kvm_host.h 39.5 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>
P
Paolo Bonzini 已提交
31
#include <linux/nospec.h>
A
Alexey Dobriyan 已提交
32
#include <asm/signal.h>
A
Avi Kivity 已提交
33 34

#include <linux/kvm.h>
I
Ingo Molnar 已提交
35
#include <linux/kvm_para.h>
A
Avi Kivity 已提交
36

37
#include <linux/kvm_types.h>
38

39
#include <asm/kvm_host.h>
40

G
Greg Kurz 已提交
41 42 43 44
#ifndef KVM_MAX_VCPU_ID
#define KVM_MAX_VCPU_ID KVM_MAX_VCPUS
#endif

45 46 47 48 49 50 51
/*
 * 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)

52
/*
53
 * Bit 63 of the memslot generation number is an "update in-progress flag",
54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70
 * e.g. is temporarily set for the duration of install_new_memslots().
 * This flag effectively creates a unique generation number that is used to
 * mark cached memslot data, e.g. MMIO accesses, as potentially being stale,
 * i.e. may (or may not) have come from the previous memslots generation.
 *
 * This is necessary because the actual memslots update is not atomic with
 * respect to the generation number update.  Updating the generation number
 * first would allow a vCPU to cache a spte from the old memslots using the
 * new generation number, and updating the generation number after switching
 * to the new memslots would allow cache hits using the old generation number
 * to reference the defunct memslots.
 *
 * This mechanism is used to prevent getting hits in KVM's caches while a
 * memslot update is in-progress, and to prevent cache hits *after* updating
 * the actual generation number against accesses that were inserted into the
 * cache *before* the memslots were updated.
 */
71
#define KVM_MEMSLOT_GEN_UPDATE_IN_PROGRESS	BIT_ULL(63)
72

73 74
/* Two fragments for cross MMIO pages. */
#define KVM_MAX_MMIO_FRAGMENTS	2
A
Avi Kivity 已提交
75

76 77 78 79
#ifndef KVM_ADDRESS_SPACE_NUM
#define KVM_ADDRESS_SPACE_NUM	1
#endif

80 81
/*
 * For the normal pfn, the highest 12 bits should be zero,
82 83
 * so we can mask bit 62 ~ bit 52  to indicate the error pfn,
 * mask bit 63 to indicate the noslot pfn.
84
 */
85 86 87
#define KVM_PFN_ERR_MASK	(0x7ffULL << 52)
#define KVM_PFN_ERR_NOSLOT_MASK	(0xfffULL << 52)
#define KVM_PFN_NOSLOT		(0x1ULL << 63)
88 89 90

#define KVM_PFN_ERR_FAULT	(KVM_PFN_ERR_MASK)
#define KVM_PFN_ERR_HWPOISON	(KVM_PFN_ERR_MASK + 1)
91
#define KVM_PFN_ERR_RO_FAULT	(KVM_PFN_ERR_MASK + 2)
92

93 94 95 96
/*
 * error pfns indicate that the gfn is in slot but faild to
 * translate it to pfn on host.
 */
D
Dan Williams 已提交
97
static inline bool is_error_pfn(kvm_pfn_t pfn)
X
Xiao Guangrong 已提交
98
{
99
	return !!(pfn & KVM_PFN_ERR_MASK);
X
Xiao Guangrong 已提交
100 101
}

102 103 104 105 106
/*
 * 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
107
static inline bool is_error_noslot_pfn(kvm_pfn_t pfn)
X
Xiao Guangrong 已提交
108
{
109
	return !!(pfn & KVM_PFN_ERR_NOSLOT_MASK);
X
Xiao Guangrong 已提交
110 111
}

112
/* noslot pfn indicates that the gfn is not in slot. */
D
Dan Williams 已提交
113
static inline bool is_noslot_pfn(kvm_pfn_t pfn)
X
Xiao Guangrong 已提交
114
{
115
	return pfn == KVM_PFN_NOSLOT;
X
Xiao Guangrong 已提交
116 117
}

118 119 120 121 122 123
/*
 * 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

124 125
#define KVM_HVA_ERR_BAD		(PAGE_OFFSET)
#define KVM_HVA_ERR_RO_BAD	(PAGE_OFFSET + PAGE_SIZE)
X
Xiao Guangrong 已提交
126 127 128

static inline bool kvm_is_error_hva(unsigned long addr)
{
129
	return addr >= PAGE_OFFSET;
X
Xiao Guangrong 已提交
130 131
}

132 133
#endif

134 135
#define KVM_ERR_PTR_BAD_PAGE	(ERR_PTR(-ENOENT))

136
static inline bool is_error_page(struct page *page)
137 138 139 140
{
	return IS_ERR(page);
}

141 142
#define KVM_REQUEST_MASK           GENMASK(7,0)
#define KVM_REQUEST_NO_WAKEUP      BIT(8)
143
#define KVM_REQUEST_WAIT           BIT(9)
144
/*
145 146
 * Architecture-independent vcpu->requests bit members
 * Bits 4-7 are reserved for more arch-independent bits.
147
 */
148 149 150 151
#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
152 153 154
#define KVM_REQUEST_ARCH_BASE     8

#define KVM_ARCH_REQ_FLAGS(nr, flags) ({ \
155
	BUILD_BUG_ON((unsigned)(nr) >= (FIELD_SIZEOF(struct kvm_vcpu, requests) * 8) - KVM_REQUEST_ARCH_BASE); \
156 157 158
	(unsigned)(((nr) + KVM_REQUEST_ARCH_BASE) | (flags)); \
})
#define KVM_ARCH_REQ(nr)           KVM_ARCH_REQ_FLAGS(nr, 0)
159

160 161
#define KVM_USERSPACE_IRQ_SOURCE_ID		0
#define KVM_IRQFD_RESAMPLE_IRQ_SOURCE_ID	1
162

163
extern struct kmem_cache *kvm_vcpu_cache;
A
Avi Kivity 已提交
164

J
Junaid Shahid 已提交
165
extern struct mutex kvm_lock;
166 167
extern struct list_head vm_list;

168 169 170 171 172 173
struct kvm_io_range {
	gpa_t addr;
	int len;
	struct kvm_io_device *dev;
};

174
#define NR_IOBUS_DEVS 1000
175

176
struct kvm_io_bus {
177 178
	int dev_count;
	int ioeventfd_count;
179
	struct kvm_io_range range[];
180 181
};

M
Marcelo Tosatti 已提交
182 183 184
enum kvm_bus {
	KVM_MMIO_BUS,
	KVM_PIO_BUS,
185
	KVM_VIRTIO_CCW_NOTIFY_BUS,
186
	KVM_FAST_MMIO_BUS,
M
Marcelo Tosatti 已提交
187 188 189
	KVM_NR_BUSES
};

190
int kvm_io_bus_write(struct kvm_vcpu *vcpu, enum kvm_bus bus_idx, gpa_t addr,
M
Marcelo Tosatti 已提交
191
		     int len, const void *val);
192 193 194 195
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);
196 197
int kvm_io_bus_register_dev(struct kvm *kvm, enum kvm_bus bus_idx, gpa_t addr,
			    int len, struct kvm_io_device *dev);
198 199
void kvm_io_bus_unregister_dev(struct kvm *kvm, enum kvm_bus bus_idx,
			       struct kvm_io_device *dev);
200 201
struct kvm_io_device *kvm_io_bus_get_dev(struct kvm *kvm, enum kvm_bus bus_idx,
					 gpa_t addr);
202

203 204 205 206 207 208 209 210 211 212
#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;
213
	bool   wakeup_all;
214 215 216 217
};

void kvm_clear_async_pf_completion_queue(struct kvm_vcpu *vcpu);
void kvm_check_async_pf_completion(struct kvm_vcpu *vcpu);
218
int kvm_setup_async_pf(struct kvm_vcpu *vcpu, gva_t gva, unsigned long hva,
219
		       struct kvm_arch_async_pf *arch);
220
int kvm_async_pf_wakeup_all(struct kvm_vcpu *vcpu);
221 222
#endif

223 224 225
enum {
	OUTSIDE_GUEST_MODE,
	IN_GUEST_MODE,
226 227
	EXITING_GUEST_MODE,
	READING_SHADOW_PAGE_TABLES,
228 229
};

230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255
#define KVM_UNMAPPED_PAGE	((void *) 0x500 + POISON_POINTER_DELTA)

struct kvm_host_map {
	/*
	 * Only valid if the 'pfn' is managed by the host kernel (i.e. There is
	 * a 'struct page' for it. When using mem= kernel parameter some memory
	 * can be used as guest memory but they are not managed by host
	 * kernel).
	 * If 'pfn' is not managed by the host kernel, this field is
	 * initialized to KVM_UNMAPPED_PAGE.
	 */
	struct page *page;
	void *hva;
	kvm_pfn_t pfn;
	kvm_pfn_t gfn;
};

/*
 * Used to check if the mapping is valid or not. Never use 'kvm_host_map'
 * directly to check for that.
 */
static inline bool kvm_vcpu_mapped(struct kvm_host_map *map)
{
	return !!map->hva;
}

A
Avi Kivity 已提交
256 257 258 259 260 261 262 263 264 265
/*
 * 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;
};

266 267
struct kvm_vcpu {
	struct kvm *kvm;
268
#ifdef CONFIG_PREEMPT_NOTIFIERS
269
	struct preempt_notifier preempt_notifier;
270
#endif
271
	int cpu;
272
	int vcpu_id;
273 274
	int srcu_idx;
	int mode;
275
	u64 requests;
J
Jan Kiszka 已提交
276
	unsigned long guest_debug;
277

278 279 280
	int pre_pcpu;
	struct list_head blocked_vcpu_list;

281 282
	struct mutex mutex;
	struct kvm_run *run;
283

284
	int guest_xcr0_loaded;
285
	struct swait_queue_head wq;
286
	struct pid __rcu *pid;
287 288 289
	int sigset_active;
	sigset_t sigset;
	struct kvm_vcpu_stat stat;
W
Wanpeng Li 已提交
290
	unsigned int halt_poll_ns;
291
	bool valid_wakeup;
292

293
#ifdef CONFIG_HAS_IOMEM
294 295 296
	int mmio_needed;
	int mmio_read_completed;
	int mmio_is_write;
A
Avi Kivity 已提交
297 298 299
	int mmio_cur_fragment;
	int mmio_nr_fragments;
	struct kvm_mmio_fragment mmio_fragments[KVM_MAX_MMIO_FRAGMENTS];
300
#endif
A
Avi Kivity 已提交
301

302 303 304 305 306 307 308 309 310
#ifdef CONFIG_KVM_ASYNC_PF
	struct {
		u32 queued;
		struct list_head queue;
		struct list_head done;
		spinlock_t lock;
	} async_pf;
#endif

311 312 313 314 315 316 317 318 319 320 321 322
#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
323
	bool preempted;
324
	struct kvm_vcpu_arch arch;
325
	struct dentry *debugfs_dentry;
326 327
};

328 329
static inline int kvm_vcpu_exiting_guest_mode(struct kvm_vcpu *vcpu)
{
330 331 332 333 334 335
	/*
	 * 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();
336 337 338
	return cmpxchg(&vcpu->mode, IN_GUEST_MODE, EXITING_GUEST_MODE);
}

339 340 341 342 343 344
/*
 * 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
345 346 347 348
struct kvm_memory_slot {
	gfn_t base_gfn;
	unsigned long npages;
	unsigned long *dirty_bitmap;
349
	struct kvm_arch_memory_slot arch;
350
	unsigned long userspace_addr;
351
	u32 flags;
352
	short id;
A
Avi Kivity 已提交
353 354
};

355 356 357 358 359
static inline unsigned long kvm_dirty_bitmap_bytes(struct kvm_memory_slot *memslot)
{
	return ALIGN(memslot->npages, BITS_PER_LONG) / 8;
}

360 361 362 363 364 365 366
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);
}

367 368 369 370 371 372 373 374
struct kvm_s390_adapter_int {
	u64 ind_addr;
	u64 summary_addr;
	u64 ind_offset;
	u32 summary_offset;
	u32 adapter_id;
};

375 376 377 378 379
struct kvm_hv_sint {
	u32 vcpu;
	u32 sint;
};

380 381
struct kvm_kernel_irq_routing_entry {
	u32 gsi;
382
	u32 type;
383
	int (*set)(struct kvm_kernel_irq_routing_entry *e,
384 385
		   struct kvm *kvm, int irq_source_id, int level,
		   bool line_status);
386 387 388 389 390
	union {
		struct {
			unsigned irqchip;
			unsigned pin;
		} irqchip;
391 392 393 394 395 396 397
		struct {
			u32 address_lo;
			u32 address_hi;
			u32 data;
			u32 flags;
			u32 devid;
		} msi;
398
		struct kvm_s390_adapter_int adapter;
399
		struct kvm_hv_sint hv_sint;
400
	};
401 402 403
	struct hlist_node link;
};

404 405 406 407 408 409 410 411 412 413 414 415
#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

416 417 418 419
#ifndef KVM_PRIVATE_MEM_SLOTS
#define KVM_PRIVATE_MEM_SLOTS 0
#endif

420
#ifndef KVM_MEM_SLOTS_NUM
421
#define KVM_MEM_SLOTS_NUM (KVM_USER_MEM_SLOTS + KVM_PRIVATE_MEM_SLOTS)
422 423
#endif

424 425 426 427 428 429 430
#ifndef __KVM_VCPU_MULTIPLE_ADDRESS_SPACE
static inline int kvm_arch_vcpu_memslots_id(struct kvm_vcpu *vcpu)
{
	return 0;
}
#endif

431 432 433 434 435
/*
 * Note:
 * memslots are not sorted by id anymore, please use id_to_memslot()
 * to get the memslot by its id.
 */
436
struct kvm_memslots {
437
	u64 generation;
438
	struct kvm_memory_slot memslots[KVM_MEM_SLOTS_NUM];
439
	/* The mapping table from slot id to the index in memslots[]. */
440
	short id_to_index[KVM_MEM_SLOTS_NUM];
441
	atomic_t lru_slot;
442
	int used_slots;
443 444
};

A
Avi Kivity 已提交
445
struct kvm {
446
	spinlock_t mmu_lock;
447
	struct mutex slots_lock;
448
	struct mm_struct *mm; /* userspace tied to this vm */
449
	struct kvm_memslots __rcu *memslots[KVM_ADDRESS_SPACE_NUM];
R
Rusty Russell 已提交
450
	struct kvm_vcpu *vcpus[KVM_MAX_VCPUS];
451 452 453 454 455 456 457

	/*
	 * 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.
	 */
458
	atomic_t online_vcpus;
459
	int created_vcpus;
460
	int last_boosted_vcpu;
461
	struct list_head vm_list;
462
	struct mutex lock;
463
	struct kvm_io_bus __rcu *buses[KVM_NR_BUSES];
G
Gregory Haskins 已提交
464 465 466 467
#ifdef CONFIG_HAVE_KVM_EVENTFD
	struct {
		spinlock_t        lock;
		struct list_head  items;
468 469
		struct list_head  resampler_list;
		struct mutex      resampler_lock;
G
Gregory Haskins 已提交
470
	} irqfds;
G
Gregory Haskins 已提交
471
	struct list_head ioeventfds;
G
Gregory Haskins 已提交
472
#endif
473
	struct kvm_vm_stat stat;
474
	struct kvm_arch arch;
475
	refcount_t users_count;
476
#ifdef CONFIG_KVM_MMIO
477
	struct kvm_coalesced_mmio_ring *coalesced_mmio_ring;
478 479
	spinlock_t ring_lock;
	struct list_head coalesced_zones;
480
#endif
481

482
	struct mutex irq_lock;
483
#ifdef CONFIG_HAVE_KVM_IRQCHIP
484
	/*
485
	 * Update side is protected by irq_lock.
486
	 */
A
Arnd Bergmann 已提交
487
	struct kvm_irq_routing_table __rcu *irq_routing;
488 489
#endif
#ifdef CONFIG_HAVE_KVM_IRQFD
490
	struct hlist_head irq_ack_notifier_list;
491 492
#endif

493
#if defined(CONFIG_MMU_NOTIFIER) && defined(KVM_ARCH_WANT_MMU_NOTIFIER)
494 495 496 497
	struct mmu_notifier mmu_notifier;
	unsigned long mmu_notifier_seq;
	long mmu_notifier_count;
#endif
498
	long tlbs_dirty;
499
	struct list_head devices;
500
	bool manual_dirty_log_protect;
501 502
	struct dentry *debugfs_dentry;
	struct kvm_stat_data **debugfs_stat_data;
503 504
	struct srcu_struct srcu;
	struct srcu_struct irq_srcu;
505
	pid_t userspace_pid;
A
Avi Kivity 已提交
506 507
};

508 509 510 511 512 513
#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__)
514 515 516
#define kvm_debug_ratelimited(fmt, ...) \
	pr_debug_ratelimited("kvm [%i]: " fmt, task_pid_nr(current), \
			     ## __VA_ARGS__)
517 518 519
#define kvm_pr_unimpl(fmt, ...) \
	pr_err_ratelimited("kvm [%i]: " fmt, \
			   task_tgid_nr(current), ## __VA_ARGS__)
520

521 522
/* The guest did something we don't support. */
#define vcpu_unimpl(vcpu, fmt, ...)					\
523 524
	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
525

526 527
#define vcpu_debug(vcpu, fmt, ...)					\
	kvm_debug("vcpu%i " fmt, (vcpu)->vcpu_id, ## __VA_ARGS__)
528 529 530
#define vcpu_debug_ratelimited(vcpu, fmt, ...)				\
	kvm_debug_ratelimited("vcpu%i " fmt, (vcpu)->vcpu_id,           \
			      ## __VA_ARGS__)
531 532
#define vcpu_err(vcpu, fmt, ...)					\
	kvm_err("vcpu%i " fmt, (vcpu)->vcpu_id, ## __VA_ARGS__)
533

534 535 536
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,
537 538
				      lockdep_is_held(&kvm->slots_lock) ||
				      !refcount_read(&kvm->users_count));
539 540
}

541 542
static inline struct kvm_vcpu *kvm_get_vcpu(struct kvm *kvm, int i)
{
P
Paolo Bonzini 已提交
543 544 545 546
	int num_vcpus = atomic_read(&kvm->online_vcpus);
	i = array_index_nospec(i, num_vcpus);

	/* Pairs with smp_wmb() in kvm_vm_ioctl_create_vcpu.  */
547 548 549 550 551
	smp_rmb();
	return kvm->vcpus[i];
}

#define kvm_for_each_vcpu(idx, vcpup, kvm) \
552 553 554 555
	for (idx = 0; \
	     idx < atomic_read(&kvm->online_vcpus) && \
	     (vcpup = kvm_get_vcpu(kvm, idx)) != NULL; \
	     idx++)
556

557 558
static inline struct kvm_vcpu *kvm_get_vcpu_by_id(struct kvm *kvm, int id)
{
559
	struct kvm_vcpu *vcpu = NULL;
560 561
	int i;

562
	if (id < 0)
563
		return NULL;
564 565
	if (id < KVM_MAX_VCPUS)
		vcpu = kvm_get_vcpu(kvm, id);
566 567
	if (vcpu && vcpu->vcpu_id == id)
		return vcpu;
568 569 570 571 572 573
	kvm_for_each_vcpu(i, vcpu, kvm)
		if (vcpu->vcpu_id == id)
			return vcpu;
	return NULL;
}

574 575 576 577 578 579 580 581 582 583 584
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();
}

585 586
#define kvm_for_each_memslot(memslot, slots)	\
	for (memslot = &slots->memslots[0];	\
587 588
	      memslot < slots->memslots + KVM_MEM_SLOTS_NUM && memslot->npages;\
		memslot++)
589

R
Rusty Russell 已提交
590 591 592
int kvm_vcpu_init(struct kvm_vcpu *vcpu, struct kvm *kvm, unsigned id);
void kvm_vcpu_uninit(struct kvm_vcpu *vcpu);

593
void vcpu_load(struct kvm_vcpu *vcpu);
594 595
void vcpu_put(struct kvm_vcpu *vcpu);

596
#ifdef __KVM_HAVE_IOAPIC
597
void kvm_arch_post_irq_ack_notifier_list_update(struct kvm *kvm);
598
void kvm_arch_post_irq_routing_update(struct kvm *kvm);
599
#else
600
static inline void kvm_arch_post_irq_ack_notifier_list_update(struct kvm *kvm)
601 602
{
}
603
static inline void kvm_arch_post_irq_routing_update(struct kvm *kvm)
604 605
{
}
606 607
#endif

608
#ifdef CONFIG_HAVE_KVM_IRQFD
609 610 611 612 613 614 615 616 617 618 619 620
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
621
int kvm_init(void *opaque, unsigned vcpu_size, unsigned vcpu_align,
622
		  struct module *module);
623
void kvm_exit(void);
A
Avi Kivity 已提交
624

I
Izik Eidus 已提交
625 626 627
void kvm_get_kvm(struct kvm *kvm);
void kvm_put_kvm(struct kvm *kvm);

628
static inline struct kvm_memslots *__kvm_memslots(struct kvm *kvm, int as_id)
629
{
P
Paolo Bonzini 已提交
630
	as_id = array_index_nospec(as_id, KVM_ADDRESS_SPACE_NUM);
631
	return srcu_dereference_check(kvm->memslots[as_id], &kvm->srcu,
632 633
			lockdep_is_held(&kvm->slots_lock) ||
			!refcount_read(&kvm->users_count));
634 635
}

636 637 638 639 640
static inline struct kvm_memslots *kvm_memslots(struct kvm *kvm)
{
	return __kvm_memslots(kvm, 0);
}

641 642
static inline struct kvm_memslots *kvm_vcpu_memslots(struct kvm_vcpu *vcpu)
{
643 644 645
	int as_id = kvm_arch_vcpu_memslots_id(vcpu);

	return __kvm_memslots(vcpu->kvm, as_id);
646 647
}

648 649 650
static inline struct kvm_memory_slot *
id_to_memslot(struct kvm_memslots *slots, int id)
{
651 652
	int index = slots->id_to_index[id];
	struct kvm_memory_slot *slot;
653

654
	slot = &slots->memslots[index];
655

656 657
	WARN_ON(slot->id != id);
	return slot;
658 659
}

660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677
/*
 * 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,
};

678
int kvm_set_memory_region(struct kvm *kvm,
679
			  const struct kvm_userspace_memory_region *mem);
680
int __kvm_set_memory_region(struct kvm *kvm,
681
			    const struct kvm_userspace_memory_region *mem);
682
void kvm_arch_free_memslot(struct kvm *kvm, struct kvm_memory_slot *free,
683
			   struct kvm_memory_slot *dont);
684 685
int kvm_arch_create_memslot(struct kvm *kvm, struct kvm_memory_slot *slot,
			    unsigned long npages);
686
void kvm_arch_memslots_updated(struct kvm *kvm, u64 gen);
687 688
int kvm_arch_prepare_memory_region(struct kvm *kvm,
				struct kvm_memory_slot *memslot,
689
				const struct kvm_userspace_memory_region *mem,
690
				enum kvm_mr_change change);
691
void kvm_arch_commit_memory_region(struct kvm *kvm,
692
				const struct kvm_userspace_memory_region *mem,
693
				const struct kvm_memory_slot *old,
694
				const struct kvm_memory_slot *new,
695
				enum kvm_mr_change change);
696
bool kvm_largepages_enabled(void);
697
void kvm_disable_largepages(void);
698 699 700 701 702
/* 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);
703

704 705
int gfn_to_page_many_atomic(struct kvm_memory_slot *slot, gfn_t gfn,
			    struct page **pages, int nr_pages);
706

A
Avi Kivity 已提交
707
struct page *gfn_to_page(struct kvm *kvm, gfn_t gfn);
M
Marcelo Tosatti 已提交
708
unsigned long gfn_to_hva(struct kvm *kvm, gfn_t gfn);
709
unsigned long gfn_to_hva_prot(struct kvm *kvm, gfn_t gfn, bool *writable);
X
Xiao Guangrong 已提交
710
unsigned long gfn_to_hva_memslot(struct kvm_memory_slot *slot, gfn_t gfn);
711 712
unsigned long gfn_to_hva_memslot_prot(struct kvm_memory_slot *slot, gfn_t gfn,
				      bool *writable);
713 714
void kvm_release_page_clean(struct page *page);
void kvm_release_page_dirty(struct page *page);
715 716
void kvm_set_page_accessed(struct page *page);

D
Dan Williams 已提交
717 718 719
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,
720
		      bool *writable);
D
Dan Williams 已提交
721 722 723 724 725
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);
726

D
Dan Williams 已提交
727
void kvm_release_pfn_clean(kvm_pfn_t pfn);
728
void kvm_release_pfn_dirty(kvm_pfn_t pfn);
D
Dan Williams 已提交
729 730 731
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);
732

733 734
int kvm_read_guest_page(struct kvm *kvm, gfn_t gfn, void *data, int offset,
			int len);
735 736
int kvm_read_guest_atomic(struct kvm *kvm, gpa_t gpa, void *data,
			  unsigned long len);
737
int kvm_read_guest(struct kvm *kvm, gpa_t gpa, void *data, unsigned long len);
738 739
int kvm_read_guest_cached(struct kvm *kvm, struct gfn_to_hva_cache *ghc,
			   void *data, unsigned long len);
740 741 742 743
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);
744 745 746
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,
747 748
				  void *data, unsigned int offset,
				  unsigned long len);
749 750
int kvm_gfn_to_hva_cache_init(struct kvm *kvm, struct gfn_to_hva_cache *ghc,
			      gpa_t gpa, unsigned long len);
751 752
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
753
struct kvm_memory_slot *gfn_to_memslot(struct kvm *kvm, gfn_t gfn);
754
bool kvm_is_visible_gfn(struct kvm *kvm, gfn_t gfn);
J
Joerg Roedel 已提交
755
unsigned long kvm_host_page_size(struct kvm *kvm, gfn_t gfn);
A
Avi Kivity 已提交
756 757
void mark_page_dirty(struct kvm *kvm, gfn_t gfn);

758 759
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
760 761
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);
762
int kvm_vcpu_map(struct kvm_vcpu *vcpu, gpa_t gpa, struct kvm_host_map *map);
763
struct page *kvm_vcpu_gfn_to_page(struct kvm_vcpu *vcpu, gfn_t gfn);
764
void kvm_vcpu_unmap(struct kvm_vcpu *vcpu, struct kvm_host_map *map, bool dirty);
765 766 767 768 769 770 771 772 773 774 775 776 777 778
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);

779 780 781
void kvm_sigset_activate(struct kvm_vcpu *vcpu);
void kvm_sigset_deactivate(struct kvm_vcpu *vcpu);

782
void kvm_vcpu_block(struct kvm_vcpu *vcpu);
783 784
void kvm_arch_vcpu_blocking(struct kvm_vcpu *vcpu);
void kvm_arch_vcpu_unblocking(struct kvm_vcpu *vcpu);
785
bool kvm_vcpu_wake_up(struct kvm_vcpu *vcpu);
786
void kvm_vcpu_kick(struct kvm_vcpu *vcpu);
787
int kvm_vcpu_yield_to(struct kvm_vcpu *target);
788
void kvm_vcpu_on_spin(struct kvm_vcpu *vcpu, bool usermode_vcpu_not_eligible);
789

790
void kvm_flush_remote_tlbs(struct kvm *kvm);
791
void kvm_reload_remote_mmus(struct kvm *kvm);
792 793 794

bool kvm_make_vcpus_request_mask(struct kvm *kvm, unsigned int req,
				 unsigned long *vcpu_bitmap, cpumask_var_t tmp);
795
bool kvm_make_all_cpus_request(struct kvm *kvm, unsigned int req);
A
Avi Kivity 已提交
796

797 798
long kvm_arch_dev_ioctl(struct file *filp,
			unsigned int ioctl, unsigned long arg);
799 800
long kvm_arch_vcpu_ioctl(struct file *filp,
			 unsigned int ioctl, unsigned long arg);
801
vm_fault_t kvm_arch_vcpu_fault(struct kvm_vcpu *vcpu, struct vm_fault *vmf);
802

803
int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext);
804

805 806
int kvm_get_dirty_log(struct kvm *kvm,
			struct kvm_dirty_log *log, int *is_dirty);
807 808

int kvm_get_dirty_log_protect(struct kvm *kvm,
809
			      struct kvm_dirty_log *log, bool *flush);
810 811
int kvm_clear_dirty_log_protect(struct kvm *kvm,
				struct kvm_clear_dirty_log *log, bool *flush);
812

813
void kvm_arch_mmu_enable_log_dirty_pt_masked(struct kvm *kvm,
814 815 816 817
					struct kvm_memory_slot *slot,
					gfn_t gfn_offset,
					unsigned long mask);

818 819
int kvm_vm_ioctl_get_dirty_log(struct kvm *kvm,
				struct kvm_dirty_log *log);
820 821
int kvm_vm_ioctl_clear_dirty_log(struct kvm *kvm,
				  struct kvm_clear_dirty_log *log);
822

823 824
int kvm_vm_ioctl_irq_line(struct kvm *kvm, struct kvm_irq_level *irq_level,
			bool line_status);
825 826
int kvm_vm_ioctl_enable_cap(struct kvm *kvm,
			    struct kvm_enable_cap *cap);
827 828
long kvm_arch_vm_ioctl(struct file *filp,
		       unsigned int ioctl, unsigned long arg);
829

830 831 832
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);

833 834 835
int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
				    struct kvm_translation *tr);

836 837 838 839 840 841
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);
842 843 844 845
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
846 847
int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
					struct kvm_guest_debug *dbg);
848 849
int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run);

850 851
int kvm_arch_init(void *opaque);
void kvm_arch_exit(void);
852

853 854 855
int kvm_arch_vcpu_init(struct kvm_vcpu *vcpu);
void kvm_arch_vcpu_uninit(struct kvm_vcpu *vcpu);

R
Radim Krčmář 已提交
856 857
void kvm_arch_sched_in(struct kvm_vcpu *vcpu, int cpu);

858 859 860 861
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);
862
int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu);
863
void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu);
864
void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu);
865

866 867 868
bool kvm_arch_has_vcpu_debugfs(void);
int kvm_arch_create_vcpu_debugfs(struct kvm_vcpu *vcpu);

869 870
int kvm_arch_hardware_enable(void);
void kvm_arch_hardware_disable(void);
871 872
int kvm_arch_hardware_setup(void);
void kvm_arch_hardware_unsetup(void);
873
int kvm_arch_check_processor_compat(void);
874
int kvm_arch_vcpu_runnable(struct kvm_vcpu *vcpu);
875
bool kvm_arch_vcpu_in_kernel(struct kvm_vcpu *vcpu);
876
int kvm_arch_vcpu_should_kick(struct kvm_vcpu *vcpu);
877

878
#ifndef __KVM_HAVE_ARCH_VM_ALLOC
879 880 881 882
/*
 * All architectures that want to use vzalloc currently also
 * need their own kvm_arch_alloc_vm implementation.
 */
883 884 885 886 887 888 889 890 891 892 893
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

894 895 896 897 898 899 900
#ifndef __KVM_HAVE_ARCH_FLUSH_REMOTE_TLB
static inline int kvm_arch_flush_remote_tlb(struct kvm *kvm)
{
	return -ENOTSUPP;
}
#endif

901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918
#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
919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936
#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
937

938
static inline struct swait_queue_head *kvm_arch_vcpu_wq(struct kvm_vcpu *vcpu)
939
{
940 941 942
#ifdef __KVM_HAVE_ARCH_WQP
	return vcpu->arch.wqp;
#else
943 944
	return &vcpu->wq;
#endif
945
}
946

947 948 949 950 951 952 953 954 955 956 957 958 959 960
#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

961
int kvm_arch_init_vm(struct kvm *kvm, unsigned long type);
962
void kvm_arch_destroy_vm(struct kvm *kvm);
963
void kvm_arch_sync_events(struct kvm *kvm);
964

965
int kvm_cpu_has_pending_timer(struct kvm_vcpu *vcpu);
966
void kvm_vcpu_kick(struct kvm_vcpu *vcpu);
967

D
Dan Williams 已提交
968
bool kvm_is_reserved_pfn(kvm_pfn_t pfn);
969

B
Ben-Ami Yassour 已提交
970 971 972 973 974 975
struct kvm_irq_ack_notifier {
	struct hlist_node link;
	unsigned gsi;
	void (*irq_acked)(struct kvm_irq_ack_notifier *kian);
};

976 977 978
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);
979

980 981
int kvm_set_irq(struct kvm *kvm, int irq_source_id, u32 irq, int level,
		bool line_status);
982
int kvm_set_msi(struct kvm_kernel_irq_routing_entry *irq_entry, struct kvm *kvm,
983
		int irq_source_id, int level, bool line_status);
984 985 986
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);
987
bool kvm_irq_has_notifier(struct kvm *kvm, unsigned irqchip, unsigned pin);
988
void kvm_notify_acked_gsi(struct kvm *kvm, int gsi);
989
void kvm_notify_acked_irq(struct kvm *kvm, unsigned irqchip, unsigned pin);
990 991
void kvm_register_irq_ack_notifier(struct kvm *kvm,
				   struct kvm_irq_ack_notifier *kian);
992 993
void kvm_unregister_irq_ack_notifier(struct kvm *kvm,
				   struct kvm_irq_ack_notifier *kian);
994 995
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
996

997 998 999 1000 1001 1002 1003 1004 1005
/*
 * 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)
{
1006
	int start = 0, end = slots->used_slots;
1007
	int slot = atomic_read(&slots->lru_slot);
1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027
	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];
	}
1028 1029 1030 1031 1032 1033 1034 1035 1036 1037

	return NULL;
}

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

1038 1039 1040 1041 1042 1043
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;
}

1044 1045 1046 1047 1048
static inline int memslot_id(struct kvm *kvm, gfn_t gfn)
{
	return gfn_to_memslot(kvm, gfn)->id;
}

1049 1050
static inline gfn_t
hva_to_gfn_memslot(unsigned long hva, struct kvm_memory_slot *slot)
1051
{
1052 1053 1054
	gfn_t gfn_offset = (hva - slot->userspace_addr) >> PAGE_SHIFT;

	return slot->base_gfn + gfn_offset;
1055 1056
}

A
Avi Kivity 已提交
1057 1058 1059 1060
static inline gpa_t gfn_to_gpa(gfn_t gfn)
{
	return (gpa_t)gfn << PAGE_SHIFT;
}
A
Avi Kivity 已提交
1061

1062 1063 1064 1065 1066
static inline gfn_t gpa_to_gfn(gpa_t gpa)
{
	return (gfn_t)(gpa >> PAGE_SHIFT);
}

D
Dan Williams 已提交
1067
static inline hpa_t pfn_to_hpa(kvm_pfn_t pfn)
B
Ben-Ami Yassour 已提交
1068 1069 1070 1071
{
	return (hpa_t)pfn << PAGE_SHIFT;
}

1072 1073 1074 1075 1076 1077
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));
}

1078 1079 1080 1081 1082 1083 1084
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);
}

1085 1086 1087 1088 1089
enum kvm_stat_kind {
	KVM_STAT_VM,
	KVM_STAT_VCPU,
};

1090 1091 1092 1093 1094
struct kvm_stat_data {
	int offset;
	struct kvm *kvm;
};

1095 1096 1097
struct kvm_stats_debugfs_item {
	const char *name;
	int offset;
1098
	enum kvm_stat_kind kind;
1099 1100
};
extern struct kvm_stats_debugfs_item debugfs_entries[];
1101
extern struct dentry *kvm_debugfs_dir;
1102

1103
#if defined(CONFIG_MMU_NOTIFIER) && defined(KVM_ARCH_WANT_MMU_NOTIFIER)
1104
static inline int mmu_notifier_retry(struct kvm *kvm, unsigned long mmu_seq)
1105
{
1106
	if (unlikely(kvm->mmu_notifier_count))
1107 1108
		return 1;
	/*
1109 1110 1111 1112 1113 1114 1115 1116
	 * 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.
1117
	 */
1118
	smp_rmb();
1119
	if (kvm->mmu_notifier_seq != mmu_seq)
1120 1121 1122 1123 1124
		return 1;
	return 0;
}
#endif

1125
#ifdef CONFIG_HAVE_KVM_IRQ_ROUTING
1126

1127
#define KVM_MAX_IRQ_ROUTES 4096 /* might need extension/rework in the future */
1128

1129
bool kvm_arch_can_set_irq_routing(struct kvm *kvm);
1130 1131 1132 1133
int kvm_set_irq_routing(struct kvm *kvm,
			const struct kvm_irq_routing_entry *entries,
			unsigned nr,
			unsigned flags);
1134 1135
int kvm_set_routing_entry(struct kvm *kvm,
			  struct kvm_kernel_irq_routing_entry *e,
1136
			  const struct kvm_irq_routing_entry *ue);
1137 1138 1139 1140 1141 1142 1143 1144
void kvm_free_irq_routing(struct kvm *kvm);

#else

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

#endif

1145 1146
int kvm_send_userspace_msi(struct kvm *kvm, struct kvm_msi *msi);

G
Gregory Haskins 已提交
1147 1148
#ifdef CONFIG_HAVE_KVM_EVENTFD

G
Gregory Haskins 已提交
1149
void kvm_eventfd_init(struct kvm *kvm);
1150 1151
int kvm_ioeventfd(struct kvm *kvm, struct kvm_ioeventfd *args);

1152
#ifdef CONFIG_HAVE_KVM_IRQFD
1153
int kvm_irqfd(struct kvm *kvm, struct kvm_irqfd *args);
G
Gregory Haskins 已提交
1154
void kvm_irqfd_release(struct kvm *kvm);
1155
void kvm_irq_routing_update(struct kvm *);
1156 1157 1158 1159 1160 1161 1162 1163
#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
1164 1165 1166

#else

G
Gregory Haskins 已提交
1167
static inline void kvm_eventfd_init(struct kvm *kvm) {}
1168

1169
static inline int kvm_irqfd(struct kvm *kvm, struct kvm_irqfd *args)
G
Gregory Haskins 已提交
1170 1171 1172 1173 1174
{
	return -EINVAL;
}

static inline void kvm_irqfd_release(struct kvm *kvm) {}
1175

A
Alexander Graf 已提交
1176
#ifdef CONFIG_HAVE_KVM_IRQCHIP
1177
static inline void kvm_irq_routing_update(struct kvm *kvm)
1178 1179
{
}
A
Alexander Graf 已提交
1180
#endif
1181

G
Gregory Haskins 已提交
1182 1183 1184 1185
static inline int kvm_ioeventfd(struct kvm *kvm, struct kvm_ioeventfd *args)
{
	return -ENOSYS;
}
G
Gregory Haskins 已提交
1186 1187 1188

#endif /* CONFIG_HAVE_KVM_EVENTFD */

1189 1190
void kvm_arch_irq_routing_update(struct kvm *kvm);

1191 1192
static inline void kvm_make_request(int req, struct kvm_vcpu *vcpu)
{
1193 1194 1195 1196 1197
	/*
	 * 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();
1198
	set_bit(req & KVM_REQUEST_MASK, (void *)&vcpu->requests);
1199 1200
}

R
Radim Krčmář 已提交
1201 1202 1203 1204 1205
static inline bool kvm_request_pending(struct kvm_vcpu *vcpu)
{
	return READ_ONCE(vcpu->requests);
}

1206 1207
static inline bool kvm_test_request(int req, struct kvm_vcpu *vcpu)
{
1208
	return test_bit(req & KVM_REQUEST_MASK, (void *)&vcpu->requests);
1209 1210 1211 1212
}

static inline void kvm_clear_request(int req, struct kvm_vcpu *vcpu)
{
1213
	clear_bit(req & KVM_REQUEST_MASK, (void *)&vcpu->requests);
1214 1215
}

1216 1217
static inline bool kvm_check_request(int req, struct kvm_vcpu *vcpu)
{
1218 1219
	if (kvm_test_request(req, vcpu)) {
		kvm_clear_request(req, vcpu);
1220 1221 1222 1223 1224 1225

		/*
		 * 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();
1226 1227 1228 1229
		return true;
	} else {
		return false;
	}
1230 1231
}

1232 1233
extern bool kvm_rebooting;

1234 1235
extern unsigned int halt_poll_ns;
extern unsigned int halt_poll_ns_grow;
1236
extern unsigned int halt_poll_ns_grow_start;
1237 1238
extern unsigned int halt_poll_ns_shrink;

S
Scott Wood 已提交
1239 1240 1241 1242
struct kvm_device {
	struct kvm_device_ops *ops;
	struct kvm *kvm;
	void *private;
1243
	struct list_head vm_node;
S
Scott Wood 已提交
1244 1245 1246 1247 1248
};

/* create, destroy, and name are mandatory */
struct kvm_device_ops {
	const char *name;
1249 1250 1251 1252 1253 1254

	/*
	 * 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
1255 1256
	int (*create)(struct kvm_device *dev, u32 type);

1257 1258 1259 1260 1261 1262
	/*
	 * 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
1263 1264 1265 1266 1267 1268 1269 1270 1271 1272
	/*
	 * 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);

1273 1274 1275 1276 1277 1278 1279 1280 1281
	/*
	 * Release is an alternative method to free the device. It is
	 * called when the device file descriptor is closed. Once
	 * release is called, the destroy method will not be called
	 * anymore as the device is removed from the device list of
	 * the VM. kvm->lock is held.
	 */
	void (*release)(struct kvm_device *dev);

S
Scott Wood 已提交
1282 1283 1284 1285 1286
	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);
1287
	int (*mmap)(struct kvm_device *dev, struct vm_area_struct *vma);
S
Scott Wood 已提交
1288 1289 1290 1291 1292
};

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);
1293
int kvm_register_device_ops(struct kvm_device_ops *ops, u32 type);
1294
void kvm_unregister_device_ops(u32 type);
S
Scott Wood 已提交
1295

1296
extern struct kvm_device_ops kvm_mpic_ops;
1297
extern struct kvm_device_ops kvm_arm_vgic_v2_ops;
1298
extern struct kvm_device_ops kvm_arm_vgic_v3_ops;
1299

1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320
#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 */
1321 1322

#ifdef CONFIG_HAVE_KVM_IRQ_BYPASS
1323
bool kvm_arch_has_irq_bypass(void);
1324 1325 1326 1327 1328 1329
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 *);
1330 1331
int kvm_arch_update_irqfd_routing(struct kvm *kvm, unsigned int host_irq,
				  uint32_t guest_irq, bool set);
1332
#endif /* CONFIG_HAVE_KVM_IRQ_BYPASS */
1333

1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347
#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 */

1348 1349 1350 1351 1352 1353 1354 1355 1356 1357
#ifdef CONFIG_HAVE_KVM_NO_POLL
/* Callback that tells if we must not poll */
bool kvm_arch_no_poll(struct kvm_vcpu *vcpu);
#else
static inline bool kvm_arch_no_poll(struct kvm_vcpu *vcpu)
{
	return false;
}
#endif /* CONFIG_HAVE_KVM_NO_POLL */

1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369
#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 */

1370 1371
int kvm_arch_mmu_notifier_invalidate_range(struct kvm *kvm,
		unsigned long start, unsigned long end, bool blockable);
1372

1373 1374 1375 1376 1377 1378 1379 1380 1381
#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 */

1382
#endif