kvm_host.h 36.5 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/vmalloc.h>
23
#include <linux/rcupdate.h>
24
#include <linux/ratelimit.h>
X
Xiao Guangrong 已提交
25
#include <linux/err.h>
26
#include <linux/irqflags.h>
27
#include <linux/context_tracking.h>
28
#include <linux/irqbypass.h>
29
#include <linux/swait.h>
30
#include <linux/refcount.h>
A
Alexey Dobriyan 已提交
31
#include <asm/signal.h>
A
Avi Kivity 已提交
32 33

#include <linux/kvm.h>
I
Ingo Molnar 已提交
34
#include <linux/kvm_para.h>
A
Avi Kivity 已提交
35

36
#include <linux/kvm_types.h>
37

38
#include <asm/kvm_host.h>
39

G
Greg Kurz 已提交
40 41 42 43
#ifndef KVM_MAX_VCPU_ID
#define KVM_MAX_VCPU_ID KVM_MAX_VCPUS
#endif

44 45 46 47 48 49 50
/*
 * 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)

51 52
/* Two fragments for cross MMIO pages. */
#define KVM_MAX_MMIO_FRAGMENTS	2
A
Avi Kivity 已提交
53

54 55 56 57
#ifndef KVM_ADDRESS_SPACE_NUM
#define KVM_ADDRESS_SPACE_NUM	1
#endif

58 59
/*
 * For the normal pfn, the highest 12 bits should be zero,
60 61
 * so we can mask bit 62 ~ bit 52  to indicate the error pfn,
 * mask bit 63 to indicate the noslot pfn.
62
 */
63 64 65
#define KVM_PFN_ERR_MASK	(0x7ffULL << 52)
#define KVM_PFN_ERR_NOSLOT_MASK	(0xfffULL << 52)
#define KVM_PFN_NOSLOT		(0x1ULL << 63)
66 67 68

#define KVM_PFN_ERR_FAULT	(KVM_PFN_ERR_MASK)
#define KVM_PFN_ERR_HWPOISON	(KVM_PFN_ERR_MASK + 1)
69
#define KVM_PFN_ERR_RO_FAULT	(KVM_PFN_ERR_MASK + 2)
70

71 72 73 74
/*
 * error pfns indicate that the gfn is in slot but faild to
 * translate it to pfn on host.
 */
D
Dan Williams 已提交
75
static inline bool is_error_pfn(kvm_pfn_t pfn)
X
Xiao Guangrong 已提交
76
{
77
	return !!(pfn & KVM_PFN_ERR_MASK);
X
Xiao Guangrong 已提交
78 79
}

80 81 82 83 84
/*
 * 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.
 */
D
Dan Williams 已提交
85
static inline bool is_error_noslot_pfn(kvm_pfn_t pfn)
X
Xiao Guangrong 已提交
86
{
87
	return !!(pfn & KVM_PFN_ERR_NOSLOT_MASK);
X
Xiao Guangrong 已提交
88 89
}

90
/* noslot pfn indicates that the gfn is not in slot. */
D
Dan Williams 已提交
91
static inline bool is_noslot_pfn(kvm_pfn_t pfn)
X
Xiao Guangrong 已提交
92
{
93
	return pfn == KVM_PFN_NOSLOT;
X
Xiao Guangrong 已提交
94 95
}

96 97 98 99 100 101
/*
 * 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

102 103
#define KVM_HVA_ERR_BAD		(PAGE_OFFSET)
#define KVM_HVA_ERR_RO_BAD	(PAGE_OFFSET + PAGE_SIZE)
X
Xiao Guangrong 已提交
104 105 106

static inline bool kvm_is_error_hva(unsigned long addr)
{
107
	return addr >= PAGE_OFFSET;
X
Xiao Guangrong 已提交
108 109
}

110 111
#endif

112 113
#define KVM_ERR_PTR_BAD_PAGE	(ERR_PTR(-ENOENT))

114
static inline bool is_error_page(struct page *page)
115 116 117 118
{
	return IS_ERR(page);
}

119 120
#define KVM_REQUEST_MASK           GENMASK(7,0)
#define KVM_REQUEST_NO_WAKEUP      BIT(8)
121
#define KVM_REQUEST_WAIT           BIT(9)
122
/*
123 124
 * Architecture-independent vcpu->requests bit members
 * Bits 4-7 are reserved for more arch-independent bits.
125
 */
126 127 128 129
#define KVM_REQ_TLB_FLUSH         (0 | KVM_REQUEST_WAIT | KVM_REQUEST_NO_WAKEUP)
#define KVM_REQ_MMU_RELOAD        (1 | KVM_REQUEST_WAIT | KVM_REQUEST_NO_WAKEUP)
#define KVM_REQ_PENDING_TIMER     2
#define KVM_REQ_UNHALT            3
130 131 132 133 134 135 136
#define KVM_REQUEST_ARCH_BASE     8

#define KVM_ARCH_REQ_FLAGS(nr, flags) ({ \
	BUILD_BUG_ON((unsigned)(nr) >= 32 - KVM_REQUEST_ARCH_BASE); \
	(unsigned)(((nr) + KVM_REQUEST_ARCH_BASE) | (flags)); \
})
#define KVM_ARCH_REQ(nr)           KVM_ARCH_REQ_FLAGS(nr, 0)
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);
176 177
void kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx,
			       struct kvm_io_device *dev);
178 179
struct kvm_io_device *kvm_io_bus_get_dev(struct kvm *kvm, enum kvm_bus bus_idx,
					 gpa_t addr);
180

181 182 183 184 185 186 187 188 189 190
#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;
191
	bool   wakeup_all;
192 193 194 195
};

void kvm_clear_async_pf_completion_queue(struct kvm_vcpu *vcpu);
void kvm_check_async_pf_completion(struct kvm_vcpu *vcpu);
196
int kvm_setup_async_pf(struct kvm_vcpu *vcpu, gva_t gva, unsigned long hva,
197
		       struct kvm_arch_async_pf *arch);
198
int kvm_async_pf_wakeup_all(struct kvm_vcpu *vcpu);
199 200
#endif

201 202 203
enum {
	OUTSIDE_GUEST_MODE,
	IN_GUEST_MODE,
204 205
	EXITING_GUEST_MODE,
	READING_SHADOW_PAGE_TABLES,
206 207
};

A
Avi Kivity 已提交
208 209 210 211 212 213 214 215 216 217
/*
 * 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;
};

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

230 231 232
	int pre_pcpu;
	struct list_head blocked_vcpu_list;

233 234
	struct mutex mutex;
	struct kvm_run *run;
235

236
	int guest_xcr0_loaded;
237
	struct swait_queue_head wq;
238
	struct pid __rcu *pid;
239 240 241
	int sigset_active;
	sigset_t sigset;
	struct kvm_vcpu_stat stat;
W
Wanpeng Li 已提交
242
	unsigned int halt_poll_ns;
243
	bool valid_wakeup;
244

245
#ifdef CONFIG_HAS_IOMEM
246 247 248
	int mmio_needed;
	int mmio_read_completed;
	int mmio_is_write;
A
Avi Kivity 已提交
249 250 251
	int mmio_cur_fragment;
	int mmio_nr_fragments;
	struct kvm_mmio_fragment mmio_fragments[KVM_MAX_MMIO_FRAGMENTS];
252
#endif
A
Avi Kivity 已提交
253

254 255 256 257 258 259 260 261 262
#ifdef CONFIG_KVM_ASYNC_PF
	struct {
		u32 queued;
		struct list_head queue;
		struct list_head done;
		spinlock_t lock;
	} async_pf;
#endif

263 264 265 266 267 268 269 270 271 272 273 274
#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
275
	bool preempted;
276
	struct kvm_vcpu_arch arch;
277
	struct dentry *debugfs_dentry;
278 279
};

280 281
static inline int kvm_vcpu_exiting_guest_mode(struct kvm_vcpu *vcpu)
{
282 283 284 285 286 287
	/*
	 * The memory barrier ensures a previous write to vcpu->requests cannot
	 * be reordered with the read of vcpu->mode.  It pairs with the general
	 * memory barrier following the write of vcpu->mode in VCPU RUN.
	 */
	smp_mb__before_atomic();
288 289 290
	return cmpxchg(&vcpu->mode, IN_GUEST_MODE, EXITING_GUEST_MODE);
}

291 292 293 294 295 296
/*
 * 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
297 298 299 300
struct kvm_memory_slot {
	gfn_t base_gfn;
	unsigned long npages;
	unsigned long *dirty_bitmap;
301
	struct kvm_arch_memory_slot arch;
302
	unsigned long userspace_addr;
303
	u32 flags;
304
	short id;
A
Avi Kivity 已提交
305 306
};

307 308 309 310 311
static inline unsigned long kvm_dirty_bitmap_bytes(struct kvm_memory_slot *memslot)
{
	return ALIGN(memslot->npages, BITS_PER_LONG) / 8;
}

312 313 314 315 316 317 318
static inline unsigned long *kvm_second_dirty_bitmap(struct kvm_memory_slot *memslot)
{
	unsigned long len = kvm_dirty_bitmap_bytes(memslot);

	return memslot->dirty_bitmap + len / sizeof(*memslot->dirty_bitmap);
}

319 320 321 322 323 324 325 326
struct kvm_s390_adapter_int {
	u64 ind_addr;
	u64 summary_addr;
	u64 ind_offset;
	u32 summary_offset;
	u32 adapter_id;
};

327 328 329 330 331
struct kvm_hv_sint {
	u32 vcpu;
	u32 sint;
};

332 333
struct kvm_kernel_irq_routing_entry {
	u32 gsi;
334
	u32 type;
335
	int (*set)(struct kvm_kernel_irq_routing_entry *e,
336 337
		   struct kvm *kvm, int irq_source_id, int level,
		   bool line_status);
338 339 340 341 342
	union {
		struct {
			unsigned irqchip;
			unsigned pin;
		} irqchip;
343 344 345 346 347 348 349
		struct {
			u32 address_lo;
			u32 address_hi;
			u32 data;
			u32 flags;
			u32 devid;
		} msi;
350
		struct kvm_s390_adapter_int adapter;
351
		struct kvm_hv_sint hv_sint;
352
	};
353 354 355
	struct hlist_node link;
};

356 357 358 359 360 361 362 363 364 365 366 367
#ifdef CONFIG_HAVE_KVM_IRQ_ROUTING
struct kvm_irq_routing_table {
	int chip[KVM_NR_IRQCHIPS][KVM_IRQCHIP_NUM_PINS];
	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];
};
#endif

368 369 370 371
#ifndef KVM_PRIVATE_MEM_SLOTS
#define KVM_PRIVATE_MEM_SLOTS 0
#endif

372
#ifndef KVM_MEM_SLOTS_NUM
373
#define KVM_MEM_SLOTS_NUM (KVM_USER_MEM_SLOTS + KVM_PRIVATE_MEM_SLOTS)
374 375
#endif

376 377 378 379 380 381 382
#ifndef __KVM_VCPU_MULTIPLE_ADDRESS_SPACE
static inline int kvm_arch_vcpu_memslots_id(struct kvm_vcpu *vcpu)
{
	return 0;
}
#endif

383 384 385 386 387
/*
 * Note:
 * memslots are not sorted by id anymore, please use id_to_memslot()
 * to get the memslot by its id.
 */
388
struct kvm_memslots {
389
	u64 generation;
390
	struct kvm_memory_slot memslots[KVM_MEM_SLOTS_NUM];
391
	/* The mapping table from slot id to the index in memslots[]. */
392
	short id_to_index[KVM_MEM_SLOTS_NUM];
393
	atomic_t lru_slot;
394
	int used_slots;
395 396
};

A
Avi Kivity 已提交
397
struct kvm {
398
	spinlock_t mmu_lock;
399
	struct mutex slots_lock;
400
	struct mm_struct *mm; /* userspace tied to this vm */
401
	struct kvm_memslots __rcu *memslots[KVM_ADDRESS_SPACE_NUM];
R
Rusty Russell 已提交
402
	struct kvm_vcpu *vcpus[KVM_MAX_VCPUS];
403 404 405 406 407 408 409

	/*
	 * created_vcpus is protected by kvm->lock, and is incremented
	 * at the beginning of KVM_CREATE_VCPU.  online_vcpus is only
	 * incremented after storing the kvm_vcpu pointer in vcpus,
	 * and is accessed atomically.
	 */
410
	atomic_t online_vcpus;
411
	int created_vcpus;
412
	int last_boosted_vcpu;
413
	struct list_head vm_list;
414
	struct mutex lock;
415
	struct kvm_io_bus __rcu *buses[KVM_NR_BUSES];
G
Gregory Haskins 已提交
416 417 418 419
#ifdef CONFIG_HAVE_KVM_EVENTFD
	struct {
		spinlock_t        lock;
		struct list_head  items;
420 421
		struct list_head  resampler_list;
		struct mutex      resampler_lock;
G
Gregory Haskins 已提交
422
	} irqfds;
G
Gregory Haskins 已提交
423
	struct list_head ioeventfds;
G
Gregory Haskins 已提交
424
#endif
425
	struct kvm_vm_stat stat;
426
	struct kvm_arch arch;
427
	refcount_t users_count;
428
#ifdef CONFIG_KVM_MMIO
429
	struct kvm_coalesced_mmio_ring *coalesced_mmio_ring;
430 431
	spinlock_t ring_lock;
	struct list_head coalesced_zones;
432
#endif
433

434
	struct mutex irq_lock;
435
#ifdef CONFIG_HAVE_KVM_IRQCHIP
436
	/*
437
	 * Update side is protected by irq_lock.
438
	 */
A
Arnd Bergmann 已提交
439
	struct kvm_irq_routing_table __rcu *irq_routing;
440 441
#endif
#ifdef CONFIG_HAVE_KVM_IRQFD
442
	struct hlist_head irq_ack_notifier_list;
443 444
#endif

445
#if defined(CONFIG_MMU_NOTIFIER) && defined(KVM_ARCH_WANT_MMU_NOTIFIER)
446 447 448 449
	struct mmu_notifier mmu_notifier;
	unsigned long mmu_notifier_seq;
	long mmu_notifier_count;
#endif
450
	long tlbs_dirty;
451
	struct list_head devices;
452 453
	struct dentry *debugfs_dentry;
	struct kvm_stat_data **debugfs_stat_data;
454 455
	struct srcu_struct srcu;
	struct srcu_struct irq_srcu;
456
	pid_t userspace_pid;
A
Avi Kivity 已提交
457 458
};

459 460 461 462 463 464
#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__)
465 466 467
#define kvm_debug_ratelimited(fmt, ...) \
	pr_debug_ratelimited("kvm [%i]: " fmt, task_pid_nr(current), \
			     ## __VA_ARGS__)
468 469 470
#define kvm_pr_unimpl(fmt, ...) \
	pr_err_ratelimited("kvm [%i]: " fmt, \
			   task_tgid_nr(current), ## __VA_ARGS__)
471

472 473
/* The guest did something we don't support. */
#define vcpu_unimpl(vcpu, fmt, ...)					\
474 475
	kvm_pr_unimpl("vcpu%i, guest rIP: 0x%lx " fmt,			\
			(vcpu)->vcpu_id, kvm_rip_read(vcpu), ## __VA_ARGS__)
A
Avi Kivity 已提交
476

477 478
#define vcpu_debug(vcpu, fmt, ...)					\
	kvm_debug("vcpu%i " fmt, (vcpu)->vcpu_id, ## __VA_ARGS__)
479 480 481
#define vcpu_debug_ratelimited(vcpu, fmt, ...)				\
	kvm_debug_ratelimited("vcpu%i " fmt, (vcpu)->vcpu_id,           \
			      ## __VA_ARGS__)
482 483
#define vcpu_err(vcpu, fmt, ...)					\
	kvm_err("vcpu%i " fmt, (vcpu)->vcpu_id, ## __VA_ARGS__)
484

485 486 487
static inline struct kvm_io_bus *kvm_get_bus(struct kvm *kvm, enum kvm_bus idx)
{
	return srcu_dereference_check(kvm->buses[idx], &kvm->srcu,
488 489
				      lockdep_is_held(&kvm->slots_lock) ||
				      !refcount_read(&kvm->users_count));
490 491
}

492 493
static inline struct kvm_vcpu *kvm_get_vcpu(struct kvm *kvm, int i)
{
494 495 496 497
	/* Pairs with smp_wmb() in kvm_vm_ioctl_create_vcpu, in case
	 * the caller has read kvm->online_vcpus before (as is the case
	 * for kvm_for_each_vcpu, for example).
	 */
498 499 500 501 502
	smp_rmb();
	return kvm->vcpus[i];
}

#define kvm_for_each_vcpu(idx, vcpup, kvm) \
503 504 505 506
	for (idx = 0; \
	     idx < atomic_read(&kvm->online_vcpus) && \
	     (vcpup = kvm_get_vcpu(kvm, idx)) != NULL; \
	     idx++)
507

508 509
static inline struct kvm_vcpu *kvm_get_vcpu_by_id(struct kvm *kvm, int id)
{
510
	struct kvm_vcpu *vcpu = NULL;
511 512
	int i;

513
	if (id < 0)
514
		return NULL;
515 516
	if (id < KVM_MAX_VCPUS)
		vcpu = kvm_get_vcpu(kvm, id);
517 518
	if (vcpu && vcpu->vcpu_id == id)
		return vcpu;
519 520 521 522 523 524
	kvm_for_each_vcpu(i, vcpu, kvm)
		if (vcpu->vcpu_id == id)
			return vcpu;
	return NULL;
}

525 526 527 528 529 530 531 532 533 534 535
static inline int kvm_vcpu_get_idx(struct kvm_vcpu *vcpu)
{
	struct kvm_vcpu *tmp;
	int idx;

	kvm_for_each_vcpu(idx, tmp, vcpu->kvm)
		if (tmp == vcpu)
			return idx;
	BUG();
}

536 537
#define kvm_for_each_memslot(memslot, slots)	\
	for (memslot = &slots->memslots[0];	\
538 539
	      memslot < slots->memslots + KVM_MEM_SLOTS_NUM && memslot->npages;\
		memslot++)
540

R
Rusty Russell 已提交
541 542 543
int kvm_vcpu_init(struct kvm_vcpu *vcpu, struct kvm *kvm, unsigned id);
void kvm_vcpu_uninit(struct kvm_vcpu *vcpu);

544
void vcpu_load(struct kvm_vcpu *vcpu);
545 546
void vcpu_put(struct kvm_vcpu *vcpu);

547
#ifdef __KVM_HAVE_IOAPIC
548
void kvm_arch_post_irq_ack_notifier_list_update(struct kvm *kvm);
549
void kvm_arch_post_irq_routing_update(struct kvm *kvm);
550
#else
551
static inline void kvm_arch_post_irq_ack_notifier_list_update(struct kvm *kvm)
552 553
{
}
554
static inline void kvm_arch_post_irq_routing_update(struct kvm *kvm)
555 556
{
}
557 558
#endif

559
#ifdef CONFIG_HAVE_KVM_IRQFD
560 561 562 563 564 565 566 567 568 569 570 571
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
572
int kvm_init(void *opaque, unsigned vcpu_size, unsigned vcpu_align,
573
		  struct module *module);
574
void kvm_exit(void);
A
Avi Kivity 已提交
575

I
Izik Eidus 已提交
576 577 578
void kvm_get_kvm(struct kvm *kvm);
void kvm_put_kvm(struct kvm *kvm);

579
static inline struct kvm_memslots *__kvm_memslots(struct kvm *kvm, int as_id)
580
{
581
	return srcu_dereference_check(kvm->memslots[as_id], &kvm->srcu,
582 583
			lockdep_is_held(&kvm->slots_lock) ||
			!refcount_read(&kvm->users_count));
584 585
}

586 587 588 589 590
static inline struct kvm_memslots *kvm_memslots(struct kvm *kvm)
{
	return __kvm_memslots(kvm, 0);
}

591 592
static inline struct kvm_memslots *kvm_vcpu_memslots(struct kvm_vcpu *vcpu)
{
593 594 595
	int as_id = kvm_arch_vcpu_memslots_id(vcpu);

	return __kvm_memslots(vcpu->kvm, as_id);
596 597
}

598 599 600
static inline struct kvm_memory_slot *
id_to_memslot(struct kvm_memslots *slots, int id)
{
601 602
	int index = slots->id_to_index[id];
	struct kvm_memory_slot *slot;
603

604
	slot = &slots->memslots[index];
605

606 607
	WARN_ON(slot->id != id);
	return slot;
608 609
}

610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627
/*
 * 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,
};

628
int kvm_set_memory_region(struct kvm *kvm,
629
			  const struct kvm_userspace_memory_region *mem);
630
int __kvm_set_memory_region(struct kvm *kvm,
631
			    const struct kvm_userspace_memory_region *mem);
632
void kvm_arch_free_memslot(struct kvm *kvm, struct kvm_memory_slot *free,
633
			   struct kvm_memory_slot *dont);
634 635
int kvm_arch_create_memslot(struct kvm *kvm, struct kvm_memory_slot *slot,
			    unsigned long npages);
636
void kvm_arch_memslots_updated(struct kvm *kvm, struct kvm_memslots *slots);
637 638
int kvm_arch_prepare_memory_region(struct kvm *kvm,
				struct kvm_memory_slot *memslot,
639
				const struct kvm_userspace_memory_region *mem,
640
				enum kvm_mr_change change);
641
void kvm_arch_commit_memory_region(struct kvm *kvm,
642
				const struct kvm_userspace_memory_region *mem,
643
				const struct kvm_memory_slot *old,
644
				const struct kvm_memory_slot *new,
645
				enum kvm_mr_change change);
646
bool kvm_largepages_enabled(void);
647
void kvm_disable_largepages(void);
648 649 650 651 652
/* 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);
653

654 655
int gfn_to_page_many_atomic(struct kvm_memory_slot *slot, gfn_t gfn,
			    struct page **pages, int nr_pages);
656

A
Avi Kivity 已提交
657
struct page *gfn_to_page(struct kvm *kvm, gfn_t gfn);
M
Marcelo Tosatti 已提交
658
unsigned long gfn_to_hva(struct kvm *kvm, gfn_t gfn);
659
unsigned long gfn_to_hva_prot(struct kvm *kvm, gfn_t gfn, bool *writable);
X
Xiao Guangrong 已提交
660
unsigned long gfn_to_hva_memslot(struct kvm_memory_slot *slot, gfn_t gfn);
661 662
unsigned long gfn_to_hva_memslot_prot(struct kvm_memory_slot *slot, gfn_t gfn,
				      bool *writable);
663 664
void kvm_release_page_clean(struct page *page);
void kvm_release_page_dirty(struct page *page);
665 666
void kvm_set_page_accessed(struct page *page);

D
Dan Williams 已提交
667 668 669
kvm_pfn_t gfn_to_pfn_atomic(struct kvm *kvm, gfn_t gfn);
kvm_pfn_t gfn_to_pfn(struct kvm *kvm, gfn_t gfn);
kvm_pfn_t gfn_to_pfn_prot(struct kvm *kvm, gfn_t gfn, bool write_fault,
670
		      bool *writable);
D
Dan Williams 已提交
671 672 673 674 675
kvm_pfn_t gfn_to_pfn_memslot(struct kvm_memory_slot *slot, gfn_t gfn);
kvm_pfn_t gfn_to_pfn_memslot_atomic(struct kvm_memory_slot *slot, gfn_t gfn);
kvm_pfn_t __gfn_to_pfn_memslot(struct kvm_memory_slot *slot, gfn_t gfn,
			       bool atomic, bool *async, bool write_fault,
			       bool *writable);
676

D
Dan Williams 已提交
677
void kvm_release_pfn_clean(kvm_pfn_t pfn);
678
void kvm_release_pfn_dirty(kvm_pfn_t pfn);
D
Dan Williams 已提交
679 680 681
void kvm_set_pfn_dirty(kvm_pfn_t pfn);
void kvm_set_pfn_accessed(kvm_pfn_t pfn);
void kvm_get_pfn(kvm_pfn_t pfn);
682

683 684
int kvm_read_guest_page(struct kvm *kvm, gfn_t gfn, void *data, int offset,
			int len);
685 686
int kvm_read_guest_atomic(struct kvm *kvm, gpa_t gpa, void *data,
			  unsigned long len);
687
int kvm_read_guest(struct kvm *kvm, gpa_t gpa, void *data, unsigned long len);
688 689
int kvm_read_guest_cached(struct kvm *kvm, struct gfn_to_hva_cache *ghc,
			   void *data, unsigned long len);
690 691 692 693
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);
694 695 696 697 698 699
int kvm_write_guest_cached(struct kvm *kvm, struct gfn_to_hva_cache *ghc,
			   void *data, unsigned long len);
int kvm_write_guest_offset_cached(struct kvm *kvm, struct gfn_to_hva_cache *ghc,
			   void *data, int offset, unsigned long len);
int kvm_gfn_to_hva_cache_init(struct kvm *kvm, struct gfn_to_hva_cache *ghc,
			      gpa_t gpa, unsigned long len);
700 701
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
702
struct kvm_memory_slot *gfn_to_memslot(struct kvm *kvm, gfn_t gfn);
703
bool kvm_is_visible_gfn(struct kvm *kvm, gfn_t gfn);
J
Joerg Roedel 已提交
704
unsigned long kvm_host_page_size(struct kvm *kvm, gfn_t gfn);
A
Avi Kivity 已提交
705 706
void mark_page_dirty(struct kvm *kvm, gfn_t gfn);

707 708
struct kvm_memslots *kvm_vcpu_memslots(struct kvm_vcpu *vcpu);
struct kvm_memory_slot *kvm_vcpu_gfn_to_memslot(struct kvm_vcpu *vcpu, gfn_t gfn);
D
Dan Williams 已提交
709 710
kvm_pfn_t kvm_vcpu_gfn_to_pfn_atomic(struct kvm_vcpu *vcpu, gfn_t gfn);
kvm_pfn_t kvm_vcpu_gfn_to_pfn(struct kvm_vcpu *vcpu, gfn_t gfn);
711 712 713 714 715 716 717 718 719 720 721 722 723 724 725
struct page *kvm_vcpu_gfn_to_page(struct kvm_vcpu *vcpu, gfn_t gfn);
unsigned long kvm_vcpu_gfn_to_hva(struct kvm_vcpu *vcpu, gfn_t gfn);
unsigned long kvm_vcpu_gfn_to_hva_prot(struct kvm_vcpu *vcpu, gfn_t gfn, bool *writable);
int kvm_vcpu_read_guest_page(struct kvm_vcpu *vcpu, gfn_t gfn, void *data, int offset,
			     int len);
int kvm_vcpu_read_guest_atomic(struct kvm_vcpu *vcpu, gpa_t gpa, void *data,
			       unsigned long len);
int kvm_vcpu_read_guest(struct kvm_vcpu *vcpu, gpa_t gpa, void *data,
			unsigned long len);
int kvm_vcpu_write_guest_page(struct kvm_vcpu *vcpu, gfn_t gfn, const void *data,
			      int offset, int len);
int kvm_vcpu_write_guest(struct kvm_vcpu *vcpu, gpa_t gpa, const void *data,
			 unsigned long len);
void kvm_vcpu_mark_page_dirty(struct kvm_vcpu *vcpu, gfn_t gfn);

726 727 728
void kvm_sigset_activate(struct kvm_vcpu *vcpu);
void kvm_sigset_deactivate(struct kvm_vcpu *vcpu);

729
void kvm_vcpu_block(struct kvm_vcpu *vcpu);
730 731
void kvm_arch_vcpu_blocking(struct kvm_vcpu *vcpu);
void kvm_arch_vcpu_unblocking(struct kvm_vcpu *vcpu);
732
bool kvm_vcpu_wake_up(struct kvm_vcpu *vcpu);
733
void kvm_vcpu_kick(struct kvm_vcpu *vcpu);
734
int kvm_vcpu_yield_to(struct kvm_vcpu *target);
735
void kvm_vcpu_on_spin(struct kvm_vcpu *vcpu, bool usermode_vcpu_not_eligible);
736 737
void kvm_load_guest_fpu(struct kvm_vcpu *vcpu);
void kvm_put_guest_fpu(struct kvm_vcpu *vcpu);
738

739
void kvm_flush_remote_tlbs(struct kvm *kvm);
740
void kvm_reload_remote_mmus(struct kvm *kvm);
741 742 743

bool kvm_make_vcpus_request_mask(struct kvm *kvm, unsigned int req,
				 unsigned long *vcpu_bitmap, cpumask_var_t tmp);
744
bool kvm_make_all_cpus_request(struct kvm *kvm, unsigned int req);
A
Avi Kivity 已提交
745

746 747
long kvm_arch_dev_ioctl(struct file *filp,
			unsigned int ioctl, unsigned long arg);
748 749
long kvm_arch_vcpu_ioctl(struct file *filp,
			 unsigned int ioctl, unsigned long arg);
750
vm_fault_t kvm_arch_vcpu_fault(struct kvm_vcpu *vcpu, struct vm_fault *vmf);
751

752
int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext);
753

754 755
int kvm_get_dirty_log(struct kvm *kvm,
			struct kvm_dirty_log *log, int *is_dirty);
756 757 758 759

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

760
void kvm_arch_mmu_enable_log_dirty_pt_masked(struct kvm *kvm,
761 762 763 764
					struct kvm_memory_slot *slot,
					gfn_t gfn_offset,
					unsigned long mask);

765 766 767
int kvm_vm_ioctl_get_dirty_log(struct kvm *kvm,
				struct kvm_dirty_log *log);

768 769
int kvm_vm_ioctl_irq_line(struct kvm *kvm, struct kvm_irq_level *irq_level,
			bool line_status);
770 771
long kvm_arch_vm_ioctl(struct file *filp,
		       unsigned int ioctl, unsigned long arg);
772

773 774 775
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);

776 777 778
int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
				    struct kvm_translation *tr);

779 780 781 782 783 784
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);
785 786 787 788
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
789 790
int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
					struct kvm_guest_debug *dbg);
791 792
int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run);

793 794
int kvm_arch_init(void *opaque);
void kvm_arch_exit(void);
795

796 797 798
int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu);
void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu);

R
Radim Krčmář 已提交
799 800
void kvm_arch_sched_in(struct kvm_vcpu *vcpu, int cpu);

801 802 803 804
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);
805
int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu);
806
void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu);
807
void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu);
808

809 810 811
bool kvm_arch_has_vcpu_debugfs(void);
int kvm_arch_create_vcpu_debugfs(struct kvm_vcpu *vcpu);

812 813
int kvm_arch_hardware_enable(void);
void kvm_arch_hardware_disable(void);
814 815 816
int kvm_arch_hardware_setup(void);
void kvm_arch_hardware_unsetup(void);
void kvm_arch_check_processor_compat(void *rtn);
817
int kvm_arch_vcpu_runnable(struct kvm_vcpu *vcpu);
818
bool kvm_arch_vcpu_in_kernel(struct kvm_vcpu *vcpu);
819
int kvm_arch_vcpu_should_kick(struct kvm_vcpu *vcpu);
820

821
#ifndef __KVM_HAVE_ARCH_VM_ALLOC
822 823 824 825
/*
 * All architectures that want to use vzalloc currently also
 * need their own kvm_arch_alloc_vm implementation.
 */
826 827 828 829 830 831 832 833 834 835 836
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

837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854
#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
855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872
#ifdef __KVM_HAVE_ARCH_ASSIGNED_DEVICE
void kvm_arch_start_assignment(struct kvm *kvm);
void kvm_arch_end_assignment(struct kvm *kvm);
bool kvm_arch_has_assigned_device(struct kvm *kvm);
#else
static inline void kvm_arch_start_assignment(struct kvm *kvm)
{
}

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

static inline bool kvm_arch_has_assigned_device(struct kvm *kvm)
{
	return false;
}
#endif
873

874
static inline struct swait_queue_head *kvm_arch_vcpu_wq(struct kvm_vcpu *vcpu)
875
{
876 877 878
#ifdef __KVM_HAVE_ARCH_WQP
	return vcpu->arch.wqp;
#else
879 880
	return &vcpu->wq;
#endif
881
}
882

883 884 885 886 887 888 889 890 891 892 893 894 895 896
#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

897
int kvm_arch_init_vm(struct kvm *kvm, unsigned long type);
898
void kvm_arch_destroy_vm(struct kvm *kvm);
899
void kvm_arch_sync_events(struct kvm *kvm);
900

901
int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu);
902
void kvm_vcpu_kick(struct kvm_vcpu *vcpu);
903

D
Dan Williams 已提交
904
bool kvm_is_reserved_pfn(kvm_pfn_t pfn);
905

B
Ben-Ami Yassour 已提交
906 907 908 909 910 911
struct kvm_irq_ack_notifier {
	struct hlist_node link;
	unsigned gsi;
	void (*irq_acked)(struct kvm_irq_ack_notifier *kian);
};

912 913 914
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);
915

916 917
int kvm_set_irq(struct kvm *kvm, int irq_source_id, u32 irq, int level,
		bool line_status);
918
int kvm_set_msi(struct kvm_kernel_irq_routing_entry *irq_entry, struct kvm *kvm,
919
		int irq_source_id, int level, bool line_status);
920 921 922
int kvm_arch_set_irq_inatomic(struct kvm_kernel_irq_routing_entry *e,
			       struct kvm *kvm, int irq_source_id,
			       int level, bool line_status);
923
bool kvm_irq_has_notifier(struct kvm *kvm, unsigned irqchip, unsigned pin);
924
void kvm_notify_acked_gsi(struct kvm *kvm, int gsi);
925
void kvm_notify_acked_irq(struct kvm *kvm, unsigned irqchip, unsigned pin);
926 927
void kvm_register_irq_ack_notifier(struct kvm *kvm,
				   struct kvm_irq_ack_notifier *kian);
928 929
void kvm_unregister_irq_ack_notifier(struct kvm *kvm,
				   struct kvm_irq_ack_notifier *kian);
930 931
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
932

933 934 935 936 937 938 939 940 941
/*
 * 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)
{
942
	int start = 0, end = slots->used_slots;
943
	int slot = atomic_read(&slots->lru_slot);
944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963
	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];
	}
964 965 966 967 968 969 970 971 972 973

	return NULL;
}

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

974 975 976 977 978 979
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;
}

980 981 982 983 984
static inline int memslot_id(struct kvm *kvm, gfn_t gfn)
{
	return gfn_to_memslot(kvm, gfn)->id;
}

985 986
static inline gfn_t
hva_to_gfn_memslot(unsigned long hva, struct kvm_memory_slot *slot)
987
{
988 989 990
	gfn_t gfn_offset = (hva - slot->userspace_addr) >> PAGE_SHIFT;

	return slot->base_gfn + gfn_offset;
991 992
}

A
Avi Kivity 已提交
993 994 995 996
static inline gpa_t gfn_to_gpa(gfn_t gfn)
{
	return (gpa_t)gfn << PAGE_SHIFT;
}
A
Avi Kivity 已提交
997

998 999 1000 1001 1002
static inline gfn_t gpa_to_gfn(gpa_t gpa)
{
	return (gfn_t)(gpa >> PAGE_SHIFT);
}

D
Dan Williams 已提交
1003
static inline hpa_t pfn_to_hpa(kvm_pfn_t pfn)
B
Ben-Ami Yassour 已提交
1004 1005 1006 1007
{
	return (hpa_t)pfn << PAGE_SHIFT;
}

1008 1009 1010 1011 1012 1013
static inline struct page *kvm_vcpu_gpa_to_page(struct kvm_vcpu *vcpu,
						gpa_t gpa)
{
	return kvm_vcpu_gfn_to_page(vcpu, gpa_to_gfn(gpa));
}

1014 1015 1016 1017 1018 1019 1020
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);
}

1021 1022 1023 1024 1025
enum kvm_stat_kind {
	KVM_STAT_VM,
	KVM_STAT_VCPU,
};

1026 1027 1028 1029 1030
struct kvm_stat_data {
	int offset;
	struct kvm *kvm;
};

1031 1032 1033
struct kvm_stats_debugfs_item {
	const char *name;
	int offset;
1034
	enum kvm_stat_kind kind;
1035 1036
};
extern struct kvm_stats_debugfs_item debugfs_entries[];
1037
extern struct dentry *kvm_debugfs_dir;
1038

1039
#if defined(CONFIG_MMU_NOTIFIER) && defined(KVM_ARCH_WANT_MMU_NOTIFIER)
1040
static inline int mmu_notifier_retry(struct kvm *kvm, unsigned long mmu_seq)
1041
{
1042
	if (unlikely(kvm->mmu_notifier_count))
1043 1044
		return 1;
	/*
1045 1046 1047 1048 1049 1050 1051 1052
	 * 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.
1053
	 */
1054
	smp_rmb();
1055
	if (kvm->mmu_notifier_seq != mmu_seq)
1056 1057 1058 1059 1060
		return 1;
	return 0;
}
#endif

1061
#ifdef CONFIG_HAVE_KVM_IRQ_ROUTING
1062

1063
#define KVM_MAX_IRQ_ROUTES 4096 /* might need extension/rework in the future */
1064

1065
bool kvm_arch_can_set_irq_routing(struct kvm *kvm);
1066 1067 1068 1069
int kvm_set_irq_routing(struct kvm *kvm,
			const struct kvm_irq_routing_entry *entries,
			unsigned nr,
			unsigned flags);
1070 1071
int kvm_set_routing_entry(struct kvm *kvm,
			  struct kvm_kernel_irq_routing_entry *e,
1072
			  const struct kvm_irq_routing_entry *ue);
1073 1074 1075 1076 1077 1078 1079 1080
void kvm_free_irq_routing(struct kvm *kvm);

#else

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

#endif

1081 1082
int kvm_send_userspace_msi(struct kvm *kvm, struct kvm_msi *msi);

G
Gregory Haskins 已提交
1083 1084
#ifdef CONFIG_HAVE_KVM_EVENTFD

G
Gregory Haskins 已提交
1085
void kvm_eventfd_init(struct kvm *kvm);
1086 1087
int kvm_ioeventfd(struct kvm *kvm, struct kvm_ioeventfd *args);

1088
#ifdef CONFIG_HAVE_KVM_IRQFD
1089
int kvm_irqfd(struct kvm *kvm, struct kvm_irqfd *args);
G
Gregory Haskins 已提交
1090
void kvm_irqfd_release(struct kvm *kvm);
1091
void kvm_irq_routing_update(struct kvm *);
1092 1093 1094 1095 1096 1097 1098 1099
#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
1100 1101 1102

#else

G
Gregory Haskins 已提交
1103
static inline void kvm_eventfd_init(struct kvm *kvm) {}
1104

1105
static inline int kvm_irqfd(struct kvm *kvm, struct kvm_irqfd *args)
G
Gregory Haskins 已提交
1106 1107 1108 1109 1110
{
	return -EINVAL;
}

static inline void kvm_irqfd_release(struct kvm *kvm) {}
1111

A
Alexander Graf 已提交
1112
#ifdef CONFIG_HAVE_KVM_IRQCHIP
1113
static inline void kvm_irq_routing_update(struct kvm *kvm)
1114 1115
{
}
A
Alexander Graf 已提交
1116
#endif
1117

G
Gregory Haskins 已提交
1118 1119 1120 1121
static inline int kvm_ioeventfd(struct kvm *kvm, struct kvm_ioeventfd *args)
{
	return -ENOSYS;
}
G
Gregory Haskins 已提交
1122 1123 1124

#endif /* CONFIG_HAVE_KVM_EVENTFD */

1125 1126
void kvm_arch_irq_routing_update(struct kvm *kvm);

1127 1128
static inline void kvm_make_request(int req, struct kvm_vcpu *vcpu)
{
1129 1130 1131 1132 1133
	/*
	 * Ensure the rest of the request is published to kvm_check_request's
	 * caller.  Paired with the smp_mb__after_atomic in kvm_check_request.
	 */
	smp_wmb();
1134
	set_bit(req & KVM_REQUEST_MASK, &vcpu->requests);
1135 1136
}

R
Radim Krčmář 已提交
1137 1138 1139 1140 1141
static inline bool kvm_request_pending(struct kvm_vcpu *vcpu)
{
	return READ_ONCE(vcpu->requests);
}

1142 1143
static inline bool kvm_test_request(int req, struct kvm_vcpu *vcpu)
{
1144
	return test_bit(req & KVM_REQUEST_MASK, &vcpu->requests);
1145 1146 1147 1148
}

static inline void kvm_clear_request(int req, struct kvm_vcpu *vcpu)
{
1149
	clear_bit(req & KVM_REQUEST_MASK, &vcpu->requests);
1150 1151
}

1152 1153
static inline bool kvm_check_request(int req, struct kvm_vcpu *vcpu)
{
1154 1155
	if (kvm_test_request(req, vcpu)) {
		kvm_clear_request(req, vcpu);
1156 1157 1158 1159 1160 1161

		/*
		 * Ensure the rest of the request is visible to kvm_check_request's
		 * caller.  Paired with the smp_wmb in kvm_make_request.
		 */
		smp_mb__after_atomic();
1162 1163 1164 1165
		return true;
	} else {
		return false;
	}
1166 1167
}

1168 1169
extern bool kvm_rebooting;

1170 1171 1172 1173
extern unsigned int halt_poll_ns;
extern unsigned int halt_poll_ns_grow;
extern unsigned int halt_poll_ns_shrink;

S
Scott Wood 已提交
1174 1175 1176 1177
struct kvm_device {
	struct kvm_device_ops *ops;
	struct kvm *kvm;
	void *private;
1178
	struct list_head vm_node;
S
Scott Wood 已提交
1179 1180 1181 1182 1183
};

/* create, destroy, and name are mandatory */
struct kvm_device_ops {
	const char *name;
1184 1185 1186 1187 1188 1189

	/*
	 * create is called holding kvm->lock and any operations not suitable
	 * to do while holding the lock should be deferred to init (see
	 * below).
	 */
S
Scott Wood 已提交
1190 1191
	int (*create)(struct kvm_device *dev, u32 type);

1192 1193 1194 1195 1196 1197
	/*
	 * init is called after create if create is successful and is called
	 * outside of holding kvm->lock.
	 */
	void (*init)(struct kvm_device *dev);

S
Scott Wood 已提交
1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217
	/*
	 * 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);
1218
int kvm_register_device_ops(struct kvm_device_ops *ops, u32 type);
1219
void kvm_unregister_device_ops(u32 type);
S
Scott Wood 已提交
1220

1221
extern struct kvm_device_ops kvm_mpic_ops;
1222
extern struct kvm_device_ops kvm_arm_vgic_v2_ops;
1223
extern struct kvm_device_ops kvm_arm_vgic_v3_ops;
1224

1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245
#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 */
1246 1247

#ifdef CONFIG_HAVE_KVM_IRQ_BYPASS
1248
bool kvm_arch_has_irq_bypass(void);
1249 1250 1251 1252 1253 1254
int kvm_arch_irq_bypass_add_producer(struct irq_bypass_consumer *,
			   struct irq_bypass_producer *);
void kvm_arch_irq_bypass_del_producer(struct irq_bypass_consumer *,
			   struct irq_bypass_producer *);
void kvm_arch_irq_bypass_stop(struct irq_bypass_consumer *);
void kvm_arch_irq_bypass_start(struct irq_bypass_consumer *);
1255 1256
int kvm_arch_update_irqfd_routing(struct kvm *kvm, unsigned int host_irq,
				  uint32_t guest_irq, bool set);
1257
#endif /* CONFIG_HAVE_KVM_IRQ_BYPASS */
1258

1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272
#ifdef CONFIG_HAVE_KVM_INVALID_WAKEUPS
/* If we wakeup during the poll time, was it a sucessful poll? */
static inline bool vcpu_valid_wakeup(struct kvm_vcpu *vcpu)
{
	return vcpu->valid_wakeup;
}

#else
static inline bool vcpu_valid_wakeup(struct kvm_vcpu *vcpu)
{
	return true;
}
#endif /* CONFIG_HAVE_KVM_INVALID_WAKEUPS */

1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284
#ifdef CONFIG_HAVE_KVM_VCPU_ASYNC_IOCTL
long kvm_arch_vcpu_async_ioctl(struct file *filp,
			       unsigned int ioctl, unsigned long arg);
#else
static inline long kvm_arch_vcpu_async_ioctl(struct file *filp,
					     unsigned int ioctl,
					     unsigned long arg)
{
	return -ENOIOCTLCMD;
}
#endif /* CONFIG_HAVE_KVM_VCPU_ASYNC_IOCTL */

1285 1286 1287
void kvm_arch_mmu_notifier_invalidate_range(struct kvm *kvm,
		unsigned long start, unsigned long end);

1288 1289 1290 1291 1292 1293 1294 1295 1296
#ifdef CONFIG_HAVE_KVM_VCPU_RUN_PID_CHANGE
int kvm_arch_vcpu_run_pid_change(struct kvm_vcpu *vcpu);
#else
static inline int kvm_arch_vcpu_run_pid_change(struct kvm_vcpu *vcpu)
{
	return 0;
}
#endif /* CONFIG_HAVE_KVM_VCPU_RUN_PID_CHANGE */

1297
#endif