kvm_host.h 28.3 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>
A
Alexey Dobriyan 已提交
26
#include <asm/signal.h>
A
Avi Kivity 已提交
27 28

#include <linux/kvm.h>
I
Ingo Molnar 已提交
29
#include <linux/kvm_para.h>
A
Avi Kivity 已提交
30

31
#include <linux/kvm_types.h>
32

33
#include <asm/kvm_host.h>
34

A
Avi Kivity 已提交
35 36 37 38
#ifndef KVM_MMIO_SIZE
#define KVM_MMIO_SIZE 8
#endif

39 40 41 42 43 44 45
/*
 * 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)

46 47
/* Two fragments for cross MMIO pages. */
#define KVM_MAX_MMIO_FRAGMENTS	2
A
Avi Kivity 已提交
48

49 50
/*
 * For the normal pfn, the highest 12 bits should be zero,
51 52
 * so we can mask bit 62 ~ bit 52  to indicate the error pfn,
 * mask bit 63 to indicate the noslot pfn.
53
 */
54 55 56
#define KVM_PFN_ERR_MASK	(0x7ffULL << 52)
#define KVM_PFN_ERR_NOSLOT_MASK	(0xfffULL << 52)
#define KVM_PFN_NOSLOT		(0x1ULL << 63)
57 58 59

#define KVM_PFN_ERR_FAULT	(KVM_PFN_ERR_MASK)
#define KVM_PFN_ERR_HWPOISON	(KVM_PFN_ERR_MASK + 1)
60
#define KVM_PFN_ERR_RO_FAULT	(KVM_PFN_ERR_MASK + 2)
61

62 63 64 65
/*
 * error pfns indicate that the gfn is in slot but faild to
 * translate it to pfn on host.
 */
66
static inline bool is_error_pfn(pfn_t pfn)
X
Xiao Guangrong 已提交
67
{
68
	return !!(pfn & KVM_PFN_ERR_MASK);
X
Xiao Guangrong 已提交
69 70
}

71 72 73 74 75 76
/*
 * 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
77
{
78
	return !!(pfn & KVM_PFN_ERR_NOSLOT_MASK);
X
Xiao Guangrong 已提交
79 80
}

81 82
/* 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
83
{
84
	return pfn == KVM_PFN_NOSLOT;
X
Xiao Guangrong 已提交
85 86
}

87 88
#define KVM_HVA_ERR_BAD		(PAGE_OFFSET)
#define KVM_HVA_ERR_RO_BAD	(PAGE_OFFSET + PAGE_SIZE)
X
Xiao Guangrong 已提交
89 90 91

static inline bool kvm_is_error_hva(unsigned long addr)
{
92
	return addr >= PAGE_OFFSET;
X
Xiao Guangrong 已提交
93 94
}

95 96
#define KVM_ERR_PTR_BAD_PAGE	(ERR_PTR(-ENOENT))

97
static inline bool is_error_page(struct page *page)
98 99 100 101
{
	return IS_ERR(page);
}

102 103 104
/*
 * vcpu->requests bit members
 */
105
#define KVM_REQ_TLB_FLUSH          0
106
#define KVM_REQ_MIGRATE_TIMER      1
107
#define KVM_REQ_REPORT_TPR_ACCESS  2
108
#define KVM_REQ_MMU_RELOAD         3
J
Joerg Roedel 已提交
109
#define KVM_REQ_TRIPLE_FAULT       4
110
#define KVM_REQ_PENDING_TIMER      5
111
#define KVM_REQ_UNHALT             6
112
#define KVM_REQ_MMU_SYNC           7
Z
Zachary Amsden 已提交
113
#define KVM_REQ_CLOCK_UPDATE       8
114
#define KVM_REQ_KICK               9
115
#define KVM_REQ_DEACTIVATE_FPU    10
116
#define KVM_REQ_EVENT             11
117
#define KVM_REQ_APF_HALT          12
G
Glauber Costa 已提交
118
#define KVM_REQ_STEAL_UPDATE      13
A
Avi Kivity 已提交
119
#define KVM_REQ_NMI               14
120
#define KVM_REQ_IMMEDIATE_EXIT    15
121 122
#define KVM_REQ_PMU               16
#define KVM_REQ_PMI               17
123
#define KVM_REQ_WATCHDOG          18
124 125
#define KVM_REQ_MASTERCLOCK_UPDATE 19
#define KVM_REQ_MCLOCK_INPROGRESS 20
126
#define KVM_REQ_EPR_EXIT          21
127
#define KVM_REQ_EOIBITMAP         22
A
Avi Kivity 已提交
128

129 130
#define KVM_USERSPACE_IRQ_SOURCE_ID		0
#define KVM_IRQFD_RESAMPLE_IRQ_SOURCE_ID	1
131

132
struct kvm;
A
Avi Kivity 已提交
133
struct kvm_vcpu;
134
extern struct kmem_cache *kvm_vcpu_cache;
A
Avi Kivity 已提交
135

136 137 138 139 140 141
struct kvm_io_range {
	gpa_t addr;
	int len;
	struct kvm_io_device *dev;
};

142
#define NR_IOBUS_DEVS 1000
143

144 145
struct kvm_io_bus {
	int                   dev_count;
146
	struct kvm_io_range range[];
147 148
};

M
Marcelo Tosatti 已提交
149 150 151 152 153 154 155 156 157
enum kvm_bus {
	KVM_MMIO_BUS,
	KVM_PIO_BUS,
	KVM_NR_BUSES
};

int kvm_io_bus_write(struct kvm *kvm, enum kvm_bus bus_idx, gpa_t addr,
		     int len, const void *val);
int kvm_io_bus_read(struct kvm *kvm, enum kvm_bus bus_idx, gpa_t addr, int len,
158
		    void *val);
159 160
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
161 162
int kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx,
			      struct kvm_io_device *dev);
163

164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181
#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;
	struct page *page;
	bool done;
};

void kvm_clear_async_pf_completion_queue(struct kvm_vcpu *vcpu);
void kvm_check_async_pf_completion(struct kvm_vcpu *vcpu);
int kvm_setup_async_pf(struct kvm_vcpu *vcpu, gva_t gva, gfn_t gfn,
		       struct kvm_arch_async_pf *arch);
182
int kvm_async_pf_wakeup_all(struct kvm_vcpu *vcpu);
183 184
#endif

185 186 187
enum {
	OUTSIDE_GUEST_MODE,
	IN_GUEST_MODE,
188 189
	EXITING_GUEST_MODE,
	READING_SHADOW_PAGE_TABLES,
190 191
};

A
Avi Kivity 已提交
192 193 194 195 196 197 198 199 200 201
/*
 * 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;
};

202 203
struct kvm_vcpu {
	struct kvm *kvm;
204
#ifdef CONFIG_PREEMPT_NOTIFIERS
205
	struct preempt_notifier preempt_notifier;
206
#endif
207
	int cpu;
208
	int vcpu_id;
209 210
	int srcu_idx;
	int mode;
211
	unsigned long requests;
J
Jan Kiszka 已提交
212
	unsigned long guest_debug;
213 214 215

	struct mutex mutex;
	struct kvm_run *run;
216

217
	int fpu_active;
218
	int guest_fpu_loaded, guest_xcr0_loaded;
219
	wait_queue_head_t wq;
220
	struct pid *pid;
221 222 223 224
	int sigset_active;
	sigset_t sigset;
	struct kvm_vcpu_stat stat;

225
#ifdef CONFIG_HAS_IOMEM
226 227 228
	int mmio_needed;
	int mmio_read_completed;
	int mmio_is_write;
A
Avi Kivity 已提交
229 230 231
	int mmio_cur_fragment;
	int mmio_nr_fragments;
	struct kvm_mmio_fragment mmio_fragments[KVM_MAX_MMIO_FRAGMENTS];
232
#endif
A
Avi Kivity 已提交
233

234 235 236 237 238 239 240 241 242
#ifdef CONFIG_KVM_ASYNC_PF
	struct {
		u32 queued;
		struct list_head queue;
		struct list_head done;
		spinlock_t lock;
	} async_pf;
#endif

243 244 245 246 247 248 249 250 251 252 253 254
#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
255 256 257
	struct kvm_vcpu_arch arch;
};

258 259 260 261 262
static inline int kvm_vcpu_exiting_guest_mode(struct kvm_vcpu *vcpu)
{
	return cmpxchg(&vcpu->mode, IN_GUEST_MODE, EXITING_GUEST_MODE);
}

263 264 265 266 267 268
/*
 * 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
269 270 271 272
struct kvm_memory_slot {
	gfn_t base_gfn;
	unsigned long npages;
	unsigned long *dirty_bitmap;
273
	struct kvm_arch_memory_slot arch;
274
	unsigned long userspace_addr;
275
	u32 flags;
276
	short id;
A
Avi Kivity 已提交
277 278
};

279 280 281 282 283
static inline unsigned long kvm_dirty_bitmap_bytes(struct kvm_memory_slot *memslot)
{
	return ALIGN(memslot->npages, BITS_PER_LONG) / 8;
}

284 285
struct kvm_kernel_irq_routing_entry {
	u32 gsi;
286
	u32 type;
287
	int (*set)(struct kvm_kernel_irq_routing_entry *e,
288
		   struct kvm *kvm, int irq_source_id, int level);
289 290 291 292 293
	union {
		struct {
			unsigned irqchip;
			unsigned pin;
		} irqchip;
S
Sheng Yang 已提交
294
		struct msi_msg msi;
295
	};
296 297 298
	struct hlist_node link;
};

299 300
#ifdef __KVM_HAVE_IOAPIC

301
struct kvm_irq_routing_table {
302
	int chip[KVM_NR_IRQCHIPS][KVM_IOAPIC_NUM_PINS];
303 304 305 306 307 308 309
	struct kvm_kernel_irq_routing_entry *rt_entries;
	u32 nr_rt_entries;
	/*
	 * Array indexed by gsi. Each entry contains list of irq chips
	 * the gsi is connected to.
	 */
	struct hlist_head map[0];
310 311
};

312 313 314 315 316 317
#else

struct kvm_irq_routing_table {};

#endif

318 319 320 321
#ifndef KVM_PRIVATE_MEM_SLOTS
#define KVM_PRIVATE_MEM_SLOTS 0
#endif

322
#ifndef KVM_MEM_SLOTS_NUM
323
#define KVM_MEM_SLOTS_NUM (KVM_USER_MEM_SLOTS + KVM_PRIVATE_MEM_SLOTS)
324 325
#endif

326 327 328 329 330
/*
 * Note:
 * memslots are not sorted by id anymore, please use id_to_memslot()
 * to get the memslot by its id.
 */
331
struct kvm_memslots {
332
	u64 generation;
333
	struct kvm_memory_slot memslots[KVM_MEM_SLOTS_NUM];
334
	/* The mapping table from slot id to the index in memslots[]. */
335
	short id_to_index[KVM_MEM_SLOTS_NUM];
336 337
};

A
Avi Kivity 已提交
338
struct kvm {
339
	spinlock_t mmu_lock;
340
	struct mutex slots_lock;
341
	struct mm_struct *mm; /* userspace tied to this vm */
342
	struct kvm_memslots *memslots;
343
	struct srcu_struct srcu;
344 345 346
#ifdef CONFIG_KVM_APIC_ARCHITECTURE
	u32 bsp_vcpu_id;
#endif
R
Rusty Russell 已提交
347
	struct kvm_vcpu *vcpus[KVM_MAX_VCPUS];
348
	atomic_t online_vcpus;
349
	int last_boosted_vcpu;
350
	struct list_head vm_list;
351
	struct mutex lock;
M
Marcelo Tosatti 已提交
352
	struct kvm_io_bus *buses[KVM_NR_BUSES];
G
Gregory Haskins 已提交
353 354 355 356
#ifdef CONFIG_HAVE_KVM_EVENTFD
	struct {
		spinlock_t        lock;
		struct list_head  items;
357 358
		struct list_head  resampler_list;
		struct mutex      resampler_lock;
G
Gregory Haskins 已提交
359
	} irqfds;
G
Gregory Haskins 已提交
360
	struct list_head ioeventfds;
G
Gregory Haskins 已提交
361
#endif
362
	struct kvm_vm_stat stat;
363
	struct kvm_arch arch;
I
Izik Eidus 已提交
364
	atomic_t users_count;
365 366
#ifdef KVM_COALESCED_MMIO_PAGE_OFFSET
	struct kvm_coalesced_mmio_ring *coalesced_mmio_ring;
367 368
	spinlock_t ring_lock;
	struct list_head coalesced_zones;
369
#endif
370

371
	struct mutex irq_lock;
372
#ifdef CONFIG_HAVE_KVM_IRQCHIP
373 374 375 376
	/*
	 * Update side is protected by irq_lock and,
	 * if configured, irqfds.lock.
	 */
A
Arnd Bergmann 已提交
377
	struct kvm_irq_routing_table __rcu *irq_routing;
378
	struct hlist_head mask_notifier_list;
379
	struct hlist_head irq_ack_notifier_list;
380 381
#endif

382
#if defined(CONFIG_MMU_NOTIFIER) && defined(KVM_ARCH_WANT_MMU_NOTIFIER)
383 384 385 386
	struct mmu_notifier mmu_notifier;
	unsigned long mmu_notifier_seq;
	long mmu_notifier_count;
#endif
387
	long tlbs_dirty;
A
Avi Kivity 已提交
388 389
};

390 391 392 393 394 395 396 397 398
#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__)
399

400 401 402
/* 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
403

404 405 406 407 408 409 410
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) \
411 412 413 414
	for (idx = 0; \
	     idx < atomic_read(&kvm->online_vcpus) && \
	     (vcpup = kvm_get_vcpu(kvm, idx)) != NULL; \
	     idx++)
415

416 417
#define kvm_for_each_memslot(memslot, slots)	\
	for (memslot = &slots->memslots[0];	\
418 419
	      memslot < slots->memslots + KVM_MEM_SLOTS_NUM && memslot->npages;\
		memslot++)
420

R
Rusty Russell 已提交
421 422 423
int kvm_vcpu_init(struct kvm_vcpu *vcpu, struct kvm *kvm, unsigned id);
void kvm_vcpu_uninit(struct kvm_vcpu *vcpu);

424
int __must_check vcpu_load(struct kvm_vcpu *vcpu);
425 426
void vcpu_put(struct kvm_vcpu *vcpu);

427
int kvm_init(void *opaque, unsigned vcpu_size, unsigned vcpu_align,
428
		  struct module *module);
429
void kvm_exit(void);
A
Avi Kivity 已提交
430

I
Izik Eidus 已提交
431 432
void kvm_get_kvm(struct kvm *kvm);
void kvm_put_kvm(struct kvm *kvm);
433 434
void update_memslots(struct kvm_memslots *slots, struct kvm_memory_slot *new,
		     u64 last_generation);
I
Izik Eidus 已提交
435

436 437 438 439 440 441 442
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));
}

443 444 445
static inline struct kvm_memory_slot *
id_to_memslot(struct kvm_memslots *slots, int id)
{
446 447
	int index = slots->id_to_index[id];
	struct kvm_memory_slot *slot;
448

449
	slot = &slots->memslots[index];
450

451 452
	WARN_ON(slot->id != id);
	return slot;
453 454
}

455 456
int kvm_set_memory_region(struct kvm *kvm,
			  struct kvm_userspace_memory_region *mem,
457
			  bool user_alloc);
458 459
int __kvm_set_memory_region(struct kvm *kvm,
			    struct kvm_userspace_memory_region *mem,
460
			    bool user_alloc);
461 462 463
void kvm_arch_free_memslot(struct kvm_memory_slot *free,
			   struct kvm_memory_slot *dont);
int kvm_arch_create_memslot(struct kvm_memory_slot *slot, unsigned long npages);
464 465 466
int kvm_arch_prepare_memory_region(struct kvm *kvm,
				struct kvm_memory_slot *memslot,
				struct kvm_memory_slot old,
467
				struct kvm_userspace_memory_region *mem);
468
void kvm_arch_commit_memory_region(struct kvm *kvm,
469
				struct kvm_userspace_memory_region *mem,
470
				struct kvm_memory_slot old);
471
bool kvm_largepages_enabled(void);
472
void kvm_disable_largepages(void);
473 474 475 476 477
/* 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);
478

479 480 481
int gfn_to_page_many_atomic(struct kvm *kvm, gfn_t gfn, struct page **pages,
			    int nr_pages);

A
Avi Kivity 已提交
482
struct page *gfn_to_page(struct kvm *kvm, gfn_t gfn);
M
Marcelo Tosatti 已提交
483
unsigned long gfn_to_hva(struct kvm *kvm, gfn_t gfn);
X
Xiao Guangrong 已提交
484
unsigned long gfn_to_hva_memslot(struct kvm_memory_slot *slot, gfn_t gfn);
485 486
void kvm_release_page_clean(struct page *page);
void kvm_release_page_dirty(struct page *page);
487 488 489
void kvm_set_page_dirty(struct page *page);
void kvm_set_page_accessed(struct page *page);

490
pfn_t gfn_to_pfn_atomic(struct kvm *kvm, gfn_t gfn);
491 492
pfn_t gfn_to_pfn_async(struct kvm *kvm, gfn_t gfn, bool *async,
		       bool write_fault, bool *writable);
493
pfn_t gfn_to_pfn(struct kvm *kvm, gfn_t gfn);
494 495
pfn_t gfn_to_pfn_prot(struct kvm *kvm, gfn_t gfn, bool write_fault,
		      bool *writable);
496
pfn_t gfn_to_pfn_memslot(struct kvm_memory_slot *slot, gfn_t gfn);
497 498
pfn_t gfn_to_pfn_memslot_atomic(struct kvm_memory_slot *slot, gfn_t gfn);

499
void kvm_release_pfn_dirty(pfn_t pfn);
500 501 502 503 504
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);

505 506
int kvm_read_guest_page(struct kvm *kvm, gfn_t gfn, void *data, int offset,
			int len);
507 508
int kvm_read_guest_atomic(struct kvm *kvm, gpa_t gpa, void *data,
			  unsigned long len);
509
int kvm_read_guest(struct kvm *kvm, gpa_t gpa, void *data, unsigned long len);
510 511
int kvm_read_guest_cached(struct kvm *kvm, struct gfn_to_hva_cache *ghc,
			   void *data, unsigned long len);
512 513 514 515
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);
516 517 518 519
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,
			      gpa_t gpa);
520 521
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
522
struct kvm_memory_slot *gfn_to_memslot(struct kvm *kvm, gfn_t gfn);
523
int kvm_is_visible_gfn(struct kvm *kvm, gfn_t gfn);
J
Joerg Roedel 已提交
524
unsigned long kvm_host_page_size(struct kvm *kvm, gfn_t gfn);
A
Avi Kivity 已提交
525
void mark_page_dirty(struct kvm *kvm, gfn_t gfn);
526 527
void mark_page_dirty_in_slot(struct kvm *kvm, struct kvm_memory_slot *memslot,
			     gfn_t gfn);
A
Avi Kivity 已提交
528

529
void kvm_vcpu_block(struct kvm_vcpu *vcpu);
530
void kvm_vcpu_kick(struct kvm_vcpu *vcpu);
531
bool kvm_vcpu_yield_to(struct kvm_vcpu *target);
Z
Zhai, Edwin 已提交
532
void kvm_vcpu_on_spin(struct kvm_vcpu *vcpu);
A
Avi Kivity 已提交
533
void kvm_resched(struct kvm_vcpu *vcpu);
534 535
void kvm_load_guest_fpu(struct kvm_vcpu *vcpu);
void kvm_put_guest_fpu(struct kvm_vcpu *vcpu);
536

537
void kvm_flush_remote_tlbs(struct kvm *kvm);
538
void kvm_reload_remote_mmus(struct kvm *kvm);
539
void kvm_make_mclock_inprogress_request(struct kvm *kvm);
540
void kvm_make_update_eoibitmap_request(struct kvm *kvm);
A
Avi Kivity 已提交
541

542 543
long kvm_arch_dev_ioctl(struct file *filp,
			unsigned int ioctl, unsigned long arg);
544 545
long kvm_arch_vcpu_ioctl(struct file *filp,
			 unsigned int ioctl, unsigned long arg);
546
int kvm_arch_vcpu_fault(struct kvm_vcpu *vcpu, struct vm_fault *vmf);
547 548 549

int kvm_dev_ioctl_check_extension(long ext);

550 551 552 553 554
int kvm_get_dirty_log(struct kvm *kvm,
			struct kvm_dirty_log *log, int *is_dirty);
int kvm_vm_ioctl_get_dirty_log(struct kvm *kvm,
				struct kvm_dirty_log *log);

555 556 557
int kvm_vm_ioctl_set_memory_region(struct kvm *kvm,
				   struct
				   kvm_userspace_memory_region *mem,
558
				   bool user_alloc);
559
int kvm_vm_ioctl_irq_line(struct kvm *kvm, struct kvm_irq_level *irq_level);
560 561
long kvm_arch_vm_ioctl(struct file *filp,
		       unsigned int ioctl, unsigned long arg);
562

563 564 565
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);

566 567 568
int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
				    struct kvm_translation *tr);

569 570 571 572 573 574
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);
575 576 577 578
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
579 580
int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
					struct kvm_guest_debug *dbg);
581 582
int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run);

583 584
int kvm_arch_init(void *opaque);
void kvm_arch_exit(void);
585

586 587 588 589 590 591 592
int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu);
void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu);

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);
593
int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu);
594
int kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu);
595
void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu);
596

597
int kvm_arch_hardware_enable(void *garbage);
598 599 600 601
void kvm_arch_hardware_disable(void *garbage);
int kvm_arch_hardware_setup(void);
void kvm_arch_hardware_unsetup(void);
void kvm_arch_check_processor_compat(void *rtn);
602
int kvm_arch_vcpu_runnable(struct kvm_vcpu *vcpu);
603
int kvm_arch_vcpu_should_kick(struct kvm_vcpu *vcpu);
604

605 606
void kvm_free_physmem(struct kvm *kvm);

607 608 609
void *kvm_kvzalloc(unsigned long size);
void kvm_kvfree(const void *addr);

610 611 612 613 614 615 616 617 618 619 620 621
#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

622 623
static inline wait_queue_head_t *kvm_arch_vcpu_wq(struct kvm_vcpu *vcpu)
{
624 625 626
#ifdef __KVM_HAVE_ARCH_WQP
	return vcpu->arch.wqp;
#else
627 628
	return &vcpu->wq;
#endif
629
}
630

631
int kvm_arch_init_vm(struct kvm *kvm, unsigned long type);
632
void kvm_arch_destroy_vm(struct kvm *kvm);
633
void kvm_free_all_assigned_devices(struct kvm *kvm);
634
void kvm_arch_sync_events(struct kvm *kvm);
635

636
int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu);
637
void kvm_vcpu_kick(struct kvm_vcpu *vcpu);
638

X
Xiao Guangrong 已提交
639
bool kvm_is_mmio_pfn(pfn_t pfn);
640

B
Ben-Ami Yassour 已提交
641 642 643 644 645 646 647 648 649 650
struct kvm_irq_ack_notifier {
	struct hlist_node link;
	unsigned gsi;
	void (*irq_acked)(struct kvm_irq_ack_notifier *kian);
};

struct kvm_assigned_dev_kernel {
	struct kvm_irq_ack_notifier ack_notifier;
	struct list_head list;
	int assigned_dev_id;
651
	int host_segnr;
B
Ben-Ami Yassour 已提交
652 653
	int host_busnr;
	int host_devfn;
S
Sheng Yang 已提交
654
	unsigned int entries_nr;
B
Ben-Ami Yassour 已提交
655
	int host_irq;
656
	bool host_irq_disabled;
657
	bool pci_2_3;
S
Sheng Yang 已提交
658
	struct msix_entry *host_msix_entries;
B
Ben-Ami Yassour 已提交
659
	int guest_irq;
660
	struct msix_entry *guest_msix_entries;
661
	unsigned long irq_requested_type;
662
	int irq_source_id;
663
	int flags;
B
Ben-Ami Yassour 已提交
664 665
	struct pci_dev *dev;
	struct kvm *kvm;
666
	spinlock_t intx_lock;
667
	spinlock_t intx_mask_lock;
668
	char irq_name[32];
669
	struct pci_saved_state *pci_saved_state;
B
Ben-Ami Yassour 已提交
670
};
671 672 673 674 675 676 677 678 679 680 681

struct kvm_irq_mask_notifier {
	void (*func)(struct kvm_irq_mask_notifier *kimn, bool masked);
	int irq;
	struct hlist_node link;
};

void kvm_register_irq_mask_notifier(struct kvm *kvm, int irq,
				    struct kvm_irq_mask_notifier *kimn);
void kvm_unregister_irq_mask_notifier(struct kvm *kvm, int irq,
				      struct kvm_irq_mask_notifier *kimn);
682 683
void kvm_fire_mask_notifiers(struct kvm *kvm, unsigned irqchip, unsigned pin,
			     bool mask);
684

685 686 687 688 689 690
#ifdef __KVM_HAVE_IOAPIC
void kvm_get_intr_delivery_bitmask(struct kvm_ioapic *ioapic,
				   union kvm_ioapic_redirect_entry *entry,
				   unsigned long *deliver_bitmask);
#endif
int kvm_set_irq(struct kvm *kvm, int irq_source_id, u32 irq, int level);
M
Michael S. Tsirkin 已提交
691
int kvm_set_irq_inatomic(struct kvm *kvm, int irq_source_id, u32 irq, int level);
692 693
int kvm_set_msi(struct kvm_kernel_irq_routing_entry *irq_entry, struct kvm *kvm,
		int irq_source_id, int level);
694
bool kvm_irq_has_notifier(struct kvm *kvm, unsigned irqchip, unsigned pin);
695
void kvm_notify_acked_irq(struct kvm *kvm, unsigned irqchip, unsigned pin);
696 697
void kvm_register_irq_ack_notifier(struct kvm *kvm,
				   struct kvm_irq_ack_notifier *kian);
698 699
void kvm_unregister_irq_ack_notifier(struct kvm *kvm,
				   struct kvm_irq_ack_notifier *kian);
700 701
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
702

703 704 705
/* For vcpu->arch.iommu_flags */
#define KVM_IOMMU_CACHE_COHERENCY	0x1

J
Joerg Roedel 已提交
706
#ifdef CONFIG_IOMMU_API
707
int kvm_iommu_map_pages(struct kvm *kvm, struct kvm_memory_slot *slot);
708
void kvm_iommu_unmap_pages(struct kvm *kvm, struct kvm_memory_slot *slot);
W
Weidong Han 已提交
709
int kvm_iommu_map_guest(struct kvm *kvm);
B
Ben-Ami Yassour 已提交
710
int kvm_iommu_unmap_guest(struct kvm *kvm);
W
Weidong Han 已提交
711 712
int kvm_assign_device(struct kvm *kvm,
		      struct kvm_assigned_dev_kernel *assigned_dev);
W
Weidong Han 已提交
713 714
int kvm_deassign_device(struct kvm *kvm,
			struct kvm_assigned_dev_kernel *assigned_dev);
J
Joerg Roedel 已提交
715
#else /* CONFIG_IOMMU_API */
B
Ben-Ami Yassour 已提交
716
static inline int kvm_iommu_map_pages(struct kvm *kvm,
717
				      struct kvm_memory_slot *slot)
B
Ben-Ami Yassour 已提交
718 719 720 721
{
	return 0;
}

722 723 724 725 726
static inline void kvm_iommu_unmap_pages(struct kvm *kvm,
					 struct kvm_memory_slot *slot)
{
}

W
Weidong Han 已提交
727
static inline int kvm_iommu_map_guest(struct kvm *kvm)
B
Ben-Ami Yassour 已提交
728 729 730 731 732 733 734 735
{
	return -ENODEV;
}

static inline int kvm_iommu_unmap_guest(struct kvm *kvm)
{
	return 0;
}
W
Weidong Han 已提交
736 737 738 739 740 741

static inline int kvm_assign_device(struct kvm *kvm,
		struct kvm_assigned_dev_kernel *assigned_dev)
{
	return 0;
}
W
Weidong Han 已提交
742 743 744 745 746 747

static inline int kvm_deassign_device(struct kvm *kvm,
		struct kvm_assigned_dev_kernel *assigned_dev)
{
	return 0;
}
J
Joerg Roedel 已提交
748
#endif /* CONFIG_IOMMU_API */
B
Ben-Ami Yassour 已提交
749

750
static inline void __guest_enter(void)
751
{
752 753 754 755
	/*
	 * This is running in ioctl context so we can avoid
	 * the call to vtime_account() with its unnecessary idle check.
	 */
756
	vtime_account_system(current);
757
	current->flags |= PF_VCPU;
758 759
}

760
static inline void __guest_exit(void)
761 762 763 764 765 766 767 768 769
{
	/*
	 * This is running in ioctl context so we can avoid
	 * the call to vtime_account() with its unnecessary idle check.
	 */
	vtime_account_system(current);
	current->flags &= ~PF_VCPU;
}

770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785
#ifdef CONFIG_CONTEXT_TRACKING
extern void guest_enter(void);
extern void guest_exit(void);

#else /* !CONFIG_CONTEXT_TRACKING */
static inline void guest_enter(void)
{
	__guest_enter();
}

static inline void guest_exit(void)
{
	__guest_exit();
}
#endif /* !CONFIG_CONTEXT_TRACKING */

786 787 788 789 790 791 792 793 794 795
static inline void kvm_guest_enter(void)
{
	unsigned long flags;

	BUG_ON(preemptible());

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

796 797 798 799 800 801 802 803
	/* 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
	 * is very similar to exiting to userspase from rcu point of view. In
	 * 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());
804 805 806 807
}

static inline void kvm_guest_exit(void)
{
808 809 810 811 812
	unsigned long flags;

	local_irq_save(flags);
	guest_exit();
	local_irq_restore(flags);
813 814
}

815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839
/*
 * 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)
{
	struct kvm_memory_slot *memslot;

	kvm_for_each_memslot(memslot, slots)
		if (gfn >= memslot->base_gfn &&
		      gfn < memslot->base_gfn + memslot->npages)
			return memslot;

	return NULL;
}

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

840 841 842 843 844 845
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;
}

846 847 848 849 850
static inline int memslot_id(struct kvm *kvm, gfn_t gfn)
{
	return gfn_to_memslot(kvm, gfn)->id;
}

851 852 853 854 855 856 857
static inline gfn_t gfn_to_index(gfn_t gfn, gfn_t base_gfn, int level)
{
	/* KVM_HPAGE_GFN_SHIFT(PT_PAGE_TABLE_LEVEL) must be 0. */
	return (gfn >> KVM_HPAGE_GFN_SHIFT(level)) -
		(base_gfn >> KVM_HPAGE_GFN_SHIFT(level));
}

858 859
static inline gfn_t
hva_to_gfn_memslot(unsigned long hva, struct kvm_memory_slot *slot)
860
{
861 862 863
	gfn_t gfn_offset = (hva - slot->userspace_addr) >> PAGE_SHIFT;

	return slot->base_gfn + gfn_offset;
864 865
}

A
Avi Kivity 已提交
866 867 868 869
static inline gpa_t gfn_to_gpa(gfn_t gfn)
{
	return (gpa_t)gfn << PAGE_SHIFT;
}
A
Avi Kivity 已提交
870

871 872 873 874 875
static inline gfn_t gpa_to_gfn(gpa_t gpa)
{
	return (gfn_t)(gpa >> PAGE_SHIFT);
}

B
Ben-Ami Yassour 已提交
876 877 878 879 880
static inline hpa_t pfn_to_hpa(pfn_t pfn)
{
	return (hpa_t)pfn << PAGE_SHIFT;
}

M
Marcelo Tosatti 已提交
881
static inline void kvm_migrate_timers(struct kvm_vcpu *vcpu)
882 883 884 885
{
	set_bit(KVM_REQ_MIGRATE_TIMER, &vcpu->requests);
}

886 887 888 889 890
enum kvm_stat_kind {
	KVM_STAT_VM,
	KVM_STAT_VCPU,
};

891 892 893
struct kvm_stats_debugfs_item {
	const char *name;
	int offset;
894
	enum kvm_stat_kind kind;
895 896 897
	struct dentry *dentry;
};
extern struct kvm_stats_debugfs_item debugfs_entries[];
898
extern struct dentry *kvm_debugfs_dir;
899

900
#if defined(CONFIG_MMU_NOTIFIER) && defined(KVM_ARCH_WANT_MMU_NOTIFIER)
901
static inline int mmu_notifier_retry(struct kvm *kvm, unsigned long mmu_seq)
902
{
903
	if (unlikely(kvm->mmu_notifier_count))
904 905
		return 1;
	/*
906 907 908 909 910 911 912 913
	 * 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.
914
	 */
915
	smp_rmb();
916
	if (kvm->mmu_notifier_seq != mmu_seq)
917 918 919 920 921
		return 1;
	return 0;
}
#endif

922
#ifdef KVM_CAP_IRQ_ROUTING
923 924 925 926 927 928 929 930 931 932

#define KVM_MAX_IRQ_ROUTES 1024

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);
void kvm_free_irq_routing(struct kvm *kvm);

933 934
int kvm_send_userspace_msi(struct kvm *kvm, struct kvm_msi *msi);

935 936 937 938 939 940
#else

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

#endif

G
Gregory Haskins 已提交
941 942
#ifdef CONFIG_HAVE_KVM_EVENTFD

G
Gregory Haskins 已提交
943
void kvm_eventfd_init(struct kvm *kvm);
944 945 946
int kvm_ioeventfd(struct kvm *kvm, struct kvm_ioeventfd *args);

#ifdef CONFIG_HAVE_KVM_IRQCHIP
947
int kvm_irqfd(struct kvm *kvm, struct kvm_irqfd *args);
G
Gregory Haskins 已提交
948
void kvm_irqfd_release(struct kvm *kvm);
949
void kvm_irq_routing_update(struct kvm *, struct kvm_irq_routing_table *);
950 951 952 953 954 955 956 957
#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
958 959 960

#else

G
Gregory Haskins 已提交
961
static inline void kvm_eventfd_init(struct kvm *kvm) {}
962

963
static inline int kvm_irqfd(struct kvm *kvm, struct kvm_irqfd *args)
G
Gregory Haskins 已提交
964 965 966 967 968
{
	return -EINVAL;
}

static inline void kvm_irqfd_release(struct kvm *kvm) {}
969

A
Alexander Graf 已提交
970
#ifdef CONFIG_HAVE_KVM_IRQCHIP
971 972 973 974 975
static inline void kvm_irq_routing_update(struct kvm *kvm,
					  struct kvm_irq_routing_table *irq_rt)
{
	rcu_assign_pointer(kvm->irq_routing, irq_rt);
}
A
Alexander Graf 已提交
976
#endif
977

G
Gregory Haskins 已提交
978 979 980 981
static inline int kvm_ioeventfd(struct kvm *kvm, struct kvm_ioeventfd *args)
{
	return -ENOSYS;
}
G
Gregory Haskins 已提交
982 983 984

#endif /* CONFIG_HAVE_KVM_EVENTFD */

985
#ifdef CONFIG_KVM_APIC_ARCHITECTURE
986 987
static inline bool kvm_vcpu_is_bsp(struct kvm_vcpu *vcpu)
{
988
	return vcpu->kvm->bsp_vcpu_id == vcpu->vcpu_id;
989
}
990 991 992 993 994 995 996

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
997
#endif
998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011

#ifdef __KVM_HAVE_DEVICE_ASSIGNMENT

long kvm_vm_ioctl_assigned_device(struct kvm *kvm, unsigned ioctl,
				  unsigned long arg);

#else

static inline long kvm_vm_ioctl_assigned_device(struct kvm *kvm, unsigned ioctl,
						unsigned long arg)
{
	return -ENOTTY;
}

1012
#endif
1013

1014 1015 1016 1017 1018 1019 1020
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)
{
1021 1022 1023 1024 1025 1026
	if (test_bit(req, &vcpu->requests)) {
		clear_bit(req, &vcpu->requests);
		return true;
	} else {
		return false;
	}
1027 1028
}

1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049
#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)
{
}

1050 1051 1052 1053 1054
static inline bool kvm_vcpu_eligible_for_directed_yield(struct kvm_vcpu *vcpu)
{
	return true;
}

1055
#endif /* CONFIG_HAVE_KVM_CPU_RELAX_INTERCEPT */
1056 1057
#endif