET_MSR. | |
1613 | ||
1614 | If KVM_X2APIC_API_USE_32BIT_IDS feature is disabled, struct kvm_lapic_state | |
1615 | always uses xAPIC format. | |
1616 | ||
414fa985 JK |
1617 | |
1618 | 4.58 KVM_SET_LAPIC | |
e7677933 AK |
1619 | |
1620 | Capability: KVM_CAP_IRQCHIP | |
1621 | Architectures: x86 | |
1622 | Type: vcpu ioctl | |
1623 | Parameters: struct kvm_lapic_state (in) | |
1624 | Returns: 0 on success, -1 on error | |
1625 | ||
1626 | #define KVM_APIC_REG_SIZE 0x400 | |
1627 | struct kvm_lapic_state { | |
1628 | char regs[KVM_APIC_REG_SIZE]; | |
1629 | }; | |
1630 | ||
df5cbb27 | 1631 | Copies the input argument into the Local APIC registers. The data format |
e7677933 AK |
1632 | and layout are the same as documented in the architecture manual. |
1633 | ||
37131313 RK |
1634 | The format of the APIC ID register (bytes 32-35 of struct kvm_lapic_state's |
1635 | regs field) depends on the state of the KVM_CAP_X2APIC_API capability. | |
1636 | See the note in KVM_GET_LAPIC. | |
1637 | ||
414fa985 JK |
1638 | |
1639 | 4.59 KVM_IOEVENTFD | |
55399a02 SL |
1640 | |
1641 | Capability: KVM_CAP_IOEVENTFD | |
1642 | Architectures: all | |
1643 | Type: vm ioctl | |
1644 | Parameters: struct kvm_ioeventfd (in) | |
1645 | Returns: 0 on success, !0 on error | |
1646 | ||
1647 | This ioctl attaches or detaches an ioeventfd to a legal pio/mmio address | |
1648 | within the guest. A guest write in the registered address will signal the | |
1649 | provided event instead of triggering an exit. | |
1650 | ||
1651 | struct kvm_ioeventfd { | |
1652 | __u64 datamatch; | |
1653 | __u64 addr; /* legal pio/mmio address */ | |
e9ea5069 | 1654 | __u32 len; /* 0, 1, 2, 4, or 8 bytes */ |
55399a02 SL |
1655 | __s32 fd; |
1656 | __u32 flags; | |
1657 | __u8 pad[36]; | |
1658 | }; | |
1659 | ||
2b83451b CH |
1660 | For the special case of virtio-ccw devices on s390, the ioevent is matched |
1661 | to a subchannel/virtqueue tuple instead. | |
1662 | ||
55399a02 SL |
1663 | The following flags are defined: |
1664 | ||
1665 | #define KVM_IOEVENTFD_FLAG_DATAMATCH (1 << kvm_ioeventfd_flag_nr_datamatch) | |
1666 | #define KVM_IOEVENTFD_FLAG_PIO (1 << kvm_ioeventfd_flag_nr_pio) | |
1667 | #define KVM_IOEVENTFD_FLAG_DEASSIGN (1 << kvm_ioeventfd_flag_nr_deassign) | |
2b83451b CH |
1668 | #define KVM_IOEVENTFD_FLAG_VIRTIO_CCW_NOTIFY \ |
1669 | (1 << kvm_ioeventfd_flag_nr_virtio_ccw_notify) | |
55399a02 SL |
1670 | |
1671 | If datamatch flag is set, the event will be signaled only if the written value | |
1672 | to the registered address is equal to datamatch in struct kvm_ioeventfd. | |
1673 | ||
2b83451b CH |
1674 | For virtio-ccw devices, addr contains the subchannel id and datamatch the |
1675 | virtqueue index. | |
1676 | ||
e9ea5069 JW |
1677 | With KVM_CAP_IOEVENTFD_ANY_LENGTH, a zero length ioeventfd is allowed, and |
1678 | the kernel will ignore the length of guest write and may get a faster vmexit. | |
1679 | The speedup may only apply to specific architectures, but the ioeventfd will | |
1680 | work anyway. | |
414fa985 JK |
1681 | |
1682 | 4.60 KVM_DIRTY_TLB | |
dc83b8bc SW |
1683 | |
1684 | Capability: KVM_CAP_SW_TLB | |
1685 | Architectures: ppc | |
1686 | Type: vcpu ioctl | |
1687 | Parameters: struct kvm_dirty_tlb (in) | |
1688 | Returns: 0 on success, -1 on error | |
1689 | ||
1690 | struct kvm_dirty_tlb { | |
1691 | __u64 bitmap; | |
1692 | __u32 num_dirty; | |
1693 | }; | |
1694 | ||
1695 | This must be called whenever userspace has changed an entry in the shared | |
1696 | TLB, prior to calling KVM_RUN on the associated vcpu. | |
1697 | ||
1698 | The "bitmap" field is the userspace address of an array. This array | |
1699 | consists of a number of bits, equal to the total number of TLB entries as | |
1700 | determined by the last successful call to KVM_CONFIG_TLB, rounded up to the | |
1701 | nearest multiple of 64. | |
1702 | ||
1703 | Each bit corresponds to one TLB entry, ordered the same as in the shared TLB | |
1704 | array. | |
1705 | ||
1706 | The array is little-endian: the bit 0 is the least significant bit of the | |
1707 | first byte, bit 8 is the least significant bit of the second byte, etc. | |
1708 | This avoids any complications with differing word sizes. | |
1709 | ||
1710 | The "num_dirty" field is a performance hint for KVM to determine whether it | |
1711 | should skip processing the bitmap and just invalidate everything. It must | |
1712 | be set to the number of set bits in the bitmap. | |
1713 | ||
414fa985 | 1714 | |
54738c09 DG |
1715 | 4.62 KVM_CREATE_SPAPR_TCE |
1716 | ||
1717 | Capability: KVM_CAP_SPAPR_TCE | |
1718 | Architectures: powerpc | |
1719 | Type: vm ioctl | |
1720 | Parameters: struct kvm_create_spapr_tce (in) | |
1721 | Returns: file descriptor for manipulating the created TCE table | |
1722 | ||
1723 | This creates a virtual TCE (translation control entry) table, which | |
1724 | is an IOMMU for PAPR-style virtual I/O. It is used to translate | |
1725 | logical addresses used in virtual I/O into guest physical addresses, | |
1726 | and provides a scatter/gather capability for PAPR virtual I/O. | |
1727 | ||
1728 | /* for KVM_CAP_SPAPR_TCE */ | |
1729 | struct kvm_create_spapr_tce { | |
1730 | __u64 liobn; | |
1731 | __u32 window_size; | |
1732 | }; | |
1733 | ||
1734 | The liobn field gives the logical IO bus number for which to create a | |
1735 | TCE table. The window_size field specifies the size of the DMA window | |
1736 | which this TCE table will translate - the table will contain one 64 | |
1737 | bit TCE entry for every 4kiB of the DMA window. | |
1738 | ||
1739 | When the guest issues an H_PUT_TCE hcall on a liobn for which a TCE | |
1740 | table has been created using this ioctl(), the kernel will handle it | |
1741 | in real mode, updating the TCE table. H_PUT_TCE calls for other | |
1742 | liobns will cause a vm exit and must be handled by userspace. | |
1743 | ||
1744 | The return value is a file descriptor which can be passed to mmap(2) | |
1745 | to map the created TCE table into userspace. This lets userspace read | |
1746 | the entries written by kernel-handled H_PUT_TCE calls, and also lets | |
1747 | userspace update the TCE table directly which is useful in some | |
1748 | circumstances. | |
1749 | ||
414fa985 | 1750 | |
aa04b4cc PM |
1751 | 4.63 KVM_ALLOCATE_RMA |
1752 | ||
1753 | Capability: KVM_CAP_PPC_RMA | |
1754 | Architectures: powerpc | |
1755 | Type: vm ioctl | |
1756 | Parameters: struct kvm_allocate_rma (out) | |
1757 | Returns: file descriptor for mapping the allocated RMA | |
1758 | ||
1759 | This allocates a Real Mode Area (RMA) from the pool allocated at boot | |
1760 | time by the kernel. An RMA is a physically-contiguous, aligned region | |
1761 | of memory used on older POWER processors to provide the memory which | |
1762 | will be accessed by real-mode (MMU off) accesses in a KVM guest. | |
1763 | POWER processors support a set of sizes for the RMA that usually | |
1764 | includes 64MB, 128MB, 256MB and some larger powers of two. | |
1765 | ||
1766 | /* for KVM_ALLOCATE_RMA */ | |
1767 | struct kvm_allocate_rma { | |
1768 | __u64 rma_size; | |
1769 | }; | |
1770 | ||
1771 | The return value is a file descriptor which can be passed to mmap(2) | |
1772 | to map the allocated RMA into userspace. The mapped area can then be | |
1773 | passed to the KVM_SET_USER_MEMORY_REGION ioctl to establish it as the | |
1774 | RMA for a virtual machine. The size of the RMA in bytes (which is | |
1775 | fixed at host kernel boot time) is returned in the rma_size field of | |
1776 | the argument structure. | |
1777 | ||
1778 | The KVM_CAP_PPC_RMA capability is 1 or 2 if the KVM_ALLOCATE_RMA ioctl | |
1779 | is supported; 2 if the processor requires all virtual machines to have | |
1780 | an RMA, or 1 if the processor can use an RMA but doesn't require it, | |
1781 | because it supports the Virtual RMA (VRMA) facility. | |
1782 | ||
414fa985 | 1783 | |
3f745f1e AK |
1784 | 4.64 KVM_NMI |
1785 | ||
1786 | Capability: KVM_CAP_USER_NMI | |
1787 | Architectures: x86 | |
1788 | Type: vcpu ioctl | |
1789 | Parameters: none | |
1790 | Returns: 0 on success, -1 on error | |
1791 | ||
1792 | Queues an NMI on the thread's vcpu. Note this is well defined only | |
1793 | when KVM_CREATE_IRQCHIP has not been called, since this is an interface | |
1794 | between the virtual cpu core and virtual local APIC. After KVM_CREATE_IRQCHIP | |
1795 | has been called, this interface is completely emulated within the kernel. | |
1796 | ||
1797 | To use this to emulate the LINT1 input with KVM_CREATE_IRQCHIP, use the | |
1798 | following algorithm: | |
1799 | ||
5d4f6f3d | 1800 | - pause the vcpu |
3f745f1e AK |
1801 | - read the local APIC's state (KVM_GET_LAPIC) |
1802 | - check whether changing LINT1 will queue an NMI (see the LVT entry for LINT1) | |
1803 | - if so, issue KVM_NMI | |
1804 | - resume the vcpu | |
1805 | ||
1806 | Some guests configure the LINT1 NMI input to cause a panic, aiding in | |
1807 | debugging. | |
1808 | ||
414fa985 | 1809 | |
e24ed81f | 1810 | 4.65 KVM_S390_UCAS_MAP |
27e0393f CO |
1811 | |
1812 | Capability: KVM_CAP_S390_UCONTROL | |
1813 | Architectures: s390 | |
1814 | Type: vcpu ioctl | |
1815 | Parameters: struct kvm_s390_ucas_mapping (in) | |
1816 | Returns: 0 in case of success | |
1817 | ||
1818 | The parameter is defined like this: | |
1819 | struct kvm_s390_ucas_mapping { | |
1820 | __u64 user_addr; | |
1821 | __u64 vcpu_addr; | |
1822 | __u64 length; | |
1823 | }; | |
1824 | ||
1825 | This ioctl maps the memory at "user_addr" with the length "length" to | |
1826 | the vcpu's address space starting at "vcpu_addr". All parameters need to | |
f884ab15 | 1827 | be aligned by 1 megabyte. |
27e0393f | 1828 | |
414fa985 | 1829 | |
e24ed81f | 1830 | 4.66 KVM_S390_UCAS_UNMAP |
27e0393f CO |
1831 | |
1832 | Capability: KVM_CAP_S390_UCONTROL | |
1833 | Architectures: s390 | |
1834 | Type: vcpu ioctl | |
1835 | Parameters: struct kvm_s390_ucas_mapping (in) | |
1836 | Returns: 0 in case of success | |
1837 | ||
1838 | The parameter is defined like this: | |
1839 | struct kvm_s390_ucas_mapping { | |
1840 | __u64 user_addr; | |
1841 | __u64 vcpu_addr; | |
1842 | __u64 length; | |
1843 | }; | |
1844 | ||
1845 | This ioctl unmaps the memory in the vcpu's address space starting at | |
1846 | "vcpu_addr" with the length "length". The field "user_addr" is ignored. | |
f884ab15 | 1847 | All parameters need to be aligned by 1 megabyte. |
27e0393f | 1848 | |
414fa985 | 1849 | |
e24ed81f | 1850 | 4.67 KVM_S390_VCPU_FAULT |
ccc7910f CO |
1851 | |
1852 | Capability: KVM_CAP_S390_UCONTROL | |
1853 | Architectures: s390 | |
1854 | Type: vcpu ioctl | |
1855 | Parameters: vcpu absolute address (in) | |
1856 | Returns: 0 in case of success | |
1857 | ||
1858 | This call creates a page table entry on the virtual cpu's address space | |
1859 | (for user controlled virtual machines) or the virtual machine's address | |
1860 | space (for regular virtual machines). This only works for minor faults, | |
1861 | thus it's recommended to access subject memory page via the user page | |
1862 | table upfront. This is useful to handle validity intercepts for user | |
1863 | controlled virtual machines to fault in the virtual cpu's lowcore pages | |
1864 | prior to calling the KVM_RUN ioctl. | |
1865 | ||
414fa985 | 1866 | |
e24ed81f AG |
1867 | 4.68 KVM_SET_ONE_REG |
1868 | ||
1869 | Capability: KVM_CAP_ONE_REG | |
1870 | Architectures: all | |
1871 | Type: vcpu ioctl | |
1872 | Parameters: struct kvm_one_reg (in) | |
1873 | Returns: 0 on success, negative value on failure | |
1874 | ||
1875 | struct kvm_one_reg { | |
1876 | __u64 id; | |
1877 | __u64 addr; | |
1878 | }; | |
1879 | ||
1880 | Using this ioctl, a single vcpu register can be set to a specific value | |
1881 | defined by user space with the passed in struct kvm_one_reg, where id | |
1882 | refers to the register identifier as described below and addr is a pointer | |
1883 | to a variable with the respective size. There can be architecture agnostic | |
1884 | and architecture specific registers. Each have their own range of operation | |
1885 | and their own constants and width. To keep track of the implemented | |
1886 | registers, find a list below: | |
1887 | ||
bf5590f3 JH |
1888 | Arch | Register | Width (bits) |
1889 | | | | |
1890 | PPC | KVM_REG_PPC_HIOR | 64 | |
1891 | PPC | KVM_REG_PPC_IAC1 | 64 | |
1892 | PPC | KVM_REG_PPC_IAC2 | 64 | |
1893 | PPC | KVM_REG_PPC_IAC3 | 64 | |
1894 | PPC | KVM_REG_PPC_IAC4 | 64 | |
1895 | PPC | KVM_REG_PPC_DAC1 | 64 | |
1896 | PPC | KVM_REG_PPC_DAC2 | 64 | |
1897 | PPC | KVM_REG_PPC_DABR | 64 | |
1898 | PPC | KVM_REG_PPC_DSCR | 64 | |
1899 | PPC | KVM_REG_PPC_PURR | 64 | |
1900 | PPC | KVM_REG_PPC_SPURR | 64 | |
1901 | PPC | KVM_REG_PPC_DAR | 64 | |
1902 | PPC | KVM_REG_PPC_DSISR | 32 | |
1903 | PPC | KVM_REG_PPC_AMR | 64 | |
1904 | PPC | KVM_REG_PPC_UAMOR | 64 | |
1905 | PPC | KVM_REG_PPC_MMCR0 | 64 | |
1906 | PPC | KVM_REG_PPC_MMCR1 | 64 | |
1907 | PPC | KVM_REG_PPC_MMCRA | 64 | |
1908 | PPC | KVM_REG_PPC_MMCR2 | 64 | |
1909 | PPC | KVM_REG_PPC_MMCRS | 64 | |
1910 | PPC | KVM_REG_PPC_SIAR | 64 | |
1911 | PPC | KVM_REG_PPC_SDAR | 64 | |
1912 | PPC | KVM_REG_PPC_SIER | 64 | |
1913 | PPC | KVM_REG_PPC_PMC1 | 32 | |
1914 | PPC | KVM_REG_PPC_PMC2 | 32 | |
1915 | PPC | KVM_REG_PPC_PMC3 | 32 | |
1916 | PPC | KVM_REG_PPC_PMC4 | 32 | |
1917 | PPC | KVM_REG_PPC_PMC5 | 32 | |
1918 | PPC | KVM_REG_PPC_PMC6 | 32 | |
1919 | PPC | KVM_REG_PPC_PMC7 | 32 | |
1920 | PPC | KVM_REG_PPC_PMC8 | 32 | |
1921 | PPC | KVM_REG_PPC_FPR0 | 64 | |
a8bd19ef | 1922 | ... |
bf5590f3 JH |
1923 | PPC | KVM_REG_PPC_FPR31 | 64 |
1924 | PPC | KVM_REG_PPC_VR0 | 128 | |
a8bd19ef | 1925 | ... |
bf5590f3 JH |
1926 | PPC | KVM_REG_PPC_VR31 | 128 |
1927 | PPC | KVM_REG_PPC_VSR0 | 128 | |
a8bd19ef | 1928 | ... |
bf5590f3 JH |
1929 | PPC | KVM_REG_PPC_VSR31 | 128 |
1930 | PPC | KVM_REG_PPC_FPSCR | 64 | |
1931 | PPC | KVM_REG_PPC_VSCR | 32 | |
1932 | PPC | KVM_REG_PPC_VPA_ADDR | 64 | |
1933 | PPC | KVM_REG_PPC_VPA_SLB | 128 | |
1934 | PPC | KVM_REG_PPC_VPA_DTL | 128 | |
1935 | PPC | KVM_REG_PPC_EPCR | 32 | |
1936 | PPC | KVM_REG_PPC_EPR | 32 | |
1937 | PPC | KVM_REG_PPC_TCR | 32 | |
1938 | PPC | KVM_REG_PPC_TSR | 32 | |
1939 | PPC | KVM_REG_PPC_OR_TSR | 32 | |
1940 | PPC | KVM_REG_PPC_CLEAR_TSR | 32 | |
1941 | PPC | KVM_REG_PPC_MAS0 | 32 | |
1942 | PPC | KVM_REG_PPC_MAS1 | 32 | |
1943 | PPC | KVM_REG_PPC_MAS2 | 64 | |
1944 | PPC | KVM_REG_PPC_MAS7_3 | 64 | |
1945 | PPC | KVM_REG_PPC_MAS4 | 32 | |
1946 | PPC | KVM_REG_PPC_MAS6 | 32 | |
1947 | PPC | KVM_REG_PPC_MMUCFG | 32 | |
1948 | PPC | KVM_REG_PPC_TLB0CFG | 32 | |
1949 | PPC | KVM_REG_PPC_TLB1CFG | 32 | |
1950 | PPC | KVM_REG_PPC_TLB2CFG | 32 | |
1951 | PPC | KVM_REG_PPC_TLB3CFG | 32 | |
1952 | PPC | KVM_REG_PPC_TLB0PS | 32 | |
1953 | PPC | KVM_REG_PPC_TLB1PS | 32 | |
1954 | PPC | KVM_REG_PPC_TLB2PS | 32 | |
1955 | PPC | KVM_REG_PPC_TLB3PS | 32 | |
1956 | PPC | KVM_REG_PPC_EPTCFG | 32 | |
1957 | PPC | KVM_REG_PPC_ICP_STATE | 64 | |
1958 | PPC | KVM_REG_PPC_TB_OFFSET | 64 | |
1959 | PPC | KVM_REG_PPC_SPMC1 | 32 | |
1960 | PPC | KVM_REG_PPC_SPMC2 | 32 | |
1961 | PPC | KVM_REG_PPC_IAMR | 64 | |
1962 | PPC | KVM_REG_PPC_TFHAR | 64 | |
1963 | PPC | KVM_REG_PPC_TFIAR | 64 | |
1964 | PPC | KVM_REG_PPC_TEXASR | 64 | |
1965 | PPC | KVM_REG_PPC_FSCR | 64 | |
1966 | PPC | KVM_REG_PPC_PSPB | 32 | |
1967 | PPC | KVM_REG_PPC_EBBHR | 64 | |
1968 | PPC | KVM_REG_PPC_EBBRR | 64 | |
1969 | PPC | KVM_REG_PPC_BESCR | 64 | |
1970 | PPC | KVM_REG_PPC_TAR | 64 | |
1971 | PPC | KVM_REG_PPC_DPDES | 64 | |
1972 | PPC | KVM_REG_PPC_DAWR | 64 | |
1973 | PPC | KVM_REG_PPC_DAWRX | 64 | |
1974 | PPC | KVM_REG_PPC_CIABR | 64 | |
1975 | PPC | KVM_REG_PPC_IC | 64 | |
1976 | PPC | KVM_REG_PPC_VTB | 64 | |
1977 | PPC | KVM_REG_PPC_CSIGR | 64 | |
1978 | PPC | KVM_REG_PPC_TACR | 64 | |
1979 | PPC | KVM_REG_PPC_TCSCR | 64 | |
1980 | PPC | KVM_REG_PPC_PID | 64 | |
1981 | PPC | KVM_REG_PPC_ACOP | 64 | |
1982 | PPC | KVM_REG_PPC_VRSAVE | 32 | |
cc568ead PB |
1983 | PPC | KVM_REG_PPC_LPCR | 32 |
1984 | PPC | KVM_REG_PPC_LPCR_64 | 64 | |
bf5590f3 JH |
1985 | PPC | KVM_REG_PPC_PPR | 64 |
1986 | PPC | KVM_REG_PPC_ARCH_COMPAT | 32 | |
1987 | PPC | KVM_REG_PPC_DABRX | 32 | |
1988 | PPC | KVM_REG_PPC_WORT | 64 | |
bc8a4e5c BB |
1989 | PPC | KVM_REG_PPC_SPRG9 | 64 |
1990 | PPC | KVM_REG_PPC_DBSR | 32 | |
e9cf1e08 PM |
1991 | PPC | KVM_REG_PPC_TIDR | 64 |
1992 | PPC | KVM_REG_PPC_PSSCR | 64 | |
5855564c | 1993 | PPC | KVM_REG_PPC_DEC_EXPIRY | 64 |
30323418 | 1994 | PPC | KVM_REG_PPC_PTCR | 64 |
bf5590f3 | 1995 | PPC | KVM_REG_PPC_TM_GPR0 | 64 |
3b783474 | 1996 | ... |
bf5590f3 JH |
1997 | PPC | KVM_REG_PPC_TM_GPR31 | 64 |
1998 | PPC | KVM_REG_PPC_TM_VSR0 | 128 | |
3b783474 | 1999 | ... |
bf5590f3 JH |
2000 | PPC | KVM_REG_PPC_TM_VSR63 | 128 |
2001 | PPC | KVM_REG_PPC_TM_CR | 64 | |
2002 | PPC | KVM_REG_PPC_TM_LR | 64 | |
2003 | PPC | KVM_REG_PPC_TM_CTR | 64 | |
2004 | PPC | KVM_REG_PPC_TM_FPSCR | 64 | |
2005 | PPC | KVM_REG_PPC_TM_AMR | 64 | |
2006 | PPC | KVM_REG_PPC_TM_PPR | 64 | |
2007 | PPC | KVM_REG_PPC_TM_VRSAVE | 64 | |
2008 | PPC | KVM_REG_PPC_TM_VSCR | 32 | |
2009 | PPC | KVM_REG_PPC_TM_DSCR | 64 | |
2010 | PPC | KVM_REG_PPC_TM_TAR | 64 | |
0d808df0 | 2011 | PPC | KVM_REG_PPC_TM_XER | 64 |
c2d2c21b JH |
2012 | | | |
2013 | MIPS | KVM_REG_MIPS_R0 | 64 | |
2014 | ... | |
2015 | MIPS | KVM_REG_MIPS_R31 | 64 | |
2016 | MIPS | KVM_REG_MIPS_HI | 64 | |
2017 | MIPS | KVM_REG_MIPS_LO | 64 | |
2018 | MIPS | KVM_REG_MIPS_PC | 64 | |
2019 | MIPS | KVM_REG_MIPS_CP0_INDEX | 32 | |
013044cc JH |
2020 | MIPS | KVM_REG_MIPS_CP0_ENTRYLO0 | 64 |
2021 | MIPS | KVM_REG_MIPS_CP0_ENTRYLO1 | 64 | |
c2d2c21b | 2022 | MIPS | KVM_REG_MIPS_CP0_CONTEXT | 64 |
dffe042f | 2023 | MIPS | KVM_REG_MIPS_CP0_CONTEXTCONFIG| 32 |
c2d2c21b | 2024 | MIPS | KVM_REG_MIPS_CP0_USERLOCAL | 64 |
dffe042f | 2025 | MIPS | KVM_REG_MIPS_CP0_XCONTEXTCONFIG| 64 |
c2d2c21b | 2026 | MIPS | KVM_REG_MIPS_CP0_PAGEMASK | 32 |
c992a4f6 | 2027 | MIPS | KVM_REG_MIPS_CP0_PAGEGRAIN | 32 |
4b7de028 JH |
2028 | MIPS | KVM_REG_MIPS_CP0_SEGCTL0 | 64 |
2029 | MIPS | KVM_REG_MIPS_CP0_SEGCTL1 | 64 | |
2030 | MIPS | KVM_REG_MIPS_CP0_SEGCTL2 | 64 | |
5a2f352f JH |
2031 | MIPS | KVM_REG_MIPS_CP0_PWBASE | 64 |
2032 | MIPS | KVM_REG_MIPS_CP0_PWFIELD | 64 | |
2033 | MIPS | KVM_REG_MIPS_CP0_PWSIZE | 64 | |
c2d2c21b | 2034 | MIPS | KVM_REG_MIPS_CP0_WIRED | 32 |
5a2f352f | 2035 | MIPS | KVM_REG_MIPS_CP0_PWCTL | 32 |
c2d2c21b JH |
2036 | MIPS | KVM_REG_MIPS_CP0_HWRENA | 32 |
2037 | MIPS | KVM_REG_MIPS_CP0_BADVADDR | 64 | |
edc89260 JH |
2038 | MIPS | KVM_REG_MIPS_CP0_BADINSTR | 32 |
2039 | MIPS | KVM_REG_MIPS_CP0_BADINSTRP | 32 | |
c2d2c21b JH |
2040 | MIPS | KVM_REG_MIPS_CP0_COUNT | 32 |
2041 | MIPS | KVM_REG_MIPS_CP0_ENTRYHI | 64 | |
2042 | MIPS | KVM_REG_MIPS_CP0_COMPARE | 32 | |
2043 | MIPS | KVM_REG_MIPS_CP0_STATUS | 32 | |
ad58d4d4 | 2044 | MIPS | KVM_REG_MIPS_CP0_INTCTL | 32 |
c2d2c21b JH |
2045 | MIPS | KVM_REG_MIPS_CP0_CAUSE | 32 |
2046 | MIPS | KVM_REG_MIPS_CP0_EPC | 64 | |
1068eaaf | 2047 | MIPS | KVM_REG_MIPS_CP0_PRID | 32 |
7801bbe1 | 2048 | MIPS | KVM_REG_MIPS_CP0_EBASE | 64 |
c2d2c21b JH |
2049 | MIPS | KVM_REG_MIPS_CP0_CONFIG | 32 |
2050 | MIPS | KVM_REG_MIPS_CP0_CONFIG1 | 32 | |
2051 | MIPS | KVM_REG_MIPS_CP0_CONFIG2 | 32 | |
2052 | MIPS | KVM_REG_MIPS_CP0_CONFIG3 | 32 | |
c771607a JH |
2053 | MIPS | KVM_REG_MIPS_CP0_CONFIG4 | 32 |
2054 | MIPS | KVM_REG_MIPS_CP0_CONFIG5 | 32 | |
c2d2c21b | 2055 | MIPS | KVM_REG_MIPS_CP0_CONFIG7 | 32 |
c992a4f6 | 2056 | MIPS | KVM_REG_MIPS_CP0_XCONTEXT | 64 |
c2d2c21b | 2057 | MIPS | KVM_REG_MIPS_CP0_ERROREPC | 64 |
05108709 JH |
2058 | MIPS | KVM_REG_MIPS_CP0_KSCRATCH1 | 64 |
2059 | MIPS | KVM_REG_MIPS_CP0_KSCRATCH2 | 64 | |
2060 | MIPS | KVM_REG_MIPS_CP0_KSCRATCH3 | 64 | |
2061 | MIPS | KVM_REG_MIPS_CP0_KSCRATCH4 | 64 | |
2062 | MIPS | KVM_REG_MIPS_CP0_KSCRATCH5 | 64 | |
2063 | MIPS | KVM_REG_MIPS_CP0_KSCRATCH6 | 64 | |
d42a008f | 2064 | MIPS | KVM_REG_MIPS_CP0_MAAR(0..63) | 64 |
c2d2c21b JH |
2065 | MIPS | KVM_REG_MIPS_COUNT_CTL | 64 |
2066 | MIPS | KVM_REG_MIPS_COUNT_RESUME | 64 | |
2067 | MIPS | KVM_REG_MIPS_COUNT_HZ | 64 | |
379245cd JH |
2068 | MIPS | KVM_REG_MIPS_FPR_32(0..31) | 32 |
2069 | MIPS | KVM_REG_MIPS_FPR_64(0..31) | 64 | |
ab86bd60 | 2070 | MIPS | KVM_REG_MIPS_VEC_128(0..31) | 128 |
379245cd JH |
2071 | MIPS | KVM_REG_MIPS_FCR_IR | 32 |
2072 | MIPS | KVM_REG_MIPS_FCR_CSR | 32 | |
ab86bd60 JH |
2073 | MIPS | KVM_REG_MIPS_MSA_IR | 32 |
2074 | MIPS | KVM_REG_MIPS_MSA_CSR | 32 | |
414fa985 | 2075 | |
749cf76c CD |
2076 | ARM registers are mapped using the lower 32 bits. The upper 16 of that |
2077 | is the register group type, or coprocessor number: | |
2078 | ||
2079 | ARM core registers have the following id bit patterns: | |
aa404ddf | 2080 | 0x4020 0000 0010 <index into the kvm_regs struct:16> |
749cf76c | 2081 | |
1138245c | 2082 | ARM 32-bit CP15 registers have the following id bit patterns: |
aa404ddf | 2083 | 0x4020 0000 000F <zero:1> <crn:4> <crm:4> <opc1:4> <opc2:3> |
1138245c CD |
2084 | |
2085 | ARM 64-bit CP15 registers have the following id bit patterns: | |
aa404ddf | 2086 | 0x4030 0000 000F <zero:1> <zero:4> <crm:4> <opc1:4> <zero:3> |
749cf76c | 2087 | |
c27581ed | 2088 | ARM CCSIDR registers are demultiplexed by CSSELR value: |
aa404ddf | 2089 | 0x4020 0000 0011 00 <csselr:8> |
749cf76c | 2090 | |
4fe21e4c | 2091 | ARM 32-bit VFP control registers have the following id bit patterns: |
aa404ddf | 2092 | 0x4020 0000 0012 1 <regno:12> |
4fe21e4c RR |
2093 | |
2094 | ARM 64-bit FP registers have the following id bit patterns: | |
aa404ddf | 2095 | 0x4030 0000 0012 0 <regno:12> |
4fe21e4c | 2096 | |
85bd0ba1 MZ |
2097 | ARM firmware pseudo-registers have the following bit pattern: |
2098 | 0x4030 0000 0014 <regno:16> | |
2099 | ||
379e04c7 MZ |
2100 | |
2101 | arm64 registers are mapped using the lower 32 bits. The upper 16 of | |
2102 | that is the register group type, or coprocessor number: | |
2103 | ||
2104 | arm64 core/FP-SIMD registers have the following id bit patterns. Note | |
2105 | that the size of the access is variable, as the kvm_regs structure | |
2106 | contains elements ranging from 32 to 128 bits. The index is a 32bit | |
2107 | value in the kvm_regs structure seen as a 32bit array. | |
2108 | 0x60x0 0000 0010 <index into the kvm_regs struct:16> | |
2109 | ||
2110 | arm64 CCSIDR registers are demultiplexed by CSSELR value: | |
2111 | 0x6020 0000 0011 00 <csselr:8> | |
2112 | ||
2113 | arm64 system registers have the following id bit patterns: | |
2114 | 0x6030 0000 0013 <op0:2> <op1:3> <crn:4> <crm:4> <op2:3> | |
2115 | ||
85bd0ba1 MZ |
2116 | arm64 firmware pseudo-registers have the following bit pattern: |
2117 | 0x6030 0000 0014 <regno:16> | |
2118 | ||
c2d2c21b JH |
2119 | |
2120 | MIPS registers are mapped using the lower 32 bits. The upper 16 of that is | |
2121 | the register group type: | |
2122 | ||
2123 | MIPS core registers (see above) have the following id bit patterns: | |
2124 | 0x7030 0000 0000 <reg:16> | |
2125 | ||
2126 | MIPS CP0 registers (see KVM_REG_MIPS_CP0_* above) have the following id bit | |
2127 | patterns depending on whether they're 32-bit or 64-bit registers: | |
2128 | 0x7020 0000 0001 00 <reg:5> <sel:3> (32-bit) | |
2129 | 0x7030 0000 0001 00 <reg:5> <sel:3> (64-bit) | |
2130 | ||
013044cc JH |
2131 | Note: KVM_REG_MIPS_CP0_ENTRYLO0 and KVM_REG_MIPS_CP0_ENTRYLO1 are the MIPS64 |
2132 | versions of the EntryLo registers regardless of the word size of the host | |
2133 | hardware, host kernel, guest, and whether XPA is present in the guest, i.e. | |
2134 | with the RI and XI bits (if they exist) in bits 63 and 62 respectively, and | |
2135 | the PFNX field starting at bit 30. | |
2136 | ||
d42a008f JH |
2137 | MIPS MAARs (see KVM_REG_MIPS_CP0_MAAR(*) above) have the following id bit |
2138 | patterns: | |
2139 | 0x7030 0000 0001 01 <reg:8> | |
2140 | ||
c2d2c21b JH |
2141 | MIPS KVM control registers (see above) have the following id bit patterns: |
2142 | 0x7030 0000 0002 <reg:16> | |
2143 | ||
379245cd JH |
2144 | MIPS FPU registers (see KVM_REG_MIPS_FPR_{32,64}() above) have the following |
2145 | id bit patterns depending on the size of the register being accessed. They are | |
2146 | always accessed according to the current guest FPU mode (Status.FR and | |
2147 | Config5.FRE), i.e. as the guest would see them, and they become unpredictable | |
ab86bd60 JH |
2148 | if the guest FPU mode is changed. MIPS SIMD Architecture (MSA) vector |
2149 | registers (see KVM_REG_MIPS_VEC_128() above) have similar patterns as they | |
2150 | overlap the FPU registers: | |
379245cd JH |
2151 | 0x7020 0000 0003 00 <0:3> <reg:5> (32-bit FPU registers) |
2152 | 0x7030 0000 0003 00 <0:3> <reg:5> (64-bit FPU registers) | |
ab86bd60 | 2153 | 0x7040 0000 0003 00 <0:3> <reg:5> (128-bit MSA vector registers) |
379245cd JH |
2154 | |
2155 | MIPS FPU control registers (see KVM_REG_MIPS_FCR_{IR,CSR} above) have the | |
2156 | following id bit patterns: | |
2157 | 0x7020 0000 0003 01 <0:3> <reg:5> | |
2158 | ||
ab86bd60 JH |
2159 | MIPS MSA control registers (see KVM_REG_MIPS_MSA_{IR,CSR} above) have the |
2160 | following id bit patterns: | |
2161 | 0x7020 0000 0003 02 <0:3> <reg:5> | |
2162 | ||
c2d2c21b | 2163 | |
e24ed81f AG |
2164 | 4.69 KVM_GET_ONE_REG |
2165 | ||
2166 | Capability: KVM_CAP_ONE_REG | |
2167 | Architectures: all | |
2168 | Type: vcpu ioctl | |
2169 | Parameters: struct kvm_one_reg (in and out) | |
2170 | Returns: 0 on success, negative value on failure | |
2171 | ||
2172 | This ioctl allows to receive the value of a single register implemented | |
2173 | in a vcpu. The register to read is indicated by the "id" field of the | |
2174 | kvm_one_reg struct passed in. On success, the register value can be found | |
2175 | at the memory location pointed to by "addr". | |
2176 | ||
2177 | The list of registers accessible using this interface is identical to the | |
2e232702 | 2178 | list in 4.68. |
e24ed81f | 2179 | |
414fa985 | 2180 | |
1c0b28c2 EM |
2181 | 4.70 KVM_KVMCLOCK_CTRL |
2182 | ||
2183 | Capability: KVM_CAP_KVMCLOCK_CTRL | |
2184 | Architectures: Any that implement pvclocks (currently x86 only) | |
2185 | Type: vcpu ioctl | |
2186 | Parameters: None | |
2187 | Returns: 0 on success, -1 on error | |
2188 | ||
2189 | This signals to the host kernel that the specified guest is being paused by | |
2190 | userspace. The host will set a flag in the pvclock structure that is checked | |
2191 | from the soft lockup watchdog. The flag is part of the pvclock structure that | |
2192 | is shared between guest and host, specifically the second bit of the flags | |
2193 | field of the pvclock_vcpu_time_info structure. It will be set exclusively by | |
2194 | the host and read/cleared exclusively by the guest. The guest operation of | |
2195 | checking and clearing the flag must an atomic operation so | |
2196 | load-link/store-conditional, or equivalent must be used. There are two cases | |
2197 | where the guest will clear the flag: when the soft lockup watchdog timer resets | |
2198 | itself or when a soft lockup is detected. This ioctl can be called any time | |
2199 | after pausing the vcpu, but before it is resumed. | |
2200 | ||
414fa985 | 2201 | |
07975ad3 JK |
2202 | 4.71 KVM_SIGNAL_MSI |
2203 | ||
2204 | Capability: KVM_CAP_SIGNAL_MSI | |
2988509d | 2205 | Architectures: x86 arm arm64 |
07975ad3 JK |
2206 | Type: vm ioctl |
2207 | Parameters: struct kvm_msi (in) | |
2208 | Returns: >0 on delivery, 0 if guest blocked the MSI, and -1 on error | |
2209 | ||
2210 | Directly inject a MSI message. Only valid with in-kernel irqchip that handles | |
2211 | MSI messages. | |
2212 | ||
2213 | struct kvm_msi { | |
2214 | __u32 address_lo; | |
2215 | __u32 address_hi; | |
2216 | __u32 data; | |
2217 | __u32 flags; | |
2b8ddd93 AP |
2218 | __u32 devid; |
2219 | __u8 pad[12]; | |
07975ad3 JK |
2220 | }; |
2221 | ||
6f49b2f3 PB |
2222 | flags: KVM_MSI_VALID_DEVID: devid contains a valid value. The per-VM |
2223 | KVM_CAP_MSI_DEVID capability advertises the requirement to provide | |
2224 | the device ID. If this capability is not available, userspace | |
2225 | should never set the KVM_MSI_VALID_DEVID flag as the ioctl might fail. | |
2b8ddd93 | 2226 | |
6f49b2f3 PB |
2227 | If KVM_MSI_VALID_DEVID is set, devid contains a unique device identifier |
2228 | for the device that wrote the MSI message. For PCI, this is usually a | |
2229 | BFD identifier in the lower 16 bits. | |
07975ad3 | 2230 | |
055b6ae9 PB |
2231 | On x86, address_hi is ignored unless the KVM_X2APIC_API_USE_32BIT_IDS |
2232 | feature of KVM_CAP_X2APIC_API capability is enabled. If it is enabled, | |
2233 | address_hi bits 31-8 provide bits 31-8 of the destination id. Bits 7-0 of | |
2234 | address_hi must be zero. | |
37131313 | 2235 | |
414fa985 | 2236 | |
0589ff6c JK |
2237 | 4.71 KVM_CREATE_PIT2 |
2238 | ||
2239 | Capability: KVM_CAP_PIT2 | |
2240 | Architectures: x86 | |
2241 | Type: vm ioctl | |
2242 | Parameters: struct kvm_pit_config (in) | |
2243 | Returns: 0 on success, -1 on error | |
2244 | ||
2245 | Creates an in-kernel device model for the i8254 PIT. This call is only valid | |
2246 | after enabling in-kernel irqchip support via KVM_CREATE_IRQCHIP. The following | |
2247 | parameters have to be passed: | |
2248 | ||
2249 | struct kvm_pit_config { | |
2250 | __u32 flags; | |
2251 | __u32 pad[15]; | |
2252 | }; | |
2253 | ||
2254 | Valid flags are: | |
2255 | ||
2256 | #define KVM_PIT_SPEAKER_DUMMY 1 /* emulate speaker port stub */ | |
2257 | ||
b6ddf05f JK |
2258 | PIT timer interrupts may use a per-VM kernel thread for injection. If it |
2259 | exists, this thread will have a name of the following pattern: | |
2260 | ||
2261 | kvm-pit/<owner-process-pid> | |
2262 | ||
2263 | When running a guest with elevated priorities, the scheduling parameters of | |
2264 | this thread may have to be adjusted accordingly. | |
2265 | ||
0589ff6c JK |
2266 | This IOCTL replaces the obsolete KVM_CREATE_PIT. |
2267 | ||
2268 | ||
2269 | 4.72 KVM_GET_PIT2 | |
2270 | ||
2271 | Capability: KVM_CAP_PIT_STATE2 | |
2272 | Architectures: x86 | |
2273 | Type: vm ioctl | |
2274 | Parameters: struct kvm_pit_state2 (out) | |
2275 | Returns: 0 on success, -1 on error | |
2276 | ||
2277 | Retrieves the state of the in-kernel PIT model. Only valid after | |
2278 | KVM_CREATE_PIT2. The state is returned in the following structure: | |
2279 | ||
2280 | struct kvm_pit_state2 { | |
2281 | struct kvm_pit_channel_state channels[3]; | |
2282 | __u32 flags; | |
2283 | __u32 reserved[9]; | |
2284 | }; | |
2285 | ||
2286 | Valid flags are: | |
2287 | ||
2288 | /* disable PIT in HPET legacy mode */ | |
2289 | #define KVM_PIT_FLAGS_HPET_LEGACY 0x00000001 | |
2290 | ||
2291 | This IOCTL replaces the obsolete KVM_GET_PIT. | |
2292 | ||
2293 | ||
2294 | 4.73 KVM_SET_PIT2 | |
2295 | ||
2296 | Capability: KVM_CAP_PIT_STATE2 | |
2297 | Architectures: x86 | |
2298 | Type: vm ioctl | |
2299 | Parameters: struct kvm_pit_state2 (in) | |
2300 | Returns: 0 on success, -1 on error | |
2301 | ||
2302 | Sets the state of the in-kernel PIT model. Only valid after KVM_CREATE_PIT2. | |
2303 | See KVM_GET_PIT2 for details on struct kvm_pit_state2. | |
2304 | ||
2305 | This IOCTL replaces the obsolete KVM_SET_PIT. | |
2306 | ||
2307 | ||
5b74716e BH |
2308 | 4.74 KVM_PPC_GET_SMMU_INFO |
2309 | ||
2310 | Capability: KVM_CAP_PPC_GET_SMMU_INFO | |
2311 | Architectures: powerpc | |
2312 | Type: vm ioctl | |
2313 | Parameters: None | |
2314 | Returns: 0 on success, -1 on error | |
2315 | ||
2316 | This populates and returns a structure describing the features of | |
2317 | the "Server" class MMU emulation supported by KVM. | |
cc22c354 | 2318 | This can in turn be used by userspace to generate the appropriate |
5b74716e BH |
2319 | device-tree properties for the guest operating system. |
2320 | ||
c98be0c9 | 2321 | The structure contains some global information, followed by an |
5b74716e BH |
2322 | array of supported segment page sizes: |
2323 | ||
2324 | struct kvm_ppc_smmu_info { | |
2325 | __u64 flags; | |
2326 | __u32 slb_size; | |
2327 | __u32 pad; | |
2328 | struct kvm_ppc_one_seg_page_size sps[KVM_PPC_PAGE_SIZES_MAX_SZ]; | |
2329 | }; | |
2330 | ||
2331 | The supported flags are: | |
2332 | ||
2333 | - KVM_PPC_PAGE_SIZES_REAL: | |
2334 | When that flag is set, guest page sizes must "fit" the backing | |
2335 | store page sizes. When not set, any page size in the list can | |
2336 | be used regardless of how they are backed by userspace. | |
2337 | ||
2338 | - KVM_PPC_1T_SEGMENTS | |
2339 | The emulated MMU supports 1T segments in addition to the | |
2340 | standard 256M ones. | |
2341 | ||
901f8c3f PM |
2342 | - KVM_PPC_NO_HASH |
2343 | This flag indicates that HPT guests are not supported by KVM, | |
2344 | thus all guests must use radix MMU mode. | |
2345 | ||
5b74716e BH |
2346 | The "slb_size" field indicates how many SLB entries are supported |
2347 | ||
2348 | The "sps" array contains 8 entries indicating the supported base | |
2349 | page sizes for a segment in increasing order. Each entry is defined | |
2350 | as follow: | |
2351 | ||
2352 | struct kvm_ppc_one_seg_page_size { | |
2353 | __u32 page_shift; /* Base page shift of segment (or 0) */ | |
2354 | __u32 slb_enc; /* SLB encoding for BookS */ | |
2355 | struct kvm_ppc_one_page_size enc[KVM_PPC_PAGE_SIZES_MAX_SZ]; | |
2356 | }; | |
2357 | ||
2358 | An entry with a "page_shift" of 0 is unused. Because the array is | |
2359 | organized in increasing order, a lookup can stop when encoutering | |
2360 | such an entry. | |
2361 | ||
2362 | The "slb_enc" field provides the encoding to use in the SLB for the | |
2363 | page size. The bits are in positions such as the value can directly | |
2364 | be OR'ed into the "vsid" argument of the slbmte instruction. | |
2365 | ||
2366 | The "enc" array is a list which for each of those segment base page | |
2367 | size provides the list of supported actual page sizes (which can be | |
2368 | only larger or equal to the base page size), along with the | |
f884ab15 | 2369 | corresponding encoding in the hash PTE. Similarly, the array is |
5b74716e BH |
2370 | 8 entries sorted by increasing sizes and an entry with a "0" shift |
2371 | is an empty entry and a terminator: | |
2372 | ||
2373 | struct kvm_ppc_one_page_size { | |
2374 | __u32 page_shift; /* Page shift (or 0) */ | |
2375 | __u32 pte_enc; /* Encoding in the HPTE (>>12) */ | |
2376 | }; | |
2377 | ||
2378 | The "pte_enc" field provides a value that can OR'ed into the hash | |
2379 | PTE's RPN field (ie, it needs to be shifted left by 12 to OR it | |
2380 | into the hash PTE second double word). | |
2381 | ||
f36992e3 AW |
2382 | 4.75 KVM_IRQFD |
2383 | ||
2384 | Capability: KVM_CAP_IRQFD | |
174178fe | 2385 | Architectures: x86 s390 arm arm64 |
f36992e3 AW |
2386 | Type: vm ioctl |
2387 | Parameters: struct kvm_irqfd (in) | |
2388 | Returns: 0 on success, -1 on error | |
2389 | ||
2390 | Allows setting an eventfd to directly trigger a guest interrupt. | |
2391 | kvm_irqfd.fd specifies the file descriptor to use as the eventfd and | |
2392 | kvm_irqfd.gsi specifies the irqchip pin toggled by this event. When | |
17180032 | 2393 | an event is triggered on the eventfd, an interrupt is injected into |
f36992e3 AW |
2394 | the guest using the specified gsi pin. The irqfd is removed using |
2395 | the KVM_IRQFD_FLAG_DEASSIGN flag, specifying both kvm_irqfd.fd | |
2396 | and kvm_irqfd.gsi. | |
2397 | ||
7a84428a AW |
2398 | With KVM_CAP_IRQFD_RESAMPLE, KVM_IRQFD supports a de-assert and notify |
2399 | mechanism allowing emulation of level-triggered, irqfd-based | |
2400 | interrupts. When KVM_IRQFD_FLAG_RESAMPLE is set the user must pass an | |
2401 | additional eventfd in the kvm_irqfd.resamplefd field. When operating | |
2402 | in resample mode, posting of an interrupt through kvm_irq.fd asserts | |
2403 | the specified gsi in the irqchip. When the irqchip is resampled, such | |
17180032 | 2404 | as from an EOI, the gsi is de-asserted and the user is notified via |
7a84428a AW |
2405 | kvm_irqfd.resamplefd. It is the user's responsibility to re-queue |
2406 | the interrupt if the device making use of it still requires service. | |
2407 | Note that closing the resamplefd is not sufficient to disable the | |
2408 | irqfd. The KVM_IRQFD_FLAG_RESAMPLE is only necessary on assignment | |
2409 | and need not be specified with KVM_IRQFD_FLAG_DEASSIGN. | |
2410 | ||
180ae7b1 EA |
2411 | On arm/arm64, gsi routing being supported, the following can happen: |
2412 | - in case no routing entry is associated to this gsi, injection fails | |
2413 | - in case the gsi is associated to an irqchip routing entry, | |
2414 | irqchip.pin + 32 corresponds to the injected SPI ID. | |
995a0ee9 EA |
2415 | - in case the gsi is associated to an MSI routing entry, the MSI |
2416 | message and device ID are translated into an LPI (support restricted | |
2417 | to GICv3 ITS in-kernel emulation). | |
174178fe | 2418 | |
5fecc9d8 | 2419 | 4.76 KVM_PPC_ALLOCATE_HTAB |
32fad281 PM |
2420 | |
2421 | Capability: KVM_CAP_PPC_ALLOC_HTAB | |
2422 | Architectures: powerpc | |
2423 | Type: vm ioctl | |
2424 | Parameters: Pointer to u32 containing hash table order (in/out) | |
2425 | Returns: 0 on success, -1 on error | |
2426 | ||
2427 | This requests the host kernel to allocate an MMU hash table for a | |
2428 | guest using the PAPR paravirtualization interface. This only does | |
2429 | anything if the kernel is configured to use the Book 3S HV style of | |
2430 | virtualization. Otherwise the capability doesn't exist and the ioctl | |
2431 | returns an ENOTTY error. The rest of this description assumes Book 3S | |
2432 | HV. | |
2433 | ||
2434 | There must be no vcpus running when this ioctl is called; if there | |
2435 | are, it will do nothing and return an EBUSY error. | |
2436 | ||
2437 | The parameter is a pointer to a 32-bit unsigned integer variable | |
2438 | containing the order (log base 2) of the desired size of the hash | |
2439 | table, which must be between 18 and 46. On successful return from the | |
f98a8bf9 | 2440 | ioctl, the value will not be changed by the kernel. |
32fad281 PM |
2441 | |
2442 | If no hash table has been allocated when any vcpu is asked to run | |
2443 | (with the KVM_RUN ioctl), the host kernel will allocate a | |
2444 | default-sized hash table (16 MB). | |
2445 | ||
2446 | If this ioctl is called when a hash table has already been allocated, | |
f98a8bf9 DG |
2447 | with a different order from the existing hash table, the existing hash |
2448 | table will be freed and a new one allocated. If this is ioctl is | |
2449 | called when a hash table has already been allocated of the same order | |
2450 | as specified, the kernel will clear out the existing hash table (zero | |
2451 | all HPTEs). In either case, if the guest is using the virtualized | |
2452 | real-mode area (VRMA) facility, the kernel will re-create the VMRA | |
2453 | HPTEs on the next KVM_RUN of any vcpu. | |
32fad281 | 2454 | |
416ad65f CH |
2455 | 4.77 KVM_S390_INTERRUPT |
2456 | ||
2457 | Capability: basic | |
2458 | Architectures: s390 | |
2459 | Type: vm ioctl, vcpu ioctl | |
2460 | Parameters: struct kvm_s390_interrupt (in) | |
2461 | Returns: 0 on success, -1 on error | |
2462 | ||
2463 | Allows to inject an interrupt to the guest. Interrupts can be floating | |
2464 | (vm ioctl) or per cpu (vcpu ioctl), depending on the interrupt type. | |
2465 | ||
2466 | Interrupt parameters are passed via kvm_s390_interrupt: | |
2467 | ||
2468 | struct kvm_s390_interrupt { | |
2469 | __u32 type; | |
2470 | __u32 parm; | |
2471 | __u64 parm64; | |
2472 | }; | |
2473 | ||
2474 | type can be one of the following: | |
2475 | ||
2822545f | 2476 | KVM_S390_SIGP_STOP (vcpu) - sigp stop; optional flags in parm |
416ad65f CH |
2477 | KVM_S390_PROGRAM_INT (vcpu) - program check; code in parm |
2478 | KVM_S390_SIGP_SET_PREFIX (vcpu) - sigp set prefix; prefix address in parm | |
2479 | KVM_S390_RESTART (vcpu) - restart | |
e029ae5b TH |
2480 | KVM_S390_INT_CLOCK_COMP (vcpu) - clock comparator interrupt |
2481 | KVM_S390_INT_CPU_TIMER (vcpu) - CPU timer interrupt | |
416ad65f CH |
2482 | KVM_S390_INT_VIRTIO (vm) - virtio external interrupt; external interrupt |
2483 | parameters in parm and parm64 | |
2484 | KVM_S390_INT_SERVICE (vm) - sclp external interrupt; sclp parameter in parm | |
2485 | KVM_S390_INT_EMERGENCY (vcpu) - sigp emergency; source cpu in parm | |
2486 | KVM_S390_INT_EXTERNAL_CALL (vcpu) - sigp external call; source cpu in parm | |
d8346b7d CH |
2487 | KVM_S390_INT_IO(ai,cssid,ssid,schid) (vm) - compound value to indicate an |
2488 | I/O interrupt (ai - adapter interrupt; cssid,ssid,schid - subchannel); | |
2489 | I/O interruption parameters in parm (subchannel) and parm64 (intparm, | |
2490 | interruption subclass) | |
48a3e950 CH |
2491 | KVM_S390_MCHK (vm, vcpu) - machine check interrupt; cr 14 bits in parm, |
2492 | machine check interrupt code in parm64 (note that | |
2493 | machine checks needing further payload are not | |
2494 | supported by this ioctl) | |
416ad65f CH |
2495 | |
2496 | Note that the vcpu ioctl is asynchronous to vcpu execution. | |
2497 | ||
a2932923 PM |
2498 | 4.78 KVM_PPC_GET_HTAB_FD |
2499 | ||
2500 | Capability: KVM_CAP_PPC_HTAB_FD | |
2501 | Architectures: powerpc | |
2502 | Type: vm ioctl | |
2503 | Parameters: Pointer to struct kvm_get_htab_fd (in) | |
2504 | Returns: file descriptor number (>= 0) on success, -1 on error | |
2505 | ||
2506 | This returns a file descriptor that can be used either to read out the | |
2507 | entries in the guest's hashed page table (HPT), or to write entries to | |
2508 | initialize the HPT. The returned fd can only be written to if the | |
2509 | KVM_GET_HTAB_WRITE bit is set in the flags field of the argument, and | |
2510 | can only be read if that bit is clear. The argument struct looks like | |
2511 | this: | |
2512 | ||
2513 | /* For KVM_PPC_GET_HTAB_FD */ | |
2514 | struct kvm_get_htab_fd { | |
2515 | __u64 flags; | |
2516 | __u64 start_index; | |
2517 | __u64 reserved[2]; | |
2518 | }; | |
2519 | ||
2520 | /* Values for kvm_get_htab_fd.flags */ | |
2521 | #define KVM_GET_HTAB_BOLTED_ONLY ((__u64)0x1) | |
2522 | #define KVM_GET_HTAB_WRITE ((__u64)0x2) | |
2523 | ||
2524 | The `start_index' field gives the index in the HPT of the entry at | |
2525 | which to start reading. It is ignored when writing. | |
2526 | ||
2527 | Reads on the fd will initially supply information about all | |
2528 | "interesting" HPT entries. Interesting entries are those with the | |
2529 | bolted bit set, if the KVM_GET_HTAB_BOLTED_ONLY bit is set, otherwise | |
2530 | all entries. When the end of the HPT is reached, the read() will | |
2531 | return. If read() is called again on the fd, it will start again from | |
2532 | the beginning of the HPT, but will only return HPT entries that have | |
2533 | changed since they were last read. | |
2534 | ||
2535 | Data read or written is structured as a header (8 bytes) followed by a | |
2536 | series of valid HPT entries (16 bytes) each. The header indicates how | |
2537 | many valid HPT entries there are and how many invalid entries follow | |
2538 | the valid entries. The invalid entries are not represented explicitly | |
2539 | in the stream. The header format is: | |
2540 | ||
2541 | struct kvm_get_htab_header { | |
2542 | __u32 index; | |
2543 | __u16 n_valid; | |
2544 | __u16 n_invalid; | |
2545 | }; | |
2546 | ||
2547 | Writes to the fd create HPT entries starting at the index given in the | |
2548 | header; first `n_valid' valid entries with contents from the data | |
2549 | written, then `n_invalid' invalid entries, invalidating any previously | |
2550 | valid entries found. | |
2551 | ||
852b6d57 SW |
2552 | 4.79 KVM_CREATE_DEVICE |
2553 | ||
2554 | Capability: KVM_CAP_DEVICE_CTRL | |
2555 | Type: vm ioctl | |
2556 | Parameters: struct kvm_create_device (in/out) | |
2557 | Returns: 0 on success, -1 on error | |
2558 | Errors: | |
2559 | ENODEV: The device type is unknown or unsupported | |
2560 | EEXIST: Device already created, and this type of device may not | |
2561 | be instantiated multiple times | |
2562 | ||
2563 | Other error conditions may be defined by individual device types or | |
2564 | have their standard meanings. | |
2565 | ||
2566 | Creates an emulated device in the kernel. The file descriptor returned | |
2567 | in fd can be used with KVM_SET/GET/HAS_DEVICE_ATTR. | |
2568 | ||
2569 | If the KVM_CREATE_DEVICE_TEST flag is set, only test whether the | |
2570 | device type is supported (not necessarily whether it can be created | |
2571 | in the current vm). | |
2572 | ||
2573 | Individual devices should not define flags. Attributes should be used | |
2574 | for specifying any behavior that is not implied by the device type | |
2575 | number. | |
2576 | ||
2577 | struct kvm_create_device { | |
2578 | __u32 type; /* in: KVM_DEV_TYPE_xxx */ | |
2579 | __u32 fd; /* out: device handle */ | |
2580 | __u32 flags; /* in: KVM_CREATE_DEVICE_xxx */ | |
2581 | }; | |
2582 | ||
2583 | 4.80 KVM_SET_DEVICE_ATTR/KVM_GET_DEVICE_ATTR | |
2584 | ||
f577f6c2 SZ |
2585 | Capability: KVM_CAP_DEVICE_CTRL, KVM_CAP_VM_ATTRIBUTES for vm device, |
2586 | KVM_CAP_VCPU_ATTRIBUTES for vcpu device | |
2587 | Type: device ioctl, vm ioctl, vcpu ioctl | |
852b6d57 SW |
2588 | Parameters: struct kvm_device_attr |
2589 | Returns: 0 on success, -1 on error | |
2590 | Errors: | |
2591 | ENXIO: The group or attribute is unknown/unsupported for this device | |
f9cbd9b0 | 2592 | or hardware support is missing. |
852b6d57 SW |
2593 | EPERM: The attribute cannot (currently) be accessed this way |
2594 | (e.g. read-only attribute, or attribute that only makes | |
2595 | sense when the device is in a different state) | |
2596 | ||
2597 | Other error conditions may be defined by individual device types. | |
2598 | ||
2599 | Gets/sets a specified piece of device configuration and/or state. The | |
2600 | semantics are device-specific. See individual device documentation in | |
2601 | the "devices" directory. As with ONE_REG, the size of the data | |
2602 | transferred is defined by the particular attribute. | |
2603 | ||
2604 | struct kvm_device_attr { | |
2605 | __u32 flags; /* no flags currently defined */ | |
2606 | __u32 group; /* device-defined */ | |
2607 | __u64 attr; /* group-defined */ | |
2608 | __u64 addr; /* userspace address of attr data */ | |
2609 | }; | |
2610 | ||
2611 | 4.81 KVM_HAS_DEVICE_ATTR | |
2612 | ||
f577f6c2 SZ |
2613 | Capability: KVM_CAP_DEVICE_CTRL, KVM_CAP_VM_ATTRIBUTES for vm device, |
2614 | KVM_CAP_VCPU_ATTRIBUTES for vcpu device | |
2615 | Type: device ioctl, vm ioctl, vcpu ioctl | |
852b6d57 SW |
2616 | Parameters: struct kvm_device_attr |
2617 | Returns: 0 on success, -1 on error | |
2618 | Errors: | |
2619 | ENXIO: The group or attribute is unknown/unsupported for this device | |
f9cbd9b0 | 2620 | or hardware support is missing. |
852b6d57 SW |
2621 | |
2622 | Tests whether a device supports a particular attribute. A successful | |
2623 | return indicates the attribute is implemented. It does not necessarily | |
2624 | indicate that the attribute can be read or written in the device's | |
2625 | current state. "addr" is ignored. | |
f36992e3 | 2626 | |
d8968f1f | 2627 | 4.82 KVM_ARM_VCPU_INIT |
749cf76c CD |
2628 | |
2629 | Capability: basic | |
379e04c7 | 2630 | Architectures: arm, arm64 |
749cf76c | 2631 | Type: vcpu ioctl |
beb11fc7 | 2632 | Parameters: struct kvm_vcpu_init (in) |
749cf76c CD |
2633 | Returns: 0 on success; -1 on error |
2634 | Errors: | |
2635 | EINVAL: the target is unknown, or the combination of features is invalid. | |
2636 | ENOENT: a features bit specified is unknown. | |
2637 | ||
2638 | This tells KVM what type of CPU to present to the guest, and what | |
2639 | optional features it should have. This will cause a reset of the cpu | |
2640 | registers to their initial values. If this is not called, KVM_RUN will | |
2641 | return ENOEXEC for that vcpu. | |
2642 | ||
2643 | Note that because some registers reflect machine topology, all vcpus | |
2644 | should be created before this ioctl is invoked. | |
2645 | ||
f7fa034d CD |
2646 | Userspace can call this function multiple times for a given vcpu, including |
2647 | after the vcpu has been run. This will reset the vcpu to its initial | |
2648 | state. All calls to this function after the initial call must use the same | |
2649 | target and same set of feature flags, otherwise EINVAL will be returned. | |
2650 | ||
aa024c2f MZ |
2651 | Possible features: |
2652 | - KVM_ARM_VCPU_POWER_OFF: Starts the CPU in a power-off state. | |
3ad8b3de CD |
2653 | Depends on KVM_CAP_ARM_PSCI. If not set, the CPU will be powered on |
2654 | and execute guest code when KVM_RUN is called. | |
379e04c7 MZ |
2655 | - KVM_ARM_VCPU_EL1_32BIT: Starts the CPU in a 32bit mode. |
2656 | Depends on KVM_CAP_ARM_EL1_32BIT (arm64 only). | |
85bd0ba1 MZ |
2657 | - KVM_ARM_VCPU_PSCI_0_2: Emulate PSCI v0.2 (or a future revision |
2658 | backward compatible with v0.2) for the CPU. | |
50bb0c94 | 2659 | Depends on KVM_CAP_ARM_PSCI_0_2. |
808e7381 SZ |
2660 | - KVM_ARM_VCPU_PMU_V3: Emulate PMUv3 for the CPU. |
2661 | Depends on KVM_CAP_ARM_PMU_V3. | |
aa024c2f | 2662 | |
749cf76c | 2663 | |
740edfc0 AP |
2664 | 4.83 KVM_ARM_PREFERRED_TARGET |
2665 | ||
2666 | Capability: basic | |
2667 | Architectures: arm, arm64 | |
2668 | Type: vm ioctl | |
2669 | Parameters: struct struct kvm_vcpu_init (out) | |
2670 | Returns: 0 on success; -1 on error | |
2671 | Errors: | |
a7265fb1 | 2672 | ENODEV: no preferred target available for the host |
740edfc0 AP |
2673 | |
2674 | This queries KVM for preferred CPU target type which can be emulated | |
2675 | by KVM on underlying host. | |
2676 | ||
2677 | The ioctl returns struct kvm_vcpu_init instance containing information | |
2678 | about preferred CPU target type and recommended features for it. The | |
2679 | kvm_vcpu_init->features bitmap returned will have feature bits set if | |
2680 | the preferred target recommends setting these features, but this is | |
2681 | not mandatory. | |
2682 | ||
2683 | The information returned by this ioctl can be used to prepare an instance | |
2684 | of struct kvm_vcpu_init for KVM_ARM_VCPU_INIT ioctl which will result in | |
2685 | in VCPU matching underlying host. | |
2686 | ||
2687 | ||
2688 | 4.84 KVM_GET_REG_LIST | |
749cf76c CD |
2689 | |
2690 | Capability: basic | |
c2d2c21b | 2691 | Architectures: arm, arm64, mips |
749cf76c CD |
2692 | Type: vcpu ioctl |
2693 | Parameters: struct kvm_reg_list (in/out) | |
2694 | Returns: 0 on success; -1 on error | |
2695 | Errors: | |
2696 | E2BIG: the reg index list is too big to fit in the array specified by | |
2697 | the user (the number required will be written into n). | |
2698 | ||
2699 | struct kvm_reg_list { | |
2700 | __u64 n; /* number of registers in reg[] */ | |
2701 | __u64 reg[0]; | |
2702 | }; | |
2703 | ||
2704 | This ioctl returns the guest registers that are supported for the | |
2705 | KVM_GET_ONE_REG/KVM_SET_ONE_REG calls. | |
2706 | ||
ce01e4e8 CD |
2707 | |
2708 | 4.85 KVM_ARM_SET_DEVICE_ADDR (deprecated) | |
3401d546 CD |
2709 | |
2710 | Capability: KVM_CAP_ARM_SET_DEVICE_ADDR | |
379e04c7 | 2711 | Architectures: arm, arm64 |
3401d546 CD |
2712 | Type: vm ioctl |
2713 | Parameters: struct kvm_arm_device_address (in) | |
2714 | Returns: 0 on success, -1 on error | |
2715 | Errors: | |
2716 | ENODEV: The device id is unknown | |
2717 | ENXIO: Device not supported on current system | |
2718 | EEXIST: Address already set | |
2719 | E2BIG: Address outside guest physical address space | |
330690cd | 2720 | EBUSY: Address overlaps with other device range |
3401d546 CD |
2721 | |
2722 | struct kvm_arm_device_addr { | |
2723 | __u64 id; | |
2724 | __u64 addr; | |
2725 | }; | |
2726 | ||
2727 | Specify a device address in the guest's physical address space where guests | |
2728 | can access emulated or directly exposed devices, which the host kernel needs | |
2729 | to know about. The id field is an architecture specific identifier for a | |
2730 | specific device. | |
2731 | ||
379e04c7 MZ |
2732 | ARM/arm64 divides the id field into two parts, a device id and an |
2733 | address type id specific to the individual device. | |
3401d546 CD |
2734 | |
2735 | bits: | 63 ... 32 | 31 ... 16 | 15 ... 0 | | |
2736 | field: | 0x00000000 | device id | addr type id | | |
2737 | ||
379e04c7 MZ |
2738 | ARM/arm64 currently only require this when using the in-kernel GIC |
2739 | support for the hardware VGIC features, using KVM_ARM_DEVICE_VGIC_V2 | |
2740 | as the device id. When setting the base address for the guest's | |
2741 | mapping of the VGIC virtual CPU and distributor interface, the ioctl | |
2742 | must be called after calling KVM_CREATE_IRQCHIP, but before calling | |
2743 | KVM_RUN on any of the VCPUs. Calling this ioctl twice for any of the | |
2744 | base addresses will return -EEXIST. | |
3401d546 | 2745 | |
ce01e4e8 CD |
2746 | Note, this IOCTL is deprecated and the more flexible SET/GET_DEVICE_ATTR API |
2747 | should be used instead. | |
2748 | ||
2749 | ||
740edfc0 | 2750 | 4.86 KVM_PPC_RTAS_DEFINE_TOKEN |
8e591cb7 ME |
2751 | |
2752 | Capability: KVM_CAP_PPC_RTAS | |
2753 | Architectures: ppc | |
2754 | Type: vm ioctl | |
2755 | Parameters: struct kvm_rtas_token_args | |
2756 | Returns: 0 on success, -1 on error | |
2757 | ||
2758 | Defines a token value for a RTAS (Run Time Abstraction Services) | |
2759 | service in order to allow it to be handled in the kernel. The | |
2760 | argument struct gives the name of the service, which must be the name | |
2761 | of a service that has a kernel-side implementation. If the token | |
2762 | value is non-zero, it will be associated with that service, and | |
2763 | subsequent RTAS calls by the guest specifying that token will be | |
2764 | handled by the kernel. If the token value is 0, then any token | |
2765 | associated with the service will be forgotten, and subsequent RTAS | |
2766 | calls by the guest for that service will be passed to userspace to be | |
2767 | handled. | |
2768 | ||
4bd9d344 AB |
2769 | 4.87 KVM_SET_GUEST_DEBUG |
2770 | ||
2771 | Capability: KVM_CAP_SET_GUEST_DEBUG | |
0e6f07f2 | 2772 | Architectures: x86, s390, ppc, arm64 |
4bd9d344 AB |
2773 | Type: vcpu ioctl |
2774 | Parameters: struct kvm_guest_debug (in) | |
2775 | Returns: 0 on success; -1 on error | |
2776 | ||
2777 | struct kvm_guest_debug { | |
2778 | __u32 control; | |
2779 | __u32 pad; | |
2780 | struct kvm_guest_debug_arch arch; | |
2781 | }; | |
2782 | ||
2783 | Set up the processor specific debug registers and configure vcpu for | |
2784 | handling guest debug events. There are two parts to the structure, the | |
2785 | first a control bitfield indicates the type of debug events to handle | |
2786 | when running. Common control bits are: | |
2787 | ||
2788 | - KVM_GUESTDBG_ENABLE: guest debugging is enabled | |
2789 | - KVM_GUESTDBG_SINGLESTEP: the next run should single-step | |
2790 | ||
2791 | The top 16 bits of the control field are architecture specific control | |
2792 | flags which can include the following: | |
2793 | ||
4bd611ca | 2794 | - KVM_GUESTDBG_USE_SW_BP: using software breakpoints [x86, arm64] |
834bf887 | 2795 | - KVM_GUESTDBG_USE_HW_BP: using hardware breakpoints [x86, s390, arm64] |
4bd9d344 AB |
2796 | - KVM_GUESTDBG_INJECT_DB: inject DB type exception [x86] |
2797 | - KVM_GUESTDBG_INJECT_BP: inject BP type exception [x86] | |
2798 | - KVM_GUESTDBG_EXIT_PENDING: trigger an immediate guest exit [s390] | |
2799 | ||
2800 | For example KVM_GUESTDBG_USE_SW_BP indicates that software breakpoints | |
2801 | are enabled in memory so we need to ensure breakpoint exceptions are | |
2802 | correctly trapped and the KVM run loop exits at the breakpoint and not | |
2803 | running off into the normal guest vector. For KVM_GUESTDBG_USE_HW_BP | |
2804 | we need to ensure the guest vCPUs architecture specific registers are | |
2805 | updated to the correct (supplied) values. | |
2806 | ||
2807 | The second part of the structure is architecture specific and | |
2808 | typically contains a set of debug registers. | |
2809 | ||
834bf887 AB |
2810 | For arm64 the number of debug registers is implementation defined and |
2811 | can be determined by querying the KVM_CAP_GUEST_DEBUG_HW_BPS and | |
2812 | KVM_CAP_GUEST_DEBUG_HW_WPS capabilities which return a positive number | |
2813 | indicating the number of supported registers. | |
2814 | ||
4bd9d344 AB |
2815 | When debug events exit the main run loop with the reason |
2816 | KVM_EXIT_DEBUG with the kvm_debug_exit_arch part of the kvm_run | |
2817 | structure containing architecture specific debug information. | |
3401d546 | 2818 | |
209cf19f AB |
2819 | 4.88 KVM_GET_EMULATED_CPUID |
2820 | ||
2821 | Capability: KVM_CAP_EXT_EMUL_CPUID | |
2822 | Architectures: x86 | |
2823 | Type: system ioctl | |
2824 | Parameters: struct kvm_cpuid2 (in/out) | |
2825 | Returns: 0 on success, -1 on error | |
2826 | ||
2827 | struct kvm_cpuid2 { | |
2828 | __u32 nent; | |
2829 | __u32 flags; | |
2830 | struct kvm_cpuid_entry2 entries[0]; | |
2831 | }; | |
2832 | ||
2833 | The member 'flags' is used for passing flags from userspace. | |
2834 | ||
2835 | #define KVM_CPUID_FLAG_SIGNIFCANT_INDEX BIT(0) | |
2836 | #define KVM_CPUID_FLAG_STATEFUL_FUNC BIT(1) | |
2837 | #define KVM_CPUID_FLAG_STATE_READ_NEXT BIT(2) | |
2838 | ||
2839 | struct kvm_cpuid_entry2 { | |
2840 | __u32 function; | |
2841 | __u32 index; | |
2842 | __u32 flags; | |
2843 | __u32 eax; | |
2844 | __u32 ebx; | |
2845 | __u32 ecx; | |
2846 | __u32 edx; | |
2847 | __u32 padding[3]; | |
2848 | }; | |
2849 | ||
2850 | This ioctl returns x86 cpuid features which are emulated by | |
2851 | kvm.Userspace can use the information returned by this ioctl to query | |
2852 | which features are emulated by kvm instead of being present natively. | |
2853 | ||
2854 | Userspace invokes KVM_GET_EMULATED_CPUID by passing a kvm_cpuid2 | |
2855 | structure with the 'nent' field indicating the number of entries in | |
2856 | the variable-size array 'entries'. If the number of entries is too low | |
2857 | to describe the cpu capabilities, an error (E2BIG) is returned. If the | |
2858 | number is too high, the 'nent' field is adjusted and an error (ENOMEM) | |
2859 | is returned. If the number is just right, the 'nent' field is adjusted | |
2860 | to the number of valid entries in the 'entries' array, which is then | |
2861 | filled. | |
2862 | ||
2863 | The entries returned are the set CPUID bits of the respective features | |
2864 | which kvm emulates, as returned by the CPUID instruction, with unknown | |
2865 | or unsupported feature bits cleared. | |
2866 | ||
2867 | Features like x2apic, for example, may not be present in the host cpu | |
2868 | but are exposed by kvm in KVM_GET_SUPPORTED_CPUID because they can be | |
2869 | emulated efficiently and thus not included here. | |
2870 | ||
2871 | The fields in each entry are defined as follows: | |
2872 | ||
2873 | function: the eax value used to obtain the entry | |
2874 | index: the ecx value used to obtain the entry (for entries that are | |
2875 | affected by ecx) | |
2876 | flags: an OR of zero or more of the following: | |
2877 | KVM_CPUID_FLAG_SIGNIFCANT_INDEX: | |
2878 | if the index field is valid | |
2879 | KVM_CPUID_FLAG_STATEFUL_FUNC: | |
2880 | if cpuid for this function returns different values for successive | |
2881 | invocations; there will be several entries with the same function, | |
2882 | all with this flag set | |
2883 | KVM_CPUID_FLAG_STATE_READ_NEXT: | |
2884 | for KVM_CPUID_FLAG_STATEFUL_FUNC entries, set if this entry is | |
2885 | the first entry to be read by a cpu | |
2886 | eax, ebx, ecx, edx: the values returned by the cpuid instruction for | |
2887 | this function/index combination | |
2888 | ||
41408c28 TH |
2889 | 4.89 KVM_S390_MEM_OP |
2890 | ||
2891 | Capability: KVM_CAP_S390_MEM_OP | |
2892 | Architectures: s390 | |
2893 | Type: vcpu ioctl | |
2894 | Parameters: struct kvm_s390_mem_op (in) | |
2895 | Returns: = 0 on success, | |
2896 | < 0 on generic error (e.g. -EFAULT or -ENOMEM), | |
2897 | > 0 if an exception occurred while walking the page tables | |
2898 | ||
5d4f6f3d | 2899 | Read or write data from/to the logical (virtual) memory of a VCPU. |
41408c28 TH |
2900 | |
2901 | Parameters are specified via the following structure: | |
2902 | ||
2903 | struct kvm_s390_mem_op { | |
2904 | __u64 gaddr; /* the guest address */ | |
2905 | __u64 flags; /* flags */ | |
2906 | __u32 size; /* amount of bytes */ | |
2907 | __u32 op; /* type of operation */ | |
2908 | __u64 buf; /* buffer in userspace */ | |
2909 | __u8 ar; /* the access register number */ | |
2910 | __u8 reserved[31]; /* should be set to 0 */ | |
2911 | }; | |
2912 | ||
2913 | The type of operation is specified in the "op" field. It is either | |
2914 | KVM_S390_MEMOP_LOGICAL_READ for reading from logical memory space or | |
2915 | KVM_S390_MEMOP_LOGICAL_WRITE for writing to logical memory space. The | |
2916 | KVM_S390_MEMOP_F_CHECK_ONLY flag can be set in the "flags" field to check | |
2917 | whether the corresponding memory access would create an access exception | |
2918 | (without touching the data in the memory at the destination). In case an | |
2919 | access exception occurred while walking the MMU tables of the guest, the | |
2920 | ioctl returns a positive error number to indicate the type of exception. | |
2921 | This exception is also raised directly at the corresponding VCPU if the | |
2922 | flag KVM_S390_MEMOP_F_INJECT_EXCEPTION is set in the "flags" field. | |
2923 | ||
2924 | The start address of the memory region has to be specified in the "gaddr" | |
2925 | field, and the length of the region in the "size" field. "buf" is the buffer | |
2926 | supplied by the userspace application where the read data should be written | |
2927 | to for KVM_S390_MEMOP_LOGICAL_READ, or where the data that should be written | |
2928 | is stored for a KVM_S390_MEMOP_LOGICAL_WRITE. "buf" is unused and can be NULL | |
2929 | when KVM_S390_MEMOP_F_CHECK_ONLY is specified. "ar" designates the access | |
2930 | register number to be used. | |
2931 | ||
2932 | The "reserved" field is meant for future extensions. It is not used by | |
2933 | KVM with the currently defined set of flags. | |
2934 | ||
30ee2a98 JH |
2935 | 4.90 KVM_S390_GET_SKEYS |
2936 | ||
2937 | Capability: KVM_CAP_S390_SKEYS | |
2938 | Architectures: s390 | |
2939 | Type: vm ioctl | |
2940 | Parameters: struct kvm_s390_skeys | |
2941 | Returns: 0 on success, KVM_S390_GET_KEYS_NONE if guest is not using storage | |
2942 | keys, negative value on error | |
2943 | ||
2944 | This ioctl is used to get guest storage key values on the s390 | |
2945 | architecture. The ioctl takes parameters via the kvm_s390_skeys struct. | |
2946 | ||
2947 | struct kvm_s390_skeys { | |
2948 | __u64 start_gfn; | |
2949 | __u64 count; | |
2950 | __u64 skeydata_addr; | |
2951 | __u32 flags; | |
2952 | __u32 reserved[9]; | |
2953 | }; | |
2954 | ||
2955 | The start_gfn field is the number of the first guest frame whose storage keys | |
2956 | you want to get. | |
2957 | ||
2958 | The count field is the number of consecutive frames (starting from start_gfn) | |
2959 | whose storage keys to get. The count field must be at least 1 and the maximum | |
2960 | allowed value is defined as KVM_S390_SKEYS_ALLOC_MAX. Values outside this range | |
2961 | will cause the ioctl to return -EINVAL. | |
2962 | ||
2963 | The skeydata_addr field is the address to a buffer large enough to hold count | |
2964 | bytes. This buffer will be filled with storage key data by the ioctl. | |
2965 | ||
2966 | 4.91 KVM_S390_SET_SKEYS | |
2967 | ||
2968 | Capability: KVM_CAP_S390_SKEYS | |
2969 | Architectures: s390 | |
2970 | Type: vm ioctl | |
2971 | Parameters: struct kvm_s390_skeys | |
2972 | Returns: 0 on success, negative value on error | |
2973 | ||
2974 | This ioctl is used to set guest storage key values on the s390 | |
2975 | architecture. The ioctl takes parameters via the kvm_s390_skeys struct. | |
2976 | See section on KVM_S390_GET_SKEYS for struct definition. | |
2977 | ||
2978 | The start_gfn field is the number of the first guest frame whose storage keys | |
2979 | you want to set. | |
2980 | ||
2981 | The count field is the number of consecutive frames (starting from start_gfn) | |
2982 | whose storage keys to get. The count field must be at least 1 and the maximum | |
2983 | allowed value is defined as KVM_S390_SKEYS_ALLOC_MAX. Values outside this range | |
2984 | will cause the ioctl to return -EINVAL. | |
2985 | ||
2986 | The skeydata_addr field is the address to a buffer containing count bytes of | |
2987 | storage keys. Each byte in the buffer will be set as the storage key for a | |
2988 | single frame starting at start_gfn for count frames. | |
2989 | ||
2990 | Note: If any architecturally invalid key value is found in the given data then | |
2991 | the ioctl will return -EINVAL. | |
2992 | ||
47b43c52 JF |
2993 | 4.92 KVM_S390_IRQ |
2994 | ||
2995 | Capability: KVM_CAP_S390_INJECT_IRQ | |
2996 | Architectures: s390 | |
2997 | Type: vcpu ioctl | |
2998 | Parameters: struct kvm_s390_irq (in) | |
2999 | Returns: 0 on success, -1 on error | |
3000 | Errors: | |
3001 | EINVAL: interrupt type is invalid | |
3002 | type is KVM_S390_SIGP_STOP and flag parameter is invalid value | |
3003 | type is KVM_S390_INT_EXTERNAL_CALL and code is bigger | |
3004 | than the maximum of VCPUs | |
3005 | EBUSY: type is KVM_S390_SIGP_SET_PREFIX and vcpu is not stopped | |
3006 | type is KVM_S390_SIGP_STOP and a stop irq is already pending | |
3007 | type is KVM_S390_INT_EXTERNAL_CALL and an external call interrupt | |
3008 | is already pending | |
3009 | ||
3010 | Allows to inject an interrupt to the guest. | |
3011 | ||
3012 | Using struct kvm_s390_irq as a parameter allows | |
3013 | to inject additional payload which is not | |
3014 | possible via KVM_S390_INTERRUPT. | |
3015 | ||
3016 | Interrupt parameters are passed via kvm_s390_irq: | |
3017 | ||
3018 | struct kvm_s390_irq { | |
3019 | __u64 type; | |
3020 | union { | |
3021 | struct kvm_s390_io_info io; | |
3022 | struct kvm_s390_ext_info ext; | |
3023 | struct kvm_s390_pgm_info pgm; | |
3024 | struct kvm_s390_emerg_info emerg; | |
3025 | struct kvm_s390_extcall_info extcall; | |
3026 | struct kvm_s390_prefix_info prefix; | |
3027 | struct kvm_s390_stop_info stop; | |
3028 | struct kvm_s390_mchk_info mchk; | |
3029 | char reserved[64]; | |
3030 | } u; | |
3031 | }; | |
3032 | ||
3033 | type can be one of the following: | |
3034 | ||
3035 | KVM_S390_SIGP_STOP - sigp stop; parameter in .stop | |
3036 | KVM_S390_PROGRAM_INT - program check; parameters in .pgm | |
3037 | KVM_S390_SIGP_SET_PREFIX - sigp set prefix; parameters in .prefix | |
3038 | KVM_S390_RESTART - restart; no parameters | |
3039 | KVM_S390_INT_CLOCK_COMP - clock comparator interrupt; no parameters | |
3040 | KVM_S390_INT_CPU_TIMER - CPU timer interrupt; no parameters | |
3041 | KVM_S390_INT_EMERGENCY - sigp emergency; parameters in .emerg | |
3042 | KVM_S390_INT_EXTERNAL_CALL - sigp external call; parameters in .extcall | |
3043 | KVM_S390_MCHK - machine check interrupt; parameters in .mchk | |
3044 | ||
3045 | ||
3046 | Note that the vcpu ioctl is asynchronous to vcpu execution. | |
3047 | ||
816c7667 JF |
3048 | 4.94 KVM_S390_GET_IRQ_STATE |
3049 | ||
3050 | Capability: KVM_CAP_S390_IRQ_STATE | |
3051 | Architectures: s390 | |
3052 | Type: vcpu ioctl | |
3053 | Parameters: struct kvm_s390_irq_state (out) | |
3054 | Returns: >= number of bytes copied into buffer, | |
3055 | -EINVAL if buffer size is 0, | |
3056 | -ENOBUFS if buffer size is too small to fit all pending interrupts, | |
3057 | -EFAULT if the buffer address was invalid | |
3058 | ||
3059 | This ioctl allows userspace to retrieve the complete state of all currently | |
3060 | pending interrupts in a single buffer. Use cases include migration | |
3061 | and introspection. The parameter structure contains the address of a | |
3062 | userspace buffer and its length: | |
3063 | ||
3064 | struct kvm_s390_irq_state { | |
3065 | __u64 buf; | |
bb64da9a | 3066 | __u32 flags; /* will stay unused for compatibility reasons */ |
816c7667 | 3067 | __u32 len; |
bb64da9a | 3068 | __u32 reserved[4]; /* will stay unused for compatibility reasons */ |
816c7667 JF |
3069 | }; |
3070 | ||
3071 | Userspace passes in the above struct and for each pending interrupt a | |
3072 | struct kvm_s390_irq is copied to the provided buffer. | |
3073 | ||
bb64da9a CB |
3074 | The structure contains a flags and a reserved field for future extensions. As |
3075 | the kernel never checked for flags == 0 and QEMU never pre-zeroed flags and | |
3076 | reserved, these fields can not be used in the future without breaking | |
3077 | compatibility. | |
3078 | ||
816c7667 JF |
3079 | If -ENOBUFS is returned the buffer provided was too small and userspace |
3080 | may retry with a bigger buffer. | |
3081 | ||
3082 | 4.95 KVM_S390_SET_IRQ_STATE | |
3083 | ||
3084 | Capability: KVM_CAP_S390_IRQ_STATE | |
3085 | Architectures: s390 | |
3086 | Type: vcpu ioctl | |
3087 | Parameters: struct kvm_s390_irq_state (in) | |
3088 | Returns: 0 on success, | |
3089 | -EFAULT if the buffer address was invalid, | |
3090 | -EINVAL for an invalid buffer length (see below), | |
3091 | -EBUSY if there were already interrupts pending, | |
3092 | errors occurring when actually injecting the | |
3093 | interrupt. See KVM_S390_IRQ. | |
3094 | ||
3095 | This ioctl allows userspace to set the complete state of all cpu-local | |
3096 | interrupts currently pending for the vcpu. It is intended for restoring | |
3097 | interrupt state after a migration. The input parameter is a userspace buffer | |
3098 | containing a struct kvm_s390_irq_state: | |
3099 | ||
3100 | struct kvm_s390_irq_state { | |
3101 | __u64 buf; | |
bb64da9a | 3102 | __u32 flags; /* will stay unused for compatibility reasons */ |
816c7667 | 3103 | __u32 len; |
bb64da9a | 3104 | __u32 reserved[4]; /* will stay unused for compatibility reasons */ |
816c7667 JF |
3105 | }; |
3106 | ||
bb64da9a CB |
3107 | The restrictions for flags and reserved apply as well. |
3108 | (see KVM_S390_GET_IRQ_STATE) | |
3109 | ||
816c7667 JF |
3110 | The userspace memory referenced by buf contains a struct kvm_s390_irq |
3111 | for each interrupt to be injected into the guest. | |
3112 | If one of the interrupts could not be injected for some reason the | |
3113 | ioctl aborts. | |
3114 | ||
3115 | len must be a multiple of sizeof(struct kvm_s390_irq). It must be > 0 | |
3116 | and it must not exceed (max_vcpus + 32) * sizeof(struct kvm_s390_irq), | |
3117 | which is the maximum number of possibly pending cpu-local interrupts. | |
47b43c52 | 3118 | |
ed8e5a24 | 3119 | 4.96 KVM_SMI |
f077825a PB |
3120 | |
3121 | Capability: KVM_CAP_X86_SMM | |
3122 | Architectures: x86 | |
3123 | Type: vcpu ioctl | |
3124 | Parameters: none | |
3125 | Returns: 0 on success, -1 on error | |
3126 | ||
3127 | Queues an SMI on the thread's vcpu. | |
3128 | ||
d3695aa4 AK |
3129 | 4.97 KVM_CAP_PPC_MULTITCE |
3130 | ||
3131 | Capability: KVM_CAP_PPC_MULTITCE | |
3132 | Architectures: ppc | |
3133 | Type: vm | |
3134 | ||
3135 | This capability means the kernel is capable of handling hypercalls | |
3136 | H_PUT_TCE_INDIRECT and H_STUFF_TCE without passing those into the user | |
3137 | space. This significantly accelerates DMA operations for PPC KVM guests. | |
3138 | User space should expect that its handlers for these hypercalls | |
3139 | are not going to be called if user space previously registered LIOBN | |
3140 | in KVM (via KVM_CREATE_SPAPR_TCE or similar calls). | |
3141 | ||
3142 | In order to enable H_PUT_TCE_INDIRECT and H_STUFF_TCE use in the guest, | |
3143 | user space might have to advertise it for the guest. For example, | |
3144 | IBM pSeries (sPAPR) guest starts using them if "hcall-multi-tce" is | |
3145 | present in the "ibm,hypertas-functions" device-tree property. | |
3146 | ||
3147 | The hypercalls mentioned above may or may not be processed successfully | |
3148 | in the kernel based fast path. If they can not be handled by the kernel, | |
3149 | they will get passed on to user space. So user space still has to have | |
3150 | an implementation for these despite the in kernel acceleration. | |
3151 | ||
3152 | This capability is always enabled. | |
3153 | ||
58ded420 AK |
3154 | 4.98 KVM_CREATE_SPAPR_TCE_64 |
3155 | ||
3156 | Capability: KVM_CAP_SPAPR_TCE_64 | |
3157 | Architectures: powerpc | |
3158 | Type: vm ioctl | |
3159 | Parameters: struct kvm_create_spapr_tce_64 (in) | |
3160 | Returns: file descriptor for manipulating the created TCE table | |
3161 | ||
3162 | This is an extension for KVM_CAP_SPAPR_TCE which only supports 32bit | |
3163 | windows, described in 4.62 KVM_CREATE_SPAPR_TCE | |
3164 | ||
3165 | This capability uses extended struct in ioctl interface: | |
3166 | ||
3167 | /* for KVM_CAP_SPAPR_TCE_64 */ | |
3168 | struct kvm_create_spapr_tce_64 { | |
3169 | __u64 liobn; | |
3170 | __u32 page_shift; | |
3171 | __u32 flags; | |
3172 | __u64 offset; /* in pages */ | |
3173 | __u64 size; /* in pages */ | |
3174 | }; | |
3175 | ||
3176 | The aim of extension is to support an additional bigger DMA window with | |
3177 | a variable page size. | |
3178 | KVM_CREATE_SPAPR_TCE_64 receives a 64bit window size, an IOMMU page shift and | |
3179 | a bus offset of the corresponding DMA window, @size and @offset are numbers | |
3180 | of IOMMU pages. | |
3181 | ||
3182 | @flags are not used at the moment. | |
3183 | ||
3184 | The rest of functionality is identical to KVM_CREATE_SPAPR_TCE. | |
3185 | ||
ccc4df4e | 3186 | 4.99 KVM_REINJECT_CONTROL |
107d44a2 RK |
3187 | |
3188 | Capability: KVM_CAP_REINJECT_CONTROL | |
3189 | Architectures: x86 | |
3190 | Type: vm ioctl | |
3191 | Parameters: struct kvm_reinject_control (in) | |
3192 | Returns: 0 on success, | |
3193 | -EFAULT if struct kvm_reinject_control cannot be read, | |
3194 | -ENXIO if KVM_CREATE_PIT or KVM_CREATE_PIT2 didn't succeed earlier. | |
3195 | ||
3196 | i8254 (PIT) has two modes, reinject and !reinject. The default is reinject, | |
3197 | where KVM queues elapsed i8254 ticks and monitors completion of interrupt from | |
3198 | vector(s) that i8254 injects. Reinject mode dequeues a tick and injects its | |
3199 | interrupt whenever there isn't a pending interrupt from i8254. | |
3200 | !reinject mode injects an interrupt as soon as a tick arrives. | |
3201 | ||
3202 | struct kvm_reinject_control { | |
3203 | __u8 pit_reinject; | |
3204 | __u8 reserved[31]; | |
3205 | }; | |
3206 | ||
3207 | pit_reinject = 0 (!reinject mode) is recommended, unless running an old | |
3208 | operating system that uses the PIT for timing (e.g. Linux 2.4.x). | |
3209 | ||
ccc4df4e | 3210 | 4.100 KVM_PPC_CONFIGURE_V3_MMU |
c9270132 PM |
3211 | |
3212 | Capability: KVM_CAP_PPC_RADIX_MMU or KVM_CAP_PPC_HASH_MMU_V3 | |
3213 | Architectures: ppc | |
3214 | Type: vm ioctl | |
3215 | Parameters: struct kvm_ppc_mmuv3_cfg (in) | |
3216 | Returns: 0 on success, | |
3217 | -EFAULT if struct kvm_ppc_mmuv3_cfg cannot be read, | |
3218 | -EINVAL if the configuration is invalid | |
3219 | ||
3220 | This ioctl controls whether the guest will use radix or HPT (hashed | |
3221 | page table) translation, and sets the pointer to the process table for | |
3222 | the guest. | |
3223 | ||
3224 | struct kvm_ppc_mmuv3_cfg { | |
3225 | __u64 flags; | |
3226 | __u64 process_table; | |
3227 | }; | |
3228 | ||
3229 | There are two bits that can be set in flags; KVM_PPC_MMUV3_RADIX and | |
3230 | KVM_PPC_MMUV3_GTSE. KVM_PPC_MMUV3_RADIX, if set, configures the guest | |
3231 | to use radix tree translation, and if clear, to use HPT translation. | |
3232 | KVM_PPC_MMUV3_GTSE, if set and if KVM permits it, configures the guest | |
3233 | to be able to use the global TLB and SLB invalidation instructions; | |
3234 | if clear, the guest may not use these instructions. | |
3235 | ||
3236 | The process_table field specifies the address and size of the guest | |
3237 | process table, which is in the guest's space. This field is formatted | |
3238 | as the second doubleword of the partition table entry, as defined in | |
3239 | the Power ISA V3.00, Book III section 5.7.6.1. | |
3240 | ||
ccc4df4e | 3241 | 4.101 KVM_PPC_GET_RMMU_INFO |
c9270132 PM |
3242 | |
3243 | Capability: KVM_CAP_PPC_RADIX_MMU | |
3244 | Architectures: ppc | |
3245 | Type: vm ioctl | |
3246 | Parameters: struct kvm_ppc_rmmu_info (out) | |
3247 | Returns: 0 on success, | |
3248 | -EFAULT if struct kvm_ppc_rmmu_info cannot be written, | |
3249 | -EINVAL if no useful information can be returned | |
3250 | ||
3251 | This ioctl returns a structure containing two things: (a) a list | |
3252 | containing supported radix tree geometries, and (b) a list that maps | |
3253 | page sizes to put in the "AP" (actual page size) field for the tlbie | |
3254 | (TLB invalidate entry) instruction. | |
3255 | ||
3256 | struct kvm_ppc_rmmu_info { | |
3257 | struct kvm_ppc_radix_geom { | |
3258 | __u8 page_shift; | |
3259 | __u8 level_bits[4]; | |
3260 | __u8 pad[3]; | |
3261 | } geometries[8]; | |
3262 | __u32 ap_encodings[8]; | |
3263 | }; | |
3264 | ||
3265 | The geometries[] field gives up to 8 supported geometries for the | |
3266 | radix page table, in terms of the log base 2 of the smallest page | |
3267 | size, and the number of bits indexed at each level of the tree, from | |
3268 | the PTE level up to the PGD level in that order. Any unused entries | |
3269 | will have 0 in the page_shift field. | |
3270 | ||
3271 | The ap_encodings gives the supported page sizes and their AP field | |
3272 | encodings, encoded with the AP value in the top 3 bits and the log | |
3273 | base 2 of the page size in the bottom 6 bits. | |
3274 | ||
ef1ead0c DG |
3275 | 4.102 KVM_PPC_RESIZE_HPT_PREPARE |
3276 | ||
3277 | Capability: KVM_CAP_SPAPR_RESIZE_HPT | |
3278 | Architectures: powerpc | |
3279 | Type: vm ioctl | |
3280 | Parameters: struct kvm_ppc_resize_hpt (in) | |
3281 | Returns: 0 on successful completion, | |
3282 | >0 if a new HPT is being prepared, the value is an estimated | |
3283 | number of milliseconds until preparation is complete | |
3284 | -EFAULT if struct kvm_reinject_control cannot be read, | |
3285 | -EINVAL if the supplied shift or flags are invalid | |
3286 | -ENOMEM if unable to allocate the new HPT | |
3287 | -ENOSPC if there was a hash collision when moving existing | |
3288 | HPT entries to the new HPT | |
3289 | -EIO on other error conditions | |
3290 | ||
3291 | Used to implement the PAPR extension for runtime resizing of a guest's | |
3292 | Hashed Page Table (HPT). Specifically this starts, stops or monitors | |
3293 | the preparation of a new potential HPT for the guest, essentially | |
3294 | implementing the H_RESIZE_HPT_PREPARE hypercall. | |
3295 | ||
3296 | If called with shift > 0 when there is no pending HPT for the guest, | |
3297 | this begins preparation of a new pending HPT of size 2^(shift) bytes. | |
3298 | It then returns a positive integer with the estimated number of | |
3299 | milliseconds until preparation is complete. | |
3300 | ||
3301 | If called when there is a pending HPT whose size does not match that | |
3302 | requested in the parameters, discards the existing pending HPT and | |
3303 | creates a new one as above. | |
3304 | ||
3305 | If called when there is a pending HPT of the size requested, will: | |
3306 | * If preparation of the pending HPT is already complete, return 0 | |
3307 | * If preparation of the pending HPT has failed, return an error | |
3308 | code, then discard the pending HPT. | |
3309 | * If preparation of the pending HPT is still in progress, return an | |
3310 | estimated number of milliseconds until preparation is complete. | |
3311 | ||
3312 | If called with shift == 0, discards any currently pending HPT and | |
3313 | returns 0 (i.e. cancels any in-progress preparation). | |
3314 | ||
3315 | flags is reserved for future expansion, currently setting any bits in | |
3316 | flags will result in an -EINVAL. | |
3317 | ||
3318 | Normally this will be called repeatedly with the same parameters until | |
3319 | it returns <= 0. The first call will initiate preparation, subsequent | |
3320 | ones will monitor preparation until it completes or fails. | |
3321 | ||
3322 | struct kvm_ppc_resize_hpt { | |
3323 | __u64 flags; | |
3324 | __u32 shift; | |
3325 | __u32 pad; | |
3326 | }; | |
3327 | ||
3328 | 4.103 KVM_PPC_RESIZE_HPT_COMMIT | |
3329 | ||
3330 | Capability: KVM_CAP_SPAPR_RESIZE_HPT | |
3331 | Architectures: powerpc | |
3332 | Type: vm ioctl | |
3333 | Parameters: struct kvm_ppc_resize_hpt (in) | |
3334 | Returns: 0 on successful completion, | |
3335 | -EFAULT if struct kvm_reinject_control cannot be read, | |
3336 | -EINVAL if the supplied shift or flags are invalid | |
3337 | -ENXIO is there is no pending HPT, or the pending HPT doesn't | |
3338 | have the requested size | |
3339 | -EBUSY if the pending HPT is not fully prepared | |
3340 | -ENOSPC if there was a hash collision when moving existing | |
3341 | HPT entries to the new HPT | |
3342 | -EIO on other error conditions | |
3343 | ||
3344 | Used to implement the PAPR extension for runtime resizing of a guest's | |
3345 | Hashed Page Table (HPT). Specifically this requests that the guest be | |
3346 | transferred to working with the new HPT, essentially implementing the | |
3347 | H_RESIZE_HPT_COMMIT hypercall. | |
3348 | ||
3349 | This should only be called after KVM_PPC_RESIZE_HPT_PREPARE has | |
3350 | returned 0 with the same parameters. In other cases | |
3351 | KVM_PPC_RESIZE_HPT_COMMIT will return an error (usually -ENXIO or | |
3352 | -EBUSY, though others may be possible if the preparation was started, | |
3353 | but failed). | |
3354 | ||
3355 | This will have undefined effects on the guest if it has not already | |
3356 | placed itself in a quiescent state where no vcpu will make MMU enabled | |
3357 | memory accesses. | |
3358 | ||
3359 | On succsful completion, the pending HPT will become the guest's active | |
3360 | HPT and the previous HPT will be discarded. | |
3361 | ||
3362 | On failure, the guest will still be operating on its previous HPT. | |
3363 | ||
3364 | struct kvm_ppc_resize_hpt { | |
3365 | __u64 flags; | |
3366 | __u32 shift; | |
3367 | __u32 pad; | |
3368 | }; | |
3369 | ||
3aa53859 LC |
3370 | 4.104 KVM_X86_GET_MCE_CAP_SUPPORTED |
3371 | ||
3372 | Capability: KVM_CAP_MCE | |
3373 | Architectures: x86 | |
3374 | Type: system ioctl | |
3375 | Parameters: u64 mce_cap (out) | |
3376 | Returns: 0 on success, -1 on error | |
3377 | ||
3378 | Returns supported MCE capabilities. The u64 mce_cap parameter | |
3379 | has the same format as the MSR_IA32_MCG_CAP register. Supported | |
3380 | capabilities will have the corresponding bits set. | |
3381 | ||
3382 | 4.105 KVM_X86_SETUP_MCE | |
3383 | ||
3384 | Capability: KVM_CAP_MCE | |
3385 | Architectures: x86 | |
3386 | Type: vcpu ioctl | |
3387 | Parameters: u64 mcg_cap (in) | |
3388 | Returns: 0 on success, | |
3389 | -EFAULT if u64 mcg_cap cannot be read, | |
3390 | -EINVAL if the requested number of banks is invalid, | |
3391 | -EINVAL if requested MCE capability is not supported. | |
3392 | ||
3393 | Initializes MCE support for use. The u64 mcg_cap parameter | |
3394 | has the same format as the MSR_IA32_MCG_CAP register and | |
3395 | specifies which capabilities should be enabled. The maximum | |
3396 | supported number of error-reporting banks can be retrieved when | |
3397 | checking for KVM_CAP_MCE. The supported capabilities can be | |
3398 | retrieved with KVM_X86_GET_MCE_CAP_SUPPORTED. | |
3399 | ||
3400 | 4.106 KVM_X86_SET_MCE | |
3401 | ||
3402 | Capability: KVM_CAP_MCE | |
3403 | Architectures: x86 | |
3404 | Type: vcpu ioctl | |
3405 | Parameters: struct kvm_x86_mce (in) | |
3406 | Returns: 0 on success, | |
3407 | -EFAULT if struct kvm_x86_mce cannot be read, | |
3408 | -EINVAL if the bank number is invalid, | |
3409 | -EINVAL if VAL bit is not set in status field. | |
3410 | ||
3411 | Inject a machine check error (MCE) into the guest. The input | |
3412 | parameter is: | |
3413 | ||
3414 | struct kvm_x86_mce { | |
3415 | __u64 status; | |
3416 | __u64 addr; | |
3417 | __u64 misc; | |
3418 | __u64 mcg_status; | |
3419 | __u8 bank; | |
3420 | __u8 pad1[7]; | |
3421 | __u64 pad2[3]; | |
3422 | }; | |
3423 | ||
3424 | If the MCE being reported is an uncorrected error, KVM will | |
3425 | inject it as an MCE exception into the guest. If the guest | |
3426 | MCG_STATUS register reports that an MCE is in progress, KVM | |
3427 | causes an KVM_EXIT_SHUTDOWN vmexit. | |
3428 | ||
3429 | Otherwise, if the MCE is a corrected error, KVM will just | |
3430 | store it in the corresponding bank (provided this bank is | |
3431 | not holding a previously reported uncorrected error). | |
3432 | ||
4036e387 CI |
3433 | 4.107 KVM_S390_GET_CMMA_BITS |
3434 | ||
3435 | Capability: KVM_CAP_S390_CMMA_MIGRATION | |
3436 | Architectures: s390 | |
3437 | Type: vm ioctl | |
3438 | Parameters: struct kvm_s390_cmma_log (in, out) | |
3439 | Returns: 0 on success, a negative value on error | |
3440 | ||
3441 | This ioctl is used to get the values of the CMMA bits on the s390 | |
3442 | architecture. It is meant to be used in two scenarios: | |
3443 | - During live migration to save the CMMA values. Live migration needs | |
3444 | to be enabled via the KVM_REQ_START_MIGRATION VM property. | |
3445 | - To non-destructively peek at the CMMA values, with the flag | |
3446 | KVM_S390_CMMA_PEEK set. | |
3447 | ||
3448 | The ioctl takes parameters via the kvm_s390_cmma_log struct. The desired | |
3449 | values are written to a buffer whose location is indicated via the "values" | |
3450 | member in the kvm_s390_cmma_log struct. The values in the input struct are | |
3451 | also updated as needed. | |
3452 | Each CMMA value takes up one byte. | |
3453 | ||
3454 | struct kvm_s390_cmma_log { | |
3455 | __u64 start_gfn; | |
3456 | __u32 count; | |
3457 | __u32 flags; | |
3458 | union { | |
3459 | __u64 remaining; | |
3460 | __u64 mask; | |
3461 | }; | |
3462 | __u64 values; | |
3463 | }; | |
3464 | ||
3465 | start_gfn is the number of the first guest frame whose CMMA values are | |
3466 | to be retrieved, | |
3467 | ||
3468 | count is the length of the buffer in bytes, | |
3469 | ||
3470 | values points to the buffer where the result will be written to. | |
3471 | ||
3472 | If count is greater than KVM_S390_SKEYS_MAX, then it is considered to be | |
3473 | KVM_S390_SKEYS_MAX. KVM_S390_SKEYS_MAX is re-used for consistency with | |
3474 | other ioctls. | |
3475 | ||
3476 | The result is written in the buffer pointed to by the field values, and | |
3477 | the values of the input parameter are updated as follows. | |
3478 | ||
3479 | Depending on the flags, different actions are performed. The only | |
3480 | supported flag so far is KVM_S390_CMMA_PEEK. | |
3481 | ||
3482 | The default behaviour if KVM_S390_CMMA_PEEK is not set is: | |
3483 | start_gfn will indicate the first page frame whose CMMA bits were dirty. | |
3484 | It is not necessarily the same as the one passed as input, as clean pages | |
3485 | are skipped. | |
3486 | ||
3487 | count will indicate the number of bytes actually written in the buffer. | |
3488 | It can (and very often will) be smaller than the input value, since the | |
3489 | buffer is only filled until 16 bytes of clean values are found (which | |
3490 | are then not copied in the buffer). Since a CMMA migration block needs | |
3491 | the base address and the length, for a total of 16 bytes, we will send | |
3492 | back some clean data if there is some dirty data afterwards, as long as | |
3493 | the size of the clean data does not exceed the size of the header. This | |
3494 | allows to minimize the amount of data to be saved or transferred over | |
3495 | the network at the expense of more roundtrips to userspace. The next | |
3496 | invocation of the ioctl will skip over all the clean values, saving | |
3497 | potentially more than just the 16 bytes we found. | |
3498 | ||
3499 | If KVM_S390_CMMA_PEEK is set: | |
3500 | the existing storage attributes are read even when not in migration | |
3501 | mode, and no other action is performed; | |
3502 | ||
3503 | the output start_gfn will be equal to the input start_gfn, | |
3504 | ||
3505 | the output count will be equal to the input count, except if the end of | |
3506 | memory has been reached. | |
3507 | ||
3508 | In both cases: | |
3509 | the field "remaining" will indicate the total number of dirty CMMA values | |
3510 | still remaining, or 0 if KVM_S390_CMMA_PEEK is set and migration mode is | |
3511 | not enabled. | |
3512 | ||
3513 | mask is unused. | |
3514 | ||
3515 | values points to the userspace buffer where the result will be stored. | |
3516 | ||
3517 | This ioctl can fail with -ENOMEM if not enough memory can be allocated to | |
3518 | complete the task, with -ENXIO if CMMA is not enabled, with -EINVAL if | |
3519 | KVM_S390_CMMA_PEEK is not set but migration mode was not enabled, with | |
3520 | -EFAULT if the userspace address is invalid or if no page table is | |
3521 | present for the addresses (e.g. when using hugepages). | |
3522 | ||
3523 | 4.108 KVM_S390_SET_CMMA_BITS | |
3524 | ||
3525 | Capability: KVM_CAP_S390_CMMA_MIGRATION | |
3526 | Architectures: s390 | |
3527 | Type: vm ioctl | |
3528 | Parameters: struct kvm_s390_cmma_log (in) | |
3529 | Returns: 0 on success, a negative value on error | |
3530 | ||
3531 | This ioctl is used to set the values of the CMMA bits on the s390 | |
3532 | architecture. It is meant to be used during live migration to restore | |
3533 | the CMMA values, but there are no restrictions on its use. | |
3534 | The ioctl takes parameters via the kvm_s390_cmma_values struct. | |
3535 | Each CMMA value takes up one byte. | |
3536 | ||
3537 | struct kvm_s390_cmma_log { | |
3538 | __u64 start_gfn; | |
3539 | __u32 count; | |
3540 | __u32 flags; | |
3541 | union { | |
3542 | __u64 remaining; | |
3543 | __u64 mask; | |
3544 | }; | |
3545 | __u64 values; | |
3546 | }; | |
3547 | ||
3548 | start_gfn indicates the starting guest frame number, | |
3549 | ||
3550 | count indicates how many values are to be considered in the buffer, | |
3551 | ||
3552 | flags is not used and must be 0. | |
3553 | ||
3554 | mask indicates which PGSTE bits are to be considered. | |
3555 | ||
3556 | remaining is not used. | |
3557 | ||
3558 | values points to the buffer in userspace where to store the values. | |
3559 | ||
3560 | This ioctl can fail with -ENOMEM if not enough memory can be allocated to | |
3561 | complete the task, with -ENXIO if CMMA is not enabled, with -EINVAL if | |
3562 | the count field is too large (e.g. more than KVM_S390_CMMA_SIZE_MAX) or | |
3563 | if the flags field was not 0, with -EFAULT if the userspace address is | |
3564 | invalid, if invalid pages are written to (e.g. after the end of memory) | |
3565 | or if no page table is present for the addresses (e.g. when using | |
3566 | hugepages). | |
3567 | ||
7bf14c28 | 3568 | 4.109 KVM_PPC_GET_CPU_CHAR |
3214d01f PM |
3569 | |
3570 | Capability: KVM_CAP_PPC_GET_CPU_CHAR | |
3571 | Architectures: powerpc | |
3572 | Type: vm ioctl | |
3573 | Parameters: struct kvm_ppc_cpu_char (out) | |
3574 | Returns: 0 on successful completion | |
3575 | -EFAULT if struct kvm_ppc_cpu_char cannot be written | |
3576 | ||
3577 | This ioctl gives userspace information about certain characteristics | |
3578 | of the CPU relating to speculative execution of instructions and | |
3579 | possible information leakage resulting from speculative execution (see | |
3580 | CVE-2017-5715, CVE-2017-5753 and CVE-2017-5754). The information is | |
3581 | returned in struct kvm_ppc_cpu_char, which looks like this: | |
3582 | ||
3583 | struct kvm_ppc_cpu_char { | |
3584 | __u64 character; /* characteristics of the CPU */ | |
3585 | __u64 behaviour; /* recommended software behaviour */ | |
3586 | __u64 character_mask; /* valid bits in character */ | |
3587 | __u64 behaviour_mask; /* valid bits in behaviour */ | |
3588 | }; | |
3589 | ||
3590 | For extensibility, the character_mask and behaviour_mask fields | |
3591 | indicate which bits of character and behaviour have been filled in by | |
3592 | the kernel. If the set of defined bits is extended in future then | |
3593 | userspace will be able to tell whether it is running on a kernel that | |
3594 | knows about the new bits. | |
3595 | ||
3596 | The character field describes attributes of the CPU which can help | |
3597 | with preventing inadvertent information disclosure - specifically, | |
3598 | whether there is an instruction to flash-invalidate the L1 data cache | |
3599 | (ori 30,30,0 or mtspr SPRN_TRIG2,rN), whether the L1 data cache is set | |
3600 | to a mode where entries can only be used by the thread that created | |
3601 | them, whether the bcctr[l] instruction prevents speculation, and | |
3602 | whether a speculation barrier instruction (ori 31,31,0) is provided. | |
3603 | ||
3604 | The behaviour field describes actions that software should take to | |
3605 | prevent inadvertent information disclosure, and thus describes which | |
3606 | vulnerabilities the hardware is subject to; specifically whether the | |
3607 | L1 data cache should be flushed when returning to user mode from the | |
3608 | kernel, and whether a speculation barrier should be placed between an | |
3609 | array bounds check and the array access. | |
3610 | ||
3611 | These fields use the same bit definitions as the new | |
3612 | H_GET_CPU_CHARACTERISTICS hypercall. | |
3613 | ||
7bf14c28 | 3614 | 4.110 KVM_MEMORY_ENCRYPT_OP |
5acc5c06 BS |
3615 | |
3616 | Capability: basic | |
3617 | Architectures: x86 | |
3618 | Type: system | |
3619 | Parameters: an opaque platform specific structure (in/out) | |
3620 | Returns: 0 on success; -1 on error | |
3621 | ||
3622 | If the platform supports creating encrypted VMs then this ioctl can be used | |
3623 | for issuing platform-specific memory encryption commands to manage those | |
3624 | encrypted VMs. | |
3625 | ||
3626 | Currently, this ioctl is used for issuing Secure Encrypted Virtualization | |
3627 | (SEV) commands on AMD Processors. The SEV commands are defined in | |
21e94aca | 3628 | Documentation/virtual/kvm/amd-memory-encryption.rst. |
5acc5c06 | 3629 | |
7bf14c28 | 3630 | 4.111 KVM_MEMORY_ENCRYPT_REG_REGION |
69eaedee BS |
3631 | |
3632 | Capability: basic | |
3633 | Architectures: x86 | |
3634 | Type: system | |
3635 | Parameters: struct kvm_enc_region (in) | |
3636 | Returns: 0 on success; -1 on error | |
3637 | ||
3638 | This ioctl can be used to register a guest memory region which may | |
3639 | contain encrypted data (e.g. guest RAM, SMRAM etc). | |
3640 | ||
3641 | It is used in the SEV-enabled guest. When encryption is enabled, a guest | |
3642 | memory region may contain encrypted data. The SEV memory encryption | |
3643 | engine uses a tweak such that two identical plaintext pages, each at | |
3644 | different locations will have differing ciphertexts. So swapping or | |
3645 | moving ciphertext of those pages will not result in plaintext being | |
3646 | swapped. So relocating (or migrating) physical backing pages for the SEV | |
3647 | guest will require some additional steps. | |
3648 | ||
3649 | Note: The current SEV key management spec does not provide commands to | |
3650 | swap or migrate (move) ciphertext pages. Hence, for now we pin the guest | |
3651 | memory region registered with the ioctl. | |
3652 | ||
7bf14c28 | 3653 | 4.112 KVM_MEMORY_ENCRYPT_UNREG_REGION |
69eaedee BS |
3654 | |
3655 | Capability: basic | |
3656 | Architectures: x86 | |
3657 | Type: system | |
3658 | Parameters: struct kvm_enc_region (in) | |
3659 | Returns: 0 on success; -1 on error | |
3660 | ||
3661 | This ioctl can be used to unregister the guest memory region registered | |
3662 | with KVM_MEMORY_ENCRYPT_REG_REGION ioctl above. | |
3663 | ||
faeb7833 RK |
3664 | 4.113 KVM_HYPERV_EVENTFD |
3665 | ||
3666 | Capability: KVM_CAP_HYPERV_EVENTFD | |
3667 | Architectures: x86 | |
3668 | Type: vm ioctl | |
3669 | Parameters: struct kvm_hyperv_eventfd (in) | |
3670 | ||
3671 | This ioctl (un)registers an eventfd to receive notifications from the guest on | |
3672 | the specified Hyper-V connection id through the SIGNAL_EVENT hypercall, without | |
3673 | causing a user exit. SIGNAL_EVENT hypercall with non-zero event flag number | |
3674 | (bits 24-31) still triggers a KVM_EXIT_HYPERV_HCALL user exit. | |
3675 | ||
3676 | struct kvm_hyperv_eventfd { | |
3677 | __u32 conn_id; | |
3678 | __s32 fd; | |
3679 | __u32 flags; | |
3680 | __u32 padding[3]; | |
3681 | }; | |
3682 | ||
3683 | The conn_id field should fit within 24 bits: | |
3684 | ||
3685 | #define KVM_HYPERV_CONN_ID_MASK 0x00ffffff | |
3686 | ||
3687 | The acceptable values for the flags field are: | |
3688 | ||
3689 | #define KVM_HYPERV_EVENTFD_DEASSIGN (1 << 0) | |
3690 | ||
3691 | Returns: 0 on success, | |
3692 | -EINVAL if conn_id or flags is outside the allowed range | |
3693 | -ENOENT on deassign if the conn_id isn't registered | |
3694 | -EEXIST on assign if the conn_id is already registered | |
3695 | ||
8fcc4b59 JM |
3696 | 4.114 KVM_GET_NESTED_STATE |
3697 | ||
3698 | Capability: KVM_CAP_NESTED_STATE | |
3699 | Architectures: x86 | |
3700 | Type: vcpu ioctl | |
3701 | Parameters: struct kvm_nested_state (in/out) | |
3702 | Returns: 0 on success, -1 on error | |
3703 | Errors: | |
3704 | E2BIG: the total state size (including the fixed-size part of struct | |
3705 | kvm_nested_state) exceeds the value of 'size' specified by | |
3706 | the user; the size required will be written into size. | |
3707 | ||
3708 | struct kvm_nested_state { | |
3709 | __u16 flags; | |
3710 | __u16 format; | |
3711 | __u32 size; | |
3712 | union { | |
3713 | struct kvm_vmx_nested_state vmx; | |
3714 | struct kvm_svm_nested_state svm; | |
3715 | __u8 pad[120]; | |
3716 | }; | |
3717 | __u8 data[0]; | |
3718 | }; | |
3719 | ||
3720 | #define KVM_STATE_NESTED_GUEST_MODE 0x00000001 | |
3721 | #define KVM_STATE_NESTED_RUN_PENDING 0x00000002 | |
3722 | ||
3723 | #define KVM_STATE_NESTED_SMM_GUEST_MODE 0x00000001 | |
3724 | #define KVM_STATE_NESTED_SMM_VMXON 0x00000002 | |
3725 | ||
3726 | struct kvm_vmx_nested_state { | |
3727 | __u64 vmxon_pa; | |
3728 | __u64 vmcs_pa; | |
3729 | ||
3730 | struct { | |
3731 | __u16 flags; | |
3732 | } smm; | |
3733 | }; | |
3734 | ||
3735 | This ioctl copies the vcpu's nested virtualization state from the kernel to | |
3736 | userspace. | |
3737 | ||
3738 | The maximum size of the state, including the fixed-size part of struct | |
3739 | kvm_nested_state, can be retrieved by passing KVM_CAP_NESTED_STATE to | |
3740 | the KVM_CHECK_EXTENSION ioctl(). | |
3741 | ||
3742 | 4.115 KVM_SET_NESTED_STATE | |
3743 | ||
3744 | Capability: KVM_CAP_NESTED_STATE | |
3745 | Architectures: x86 | |
3746 | Type: vcpu ioctl | |
3747 | Parameters: struct kvm_nested_state (in) | |
3748 | Returns: 0 on success, -1 on error | |
3749 | ||
3750 | This copies the vcpu's kvm_nested_state struct from userspace to the kernel. For | |
3751 | the definition of struct kvm_nested_state, see KVM_GET_NESTED_STATE. | |
7bf14c28 | 3752 | |
9943450b PH |
3753 | 4.116 KVM_(UN)REGISTER_COALESCED_MMIO |
3754 | ||
0804c849 PH |
3755 | Capability: KVM_CAP_COALESCED_MMIO (for coalesced mmio) |
3756 | KVM_CAP_COALESCED_PIO (for coalesced pio) | |
9943450b PH |
3757 | Architectures: all |
3758 | Type: vm ioctl | |
3759 | Parameters: struct kvm_coalesced_mmio_zone | |
3760 | Returns: 0 on success, < 0 on error | |
3761 | ||
0804c849 | 3762 | Coalesced I/O is a performance optimization that defers hardware |
9943450b PH |
3763 | register write emulation so that userspace exits are avoided. It is |
3764 | typically used to reduce the overhead of emulating frequently accessed | |
3765 | hardware registers. | |
3766 | ||
0804c849 | 3767 | When a hardware register is configured for coalesced I/O, write accesses |
9943450b PH |
3768 | do not exit to userspace and their value is recorded in a ring buffer |
3769 | that is shared between kernel and userspace. | |
3770 | ||
0804c849 | 3771 | Coalesced I/O is used if one or more write accesses to a hardware |
9943450b PH |
3772 | register can be deferred until a read or a write to another hardware |
3773 | register on the same device. This last access will cause a vmexit and | |
3774 | userspace will process accesses from the ring buffer before emulating | |
0804c849 PH |
3775 | it. That will avoid exiting to userspace on repeated writes. |
3776 | ||
3777 | Coalesced pio is based on coalesced mmio. There is little difference | |
3778 | between coalesced mmio and pio except that coalesced pio records accesses | |
3779 | to I/O ports. | |
9943450b | 3780 | |
2a31b9db PB |
3781 | 4.117 KVM_CLEAR_DIRTY_LOG (vm ioctl) |
3782 | ||
3783 | Capability: KVM_CAP_MANUAL_DIRTY_LOG_PROTECT | |
3784 | Architectures: x86 | |
3785 | Type: vm ioctl | |
3786 | Parameters: struct kvm_dirty_log (in) | |
3787 | Returns: 0 on success, -1 on error | |
3788 | ||
3789 | /* for KVM_CLEAR_DIRTY_LOG */ | |
3790 | struct kvm_clear_dirty_log { | |
3791 | __u32 slot; | |
3792 | __u32 num_pages; | |
3793 | __u64 first_page; | |
3794 | union { | |
3795 | void __user *dirty_bitmap; /* one bit per page */ | |
3796 | __u64 padding; | |
3797 | }; | |
3798 | }; | |
3799 | ||
3800 | The ioctl clears the dirty status of pages in a memory slot, according to | |
3801 | the bitmap that is passed in struct kvm_clear_dirty_log's dirty_bitmap | |
3802 | field. Bit 0 of the bitmap corresponds to page "first_page" in the | |
3803 | memory slot, and num_pages is the size in bits of the input bitmap. | |
3804 | Both first_page and num_pages must be a multiple of 64. For each bit | |
3805 | that is set in the input bitmap, the corresponding page is marked "clean" | |
3806 | in KVM's dirty bitmap, and dirty tracking is re-enabled for that page | |
3807 | (for example via write-protection, or by clearing the dirty bit in | |
3808 | a page table entry). | |
3809 | ||
3810 | If KVM_CAP_MULTI_ADDRESS_SPACE is available, bits 16-31 specifies | |
3811 | the address space for which you want to return the dirty bitmap. | |
3812 | They must be less than the value that KVM_CHECK_EXTENSION returns for | |
3813 | the KVM_CAP_MULTI_ADDRESS_SPACE capability. | |
3814 | ||
3815 | This ioctl is mostly useful when KVM_CAP_MANUAL_DIRTY_LOG_PROTECT | |
3816 | is enabled; for more information, see the description of the capability. | |
3817 | However, it can always be used as long as KVM_CHECK_EXTENSION confirms | |
3818 | that KVM_CAP_MANUAL_DIRTY_LOG_PROTECT is present. | |
3819 | ||
2bc39970 VK |
3820 | 4.118 KVM_GET_SUPPORTED_HV_CPUID |
3821 | ||
3822 | Capability: KVM_CAP_HYPERV_CPUID | |
3823 | Architectures: x86 | |
3824 | Type: vcpu ioctl | |
3825 | Parameters: struct kvm_cpuid2 (in/out) | |
3826 | Returns: 0 on success, -1 on error | |
3827 | ||
3828 | struct kvm_cpuid2 { | |
3829 | __u32 nent; | |
3830 | __u32 padding; | |
3831 | struct kvm_cpuid_entry2 entries[0]; | |
3832 | }; | |
3833 | ||
3834 | struct kvm_cpuid_entry2 { | |
3835 | __u32 function; | |
3836 | __u32 index; | |
3837 | __u32 flags; | |
3838 | __u32 eax; | |
3839 | __u32 ebx; | |
3840 | __u32 ecx; | |
3841 | __u32 edx; | |
3842 | __u32 padding[3]; | |
3843 | }; | |
3844 | ||
3845 | This ioctl returns x86 cpuid features leaves related to Hyper-V emulation in | |
3846 | KVM. Userspace can use the information returned by this ioctl to construct | |
3847 | cpuid information presented to guests consuming Hyper-V enlightenments (e.g. | |
3848 | Windows or Hyper-V guests). | |
3849 | ||
3850 | CPUID feature leaves returned by this ioctl are defined by Hyper-V Top Level | |
3851 | Functional Specification (TLFS). These leaves can't be obtained with | |
3852 | KVM_GET_SUPPORTED_CPUID ioctl because some of them intersect with KVM feature | |
3853 | leaves (0x40000000, 0x40000001). | |
3854 | ||
3855 | Currently, the following list of CPUID leaves are returned: | |
3856 | HYPERV_CPUID_VENDOR_AND_MAX_FUNCTIONS | |
3857 | HYPERV_CPUID_INTERFACE | |
3858 | HYPERV_CPUID_VERSION | |
3859 | HYPERV_CPUID_FEATURES | |
3860 | HYPERV_CPUID_ENLIGHTMENT_INFO | |
3861 | HYPERV_CPUID_IMPLEMENT_LIMITS | |
3862 | HYPERV_CPUID_NESTED_FEATURES | |
3863 | ||
3864 | HYPERV_CPUID_NESTED_FEATURES leaf is only exposed when Enlightened VMCS was | |
3865 | enabled on the corresponding vCPU (KVM_CAP_HYPERV_ENLIGHTENED_VMCS). | |
3866 | ||
3867 | Userspace invokes KVM_GET_SUPPORTED_CPUID by passing a kvm_cpuid2 structure | |
3868 | with the 'nent' field indicating the number of entries in the variable-size | |
3869 | array 'entries'. If the number of entries is too low to describe all Hyper-V | |
3870 | feature leaves, an error (E2BIG) is returned. If the number is more or equal | |
3871 | to the number of Hyper-V feature leaves, the 'nent' field is adjusted to the | |
3872 | number of valid entries in the 'entries' array, which is then filled. | |
3873 | ||
3874 | 'index' and 'flags' fields in 'struct kvm_cpuid_entry2' are currently reserved, | |
3875 | userspace should not expect to get any particular value there. | |
2a31b9db | 3876 | |
9c1b96e3 | 3877 | 5. The kvm_run structure |
414fa985 | 3878 | ------------------------ |
9c1b96e3 AK |
3879 | |
3880 | Application code obtains a pointer to the kvm_run structure by | |
3881 | mmap()ing a vcpu fd. From that point, application code can control | |
3882 | execution by changing fields in kvm_run prior to calling the KVM_RUN | |
3883 | ioctl, and obtain information about the reason KVM_RUN returned by | |
3884 | looking up structure members. | |
3885 | ||
3886 | struct kvm_run { | |
3887 | /* in */ | |
3888 | __u8 request_interrupt_window; | |
3889 | ||
3890 | Request that KVM_RUN return when it becomes possible to inject external | |
3891 | interrupts into the guest. Useful in conjunction with KVM_INTERRUPT. | |
3892 | ||
460df4c1 PB |
3893 | __u8 immediate_exit; |
3894 | ||
3895 | This field is polled once when KVM_RUN starts; if non-zero, KVM_RUN | |
3896 | exits immediately, returning -EINTR. In the common scenario where a | |
3897 | signal is used to "kick" a VCPU out of KVM_RUN, this field can be used | |
3898 | to avoid usage of KVM_SET_SIGNAL_MASK, which has worse scalability. | |
3899 | Rather than blocking the signal outside KVM_RUN, userspace can set up | |
3900 | a signal handler that sets run->immediate_exit to a non-zero value. | |
3901 | ||
3902 | This field is ignored if KVM_CAP_IMMEDIATE_EXIT is not available. | |
3903 | ||
3904 | __u8 padding1[6]; | |
9c1b96e3 AK |
3905 | |
3906 | /* out */ | |
3907 | __u32 exit_reason; | |
3908 | ||
3909 | When KVM_RUN has returned successfully (return value 0), this informs | |
3910 | application code why KVM_RUN has returned. Allowable values for this | |
3911 | field are detailed below. | |
3912 | ||
3913 | __u8 ready_for_interrupt_injection; | |
3914 | ||
3915 | If request_interrupt_window has been specified, this field indicates | |
3916 | an interrupt can be injected now with KVM_INTERRUPT. | |
3917 | ||
3918 | __u8 if_flag; | |
3919 | ||
3920 | The value of the current interrupt flag. Only valid if in-kernel | |
3921 | local APIC is not used. | |
3922 | ||
f077825a PB |
3923 | __u16 flags; |
3924 | ||
3925 | More architecture-specific flags detailing state of the VCPU that may | |
3926 | affect the device's behavior. The only currently defined flag is | |
3927 | KVM_RUN_X86_SMM, which is valid on x86 machines and is set if the | |
3928 | VCPU is in system management mode. | |
9c1b96e3 AK |
3929 | |
3930 | /* in (pre_kvm_run), out (post_kvm_run) */ | |
3931 | __u64 cr8; | |
3932 | ||
3933 | The value of the cr8 register. Only valid if in-kernel local APIC is | |
3934 | not used. Both input and output. | |
3935 | ||
3936 | __u64 apic_base; | |
3937 | ||
3938 | The value of the APIC BASE msr. Only valid if in-kernel local | |
3939 | APIC is not used. Both input and output. | |
3940 | ||
3941 | union { | |
3942 | /* KVM_EXIT_UNKNOWN */ | |
3943 | struct { | |
3944 | __u64 hardware_exit_reason; | |
3945 | } hw; | |
3946 | ||
3947 | If exit_reason is KVM_EXIT_UNKNOWN, the vcpu has exited due to unknown | |
3948 | reasons. Further architecture-specific information is available in | |
3949 | hardware_exit_reason. | |
3950 | ||
3951 | /* KVM_EXIT_FAIL_ENTRY */ | |
3952 | struct { | |
3953 | __u64 hardware_entry_failure_reason; | |
3954 | } fail_entry; | |
3955 | ||
3956 | If exit_reason is KVM_EXIT_FAIL_ENTRY, the vcpu could not be run due | |
3957 | to unknown reasons. Further architecture-specific information is | |
3958 | available in hardware_entry_failure_reason. | |
3959 | ||
3960 | /* KVM_EXIT_EXCEPTION */ | |
3961 | struct { | |
3962 | __u32 exception; | |
3963 | __u32 error_code; | |
3964 | } ex; | |
3965 | ||
3966 | Unused. | |
3967 | ||
3968 | /* KVM_EXIT_IO */ | |
3969 | struct { | |
3970 | #define KVM_EXIT_IO_IN 0 | |
3971 | #define KVM_EXIT_IO_OUT 1 | |
3972 | __u8 direction; | |
3973 | __u8 size; /* bytes */ | |
3974 | __u16 port; | |
3975 | __u32 count; | |
3976 | __u64 data_offset; /* relative to kvm_run start */ | |
3977 | } io; | |
3978 | ||
2044892d | 3979 | If exit_reason is KVM_EXIT_IO, then the vcpu has |
9c1b96e3 AK |
3980 | executed a port I/O instruction which could not be satisfied by kvm. |
3981 | data_offset describes where the data is located (KVM_EXIT_IO_OUT) or | |
3982 | where kvm expects application code to place the data for the next | |
2044892d | 3983 | KVM_RUN invocation (KVM_EXIT_IO_IN). Data format is a packed array. |
9c1b96e3 | 3984 | |
8ab30c15 | 3985 | /* KVM_EXIT_DEBUG */ |
9c1b96e3 AK |
3986 | struct { |
3987 | struct kvm_debug_exit_arch arch; | |
3988 | } debug; | |
3989 | ||
8ab30c15 AB |
3990 | If the exit_reason is KVM_EXIT_DEBUG, then a vcpu is processing a debug event |
3991 | for which architecture specific information is returned. | |
9c1b96e3 AK |
3992 | |
3993 | /* KVM_EXIT_MMIO */ | |
3994 | struct { | |
3995 | __u64 phys_addr; | |
3996 | __u8 data[8]; | |
3997 | __u32 len; | |
3998 | __u8 is_write; | |
3999 | } mmio; | |
4000 | ||
2044892d | 4001 | If exit_reason is KVM_EXIT_MMIO, then the vcpu has |
9c1b96e3 AK |
4002 | executed a memory-mapped I/O instruction which could not be satisfied |
4003 | by kvm. The 'data' member contains the written data if 'is_write' is | |
4004 | true, and should be filled by application code otherwise. | |
4005 | ||
6acdb160 CD |
4006 | The 'data' member contains, in its first 'len' bytes, the value as it would |
4007 | appear if the VCPU performed a load or store of the appropriate width directly | |
4008 | to the byte array. | |
4009 | ||
cc568ead | 4010 | NOTE: For KVM_EXIT_IO, KVM_EXIT_MMIO, KVM_EXIT_OSI, KVM_EXIT_PAPR and |
ce91ddc4 | 4011 | KVM_EXIT_EPR the corresponding |
ad0a048b AG |
4012 | operations are complete (and guest state is consistent) only after userspace |
4013 | has re-entered the kernel with KVM_RUN. The kernel side will first finish | |
67961344 MT |
4014 | incomplete operations and then check for pending signals. Userspace |
4015 | can re-enter the guest with an unmasked signal pending to complete | |
4016 | pending operations. | |
4017 | ||
9c1b96e3 AK |
4018 | /* KVM_EXIT_HYPERCALL */ |
4019 | struct { | |
4020 | __u64 nr; | |
4021 | __u64 args[6]; | |
4022 | __u64 ret; | |
4023 | __u32 longmode; | |
4024 | __u32 pad; | |
4025 | } hypercall; | |
4026 | ||
647dc49e AK |
4027 | Unused. This was once used for 'hypercall to userspace'. To implement |
4028 | such functionality, use KVM_EXIT_IO (x86) or KVM_EXIT_MMIO (all except s390). | |
4029 | Note KVM_EXIT_IO is significantly faster than KVM_EXIT_MMIO. | |
9c1b96e3 AK |
4030 | |
4031 | /* KVM_EXIT_TPR_ACCESS */ | |
4032 | struct { | |
4033 | __u64 rip; | |
4034 | __u32 is_write; | |
4035 | __u32 pad; | |
4036 | } tpr_access; | |
4037 | ||
4038 | To be documented (KVM_TPR_ACCESS_REPORTING). | |
4039 | ||
4040 | /* KVM_EXIT_S390_SIEIC */ | |
4041 | struct { | |
4042 | __u8 icptcode; | |
4043 | __u64 mask; /* psw upper half */ | |
4044 | __u64 addr; /* psw lower half */ | |
4045 | __u16 ipa; | |
4046 | __u32 ipb; | |
4047 | } s390_sieic; | |
4048 | ||
4049 | s390 specific. | |
4050 | ||
4051 | /* KVM_EXIT_S390_RESET */ | |
4052 | #define KVM_S390_RESET_POR 1 | |
4053 | #define KVM_S390_RESET_CLEAR 2 | |
4054 | #define KVM_S390_RESET_SUBSYSTEM 4 | |
4055 | #define KVM_S390_RESET_CPU_INIT 8 | |
4056 | #define KVM_S390_RESET_IPL 16 | |
4057 | __u64 s390_reset_flags; | |
4058 | ||
4059 | s390 specific. | |
4060 | ||
e168bf8d CO |
4061 | /* KVM_EXIT_S390_UCONTROL */ |
4062 | struct { | |
4063 | __u64 trans_exc_code; | |
4064 | __u32 pgm_code; | |
4065 | } s390_ucontrol; | |
4066 | ||
4067 | s390 specific. A page fault has occurred for a user controlled virtual | |
4068 | machine (KVM_VM_S390_UNCONTROL) on it's host page table that cannot be | |
4069 | resolved by the kernel. | |
4070 | The program code and the translation exception code that were placed | |
4071 | in the cpu's lowcore are presented here as defined by the z Architecture | |
4072 | Principles of Operation Book in the Chapter for Dynamic Address Translation | |
4073 | (DAT) | |
4074 | ||
9c1b96e3 AK |
4075 | /* KVM_EXIT_DCR */ |
4076 | struct { | |
4077 | __u32 dcrn; | |
4078 | __u32 data; | |
4079 | __u8 is_write; | |
4080 | } dcr; | |
4081 | ||
ce91ddc4 | 4082 | Deprecated - was used for 440 KVM. |
9c1b96e3 | 4083 | |
ad0a048b AG |
4084 | /* KVM_EXIT_OSI */ |
4085 | struct { | |
4086 | __u64 gprs[32]; | |
4087 | } osi; | |
4088 | ||
4089 | MOL uses a special hypercall interface it calls 'OSI'. To enable it, we catch | |
4090 | hypercalls and exit with this exit struct that contains all the guest gprs. | |
4091 | ||
4092 | If exit_reason is KVM_EXIT_OSI, then the vcpu has triggered such a hypercall. | |
4093 | Userspace can now handle the hypercall and when it's done modify the gprs as | |
4094 | necessary. Upon guest entry all guest GPRs will then be replaced by the values | |
4095 | in this struct. | |
4096 | ||
de56a948 PM |
4097 | /* KVM_EXIT_PAPR_HCALL */ |
4098 | struct { | |
4099 | __u64 nr; | |
4100 | __u64 ret; | |
4101 | __u64 args[9]; | |
4102 | } papr_hcall; | |
4103 | ||
4104 | This is used on 64-bit PowerPC when emulating a pSeries partition, | |
4105 | e.g. with the 'pseries' machine type in qemu. It occurs when the | |
4106 | guest does a hypercall using the 'sc 1' instruction. The 'nr' field | |
4107 | contains the hypercall number (from the guest R3), and 'args' contains | |
4108 | the arguments (from the guest R4 - R12). Userspace should put the | |
4109 | return code in 'ret' and any extra returned values in args[]. | |
4110 | The possible hypercalls are defined in the Power Architecture Platform | |
4111 | Requirements (PAPR) document available from www.power.org (free | |
4112 | developer registration required to access it). | |
4113 | ||
fa6b7fe9 CH |
4114 | /* KVM_EXIT_S390_TSCH */ |
4115 | struct { | |
4116 | __u16 subchannel_id; | |
4117 | __u16 subchannel_nr; | |
4118 | __u32 io_int_parm; | |
4119 | __u32 io_int_word; | |
4120 | __u32 ipb; | |
4121 | __u8 dequeued; | |
4122 | } s390_tsch; | |
4123 | ||
4124 | s390 specific. This exit occurs when KVM_CAP_S390_CSS_SUPPORT has been enabled | |
4125 | and TEST SUBCHANNEL was intercepted. If dequeued is set, a pending I/O | |
4126 | interrupt for the target subchannel has been dequeued and subchannel_id, | |
4127 | subchannel_nr, io_int_parm and io_int_word contain the parameters for that | |
4128 | interrupt. ipb is needed for instruction parameter decoding. | |
4129 | ||
1c810636 AG |
4130 | /* KVM_EXIT_EPR */ |
4131 | struct { | |
4132 | __u32 epr; | |
4133 | } epr; | |
4134 | ||
4135 | On FSL BookE PowerPC chips, the interrupt controller has a fast patch | |
4136 | interrupt acknowledge path to the core. When the core successfully | |
4137 | delivers an interrupt, it automatically populates the EPR register with | |
4138 | the interrupt vector number and acknowledges the interrupt inside | |
4139 | the interrupt controller. | |
4140 | ||
4141 | In case the interrupt controller lives in user space, we need to do | |
4142 | the interrupt acknowledge cycle through it to fetch the next to be | |
4143 | delivered interrupt vector using this exit. | |
4144 | ||
4145 | It gets triggered whenever both KVM_CAP_PPC_EPR are enabled and an | |
4146 | external interrupt has just been delivered into the guest. User space | |
4147 | should put the acknowledged interrupt vector into the 'epr' field. | |
4148 | ||
8ad6b634 AP |
4149 | /* KVM_EXIT_SYSTEM_EVENT */ |
4150 | struct { | |
4151 | #define KVM_SYSTEM_EVENT_SHUTDOWN 1 | |
4152 | #define KVM_SYSTEM_EVENT_RESET 2 | |
2ce79189 | 4153 | #define KVM_SYSTEM_EVENT_CRASH 3 |
8ad6b634 AP |
4154 | __u32 type; |
4155 | __u64 flags; | |
4156 | } system_event; | |
4157 | ||
4158 | If exit_reason is KVM_EXIT_SYSTEM_EVENT then the vcpu has triggered | |
4159 | a system-level event using some architecture specific mechanism (hypercall | |
4160 | or some special instruction). In case of ARM/ARM64, this is triggered using | |
4161 | HVC instruction based PSCI call from the vcpu. The 'type' field describes | |
4162 | the system-level event type. The 'flags' field describes architecture | |
4163 | specific flags for the system-level event. | |
4164 | ||
cf5d3188 CD |
4165 | Valid values for 'type' are: |
4166 | KVM_SYSTEM_EVENT_SHUTDOWN -- the guest has requested a shutdown of the | |
4167 | VM. Userspace is not obliged to honour this, and if it does honour | |
4168 | this does not need to destroy the VM synchronously (ie it may call | |
4169 | KVM_RUN again before shutdown finally occurs). | |
4170 | KVM_SYSTEM_EVENT_RESET -- the guest has requested a reset of the VM. | |
4171 | As with SHUTDOWN, userspace can choose to ignore the request, or | |
4172 | to schedule the reset to occur in the future and may call KVM_RUN again. | |
2ce79189 AS |
4173 | KVM_SYSTEM_EVENT_CRASH -- the guest crash occurred and the guest |
4174 | has requested a crash condition maintenance. Userspace can choose | |
4175 | to ignore the request, or to gather VM memory core dump and/or | |
4176 | reset/shutdown of the VM. | |
cf5d3188 | 4177 | |
7543a635 SR |
4178 | /* KVM_EXIT_IOAPIC_EOI */ |
4179 | struct { | |
4180 | __u8 vector; | |
4181 | } eoi; | |
4182 | ||
4183 | Indicates that the VCPU's in-kernel local APIC received an EOI for a | |
4184 | level-triggered IOAPIC interrupt. This exit only triggers when the | |
4185 | IOAPIC is implemented in userspace (i.e. KVM_CAP_SPLIT_IRQCHIP is enabled); | |
4186 | the userspace IOAPIC should process the EOI and retrigger the interrupt if | |
4187 | it is still asserted. Vector is the LAPIC interrupt vector for which the | |
4188 | EOI was received. | |
4189 | ||
db397571 AS |
4190 | struct kvm_hyperv_exit { |
4191 | #define KVM_EXIT_HYPERV_SYNIC 1 | |
83326e43 | 4192 | #define KVM_EXIT_HYPERV_HCALL 2 |
db397571 AS |
4193 | __u32 type; |
4194 | union { | |
4195 | struct { | |
4196 | __u32 msr; | |
4197 | __u64 control; | |
4198 | __u64 evt_page; | |
4199 | __u64 msg_page; | |
4200 | } synic; | |
83326e43 AS |
4201 | struct { |
4202 | __u64 input; | |
4203 | __u64 result; | |
4204 | __u64 params[2]; | |
4205 | } hcall; | |
db397571 AS |
4206 | } u; |
4207 | }; | |
4208 | /* KVM_EXIT_HYPERV */ | |
4209 | struct kvm_hyperv_exit hyperv; | |
4210 | Indicates that the VCPU exits into userspace to process some tasks | |
4211 | related to Hyper-V emulation. | |
4212 | Valid values for 'type' are: | |
4213 | KVM_EXIT_HYPERV_SYNIC -- synchronously notify user-space about | |
4214 | Hyper-V SynIC state change. Notification is used to remap SynIC | |
4215 | event/message pages and to enable/disable SynIC messages/events processing | |
4216 | in userspace. | |
4217 | ||
9c1b96e3 AK |
4218 | /* Fix the size of the union. */ |
4219 | char padding[256]; | |
4220 | }; | |
b9e5dc8d CB |
4221 | |
4222 | /* | |
4223 | * shared registers between kvm and userspace. | |
4224 | * kvm_valid_regs specifies the register classes set by the host | |
4225 | * kvm_dirty_regs specified the register classes dirtied by userspace | |
4226 | * struct kvm_sync_regs is architecture specific, as well as the | |
4227 | * bits for kvm_valid_regs and kvm_dirty_regs | |
4228 | */ | |
4229 | __u64 kvm_valid_regs; | |
4230 | __u64 kvm_dirty_regs; | |
4231 | union { | |
4232 | struct kvm_sync_regs regs; | |
7b7e3952 | 4233 | char padding[SYNC_REGS_SIZE_BYTES]; |
b9e5dc8d CB |
4234 | } s; |
4235 | ||
4236 | If KVM_CAP_SYNC_REGS is defined, these fields allow userspace to access | |
4237 | certain guest registers without having to call SET/GET_*REGS. Thus we can | |
4238 | avoid some system call overhead if userspace has to handle the exit. | |
4239 | Userspace can query the validity of the structure by checking | |
4240 | kvm_valid_regs for specific bits. These bits are architecture specific | |
4241 | and usually define the validity of a groups of registers. (e.g. one bit | |
4242 | for general purpose registers) | |
4243 | ||
d8482c0d DH |
4244 | Please note that the kernel is allowed to use the kvm_run structure as the |
4245 | primary storage for certain register types. Therefore, the kernel may use the | |
4246 | values in kvm_run even if the corresponding bit in kvm_dirty_regs is not set. | |
4247 | ||
9c1b96e3 | 4248 | }; |
821246a5 | 4249 | |
414fa985 | 4250 | |
9c15bb1d | 4251 | |
699a0ea0 PM |
4252 | 6. Capabilities that can be enabled on vCPUs |
4253 | -------------------------------------------- | |
821246a5 | 4254 | |
0907c855 CH |
4255 | There are certain capabilities that change the behavior of the virtual CPU or |
4256 | the virtual machine when enabled. To enable them, please see section 4.37. | |
4257 | Below you can find a list of capabilities and what their effect on the vCPU or | |
4258 | the virtual machine is when enabling them. | |
821246a5 AG |
4259 | |
4260 | The following information is provided along with the description: | |
4261 | ||
4262 | Architectures: which instruction set architectures provide this ioctl. | |
4263 | x86 includes both i386 and x86_64. | |
4264 | ||
0907c855 CH |
4265 | Target: whether this is a per-vcpu or per-vm capability. |
4266 | ||
821246a5 AG |
4267 | Parameters: what parameters are accepted by the capability. |
4268 | ||
4269 | Returns: the return value. General error numbers (EBADF, ENOMEM, EINVAL) | |
4270 | are not detailed, but errors with specific meanings are. | |
4271 | ||
414fa985 | 4272 | |
821246a5 AG |
4273 | 6.1 KVM_CAP_PPC_OSI |
4274 | ||
4275 | Architectures: ppc | |
0907c855 | 4276 | Target: vcpu |
821246a5 AG |
4277 | Parameters: none |
4278 | Returns: 0 on success; -1 on error | |
4279 | ||
4280 | This capability enables interception of OSI hypercalls that otherwise would | |
4281 | be treated as normal system calls to be injected into the guest. OSI hypercalls | |
4282 | were invented by Mac-on-Linux to have a standardized communication mechanism | |
4283 | between the guest and the host. | |
4284 | ||
4285 | When this capability is enabled, KVM_EXIT_OSI can occur. | |
4286 | ||
414fa985 | 4287 | |
821246a5 AG |
4288 | 6.2 KVM_CAP_PPC_PAPR |
4289 | ||
4290 | Architectures: ppc | |
0907c855 | 4291 | Target: vcpu |
821246a5 AG |
4292 | Parameters: none |
4293 | Returns: 0 on success; -1 on error | |
4294 | ||
4295 | This capability enables interception of PAPR hypercalls. PAPR hypercalls are | |
4296 | done using the hypercall instruction "sc 1". | |
4297 | ||
4298 | It also sets the guest privilege level to "supervisor" mode. Usually the guest | |
4299 | runs in "hypervisor" privilege mode with a few missing features. | |
4300 | ||
4301 | In addition to the above, it changes the semantics of SDR1. In this mode, the | |
4302 | HTAB address part of SDR1 contains an HVA instead of a GPA, as PAPR keeps the | |
4303 | HTAB invisible to the guest. | |
4304 | ||
4305 | When this capability is enabled, KVM_EXIT_PAPR_HCALL can occur. | |
dc83b8bc | 4306 | |
414fa985 | 4307 | |
dc83b8bc SW |
4308 | 6.3 KVM_CAP_SW_TLB |
4309 | ||
4310 | Architectures: ppc | |
0907c855 | 4311 | Target: vcpu |
dc83b8bc SW |
4312 | Parameters: args[0] is the address of a struct kvm_config_tlb |
4313 | Returns: 0 on success; -1 on error | |
4314 | ||
4315 | struct kvm_config_tlb { | |
4316 | __u64 params; | |
4317 | __u64 array; | |
4318 | __u32 mmu_type; | |
4319 | __u32 array_len; | |
4320 | }; | |
4321 | ||
4322 | Configures the virtual CPU's TLB array, establishing a shared memory area | |
4323 | between userspace and KVM. The "params" and "array" fields are userspace | |
4324 | addresses of mmu-type-specific data structures. The "array_len" field is an | |
4325 | safety mechanism, and should be set to the size in bytes of the memory that | |
4326 | userspace has reserved for the array. It must be at least the size dictated | |
4327 | by "mmu_type" and "params". | |
4328 | ||
4329 | While KVM_RUN is active, the shared region is under control of KVM. Its | |
4330 | contents are undefined, and any modification by userspace results in | |
4331 | boundedly undefined behavior. | |
4332 | ||
4333 | On return from KVM_RUN, the shared region will reflect the current state of | |
4334 | the guest's TLB. If userspace makes any changes, it must call KVM_DIRTY_TLB | |
4335 | to tell KVM which entries have been changed, prior to calling KVM_RUN again | |
4336 | on this vcpu. | |
4337 | ||
4338 | For mmu types KVM_MMU_FSL_BOOKE_NOHV and KVM_MMU_FSL_BOOKE_HV: | |
4339 | - The "params" field is of type "struct kvm_book3e_206_tlb_params". | |
4340 | - The "array" field points to an array of type "struct | |
4341 | kvm_book3e_206_tlb_entry". | |
4342 | - The array consists of all entries in the first TLB, followed by all | |
4343 | entries in the second TLB. | |
4344 | - Within a TLB, entries are ordered first by increasing set number. Within a | |
4345 | set, entries are ordered by way (increasing ESEL). | |
4346 | - The hash for determining set number in TLB0 is: (MAS2 >> 12) & (num_sets - 1) | |
4347 | where "num_sets" is the tlb_sizes[] value divided by the tlb_ways[] value. | |
4348 | - The tsize field of mas1 shall be set to 4K on TLB0, even though the | |
4349 | hardware ignores this value for TLB0. | |
fa6b7fe9 CH |
4350 | |
4351 | 6.4 KVM_CAP_S390_CSS_SUPPORT | |
4352 | ||
4353 | Architectures: s390 | |
0907c855 | 4354 | Target: vcpu |
fa6b7fe9 CH |
4355 | Parameters: none |
4356 | Returns: 0 on success; -1 on error | |
4357 | ||
4358 | This capability enables support for handling of channel I/O instructions. | |
4359 | ||
4360 | TEST PENDING INTERRUPTION and the interrupt portion of TEST SUBCHANNEL are | |
4361 | handled in-kernel, while the other I/O instructions are passed to userspace. | |
4362 | ||
4363 | When this capability is enabled, KVM_EXIT_S390_TSCH will occur on TEST | |
4364 | SUBCHANNEL intercepts. | |
1c810636 | 4365 | |
0907c855 CH |
4366 | Note that even though this capability is enabled per-vcpu, the complete |
4367 | virtual machine is affected. | |
4368 | ||
1c810636 AG |
4369 | 6.5 KVM_CAP_PPC_EPR |
4370 | ||
4371 | Architectures: ppc | |
0907c855 | 4372 | Target: vcpu |
1c810636 AG |
4373 | Parameters: args[0] defines whether the proxy facility is active |
4374 | Returns: 0 on success; -1 on error | |
4375 | ||
4376 | This capability enables or disables the delivery of interrupts through the | |
4377 | external proxy facility. | |
4378 | ||
4379 | When enabled (args[0] != 0), every time the guest gets an external interrupt | |
4380 | delivered, it automatically exits into user space with a KVM_EXIT_EPR exit | |
4381 | to receive the topmost interrupt vector. | |
4382 | ||
4383 | When disabled (args[0] == 0), behavior is as if this facility is unsupported. | |
4384 | ||
4385 | When this capability is enabled, KVM_EXIT_EPR can occur. | |
eb1e4f43 SW |
4386 | |
4387 | 6.6 KVM_CAP_IRQ_MPIC | |
4388 | ||
4389 | Architectures: ppc | |
4390 | Parameters: args[0] is the MPIC device fd | |
4391 | args[1] is the MPIC CPU number for this vcpu | |
4392 | ||
4393 | This capability connects the vcpu to an in-kernel MPIC device. | |
5975a2e0 PM |
4394 | |
4395 | 6.7 KVM_CAP_IRQ_XICS | |
4396 | ||
4397 | Architectures: ppc | |
0907c855 | 4398 | Target: vcpu |
5975a2e0 PM |
4399 | Parameters: args[0] is the XICS device fd |
4400 | args[1] is the XICS CPU number (server ID) for this vcpu | |
4401 | ||
4402 | This capability connects the vcpu to an in-kernel XICS device. | |
8a366a4b CH |
4403 | |
4404 | 6.8 KVM_CAP_S390_IRQCHIP | |
4405 | ||
4406 | Architectures: s390 | |
4407 | Target: vm | |
4408 | Parameters: none | |
4409 | ||
4410 | This capability enables the in-kernel irqchip for s390. Please refer to | |
4411 | "4.24 KVM_CREATE_IRQCHIP" for details. | |
699a0ea0 | 4412 | |
5fafd874 JH |
4413 | 6.9 KVM_CAP_MIPS_FPU |
4414 | ||
4415 | Architectures: mips | |
4416 | Target: vcpu | |
4417 | Parameters: args[0] is reserved for future use (should be 0). | |
4418 | ||
4419 | This capability allows the use of the host Floating Point Unit by the guest. It | |
4420 | allows the Config1.FP bit to be set to enable the FPU in the guest. Once this is | |
4421 | done the KVM_REG_MIPS_FPR_* and KVM_REG_MIPS_FCR_* registers can be accessed | |
4422 | (depending on the current guest FPU register mode), and the Status.FR, | |
4423 | Config5.FRE bits are accessible via the KVM API and also from the guest, | |
4424 | depending on them being supported by the FPU. | |
4425 | ||
d952bd07 JH |
4426 | 6.10 KVM_CAP_MIPS_MSA |
4427 | ||
4428 | Architectures: mips | |
4429 | Target: vcpu | |
4430 | Parameters: args[0] is reserved for future use (should be 0). | |
4431 | ||
4432 | This capability allows the use of the MIPS SIMD Architecture (MSA) by the guest. | |
4433 | It allows the Config3.MSAP bit to be set to enable the use of MSA by the guest. | |
4434 | Once this is done the KVM_REG_MIPS_VEC_* and KVM_REG_MIPS_MSA_* registers can be | |
4435 | accessed, and the Config5.MSAEn bit is accessible via the KVM API and also from | |
4436 | the guest. | |
4437 | ||
01643c51 KH |
4438 | 6.74 KVM_CAP_SYNC_REGS |
4439 | Architectures: s390, x86 | |
4440 | Target: s390: always enabled, x86: vcpu | |
4441 | Parameters: none | |
4442 | Returns: x86: KVM_CHECK_EXTENSION returns a bit-array indicating which register | |
4443 | sets are supported (bitfields defined in arch/x86/include/uapi/asm/kvm.h). | |
4444 | ||
4445 | As described above in the kvm_sync_regs struct info in section 5 (kvm_run): | |
4446 | KVM_CAP_SYNC_REGS "allow[s] userspace to access certain guest registers | |
4447 | without having to call SET/GET_*REGS". This reduces overhead by eliminating | |
4448 | repeated ioctl calls for setting and/or getting register values. This is | |
4449 | particularly important when userspace is making synchronous guest state | |
4450 | modifications, e.g. when emulating and/or intercepting instructions in | |
4451 | userspace. | |
4452 | ||
4453 | For s390 specifics, please refer to the source code. | |
4454 | ||
4455 | For x86: | |
4456 | - the register sets to be copied out to kvm_run are selectable | |
4457 | by userspace (rather that all sets being copied out for every exit). | |
4458 | - vcpu_events are available in addition to regs and sregs. | |
4459 | ||
4460 | For x86, the 'kvm_valid_regs' field of struct kvm_run is overloaded to | |
4461 | function as an input bit-array field set by userspace to indicate the | |
4462 | specific register sets to be copied out on the next exit. | |
4463 | ||
4464 | To indicate when userspace has modified values that should be copied into | |
4465 | the vCPU, the all architecture bitarray field, 'kvm_dirty_regs' must be set. | |
4466 | This is done using the same bitflags as for the 'kvm_valid_regs' field. | |
4467 | If the dirty bit is not set, then the register set values will not be copied | |
4468 | into the vCPU even if they've been modified. | |
4469 | ||
4470 | Unused bitfields in the bitarrays must be set to zero. | |
4471 | ||
4472 | struct kvm_sync_regs { | |
4473 | struct kvm_regs regs; | |
4474 | struct kvm_sregs sregs; | |
4475 | struct kvm_vcpu_events events; | |
4476 | }; | |
4477 | ||
699a0ea0 PM |
4478 | 7. Capabilities that can be enabled on VMs |
4479 | ------------------------------------------ | |
4480 | ||
4481 | There are certain capabilities that change the behavior of the virtual | |
4482 | machine when enabled. To enable them, please see section 4.37. Below | |
4483 | you can find a list of capabilities and what their effect on the VM | |
4484 | is when enabling them. | |
4485 | ||
4486 | The following information is provided along with the description: | |
4487 | ||
4488 | Architectures: which instruction set architectures provide this ioctl. | |
4489 | x86 includes both i386 and x86_64. | |
4490 | ||
4491 | Parameters: what parameters are accepted by the capability. | |
4492 | ||
4493 | Returns: the return value. General error numbers (EBADF, ENOMEM, EINVAL) | |
4494 | are not detailed, but errors with specific meanings are. | |
4495 | ||
4496 | ||
4497 | 7.1 KVM_CAP_PPC_ENABLE_HCALL | |
4498 | ||
4499 | Architectures: ppc | |
4500 | Parameters: args[0] is the sPAPR hcall number | |
4501 | args[1] is 0 to disable, 1 to enable in-kernel handling | |
4502 | ||
4503 | This capability controls whether individual sPAPR hypercalls (hcalls) | |
4504 | get handled by the kernel or not. Enabling or disabling in-kernel | |
4505 | handling of an hcall is effective across the VM. On creation, an | |
4506 | initial set of hcalls are enabled for in-kernel handling, which | |
4507 | consists of those hcalls for which in-kernel handlers were implemented | |
4508 | before this capability was implemented. If disabled, the kernel will | |
4509 | not to attempt to handle the hcall, but will always exit to userspace | |
4510 | to handle it. Note that it may not make sense to enable some and | |
4511 | disable others of a group of related hcalls, but KVM does not prevent | |
4512 | userspace from doing that. | |
ae2113a4 PM |
4513 | |
4514 | If the hcall number specified is not one that has an in-kernel | |
4515 | implementation, the KVM_ENABLE_CAP ioctl will fail with an EINVAL | |
4516 | error. | |
2444b352 DH |
4517 | |
4518 | 7.2 KVM_CAP_S390_USER_SIGP | |
4519 | ||
4520 | Architectures: s390 | |
4521 | Parameters: none | |
4522 | ||
4523 | This capability controls which SIGP orders will be handled completely in user | |
4524 | space. With this capability enabled, all fast orders will be handled completely | |
4525 | in the kernel: | |
4526 | - SENSE | |
4527 | - SENSE RUNNING | |
4528 | - EXTERNAL CALL | |
4529 | - EMERGENCY SIGNAL | |
4530 | - CONDITIONAL EMERGENCY SIGNAL | |
4531 | ||
4532 | All other orders will be handled completely in user space. | |
4533 | ||
4534 | Only privileged operation exceptions will be checked for in the kernel (or even | |
4535 | in the hardware prior to interception). If this capability is not enabled, the | |
4536 | old way of handling SIGP orders is used (partially in kernel and user space). | |
68c55750 EF |
4537 | |
4538 | 7.3 KVM_CAP_S390_VECTOR_REGISTERS | |
4539 | ||
4540 | Architectures: s390 | |
4541 | Parameters: none | |
4542 | Returns: 0 on success, negative value on error | |
4543 | ||
4544 | Allows use of the vector registers introduced with z13 processor, and | |
4545 | provides for the synchronization between host and user space. Will | |
4546 | return -EINVAL if the machine does not support vectors. | |
e44fc8c9 ET |
4547 | |
4548 | 7.4 KVM_CAP_S390_USER_STSI | |
4549 | ||
4550 | Architectures: s390 | |
4551 | Parameters: none | |
4552 | ||
4553 | This capability allows post-handlers for the STSI instruction. After | |
4554 | initial handling in the kernel, KVM exits to user space with | |
4555 | KVM_EXIT_S390_STSI to allow user space to insert further data. | |
4556 | ||
4557 | Before exiting to userspace, kvm handlers should fill in s390_stsi field of | |
4558 | vcpu->run: | |
4559 | struct { | |
4560 | __u64 addr; | |
4561 | __u8 ar; | |
4562 | __u8 reserved; | |
4563 | __u8 fc; | |
4564 | __u8 sel1; | |
4565 | __u16 sel2; | |
4566 | } s390_stsi; | |
4567 | ||
4568 | @addr - guest address of STSI SYSIB | |
4569 | @fc - function code | |
4570 | @sel1 - selector 1 | |
4571 | @sel2 - selector 2 | |
4572 | @ar - access register number | |
4573 | ||
4574 | KVM handlers should exit to userspace with rc = -EREMOTE. | |
e928e9cb | 4575 | |
49df6397 SR |
4576 | 7.5 KVM_CAP_SPLIT_IRQCHIP |
4577 | ||
4578 | Architectures: x86 | |
b053b2ae | 4579 | Parameters: args[0] - number of routes reserved for userspace IOAPICs |
49df6397 SR |
4580 | Returns: 0 on success, -1 on error |
4581 | ||
4582 | Create a local apic for each processor in the kernel. This can be used | |
4583 | instead of KVM_CREATE_IRQCHIP if the userspace VMM wishes to emulate the | |
4584 | IOAPIC and PIC (and also the PIT, even though this has to be enabled | |
4585 | separately). | |
4586 | ||
b053b2ae SR |
4587 | This capability also enables in kernel routing of interrupt requests; |
4588 | when KVM_CAP_SPLIT_IRQCHIP only routes of KVM_IRQ_ROUTING_MSI type are | |
4589 | used in the IRQ routing table. The first args[0] MSI routes are reserved | |
4590 | for the IOAPIC pins. Whenever the LAPIC receives an EOI for these routes, | |
4591 | a KVM_EXIT_IOAPIC_EOI vmexit will be reported to userspace. | |
49df6397 SR |
4592 | |
4593 | Fails if VCPU has already been created, or if the irqchip is already in the | |
4594 | kernel (i.e. KVM_CREATE_IRQCHIP has already been called). | |
4595 | ||
051c87f7 DH |
4596 | 7.6 KVM_CAP_S390_RI |
4597 | ||
4598 | Architectures: s390 | |
4599 | Parameters: none | |
4600 | ||
4601 | Allows use of runtime-instrumentation introduced with zEC12 processor. | |
4602 | Will return -EINVAL if the machine does not support runtime-instrumentation. | |
4603 | Will return -EBUSY if a VCPU has already been created. | |
e928e9cb | 4604 | |
37131313 RK |
4605 | 7.7 KVM_CAP_X2APIC_API |
4606 | ||
4607 | Architectures: x86 | |
4608 | Parameters: args[0] - features that should be enabled | |
4609 | Returns: 0 on success, -EINVAL when args[0] contains invalid features | |
4610 | ||
4611 | Valid feature flags in args[0] are | |
4612 | ||
4613 | #define KVM_X2APIC_API_USE_32BIT_IDS (1ULL << 0) | |
c519265f | 4614 | #define KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK (1ULL << 1) |
37131313 RK |
4615 | |
4616 | Enabling KVM_X2APIC_API_USE_32BIT_IDS changes the behavior of | |
4617 | KVM_SET_GSI_ROUTING, KVM_SIGNAL_MSI, KVM_SET_LAPIC, and KVM_GET_LAPIC, | |
4618 | allowing the use of 32-bit APIC IDs. See KVM_CAP_X2APIC_API in their | |
4619 | respective sections. | |
4620 | ||
c519265f RK |
4621 | KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK must be enabled for x2APIC to work |
4622 | in logical mode or with more than 255 VCPUs. Otherwise, KVM treats 0xff | |
4623 | as a broadcast even in x2APIC mode in order to support physical x2APIC | |
4624 | without interrupt remapping. This is undesirable in logical mode, | |
4625 | where 0xff represents CPUs 0-7 in cluster 0. | |
37131313 | 4626 | |
6502a34c DH |
4627 | 7.8 KVM_CAP_S390_USER_INSTR0 |
4628 | ||
4629 | Architectures: s390 | |
4630 | Parameters: none | |
4631 | ||
4632 | With this capability enabled, all illegal instructions 0x0000 (2 bytes) will | |
4633 | be intercepted and forwarded to user space. User space can use this | |
4634 | mechanism e.g. to realize 2-byte software breakpoints. The kernel will | |
4635 | not inject an operating exception for these instructions, user space has | |
4636 | to take care of that. | |
4637 | ||
4638 | This capability can be enabled dynamically even if VCPUs were already | |
4639 | created and are running. | |
37131313 | 4640 | |
4e0b1ab7 FZ |
4641 | 7.9 KVM_CAP_S390_GS |
4642 | ||
4643 | Architectures: s390 | |
4644 | Parameters: none | |
4645 | Returns: 0 on success; -EINVAL if the machine does not support | |
4646 | guarded storage; -EBUSY if a VCPU has already been created. | |
4647 | ||
4648 | Allows use of guarded storage for the KVM guest. | |
4649 | ||
47a4693e YMZ |
4650 | 7.10 KVM_CAP_S390_AIS |
4651 | ||
4652 | Architectures: s390 | |
4653 | Parameters: none | |
4654 | ||
4655 | Allow use of adapter-interruption suppression. | |
4656 | Returns: 0 on success; -EBUSY if a VCPU has already been created. | |
4657 | ||
3c313524 PM |
4658 | 7.11 KVM_CAP_PPC_SMT |
4659 | ||
4660 | Architectures: ppc | |
4661 | Parameters: vsmt_mode, flags | |
4662 | ||
4663 | Enabling this capability on a VM provides userspace with a way to set | |
4664 | the desired virtual SMT mode (i.e. the number of virtual CPUs per | |
4665 | virtual core). The virtual SMT mode, vsmt_mode, must be a power of 2 | |
4666 | between 1 and 8. On POWER8, vsmt_mode must also be no greater than | |
4667 | the number of threads per subcore for the host. Currently flags must | |
4668 | be 0. A successful call to enable this capability will result in | |
4669 | vsmt_mode being returned when the KVM_CAP_PPC_SMT capability is | |
4670 | subsequently queried for the VM. This capability is only supported by | |
4671 | HV KVM, and can only be set before any VCPUs have been created. | |
2ed4f9dd PM |
4672 | The KVM_CAP_PPC_SMT_POSSIBLE capability indicates which virtual SMT |
4673 | modes are available. | |
3c313524 | 4674 | |
134764ed AP |
4675 | 7.12 KVM_CAP_PPC_FWNMI |
4676 | ||
4677 | Architectures: ppc | |
4678 | Parameters: none | |
4679 | ||
4680 | With this capability a machine check exception in the guest address | |
4681 | space will cause KVM to exit the guest with NMI exit reason. This | |
4682 | enables QEMU to build error log and branch to guest kernel registered | |
4683 | machine check handling routine. Without this capability KVM will | |
4684 | branch to guests' 0x200 interrupt vector. | |
4685 | ||
4d5422ce WL |
4686 | 7.13 KVM_CAP_X86_DISABLE_EXITS |
4687 | ||
4688 | Architectures: x86 | |
4689 | Parameters: args[0] defines which exits are disabled | |
4690 | Returns: 0 on success, -EINVAL when args[0] contains invalid exits | |
4691 | ||
4692 | Valid bits in args[0] are | |
4693 | ||
4694 | #define KVM_X86_DISABLE_EXITS_MWAIT (1 << 0) | |
caa057a2 | 4695 | #define KVM_X86_DISABLE_EXITS_HLT (1 << 1) |
4d5422ce WL |
4696 | |
4697 | Enabling this capability on a VM provides userspace with a way to no | |
4698 | longer intercept some instructions for improved latency in some | |
4699 | workloads, and is suggested when vCPUs are associated to dedicated | |
4700 | physical CPUs. More bits can be added in the future; userspace can | |
4701 | just pass the KVM_CHECK_EXTENSION result to KVM_ENABLE_CAP to disable | |
4702 | all such vmexits. | |
4703 | ||
caa057a2 | 4704 | Do not enable KVM_FEATURE_PV_UNHALT if you disable HLT exits. |
4d5422ce | 4705 | |
a4499382 JF |
4706 | 7.14 KVM_CAP_S390_HPAGE_1M |
4707 | ||
4708 | Architectures: s390 | |
4709 | Parameters: none | |
4710 | Returns: 0 on success, -EINVAL if hpage module parameter was not set | |
40ebdb8e JF |
4711 | or cmma is enabled, or the VM has the KVM_VM_S390_UCONTROL |
4712 | flag set | |
a4499382 JF |
4713 | |
4714 | With this capability the KVM support for memory backing with 1m pages | |
4715 | through hugetlbfs can be enabled for a VM. After the capability is | |
4716 | enabled, cmma can't be enabled anymore and pfmfi and the storage key | |
4717 | interpretation are disabled. If cmma has already been enabled or the | |
4718 | hpage module parameter is not set to 1, -EINVAL is returned. | |
4719 | ||
4720 | While it is generally possible to create a huge page backed VM without | |
4721 | this capability, the VM will not be able to run. | |
4722 | ||
c4f55198 | 4723 | 7.15 KVM_CAP_MSR_PLATFORM_INFO |
6fbbde9a DS |
4724 | |
4725 | Architectures: x86 | |
4726 | Parameters: args[0] whether feature should be enabled or not | |
4727 | ||
4728 | With this capability, a guest may read the MSR_PLATFORM_INFO MSR. Otherwise, | |
4729 | a #GP would be raised when the guest tries to access. Currently, this | |
4730 | capability does not enable write permissions of this MSR for the guest. | |
4731 | ||
aa069a99 PM |
4732 | 7.16 KVM_CAP_PPC_NESTED_HV |
4733 | ||
4734 | Architectures: ppc | |
4735 | Parameters: none | |
4736 | Returns: 0 on success, -EINVAL when the implementation doesn't support | |
4737 | nested-HV virtualization. | |
4738 | ||
4739 | HV-KVM on POWER9 and later systems allows for "nested-HV" | |
4740 | virtualization, which provides a way for a guest VM to run guests that | |
4741 | can run using the CPU's supervisor mode (privileged non-hypervisor | |
4742 | state). Enabling this capability on a VM depends on the CPU having | |
4743 | the necessary functionality and on the facility being enabled with a | |
4744 | kvm-hv module parameter. | |
4745 | ||
c4f55198 JM |
4746 | 7.17 KVM_CAP_EXCEPTION_PAYLOAD |
4747 | ||
4748 | Architectures: x86 | |
4749 | Parameters: args[0] whether feature should be enabled or not | |
4750 | ||
4751 | With this capability enabled, CR2 will not be modified prior to the | |
4752 | emulated VM-exit when L1 intercepts a #PF exception that occurs in | |
4753 | L2. Similarly, for kvm-intel only, DR6 will not be modified prior to | |
4754 | the emulated VM-exit when L1 intercepts a #DB exception that occurs in | |
4755 | L2. As a result, when KVM_GET_VCPU_EVENTS reports a pending #PF (or | |
4756 | #DB) exception for L2, exception.has_payload will be set and the | |
4757 | faulting address (or the new DR6 bits*) will be reported in the | |
4758 | exception_payload field. Similarly, when userspace injects a #PF (or | |
4759 | #DB) into L2 using KVM_SET_VCPU_EVENTS, it is expected to set | |
4760 | exception.has_payload and to put the faulting address (or the new DR6 | |
4761 | bits*) in the exception_payload field. | |
4762 | ||
4763 | This capability also enables exception.pending in struct | |
4764 | kvm_vcpu_events, which allows userspace to distinguish between pending | |
4765 | and injected exceptions. | |
4766 | ||
4767 | ||
4768 | * For the new DR6 bits, note that bit 16 is set iff the #DB exception | |
4769 | will clear DR6.RTM. | |
4770 | ||
2a31b9db PB |
4771 | 7.18 KVM_CAP_MANUAL_DIRTY_LOG_PROTECT |
4772 | ||
4773 | Architectures: all | |
4774 | Parameters: args[0] whether feature should be enabled or not | |
4775 | ||
4776 | With this capability enabled, KVM_GET_DIRTY_LOG will not automatically | |
4777 | clear and write-protect all pages that are returned as dirty. | |
4778 | Rather, userspace will have to do this operation separately using | |
4779 | KVM_CLEAR_DIRTY_LOG. | |
4780 | ||
4781 | At the cost of a slightly more complicated operation, this provides better | |
4782 | scalability and responsiveness for two reasons. First, | |
4783 | KVM_CLEAR_DIRTY_LOG ioctl can operate on a 64-page granularity rather | |
4784 | than requiring to sync a full memslot; this ensures that KVM does not | |
4785 | take spinlocks for an extended period of time. Second, in some cases a | |
4786 | large amount of time can pass between a call to KVM_GET_DIRTY_LOG and | |
4787 | userspace actually using the data in the page. Pages can be modified | |
4788 | during this time, which is inefficint for both the guest and userspace: | |
4789 | the guest will incur a higher penalty due to write protection faults, | |
4790 | while userspace can see false reports of dirty pages. Manual reprotection | |
4791 | helps reducing this time, improving guest performance and reducing the | |
4792 | number of dirty log false positives. | |
4793 | ||
4794 | ||
e928e9cb ME |
4795 | 8. Other capabilities. |
4796 | ---------------------- | |
4797 | ||
4798 | This section lists capabilities that give information about other | |
4799 | features of the KVM implementation. | |
4800 | ||
4801 | 8.1 KVM_CAP_PPC_HWRNG | |
4802 | ||
4803 | Architectures: ppc | |
4804 | ||
4805 | This capability, if KVM_CHECK_EXTENSION indicates that it is | |
4806 | available, means that that the kernel has an implementation of the | |
4807 | H_RANDOM hypercall backed by a hardware random-number generator. | |
4808 | If present, the kernel H_RANDOM handler can be enabled for guest use | |
4809 | with the KVM_CAP_PPC_ENABLE_HCALL capability. | |
5c919412 AS |
4810 | |
4811 | 8.2 KVM_CAP_HYPERV_SYNIC | |
4812 | ||
4813 | Architectures: x86 | |
4814 | This capability, if KVM_CHECK_EXTENSION indicates that it is | |
4815 | available, means that that the kernel has an implementation of the | |
4816 | Hyper-V Synthetic interrupt controller(SynIC). Hyper-V SynIC is | |
4817 | used to support Windows Hyper-V based guest paravirt drivers(VMBus). | |
4818 | ||
4819 | In order to use SynIC, it has to be activated by setting this | |
4820 | capability via KVM_ENABLE_CAP ioctl on the vcpu fd. Note that this | |
4821 | will disable the use of APIC hardware virtualization even if supported | |
4822 | by the CPU, as it's incompatible with SynIC auto-EOI behavior. | |
c9270132 PM |
4823 | |
4824 | 8.3 KVM_CAP_PPC_RADIX_MMU | |
4825 | ||
4826 | Architectures: ppc | |
4827 | ||
4828 | This capability, if KVM_CHECK_EXTENSION indicates that it is | |
4829 | available, means that that the kernel can support guests using the | |
4830 | radix MMU defined in Power ISA V3.00 (as implemented in the POWER9 | |
4831 | processor). | |
4832 | ||
4833 | 8.4 KVM_CAP_PPC_HASH_MMU_V3 | |
4834 | ||
4835 | Architectures: ppc | |
4836 | ||
4837 | This capability, if KVM_CHECK_EXTENSION indicates that it is | |
4838 | available, means that that the kernel can support guests using the | |
4839 | hashed page table MMU defined in Power ISA V3.00 (as implemented in | |
4840 | the POWER9 processor), including in-memory segment tables. | |
a8a3c426 JH |
4841 | |
4842 | 8.5 KVM_CAP_MIPS_VZ | |
4843 | ||
4844 | Architectures: mips | |
4845 | ||
4846 | This capability, if KVM_CHECK_EXTENSION on the main kvm handle indicates that | |
4847 | it is available, means that full hardware assisted virtualization capabilities | |
4848 | of the hardware are available for use through KVM. An appropriate | |
4849 | KVM_VM_MIPS_* type must be passed to KVM_CREATE_VM to create a VM which | |
4850 | utilises it. | |
4851 | ||
4852 | If KVM_CHECK_EXTENSION on a kvm VM handle indicates that this capability is | |
4853 | available, it means that the VM is using full hardware assisted virtualization | |
4854 | capabilities of the hardware. This is useful to check after creating a VM with | |
4855 | KVM_VM_MIPS_DEFAULT. | |
4856 | ||
4857 | The value returned by KVM_CHECK_EXTENSION should be compared against known | |
4858 | values (see below). All other values are reserved. This is to allow for the | |
4859 | possibility of other hardware assisted virtualization implementations which | |
4860 | may be incompatible with the MIPS VZ ASE. | |
4861 | ||
4862 | 0: The trap & emulate implementation is in use to run guest code in user | |
4863 | mode. Guest virtual memory segments are rearranged to fit the guest in the | |
4864 | user mode address space. | |
4865 | ||
4866 | 1: The MIPS VZ ASE is in use, providing full hardware assisted | |
4867 | virtualization, including standard guest virtual memory segments. | |
4868 | ||
4869 | 8.6 KVM_CAP_MIPS_TE | |
4870 | ||
4871 | Architectures: mips | |
4872 | ||
4873 | This capability, if KVM_CHECK_EXTENSION on the main kvm handle indicates that | |
4874 | it is available, means that the trap & emulate implementation is available to | |
4875 | run guest code in user mode, even if KVM_CAP_MIPS_VZ indicates that hardware | |
4876 | assisted virtualisation is also available. KVM_VM_MIPS_TE (0) must be passed | |
4877 | to KVM_CREATE_VM to create a VM which utilises it. | |
4878 | ||
4879 | If KVM_CHECK_EXTENSION on a kvm VM handle indicates that this capability is | |
4880 | available, it means that the VM is using trap & emulate. | |
578fd61d JH |
4881 | |
4882 | 8.7 KVM_CAP_MIPS_64BIT | |
4883 | ||
4884 | Architectures: mips | |
4885 | ||
4886 | This capability indicates the supported architecture type of the guest, i.e. the | |
4887 | supported register and address width. | |
4888 | ||
4889 | The values returned when this capability is checked by KVM_CHECK_EXTENSION on a | |
4890 | kvm VM handle correspond roughly to the CP0_Config.AT register field, and should | |
4891 | be checked specifically against known values (see below). All other values are | |
4892 | reserved. | |
4893 | ||
4894 | 0: MIPS32 or microMIPS32. | |
4895 | Both registers and addresses are 32-bits wide. | |
4896 | It will only be possible to run 32-bit guest code. | |
4897 | ||
4898 | 1: MIPS64 or microMIPS64 with access only to 32-bit compatibility segments. | |
4899 | Registers are 64-bits wide, but addresses are 32-bits wide. | |
4900 | 64-bit guest code may run but cannot access MIPS64 memory segments. | |
4901 | It will also be possible to run 32-bit guest code. | |
4902 | ||
4903 | 2: MIPS64 or microMIPS64 with access to all address segments. | |
4904 | Both registers and addresses are 64-bits wide. | |
4905 | It will be possible to run 64-bit or 32-bit guest code. | |
668fffa3 | 4906 | |
c24a7be2 | 4907 | 8.9 KVM_CAP_ARM_USER_IRQ |
3fe17e68 AG |
4908 | |
4909 | Architectures: arm, arm64 | |
4910 | This capability, if KVM_CHECK_EXTENSION indicates that it is available, means | |
4911 | that if userspace creates a VM without an in-kernel interrupt controller, it | |
4912 | will be notified of changes to the output level of in-kernel emulated devices, | |
4913 | which can generate virtual interrupts, presented to the VM. | |
4914 | For such VMs, on every return to userspace, the kernel | |
4915 | updates the vcpu's run->s.regs.device_irq_level field to represent the actual | |
4916 | output level of the device. | |
4917 | ||
4918 | Whenever kvm detects a change in the device output level, kvm guarantees at | |
4919 | least one return to userspace before running the VM. This exit could either | |
4920 | be a KVM_EXIT_INTR or any other exit event, like KVM_EXIT_MMIO. This way, | |
4921 | userspace can always sample the device output level and re-compute the state of | |
4922 | the userspace interrupt controller. Userspace should always check the state | |
4923 | of run->s.regs.device_irq_level on every kvm exit. | |
4924 | The value in run->s.regs.device_irq_level can represent both level and edge | |
4925 | triggered interrupt signals, depending on the device. Edge triggered interrupt | |
4926 | signals will exit to userspace with the bit in run->s.regs.device_irq_level | |
4927 | set exactly once per edge signal. | |
4928 | ||
4929 | The field run->s.regs.device_irq_level is available independent of | |
4930 | run->kvm_valid_regs or run->kvm_dirty_regs bits. | |
4931 | ||
4932 | If KVM_CAP_ARM_USER_IRQ is supported, the KVM_CHECK_EXTENSION ioctl returns a | |
4933 | number larger than 0 indicating the version of this capability is implemented | |
4934 | and thereby which bits in in run->s.regs.device_irq_level can signal values. | |
4935 | ||
4936 | Currently the following bits are defined for the device_irq_level bitmap: | |
4937 | ||
4938 | KVM_CAP_ARM_USER_IRQ >= 1: | |
4939 | ||
4940 | KVM_ARM_DEV_EL1_VTIMER - EL1 virtual timer | |
4941 | KVM_ARM_DEV_EL1_PTIMER - EL1 physical timer | |
4942 | KVM_ARM_DEV_PMU - ARM PMU overflow interrupt signal | |
4943 | ||
4944 | Future versions of kvm may implement additional events. These will get | |
4945 | indicated by returning a higher number from KVM_CHECK_EXTENSION and will be | |
4946 | listed above. | |
2ed4f9dd PM |
4947 | |
4948 | 8.10 KVM_CAP_PPC_SMT_POSSIBLE | |
4949 | ||
4950 | Architectures: ppc | |
4951 | ||
4952 | Querying this capability returns a bitmap indicating the possible | |
4953 | virtual SMT modes that can be set using KVM_CAP_PPC_SMT. If bit N | |
4954 | (counting from the right) is set, then a virtual SMT mode of 2^N is | |
4955 | available. | |
efc479e6 RK |
4956 | |
4957 | 8.11 KVM_CAP_HYPERV_SYNIC2 | |
4958 | ||
4959 | Architectures: x86 | |
4960 | ||
4961 | This capability enables a newer version of Hyper-V Synthetic interrupt | |
4962 | controller (SynIC). The only difference with KVM_CAP_HYPERV_SYNIC is that KVM | |
4963 | doesn't clear SynIC message and event flags pages when they are enabled by | |
4964 | writing to the respective MSRs. | |
d3457c87 RK |
4965 | |
4966 | 8.12 KVM_CAP_HYPERV_VP_INDEX | |
4967 | ||
4968 | Architectures: x86 | |
4969 | ||
4970 | This capability indicates that userspace can load HV_X64_MSR_VP_INDEX msr. Its | |
4971 | value is used to denote the target vcpu for a SynIC interrupt. For | |
4972 | compatibilty, KVM initializes this msr to KVM's internal vcpu index. When this | |
4973 | capability is absent, userspace can still query this msr's value. | |
da9a1446 CB |
4974 | |
4975 | 8.13 KVM_CAP_S390_AIS_MIGRATION | |
4976 | ||
4977 | Architectures: s390 | |
4978 | Parameters: none | |
4979 | ||
4980 | This capability indicates if the flic device will be able to get/set the | |
4981 | AIS states for migration via the KVM_DEV_FLIC_AISM_ALL attribute and allows | |
4982 | to discover this without having to create a flic device. | |
5c2b4d5b CB |
4983 | |
4984 | 8.14 KVM_CAP_S390_PSW | |
4985 | ||
4986 | Architectures: s390 | |
4987 | ||
4988 | This capability indicates that the PSW is exposed via the kvm_run structure. | |
4989 | ||
4990 | 8.15 KVM_CAP_S390_GMAP | |
4991 | ||
4992 | Architectures: s390 | |
4993 | ||
4994 | This capability indicates that the user space memory used as guest mapping can | |
4995 | be anywhere in the user memory address space, as long as the memory slots are | |
4996 | aligned and sized to a segment (1MB) boundary. | |
4997 | ||
4998 | 8.16 KVM_CAP_S390_COW | |
4999 | ||
5000 | Architectures: s390 | |
5001 | ||
5002 | This capability indicates that the user space memory used as guest mapping can | |
5003 | use copy-on-write semantics as well as dirty pages tracking via read-only page | |
5004 | tables. | |
5005 | ||
5006 | 8.17 KVM_CAP_S390_BPB | |
5007 | ||
5008 | Architectures: s390 | |
5009 | ||
5010 | This capability indicates that kvm will implement the interfaces to handle | |
5011 | reset, migration and nested KVM for branch prediction blocking. The stfle | |
5012 | facility 82 should not be provided to the guest without this capability. | |
c1aea919 | 5013 | |
2ddc6498 | 5014 | 8.18 KVM_CAP_HYPERV_TLBFLUSH |
c1aea919 VK |
5015 | |
5016 | Architectures: x86 | |
5017 | ||
5018 | This capability indicates that KVM supports paravirtualized Hyper-V TLB Flush | |
5019 | hypercalls: | |
5020 | HvFlushVirtualAddressSpace, HvFlushVirtualAddressSpaceEx, | |
5021 | HvFlushVirtualAddressList, HvFlushVirtualAddressListEx. | |
be26b3a7 | 5022 | |
688e0581 | 5023 | 8.19 KVM_CAP_ARM_INJECT_SERROR_ESR |
be26b3a7 DG |
5024 | |
5025 | Architectures: arm, arm64 | |
5026 | ||
5027 | This capability indicates that userspace can specify (via the | |
5028 | KVM_SET_VCPU_EVENTS ioctl) the syndrome value reported to the guest when it | |
5029 | takes a virtual SError interrupt exception. | |
5030 | If KVM advertises this capability, userspace can only specify the ISS field for | |
5031 | the ESR syndrome. Other parts of the ESR, such as the EC are generated by the | |
5032 | CPU when the exception is taken. If this virtual SError is taken to EL1 using | |
5033 | AArch64, this value will be reported in the ISS field of ESR_ELx. | |
5034 | ||
5035 | See KVM_CAP_VCPU_EVENTS for more details. | |
214ff83d VK |
5036 | 8.20 KVM_CAP_HYPERV_SEND_IPI |
5037 | ||
5038 | Architectures: x86 | |
5039 | ||
5040 | This capability indicates that KVM supports paravirtualized Hyper-V IPI send | |
5041 | hypercalls: | |
5042 | HvCallSendSyntheticClusterIpi, HvCallSendSyntheticClusterIpiEx. |