kvm_host.h 29.4 KB
Newer Older
1 2
#ifndef __KVM_HOST_H
#define __KVM_HOST_H
A
Avi Kivity 已提交
3 4 5 6 7 8 9

/*
 * This work is licensed under the terms of the GNU GPL, version 2.  See
 * the COPYING file in the top-level directory.
 */

#include <linux/types.h>
10
#include <linux/hardirq.h>
A
Avi Kivity 已提交
11 12 13
#include <linux/list.h>
#include <linux/mutex.h>
#include <linux/spinlock.h>
M
Markus Rechberger 已提交
14 15
#include <linux/signal.h>
#include <linux/sched.h>
16
#include <linux/bug.h>
A
Avi Kivity 已提交
17
#include <linux/mm.h>
18
#include <linux/mmu_notifier.h>
19
#include <linux/preempt.h>
20
#include <linux/msi.h>
21
#include <linux/slab.h>
22
#include <linux/rcupdate.h>
23
#include <linux/ratelimit.h>
X
Xiao Guangrong 已提交
24
#include <linux/err.h>
25
#include <linux/irqflags.h>
26
#include <linux/context_tracking.h>
A
Alexey Dobriyan 已提交
27
#include <asm/signal.h>
A
Avi Kivity 已提交
28 29

#include <linux/kvm.h>
I
Ingo Molnar 已提交
30
#include <linux/kvm_para.h>
A
Avi Kivity 已提交
31

32
#include <linux/kvm_types.h>
33

34
#include <asm/kvm_host.h>
35

36 37 38 39 40 41
/*
 * The bit 16 ~ bit 31 of kvm_memory_region::flags are internally used
 * in kvm, other bits are visible for userspace which are defined in
 * include/linux/kvm_h.
 */
#define KVM_MEMSLOT_INVALID	(1UL << 16)
42
#define KVM_MEMSLOT_INCOHERENT	(1UL << 17)
43

44 45
/* Two fragments for cross MMIO pages. */
#define KVM_MAX_MMIO_FRAGMENTS	2
A
Avi Kivity 已提交
46

47 48
/*
 * For the normal pfn, the highest 12 bits should be zero,
49 50
 * so we can mask bit 62 ~ bit 52  to indicate the error pfn,
 * mask bit 63 to indicate the noslot pfn.
51
 */
52 53 54
#define KVM_PFN_ERR_MASK	(0x7ffULL << 52)
#define KVM_PFN_ERR_NOSLOT_MASK	(0xfffULL << 52)
#define KVM_PFN_NOSLOT		(0x1ULL << 63)
55 56 57

#define KVM_PFN_ERR_FAULT	(KVM_PFN_ERR_MASK)
#define KVM_PFN_ERR_HWPOISON	(KVM_PFN_ERR_MASK + 1)
58
#define KVM_PFN_ERR_RO_FAULT	(KVM_PFN_ERR_MASK + 2)
59

60 61 62 63
/*
 * error pfns indicate that the gfn is in slot but faild to
 * translate it to pfn on host.
 */
64
static inline bool is_error_pfn(pfn_t pfn)
X
Xiao Guangrong 已提交
65
{
66
	return !!(pfn & KVM_PFN_ERR_MASK);
X
Xiao Guangrong 已提交
67 68
}

69 70 71 72 73 74
/*
 * error_noslot pfns indicate that the gfn can not be
 * translated to pfn - it is not in slot or failed to
 * translate it to pfn.
 */
static inline bool is_error_noslot_pfn(pfn_t pfn)
X
Xiao Guangrong 已提交
75
{
76
	return !!(pfn & KVM_PFN_ERR_NOSLOT_MASK);
X
Xiao Guangrong 已提交
77 78
}

79 80
/* noslot pfn indicates that the gfn is not in slot. */
static inline bool is_noslot_pfn(pfn_t pfn)
X
Xiao Guangrong 已提交
81
{
82
	return pfn == KVM_PFN_NOSLOT;
X
Xiao Guangrong 已提交
83 84
}

85 86 87 88 89 90
/*
 * architectures with KVM_HVA_ERR_BAD other than PAGE_OFFSET (e.g. s390)
 * provide own defines and kvm_is_error_hva
 */
#ifndef KVM_HVA_ERR_BAD

91 92
#define KVM_HVA_ERR_BAD		(PAGE_OFFSET)
#define KVM_HVA_ERR_RO_BAD	(PAGE_OFFSET + PAGE_SIZE)
X
Xiao Guangrong 已提交
93 94 95

static inline bool kvm_is_error_hva(unsigned long addr)
{
96
	return addr >= PAGE_OFFSET;
X
Xiao Guangrong 已提交
97 98
}

99 100
#endif

101 102
#define KVM_ERR_PTR_BAD_PAGE	(ERR_PTR(-ENOENT))

103
static inline bool is_error_page(struct page *page)
104 105 106 107
{
	return IS_ERR(page);
}

108 109 110
/*
 * vcpu->requests bit members
 */
111
#define KVM_REQ_TLB_FLUSH          0
112
#define KVM_REQ_MIGRATE_TIMER      1
113
#define KVM_REQ_REPORT_TPR_ACCESS  2
114
#define KVM_REQ_MMU_RELOAD         3
J
Joerg Roedel 已提交
115
#define KVM_REQ_TRIPLE_FAULT       4
116
#define KVM_REQ_PENDING_TIMER      5
117
#define KVM_REQ_UNHALT             6
118
#define KVM_REQ_MMU_SYNC           7
Z
Zachary Amsden 已提交
119
#define KVM_REQ_CLOCK_UPDATE       8
120
#define KVM_REQ_KICK               9
121
#define KVM_REQ_DEACTIVATE_FPU    10
122
#define KVM_REQ_EVENT             11
123
#define KVM_REQ_APF_HALT          12
G
Glauber Costa 已提交
124
#define KVM_REQ_STEAL_UPDATE      13
A
Avi Kivity 已提交
125
#define KVM_REQ_NMI               14
126 127 128 129 130 131 132
#define KVM_REQ_PMU               15
#define KVM_REQ_PMI               16
#define KVM_REQ_WATCHDOG          17
#define KVM_REQ_MASTERCLOCK_UPDATE 18
#define KVM_REQ_MCLOCK_INPROGRESS 19
#define KVM_REQ_EPR_EXIT          20
#define KVM_REQ_SCAN_IOAPIC       21
133
#define KVM_REQ_GLOBAL_CLOCK_UPDATE 22
134 135
#define KVM_REQ_ENABLE_IBS        23
#define KVM_REQ_DISABLE_IBS       24
136
#define KVM_REQ_APIC_PAGE_RELOAD  25
A
Avi Kivity 已提交
137

138 139
#define KVM_USERSPACE_IRQ_SOURCE_ID		0
#define KVM_IRQFD_RESAMPLE_IRQ_SOURCE_ID	1
140

141
extern struct kmem_cache *kvm_vcpu_cache;
A
Avi Kivity 已提交
142

143
extern spinlock_t kvm_lock;
144 145
extern struct list_head vm_list;

146 147 148 149 150 151
struct kvm_io_range {
	gpa_t addr;
	int len;
	struct kvm_io_device *dev;
};

152
#define NR_IOBUS_DEVS 1000
153

154
struct kvm_io_bus {
155 156
	int dev_count;
	int ioeventfd_count;
157
	struct kvm_io_range range[];
158 159
};

M
Marcelo Tosatti 已提交
160 161 162
enum kvm_bus {
	KVM_MMIO_BUS,
	KVM_PIO_BUS,
163
	KVM_VIRTIO_CCW_NOTIFY_BUS,
164
	KVM_FAST_MMIO_BUS,
M
Marcelo Tosatti 已提交
165 166 167
	KVM_NR_BUSES
};

168
int kvm_io_bus_write(struct kvm_vcpu *vcpu, enum kvm_bus bus_idx, gpa_t addr,
M
Marcelo Tosatti 已提交
169
		     int len, const void *val);
170 171 172 173
int kvm_io_bus_write_cookie(struct kvm_vcpu *vcpu, enum kvm_bus bus_idx,
			    gpa_t addr, int len, const void *val, long cookie);
int kvm_io_bus_read(struct kvm_vcpu *vcpu, enum kvm_bus bus_idx, gpa_t addr,
		    int len, void *val);
174 175
int kvm_io_bus_register_dev(struct kvm *kvm, enum kvm_bus bus_idx, gpa_t addr,
			    int len, struct kvm_io_device *dev);
M
Marcelo Tosatti 已提交
176 177
int kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx,
			      struct kvm_io_device *dev);
178

179 180 181 182 183 184 185 186 187 188
#ifdef CONFIG_KVM_ASYNC_PF
struct kvm_async_pf {
	struct work_struct work;
	struct list_head link;
	struct list_head queue;
	struct kvm_vcpu *vcpu;
	struct mm_struct *mm;
	gva_t gva;
	unsigned long addr;
	struct kvm_arch_async_pf arch;
189
	bool   wakeup_all;
190 191 192 193
};

void kvm_clear_async_pf_completion_queue(struct kvm_vcpu *vcpu);
void kvm_check_async_pf_completion(struct kvm_vcpu *vcpu);
194
int kvm_setup_async_pf(struct kvm_vcpu *vcpu, gva_t gva, unsigned long hva,
195
		       struct kvm_arch_async_pf *arch);
196
int kvm_async_pf_wakeup_all(struct kvm_vcpu *vcpu);
197 198
#endif

199 200 201
enum {
	OUTSIDE_GUEST_MODE,
	IN_GUEST_MODE,
202 203
	EXITING_GUEST_MODE,
	READING_SHADOW_PAGE_TABLES,
204 205
};

A
Avi Kivity 已提交
206 207 208 209 210 211 212 213 214 215
/*
 * Sometimes a large or cross-page mmio needs to be broken up into separate
 * exits for userspace servicing.
 */
struct kvm_mmio_fragment {
	gpa_t gpa;
	void *data;
	unsigned len;
};

216 217
struct kvm_vcpu {
	struct kvm *kvm;
218
#ifdef CONFIG_PREEMPT_NOTIFIERS
219
	struct preempt_notifier preempt_notifier;
220
#endif
221
	int cpu;
222
	int vcpu_id;
223 224
	int srcu_idx;
	int mode;
225
	unsigned long requests;
J
Jan Kiszka 已提交
226
	unsigned long guest_debug;
227 228 229

	struct mutex mutex;
	struct kvm_run *run;
230

231
	int fpu_active;
232
	int guest_fpu_loaded, guest_xcr0_loaded;
233
	wait_queue_head_t wq;
234
	struct pid *pid;
235 236 237 238
	int sigset_active;
	sigset_t sigset;
	struct kvm_vcpu_stat stat;

239
#ifdef CONFIG_HAS_IOMEM
240 241 242
	int mmio_needed;
	int mmio_read_completed;
	int mmio_is_write;
A
Avi Kivity 已提交
243 244 245
	int mmio_cur_fragment;
	int mmio_nr_fragments;
	struct kvm_mmio_fragment mmio_fragments[KVM_MAX_MMIO_FRAGMENTS];
246
#endif
A
Avi Kivity 已提交
247

248 249 250 251 252 253 254 255 256
#ifdef CONFIG_KVM_ASYNC_PF
	struct {
		u32 queued;
		struct list_head queue;
		struct list_head done;
		spinlock_t lock;
	} async_pf;
#endif

257 258 259 260 261 262 263 264 265 266 267 268
#ifdef CONFIG_HAVE_KVM_CPU_RELAX_INTERCEPT
	/*
	 * Cpu relax intercept or pause loop exit optimization
	 * in_spin_loop: set when a vcpu does a pause loop exit
	 *  or cpu relax intercepted.
	 * dy_eligible: indicates whether vcpu is eligible for directed yield.
	 */
	struct {
		bool in_spin_loop;
		bool dy_eligible;
	} spin_loop;
#endif
269
	bool preempted;
270 271 272
	struct kvm_vcpu_arch arch;
};

273 274 275 276 277
static inline int kvm_vcpu_exiting_guest_mode(struct kvm_vcpu *vcpu)
{
	return cmpxchg(&vcpu->mode, IN_GUEST_MODE, EXITING_GUEST_MODE);
}

278 279 280 281 282 283
/*
 * Some of the bitops functions do not support too long bitmaps.
 * This number must be determined not to exceed such limits.
 */
#define KVM_MEM_MAX_NR_PAGES ((1UL << 31) - 1)

A
Avi Kivity 已提交
284 285 286 287
struct kvm_memory_slot {
	gfn_t base_gfn;
	unsigned long npages;
	unsigned long *dirty_bitmap;
288
	struct kvm_arch_memory_slot arch;
289
	unsigned long userspace_addr;
290
	u32 flags;
291
	short id;
A
Avi Kivity 已提交
292 293
};

294 295 296 297 298
static inline unsigned long kvm_dirty_bitmap_bytes(struct kvm_memory_slot *memslot)
{
	return ALIGN(memslot->npages, BITS_PER_LONG) / 8;
}

299 300 301 302 303 304 305 306
struct kvm_s390_adapter_int {
	u64 ind_addr;
	u64 summary_addr;
	u64 ind_offset;
	u32 summary_offset;
	u32 adapter_id;
};

307 308
struct kvm_kernel_irq_routing_entry {
	u32 gsi;
309
	u32 type;
310
	int (*set)(struct kvm_kernel_irq_routing_entry *e,
311 312
		   struct kvm *kvm, int irq_source_id, int level,
		   bool line_status);
313 314 315 316 317
	union {
		struct {
			unsigned irqchip;
			unsigned pin;
		} irqchip;
S
Sheng Yang 已提交
318
		struct msi_msg msi;
319
		struct kvm_s390_adapter_int adapter;
320
	};
321 322 323
	struct hlist_node link;
};

324 325 326 327
#ifndef KVM_PRIVATE_MEM_SLOTS
#define KVM_PRIVATE_MEM_SLOTS 0
#endif

328
#ifndef KVM_MEM_SLOTS_NUM
329
#define KVM_MEM_SLOTS_NUM (KVM_USER_MEM_SLOTS + KVM_PRIVATE_MEM_SLOTS)
330 331
#endif

332 333 334 335 336
/*
 * Note:
 * memslots are not sorted by id anymore, please use id_to_memslot()
 * to get the memslot by its id.
 */
337
struct kvm_memslots {
338
	u64 generation;
339
	struct kvm_memory_slot memslots[KVM_MEM_SLOTS_NUM];
340
	/* The mapping table from slot id to the index in memslots[]. */
341
	short id_to_index[KVM_MEM_SLOTS_NUM];
342
	atomic_t lru_slot;
343
	int used_slots;
344 345
};

A
Avi Kivity 已提交
346
struct kvm {
347
	spinlock_t mmu_lock;
348
	struct mutex slots_lock;
349
	struct mm_struct *mm; /* userspace tied to this vm */
350
	struct kvm_memslots *memslots;
351
	struct srcu_struct srcu;
352
	struct srcu_struct irq_srcu;
353 354 355
#ifdef CONFIG_KVM_APIC_ARCHITECTURE
	u32 bsp_vcpu_id;
#endif
R
Rusty Russell 已提交
356
	struct kvm_vcpu *vcpus[KVM_MAX_VCPUS];
357
	atomic_t online_vcpus;
358
	int last_boosted_vcpu;
359
	struct list_head vm_list;
360
	struct mutex lock;
M
Marcelo Tosatti 已提交
361
	struct kvm_io_bus *buses[KVM_NR_BUSES];
G
Gregory Haskins 已提交
362 363 364 365
#ifdef CONFIG_HAVE_KVM_EVENTFD
	struct {
		spinlock_t        lock;
		struct list_head  items;
366 367
		struct list_head  resampler_list;
		struct mutex      resampler_lock;
G
Gregory Haskins 已提交
368
	} irqfds;
G
Gregory Haskins 已提交
369
	struct list_head ioeventfds;
G
Gregory Haskins 已提交
370
#endif
371
	struct kvm_vm_stat stat;
372
	struct kvm_arch arch;
I
Izik Eidus 已提交
373
	atomic_t users_count;
374 375
#ifdef KVM_COALESCED_MMIO_PAGE_OFFSET
	struct kvm_coalesced_mmio_ring *coalesced_mmio_ring;
376 377
	spinlock_t ring_lock;
	struct list_head coalesced_zones;
378
#endif
379

380
	struct mutex irq_lock;
381
#ifdef CONFIG_HAVE_KVM_IRQCHIP
382
	/*
383
	 * Update side is protected by irq_lock.
384
	 */
A
Arnd Bergmann 已提交
385
	struct kvm_irq_routing_table __rcu *irq_routing;
386 387
#endif
#ifdef CONFIG_HAVE_KVM_IRQFD
388
	struct hlist_head irq_ack_notifier_list;
389 390
#endif

391
#if defined(CONFIG_MMU_NOTIFIER) && defined(KVM_ARCH_WANT_MMU_NOTIFIER)
392 393 394 395
	struct mmu_notifier mmu_notifier;
	unsigned long mmu_notifier_seq;
	long mmu_notifier_count;
#endif
396
	long tlbs_dirty;
397
	struct list_head devices;
A
Avi Kivity 已提交
398 399
};

400 401 402 403 404 405 406 407 408
#define kvm_err(fmt, ...) \
	pr_err("kvm [%i]: " fmt, task_pid_nr(current), ## __VA_ARGS__)
#define kvm_info(fmt, ...) \
	pr_info("kvm [%i]: " fmt, task_pid_nr(current), ## __VA_ARGS__)
#define kvm_debug(fmt, ...) \
	pr_debug("kvm [%i]: " fmt, task_pid_nr(current), ## __VA_ARGS__)
#define kvm_pr_unimpl(fmt, ...) \
	pr_err_ratelimited("kvm [%i]: " fmt, \
			   task_tgid_nr(current), ## __VA_ARGS__)
409

410 411 412
/* The guest did something we don't support. */
#define vcpu_unimpl(vcpu, fmt, ...)					\
	kvm_pr_unimpl("vcpu%i " fmt, (vcpu)->vcpu_id, ## __VA_ARGS__)
A
Avi Kivity 已提交
413

414 415 416 417 418 419 420
static inline struct kvm_vcpu *kvm_get_vcpu(struct kvm *kvm, int i)
{
	smp_rmb();
	return kvm->vcpus[i];
}

#define kvm_for_each_vcpu(idx, vcpup, kvm) \
421 422 423 424
	for (idx = 0; \
	     idx < atomic_read(&kvm->online_vcpus) && \
	     (vcpup = kvm_get_vcpu(kvm, idx)) != NULL; \
	     idx++)
425

426 427
#define kvm_for_each_memslot(memslot, slots)	\
	for (memslot = &slots->memslots[0];	\
428 429
	      memslot < slots->memslots + KVM_MEM_SLOTS_NUM && memslot->npages;\
		memslot++)
430

R
Rusty Russell 已提交
431 432 433
int kvm_vcpu_init(struct kvm_vcpu *vcpu, struct kvm *kvm, unsigned id);
void kvm_vcpu_uninit(struct kvm_vcpu *vcpu);

434
int __must_check vcpu_load(struct kvm_vcpu *vcpu);
435 436
void vcpu_put(struct kvm_vcpu *vcpu);

437 438 439 440 441 442 443 444
#ifdef __KVM_HAVE_IOAPIC
void kvm_vcpu_request_scan_ioapic(struct kvm *kvm);
#else
static inline void kvm_vcpu_request_scan_ioapic(struct kvm *kvm)
{
}
#endif

445
#ifdef CONFIG_HAVE_KVM_IRQFD
446 447 448 449 450 451 452 453 454 455 456 457
int kvm_irqfd_init(void);
void kvm_irqfd_exit(void);
#else
static inline int kvm_irqfd_init(void)
{
	return 0;
}

static inline void kvm_irqfd_exit(void)
{
}
#endif
458
int kvm_init(void *opaque, unsigned vcpu_size, unsigned vcpu_align,
459
		  struct module *module);
460
void kvm_exit(void);
A
Avi Kivity 已提交
461

I
Izik Eidus 已提交
462 463 464
void kvm_get_kvm(struct kvm *kvm);
void kvm_put_kvm(struct kvm *kvm);

465 466 467 468 469 470 471
static inline struct kvm_memslots *kvm_memslots(struct kvm *kvm)
{
	return rcu_dereference_check(kvm->memslots,
			srcu_read_lock_held(&kvm->srcu)
			|| lockdep_is_held(&kvm->slots_lock));
}

472 473 474
static inline struct kvm_memory_slot *
id_to_memslot(struct kvm_memslots *slots, int id)
{
475 476
	int index = slots->id_to_index[id];
	struct kvm_memory_slot *slot;
477

478
	slot = &slots->memslots[index];
479

480 481
	WARN_ON(slot->id != id);
	return slot;
482 483
}

484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501
/*
 * KVM_SET_USER_MEMORY_REGION ioctl allows the following operations:
 * - create a new memory slot
 * - delete an existing memory slot
 * - modify an existing memory slot
 *   -- move it in the guest physical memory space
 *   -- just change its flags
 *
 * Since flags can be changed by some of these operations, the following
 * differentiation is the best we can do for __kvm_set_memory_region():
 */
enum kvm_mr_change {
	KVM_MR_CREATE,
	KVM_MR_DELETE,
	KVM_MR_MOVE,
	KVM_MR_FLAGS_ONLY,
};

502
int kvm_set_memory_region(struct kvm *kvm,
503
			  struct kvm_userspace_memory_region *mem);
504
int __kvm_set_memory_region(struct kvm *kvm,
505
			    struct kvm_userspace_memory_region *mem);
506
void kvm_arch_free_memslot(struct kvm *kvm, struct kvm_memory_slot *free,
507
			   struct kvm_memory_slot *dont);
508 509
int kvm_arch_create_memslot(struct kvm *kvm, struct kvm_memory_slot *slot,
			    unsigned long npages);
510
void kvm_arch_memslots_updated(struct kvm *kvm);
511 512 513
int kvm_arch_prepare_memory_region(struct kvm *kvm,
				struct kvm_memory_slot *memslot,
				struct kvm_userspace_memory_region *mem,
514
				enum kvm_mr_change change);
515
void kvm_arch_commit_memory_region(struct kvm *kvm,
516
				struct kvm_userspace_memory_region *mem,
517 518
				const struct kvm_memory_slot *old,
				enum kvm_mr_change change);
519
bool kvm_largepages_enabled(void);
520
void kvm_disable_largepages(void);
521 522 523 524 525
/* flush all memory translations */
void kvm_arch_flush_shadow_all(struct kvm *kvm);
/* flush memory translations pointing to 'slot' */
void kvm_arch_flush_shadow_memslot(struct kvm *kvm,
				   struct kvm_memory_slot *slot);
526

527 528 529
int gfn_to_page_many_atomic(struct kvm *kvm, gfn_t gfn, struct page **pages,
			    int nr_pages);

A
Avi Kivity 已提交
530
struct page *gfn_to_page(struct kvm *kvm, gfn_t gfn);
M
Marcelo Tosatti 已提交
531
unsigned long gfn_to_hva(struct kvm *kvm, gfn_t gfn);
532
unsigned long gfn_to_hva_prot(struct kvm *kvm, gfn_t gfn, bool *writable);
X
Xiao Guangrong 已提交
533
unsigned long gfn_to_hva_memslot(struct kvm_memory_slot *slot, gfn_t gfn);
534 535
unsigned long gfn_to_hva_memslot_prot(struct kvm_memory_slot *slot, gfn_t gfn,
				      bool *writable);
536 537
void kvm_release_page_clean(struct page *page);
void kvm_release_page_dirty(struct page *page);
538 539
void kvm_set_page_accessed(struct page *page);

540
pfn_t gfn_to_pfn_atomic(struct kvm *kvm, gfn_t gfn);
541 542
pfn_t gfn_to_pfn_async(struct kvm *kvm, gfn_t gfn, bool *async,
		       bool write_fault, bool *writable);
543
pfn_t gfn_to_pfn(struct kvm *kvm, gfn_t gfn);
544 545
pfn_t gfn_to_pfn_prot(struct kvm *kvm, gfn_t gfn, bool write_fault,
		      bool *writable);
546
pfn_t gfn_to_pfn_memslot(struct kvm_memory_slot *slot, gfn_t gfn);
547 548
pfn_t gfn_to_pfn_memslot_atomic(struct kvm_memory_slot *slot, gfn_t gfn);

549 550 551 552 553
void kvm_release_pfn_clean(pfn_t pfn);
void kvm_set_pfn_dirty(pfn_t pfn);
void kvm_set_pfn_accessed(pfn_t pfn);
void kvm_get_pfn(pfn_t pfn);

554 555
int kvm_read_guest_page(struct kvm *kvm, gfn_t gfn, void *data, int offset,
			int len);
556 557
int kvm_read_guest_atomic(struct kvm *kvm, gpa_t gpa, void *data,
			  unsigned long len);
558
int kvm_read_guest(struct kvm *kvm, gpa_t gpa, void *data, unsigned long len);
559 560
int kvm_read_guest_cached(struct kvm *kvm, struct gfn_to_hva_cache *ghc,
			   void *data, unsigned long len);
561 562 563 564
int kvm_write_guest_page(struct kvm *kvm, gfn_t gfn, const void *data,
			 int offset, int len);
int kvm_write_guest(struct kvm *kvm, gpa_t gpa, const void *data,
		    unsigned long len);
565 566 567
int kvm_write_guest_cached(struct kvm *kvm, struct gfn_to_hva_cache *ghc,
			   void *data, unsigned long len);
int kvm_gfn_to_hva_cache_init(struct kvm *kvm, struct gfn_to_hva_cache *ghc,
568
			      gpa_t gpa, unsigned long len);
569 570
int kvm_clear_guest_page(struct kvm *kvm, gfn_t gfn, int offset, int len);
int kvm_clear_guest(struct kvm *kvm, gpa_t gpa, unsigned long len);
A
Avi Kivity 已提交
571
struct kvm_memory_slot *gfn_to_memslot(struct kvm *kvm, gfn_t gfn);
572
int kvm_is_visible_gfn(struct kvm *kvm, gfn_t gfn);
J
Joerg Roedel 已提交
573
unsigned long kvm_host_page_size(struct kvm *kvm, gfn_t gfn);
A
Avi Kivity 已提交
574 575
void mark_page_dirty(struct kvm *kvm, gfn_t gfn);

576
void kvm_vcpu_block(struct kvm_vcpu *vcpu);
577
void kvm_vcpu_kick(struct kvm_vcpu *vcpu);
578
int kvm_vcpu_yield_to(struct kvm_vcpu *target);
Z
Zhai, Edwin 已提交
579
void kvm_vcpu_on_spin(struct kvm_vcpu *vcpu);
580 581
void kvm_load_guest_fpu(struct kvm_vcpu *vcpu);
void kvm_put_guest_fpu(struct kvm_vcpu *vcpu);
582

583
void kvm_flush_remote_tlbs(struct kvm *kvm);
584
void kvm_reload_remote_mmus(struct kvm *kvm);
585
void kvm_make_mclock_inprogress_request(struct kvm *kvm);
586
void kvm_make_scan_ioapic_request(struct kvm *kvm);
587
bool kvm_make_all_cpus_request(struct kvm *kvm, unsigned int req);
A
Avi Kivity 已提交
588

589 590
long kvm_arch_dev_ioctl(struct file *filp,
			unsigned int ioctl, unsigned long arg);
591 592
long kvm_arch_vcpu_ioctl(struct file *filp,
			 unsigned int ioctl, unsigned long arg);
593
int kvm_arch_vcpu_fault(struct kvm_vcpu *vcpu, struct vm_fault *vmf);
594

595
int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext);
596

597 598
int kvm_get_dirty_log(struct kvm *kvm,
			struct kvm_dirty_log *log, int *is_dirty);
599 600 601 602

int kvm_get_dirty_log_protect(struct kvm *kvm,
			struct kvm_dirty_log *log, bool *is_dirty);

603
void kvm_arch_mmu_enable_log_dirty_pt_masked(struct kvm *kvm,
604 605 606 607
					struct kvm_memory_slot *slot,
					gfn_t gfn_offset,
					unsigned long mask);

608 609 610
int kvm_vm_ioctl_get_dirty_log(struct kvm *kvm,
				struct kvm_dirty_log *log);

611 612
int kvm_vm_ioctl_irq_line(struct kvm *kvm, struct kvm_irq_level *irq_level,
			bool line_status);
613 614
long kvm_arch_vm_ioctl(struct file *filp,
		       unsigned int ioctl, unsigned long arg);
615

616 617 618
int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu);
int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu);

619 620 621
int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
				    struct kvm_translation *tr);

622 623 624 625 626 627
int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs);
int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs);
int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
				  struct kvm_sregs *sregs);
int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
				  struct kvm_sregs *sregs);
628 629 630 631
int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
				    struct kvm_mp_state *mp_state);
int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
				    struct kvm_mp_state *mp_state);
J
Jan Kiszka 已提交
632 633
int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
					struct kvm_guest_debug *dbg);
634 635
int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run);

636 637
int kvm_arch_init(void *opaque);
void kvm_arch_exit(void);
638

639 640 641
int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu);
void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu);

R
Radim Krčmář 已提交
642 643
void kvm_arch_sched_in(struct kvm_vcpu *vcpu, int cpu);

644 645 646 647
void kvm_arch_vcpu_free(struct kvm_vcpu *vcpu);
void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu);
void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu);
struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm, unsigned int id);
648
int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu);
649
void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu);
650
void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu);
651

652 653
int kvm_arch_hardware_enable(void);
void kvm_arch_hardware_disable(void);
654 655 656
int kvm_arch_hardware_setup(void);
void kvm_arch_hardware_unsetup(void);
void kvm_arch_check_processor_compat(void *rtn);
657
int kvm_arch_vcpu_runnable(struct kvm_vcpu *vcpu);
658
int kvm_arch_vcpu_should_kick(struct kvm_vcpu *vcpu);
659

660 661
void *kvm_kvzalloc(unsigned long size);

662 663 664 665 666 667 668 669 670 671 672 673
#ifndef __KVM_HAVE_ARCH_VM_ALLOC
static inline struct kvm *kvm_arch_alloc_vm(void)
{
	return kzalloc(sizeof(struct kvm), GFP_KERNEL);
}

static inline void kvm_arch_free_vm(struct kvm *kvm)
{
	kfree(kvm);
}
#endif

674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692
#ifdef __KVM_HAVE_ARCH_NONCOHERENT_DMA
void kvm_arch_register_noncoherent_dma(struct kvm *kvm);
void kvm_arch_unregister_noncoherent_dma(struct kvm *kvm);
bool kvm_arch_has_noncoherent_dma(struct kvm *kvm);
#else
static inline void kvm_arch_register_noncoherent_dma(struct kvm *kvm)
{
}

static inline void kvm_arch_unregister_noncoherent_dma(struct kvm *kvm)
{
}

static inline bool kvm_arch_has_noncoherent_dma(struct kvm *kvm)
{
	return false;
}
#endif

693 694
static inline wait_queue_head_t *kvm_arch_vcpu_wq(struct kvm_vcpu *vcpu)
{
695 696 697
#ifdef __KVM_HAVE_ARCH_WQP
	return vcpu->arch.wqp;
#else
698 699
	return &vcpu->wq;
#endif
700
}
701

702 703 704 705 706 707 708 709 710 711 712 713 714 715
#ifdef __KVM_HAVE_ARCH_INTC_INITIALIZED
/*
 * returns true if the virtual interrupt controller is initialized and
 * ready to accept virtual IRQ. On some architectures the virtual interrupt
 * controller is dynamically instantiated and this is not always true.
 */
bool kvm_arch_intc_initialized(struct kvm *kvm);
#else
static inline bool kvm_arch_intc_initialized(struct kvm *kvm)
{
	return true;
}
#endif

716
int kvm_arch_init_vm(struct kvm *kvm, unsigned long type);
717
void kvm_arch_destroy_vm(struct kvm *kvm);
718
void kvm_arch_sync_events(struct kvm *kvm);
719

720
int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu);
721
void kvm_vcpu_kick(struct kvm_vcpu *vcpu);
722

723
bool kvm_is_reserved_pfn(pfn_t pfn);
724

B
Ben-Ami Yassour 已提交
725 726 727 728 729 730
struct kvm_irq_ack_notifier {
	struct hlist_node link;
	unsigned gsi;
	void (*irq_acked)(struct kvm_irq_ack_notifier *kian);
};

731 732 733
int kvm_irq_map_gsi(struct kvm *kvm,
		    struct kvm_kernel_irq_routing_entry *entries, int gsi);
int kvm_irq_map_chip_pin(struct kvm *kvm, unsigned irqchip, unsigned pin);
734

735 736
int kvm_set_irq(struct kvm *kvm, int irq_source_id, u32 irq, int level,
		bool line_status);
M
Michael S. Tsirkin 已提交
737
int kvm_set_irq_inatomic(struct kvm *kvm, int irq_source_id, u32 irq, int level);
738
int kvm_set_msi(struct kvm_kernel_irq_routing_entry *irq_entry, struct kvm *kvm,
739
		int irq_source_id, int level, bool line_status);
740
bool kvm_irq_has_notifier(struct kvm *kvm, unsigned irqchip, unsigned pin);
741
void kvm_notify_acked_irq(struct kvm *kvm, unsigned irqchip, unsigned pin);
742 743
void kvm_register_irq_ack_notifier(struct kvm *kvm,
				   struct kvm_irq_ack_notifier *kian);
744 745
void kvm_unregister_irq_ack_notifier(struct kvm *kvm,
				   struct kvm_irq_ack_notifier *kian);
746 747
int kvm_request_irq_source_id(struct kvm *kvm);
void kvm_free_irq_source_id(struct kvm *kvm, int irq_source_id);
B
Ben-Ami Yassour 已提交
748

749
#ifdef CONFIG_KVM_DEVICE_ASSIGNMENT
750
int kvm_iommu_map_pages(struct kvm *kvm, struct kvm_memory_slot *slot);
751
void kvm_iommu_unmap_pages(struct kvm *kvm, struct kvm_memory_slot *slot);
752
#else
B
Ben-Ami Yassour 已提交
753
static inline int kvm_iommu_map_pages(struct kvm *kvm,
754
				      struct kvm_memory_slot *slot)
B
Ben-Ami Yassour 已提交
755 756 757 758
{
	return 0;
}

759 760 761 762
static inline void kvm_iommu_unmap_pages(struct kvm *kvm,
					 struct kvm_memory_slot *slot)
{
}
763
#endif
B
Ben-Ami Yassour 已提交
764

765 766 767 768 769 770 771 772 773 774
static inline void kvm_guest_enter(void)
{
	unsigned long flags;

	BUG_ON(preemptible());

	local_irq_save(flags);
	guest_enter();
	local_irq_restore(flags);

775 776
	/* KVM does not hold any references to rcu protected data when it
	 * switches CPU into a guest mode. In fact switching to a guest mode
M
Michael S. Tsirkin 已提交
777
	 * is very similar to exiting to userspace from rcu point of view. In
778 779 780 781 782
	 * addition CPU may stay in a guest mode for quite a long time (up to
	 * one time slice). Lets treat guest mode as quiescent state, just like
	 * we do with user-mode execution.
	 */
	rcu_virt_note_context_switch(smp_processor_id());
783 784 785 786
}

static inline void kvm_guest_exit(void)
{
787 788 789 790 791
	unsigned long flags;

	local_irq_save(flags);
	guest_exit();
	local_irq_restore(flags);
792 793
}

794 795 796 797 798 799 800 801 802
/*
 * search_memslots() and __gfn_to_memslot() are here because they are
 * used in non-modular code in arch/powerpc/kvm/book3s_hv_rm_mmu.c.
 * gfn_to_memslot() itself isn't here as an inline because that would
 * bloat other code too much.
 */
static inline struct kvm_memory_slot *
search_memslots(struct kvm_memslots *slots, gfn_t gfn)
{
803
	int start = 0, end = slots->used_slots;
804
	int slot = atomic_read(&slots->lru_slot);
805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824
	struct kvm_memory_slot *memslots = slots->memslots;

	if (gfn >= memslots[slot].base_gfn &&
	    gfn < memslots[slot].base_gfn + memslots[slot].npages)
		return &memslots[slot];

	while (start < end) {
		slot = start + (end - start) / 2;

		if (gfn >= memslots[slot].base_gfn)
			end = slot;
		else
			start = slot + 1;
	}

	if (gfn >= memslots[start].base_gfn &&
	    gfn < memslots[start].base_gfn + memslots[start].npages) {
		atomic_set(&slots->lru_slot, start);
		return &memslots[start];
	}
825 826 827 828 829 830 831 832 833 834

	return NULL;
}

static inline struct kvm_memory_slot *
__gfn_to_memslot(struct kvm_memslots *slots, gfn_t gfn)
{
	return search_memslots(slots, gfn);
}

835 836 837 838 839 840
static inline unsigned long
__gfn_to_hva_memslot(struct kvm_memory_slot *slot, gfn_t gfn)
{
	return slot->userspace_addr + (gfn - slot->base_gfn) * PAGE_SIZE;
}

841 842 843 844 845
static inline int memslot_id(struct kvm *kvm, gfn_t gfn)
{
	return gfn_to_memslot(kvm, gfn)->id;
}

846 847
static inline gfn_t
hva_to_gfn_memslot(unsigned long hva, struct kvm_memory_slot *slot)
848
{
849 850 851
	gfn_t gfn_offset = (hva - slot->userspace_addr) >> PAGE_SHIFT;

	return slot->base_gfn + gfn_offset;
852 853
}

A
Avi Kivity 已提交
854 855 856 857
static inline gpa_t gfn_to_gpa(gfn_t gfn)
{
	return (gpa_t)gfn << PAGE_SHIFT;
}
A
Avi Kivity 已提交
858

859 860 861 862 863
static inline gfn_t gpa_to_gfn(gpa_t gpa)
{
	return (gfn_t)(gpa >> PAGE_SHIFT);
}

B
Ben-Ami Yassour 已提交
864 865 866 867 868
static inline hpa_t pfn_to_hpa(pfn_t pfn)
{
	return (hpa_t)pfn << PAGE_SHIFT;
}

869 870 871 872 873 874 875
static inline bool kvm_is_error_gpa(struct kvm *kvm, gpa_t gpa)
{
	unsigned long hva = gfn_to_hva(kvm, gpa_to_gfn(gpa));

	return kvm_is_error_hva(hva);
}

M
Marcelo Tosatti 已提交
876
static inline void kvm_migrate_timers(struct kvm_vcpu *vcpu)
877 878 879 880
{
	set_bit(KVM_REQ_MIGRATE_TIMER, &vcpu->requests);
}

881 882 883 884 885
enum kvm_stat_kind {
	KVM_STAT_VM,
	KVM_STAT_VCPU,
};

886 887 888
struct kvm_stats_debugfs_item {
	const char *name;
	int offset;
889
	enum kvm_stat_kind kind;
890 891 892
	struct dentry *dentry;
};
extern struct kvm_stats_debugfs_item debugfs_entries[];
893
extern struct dentry *kvm_debugfs_dir;
894

895
#if defined(CONFIG_MMU_NOTIFIER) && defined(KVM_ARCH_WANT_MMU_NOTIFIER)
896
static inline int mmu_notifier_retry(struct kvm *kvm, unsigned long mmu_seq)
897
{
898
	if (unlikely(kvm->mmu_notifier_count))
899 900
		return 1;
	/*
901 902 903 904 905 906 907 908
	 * Ensure the read of mmu_notifier_count happens before the read
	 * of mmu_notifier_seq.  This interacts with the smp_wmb() in
	 * mmu_notifier_invalidate_range_end to make sure that the caller
	 * either sees the old (non-zero) value of mmu_notifier_count or
	 * the new (incremented) value of mmu_notifier_seq.
	 * PowerPC Book3s HV KVM calls this under a per-page lock
	 * rather than under kvm->mmu_lock, for scalability, so
	 * can't rely on kvm->mmu_lock to keep things ordered.
909
	 */
910
	smp_rmb();
911
	if (kvm->mmu_notifier_seq != mmu_seq)
912 913 914 915 916
		return 1;
	return 0;
}
#endif

917
#ifdef CONFIG_HAVE_KVM_IRQ_ROUTING
918

919 920 921
#ifdef CONFIG_S390
#define KVM_MAX_IRQ_ROUTES 4096 //FIXME: we can have more than that...
#else
922
#define KVM_MAX_IRQ_ROUTES 1024
923
#endif
924 925 926 927 928 929

int kvm_setup_default_irq_routing(struct kvm *kvm);
int kvm_set_irq_routing(struct kvm *kvm,
			const struct kvm_irq_routing_entry *entries,
			unsigned nr,
			unsigned flags);
930
int kvm_set_routing_entry(struct kvm_kernel_irq_routing_entry *e,
931
			  const struct kvm_irq_routing_entry *ue);
932 933 934 935 936 937 938 939
void kvm_free_irq_routing(struct kvm *kvm);

#else

static inline void kvm_free_irq_routing(struct kvm *kvm) {}

#endif

940 941
int kvm_send_userspace_msi(struct kvm *kvm, struct kvm_msi *msi);

G
Gregory Haskins 已提交
942 943
#ifdef CONFIG_HAVE_KVM_EVENTFD

G
Gregory Haskins 已提交
944
void kvm_eventfd_init(struct kvm *kvm);
945 946
int kvm_ioeventfd(struct kvm *kvm, struct kvm_ioeventfd *args);

947
#ifdef CONFIG_HAVE_KVM_IRQFD
948
int kvm_irqfd(struct kvm *kvm, struct kvm_irqfd *args);
G
Gregory Haskins 已提交
949
void kvm_irqfd_release(struct kvm *kvm);
950
void kvm_irq_routing_update(struct kvm *);
951 952 953 954 955 956 957 958
#else
static inline int kvm_irqfd(struct kvm *kvm, struct kvm_irqfd *args)
{
	return -EINVAL;
}

static inline void kvm_irqfd_release(struct kvm *kvm) {}
#endif
G
Gregory Haskins 已提交
959 960 961

#else

G
Gregory Haskins 已提交
962
static inline void kvm_eventfd_init(struct kvm *kvm) {}
963

964
static inline int kvm_irqfd(struct kvm *kvm, struct kvm_irqfd *args)
G
Gregory Haskins 已提交
965 966 967 968 969
{
	return -EINVAL;
}

static inline void kvm_irqfd_release(struct kvm *kvm) {}
970

A
Alexander Graf 已提交
971
#ifdef CONFIG_HAVE_KVM_IRQCHIP
972
static inline void kvm_irq_routing_update(struct kvm *kvm)
973 974
{
}
A
Alexander Graf 已提交
975
#endif
976

G
Gregory Haskins 已提交
977 978 979 980
static inline int kvm_ioeventfd(struct kvm *kvm, struct kvm_ioeventfd *args)
{
	return -ENOSYS;
}
G
Gregory Haskins 已提交
981 982 983

#endif /* CONFIG_HAVE_KVM_EVENTFD */

984
#ifdef CONFIG_KVM_APIC_ARCHITECTURE
985 986
static inline bool kvm_vcpu_is_bsp(struct kvm_vcpu *vcpu)
{
987
	return vcpu->kvm->bsp_vcpu_id == vcpu->vcpu_id;
988
}
989 990 991 992 993 994 995

bool kvm_vcpu_compatible(struct kvm_vcpu *vcpu);

#else

static inline bool kvm_vcpu_compatible(struct kvm_vcpu *vcpu) { return true; }

A
Avi Kivity 已提交
996
#endif
997

998 999 1000 1001 1002 1003 1004
static inline void kvm_make_request(int req, struct kvm_vcpu *vcpu)
{
	set_bit(req, &vcpu->requests);
}

static inline bool kvm_check_request(int req, struct kvm_vcpu *vcpu)
{
1005 1006 1007 1008 1009 1010
	if (test_bit(req, &vcpu->requests)) {
		clear_bit(req, &vcpu->requests);
		return true;
	} else {
		return false;
	}
1011 1012
}

1013 1014
extern bool kvm_rebooting;

S
Scott Wood 已提交
1015 1016 1017 1018
struct kvm_device {
	struct kvm_device_ops *ops;
	struct kvm *kvm;
	void *private;
1019
	struct list_head vm_node;
S
Scott Wood 已提交
1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046
};

/* create, destroy, and name are mandatory */
struct kvm_device_ops {
	const char *name;
	int (*create)(struct kvm_device *dev, u32 type);

	/*
	 * Destroy is responsible for freeing dev.
	 *
	 * Destroy may be called before or after destructors are called
	 * on emulated I/O regions, depending on whether a reference is
	 * held by a vcpu or other kvm component that gets destroyed
	 * after the emulated I/O.
	 */
	void (*destroy)(struct kvm_device *dev);

	int (*set_attr)(struct kvm_device *dev, struct kvm_device_attr *attr);
	int (*get_attr)(struct kvm_device *dev, struct kvm_device_attr *attr);
	int (*has_attr)(struct kvm_device *dev, struct kvm_device_attr *attr);
	long (*ioctl)(struct kvm_device *dev, unsigned int ioctl,
		      unsigned long arg);
};

void kvm_device_get(struct kvm_device *dev);
void kvm_device_put(struct kvm_device *dev);
struct kvm_device *kvm_device_from_filp(struct file *filp);
1047
int kvm_register_device_ops(struct kvm_device_ops *ops, u32 type);
1048
void kvm_unregister_device_ops(u32 type);
S
Scott Wood 已提交
1049

1050
extern struct kvm_device_ops kvm_mpic_ops;
1051
extern struct kvm_device_ops kvm_xics_ops;
1052
extern struct kvm_device_ops kvm_arm_vgic_v2_ops;
1053
extern struct kvm_device_ops kvm_arm_vgic_v3_ops;
1054

1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075
#ifdef CONFIG_HAVE_KVM_CPU_RELAX_INTERCEPT

static inline void kvm_vcpu_set_in_spin_loop(struct kvm_vcpu *vcpu, bool val)
{
	vcpu->spin_loop.in_spin_loop = val;
}
static inline void kvm_vcpu_set_dy_eligible(struct kvm_vcpu *vcpu, bool val)
{
	vcpu->spin_loop.dy_eligible = val;
}

#else /* !CONFIG_HAVE_KVM_CPU_RELAX_INTERCEPT */

static inline void kvm_vcpu_set_in_spin_loop(struct kvm_vcpu *vcpu, bool val)
{
}

static inline void kvm_vcpu_set_dy_eligible(struct kvm_vcpu *vcpu, bool val)
{
}
#endif /* CONFIG_HAVE_KVM_CPU_RELAX_INTERCEPT */
1076
#endif
1077